Ricardo - The New View: Collected Essays I

Citation preview

RICARDO – THE NEW VIEW

RICARDO - THE NEW VIEW Collected Essays I

Samuel Hollander

London and New York

First published 1995 by Routledge 11 New Fetter Lane, London EC4P 4EE Simultaneously published in the USA and Canada by Routledge 29 West 35th Street, New York, NY 10001 This edition published in the Taylor & Francis e-Library, 2001. ©1995 Samuel Hollander All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. British Library Cataloging in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data Hollander, Samuel, Ricardo – the New View: collected essays/Samuel Hollander. p. cm. Includes bibliographical references and index. 1. Ricardo, David. 1772-1823. 2. Economists - - Great Britain. HB103.R5H723 1995 330.15 E´ 3 - - dc20 95-7939 CIP ISBN 0–415–11582–5 (v. 1) ISBN 0-203-00649-6 Master e-book ISBN ISBN 0-203-15988-8 (Glassbook Format)

Dedicated to my doctoral students who have had the courage – in this day and age – to commit themselves to the History of Economics: Margaret Schabas, Evelyn Forget, Sandra Peart, Tom Kompas, Richard Kleer, Nancy Churchman, Tim Davis, Ingrid Peters-Fransen and Masazumi Wakatabe

CONTENTS

Preface Acknowledgements Introduction

xi xiii 1

Part I. On the interpretation of the early Ricardo 1. 2. 3. 4. 5.

RICARDO’S ANALYSIS OF THE PROFIT RATE, 1813–15 [1973]

19

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL [1975]

44

PROFESSOR GAREGNANI’S DEFENCE OF SRAFFA ON THE MATERIAL RATE OF PROFIT [1983]

60

ON A ‘NEW INTERPRETATION’ OF RICARDO’S EARLY TREATMENT OF PROFITABILITY [1986]

71

SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO: ON THREE CONTRIBUTIONS TO THE CAMBRIDGE JOURNAL OF ECONOMICS [1995]

79

Part II. Responses to critics of The Economics of David Ricardo 6.

THE ECONOMICS OF DAVID RICARDO: A RESPONSE TO PROFESSOR O’BRIEN [1982]

91

7.

A REPLY TO PROFESSOR RONCAGLIA [1982]

116

8.

‘PROFESSOR HOLLANDER AND RICARDIAN ECONOMICS’; A REPLY TO PROFESSOR MOSS [1982]

127

CONTENTS

Part III. Ricardian micro-economics 9.

ON THE SUBSTANTIVE IDENTITY OF THE RICARDIAN AND NEO-CLASSICAL CONCEPTIONS OF ECONOMIC ORGANIZATION: THE FRENCH CONNECTION IN BRITISH CLASSICISM [1982]

135

10. WHY MARSHALL WAS RIGHT ABOUT RICARDO [1989]

167

11. ON COMPOSITION OF DEMAND AND INCOME DISTRIBUTION IN CLASSICAL ECONOMICS [1989]

195

12. ON THE ENDOGENEITY OF THE MARGIN AND RELATED ISSUES IN RICARDIAN ECONOMICS [1991]

202

Part IV. The Ricardian growth model 13. ON THE INTERPRETATION OF RICARDIAN ECONOMICS: THE ASSUMPTION REGARDING WAGES [1983]

219

14. THE WAGE PATH IN CLASSICAL GROWTH MODELS: RICARDO, MALTHUS AND MILL [1984]

226

15. RICARDIAN GROWTH THEORY: A RESOLUTION OF SOME PROBLEMS IN TEXTUAL INTERPRETATION [1990]

241

16. A REPLY TO PROFESSOR STIGLER AND DR PEACH [1990]

265

17. ON THE TEXTUAL INTERPRETATION OF RICARDIAN GROWTH THEORY: THE ‘NEW VIEW’ CONFIRMED (AGAIN) [1994]

268

Part V. Further intellectual linkages 18. THE RECEPTION OF RICARDIAN ECONOMICS [1977]

283

19. THE ROLE OF BENTHAM IN THE EARLY DEVELOPMENT OF RICARDIAN THEORY: A SPECULATIVE ESSAY [1979]

323

viii

CONTENTS

20. ON PROFESSOR SAMUELSON’S CANONICAL CLASSICAL MODEL OF POLITICAL ECONOMY [1980] Index

ix

342 363

PREFACE

There is more repetition of evidence and argument in this book than would be appropriate were it probable that readers would choose to work through it from start to finish, rather than read the essays (that stand by themselves) selectively. And I have found it preferable to retain the originals in their entirety rather than devise a complex system of cross references to accommodate deletions and contractions. The overlap is probably greatest in the case of Chapter 9 and Chapter 10 but I chose to reproduce both since each was given as a public lecture (carrying with it fond memories) while the second has never appeared in English in its entirety and contains material not found in the first. Finally, the full flavour of the campaign that I have in effect been waging for more than two decades only becomes evident – it only became fully evident to me – when the articles appear as they now do. I note here, as a late bulletin, that the mystery posed in Chapter 19 may now have been solved with the apparent discovery of a Bentham Manuscript in the Sraffa papers held at Trinity College, Cambridge. On this matter, see my forthcoming ‘Notes on a Probable Bentham Manuscript’, in the Cambridge Journal of Economics.

ACKNOWLEDGEMENTS

I am indebted to the following for permission to reproduce the materials in this volume: 1 Blackwell publishers, for ‘Ricardo’s Analysis of the Profit Rate, 1813–15’, Economica, August 1973, pp. 261–82 (Chapter 1); ‘Ricardo and the Corn Profit Model: Reply to Eatwell’, Economica, May 1975, pp. 188–202 (Chapter 2); ‘On a “New Interpretation” of Ricardo’s Early Treatment of Profitability’, The Economic Journal, December 1986, pp. 1091–7 (Chapter 4). 2 Academic Press Ltd, for ‘Professor Garegnani’s Defence of Sraffa on the Material Rate of Profit’, Cambridge Journal of Economics, 1983, pp. 167–74 (Chapter 3); ‘Sraffa’s Rational Reconstruction of Ricardo’, Cambridge Journal of Economics, June 1995, pp. 483–9 (Chapter 5). 3 Oxford University Press, for ‘The Economics of David Ricardo: A Response to Professor O’Brien’, Oxford Economic Papers, March 1982, pp. 224–46 (Chapter 6); ‘The Wage Path in Classical Growth Models’, Oxford Economic Papers, 1984, pp. 200–12 (Chapter 14); ‘Ricardian Growth Theory: A Resolution of Some Problems in Textual Interpretation’, Oxford Economic Papers, 1990, pp. 730–50 (Chapter 15); ‘A Reply to Professor Stigler and Dr Peach’, Oxford Economic Papers, 1990, pp. 769–71 (Chapter 16); ‘The Reception of Ricardian Economics’, Oxford Economic Papers, 1977, pp. 221– 57 (Chapter 18). 4 M.E. Sharpe Inc., for ‘A Reply to Professor Roncaglia’, Journal of Post-Keynesian Economics, 1982, pp. 360–72 (Chapter 7). 5 Eastern Economic Journal, for ‘A Reply to Professor Moss’, July 1982, pp. 237–41 (Chapter 8). xiii

RICARDO - THE NEW VIEW

6 The Canadian Economics Association, for ‘On the Substantive Identity of Ricardian and Neo-Classical Conceptions of Economic Organization’, Canadian Journal of Economics, 1982, pp. 586–612 (Chapter 9). 7 The History of Economics Society, for ‘On Composition of Demand and Income Distribution in Classical Economics’, History of Economics Society Bulletin, Fall 1989, pp. 216–21 (Chapter 11); ‘On the Endogeneity of the Margin’, Journal of the History of Economic Thought, Fall 1991, pp. 159–73 (Chapter 12). 8 The American Economic Association, for ‘On the Interpretation of Ricardian Economics’, The American Economic Review, May 1983, pp. 314–18 (Chapter 13); ‘On Professor Samuelson’s Canonical Classical Model’, Journal of Economic Literature, June 1980, pp. 559–74 (Chapter 20). 9 Duke University Press, for ‘On the Textual Interpretation of Ricardian Theory: the “New View” Confirmed (Again)’, History of Political Economy, 1994, pp. 487–99 (Chapter 17). 10 The Bentham Project, for ‘The Role of Bentham in the Early Development of Ricardian Theory’, The Bentham Newsletter, December 1979, pp. 2–17 (Chapter 19). I also wish to acknowledge the award of a Grant by the Social Science and Humanities Research Council of Canada which aided me in the preparation of the Introduction and Chapter 5 and Chapter 17.

xiv

INTRODUCTION

When approached by a publishing house to preface a collection of articles and papers by an autobiographical introduction, it is difficult not to imagine the sound of fluttering wings, harps and such like. Since I am feeling in reasonably good health and wish to speculate a little further on the manner in which the fallibility and subjectivity of the historian might affect – who knows, perhaps even improve – his work, I shall take the risk and postpone a full-fledged confession to a later volume in this series of Collected Essays, limiting myself here to the minimum required to provide the present selection with an appropriate background. Those familiar with Dr Donald Trefusis, Regius Professor of Philology at the University of Cambridge and Extraordinary Fellow of St Matthew’s College, may recall that in one version of his obituary – College members traditionally write their own obituaries to ‘save ... family and friends the pain and embarrassment of having to make up lies themselves’ – he wrote of himself that ‘he was as reviled, scorned and despised as any pure academic’ (Fry 1993, 43–4). I am able to claim that my work on David Ricardo and the classical economists has attracted reactions that would make Trefusis green with envy – ‘bizarre’, ‘fantastic’, ‘ludicrous’, even ‘tragic’, are only some of the friendlier examples. Trefusis might have been inclined to cite Milton on my behalf: ‘Truth never comes into the world but like a bastard, to the ignominy of him that brought her forth.’ He could say that; I could not possibly comment. Of course no one is perfect. I have to admit to very strong support indeed from other quarters, even some ‘conversions’. To give the flavour of the extraordinarily disparate opinions I cannot do better than cite two polar extreme evaluations, each involving a Decalogue. The first is by Professor Mark Blaug in his celebrated Economic Theory in Retrospect : I have left to the last the most recent and the most exhaustive study of the Ricardian model and Ricardo’s views on just about everything: S. Hollander, The Economics of David Ricardo (1979). This massive book is nothing less than a full-scale frontal attack on the entire body of Ricardian scholarship, arguing that absolutely everybody else has more or less misinterpreted Ricardo. 1

RICARDO - THE NEW VIEW

Consider just some of the iconoclastic themes of Hollander’s opus: (1) Ricardo’s method of analysis was identical to that of Adam Smith; (2) Ricardo’s work was basically in the tradition of general equilibrium analysis that runs from Smith to Walras to modern days and, in particular, Ricardo treated pricing and distribution as interdependent; (3) Ricardo’s profit theory did not originate in a concern over the Corn Laws and Ricardo never believed, even in his earlier writings, that profits in agriculture determine the general rate of profit in the economy; (4) Ricardo’s value theory was essentially the same as that of Marshall in that it paid as much attention to demand as to supply, and Ricardo never regarded the invariable measure of value as an important element in his theory; (5) Ricardo could have established his fundamental theorem that ‘profits vary inversely with wages’ without his invariable yardstick and frequently took the short-cut of assuming identical capital–labour ratios in all industries to give him the answers he looked for; (6) wages in Ricardo are never conceived as constant or fixed at subsistence levels; (7) Ricardo never assumed a zero price-elasticity of demand for corn, in effect making the demand for agricultural produce a simple function of the size of the population; (8) Ricardo was not a quantity theorist in the conventional sense, nor a rigid Bullionist, nor did he hold a monetary theory that was very different from that of Adam Smith; (9) Ricardo did not predict a rising rental share, nor did he ever commit himself to any clear-cut predictions about any economic variable, least of all the rate of profit; and (10) Ricardo was never seriously concerned about the possibility of class conflict between landowners, on the one hand, and workers and capitalists, on the other. I believe that every one of these ten assertions is false but readers will have to consult the book and to make up their own minds. (Blaug 1985, 147) I actually go a long way with Professor Blaug. Many of the assertions (especially those containing ‘never’ or the equivalent) are false. But these I do not recognize as my own, claiming for myself and attributing to Ricardo a modicum of subtlety and sophistication. A fairer Decalogue – we may have proof here for Manicheism – is by Professor Martin Bronfenbrenner in a review of my Classical Economics (1987, 1992) in which Ricardo figures large. This version has the merit of defining the context of my work with great clarity, saving me considerable effort: The challenge of Hollander is on two fronts at once, which is to say, that it is directed against two categories of more or less complacent opponents: conventional marginalists to his political Right as well as ... Italo-Cambridge Neo-Keynesians to his political Left. Let me summarize in a decalogue of propositions his principal challenges to conventional marginalism, in which 2

INTRODUCTION

the present generation of economists (including the present writer) was largely taught, and largely trained to consider classical economics outmoded in the fashion of Grandma’s hoop skirt and Grandpa’s stovepipe hat: 1. The classical economists did not ignore the influence of demand on value and price, either of inputs or of outputs, either in the short or the long run. Rather, their entire analysis was a development of supply and demand, seen as functional relationships and not as fixed quantities. 2. The classical economists did not accept either a subsistence ‘iron law’ or a wage-fund theory of wage rates. 3. Nor did they maintain that the distribution of income would necessarily shift in favour of the landlord class as a consequence of economic progress, whether or not such progress eventuated in a stationary state. 4. The classical economists had vague, unformed general equilibrium systems in mind, as products of supply and demand functions for both inputs and outputs. 5. Turning to macroeconomics: The classical economists espoused neither a mechanical form of the quantity theory of money nor Say’s Law in the ‘Identity’ sense which does not specify any equilibrium price level. 6. The classical economists were not ‘savage deflationists’. Given a wartime fiat-money inflation to which the economy had largely adjusted, restoration of a metallic standard should not have required return to the prewar prices of the precious metals. 7. The classical method of economic analysis is better described as an interaction and reinforcement between inductive and deductive methods than as pure deduction derived solely from introspection –Schumpeter’s ‘Ricardian vice’. Their use of historical examples was more than windowdressing, and they sometimes went out of their way to consider apparent paradoxes and anomalies. (Ricardo’s ‘misdemeanor’, falling short of ‘vice’, was to include so few historical illustrations in his Principles of Political Economy and Taxation. He included them plentifully in his other writings.) 8. In view of (1–7) above, the ‘marginal revolution’ of the 1870s was not in fact a revolution at all, but ‘nothing more than a change in “concentration of attention”’. 9. Rather than a combinatory reconciliation of classical and marginalist conceptions, Alfred Marshall’s Principles of Economics should be looked upon as a modernized restatement of classical economics – the economics of Ricardo, in particular. 10. Rather than ‘It’s all in Marshall’ – battle-cry of the pre-Keynesian ‘old fogeys’ – they should have insisted that ‘It’s all in Ricardo’, meaning by ‘Ricardo’ his essays on policy problems along with his Principles. 3

RICARDO - THE NEW VIEW

Professor Hollander has convinced me that a set of propositions much like the above decalogue – if not that decalogue as it stands – is basically correct. (Bronfenbrenner 1989, 36–8) As will be clear from the Tablets cited above, a principal theme running through The Economics of David Ricardo and Classical Economics – it is true of the essays reprinted here – is my contention that Ricardo’s economics constitutes – in its fundamentals, not merely tangentially or in some inconsequential way – the economics of allocation involving the information conveying and signalling role of prices, the notion of alternative costs, the principle of maximizing net returns and the market interaction of goods and services – in brief, the theory of the coordination of decentralized economic activities. This is what I refer to by the designation ‘general equilibrium’ in the Ricardo context. It is an unfortunate term if it conjures up Arrow, Debreu or Hahn; but I had presumed it would be self-evident that this was not my intention. I have also made it clear throughout that Ricardo’s main objective was not that of Walras, but rather to demonstrate the inverse wage–profit relation. 1 There were, moreover, contemporaries or near-contemporaries of Ricardo who developed aspects of ‘general equilibrium’ not clear to Ricardo himself – Longfield, in particular. 2 I have never claimed that Ricardo said the last word. There is a second theme – strictly it is this that seems to be usually intended by the designation ‘New View’ – namely, that the process of growth subject to land scarcity entails a necessarily downward trend in both the real (commodity) wage and the profit rate, the incidence of increasing land scarcity being shared between labour and capital in a ratio depending on capital-supply and labour-supply conditions. These are not independent exercises. To express the latter in value terms, and demonstrate the necessary rise in the proportionate share of the value of the marginal product going to labour despite the fall in the real wage, necessitated a preliminary investigation of value theory. The allocation dimension implies that there was no paradigmatic or ‘revolutionary’ break between Smithian and Ricardian theory, and also none in the early 1870s with the work of the so-called ‘marginalist revolutionaries’. (This is sometimes referred to as the Continuity Thesis.) The growth dimension implies the undermining of the notion of profit as ‘surplus’. Let me make myself perfectly clear. (1) I do not deny a concern with the source of investible funds – the disposable surplus; but that might be found in all incomes, wages included. Thus with decreased land scarcity – ‘an added tract of fertile land’ – real income and part of the disposable surplus would be transferred to labour and capital. (2) Profits could not be reduced, as by taxation, without an impact on the rate of accumulation, and, accordingly, could not be represented entirely as surplus in the sense of ‘economic rent’. In the context of the inverse profit–wage relation, both incomes are exactly on a par since the real wage is 4

INTRODUCTION

not a datum and governs the population growth rate, just as the profit rate governs that of capital. My Ricardo and Classical Economics contain much else including disquisitions on policy, money and banking and method; but it would be fair to say that the foregoing résumé covers the themes that have attracted most attention. (Rather a pity.) These themes are developed and defended in Parts III and IV. Professor Bronfenbrenner closes his review by asking: Is the Hollander rehabilitation of classical economics a success? I think that it is, both in presenting what classical economics was and in stressing how much of it remains in both modern Micro- and modern Macro-economics. It represents an advance over, say, Mark Blaug’s Ricardian Economics (New Haven, 1958) and Thomas Sowell’s Classical Economics Reconsidered (Princeton, 1974) primarily in its superior treatment of the ‘Cambridge critique’. It is too much to hope for this volume the status of last word on classical economics, but it surely defines a new orthodoxy and becomes the next target – not to be confused with a sitting duck – for dissidents to attack. (Bronfenbrenner 1989, 41) He has been proven correct in his forecast. One ‘Sraffian’ reviewer of Classical Economics complained that it was ‘a tragic book. It shows a reluctance on Hollander’s part to benefit from the criticism his earlier work so abundantly received’ (Groenewegen 1988, 89). This, of course, is academese for my unwillingness to abandon my position in favour of the reviewer’s. I leave it to Professor Blaug to come to my defence. He makes the point that ‘Sam Hollander is notorious for writing rejoinders to every adverse review of one of his books’ (Blaug 1987a, 19). Precisely. Some authors maintain a dignified silence in the face of criticism especially since – as Stigler has observed –scholars ‘seldom change their minds’ (1988, 210). For better or for worse, I have chosen a different line. I have listened carefully to the criticisms, found them wanting and given my reasons for standing firm. This will become clear not only in Part II which includes formal replies to critics of my Ricardo but throughout, since many of the papers entail responses. Nor can one be absolutely sure that there will be no converts; there is some evidence of weakening even on the part of my severest critic (see Chapter 17). 3 It is also my hope that students will benefit from observing, or at least be amused by, the antics of the contending parties. A particularly serious matter is raised by Professor Bronfenbrenner. If my position is correct, as he believes it is, ‘how ... could so much eminent and honest scholarship have gone so awry over the generations which separated modern from classical economics? How could so many great men in so many countries have so badly confused first approximation and “pedagogic clarities” with wrong-headedness, onesidedness, narrow-mindedness, and sheer stupidity?’ (1989, 38). I shall try to give a (partial) answer to that question presently, pointing out now that, pace Blaug’s rhetorical 5

RICARDO - THE NEW VIEW

assertion that, in my view, ‘absolutely everybody else has more or less misinterpreted Ricardo’, I have in fact done my best to give credit where credit is due for all or part of the so-called New View.4 And the fact that Professor Morishima (1989) reaches much the same perspective on Ricardo as I do, though by a very different path, provides a most pleasing corroboration. I also find my position confirmed and reinforced by my researches for the Economics of Thomas Robert Malthus (in press). There I show (1) that the simultaneous decline in wage and profit rates was adopted by Ricardo before Malthus took it up; (2) that Malthus set forth the corn profit model that figures so large in Part I of this selection, and developed a price theory that is compatible with Sraffa’s in his Production of Commodities (1960); (3) that Ricardo insisted against Malthus on treating corn precisely as any other product in terms of independent demand and supply functions; and (4) that Ricardo rejected the superior productivity attributed by Malthus to agriculture, a physiocratic concept appearing in the Wealth of Nations side by side with – from Ricardo’s perspective – a valid body of allocation theory. 5 I shall not spell out here the potentially radical implications for one’s perspective on the development of nineteenth-century economics. As George Stigler put it so disarmingly in his memoirs: ‘One surprising feature taught by intellectual history is the persistence of uncertainty over what a person really meant. One might think that intellectual competence and goodwill are all that are required to understand what a scholar intends to say, but the study of any important scholar of the past will show that belief to be most naive’ (1988, 216). Here Stigler apparently takes for granted that our concern should be with ‘what a scholar intends to say’ though elsewhere he relegates this to a secondary category of ‘personal’ rather than ‘scientific exegesis’ (below, Chapter 15). I wholeheartedly share this objective, and would regret the infiltration of ‘literary theory’ of the deconstructionist variety into our discipline. It is the rejection of ‘authorial intent’ as a valid concern which I cannot appreciate, though fortunately self-declared literary theorists do not always practise what they preach. (I certainly do not mean by all this that one should ignore matters of style and ‘rhetoric’; on the contrary, these can provide the clue to the author’s intentions.) This leads me back to Professor Bronfenbrenner’s question – if my position on Ricardo is correct, how to explain honest misreadings? It is not, I would say, strictly a matter of misunderstanding that has emerged ‘over the generations which separated modern from classical economics’. Ricardo was misunderstood from the outset and himself vigorously protested at some of the worst cases, at one point uncharacteristically exclaiming against Malthus: ‘This is disingenuous.’ If Ricardo himself had problems making himself understood to his own friends, any later commentators are going to have their work cut out to get things straight. 6

INTRODUCTION

I illustrate from Malthus’s reading of Ricardo on profits: ‘to attempt to estimate the rate of profits in any country by a reference to this cause alone [diminishing agricultural returns], for ten, twenty, or even fifty years together, that is for periods of sufficient length to produce the most important effects on national prosperity, would inevitably lead to the greatest practical errors. Yet notwithstanding the utter inadequacy of this single cause to account for existing phenomena, Mr. Ricardo in his very ingenious chapter on profits, has dwelt on no other’ (1820, 308); or again, the facts were ‘diametrically opposed to the theory of profits founded on the natural quality of the last land taken into cultivation’ (320). Ricardo protested: ‘Mr. Malthus here brings a charge against me which he would find it very difficult to prove. He has himself Page 294 Section I of this Chapter stated his causes for the fall of profits’ – agricultural productivity and variable corn wages. ‘I fully concur with him in thinking that profits never vary but from one or other of those causes’ (1951, II, 264). As for the second formulation, he could scarcely contain himself: This is disingenuous. Who has advanced a ‘theory of profits founded on the natural quality of the last land taken into cultivation’. The theory is that profits depend on the productiveness of the last land taken into cultivation, whether that productiveness be owing to the natural quality of the land, or the economy and skill with which labour may be applied to it. Profits are increased, either by diminishing the quantity of labour bestowed on the last land which yields a given produce, or by increasing the produce with a given quantity of labour. Mr. Malthus will I am sure not say that I have ever denied this principle – he will not say that I have not distinctly advanced it. (Ricardo 1951, II, 276–7) A second instance of ‘misrepresentation’ relates to value theory: ‘[W]hen you reject the consideration of demand and supply in the price of commodities and refer only to the means of supply, you appear to me to look only at the half of your subject. No wealth can exist unless the demand, or the estimation in which the commodity is held exceeds the cost of production: and with regard to a vast mass of commodities does not the demand actually determine the cost?’ (Malthus to Ricardo, 26 October 1820; in Ricardo 1951, VIII, 285). Ricardo immediately protested: ‘I have never disputed this’ (24 November; VIII, 302). And again he explained his position, in this case the weight placed on production costs as the determinant of supply conditions in the face of changes in demand: ‘I do not dispute either the influence of demand on the price of corn and on the price of other things, but supply follows close on its heels, and soon takes the power of regulating price in his own hands, and in regulating it he is determined by cost of production. I acknowledge 7

RICARDO - THE NEW VIEW

the intervals on which you so exclusively dwell, but still they are only intervals.’ All this was in line with his Principles, where Ricardo had spelled out the process of longrun cost determination of the corn price assuming increases of demand in the face both of constant-cost and increasing-cost conditions (I, 163). 6 Part of the blame for the propensity to misunderstand Ricardo reflects the linguistic complexity, though I find that one quickly becomes accustomed to Ricardo’s specialized terminology and learns to translate. A focus on his strong cases – including forms of expression designed to convey the ‘primacy’ of supply in value formation (see Chapter 12) – explains some of the error, though again I find the Principles itself to be clear on the main lines of my interpretation with respect to allocation and growth and the interdependency between them. As Jacob Viner put it, a concentration on the long run and a tendency to omit ‘explicit mention of qualifications whose validity he was prepared to acknowledge’ enabled critics ‘to expose him to rebuttal often more damaging in appearance than in fact’ (1937, 140). 7 All this is further compounded by a tendency to take the strong statements out of context, encouraging a confusion between ‘rational’ and ‘historical’ reconstruction. There is also the matter of technical error. For example, at one point in a recent account of Ricardo on policy, Mark Blaug attempts an ecumenical exercise suggesting that while Pasinetti – who develops the constant subsistence wage growth interpretation: is unable to account for those passages in which Ricardo more or less clearly says that real wages, expressed in terms of a basket of physical commodities, can fall alongside the falling rate of profit well before the economy has reached the stationary state ... even the ‘new’ view has difficulty in making sense of passages in which Ricardo declares unambiguously that the rate of profit depends only on the cost of producing wage goods, and on nothing else; such passages are easy to interpret, however, if we stick with the Pasinetti version of Ricardo .... It is not possible, therefore, to square everything that Ricardo said with any totally consistent formulation of the entire Ricardian system. (Blaug 1987b, 121). Here Blaug repeats a perennial error. The New View can and does accommodate Ricardo’s ‘unambiguous’ statement, since the falling real wage is part and parcel of that same process which drives down the profit rate, and the profit rate falls precisely because the cost of producing the basket necessarily rises albeit that the magnitude of the basket declines. Blaug is not alone. Even Stigler was not immune (see Chapter 15, Chapter 16). It was Ricardo, as mentioned already, who taught Malthus that the falling real wage cannot prevent the fall in the profit rate, but the lesson has not yet been absorbed. 8

INTRODUCTION

Misinterpretation of Ricardo has, I believe, also been compounded by a presumption that, to avoid anachronistic readings, the researcher must pretend to lack knowledge of the future beyond the period under study. Of course we must never superimpose on an early writer reference frameworks that were developed only after his death or even after a particular moment in his career (Skinner 1969, 6). But this does not require that we avoid modern vocabulary and categories. One is writing, after all, for modern readers. More important, there is also a danger of denying the presence in an early writer’s work of modern concepts merely because they are expressed differently – a sophisticated form of anachronism. Tracing the filiation of ideas may actually be impeded by a pretence of ignorance, a pretence implicit in a complaint that the debate over Ricardo has ‘taken on all the characteristics of a bitterly contested paternity suit, with one side battling for Ricardo’s custody within the “neo-classical” home, the other for his adoption within the “Sraffian” home. My contribution to this debate is simple. Leave him within his own surroundings’ (Peach 1993, 303). 8 For my part, I disclaim any attempt to capture Ricardo for the neo-classicals in any sense other than that specified above – the fact that the price-signalling mechanism of resource allocation is central, and explicitly so, to Ricardo’s economics. If one were to imagine that the intellectual clock stopped on 11 September 1823 with Ricardo’s last breath – that there were no more advances beyond his oeuvre – those features would still be there and recognizable. That they are absorbed into the ‘neo-classical’ paradigm is another historical fact attested to by Marshall. 9 The recommendation to leave Ricardo ‘within his own surroundings’, if it is intended to convey anything more than the valid Skinner caution, is likely to blind one to perspectives actually extant and glaringly so in the Ricardo texts. In any event, why the surprise at my reading of Ricardo? The allocative function of the price system is already clearly present in Turgot. 10 And, of course, we must not forget Adam Smith. 11 There is a necessary, though insufficient, test for accurate interpretation. Any reading which captures the author’s main theoretical conclusions inevitably leaves formulations conflicting with that interpretation. The choice between readings will turn in part on how convincingly these residuals are treated. My efforts to lay bare the residual features in my own interpretation – to state frankly the apparent weaknesses of my own interpretation – have given rise to charges of ‘flights of negative imagination’. I suggest that those who oppose the New View should try to deal with their own residuals – those created by insistence that Ricardo’s economics excluded (1) a powerful price-theoretic or scarcity dimension, and/or (2) market processes in the determination of the wage rate trend in a growing system subject to land scarcity. 12 Examples of the burial of residuals range far beyond the forementioned matters. One striking case relates to Ricardo’s concern that corn law reform had to be carefully timed and slowly phased in: 9

RICARDO - THE NEW VIEW

We all have to lament the present distressed situation of the labouring classes in this country, but the remedy is not very apparent to me. The correcting of our errors in legislation with regard to trade would ultimately be of considerable service to all classes of the community, but it would afford no immediate relief: on the contrary I should expect that it would plunge us into additional difficulties. If all the prohibitions were removed from the importation of corn and many other articles, the sudden fall in the price of corn and those other articles, which could not fail to follow, would ruin most of the farmers, and many of the manufacturers; and although others would be benefited, the derangement which such measures would occasion in the actual employment of capital, and the changes which would become necessary, would rather aggravate than relieve the distress under which we are now labouring. (13 October 1819; Ricardo 1951, VIII, 103) Clearly Ricardo wished to avoid any further disturbances that would magnify the need for reallocation of resources, and therefore ruled out corn law repeal as a helpful step. He called for legislation only in 1821 after the return of prosperity in the manufacturing sector. Moreover, repeal – which was not to be total in any event – should be brought about by gradual steps and after due warning with recognition of the possible need for compensation of those (other than landlords) adversely affected. The foregoing passage is cited by Professor Hutchison as evidence of ‘great wisdom ... in the realization of ignorance and the scepticism regarding conclusions deduced from long-run, rapidly self-equilibrating models’ (1978, 49–50n). But it is read as an exception – the exception that proved the rule: ‘It amounts to an extreme contrast with the more typical kinds of Millian–Ricardian policy analysis.’ It is not at all clear to me why at this particular point, and nowhere else, Ricardo should have shown such wisdom, and Hutchison offers no suggestions. I find Ricardo similarly responsible on the return to gold, poor law reform and – this I owe to my doctoral student Nancy Churchman – on the national debt. Had Hutchison bothered to ponder the ‘exception’, he might have been led to reconsider his misleading representations of Ricardian method. Let me address next a common complaint directed against my work: Much of the vaunted ‘interdependence of pricing and distribution’ is, by Hollander’s admission, not explicit in ‘classical analysis’. ... It is only by retrospectively filling this lacuna, asking, for example, how a change in tastes would have been analysed had the ‘classical’ writers been compelled to use ‘process analysis’ that Hollander’s arguments acquire their thin glaze of plausibility. To use the phrase beloved by him, this strategy is a fine example of ‘altered concentrations of attention’, where the shift is away from the 10

INTRODUCTION

textual preoccupations and methods of analysis to those supplied and adapted by Hollander. (Peach 1988, 175) The criticism merits serious consideration. In the course of my work, I point out that certain disturbances may not formally have been dealt with by the original writer, but once having established that writer’s analysis of similar cases it is surely fair to say how, if presented with such a disturbance, he would have been obliged to deal with it consistently with the system already established. I see nothing patently unreasonable about this procedure which is designed to recognize the status of various propositions. In any event, the specific illustration is poorly chosen since – as I subsequently became aware – Ricardo did, in fact, explicitly consider the impact of a change in tastes on distribution precisely in the manner to be expected, thereby confirming my hypothesis (see Chapter 11). I turn briefly to the general matter of ‘bias’ in interpretation. Bias, of course, need not be political; it extends to one’s personal sympathy for an author, and may lead to the discounting of features that less sympathetic readers put down as contradictory or meaningless. No historian is immune and the best one can do is recognize the inevitable degree of subjectivity and try to compensate. In my own case, I admit to a fondness for Ricardo as a person; his candour, incorruptibility and sense of public duty shine through so much of his writing. Moreover, the two years that I spent in a theological academy as a teenager reinforced an already-present concern with consistency rigorously instilled since about age 7. I am aware of these ‘biases’. But both can serve well by encouraging recourse to the charge of inconsistency as a last not first resort. Ideological bias in Ricardo interpretation is too big an issue to be dealt with here. I refer to the problem in Chapter 12 with respect to the Sraffian reading and there I refer to Bronfenbrenner’s hypothesis – it appears in the review cited above – that Sraffa may have confused Ricardo with Marx. But I also cannot help feeling that Hutchison – the Sraffian’s bête noire –attributed to Ricardo the sins of Stalin and that this perhaps somewhat coloured his interpretation (1978, 240–76). One needs to stand back from one’s subject to allow for a minimum of objectivity. I shall leave these matters to a later volume in this series. The essays printed below constitute contributions to the journals on Ricardo and ‘Ricardian economics’ which touch directly or indirectly on the ‘New View’. That the whole issue will not lie down is clear from the fact that the ink is scarcely dry on quite a few of the articles and the criticisms to which they respond. The preceding remarks have, I trust, placed Parts I through IV in perspective. Part V is a catch-all with a common theme in a concern with intellectual linkages, supplementing in part the evidence for the ‘continuity thesis’ in Chapter 9 and Chapter 10. 11

RICARDO - THE NEW VIEW

NOTES 1. For this purpose the ‘exclusion’ of rent from marginal costs by the differential rent theory applied to land was crucial, a procedure that would be questionable in the case of multi-use land (see Chapter 12). In several of my papers, I have maintained that Ricardo appreciated the distinction (see Chapter 9, Chapter 10). I now think this goes too far; he on occasion extended the principle to the multi-use case (see 1951, I, 250; II, 70–1). 2. See R.D.C. Black (1984). This paper is a model of fair-minded criticism. 3. I myself once taught the ‘old view’ (below, pp. 24, 29-30, 35-6). Sir John Hicks was seduced away from the fix-wage interpretations (see the exchange between J.H. and S.H. prefacing Hicks and Hollander 1977). And the reviewer of Classical Economics for History of Political Economy recounts his personal odyssey (Young 1991). 4. My list would include inter alia Carlo Casarosa, David Levy, Pier Luigi Porta, S.C. Rankin, Lord Robbins, P.A. Samuelson, G.L.S. Tucker and Jacob Viner. And, of course, there is Alfred Marshall. I also might remind readers of at least two other commentators who have stated the so-called ‘continuity thesis’ much more strongly than I have done: ‘economics has never had a major revolution: its basic maximizing model has never been replaced. Whether this is good or bad (and the fact is the despair of some), it is, I think, remarkable when compared to the physical sciences that an economist’s fundamental way of viewing the world has remained unchanged since the eighteenth century’ (Gordon 1965, 124); ‘economics may be regarded as more “uniformitarian” than the natural sciences, for despite persistent and often penetrating criticism by a stream of heterodox writers ... it has been dominated throughout its history by a single paradigm—the theory of economic equilibrium via the market mechanism’ (Coats 1969, 292). 5. The presence in Malthus’s writings of both concepts repeats the confused picture of 1776 with the principal difference that the notion of agricultural surplus contributed towards Malthus’s support for agricultural protection—as it had led the French to recommend free trade in agricultural products when France was a net corn exporter— whereas Smith never went that far. 6. A third instance of ‘misrepresentation’ that sorely troubled Ricardo relates to his chapter ‘On Gross and Net Revenue’. Malthus read this chapter as arguing that a national advantage resulted from obtaining a given net income with a lower work force notwithstanding a massive loss of employment. Ricardo protested that his sole objective was to argue that no national advantage flowed from generating a given net revenue by means of seven rather than five million men, having in mind specifically the magnitude of net revenue with an eye to national ‘power’: ‘The employing a greater number of men would enable us neither to add a man to our army and navy, nor to contribute one guinea more in taxes’ (1951, II, 381). He had no intention of recommending a policy involving actual creation of massive unemployment. 7. Comparing James Mill’s Elements with Ricardo’s Principles, McCulloch rightly pointed out that what is valid procedure in the one case may be invalid in the other: This work, by the distinguished author of the ‘History of British India’ is a résumé of the doctrines of Smith and of Ricardo with respect to the production and distribution of wealth, and of those of Malthus with respect to population. But it is of too abstract a character to be either popular or of much utility. Those secondary principles and modifying circumstances, which exert so powerful an influence over general principles, are wholly, or

12

INTRODUCTION almost wholly, overlooked by Mill. But though their consideration might be omitted in an original work like that of Ricardo, it is not so easily excused in an elementary treatise. (McCulloch 1845, 17)

8. Blaug cites these words with utter rapture (1993, 2). But I do not take this commendation from the author of Economic Theory in Retrospect seriously. Almost at random I pick out from that work the following interpretation of Smith’s Book 1, chapter vii on price formation: ‘It will be noticed that Smith thinks of demand and supply as referring to people’s willingness to buy or sell at a particular price rather than at all possible prices; the former is expressed in actual amounts desired or offered, the latter in a schedule of amounts, each corresponding to a different price. Still, the whole of the passage given above has no real meaning unless demand at any rate is interpreted in the sense of a schedule, and a negatively inclined schedule at that. Here and elsewhere Smith intuitively gropes his way toward the right answer’ (1985, 43). 9. If one wishes to call this ‘Whig history’ (Samuelson 1988, 1992) there is no harm, since Whig history—if conducted responsibly—does not read into the record what is not there, but seeks to provide criteria for evaluation. But the Whig historian can go too far. For example, Samuelson: ‘If land is unaugmentable and permanent, and labor supply is procurable at a subsistence wage, then there is a genuine sense in which Quesnay’s Produit Net (and Ricardo’s Neat Product) is equal to Georgian land rent: the rest of national Gross Product is the wage fodder of labor’ (1988, 164). This is a valid insight as a positive statement regarding analysis; but it is not Quesnay. I touch on the contrast between ‘rational’ and ‘historical’ reconstruction. 10. See Schumpeter on Turgot’s Reflections as ‘first of all the treatises on Value and Distribution that were to become so popular in the later decades of the nineteenth century’ (1954, 249). This opinion is confirmed by Groenewegen (1970) with parallels drawn between Turgot’s reasoning and that of Walras and Wicksell in the development of exchange models. Groenewegen is apparently ashamed of the sins of his youth. For he makes no mention of this paper in his contribution to the New Palgrave, where he rejects Schumpeter’s view that Turgot anticipated much of what became important after the ‘marginal revolution’, claiming to the contrary that Turgot is ‘firmly in the classical tradition rehabilitated by Sraffa’ (1987, 711). This is questionable, considering Turgot’s explicit account of the principle of competition assuring that market prices tend to cost prices via the adjustments of supplies ( Reflections, Section LXXXVIII). Smithian and Ricardian, yes; Sraffian, no. Groenewegen himself points out that Turgot’s value theory ‘rested on a relationship between current (market) price and fundamental value dependent on competition and resource mobility’, which relationship is characteristically non- or anti-Sraffa. 11. I am delighted that Paul Samuelson comes to the defence of my position, which ‘has been lampooned rather than praised for seeing supply and demand mechanisms everywhere in the classical writers. Even Mark Blaug ... writes of Hollander as making Smith into more of a Léon Walras than a Ricardo, which I deem not to be a reductio ad absurdum, but rather a merited compliment for Smith’ (1992, 5). I note the unfortunate implication here that Ricardo, for Samuelson, is not in the Smithian tradition with respect to general equilibrium; but there is also his compensatory observation that ‘[e]ven before the 1951 and 1960 eruption into print of Sraffa, my reading of every one of the classical masters found in them an earlier and more primitive version of the Walras–Marshall–Wicksell system’ (5–6; emphasis added). Powerful new evidence for Adam Smith as anticipator of general-equilibrium modelling will be found in West, 1994. 13

RICARDO - THE NEW VIEW

12. Some commentators have just given up and concluded that Ricardo is so riddled with ‘startling inconsistencies’ that he is not susceptible to coherent formulation on value or distribution (Peach 1993, xii) .

REFERENCES Black, R.D.C. (1984) ‘The Irish Dissenters and Nineteenth-century Political Economy’, in Antoin E. Murphy (ed.) Economists and the Irish Economy, Dublin: Irish Academic Press, 120–37. Blaug, Mark (1985) Economic Theory in Retrospect, 4th edn, Cambridge: Cambridge University Press. ——(1987a) ‘On a Severe Case of Paranoia in the History of Economic Thought’,History of Economic Thought Newsletter 38 (Spring): 19–21. ——(1987b) ‘Ricardo and the Problem of Public Policy’, inEconomic History and the History of Economics, New York: New York University Press, 115–27. History of Economic Thought Newsletter 50 (Spring): 1–2. ——(1993) Review of Peach, 1993, Bronfenbrenner, Martin (1989) ‘A Rehabilitation of Classical Economics’, Aoyama Kokusai Seikei Ronshu [Aoyama University Journal of International Political Economy] 13: 35–41. Coats, A.W. (1969) ‘Is there a “Structure of Scientific Revolutions” in Economics?’ Kyklos 22: 289–96. Fry, Stephen (1993) ‘Trefusis’s Obituary’, Paperweight, London: Mandarin, 43–5. Gordon, D.F. (1965) ‘The Role of the History of Economic Thought in the Understanding of Modern Economic Theory’, American Economic Review 55 (May): 119–27. Groenewegen, Peter D. (1970) ‘A Reappraisal of Turgot’s Theory of Value, Exchange, and Price Determination’, History of Political Economy 2 (Spring): 177–96. ——(1987) ‘Turgot, Anne Robert Jacques, Baron de L’Aulne’, in J. Eatwell, M. Milgate and P. Newman (eds) The New Palgrave Dictionary of Economics, Vol. 4, London: Macmillan, 707– 12. ——(1988) Review of Hollander, 1987,Economic Record 28 (September): 88–9. Hicks, John and Hollander, Samuel (1977) ‘Mr Ricardo and the Moderns’, Quarterly Journal of Economics 91 (August): 351–69. Hollander, Samuel (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. Classical Economics, Toronto: University of Toronto Press, 1992. ——(1987) Cambridge: Hutchison, T.W. (1978) On Revolutions and Progress in Economic Knowledge, Cambridge University Press. Malthus, T.R. (1820) Principles of Political Economy, London: John Murray. McCulloch, J.R. (1845) The Literature of Political Economy, London: Longman, Brown, Green, and Longmans. Morishima, Michio (1989) Ricardo’s Economics: A General Equilibrium Theory of Distribution and Growth, Cambridge: Cambridge University Press. Peach, Terry (1988) ‘Samuel Hollander’s Classical Economics: A Review Article’, The Manchester School 56 (June): 167–75. Interpreting Ricardo, Cambridge: Cambridge University Press. ——(1993) Ricardo, David (1951–73) The Works and Correspondence of David Ricardo ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. 14

INTRODUCTION

Samuelson, Paul A. (1988) ‘Keeping Whig History Honest’, History of Economics Society Bulletin 10 (Fall): 161–8. ——(1992) ‘The Overdue Recovery of Adam Smith’s Reputation as an Economic Theorist’, in Michael Fry (ed.) Adam Smith’s Legacy: His Place in the Development of Modern Economics, London: Routledge: 1–14. Schumpeter, J.A. (1954) History of Economic Analysis, New York: Oxford University Press. Skinner, Quentin (1969) ‘Meaning and Understanding in the History of Ideas’, History and Theory 8, 3–33. Cambridge: Cambridge Sraffa, Piero (1960) Production of Commodities by Means of Commodities, University Press. Stigler, George J. (1988) Memoirs of an Unregulated Economist, New York: Basic Books. Viner, Jacob (1937) Studies in the Theory of International Trade, New York: Harper and Brothers. West, Edwin G. (1994) ‘Joint Supply Theory before Mill’, History of Political Economy 26 (Summer): 267–78. Young, Jeffrey T. (1991) Review of Hollander, 1987, History of Political Economy 23 (Spring): 167–70.

15

Part I ON THE INTERPRETATION OF THE EARLY RICARDO

1 RICARDO’S ANALYSIS OF THE PROFIT RATE, 1813–15

This essay is concerned with the early account (1813–15) provided by Ricardo of the principles governing the general rate of profit, the characteristic feature of which is defined in the celebrated proposition that ‘it is the profits of the farmer which regulate the profits of all other trades’. I shall attempt to justify the contention that the widely held explanatory hypothesis suggested by Piero Sraffa (cf. 1951, xxxi–xxxii) – according to which Ricardo’s proposition derives from the assumption that in agriculture both input and output consist of the same commodity (‘corn’) – does not in fact accurately reflect Ricardo’s intentions. It follows from the argument of this paper that substantially the same position as that ultimately appearing in the Principles was maintained from the very outset, namely that variations in the money-wage rate, in consequence of changing prices of wage goods, will be accompanied by inverse movements in the general rate of profit. We consider below the correspondence between Ricardo and Malthus on the rate of profit during the period 1813–14, and the Essay on Profits (1815). I. THE THEORY OF PROFITS (1813–14) The explanatory principle adopted in the Principles to account for the determination and secular movement of the general rate of profit runs in terms of the real cost of producing wage goods: ‘It has been my endeavour to shew throughout this work, that the rate of profits can never be increased but by a fall in wages, and that there can be no permanent fall of wages but in consequence of a fall of the necessaries on which wages are expended’ (Ricardo 1951, I, 132). This fundamental relationship can in fact be found as early as mid-1813 in correspondence with Malthus, wherein Ricardo rejected Malthus’s contention that either extensions of foreign trade, 1 or capital accumulation as such would have any effect upon the rate of profits. Attention rather should be directed at productivity in agriculture: That we have experienced a great increase of wealth and prosperity since the commencement of the war I am amongst the foremost to believe; but it is 19

RICARDO - THE NEW VIEW

not certain that such increase must have been attended by increased profits, or rather an increased rate of profits, for that is the question between us. I have little doubt however that for a long period, during the interval you mention [1793–1813], there has been an increased rate of profits, but it has been accompanied with such decided improvements of agriculture both here and abroad, – for the French revolution was exceedingly favorable to the increased production of food, that it is perfectly reconcileable to my theory. My conclusion is that there has been a rapid increase of Capital which has been prevented from shewing 2 itself in a low rate of interest by new facilities in the production of food. (17 August 1813; Ricardo 1951, VI, 94–5; emphasis added) Implicit in the argument is the presumption that but for improved agricultural technology during the Revolutionary and Napoleonic period the profit rate would have declined with accumulation. 3 In the following year Ricardo defined the conflict between himself and Malthus on the determination of the profit rate with particular care. All influences apart from the productivity of agricultural resources are again excluded, and the key role of the principle of diminishing returns in conditions of constant agricultural technology carefully defined. What must be noted above all is the formal proposition that ‘it is the profits of the farmer which regulate the profits of all other trades’: I contend that the arena for the employment of new Capital cannot increase in any country in the same or greater proportion than the Capital itself – unless Capital be withdrawn from the land – unless there be improvements in husbandry, – or new facilities be offered for the introduction of food from foreign countries; – that in short it is the profits of the farmer which regulate the profits of all other trades, – and as the profits of the farmer must necessarily decrease with every Augmentation of Capital employed on the land, provided no improvements be at the same time made in husbandry, all other profits must diminish and therefore the rate of interest must fall. (To Hutches Trower, 8 March 1814; Ricardo 1951, VI, 103–4; emphasis added) Malthus’s position – implying a mutual relationship between sectors – is then given by contrast: To this proposition Mr. Malthus does not agree. He thinks that the arena for the employment of Capital may increase, and consequently profits and interest may rise, altho’ there should be no new facilities, either by importation, or improved tillage, for the production of food; – that the profits of the farmer no more regulate the profits of other trades, than the profits of other trades regulate the 20

RICARDO’S ANALYSIS OF THE PROFIT RATE

profits of the farmer, and consequently if new markets are discovered, in which we can obtain a greater quantity of foreign commodities in exchange for our commodities, than before the discovery of such markets, profits will increase and interest will rise .... Nothing, I say, can increase the profits permanently on trade, with the same or an increased Capital, but a really cheaper mode of obtaining food. A cheaper mode of obtaining food will undoubtedly increase profits says Mr. Malthus but there are many other circumstances which may also increase profits with an increase of Capital. The discovery of a new market where there will be a great demand for our manufactures is one. (Ricardo 1951, VI, 104–5; emphasis added) 4 In a letter to Malthus himself, 26 June 1814, Ricardo wrote that ‘the rate of profits and of interest must depend on the proportion of production to the consumption necessary to such production’ (VI, 108). It is this particular statement which, Sraffa suggests, implies a model yielding the determining role of agricultural profits by assuming that agricultural capital and output both consist of the same physical commodity, namely corn, so that total profits and the rate of profit on capital can be determined in physical terms without reference to valuation. Since all other sectors utilize corn as input only it is the exchangeable value of their products which must be adjusted to yield the same profit rate as in agriculture. Ricardo, however, himself stated in this letter what he had in mind, and his explanation does not relate to the preceding argument. Whatever the surface manifestation, the ultimate determinant of the general rate of profit was, in his view, the level of money wages, governed in turn by food prices, so that the secular behaviour of the profit rate could be explained by inverse movements of money wages. The ‘proportion of production to the consumption’ referred to above, he explained, ‘essentially depends upon the cheapness of provisions, which is after all, whatever intervals we may be willing to allow, the great regulator of the wages of labour’. (Logically, the effect of import restrictions upon the profit rate would depend upon whether or not the commodities involved entered into the wage basket.) It must be carefully noted that the argument is stated in objection to Malthus’s view that the effects of trade restriction would be a reduction in aggregate capital and a consequent increase in the rate of profits. 5 This is not the occasion to enter into a detailed account of the Malthusian theory of capital accumulation and of ‘effectual demand’ which envisaged a potential divergence between aggregate demand and supply, but the full import of Ricardo’s treatment of the rate of profit cannot be appreciated unless the position of his opponent is kept in mind. Ricardo insisted in his letter upon the identity of aggregate demand and supply: ‘Demand has no other limits but the want of power of paying for the commodities 21

RICARDO - THE NEW VIEW

demanded. Every thing which tends to diminish production tends to diminish this power’ (VI, 108). To Malthus’s formal criticism, several weeks later, of James Mill’s ‘ingenious position which he lays down in his answer to Mr. Spence, that in reference to a nation, supply can never exceed demand’, 6 Ricardo replied: ‘The desire of accumulation will occasion demand just as effectually as a desire to consume, it will only change the objects on which the demand will exercise itself ..... We all wish to add to our enjoyments or to our power. Consumption adds to our enjoyments, – accumulation to our power, and they equally promote demand.’ 7 While Malthus presumed that corn restriction must have the effect of reducing aggregate capital, Ricardo for his part, saw no necessary connection. 8 On the contrary, the effect of corn restriction would be much the same as that of capital accumulation since extensions to increasingly inferior land would be required in both cases. Accordingly, Ricardo directed his main attack against the Malthusian theory of effective demand, whereby the profit rate would be influenced by capital accumulation because of a relative variation in aggregate demand. Ricardo was willing to concede at this stage of the correspondence only a decline in the level of activity – although not in aggregate capital – both supply and demand falling proportionately. The reason offered is worth consideration since he was to alter his position subsequently in a fundamental sense. The rise in the price of corn upon restriction of importation would tend to reduce the demand for manufactured goods in the sense both of a shift of the demand curve and a movement along the curve: ‘The rise of the price or rather the value of corn ... must necessarily diminish the demand for other things even if the prices of those commodities did not rise with the price of corn, which they would (tho’ slowly) certainly do. With the same Capital there would be less production, and less demand.’ 9 In his next letter to Malthus (25 July) Ricardo put even greater emphasis upon reduced demand for manufactured goods due to the rise in their cost prices : You say that ‘the proportion of production to the consumption necessary to such production, seems to be determined by the quantity of accumulated capital compared with the demand for the products of capital, and not by the mere difficulty and expence of procuring corn’. It appears to me that the difficulty and expence of procuring corn will necessarily regulate the demand for the products of capital, for the demand must essentially depend on the price at which they can be afforded, and the prices of all commodities must increase if the price of corn be increased. (Ricardo 1951, VI, 114; emphasis added) Let us take stock of Ricardo’s argument. The fundamental proposition is that the general profit rate tends to fall upon restriction of corn imports because of the effect of high corn prices upon money wages. At the same time Ricardo’s attempt to deal 22

RICARDO’S ANALYSIS OF THE PROFIT RATE

with Malthus’s contentions leads him to adopt the view that the prices of manufactured products would also rise. He did not, evidently, realize the implications of his concession that higher wage costs might be passed on to consumers in the form of higher prices. 10 The problem, once it was recognized, came to represent a serious stumbling block, but at this stage, Ricardo’s fundamental objection to Smith’s analysis whereby higher wages result in higher prices of manufacturers (and lower rents) was not yet formalized. 11 We turn next to a second item of evidence offered in support of the ‘corn model’ as the fundamental theoretical construct utilized by Ricardo. In a further letter to Ricardo, Malthus objected: ‘In no case of production, is the produce exactly of the same nature as the capital advanced. Consequently we can never properly refer to a material rate of produce, independent of demand, and of the abundance or scarcity of capital.’ 12 This formulation, it is said, suggests that Ricardo must have stated the contrary argument either in certain ‘lost “papers on the profits of Capital” ’ or in conversation since we find nothing in any of his extant letters and papers (Sraffa 1951, xxxi–xxxii). But Ricardo’s argument is perhaps in fact extant. It may well be contained in the letter of 25 July, just discussed, upon which Malthus was commenting. After explaining the expected reduced demand for manufactures as a consequence of rising corn prices, Ricardo restated his own position on profit–rate determination in terms of a corn surplus relative to a corn capital: The capitalist ‘who may find it necessary to employ a hundred days labour instead of fifty in order to produce a certain quantity of corn’ cannot retain the same share for himself unless the labourers who are employed for a hundred days will be satisfied with the same quantity of corn for their subsistence that the labourers employed for fifty had before. If you suppose the price of corn doubled, the capital to be employed estimated in money will probably be also nearly doubled, –or at any rate will be greatly augmented and if his monied income is to arise from the sale of the corn which remains to him after defraying the charges of production how is it possible to conceive that the rate of his profits – will not be diminished. (Ricardo 1951, VI, 114–15; emphasis added)13 What weight are we to place upon the corn calculation in this letter of July, or the estimate in terms of ‘a material rate of produce’ to use Malthus’s terminology? In the first place, we have already seen that in Ricardo’s earlier letter to Malthus of 26 June, the decline of the general profit rate was attributed to upward pressure on money wages due to the rising costs of wage goods. But secondly, we draw attention to a further letter to Malthus in August, wherein Ricardo clarifies significantly the 23

RICARDO - THE NEW VIEW

operation of this mechanism in the manufacturing sector. The manufacturer must pay increased money wages, while the price of his product does not rise proportionately: It is true that the Woollen or Cotton manufacturer will not be able to work up the same quantity of goods with the same capital if he is obliged to pay more for the labour which he employs, but his profits will depend on the price at which goods when manufactured will sell. [But] ... the rise of his goods will not be in the same proportion as the rise of labour, and consequently his percentage of profit will be diminished if he values his capital, which he must do, in money at the increased value to which all goods would rise in consequence of the rise of the wages of labour. (11 August 1814; Ricardo 1951, VI, 119–20; emphasis added) 14 In light of the consistent emphasis upon the role of the money wage rate in determining the general rate of profit – both immediately before and after the letter containing the corn calculation – it appears unjustified to interpret Ricardo’s intentions in terms of a corn model. Ricardo’s formulation of July represents a rather casual and inadequate restatement – the significance of which should not be exaggerated – of his basic and consistently-maintained position. Moreover, it appears to be the case that when Malthus criticized Ricardo on grounds that ‘in no case of production, is the produce exactly of the same nature as the capital advanced’ he was objecting to Ricardo’s refusal to accept his own ill-defined theory of effective demand, that is Ricardo’s tendency – in Malthus’s terms – ‘to underrate the wants and tastes of mankind in affecting prices, and consequently in affecting the means of profitably employing capital’ (see note 7). We should read no more into the criticism. Ricardo was, however, willing to make the revealing concession that his difference with Malthus regarding corn restriction was fundamentally ‘about the permanence of the effects’ (VI, 128).15 In other words, his insistence upon the governing role of agricultural productivity implied full operation of the population mechanism which assured constancy of real wages per head, while during an initial period average real wages might indeed decline and permit an increased rate of profit: ‘Profits are sometimes high when corn is scarce and dear, but this arises from the stimulus which the high price gives to industry. If the population could immediately accommodate itself to the scanty supply no such effects would follow; and in fact they only continue till time has gradually equalized them’ (VI, 129). We must keep this in mind in interpreting the subsequent comment of Ricardo that ‘the state of the cultivation of the land is almost the only great permanent cause. ... The state of production from the land compared with the means necessary to make it produce operates on all [trades], and is alone lasting in its effects.’ 16 24

RICARDO’S ANALYSIS OF THE PROFIT RATE

We introduce now a letter of October 1814 which brings together several strands of Ricardo’s argument – in terms very similar to those used in the Principles itself – defines Ricardo’s implicit assumptions, and isolates the remaining difficulties, particularly those relating to ‘value’. In the letter in question Ricardo analysed the effects on the profit rate of four possible causes of rising agricultural prices, namely capital accumulation, bad seasons, general price inflation, and import restrictions, allowing at the same time that ‘there may be other causes of high price which do not at present occur to me’. 17 In the main case – the effects of capital accumulation – we should note two fundamental features of the argument. First, the increase in corn prices reflects the reduction in agricultural productivity in the long run only, after population expansion has been allowed for. Secondly, the increase in price is not itself induced, or even reinforced, by higher money wages: ‘A rise in the price of raw produce may be occasioned by a gradual accumulation of capital which by creating new demands for labour may give a stimulus to population and consequently promote the cultivation or improvement of inferior lands, – but this will not cause profits to rise but to fall, because not only will the rate of wages rise, but more labourers will be employed without affording a proportional return of raw produce. The whole value of the wages paid will be greater compared with the whole value of the raw produce obtained’ (VI, 146)18 Thus far the argument applies to agriculture only. But it is extended, and the context once again is the rejection of Malthus’s view according to which accumulation entails a reduced price level:

Instead of anticipating a fall in the price of commodities we should expect a rise, because the fall of profits which generally follows accumulation is in consequence of the increase in the price of production [costs], compared with the price of produce [revenue]; although they would both undoubtedly rise. You appear to think, –indeed you say that you ‘know no other cause for the fall of profits which generally takes place from accumulation than that the price of produce falls compared with the expence of production, or in other words that the effective demand is diminished’, and by what follows you seem to infer that commodities will not only be relatively lower but really lower, and this is in fact the formulation of our difference with regard to the theory of Mr. Mill. (Ricardo 1951, VI, 148–9; emphasis added)19 The crucial step in Ricardo’s argument – apart from the initial acceptance of ‘Say’s Law’ (or rather Mill’s Law) and the Malthusian population mechanism – is the failure of the prices of manufactured goods to rise in proportion with money wages. In 25

RICARDO - THE NEW VIEW

the case of agricultural produce we have seen that Ricardo did not allow for any price increase due to higher money wages although quite possibly this was an oversight. The Ricardian theory of a declining profit rate due to rising corn prices required a demonstration that higher money wages are not entirely passed on in the form of higher prices. In brief, what was needed was an explanation of the determination of the price level. Ricardo himself, in a context relating to the process of accumulation, put his finger on this issue: ‘It appears to me that the consideration of money value may be the foundation of our difference on this point.’ 20 The problem of ‘value’ reappears in a different context subsequently in the correspondence of 1814. To a claim by Malthus (discussed below) that the general profit rate would rise with new lines of foreign trade Ricardo objected in terms similar to those subsequently developed in the chapter ‘On Foreign Trade’ in the Principles. Trade as such increases the quantity of commodities available and community well-being, but there will ‘not be a greater value of commodities to be exchanged for the raw produce, or for money’ and accordingly, he concluded, ‘no increased profits will any where be made’. By this he had in mind the notion that if certain commodities are increased in supply it will be their exchange values only which would decline relatively to those whose supply remained unchanged but the exchange value of these latter or of goods-in-general – relative to either corn or money – is not altered. ‘It is here I think, that our difference rests.’ 21 According to our understanding of Ricardo’s position in the correspondence of 1814 the profit rate in agriculture does not strictly determine the profit rate elsewhere. Rather the general profit rate varies inversely with the movements in money wages. In this case all depends upon the contents of the wage basket. If it contains solely agricultural produce (corn) then productivity in agriculture alone will govern the money wage. But if it contains manufactured goods then changes in manufacturing productivity might, in principle, affect the money wage and accordingly the general rate of profits. That Ricardo himself made this extension will now be demonstrated. It is true that Ricardo had expressed himself in terms of the strong proposition: ‘it is the profits of the farmer which regulate the profits of all other trades’ (above, p. 20). And late in 1814 Malthus defined the issue between himself and Ricardo in terms of whether the agricultural rate does or does not ‘take the lead’. If it could be shown that events outside the agricultural sector can (permanently) effect the general profit rate, then Ricardo’s proposition – in either form, it may be remarked – would be jeopardized: ‘It is of course by no means enough to say that from the state of production from the land, compared with the means necessary to make it produce, you can infer with certainty the state of general profits; as this is merely saying what every body knows, that all profits must ceteris paribus be on a level. But the question 26

RICARDO’S ANALYSIS OF THE PROFIT RATE

is whether agriculture always takes the lead in the determination? and I should certainly say that it did not.’ (23 November 1814; Malthus, in Ricardo 1951, VI, 152–3). To illustrate the argument Malthus refers to the opening of ‘a new foreign commerce’ –providing for the home market new and highly desirable products – whereby the profit rate on commercial capital is raised. ‘You allow’, he wrote, ‘that in this case capital may be taken from the land. But to allow this is at once to allow that the profits of foreign commerce determine in this case the profits on the land and that whichever is the highest will take the lead of the other.’ 22 In his reply Ricardo categorically denied Malthus’s attribution: ‘I do not recollect ever having allowed that an extension of foreign commerce will take capital from the land, unless we were an exporting country as far as regards corn, in which case my proposition would be true, namely that the rate of profits can never permanently rise unless capital be withdrawn from the land.’ 23 More significantly, however, Ricardo conceded that the general profit rate might be affected by events outside the agriculture He thereby widely extended the sector, if these events occur in the wage goods sector. significance of his propositions:

I have been endeavoring to get you to admit that the profits on stock employed in Manufactures and commerce are seldom permanently lowered or raised by any other cause than by the cheapness or dearness of necessaries, or of those objects on which the wages of labour are expended. Accumulation of capital has a tendency to lower profits. Why? because every accumulation is attended with increased difficulty in obtaining food, unless it is accompanied with improvement in agriculture; in which case it has no tendency to diminish profits. If there were no increased difficulty, profits would never fall, because there are no other limits to the profitable production of manufactures but the rise of wages. If with every accumulation of capital we could tack a piece of fresh fertile land to our Island, profits would never fall. I admit at the same time that commerce, or machinery, may produce an abundance and cheapness of commodities, and if they affect the prices of those commodities on which the wages of labour are expended they will so far raise profits; – but then it will be true that less capital will be employed on the land, for the wages paid for labour form a part of that capital. (Ricardo 1951, VI, 162; emphasis added) 24 Ricardo has thus clarified his position: He was not claiming that the agricultural profit rate determines the general rate but rather that agricultural productivity which determines the price of corn, playing upon the money wage rate, influences profits generally. Should the wage basket contain manufactured products, then changes in productivity in manufacturing will have similar consequences. But Ricardo without 27

RICARDO - THE NEW VIEW

doubt felt his position to be seriously threatened for he closed his argument, it will be noted, by insisting that ‘less capital will be employed on the land’ following an alteration in the cost of manufactured wage goods. This conclusion allowed him to claim formally that a reduction in agricultural capital accompanies the increase in the general profit rate; but the relationship is not a causal one, for it is not the reduction in capital which generates the increase in the general profit rate, but rather the converse.25 During the first week of February 1815, Ricardo read Malthus’s newly published Inquiry into the Nature and Progress of Rent. The theory of differential rent developed therein met with Ricardo’s approval, fitting perfectly as it did into a general structure which was already prepared: ‘It is no praise to say that all the leading principles in it meet with my perfect assent.’ Yet while Ricardo accepted the broad principle he pointed out certain applications of the theory which Malthus had deliberately or unwittingly failed to make, or about which Malthus had expressed uncertainty. The rapidity with which Ricardo absorbed the implications of the theory and incorporated it into his own vision of the economic process is striking. For example, since ‘rents are always withdrawn from the profits of stock’ the importation of cheap foreign corn results in a social advantage measured by the price differential alone, there being no need to correct for the loss of rents; the taxation of necessaries cannot fall on the landlord since ‘the last portion of land cultivated, yields nothing more than the profits of stock’; land improvements (which raise output for a given outlay or reduce the outlay for a given output) will benefit the landlord only in so far as the farmer brings poorer land into cultivation but in themselves are advantageous only to the farmer. 26 II. THE ESSAY ON PROFITS The Essay on the Profits of Stock 27 represents Ricardo’s contribution to the monograph literature of February 1815 dealing with the potential effects of the proposed restrictions upon corn importation. Ricardo publicly recognized a debt to Malthus for the theory of differential rent, 28 although he differed sharply on the deductions to be drawn therefrom. In particular, he insisted that Malthus did not utilize logically and consistently the property of rent as a transfer payment. 29 However, the main variable under investigation in the Essay was the general rate of profits, and the fundamental determinant thereof was as it had been in the correspondence of 1813 and 1814 when the theory of differential rent was still unknown, namely the cost of production of food: Profits of stock fall because land equally fertile cannot be obtained, and through the whole progress of society, profits are regulated by the difficulty or 28

RICARDO’S ANALYSIS OF THE PROFIT RATE

facility of procuring food. This is a principle of great importance, and has been almost overlooked in the writings of Political Economists. They appear to think that profits of stock can be raised by commercial causes, independently of the supply of food. (Ricardo 1951, IV, 13n) Profits then depend on the price, or rather on the value of food. Every thing which gives facility to the production of food, however scarce, or however abundant commodities may become [by contractions or extensions of foreign trade, or the use of ‘machinery’], will raise the rate of profits, whilst on the contrary, every thing which shall augment the cost of production without augmenting the quantity of food, will, under every circumstance, lower the general rate of profits. (Ricardo 1951, IV, 26)30 The theory of differential rent was incorporated effortlessly into an already prepared general structure. At the outset Ricardo stated his assumptions, namely that real wages per man are constant (at subsistence) and that agricultural technology is unchanged: ‘We will, however, suppose that no improvements take place in agriculture, and that capital and population advance in the proper proportion, so that the real wages of labour, continue uniformly the same; – that we may know what peculiar effects are to be ascribed to the growth of capital, the increase of population, and the extension of cultivation, to the more remote, and less fertile land’ (IV, 12). 31 On this basis he constructed a well-known model – relating uniquely to the agricultural sector – which yields a declining profit rate as capital and population expand and are applied to increasingly disadvantageous plots of land. In the model the plots of land are equally fertile but at increasing distances so that additional capital is required to yield the same results. Specifically, it is assumed that the capital of an (initial) farmer consists of the ‘value of two hundred quarters of wheat’ (of which 100 represents fixed capital and 100 circulating capital), and that ‘the value of the remaining produce’ – after maintenance of capital – amounts to ‘one hundred quarters of wheat, or [is] of equal value with one hundred quarters’ so that the rate of profit equals 100/200 or 50 per cent (IV, 10). A further addition of 300 quarters requires a capital of 210 quarters or rather the ‘value’ of 210 quarters so that the profit rate declines to 90/ 210 or 43 per cent. And since the common rate is ‘regulated by the profits made on the least profitable employment of capital in agriculture’, 14 quarters of the original farmer’s profits are transferred by the competitive process to the landlord as rent such that the profit rate on intra-marginal capital falls to 43 per cent (IV, 13). 32 In light of Ricardo’s references to the value of wheat it would be misleading to designate the illustration as a ‘corn model’ in the strict sense; corn is merely the numéraire. 33 29

RICARDO - THE NEW VIEW

Ricardo emphasized in particular the following results yielded by the model: that the net aggregate agricultural output (rent and profits only) rises with capital accumulation; that aggregate profits rise initially, but subsequently fall – ‘a view of the effects of accumulation which is exceedingly curious, and has, I believe, never before been noticed’; that the profit rate (‘relative profits’) declines from the outset; and that the proportion of rent to agricultural capital (the ‘share of the landlord’) rises (IV, 16f). He is in fact mistaken in this final unqualified statement; beyond the eighth period the fraction actually falls – an early example of the dangers inherent in the arithmetic method. And the fact that Ricardo did not emphasize the behaviour of the rental and profit shares in net or gross produce, but rather chose to examine the ratio of rent and profits to capital ( both fixed and circulating ), suggests that his primary concern was not with the aggregative shares. Ricardo was careful to observe that the data ‘are assumed, and are probably very far from the truth’. In particular, ‘in proportion as the capital employed on the land, consisted more of fixed capital, and less of circulating capital, would rent advance, and [profits] fall less rapidly’ (IV, 15–16n). Doubtless he has in mind that a given increase in aggregate capital, if the fixed capital constituent is relatively high, would entail a relatively smaller circulating capital (wages) increment and support accordingly a lesser increase in population, at the subsistence wage, than if the fixed capital constituent were lower. 34 Ricardo here touches, for the first time, upon the thorny question of the relationships between different categories of capital, and between capital and labour. And it is clear that these relationships are presumed to be technological data. It seems to have been Ricardo’s intention to apply the model specifically to the agricultural sector; 35 but, as we shall see, he was well aware that the fundamental result – the decline in the rate of profit with capital accumulation – was irrelevant unless some explanation for the behaviour of the rate of profit outside the agricultural sector could be provided. Only if it is satisfactorily demonstrated that the external rate moves pari passu with the internal rate can the model be said to reflect the economy as a whole. Any such demonstration, we may note, is independent of, and must be evaluated apart from, the model itself. Even in the event that an adequate relationship is defined between the agricultural and general rates of profit the model would be unsuitable for the treatment of the income shares, and in fact Ricardo nowhere states the belief that there is a parallel movement between the aggregate income shares and the shares in agricultural output as yielded by the model. We must recall at this stage of the argument a distinction between the ‘strong’ proposition that the agricultural profit rate determines the profit rate elsewhere or a more sophisticated variation thereof that the state of agricultural productivity on the margin of cultivation is the unique determinant of the general profit rate; and the 30

RICARDO’S ANALYSIS OF THE PROFIT RATE

‘weaker’ proposition that the state of agricultural productivity exerts an influence on the general profit rate although not to the exclusion of other forces. We have already demonstrated that Ricardo’s position in the correspondence of 1814 implies the ‘weak’ formulation. We turn now to consider the position adopted in the Essay. A preparatory statement by Ricardo is relevant for an evaluation of the ‘weights’ attached to these propositions in the Essay: ‘I am only desirous of proving that the profits on agricultural capital cannot materially vary, without occasioning a similar variation in the profits on capital, employed on manufactures and commerce’ (IV, 12n).36 This remark – which does not imply that the general rate is governed by the agricultural rate – is attached to the ‘proof’ itself: ‘In this state of society, when the profits on agricultural stock, by the supposition, are fifty per cent. the profits on all other capital, employed either in the rude manufactures ... or in foreign commerce ... will be also, fifty per cent. If the profits on capital employed in trade were more than fifty per cent. capital would be withdrawn from the land to be employed in trade. If they were less, capital would be taken from trade to agriculture’ (IV, 12). 37 As the so-called ‘proof’ stands it is, in fact, nothing more than an assertion that the profit rate cannot, in the long run (assuming competition), differ between sectors. It would be quite inadequate if intended as justification for the proposition that the general rate is governed by the agricultural rate since it does not preclude the possibility of variations induced by changes emanating outside the agricultural sector. Subsequently in the Essay Ricardo reconsidered the relationship between the agricultural and general rates of profit in a more sophisticated manner, reminiscent in certain important respects of the correspondence with Malthus of 1814. According to the argument, the agricultural profit rate tends to decline during the process of capital accumulation because of declining agricultural productivity; while the manufacturing rate falls pari passu as a result of rising money wages – in turn due to the rising corn price – in the face of constant commodity prices. The argument takes for granted that the exchange value of a commodity reflects the ‘difficulty of production’ and thus accounts for the rising price of corn, and the stable price of manufactured goods: If the money price of corn, and the [money] wages of labour, did not vary in price in the least degree, during the progress of the country in wealth [capital] and population, still profits would fall and rents would rise; because more labourers would be employed on the more distant or less fertile land, in order to obtain the same supply of raw produce; and therefore the cost of production would have increased whilst the value of the produce continued the same. ... [But] the exchangeable value of all commodities, rises as the difficulties of 31

RICARDO - THE NEW VIEW

their production increase. If then new difficulties occur in the production of corn, from more labour being necessary, whilst no more labour is required to produce gold, silver, cloth, linen, &c. the exchangeable value of corn will necessarily rise, as compared with those things. ... Wherever competition can have its full effect ... the difficulty or facility of their production will ultimately regulate their exchangeable value. The sole effect then of the progress of wealth on prices ... appears to be to raise the price of raw produce and of labour, leaving all other commodities at their original prices, and to lower general profits in consequence of the general rise of wages. (Ricardo 1951, IV, 18–20; emphasis added)38 The present argument in the Essay appears to contain an error which Ricardo was ultimately to correct. The increase in the price of corn reflects the fall in agricultural productivity and to this extent any fall in agricultural profits is prevented. Profits decline in agriculture for precisely the same reason that they decline in manufacturing, namely as a consequence of rising money wages. But Ricardo, at this stage, attributes the decline in the agricultural profit rate to falling productivity and the decline in the manufacturing profit rate to rising money wages. The formal argument, although taken a step further than in the correspondence was, therefore, still not satisfactory. 39 It will be noted that Ricardo has here, once again, moved from the ‘strong’ proposition that the agricultural profit rate determines the general rate – which we have seen was never adequately justified – to the ‘weaker’ formulation since, in principle, any force generating a change in money wages would affect the profit rate whether or not it has its origin in the agricultural sector. Yet in one vital respect the argument has progressed compared with that in the correspondence. For Ricardo now presumes the manufacturing price level to be unchanged during the process of capital accumulation. 40 The increase in the money wages rate itself in no way disturbs either the price structure – particularly the relative prices of agricultural and manufactured products – or the price level. It remained to be shown precisely why, or under which conditions, this will be the case. A preliminary attempt is made to deal with the issue in a brief criticism of a proposition by Adam Smith to the effect that an increase in the price of corn – regarded as the ultimate regulator of wages – will lead to a general price increase. Ricardo objected, distinguishing between a variety of causal influences: It has been thought that the price of corn regulates the prices of all other things. This appears to me to be a mistake. If the price of corn is affected by the rise or fall of the value of the precious metals themselves, then indeed will the price of commodities be also affected, but they vary, because the value of money varies, not because the value of corn is altered. Commodities, I think, 32

RICARDO’S ANALYSIS OF THE PROFIT RATE

cannot materially rise or fall, whilst money and commodities continue in the same proportions, or rather whilst the cost of production of both estimated in corn continues the same. (Ricardo 1951, VI, 21n) 41 Clearly Ricardo has, to all intents and purposes, recognized the need for a ‘measure’ of the ‘value’ or the ‘difficulty of production’ of commodities in dealing with the issue of the effect of wage-rate changes on prices. We may refer to two applications of the general argument in order to illustrate the significance thereof for Ricardo. The analysis is applied to the general theory of foreign trade. A relaxation of restrictions upon the importation of corn would not alter the aggregate amount of trade, but the profitability of trade would be increased. This is because a low price of corn allows reductions in manufacturing costs while selling prices remain unchanged : ‘Our commodities would not sell abroad for more or for less in consequence of a free trade, and a cheap price of corn; but the cost of production to our manufacturers would be very different if the price of corn was eighty, or was sixty shillings per quarter’ (IV, 36). A second application is made to Malthus’s argument – based on David Hume – that a low price of corn would have depressing effects on the economy. Ricardo objected on the grounds that there will be no general decline in prices, but rather a general increase in profits: A recurrence to a better monetary system, it is said, though highly desirable, tends to give a temporary discouragement to accumulation and industry, by depressing the commercial part of the community, and is the effect of a fall of prices: Mr. Malthus supposes that such an effect will be produced by the fall of the price of corn. If the observation made by Hume were well founded, still it would not apply to the present instance: – for every thing that the manufacturer would have to sell, would be as dear as ever: it is only what he would buy that would be cheap, namely, corn and labour by which his gains would be increased. I must again observe, that a rise in the value of money lowers all things; whereas a fall in the price of corn, only lowers the wages of labour, and therefore raises profits. (Ricardo 1951, IV, 37; emphasis added) 42 III. CONCLUSION One final issue requires comment. Ricardo had insisted during 1814 that the general profit rate could not rise ‘unless capital be withdrawn from the land’ (see above, p. 20) but in the Essay he formulates this condition more rigorously. On the assumption that (domestic) demand for corn is a function of population size only, and 33

RICARDO - THE NEW VIEW

therefore completely inelastic with respect to price, the margin of cultivation is inexorably determined given population and so, accordingly, is the agricultural profit rate. 43 Since in the long run there can be only one rate, the general rate must come into line following a disturbance in commerce or manufacturing: It is contended, that they [the rates of profit in various branches of activity] alternately take the lead; and, if the profits of commerce rise, which it is said they do, when new markets are discovered, the profits of agriculture will also rise; for it is admitted, that if they did not do so, capital would be withdrawn from the land to be employed in the more profitable trade. But if the principles respecting the progress of rent be correct, it is evident, that with the same population and capital, whilst none of the agricultural capital is withdrawn from the cultivation of the land, agricultural profits cannot rise, nor can rent fall: either then it must be contended, which is at variance with all the principles of political economy, that the profits on commercial capital will rise considerably, whilst the profits on agricultural capital suffer no alteration, or, that under such circumstances the profits on commerce will not rise. (Ricardo 1951, IV, 23–4)44 According to this argument – which takes for granted that once the margin of cultivation is given the agricultural profit rate is immediately determined – the commercial rate must come into line, following an initial increase thereof, by new capital investment in commerce (presumably financed by higher retained earnings) rather than by transfers from agriculture: ‘there will be such a fall in the price of the foreign commodity in the importing country, in consequence of its increased abundance, and the greater facility with which it is procured, that its sale will afford only the common rate of profit – that so far from the high profits obtained by the few who first engaged in the new trade elevating the general rate of profits – those profits will themselves sink to the ordinary level’ (IV, 24–5). 45 But it may be asked why equality between the two rates cannot be achieved simply as a consequence of an attempt to transfer capital from agriculture to trade, in the face of zero demand elasticity; such an attempt would force up the price of corn relative to that of other goods and tend to bring the rates into equality. 46 If the gap is to be closed, the presumption that the agricultural rate is given once the margin is determined must be justified. According to the suggestion by Sraffa, it will be recalled, Ricardo had in mind in the Essay a model wherein both agricultural inputs and outputs consist of the same physical commodity (corn) so that total profits and the rate of profit on capital (a corn wages fund) can be determined in physical terms without reference to valuation. Since all other sectors utilize corn as input only, it is the exchangeable value of their products relative to their (corn) capitals which must be adjusted to yield the same profit rate as in 34

RICARDO’S ANALYSIS OF THE PROFIT RATE

agriculture. If this is a justified attribution, then even though an attempt is made to transfer capital from agriculture and the price of corn consequently rises, the agricultural profit rate remains unchanged. Although this interpretation of Ricardo’s implicit logic is an attractive one, it was (as Sraffa himself remarks) never ‘explicitly stated by Ricardo’. Our study of the pertinent letters in 1814, which it is said lend support to this interpretation, suggests that it is unjustified to regard the arguments therein as implying an analysis in terms of the corn model. Moreover, the fact that the tabular formulations of the Essay do indeed reflect a corn model is inadequate proof since we are dealing now precisely with the justification for the use of such formulations. The basis for Ricardo’s argument seems rather to consist of the assumption that the exchangeable value of corn (as of any commodity) depends upon real costs – the ‘difficulty of production’. If the agricultural margin is fixed, the productivity of labour (and capital) on land is unchanged and so accordingly is the price of corn. If then the money wage rate is dependent upon the corn price, it too is constant, so that the agricultural profit rate remains unchanged. In this case the only way equality across the board can be achieved is if the profit rate elsewhere comes into line as a result of expansion of commerce and manufacturing. 47 If our analysis of the Essay is an accurate reflection of Ricardo’s intentions it follows that the ‘corn model’ was of little significance to him. For the results yielded thereby are of relevance only to the extent that we accept the arguments devised to account for the relationship between the agricultural and the general rate. Ricardo was well aware of this. But in the event that these arguments are accepted the model is not strictly required; it serves merely as a convenient method of portraying some of the principles in simple arithmetical form. The essence of the Essay is rather that variations in the money-wage rate – in turn largely (though not necessarily wholly) determined by the price of corn – are not reflected in final prices so that the profit rate moves inversely. This is the single significant ‘advance’ over the position already achieved in the correspondence of 1814 where some increase in the price level – albeit in lesser proportion than that in the wage rate – was allowed. All that was now required was to isolate the precise conditions for the assurance of a constant price level in the face of rising money wages. 48 It may also be convenient to contrast our conclusions with those in the wellknown account of the Ricardian theory by Professor Stigler. According to this account (Stigler 1965, 187), the Essay contains only two of the main elements which constitute the complete Ricardian system, namely the theory of rent and the dominant influence of diminishing returns in agriculture upon the general profit rate. The full system as presented in the Principles includes also the subsistence theory of wages and the measure of value. According to our view of the matter, the subsistence theory (involving the population mechanism) and the need for a ‘measure of value’ 35

RICARDO - THE NEW VIEW

(although not, of course, its actual construction) are both apparent not only in the Essay but also in the correspondence of 1814. NOTES 1. Letter dated 10 August 1813 (Ricardo 1951, VI, 93): ‘On further reflection I am confirmed in the opinion which I gave with regard to the effect of opening new markets or extending the old. ... This would again swell the value of our exports and imports but does not prove a general increase of profits nor any material growth of prosperity.’ Malthus and Ricardo met frequently during 1813, and the issue of profit-rate determination was doubtless discussed in conversation. (Malthus’s letter to which Ricardo’s of 10 August is a reply, and Malthus’s reply to Ricardo, are both wanting.) None of the published letters prior to that of 10 August touches on the issue. 2. For a discussion of the background to these letters, see Tucker 1954, 320–33. 3. The principle of diminishing returns was defined by Ricardo several years earlier in his Notes on Bentham’s ‘Sur les Prix’ (1810–11): ‘It appears to me that the possession of new Land would add to our sum of riches without additional labour, because the same labour employed on double the quantity of equally good land now in cultivation in England would produce a greater return. This opinion is founded on the decreasing power of the land to produce in proportion to the labour and capital employed on it’ (Ricardo 1951, III, 287). 4. The reference to the ‘same or an increased Capital’ doubtless should be understood as referring to agricultural investment. 5. The relevant letter of Malthus, to which Ricardo’s of 26 June was written in reply, is wanting. But Malthus subsequently restated his position regarding the reduction in capital and the elevating effect on the rate of profits in a letter dated 6 July (Ricardo 1951, VI, 110–11): You observe that in the case supposed [of corn restriction], there would be less production and less demand with the same capital; but surely there would be much less capital. There would be a smaller quantity both of corn, and of all other commodities, and every monied accumulation would command less labour and less produce. The question then seems to be whether production or demand would decrease the fastest? and as in my opinion the dearness of labour [money wages] would have more effect in diminishing capital than in diminishing revenue, particularly, rents, I do not see why the usual effects of a diminution of capital should not take place. I can by no means agree with you in thinking that every thing which diminishes produce, tends to diminish the power of paying for the commodities wanted, or as you intimate, to diminish the effective demand. ... [The] rate of production, or more definitely speaking, the proportion of production to consumption necessary to such production, seems to be determined by the quantity of accumulated capital compared with the demand for the products of capital, and not by the mere difficulty and expence of procuring corn. ... The effects of a great difficulty in procuring corn would in my opinion real wages be, a diminution of capital, a diminution of produce, and a diminution in the of labour, or their price in corn; but not a diminution of profits.

The argument is expanded in a further letter of 5 August (VI, 116–17): Since the manufacturer must ‘pay more [money wages] for the labour which he employs, owing to restrictions upon importation, he will not be able to work up the same quantity of goods with his capital; the goods will in consequence rise in price, and his profits, from the general scarcity of capital, will be increased’. 36

RICARDO’S ANALYSIS OF THE PROFIT RATE

It must be noted that the consequence of a ‘retrograde capital’ would be a decline in the real (corn) wage rate, since the price of produce tends to rise ‘without a proportionate rise of labour [money wages]’. According to Malthus, upward pressure on the profit rate would swamp any depressing effects of extensions to marginal land: This ‘most natural and frequent occurrance will allow of great variations in the rate of produce on land, and easily make up for some increase of difficulty in procuring corn’ (VI, 118). 6. Letter dated 11 September 1814, VI 132. 7. Letter dated 16 September 1814 (VI, 133–5). In reply, Malthus wrote (letter dated 9 October 1814, VI, 141): In stating the cause of high profits you seem to me to consider almost exclusively the expence of production, without attending sufficiently to the price of produce, and greatly to underrate the wants and tastes of mankind in affecting prices, and consequently in affecting the means of profitably employing capital. ... It is in considering merely of the proportions of commodities to one another, and not of their proportions to the wants and tastes of mankind that the error of Mr. Mill, in my opinion, consists. I cannot by any means agree with you in your observation that ‘the desire of accumulation will occasion demand just as effectually as a desire to consume’ and that ‘consumption and accumulation equally promote demand’.

8. 9. 10.

11. 12.

Ricardo again insisted in reply (letter dated 23 October 1814, VI, 147–8): ‘I am not aware that I have under-rated the effect of the wants and tastes of mankind on profits,—they frequently, occasion large profits on particular commodities for short periods,—but they do not I think often operate on general profits because they do not often influence the growth of raw produce. ... I go much further than you in ascribing effects to the wants and tastes of mankind,—I believe them to be unlimited. Give men but the means of purchasing and their wants are insatiable. Mr. Mill’s theory is built on this assumption.’ More specifically ‘the very term accumulation of capital supposes a power somewhere to employ more labour,—it supposes the total income of the society to be increased and therefore to create a demand for more food and more commodities.’ Letter dated 25 July 1814 (Ricardo 1951, VI, 113). Letter dated 26 June 1814 (Ricardo 1951, VI, 108). (By ‘value’ Ricardo means difficulty of production.) There is a possibility that Ricardo was referring not to the increased prices of manufactured goods due to higher money wages—in turn resulting from higher corn prices—but rather to the use of raw materials produced at higher cost. But the references to corn suggest that, in fact, he had in mind the effect of rising money wages on the prices of manufactured goods. Cf. Jacob H. Hollander 1968, 85–6. Letter dated 5 August 1814 (Ricardo 1951, VI, 117–18). (Malthus concluded: ‘The more I reflect on the subject, the more firmly I feel convinced, that it is the state of capital, or the general profits of stock and interest of money, which determines the particular profit upon the land; and that it is not the particular profits or rate of produce upon the land which determines the general profits of stock and the interest of money.’) See also the letter dated 9 October 1814 (VI, 140–1) wherein Malthus concedes that the cost of food affects the profit rate in a limiting sense only, and attributes to Ricardo an analysis in physical terms: It appears to me that ... in the interval between the two extremes [of high and low profits due to good and poor land resources] considerable variations may take place; and that practically no country was ever in such a state as not to admit of increase of profits on the land, for a period of some duration, from the advanced price of raw produce.

37

RICARDO - THE NEW VIEW The Profits of stock, or the means of employing capital advantageously may be said to be accurately equal to the price of produce, minus the expence of production. And consequently whenever the price of produce keeps a head of the price of production the profits of stock must rise. ... It is not the quantity of produce compared with the expence of production that determines profits, (which I think is your proposition) but the exchangeable value or money price of that produce, compared with the money expence of production. ...

13. The quotation at the outset is from Malthus’s letter of 6 July (Ricardo 1951, VI, 111). There is apparently a slight error in Ricardo’s argument. He must have meant that capital nearly doubles estimated in corn, not in money, for money capital must more than double if the rate of profit is to fall, given the assumption which is made of a doubled money revenue. The argument allows for a reduction in the corn wage per man, although not in proportion to the increase in the number of men since the total corn costs rise. 14. For Malthus’s position, according to which a rise in money costs could actually be accompanied by an increased profit rate, see note 5. 15. Letter dated 30 August 1814. 16. Letter dated 16 September 1814 (Ricardo 1951, VI, 133). Cf. letter dated 23 October 1814 (VI, 144f), in which Ricardo insists that Malthus had falsely ascribed to him the view that the state of agricultural productivity was the only determinant of the general profit rate. Temporary variations were possible for other reasons, but ‘even during these temporary variations, the great cause namely the accumulation of capital may be paving the way for permanently diminished profits’ (VI, 145–6). 17. Letter dated 23 October 1814 (Ricardo 1951, VI, 146–7). (Cf. Principles of Political Economy, I, 161 where Ricardo lists bad harvests, capital accumulation, general price inflation, and taxation of wage goods as causes of increase in corn prices.) ‘Bad seasons’ would raise the profit rate as a consequence of high demand inelasticity for corn: the ‘price of produce would rise considerably more than in the proportion of the deficient quantity, and would therefore be much ahead of the price of production’; a ‘fall in the value of the currency’ would raise profits because of a lag in money wages behind prices. But these are temporary phenomena only. Corn restriction too will raise the profit rate in the first instance ‘but they will ultimately fall below their former level’ (VI, 146). Presumably he has in mind, in the latter case, the proposition that the decline in the profit rate following restriction requires the full operation of the population mechanism. 18. By ‘value’ Ricardo doubtless means difficulty of production. 19. The reference is to Malthus’s letter of 9 October (Ricardo 1951, VI, 142). 20. Letter dated 18 December 1814 (Ricardo 1951, VI, 164); emphasis added. 21. Letter dated 18 December 1814 (Ricardo 1951, VI, 163). It is important to note that Ricardo defines his terms carefully to imply a long-run cost theory: ‘If we double the quantity, or rather double the facility of making stockings, we diminish their value one half, as compared with all other commodities.’ The argument is developed in a subsequent letter, dated 13 January 1815 (VI, 170–1), (although questions appear regarding the precise mechanism by which exchange value reflects costs since a price reduction without quantity increase is allowed for): If with the same labour we could obtain double the quantity of tin from the mines in Cornwall, after prices have found their level, would the value of the whole mass of commodities be increased in England? Should we obtain the same quantity of deals from Norway in exchange for a given quantity of tin as we now do? Although the mass of commodities both in the markets of Norway and in those of England would increase by the

38

RICARDO’S ANALYSIS OF THE PROFIT RATE greater abundance of tin, or of some other commodity, if the labour employed in procuring tin were diverted to other objects, yet the estimated value of all their commodities, in corn, money, or any article but tin, would, it appears to me continue unaltered. It is sufficient that deals can be purchased cheaper in Norway than elsewhere to determine a portion of foreign trade to that quarter, although it should yield no more profits than those of other trades.

22. It may be noted that this argument was designed to ‘trap’ Ricardo. Malthus himself believed that even if capital is not withdrawn from agriculture the effect of ‘throwing new objects of desire into the market’ would be to increase ‘the value of the whole mass of commodities in the country, estimated either in money, or in corn and labour, and there will in consequence be a greater value of commodities to be exchanged for the raw produce. This increase in the value of the raw produce must raise the profits of farming for a period of some duration’ (letter 23 November 1814). In his next letter (of 29 December 1814, Ricardo 1951, VI, 167) Malthus qualified himself by conceding once again that ultimately the effects predicted by Ricardo would be felt: ‘Of course I never mean to say that the high price of raw produce compared with the price of production, occasioned, (in my opinion) by a prosperous commerce can be really permanent. It is the nature of all such means of employing capital to become less and less advantageous. And all that I contend for is that a period of some duration may occur (20 years for instance) when the profits of commerce will take the lead, and regulate the profits of agriculture.’ Ricardo jumped upon the concession (in a letter dated 13 January 1815, 170), since he too had admitted the possibility of temporary variations in the profit rate for various reasons (see above): I thought you maintained, that the high or low profits on commerce were totally independent of the amount of capital which might be employed on the land; consequently that high profits might continue as long as commerce was prosperous, whether that was for 20 or for 100 years. I now understand you to say, that the profits of commerce may take the lead, and may regulate the profits of agriculture for a period of some duration, possibly for 20 years. I have always allowed that under certain circumstances profits on agriculture might be diverted from their regular course for short periods, so that we only appear to differ with respect to the duration of such profits; instead of 20 years I should limit it to about 4 or 5.

Despite the conciliatory tone of this exchange the fact is that their respective ‘models’ differed substantially and the debate was not merely one concerning the meaning of ‘temporary’. If a variety of forces could influence the profit rate ‘temporarily’ for Ricardo it was by relationships peculiar to his model alone. Similarly, when Malthus allows for ‘temporary’ variations of the general profit rate due to events in manufacturing and commerce he had in mind phenomena ruled out by Ricardo. 23. Letter dated 18 December 1814 (Ricardo 1951, VI, 163). For the ‘proposition’ referred to, see for example above, p. 20. 24. Malthus’s comment (29 December 1814, Ricardo 1951, VI, 168–9) on the effects of ‘a piece of fresh fertile land’ is revealing, for it suggests that he appreciated the role of money wages in Ricardo’s argument: ‘I quite agree with you that a piece of fertile land added to the country upon every increase of capital would prevent the fall of profits, but more in my opinion from its increasing the demand for manufactures by increasing the number of people than by its preventing the rise of wages, an effect which it would probably not have.’ 25. Subsequently Malthus (letter dated 29 December 1814, Ricardo 1951, VI, 168) challenged Ricardo’s position from another direction. He asked Ricardo to presume, for the sake of argument, an increase in the foreign demand for home corn: ‘would not the profits of 39

RICARDO - THE NEW VIEW

26.

27. 28.

29. 30. 31. 32.

33. 34.

35. 36. 37. 38.

capital employed in agriculture be increased, although certainly more rather than less capital would be employed upon the land[?] The instruments of production compared with the price of produce would be cheaper, but it could not with any propriety be said that capital was withdrawn from the land.’ But Ricardo answered quite shortly: ‘there can be no question that more capital would be employed on the land, and I think profits would fall’ (VI, 171). Letter dated 6 February 1815 (Ricardo 1951, VI, 172–4). On the matter of the effects of taxes on necessaries Malthus had stated in the Inquiry that the burden fell on the landlord not the consumer. He conceded to Ricardo, in his reply (12 February 1815, VI, 176) that in this respect he was in error. He continued to maintain strenuously that agricultural improvements will raise rents despite the fact that they are ‘undoubtedly a part of the wealth already created’ (VI, 174). Essay on the Influences of a Low Price of Corn on the Profits of Stock (Ricardo 1951, IV, 9–41). Published 24 February 1815. (Cf. Sraffa (ed.), Ricardo 1951, IV, 5, for this date and those in the next note.) Sraffa (ed.), Ricardo 1951, IV, 9, 15n, regarding the Inquiry into the Nature and Progress of Rent, and Grounds of an Opinion on the Policy of Restricting the Importation of Foreign Corn, published 3 February and 10 February 1815 respectively. (Other contributors to the literature were Edward West, Essay on the Application of Capital to Land, 13 February, and Robert Torrens, Essay on the External Corn Trade, 24 February.) ‘It is never a new creation of revenue, but always part of a revenue already created’ (Ricardo 1951, IV, 18). Value evidently means ‘difficulty of production’. With regard to the constancy of real wages per head, see also pp. 13, 18n, 23. Ricardo briefly refers also to an intensive margin (1951, IV, 14). He was on unsafe ground when he criticized Malthus (letter dated 6 February, VI, 172) for not considering separately in his Inquiry ‘the relations of rent with the profits of Stock and the wages of labour’. His stricture that ‘by treating of the joint effect of the two latter on rents’ Malthus had ‘not made the subject so clear as it might have been’ can equally well be directed at the Essay where wages are rather casually included within capital. In the ‘Table shewing the Progress of Rent and Profit’ (Ricardo 1951, IV, 17) reference is made to ‘capital estimated in quarters of wheat’. The argument is probably based on Ricardo 1951, IV, 13: ‘if capital and population increased, more food would be required, and it could only be procured from land not so advantageously situated .... The necessity of employing more labourers, horses, &c ... although no alteration were to take place in the wages of labour, would make it necessary that more capital should be permanently employed to obtain the same produce.’ The extent of population increase with wages per head at subsistence will depend upon the fraction of any given capital devoted to circulating capital relative to fixed capital. For example, at the outset Ricardo refers to ‘the usual and ordinary rate of the profits of agricultural stock’ (Ricardo 1951, IV, 10) and also the precise formulation of the results by the model would suggest this. Cf. the assertion, p. 14, that as the profit rate in agriculture declines from 50 to 43 per cent so the profits on all capital employed in trade would fall to 43 per cent. Ricardo had in mind a ‘state of society’ where agricultural improvements are ruled out and real wages are constant, capital and population advancing ‘in the proper proportion’. In a note attached to the passage Ricardo remarks that commodity prices are ‘ultimately regulated by’ and ‘always tending to, the cost of their production, including the general 40

RICARDO’S ANALYSIS OF THE PROFIT RATE

39.

40.

41.

42.

profits of stock’. In brief the ‘difficulty of production’ includes not only labour but capital costs too. But in so far as Ricardo is attempting to explain the general rate of profits it cannot be taken for granted as a datum ; he is, therefore, entrapped in circular reasoning. A similar argument is developed later in the Essay (IV, 35–6). Technological improvement in agriculture or a relaxation of corn import restrictions would lead to a fall in the price of corn only and therefore of money wages but not of manufactures, and accordingly the general profit rate would rise: ‘A fall in the price of corn, in consequence of improvements in agriculture or of importation, will lower the exchangeable value of corn only, the price of no other commodity will be affected. If, then, the price of labour falls, which it must do when the price of corn is lowered, the real profits of all descriptions must rise; and no person will be so materially benefited as the manufacturing and commercial part of society.’ The margin of cultivation would in the latter case be pushed back, and capital and labour released from agriculture would be available for use in the manufacturing sector. See p. 32: ‘If we were left to ourselves, unfettered by legislative enactments, we should gradually withdraw our capital from the cultivation of such [inferior] lands, and import the produce which is at present raised upon them. The capital withdrawn would be employed in the manufacture of such commodities as would be exported in return for the corn.’ In fact, Ricardo argued, opposition to easier corn imports on grounds that agricultural capital would be withdrawn is no more reasonable than opposition to improved machinery. (Cf. also p. 35.) In this regard we note a further statement (Ricardo 1951, IV, 26n): ‘If by foreign commerce, or the discovery of machinery, the commodities consumed by the labourer should become much cheaper, wages would fall, and this, as we have before observed, would raise the profits of the farmer, and therefore, all other profits.’ What is revealing is Ricardo’s formulation to the effect that a change in the money wage rate influences the agricultural profit rate ‘and therefore all other profits’. In fact the agricultural rate rises for the same reason that the rate elsewhere rises, because of the money wage reduction due in turn to productivity increase in agriculture. See above, p. 24, regarding Ricardo’s position in the correspondence of 1814. Prices of manufactured goods were said to rise although not in proportion with the rise in money wages despite the fact that there was no change in the ‘difficulty of production’ or ‘value’ as capital accumulation occurred. ‘Value’ once more refers to ‘difficulty of production’. The calculation, it will be noted, is in terms of corn, but a parallel calculation in terms of resources used up might have been used since the productivity of labour (or labour-and-capital) is unchanged in money and manufactured goods. It is this latter phenomenon which Ricardo evidently intended to emphasize. According to the above analysis Ricardo had attempted to demonstrate that the general profit rate was governed by movements of the money wage rate. Although in the correspondence he allowed explicitly for variations in money wages due to productivity increases in manufacturing (above p. 27), he did not formally do so in the Essay, but continually formulated variations of the theme that ‘profits depend on the price or rather on the value of food’. (Ricardo 1951, IV, 26; cf. 22: ‘If then, the principles here stated as governing rent and profit [the law of diminishing returns] be correct, general profits on capital, can only be raised by a fall in the exchangeable value of food’; or p. 23: 41

RICARDO - THE NEW VIEW

‘in every society advancing in wealth [capital] and population, independently of the effect produced by liberal or scanty wages, general profits must fall, unless there be improvements in agriculture, or corn can be imported at a cheaper price.’) Yet we are not obliged to understand by such statements an insistence that the wage basket consists solely of food; they are more likely intended as a rejection of Malthus’s view according to which the general profit rate may be affected by events which do not play upon the money wage at all. 43. In the correspondence of 1814, the notion of zero demand elasticity was not given a formal role in Ricardo’s argument. In fact, when the demand curve for corn was considered Ricardo presumed it to be relatively, but not completely, inelastic. See, for example, letters to Malthus, Ricardo 1951, VI, 129: ‘I sometimes suspect that we do not attach the same meaning to the word demand. If corn rises in price, you perhaps attribute it to a greater demand, – I should call it a greater competition. The demand cannot I think be said to increase if the quantity consumed be diminished, altho much more money may be required to purchase the smaller than the large quantity’; and p. 146: in the case of a poor harvest ‘the price of produce would rise considerably more than in the proportion of the deficient quantity’. 44. To this passage Ricardo attached the following note which confirms the role of zero demand elasticity for corn: Mr. Malthus has supplied me with a happy illustration – he has correctly compared ‘the soil to a great number of machines, all susceptible of continued improvement by the application of capital to them, but yet of very different original qualities and powers.’ How, I would ask, can profits rise whilst we are obliged to make use of that machine which has the worst original qualities and powers? We cannot abandon the use of it; for it is the condition on which we obtain the food necessary for our population, and the demand for food is by the supposition not diminished – but who would consent to use it if he could make greater profits elsewhere?

45. A version of the same argument is also applied to the case of a technological improvement introduced by one or a few firms in a particular manufacturing industry. In the first instance profits will rise for these firms as their costs fall. But with increased output permitted by the increasing adoption of the invention price will decline ‘to the actual cost of production, leaving only the usual and ordinary profits’ (Ricardo 1951, IV, 25). In this case too there ought to occur no flow of capital between sectors if Ricardo is to be consistent. Yet he inserted a paragraph immediately which does suggest a process of transfer: ‘During the period of capital moving from one employment to another, the profits on that to which capital is flowing will be relatively high, but will continue so no longer than till the requisite capital is obtained.’ The difficulty, once again, is that any such mechanism implies that the new general rate will be somewhat higher than the original level which is precisely the result which Ricardo was attempting to avoid, and indeed to disprove. 46. Cf. Stigler 1965, 186. The modern economist might be inclined to analyse the problem in terms of an upward shift in the (positively sloping) supply curve of corn reflecting the higher alternative rate of return, in the face of a given, inelastic, demand curve. Assuming a given aggregate capital, since there will be no actual transfer of capital from agriculture, there can be no increase in manufacturing output either and the profit rate in agriculture rises to equality with the rate in commerce. 47. The analysis of the effects of technological change in manufacturing (see note 45) should be contrasted with that of technical change in agriculture, where the price of corn 42

RICARDO’S ANALYSIS OF THE PROFIT RATE

falls and where, accordingly, the money wage rate and the general profit rate are affected. (See note 38 regarding the Essay, pp. 35–6.) Unfortunately, Ricardo never explained exactly how the price of corn in conditions of zero demand elasticity given population comes to reflect long-run costs with agricultural improvement. (A similar weakness appears in the correspondence of 1814; see note 21.) The argument that the money-wage rate will be determined by the subsistence level, is also made in too cavalier a fashion, at least in the present context. The argument implies the operation of the population mechanism while in fact population is assumed constant. 48. It seems, therefore, misleading to argue that ‘after the publication of the Essay, Ricardo retreated from the idea that “it is the profits of the farmer which regulate the profits of all other trades”. The general idea that profits “depend on the price or rather on the value of food” was retained in the Principles, but emphasis was now laid almost exclusively on the effect which the changes in the value of food associated with accumulation would have on profits through the medium of the consequential changes in wages’ (Meek 1967, 61–2). Sraffa himself, it may also be noted, allows that even in the Essay ‘there are passages which foreshadow [Ricardo’s] full theory of value and already link it with the theory of profits’ (Sraffa 1951, xxxiii) .

REFERENCES Hollander, Jacob H. (1968) [1910] David Ricardo: A Centenary Estimate, Baltimore: The Johns Hopkins Press. Meek, R.L. (1967) ‘The Decline of Ricardian Economics in England’, in Meek, Economics and Ideology and Other Essays, London: Chapman and Hall. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Sraffa, P. (1951) ‘Introduction’ to Principles of Political Economy, The Works and Correspondence of David Ricardo, vol. I, Cambridge: Cambridge University Press. Stigler, G.J. (1965) ‘The Ricardian Theory of Value and Distribution’ ( 1‘ Journal of Political Economy LX, June 1952), reprinted in Stigler, Essays in the History of Economics, Chicago: University of Chicago Press. Tucker, G.S.L. (1954) ‘The Origin of Ricardo’s Theory of Profits’, Economica XXI (November): 320–33.

43

2 RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

I. I am grateful to Mr Eatwell for obliging me to reconsider my article on Ricardo’s early position [Chapter 1]. Ricardo criticism is notoriously tricky and a review from time to time can be salutary. Upon careful consideration, much of Eatwell’s statement turns out to be little more than assertion, apparently based upon a preconceived notion of the nature and content of Ricardian profit theory and of Ricardo’s place in the history of economic thought; I am given no reason, by way of a demonstration of the inaccuracy or the inadequacy in other respects of my citations, to believe that I erred in my interpretation. Nevertheless, the exercise has turned out to be a fruitful one. I now believe, and hope to demonstrate in this reply, that my case is stronger than originally formulated, in the light of a body of additional evidence contained in the Ricardo–Malthus correspondence immediately following the publication of the Essay on Profits. 1 II. In this section I wish to set the issues involved in accurate perspective. It is essential to have explicitly before us the following elements which constitute Professor Sraffa’s position and with which I took issue in my paper: At first, both in the Essay and in Ricardo’s letters of 1814 and early 1815, a basic principle has been that ‘it is the profits of the farmer that regulate the profits of all other trades ...’. After the Essay this principle disappears from view, and is not to be found in the Principles. The rational foundation of the principle of the determining role of the profits of agriculture, which is never explicitly stated by Ricardo, is that in 44

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

agriculture the same commodity, namely corn, forms both the capital (conceived as composed of the subsistence necessary for workers) and the product; so that the determination of profit by the difference between total product and capital advanced, and also the determination of the ratio of this profit to the capital, is done directly between quantities of corn without any question of valuation. It is obvious that only one trade can be in the special position of not employing the product of other trades while all the others must employ its product as capital. It follows that if there is to be a uniform rate of profit in all trades it is the exchangeable values of the products of other trades relatively to their own capitals (i.e. relatively to corn) that must be adjusted so as to yield the same rate of profit as has been established in the growing of corn; since in the latter no value changes can alter the ratio of product to capital, both consisting of the same commodity. (Sraffa 1951, xxxi) With the exception of the criticisms by Malthus of Ricardo’s Table made in March 1815, which will be taken up in detail below, all the evidence brought by Professor Sraffa in support of his case was taken into consideration before reaching my conclusions. In the article the existence of formal statements of the principle of the determining role of agricultural profits was obviously at no time questioned. But my investigation led me to conclude, first, that Ricardo did not intend by his formal statements to maintain that the profit rate in agriculture literally determines the rate elsewhere, but rather that agricultural productivity alone influences profits generally in the event that corn alone enters the wage basket; and, secondly, that it does so by way of the effect of the price of corn upon money wages. According to this view, in the event that the wage basket contains manufactured goods, an alteration in the prices of manufactures also may alter the general profit rate by way of precisely the same mechanism. The argument depended upon the supposition that changes in the money wage rate do not generate changes in final price, a supposition which, in Ricardo’s case, hinged upon a labour theory of exchange value. According to my argument, the position adopted in the Essay – indeed earlier – is precisely that of the Principles, although clarification, correction, the explicit formulation of assumptions that were only implicit in the early versions, and specification of the conditions required of the money medium were required. In brief there was no ‘transition’ between an early model and a later and different model. The restatement of Professor Sraffa’s position by Mr Eatwell leaves a great deal to be desired. I have in mind particularly his cryptic comment in parentheses on p. 184 that ‘to attain the required result it is only strictly necessary that the wage in other sectors should contain corn. But it is clear that the wage in this context is what was to be called the “natural price of labour” ( Principles, p. 93), which was assumed to be 45

RICARDO - THE NEW VIEW

everywhere the same’. The fact is that Sraffa’s ‘rational foundation’ for the principle of the determining role of the agricultural profit rate depends in a very central way not only on the assumption that the wage basket consists entirely of corn but also on the assumption that the corn wage is fixed in all industries. This latter condition is implied by the fact that in agriculture total profits and the profit rate are said to be determined without reference to valuation, thus precluding the possibility of falling per capita corn wages due to a relative rise in the price of corn. It is also implied by the argument that the profits in trades outside agriculture come into line simply by way of alteration in the exchange values of their products to that of corn, which rules out quantity variations from their input accounts. 2 It is unfortunate that Mr Eatwell deals with so important an issue in such an ambiguous manner. Ricardo’s treatment of the two assumptions in question is the ultimate test of the validity of the Sraffa interpretation. In the Essay, unlike the early correspondence, Ricardo reverted in general to the assumption that corn alone enters the wage basket; since the commodity wage is presumed for the argument to be constant this implies that the corn wage is constant. I do not believe that the formal assumption carries with it any substantive implications. In particular, it does not, I think, imply that he had in mind the model which Professor Sraffa attributes to him for the following reasons. In the first place, when Ricardo deals specifically with the question of the link between the agricultural and general rates of return, his argument – as in the earlier correspondence – turns upon the relationship between money wages and money prices (see, for example, the evidence cited in Hollander 1973, 276–7, 277n [above, pp. 31–2]). Corn is not, as a matter of principle, the only ‘basic’ good and, logically, a change 3 Here, in the price of any wage good would carry with it the same consequences for profits. Ricardo’s comment of 18 December 1814 must be recalled: ‘I admit ... that commerce, or machinery, may produce an abundance and cheapness of commodities, and if they affect the prices of those commodities on which the wages of labour are expended they will so far raise profits’ (Ricardo 1951, VI, 162). Secondly, we must bear in mind Ricardo’s insistence in 1814, already discussed in the original article, that the profit rate will decline upon resort to increasingly inferior land despite declining corn wages (Hollander 1973, 266 [above, p. 23]). Thirdly – and this matter is taken up in Section IV below – Ricardo’s post-publication defence of his Table without question takes for granted a mixed wage basket and falling corn wages. III. What then are Mr Eatwell’s principal objections to my article? He finds (1975, 183) ‘most extraordinary’ my claim that Ricardo’s theory ‘required a demonstration that 46

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

higher money wages are not entirely passed on in the form of higher prices’, and he speaks of my belief that wages were not assumed to consist only of corn as an ‘assertion’. All I can say to this is that I have asserted nothing for which there is no evidence, while Mr Eatwell chooses not to address himself to my use of sources. One exception to this, the second paragraph of the comment, is not well taken. The whole point of my paper is that the money wage rate – the cost of production of money of course unchanged – varies with the price of wages goods in general, not just corn; moreover, Ricardo did not maintain as a matter of principle that the corn wage is constant. I suspect, therefore, that my claims appear extraordinary simply because Mr Eatwell does not believe that they accord with his preconception of Ricardian economics and its place in the history of economic thought. Consider in particular the closing sentences of his paper. If my argument indeed obscures the significance of the Essay as the first complete statement of a theory of distribution based on the concept of surplus, but if it is soundly and accurately based on textual evidence, the logical step, I suggest, would be to re-examine the initial preconception. For my part I do not think that my interpretation carries with it any dire consequences, one way or the other, for Ricardo’s conception of the nature of profits. All I am saying is that the position of the Principles appears, in embryo, much earlier than is usually believed – that Ricardo was preoccupied by the link between the money price of corn, the money wage and the money price of manufactures (on the assumption that the ‘value’ of money is constant) at a very early stage (see the evidence cited in Hollander 1973, 266–8 [above, pp. 23–5]). I find Mr Eatwell’s reaction exaggerated in the light of the extensive analysis of the effects of moneywage variations on prices and profits in the Wealth of Nations which constituted the ‘received doctrine’ Ricardo felt obliged to reject, and which preoccupied him throughout the Principles. That Ricardo ‘located his theory [of profits] within the conditions of production’, as Mr Eatwell insists (183), is not at issue if by this is meant that he rejects the Malthusian conception of things. 4 I certainly nowhere maintain, as Eatwell suggests, and it is not implied by my argument, that Ricardo yielded to Malthus something of substance, or that his position carried with it the logical necessity to yield something of substance, regarding demand variations as a determinant of profits. Events in the manufacturing or commercial sector certainly could affect the profit rate in Ricardo’s view, but positively not in the manner envisaged by Malthus. Mr Eatwell also talks (185) of the ‘successful demonstration’ in the Table of the Essay on Profits of Ricardo’s propositions, asserting ‘that the Table was expressed in terms of a single commodity ... on the basis of analytical insight, coupled with what might be believed to be a reasonable approximation to the real state of affairs’, and 47

RICARDO - THE NEW VIEW

that ‘the success of the Essay on Profits in demonstrating Ricardo’s propositions and the formal basis of that success are presumably indisputable’. Scarcely a word is said to support these assertions, which I take to mean that Ricardo, quite deliberately, formulated the principle that it is the profits of the farmer that regulate the profits of all other trades on the ‘rational foundation’ described by Sraffa. (Eatwell presumably is not merely expressing a view regarding the logical validity of such rational foundation for a principle of the determining role of agricultural profit, since this latter issue is not in dispute.) What I find troublesome is the concession immediately following, that ‘what may be disputed is whether Ricardo deliberately presented his analysis in terms of physical ratios’. If this is the case and if there is room for argument about this central matter, as I believe there is, then it is hard to see why we are expected to 5 In any event, as ‘conclusive take for granted as ‘indisputable’ Ricardo’s ‘analytical insight’. evidence’ for what is described as ‘the importance of a “corn wage” assumption in Ricardo’s analysis, and his deliberate framing of the argument in physical terms’ (185), we are referred to a letter by Malthus (to Horner) dated 14 March 1815 and a letter from Ricardo to Malthus of 27 March. Reference is also made in a note (186) to a letter of 8 May. 6 The letters to which Eatwell refers are part of a complex exchange lasting for several weeks and comprising more than a dozen items of relevance for the interpretation of the Table. Ricardo’s defence of his procedures during this exchange makes it clear that he did not base his Table upon the assumption that corn is the only wage good and that the wage is fixed in corn terms. Moreover, when obliged under pressure from Malthus to spell out the implied assumptions, a variety of price calculations which underlie the Table come to the fore. The citation by Mr Eatwell of isolated passages from an exchange which, considered as a whole, yields conclusions which are the precise reverse of those he suggests is, to put it mildly, questionable scholarly procedure. IV. Malthus’s principal objection to the Table was the failure of Ricardo, as he saw it, to recognize that a relative increase in the price of corn to those of manufactures will reduce total production costs in terms of corn in light of the presence of manufactured items in capital, including the wages component thereof. Professor Sraffa (1951, xxxii n) emphasized the objection, but he neglected to consider Ricardo’s defence. The point to note is that Ricardo positively did not rest his case upon the assumption that corn was the only basic good, and he agreed willingly that the per capita corn wage would fall as the price of corn increased secularly (a result later emphasized in the Principles ).

48

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

Ricardo conceded at the outset of the debate that the consequences of a rise in the price of corn would indeed be, as Malthus insisted, reduced total costs in corn terms and consequently an increase in the corn surplus and in the rate of surplus – but only in the case that agricultural extensions are ruled out: ‘I cannot hesitate in agreeing with you’, he wrote, ‘that if from a rise in the relative value of corn less is paid [in terms of corn] for fixed capital and wages, —more of the produce must remain for the landlord and farmer together, – this is indeed self-evident, but is really not the matter in dispute between us’ (letter of 14 March 1815; Ricardo 1951, VI, 189). Beyond this he refused to go, for his attention was entirely on the longrun cost determination of prices. Accordingly, he rejected Malthus’s contentions on the grounds that any rise in the price of corn reflected resort to inferior land, and necessarily implied an increase in total labour and fixed capital costs (in terms of corn) relative to the increase in the surplus. At the same time, and this is quite crucial for an accurate appreciation of Ricardo, he agreed quite readily that the per capita corn wage will indeed decline during secular advance : A rise of the price of corn and a fall in the corn price of labour is in my opinion incompatible, unless it be owing to something in the currency and it is not necessary to enquire here into what effects that would produce. Observe that I do not question that each individual labourer might receive a less corn price of labour because I believe that would be the case, but I question whether the whole corn amount of wages &c. paid for the cultivation of the land can be diminished with an increase of the exchangeable value of corn. If no more labourers were employed and the price of corn rose your proposition could not be disputed, but the cause of the rise of the price of corn is solely on account of the increased expence of production. (Ricardo 1951, VI, 189; emphasis added) This position is consistent with the tabular formulation of the Essay, according to which – despite an increase in the corn surplus or ‘total produce in quarters of wheat, after paying the cost of production’ – the ratio of the corn surplus to corn costs declines steadily. But Ricardo’s defence on the grounds that a rise in the corn price necessarily implies a reduced ratio can scarcely be regarded as a rigorous one; we turn now to consider this issue in some detail for the various arithmetical illustrations devised by Ricardo in defence of his position provide us with a splendid opportunity to pin down the precise assumptions of the analysis, and thus arrive at an accurate evaluation of the Table. In the first place, a very brief example was given to support the statement already alluded to ‘that corn can only permanently rise in its exchangeable value when the real expences of its production increase’. In the event that 5,000 quarters require an input consisting of ‘wages &c.’ amounting to 2,500 quarters, while 10,000 quarters 49

RICARDO - THE NEW VIEW

require 5,500 quarters, the price of corn will increase by 10 per cent (VI, 189). This conclusion is based upon a principle of ‘average-cost’ rather than ‘marginal-cost’ pricing, and the proposition that an increase in the (average) cost of corn production (in corn terms) leads to a proportionate increase in the money price of corn, implies that money prices are determined by relative (average) corn costs of production and that the corn cost of producing a unit of money is unchanged. 7 Now Malthus commented upon Ricardo’s failure to emphasize ‘marginal’ costs in his calculations: ‘You seem to forget ... that it is only the last 500,000 quarters of corn that have been added to the mass, which require for their production a greater quantity of capital’ (letter of 15 March; in Ricardo 1951, VI, 190–1). To this criticism Ricardo replied: I think you err in supposing it possible that the proportion of the whole corn expenditure, to the produce obtained, can fall, with an increase of the price of corn. The two are incompatible, either the whole corn expences of production will be increased or not. If they be the price of corn will rise, – but if they be not I can see no reason for a rise in the price of corn. I admit that it is only the last portion of capital employed on the land which will be attended with an increased corn expence, but unless it renders the whole produce together at an increased expence the price of produce will not rise. (Letter of 17 March; Ricardo 1951, VI, 192–3) The emphasis thus remains upon average-cost pricing, but Ricardo relies upon the proposition that in light of diminishing returns at the ‘margin’ average costs must increase. The illustration devised to explain the case more fully (see Table 1) explicitly assumes the wage basket to be composed of a fixed-proportion mix of corn and manufactured necessaries, the commodity wage to be constant, and the corn wage to fall as the relative price of corn rises. Pricing is still on an average productivity basis, but in the present case Ricardo utilizes, at least provisionally, a labour-quantity rather than a labour-cost theory of relative price. Ricardo’s objective in devising the illustration was to demonstrate that an increase in the corn price necessarily implies a rise in the average corn costs of corn production, or a rise in ‘the proportion of the whole corn expenditure to the [corn] produce obtained’. Since he utilized a labour-quantity theory of pricing, the price of corn rises by 20 per cent while average [corn] costs rise by only 5 per cent. But even this latter consequence simply follows from the particular assumption made regarding the labourer’s expenditure pattern; it is quite possible for average corn costs in corn terms to decline in the event that the corn component in the basket is accorded a somewhat higher weight so that the per capita decline in corn wages takes on greater significance. In brief, Ricardo completely failed to make his case on the basis of a labour-quantity theory of value. 50

2.5

4.5

£4

£4.16.0

I

II

15

10

3 Output (million quarter)

1

1

4 Per capita corn consumption (quarters)

£4

£4

5 Per capita consumption of manufactures 1.75a

2

6 Per capita corn wages (qs)

Source : Ricardo 1951, VI, 193. Note: Ricardo’s arithmetical calculations are slightly faulty. The following corrections are required: a. 14.64 bushels = 1.83 qs rather than 14 bushels = 1.75 qs b. 8,235,000 c. 5,490,000 d. 0.549

2 Employment (million men)

1 Corn price per quarter

Table 1

7875b

5000

7 Total corn w ages (000 qs)

5250 c

5000

0.525d

0.5

8 9 Total corn Corn costs wages for per quarters output of 10 m. (000 qs)

RICARDO - THE NEW VIEW

Malthus objected strongly to the illustration. He pointed to the irrelevance of an argument which involved an enormous addition to output of some 50 per cent. (letter of 19 March; in Ricardo 1951, VI, 195. He was, however, himself obliged to concede that his own case held good only when agricultural expansions are truly ‘marginal’.) And with regard to Ricardo’s pricing principle Malthus insisted that ‘the natural price of corn depends entirely upon the price of the last additions, and it does not matter whether with regard to the old land, a capital yields 50 per cent (rent and profit together), or 20 per cent. In either case the price of corn on such land has nothing to do with the cost of production.’ In his reply in a letter of 21 March (VI, 197–8) Ricardo maintained that the size of the incremental addition to output was irrelevant – which is perfectly true since his calculations were based on average costs; and that while he agreed with Malthus’s second observation he could not see how ‘the admission of this fact can assist [Malthus’s] argument, which relates only to the ratio of the surplus produce to the whole capital employed’. Ricardo inexplicably failed to see the relevance of the suggested pricing rule for his own calculations. Ricardo took the opportunity in his next letter to point out that his adversary made too cavalier a transfer from the proposition that the ratio of corn surplus to corn costs tended to rise secularly to the proposition that the money value of the surplus relative to money costs tended to rise. Ricardo observed first, that the latter proposition might hold good while the former was false; secondly, that he could accept the proposition expressed in money terms, since, when allowance is made for rent, the profit rate might still decline; and finally, that the ratio of corn surplus to corn costs must in fact fall : You have I think totally changed your proposition. You before contended that in consequence of increasing wealth and the cultivation of poorer land, the whole corn cost of production on the land would bear a less proportion to the whole corn produce, – but now you say that the money cost of production on the land will bear a less proportion to the money value of the whole produce. Between these propositions there is a very material difference, as the latter might be true, at the very time that the former was false. To admit what you now contend for would not affect my theory, as though it would prove that the landlord and tenant (together) got more money revenue, or if you will a greater proportion of money revenue as compared to the money capital employed, yet the tenant might and I think would get a less proportion, and therefore the rate of profit would fall. Such a state of price is quite compatible with a greater proportion of men, as compared with the produce obtained, being employed on the land; – but it is wholly 52

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

irreconcileable with the net corn produce bearing a larger proportion to the gross corn produce, – which was the principle before contended for. (Letter of 27 March; Ricardo 1951, VI, 204) In order to justify the foregoing propositions Ricardo devised a further arithmetical illustration (letter of 17 April, VI, 213). As a preliminary, he formally assumed that ‘the prices of all commodities rise, with the rise of the price of corn, excepting those only on which the wages of labour are expended, and that in consequence the corn wages of labour fall’ – his objective being to demonstrate that even on Malthus’s Smith-like assumptions regarding the effects of corn price increases on general prices, the agricultural rate must decline. However, the assumption that manufacturing goods, other than wage goods, rise in price plays no part in the illustration at all; while the assumption that the prices of manufactured wage goods are constant, so that the corn wage declines, is in fact acceptable to Ricardo as we have seen. The example is in effect based on a thoroughly ‘Ricardian’ foundation – a constant commodity wage, a rising secular price of corn but constant prices of all other goods, including manufactured wage goods, and a falling corn wage. This illustration (Table 2) is more complex than the first for it explicitly distinguishes between rent and profit and focuses attention upon the margin of cultivation in calculating movements in the rate of profit. A second characteristic is the formal emphasis upon money revenue and money costs.8 But there remain distinct contrasts with the Principles and a number of serious errors. The corn price increase of 12.5 per cent which is envisaged reflects the rise in average wage costs expressed in money terms; this, however, involves a case of circularity since the money costs of labour depend upon the price of corn relative to that of manufactures in the first place. 9 And we are faced with the problem that the wage costs per unit of the monetary commodity cannot be taken as constant. The contrast with the Principles, it will be observed, is defined particularly sharply by the fact that the money value of the revenue yielded on the marginal land rises, from 320 to 360, while in the Principles the objective of Ricardo’s calculations is to assure the constancy of the value of the output yielded at the margin of cultivation. Ricardo thus failed to demonstrate that his result was a necessary outcome, in the sense that increased money wages in an output of constant money value must necessarily reduce profits. 10 What then can be deduced from this post-publication defence of the Table of the Essay? Clearly in defending the principles of his table, Ricardo did not assume the wage rate to be fixed in terms of corn and the wage basket to consist only of corn. Ricardo upheld the propositions, particularly the supposed secular fall in the ratio of corn surplus to corn costs, even when full allowance is made for a mixed wage basket and reductions in per capita corn wages. There are no price calculations in 53

RICARDO - THE NEW VIEW

the table of the Essay, it is true, but this it now seems clear is merely because Ricardo believed his main results followed notwithstanding a decline in the corn wage in consequence of the rise in the corn price. When obliged to spell out his position, he found it necessary to make explicit the price calculations which underlie the table. (Eatwell’s note 1, p. 185, is unjustified for this reason.) These facts suggest to me that the corn calculation of the table cannot carry the weight which Professor Sraffa suggested might be placed upon it. The table merely served to portray Ricardo’s main themes in the simplest conceivable arithmetical form. This is my conclusion in the article (Hollander 1973, 281–2 [above, pp. 35–6]) and the new evidence leads me to stand by it.

Table 2 Land I (marginal)

Land I

4 8 4

412 8 4

412 812 4

16 16

18 16

18 16

32 8 256 (64 qs) 360 64 (16 qs) 25 0

34 c.712 272 (60.5 qs) 360 88 (19.5 qs) 32 12.3 b

34 c.712 289 (64.2 qs)

1 2 3

Corn price per quarter (£) Labour input to produce 80 qs Per capita corn consumption by labour (qs) 4 Money equivalent [3×1] (£) 5 Per capita consumption of manufactures (£) 6 Per capita money wage [4+5] 7 Per capita corn wage [6/1] 8 Total wage costs [2×6] (£) 9 Revenue [80 qs × 1] (£) 320 10 Surplus [9–8] (£) 11 Surplus/wages [10/8] (%) 12 Rent per £100 wages

Land II (marginal)

71 (15.8 qs) 19.7 a 0

Source : Ricardo 1951, VI, 213. Note : Ricardo’s arithmetical calculation is faulty. The following corrections are required: a. 24.5 per cent b. 7.5 per cent

It is clear that the letter of Malthus to Horner dated 14 March, cited both by Sraffa and Eatwell, and Ricardo’s letter of 27 March, cited by Eatwell in his comment, do not carry the implications attributed to them; in Ricardo’s own view his model did not turn upon the assumption that the same physical product appears both in the numerator and denominator of the profit rate expression or that the wage is 54

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

fixed in corn terms, whatever Malthus might have thought. There remains, however, one more item which requires consideration. Malthus also raised the issue of the effects of a reduction in the corn price due to free trade. In this event he forecast a decrease in the rate of profit, as well as in rent. His argument turns on the rise of wages and other capital items expressed in corn terms when the price of corn declines – the mirror image of the case discussed hitherto. Ricardo’s reply must be given in full: I have an account before me of the Capital actually employed on a farm of 200 Acres in Essex. It amounts to £3433 – or about £17 pr. Acre, of which not more than £1100, or £1200 is of that description which is not subject to the same variation of value as the produce of the land itself; for £2200 – consists of the value of the seeds in the ground, the advances for labour, – the horses and live stock, &c. &c. If then the money value of the produce from the land should fall, from facility of production, it must ever continue to bear a greater ratio to the whole money value of the capital employed on the land, for there will be a great increase of average produce per acre, whilst the fall in money value will be common to both capital and produce and it cannot therefore be true that rent, profits and wages can all really fall at the same time. (Letter of 8 May; Ricardo 1951, VI, 226) Mr Eatwell in his note 1, p. 186, refers to Ricardo’s assertion in the final sentence that when the corn price falls due to increased productivity ‘the fall in money value will be common to both capital and produce’ as evidence for his opinion that ‘the idea of the rate of profit as a purely material ratio lingers on in the letters of 1815’. I suggest that the comment implies no such thing. In the first place, the data before Ricardo indicate that two-thirds only of the capital stock in agriculture is made up of farm produce; this, however, is not the important point. What is important is the fact that Ricardo’s answer flies in the face of his position as stated up till that moment in correspondence; in the case of a rising corn price due to reduced productivity he did not make a similar defence, as we have seen at length. In isolation the statement cited by Eatwell seems to be unambiguous in its implications; placed in due context it constitutes clutching at straws – an ill-considered reply on Ricardo’s part which contrasts totally with the preceding sustained defence of the table, and which plays no part in the argument thereafter. V. The correspondence following the Essay is important for much else apart from the defence of the table, and contains analyses of technical change in the production of manufactured wage goods; 11 the process whereby the general profit rate comes into 55

RICARDO - THE NEW VIEW

line with changes in the price of corn by way of money-wage movement; and the effects of money-wages on prices.12 But on my reading all of this merely takes further the line of thought already present in 1814. Eatwell (186) reiterates Professor Sraffa’s opinion that the transition between the Essay and the Principles came at the end of 1815 when having set to work on the Principles Ricardo wrote to Mill: ‘I know I shall soon be stopped by the word price.’ I believe (and have demonstrated in my paper) that the early references to the problem of value are more extensive and systematic. Let me, however, clarify what I hoped was clear in my paper. Corn is the numéraire of the Table, but when Ricardo considers the central question of the determination of the general profit rate it is ‘money’ which is chosen as standard as in the Principles. But obviously I am not asserting that the doctrine of the Principles leapt fully developed from Ricardo’s brain in 1814 or 1815. Apart from the fact that Ricardo frequently assumed in his various letters at this early stage an average productivity rather than a marginal productivity basis for his price calculations, and in addition sometimes confused labour cost and labour quantity, the full specification of the standard had still to be formulated. Throughout 1815 and 1816 Ricardo was preoccupied with the possibility of changes in the ‘value’ of the metal itself, and also with the precise technical coefficients required of the medium. But there was no ‘transition’ during this period, between two systems from the Essay to the Principles such as Sraffa envisages, for these problems were raised directly by the analysis of 1814 and early 1815. VI. I wish finally to emphasize that my interpretation in no way touches upon the ‘usefulness’ of the corn profit model as an analytical structure. It is the historical question of whether we can legitimately attribute such a structure to Ricardo which is the subject matter of my article. The record suggests that we cannot. I would like to point out a most interesting passage of Professor Sraffa’s regarding the temporal sequence of his own theoretical contribution and of his interpretation of Ricardo: ‘It should perhaps be stated that it was only when the standard system and the distinction between basics and non-basics had emerged in the course of the present investigation that the above interpretation of Ricardo’s theory [that of Works, pp. xxxi–xxxii] suggested itself as a natural consequence’ (Sraffa 1960, 93). I believe, in the light of the available documents, that Professor Sraffa read too much into Ricardo. The fact that Ricardo failed to make his case – except on the basis of the particular data artificially inserted into his table – should warn us against attributing to an author recognition of all the assumptions which may be required to make sense of his statements.

56

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

NOTES 1. Mr Eatwell is, however, perfectly correct in taking me to task in his note 1(ii) (187). The observation in note 1(i), while demonstrating a most commendable knowledge of the literature, seems to me irrelevant for my purposes. Other objections will be taken up below. 2. I would bring to the reader’s attention that the assumption of a fixed corn wage creeps into the third paragraph of Eatwell’s text where reference is made to ‘the greater (corn) cost of producing the (corn) wage’ which presumably refers to the given corn wage. 3. I would suggest, as a tentative hypothesis, that the emphasis in the Essay upon corn in the wage basket has its source in two factors. The first is the policy orientation of the pamphlet which may have led Ricardo to work with strong cases to get the message across regarding the effects of corn import restriction. The second is the very strong objection Ricardo took to Malthus’s view that the profit rate is affected by events which do not work their way through changes in the cost of producing wage goods at all, but rather operate in terms of the relationship between aggregate demand and supply. For this purpose it was unnecessary to complicate the argument. 4. The citation of Ricardo’s statement to the effect that ‘the rate of profits and of interest must depend on the proportion of production to the consumption necessary to such production’ is unhelpful. Everything turns upon the phenomena which generate changes in the proportion in question. According to Ricardo ‘the proportion of production to the consumption ... essentially depends upon the cheapness of provisions, which is after all, whatever intervals we may be willing to allow, the great regulator of the wages of labour’; according to Malthus ‘the proportion of production to the consumption necessary to such production seems to be determined by the quantity of accumulated capital compared with the demand for the products of capital, and not by the mere difficulty and expence of procuring corn’ (in Ricardo 1951, VI: 108, 111). The phrase by itself, while catchy, is an empty one. 5. I confess to being a little confused as to who precisely Mr Eatwell considers a hero. Compare the following statements: (a) ‘It was in the search for the correct specification of “necessaries” and of the determination of the rate of profit in their conditions of production, that Ricardo was to make his ingenious simplification in the criticism of which Malthus was to enjoy some success’ (184); (b) ‘In the face of the success of Ricardo’s “material ratio” Malthus was to be forced to rely on the complications of a many sector analysis to obscure the deficiencies of his own argument, a fact which Ricardo was later to point out’ (183, n 1). 6. Professor Sraffa’s suggestion regarding Malthus’s comment of 5 August 1814 has been already dealt with at length in my paper. The quotation given from Malthus’s letter of 12 February 1815 tells us nothing; Malthus himself accepted that ‘it sometimes tends to clear up these matters to put money out of the question’. Mr Eatwell is therefore correct not to regard these items as ‘conclusive evidence’ for his case (see p. 185). 7. This last assumption is not satisfactory since Ricardo’s allowance for a decline in the per capita corn wage implies a decline in corn costs of producing money – even when no changes occur in labour input per unit of the monetary commodity. In brief, Ricardo’s implicit use of a labour-cost, as distinct from a labour-quantity theory involves difficulties which he did not consider. It might, however, be argued that since the fall in per capita corn wages affects labour in all commodities equally this variation can be ignored as a cause of relative price variations so that any changes in money prices reflect an altered labour input only. 57

RICARDO - THE NEW VIEW

8. It is worth noting that while the ratio of total money surplus to total money costs rises from 25 to 28 per cent, that between the total corn surplus and total corn costs declines from 25 to 22 per cent, illustrating Ricardo’s continued adherence to the Table in the Essay,although it is now upon the money calculation that he places the emphasis. 9. Ricardo in fact conceded that if he had chosen a much larger corn price increase he would have obtained a very different result but insisted that ‘some adequate cause must be shown for such a rise and it cannot be arbitrarily assumed’ (1951, VI, 214). 10. It is noteworthy in this regard that Ricardo’s calculation of a profit rate decline from 25 to 19.7 per cent is in error; a correct estimate involves a slight reduction to 24.5 per cent only. The error arises from calculating profits, at the margin, as the ratio of net revenue to revenue rather than to costs. We may also bring to the reader’s attention Ricardo’s attempt to prove that even in the event of general prices rising the rate of return must decline (1951, VI, 213). The proof is a partial one only – representing a reply to Malthus’s challenge to the Table – for it limits the discussion to the agricultural rate alone: ‘I will however suppose that you and Mr. Torrens are correct, and that commodities do rise in price with every increased price of corn. The value of fixed capital as well as of circulating capital employed on the land will then rise also, and altho’ the money value of the produce should be increased on the old land it will still bear the same proportion to the money value of the capital employed and as this produce will be divided in different proportions between the landlord and the farmer the rate of profits of the latter will fall.’ While mention is made of fixed capital Ricardo neglects to take it into account in the actual illustration; no allowance is made for depreciation in costs; and profits are calculated on wages capital only. 11. See in particular Ricardo to Malthus, 17 March 1815 (1951, VI, 194), where increased profits are accounted for by the supposition ‘that wages have been kept moderate by the improvements in those manufactures which supply the poor with the necessaries on which a part of their wages are expended’. 12. The appearance of Robert Torrens’s Essay on the External Corn Trade in February 1815 provided an ideal opportunity for Ricardo to restate his position. Torrens repeated Adam Smith’s fundamental belief that increases in the price of corn are passed on to consumers, by way of rising money wages, and do not affect profits. Ricardo simply replied that Torrens ‘on this part of the subject appears to me defective, as I think that the price of commodities will be very slightly affected either by a rise or fall in the price of corn. If so, every rise in the price of corn, must affect profits on manufactures, and it is impossible that agricultural profits can materially deviate from them’ (letter of 17 April 1815: 1951, VI, 213).

REFERENCES Eatwell, J. (1975) ‘The interpretation of Ricardo’s Essay on Profits’, Economica 42: 182–7. Hollander, S. (1973) ‘Ricardo’s Analysis of the Profit Rate, 1813–15’, Economica 40: 260–82. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol. I: The Principles of Political Economy and Taxation; Vol. VI: Letters, 1810–1815; Cambridge: Cambridge University Press. 58

RICARDO AND THE CORN PROFIT MODEL: REPLY TO EATWELL

Sraffa, P. (1951) ‘Introduction’ to The Works and Correspondence of David Ricardo, Vol. I: The Principles of Political Economy and Taxation, Cambridge: Cambridge University Press. —(1960)Production of Commodities by Means of Commodities, Cambridge: Cambridge University Press.

59

3 PROFESSOR GAREGNANI’S DEFENCE OF SRAFFA ON THE MATERIAL RATE OF PROFIT

I. In two articles (1973, 1975) and in my Economics of David Ricardo (1979) I have given a blow-by-blow account of the evidence relating to Ricardo’s position on profits. Professor Garegnani’s (1982) criticisms do not lead me to retract. Much of his argument is based on circumstantial evidence and fails to make accurate use of what Ricardo actually wrote on the matter, although he is aware of the materials available. The most useful path for me now is to review some of the key issues, limiting myself, however, to an examination of the evidence presented by Sraffa, who originated the hypothesis which is in dispute. There is no point in going further if this matter is not resolved. II. We must have before us the Sraffa position in support of the material theory of profits, ‘the rational foundation’ of the principle expressed by Ricardo in his letter to Trower of 8 March 1814 that ‘it is the profits of the farmer which regulate the profits of all other trades’ (Ricardo 1951, VI, 104): Although this argument is never stated by Ricardo in any of his extant letters and papers, he must have formulated it either in his lost ‘papers on the profits of Capital’ of March 1814 or in conversation, since Malthus opposes him in the following terms which are no doubt an echo of Ricardo’s own formulation: ‘In no case of production, is the produce exactly of the same nature as the capital advanced. Consequently we can never properly refer to a material rate of produce. ... It is not the particular profits or rate 60

PROFESSOR GAREGNANI’S DEFENCE

of produce upon the land which determines the general profits of stock; and the interest of money’ [5 August, 1814]. The nearest that Ricardo comes to an explicit statement on these lines is in a striking passage in a letter of [26] June 1814: ‘The rate of profits and of interest must depend on the proportion of production to the consumption necessary to such production.’ The numerical examples in the Essay reflect this approach; and particularly in the well-known Table which shows the effects of an increase of capital, both capital and the ‘neat produce’ are expressed in corn, and thus the profit per cent is calculated without need to mention price. (Sraffa 1951, xxxi–xxxii) Sraffa, it will be seen, placed considerable reliance when making his case on the formulation in Malthus’s letter of 5 August 1814 which he claims is ‘no doubt an echo of Ricardo’s own formulations’ in the lost manuscript on profits or in conversation. But Sraffa’s argument is entirely unconvincing, for it neglects the context of Malthus’s statement. The statement is in fact preceded by the following: ‘If the nominal price of corn be doubled, and the nominal amount of capital employed, be not quite doubled which you seem to allow might be the case, instead of saying “how is it possible to conceive that the rate of profits will not be diminished” I should say how is it possible to conceive that it should not be increased? In no case etc.’ (Malthus, in Ricardo 1951, VI, 117). Malthus obviously had in mind neither Ricardo’s manuscript nor his conversation as Sraffa suggests, but specifically Ricardo’s letter of 25 July to which he was replying. There is no need then for any further speculation regarding the origin in Ricardo of Malthus’s formulation. Let us then look at the extract in the Ricardo letter of 25 July to which Malthus referred: The capitalist ‘who may find it necessary to employ a hundred days labour instead of fifty in order to produce a certain quantity of corn’ [Malthus, 6 July, VI, 111] cannot retain the same share for himself unless the labourers who are employed for a hundred days will be satisfied with the same quantity of corn for their subsistence that the labourers employed for fifty had before. If you suppose the price of corn doubled, the capital to be employed estimated in money will probably be also doubled, – or at any rate will be greatly augmented and if his monied income is to arise from the sale of the corn which remains to him after defraying the charges of production how is it possible to conceive that the rate of his profits will not be diminished. (Ricardo 1951, VI, 114–15) There is an error to be noted first in Ricardo’s formulation (alluded to, it will be noted, by Malthus). In place of the capital nearly doubling, Ricardo should have 61

RICARDO - THE NEW VIEW

written nearly quadrupling to reflect the two forces at work on costs which reflect the supposed decline in agricultural productivity: (1) a doubled labour input per unit of corn, and (2) a doubled money wage which compensates for the doubled price of corn.1 As Ricardo pointed out, only if the corn wage should fall, and fall in proportion to the reduced labour productivity, would (money) costs merely rise in proportion to the rise in revenue thus preventing a squeeze on profits. The correction, we shall see, was to be made by Ricardo himself. In this passage it is clear enough that corn alone enters into costs: the case is limited in its axioms. 2 But the actual mechanism at work on profits, it is also clear, involves the upward pressure exerted by rising money wages. The fall in corn surplus relative to corn input is merely the outcome of this mechanism. Before elaborating upon this matter, I wish to point out that the ellipses in the citation by Sraffa from Malthus’s letter of 5 August conceal a revealing observation. What Malthus wrote in fact was this, that ‘In no case of production, is the produce exactly of the same nature as the capital advanced. Consequently we can never properly refer to a material rate of produce, independently of demand, and of the abundance or scarcity of capital’ (VI, 117; emphasis added). It is his own theory of profit-rate determination involving the aggregate demand problem that he had in mind: the possibility due to aggregate demand variation of increase (decrease) in money revenue but no proportionate increase (decrease) in wage costs that he believed Ricardo had ruled out. He may have objected to Ricardo’s assumption that expenses amount entirely to wages and that wages are spent on corn, but this says nothing about the mechanism Ricardo had in mind: that involving upward pressure of money wages. Ricardo’s mechanism again appears very clearly later in 1814. And in this case Ricardo carefully corrected what he saw as a misreading by Malthus of his position and thereby cast a flood of light on his precise intentions. Malthus’s statement must be given first: The Profits of stock, or the means of employing capital advantageously may be said to be accurately equal to the price of produce, minus the expence of production. And consequently whenever the price of produce keeps a head of the price of production the profits of stock must rise. ... It is not the quantity of produce compared with the expence of production that determines profits, (which I think is your proposition) but the exchangeable value or money price of that produce compared with the money expence of production. And the exchangeable value of produce is not of course always proportioned to its quantity. ... In stating the cause of high profits you seem to me to consider almost exclusively the expence of production, without attending sufficiently to the price of produce, and greatly to underrate the wants and tastes of mankind in affecting prices, and consequently in affecting 62

PROFESSOR GAREGNANI’S DEFENCE

the means of profitably employing capital. (9 October 1814; Malthus, in Ricardo 1951, VI, 140–1) Again we have Malthus’s complaint that Ricardo neglected price of output and revenue, and did so by focusing on the physical output of corn. Malthus’s observation to Ricardo regarding physical quantities – ‘which I think is your proposition’ – is the first point to be noted. Had Ricardo explicitly stated a material theory in the lost papers on profits or in conversation this is surely not how Malthus would have phrased the matter. There would have been no room for doubt. But there is no need at all for surmise; for Ricardo explicitly denied that Malthus’s attribution was an accurate one : It does not appear to me that we very materially differ in our ideas of the effects of the facility, or difficulty, of procuring food, on the profits of Stock. You say ‘that I seem to think that the state of production from the land, compared with the means necessary to make it produce, is almost the sole cause which regulates the profit of stock, and the means of advantageously employing capital.’ This is a correct statement of my opinion, and not as you have said in another part of your letter, and which essentially differs from it, ‘that it is the quantity of produce compared with the expence of production, that determines profits’. You, instead of allowing the facility of obtaining food to be almost the sole cause of high profits, think it may be safely said to be the main cause, and also a difficulty of acquiring food the main cause of low profits. There appears to me to be very little difference in these statements. (23 October, Ricardo 1951, VI, 144–5) Ricardo thus accepts responsibility for the general notion that (setting aside shortrun fluctuations) profits are governed by agricultural productivity, but no more than that. Now the operation of agricultural productivity works its way on profits through the money prices of corn and labour ; Malthus was wrong in attributing to him a neglect of corn prices. This is clear from the same letter, where Ricardo wrote regarding the effects of accumulation: ‘A rise in the price of raw produce may be occasioned by a gradual accumulation of capital which by creating new demands for labour may give a stimulus to population and consequently promote the cultivation or improvement of inferior lands, – but this will not cause profits to rise but to fall, because not only will the rate of wages rise, but more labourers will be employed without affording a proportional return of raw produce. The whole value of the wages paid will be greater compared with the whole value of the raw produce obtained’ (VI, 146). To reiterate ‘the principle’ here formulated: the agricultural profit rate falls with 63

RICARDO - THE NEW VIEW

diminishing returns because while the price of corn rises and revenue (for any given output) increases to reflect the lower productivity, there are two (related) forces playing on the costs required to produce that output: an increased labour input, which is paid a higher money wage to reflect the higher corn price. This sequence is exactly that earlier expressed on 25 July with the correction required inserted. Ricardo, one may add, did not deny the consequences for profits of short-run variations in the corn price, due for example, to bad seasons: profits rise since ‘the price of produce would rise considerably more than in the proportion to the deficient quantity, and would therefore be much a head of the price of production’, i.e. money costs; similarly, general price inflation ‘would raise the price of produce, for a time, more than it would wages, and would therefore raise profits’. But these were ‘temporary causes, no way affecting the principle itself, but merely disturbing it in its progress’ (VI, 146). In all cases, it is the behaviour of the money price of corn and money wage costs that provides the key. It will be noted how very close the general system of 1814 is to that of 1817. The difference in approach is a formal one only: in 1817 Ricardo holds labour input unchanged and assumes the yield to fall; profits are squeezed since the money value of the (lower) output remains constant while the money costs of production (the money wages paid the given work force) rise. In the earlier version of 1814, as noted, the money value of a given output rises, but the money costs, representing a greater labour input, rise in greater proportion. But in both cases the price of corn increases to reflect reduced labour productivity alone and not the increased money wage rate so that profits are forced down. There is much more pointing to my conclusions. The very statement in the letter of 26 June upon which Sraffa put so much weight as ‘the nearest that Ricardo comes to an explicit statement on [material] lines’ (above, p. 61) – ‘the rate of profits and of interest must depend on the proportion of production to the consumption necessary to such production’ – surely does not support Sraffa. For Ricardo himself tells us on the spot that this proportion ‘essentially depends upon the cheapness of provisions, which is after all, whatever intervals we may be willing to allow, the great regulator of the wages of labour’ (VI, 108). III. I do not see how much clearer Ricardo could have been that his doctrine of 1814 regarding the squeeze on profits involved the pressure exerted by money wage. But how does Garegnani deal with the evidence? He engages in impressive shadow boxing in his efforts to avoid my conclusion; he does not hit the real target. He concedes the existence of Ricardo’s references to money wages as the regulator of profits (e.g. 64

PROFESSOR GAREGNANI’S DEFENCE

p. 72); indeed it would be impossible to do otherwise. I do not, however, find that he provides any general rationalization of Ricardo’s repeated insistence upon the role of money wages. He only attempts, if I understand him correctly (73), to avoid the obvious conclusion by contending that Ricardo maintained the material theory of profits – ‘reasoning on the basis of the physical quantities of agriculture’ – but (legitimately) switched to money value terms without compromising the Sraffa logic. What is supposedly involved is ‘an underlying reasoning in physical terms [which] is used to argue about the behaviour of money values’ (77). To take this position is surely to confuse cause and effect. A corn calculation can always be made even within the fullfledged Ricardian doctrine of the Principles ; assuming a simple case where corn alone enters the wage basket the calculation is very easy. If the profit rate falls in agriculture, it will then be the case that the ratio of corn surplus to corn costs declines. But this is not a causal matter; it is not this that determines the profit rate, it is a consequence. The causal mechanism involved for Ricardo the relation between corn prices and money wages; as he phrased it so very simply, the physical proportion ‘essentially depends upon the cheapness of provisions, which is ... the great regulator of the wages of labour’. IV. I shall return to the foregoing matter when I take up the precise content of the wage basket. Let us now look more closely at the general profit rate. I take note first of Garegnani’s objection (68) that I illegitimately equate Ricardo’s comment in the letter to Trower of 8 March that ‘it is the profits of the farmer which regulate the profits of all other trades’ (VI, 104) with ‘the state of agricultural productivity’ as the determining factor, thereby arriving at my more general interpretation. My reply is that there is nothing illegitimate about the identification, since I merely follow Ricardo’s own practice. In that same letter to Trower Ricardo himself emphasizes that general profits vary with ‘improvements in husbandry, – or new facilities ... for the introduction of food from foreign countries’ (obviously features defining the state of agricultural productivity), and then restates this formulation in terms of what is obviously nothing more than a catch phrase: ‘in short it is the profits of the farmer which regulate the profits of all other trades’ (104). Conversely, he objects to Malthus’s view ‘that the profits of the farmer no more regulate the profits of other trades, than the profits of other trades regulate the profits of the farmer’, and identifies this proposition with a denial of his own insistence upon the role played by agricultural productivity (profits possibly rising for Malthus ‘altho’ there should be no new facilities either by importation, or improved tillage, for the production of food’). 65

RICARDO - THE NEW VIEW

Now to substantive matters. In his letters to Malthus of 26 June and 25 July – both of which contain the money-wage mechanism as we have seen – the effect of a higher corn price on the manufacturing sector is taken up. 3 While increased money wages (due to the higher corn price) will cause manufacturing prices to rise, demand for these products is constrained: ‘The rise of the price or rather the value of corn ... must necessarily diminish the demand for other things even if the prices of those commodities did not rise with the price of corn, which they would (tho’ slowly) certainly do. With the same Capital there would be less production, and less demand’ (VI, 108). Similarly: You say that ‘the proportion of production to the consumption necessary to such production, seems to be determined by the quantity of accumulated capital compared with the demand for the product of capital, and not by the mere difficulty and expence of procuring corn’. It appears to me that the difficulty and expence of procuring corn will necessarily regulate the demand for the products of capital, for the demand must essentially depend on the price at which they can be afforded, and the prices of all commodities must increase if the price of corn be increased. (Ricardo 1951, VI, 114) On 11 August Ricardo explained further – always in the same context of import restriction – how reduced demand for manufactures constrained the rise in manufacturing prices behind the increase in money wage costs: It is true that the Woollen or Cotton manufacturer will not be able to work up the same quantity of goods with the same capital if he is obliged to pay more for the labour which he employs, but his profits will depend on the price at which his goods when manufactured will sell. If every person is determined to live on his revenue or income, without infringing on his capital, the rise of his goods will not be in the same proportion as the rise of labour, and consequently his percentage of profit will be diminished if he values his capital, which he must do, in money at the increased value to which all goods would rise in consequence of the rise of the wages of labour. (Ricardo 1951, VI, 119–20) I have always emphasized that Ricardo’s argument regarding the effect of higher wages on manufacturing prices was to undergo further changes and correction. But the point will be clear that the mechanism whereby manufacturing profits were already envisaged in 1814 as coming into line with those in agriculture entails upward pressure on money wages due to higher corn prices, and not that claimed by Sraffa. This general principle is that of the Essay on Profits and the Principles. That it is indeed the principle of the Essay requires a word of support, for Sraffa in his 66

PROFESSOR GAREGNANI’S DEFENCE

interpretation denied it, arguing that in Ricardo’s statement of 1815 ‘the profit rate is calculated without the need to mention price’ (above, p. 61). What Sraffa neglected, however, is that while this is true of the Table it is not true of the essay as a whole, the Table constituting a simplified pedagogical device. Thus we read in the Essay: ‘The sole effect then of the progress of wealth on prices, independently of all improvements, either in agriculture or manufactures, appears to be to raise the price of raw produce and of labour, leaving all other commodities at their original prices, and to lower general profits in consequence of the general rise of wages’ (Ricardo 1951, IV, 20). Could anything be more explicit? We return now to 1814 to deal with one further issue: the content of the wage basket. In all the citations so far from Ricardo’s correspondence it appeared that corn alone is included. But Ricardo allowed that this was not a necessary part of the argument and explained that should manufacturing goods be consumed by labour, a change in their prices will also affect profits, again by acting on the money wage: I have been endeavoring to get you to admit that the profits on stock employed in Manufactures and commerce are seldom permanently lowered or raised by any other cause than by the cheapness or dearness of necessaries, or of those objects on which the wages of labour are expended. Accumulation of capital has a tendency to lower profits. Why? because every accumulation is attended with increased difficulty in obtaining food, unless it is accompanied with improvement in agriculture; in which case it has no tendency to diminish profits. If there were no increased difficulty, profits would never fall, because there are no other limits to the profitable production of manufactures but the rise of wages. If with every accumulation of capital we could tack a piece of fresh fertile land to our Island, profits would never fall. I admit at the same time that commerce, or machinery, may produce an abundance and cheapness of commodities, and if they affect the prices of those commodities on which the wages of labour are expended they will so far raise profits; – but then it will be true that less capital will be employed on the land, for the wages paid for labour form a part of that capital. (18 December 1814, Ricardo 1951, VI, 162) V. I cannot see how there can be any doubt whatsoever that it is changing money wages, whatever the components of the wage basket, upon which Ricardo relied for his analysis of profits in 1814. There is no basis for the Sraffa interpretation. But what has Garegnani to say about all this? 67

RICARDO - THE NEW VIEW

Just as he does not deny Ricardo’s references to the action of money wages in the agricultural sector, he also does not deny that ‘occasionally’ Ricardo allowed for the presence of non-wage items in the wage basket. Indeed, he is troubled by the fact that Ricardo frequently switches from the simple to the complex wage basket, referring to this practice as an ‘inconsistency’. But it is, he asserts, an inconsistency which disappears when ‘we return to Sraffa’s interpretation of the principle of the determining role of the profits of the farmer’. For, the basis of Ricardo’s argument lay in the physical corn quantities of agriculture and therefore, implicitly, on the simplification of wages consisting entirely of corn with its approximate correspondence to reality. An analogy with the corn argument would, however, allow Ricardo to see easily, and occasionally admit, that productivity in the non-agricultural wage goods sectors could also influence the rate of profit. In fact the fall in the general profit rate resulting from decreasing agricultural productivity, which Ricardo could determine on the basis of the quantities of corn in agriculture, would have to operate through a fall in the ratio between value of product and value of wages: it was then easy for Ricardo to think that the analogous changes in labour productivity in non-corn wage goods would act similarly. However, this analogy would in no way alter Ricardo’s need to hold on to the argument founded on physical corn quantities – and therefore to the principle of the determining role of farmers’ profits – as the only firm logical foundation of his admittedly approximate conclusions. (Garegnani 1982, 69–70) [Ricardo] was, in fact, concerned with conclusions applicable to reality and not, merely, with correct deductions from an assumption – that of wages consisting entirely of corn – the realism of which could, and would, be immediately disputed. Given the decisive importance of agricultural products in agricultural capital, he would feel confident that conclusions reached by adopting the simplification of wages consisting entirely of ‘corn’ would be of general validity: and he would argue these conclusions with reference to a reality where agricultural capital consisted mostly, though not exclusively, of corn. Indeed ... he never rigorously adhered to the simplification of corn wages and, occasionally, even went on to consider the effects on the rate of profit of improvements in the production of wage goods other than corn. (Garegnani 1982, 71-2; cf. 74) Now, with respect, this represents a very forced defence of Sraffa; Garegnani presumes the legitimacy of the Sraffa interpretation and then (not surprisingly) finds 68

PROFESSOR GAREGNANI’S DEFENCE

himself obliged to engage in pure speculation regarding Ricardo’s mental process. And it is a defence that actually undermines the logic which Sraffa ascribes to Ricardo. Let us recall that logic: The rational foundation of the principle of the determining role of the profits of agriculture, which is never explicitly stated by Ricardo, is that in agriculture the same commodity, namely corn, forms both the capital (conceived as composed of the subsistence necessary for workers) and the product; so that the determination of profit by the difference between total product and capital advanced, and also the determination of the ratio of this profit to the capital, is done directly between quantities of corn without any question of valuation. It is obvious that only one trade can be in the special position of not employing the products of other trades while all the others must employ its product as capital. (Sraffa 1951, I, xxxi) For Sraffa, therefore, it is clearly essential to the presumed ‘rational foundation’ that agricultural capital consist entirely of corn wages; without that assumption the argument collapses. Worse still on Garegnani’s reading, when allowance is made for noncorn wages, a fall in the general profit rate must occur ‘through a fall in the ratio between value of product and value of wages’. But this too flies in the face of Sraffa’s statement of the mechanism, namely that ‘if there is to be a uniform rate of profit in all trades it is the exchangeable values of the products of other trades relatively to their own capitals (i.e. relatively to corn) that must be adjusted so as to yield the same rate of profit as has been established in the growing of corn’ (xxxi). Since it is so apparent that Ricardo does not insist on a corn wage basket, and since this creates no difficulties whatsoever if understood within a general approach to profits involving the effects of money-wage changes, I suggest Garegnani follow me in abandoning the Sraffa argument. In any event, there is something very wrong when unnecessarily complex arguments are used to force Ricardo’s relaxation of the corn-capital assumption into a particular mould which is not logically required. VI. There is much else that could be said in response to Garegnani’s challenge. But by focusing upon the evidence presented by Sraffa we deal with the essentials of the problem. I submit that Sraffa’s evidence is unconvincing for the reasons given, and that Garegnani’s defence does not in fact serve its purpose. My interpretation, by contrast, has the merit that it does not turn on the possible content of some apparently no longer extant ‘more explicit statements’ by Ricardo of his position. I base my case on the open books before us. 69

RICARDO - THE NEW VIEW

NOTES 1. I failed to note this simple correction in Hollander 1973, 266n [above, Chapter 1, n 13]; and, as will become clear, I was needlessly preoccupied with this formulation. 2. Garegnani (1982, 71) unnecessarily weakens his own case by allowing for the possibility of non-corn inputs in this context. 3. The question at hand relates to the consequences of corn import restriction.

REFERENCES Garegnani, P. (1982) ‘On Hollander’s Interpretation of Ricardo’s Early Theory of Profits’, Cambridge Journal of Economics 6, 1 (March). Hollander, S. (1973) ‘Ricardo’s Analysis of the Profit Rate, 1813–15’, Economica 40 (August); 260–82. ——(1975) ‘Ricardo and the Corn Profit Model: Reply to Eatwell’,Economica 42 (May); 188– 202. The Economics of David Ricardo, Toronto: University of Toronto Press. ——(1979) Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol.IV: Pamphlets, 1815–1823;Vol.VI: Letters, 1810–1815;Cambridge: Cambridge University Press. Sraffa, P. (1951) ‘Introduction’ to The Works and Correspondence of David Ricardo, Vol. I: The Principles of Political Economy and Taxation, Cambridge: Cambridge University Press.

70

4 ON A ‘NEW INTERPRETATION’ OF RICARDO’S EARLY TREATMENT OF PROFITABILITY

In a recent paper Dr Terry Peach (1984) has gravely misrepresented my analysis of Ricardo’s early treatment of profitability in The Economics of David Ricardo (1979) (henceforth EDR). Peach asserts that on my account ‘Ricardo’s treatment must be characterized in terms of the manipulation of money variables; that is in terms of the relationship between money wages (which may be influenced by changes in the price of corn, although this is only one possibility), prices and profits’ (1984, 734), and he refers to ‘Hollander’s belief that Ricardo adopted an undifferentiated money variables approach to the analysis of both the agricultural and manufacturing sectors’ (741n). In making these global assertions Peach distorts my position. For I have always emphasized that Ricardo proceeded in his early writings by distinguishing the agricultural and the manufacturing sectors. Thus, conspicuously, in discussing the Essay on Profits (1815), I refer ‘to his well-known model, relating to the agricultural sector’ in which corn is used as numéraire (EDR,136–7), and reiterate that ‘it seems to have been Ricardo’s intention to apply the model specifically to the agricultural sector’ (138). And I could not have been more explicit that ‘there are no price calculations in the table of the Essay’ (162). I have sought, however, to demonstrate that when Ricardo turned to analyse the economy-wide return on capital he worked within a money-variables framework. The agricultural model served an introductory purpose only; in particular, the material rate does not determine the general rate in the manner perceived by Sraffa. I have also sought to show how the argument matured over time culminating in the Principles, for in the early accounts the ‘general’ theory – involving the impact on profits of upward pressure on money wages reflecting diminishing agricultural returns – is not yet fully developed (140). Apart from the foregoing misrepresentation, Peach also charges me with ‘an unfortunate and ... distorting participation in the ... unhelpful game ... of striving 71

RICARDO - THE NEW VIEW

to establish lineages’ (1984, 751) – a ‘neoclassical lineage’ in my case – the ‘major implication’ of his own paper being, he writes, a demonstration ‘that the interest in assimilating [Ricardo’s] writings to a type of economics is nugatory, at best’ (733). A careful reading of Peach’s paper provides scarcely a hint of the basis for this charge against me – what there is has no justification – nor, for that matter, of the alleged ‘implication’. Limited space precludes further elaboration of these matters. I shall attend rather to the specifics of Peach’s exegesis. In my view Peach fails to substantiate his assertions regarding Ricardo’s analysis of 1814. Furthermore, he adds nothing to what is in my EDR regarding the analysis of the Essay; on the contrary, he neglects much of relevance for the accurate interpretation of that analysis. II. The main theme of Peach’s paper relates to Ricardo’s treatment of the agricultural sector. It is proposed that in 1813–14, before the Essay on Profits, Ricardo assumed that all prices (including the prices of manufactured inputs) rise in proportion with the price of corn – a proposition identified with Smith’s ‘adding-up’ theory; 1 and that in the 1815 Essay itself all prices are constant (notwithstanding extensions to marginal land). By either of these assumptions Ricardo was enabled, so runs the argument, to calculate the material profit/capital ratio (using corn as numéraire ) without assuming literal homogeneity of input and output – the logical assumption Sraffa ascribed to Ricardo’s model – since relative price changes are precluded. That Ricardo sought to estimate the profit ratio in material terms is a presumption turning on Malthus’s objection in correspondence with Ricardo that ‘in no case of production, is the produce of the same nature as the capital advanced. Consequently we can never properly refer to a material rate of produce, independent of demand, and of the abundance or scarcity of capital’ (1984, 737; citing letter of 5 August 1814; Ricardo 1951, VI, 117). But unlike Sraffa, who presumed that Ricardo must have asserted a material rate of produce in the sense that corn is both input and output, Peach has it that Ricardo ‘abstract[ed] from monetary disturbances which would mask the true magnitude of the materially specified profit share’ by assuming equiproportionate price movements of corn and manufactures (737, 741–3). Peach makes heavy weather of Ricardo’s use of a corn calculation. He puts far too much weight on a passage in Ricardo’s reply to Malthus of 11 August, namely ‘Individuals do not estimate their profits by the material production, but nations invariably do. If we had precisely the same amount of commodities of all descriptions in the year 1815 that we now have in 1814 as a nation we should be no richer, but if money had sunk in value they would be represented by a greater quantity of money, and individuals would be apt to think themselves richer’ (Ricardo VI, 121; 72

O N A ‘ N E W I N T E R P R E T A T I O N’

cited Peach 1984, 737). This passage in fact says nothing about the profit rate or profit share – ‘a rate of produce’; it merely insists that wealth is not well represented by money, variable as money is or may be in its purchasing power, but can be measured unambiguously if the mix of commodities remains unchanged – constancy of wealth implying ‘precisely the same amount of commodities of all descriptions’. This general (and non-controversial) observation carries no necessary implications whatsoever for the representation of the profit rate in material terms. Secondly, the passage is not limited to the agricultural sector. This latter point is actually recognized by Peach himself – ‘this technique was not believed to be limited in application to the agricultural sector’ (737) – so that it is difficult to appreciate how he can use the passage as prologue to his attribution to Ricardo of an approach in the agricultural sector based upon the profit share in material terms. Fortunately, all this tortuous meandering is unnecessary. 2 For we find Ricardo explicitly referring to the agricultural profit rate in terms of a corn surplus relative to a corn capital in a letter of 25 July 1814 (Ricardo 1951, VI, 114–15) – the letter to which Malthus was replying on 5 August – and there is also a letter of June 14: ‘the rate of profits and of interest must depend on the proportion of production to the consumption necessary to such production’ (108). In the light of these letters I readily accept the corn calculation (see EDR, 127–8). What, however, is quite unconvincing is Peach’s attribution to Ricardo at this period of ‘contemporaneous and proportional price rises’ (1984, 742), and ‘equi-proportional price changes’ (743). For the outstanding feature of the 1814 correspondence is Ricardo’s new insistence on a less-than-proportional increase in manufacturing prices relative to the presumed money wage and ( a fortiori ) corn price increase upon agricultural extensions: ‘the rise of [the manufacturer’s] goods will not be in the same proportion as the rise of labour’ (11 August; Ricardo 1951, VI, 120; cited EDR,129; a lag in the money wage reaction behind the price of corn is alluded to, 6 June; VI, 108; cited EDR,125). Yet Peach wants us to believe that these facts – of which he is very well aware (1984, 746) – were set aside by Ricardo in the treatment of the agricultural sector ‘for the sake of analytical convenience’ (741, 746). 3 For the attribution of equiproportional price movements not a shred of evidence is provided. The most explicit statement regarding the agricultural sector in fact points away from the attribution: ‘If you suppose the price of corn doubled, the capital to be employed estimated in money will probably be also nearly doubled – or at any rate will be greatly augmented; and if his money income is to arise from the sale of his corn which remains to him after defraying the charges of production how is it possible to conceive that the rate of his profits will not be diminished?’ (25 July 1814; Ricardo 1951, VI, 115). I am not concerned now with the ambiguities of this passage, but with Peach’s admirably candid observation that it ‘apparently conflicts 73

RICARDO - THE NEW VIEW

with [his own] interpretation that Ricardo assumed the prices of manufactures to rise pari passu with the price of corn’ (1984, 745). Yet he still insists on the ‘conjecture’ – albeit ‘only for analytical convenience in formal analysis’. There is one further piece of ‘evidence’ which scarcely merits consideration. Peach refers to the post-Essay correspondence and draws upon a passage in which Ricardo – for the sake of argument – assumes equiproportionate price increases (‘I will ... suppose that you [Malthus] and Mr. Torrens are correct, and that commodities do rise in price with every increased price of corn’) to support his case regarding Ricardo’s pre-Essay position when Ricardo ‘maintained the positive relationship between movements in corn prices and movements in general prices’ (1984, 743). This later formulation is said to provide ‘further support for the interpretation that he went through a stage of assuming equiproportional price changes between corn and other commodities’. It is beyond me how adoption of an assumption in 1815 purely for argument’s sake can be said to indicate that the same assumption had been adopted in 1814. In any event the fact is that the post- Essay ‘argument’ assumes proportionate price increases, rather than the less-than-proportionate increases of 1814. The argument breaks down totally. Peach, in brief, fails to justify his assertion that Ricardo based his statements of 1814 involving a calculation in material terms upon equiproportionate price increases. I think it unlikely for the reasons already given. There is a further matter. A conspicuous feature of Ricardo’s letter of 25 July 1814 is the insistence that the corn surplus relative to corn capital falls with agricultural extensions notwithstanding a decline in the average corn wage – this in contrast to Malthus’s earlier insistence that ‘the effects of a great difficulty in procuring corn would ... be a diminution in the real wages of labour, or their price in corn; but not a diminution of profits’ (6 July; Ricardo 1951, VI, 111; cited EDR, 128). This difference of opinion was to surface again in discussion of the Essay. I mention it now because the fall in the corn wage (accepted by both correspondents) does not suggest equiproportionate increases in all money variables; it is consistent rather with a rise in the corn price relative to manufactured prices which (assuming a given mixed wage basket) will imply a lower corn wage. III. In his discussion of the Essay, Peach (1984, 739) focuses upon the fact that in the famous Table (which summarizes the discussion of the agricultural sector) the ‘capital’ on the first portion of land (estimated in quarters of wheat) remains unchanged as further capital is applied to less productive or less advantageously situated plots. This procedure (supposing, as Ricardo indeed supposed, that capital is not comprised of corn alone) would be unacceptable should there occur a relative increase in the 74

O N A ‘ N E W I N T E R P R E T A T I O N’

corn/manufactures price ratio. But now the implicit constancy of relative prices is no longer attributed to an assumption of equiproportional price increases, as in 1814, but rather to constancy of all prices – notwithstanding Ricardo’s insistence in the Essay itself that the impact of diminishing returns is ‘to raise the price of raw produce and of labour, leaving all other commodities at their original prices, and to lower general profits in consequence of the general rise of wages’ (Ricardo 1951, IV, 19–20). The assumption of constant prices in agriculture, is said to be ‘confirmed’ by a passage which represents the ‘conclusion’ to the agricultural analysis: ‘If the money price of corn, and the wages of labour, did not vary in price in the least degree, during the progress of the country in wealth and population, still profits would fall and rents would rise; because more labourers would be employed on the more distant or less fertile land, in order to obtain the same supply of raw produce; and therefore the cost of production would be increased, whilst the value of the produce continued the same’ (Ricardo IV, 18; cited 1984, 741). Peach is here largely concerned with controverting Sraffa who had asserted (with homogeneity of input and output in mind) that ‘the profit rate was calculated [by Ricardo] without need to mention price’. For Peach the absence of price considerations follows ‘because of a tacit assumption about price behaviour’ rather than the ‘technical identity between product and capital’. A ‘tacit assumption about price behaviour’ is a high-faluting way of saying that Ricardo chose not to take the relative increase in the corn price into account when devising the Table, or (as I put it in my book) that ‘there are no price calculations in the table of the Essay’ ( EDR,162). The question Peach neglects to ask is why Ricardo should proceed thus – unless, heaven forbid, Sraffa should be right – when he himself was so insistent that the corn price does rise while manufactures remain constant (Ricardo 1951, IV, 19). That it was a matter of ‘analytical convenience’ (Peach 1984, 145) scarcely answers the question. An answer is given in my EDR,154f. It is that formal allowance for an increased price of corn would not, Ricardo believed, change the direction of the tendencies summarized in the Table and could, on a first view, be neglected without distortion. A brief summary of this argument follows. It is Malthus who – much as in 1814 – set the ball rolling by raising the problem that corn price increases upon agricultural extensions will reduce the corn value of the capital stock applied to intra-marginal land in consequence of a reduced corn value of fixed capital and a reduced corn wage (the given wage basket composed of a fixed-proportions mix of food and manufactures), with a resultant increase in the corn surplus: ‘Pray think once more of the effect of a rise in the relative price of corn, upon the whole surplus derived from land already in cultivation. It appears to me, I confess, as clear as possible that it must be increased. The expences estimated in Corn will be less, owing to the power of purchasing with a less quantity of corn, the same quantity of fixed capital, and of the circulating capital of tea sugar cloaths &c: 75

RICARDO - THE NEW VIEW

for the labourers; and consequently more clear surplus will remain in the shape of rent and profits together (no matter which) ’ (12 March 1815; Malthus, in Ricardo 1951, VI, 185). Two days earlier he had written: ‘I confess I think that the kind of calculation which I mentioned to you in Town, shews in what manner profits on land may rise decidedly, from the alteration in the relative value of corn’ (10 March 1815; VI, 182). These formulations imply that Ricardo in discussing his table had not defended himself by asserting constancy of the corn price. And in fact this impression is confirmed by an explicit statement that ‘my table is applicable to all cases in which the relative price of corn rises from more labour being required to produce it’ (21 April 1815; VI, 220). 4 Ricardo’s position is not difficult to appreciate. He accepted readily ‘that if from a rise in the relative value of corn less is paid [in corn terms] for fixed capital and wages, – more of the produce must remain for the landlord and the farmer together,– this is indeed self evident’ (14 March 1815; VI, 189). But this ‘self evident’ proposition could not be automatically applied in the event that the corn price increase reflects reduced agricultural productivity, for then (notwithstanding lower per capita corn wages) total costs (measured in corn) must increase: A rise of the price of corn and a fall in the corn price of labour is in my opinion incompatible ... observe that I do not question that each individual labourer might receive a less corn price of labour ... but I question whether the whole amount of wages &c. paid for the cultivation of the land can be diminished with an increase of the exchangeable value of corn. If no more labourers were employed and the price of corn rose your proposition could not be disputed, but the cause of the rise of the price of corn is solely on account of the increased expence of production. Since corn costs rise absolutely with agricultural extensions the possibility of (relatively) increasing net returns could be excluded: ‘You will see by what I have said, that a rise in the price of corn is always in my opinion accompanied by a less material surplus produce. ... Of this produce the landlord gets so large a share that in spite of the rise of produce [in price] the situation of the farmer is constantly getting worse’ (VI, 190). Evidently Ricardo is not denying that the corn surplus rises absolutely, but is pointing to a constantly diminishing increase in the total surplus reflecting the continued upward pressure on capital costs measured in corn. 5 This is confirmed by a comment to Malthus of 21 March: ‘if you meant only that the surplus produce would increase with every accumulation of capital on the land, though in a diminishing ratio to the capital employed on the land, that is not only advanced, but strenuously maintained as the groundwork of my theory, and is the basis on which my table is formed’ (VI, 197). This latter statement is very revealing, especially when coupled with that of 21 April asserting that the table applied ‘to all cases in which the relative price of corn 76

O N A ‘ N E W I N T E R P R E T A T I O N’

rises from more labour being required to produce it’. Observation of the table itself (IV, 17) confirms that ‘total produce in quarters of wheat, after paying the cost of production’ (i.e. the surplus) rises but in diminishing increments and relative to total capital; that ‘profit per cent on the whole capital’ falls; and ‘rent per cent of the whole capital’ rises. Ricardo does not revalue, as it were, the earlier applications of capital in corn terms. But it is clear why; in Ricardo’s view, notwithstanding the impact of a relative rise in the price of corn in reducing the corn wage rate, total corn costs relatively to the total surplus (i.e. average costs) necessarily rise with expansion. It is that increase which counted and any ‘revaluations’ would not reverse the general tendencies portrayed in the table. 6 Since Peach is confessedly working in the realm of ‘conjecture’ (1984, 745), and ‘inference’ (742), let me be permitted a conjecture myself. Ricardo in 1814 (see above, p. 74) had allowed for a reduced corn wage, insisting none the less upon a fall in the corn surplus relative to corn costs. My conjecture is that he could do so for precisely the same reason as that just given, and not because he ‘tacitly assumed’ equiproportionate price increases. For in both 1814 and 1815 Ricardo insisted upon a relative increase in the corn price, the sole difference being that in 1814 he still allowed for some increase in the manufacturing price whereas in 1815 he insisted that manufacturing prices remain constant. One and the same rationale accounts for the decision to neglect the complication. IV. There remains one further matter. In a recent ‘overview’ of ‘What Ricardo Said and What Ricardo Meant’ Professor Mark Blaug asserts that I err in proceeding ‘from the denial of the corn model interpretation to deny that Ricardo ever believed in an “agricultural” theory of profits, whereas Ricardo clearly held the view, in his early writings that “the profits of the farmer” determine the general rate of profit’ (Blaug 1985, 7n). On this we are referred to ‘the definitive exegesis of Peach (1984)’. Yet a principal feature of Peach’s article happens to be the insistence (which I share) that by the ‘regulatory’ role of farmers’ profits (in the famous letter of 8 March 1814) Ricardo did not intend ‘the mathematically precise concept of “unique determination” ’, i.e. that a ‘regulatory’ role must not be identified with a ‘determining role’ (1984, 735)! Professor Blaug has misread Dr Peach and reversed the message in the process. NOTES 1. This identification is actually misleading. The ‘adding-up’ theory was thus labelled by Sraffa (1951, xxxv), who keeps it distinct from the positive impact of wages on prices. 77

RICARDO - THE NEW VIEW

2. Peach himself seems doubtful of the material calculation; he refers not to ‘the previous reference by Ricardo to a “material rate of produce”’ but to ‘any previous reference’. 3. Yet more strongly, Peach insists that ‘the [Smithian] “adding-up” approach was central to the treatment of profitability prior to the Essay’ (1984, 734); and ‘before the Essay, the argument seems to have been pure Adam Smith’ (745). This is simply wrong. Smith’s position was that all prices rise in proportion to the corn price; Ricardo’s position in 1814 insisted on less-than-proportional general price increases. 4. Cf. also 17 March 1815: ‘It can never happen ... that profits can fall, and encourage the cultivation of poor land in the manner assumed in my table, without a rise in the price of corn’ (Ricardo 1951, VI, 194). 5. I make unnecessary difficulties for myself regarding this matter in EDR, 156n. 6. As I show in EDR,154f, Ricardo was to make explicit some of the price calculations in a formal attempt to justify his position. In these extensions the problem of marginal versus average pricing arises.

REFERENCES Blaug, M. (1985) ‘What Ricardo Said and What Ricardo Meant’ in G.A. Caravale (ed.) The Legacy of Ricardo, Oxford: Basil Blackwell, 3–10. Hollander, S. (1979) The Economics of David Ricardo, Studies in Classical Economics, Vol. II, Toronto: University of Toronto Press. Peach, T. (1984) ‘David Ricardo’s Early Treatment of Profitability: A New Interpretation’, Economic Journal 94: 733–51. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol. IV: Pamphlets, 1815–1823, Vol. VI: Letters, 1810–1815,Cambridge: Cambridge University Press. Sraffa, P. (1951) ‘Introduction’ to The Works and Correspondence of David Ricardo, Vol. I: The Principles of Political Economy and Taxation, Cambridge: Cambridge University Press.

78

5 SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO: ON THREE CONTRIBUTIONS TO THE CAMBRIDGE JOURNAL OF ECONOMICS

I. In this Note I consider some circumstantial evidence offered in three contributions to the Cambridge Journal of Economics (Langer 1982; de Vivo 1985; Prendergast 1986) in support of Sraffa’s interpretation of Ricardo’s early theory of profit as a material or corn ratio model, resting on a ratio of corn input to corn output and relying on variation in the corn price of manufactured goods to bring other sectors into line (Sraffa 1951, xxxi). These contributions were all directed against my own denial of the validity of Sraffa’s rational reconstruction in favour of an interpretation involving the impact on the profit rate of money-wage variations, such that the general (including the agricultural) profit rate entails a value calculation (e.g. Hollander 1979, 162–3, 183–4, 685–6). Groenewegen (1986, 456) observes regarding the evidence by Langer and de Vivo that none of it ‘has been effectively rebutted’ by critics of the Sraffa interpretation. It is high time to provide an effective rebuttal. 1 II. Langer ascribes an explicit corn-profit model to Robert Torrens and points to Ricardo’s approval: ‘I am very much pleased with Col. Torrens essay in the last Edinb. Review’ [October 1819]; ‘I do not think there is more than one proposition in it which I should be disposed to dispute’ (letter to McCulloch, 28 February 1820; 1951, VIII, 159). Since Torrens makes no mention of the money-wage mechanism in profit-rate determination (which I ascribe to Ricardo), Langer concludes that Ricardo’s 79

RICARDO - THE NEW VIEW

praise for the paper must allude to the determination of the profit rate as a physical ratio. Now to the substance of Langer’s case. The first citation that he gives from Torrens (1819, 455–6) describes a single-sector (agricultural) case with no implications whatsoever for Sraffa’s interpretation which involves both agriculture and manufacturing. But there is also a second ‘cause’ influencing profits apart from the ‘quality of the soil’, 2 which Langer mentions but does not deal with, and this does constitute a test case. In the course of the discussion of ‘the degree of Skill and economy with which labour is employed, whether in agriculture, or in manufactures’ we find (a) that Torrens refers to expenditure by a farmer of 100 quarters of corn, or the value of 100 quarters; and (b) that Torrens assumes a mixed wage basket in dealing with the impact of manufacturing improvements. Evidently a value relationship is required to obtain the corn ratio: The second circumstance which influences the rate of profit, is the d egree of Skill and economy with which labour is employed, whether in agriculture, or in manufactures. If a farmer expend 100 quarters of corn, or the value of 100 quarters, in cultivation, and obtain a reproduction of 200 quarters, it makes not the smallest difference with respect to the rate of profit, whether this return be raised from a very fertile soil unskilfully managed, or from one of inferior quality judiciously managed. In either case there are 100 quarters expended, and 200 produced; and though the cause of the increased proportion in which the return exceeds the advance is different, the effect is the same, and the rate of profit in either case is cent. per cent. Improvements in manufactures have the same influence on the rate of profit as improvements in agriculture. If a farmer were to employ 50 labourers in cultivating fields which yielded 150 quarters of corn, and were to expend 60 quarters on the food and seed, and 60 on the clothing and implements they consumed while at work, his total surplus or profit would be 25 per cent.; but if an improvement in manufacturing industry were to take place, which so reduced the productive cost, and consequently the exchangeable value for wrought necessaries, that the farmer could purchase the clothing and implements consumed by his 50 labourers for 40, instead of 60 quarters of corn, his profit would rise from 25 to 50 per cent.; for, in this case, the reproduction of 150 quarters would be obtained by an advance of 60 quarters for food and seed, and 40 for clothing and implements. (Torrens 1819, 456) Secondly, both the quality of land and degree of skill (technology) are said to have ‘precisely the same effect in regulating the rate of manufacturing [profit] that they 80

SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO

have in regulating the rate of agricultural profit’. In the calculation he makes precisely the same sort of estimation for the industrialist as for the farmer, thereby precluding a Sraffian perspective: Now, the two causes which we have mentioned, namely, the quality of the land under cultivation, and the degree of skill with which labour is applied, have precisely the same effect in regulating the rate of manufacturing that they have in regulating the rate of agricultural profit. When a master-manufacturer, by advancing to 100 labourers their clothing and tools, with food and material equal to these in value, can fabricate clothing and tools for 300, his surplus of product above expenditure will be 50 per cent. But if, in consequence of cultivating inferior soils, or of pursuing a less skilful mode of husbandry, the productive cost, and consequently the exchangeable value of agricultural produce should be so increased, that for the food and raw material furnished to his 100 labourers he is obliged give clothing and tools for 150 instead of for 100, his profit would sink to 20 per cent.; because, in producing clothing and tools for 300, he would have expended the clothing and tools of 250. (Torrens 1819, 456–7) That there is no Sraffian technical model at play is further confirmed by the elaboration of the impact of growth on the profit rate, where we have allusion to the rising Pc/Pm ratio reflecting diminishing returns in corn production and increasing returns in manufacturing. In the Sraffa model an increase in Pc/Pm reduces the manufacturing rate in line with the falling agricultural rate; but for Torrens the decline in Pm acts to increase the general rate, though he believed it to be outweighed by the impact of increasing Pc: Thus, as population and improvement advance, manufactured articles are constantly falling in value, as compared with agricultural produce. But, on the principles already unfolded, increased facility in producing wrought necessaries, has the same effect in raising the rate of profit, which diminished facility in producing food and material has in lowering it. And hence it will frequently happen, that a greater degree of economy and skill in the application of labour may completely counteract the effects of resorting to inferior soils; and that the return upon productive capital may rise, on the whole, though the difficulty of obtaining food and material should increase. Such a process, however, could not continue long. Under any given degree of skill and economy in the application of labour, the return upon capital will be determined by the quality of the land in cultivation; and as inferior soils are resorted to, the rate of profit will constantly diminish, until that stationary 81

RICARDO - THE NEW VIEW

state is attained, in which no additional capital can be employed, and all tendency to increased population must be checked by famine. (Torrens 1819, 459; cf. Torrens 1829, 119–20)3 I conclude that Torrens’s little arithmetical examples – including that in 1819 (457– 8, cited in Langer 1982, 399), involving a joint manufacturer– farmer – are no more than extreme simplifications designed to convey specific points and without the analytical content ascribed to them by Langer. But what of Ricardo’s high praise for the article? Langer presumes that it must refer to the corn-profit model. Even were such a model present, which for the reasons given above there is reason to doubt, how can one be sure that Ricardo had it in mind? There is so much else, including, quite apart from the general impact of diminishing returns on the profit rate (e.g. 1819, 461, 464, 465), (a) the dependence of employment on capital accumulation and of accumulation on the rate of return (466); (b) the criticism of Owen’s schemes on grounds of Say’s or Mill’s Law of Markets – the review after all has the running head ‘Mr. Owen’s Plans for Relieving the National Distress’ (470–5); (c) the case for a free corn trade to counteract diminishing returns (461–3, 475–6); and (d) objections to heavy taxation (462). III. De Vivo purports to bring stronger evidence that Torrens ascribed to the logic of the Sraffian reconstruction, and argues further that Torrens formally acknowledged his indebtedness to Ricardo regarding profit-rate determination (1985, 90–1). The references are (a) to the second edition (1820) – the new Fourth Part – of Torrens’s Essay on the External Corn Trade especially the ‘general principle, that in whatever proportion the quantity of produce obtained from the soil exceeds the quantity employed in raising it, in that proportion the value of manufactured goods will exceed the values of the food and material expended in preparing them’ (1820, 362), and (b) to the derivation of the appropriate manufacturing/corn price ratio which brings the manufacturing rate into line (364–5). That Torrens allows for ‘materials’ as well as corn in advances, so that value is involved in estimating advances in corn, is referred to by de Vivo as a ‘weakness’ (1985, 91). But this surely is another way of admitting that we do not have a Sraffa model. A passage added to the third (1826) edition of the Essay, said by de Vivo (91) to confirm the Sraffa construction, is subject to the same qualification: In agriculture, some of the things produced are generally homogeneous with some of the things expended in production; and, to whatever extent this may be the case, the exchangeable value of the production will have no influence on profit. In manufactures, however, it very frequently happens, that the advances and the reproduction are altogether heterogeneous, and that no part of the former can be directly replaced by the latter. In these branches of 82

SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO

manufacture, therefore, the profit of the capitalist must depend entirely upon the proportion which the value of the whole of the reproduction bears to the value of the whole of the advance. (Torrens 1826, 105; also in 1827, 1829 editions) The problem for the corn model interpretation is that in his text Torrens assumes that the wage in all sectors, agriculture included, comprises clothing as well as food; and the 1820 allowance for ‘materials’ amongst the advances is retained. There is, moreover, de Vivo’s failure to demonstrate that ‘Torrens’s “corn ratio” theory of profits’ – subject to the foregoing qualifications – ‘did in fact derive from Ricardo’ (1985, 92). First and foremost, he fails to cite Torrens’s own claim to priority on the ‘general principle’ in question: ‘The laws, indeed, which determine the difference between the value of produce in the raw, and in the manufactured state, have not, as far as I know, been traced out by any preceding writers; and I must therefore entreat the indulgence of the reader while I endeavor to supply what, as it appears to me, has hitherto remained a desideratum in economical science’ (Torrens 1820, 360). This compromises the view that Torrens drew on Ricardo for his (‘weak’) Sraffa corn-profit model. But the same conclusion emerges if we consider the positive evidence relied on by de Vivo for a Ricardo connection (1985, 90–1). This comprises (a) Ricardo’s comment to Malthus that Torrens had adopted ‘all my views respecting profits and rent’ (23 February 1816; 1951, VII, 24); (b) his comment, also to Malthus, that Torrens is ‘quite a convert to all what you have called my peculiar opinions on profits, rent, &c. &c.’ (28 May 1816; VII, 36); and (c) the appearance of the second edition of the Essay – written, de Vivo asserts, ‘while under the influence of Ricardo’s Essay on Profits ’ (but delayed by extraneous circumstances) ‘where we find both the determination of the agricultural rate of profit in physical terms and the determination of the value of manufactured goods relative to corn using the given agricultural rate of profit’ and a prefatorial expression of indebtedness to Ricardo’s ‘original and profound inquiry into the laws by which the rate of profit is determined’ (Torrens 1820, xix). We shall deal with the foregoing arguments in sequence. De Vivo (1985, 91) himself recognizes that Ricardo’s original objections to Torrens relate to the latter’s Smithian view expressed in the first edition of the Essay on the External Corn Trade – that ‘when corn and wages rise, a universal rise in commodities will take place’ (1815, 243; de Vivo might also have cited pp. 81–93 which establish the proposition). Regarding this Smithian linkage Ricardo wrote to Malthus on 17 April 1815: ‘You, I think agree with Mr. Torrens that a rise in the price of corn will be followed by a rise in the price of home commodities ... Mr. Torrens theory however on this part of the subject appears to me defective, as I think that the price of commodities will be very slightly affected 83

RICARDO - THE NEW VIEW

by a rise or fall in the price of corn. If so every rise in the price of corn must affect profits on manufactures, and it is impossible that agricultural profits can materially deviate from them’ (Ricardo 1951, VI, 212–13). As mentioned, the letter of 23 February 1816 (to which de Vivo alludes) refers to Torrens’s adoption of ‘all my views respecting profits and rent’. But we cannot simply presume that Ricardo had a corn model in mind; and in fact, the one specific instance given by Ricardo of Torrens’s conversion relates to the constancy of manufacturing prices in the face of a rise in the wage, i.e. to Torrens’s abandonment of the Smithian position on the impact of the wage on prices: ‘Have you seen Torrens’ Letter to Lord Liverpool? – He appears to me to have adopted all my views respecting profits and rent; and in some conversations which I had with him a few days ago, he unequivocally avowed that he was now of my opinion that the price of labour, arising from a difficulty of procuring food, did not affect the prices of commodities. He confessed that his former view of that subject was erroneous’ (VII, 24).4 The Letter to Lord Liverpool itself, it is true, makes no mention of the new opinion and, in fact, brief remarks pertaining to the impact of wages on prices, suggests the old Smithian position (Torrens 1816, 31–2); but its commendation of the ‘very able and original publication on the Profits of Stock, by D. Ricardo, Esq’ (30n) relates in the most general terms to that work, disallowing de Vivo’s presumptive reading – at least without additional justification for it: ‘The influence which the high price of corn, and the bringing in of inferior land, has upon the profits of stock, is one of the most interesting and important topics connected with the science of wealth. From the invariable law of competition, that which lowers agricultural profit, must also lower profit in every other occupation. At present, however, it is not my intention to consider the effect of an artificial scale of prices upon the general industry of the country’ (30–1). 5 As for the second letter to Malthus (28 May 1816), this too tells us too little. ‘All [Ricardo’s] peculiar opinions on profits, rent, etc.’ might well have been a reference to the mechanism involving the relationship between money wages and commodity prices and the differential rent doctrine, especially when we have in mind the views expressed in conversations alluded to in the letter of 23 February. Finally, what of the statement of indebtedness by Torrens in 1820 for Ricardo’s instruction? De Vivo presumes that this refers to the Essay on Profits. In fact Torrens alludes to both the Essay and the Principles (1820, xviii–xix), and does not specify which he had in mind in making his commendation, referring only to ‘the works of Mr. Ricardo’. In any event, it cannot be taken for granted that he interpreted the Essay on Profits in terms of Sraffa’s corn-profit model. After all, it is in the Essay that Ricardo maintained: ‘The sole effect then of the progress of wealth on prices ... appears to be to raise the price of raw produce and of labour, leaving all other commodities at their original price, and to lower general profits in 84

SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO

consequence of a general rise of wages’ (1951, IV, 20). It is also pertinent that in 1820 Torrens refused to surrender priority for several of the themes common to himself, Ricardo and Malthus in 1815, none of which specify the corn-profit perspective. Moreover, a statement which insists on priority in the first edition of 1815 scarcely supports de Vivo’s claim that the revised edition was written ‘while under the influence of Ricardo’s Essay on Profits ’: [The author] now feels that he was less successful in tracing the effects of this commerce upon the distribution of wealth. The greater part of the work was composed and printed before he had read Mr. Malthus’s Inquiry into the Nature and Progress of Rent; and the whole was printed and published before he saw Mr. Ricardo’s Essay on the Influence of the Price of Corn upon the Profits of Stock. Previously, indeed, to his perusing these able and original works, the Author, in the First Edition of the Essay on the Corn Trade [1815], had stated, that the value and the rent of the better soils are in proportion to their superiority to the inferior soils which may be profitably tilled; and that an increased difficulty in raising corn diminishes the net produce or profit obtained in manufacturing industry. In a tract, also, entitled, The Economists Refuted, published several years before, the Author had pointed out the influence of low or high wages in raising or lowering the rate of profit, as well as the tendency of profits to conform to a common level, and of capital to emigrate from countries where the rate of profit is lower, to those in which it is high. (Torrens 1820, xv–xvii) In this retrospective view the closest we come to our topic is the proposition ‘that an increased difficulty in raising corn diminishes the net produce or profit obtained in manufacturing industry’. But the full analysis of this proposition (1815, 69f) entails a mixed wage basket disallowing a material profit rate. IV. Prendergast (1986, 188–9) maintains that Malthus rejected the corn-ratio theory of profits in the second edition (1836) of his Principles of Political Economy on the grounds that to estimate profits by quantity is to neglect supply and demand. On her reading, Malthus insisted (even in the case of homogeneous input and output) on a value calculation. She suggests that this rejection of the estimation of profits by quantity may have been motivated by Torrens’s objection to the formal definition of the profit rate in terms of value ratios in the 1820 Principles, on the grounds that profit is a surplus which would exist ‘quite independently of value’ originating ‘not in the interchange of commodities nor in the quality of value which wealth therby acquires,

85

RICARDO - THE NEW VIEW

but in the power of human industry to produce a greater quantity of the necessaries of life than is sufficient to support the labourers by whom it is carried on’. 6 That Malthus read Torrens I do not dispute. But one cannot presume that he read Torrens as proposing a corn output–input solution (which he himself rejected). Torrens seems to have been making a more general point about the source of profits, in fact of all non-wage incomes.7 But let us assume, for our purposes in this paper, that Prendergast is right on all the above scores. 8 It is difficult to appreciate her further assertion that Malthus’s rejection of the corn-ratio theory of profits in his revised Principles ‘seems to support Sraffa’s conjecture that “the rational foundation of the principle of the determining role of the profits of agriculture, in Ricardo’s early theory of profits, is that in agriculture the same commodity, namely corn, forms both the capital and the product”’ (1986: 189). Malthus himself made no such claim when undertaking his revisions; and it is difficult to see how a position on profit-rate determination taken by one economist in the 1820s casts light on the position taken by another economist in 1815. NOTES 1. Mention should also be made of Skourtos (1991). This paper maintains the debatable proposition that ‘The question of legitimately attributing to Ricardo’s Essay on Profits the implicit use of a corn model cannot be answered independently of the answer to the more general question: was such an analytical construct a part of the classical tradition in general?’ (215), a construct which ‘was well acknowledged and widely discussed in the classical tradition although not unanimously adopted’ (217). But Skourtos (who evidently takes for granted the validity of Sraffa’s interpretation) goes no further than to assert regarding his text that it provides ‘hints ... which do, if not establish, foster the impression of the existence of the Ricardian link’, i.e., the impression that Ricardo himself did rely on the corn-profit model in his early writings. This weak claim comes as a surprise considering the initial proposition; and in fact the paper casts no light at all on Ricardo’s position. 2. According to Torrens there were three ‘causes’ which determine the profit rate: ‘the quality of the soil under cultivation’, ‘the degree of skill with which labour is applied’, and ‘the proportion of the produce absorbed as wages’ (Torrens 1819, 458; cited Langer 1982, 398). 3. See also Torrens 1819, 468–9: If a piece of ground will produce 100 quarters of corn, and if the labourers employed upon it expend 50 quarters for seed and food, with clothing and implements which cost 50 quarters more, then it is evident that such land will not be cultivated; and for the plain reason, that its cultivation will afford the farmer no profit. But if improved machinery were to lower the price of manufactured goods, until the farmer could purchase for 30 quarters the same quantity of necessary clothing and implements which formerly cost him 50, then this land would be eagerly sought, for the purpose of tillage; because in this case the diminished expenditure of 30 instead of 50 quarters for the purchase of clothing and implements, would yield the farmer a profit of 25 per cent.

4. The change of mind to which Ricardo referred left no mark on the first three parts of the Essay including pages 81–93 and 243 expressing the Smithian position. But the new Part 86

SRAFFA’S RATIONAL RECONSTRUCTION OF RICARDO

5. 6. 7.

8.

Four does reflect the change: ‘though the price of corn were the only circumstance which regulated the price of materials and of labour, it would by no means follow, as [Smith] asserts, that it thereby regulated the price of all home-made goods’ (Torrens 1820, 351–2). As for the analysis of ‘the effects which the high price of the produce of the soil’ has on the agricultural profit rate, that runs in terms of quarters of corn or ‘what is the same thing’ their value (Torrens 1816, 28–30). R. Torrens, review in The Traveller, No. 6624, 26 April 1820. In a continuation of the review of the Principles for the Traveller of 1 May 1820, Torrens extended to rent his denial that value is essential, making out a case that rent could exist even were institutions not of the exchange variety: ‘Rent, like profit has its origin in the power of human industry to produce a greater quantity of wealth than is necessary to support the labour by which it is carried on; and may appear though there should be neither markets nor market prices, neither exchange nor exchangeable value.’ It would exist ‘in a state of society, in which there was no division of employment, and in which each capitalist engaged his labourers in the immediate production of the several articles he consumed’ (see Robbins 1958, 282–3). I would dispute Prendergast’s interpretation of Malthus. The evidence suggests that Malthus did maintain the corn-ratio theory (Hollander 1994).

REFERENCES de Vivo, G. (1985) ‘Robert Torrens and Ricardo’s “Corn-Ratio” Theory of Profits’, Cambridge Journal of Economics 9: 89–92. Groenewegen, P. (1986) ‘Professor Porta on the Significance of Understanding Sraffa’s Standard Commodity and the Marxian Theory of Surplus’, History of Political Economy 18: 455–62. Hollander, S. (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. ——[1994] forthcoming, ‘Malthus and the Corn Profit Model’, in Heinz Kurz (ed.) Piero Sraffa’s Contributions. Langer, G.F. (1982) ‘Further Evidence for Sraffa’s Interpretation of Ricardo’, Cambridge Journal of Economics 6: 397–400. Prendergast, R. (1986) ‘Malthus’s Discussion of the Corn Ratio Theory of Profits’, Cambridge Journal of Economics 10: 187–9. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol. I: The Principles of Political Economy and Taxation, Vol. IV: Pamphlets, 1815–1823, Vols VI–VIII: Letters,Cambridge: Cambridge University Press. Robbins, L.C. (1958) Robert Torrens and the Evolution of Classical Economics, London: Macmillan. Skourtos, M. (1991) ‘Corn Models in the Classical Tradition: P. Sraffa Considered Historically’, Cambridge Journal of Economics 15: 215–28. Sraffa, P. (1951) ‘Introduction’ to The Works and Correspondence of David Ricardo, Vol. I, Cambridge: Cambridge University Press. Torrens, R. (1815) An Essay on the External Corn Trade, London: Hatchard. A Letter to the Rt. Hon. the Earl of Liverpool on the State of the Agriculture of the United ——(1816) Kingdom, London: Hatchard.

87

RICARDO - THE NEW VIEW

[Torrens, R.] (1819) ‘Mr Owen’s Plans for Relieving the National Distress’, Edinburgh Review, October. Torrens, R. (1820) An Essay on the Influence of the External Corn Trade upon the Production and Distribution of National Wealth, 2nd edn, London: Hatchard. ——(1826)An Essay on the External Corn Trade, 3rd edn, London: Longman, Rees, Orme and Brown. ——(1827)An Essay on the External Corn Trade, 4th edn, London: Longman, Rees, Orme and Brown. ——(1829)An Essay on the External Corn Trade. A New Edition, London: Longman, Rees, Orme and Brown .

88

Part II

RESPONSES TO CRITICS OF THE ECONOMICS OF DAVID RICARDO

6

THE ECONOMICS OF DAVID RICARDO: A RESPONSE TO PROFESSOR O’BRIEN

I. INTRODUCTION Professor O’Brien has paid me the high compliment of subjecting my Economics of David Ricardo (1979) to a very close examination. I am grateful to him despite the fact that he takes strong exception to what he calls my ‘frontal assault upon the accumulated body of Ricardo scholarship’ (1981, 385). I have given his objections the serious consideration which they merit, and find, in my turn, his defence of the standard textbook interpretation of Ricardo’s economics to be unconvincing. I do not see that he adds to the documentation, primary and secondary, that was available and taken into account when I composed my volume or to the appropriate interpretation of that documentation. On the contrary, what he has to say convinces me that the consensus view does far less justice to Ricardo than the historical record dictates. Before turning to substantive matters I wish to correct a misapprehension. O’Brien suggests that my rejection of the ‘standard’ or ‘majority’ or ‘unanimous’ interpretation of Ricardo – I use his terms (365, 377, 385) – stems from ‘disagreement with Cambridge’ (385). This is, in fact, not so: it is my belief that several of the Cambridge criticisms of neo-classical equilibrium economics – ‘substitutionist’ economics as I have heard it called by Sir John Hicks - have merit. But the question of the desirable content of modern economics must be kept apart from the historical record and it is solely the Cambridge reading of the history of nineteenth-century economics to which I object in my book. I might add that this is not the position with which I commenced my researches (see Hollander 1973, 11–17); it was the outcome, not the starting point, of my analysis. I refer now, of course, to the illegitimacy of that interpretation which attributes to Ricardo an approach involving the treatment of distribution, ‘natural’ prices and production levels 91

RICARDO - THE NEW VIEW

by means of separate models – purportedly reflecting the isolation of ‘one-way direction’ relationships or sequential analysis or the ‘causal ordering’ of variables (Dobb 1973; Pasinetti 1974; Garegnani 1976; Roncaglia 1978). It is not only Cambridge that denies Ricardian economics the characteristic ‘neoclassical’ interdependency of distribution and pricing. This same notion pervades the orthodox textbooks, including O’Brien’s own The Classical Economists (1975). In his review O’Brien subscribes explicitly to ‘one-way models’ as distinct from ‘two (or n ) way causation’ (1981, 359) in Ricardo interpretation – and it is apparent that the only economics that has merit for O’Brien is that entailing one-way causality which, he believes, allows clear-cut prediction subject to unambiguous empirical testing; mutual causation is apparently anathema to him. This is clear from the following summary passage which also states his main objections to my book: Hollander feels constrained to throw out all the work that has been done to make sense of Ricardo, and instead to reinterpret Ricardo in a kind of loose general equilibrium framework. Now one of the key features of general equilibrium – which renders it virtually useless – is that it results in a model of complete generality in which virtually anything can happen, and which is almost completely devoid of predictions that can be subject to empirical testing. And this is more or less the end-product of Hollander’s reinterpretation of Ricardo. The Hollander version goes something like this. ‘Wages will no longer be necessarily at subsistence. They may or they may not be. Rental share may or may not increase. It is not clear that Ricardo thought about this. Profits can decline it is true. But this is not an immediate problem.’ Almost all the predictive element (and all the rigour) of Ricardo’s own thinking have gone. The Ricardian Vice therefore never existed. Ricardo did not really even build models. (O’Brien 1981, 385) To appreciate fully the issues at stake between us this summary statement should be read bearing in mind the version of Ricardo outlined in O’Brien’s The Classical Economists. ‘Ricardo’s system’, we are there instructed, ‘was, if not entirely the first, certainly the first sweepingly successful example of economic model building. Its essence was the “corn model” of aggregate economic relationships’ (1975, 37). Then follows the famous Kaldor diagram with its emphasis upon the fix-wage or subsistencewage secular path from which can be seen, runs the text (40), ‘how it was that Ricardo was able to draw from his model precise predictions (none of which was in fact borne out by events) about the distributive shares’, and the downward trend of the profit rate to some minimum level: 92

A RESPONSE TO PROFESSOR O’BRIEN

Now this is the core of Ricardo’s system and it is important to be clear about this. The corn model is quite clearly the fundamental concern of Ricardo’s principles. It is a model designed to prove one central proposition: that the existence of the Corn Laws, which hindered the import of corn, caused resort to inferior land at home, involving a fall in the average and marginal products of labour and capital and hence bringing about a stationary state in the not very distant future. Smith had envisaged a stationary state, but it was nothing like imminent and its achievement was not mechanical. Ricardo on the other hand had provided a mechanical model to establish its fairly immediate likelihood. (O’Brien 1975, 41; see also O’Brien 1970, 296) And on method let us also have before us the following elaboration from O’Brien’s text: But if the Corn Model does constitute the core of Ricardo’s system there is another characteristic of his work which is of great importance and which has to be taken account of in estimating his contribution to Classical economics. For Ricardo’s deductive method, his model-building, was to be of the greatest importance. Ricardo essentially invented these techniques. His procedure not only contrasts strongly with Adam Smith’s basically inductive approach, but, as a process of heroic abstraction, it not only neglects the frictions in the economic system but also habitually reasons in terms of the immediate relevance of the long run. (O’Brien 1975, 42) I repeat: this is the version of Ricardo with which I take issue in my book; O’Brien’s review constitutes a restatement of that version. Reference is made to ‘Ricardo’s remarkable tendency to carve up the economy into one giant farm and one giant firm for the purposes of the corn model’ (1981, 365) and ‘without a corn model’, it is asserted, ‘Ricardo’s inverse relationship between wages and profits is much more difficult to establish’ (356). There is the same insistence upon a subsistence or fixed commodity wage (365–6, 368). We have repeated allusions to the notion that Ricardo’s model was designed for specific prediction and the observation that ‘once Ricardo stepped outside his model his predictions did not always follow’ (356). And the sharpest distinctions are again drawn between Smithian and Ricardian method: ‘Smith used short chains of reasoning in the Scottish tradition, while Ricardo essentially used “as if” methodology but with the status of verification somewhat unclear. Smith had continuous resort to factual material while Ricardo’s 1815 Essay and 1817 Principles contain no facts at all’ (358). In the same spirit, emphasis is placed on the ‘mechanical nature’ of Ricardian economics (358) and on ‘Ricardo’s habit of telescoping the long run and the short run’ (373). 93

RICARDO - THE NEW VIEW

Since I evidently failed in my book to divert my critic from this erroneous interpretation, I am not very hopeful that I shall succeed now. But I am duty-bound to make another attempt to seduce him a little closer to my position. I cannot hope to respond to every one of the reviewer’s criticisms but will attempt to cover the primary issues raised. II. RICARDO AS MODEL BUILDER As we have seen, O’Brien – apparently because I reject the corn model interpretation with its subsistence wage and its sequential approach to distribution and pricing – attributes to me the view that ‘Ricardo did not really even build models’ (385). This is no rhetorical flourish to close his article. This theme is stated at the outset: ‘the algebraic models of Ricardo are ... comprehensively rejected’ (352); it is an attribution to me formally listed amongst thirty-two Hollander propositions (I would have preferred thirty-nine): ‘Ricardo made free use of special assumptions which should not be taken to represent a model’ and ‘the use of models in interpreting Ricardo is not helpful’ (354); twice I read of my ‘throwing away the explicit Ricardian model’ (355; cf. 358); I am told that ‘I am not looking for a consistent model’ (357) and also that, on my interpretation ‘there is no specific model’ (359). I must be a very poor expositor indeed if this is the impression I have left. Let me therefore reiterate to avoid any chance of further misunderstanding on this central issue: I do ascribe model-building to Ricardo, but not the simple-minded model-building of the textbooks. Thus I attribute to Ricardo the growth model described geometrically in a joint article with Sir John Hicks (Hicks and Hollander 1977) and elaborated verbally in my book ( EDR,308–10, 395–404), to the algebraic rendition by Carlo Casarosa (1978) and to the version by David Levy (1976). I have also taken an entire chapter to outline Ricardo’s model of allocation theory in a static framework. The formal combination of the static and the growth models involves technical problems which Ricardo, of course, did not fully address. But to the extent that the allocation mechanisms can be grafted on to a growing system they will provide the key to an appreciation of the transition between ‘equilibrium’ positions on Ricardo’s secular path of wages, since the inverse wage– profit relation is at play along the road to stationariness and this relation ( EDR,xi, 302f) entails the mechanisms of allocative adjustment. In this context the cracking of the ice over which Professor O’Brien is skating can be distinctly heard. I have in mind his statement already cited that ‘without a corn model, Ricardo’s inverse relationship between wages and profits is much more difficult to establish’. More difficult – true, but can this possibly be a valid reason to deny the existence of such a relationship in theoretical contexts other than the corn model as O’Brien seems to do? If so we have a new and extraordinary principle 94

A RESPONSE TO PROFESSOR O’BRIEN

whereby historical evidence can be set aside for the sake of theoretical simplicity. In any event, the modern readings alluded to above have demonstrated the technical legitimacy of Ricardo’s insistence upon the inverse wage-profit relation in the complex cases – including that of contemporaneous decline in both the real wage and the profit rates. I shall have more to say about this matter in the next section. It is also appropriate to correct, at this juncture, a theoretical error regarding the wage–profit relation – the inverse relation between the money wage and the profit rates – the appearance of which surprises me: ‘It is not clear to me that Ricardo would have been very happy about reliance on money wages’, writes O’Brien (1981, 362), referring to Ricardo’s statement that ‘profits ... depend on wages; not on nominal, but real wages; not on the number of pounds that may be annually paid to the labourer, but on the number of days’ work necessary to obtain those pounds’. It really should not be necessary at this late stage to have to explain that when one talks of ‘money’ wages in the context of the theorem on distribution one has in mind money of constant general purchasing power – to which end Ricardo devised his ‘gold’, a unit designed to reflect labour embodiment in the wage basket (and, of course, proportionate wages). III. SOME PROBLEMS OF INTERPRETATION In this section I carry the matter of Ricardian model-building a step further, addressing myself to the central complaint in the review, namely that ‘by presenting Ricardo’s qualifications as the main argument and the main argument as the qualifications, Hollander manages to achieve a picture which is a complete reversal of the normal – and, it seems to me, correct – view of Ricardo’ (O’Brien 1981, 371). Similarly ‘I feel that Hollander has read too much into Ricardo’s qualifications and hints in correspondence’ (384). This theme is reiterated repeatedly (e.g. 355, 357, 373, 374, 378, 381) and deserves careful consideration. The isolation of the ‘main argument’ in a writer’s work can be a tricky problem. My complaint against O’Brien is that he does not adequately appreciate the difficulties involved; he takes the easy way out by representing the most uncompromising of Ricardo’s strong cases as the main picture. On his reading the details and fine points of the analysis appear as unabsorbed and often unabsorbable afterthoughts frequently forced on an unwilling Ricardo – ‘qualifications’ for which there is neither rhyme nor reason. As O’Brien himself puts it, ‘there was a sort of guerrilla warfare in which every time Malthus succeeded in drawing Ricardo from the stronghold of his model he inflicted casualties’ (356). Yet quite inconsistently the review opens with the resounding theme that ‘a picture of some internal consistency, which showed Ricardo to be a very powerful theorist, had eventually emerged’ (352). Some ‘powerful 95

RICARDO - THE NEW VIEW

theorist’! In my judgment O’Brien’s perspective casts Ricardo in a poor light which does him less than justice. Let me illustrate from monetary theory before turning to the growth model itself. Ricardo’s remarkably forceful statements of the case that can be made for devaluation of the currency in order to avoid the deflationary pressures on output and employment typically associated with resumption of gold payments ( EDR,488f) are simply dismissed by O’Brien with the cryptic comment that the case is ‘greatly overplayed’ in my work (1981, 371). Or consider my citation of the following clear statement of temporary excess demand for money reflecting an initial reduction in the money supply: [A] reduction in the amount of the circulating medium should speedily operate on prices, but the resistance which is offered – the unwillingness that every man feels to sell his goods at a reduced price, induces him to borrow at a high interest and to have recourse to other shifts to postpone the necessity of selling. The effect is however certain at last, but the duration of the resistance depends on the degree of information, or the strength of the prejudices of those who offer it, and therefore, it cannot be the subject of anything like accurate calculation. (EDR,512–13). In this context I also cite a lengthy passage from the Principles which constitutes an embryonic statement of J.S. Mill’s famous essay ‘Of the Influence of Consumption on Production’. O’Brien merely dismisses my belief that Ricardo allowed for temporary excess demand for money with the comment that it is ‘based upon two quotations’ – he overlooks the case for devaluation – ‘which only [sic] show that, given inelastic expectations, prices can be sticky downwards’ (1981, 372). Similarly, when I demonstrate Ricardo’s recognition that the import of corn may be paid for in gold rather than commodities, my reviewer responds ‘If Ricardo had really recognized such a possibility he could no longer have remained the strict Bullionist he was’ (373). I fear that O’Brien has closed his mind to the import of some of Ricardo’s most interesting formulations. He has decided that Ricardian economics involves the ‘telescoping’ of the long and short run and an unconcern with ‘transitional effects’ (373–4) and henceforth – this is conspicuous not only in the discussion of money but also in that of the Corn Laws (377) and the Poor Laws (374) – is obliged to set at nothing much of the evidence pointing to a different conclusion by either denying it outright or representing it as involving insignificant ‘exceptions’ to the rule. My next illustration of the errors, as I see them, into which Professor O’Brien is drawn comes from the theory of growth – particularly his subscription to the fixwage interpretation of the corn model. Ricardo’s growth model is expounded in the Essay on Profits on the basis of a constant real wage and constant technology ( EDR, 96

A RESPONSE TO PROFESSOR O’BRIEN

136f). This I do not question. But Ricardo himself explains precisely why he proceeds thus: ‘We will, however, suppose ... that no improvements take place in agriculture, and that capital and population advance in proper proportion, so that the real wages of labour continue uniformly the same; – that we may know what peculiar effects are to be ascribed to the growth of capital, the increase of population, and the extension of cultivation, to the more remote, and less fertile lands’ (1951, IV, 12; EDR, 136). Quite obviously he intended nothing more than two simplifying assumptions for the sake of clear exposition of the principle at hand. There is no reason to believe that anything more was at stake when he devised the famous numerical illustrations of the fifth and sixth chapters of the Principles involving the constant wage assumption. The method involved was elaborated quite generally in correspondence with Malthus in the following well-known terms: ‘After the frequent debates between us, you will not be surprised at my saying that I am not convinced by your arguments on those subjects on which we have long differed. Our differences may in some respects, I think, be ascribed to your considering my book as more practical than I intended it to be. My object was to elucidate principles, and to do this I imagined strong cases that I might shew the operation of those principles’ (VIII, 184). Ricardo was evidently conscious of the scope and method of his own work and recognized the imminent danger of misunderstanding by his readers. With respect, O’Brien would have done well to heed Ricardo’s warning not to confuse the strong-case simplifying assumptions with the entire analysis. He would then not have committed himself to the statement that ‘Ricardo believed in simple models more unreservedly than almost any other economist in the nineteenth century’ (1981, 376). I do not deny for one moment that when one follows my path and attempts to take seriously what an author writes in a variety of contexts and occasions without prejudgment one is often faced with the need to resolve apparent conundrums: Ricardo’s analysis of the postwar depression despite his recognition of the phenomenon of excess commodity supply is one such and requires attention ( EDR,514f). But this is the nature of any serious exercise in intellectual history. If an author is worth treating the attempt must be made to see his work as a whole on the presumption that – unless proven otherwise – his ideas are consistent. There are difficulties with my procedure but I submit that it carries us further and deeper than does the practice of ignoring or treating as unabsorbed ‘qualification’ evidence which does not fit well into the preconceived model. It so happens that the technical and methodological details essential for an appreciation of the full picture can often be gleaned from a careful reading of the Principles itself – this is indeed so in the case of the assumption regarding wages in the growth context. Sometimes, however, one is obliged to go further afield. For we 97

RICARDO - THE NEW VIEW

must not forget the relatively narrow scope of the main text: ‘it will be easier for me’, Ricardo explained to James Mill late in 1816, ‘to publish only those parts of the science which have particularly engaged my attention’ – this in reply to Mill’s advice that he decide ‘whether ... to include in it a view of the whole science’ or only ‘those parts of the science which you yourself have improved’ (Ricardo 1951, VII, 112; EDR, 5). Similarly (also to Mill): ‘I hope I shall be able to convince you of the general correctness of my principles. I have dwelt very little on the effect of those taxes on which there can be no difference of opinion, and have not mentioned many which have been ably handled by Adam Smith’ (VII, 88). The Principles was not designed as an all-inclusive textbook but as a technical monograph with a very particular objective – to correct an illustrious predecessor on specific points – from which kind of work the author’s full position cannot possibly be appreciated. When, as in our case, we have at hand a rich correspondence wherein Ricardo more fully expounds his position it is surely foolhardy to minimize the opportunity to fill in the gaps. And this source is no less significant when ‘concessions’ are drawn from our author as part of the learning process generated by the cut and thrust of debate. For we then have indications of the extent various relationships can be absorbed by the model even if they had not been part of the author’s original intellectual baggage. (The correspondence with Malthus following the Essay on Profits is supremely worthwhile in this respect; cf. EDR, 146f.) My case thus far in effect calls upon what Stigler has labelled ‘personal exegesis’ – the biographical information we have regarding Ricardo’s style and purpose. But the same conclusion follows from Stigler’s criterion of ‘scientific exegesis’ referred to by O’Brien (1981, 355), namely that ‘we increase our confidence in the interpretation of an author by increasing the number of his main theoretical conclusions which we can deduce from (our interpretation of) his analytical system’ (1965, 448). The fact is that many of O’Brien’s so-called ‘qualifications’ are far more easily absorbed within my version of the growth model than his. Most important, on his reading the secular path of real wages described by Ricardo (1951, I, 101–2) – an initial stage of rising real wages and a subsequent stage of declining real wages, each accompanied by a falling profit rate – simply cannot be accommodated. The corn model is at a loss to deal with a central theme of Ricardian growth theory – against which fact I set O’Brien’s assertion, already noted, that ‘without a corn model, Ricardo’s inverse relationship between wages and profits is much more difficult to establish’. IV. ON PREDICTION From my denial that Ricardo designed his economics for purposes of specific historical prediction O’Brien draws further support for his assertion that I refuse to attribute to Ricardo model-building (1981, 376). This is a non sequitur reflecting 98

A RESPONSE TO PROFESSOR O’BRIEN

O’Brien’s view of appropriate methodology. G.L.S. Shackle has urged that a theory ought to be ‘a classificatory one, putting situations in this box or that according to what can happen as a sequel to it. Theories which tell us what will happen are claiming too much’ (1972, 72– 3). My investigation suggests to me that this was Ricardo’s position ( EDR,ch. 11). His repeated insistence that changes in the costs and hence prices of goods which enter the wage basket can affect profits, whilst changes in the costs and prices of those not so included cannot, is precisely of this order. And he implied in an early reaction to Malthus that he did not charge his growth model with the task of generating specific historical prediction, by retaining faith in its validity despite an observed upward trend in the profit rate – and this because of changes in the ceteris paribus condition relating to new technology: ‘I have little doubt ... that for a long period, during the interval you mention [1793–1813], there has been an increased rate of profits, but it has been accompanied with such decided improvements of agriculture both here and abroad ... that it is perfectly reconcileable with my theory. My conclusion is that there has been a rapid increase of Capital which has been prevented from shewing itself in a low rate of interest by new facilities in the production of food’ (Ricardo 1951, VI, 94). I can see no reason to believe that Ricardo’s reaction would have been any different had the period involved extended beyond twenty years. As I also carefully explain in my book ( EDR, 600–1) we must distinguish the (secular or historical) prediction that in a protected economy the rate of profit will necessarily undergo a decline, from the (comparative statics) proposition that the Corn Laws rendered the rate of profit lower at any and every point of time than it would otherwise be. Ricardo certainly engaged in this kind of exercise. Similarly he used his model in this manner to deal with the distributional effects of agricultural innovation and public finance ( EDR, 642). O’Brien would surely admit that there are uses to which models can be put apart from specific historical prediction. It is also regrettable that O’Brien does not explain his charge (1981, 379) that I have taken out of context Ricardo’s letter to McCulloch following his Parliamentary speech of 7 March 1821 which I bring in support of my interpretation ( EDR,615– 21). It would also have been helpful had he given us his analysis of the detailed exchange wherein McCulloch makes a case for a continually declining rate of return and consequential loss of capital abroad and protests to Ricardo for taking a much more optimistic line. O’Brien doubtless would designate Ricardo’s response to Malthus, cited above, as a face-saving concession. But much more is involved. This was standard ‘classical’ method. Thus J.S. Mill insisted upon the distinction between ‘verification’ and ‘prediction’: a model cannot be expected to make accurate historical predictions and may be retained provided it is ‘verified’ – allowance made for changes in the ceteris

99

RICARDO - THE NEW VIEW

paribus conditions. More pertinent for us here, however, is the fact that Ricardo was writing in the Smithian tradition. For it is Smith who, when faced by a fall in the price of corn since the corn export bounty was established, insisted that ‘the price of corn had not been lowered in consequence of the bounty, but in spite of it’ (1937, 474) and refused to question the validity of his model, seeking rather for the disturbing cause at work. (I shall return to this matter presently.) V. THE SMITHIAN LEGACY AND THE RICARDIAN REACTION It is O’Brien’s position that Ricardo’s historical pessimism provides the raison d’être of his theorizing: if my interpretation of Ricardo as optimist were correct, he states, ‘Ricardo’s theoretical work ... would have been pointless’ (1981, 375–6). This matter has been in part dealt with in the foregoing paragraphs. But I wish to carry it further. By taking this position I suggest that O’Brien neglects an entire body of historical evidence relating to the late eighteenth and early nineteenth centuries which casts a bright light on the nature of Ricardo’s objectives in the Principles. There is evidence pointing to the preoccupation of economists in the postSmithian period with the relationship Smith envisaged between corn prices, money wages and general prices, particularly the notion that a rise in corn prices – due (say) to a subsidy on the export of corn – will be passed on in the form of higher manufacturing prices by way of upward pressure on money wages: [The] money price of corn regulates that of all other home-made commodities. It regulates the money price of labour, which must always be such as to enable the labourer to purchase a quantity of corn sufficient to maintain him and his family either in the liberal, moderate, or scanty manner in which the advancing, stationary or declining circumstances of the society oblige his employers to maintain him. It regulates the money price of all the other parts of the rude produce of land, which, in every period of improvement, must bear a certain proportion to that of corn, though this proportion is different in different periods. It regulates, for example, the money price of grass and hay, of butcher’s meat, of horses, and the maintenance of horses, of land carriage consequently, or the greater part of the inland commerce of the country. By regulating the money price of all the other parts of the rude produce of land, it regulates that of the materials of almost all manufactures. By regulating the money price of labour, it regulates that of manufacturing art and industry. And by regulating both, it regulates that of the complete manufacture. The money price of labour, and of every thing that is the 100

A RESPONSE TO PROFESSOR O’BRIEN

produce either of land or labour, must necessarily either rise or fall in proportion to the money price of corn. (Smith 1937, 476–7; EDR,21) This complex set of relationships, which precludes any change in the relative price and profitability of agricultural produce by acting on the price of corn is encapsulated in Smith’s famous statement that ‘the nature of things has stamped upon corn a real value which cannot be altered merely by altering its money price. No bounty upon exportation, no monopoly of the home market, can raise that value. The freest competition cannot lower it’ (1937, 482; EDR,31–2n). Similarly ‘the real effect of the [corn export] bounty is not so much to raise the real value of corn, as to degrade the real value of silver; or to make an equal quantity of it exchange for a smaller quantity, not only of corn, but of all other home-made commodities’ (1937, 476). Now Ricardo had originally adopted the Smithian position. He maintained, for example, that a change in the profit rate can affect the general level of prices – Smith’s ‘adding-up’ cost approach implied that whichever of the cost elements change generated a corresponding change in general prices; and he conceded to Bosanquet that rising corn prices due to the poor harvest of 1800 and 1801 ‘must have produced some rise in the price of commodities’ ( EDR, 106, 109, 112). By mid-1813 he had become aware that the Smithian principle of ‘competition of capitals’, whereby the discovery of new markets would assure a rise in profits and the value of commodities in general, clashed with the requirement for an expanded money supply that would not be forthcoming ( EDR, 114–15, 118); and by late 1814 Smith’s positive relation between corn prices and general prices via money wages came to be questioned in the light of ‘considerations of money value’ ( EDR,131). The Essay on Profits states the new view succinctly: It has been thought that the price of corn regulates the price of other things. This appears to me to be a mistake. If the price of corn is affected by the rise or fall in the value of the precious metals themselves, then indeed will the price of commodities be also affected, but they vary, because the value of money varies, not because the value of corn is altered. Commodities, I think, cannot materially rise or fall, whilst money and commodities continue in the same proportion, or rather whilst the cost of production of both estimated in corn continues the same. (Ricardo 1951, IV, 21n; EDR, 141) From this point on Ricardo was prepared to base his theorem on distribution – that a rise in money wages is accompanied by a fall in the profit rate rather than a general price rise – alternatively upon a commodity (real-cost) theory or a quantity theory of the value of money. 101

RICARDO - THE NEW VIEW

The extent of Ricardo’s departure may be gauged from a letter to James Mill prior to publication of the Principles : ‘In reading Adam Smith, again, I find many opinions to question, all I believe founded on his original error respecting value. He is particularly faulty in the chapter on bounties’ (Ricardo 1951, VII, 100; EDR, 5). The ‘original error’ in question is explicitly referred to in correspondence with McCulloch: ‘Your system proceeds upon the supposition that the price of corn regulates the price of all other things, and that where corn rises or falls, commodities also rise or fall, – but this I hold to be an erroneous system although you have great authorities in your favour, no less than Adam Smith, Mr. Malthus and Mr. Say’ (VII, 105). Or equally we may refer to the Principles : ‘Dr. Smith’s error throughout his whole work lies in supposing that the value of corn is constant; that though the value of all other things may, the value of corn never can be raised’ (I, 374; EDR, 204). Ricardo’s decision to attempt to settle the problem of the effect of a change in wages upon prices in the very first chapter as prelude to the establishment of the inverse relation is easily understood in this context. What is Professor O’Brien’s reaction to all this? First he minimizes the significance of the Smithian cost-push model of inflation to the point almost of actually denying its presence in the Wealth of Nations : ‘I found Hollander’s account of Smith’s own work rather odd. l cannot recognize the account of the key features of the Smithian system given by Hollander. ... The emphasis upon the corn price and the general price level in Hollander’s account of Smith makes it possible to represent Ricardo as continuing the work of Smith but correcting the mistaken analysis. This allows the rest of Smith’s work to be smuggled in largely by implication. It is however upon this last that Hollander bases his “Walrasian” claims for Ricardo’s analysis’ (O’Brien 1981, 359). I am a little taken aback by the fact that O’Brien apparently finds me so disingenuous. I must emphasize that I do not ‘smuggle’ Ricardo’s Smithian heritage in at all. I carry it quite unselfconsciously through the Nothing to Declare Exit: Ricardo’s subscription to Smith’s allocation mechanisms is an explicit part of my argument in Chapter 6 upon which I base my ‘Walrasian claims’. But quite enough has been said already about Ricardo’s own designation of the Principles as a work presuming continuity. What I find truly difficult to comprehend is how the Smithian model relating general prices to the price of corn can be swept aside so casually in O’Brien’s fashion. I am forced to conclude that he cannot escape from the belief that Smith used ‘short chains of reasoning’. This leads me to the second prong of O’Brien’s approach. My account of the complaints by Smith’s contemporaries that he engaged in excessive abstraction and was prone to leap prematurely to policy conclusions on the basis of such abstraction is dismissed as involving ‘a ragbag of quotations from around 1800 concerning contemporary views of Smith’s methodology which seem to prove nothing’ (1981, 358). 102

A RESPONSE TO PROFESSOR O’BRIEN

Since O’Brien (quite rightly) has much respect for James Anderson1 let me ask him how he feels able to throw into the ‘ragbag’ Anderson’s criticism of what he understood to be Smithian procedure: Soon after Dr. Adam Smith’s excellent work on the wealth of nations was published, it fell into my hands. I read it with the attention that it deserved, and with no small degree of satisfaction and improvement in many respects; but the satisfaction was not without alloy. While I admired the liberality of his general turn of thinking, the ingenuity of his arguments, the perspicuity of arrangement, and the acute accuracy of discrimination, that are eminently conspicuous in that work, I could not help lamenting that in it, as in all others that I have seen on practical subjects, which have been written by speculative men, the conclusions were, on too many occasions, deduced from theoretical principles, rather than from a patient induction of actual facts. It is much to be lamented, that, when the mind is thus occupied with preconceived notions, many important facts are suffered to escape notice; and those only are deemed worthy of attention which serve to corroborate the favourite hypothesis. When a mind is under the influence of this disposition, acuteness of talents only tends to lead it the farther astray, by suggesting plausible combinations, that serve only to prevent the objects from appearing in their true light to the inquirer himself; and, of course, he will find no difficulty in representing them under false colours to others. Few writers, of eminence, in modern times have been more frequently under the influence of this fascinating power than the author of the Wealth of Nations. (Anderson 1801, 16-17; EDR,39–40). It is Anderson, be it also noted, who made reference to Smith’s quest for disturbing causes long before I did: ‘when Dr. Smith found himself pressed by the obvious argument, that the price of corn had declined ever since the bounty was [established], while that of other commodities had been almost universally advancing, he attributes this circumstance solely to the general prosperity of the nation, which resulted from the revolution, and he peevishly asserts, “that the price of corn had not been lowered in consequence of the bounty, but in spite of it”. As if the same argument would not equally apply to the price of all other commodities.’ How too does O’Brien feel able to throw into his ‘ragbag’ Malthus’s criticisms of Smith’s abstract modelling and hasty application. I refer to the Essay on Population where Malthus wrote with reference to the corn export bounty in these terms: The most plausible argument that Dr. Smith advances against the corn laws, is, that, as the money price of corn regulates that of all other home-made commodities, the advantage to the proprietor from the increased money 103

RICARDO - THE NEW VIEW

price is merely apparent, and not real; since what he gains in his sales he loses in his purchases. This position, however, is not true, without many limitations. The money price of corn, in a particular country, is undoubtedly by far the most powerful ingredient in regulating the price of labour, and of all other commodities; but it is not the sole ingredient. Many parts of the raw produce of land, though affected by the price of corn, do not, by any means, rise and fall exactly in proportion to this price. When great improvements in manufacturing machinery have taken place in any country, the part of the expense arising from the wages of labour will bear a comparatively small proportion to the whole value of the wrought commodity, and consequently, the price of it, though affected by the price of corn, will not be affected proportionally. (Malthus 1803, 458–9; EDR,46) Similarly, in his Observations on the Effects of the Corn Laws (1814) Malthus turned again to Smith’s analysis of the corn export bounty and in particular the special treatment accorded corn pricing: ‘I have always thought, and still think, that this peculiar argument of Dr. Smith, is fundamentally erroneous’ (1970 [1814], 96–7; EDR,48). Once more he insisted that the money-wage rate does not rise proportionately with the corn price, and this because of the existence in the wage basket of home and imported goods other than corn (98–9). Moreover, the reaction of money wages to a rise in the corn price is lagged, not immediate: ‘corn and labour rarely keep an even pace together; but must often be separated at a sufficient distance and for a sufficient time, to change the direction of capital’. In brief, ‘the real price of corn is capable of varying for periods of sufficient length to give a decided stimulus or discouragement to agriculture’ (106); Smith’s extreme theoretical case had led him to violate his own ‘great principles of supply and demand’ and contradict ‘the general spirit and scope of the reasonings, which pervade “the Wealth of Nations”’ (97). Surely O’Brien must admit the abstract nature of Smith’s complex analysis of the consequences of an increase in the corn price and the fact that it attracted the adverse criticism of competent contemporaries. This leads me to O’Brien’s third point. It is to minimize the significance of the inverse relationship so that Ricardo’s correction of Smith appears inconsequential: this inverse relationship [between wages and profits] is virtually all that is left after Hollander has thrown out everything else. Of course if that is all that is left, it is easier to dismiss Ricardo’s critics ... and argue that his work leads directly to that of Marshall. (O’Brien 1981, 356–7). Of course it is highly likely that Ricardo decided that wages and profits 104

A RESPONSE TO PROFESSOR O’BRIEN

moved inversely at quite an early stage. But this hardly amounted to a new theory of profit. The obvious Bullionist objection (that either the money supply or velocity must rise if cost increases are to raise the price level) to Smith’s cost-push view of inflation hardly amounts to laying the foundation of Ricardo’s later models. Because Ricardo was a Bullionist he would, in his typically single-minded way, have rejected all explanations for price rises that did not involve increases in the money supply. (O’Brien 1981, 360–1) Here we have the classic error of reading history backwards. I have shown above that the ‘Bullionist’ objection to the Smithian model was by no means so obvious to Ricardo at the time. A lengthy and difficult intellectual effort was required before he realized the disaccord between the notion that a rise in wages leads to a general price increase and the quantity or cost theory of the value of money and, thus, arrived at the theory of proportionate wages – the inverse profit–wage relation. O’Brien has simply failed to take into account the available evidence. This evidence extends beyond the record of Ricardo’s intellectual history during the period 1809–17 summarized above. After the event Malthus himself admitted the significance of the Ricardian innovation. I cannot imagine that O’Brien has taken into account Malthus’s statement: ‘Of all the truths which Mr Ricardo has established, one of the most useful and important is that profits are determined by the proportion of the whole produce which goes to labour. It is, indeed, a direct corollary from the proposition, that the value of commodities is resolvable into wages and profits; but its simplicity and apparent obviousness do not detract from its utility’ (1963 [1824], 189; EDR, 229). Similarly, ‘we fully agree with the author of the present treatise’ – this a reference to McCulloch’s Principles – ‘that when it is said that profits depend on wages, they must be understood to depend on ... proportional wages, that is, on the share of the commodities produced by the labourer, or of their value, which is given to him’ (1963, 199). Evidently we are not dealing with a matter of minor theoretical importance as O’Brien would have it. But if all this is still not enough let him compare McCulloch’s position after 1817 with that maintained earlier: ‘a forced rise in the price of corn being attended with a corresponding rise in the price of labour, must equally augment the price of almost every article the farmer purchases ... they receive more money with one hand, but they are forced to pay away more with the other’ (1816, 138; EDR, 59). Alternatively, to convince himself that more was involved than he has admitted so far let O’Brien compare James Mill’s Elements with his characteristically uncompromising statement of the Smithian theorem in 1804: ‘No proposition is established more thoroughly to the conviction of those who have studied the scientific 105

RICARDO - THE NEW VIEW

principles of political economy than this; that the money price of corn, regulates the money price of everything’ (1804, 36; EDR, 57). I rest my case on these matters bar one final observation. The falling profit rate in consequence of the pressures exerted by land scarcity involves one application only of the inverse wage–profit relationship. Why is O’Brien so unwilling to admit to others? Ricardo himself did: ‘I have invariably insisted that high or low profits depend on low and high wages, how then can it be justly said of me that the only cause which I have recognized of high or low profits is the facility or difficulty of providing food for the labourer. I contend that I have also recognized the other cause, the relative amount of population to capital, which is another of the great regulators of wages’ (Ricardo 1951, II, 264–5; EDR, 640). But we surely are not bound to limit ourselves forever to the applications the innovating theorist happened to make. The Ricardian analysis is pertinent to the investigation of the consequences for prices and profits of a wage increase whatever its cause. Thus it proved the key for Mill’s analysis of trade unions as late as 1872: You must have been struck as I have been, by the thoroughly confused and erroneous ideas respecting the relation of wages to price, which have shewn themselves to be almost universal in the discussions about the recent strikes. The notion that a general rise of wages must produce a general rise of prices, is preached universally not only by the newspapers but by political economists, as a certain and admitted economical truth; and political economy has to bear the responsibility of a self-contradicting absurdity which it is one of the achievements of political economy to have exploded. It provokes one to see such ignorance of political economy in the whole body of its self-selected teachers. The Times joins in the chorus .... Certainly no one who knows, even imperfectly, what the Ricardo political economy is, whether he agrees with or not, can suppose this to be it. I hope you will come down upon it with all the weight of your clear scientific intellect, your remarkable power of exposition, and the authority of your name as a political economist. (To Cairnes, Mill 1972, XVII, 1909–10) And it was the model used by Mill in his treatment of differential rates of return on capital in Great Britain and the USA. My point is that the Ricardian model had implications extending far beyond the Corn Law issue. O’Brien’s perspective cannot accommodate the post-Ricardian record, particularly the Ricardo–Mill relationship.

106

A RESPONSE TO PROFESSOR O’BRIEN

VI. ON PRICING AND ALLOCATION In what follows I give my reaction to a range of rather more specific objections on O’Brien’s part. First the matter of value and allocation. I do not know upon what O’Brien bases his assertion that in my account ‘Ricardo did not attach very much importance to the Invariable Measure’ (1981, 363; cf. 354, 370). It is in fact given a central place in my account of the analytical core of Ricardian economics, particularly the derivation of the fundamental theorem on distribution ( EDR, 5f) and the investigation of Ricardo’s legacy ( EDR,660f; see also my article (below, Chapter 18) on the topic for this Journal, July 1977). Moreover, some fifty pages (7 per cent of the text) are devoted to the technical details of the measure ( EDR, 218–69). I do argue that Ricardo took the sensible position that ‘even in the event an adequate measure could not be achieved, he was none the less prepared to maintain the inverse wage–profit relation’ ( EDR,227) and did so sometimes in terms of a quantity theory of the value of money as distinct from a realcost theory ( EDR,243). And in my Chapter 6 I argue further that the allocation process involved when we trace through the consequences of a change of wages for relative prices and the profit rate can be appreciated apart from the measuring device ( EDR,302f). All this I certainly believe to be true. But nowhere do I claim, nor do my arguments imply, that the Invariable Measure was unimportant to Ricardo. I shall not repeat here my full case for Ricardo’s appreciation of demand– supply technique upon which O’Brien heaps scorn (1981, 363–4). I must, however, say that a reader who relies on the skimpy account given of my Chapter 6 on Allocation Mechanisms would not have a fair indication of the available evidence which helps explain what has hitherto been a mystery – Marshall’s insistence upon the significance of demand–supply analysis in Ricardo’ thought. 2 Secondly, for O’Brien to emphasize on one page (363) differential factor ratios and consequential change in the average upon alteration in the composition of output only to insist on the next that demand has no role to play in Ricardian economics other than to select the marginal unit in agriculture, suggests a failure to appreciate the theoretical issues at stake: it is upon the alteration in the average fixed–circulating capital ratio with change in the configuration of final demand that a case can be made for the interdependence of distribution and pricing – a main theme of the chapter. It is appropriate to allude here to O’Brien’s observations on agricultural improvement in the context of the principle of profit-rate equalization (364–5). O’Brien appeals to the famous Marshall–Stigler diagram to demonstrate that the improvement leads to ‘increased factor rewards in agriculture’ and thus to two rates of profit in the system as released factors flow to manufacturing. This argument is faulty as I explained ( EDR,293f). The increase in marginal product, given population (and thus the demand for corn) implies a corresponding fall in the corn price. What 107

RICARDO - THE NEW VIEW

follows will depend upon the assumption made regarding the commodity wage. (i) If the commodity wage is assumed fixed, then of course the money wage falls generating the familiar inverse movement in the profit rate. But the fall in wages is common to all sectors, not agriculture alone – there are thus no differential effects (unless fortuitously differing factor ratios pertain in the different sectors, in which case allocative readjustments would be set in motion until equilibrium is achieved). (ii) If real wages rise upon the fall in the price of corn – if we assume constant money wages – there will be no change in the general profit rate at all in consequence of the improvement. In both cases labour and capital released from agriculture can be reabsorbed in manufacturing without difficulty in consequence of the operation of the law of markets. I cannot make the matter clearer than this. One further substantive matter falling within the present range of topics merits consideration: the theory of rent. O’Brien refers to my statement in the concluding chapter that ‘Ricardo had shown a thorough awareness that differential rent is but a special case of a more general phenomenon’ as a ‘sweeping assertion’ for which I provide no evidence (1981, 384). The evidence will be found on the same page as the assertion. I shall repeat it here (from EDR,665–6): no one would pay for the use of land, when there was an abundant quantity not yet appropriated, and, therefore, at the disposal of whosoever might choose to cultivate it. On the common principles of supply and demand, no rent could be paid for such land, for the reason stated why nothing is given for the use of air and water, or for any other of the gifts of nature which exist in boundless quantity ... as the supply is boundless, they bear no price. If all land had the same properties, if it were unlimited in quantity, and uniform in quality, no charge could be made for its use, unless it possessed peculiar advantages of situation. (Ricardo 1951, I, 69). [if] air, water, the elasticity of steam, and the pressure of the atmosphere, were of various qualities; if they could be appropriated, and each quality existed only in moderate abundance, they, as well as the land, would afford a rent, as the successive qualities were brought into use. (Ricardo 1951, I, 75) Dr. Smith does not reflect that rent is the effect of high price. ... It is ... from the price at which the produce is sold, that the rent is derived; and this price is got not because nature assists in the production, but because it is the price which suits the consumption to the supply. (Ricardo 1951, I, 77n; citing Buchanan) 108

A RESPONSE TO PROFESSOR O’BRIEN

I see here a very obvious recognition of land scarcity as the underlying source of differential rent. That such recognition ‘would rob Bailey and Senior of credit for generalising the rent concept’ bothers O’Brien for some reason (1981, 384). It does not bother me. It is part of the historical record so we have no alternative but to recognize it and follow its implications wherever they may lead us. Finally, I note that it is in the context of allocation theory that we find another of O’Brien’s somewhat extraordinary historiographical principles. Against my demonstration of the significance attached by Ricardo to the allocative losses of agricultural protection O’Brien asserts that ‘it was hardly in the forefront of Ricardo’s picture’, adding ‘if it had been it would diminish his independent standing as an original thinker’ given Smith’s earlier preoccupation with allocative loss (381). I am quite at a loss. This principle, if taken seriously, would sweep away all textual evidence of intellectual continuity in the interest of novelty. What kind of history is this? Nor is it an accidental remark. The same reasoning is implied by O’Brien’s earlier reference (above p. 102) to my ‘smuggling in’ of Smith’s work to justify my Walrasian claims for Ricardo’s analysis. VII. WAGES AND TAXATION At one point in his review O’Brien makes the very pertinent comment that ‘Failure to solve a theoretical problem does not mean that the problem is automatically regarded as unimportant’ (1981, 363). It would have been nice had he applied this valid principle to the problem of Ricardian wages in the taxation context. O’Brien subscribes to the standard view that this analysis hinges on the subsistence wage assumption. I do not. But I do not in my book ( EDR, 375f) hide the analytical problems entailed by the more general perspective I believe Ricardo to have taken, particularly the precise mechanisms of population adjustment involved where wages are above subsistence. Yet focusing on these problems O’Brien (1981, 366f) leaps to the conclusion that Ricardo’s analysis was based on the simplest case. And this will not do. As I explain in my book ( EDR,383) Ricardo’s taxation theorems ‘can be interpreted as applying to the case where a subsistence wage rules’. I provide textual evidence from the chapter ‘On Wages’ and elsewhere to this effect. But I then show that the wage taxation theorems do not stand or fall with the subsistence assumption (386). For during the course of his discussion (in the chapter on ‘Taxes on Raw Produce’) of the effect upon the money-wage rate induced by the taxation of necessaries, Ricardo introduced the qualification that the ‘rate of progression’ of the economy is throughout taken for granted, implying that the analysis is intended to apply whether or not wages are initially at ‘subsistence’: ‘Those who maintain that it is the price of necessaries which regulates the price of labour, always allowing for the 109

RICARDO - THE NEW VIEW

particular state of progression in which the society may be, seem to have conceded too readily, that a rise or fall in the price of necessaries will be very slowly succeeded by a rise or fall of wages’ (Ricardo 1951, I, 161). If this statement were the only one of its kind, it might perhaps be dismissed as unrepresentative. But the fact is that the chapter ‘On Taxation of Wages’ itself is formally contingent upon it. The chapter ‘On Taxation of Wages’ unlike that ‘On Wages’ is based upon Smith’s proposition that ‘the demand for labour, according as it happens to be either increasing, stationary, or declining, or to require an increasing, stationary, or declining population, regulates the subsistence of the labourer, and determines in what degree it shall be either liberal, moderate, or scanty’ (215). That Ricardo should have proceeded in the chapter ‘On Wages’ and in other contexts on the assumption of a subsistence wage rate, despite his acceptance of the Smithian position according to which the equilibrium wage is that wage which assures an appropriate positive rather than a zero growth rate of labour supply, is not difficult to appreciate. It is another example of his general method which is to simplify the analysis wherever this can be done without loss. The broader application of the taxation theorems means that the mechanism of population adjustment to wages above or below a given ‘subsistence’ rate must now be applied to wage variations about a long-run labour supply curve relating the wage rate to the growth rate of population. There are technical difficulties but these do not imply that the foregoing texts vanish into thin air. VIII. MONETARY THEORY O’Brien’s very critical stance leads him to make irrelevant complaints in the context of monetary thought once again charging me with giving a false impression of continuity between Smith and Ricardo: My severest reservations about Hollander’s treatment concern his discussion of monetary thought. ... I am particularly concerned, however, about Hollander’s attempt to establish that Ricardo’s monetary theory was somehow Smithian. This involves playing down the importance of David Hume in the history of monetary thought to the point where he virtually disappears. Thus the so-called ‘Smithian’ ( EDR,105–6) position that the rate of interest is not affected by increases in the money supply is not only Hume’s (and thus derived from the same place as the rest of Ricardo’s monetary thought) but also available in the works of Massie. By inflating the role of Adam Smith vis-à-vis classical monetary theory Hollander is able to give the impression of a continuity between Smith and Ricardo which otherwise does not exist. The account of pre-Bullion-Report monetary theory I found strange. It leaves out Hume (apart from a few oblique references to Hume-type mechanisms) 110

A RESPONSE TO PROFESSOR O’BRIEN

placing Adam Smith at the forefront of pre-Ricardian monetary theory (where he certainly does not belong) and considerably overrating Ricardo’s own importance as a monetary theorist vis-à-vis Thornton and others. (O’Brien 1981, 370) This amazes me. The reader who had not examined my book and the references there given ( EDR,105–6) would be excused for concluding that the following statements by Ricardo are figments of my fruitful imagination: It has been shewn incontrovertibly by that able Writer, Dr. Adam Smith, that the rate of interest for money is regulated by the rate of profits on that part of capital only which does not consist of circulating medium, and that those profits are not regulated but are wholly independent of the greater or lesser quantity of money which may be employed for the purposes of circulation; that the increase of circulating medium will increase the prices of all commodities, but will not lower the rate of interest. (Ricardo 1951, III, 25–6) If Sir John [Sinclair] will take the trouble to consult the 4th chap. 2d book, of Dr. A. Smith’s celebrated work, he will there see it undeniably demonstrated, that the rate of interest for money is totally independent of the nominal amount of the circulating medium. It is regulated solely by the competition of capital, not consisting of money. (Ricardo 1951, III, 143) In the 4th chapter of the 2d book of the Wealth of Nations, to which I, in my last letter referred, it is demonstrated that the rate of interest depends on the rate of profits, which again is totally independent of the nominal amount of the circulating medium. (Ricardo 1951, III, 150) If after the able exposition of Dr. Smith any further argument were necessary to prove that the rate of interest is governed wholly by the relation of the amount of capital with the means of employing it, and is entirely independent of the abundance or scarcity of the circulating medium, this illustration, would, I think afford it. (Ricardo 1951, III, 194) All of this from the letters to the Morning Chronicle and the Reply to Bosanquet. There is no mention of Hume. It is incontrovertible that Ricardo had before him the text of the Wealth of Nations. That Hume may indeed have been the source for Smith is a quite irrelevant matter for the purpose at hand in my discussion which was to establish the identity of position between Ricardo and Smith at the early period in question. 111

RICARDO - THE NEW VIEW

Let me add that in the High Price of Bullion (Ricardo 1951, III, 88f) also there are extensive verbatim quotations from the Wealth of Nations to the effect that ‘the rate of interest is not regulated by the abundance or scarcity of money, but by the abundance or scarcity of that part of capital not consisting of money’ followed only by the brief statement that ‘Mr Hume has supported the same opinion’. I do not think I have distorted the record. IX. ON CLASS CONFLICT AND THE SECULAR SHARES O’Brien makes a great deal of Ricardo’s supposed emphasis upon class conflict, particularly that between landlords and other classes and complains of my position on this issue (1981, 355, 375, 382). He finds it ‘very surprising that in the section (EDR, 586–93) “Ricardo as Radical” his attitude towards improvements is not mentioned’ (382). I do not know how to satisfy so severe a critic since Ricardo’s attitude to the class of landlords is examined in this section in detail specifically in terms of the problem of improvements. I certainly do ‘refuse to recognize the importance of class conflict in Ricardo’ for my reading reveals no evidence of this order. Ricardo himself denied it firmly. I also raise the question in the context of agricultural improvement: ‘can it be said that Ricardo, despite his protestations, had in fact unduly emphasized the lag in population response, uncharacteristically weighing the short-term consequences of technical change on the landlords’ interests as Malthus indeed charged?’ ( EDR, 591) – a question answered in the negative in the light of my demonstrations in other contexts of his concern with the consequences of disturbances prior to population expansion. I would like to call attention here to the fact that the characteristic approach adopted by Ricardo in his formal work – his deliberate devising of ‘strong cases’ better to portray his principles (cited above, p. 97, from Ricardo 1951, VIII, 184) – is specifically illustrated from the treatment of the effects of agricultural improvements: ‘I never thought for example that practically any improvements took place on the land which would at once double its produce, but to shew what the effect of improvements would be undisturbed by any other operating cause, I supposed an improvement to that extent to be adopted, and I think I have reasoned correctly from such premises. I am sure I do not undervalue the importance of improvements in Agriculture to landlords, though it is possible that I may not have stated it so strongly as I ought to have done.’ This, we know, is standard Ricardian method. Ricardo’s protestations ring true. It is appropriate also to recall his insistence upon his method in the more general context, ‘On Machinery’: ‘The statements which I have made will not, I hope, lead to the inference that machinery should not be encouraged. To elucidate the principle, I have been supposing, that improved machinery is suddenly discovered, and extensively 112

A RESPONSE TO PROFESSOR O’BRIEN

used but the truth is, that these discoveries are gradual, and rather operate in determining the employment of the capital which is saved and accumulated, than in diverting capital from its actual employment’ (I, 395). I emphasize this because O’Brien has insisted (1975, 130) that recognition that population and improvements tend to go hand in hand, or at least that any fall in rent due to improvement is merely a short-run phenomenon and reversed in the long run upon population expansion, are post-Ricardian arguments. They are in fact Ricardo’s arguments, and not ‘fleetingly’ and unwillingly made as O’Brien has it (see also Ricardo 1951, I, 79– 80, 412; IV, 19, 81n). It would also help if we follow Ricardo and distinguish between agricultural innovation and the protection issue. Landlords did not have it in their power to prevent the former: I have indeed observed that improvements in agriculture were in their immediate effects injurious to the landlord, and beneficial to consumers, but that ultimately when population increased, the advantage of the improvement was transferred to the landlord. To this opinion I also adhere, but in saying so I cast no reproach on landlords – they have not the power to arrest improvements, nor would it be their interest to do so if they could. Great improvements in any branch of production are in their first effects injurious to the class who are engaged in that branch, but this is a statement of a fact or of an opinion, and cannot be supposed to cast any injurious reflections. Mr. Malthus is not justified by any thing I have said in pointing me out as the enemy of the landlords, or as holding any less favorable opinion of them, than of any other class of the community. (Ricardo 1951, II, 118–19) Landlords, however, were in a position to lobby effectively for protection against foreign imports; it is to this of course that Ricardo strongly objected in the Parliamentary speeches cited by O’Brien (1981, 383; see my EDR, 625). But I do not see that this attitude is a matter of class bias. Ricardo typically objected to any special interest group exerting pressure at the expense of the community as a whole. O’Brien’s case still does not convince me. Finally, it is appropriate to raise in this context the matter of the secular shares. O’Brien insists upon a wealth of evidence suggesting that Ricardo made a case for a rising rental share and suggests such evidence supports his view of class conflict (1981, 374–5). He alludes especially to the Table in the Essay on Profits indicating (1) that rent as a share of net output rises from 7.3 per cent to 60.4 per cent; (2) that rent as a share of net output plus capital rises from 2.3 per cent to 13 per cent; and (3) that rent as a proportion of capital rises (from 3.5 per cent to 16.5 per cent). The fact is that none of these measures refers to an index appropriate for class income 113

RICARDO - THE NEW VIEW

distribution: ‘net output’ in the Table includes profits and rent only excluding wages; while ‘capital’ includes more than wages fund. I stand by my conclusion that Ricardo was not concerned with three-fold class income distribution ( EDR,404–11). X. CONCLUSION In my view Professor O’Brien’s version of Ricardo rests upon error – error which he shares with others. He is totally committed to the corn model of growth, to the vision of Ricardo as an extreme abstractionist, to a sharp discontinuity in doctrine and method between Smith and Ricardo, to the notion of Ricardian economics as a ‘detour’ in nineteenth-century thought. My views on these matters differ from those of O’Brien and similar thinkers. I have tried to present Ricardo in all his complexity in order to build up a consistent picture, as far as possible, from the available evidence – the texts, the letters, the speeches, the pamphlets – without being strait-jacketed from the outset by the version of sundry textbooks. The full evidence does not, I maintain, support the ‘majority’ view. NOTES 1. I have a harsh critic. He charges me at one point with ‘fail[ing] to quote and discuss the vital material on diminishing returns which McCulloch believed meant that Anderson should be credited with inventing rent theory’ except ‘grudgingly and in criticism of Adam Smith’ (1981, 384). This despite my detailed discussion of Anderson’s work which concludes ‘The notion of no-rent land at the (extensive) margin of cultivation is expounded with most impressive clarity in an account that has much in common with that regarded as “revolutionary” nearly forty years later’ – an account which ‘has much in common with that of Ricardo’ and which J.R. McCulloch described as marking ‘an-important era in the history of economic science, etc.’ ( EDR, 43). 2. I am very surprised not to find a reference to the paper by Rankin (1980) which independently reached many of the same conclusions as myself.

REFERENCES Anderson, J. (1801) A Calm Investigation of the Circumstances that Have Led to the Present Scarcity of Grain in Britain, London: J. Cumming. Casarosa, C. (1978) ‘A New Foundation of the Ricardian System’, Oxford Economic Papers XXX: 38–63. Dobb, M. (1973) Theories of Value and Distribution since Adam Smith, Cambridge: Cambridge University Press. Garegnani, P. (1976) ‘On a Change in the Notion of Equilibrium in Recent Work on Value and Distribution’, in M. Brown, K. Sato and P. Zarembka (eds) Essays in Modern Capital Theory, Amsterdam: North-Holland Hicks, J.R. and Hollander, S. (1977) ‘Mr. Ricardo and the Moderns’, Quarterly Journal of Economics XCI: 351–69. 114

A RESPONSE TO PROFESSOR O’BRIEN

Hollander, S. (1973) The Economics of Adam Smith, Toronto: University of Toronto Press. ——(1977) ‘The Reception of Ricardian Economics’,Oxford Economic Papers XXIX: 221–57. ——(1979) The Economics of David Ricardo, Toronto: University of Toronto Press. Levy, D. (1976) ‘Ricardo and the Iron Law: A Correction of the Record’, History of Political Economy VIII: 235–52. McCulloch, J.R. (1816) Essay on the Question of Reducing the Interest on the National Debt, Edinburgh: David Brown, and Adam Black. Malthus, T.R. (1803) An Essay on the Principle of Population, London: Joseph Johnson. ——(1970) [1814]Observations on the Effects of the Corn Laws (London) in Pamphlets of T.R. Malthus, New York: Augustus M. Kelley. ——(1963) [1824] ‘Political Economy’ in B. Semmel (ed.)Occasional Papers, New York: Burt Franklin. London: C. and Mill, J. (1804) An Essay on the Impolicy of a Bounty on the Exportation of Grain, R. Baldwin. Mill, J.S. (1972) Collected Works, vol. XVII, Toronto: Toronto University Press. O’Brien, D.P. (1970) J.R. McCulloch: A Study in Classical Economics, London: George Allen and Unwin. The Classical Economists, Oxford: Clarendon Press. ——(1975) ——(1981) ‘Ricardian Economics and the Economics of David Ricardo’, Oxford Economic Papers XXXIII: 352–86. Pasinetti, L. (1974) Growth and Income Distribution: Essays in Economic Theory, Cambridge: Cambridge University Press. Rankin, S. (1980) ‘Supply and Demand in Ricardian Price Theory: A Re-interpretation’, Oxford Economic Papers XXXII: 241–62. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Roncaglia, A. (1978) Sraffa and the Theory of Prices, Chichester and New York: John Wiley. Shackle, G.L.S. (1972) Epistemics and Economics: A Critique of Economic Doctrine, Cambridge: Cambridge University Press. Smith, A. (1937) [1776] The Wealth of Nations, New York: Modern Library. Stigler, G. (1965) ‘Textual Exegesis as a Scientific Problem’, Economica XXXII: 447– 50.

115

7 A REPLY TO PROFESSOR RONCAGLIA

Professor Roncaglia and I are still far apart in our conceptions of Ricardian economics. Yet I myself once adopted Roncaglia’s version: ‘The classification of Ricardian economics as a “detour” ... might be valid if it can be demonstrated that the characteristically “Ricardian” features lie within the tradition of allocation economics. That this is so is by no means certain. To the extent that Ricardo was in fact developing an alternative “basic theory” – whereby distribution and value are treated separately – the view of Knight and Schumpeter ... is seriously misleading’ (Hollander 1973, 14). This perspective coloured my view of the nature of subsequent developments in classical economics and I again committed myself to a sharp conceptual distinction in writing of J.S. Mill as attempting an impossible balancing trick (1976). It was only after living with Ricardo for an extended period of time that the implications registered with me of his characteristic methodology – deliberate simplification where appropriate for purely pedagogic purposes, pre-eminently constancy of the real wage and uniform factor ratios. Here will be found a principal source, so it now seems, of the vastly different interpretations yielded by honest readings of Ricardo’s Principles. I. THE MAJOR ISSUE The main theme of Roncaglia’s review is that in Ricardian (classical) theory, ‘technology and the wage rate are the necessary and sufficient data for determining the rate of profits (and “exchange values”)’ (1982, 342; cf. 354). As for ‘technology’ (which includes the levels of production in the various sectors): ‘In the classical (“surplus”) approach the economic system reproduces itself over time, and the study of the forces modifying the structure of production is severed from the analytical “core” of the system. Ricardo, in particular, does not consider “the choice” of the structure of production; this is visualized as a historical problem, with the production structure evolving from social habits and technological developments’ (355). The wage rate – given at subsistence (341) – is similarly treated as ‘exogenous’ or ‘external to the 116

A REPLY TO PROFESSOR RONCAGLIA

analytical core’ (356). On this ‘Cambridge’ view value and output levels are not determined within a single model. But consumer demand and thus output might influence cost prices by acting upon the relative scarcity of the factors and thus their returns. To divorce value and output level is to exclude this possibility, thereby implying the divorce of value and distribution. The ‘classical’ perspective is then set in contrast to that of the neoclassicist or marginalist ‘allocation’ theory, which treats the optimum allocation of scarce resources as a problem ‘typically solved by the simultaneous determination of equilibrium prices and quantities’ (355). Similarly: ‘In the Walrasian model ... once technology, consumers’ tastes, and the initial endowments of resources are given, all prices, including the services of the factors of production, are the solution to a general equilibrium system. Here “distribution and pricing are interdependent”, for they are one and the same thing’ (356). If the wage rate and the structure of production are treated exogenously, Roncaglia’s ‘Cambridge’ results immediately fall into place. It is my argument in The Economics of David Ricardo (1979), by contrast, that there is a symmetrical explanation of the returns to labour and capital in terms of equilibrium between the demand and supply for the factors, reflecting an interdependence of the two systems of markets – that for factors and for products. Specifically, consumer choices (and hence outputs) can affect the relative scarcity of labour and capital, and thus the wage and interest rates; and distributional changes in turn can alter relative costs and thus prices and outputs, given the pattern of demand schedules. From my argument there follows the identity of the conceptions of classical ‘natural’ and neoclassical long-run ‘normal’ prices reflecting uniform rates of wages and uniform rates of profit on the supply prices of capital goods, and the identity in the two general systems of adjustments to such prices. The centres of gravitation to which even Ricardo’s prices tend are genuine ‘equilibria’ between supply and demand. This is the general perspective from which I approach Roncaglia’s review. 1 II. THE WAGE RATE AS ENDOGENOUS VARIABLE I turn first to Roncaglia’s rejection of the Ricardian model of growth by Hicks and Hollander and Casarosa. In these renditions population growth is tied to the wage rate, treated as an endogenous demand-and-supply variable. To Roncaglia, however, ‘Ricardo does not state that the greater the gap between actual and subsistence wage rates, the faster the growth of population; he avoids even a hint of a tight leash between wage rate and population growth’ (1982, 349). I shall now briefly review some of the contrary evidence supporting a Ricardian labour-supply function. We have first the standard ‘Malthusian’ case, stated even earlier by Adam Smith, who defined the same general functional dependency (1937, 80). This was later taken over by Ricardo, whose chapter ‘On Taxation of Wages’ is based upon Smith’s 117

RICARDO - THE NEW VIEW

proposition that ‘the demand for labour, according as it happens to be either increasing, stationary, or declining, or to require an increasing, stationary, or declining population, regulates the subsistence of the labourer and determines in what degree it shall be either liberal, moderate, or scanty’ (1951, I, 215). Accordingly, the wage will be just sufficient to support the population, which at that time the state of the funds for the maintenance of labourers, requires. If the labourer’s wages were before only adequate to supply the requisite population, they will, after the tax, be inadequate to that supply, for he will not have the same funds to expend on his family. Labour will, therefore, rise, because the demand continues, and it is only by raising the price, that the supply is not checked. ... Suppose the circumstances of the country to be such, that the lowest labourers are not only called upon to continue their race, but to increase it; their wages would be regulated accordingly. Can they multiply [in the degree required], if a tax has taken from them a part of their wages, and reduced them to bare necessaries? (Ricardo 1951, I, 219–20) I do not see how there could be a clearer expression of the functional relationship between population and the wage rate. Ricardo’s position, in essentials, is precisely that of Smith and Malthus. The rate of capital accumulation acts as an ‘independent variable’ which determines what the commodity wage rate must be to guarantee an equivalent growth rate of population. The assumption that the rate of capital accumulation is an independent variable – totally unaffected by the reduction in profits corresponding to an increase in money wages – is, however, no more than a first approximation. A reduction in profits would, Ricardo conceded, probably have some effect on accumulation, so that the compensatory increase in money wages would not entirely prevent a fall in real wages in consequence of taxation (I, 221–2). Most important for us here is the fact that once it is recognized that wages, for Ricardo, are an endogenous variable, it becomes difficult to represent profits in terms of ‘surplus’ (Roncaglia, 1982, 341, 350–1). Profits, of course, ‘depend on wages’, but labour demand and thus the wage rate are in turn affected by changes in the profit rate. There is a mutuality of interrelation between profits and wages (within a supply and demand framework) – well-expressed by Ricardo himself, I emphasize, in terms of a ‘natural equilibrium’ (1951, I, 226). III. PROFITS: A REWARD FOR ‘WAITING’ I am sceptical of the notion of surplus if a change in the profit rate acts upon the rate of accumulation via the ‘motive’ to save. What was Ricardo’s position in fact? 118

A REPLY TO PROFESSOR RONCAGLIA

Roncaglia denies Marshall’s belief that the rate of profit is related by Ricardo to ‘waiting’ (1982, 345). It can always be said (with Roncaglia) that ‘ given the existence of profits and the level of the rate of profits, prices must behave to accommodate them, without implying anything by way of explanation’. But I have tried to resolve this issue (1979, 211, 214, 317ff, 326, 672) and stand by my previous conclusion upholding Marshall’s interpretation. Ricardo’s famous letter to McCulloch (13 June 1820) places ‘compensation for time’ on a par with ‘compensation for labour’: ‘When the times are unequal, the relative quantity of labour bestowed on [commodities] is still the main ingredient which regulates their relative value, but it is not the only ingredient, for besides compensating for the labour, the price of the commodity, must also compensate for the length of time that elapses before it can be brought to market. All the exceptions to the general rule come under this one of time’ (1951, VIII, 193). This formulation is incorporated into the third edition of the Principles with an important addendum: the values of commodities (cloth or cotton) produced in a two-year process are shown to be more than double that of a commodity produced in one year (corn), although precisely double the labour is assumed to be embodied in the former cases, ‘for the profit on the clothier’s and cotton manufacturer’s capital for the first year has been added to their capitals, while that of the farmer has been expended and enjoyed’ (I, 34). Does not Ricardo imply a pain cost that requires compensation if the period of investment of capital is to be lengthened? The observation in the Notes on Malthus to the effect that ‘the power and will to save ... must depend upon the share of the produce allotted to the farmer or manufacturer’ (II, 303; emphasis added) seems to clinch the matter. The evidence bears out Marshall’s position. To summarize: if, as Ricardo seemed to believe, profit is a compensation for ‘postponed enjoyment’, its representation as ‘surplus’ in the sense of ‘residual’ (Roncaglia 1982, 351) is unhelpful. It may be retained to describe an income potentially available for accumulation or taxation, and also (as I explain, 1979, 265ff) in the sense that the sole contractual payment in the system is that made to labour. I would emphasize too that allowing a functional relationship between savings decisions and the interest rate conflicts in no way with the basic inverse profit–wage relationship, as Roncaglia appears to believe. IV. RICARDO ON DEMAND AND THE ‘CIRCULAR-FLOW’ PROCESS I turn now to demand theory, with at least a mild protest at Roncaglia’s assertion that ‘all Hollander can find in Ricardo’ on demand is the proposition that ‘consumption is the sole end and purpose of all production’ (1982, 342). My book (1979, ch. 6) demonstrates the rationalization of the elasticity property, the use of that property (together with supply elasticity) in the derivation of the ‘Marshallian’ 119

RICARDO - THE NEW VIEW

principles of tax incidence, and the role accorded demand–supply analysis in costprice determination, costs including the average (uniform) profit rate. 2 I rehearse here merely a fraction of the evidence supporting Marshall’s insistence that Ricardo ‘seems to be feeling his way towards the distinction between marginal and total utility’ (1920, 814), a position strongly denied by Roncaglia (1982, 344). Consider first Ricardo’s defence of Smith against Say: M. Say accuses Dr. Smith of having overlooked the value which is given to commodities by natural agents, and by machinery, because he considered that the value of all things was derived from the labour of man; but it does not appear to me, that this charge is made out; for Adam Smith no where undervalues the services which these natural agents and machinery perform for us, but he very justly distinguishes the nature of the value which they add to commodities – they are serviceable to us, by increasing the abundance of productions, by making men richer, by adding to value in use; but as they perform their work gratuitously, as nothing is paid for the use of air, of heat, and of water, the assistance which they afford us, adds nothing to value in exchange. (Ricardo 1951, I, 286–7) I read this statement as a recognition of the distinction between total utility and marginal utility – in this case a marginal utility of zero. But let us simply add Ricardo’s statement a few pages before that ‘by scarcity the value of commodities is raised’ (I, 276; see also 69–70 cited below).3 On p. 354 of the review I read ‘that neither Ricardo nor Smith uses supply and demand schedules’. I strongly disagree. In my Economics of Adam Smith (1973, 117ff), I demonstrate Smith’s use of demand and supply schedules in his famous chapter ‘Of the Natural and Market Price of Commodities’ in the context of profit-rate equalization. Ricardo fully subscribed to all this (1951, I, 91). Indeed, the analysis of market price and its tendency to natural price is incomprehensible without demand schedules, for the output movements, which Roncaglia (1982, 354) allows occur when market and natural prices diverge, are surely responsible for the inverse movements of market prices toward their long-run equilibria, implying variation up or down demand curves. In Ricardian theory there is no formal discussion of the imputation from the value of the final product to that of the productive service (derived demand); indeed there is no conception of incremental variation of individual productive factors. Do not these characteristics set Ricardian economics apart from the later neoclassical variety? I do not think so. Recall that along with his fourth proposition on capital, that ‘demand for commodities is not demand for labour’, J.S. Mill explicitly adopted the view that ‘all 120

A REPLY TO PROFESSOR RONCAGLIA

concerned in production depend for their remuneration on the price of a particular commodity’ and ‘derive their remuneration from the ultimate product’ (Mill 1965, 32–3, 455). There is no logical incompatibility between the two perspectives: the fourth proposition on capital was not designed to deal with the return to labour in particular uses but rather with aggregate wages and aggregate employment. Accordingly, the later neoclassical developments along imputation lines are best envisaged as an elaboration of a relatively neglected aspect of distribution rather than a paradigmatic displacement of the earlier position. 4 Mill’s position is no different from Ricardo’s. Consider first a statement by J.B. Say in the Traité d’économie politique : It is utility which determines the demand for a commodity, but it is the cost of its production which limits the extent of its demand. When its utility does not elevate its value to the level of the cost of production, the thing is not worth what it cost; it is a proof that the productive services might be employed to create a commodity of a superior value. The possessors of productive funds, that is to say, those who have the disposal of labour, of capital or land, are perpetually occupied in comparing the cost of production with the value of the things produced, or which comes to the same thing, in comparing the value of different commodities with each other; because the cost of production is nothing else but the value of productive services, consumed in forming a production; and the value of a productive service is nothing else than the value of the commodity, which is the result. The value of a commodity, the value of a productive service, the value of the cost of production are all, then, similar values when every thing is left to its natural course. (Say; cited by Ricardo 1951, I, 281–3; emphasis added) Not unexpectedly Walras much later believed Say to have been on the right road (1954, 425). But so did Ricardo, who commented on this passage: ‘M. Say maintains with scarcely any variation, the doctrine which I hold concerning value.’ The sole complaint related to Say’s treatment of the services of land (1951, I, 283–4). 5 I fail therefore to appreciate Roncaglia’s representation of ‘modern or marginalist theory’ as a one-way avenue that leads from ‘factors of production’ to ‘consumption goods’ in contrast to the classical ‘system of production and consumption as a circular process’ (1982, 342; cf. 354). Every orthodox ‘Principles of Economics’ commences by expounding a ‘circularflow’ process of production and consumption. And, as we have just shown, this approach, formally stated by J.B. Say, was commended by Ricardo. Also relevant here is Ricardo’s representation of rent as ‘that compensation, which is paid to the owner of land for the use of its original and indestructible powers’ (1951, I, 67). The exposition of rent is in terms of scarcity considerations: ‘no one 121

RICARDO - THE NEW VIEW

would pay for the use of land [in new settlements], when there was an abundant quantity not yet appropriated, and, therefore, at the disposal of whosoever might choose to cultivate it’, and draws on J.B. Say: ‘On the common principle of supply and demand, no rent would be paid for such land, for the reason stated why nothing is given for the use of air and water ... as the supply is boundless they bear no price’ (I, 69–70). There is too the insistence (against Adam Smith) that it is ‘from the price at which the produce is sold, that rent is derived; and this price is got ... because it is the price which suits the consumption to the supply’ (I, 77n). We are here on the way toward a productivity explanation of factor returns with an eye on ‘derived’ demand. The final-demand dimension is certainly most obvious in the case of land; but, by accepting Say’s formulation of circular flow, Ricardo recognized its pervasiveness and applicability to factors whose returns do enter into the calculation of cost prices in the form of wages and profits. Roncaglia insists that for Ricardo ‘profits are generated in production’ (1982, 342; cf. 354). This is true, and nothing that has been said thus far controverts the point. But he seems also to believe that Ricardo’s rejection of Smith’s approach to profits in terms of ‘competition of capitals’ implies rejection of demand–supply analysis in factor markets, and this is a misconception. Ricardo’s rejection of Smith’s position turned upon the law of markets, for it seemed that Smith allowed excess overall supply by his formulations (Hollander 1979, 509). This reaction fits perfectly with our demonstration of Ricardo’s adherence to Say’s conception of economic organization. V. RICARDIAN COST PRICE AND THE SCARCITY PRINCIPLE The absence of a generalized conception of the incremental variation of factor proportions does not constitute an irreparable constraint on the applicability of Ricardian theory to standard neoclassical ‘rationing’ problems. In effect, economywide substitution between capital and labour was incorporated by Ricardo even though neglected within the enterprise–the procedure in the early editions of Walras’ Elements. To this matter I turn next, having in mind Roncaglia’s assertions that Ricardo ‘does not consider “the choice” of the structure of production’ which is ‘a historical problem ... evolving from social habits, and technological developments’ (1982, 355); that ‘natural prices and production structure are clearly understood [by Ricardo] as previously determined data for the adjustment process [of market to natural price]’ (355); and that ‘for the classical theory, relative values reside in the objective “difficulty of production”, be it represented by a single magnitude (laborcontained) or by the technical coefficients of production’ (343). 122

6

A REPLY TO PROFESSOR RONCAGLIA

This position can easily be shown to be defective: relative values (reflecting cost prices) are for Ricardo a function of distribution and not only of technology or production techniques. In fact, Ricardo’s discussion of longrun cost price determination provides a beautiful illustration of the operation of the scarcity principle as it pertains to capital and labour. Ricardo makes it clear that the process of market-price adjustment to cost price occurs by way of supply variation: ‘I do not say that the value of a commodity will always conform to its natural price without an additional supply, but I say that the cost of production regulates the supply, and therefore regulates the price’ (1951, II, 48–9). Consider now the case of differential factor proportions and Ricardo’s first chapter, where it is explained that the consequence of a rise in wages assuming an initial equilibrium (uniform rate of profit) is an altered structure of cost prices, labour-intensive commodities rising relative to capital-intensive commodities. I ask one simple question: What else can possibly be involved here but a rise in costs of the one category relative to the other in consequence of the supposed disturbance to the initial equilibrium, followed by a contraction in output of the first category and an expansion of the second until market prices reach the new cost levels? And if this is so, what is at work if not a rationing of labour – an economy-wide substitution against labour – by way of alterations in the outputs of the various commodities? The implications of this perspective are extensive. Wages, we have seen, are treated by Ricardo as a price, determined by demand – supply relations. The analysis proceeds at the aggregate level, demand represented by part of the capital supply and supply by the work force; it is the average wage that is at stake, not the wage rate paid to particular categories of workers. Now Ricardo followed the Smithian procedure in his approach to wage-rate determination, but took an important step forward in the chapter ‘On Machinery’, where variations in the circulating–fixed capital division were introduced and the implications for labour demand and the wage rate traced out. It is true that his formal analysis of the allocative effects of changes in the pattern of demand implies uniformity of factor proportions (Hollander 1979, 287– 8).7 But once this restrictive assumption is relaxed, and the principle developed in the discussion of machinery extended generally, there is no way of avoiding the conclusion that changes in the pattern of final demand may affect the demand for labour and thus the general wage rate (by altering the overall circulating–fixed capital ratio), leading in turn to a reorganization of the commodity mix to reflect the altered cost relativities. I am not claiming that Ricardo himself spelled out this mechanism, but that the mechanism is consistent with the existing corpus of Ricardian theory; there are no ‘paradigmatic’ differences between Ricardian and neoclassical theory in so far as concerns the effects upon distribution of a change in the pattern of final demand. The structure of production – like the wage rate itself – is a key variable of 123

RICARDO - THE NEW VIEW

Ricardo’s theoretical system, and the effort to divorce Ricardo from later nineteenthcentury neoclassical developments is unconvincing. To my mind it is Ricardo’s great achievement to have demonstrated that a rise in the wage rate can alter the structure of cost prices (and outputs) – a correction of the Smithian ‘adding-up’ analysis, using allocation tools learned largely from the Wealth of Nations itself. To see things in this light does not, I think, ‘fog communication’ (Roncaglia 1982, 325). VI. AN ECUMENICAL POSTSCRIPT The differences between Roncaglia and myself are perhaps less great than would appear. Throughout the review I find concessions which narrow the gap. For example: The distinction between the classical surplus and the marginalist– neoclassical themes has been characterized as a counterposition of reproducibility versus scarcity. ... Yet, both aspects are present in each approach: in the classical theory scarcity is a prerequisite for a good to be a commodity, i.e. to be economically relevant – air is not a commodity. For marginalists the productive transformation of primary resources into useful goods is taken into account as a complication which does not imply substantial modification of the structure of the analysis. (Roncaglia 1982, 343) Roncaglia is too hesitant in his allowance regarding the scarcity dimension of classicism; and I do not quite understand his point regarding the marginalists. But I commend the spirit of the observation. Roncaglia says of my demonstration of the interdependence of distribution and pricing in the Ricardian system that it is ‘correct, in part’, appending the qualification that the wage rate is ‘external to the analytical core’ (356). I do not mind so much whether one defines the wage rate as ‘external’ if it is conceded that final demand patterns play a part in its determination. On one occasion, Roncaglia, in the context of market-price movements, seems to accept Marshall’s position regarding the complementarity of objective and subjective theories of value, but insists that ‘the analysis of the adjustment process is clearly distinguished by Ricardo from his main theoretical preoccupation on the determination of “natural” or “normal” prices’ (344). It is a step in the right direction to recognize Ricardo’s use of demand–supply analysis in the market case, since the Cambridge view denies this. What remains is to recall that the classical natural price is nothing but the Marshallian, and as can be shown, Walrasian (cf. Hollander 1981 [see Chapter 9]), long-run supply price; and second, that an economy-wide wage change can affect the entire structure of cost prices. Cost prices are the prices to 124

A REPLY TO PROFESSOR RONCAGLIA

which market prices ‘tend’, but we must not forget that they are themselves subject to disturbance by distributive as well as technological changes. Finally, I am gratified by Roncaglia’s commendation of my position that the key theorem of Ricardian economics is the inverse wage–profit relationship while the analysis of the secular fall in the profit rate is only a A REPLY TO PROFESSOR RONCAGLIA writes with reference to the inverse relationship of ‘a certain set of analytical propositions, as being somehow separable from the others as the “core” to which the term “science” is fully applicable’ (350). I agree with this observation, although as it stands it is rather vague. As I see it, there is no question that Ricardo constructed a growth model and that this was part of his analytical endeavour. But as I have demonstrated (1979, ch. 9), this model was not intended to yield a specific historical prediction relating to the wage and profit rates, considering the likelihood of changes in technology; yet the movement in the profit rate, whatever it might be, is always explicable in terms of the inverse profit–wage theorem, which holds good whatever the particular values of the distributive variables that may be historically relevant. This formulation may help define the nature of the ‘scientific core’ that Roncaglia is seeking. NOTES 1. Roncaglia (1982, 351) leaves aside my criticism of the ‘corn model’ and the ‘physical rate of profit’ attributed to Ricardo, relying on an ‘apt refutation’ by Eatwell. See, however, my reply to Eatwell (Hollander 1975). 2. Roncaglia (1982, 355) seeks to make much of Ricardo’s statement that ‘no general rule can be laid down for the variations of price in proportion to quantity’. But here Ricardo (very reasonably) is simply rejecting the possibility of specifying a particular elasticity coefficient pertinent to all commodities at all times (see Hollander 1979, 276). 3. In this same context Roncaglia (1982, 344) objects to my representing Ricardo’s concern for real per capita income as a concern for utility. How then are we to explain Ricardo’s statements that ‘consumption is the sole end and purpose of all production’ (1951, V, 271), or that ‘by increasing production ... you increase the general happiness’ (I, 271), or (on the disadvantages of a closed system) ‘I do not want to know what value we could have obtained ... but what riches we might have got – what means of happiness to the community’ (II, 203), or similarly, the representation of the gain from trade following J.B. Say as ‘an utilité produite ’ (I, 320)? On all this see my book, 1979, 544ff. 4. Roncaglia is in error when he asserts (1982, 341) that the classical ‘vision’ of economic process entails the notion of literal pre-accumulations of subsistence. Ricardo, for one, rejected the notion; see Hollander 1979, 274, 326ff. So did Mill, but I cannot go into that matter here. 5. See also the doctrine of productive services in Say (1967, 12ff) regarding which Ricardo wrote to Malthus: ‘if he would give up rent, he and I should not differ very materially on that subject’ (1951, VIII, 277), and to Say: ‘In your doctrine of productive services I almost fully agree, but I submit to you, whether, as rent is the effect of high price, and not the 125

RICARDO - THE NEW VIEW

cause of it, it should not be rejected when we estimate the comparative value of commodities’ (VIII, 279). 6. ‘It is in connection with rent that we find the nearest approach in the classical writings to a process of imputation on the basis of final increments’ (Knight 1956, 69). 7. The passage cited by Roncaglia (1982, 354) from Ricardo (1951, I, 382) applies to this case.

REFERENCES Hollander, Samuel (1973) The Economics of Adam Smith, London and Toronto: University of Toronto Press. ——(1975) ‘Ricardo and the Corn Profit Model: Reply to Eatwell’,Economica 42. ——(1976) ‘Ricardianism, J.S. Mill, and the Neo-classical Challenge’ in J.M. Robson and M. Toronto: University Laine (eds) James and John Stuart Mill: Papers on the Centenary Conference, of Toronto Press. ——(1979) The Economics of David Ricardo, London and Toronto: University of Toronto Press. ——(1981) ‘On the Substantive Identity of the Classical and Neo-classical Conceptions of Economic Organization’ (unpublished manuscript). Knight, F.H. (1956) On the History and Method of Economics, Chicago: University of Chicago Press. Marshall, Alfred (1920) Principles of Economics, 8th edn, London: Macmillan. Mill, J.S. (1965) [1848] Principles of Political Economy, Collected Works, vols II and III, Toronto: University of Toronto Press. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Roncaglia, A. (1982) ‘Hollander’s Ricardo’, Journal of Post Keynesian Economics IV, 3 (Spring): 339–59. Say, J.B. (1967) [1821] Letters to Mr. Malthus, London: Sherwood, Neely, and Jones. Smith, Adam (1937) [1776] The Wealth of Nations, New York: Modern Library. Walras, Léon (1954) Elements of Pure Economics, ed. W. Jaffé, 4th definitive edn (1926), London: George Allen and Unwin.

126

8 ‘PROFESSOR HOLLANDER AND RICARDIAN ECONOMICS’: A REPLY TO PROFESSOR MOSS

I. Professor Moss (1979) provides a helpful and commendably accurate summary of the various interpretations to be found in the literature on Ricardian economics including my position which gives pride of place to the inverse profit–wage relationship and regards changes in the price of corn simply as one possible cause of variation in the money wage. Moss (505–6) draws the valid distinction between two forms of the inverse profit– wage relation: first, what he calls the ‘tautological version that, if wages plus profits equal output and output is held constant, then a rise in wages will lower profits’; and secondly the version involving market process. Now I do not claim in my book that the second version was spelled out in any detail by Ricardo; what I do claim is that competitive allocation analysis is a central part of Ricardian theory, and that the consequences of a change in the wage rate for prices and profits outlined by Ricardo imply the operation of the market process even though he left the exercise to his readers. In making my case for a significant influence upon his successors, however, it is the first version that takes pride of place since I do not suggest that Ricardo’s readers appreciated the implicit process analysis entailed. Moss subscribes to the ‘consensus view’ that the essence of Ricardianism lies specifically in the agricultural theory of profit (the corn model as it is sometimes called) rather than the general inverse profit–wage relation (503). The reason for this, in part, seems to be his belief that the inverse wage– profit relation Mark I is quite insignificant. On this we must part company. Even the first version I regard as of the highest significance so that if, as I believe can be demonstrated, Ricardo persuaded his successors to adopt it one can speak of a meaningful legacy. 127

RICARDO - THE NEW VIEW

Does not Moss’s historical perspective go awry when he dismisses the first version as ‘tautological’ – ‘to deny the validity of this proposition is surely tantamount to denying the basic logical law of noncontradiction’ – and ‘unimportant’? I am not sure why he takes for granted that what is ‘tautological’ is necessarily ‘unimportant’ as he seems to do; on the contrary, surely the growth of knowledge involves in part a growing awareness of ‘tautological’ relationships. This indeed seems to be the case in the present instance. It is obviously true that an increase in a proportionate share must imply a decrease in the residual proportionate share. But Ricardo’s efforts were not devoted to proving this proposition – his concern was with the preceding stages of the argument (cf. my The Economics of David Ricardo, 267f). In the first place, he struggled (unsuccessfully as he realized) to devise a means of assuring constancy in the value of the total to be shared between wages and profits. Secondly, if this problem is put aside, or assumed away, and we presume in principle (if not in practice) an approximately unchanged value of the produce to be distributed, Ricardo sought to demonstrate that only certain categories of wage-rate variation may be identified with variations in the wage share. Specifically, a wage-rate increase which reflects an increase in the cost of producing wage goods or a larger wage basket (in conditions of unchanged productivity) is identifiable with an increased wage share of output (net of rent), while a wage-rate increase which is purely nominal – reflecting an alteration in the cost (or quantity) conditions of the monetary medium – leaves the share of wages unchanged, a constructive distinction of the very first importance. Thirdly, the identification of an increase in the ‘real’ wage with an increase in the wage share in output net of rent is itself not a truism because it does not hold good universally. An increase in the labour embodied in wages may not involve an increase in the wage share if the value of output-as-a-whole (net of rent) should be increasing. An increase in the share of wages is implied unambiguously only if the value of per capita wages rises. The fact must also be taken into account that Ricardo did not exert all his efforts merely to arrive at a statement of the inverse relation between wages and profits envisaged as proportions. This was only a step towards the ultimate objective which was to define the determinants of the rate of return on capital. The rate of profit is said to be dependent upon, but is certainly not identified with, the share of profits in per capita or total output (net of rent), and the relationship is certainly not a ‘truistic’ one since the capital stock considered by Ricardo in his calculations – including those of the chapter ‘On Profits’ – is not generally reduced entirely to wage advances. However, when all is said and done, the full significance of the inverse profit– wage nexus can be evaluated, and accordingly Ricardo’s place in the development of 128

A REPLY TO PROFESSOR MOSS

economic thought identified, only if full attention is paid to the instigating force behind his investigations. The relevant consideration in this regard is Ricardo’s objection to received doctrine based upon the Smithian analysis. The charge that Ricardo merely formulated a ‘truism’ will be seen to be without foundation if it is remembered that contemporary thought did not distinguish between those (‘nominal’) wage increases which do not and those (‘real’) wage-rate increases which do involve a change in the distributive shares, and in fact implied that even wage increases falling within the latter category may be passed on by employers in the form of higher prices in the manufacturing sector and lower rents in agriculture. I have traced in detail Ricardo’s odyssey, lasting several years from subscription to the Smithian position (in 1809–10) until his own position was fully formulated (EDR, ch. 3). The demonstration of the fallacy of Smith’s position took various forms. At the simplest level there is the argument that it is logically self-contradictory to maintain that all prices (including the medium) rise simultaneously upon a rise in wages – a truly general wage increase can only affect the structure of prices. If then we assume constancy of the general purchasing power of money, a money wage increase cannot but force down the profit rate since the level of prices is constant. The ‘measure of value’ was, in effect, designed precisely to preclude purely nominal changes in the medium of account, but as shown in my book an equivalent solution was based upon the quantity theory: ‘All commodities cannot rise at the same time without an addition to the quantity of money. This addition could not be obtained at home ... nor could it be imported from abroad. To purchase any additional quantity from abroad, commodities at home must be cheap, not dear. The importation of gold, and a rise in the price of all home-made commodities with which gold is purchased or paid for, are effects absolutely incompatible’ (Ricardo 1951, I, 104–5; EDR,244). To all of this Ricardo added that even were the price level to rise, ‘the proposition would not be less true, which asserts that high wages invariably affect the employers of labour, by depriving them of a portion of their real profits. ... I have endeavoured to show, first, that rise of wages would not raise the price of commodities but would invariably lower profits; and secondly that if the prices of all commodities could be raised, still the effect on profits would be the same, and that in fact the value of the medium only in which prices and profits are estimated would be lowered’ (I, 126–7; EDR, 245). Here we have the allusion to erosion of the real value of profits. These arguments do not in themselves explain the precise adjustment mechanism at work following a wage variation. They operate, as it were, purely on the logical plane. Inverse relation Mark II is required to provide economic content to the argument. Nevertheless, to dismiss the matter as ‘tautological’ and ‘unimportant’ misses the point. Ricardo did not exert such mental effort to establish something self-evident to any layman. 129

RICARDO - THE NEW VIEW

II. Let us turn briefly to the Ricardian ‘legacy’ and the assertion that ‘most (if not all) of the economists of the day who defended Ricardo’s fundamental theorem, defended only the definitional and unimportant version of the theorem’ (Moss 1979, 505). That most economists defended the inverse relationship Mark I is not at issue. It is the downplaying of Ricardo’s influence which I challenge. How does Moss explain Malthus’s explicit insistence upon the ‘importance’ of the theorem despite its ‘apparent obviousness’? ‘Of all the truths which Mr. Ricardo has established, one of the most useful and important is, that profits are determined by the proportion of the whole produce which goes to labour. It is, indeed, a direct corollary from the proposition, that the value of commodities is resolvable into wages and profits; but its simplicity and apparent obviousness do not detract from its utility’ (1963 [1824], 189). How also does he explain Torrens’s initial description of the inverse relationship as ‘equally untenable, whether the terms alteration of wages, alteration of profits are employed with a reference to proportions, or whether they are used in relation to quantities’ (1826, xv–xvi). Some tautology! Or his later concession (after reading Longfield) that since Ricardo ‘explained and explained correctly the laws which regulate rent, wages and profits, under the circumstances which he assumed ’ he ‘rendered important service to the science of Political Economy’ (1844, xxvi) or his description of the device of the ‘ proportions-measuring money ’ in the derivation of the relationship as ‘of great practical importance’ (xxiii). Similarly, I do not see that Moss shows Longfield to have minimized the significance of the relationship. I cannot help noting, incidentally, that Moss himself in his own book on Longfield, at least at one point, seems to take my position on what constitutes the main theme of Ricardian theory: ‘It is ... ironic that, when Torrens did mention Longfield’s name it was to credit him with having removed “the main objection to ... Ricardo’s theory of profit” by explaining in what sense it is to be understood that wages and profits vary inversely. What a curious turn of events that Longfield, who set out to supplant the English theory of wages and profits with a new one of his own, should be hailed in England as one who defended the principal pillar of the Ricardian theory of distribution ’ (1976, 120–1; emphasis added). III. Moss (1979, 507) asks me to explain why Ricardo was so eager to establish that an exogenous increase in the commercial or manufacturing profit rate will be corrected to assure equality with the going rate in agriculture if my argument is correct that the core of his theory is not the determination of the general rate in the agricultural

130

A REPLY TO PROFESSOR MOSS

sector. Surely Moss has given my answer on that same page: if the agricultural rate comes into line with the rate elsewhere the legitimacy of the inverse relationship is at stake since the general profit rate then varies without any alteration in the money wage having occurred. What more is to be said? Moss’s assertion that an innovation in textiles could have been easily accommodated by assuming that the money wage is affected (textiles entering the wage basket) is quite beside the point. If indeed textiles do enter the wage basket and the money wage falls then certainly the general profit rate will rise – this in fact is a central argument of my book, namely that for Ricardo it is immaterial which category of wage goods varies in price. But how would the differential in rates initially presupposed be corrected by a general wage reduction? Of course Ricardo did not take this route! (In any event what if manufactured wage goods are not involved?) I am not at all convinced by Moss’s observation regarding the consistency of my position and that of those who attribute the agricultural profit theory to Ricardo (509). I am fussy about my bedfellows. There are key experiments, the outcome of which vary totally according to the two views. For example, as I show in considerable detail ( EDR, ch. 6), technological improvement in the agricultural sector releases labour and capital for employment in other sectors which are reabsorbed elsewhere with no alteration in their respective returns in consequence of the operation of the law of markets; the price of corn falls to the lower cost level and the return in agriculture (temporarily raised) comes back into line with the given general rate. Thus, despite a change in the ‘margin of cultivation’, the profit rate remains constant. Similarly, freer corn importation leaves the general profit rate unchanged despite a contraction of the domestic margin. The process involves a fall in the price of corn and the transfer of resources to the manufacturing sector with no effect on the general profit rate. These are not the outcomes predicted by the agricultural profit model. Now Ricardo certainly insisted that if the price of luxury goods (silks, velvets, etc.) rises there would be no effect on profits, ‘for nothing can affect profits but a rise in wages; silks and velvets are not consumed by the labourer, and therefore cannot raise wages’ (1951, I, 118). But this is a quite separate analytical issue. Ricardo himself tried to keep the issues separate. Indeed he felt able to proceed in his treatment of the corn-export bounty in terms of an exposition of the fundamental principles at stake involving the (provisional) assumption that agriculture is a constant-cost industry; in this case there is no long-run change in the corn price at all, and thus no change in the ‘money’ wage. The subsequent allowance for rising money wages in consequence of increasing costs, is a separate complication. In a different context Ricardo recognized the possibility that technical change might reduce the cost and price of corn and yet leave ‘money’ wages unaffected – in which case the profit rate remains unchanged (although the commodity wage rises). Similarly, an increase in 131

RICARDO - THE NEW VIEW

the price of corn might leave the money wage unchanged with labourers reducing their consumption of other goods, in which case again the profit rate is unaffected. With such a wide variety of possibilities it is essential not to confuse the effects on the profit rate induced by a change in the margin of agriculture itself – and I have argued there are none – and the effects of a change in the price of corn working upon the general profit rate by way of money wages. It is the attribution to Ricardo of a fixed (real) wage assumption which precludes this essential distinction, and this assumption is the essence of the agricultural model. IV. Ricardian economics, I have argued in EDR, has relevance quite apart from the Corn Laws (as J.S. Mill so well explained) for the inverse relation is at play no matter what the reason for a change in wages. But I do not maintain that the secular behaviour of corn prices was an irrelevant matter for Ricardo; or deny that diminishing returns constituted the key to his growth model. I simply do not believe that model to have been the corn model; nor does one require the corn model to allow a significant role to movements in the price of corn in profit-rate determination. Moss’s quest for ‘a “motive” on Ricardo’s part for holding on to [the agricultural] theory’ (1979, 508) is thus based on a false premise. REFERENCES Hollander, S. (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. Malthus, T.R. (1963) [1824] ‘Political Economy’, in B. Semmel (ed.) Occasional Papers, New York: Burt Franklin. Ottawa, Moss, L.S. (1976) Mountifort Longfield: Ireland’s First Professor of Political Economy, Illinois: Green Hill Publishers. —— (1979) ‘Professor Hollander and Ricardian Economics’,Eastern Economic Journal, December. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol. I: The Principles of Political Economy and Taxation, Cambridge: Cambridge University Press. Torrens, R. (1826) An Essay on the External Corn Trade, 3rd edn, London: Longman, Rees, Orme and Brown. The Budget: On Commercial and Colonial Policy, London: Smith, Elder. —— (1844)

132

Part III RICARDIAN MICRO-ECONOMICS

9 ON THE SUBSTANTIVE IDENTITY OF THE RICARDIAN AND NEO-CLASSICAL CONCEPTIONS OF ECONOMIC ORGANIZATION: THE FRENCH CONNECTION IN BRITISH CLASSICISM*

STATEMENT OF ISSUES My paper is a contribution to the ongoing debate regarding the nature of the neoclassical developments of the 1870s, particularly the legitimacy of the term ‘revolution’, which implies analytical discontinuity, as a valid description of those developments. This representation has become particularly topical, since the notion of a neo-classical or marginalist economics, contrasting sharply in analytical essentials with Ricardian classicism, constitutes a central theme of the historiography of the modern Cambridge (UK) School. The evidence discussed in this paper suggests, on the contrary, how useful in the present context is the notion of altered ‘concentrations of attention’ (Hicks 1976, 208–9), which avoids a revolutionary connotation. For what seems to have occurred in the 1870s was a narrowing of focus, specifically a greater concern with exchange and allocation in their own right; a sharpening of theoretical tools, particularly those relating to consumer choice; and the algebraic formulation of general-equilibrium relationships. These are developments which could have been absorbed by the traditional corpus of analysis, whereas the impatience of the marginalists and their apparent wish to wipe the slate clean meant that much of great import in classical theory for their own chosen and relatively narrow sphere of discourse was not recognized, and spurious analytical distinctions were artificially reinforced. My evidence, in short, suggests how justified was Marshall’s insistence, 135

RICARDO - THE NEW VIEW

against both Jevons and Walras, upon the essential continuity of nineteenth-century doctrine: ‘Under the honest belief that Ricardo and his followers had rendered their account of the causes that determine value hopelessly wrong by omitting to lay stress on the law of satiable wants, [Jevons] led many to think he was correcting great errors; whereas he was really only adding very important explanations’ (Marshall 1920, 101n). Indeed Marshall found Ricardo’s formulation of pricing preferable to that of Jevons, who ‘substitutes a catena of causes for mutual causation’ (818). Gerald Shove’s estimate of four decades ago stands the test of time: the analytical backbone of Marshall’s Principles is nothing more nor less than a completion and generalization, by means of a mathematical apparatus, of Ricardo’s theory of value and distribution as expounded by J.S. Mill. It is not ... a conflation of Ricardian notions with those of the ‘marginal utility’ school. Nor is it an attempt to substitute for Ricardian doctrine a new system of ideas arrived at by a different line of approach. ... [So] far as its strictly analytical content is concerned, the Principles is in the direct line of descent through Mill from Ricardo. (Shove 1960 [1942], 712) A preliminary word on the contrary positions may be helpful, first and foremost that of the marginalists themselves. Distribution was envisaged by Jevons (ideally) as a matter of service pricing ‘entirely subject to the principles of value and the laws of supply and demand’, with input prices ‘the effect and not the cause of the value of the produce’ – ‘I hold labour to be essentially variable, so that its value must be determined by the value of the produce, not the value of the produce by that of labour’ ; and cost of production as a reflection of opportunities forgone (Jevons 1924: xliii f, 186). He accordingly directed his criticisms at the wage fund and subsistence approaches to wage-rate determination and the cost approach to value – as he understood them – paying tribute to the French tradition: ‘the only hope of attaining a true system of Economics is to fling aside, once and for ever after, the mazy and preposterous assumptions of the Ricardian School. Our English Economists have been living in a fool’s paradise. The truth is with the French School’ (xliv–xlv). 1 Jevons recognized elements of the ‘correct’ position in Mill’s Principles – that rent enters into cost where land has alternative uses, that all inequalities (whether natural or artificial) generate economic rents, and the representation of demand and supply as a law ‘anterior’ to costs (xlviii, li, 197) – but could not resist remarking (in the context of the generalization of the rent concept) that ‘those who have studied Mill’s philosophic character as long and minutely as I have done, will not for a moment suppose that the occurance of this section of Mill’s book tends to establish its consistency with other positions in the same treatise’ (li). 136

T H E S U B S T ANT I VE I DE N T I T Y

Walras, whose intellectual origins include par excellence J.B. Say (Schumpeter 1954, 828), similarly objected to the classical pricing and distribution model (as he understood it) – particularly the cost orientation and the natural-wage approach. By neglecting a final demand dimension, and accordingly derived demand, the English had constructed an underdetermined system (Walras 1954, 434–5; cf. Jevons 1924, 269). In more recent times we have the famous criticism of classicism along similar lines by Knight (1956). For Knight, of course, prices depend on the relative subjective appeal to consumers, the flow of goods and thus their marginal utilities governed by cost consideration, where ‘costs’ reflect alternatives surrendered rather than ‘pain’ in the sense of labour or abstinence. On this view the economizing principle involves maximizing the total return from any resource, by equalizing the increments of return at the margin to the scarce resource in alternative uses. It has been suggested (Arrow and Starrett 1973, 132–3) that once the subsistence theory of wages broke down, ‘the most natural alternative was to explain wages by the productivity of labor, an explanation only useful if labor was intrinsically scarce. In short labor had to be treated like land.’ Moreover, recognition of the phenomenon of non-competing groups implied a multiplicity of primary factors which, so it is argued, ‘required a new theory’. The founders of the neo-classical school ‘understood the glaring omission of demand from the classical model’. In his Nobel lecture (1970) Ragnar Frisch neatly stated precisely that reading of the record that I dispute: The classical theory of value – as we find it streamlined in Stuart Mill – was essentially a theory of production costs ... there emerges a sort of gravitational force that pulls prices down. The cost of production is so to speak the solid base on to which the prices fall down and remain .... This theory contains, of course, an irrefutable element of truth. But it is too simple to give even a crude presentation of the forces at play. The economic process is an equilibrium affair where both technological and subjective forces are at play. The subjective element was nearly left out by the classicists. On this point economic theory was completely renewed in the years between 1870 and 1890. ... (Frisch 1981, 5) As remarked above, the theme of a revolutionary break by the general equilibrium economists from classicism is also a feature of modern ‘Cambridge’ historiography. Thus Joan Robinson: either there may be a tendency towards uniformity of wages and the rate of profit in different lines of production [ – the classical position – ] or prices 137

RICARDO - THE NEW VIEW

may be governed by supply and demand, but not both. Where supply and demand rule, there is no room for uniform levels of wages and the rate of profits. The Walrasian system makes sense if we interpret it in terms of an artisan economy, where each producer is committed to a particular product, so that his income depends on his output and its price. Each can have a prospective rate of return on investment in his own line, but there is no mechanism to equalize profits between one line and another. (Robinson 1961, 57) This observation is apparently based on the supposition that among the data of the Walrasian system are included the quantities of each and every specific kind of labour, capital good and land. Following Piero Sraffa, Professor Roncaglia (1982, 341–3) similarly represents the analytical core of neo-classicism as ‘the model of pure exchange’, whereby perfect competition guides us to the optimal allocation of scarce resources. Prices are ‘indexes of resource scarcity relative to wants; income distribution comes out as a by-product of price-determination, distributive variables being but the prices for the services of the so-called “factors of production”. Production processes are only an intermediate stage connecting final consumers’ tastes to the initial scarce resource endowments.’ By contrast, ‘classicism’ is represented as a reproductive process (involving the ‘production of commodities by means of commodities’) wherein ‘at the beginning of the production period, specific quantities of commodities are advanced, as means of production or as subsistence for the workers employed’ – both technology (including the structure of production) and wages are exogenously determined – the utilization of which yields outputs exceeding the initial stocks, a surplus ‘consisting of a heterogeneous set of commodities’. In this system ‘relative prices must be such as to allow all sectors a profit inducement to repeat the production sequence’; the spread between product prices and costs must generate a uniform rate of profit in all sectors, the average profit rate itself being determined solely by the exogenously given wage rate and technology. 2 It is an essential part of the foregoing argument that in the classical system value and output levels are not determined simultaneously by the forces of demand and supply. 3 This separation of value and output precludes the possibility that a change in the pattern of consumer demand can influence factor returns, and thus costs, by playing upon the relative scarcity of the factors: the divorce of value and output implies a divorce of value and distribution (cf. Dobb 1973, 261; Garegnani 1972, 278f; 1976, 24–45; Pasinetti 1974, 43–4; Roncaglia 1978, 119f). All this is in contrast to Walrasian theory. Indeed, the paradigmatic contrasts have led to the charge that Walras was seriously inconsistent for conceiving the capital endowment of the community as a set of given quantities of ‘capital-goods 138

T H E S U B S T ANT I VE I DE N T I T Y

proper’ yet also (in parts of his Elements ) adopting the notion of uniformity of profit rates (Garegnani 1976, 34, 36). It is my contention that Ricardian economics – the economics of Ricardo and J.S. Mill – in fact comprises in its essentials an exchange system fully consistent with the marginalist elaborations. In particular, their cost-price analysis is pre-eminently an analysis of the allocation of scarce resources, proceeding in terms of general equilibrium, with allowance for final demand, and the interdependence of factor and commodity markets. 4 Serious and long-lived misconceptions regarding classicism flow from a failure to recognize that the classical notions of wages and interest as compensation for effort and abstinence were pertinent only at the macro-economic level where the determinants of aggregate factor supplies are under investigation and not in the micro-economic context where costs referred to forgone opportunities. 5 My perspective places J.S. Mill directly in the Ricardian theoretical tradition. That we find simultaneously in his Principles both ‘neo-classical’ and ‘Ricardian’ features implies neither inconsistency (Hollander 1976) – or no more inconsistency than in Ricardo himself – nor a process of escape, or attempted escape, from his Ricardian heritage, a view expressed recently in the following terms: A silent revolution in the direction of the marginalist supply-and-demand theory was brought about [by Marshall] in the course of adopting, extending and transforming some ideas in Mill. As Mill himself had departed considerably from Ricardo, Marshall was thus moving even further from the Ricardian source. (Bharadwaj 1978, 254) It was precisely the beginnings in Mill of considerable deviations from Ricardo’s theory of distribution that called for and received at Marshall’s hands ... extensions and refinement, so that Marshall’s deliberations on value and distribution departed systematically from questions Ricardo posed and the framework of analysis he employed. ... What Shove regarded as extensions and generalisations of Ricardo in [Marshall’s] Principles (the introduction of the demand side, the functional relation between costs and output, the supply and demand determination of wages and profits) are radical departures from the Ricardian standpoint. (Bharadwaj 1978, 269) 6 My perspective is one that avoids the difficult psychological problems posed by interpretations that refuse to accept at face value Marshall’s statements of his relationship with his classical forebears, or those of Mill regarding his intellectual relationship with Ricardo – his repeated insistence that he was elaborating upon Ricardian themes. The 139

RICARDO - THE NEW VIEW

demand side, the functional relation between cost and output, and the supply and demand determination of wages and profits, far from being ‘radical departures’ from Ricardianism, are central to that doctrine without which neither the cost theory of price nor the inverse wage–profit relation can be understood. 7 The second and third sections of this paper will demonstrate the key role accorded by Ricardo and J.S. Mill to opportunity cost and derived factor demand: their simultaneous and consistent attachment to cost theories of value and to the generalequilibrium conception of economic organization as formulated by J.B. Say and much admired by Walras. Such demonstration clearly has important implications for the nature of the ‘neo-classical’ developments of the 1870s. But we must also consider the reverse side of the coin, from which perspective it again becomes clear that the term ‘revolution’ to describe that doctrine is unhelpful. In his criticisms of Ricardo, Walras wrote that it is ‘the price of the products which determines the price of productive services’ (1954, 425). Similarly, he praised Jevons’s statement of the ideal procedure according to which ‘the formula of the English school, in any case the school of Ricardo and Mill, must be reversed, for the prices of productive services are determined by the prices of the products, and not the other way round’ (45). This clearly does not constitute a picture of mutual interdependence between factor and product markets. It is in fact a statement that emphasizes what the classics had supposedly omitted, and does so by implicitly adopting a short-run perspective. My fourth section is devoted to a demonstration that Walras accepted the ‘classical’ conception of long-run cost prices – ‘costs’ incorporating profits at a uniform rate on the supply prices of capital goods – employing the Ricardian or Marshallian adjustment mechanism of output response to deviations between demand and supply prices. It also becomes clear that the charge of inconsistency for so doing is unfounded; he insisted upon profit-rate uniformity, as Ricardo had done and as Marshall was to do, only when allowance is made for changes in the outputs of the different types of capital goods. (The same applies to labour.) Walras adhered to classical cost-price analysis given the appropriate long-run assumptions and, like Ricardo and Marshall, distinguished between maximizing decisions regarding new investments and the actual return on capital goods once constructed. The fifth section will draw the threads of our analysis together. Brief consideration will then be given to the sources of some of the erroneous views regarding classicism described above. RICARDO ON ECONOMIC ORGANIZATION: THE SAY TRADITION We set out with J.B. Say’s well-known statement in the Traité d’économie politique of mutual interdependence between product and factor markets incorporating the 140

T H E S U B S T ANT I VE I DE N T I T Y

principles of opportunity cost and of imputing the values of factors from the values of their products – in broad terms only because of the absence of a marginal conception whereby the physical contributions of individual factors can be isolated: It is utility which determines the demand for a commodity, but it is the cost of its production which limits the extent of its demand. When its utility does not elevate its value to the level of the cost of production, the thing is not worth what it cost; it is a proof that the productive services might be employed to create a commodity of a superior value. The possessors of productive funds, that is to say, those who have the disposal of labour, of capital or land, are perpetually occupied in comparing the cost of production with the value of the things produced, or which comes to the same thing, in comparing the value of different commodities with each other; because the cost of production is nothing else but the value of productive services consumed in forming a production; and the value of a productive service is nothing else than the value of the commodity, which is the result. The value of a commodity, the value of a productive service, the value of the cost of production are all, then, similar values when every thing is left to its natural course. (Say; cited in Ricardo 1951, I, 282–3) Now Walras certainly believed Say to have been on the right road by this formulation of general interdependency (1954, 425). But so did Ricardo, who commented on the passage: ‘M. Say maintains with scarcely any variation, the doctrine which I hold concerning value.’ His sole complaint related to Say’s treatment of the services of land on a par with those of capital and labour, in the light of his own (implied) presumption of one-use land (to be elaborated presently) whereby rent is excluded from (marginal) cost (1951, I, 283–4). 8 Ricardo’s subscription to Say’s position cannot easily be appreciated in terms of those interpretations that envisage a sharp divergence between the ‘British’ and the ‘French’ traditions. Yet the notion of opportunity cost pervades Ricardo’s work. Indeed his cost prices make no sense except in these terms. To this matter we now turn. It is inviting to identify Ricardian cost price with labour embodied, as Malthus in fact did, but it would be incorrect to do so: It is necessary for me also to remark, that I have not said, because one commodity has so much labour bestowed upon it as will cost 1000 l. and another so much as will cost 2000 l. that therefore one would be of the value of 1000 l. and the other of the value of 2000 l. but I have said that their value will be to each other as two to one, and that in those proportions they will be exchanged. (Ricardo 1951, I, 46–7) 141

RICARDO - THE NEW VIEW

Mr. M. ... misunderstands me ... I say its whole value will be in proportion to a portion of its cost, and I do not say this without allowing for modifications and exceptions – though I consider these of no great magnitude. I have said that the relative value of commodities is in proportion to the quantity of labour bestowed on them. That value may be double what the labour cost. (Ricardo 1951, II, 100–2) ‘Cost of production’ or natural price thus includes profits as well as wages each at its average or ordinary rate as Smith had explained (I, 291). Under the appropriate technological conditions defined in the first chapter of the Principles (namely uniform factor proportions) a state of general equilibrium, such that prices reflect costs throughout the system, will be one satisfying the principles of profit rate (and wage rate) equalization and proportionality of prices to labour inputs. More accurately, under the stated circumstances uniformity of profit rates (and wage rates) require that proportionality. The following passage (drawn from a discussion of subsidized labour for some firms in a manufacturing industry) beautifully summarizes the point, and does so in a context expressing that what is relevant is marginal labour input: ‘The manufacturer enjoying none of these facilities might indeed be driven altogether from the market, if the supply afforded by these favoured workmen were equal to all the wants of the community; but if he continued in the trade, it would be only on condition that he should derive from it the usual and general rate of profits on stock, and that could only happen when his commodity sold for a price proportional to the quantity of labour bestowed on its production ’ (I, 73n; emphasis added). Now it is the possibility of capital (and labour) movement between uses or commodity-supply adjustment that assures the tendency to cost price and proportionality to labour input – a matter of great importance that is apparently denied by Cambridge writers and others. This can be illustrated from the discussion of an exogenous change in tastes: Let us suppose that all commodities are at their natural price, and consequently that the profits of capital in all employments are exactly at the same rate. ... Suppose now that a change of fashion should increase the demand for silks, and lessen that of woollens; their natural price, the quantity of labour necessary for their production, would continue unaltered, but the market price of silks would rise, and that of woollens would fall; and consequently the profits of the silk manufacturer would be above, whilst those of the woollen manufacturer would be below, the general and adjusted rate of profits. Not only the profits, but the wages of the workmen, would be affected in these employments. This increased demand for silks would however soon be supplied, by the 142

T H E S U B S T ANT I VE I DE N T I T Y

transference of capital and labour from the woollen to the silk manufacture ; when the market prices of silks and woollens would again approach their natural prices, and then the usual profits would be obtained by the respective manufacturers of those commodities. It is then the desire, which every capitalist has, of diverting his funds from a less to a more profitable employment, that prevents the market price of commodities from continuing for any length of time either much above, or much below their natural price. It is this competition, which so adjusts the exchangeable value of commodities, that after paying the wages necessary to their production, and all other expences required to put the capital employed in its original state of efficiency, the remaining value or overplus will in each trade be in proportion to the value of the capital employed. (Ricardo 1951, I, 90–1; emphasis added) In circumstances of differential factor ratios the same assumption of factor mobility dictates a divergence of (relative) cost prices from (relative) labour inputs as Ricardo explained at length in his first chapter. But the entire notion of cost price presumes factors that have alternative uses; and in all cases, whether or not costs are proportional to labour inputs, only those returns that reflect alternative opportunities are allowed for in costs. Embodiment of labour or for that matter the pain cost attached to labour and abstinence are not in themselves the relevant consideration, 9 as is clear from the fact that since ‘it is through the inequality of profits that capital is moved from one employement to another’ (I, 119), an economy-wide change in labour productivity or the wage rate or any other disturbance will leave cost prices unaltered should all commodities be impinged upon equally. The implications of this perspective are legion. For example, from the context of public finance: From this circumstance [the differential impact on agriculture of the poor rates] it follows, that the farmer will be enabled to raise the price of his produce by the whole difference. For since the tax falls unequally, and peculiarly on his profits, he would have less motive to devote his capital to the land, than to employ it in some other trade, were not the price of raw produce raised. If on the contrary, the rate had fallen with greater weight on the manufacturer than on the farmer he would have been enabled to raise the price of his goods by the amount of the difference, for the same reason that the farmer under similar circumstances could raise the price of raw produce ...; for there can be no reason why their profits should be 143

RICARDO - THE NEW VIEW

reduced below the general rate of profits, when their capitals might be easily removed to agriculture. (Ricardo 1951, I, 260–1) The conception in question bears strategically upon the nature of trade: ‘If any cause should raise the price of a few manufactured commodities, it would prevent or check their exportation; but if the same cause operated generally on all, the effect would be merely nominal, and would neither interfere with their relative [cost] value, nor in any degree diminish the stimulus to a trade of barter, which all commerce, both foreign and domestic really is’ (I, 228). And the principle that cost prices reflect alternative opportunities provides the rationale for the fundamental theorem of distribution itself – the inverse wage–profit relation. A general wage increase either leaves cost prices entirely unaffected, thus forcing down profits (the special case of identical factor proportions), or, by disturbing the structure of profit rates, sets in motion appropriate supply adjustments that lead to the establishment of a new cost structure, assuring again uniform profit rates in all industries albeit at a lower general level (Hollander 1979, 302f). One also should not forget that although disturbances to the wage structure were largely set aside by Ricardo, the rationale for this procedure was reliance upon the operation of ‘competition’ – the demand–supply mechanism – which assured a pattern of relativities reflecting ‘the comparative skill of the labourer, and intensity of the labour performed’ (1951, I, 20). It will be apparent that Ricardo’s cost prices make no sense except within a demand–supply framework allowing for alternative uses of resources. Yet Ricardo is frequently said to have rejected demand–supply analysis (except perhaps for market price determination). This is a misconception. What he actually complained of was ‘the opinion that the price of commodities depends solely on the proportion of supply to demand or demand to supply’ (I, 382; emphasis added), a complaint alluding to those formulations that appeared to exclude a role for cost conditions in the mechanism. Thus his observation to Say: ‘You say demand and supply regulates the price of bread; that is true, but what regulates supply? the cost of production’ (IX, 172). Indeed, Ricardo’s point (as is apparent from the following reaction to the treatment by Say of a commodity tax) was precisely that Say had failed to follow out the logic of his own approach to pricing which runs in terms of alternative uses: It is observed by M. Say, ‘that a manufacturer is not enabled to make the consumer pay the whole tax levied on his commodity, because its increased price will diminish its consumption’. Should this be the case, should the consumption be diminished, will not the supply also speedily be diminished? Why should the manufacturer continue in the trade if his profits are below 144

T H E S U B S T ANT I VE I DE N T I T Y

the general level? M. Say appears here also to have forgotten the doctrine which he elsewhere supports, ‘that the cost of production determines the price, below which commodities cannot fall for any length of time, because production would be then either suspended or diminished’. (Ricardo 1951, I, 243n) The essence of the matter is captured exquisitely in a letter to Malthus: ‘You say demand and supply regulates value – this I think is saying nothing ... – it is supply which regulates value – and supply is itself controlled by comparative cost of production’ (VIII, 279; emphasis added). On matters of principle Ricardo’s line is that of Lausanne: equality of wage rates and of profit rates maximizes the return to the factors ‘capital’ and ‘labour’. 10 Moreover, it must be emphatically stated that these average returns themselves are, in principle, variables governed by the relative scarcity of the factors: ‘I have invariably insisted, that high or low profits depend on low and high wages, how then can it be justly said of me that the only cause which I have recognized of high or low profits is the facility or difficulty of providing food for the labourer[?] I contend that I have also recognized the other cause, the relative amount of population to capital, which is another of the great regulators of wages’ (II, 264–5). 11 On this view, and keeping in mind Ricardo’s further insistence that ‘the power of employing labour depends on the increase of a particular part of capital, not on the increase of the whole capital’ (IX, 127), the way is open for an allowance that the adjustment process following a variation in the pattern of final demand itself will affect the average factor returns. In short, output levels can affect relative cost prices by playing upon the relative scarcity of labour and capital, an outcome in line with neo-classical theorizing. We can sharpen our understanding of Ricardo’s position on the nature of cost price by considering the contrast between costs and rents. I shall first establish Ricardo’s awareness that the phenomenon of differential rent, which plays so large a part in his system, is but a special case of a more general phenomenon – land scarcity. Rent is provisionally defined as the payment to the landlord ‘for the use of the original and indestructible powers of the soil’ (I, 67) – a productivity phenomenon. But in the following passage the ultimate rationale is more specifically expressed in terms of productivity and scarcity (demand and supply): On the first settling of a country, in which there is an abundance of rich and fertile land, a very small proportion of which is required to be cultivated for the support of the actual population or indeed can be cultivated with the capital which the population can command, there will be no rent; for no 145

RICARDO - THE NEW VIEW

one would pay for the use of land, when there was an abundant quantity not yet appropriated, and, therefore, at the disposal of whosoever might choose to cultivate it. On the common principles of supply and demand, no rent could be paid for such land, for the reason stated why nothing is given for the use of air and water, or for any other of the gifts of nature which exist in boundless quantity. With a given quantity of materials, and with the assistance of the pressure of the atmosphere, and the elasticity of steam, engines may perform work and abridge human labour to a very great extent; but no charge is made for the use of these natural aids, because they are inexaustible, and at every man’s disposal. In the same manner the brewer, the distiller, the dyer, make incessant use of the air and water for the production of their commodities; but as the supply is boundless, they bear no price. (Ricardo 1951, I, 69) Here J.B. Say is cited to the same effect: ‘The earth ... is not the only agent of nature which has a productive power; but it is the only one, or nearly so, that one set of men take to themselves, to the exclusion of others; and of which, consequently, they can appropriate the benefits.’ Using these principles Ricardo rejected physiocratic residues in the Wealth of Nations, specifically the notion that in manufactures ‘nature does nothing, man does all; and the reproduction must always be in proportion to the strength of the agents that occasion It’. Factor productivity, runs Ricardo’s argument, is a necessary but insufficient condition for a positive return: Does nature nothing for man in manufactures? Are the powers of wind and water, which move our machinery, and assist navigation, nothing? The pressure of the atmosphere and the elasticity of steam, which enable us to work the most stupendous engines – are they not the gifts of nature? to say nothing of the effects of the matter of heat in softening and melting metals, of the decomposition of the atmosphere in the process of dyeing and fermentation. There is not a manufacture which can be mentioned, in which nature does not give her assistance to man, and give it too generously and gratuitously. (Ricardo 1951, I, 76n) Differential rent (reflecting productivity differentials) is simply a special case of the genus, scarcity rent: ‘If all land had the same properties, if it were all unlimited in quantity, and uniform in quality, no charge could be made for its use, unless it possesses peculiar advantages of situation’ (I, 69). Similarly: ‘[if] air, water, the elasticity of steam, and the pressure of the atmosphere, were of various qualities; if they could be appropriated, and each quality existed only in moderate abundance, they, as well 146

T H E S U B S T ANT I VE I DE N T I T Y

as the land, would afford a rent, as the successive qualities were brought into use’ (I, 75). Differential rent is not to be understood as falling outside the general demand– supply framework. Secondly, Ricardo stated clearly the conditions under which rent would appear even on marginal units of output (absolute rent) – namely, where scarcity manifests itself in an extreme form, the supply curve of agricultural produce becoming, as it were, vertical (the functional relation between output and costs terminating); and he traced through some of the analytical consequences: The corn and raw produce of a country may, indeed, for a time sell at a monopoly price; but they can do so permanently only when no more capital can be profitably employed on the lands, and when, therefore, their produce cannot be increased. At such time, every portion of land in cultivation, and every portion of capital employed on the land will yield a rent, differing, indeed, in proportion to the difference in the return. At such time too, any tax which may be imposed on the farmer, will fall on rent, not on the consumer. He cannot raise the price of his corn, because by the supposition, it is already at the highest price at which the purchasers will or can buy it. He will not be satisfied with a lower rate of profits, than that obtained by other capitalists, and, therefore, his only alternative will be to obtain a reduction of rent or to quit his employment. (Ricardo 1951, I, 250–1)12 I spell this out in order to emphasize Ricardo’s comprehension of the general principle of scarcity price both where land differentials exist and where they do not; he went along with Buchanan’s statement that ‘rent is the effect of high price. ... It is ... from the price which the produce is sold, that the rent is derived; and this price is got not because nature assists in the production, but because it is the price which suits the consumption to the supply’ (I, 77n). To my knowledge Ricardo nowhere explicitly states that rent cannot be excluded from cost, notwithstanding the fact that the aggregate supply conditions of land differ from those of capital and labour, in the case of multi-use land. Yet I believe that he appreciated the logic. Consider in particular the generalization of the rent concept from land to capital – the very explicit recognition that once the assumption of capital mobility between uses is abandoned it is the rent analysis that becomes appropriate: As a part of this capital, when once expended in the improvement of a farm, is inseparably amalgamated with the land, and tends to increase its productive powers, the remuneration paid to the landlord for its use is strictly of the nature of rent, and is subject to all the laws of rent. Whether the improvement 147

RICARDO - THE NEW VIEW

be made at the expense of the landlord or the tenant, it will not be undertaken in the first instance, unless there is a strong probability that the return will at least be equal to the profit that can be made by the disposition of any other equal capital, but when once made, the return obtained will ever after be wholly of the nature of rent, and will be subject to all the variations of rent. (Ricardo 1951, I, 262) This principle was in fact utilized in an argument against those who opposed freer corn importation on grounds of capital immobility: ‘Suppose’, Ricardo reasoned, ‘that the fact be as stated and that no part of the capital could be withdrawn; the farmer would continue to raise corn, and precisely the same quantity too, at whatever price it might sell; for it could not be his interest to produce less, and if he did not so employ his capital, he would obtain from it no return whatever’ (I, 269). There are also the similar effects of a tax on rent (I, 257) and a tax on profits-ingeneral – in neither case (even allowing capital mobility) can the tax be escaped by way of allocative readjustments. Of course, a fall in the (net) profit rate may subsequently play on the rate of capital accumulation and thus the wage rate, but this is a matter relating to long-run aggregate supply conditions. Also relevant is the recognition in the context of foreign trade that once the possibility of resource mobility between uses is ruled out, the general rules of cost price break down. Even assuming uniform capital–labour ratios, commodities will no longer exchange in proportion to relative labour inputs: ‘The difference in this respect, between a single country and many, is easily accounted for, by considering the difficulty with which capital moves from one country to another, to seek a more profitable employment, and the activity with which it invariably passes from one province to another in the same country’ (I, 135–6). We return now to the passage cited above from J.B. Say, which Ricardo applauded: the doctrine of productive services. This passage explicitly spells out the principle of opportunity costs and the closely related argument that the source of factor returns (and the motive for the use of factors) in any use is final demand. While the final demand dimension leaps to the eye in the case of (single-use) land, it is no less pertinent in the case of those returns that enter into cost price. 13 For to refer to cost price is merely to say that demand in any sector is sufficient to meet the competition of demand elsewhere for the use of resources. There should be no surprise at Ricardo’s acceptance of the Say formulation. What, however, are we to make of Ricardo’s subscription to the wages fund theory and its corollary, as expressed by Mill’s fourth proposition on capital, that ‘demand for commodities is not demand for labour’? For it was Jevons’s complaint 148

T H E S U B S T ANT I VE I DE N T I T Y

precisely that according to this proposition (which he rightly observed originated with Ricardo), capitalists ‘maintain and pay for labour whether or not there is a demand for the commodities produced’ and ‘production goes on independently of the use to which the produce is to be put’ (Jevons 1905, 127). This complaint is unjustified. Nothing in the fourth proposition conflicts with derived demand, since it relates in no way to the individual capitalist’s motivation in offering employment. It is a description of the manner in which the aggregate demand for labour and thus the aggregate wage bill (reflecting either higher average earnings or higher employment or both) is expanded. Thus it is that for the reabsorption of labour displaced by machinery Ricardo relied in part upon increased demand for service labour out of net revenue and in part upon net accumulation. That a transfer by capitalists from consumption expenditure to investment raises labour demand is easily demonstrable (Hollander 1979, 326f). That the same holds true of a transfer from consumption to expenditure on services is clarified in Ricardo’s chapter ‘On Machinery’. The altered pattern of consumption from commodities to services encourages an expansion of the agricultural sector; labourers displaced in the consumer-goods sector are not, as it were, simply reabsorbed in the service sector but are reabsorbed in expanding agricultural production to meet the consumption requirements of the (additional) service labour, a sequence of events corresponding to that entailed by ‘savings from revenue to add to capital’ (Hollander 1979, 373f). Similarly, it can be shown that Ricardo’s discussion of the demand for labour in no way precludes an approach to economic organization in terms of ‘synchronized’ activity as some writers during the 1870s believed. 14 Ricardo firmly rejected the notion of a literal pre-accumulation of stocks of wage goods advanced by employers and constituting the demand for labour. Workers, he insisted, are paid in money that is disbursed by them directly at retail in the manner of all consumers, the quantity and character of the commodities produced reflecting that demand: ‘I dispute your position’, he wrote to Malthus, ‘that a demand for labour is the same thing as a supply of necessaries’ (1951, VIII, 258). What was involved, rather, was a direction of activity towards wage-goods production in appropriate response to the volume and pattern of labourers’ consumption – the volume alone governed by capitalists’ savings decisions. J.S. MILL ON ECONOMIC ORGANIZATION On all these matters J.S. Mill was entirely at one with Ricardo. The theory of costs was treated from a micro-economic perspective involving relative value and the motives underlying resource allocation; indeed, his appreciation of the relativity of exchange value is nowhere expressed more clearly than in the context of costs of 149

RICARDO - THE NEW VIEW

production. ‘Value is a relative term’, he wrote in his chapter on the ‘Ultimate Analysis of Cost of Production’, ‘not a name for an inherent and substantive quality of the thing itself’ (Mill 1965, 479). Accordingly, he defended the emphasis upon labour in Ricardo’s treatment of value – despite the fact that the primary costs to be met by the capitalist–employer are wage costs – on the grounds that ‘in considering ... the causes of variations in value, quantity of labour is the thing of chief importance, for when that varies, it is generally in one or a few commodities at a time, but the variations of wages (except passing fluctuations) are usually general, and have no considerable effect on value’ (481). None the less, wage differentials will be reflected in the price structure, as well as relative labour inputs, and changes in wage differentials will generate changes in the price structure: ‘Although, however, general wages, whether high or low, do not affect values, yet if wages are higher in one employment than another, or if they rise and fall permanently in one employment without doing so in others, these inequalities do really operate on values’ (480; also 692). The same principles applied to profits: ‘Values ... being purely relative, cannot depend upon absolute profits, no more than upon absolute wages, but upon relative profits only. ... Insofar as profits enter into the cost of production of all things [equally] they cannot affect the value of any. It is only by entering in a greater degree into the cost of production of some things than of others, that they can have any influence on value’ (482). By this latter allowance Mill had in mind more than the consequence for the price structure of profit-rate differentials reflecting unequal risk and so forth. The allowance covered compensation for differential time periods of production from industry to industry: ‘one commodity may be called upon to yield profit during a longer period than the other’. In consequence of differential factor proportions it followed that ‘commodities do not exchange in the ratio simply of the quantities of labour required to produce them’ (484), and this quite apart from the complication created by partial wage changes. Even general wage changes might influence the structure of prices: ‘even a general rise of wages, when it involves a real increase in the cost of labour, does in some degree influence values. It does not affect them in the manner vulgarly supposed, by raising them universally. But an increase in the cost of labour lowers profits; and therefore lowers in natural value the things into which profits enter in a greater proportion than the average, and raises those into which they enter in a less proportion than the average’ (485). We must here have in mind Mill’s repeated insistence, always following Ricardo, that changes in production costs exert their influence by way of supply variation. Let us recall the opinion, expressed by Malthus in 1824 (and repeated ever since), that Ricardo had limited demand–supply analysis solely to the market period and cases of monopoly, treating long-run cost price quite independently; the two theories were mutually exclusive (Malthus 1963, 181–2). Mill rejected Malthus’s attribution as 150

T H E S U B S T ANT I VE I DE N T I T Y

soon as it appeared, insisting that in the opinion of the Ricardo school long-run cost prices were arrived at by way of supply variation (Mill 1967, 33–4). In the chapter in the Principles, ‘Of Demand and Supply, in their Relation to Value’, reference is indeed made to ‘another law [than that of demand and supply] for that much larger class of things, which admit of indefinite multiplication’; but immediately thereafter we find a caution that in dealing with production costs, ‘it is not less necessary to conceive distinctly and grasp firmly the theory of this exceptional case’ (that of given supplies), which ‘will be found to be of great assistance in rendering the more common case intelligible’ (Mill 1965, 468). The point is clear enough: ‘The value at any particular time is the result of supply and demand’, but ‘unless that value is sufficient to repay the Cost of Production, and to afford, besides, the ordinary expectation of profit, the commodity will not continue to be produced’. Necessary price, in brief, includes a return on capital ‘as great ... as can be hoped for in any other occupation at that time and place’, and in the event of a return in excess of the going rate ‘capital rushes to share in this extra gain, and by increasing the supply of the article, reduces its value’; conversely, in the reverse case output is restricted (471–2). And it is in this sense that one may easily appreciate the famous reference to ‘a law of value anterior to cost of production, and more fundamental, the law of demand and supply’ (583). Here Mill did not intend to deny, any more than Ricardo, that cost of production works its influence by way of supply variations. His point was that demand–supply analysis applied to all cases, even where cost analysis was irrelevant – an appropriate perspective in a chapter dealing with ‘Some Peculiar Cases of Value’. The general conclusion regarding the central role of supply variation in the establishment of cost price is scarcely surprising in the light of Mill’s observation that the pertinent perspective in cost-price analysis is one involving ‘the motives by which the exchange of commodities against one another is immediately determined’. 15 Numerous explicit allusions to relative supply variation will be found precisely in this context; the higher value of commodities produced in industries entailing a relatively high degree of unpleasantness or a relatively longer period of production, or a differential tax are all explained in terms of appropriate supply restraints (482f). It is indeed in the ‘general-equilibrium’ context that the dependency of cost price upon demand–supply can be seen most comprehensively. Cost of production, we have already observed, includes normal profit; the equilibrium price structure will thus be one that yields the going return on capital in all sectors. Mill described at great length (407) the institutional arrangements whereby returns, or expectation of returns, are equalized (‘the method of accommodating production to demand’) leaving no doubt of the conspicuous role accorded supply variation in the establishment of equilibrium prices, following exactly the lines laid down in Ricardo’s Principles. 151

RICARDO - THE NEW VIEW

Also in line with Ricardo, Mill used the principles of allocation theory in the rationalization of the inverse wage–profit relation. Thus, in contrast to a wage increase affecting one sector, where price will rise to assure equality of profit rates across the board, there exists no mechanism whereby general prices would be forced upwards in the event of an economy-wide wage increase, when all firms throughout the system are affected equally by the change. From this perspective the fundamental theorem on distribution is founded squarely upon the theory of allocation: a general rise of wages would not raise prices but would be taken out of the profits of the employers; always supposing that those profits were sufficient to bear the reduction. The case is different with a rise of wages confined to a single or a small number of employments. That rise if taken out of profits, would place a particular class of employers at a disadvantage compared with other employers: & as soon as they ceased to hope that the loss would be only temporary, they would withdraw part of their capital, or at all events, all new capital would avoid those trades & go into others. Consequently the supply of these particular articles would fall short, and their prices would rise so as to indemnify the employers for the rise of wages. But this would not happen in case of a rise of all wages, for as all capitalists would be affected nearly alike they could not as a body relieve themselves by turning their capital into another employment. (Mill 1972, 1735) Expenses which affect all commodities equally, have no influence on prices. If the maker of broadcloth or cutlery, and nobody else, had to pay higher wages, the price of his commodity would rise, just as it would if he had to employ more labour; because otherwise he would gain less profit than other producers, and nobody would engage in the employment. But if everybody has to pay higher wages, or everybody to employ more labour, the loss must be submitted to; as it affects everybody alike, no one can hope to get rid of it by a change of employment, each therefore resigns himself to a diminution of profits and prices remain as they were. In like manner, general low wages, or a general increase in the productiveness of labour, does not make prices low, but profits high. If wages fall, (meaning here by wages the cost of labour), why, on that account, should the producer lower his price? He will be forced, it may be said, by the competition of other capitalists who will crowd into his employment. But other capitalists are also paying lower wages, and by entering into competition with him they would gain nothing but what they are gaining already. (Mill 1965, 692) 152

T H E S U B S T ANT I VE I DE N T I T Y

A statement of this position will also be found in the context of Mill’s discussion of the nature of costs when focusing upon the distinction between costs from an aggregative and industry perspective. It is indeed precisely in the course of expounding the principles of allocation theory that Mill insisted on the inverse wage–profit relation: ‘There is no mode in which capitalists can compensate themselves for a high cost of labour, through any action, on values or prices. It cannot be prevented from taking its place on low profits’ (479). It is significant, too, that in his correspondence with Cairnes regarding the nature of costs Mill immediately saw the implications of the argument for the Ricardian inverse wage–profit relationship: Your discussion of the question whether wages ought in any sense to be considered as cost of production, or whether that term should be exclusively predicated of labour and abstinence, was always likely to be scientifically instructive, but I now perceive that it will have a special value de circonstance. You must have been struck as I have been, by the thoroughly confused and erroneous ideas respecting the relation of wages to prices, which have shewn themselves to be almost universal in the discussions of the recent strikes. The notion that a general rise of wages must produce a general rise of prices, is preached universally. ... Certainly no one who knows, even imperfectly, what the Ricardo political economy is ... can suppose this to be it. (Mill 1972, 1909–10) The appreciation of the equilibrating function of price, the extension of demand– supply analysis from the ‘market’ to the ‘long-run’ period involving the process of profit-rate equalization, and the application of these principles to the basic theorem of distribution were all part and parcel of Ricardian analysis. In some respects, doubtless, Mill’s formulations constituted an improvement in rigour – particularly the formal statement of an equation of demand and supply and the distinction between displacements of the demand schedules and movements from one position to another on the same schedule (1965, 466). But their merit lies less in substantive content than in their location at a conspicuous juncture among the basic theoretical principles; for Ricardo made many of his statements regarding the theory of allocation in various informal contexts relating to applied problems. No good purpose is served, however, by invidious comparisons regarding ‘quality’ of analysis. And that is not my concern. The point is simply that Mill’s theory of allocation does not constitute a breakaway 16 from Ricardian doctrine, but is a reiteration thereof. I follow the pattern of the previous section by considering Mill’s approach to rent. In the aggregate, rent (‘the price paid for the use of an appropriated natural agent’) differed from the other factor returns in consequence of the conditions of land supply. The agent is ‘as indispensable (and even more so) as any implement: but the 153

RICARDO - THE NEW VIEW

having to pay a price for it, is not. In the case of the implement (a thing produced by labour) a price of some sort is the necessary condition for its existence: but the land exists by nature. The payment for it, therefore, is not one of the expenses of production’ (Mill 1965, 58). Allowing for differentials complicated the issue only slightly – ‘the real expenses of production are those incurred on the worst land, or by the capital employed in the least favourable circumstances’ (429). It should be noted also that Mill, following Ricardo, realized that differential rent entails a special case of scarcity value, and that rent might be generated even in the absence of differentials in the event of an absolute constraint on farm output: ‘It is also distinctly a portion of Ricardo’s doctrine, that ... the land of a country supposed to be of uniform fertility would, all of it, on a certain supposition, pay rent: namely, if the demand of the community required that it should all be cultivated, and cultivated beyond the point at which a further application of capital begins to be attended with a smaller proportionate return’ (428). 17 When Mill focused upon individual sectors – and here Mill made explicit what was only implicit in Ricardo – the picture is a very different one: The question ... respecting the influence which the appropriation of natural agents produces on values, is often stated in this form: Does Rent enter into Cost of Production? and the answer of the best political economists is in the negative. The temptation is strong to the adoption of these sweeping expressions, even by those who are aware of the restrictions with which they must be taken, for there is no denying that they stamp a general principle more firmly on the mind, than if they were hedged round in theory with all its practical limitations. But they also puzzle and mislead, and create an impression unfavourable to political economy, as if it disregarded the evidence of facts. No one can deny that rent sometimes enters into cost of production. If I buy or rent a piece of ground, and build a cloth manufactory on it, the ground-rent forms legitimately a part of my expenses of production, which must be repaid by the product. And since all factories are built on ground, and most of them in places where ground is peculiarly valuable, the rent for it must, on the average be compensated in the values of all things made in factories. (Mill 1965, 487) The consequence of multi-use land for cost pricing is laid down clearly in the ‘summary of the theory of value’, namely, that ‘when land capable of yielding rent in agriculture is applied to some other purpose, the rent which it would have yielded is an element in the cost of production of the commodity which it is employed to produce’ (498; cf. 484). 154

T H E S U B S T ANT I VE I DE N T I T Y

Consistent with Mill’s ‘Ricardian’ approach to cost price is the Say conception of organization, which emphasizes that the ultimate source of factor remuneration is in sales proceeds and the motive for factor employment in the revenue product. We encounter the relationship in question in reference to ‘the present system of industrial life, in which employments are minutely subdivided, and all concerned in production depend for their remuneration on the price of a particular commodity’ (Mill 1965, 455). The principle is elaborated in a chapter dealing with indirect inputs of labour in lengthy processes of production: ‘All these people ultimately derive the remuneration of their labour from the bread, or its price: the ploughmaker as much as the rest; for since ploughs are of no use except for tilling the soil, no one would make or use ploughs for any other reason than because the increased returns thereby obtained from the ground, afforded a source from which an adequate equivalent could be assigned for the labour of the ploughmaker. If the produce is to be used or consumed in the form of bread, it is from the bread that this equivalent must come’ (31; it is presumably the expectation of future yield that provides the motive for the use of the input). 18 This perspective completely confounds Jevons’s reading of his predecessors. And there is no need to repeat what was said above regarding the fourth proposition on capital; Mill was crystal clear about its intended application to aggregate wages and employment alone (87). Finally, Mill’s pronouncements on the wages fund yield a vision of capitalist organization far removed from one wherein workers consume a distinct class of commodities produced in annual jets – a vision fully in line with that of Ricardo elaborated above (cf. Hollander 1984). WALRAS’S CLASSICAL COST-PRICE ANALYSIS It is not difficult to show that Walras subscribed to a cost-price analysis of the classical order. 19 In fact, it is the so-called ‘Marshallian’ adjustment mechanism entailing comparisons of demand and supply prices – the Smith–Ricardo–Mill tradition and also formulated so clearly by Say, as we have seen – rather than the so-called ‘Walrasian’ adjustment mechanism entailing comparisons of demand and supply quantities (pertinent in the simple exchange model) that he adopted in the production context: under free competition, if the selling price of a product exceeds the cost of the productive services for certain firms and a profit results, entrepreneurs will flow towards this branch of production or expand their output, so that the quantity of the product [on the market] will increase, its price will fall, and the difference between price and cost will be reduced; and, if [on the contrary], the cost of the productive services exceeds the selling price for certain firms, so that a loss results, entrepreneurs will leave this branch of production, or curtail their output, so that the quantity of the product [on 155

RICARDO - THE NEW VIEW

the market] will decrease, its price will rise and the difference between price and cost will again be reduced. (Walras 1954, 225) 20 This process involves not merely the transfer of services between sectors without alteration in the supplies of factors from which they derive, although doubtless to some extent this may be involved. For such transfer would be almost instantaneous, whereas Walras emphasized the slowness of adjustment to disturbance. We consider first remarks made regarding labour that imply the possibility of altering the types of personal capital in response to market pressures, in the long run; that is, upon retraining (or even perhaps with the renewal of the population stock). The case of ‘unspecialized productive services’ was ‘the most frequent case’, Walras conceded to the classics, particularly in the market for labour: Apart from certain individuals naturally gifted with the voice of a great tenor, the limbs of an acrobat, the eye of a painter or the ear of a musician, the great mass of men are capable of performing a wide variety of tasks, just because they are not especially qualified for the performance of any one of them. A man educated to be a lawyer might often just as well have been a manager; and certainly a person trained as a carpenter could have been a locksmith. What do most men inquire into when they come to choose their occupation? Surely, it is the wages they can earn in it, in other words, the value of their productive services in that occupation. The unspecialized productive services, in contradistinction to specific services have competition to fear. (Walras 1954, 401) Clearly there is a long-run tendency towards wage-rate equalization, contingent upon training and education, very much like that of Adam Smith who, of course, had also minimized innate differences of character and ability. The same conception appears in the famous chapter on the ‘Continuous Market’, where we find in effect a more general description of the ‘tendency’ of market to natural price mediated by credit – an account that would have been at home in any classical text. The emphasis is upon the slowness of adjustment to changes in data – including patterns of final demand and technology – having in mind the differential rates of replacement of circulating capital, personal capital and capital goods proper. But that such a ‘tendency’ is at work, albeit ever disturbed, is quite clear: Every hour, nay, every minute, portions of [the] different classes of circulating capital are disappearing and reappearing. Personal capital, capital goods proper and money also disappear and reappear, in a similar manner, but much 156

T H E S U B S T ANT I VE I DE N T I T Y

more slowly. Only landed capital escapes this process of renewal. Such is the continuous market, which is perpetually tending towards equilibrium without ever actually attaining it, because the market has no other way of approaching equilibrium except by groping, and, before the goal is reached, it has to renew its efforts and start over again, all the basic data of the problem, e.g. the initial quantities possessed, the utilities of goods and services, the technical coefficients, the excess of income over consumption, the working capital requirements, etc., having changed in the meantime. Viewed in this way, the market is like a lake agitated by the wind, where the water is incessantly seeking its level without ever reaching it. But whereas there are days when the surface of a lake is almost smooth, there never is a day when the effective demand for products and services equals their effective supply and when the selling price of products equals the cost of productive services used in making them. The diversion of productive services from enterprises that are losing money to profitable enterprises takes place in various ways, the most important being through credit operations, but at best these are slow. (Walras 1954, 380) It will be instructive to consider briefly at this point the formal model itself. A feature of Walras’s theory of capital is the determination of the rate of net revenue in terms of the exchange of net savings (incomes exceeding consumption and allowance for depreciation) for new additions to the stock of capital goods, the quantities of the different types of capital goods satisfying the condition of equality between their demand and supply prices. 21 This condition also reflects the principle of uniformity of net return on the new investments; for if the condition of uniformity ‘is not fulfilled with respect to any two capital goods, it will be advantageous to produce less of the capital good for which the ratio is smaller, and more of the capital good for which this ratio is larger’ (Walras 1954, 276; cf. 305). The emphasis is thus upon uniformity of return on new investments; but it would be difficult to appreciate the exclusion of the possibility that a change in the structure of final demand patterns or in technology such as Walras emphasizes, as we have seen, might lower the rentals on certain types of existing capital goods and thus require a contraction in their quantities by the non-investment of available depreciation allowances. In short, the principle of uniformity of the return on capital extends to investment decisions in general – to replacement demand as well as the net demand for new capital goods – although at any particular moment of time it is highly unlikely that uniformity across-the-board will be satisfied in the light of the disturbances in question: ‘In an economy like the one we have imagined, which establishes its economic equilibrium ab ovo, it is probable that there would be no equality of rates of net 157

RICARDO - THE NEW VIEW

income. Nor would such an equality be likely to exist in an economy which had just been disrupted by a war, a revolution or a business crisis’ (308). As we concluded, we are dealing only with a tendency to uniformity on capital-in-general. It is clear that equality of net interest is treated by Walras as a ‘point of reference’ only. It is in the course of adding to and replacing capital goods that decisions are made with an eye on prospective earnings – as we have seen to be the case also in the labour market – but expectations are continually disappointed, so that actual yields diverge from those expected when the investments are undertaken. This was Marshall’s position, too. He maintained that interest is earned on ‘free’ or ‘floating’ capital, and the phrase ‘rate of interest is applicable to old investments of capital only in a very limited sense’; given complexes earn quasi-rents, and in the polar extreme case of permanent investments the term interest on capital is totally inapplicable (Marshall 1920, 592–3; also 411–12, 418–19, 533). Nonetheless, the tendency to equalization is continually at play to the extent that complexes do wear out more or less rapidly. Indeed Marshall estimated that as much as 25 per cent of existing capital goods is replaced annually ‘even in a country in which the prevailing forms of capital are as durable as in England’, and he was prepared for some analytical purposes (particularly in the context of accumulation) to assume ‘that the owners of capital in general have been able in the main to adapt its forms to the normal conditions of the time, so as to derive as good a net income from their investments in one way as another’ (592). This has been termed a ‘sort of development of Walrasian theory’ – that there is uniformity of the rate of profits on ‘free capital’, which points ‘not towards longrun equilibrium analysis (in its stationary state, growth theory sense)’ but to ‘a sort of long-run equilibrium – in its proper sense of new capital gradually and actually flowing towards where quasi-rents have turned out best’ (Harcourt 1975, 351). 22 Surely this is the Walras–Marshall line itself? Now I can discern no difference between this line and that of Ricardo. For some analytical purposes Ricardo certainly presumed across-the-board equality of the return on capital. Thus, to investigate the consequences of a change in demand patterns or in technology he would set off from an assumed state of equilibrium in the sense that we ‘suppose that all commodities are at their natural price, and consequently that the profits of capital in all employments are exactly at the same rate’ (Ricardo 1951, I, 90). But he was perfectly well aware, first, that what matters in the (practical) profitability calculations that govern allocation is the return on new investments; for once investments are embodied in an actual capital structure, we are dealing with rentals. And secondly that in the polar case of permanent embodiments it is no longer pertinent to talk of a return on capital at all (I, 262; see above, p. 148). Finally, like both Walras and Marshall, Ricardo appreciated that adjustments are never 158

T H E S U B S T ANT I VE I DE N T I T Y

instantaneous even where the ‘withdrawal of capital’ is physically possible: ‘it always becomes a matter of calculation, whether these [capital goods] shall continue to be employed on the land, notwithstanding the low [the unexpectedly low] price of corn, or whether they shall be sold, and their value transferred to another employment’ (I, 269) – a calculation that is characterized by ‘the prejudices and obstinacy with which men persevere in their old employments’; for ‘they expect daily a change for the better, and therefore continue to produce commodities for which there is no adequate demand’ (VIII, 277). SOME IMPLICATION FOR THE ‘MARGINAL R E V O L U T I O N’ A major implication of the foregoing analysis is that the neo-classical developments of the 1870s involved pre-eminently an altered weighting in the selection of axioms, and in that sense a changed ‘focus of attention’ as well as a sharpening of analytical tools, but not a paradigmatic displacement. Thus, the principle of marginal utility merely added ‘very important explanations’ (Marshall, see above, p. 136) for the negative slope of the demand curve; and the marginal productivity principle added to the better understanding of factor demand in particular uses – elaborations required as much by Say as by Ricardo and Mill. To assert that recognition of a multiplicity of primary factors required a new theory based upon the principles of demand (Arrow and Starrett, see above, p. 137) or that classical theory could not solve ‘the logical problem of explaining the relative wages of heterogeneous types of labour’ (Arrow; cited in Hutchison 1978, 69) is historically unjustified. 23 The wage structure had long been analysed in terms of demand–supply with recognition of a productivity dimension on the demand side; and while value productivity is more conspicuous the more specialized to a particular use individual factors are, those same considerations are no less relevant when allowance is made for factor mobility between uses, although now strict limits are placed on the extent returns in different uses can diverge. In any event, although factor specificity is indeed a neo-classical preoccupation, both Ricardo and Mill carried this very matter far in their generalizations of the rent doctrine. If we take into account foreign trade and classical analysis in that case – the conspicuous role accorded demand considerations where the mobility axioms are abandoned – it becomes yet clearer that the notion of a paradigmatic transformation in the 1870s is not helpful. And while Mill was largely responsible for the analysis, it must be remembered that Ricardo had left the door open by his formulation (see above, p. 148) so that the elaborations, brilliant as they were, were consistent with existing doctrines. 159

RICARDO - THE NEW VIEW

Thus far we have considered the neo-classical ‘relaxation’ of the general (but far from universal) classical assumptions regarding factor mobility between uses. There is also the reverse of the medal to consider – the Ricardian assumption of single-use land. By adopting this assumption Ricardo indicates a preoccupation with the macro dimension; yet the ‘class’ relationship that concerned him, pre-eminently the inverse wage–profit relationship, could not be understood except in terms of allocation theory. He insisted, as Mill did, upon a micro-foundation for macro-analysis which seems eminently sensible if capitalist-exchange institutions are taken seriously. To trace through the consequences of a variation in the general wage in the case where (marginal) cost price incorporates land rent in the Smith– Say manner would have been technically impossible, given the state of the science. This is also true of J.S. Mill. (It is doubtful whether a specific outcome could be generated in the present day and age, for which reason so much analysis proceeds on the basis of two-factors and two-products.) For all that, the severe problems created for Ricardo’s theorem by allowance for multi-use land derive from an analytical model of allocation with which Ricardo himself was familiar. That he did not apply the assumptions of the model universally can be easily appreciated. That there was indeed no paradigmatic displacement is also quite evident from our investigation of Walras. For when he set aside his own restrictive assumption of factor immobility between uses, he was led to formulations of cost price identical to those of Ricardo and Mill. This theme can be extended. The appropriate axiomatic base depends in part on the context. The early and later nineteenth-century economists were concerned (as was Adam Smith) with both growth and allocation, although the weighting of their preoccupations certainly differed. Depending upon the context, it was appropriate to emphasize factor ‘scarcity’ or factor ‘reproducibility’ or various combinations. Thus Ricardo frequently dealt with disturbances (demand changes, innovation, taxation) within a static framework, although there is no question of his predominant concern with a broad range of analytical issues relating to the growth process – the aggregate factors (capital and labour) treated as variables. And conversely, Walras extended his own analysis in the Elements to the ‘Conditions and Consequences of Economic Progress’, which deals with the distributional implications of growth in labour and capital supplies (given land): ‘What does need to be discussed ... in view of its extremely weighty consequences, is the fact ... that the quantity of land cannot possibly increase though it is possible to increase the number of persons and the quantity of capital goods proper in an economy that saves and converts its savings into capital’ (Walras 1954, 382). 24 Similarly, he recognized the ‘excess of income over consumption in the aggregate’ – the matter of surplus and accumulation – as the condition of progress (264). Clearly, the classical growth model was not superannuated by the marginalists. 160

T H E S U B S T ANT I VE I DE N T I T Y

J.S. Mill’s continued preoccupation with growth issues (and with the general profit rate) despite his sharp awareness of the problem of factor immobility, also requires consideration from this perspective. For it is not obviously true that concern with the general wage and profit rates dissipates with recognition of the phenomenon of non-competing groups. There may yet be disturbances affecting all types of labour more or less equally, and for analysis of the growth process the standard classical mobility axiom (subject to the qualifications made regarding speed of adjustment) may be most appropriate. A NOTE ON SOURCES OF MISINTERPRETATION Professor Samuelson has written of the ‘sophisticated-anthropomorphic sin’ of not recognizing the equivalent content in older writers, because they did not use the same terminology and symbols as we do (Samuelson 1949, 373). This certainly explains to a large extent the general failure to recognize a key allocative dimension to British classicism. (To some extent this error was invited by Ricardo’s sometimes opaque formulations; since Ricardo was perhaps the first formally to contrast disturbances that have allocative implications and those that exert influence only at the aggregate level, this is scarcely surprising.) But there is also a matter of theoretical misunderstanding. It is sometimes presumed that since, in Ricardian theory, ‘profits are generated in production’, the kind of perspective we have adopted in this paper must be erroneous (Roncaglia 1982). This is a non sequitur. Ricardo’s achievement was to correct Smith’s ‘adding-up’ approach to cost, whereby a change in either wages or profits (or rents) simply generates a corresponding change in general prices: a wage increase must imply reduced profits since the total is constrained in real terms. Nothing that we have said regarding Ricardo’s acceptance of J.B. Say’s circular-flow conception of economic organization with its allowance for final demand controverts this point. For the volume of final demand itself, of course, is governed by the income flow, which in turn is generated in production. NOTES *The Harold A. Innis lecture given at the meetings of the Canadian Economics Association, June 1982, University of Ottawa. For their comments and advice thanks are due J.K. Whitaker (University of Virginia) and Irene M. Spry (University of Ottawa). I owe a special debt to Tom Kompas (Western Ontario) for most helpful criticism of the various drafts, particularly the discussion on Walrasian pricing. 1. Cf. Stigler (1965, 304): ‘[J.B.] Say’s approach was fundamentally much more modern than that of his English contemporaries’; for an elaboration of this position, see Hutchison 1978, 84f. 161

RICARDO - THE NEW VIEW

2. See also the emphasis upon given endowments of resources as the peculiar characteristic of ‘neo-classical’ theory, in Walsh and Gram (1980, 152). Here, too, Walrasian economics is represented as an exchange system extended to allow for production, capital formation, and money, but isomorphically, remaining faithful always to catallactics (123). And a sharp analytical distinction is made between a classical economics concerned with the creation, extraction, and division of the surplus between accumulation and luxury consumption by the capitalist class, and a neo-classical economics, wherein social class is irrelevant, focusing upon the allocation of given resources among alternative uses by means of competitive prices (9–10, 125–6). 3. Cf. Pasinetti (1974, 12) on the absence of a demand theory in the context of profit-rate equalization: ‘[Ricardo] does not find it useful to enter into complicated details (and in his case they would have been very complicated indeed for him, who did not possess a demand theory).’ 4. As far as Ricardo is concerned, my argument here is an elaboration of that given in Hollander 1979, ch. 6. I shall take for granted throughout the demonstration there given of Ricardo’s appreciation of the ‘demand schedule’ – and the variability of the wage rate – although his position in this regard will be apparent in the citations below. 5. In the sense of forgone products alone, and excluding forgone leisure (cf. Robbins 1970, 18). 6. That J.S. Mill in his Principles had turned or was in the process of turning his back on Ricardianism is a widespread belief; cf. for example Schumpeter (1954, 529): ‘the economics of [Mill’s] Principles are no longer Ricardian. ... From Marshall’s Principles Ricardianism can be removed without being missed at all. From Mill’s Principles, it could be dropped without being missed very greatly.’ 7. It is pertinent to refer also to the opinion that Thomas De Quincey corrected Ricardian value theory by stressing the mutual determination of exchange value by ‘intrinsic utility’ and ‘difficulty of attainment’, a ‘correction’ which ‘greatly influenced J.S. Mill’s treatment of value in the Principles and which is in the Hutcheson–Smith tradition of value theory’ (Groenewegen 1974, 193) This is not to my mind a convincing evaluation. The ‘mutual determination’ of exchange value by demand and cost considerations was a thoroughly central aspect of Ricardian doctrine. De Quincey may have believed he was ‘correcting’ Ricardo, but it is unlikely that Mill was convinced. It must be stated that Mill studies are in a state of confusion. For the literature also provides assertions to the effect that Mill ‘put the clock back’ by subscribing to cost of production theories (Hutchison 1978, 64–5n, citing Sowell 1972), a view qualified by admiration for Mill’s contribution to the theory of international trade with its evident demand dimension (Sowell 1972, 159–60). 8. See also Say’s (1820, 12f) doctrine of productive services, regarding which Ricardo wrote to Malthus: ‘if he would give up rent, he and I should not differ very materially on that subject’ (1951, VIII, 277) and to Say: ‘In your doctrine of productive services I almost fully agree, but I submit to you, whether, as rent is the effect of high price, and not the cause of it, it should not be rejected when we estimate the comparative value of commodities’ (VIII, 279). 9. Although a good case can be made whereby Ricardo allowed that the profit rate contains a reward for abstinence and thus acts upon accumulation; Hollander 1979, ch. 7. 10. This is sometimes conceded by Knight (1956, 41–2n, 63) and Schumpeter (1954, 590), despite their generally critical approaches to Ricardian economics from a neo-classical perspective. 162

T H E S U B S T ANT I VE I DE N T I T Y

11. And Ricardo himself referred to the relation between wages and profits in terms of ‘equilibrium’ (1951, I, 226). 12. Cf. Ricardo 1951, I, 252: ‘I hope I have made it sufficiently clear, that until a country is cultivated in every part’ (an allusion to the extensive margin) ‘and up to the highest degree’ (an allusion to the intensive margin), ‘there is always a portion of capital employed on the land which yields no rent, and that it is this portion of capital, the result of which, as in manufactures, is divided between profits and wages that regulates the price of corn.’ 13. This point is frequently neglected. See below, p. 161. 14. ‘It is not necessary to the production of things that cannot be used as subsistence or cannot be immediately utilized, that there should have been a previous production of the wealth required for the maintenance of the labourers while the production is going on. It is only necessary that there should be, somewhere within the circle of exchange, a contemporaneous production of sufficient subsistence for the labourers, and a willingness to exchange this subsistence for the thing on which the labour is being bestowed’ (George 1879, 24). 15. In his Leading Principles (1874, 48–9) J.E. Cairnes criticized Mill’s inclusion of wages and profits within costs, although that position was ‘generally accepted by economists’. Wages and profits, he argued, were not ‘costs’ in the legitimate sense of that term – namely, ‘sacrifices incurred by man in productive industry’ – but, on the contrary, they constituted ‘the return made by nature to man upon that sacrifice’; ‘labour’ and ‘abstinence’ were the true costs of production. Mill (who already had some idea of the nature of the criticism) observed to his friend that both forms were legitimate depending on context – whether it involved the economic system as a whole or the motives of the individual participants in activity: ‘the cost to society, as a whole, of any production, consists in the labour and abstinence required for it. But, as concerns individuals and their mutual transactions, wages and profits are the measure of that labour and abstinence, and constitute the motives by which the exchange of commodities against one another is immediately determined’ (Mill 1972, 1894–5). 16. It must at the same time be remembered that in introducing his account of the ‘equation of demand and supply’ Mill did insist upon his own priority, with the exception of J.B. Say, regarding the solution to the ‘paradox of two things, each depending upon the other’. But what we know of Mill’s general reaction to Ricardianism suggests that he regarded – and rightly so – this analysis as a clarification of sometimes obscure or ambiguous or incomplete formulations in the original statements of 1817. 17. Pertinent, too, is Mill’s early defence of the differential rent theory against the strictures of Senior and others ‘who affect to suppose that Sir Edward West, Mr. Malthus, and Mr. Ricardo, considered the cultivation of inferior land as the cause of a high price of corn’. The reverse was the case, Mill insisted in 1828; that ‘the cultivation of inferior soils’ was the effect of high price ‘itself the effect of demand’ was a doctrine ‘explicitly laid down by the distinguished authors previously referred to, and particularly by Mr. Ricardo’ (Mill 1967, 174) – a perfectly justified defence of the Ricardian position. 18. Cf. Mill 1965, 32: the labourers producing fixed capital ‘do not depend for their remuneration upon the bread made from the produce of a single harvest, but upon that made from the produce of all the harvests which are successively gathered until the plough, or the buildings and fences, are worn out’. 19. I have found Milgate (1979) most helpful. 163

RICARDO - THE NEW VIEW

20. I understand ‘profit’ as super-normal profit in excess of the normal return earned on capital goods proper; cf. Walras 1954, 267. 21. ‘New capital goods are exchanged against the excess of income over consumption; and the condition of equality between the value of the new capital goods and the value of the excess gives us the equation required for the determination of the rate of net income and consequently for the determination of the prices of capital goods. Moreover, new capital goods are products; and the condition of equality between their selling price and their cost of production gives us the equations required for the determination of the quantities manufactured’ (Walras 1954, 269–70). 22. See also Hicks 1973, 340–1: ‘it is quite unnecessary, because we use terms like “capitalintensive” and “rate of profit”, to trouble ourselves about the valuation of the capital stock as a whole (as we appeared to have to do in the production function method). What matters is not the average rate of profit on the whole capital stock (which cannot be determined without such valuation); what matters is the rate of profit on new investment. When the new investment is undertaken, that profit is no more than an expected profit, and what is realized may not be the same as what is expected’. 23. It may be agreed that the neo-classics ‘took as an expository point of departure a model which was the polar opposite of the classical, the model of pure exchange’ (Arrow and Starrett 1973, 133). Clear exposition may require extreme assumptions, as Ricardo repeatedly insisted. 24. Professor Morishima indeed goes so far as to consider this dynamic part as the capstone of the entire general-equilibrium structure, from which perspective he regards the Elements as a whole ‘as providing a general equilibrium foundation for the Ricardian neo-classical macroeconomics’ (Morishima 1980, 558). Even W. Jaffé, who denied that Part VII was intended as ‘an integral part of Walras’s general equilibrium edifice’, conceded that Walras’s intentions were to show ‘how the relations analyzed in the static theory could be used to elucidate such dynamic tendencies as the rise in land-rent, and the fall in the rate of profit in an expanding economy’ (Jaffé 1980, 546–7).

REFERENCES Arrow, K.J. and Starrett, D.A. (1973) ‘Cost- and Demand-Theoretical Approaches to the Theory of Price Determination’, in J.R. Hicks and W. Weber, Carl Menger and the Austrian School of Economics, Oxford: Oxford University Press. Bharadwaj, D. (1978) ‘The Subversion of Classical Analysis: Alfred Marshall’s Early Writing on Value’, Cambridge Journal of Economics, 253–71. Cairnes, J.E. (1874) Some Leading Principles of Political Economy, London: Macmillan. Dobb, M. (1973) Theories of Value and Distribution, Cambridge: Cambridge University Press. Frisch, R. (1981) ‘From Utopian Theory to Practical Application: the Case of Econometrics’, American Economic Review LXXI (June): 1–16. Garegnani, P. (1972) ‘Heterogeneous Capital, the Production Function and the Theory of Distribution’, in E.K. Hunt and J.G. Schwartz (eds) A Critique of Economic Theory, Harmondsworth: Penguin. — (1976) ‘On a Change in the Notion of Equilibrium in Recent Work on Value and Distribution’, in M. Brown, K. Sato and P. Zarembka (eds) Essays in Modern Capital Theory, Amsterdam: North-Holland. 164

T H E S U B S T ANT I VE I DE N T I T Y

George, H. (1879) Progress and Poverty, San Francisco: W.M. Hinton. Groenewegen, P. (1974) Review, Economic Journal LXXXIII (March): 192–3. Harcourt, G.C. (1975) ‘Decline and Rise: the Revival of (Classical) Political Economy’, Economic Record LI: 339–55. Hicks, John (1973) ‘The Mainspring of Economic Growth’, Swedish Journal of Economics LXXV: 336–48. —— (1976) ‘ “Revolutions” in Economics’, in S. Latsis,Method and Appraisal in Economics, Cambridge: Cambridge University Press. Hollander, S. (1976) ‘Ricardianism, J.S. Mill, and the Neo-Classical Challenge’, in J.M. Robson and M. Laine (eds) James and John Stuart Mill: Papers on the Centenary Conference, Toronto: University of Toronto Press. The Economics of David Ricardo, Toronto: University of Toronto Press. —— (1979) —— (1984) ‘J.S. Mill on “Derived Demand” and the Wage-Fund Theory Recantation’,Eastern Economic Journal IX,1: 87–98. Hutchison, T.W. (1978) On Revolutions and Progress in Economic Knowledge, Cambridge: Cambridge University Press. Jaffé, W. (1980) ‘Walras’s Economics as Others See It’, Journal of Economic Literature XVIII: 528–49. Jevons, W.S. (1905) Principles of Economics, London: Macmillan. —— (1924) Theory of Political Economy, 4th edn, London: Macmillan. Knight, F.H. (1956) On the History and Method of Economics, Chicago: University of Chicago Press. Malthus, T.R. (1963) [1824] ‘Political Economy’, in B. Semmel (ed.) Occasional Papers, New York: Burt Franklin. Marshall, Alfred (1920) Principles of Economics, 8th edn, London: Macmillan. Milgate, M. (1979) ‘On the Origin of the Notion of “Intertemporal” Equilibrium’, Economica XLVI: 1–10. Mill, J.S. (1965) Principles of Political Economy, Collected Works, vols II, III, Toronto: University of Toronto Press. —— (1967) Essays on Economics and Society, Collected Works, vols IV, V, Toronto: University of Toronto Press. —— (1972)The Later Letters, Collected Works, vols XIV, XV, Toronto: University of Toronto Press. Morishima, M. (1980) ‘W. Jaffé on Léon Walras’, Journal of Economic Literature XV: 550–8. Pasinetti, L.L. (1974) Growth and Income Distribution, Cambridge: Cambridge University Press. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols) Cambridge: Cambridge University Press. Robbins, L.C. (1970) The Evolution of Modern Economic Theory, London: Macmillan. Robinson, Joan (1961) ‘Prelude to a Critique of Economic Theory’, Oxford Economic Papers XII: 53–8. Roncaglia, A. (1978) Sraffa and the Theory of Prices, Chichester and New York: John Wiley. —— (1982) ‘Hollander’s Ricardo’,Journal of Post-Keynesian Economics IV: 339–59. Samuelson, P.A. (1949) Review, Economica XV: 373–4. Say, J.B. (1820) Lettres à M. Malthus sur différens sujets d’économie politique, Paris: Bossange. Schumpeter, J.A. (1954) History of Economic Analysis, New York: Oxford University Press. Shove, G. (1960) [1942] ‘The Place of Marshall’s Principles in the Development of Economic 165

RICARDO - THE NEW VIEW

Theory’, in J.J. Spengler and W.R. Allen, Essays in Economic Thought, Chicago: Rand McNally. Sowell, T. (1972) Say’s Law, Princeton: Princeton University Press. Cambridge: Cambridge Sraffa, P. (1960) Production of Commodities by Means of Commodities, University Press. Stigler, G.J. (1965) Essays in the History of Economics, Chicago: University of Chicago Press. Walras, Léon (1954) Elements of Pure Economics, ed. W. Jaffé, 4th definitive edn (1926) London: George Allen and Unwin. Walsh, V. and Gram, H. (1980) Classical and Neo-classical Theories of General Equilibrium, New York: Oxford University Press.

166

10 WHY MARSHALL WAS RIGHT ABOUT RICARDO*

I. INTRODUCTION I propose to demonstrate in this lecture the validity of Alfred Marshall’s defence of Ricardo against the ‘marginalist’ reaction of the 1870s. In my view Marshall was wholly right as far as he went in his defence, but perhaps did not go far enough. For he stayed rather too much within a partial equilibrium framework, whereas Ricardo’s economics reflect pre-eminently general equilibrium. By this I intend appreciation of the information conveying and signalling role of prices, the notion of alternative costs, the principle of maximizing net returns and market interaction of goods and services – in brief the theory of the coordination of decentralized economic activities. 1 Here I should specify a distinction between concern with allocative problems based on market mechanisms, and concern with individual choice. We are conditioned nowadays to think of rigorous individual choice theory as necessary to a theory of resource allocation. 2 Ricardo, on the other hand, was less concerned with the problem of individual choice than with market process analysis. Still, it would be misleading to draw the contrast too sharply. Thus even Schumpeter, who was foremost in condemning Ricardo for failing to appreciate the production–distribution process as a web of exchanges, was obliged to concede that ‘the “classics” sensed the existence of what we now call economic equilibrium and if they did not try to prove its existence, they made it, as it were, plausible, embodying their intuition in certain empirical rules, such as the tendency of “profits” to be roughly equal in different but similarly conditioned lines of business. We derive a similar proposition from the principle of maximizing net returns and associate it with the principle of substitution’ (Schumpeter 1954, 590). 3 A second preparatory clarification relates to the fact that Ricardo’s primary concern was the time path of the average returns to labour and capital in conditions of increasing land scarcity. Allocation theory was not his preeminent interest. Yet the 167

RICARDO - THE NEW VIEW

growth process is driven from the side of demand. When he lets the marginal cost of food production rise under pressure of population, it is the increased demand for food which forces the extension of cultivation, and there is much of interest in this context regarding budget allocations by the individual with income increase. Moreover, by conceiving the average factor returns as endogenous variables and elaborating upon the conditions of demand and supply in the labour market he opened the door to the interdependence of income distribution and final expenditure patterns, thus confuting current ‘Italo-Cambridge’ and sundry other critics of his general Marshallian perspective. This paper, then, argues the case that Ricardian economics comprises an exchange system consistent with later ‘neo-classical’ or ‘general-equilibrium’ or ‘marginalist’ elaborations – that Marshall’s position on nineteenth-century historiography is substantially accurate. 4 In the course of my demonstration, and with Marshall’s valid protest against fixed-wage interpretations of Ricardo in mind, I shall take up the relation between distribution and final expenditure patterns. II. THE ‘MARGINALIST’ AND THE ‘CAMBRIDGE’ POSITIONS Distribution was envisaged by W.S. Jevons as (ideally) a matter of service pricing, ‘entirely subject to the principles of value and the laws of supply and demand’ (1924, xlvi), with input prices ‘the effect not the cause of the value of the produce’ (1) – ‘I hold labour to be essentially variable, so that its value must be determined by the value of the produce, not the value of the produce by that of the labour ’ (166) – and cost of production a reflection of opportunities forgone (xlviii–l). He accordingly directed his criticisms at the wage fund and subsistence approaches to wage-rate determination and the cost approach to value – as he understood them – paying tribute to the French tradition of J.B. Say and others: ‘the only hope of attaining a true system of Economics is to fling aside, once and for ever, the mazy and preposterous assumptions of the Ricardian School. Our English Economists have been living in a fool’s paradise. The truth is with the French School’ (xliv–xlv). Léon Walras, whose intellectual origins include J.B. Say, similarly objected to the classical pricing and distribution model (again, as he understood it), particularly the cost orientation and the natural wage approach. By neglecting a final demand dimension and accordingly derived demand, so he charged, the English had constructed an under-determined system: ‘It is clear that the English economists are completely baffled by the problem of price determination; for it is impossible for I [interest] to determine P [price] at the same time that P determines I. In the language of mathematics one equation cannot be used to determine two unknowns’ (Walras 1954 [1874] 424–5). 168

WAY MARSHALL WAS RIGHT ABOUT RICARDO

In more recent times we have the famous criticism of classicism by Schumpeter and Knight, already mentioned. From their perspective prices depend on the relative subjective appeal to consumers, the flow of goods and thus their marginal utilities governed by cost considerations, where ‘costs’ reflect alternatives surrendered rather than the ‘pain’ of labour or abstinence. On this view, the economizing principle involves maximizing the total return from any resource by equalizing the increments of return at the margin to the scarce resource in alternative uses. Taking this perspective, the developments of the 1870s are, obviously, ‘revolutionary’. But for a pristine formulation of the supposedly revolutionary status of the developments of the 1870s – especially respecting the role of final demand – the following statement by Frisch is difficult to improve upon: The classical theory of value – as we find it streamlined in Stuart Mill – was essentially a theory of production costs based on the thinking of the private entrepreneur. ... This theory contains, of course, an irrefutable element of truth. But it is too simple to give even a crude presentation of the forces of play. The economic process is an equilibrium affair where both technological and subjective forces are at play. The subjective element was nearly left out by the classicists. On this point economic theory was completely renewed, in the years between 1870 and 1890 when a number of Austrian economists headed by Carl Menger (1840–1921) undertook a systematic study of the human wants and their place in a theory of prices. Similar thoughts were expressed also by the Swiss Léon Walras (1834–1910) and the Englishman Stanley Jevons (1835– 1882). This was the first break-through since Stuart Mill. The Englishman, Alfred Marshall (1842–1924) subsequently did much to combine the subjective viewpoint and the cost of production viewpoint. This led to what we now speak of as the neo-classical theory. (Frisch 1981, 5) It has also been suggested (Arrow and Starrett 1973, 132–3) that Mill’s recognition of non-competing groups implied a multiplicity of primary factors which ‘required a new theory’. The founders of the neo-classical school ‘understood the glaring omission of demand from the classical model’. Classical theory, runs the argument, could not solve the logical problem of explaining the relative wages of heterogeneous types of labour. This summary should suffice to set the stage for Marshall’s own, very different, reading of nineteenth-century economic thought. But before proceeding we take the occasion to spell out the ‘Italo-Cambridge’ or ‘neo-Ricardian’ tradition which 169

RICARDO - THE NEW VIEW

ascribes to Ricardo many of the features just outlined, but by way of mirror image – where neo-classical critics find error or lacunae Cambridge finds merit. The characteristic of this tradition is its championship of models entailing a priority of distribution over pricing in which one or other of the key distributive variables – either the wage or the profit rate – is exogenously given to the economic system; and in which demand–supply analysis in factor and commodity markets is to all interests and purposes precluded. Under certain precisely defined conditions a given real wage will dictate the return on capital and the set of cost prices. Maurice Dobb has phrased it thus: It should be fairly clear ... that a system which determines distribution in terms of exchange and its emergent prices must, in one way or another, with possibility of varying emphasis, be cast in terms of supply and demand; but au contraire, the Ricardian system, which explains exchange in terms of distribution, and distribution itself in terms of productivity and conditions of production in one industry or sector of industry (given the real-wage), has no place for the relation of supply and demand – at least, until it comes to movements in relative prices, and in particular of Smithian market prices. (Dobb 1973, 118–19) Similarly: ‘The nature of [Marx’s] approach required him to start from the postulation of a certain rate of exploitation or of surplus-value (or profit– wage ratio in Ricardo’s terms); since this was prior to the formation of exchange-values or prices and was not derived from them. In other words, this needed to be expressed in terms of production, before bringing in circulation or exchange’ (148). For a very recent statement of the Cambridge view, we have the contribution on ‘Ricardo’ for the New Palgrave Dictionary of Economics (1987). G. de Vivo, the contributor, here asserts that Marshall’s position on Ricardo – his attribution to him of a variable wage doctrine and a generalized cost theory of value – was based on ‘flimsy evidence’, and proceeds to maintain (on the basis of no evidence at all) that Ricardian theory ‘basically conflicted with marginalism’: ‘given the rate of wages and productive techniques (and assuming no rent), the rate of profits would be wholly determined, and therefore the cost of production and the price of the product. No role would be left to play for the marginalist mechanism of supply and demand: the proportions in which goods are demanded become irrelevant to the determination of value and distribution, and there would not be any mechanism to bring the supply and demand of factors to equality’ (de Vivo 1987, 197). In what follows I shall provide the answer to both the early or marginalist critics of Ricardo and the self-styled ‘neo-Ricardians’. The textual evidence fully supports Marshall’s reading of Ricardo, though as already mentioned, that line extends yet further than Marshall indicated. 170

WAY MARSHALL WAS RIGHT ABOUT RICARDO

III. MARSHALL ON RICARDO’S VALUE THEORY Marshall’s famous Appendix I to his Principles of Economics addresses the criticisms by Jevons of Ricardo (and J.S. Mill) on value theory. ‘Though obscurely expressed, it anticipated more of the modern doctrine of the relations between cost, utility and value, than has been recognized by Jevons and some other critics’ (1920, xxxiii). He believed that Jevons had judged Ricardo and Mill harshly, attributing to them ‘doctrines narrower and less scientific than those which they really held’ (817). Ricardo’s Principles of Political Economy was admittedly an unsystematic work, deliberately incomplete with much taken for granted from his readers, necessitating reference to other parts of his work (including correspondence) to settle ambiguities (813). Ricardo also ‘delighted in short phrases, and he thought his readers would always supply for themselves the explanations of which he had given them a hint’ (816). But when Ricardo’s text is approached in an effort to ‘ascertain what he really meant’, it emerges that ‘his doctrines, though very far from complete, are free from many of the errors that are commonly attributed to him’ (813). In fact, Ricardo’s position on value ‘though unsystematic and open to many objections, seems to be more philosophic in principle and closer to the actual facts of life’ than that of Jevons (819).5 Here Marshall had in mind the famous catena or chain of causation which he saw to be Jevons’s ‘central position’: Cost of production determines supply; Supply determines final degree of utility; Final degree of utility determines value. (Jevons 1924 [1871], 165)6 Marshall objected first to the ambiguity of ‘cost’ and ‘supply’ for which there was little excuse since Jevons had the technical apparatus required to deal with that matter. Secondly, and more serious, was the notion that marginal utility ‘determines value’: ‘For the price which the various purchasers in a market will pay for a thing, is determined not solely by the final degrees of its utility to them, but by these in conjunction with the amounts of purchasing power severally at their disposal. The exchange value of a thing is the same all over a market; but the final degrees of utility to which it corresponds are not equal at any two parts’ (1920, 818). Marginal demand price was the appropriate concept. And he proceeds subsequently: ‘Perhaps Jevons’s antagonism to Ricardo and Mill would have been less if he had not himself fallen into the habit of speaking of relations which really exist only between demand price and value as though they held between utility and value’ (820). Thirdly, and most serious of all, was the failure to convey the notion of mutual causation : ‘But the greatest objection of all to his formal statement of his central doctrine is that it does 171

RICARDO - THE NEW VIEW

not represent supply price, demand price and amount produced as mutually determining one another (subject to certain other conditions), but as determined one by another in a series’ (818). In fact, if a catena is insisted upon, a reversed order would be ‘rather less untrue’ than that of Jevons, namely Utility determines the amount that has to be supplied, The amount that has to be supplied determines cost of production, Cost of production determines value, because cost determines the supply price which is required to make the producers keep to their work. (Marshall 1920, 818–19) Jevons had failed to convey what Cournot long before had emphasized and as the use of mathematics should have suggested, namely ‘that fundamental symmetry of the general relations in which demand and supply stand to value’ (820). 7 Marshall admitted that Ricardo ‘does not state clearly, and in some cases perhaps did not fully and clearly perceive how, in the problem of normal value, the various elements govern one another mutually not successively in a long chain of causation’ (816). But, as mentioned, he found Ricardo’s formulations to be ‘more philosophic in principle’ than those of Jevons. The dependence of value on costs reflected ‘careless brevity’, not principle, and was ‘part of a larger doctrine, the rest of which he had tried to explain’ (817). In particular, Jevons had erroneously assumed that Ricardo thought of value as governed by cost of production without reference to demand (821n). As for the details of Ricardo’s position regarding the supply side, Marshall lambasted the fallaciousness of attributions to him of a simple-minded labour theory: ‘it seems difficult to imagine how he could have more strongly emphasized the fact that Time or Waiting as well as Labour is an element of cost of production’ (816). At most, it was a strategic error of judgment on Ricardo’s part not to continually repeat his position that prices are proportionate to labour values only under restrictive conditions – all of which he laid down in the first chapter of the Principles. 8 A second matter relates to the demand side. Ricardo had far less to say on demand than costs in the analysis of value; the misunderstandings he sought to correct related largely to costs (814). Still, in his first chapter he specified that utility was ‘absolutely necessary’ to normal value if not its measure; and more important, in his chapter ‘Value and Riches’ he sought to convey the contrast between marginal and total utility and the differential impact on them of a reduction in supply: Again, in a profound, though very incomplete, discussion of the difference between ‘Value and Riches’ he seems to be feeling his way towards the distinction between marginal and total utility. For by Riches he means total 172

WAY MARSHALL WAS RIGHT ABOUT RICARDO

utility, and he seems to be always on the point of stating that value corresponds to the increment of riches which results from that part of the commodity which it is only just worth the while of purchasers to buy; and that when the supply runs short, whether temporarily in consequence of a passing accident, or permanently in consequence of an increase in cost of production, there is a rise in that marginal increment of riches which is measured by value, at the same time that there is a diminution in the aggregate riches, the total utility, derived from the commodity. Throughout the whole discussion he is trying to say, though (being ignorant of the terse language of the differential calculus) he did not get hold of the right words in which to say it neatly, that marginal utility is raised and total utility is lessened by any check to supply. (Marshall 1920, 814) In the body of Marshall’s Principles the same point emerges in the course of an important claim regarding the continuity of nineteenth-century doctrine: A great change in the manner of economic thought has been brought about during the present generation by the general adoption of semi-mathematical language for expressing the relation between small increments of a commodity on the one hand, and on the other hand small increments in the aggregate price that will be paid for it: and by formally describing these small increments of price as measuring corresponding small increments of pleasure. The former, and by far the most important, step was taken by Cournot (1838); the latter by Dupuit (1844), and by Gossen (1854). But their work was forgotten; part of it was done over again, developed and published almost simultaneously by Jevons and by Carl Menger in 1871, and by Walras a little later. Jevons almost at once arrested public attention by his brilliant lucidity and interesting style. He applied the new name final utility so ingeniously as to enable people who knew nothing of mathematical science to get clear ideas of the general relations between the small increments of two things that are gradually changing in causal connection with one another. His success was aided even by his faults. For under the honest belief that Ricardo and his followers had rendered their account of the causes that determine value hopelessly wrong by omitting to lay stress on the law of satiable wants, he led many to think he was correcting great errors; where he was really only adding very important explanations. (Marshall 1920, 101n) Marshall readily conceded that at the time Jevons wrote ‘the demand side of the theory of value has been much neglected; and that [Jevons] did excellent service by 173

RICARDO - THE NEW VIEW

calling attention to it and developing it’ (820). But while Ricardo took the operation of demand ‘too much for granted’ (525), the reaction had gone too far in the other direction and in some cases had encouraged an approach to distribution theory implying that inputs are specific to particular products, i.e. had no alternative uses, so that their returns are in the nature of rent or demand-determined surpluses, with inadequate attention given to conditions of factor supply: In the reaction, too much insistence has been laid on the fact that the earnings of every agent of production come from, and are for the time mainly governed by the value of the product which it takes part in producing; its earnings being so far governed on the same principle as the rent of land; and some have even thought it possible to constitute a complete theory of Distribution out of multifold applications of the law of rent. But they will not reach to that end. Ricardo and his followers seem to have been rightly guided by their intuitions, when they silently determined that the forces of supply were those, the study of which is the more urgent and involves the greater difficulty. Marshall was too harsh. Jevons’s catena should also be seen as ‘part of a larger doctrine’. Consider the section ‘Relation of the Theories of Labour and Exchange’: ‘It may tend to give the reader confidence in the preceding theories when he finds that they lead directly to the well-known law, as stated in the ordinary language of economists, that value is proportional to the cost of production’ (Jevons 1924, 186). Or again: ‘It is not easy to express in words how the ratios of exchange are finally determined. They depend upon a general balance of producing power and of demand as measured by the final degree of utility. Every additional supply tends to lower the degree of utility; but whether that supply will be forthcoming from any country depends on its comparative powers of producing different commodities’ (188). H. Stanley Jevons (1911) defended his father against Marshall’s charge (in the context of interest rate determination) in terms of Jevons’s habit – his Ricardian habit one might add – of deliberate simplification in conveying particular propositions (H.S. Jevons 1911, 280). But my concern is Marshall on Ricardo. I shall show that Marshall was right in what he said but could have gone yet further in allowing for Ricardo’s appreciation of the demand side and in emphasizing the general-equilibrium implications of Ricardo’s demand–supply orientation. IV. RICARDO AND THE THEORY OF DEMAND Nothing could be further from the truth than the tradition that demand theory played a small part, if any at all, in Ricardian analysis. Ricardo went to some lengths to make the point ‘that the production of no commodity, except from miscalculation, precedes the demand or anticipated demand for it’ (Ricardo 1951, VIII, 273–4). This 174

WAY MARSHALL WAS RIGHT ABOUT RICARDO

apparently obvious point had to be made, since in T.R. Malthus’s opinion increased food production preceded increased population: ‘The point in dispute is this, Does the supply of corn precede the demand for it, or does it follow such demand? ... I [am] of the latter opinion’ (255). But much more is involved than recognition of demand as a necessary condition for exchange value. The responsiveness of quantity demanded to price variation was clearly recognized by Ricardo in the Principles, although there would perhaps be ranges of inelasticity: ‘Whatever habit has rendered delightful, will be relinquished with reluctance, and will continue to be consumed notwithstanding a very heavy tax; but this reluctance has it limits, and experience every day demonstrates that an increase in the nominal amount of taxation, often diminishes the produce’ (I, 241). Ricardo recognized elsewhere that price elasticity – defined in terms of the proportionate response of quantity demanded to price and consequential changes in total expenditure – varies over the range of prices, exceeding unity at high levels and declining to values less than unity at lower levels; thus ‘if he went on in this way increasing the quantity of the cloth, until it came within the reach of the purchase of every class in the country, from that time any addition to its quantity would diminish the aggregate value’ (speech, 1822; V, 171). Ricardo’s logic turns on the notion that at low prices the opportunities to attract additional purchasers by further price reductions are limited – a clear notion of marginal demand price. This rationalization of the inelastic range appeared also in early correspondence of 1813 involving the differential characteristics of regular commodities and of the monetary metals: Coffee, Sugar, and Indigo, are commodities for which, although there would be an increased use, if they were to sink much in value, still as they are not applicable to a great variety of new purposes, the demand would necessarily be limited; not so with gold and silver. These metals exist in a degree of scarcity, and are applicable to a great variety of new uses – the fall of their price, in consequence of augmented quantity, would always be checked, not only by an increased demand for those purposes to which they had before been applied, but to the want of them for entirely new employment. (Ricardo 1951, VI, 91–2) It does not seem that Ricardo appreciated the modern ‘substitution effect’, the response of quantity demanded to relative price changes assuming unchanged purchasing power (J.S. Mill did; cf. Hollander 1987a, 123). Variation of quantity demanded upon price change he attributed to our ‘income effect’. The income elasticity of demand for corn he presumed to be very low and sometimes zero for simplicity; and he maintained with pellucid clarity in the Principles that it is the zero income elasticity of demand which accounts for zero price elasticity: ‘An increase in the cost of production of a commodity, if it be an article of the first necessity, will 175

RICARDO - THE NEW VIEW

not necessarily diminish its consumption; for although the general power of the purchasers to consume, is diminished by the rise of any one commodity, yet they may relinquish the consumption of some other commodity whose cost of production has not risen’ (Ricardo 1951, I, 343–4). That Ricardo lacked the formal conception of marginal utility is clear from an observation regarding ‘value-in-use’ and ‘value-in-exchange’: ‘If by an improved machine I can with the same quantity of labour, make two pair of stockings instead of one, I in no way impair the utility of one pair of stockings, though I diminish their value’ (I, 208n). In this passage the emphasis is on the physiological properties of commodities, which suggests a hierarchical ranking of utility, and rules out the incremental dimension. It is, however, true that Ricardo also adopted the psychological dimension, or a notion of utility in terms of satisfaction or ‘gratification’ (e.g. I, 11). His slipping into physiology cannot explain everything. At the same time, it must not be forgotten that Say too did not spell out the incremental dimension (see letter of 19 July, 1821, IX, 31f; cf. Hollander 1979, 277 n19). Moreover, it is a conspicuous fact that, notwithstanding the neglect of the incremental dimension to utility, Ricardo (like Adam Smith before him) skilfully applied the scarcity property of valuable goods to a wide variety of cases. Consider first his reply (in a letter to Malthus) to a charge that the British economists neglected the scarcity property of valuable goods: [J.B. Say’s brother] Louis Say has published a thick volume of remarks upon Adam Smith’s, his brother’s, Your and my opinions. He is not satisfied with any of us. His principle object is to shew that wealth consists in the abundance of enjoyable commodities, – he accuses us all of wishing to heap up what we call valuable commodities, without any regard to quantity, about which only the Polit. Economist should be anxious. I do not believe that any of us will plead guilty to this charge. I feel fully assured that I do not merit it should be made against me. (Ricardo 1951, IX, 248–9) Quite evidently, Ricardo placed himself with Smith, Malthus and Say amongst those who adopted the consumer’s concept of wealth with all that it implies from the scarcity perspective.9 It is hardly surprising then that Ricardo was perfectly well equipped to solve the so-called ‘paradox of value’, an application of which solution is made in the context of land rent: On the first settling of a country, in which there is an abundance of rich and fertile land ... there will be no rent; ... 176

WAY MARSHALL WAS RIGHT ABOUT RICARDO

On the common principles of supply and demand, no rent could be paid for such land, for the reason stated why nothing is given for the use of air and water, or for any other of the gifts of nature which exist in boundless quantity ... [no] charge is made for the use of these natural aids [in production], because they are inexhaustible, and at every man’s disposal. In the same manner the brewer, the distiller, the dyer, make incessant use of the air and water for the production of their commodities; but as the supply is boundless, they bear no price. (Ricardo 1951, I, 69) The paradox is resolved elsewhere in a particularly clear manner: ‘Why is water [though ‘abundantly useful’ as explained on the opening page of the Principles ] without value [in exchange], but because of its abundance? If corn were equally plenty [sic] it would have no greater value, whatever quantity of labour might have been bestowed on its production’ (IV, 221). 10 Significantly, the scarcity principle is used to defend Adam Smith against a charge by Say that Smith has neglected ‘the value which is given to commodities by natural agents, and by machinery’ by relating value to labour: it does not appear to me, that this charge is made out; for Adam Smith no where undervalues the services which these natural agents and machinery perform for us, but he very justly distinguishes the nature of the value which they add to commodities – they are serviceable to us, by increasing the abundance of productions, by making men richer, by adding to value in use; but as they perform their work gratuitously, as nothing is paid for the use of air, of heat, and of water, the assistance which they afforded us, adds nothing to value in exchange. (Ricardo 1951, IV, 286–7) A sufficient contraction in the supply of a natural agent would generate a positive price for the service, which evidently would be passed on in turn to consumers by way of output contraction. All this provides a splendid illustration of the fact that scarce labour – in contrast to free natural agents – is relevant to cost price only by way of its impact on commodity supply. We come now (using Marshall’s terms, above, p. 172), to the ‘profound, though very incomplete, discussion, of the difference between Value and Riches’, in which Ricardo ‘seems to be feeling his way towards the distinction between marginal and total utility’: A man is rich or poor, according to the abundance of necessaries and luxuries which he can command; and whether the exchangeable value of these for money, for corn, or for labour, be high or low, they will equally contribute 177

RICARDO - THE NEW VIEW

to the enjoyment of their possessor. It is through confounding the ideas of value and wealth, or riches [,] that it has been asserted, that by diminishing the quantity of commodities, that is to say of the necessaries, conveniences, and enjoyments of human life, riches may be increased. If value were the measure of riches, this could not be denied, because by scarcity the value of commodities is raised; but if Adam Smith be correct, if riches consist in necessaries and enjoyments, then they cannot be increased by a diminution of quantity. (Ricardo 1951, IV, 275–6)11 This passage asserts unmistakably (1) that total utility is lessened by a reduction in supply, i.e. by increased scarcity, and (2) that ‘value’ per unit is raised by this increased scarcity; it follows immediately that the increase in value corresponds to the reduction in total utility, i.e. value measures ‘riches’ – at the margin. Ricardo did not state this definition in so many words. But Marshall’s very careful formulation captures to perfection what is entailed. I conclude with Marshall that if ‘the terse language of the differential calculus’ had been pointed out to Ricardo – had Say recognized the incremental dimension to utility and formulated it in the debate – a place was available in Ricardo’s model for its absorption. The notion of marginal utility did not require a new ‘paradigm’ or logical structure.12 This evaluation turns on Ricardo’s appreciation and widespread application of the scarcity principle; it does not stand or fall with Marshall’s interpretation of the passage regarding Value and Riches, valid though it appears to be. V. SUPPLY–DEMAND ANALYSIS AND COST PRICE Ricardo is frequently said to have rejected the demand–supply analysis of natural or long-run competitive price (cf. Schumpeter 1954, 220, 482, 569– 70, 604, 611, 684, 921). From the above discussion, this is obviously a total misconception. What he actually complained of was ‘the opinion that the price of commodities depends solely on the proportion of supply to demand, or demand to supply’ (1951, I, 382; emphasis added), a complaint alluding to those formulations which appeared to exclude a role for cost conditions in the mechanism. As Marshall pointed out, his chief concern was with costs and the supply side of the market. Thus his observation to Say: ‘You say demand and supply regulates the price of bread; that is true, but what regulates supply? the cost of production’ (IX, 172). The essence of the matter is also captured in a letter to Malthus: ‘You say demand and supply regulates value – this I think is saying nothing ... – it is supply which regulates value – and supply is itself controlled by comparative cost of production’ (VIII, 279). 13 It is worth noting that Robert Torrens (1790–1864) regarded all of this as received doctrine: ‘Political Economists seem on all hands agreed, that the quantity in which 178

WAY MARSHALL WAS RIGHT ABOUT RICARDO

commodities exchange for one another depends, in any given instance, upon the proportion of demand and supply. It is also on all hands agreed, that with respect to all commodities which industry can indefinitely increase, the cost of production is the circumstance which, by limiting the quantity of them brought to market, regulates the proportion of supply to demand, and ultimately determines the exchangeable value’ (Torrens 1936 [1822], 9). In fact Ricardo himself said the same in commenting on a statement by Malthus to the effect that ‘the relation of the supply to the demand ... is the dominant principle in the determination of prices whether market or natural, and that the cost of production can do nothing but in subordination to it, that is, merely as this cost affects ... the relation which the supply bears to the demand’ (in Ricardo 1951, II, 47). On this Ricardo remarked that it is ‘not that I know of disputed by any body’. 14 We turn now to the precise nature of the Ricardian supply adjustments. If one’s concern is capital accumulation or the growth of population, it is often sensible to assume as a first approximation that profit-rate and wage-rate uniformity have both been achieved, i.e. that the various sectors of the economy are throughout in equilibrium. If allocation is the issue we must look at the motivation for the use of factors in one sector rather than another. That hinges on differential wages and differential profits – the fact that capitalists (labourers) have an eye on what their capital (labour) can yield elsewhere. To talk of ‘costs’ is to take account of precisely this aspect of the economic problem. The classical notions of wages and interest as compensation for effort and abstinence respectively, are pertinent at the macroeconomic level where the determinants of aggregate factor supplies are under investigation; in the micro-economic context ‘costs’ reflect alternative opportunities. The cost price analysis of Ricardo is pre-eminently an analysis of the allocation of scarce resources proceeding in terms of general equilibrium. To demonstrate this position let us consider Ricardo’s reaction to a statement by J.B. Say, in his Traité d’économie politique (4th edition, 1819) of mutual interdependence between product and factor markets. This statement incorporates the principle of opportunity cost and also that of imputing the values of factors from the values of their products (or derived demand) – in broad terms only because of the absence of a formal marginal conception whereby the physical contributions of individual factors can be isolated or, indeed, whereby to specify marginal utility: It is utility which determines the demand for a commodity, but it is the cost of its production which limits the extent of its demand. When its utility does not elevate its value to the level of the cost of production, the thing is not worth what it costs; it is a proof that the productive services might be employed to create a commodity of a superior value. The possessors of 179

RICARDO - THE NEW VIEW

productive funds, that is to say, those who have the disposal of labour, of capital or land, are perpetually occupied in comparing the cost of production with the value of the things produced, or which comes to the same thing, in comparing the value of different commodities with each other; because the cost of production is nothing else but the value of productive services, consumed in forming a production; and the value of a productive service is nothing else than the value of the commodity, which is the result. The value of a commodity, the value of a productive service, the value of the cost of production are all, then, similar values when every thing is left to its natural course. (Say; cited in Ricardo 1951, I, 282–3) Léon Walras (1874), who developed the algebraic formulation of general-equilibrium relations, believed Say to have been on the right path by this formulation of general interdependency (1954, 425). 15 But so did Ricardo, who commented on the passage: ‘M. Say maintains with scarcely any variation, the doctrine which I hold concerning value’ (1951, I, 283). He objected only to Say’s treatment of the services of land on a par with those of capital and labour, given his own presumption that rent must be excluded from marginal cost (cf. IX, 172); 16 and to the various other statements by Say which, by relating value solely to utility, seemed to exclude any role for costs: ‘I think more may be said in defence of this doctrine’, he wrote to Malthus regarding Say on services: ‘[They] are I think the regulators of value, and if he would give up rent, he and I should not differ very materially on that subject. In what he says of services he is quite inconsistent with his other doctrine about utility’ (VIII, 277).17 The perception of factor demand as entailing a productivity dimension emerges also, independently of the Say exchanges, in Ricardo’s observations regarding the return to permanent land improvement: ‘a portion only of the money annually to be paid for the improved farm, would be given for the original and indestructible powers of the soil; the other portion would be paid for the use of capital which had been employed in ameliorating the quality of the land, and in erecting such buildings as were necessary to secure and preserve the product’ (I, 67; emphasis added); a little later rent is again defined as ‘that compensation, which is paid to the owner of land for the use of its original and indestructible powers’ (I, 69). 18 A productivity dimension is thus clearly implied in the return to capital as well as to land. VI. ON FACTOR SUBSTITUTION AND MARGINAL PRODUCTIVITY Jevons phrased the ‘Economic Problem’ thus in his Concluding Remarks to the Theory of Political Economy : ‘Given, a certain population, with various needs and powers of 180

WAY MARSHALL WAS RIGHT ABOUT RICARDO

production, in possession of certain lands and other sources of material: required, the mode of (1924, 266–7). employing that labour which will maximise the utility of the product’ It is certainly true that the chief preoccupation of the classics did not turn on the case of given aggregate amounts of the factors and their optimal allocation. But this is equally the case of Marshall. And in any event the economic problem as stated by Jevons constitutes a subset of the much broader concerns of the classics – concerns with growth – which generate complex allocative questions. 19 We can, however, go further. The absence of the formal conception of incremental variation of factor proportions whereby the physical contribution of individual factors may be isolated does not preclude the applicability of Ricardian theory to standard neo-classical problems, particularly allocative and the pricing implications for factors and products of disturbances to demand or to costs within the context of given aggregate factors supplies. Substitution between capital and labour is incorporated by Ricardo by way of variation in the commodity mix, even when ruled out at the technical level. Moreover, we obtain a distorted picture if we forget (1) that the first generation of ‘marginalists’ also lacked the principle of factor substitution or technical variation – though Menger went further, the notion of diminishing marginal product eluded him – and (2) that later neoclassicists were obliged to recognize the limited status of the principle in analysing the properties of generalequilibrium systems. Classical inter-industry substitution was never superannuated. 20 VII. THE LABOUR THEORY OF VALUE AND ALLOCATION ANALYSIS Ricardo’s subscription to Say’s general-equilibrium perspective is no mere formality. The notion of opportunity costs pervades Ricardo’s work, including the labour theory of value. ‘Cost of production’ or ‘natural price’ includes profits as well as wages, each at its average or ordinary rate (Ricardo 1951, I, 291). Assuming uniform factor proportions, a state of general equilibrium such that prices reflect costs throughout the system, will satisfy both profit-rate uniformity and proportionality of prices to labour inputs. More accurately, under the stated circumstances, uniformity of profit rates requires that proportionality. The following passage (drawn from a discussion of subsidized labour for some firms in a manufacturing industry) summarizes the point beautifully, and does so in a context expressing that what is relevant is marginal labour input: ‘The manufacturer enjoying none of these facilities might indeed be driven altogether from the market, if the supply afforded by these favoured workmen were equal to all the wants of the community; but if he continued in the trade, it would be only on condition that he should derive from it the usual and general rate 181

RICARDO - THE NEW VIEW

of profits on stock. and that could only happen when his commodity sold for a price proportional to the quantity of labour bestowed on its production’ (I, 73n). It is the possibility of capital and labour movement between uses, i.e. commoditysupply adjustment, which assures the tendency to cost price and proportionality of prices to labour input. This can be illustrated from the discussion of an exogenous change in tastes, starting out from equilibrium with ‘all commodities ... at their natural price’, implying by this (1) that ‘the profits of capital in all employments are exactly at the same rate’ and (2) that prices are proportionate to labour input (I, 90). Market and natural prices diverge temporarily following the disturbance, but the corrective mechanism of ‘competition’, i.e. supply adjustment, re-establishes equilibrium: It is ... the desire, which every capitalist has, of diverting his funds from a less to a more profitable employment, that prevents the market price of commodities from continuing for any length of time either much above, or much below their natural price. It is this competition, which so adjusts the exchangeable value of commodities, that after paying the wages necessary to their production, and all other expenses required to put the capital employed in its original state of efficiency, the remaining value or overplus will in each trade be in proportion to the value of the capital employed. (Ricardo 1951, I, 91). It is supply adjustment, given the pattern of demand, that assures re-establishment of uniformity of surplus to capital across industries, and proportionality of prices to labour values. Whether or not costs are proportional to labour inputs – this depends on the uniformity or otherwise of factor proportions – only the returns to factors that have alternative uses are allowed for. In circumstances of differential factor proportions, the assumption of factor mobility dictates a divergence of equilibrium cost prices from labour inputs, some prices rising and some falling (by way of appropriate supply adjustments) relative to a commodity produced with mean factor proportions. Ricardo’s potent contribution to allocation theory is reflected in his recognition that an increase in the wage will generate supply expansions (and thus generate lower relative prices) in the case of commodities produced with above-average capital– labour ratios. 21 Here in effect he warns against the myopia engendered by excessive concentration on partial or industry analysis which threatens to yield false results in the case even of the simplest disturbance such as an experimental wage increase. More generally, since ‘it is through the inequality of profits, that capital is moved from one employment to another’ (I, 119), an economy-wide change in labour productivity or any other disturbance impinging equally on all commodities, has 182

WAY MARSHALL WAS RIGHT ABOUT RICARDO

no differential effects on profitability at the initial longrun cost prices and therefore will generate no changes in supply, leaving those prices unchanged. Evidently embodiment of labour (or pain cost of some sort) as such is not the relevant microeconomic consideration. By contrast, a disturbance limited to a single industry, such as a change in input coefficients or a tax or subsidy, will generate an alteration in output and consequent alteration in price, to re-establish the original return on capital in that industry. Applications of the distinction in question extend far and wide, one of the most important bearing upon the fundamental nature of trade: ‘If any cause should raise the price of a few manufactured commodities, it would prevent or check their exportation; but if the same cause operated generally on all, the effect would be merely nominal, and would neither interfere with their relative [cost] value, nor in any degree diminish the stimulus to a trade of barter, which all commerce, both foreign and domestic really is’ (I, 228). In all this Ricardo appreciated the contrast – with an eye to motivation – between allocation theory and the determinants of aggregate factor supply: ‘When profits are universally high, the temptation to produce an increased quantity of commodities is very different from that which a high market price of a particular commodity affords, for the production of that particular commodity’ (II, 372). It is profit-rate differentials that govern resource allocation; the general rate governs accumulation. VIII. COST AND RENT We can sharpen our understanding of Ricardo’s position on cost price by considering the contrast between cost and rent. I shall first establish Ricardo’s awareness that the phenomenon of differential rent, which plays so large a role in his system, is but a special case of a more general phenomenon, namely land scarcity. Rent is provisionally defined as payment to the landlord ‘for the use of the original and indestructible powers of the soil’ (I, 67), a productivity phenomenon. But in that passage already cited alluding to the solution of the ‘paradox of value’ in terms of demand–supply (I, 69; see above, p. 177), the ultimate rationale is more specifically expressed in terms both of productivity and scarcity. And using these principles Ricardo rejected a physiocratic residue in the Wealth of Nations, namely Smith’s assertion that in manufactures ‘ nature does nothing, man does all ; and the reproduction must always be in proportion to the strength of the agents that occasion it’ (cited I, 76n). Factor productivity, Ricardo insisted, is a necessary but insufficient condition for a positive return: ‘Does nature nothing for man in manufactures? ... There is not a manufacture which can be mentioned, in which nature does not give her assistance to man, and give it too, generously and gratuitously.’ 183

RICARDO - THE NEW VIEW

Now differential rent is simply a special case of scarcity rent, falling within the general demand–supply framework: ‘If all land had the same properties, if it were all unlimited in quantity, and uniform in quality, no charge could be made for its use, unless it possesses peculiar advantages of situation’ (I, 69); conversely, ‘[if] air, water, the elasticity of steam, and the pressure of the atmosphere, were of various qualities; if they could be appropriated, and each quality existed only in moderate abundance, they, as well as the land, would afford a rent as the successive qualities were brought into use’ (I, 75). Ricardo also understood that rent would be generated even on marginal units of output where scarcity manifests itself in an extreme form – the functional relation between output and costs terminating, i.e. the supply curve becoming vertical; and he traced through some of the analytical consequences – typically ‘Marshallian’ consequences – for taxation: The corn and raw produce of a country may, indeed, for a time sell at a monopoly price; but they can do so permanently only when no more capital can be profitably employed on the lands, and when, therefore, their produce cannot be increased. At such time, every portion of land in cultivation, and every portion of capital employed on the land will yield a rent, differing, indeed, in proportion to the difference in the return. At such time too, any tax which may be imposed on the farmer, will fall on rent, not on the consumer. He cannot raise the price of his corn because by the supposition, it is already at the highest price at which the purchasers will or can buy it. He will not be satisfied with a lower rate of profits, than that obtained by other capitalists, and, therefore, his only alternative will be to obtain a reduction of rent or to quit his employment. (Ricardo 1951, I, 250–1) This analysis presumes throughout a negatively sloped demand curve. It is scarcely surprising, in the light of all this, to find Ricardo maintaining that it is ‘from the price at which the produce is sold, that rent is derived; and this price is got not because nature assists in the production, but because it is the price which suits the consumption to the supply’ (I, 77n). Rent appears even in the price covering the last unit, if the demand curve is high enough, as a pure demand-determined surplus. In the usual, less extreme, cases rent emerges entirely as producer’s surplus, dependent upon the location of the demand curve and is not paid on marginal units. (Rent also appears within cost price if alternative land uses are recognized; see note 16.) IX. FURTHER APPLICATIONS The tools of allocation theory outlined above, it can be shown, govern Ricardo’s socalled ‘fundamental theorem of distribution’, or inverse wage–profit relation. (This 184

WAY MARSHALL WAS RIGHT ABOUT RICARDO

applies also to Marxian economics.) They are used by Ricardo to analyse the process of reabsorption into the economic system of factors displaced by various disturbances including new technology and taxation; and conversely, the impact of disturbances which attract factors into particular sectors. In these applications we find incorporated the law of markets, as part of the full-fledged general-equilibrium system. And, of course, there is the comparative-cost approach to trade which is pre-eminently an exercise in allocation theory. (On these matters, see Hollander 1987a, ch. 5.) But my purpose will have been satisfied if I have been able to convey the elementary essentials of these applications. X. WAGE VARIABILITY AND ‘SURPLUS’ A principal issue that arises in discussions of Ricardian growth theory is the variability or fixity of the average wage. Marshall insisted on the former. 22 And so do I, for Ricardo’s model, involving population and capital expansion given land, requires a variable wage (Hollander 1987a, ch. 8). This issue has the most profound implications for nineteenth-century historiography, having in mind the Italo-Cambridge perspective outlined earlier. That perspective I find to be almost entirely based on wishful thinking. Demand– supply analysis is the basis for Ricardian cost prices (this I have demonstrated above) and also for the average returns to labour and capital (see preceding reference). But to take this view undermines the whole concept of interest as ‘surplus’ and points towards the Marshallian position according to which ‘in the long run, the earnings of each agent are, as a rule, sufficient only to recompense at their marginal rates the sum total of the efforts and sacrifices required to produce them’ (Marshall 1920, 832). All this must be placed in appropriate context. The formal conception of the source of profits in surplus – actually surplus labour time – is adopted by J.S. Mill (not only Marx): the reason why capital yields a profit, is because food, clothing, materials, and tools, last longer than the time which was required to produce them; so that if a capitalist supplies a party of labourers with these things, on condition of receiving all they produce, they will, in addition to reproducing their own necessaries and instruments, have a portion of their time remaining, to work for the capitalist. We thus see that profit arises, not from the incident of exchange, but from the productive power of labour; and the general profit of the country is always what the productive power of labour makes it, whether any exchange takes place or not. If there were no division of employments, there would be no buying or selling, but there would still be profit. If the labourers of the country collectively produce twenty percent 185

RICARDO - THE NEW VIEW

more than their wages, profits will be twenty per cent, whatever prices may or may not be. The accidents of price may for a time make one set of producers get more than the twenty per cent, and another less, the one commodity being rated above the natural value in relation to other commodities, and the other below, until prices have again adjusted themselves; but there will always be just twenty per cent divided among them all. (Mill 1965, II, 411; first introduced into the 4th edn, 1857) This passage conveys the notion that productivity must be sufficiently high to leave part of labour’s time free for the production of commodities other than wage goods. Ricardo broaches the same issue but indirectly in formulating the inverse wage–profit relation – the notion that the profit rate turns on the wage only, insofar as the latter reflects the share of labour in the product to be distributed between labour and capital. Under ideal conditions relating to the invariable measure this theorem can be specified in terms of the fraction of the work day devoted to the production of wage goods (Hollander 1987a, 111), though Ricardo (and Mill) stood by the theorem even if expressed in terms of real national income, i.e. income of constant purchasing power (114).23 Assume now that the real wage and thus the cost of producing labour is a variable. The notion of surplus then breaks down, for we can say nothing about the ‘surplus’ until the ‘necessary’ labour time (the commodity wage corrected for productivity change) is specified. But the rate of capital accumulation (and thus of labour demand) can vary, and with it the commodity wage. It will vary, according to Ricardo, with variation in the return on capital: ‘The motive for accumulation will diminish with every diminution of profits’ (Ricardo I, 111), a perspective fully consistent with the theory of abstinence formally expressed by Senior, Mill, and later Marshall. To this Marx reacted harshly, for he realized the implication: It is incomprehensible how economists like John Stuart Mill, who are Ricardians and even express the principle that profit is equal to surplusvalue, surplus labour, in the form that the rate of profit and wages stand in inverse ratio to one another and that the rate of wages determines the rate of profit (which is incorrect when put in this form), suddenly convert industrial profit into the individual labour of the capitalist instead of into the surplus labour of the worker, unless the function of exploitation of other people’s labour is called labour by them. (Marx 1971, [1862–3] III, 506) To isolate the source of profit in surplus labour time does not ( pace Marx) rule out the conception of interest as a necessary reward. That productivity assures an excess over the output consumed by labour is the basis for investment demand; 186

WAY MARSHALL WAS RIGHT ABOUT RICARDO

while abstinence relates to capital-supply conditions. However, since the wage rate is a variable partly governed by capitalsupply conditions, it follows that the ‘necessary’ part of the work day is as much a consequence as a determinant of the ‘surplus’ part. As Ricardo put the matter, there is a ‘natural equilibrium between profits and wages’ (1951, I, 226). We are in a Marshallian world. The undermining of any meaningful notion of surplus is compounded if we allow interdependence between distribution and final expenditure patterns. Ricardo recognized such a two-way relation: an impact of distribution on expenditure patterns and a reverse impact of expenditure patterns on distribution (see Hollander 1989 [below, Chapter 11]); so for that matter did Marx, who indeed insisted that expenditure patterns are governed by income distribution (Hollander 1987a, 370–1). In this case it is misleading to assert – as Marx and Cambridge writers do assert – that knowledge of the wage allows the ‘prediction’ of the profit-rate and the set of equilibrium cost prices; for the wage cannot be ‘fixed’ independently of the pattern of final demand. The notion of the allocation of a ‘pre-existing’ surplus across industries by a process of ‘circulation’ such as Marx purports to undertake in his ‘transformation of values into prices’ is invalid. 24 In brief, to allow that a market process is involved merely in the equalization of returns on capital, while the average profit rate is pre-established requires that the strategic variables be given exogenously – pre-eminently the commodity wage per unit of time (and productivity) thereby fixing the ‘cost of producing labour’. This requirement is ruled out by Ricardo’s (and also Marx’s) assumption of variable wages dependent (a) on the rate of capital accumulation and therefore on the return on capital, and (b) on the pattern of expenditure. Our demonstration in this section thus reinforces the notion that Ricardian theory cannot be represented in terms of Cambridge one-way causal relations. XI. SUMMARY AND CONCLUSION It is unjustified to say of Ricardo that he had little comprehension of the way markets work and relative prices operate in allocating resources. Neither can it be said that there is a crucial sense in which distribution is ‘prior to’ exchange; the Ricardian system is one of mutual dependence between distribution and pricing. Marshall was then right to reject the Jevonian (and Walrasian) call for a postclassical ‘reconstruction’ of economic theory, on the grounds that ‘the foundations of the theory as they were left by Ricardo remain intact; that much has been added to them, and that very much has been built upon them, but that little has been taken from them’ (1920, 503). As I see it, there occurred after 1870 a narrowing of focus from economic growth to exchange and allocation in their own right, and more positively, the elaboration of mathematical theories of the maximizing consumer 187

RICARDO - THE NEW VIEW

(and ultimately the firm) and general-equilibrium relations. The marginalists made spurious analytical distinctions failing to recognize how much the classics had contributed to the theory of allocation; and thus they also failed to represent their own distinctive contribution to individual choice analysis as a sup-plement to or strengthening of that theory. 25 As for the matter of marginal utility, Marshall put it accurately in his comment that Jevons added very important ‘explanations’ but no more. These explanations were, I would add, as much pertinent to Say’s account as to Ricardo’s. What of the marginal productivity principle, adopted a generation after the so-called ‘marginal revolution’ itself? This certainly clarified the understanding of factor demand by the firm (and thus the industry) in particular uses. But again the elaboration was required as much by Say as by Ricardo and Mill. Moreover, it is historically unjustified to assert that recognition of a multiplicity of primary factors required a new theory based upon the principles of demand, or that classical theory could not explain the relative wages of heterogeneous types of labour (see above, p. 169). Since Smith’s time and before the wage structure had been analysed in demand– supply terms. Smith and his successors had all recognized a productivity dimension on the demand side of the labour market, although in dealing with the wage structure they typically chose to focus on the supply side; and while value productivity is more conspicuous the more specialized to a particular use are individual factors, those same considerations are equally relevant when allowance is made for factor mobility between uses – the characteristic classical allowance – although now strict limits are placed on the extent to which returns in different uses can diverge. Marshall, we have seen, protested against that over-reaction from Ricardo which involves a wholly generalized application of rent theory – what some call today ‘neo-neo-classicism’. At the same time, it remains true that Ricardo (and Mill) appreciated the technical implications of factor specificity, as is clear from the various applications of the rent doctrine discussed above.26 NOTES *Public lecture given at Tinbergen Institute, Erasmus University, Rotterdam, Holland, April 11, 1989. This paper was published in a Spanish translation ‘Mercados precios y distribucion: por que Marshall estaba en lo correcto con respecto a Ricardo’, Economia (Pontificia Universidad Catolica del Peru), 13 (June 1990): 9–46. An abbreviated version appears as ‘Alfred Marshall in Historical Perspective: Why Marshall was Right about Ricardo’, European Economic Review 35 (May 1991): 313–22. 1. Schumpeter took precisely the opposite view to mine: ‘No unbiased reader can fail to perceive ... that Marshall’s theoretical structure, barring its technical superiority and various developments of detail, is fundamentally the same as that of Jevons, Menger and especially Walras, but that the rooms in this new house are unnecessarily cluttered up with 188

WAY MARSHALL WAS RIGHT ABOUT RICARDO

2.

3.

4.

5.

6.

7.

Ricardian heirlooms, which receive emphasis out of proportion to their operational importance’ (1954, 837). The contrast was suggested to me by Professor David Laidler who sought to understand the mutual incomprehension of protagonists in the debates of the past decade regarding ‘classical’ economics. Frank Knight, the arch-critic of Ricardo, had high praise for Adam Smith’s adjustment process involving the transfer of resources ‘from less to more remunerative uses until the remuneration in all competing fields is equalized’ (1956, 64). The force of his hostility blinded Knight to the fact that Ricardo readily adopted Smith’s analysis. Sir John Hicks has argued for the replacement of the term ‘marginalism’ by ‘catallactics’, admitting that ‘Ricardo himself could be quite marginalist at times’ (1974, 212). My case is that Ricardo was both ‘marginalist’ and ‘catallactician’, envisaging the economic system as a system of interrelated markets. (Hicks contrasts ‘catallactics’ with ‘plutology’ or the science of wealth, drawing a line between neo-classicism and classicism in these terms; but this neglects that Walras, for one, identified pure economics with the theory of social wealth on the grounds that ‘the sum total of all things, material or immaterial, on which a price can be set because they are scarce (i.e. both useful and limited in quantity ) constitutes social wealth’ (1954, 40). My position seems close to that of Blaug (1976) though I would wish to avoid linking Menger too closely with Jevons and Walras: ‘the innovations of Menger, Jevons and Walras are more suitably described, not as a new SRP [scientific research programme], but as a “progressive problem-shift” in the older research programme of classical political economy’ (165; see also p. 161 regarding a common ‘hard core’). I have been unable to pin down with absolute assurance Blaug’s position from his numerous other statements regarding the nature of the ‘marginal revolution’, although I have the impression that he continues to maintain that it amounted to a ‘shift of emphasis’. (On this issue see Hollander 1987b.) A recent volume devoted to demonstrating the opposing view (Fisher 1986) is vitiated by the conspicuous neglect of the nature of classicism. Marshall cited the German historicist Roscher to the same effect: ‘As Roscher says ( Political Economy, Sect. CLV), “In judging Ricardo, it must not be forgotten that it was not his intention to write a text-book on the science of Political Economy, but only to communicate to those versed in it the result of his researches in as brief a manner as possible. Hence he writes so frequently making certain assumptions, and his words are to be extended to other cases only after due consideration, or rather re-written to suit the changed case” ’ (1920, 163n). See also Marshall’s comment regarding Ricardo’s misleading modes of expression (503, 761n, 834). On the opening page of the Theory of Political Economy Jevons asserted: ‘Repeated reflection and inquiry have led me to the somewhat novel opinion, that value depends entirely upon utility ’ (1924, 5). Similar charges appear in Marshall’s famous Academy review (1872) of The Theory of Political Economy : Jevons ‘was so encumbered by his mathematics in his central argument, that he tried to draw nature’s actions out into a long queue’ (1925, 99). On Jevons’s position that ‘the wages of a working man are ultimately coincident with what he produces after the deduction of rent, taxes and the interest on capital’, Marshall notes: ‘He does not see that, since rent, taxes, etc. are not paid in kind, we must have before us a complete theory of value in order that we may perform this subtraction. He 189

RICARDO - THE NEW VIEW

8.

9.

10.

11. 12.

does not speak of the amount of the wages, and the exchange value of the products as varying elements, the variations of each of which affect those of the other’ (94) . Marshall lists the assumptions of equal skill, equal capital–labour ratios including the period of production and equal profit rates. This latter is problematic, since Ricardo never allows differential profit rates as a source of deviation of prices from values (see also Marshall 1920, 814–15). The Ricardian perspective points directly away from ‘old-style’ Soviet doctrine. T.W. Hutchison mistakenly found the source of that doctrine in Ricardian classicism: ‘It was, of course, in Marxist economic theorizing that some of the old ideas and concepts of English “classical” orthodoxy continued a form of existence. Certainly, in the economic regimes of Eastern Europe, claiming somewhat questionably to be following out the economic theories of Marx, the neglect of consumer demand, utility, and choice of the earlier English theories had a twentieth-century practical, political counterpart’ (1972, 468). The failure to specify an incremental dimension to utility, varying in magnitude with quantity, may be partly explained in definitional terms. For if total utility is identified with ‘riches’ – and thus linked in a one-to-one relationship – the isolation of a (variable) marginal utility is precluded. This seeming ‘identification’ emerges, for example, in discussion of an individual who, with his resources, commands twice the amount of corn upon a halving of cost price: ‘If two sacks be of the value that one was before, he evidently obtains the same value and no more, – he gets, indeed, double the quantity of riches – double the quantity of utility – double the quantity of what Adam Smith calls value in use, but not double the quantity of value, and therefore M. Say cannot be right in considering value, riches, and utility to be synonymous’ (Ricardo 1951, I, 281). But a one-to-one relation is precluded by Ricardo’s own solution to the paradox of value – should corn increase sufficiently some of it (like water) will be actually discarded as useless. Cf. Ricardo 1951, IX, 169: ‘I ... do not estimate riches by value, but by the whole quantity of utility which the commodities which constitute riches possess.’ I have phrased this with deliberation. It is impossible to say whether Ricardo would have wished to formulate a law of demand based on diminishing marginal utility. To do so would, of course, have entailed all the later problems of relating marginal utility to demand price. Moreover, Ricardo himself pointed out that ‘Every man has some standard in his own mind by which he estimates the value of his enjoyments, but that standard is as various as the human character’ (1951, I, 241); ‘value in use cannot be measured by any known standard; it is differently estimated by different persons’ (I, 429). Although Jevons objected that Cournot ‘does not recede to any theory of utility, but commences with the phenomenal laws of supply and demand’ (1924, xxxi), he himself admitted: ‘I hesitate to say that man will ever have the means of measuring directly the feelings of the human heart. A unit of pleasure or of pain is difficult even to conceive’ (11). And it was, ‘in practice’, an average or representative individual which concerned him – an important point casting doubt on Stigler’s representation of the ‘law of utility’ in the form: ‘the marginal utility of every commodity diminishes for every man, and this phenomenon underlies his demand curve for each commodity’ (1972, 579; emphasis added). Jevons wrote as follows: I must here point out that, though the theory presumes to investigate the condition of a mind, and bases upon this investigation the whole of Economics, practically it is an aggregate of individuals which will be treated. The general forms of the laws of Economics are the same in the case of individuals and nations; and, in reality, it is a law operating in

190

WAY MARSHALL WAS RIGHT ABOUT RICARDO the case of multitudes of individuals which gives rise to the aggregate represented in the transactions of a nation. Practically, however, it is quite impossible to detect the operation of general laws of this kind in the actions of one or a few individuals. The motives and conditions are so numerous and complicated, that the resulting actions have the appearance of caprice, and are beyond the analytic powers of science. ... It would be by examining the average consumption of sugar in a large population that we might detect a continuous variation, connected with the variation of price by a constant law. (Jevons 1924, 14–15)

This perspective lends support for the opinion that Jevons’s representation of economics as a science of pleasures and pains was a ‘façon de parler’; his concern was the mathematical logic of the maximization process not the psychological assumptions of hedonism as such (see Black 1972, 372–4). Similarly, Jaffé emphasizes Walras’s concern with marketplace satisfactions thereby avoiding the need for a true theory of consumption (cf. Jaffé 1972, 384–5). Apart from this sort of complexity – perhaps because of it – Ricardo might have been happy enough to stay with an empirical law of market demand reflecting the principle of satiety and generating a theory of scarcity value into which even the cost perspective is incorporated. There was no burning need to go further. That to my mind is one reason why J.S. Mill ‘failed’ to enunciate the principle of diminishing marginal utility. In Mill’s case there is also evidence of revealed-preference reasoning which avoids many of the difficult psychological issues raised by a formal utility approach. 13. Against Schumpeter’s position, Viner makes the following point: In using demand-and-supply terminology for the determination of ‘market price’, that is, actual price, or temporary price, or instantaneous price, but rejecting it in his explanation of ‘natural price’, Ricardo was not innovating. This practice goes back to the seventeenth century at least. It can be justified on the ground that it was semantically unfortunate that it later on became common to use the same term ‘demand’ (and correspondingly for ‘supply’) both for the quantity that would actually be taken in a given historical market at a given actual price in a given actual moment of time and for ‘that highly abstract creation of the observer’s mind’, the long-run normal demand function. (Viner 1958, 357)

14. Viner has this to say on this episode: Terminology aside, there did not exist that dual line of price analysis, Ricardian and Malthusian, which Schumpeter insists upon. If there was any significant difference between the two, it was that Ricardo’s concentration on constant-cost cases kept him inadvertently from working out an adequate apparatus for explaining the determination of long-run price where both quantity demanded and quantity offered were variables dependent on price. It does not follow, in the absence of supporting evidence, that if such a case were presented to Ricardo, or to any one of his followers, he would have handled it any differently than Malthus. (Viner 1958, 358)

15. Walras was nonetheless critical; Say traced the origin of value to utility which was insufficiently precise – scarcity was the key (1954, 201). This was to be Ricardo’s complaint against Say as we shall shortly see. 16. Ricardo’s exclusion of rent from costs turns on an implicit assumption of one-use land. Where he abandons this assumption rent enters into costs – as for Smith – with an eye to alternative opportunities (1951, I, 252). I am grateful to my undergraduate student Susan Ecclestone for reminding me of this particular text. 17. In his posthumously published paper ‘Absolute and Exchangeable Value’ Ricardo expresses the source of the returns to the factors labour and capital very clearly: ‘One class gives its labour only to assist towards the production of the commodity and must be paid out of its value the compensation to which it is entitled, the other class makes the advances 191

RICARDO - THE NEW VIEW

required in the shape of capital and must receive remuneration from the same source’ (1951, IV, 365). 18. Cf. Essays on Profits, Ricardo 1951, IV, 18n: ‘If either the landlord expends capital on his own land, or the capital of a preceding tenant is left upon it at the expiration of his lease, he may obtain what is indeed called a larger rent, but a portion of this is evidently paid for the use of capital. The other portion only is paid for the use of the original power of the land.’ 19. The point has been nicely made by Jacob Viner in his answer to Schumpeter’s negative overview of Ricardo’s economics: Many of Schumpeter’s ... critical comments on Ricardo’s analysis lose their point if Ricardo’s major concern in his value theorising was not the explanation of how a given structure of prices had come to be what it was but the explanation of (a) the effect on a given structure of prices of divergent changes in the amounts of the respective factors, and (b) the effect on the relative amounts of the factors of changes in the structure of prices. This interpretation of Ricardo involves the question of the role of the supply functions of the factors in ‘Ricardian’ as contrasted with ‘Austrian’ theory. (Viner 1958, 358; see also 361–2)

20.

21.

22.

23.

24.

Bearing in mind our demonstration that changes in the price structure for Ricardo turn on output adjustments, what we have here is recognition of the interdependence in Ricardo’s system between distribution and expenditure patterns, namely, (a) the effect on the structure of outputs of alternative changes in aggregate factor supply; and (b) the effect on the relative amounts of the factors – and therefore the average returns to the factors – of changes in the output structure. We may add to this that Ricardo did, in fact, recognize technical capital–labour substitution in the growth context: ‘every rise of wages will have a tendency to determine the saved capital in a greater proportion than before to the employment of machinery. Machinery and labour are in constant competition, and the former can frequently not be employed until labour rises’ (1951, I, 395; cf. 266). Lord Robbins refers to the demonstration as reflecting ‘the insight of genius’ and ‘one of the intellectual triumphs of his system’ (1970, 20). This mechanism assures that even in the absence of factor substitution the economywide demand for labour will have a negative slope. ‘The persistency with which many writers continue to attribute to him a belief in the “iron law” can be accounted for only by his delight “in imagining strong cases”, and his habit of not repeating a hint, which he had once given, that he was omitting for the sake of simplicity the conditions and limitations that was needed to make his results applicable to real life’ (Marshall 1920, 509). The intention of Mill’s formulation given above (p. 185) was to escape the ‘mercantilist’ notion that profits could be accounted for by reference to ‘the incident of exchange’ or an excess of sales price over costs – profits on alienation as Marx called it. National income is constrained in real terms and the profit rate is governed by the wage insofar as that wage reflects the proportionate share of output going to labour. From this perspective the source of profits is not in surplus labour time as such ; but rather in surplus output or real income. And in the case of a single-product world there could be no need to bring in labour, i.e. the classical themes can be specified, in elementary forms, in commodity terms. Marx’s procedures in Capital are totally misleading. Identification of the industries in the value scheme of Capital, Volume I, requires information derived from the ‘sphere of 192

WAY MARSHALL WAS RIGHT ABOUT RICARDO

circulation’. The problem is further aggravated by the impossibility even of defining the labour input without having recourse to a market process for the reduction of different qualities of labour to some base unit. 25. There is the further point that when Jevons and Walras approached the ‘long run’ problems of cost price and growth they reverted to the classical solutions; see Hollander 1987a, 430–4. It is interesting to find that Ronald Meek, writing from a (post-Stalinist) Marxian perspective, retracted some earlier views on this matter: ‘The idea that economics must always necessarily be confined to the economics of scarcity was not really an inherent part of the new [post 1870] philosophy’ (1972, 505). For a new study of Jevons as growth theorist, see Peart 1990. 26. According to a famous interpretation of Ricardo by Knut Wicksell, there is discernible in the Principles a conception of the greater productivity of longer ‘construction periods’ – extensions of which are determined by money-wage movements – wherein the rate of interest is governed by the declining (marginal) product of capital. According to this view, an increase in the wage rate encourages extensions of capital investment which in turn brings about a fall in the rate of return. (On Wicksell’s interpretation, see Stigler 1941, 284.) It is in this causal sense, runs the contention, that Ricardo’s dependency of profits upon wages must be understood. On this view the general profit rate is determined in the Ricardian structure by marginal productivity considerations. There is something to be said for Wicksell’s interpretation. For Ricardo, a new, lower, equilibrium profit rate comes into effect after a reallocation of resources set in motion by an initial rise in the wage rate. This reallocation entails an increase in the average capital–labour ratio which may, following Ricardo’s own practice, be identified with an extension of the period of production: ‘all the questions of fixed capital come under the second rule’ – alluding to ‘the relative times that must elapse before the result of ... labour can be brought to market’ (letter to McCulloch, Ricardo 1951, VIII, 180; cf. 193). A word of caution is, however, in order. Ricardo nowhere formulated a specific statement of the effects on output of extensions in the time-period of production. And there is certainly no specific statement of a declining (marginal) product of such extensions. My point is that here is an instance where one leading ‘neo-classical’ considered his approach to interest rate determination as falling on a line of approach originating with Ricardo.

REFERENCES Arrow, K.J. and D.A. Starrett (1973) ‘Cost and Demand Theoretical Approaches to the Theory of Price Determination’ in J.R. Hicks and W. Weber (eds) Carl Menger and the Austrian School of Economics, Oxford: Oxford University Press, 129–48. Black, R.D.C. (1972) ‘W.S. Jevons and the Foundation of Modern Economics’, History of Political Economy 4: 364–78. Blaug, M. (1976) ‘Kuhn versus Lakatos or Paradigms versus Research Programmes in the History of Economics’, in S.J. Latsis (ed.) Method and Appraisal in Economics, Cambridge: Cambridge University Press, 149–80. de Vivo, G. (1987) ‘Ricardo’, The New Palgrave: A Dictionary of Economics, London and Basingstoke: Macmillan, 4, 183–98. Dobb, M. (1973) Theories of Value and Distribution since Adam Smith, Cambridge: Cambridge University Press. 193

RICARDO - THE NEW VIEW

Fisher, R.M. (1986) The Logic of Economic Discovery: Neoclassical Economics and the Marginal Revolution, New York: New York University Press. Frisch, R. (1981) ‘From Utopian Theory to Practical Applications: the Case of Econometrics’, American Economic Review 71: 1–16. Hicks, John (1974) ‘ “Revolutions” in Economics’, in S.J. Latsis (ed.) Method and Appraisal in Economics, Cambridge: Cambridge University Press, 207–18. Hollander, S. (1979) The Economics of David Ricardo, Toronto and London: University of Toronto Press. —— (1987a) Classical Economics, Oxford: Basil Blackwell. —— (1987b) ‘On a Severe Case of Schizophrenia in the History of Economic Thought’,History of Economic Thought Newsletter 38: 17–19. —— (1989) ‘On Composition of Demand and Income Distribution in Classical Economics’, History of Economics Society Bulletin 11 (Fall): 216–21. Hutchison, T.W. (1972) ‘The “Marginal Revolution” and the Decline and Fall of English Classical Political Economy’, History of Political Economy 4: 442–68. Jaffé, W. (1972) ‘Léon Walras’s Role in the “Marginal Revolution” of the 1870s’, History of Political Economy 4: 379–405. Jevons, H.S. (1911) Appendix I to W.S. Jevons (1924). Jevons. W.S. (1924) [1871] The Theory of Political Economy, 4th edn, London: Macmillan. Knight, F.H. (1956) On the History and Method of Economics, Chicago: University of Chicago Press. Marshall, A. (1920) Principles of Economics, 8th edn, London: Macmillan. —— (1925) Memories of Alfred Marshall, ed. A.C. Pigou, London: Macmillan. Marx, K. (1971) [1862–3] Theories of Surplus Value, III, Moscow: Progress Publishers. Meek, R.L. (1972) ‘Marginalism and Marxism’, History of Political Economy 4: 499– 511. Mill, J.S. (1965) Principles of Political Economy, Collected Works, vols II, III, Toronto: University of Toronto Press. Peart, S. (1990) ‘The Population Mechanism in W.S. Jevons’s Applied Economics’, The Manchester School of Economic and Social Studies 58: 32–53. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Robbins, L.C. (1970) The Evolution of Modern Economic Theory, London: Macmillan. Schumpeter, J.A. (1954) History of Economic Analysis, Oxford: Oxford University Press. Stigler, G.J. (1941) Production and Distribution Theories: The Formative Years, New York: Macmillan. —— (1972) ‘The Adoption of the Marginal Utility Theory’,History of Political Economy 4: 570– 86. Torrens, R, (1936) [1822] ‘Three Editorial Notes on Value, Contributed to The Traveller in December 1822’, in Jacob Hollander (ed.) Two Letters on the Measure of Value by Jhons Stuart Mill. A Reprint of Economic Tracts, Baltimore: The Johns Hopkins Press. Viner, J. (1958) The Long View and the Short, Glencoe, Ill.: Free Press. Walras, L. (1954) [1874] Elements of Pure Economics, ed. W. Jaffé, 4th definitive edn (1926), London: George Allen and Unwin.

194

11 ON COMPOSITION OF DEMAND AND INCOME DISTRIBUTION IN CLASSICAL ECONOMICS

I. In several reviews of my Classical Economics (1987; henceforth CE) a criticism recurs relating to my proposition that distribution in Ricardian economics is dependent upon the pattern of final demand. Anthony Brewer, who is convinced by the demonstration in the book of ‘a fundamentally important core of general equilibrium economics accounting for resource allocation in terms of the rationing function of relative prices’, has stated the objection fairly and his formulation invites and deserves a response: [Hollander] does overstate his case at times. For example, he claims that, in Ricardo’s theory, changes in the pattern of demand should react on the demand for labour, and thus on wages, while admitting that ‘Ricardo himself never formally made’ this extension [ CE, 104]. He later uses exactly this interaction of demand and wages to support his interpretation of Ricardo against Dobb [ CE,360]. Surely, the fact that Ricardo did not ‘formally make’ this point (i.e., did not make it at all) is an argument against Hollander’s reading, not for it. (Brewer 1988, 555) In approaching this criticism I must lay out the context of my claim. That relates to the impact of various disturbances on distribution by way of the effect on aggregate labour demand. Ricardo in his 1821 chapter ‘On Machinery’ clarifies that aggregate labour demand cannot be taken as proportional to total capital accumulation considering variations of the ‘fixed– circulating’ breakdown, variations which may be either exogenous (Ricardo 1951, I, 386–90), 1 or endogenous: ‘every rise of wages will have a tendency to determine the saved capital in a greater proportion than 195

RICARDO - THE NEW VIEW

before to the employment of machinery. Machinery and labour are in constant compe-tition, and the former can frequently not be employed until labour rises’ (I, 395). The ‘extension’ I refer to ( CE,104) relates to an exogenous change in the fixed– circulating capital ratio reflecting not technology but altered tastes assuming nonuniform factor proportions. The point in CE is that Ricardo himself, by introducing the impact on distribution of variation in the components of total capital in one application, opened the door for his readers to undertake further applications of a similar order within the framework of his own ‘system’. This concern with ‘systems’ is brought out in the penultimate paragraph of the book: ‘There is nothing “anti-classical” in adding that changes in the structure of industry ... may play back on the wage by influencing labour market conditions in so far as such changes affect the breakdown between constant and variable capital’ (CE, 439). And it lies behind my statement that ‘there is nothing in Ricardian logic to preclude a playback from the pattern of activity upon distribution’ ( CE, 360–1). But is it true that Ricardo never addressed the issue of the impact of expenditure patterns on distribution? The fact is that he did – though not in dealing with ‘a change of fashion [which] should increase the demand for silks and lessen that for woollens’ (CE,95 regarding Ricardo 1951, I, 90–1). That analysis proceeds in Smithian fashion on the assumption that the general levels of wages and profits are unaffected by the adjustment of market to cost prices. The impact of expenditure patterns on distribution emerges rather in two other (closely related) contexts. The first involves the positive effect on aggregate labour demand of a switch in expenditure from ‘luxury’ goods to services and conversely in the reverse case: If my revenue were 10,000 l., the same quantity nearly of productive labour would be employed, whether I realised it in fine clothes and costly furniture, &c.&c. or in a quantity of food and clothing of the same value. If, however, I realised my revenue in the first set of commodities, no more labour would be consequently employed: – I should enjoy my furniture and my clothes, and there would be an end of them; but if I realised my revenue in food and clothing, and my desire was to employ menial servants, all those whom I could so employ with my revenue of 10,000 l., or with the food and clothing, which it would purchase, would be added to the former demand for labourers, and this addition would take place only because I chose this mode of expending my revenue. As the labourers, then, are interested in the demand for labour, they must naturally desire that as much of the revenue as possible should be diverted from expenditure on luxuries, to be expended in the support of menial servants. (Ricardo 1951, I, 393) 2 The proposition is applied by Ricardo to wartime expansion of employment and 196

ON COMPOSITION OF DEMAND

also to postwar contraction of employment and reduced wages: ‘At the termination of the war, when part of my revenue reverts to me, and is employed as before in the purchase of wine, furniture, or other luxuries, the population which it before supported, and which the war called into existence, will become redundant, and by its effect on the rest of the population, and its competition with it for employment, will sink the value of wages, and very materially deteriorate the condition of the labouring classes’ (I, 393–4). From this it followed that ‘the labouring class have no small interest in the manner in which the net income of the country is expended’ (I, 392); ‘they must naturally desire that as much of the revenue as possible should be diverted from expenditure on luxuries, to be expended in the support of menial servants’ (I, 393). A recent article in this Bulletin by Professor Paul Samuelson has a pretty footnote on this issue: ‘Good Whig History also refuses to whitewash an author just because he has momentarily lapsed into good sense. Ricardo recognized in his famous new chapter on machinery that composition of demand can alter the distribution of income – as when the Napoleonic wars shifted demand from civilian goods to more labor-intensive warfare. Cheers. That perception does not extenuate but rather rebuts the fallacious classical and neo-Ricardian attempt to separate the distribution of income from the process of value and demand ’ (1988, 164n). Ricardo’s recognition that composition of demand can alter income distribution is here treated as an instance of ‘momentary good sense’. To my mind Professor Samuelson is trapped by a habit – fortunately not incurable – of confounding Ricardo with the appallingly misnamed ‘neo-Ricardians’. I see the allowance as lending very considerable support to my argument that ‘there is nothing in Ricardian logic to preclude a playback from the pattern of activity upon distribution’. 3 The second case relates to alternative capital compositions induced by alternative taste patterns, and provides an extraordinarily clear statement of the impact of such patterns on labour demand. It is, in fact, what I thought was my ‘extension’ precisely. The analysis is found in the posthumously published Notes on Malthus and supplements the case mentioned above of induced variation in capital composition recognized in the Principles : The effective demand for labour must depend upon the increase of that part of capital, in which the wages of labour are paid. If I have a revenue of £2000 – in the expenditure of that revenue I necessarily employ labour. If I turn this revenue into capital, I at first employ the same labour as before, but productively instead of unproductively. This labour may be employed in making a machine, the machine becomes a capital, and all that it produces is the revenue derived from that capital. Or this labour may be employed on the land, and the corn which it produces may be a capital to enable me to 197

RICARDO - THE NEW VIEW

employ an additional quantity of labour. A society does one or the other in proportion to the demand for either the objects of men’s work; or for objects which are almost exclusively produced by machinery : – in general the capital accumulated will consist of a mixture of both, of fixed and of circulating capital ... [To] the capitalist it can be of no importance whether his capital consists of fixed or of circulating capital, but it is of the greatest importance to those who live by the wages of labour; they are greatly interested in increasing the gross revenue, as it is on the gross revenue that must depend the means of providing for the population. If capital is realized in machinery, there will be little demand for an increased quantity of labour, – if it create an additional demand for labour it will necessarily be realized in those things which are consumed by the labourer. (Ricardo 1951, II, 234–6; emphasis added) So explicit is the role of final expenditure patterns in determining aggregate labour demand that I have to explain why I did not reinforce my argument in the book itself at least by reference to the case noted by Professor Samuelson. The simple response is that I should have done so. I was excessively cautious, concentrating unduly upon the fact that in the discussion of the tendency of market to cost prices the wage and profit rates are assumed unchanged throughout. To attribute to Ricardo explicit appreciation of an impact on distribution of altered expenditure between commodities did not seem justified solely on the basis of the case entailing reallocation of demand between commodities and services. I now see that such caution was unnecessary, considering Ricardo’s own confirmation in the Notes on Malthus that capital composition and thus aggregate labour demand can be affected by the pattern of final demand for commodities. 4 Ricardo himself has, from the grave, confirmed my interpretation and the validity of my ‘extension’. The principle of valid extensions remains intact. I would, therefore, like to suggest another. Consider Ricardo’s recognition in correspondence of a possible impact of variations in distribution on the pattern of demand: ‘if in the division of the gross produce, the labourers commanded a great proportion, the demand would be for one set of commodities – if the masters had more than a usual share, the demand would be for another set’ (VIII, 272–3; see on this issue CE, 97, 439). Evidently, should such (endogenous) changes in demand involve commodities with different factor ratios, there will be a playback on distribution. Ricardo himself did not specifically combine his two relationships to obtain a mutual linkage between distribution and final demand, but for us to do so is to remain faithful to Ricardo since he specified each relationship apart and provided a framework for the treatment of their interdependence. 5 198

ON COMPOSITION OF DEMAND

II. A word is in order regarding J.S. Mill, for the fact is that Ricardo’s analysis, under the rubric of Mill’s fourth proposition on capital ‘demand for commodities is not demand for labour’ has been read by Dobb and others deliberately to exclude the impact of demand composition on distribution. 6 The proposition is elaborated thus: The demand for commodities determines in what particular branch of production the labour and capital shall be employed; it determines the direction of the labour: but not the more or less of the labour itself, or the maintenance or payment of the labour. These depend on the amount of capital, or other funds directly devoted to the sustenance and remuneration of labour. (Mill 1965, II, 78) if by the demand for labour be meant the demand by which wages are raised, or the number of labourers in employment increased, demand for commodities does not constitute demand for labour. I conceive that a person who buys commodities and consumes them himself, does no good to the labouring classes; and that it is only by what he abstains from consuming, and expends in direct payments to labourers in exchange for labour, that he benefits the labouring classes, or adds anything to the amount of their employment. (Mill 1965, II, 80) These formulations reflect Mill’s concern to convey the determinants of aggregate labour demand with special reference to the positive impact on the wages fund of capital accumulation–and direct expenditure on services even of a consumption type (‘other funds directly devoted to ... labour’) – as against consumption or ‘demand for commodities’. In conveying this notion, Mill drew the contrast between investment and consumption sharply and excluded variation in the wages fund component in any given aggregate capital such as would result from changed taste patterns in the event of differential factor ratios. But the fact is that there is also no mention in the chapter at hand, ‘Fundamental Propositions Regarding Capital’, of the impact of ‘machinery’ on the wages fund component of a given amount of capital, despite the fact that the next chapter, ‘On Circulating and Fixed Capital’, adopts Ricardo’s position (93–4). Mill’s attention in the former chapter was evidently diverted from the entire issue of factor proportions. However, once the principle is allowed that aggregate labour demand is not proportional to total capital because of possible variation in capital composition, there is no technical reason – that is, one flowing from the structure of the model – to preclude any other application of that principle even if Mill himself neglected to carry out the exercise. 7 But this is merely to repeat what was said earlier – and confirmed – of Ricardo. 199

RICARDO - THE NEW VIEW

NOTES 1. The analysis designed to show that ‘the employment of machinery is frequently deterimental to [labour’s] interests’ (Ricardo 1951, I, 392) implicitly takes the real wage as given so that the impact is on employment. But the purpose of the analysis is satisfied should the negative impact be entirely or partly on wages. This is made clear in correspondence: ‘Labour will fall because there will be a diminished demand for it’ (VIII, 399). Moreover, in the Principles itself, in the discussion of alternative expenditure patterns between services and luxury goods, flexible wages are explicitly assumed. 2. The altered pattern of expenditure from commodities to services generates an expansion of the wage-goods sector: labourers displaced in the consumer-goods sector are not simply reabsorbed in the service sector but are reabsorbed in expanding wage-goods production to meet the consumption requirements of the (additional) service labour. 3. I have dealt with the problem of recognizing when an author’s statement is ‘unintegrated knowledge’ with special reference to Ricardo’s wage theory, in Hollander 1988 [below, Chapter 15]. 4. Even the discussion of services suggests my excessive caution. Consider the opening of the main statement cited above: ‘If my revenue were 10,000 l., the same quantity nearly of productive labour would be employed whether I realised it in fine clothes and costly furniture, &c.&c. or in a quantity of food and clothing of the same value.’ This assertion relates to alternative commodity expenditure patterns and the impression is that Ricardo does allow a differential impact on labour demand but considers it too small empirically to merit attention – at least in a context concerning the contrasting impact on labour demand of expenditure on services versus expenditure on final goods. 5. The impact of changing distribution on the configuration of demand alluded to here involves shifting demand curves. There is also an impact on supply conditions, and therefore on commodity outputs and prices, given the pattern of demand. This matter, central to the Principles, needs no elaboration since my reviewer has been convinced by my reading (Brewer 1988, 554–5). 6. Cf. Dobb 1940, 44–5: ‘[Mill] apparently intended to imply ... that a demand for some particular commodity as compared with another exerted no appreciable influence on the level of wages ... a repetition of the familiar classical doctrine that the configuration of demand was irrelevant to the distribution of the product between profit and wages .... Like so much of Ricardian reasoning, it rested on a particular assumption: namely, that the proportions between capital and labour were equal in all industries. Without this assumption, the statement would no longer be valid.’ 7. Allowance for differential factor ratios is in fact suggested at the close of the discussion of Circulating and Fixed Capital, though unfortunately without an explicit application: ‘With the proceeds of his finished goods, a manufacturer will partly pay his work-people, partly replenish his stock of the materials of his manufacturer, and partly provide new buildings and machinery, or repair the old; but how much will be devoted to one purpose, and how much to another, depends on the nature of the manufacturer, and the requirement of the particular moment’ (Mill 1965, 99; emphasis added). Note that ‘the portion of capital consumed in the form of ... material ... stands yet in the same relation to the employment of labour, as fixed capital does’ . 200

ON COMPOSITION OF DEMAND

REFERENCES Brewer, A. (1988) Review, Economica 55: 554–5. Dobb, M.H. (1940) Political Economy and Capitalism, revised edn, London: Routledge and Kegan Paul. Hollander, S. (1987) Classical Economics, Oxford: Basil Blackwell. —— (1988) ‘Principles of Textual Interpretation: Illustrated by Ricardian Growth Theory’. (Unpublished manuscript prepared for H.E.S. meeting, Richmond, Virginia.) Mill, J.S. (1965) Principles of Political Economy, Collected Works, vols II and III, Toronto: University of Toronto Press. Ricardo, D. (1951–73) The Works and Correspondence of David Ricarrdo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Samuelson, P.A. (1988) ‘Keeping Whig History Honest’, History of Economic Society Bulletin 10 (Fall) 161–7.

201

12 ON THE ENDOGENEITY OF THE MARGIN AND RELATED ISSUES IN RICARDIAN ECONOMICS*

I. INTRODUCTION As Kenneth Arrow has pointed out in a recent paper, ‘David Ricardo was a peaceful man’ (Arrow 1991, 70). Indeed he was – during his lifetime. I am not so sure he is resting peacefully given the further assertion that his system was ‘a bold attempt to determine values independent of demand considerations’ (75). Arrow adds, by way of qualification, that he does ‘not think, as some neo-Ricardians seem to, that there was in any sense an intended repudiation of the demand schedule’; rather Ricardo did not conceive of such a schedule even though ‘some of [his] analysis can only be made sensible on the basis of such a concept’. Fortuitously, at precisely the same time, there has appeared much the same perspective on Ricardo from Paul Samuelson’s pen. Professor Samuelson condemns Ricardo, along with Sraffa, for working with ‘one-legged’ pricing models, prices emerging ‘ autonomously in terms of technology and costs alone’ (Samuelson 1991, 570). Ricardo ‘positively fails to understand that where the extensive margin falls (and the intensive) is an endogenous unknown in the classical system ’ (571n). In Samuelson’s opinion Smith had done better; Ricardo put back the analytic clock; and Sraffa ought to have taken advice long available in Wicksell’s Lectures in this regard. As Samuelson expresses matters, Ricardo (1951, p. 78) scolds, erroneously scolds, Smith for believing that recourse to scarce land alters the validity of the embodied-labour theory of value. Ricardo wrongly supposes that recourse to external-margin zero-rent land exorcises deviations from the labour theory of value occasioned by enhanced consumer demand for land-intensive goods. He foolishly fails to understand that, where the extensive margin falls (and the intensive), is an 202

ON THE ENDOGENEITY OF THE MARGIN

endogenous unknown of the classical system ! Had Sraffa reflected on Knut Wicksell’s Lectures (1913), he would have been alerted to Ricardo’s fatal misunderstanding of his own model. (Samuelson 1991, 571n) Failure to recognize the endogeneity of the margin had indeed been regarded by Wicksell as ‘the fundamental error’ of Ricardo’s theory of value; Ricardo had failed to appreciate the demand dimension and the ‘mutual conditioning’ of prices by demand as well as supply factors (Wicksell 1934 [1913], I, 24–6). Wicksell does not actually criticize Ricardo for maintaining (on his assumptions) that relative prices in equilibrium will reflect marginal labour costs, but only with failing to recognize that the magnitude of that ratio is dependent on the pattern of demand. This, Wicksell suggests, the classics failed to realize because ‘in the case of one of the most important groups of commodities, the means of subsistence, they regarded demand, or consumption (and therefore also the extension of the margin of production), as given by the size of the population ’ (26).1 Paul Samuelson seems to go much beyond Wicksell. As his note cited above indicates, he also charges Ricardo with error in defending the labour theory by recourse to his differential rent doctrine. 2 Moreover, the passage I have cited continues with the observation that ‘the example of one homogeneous land, in limited supply and used along with labour for only one of two goods, could have alerted Sraffa to the fallacy sans any knowledge of Wicksell’, which suggests that in such a case equilibrium prices would no longer be proportionate to marginal labour inputs. Similarly, other of his remarks point to this conclusion: ‘I was dumbfounded to read that Piero Sraffa evidently believed that Ricardo could succeed in getting rid of land and rent as a complication to the (labour) theory of value. ... How nonsensical is the vulgar belief that zero-rent marginal land can be employed to get rid of land’s effects on the relative prices of newly-produced land-intensive goods’ (Samuelson 1991, 571) If I understand Samuelson correctly, he charges Ricardo with two sins: a failure to recognize the endogeneity of the margin with respect to demand and a failure to realize that the marginal labour theory must be abandoned in the presence of scarce land. 3 There are then two issues: First, can Ricardian rent theory legitimately be used to support the labour theory? This matter should be treated in terms of the logic of the case, and Ricardo’s understanding of that logic, since conceivably Ricardo did not appreciate all aspects of his own model as Samuelson and Arrow suggest is the case. Secondly, does Ricardian theory deny the endogeneity of the margin with respect to demand. Here, too, we should also address Ricardo’s own perceptions of the matter. To summarize my conclusions in this paper: Ricardian rent theory may be shown logically to support the labour theory as Sraffa apparently maintained (Ricardo 203

RICARDO - THE NEW VIEW

1951, I, xxiii), at least on Ricardo’s usual assumption of one-use land; 4 on the other hand – and this is my primary concern – market circumstances govern the relevant labour cost margins and consequently the structure of equilibrium prices. In short, the labour theory as used by Ricardo incorporates allowance for land scarcity. But the endogeneity of the cost margin with respect to demand patterns is central not only to Ricardian theory in the abstract, but also to Ricardo’s policy applications. He well understood its practical significance. 5 The historiographical issue at stake can be easily stated. The Sraffians, Arrow and Samuelson all persist in the belief that Ricardo’s labour theory goes counter to the principles (the so-called ‘neo-classical’ principles) of the allocation of scarce resources – a formal rejection of those principles for the Sraffians, a failure to appreciate them for Arrow and Samuelson (following in Wicksell’s footsteps). In point of fact, Ricardo’s labour theory and the doctrine of rent used in support are part and parcel of that body of analysis. 6 At the same time, Ricardo’s success as allocation theorist is subject to the constraint that land cannot vary optimally – the only technological possibility is variable labour on fixed land, not the reverse. 7 II. RICARDIAN RENT THEORY Ricardo’s rent doctrine, of course, makes use of both the extensive and intensive margins and perhaps more often than not the former. But the general principle is insisted upon where only the more intensive use of given land is involved: It often, and, indeed, commonly happens, that before No. 2, 3, 4, or 5, or the inferior lands are cultivated, capital can be employed more productively on those lands which are already in cultivation. It may perhaps be found, that by doubling the original capital employed on No. 1, though the produce will not be doubled, will not be increased by 100 quarters, it may be increased by eighty-five quarters, and that this quantity exceeds what could be obtained by employing the same capital, on land No. 3. In such case, capital will be preferably employed on the old land, and will equally create a rent; for rent is always the difference between the produce obtained by the employment of two equal quantities of capital and labour. (Ricardo 1951, I, 71–2)8 In what follows I assume solely the possibility of expanding output by the more intensive farming of scarce land, and also single-use land since this is usually Ricardo’s implicit assumption. This amounts to Samuelson’s example of ‘one homogeneous land, in limited supply and used along with labour for only one of two goods’. In this case a switch of demand between manufactures (which is assumed to require no land) and corn will alter the corn-manufacturing exchange rate as given land is used 204

ON THE ENDOGENEITY OF THE MARGIN

more or less intensively, but still it will be true that the new price ratio is proportionate to marginal labour cost. 9 P /P is certainly dependent on the composition of final c m demand, but still the ratio is proportional to relative marginal labour inputs (just as Wicksell has it). The essence of the Ricardian doctrine is that the magnitude of labour costs is determined (partly) by the degree of land scarcity. Accordingly, should pressure on land be relaxed in consequence of a fall in the demand for corn, or an increase in land supply (literally or effectively in consequence of technical progress), marginal labour costs would be lower than they actually are. It is desirable to establish that this perspective is not a modern and anachronistic reading into an early nineteenth-century text. It is Ricardo’s perception of his own model in the Principles. The scarcity property of land rent is stated thus at the outset of the rent chapter: ‘If all land had the same properties, if it were all unlimited in quantity, and uniform in quality, no charge could be made for its use, unless it possesses peculiar advantages of situation’ (I, 69). And though the price of corn equals marginal labour cost – it is more accurate to express this in terms of the proportionality of the corn:gold exchange ratio to the ratio of respective marginal labour costs 10 – that price yields a rent and satisfies the condition that the market for corn clears. The corn price and land rent are as much demand as supply determined: ‘It is ... from the price at which the produce is sold, that rent is derived; and this price is got ... because it is the price which suits the consumption to the supply’ (I, 77n). Accordingly, a relaxation of land scarcity dictates a fall in price and in rent and it does so via a fall in marginal (labour) costs. Reduced demand for corn is specifically mentioned: ‘Any circumstances in the society which should make it unnecessary to employ the same amount of capital on the land, and which should therefore make the portion last employed more productive, would lower rent. ... Every reduction of capital is ... necessarily followed by a less effective demand for corn, by a fall of price, and by diminished cultivation’ (I, 78). 11 Evidently, and happily, Ricardo was at least as competent as any ‘beginning student of economics’ in his appreciation that the transformation terms between labourintensive manufactures and land-intensive corn varies with demand – and with technology. But on his assumptions the new set of equilibrium prices will be proportional to the new ratio of relative marginal labour costs. That is all he insisted upon. That his main question at the outset of the rent chapter – ‘whether the appropriation of land and the consequent creation of rent, will occasion any variation in the relative value of commodities, independently of the quantity of labour necessary to production’ (I, 67) – is answered in the negative, does not mean that the requirement to pay for scarce land has no effect on relative values. It has an effect, by playing on relative marginal labour costs. 205

RICARDO - THE NEW VIEW

So much for the theory in the simplest case. But the principles at issue are applied to policy in an analysis which explicitly points to the dependency of the price ratio on the composition of the consumer’s final demand. Consider a net increase in the demand for domestic corn due to the granting of a corn export subsidy. If agriculture is a constant-cost industry (as Ricardo here provisionally assumes), the adjustment proceeds until the corn price, initially raised above costs, falls to its original level: A bounty then, which should lower the price of British corn in the foreign country, below the cost of producing corn in that country, would naturally extend the demand for British, and diminish the demand for their own corn. This extension of demand for British corn could not fail to raise its price for a time in the home market. ... But the causes which would thus operate on the market price of corn in England would produce no effect whatever on its natural price, or its real cost of production. ... By raising the profits of the farmer’s stock, the bounty will operate as an encouragement to agriculture, and capital will be withdrawn from manufactures to be employed on the land, till the enlarged demand for the foreign market has been supplied, when the price of corn will again fall in the home market to its natural and necessary price, and profits will be again at their ordinary and accustomed level. (Ricardo 1951, I, 301–2; emphasis added) In the more usual case, of course, the equilibrium outcome reflects a higher marginal labour cost and corn price: I have already attempted to show, that the market price of corn would, under an increased demand from the effects of a bounty, exceed its natural price, till the requisite additional supply was obtained, and that then it would again fall to its natural price. But the natural price of corn is not so fixed as the natural price of commodities; because, with any great additional demand for corn, land of a worse quality must be taken into cultivation, on which more labour will be required to produce a given quantity, and the natural price of corn will be raised. (Ricardo 1951, I, 312; emphasis added) Here Pc/Pm rises with a switch in the policy-induced pattern of demand favouring agriculture. 12 A further indication of the role Ricardo accorded demand in determining the margin, and the relevant marginal labour costs, is provided by the analysis of a contemporary case where certain firms in an industry had their labour costs subsidized, thus generating a (discrete) increasing-cost supply schedule. The appropriate 206

ON THE ENDOGENEITY OF THE MARGIN

margin and corresponding marginal labour cost are explicitly assumed to be governed by ‘the quantity of produce required’, i.e. by demand considerations. Should the level of demand be sufficiently low for the output of the subsidized firms alone to be ‘equal to all the wants of the community’, the unsubsidized firms would be excluded and the equilibrium will be determined by the low cost conditions; at a higher demand the higher marginal cost becomes pertinent (I, 73). III. A SPECIAL CASE: CAPACITY OUTPUT The distinctive treatment accorded rent – its so-called ‘exclusion’ from marginal costs – reflects the assumption that land is subject to more or less intensive use by labour (labour-cum-capital), whereas the reverse relation involving a marginal product to variable land given labour-and-capital does not hold. Within this framework exchange ratios, proportionate to relative marginal labour costs, nonetheless assure rents per acre which reflect the degree of land scarcity dictated by final demand pressures. Ricardo takes for granted that a positive marginal labour product can be defined: ‘Is it possible ... seriously [to] assert, that the produce of the land cannot be increased, if the demand increases?’ (I, 252n). But what if capacity output has been achieved? Here the more intensive use of land is ruled out, so that the marginal principle breaks down and rent emerges as a demand-determined surplus on every unit of corn produced including the last. 13 In this case the incidence of a tax on corn, Ricardo observes, is borne entirely by producers, which suggests ( pace Arrow) that we must attribute to Ricardo himself appreciation of a negatively-sloped ‘demand curve’ cutting across the vertical supply. 14 Thus although he usually assumes zero price elasticity of demand for corn this is not his practice when a negative slope is required for stability of equilibrium: The corn and raw produce of a country may, indeed, for a time sell at a monopoly price; but they can do so permanently only when no more capital can be profitably employed on the lands, and when, therefore, their produce cannot be increased. At such time, every portion of land in cultivation, and every portion of capital employed on the land will yield a rent, differing, indeed, in proportion to the differences in the return. At such a time too, any tax which may be imposed on the farmer, will fall on rent, and not on the consumer. He cannot raise the price of his corn, because, by the supposition, it is already at the highest price at which the purchasers will or can buy it. He will not be satisfied with a lower rate of profits, than that obtained by other capitalists, and, therefore, his only alternative will be to obtain a reduction of rent, or to quit his employment. (Ricardo 1951, I, 250–51; emphasis added) 207

RICARDO - THE NEW VIEW

IV. GROWTH AND THE DEMAND-DETERMINED MARGIN My demonstration that the endogeneity of the margin with respect to demand was a central part not only of Ricardo’s model but of his own understanding of that model has deliberately concentrated on cases involving alterations in the structure of activity, population given. But the role of demand in determining the margin and marginal labour cost emerges equally clearly in the analysis of population expansion. This is worthwhile demonstrating because of high authority to the contrary, Sir John Hicks going beyond Wicksell or Arrow or Samuelson in denying a role for demand even in the growth process: Ricardo was no Marshallian. He maintained consistently that prices are determined by cost; demand has nothing to do with them. It may indeed be objected that when he lets the (marginal) cost for food production rise, under pressure of population, he is admitting demand, it is increased demand for food which forces the extension of cultivation. I do not believe that Ricardo looked at the matter like that. It is not a change in demand which marks the transition from one equilibrium to its successor; it is the increase in population itself. (Hicks 1985, 317) That Ricardo in fact insisted on the role of demand in dictating the extent of agricultural expansion is clear, for example, from objections he made to Malthus’s frequent linkage of population growth to preceding accumulations of food. Only if improved living conditions with higher demand for labour should generate higher marriage and birth rates will there be an increased demand for food – in place of workers’ demand for luxuries – and in that case the agricultural margin is extended in response:

When a high price of corn is the effect of an increasing demand, it is always preceded by an increase of wages, for demand cannot increase, without an increase of means in the people to pay for that which they desire. An accumulation of capital naturally produces an increased competition among the employers of labour, and a consequent rise in its price. The increased wages are not [‘not always’ in the third edition] immediately expended on food, but are first made to contribute to the other enjoyments of the labourer. His improved condition however induces, and enables him to marry, and then the demand for food for the support of his family naturally supersedes that of those other enjoyments on which his wages were temporarily expended. Corn rises then because the demand for it increases, because there are those in the society who have improved means of paying for it ; and the profits of the farmer will be raised above the general level of profits, till the requisite quantity of capital has been employed on its production. Whether, 208

ON THE ENDOGENEITY OF THE MARGIN

after this has taken place, corn shall again fall to its former price, or shall continue permanently higher, will depend on the quality of the land from which the increased quantity of corn has been supplied. If it be obtained from land of the same fertility, as that which was last in cultivation, and with no greater cost of labour, the price will fall to its former state; if from poorer land, it will continue permanently higher. (Ricardo 1951, I, 162–3; emphasis added) Merely to accumulate food would be disastrous precisely because of neglect of the demand principle. This was a matter of fundamental concern from the perspective of the law of markets (I, 293). Although evidence for my case might be drawn from the entire corpus of Ricardo’s work, I have deliberately stayed within the confines of the Principles of Political Economy. But one particular letter, dated 26 September 1820, should be noted because it provides especially striking confirmation. In this letter Ricardo again objects to the sequence of increased food supply prior to population growth (the Malthusian position as expressed by Hutches Trower). Only in a control-economy can demand considerations be dispensed with; in a market-economy expansion of food supplies, like that of any commodity, turns on ‘an actual or expected demand’ for food: It is undoubtedly true that if production were wholly under the control of one individual, whose object it was to increase population, he could not better effect his object than by growing more corn in the country than the existing community could consume – it would in that case be at a low price, and the greatest stimulus would be given to population. [But] ... what we want to know is, whether, in the present distribution of property, and under the influence of the motives which invite to production, corn is produced for any other reason than that iron, silk, wine &c. &c. are produced – whether they are not all produced on account of an actual or expected demand for them. (Ricardo 1951, VIII, 255–6) In this letter too, as in the Principles, the role of demand is explicitly elucidated in the analysis of the impact of net capital accumulation. It is only to the extent that such accumulation generates increased demand for food by labour that corn output expands: ‘The aggregate capitals will be increased! If labour cannot be procured no more work will be done with the additional capital, but wages will rise, and the distribution of the produce will be favourable to the workmen. In this case no more food will be produced if the workmen were well fed before, their demand will be for conveniences, and luxuries. But the number of labourers are increased, or the children of labourers! Then indeed the demand for food will increase, and food will be produced in consequence of such demand ’ (VIII, 258). The italics here, for once, are Ricardo’s, not mine. 209

RICARDO - THE NEW VIEW

Finally, there will be found explicit insistence on the demand factor in Ricardo’s Notes on a Malthusian account of agricultural expansion during the growth process: Note 77: This is correct provided there is a demand for the produce – that is absolutely essential to increased cultivation. Mere quantity of produce will not compensate the producer. Note 78: Here expences of producing are compared with the price of produce – this supposes an adequate demand for the produce. The question in dispute is taken for granted. (Ricardo 1951, II, 139) A word of caution may be in order. Ricardo frequently engaged in comparativestatic exercises tracing out the consequences of once-and-for-all increases in capital from initial full equilibrium states. Several of the foregoing extracts illustrate the procedure. But the full-fledged growth model involves ongoing capital accumulation and population growth until stationariness is achieved, attributed to an excess of the returns to capital and labour above their respective minima (Ricardo 1951, I, 101–2). The principles derived from the exercises must, therefore, be applied to the more complex case; for to say that the real wage exceeds ‘subsistence’ assuring population growth and agricultural extension is merely a shorthand expression for the full process. That process entails ongoing net capital accumulation such that the real wage suffices, period after period, to provide a continuous (though decelerating) stimulus for population expansion, and therefore for the increase in demand for food required to motivate continuous extension of the agricultural margin until the stationary state. V. RENT AND THE INVERSE WAGE-PROFIT RELATION Wicksell suggested that by assuming the demand for corn to be a function of population size alone, Ricardo was blinded to the impact of alternative patterns of final demand on the margin. We have shown how Ricardo did deal with that specific issue, even given population, in the course of his treatment of policy-induced changes in activity. A change in favour of land-biased products raises rent, and conversely, even though long-run equilibrium prices reflect relative marginal labour inputs; scarcity of land plays on relative prices, but indirectly by affecting the location of the margin. I can see no cause for complaint here. The demand for corn has multiple determinants. For the specific purposes of his analysis Ricardo had no need to spell them all out, for it sufficed to make the simplest assumption holding all else constant in the background. But when he did address the issue it emerges explicitly that prices are determined by both demand and supply factors as we have shown. 210

ON THE ENDOGENEITY OF THE MARGIN

But is the labour theory essential to Ricardo’s main purposes? The answer must be positive if it is required for the inverse wage–profit relation. Otherwise it serves only as a convenient simplification. Now we know that Ricardo stood by his fundamental theorem even in the event of non-uniform capital–labour ratios. An increase in the wage rate then disturbs the price structure, but still generates an across-the-board decline in the profit rate. To that extent a labour theory is not required. But this procedure does call for the treatment of rent as residual to allow a focus specifically on the wage–profit relation. Does the latter device adequately serve the purpose? It has been my argument that Ricardo’s rent theory does not exclude land scarcity relative to the pattern of final demand from exerting an influence on the price structure, but it does so indirectly by determining the locus of the agricultural margin. The question can therefore be made more specific: What happens to the inverse wage–profit relation should a wage increase, by raising the demand for landusing products, affect the locus of the margin? One of the passages cited earlier (p. 208) directly addresses this possibility, though Ricardo set it aside – there is some swaying here in the third edition – by assuming that (until population rises) demand will increase for ‘other enjoyments’, i.e. that the labourer’s demand for food has zero income elasticity: ‘The increased wages are not [‘not always’ in 1821] immediately expended on food.’ Nevertheless, this particular disturbance would not have been ruinous since the margin shifts out raising rent per acre and reducing the profit rate more sharply than if the margin is fixed. But even in the unlikely case that the wage increase results in a reduced demand for agricultural products – those products being inferior goods – and allowing also that the profit rate rises with a contraction of the margin, the inverse relation holds firm. For we must not forget that it refers specifically to proportionate shares in the product to be divided between labour and capital, not to absolute quantities; that both the real wage and the profit rate may rise with an increase in the marginal product does not compromise the fundamental theorem on distribution. 15 VI. THE PRIMACY OF SUPPLY Faulty though the tradition is that Ricardo neglected consumer-demand factors, it is not entirely based on thin air. One of his chief concerns was the long-run equilibrium pattern of competitive exchange rates and the determinants of changes in that pattern. Within this frame of reference supply conditions are as it were ‘primary’ since no matter what the absolute magnitude of demand for a product may be its price will settle at a level reflecting the relevant cost conditions, unless we posit the special case of capacity output strictly defined (above, p. 207). That, after all, is the implication 211

RICARDO - THE NEW VIEW

of an equilibrium price structure characterized by uniformity of wage and profit rates. Unfortunately, the primacy accorded supply conditions has been misread either as a deliberate attempt to construct a system solely in terms of supply factors or a failure to perceive the need to consider the role of demand factors. The primacy of supply emerges throughout the Principles, but a summary exchange with Malthus reveals the matter particularly effectively since it directly addresses our main issue. Malthus at one stage explicitly raised the matter of the endogeneity of the margin: ‘when you reject the consideration of demand and supply in the price of commodities and refer only to the means of supply, you appear to me to look only at the half of your subject. No wealth can exist unless the demand, or the estimation in which the commodity is held exceeds the cost of production: and with regard to a vast mass of commodities does not demand actually determine the cost? How is the price of corn, and the quality of the last land taken into cultivation determined but by the state of the population and the demand’ (Ricardo 1951, VIII, 286). Ricardo disputed none of this but still insisted on the primacy of supply: ‘I do not dispute ... the influence of demand on the price of corn and on the price of all other things, but supply follows close at its heels, and soon takes the power of regulating price in his own hands, and in regulating it he is determined by cost of production’ (VIII, 302).16 Following an increase in demand for a product the price would settle at a level determined by the appropriate cost conditions. 17 The concern to convey the primacy of supply emerges also in international price comparisons. For conceivably a ‘great’ demand might be accompanied by a ‘low’ price in equilibrium; for example, though the French demand for domestic corn was greater than the British, British corn prices were higher: ‘It is admitted by everybody that demand and supply govern market price, but what is it that determines supply at a particular price? cost of production. Why is corn almost invariably higher here than in France? not on account of the greater demand for it, but on account of its superior cost of production in this country’ (II, 45). A careless reading of this passage, especially of the last sentence, might suggest that the level of demand in Britain (or in France) is irrelevant in determining the margin of cultivation, whereas the message in fact is simply that since long-run supply is governed by cost conditions, it follows that even if demand should be ‘high’ the price might be ‘low’. VII. SUMMARY AND CONCLUSION As Wicksell suggested, it may well be Ricardo’s usual practice of relating the demand for food to population size alone that diverted him from formally examining the impact of changing patterns of demand, population assumed to be constant, on the margin of cultivation. I have argued here that to conclude from this either that he denied in principle or that he failed to appreciate such impact is unjustified; one 212

ON THE ENDOGENEITY OF THE MARGIN

distinctly sees the impact of demand patterns on the margin in policy applications. Secondly, I have confirmed that Ricardo assigned the demand factor an essential causal role in the analysis of population expansion. Finally, I have demonstrated that the Ricardian inverse wage–profit relation is quite able to incorporate distributioninduced changes in demand patterns. The erroneous perspective taken of Ricardo reflects to some extent misunderstanding of the primacy he accorded supply factors. This is compounded in our day by the unfortunate habit of confounding Ricardo with the self-styled Cambridge ‘Ricardians’ who are one-legged on principle, and who use Ricardo ‘as a bludgeon against neo-classical economics’ (Arrow 1991, 72). A primary aspect of the Sraffian version of Ricardianism is the proposition that income distribution is independent of the pattern of consumer demand, a proposition which (I think) Samuelson also attributes to Ricardo. 18 But this does less than justice to Ricardo. As we have shown above, an altered pattern of demand between agriculture and manufactures causes changes in rent relative to profits-cum-wages notwithstanding that equilibrium prices are proportional to marginal labour inputs. It is surprising that any such failure as we are discussing could have been attributed to Ricardo considering his contribution to the Corn Law debate. 19 Arrow for his part conjectures that the Cambridge use of Ricardo ‘derives at least the aura of respectability and legitimacy from Marshall’s emotional pro-Ricardian bias’ – having in mind the ‘supreme authority of Marshall at Cambridge well beyond his active career’. On the grounds intimated in this paper, I cannot accept that Marshall’s bias was ‘emotional’ and without substance; and I doubt whether his authority could extend as far as Arrow suggests, considering Marshall’s attribution of an appreciation of the demand side to Ricardo. 20 NOTES * For comments and criticism I am grateful to W.J. Baumol, M. Bronfenbrenner, T. Kompas, J.I. McDonald, P.L. Porta and P.A. Samuelson. 1. Later in the Lectures Wicksell gave his famous demonstration that the conception of an increase of labour applied to fixed land to derive the marginal product of labour (and rent as a residual surplus) can be reversed such that rent is determined by the marginal productivity of land (and wages as a residual) (Wicksell 1934 [1913], I, 124–33). But he draws no implications from this generalization in the course of his earlier criticism of Ricardo on value theory. 2. Arrow is apparently satisfied with Ricardo’s use of the rent doctrine as a way of supporting a labour theory (Arrow 1991, 75–6). 3. In his ‘Modern Treatment of the Ricardian Economy’ Samuelson also asserts that in equilibrium relative prices would diverge from relative marginal labour inputs. For he complains that whereas ‘beginning students of economics today learn in their first week that the transformation terms between a labour-intensive and a land-intensive good vary with demand ... yet in reading a thousand pages on the labour theory of value, I can 213

RICARDO - THE NEW VIEW

4. 5.

6.

7.

8. 9. 10.

11.

12.

remember but one author who comes close to emphasizing that land – and not merely capital – vitiates a simple labour theory and that no tricks with no-rent marginal land can change this’ (Samuelson 1966 [1959], 380n). But some Samuelsonian objections to Ricardo can perhaps be read simply as a criticism of the usefulness of a theory of value in terms of marginal labour costs, if the location of the margin is governed by taste patterns: ‘He also blunders if he thinks that he can “get rid of land and rent as a complication for pricing” by concentrating on the external margin of no-rent land: where that external margin falls is an endogenous variable that shifts with tastes and demand patterns so as to vitiate a hoped-for labour theory of value or a wage- cum -profit rate theory of value’ (Samuelson 1978, 1420) In the multi-use land case, the proportionality of prices to marginal labour costs may still hold good, but as a formality. Morishima comes to the same conclusion by a formal mathematical route: ‘It is especially important to observe that labour values [for Ricardo] are not constants determined solely by technology: they fluctuate economically according to whether the circumstances of the market require the intensity of cultivation to change’ (Morishima 1989, 33). I ask myself why I am writing this paper, having already set out my view of Ricardo as a ‘demand–supply’ value theorist on several occasions (most recently 1991). I follow the advice given to Neville Chamberlain by his father (which, regrettably, he took at Munich): ‘If at first you don’t succeed, try, try, try again.’ More seriously, the present paper corrects certain misconceptions regarding the labour theory and explores further the implications of Ricardian rent doctrine for the inverse wage–profit relation. Wicksell merely noted the presumed lack of symmetry, whereas Walras asked ‘why the English School determines rent by the quantities of labour and capital-services employed, rather than wages and interest by the quantities of land-services employed; or why this school does not try to formulate a unified general theory to determine the prices of all productive services in the same way’ (Walras 1954, 416). I suggest the lack of symmetry reflected the presumption of one-use land. With multi-use land, the notion of land fixity breaks down as soon as it concerns particular industries. Ricardian rent theory can dispense with the extensive margin, but not with the intensive margin. Of course we assume uniform capital–labour ratios to focus on the primary issue. See Ricardo’s clarificatory observation in the profits chapter: ‘The reader is desired to bear in mind, that for the purpose of making the subject more clear, I consider money to be invariable in value [i.e. in its supply conditions], and therefore every variation of price to be referable to an alteration in the value [labour input] of the commodity’ (Ricardo 1951, I, 110n). A similar relaxation of land scarcity may be generated by technical progress: ‘The same effects may however be produced, when the wealth and population of a country are increased, if that increase is accompanied by such marked improvements in agriculture, as shall have the same effect of diminishing the necessity of cultivating the poorer lands, or of expending the same amount of capital on the cultivation of the more fertile portions’ (Ricardo 1951, I, 79). Ricardo explains why the manufacturing price is not pulled up as resources flow to agriculture: ‘Manufactures would not rise, because fewer would be manufactured, for a supply of them would be obtained in exchange for the exported corn’ (Ricardo 1951, I, 307). 214

ON THE ENDOGENEITY OF THE MARGIN

13.

14. 15.

16.

17. 18. 19. 20.

The relaxation of corn-import restrictions amounts in effect to a contraction of demand for home corn, and here the reverse result emerges, namely a reduction in Pc/ Pm: ‘If corn, in consequence of permanent abundance, fell to 3l.10s., the capital employed on No. 6 [land] would cease to be employed; for it was only when corn was at 4l. that it could obtain the general profits, even without paying rent; it would, therefore, be withdrawn to manufacture those commodities with which all the corn grown on No. 6 would be purchased and imported’ (I, 268). The classicists called this a case of ‘monopoly’. As Viner pointed out, ‘in Ricardo’s time the term was widely used to cover (1) ownership of a scarce commodity by a single holder; (2) ownership by a few; and (3) scarcity of a commodity, which because of zero-elasticity of supply, a rise in price would not ameliorate’ (Viner 1958, 360). Where Ricardo does assume zero demand elasticity problems are created for the adjustment of price to cost. On these complexities, and Ricardo’s solutions, see Hollander 1979, 291–3. The threat to Ricardian theory comes from another quarter. In the event that materials are required in all sectors of the economy, a rise in their supply price would depress the general profit rate independently of wages. As mentioned above, Malthus frequently insisted on increased corn production prior to increased population and demand. Ricardo, by contrast, had always insisted on the demand requirement as we have also seen. His reply is, therefore, a very restrained one. Ricardo evidently feared that the impact of costs on long-run price was clouded, even denied, by some demand–supply formulations (1951, VII, 250–1). ‘[Any] classicist who thinks he can separate “value” from “distribution” commits a logical blunder’ (Samuelson 1978, 1420). I go further and attribute to Ricardo appreciation of an impact exerted by demand patterns on the profits/wages breakdown (Hollander 1989 [above, Chapter 11]). More promising is a conjecture by Bronfenbrenner, that Sraffa’s attributions reflect his Marxism: ‘Sraffa was a Marxist, a refuge from Mussolini’s Fascist regime .... Since Marx professed himself an admirer of Ricardo – as nearly a disciple as Marx could ever admit being of any predecessor – it may have seemed natural to attribute the same system ... to Ricardo’ (Bronfenbrenner 1989, 40–1). For a more detailed argument that the Sraffa reading of Ricardo (especially the corn-ratio theory of profits) involves reading into Ricardo a perspective derived from developing Marx’s surplus doctrine, see Porta 1986. A defence of Sraffa is given in Groenewegen 1986.

REFERENCES Arrow, K.J. (1991) ‘Ricardo’s Works as Viewed by Later Economists’, Journal of the History of Economic Thought 13 (Spring): 70–7. Bronfenbrenner, M. (1989) ‘A Rehabilitation of Classical Economics’, Aoyama Kokusai Seikei Ronshu 13 (June): 35–41. Groenewegen, P. (1986) ‘Porta on Sraffa: A Comment’, History of Political Economy 18: 455– 62. Hicks, J. (1985) ‘Sraffa and Ricardo: A Critical View’, in G. Caravale (ed.) The Legacy of Ricardo, Oxford: Basil Blackwell, 305–19. Hicks, J. and Hollander, S. (1977) ‘Mr. Ricardo and the Moderns’, Quarterly Journal of Economics 91 (August): 351–69. 215

RICARDO - THE NEW VIEW

Hollander, S. (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. ——(1989) ‘On Composition of Demand and Income Distribution in Classical Economics’, History of Economics Society Bulletin 11 (Fall): 216–21. ——(1991) ‘Alfred Marshall in Historical Perspective: Why Marshall was Right about Ricardo’, European Economic Review 35: 313–22. Morishima, M. (1989) Ricardo’s Economics. A General Equilibrium Theory of Distribution and Growth, Cambridge: Cambridge University Press. Porta, P.L. (1986) ‘Understanding the Significance of Piero Sraffa’s Standard Commodity: A Note on the Marxian Notion of Surplus’, History of Political Economy 18: 443–54. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Samuelson, P.A. (1966) [1959] ‘A Modern Treatment of the Ricardian Economy’, Collected Scientific Papers, I, Cambridge, Mass.: MIT Press, 373–422. – -(1978) ‘The Canonical Classical Model of Political Economy’, Journal of Economic Literature 16: 1415–78. – -(1991) ‘Sraffa’s Other Leg’, Economic Journal 101 (May): 570–4. Viner, J. (1958) The Long View and the Short, Glencoe, Illinois: Free Press. Walras, L. (1954) Elements of Pure Economics, ed. William Jaffé, 4th definitive edn (1926), London: George Allen and Unwin. Wicksell, K. (1934) [1913] Lectures on Political Economy, ed. L.C. Robbins, London: Routledge .

216

Part IV THE RICARDIAN GROWTH MODEL

13 ON THE INTERPRETATION OF RICARDIAN ECONOMICS: THE ASSUMPTION REGARDING WAGES

Some thirty years ago, George Stigler published what was to become a standard reference for students of the classical period – ‘The Ricardian Theory of Value and Distribution’. A key feature of the theoretical system developed in the Principles of Political Economy is there said to be the subsistence wage theory, which (together with the measure of value) was added to those elements already present in the earlier Essay on Profits, namely, the theory of rent and the dominant influence of diminishing returns in agriculture upon the rate of profit (1965a, 187). From these elements there followed the ‘great conclusion’ of the model: ‘With the growth of population, the rate of wages rises’ (reflecting the increasing real cost of producing the given basket), ‘the rate of profit falls, and aggregate rents rise – all in terms of the measure of value’ (190). In a review of my recent study of David Ricardo (1979), Stigler reiterates by implication this evaluation of the nature of Ricardo’s contribution and explicitly repeats the attribution to Ricardo of the subsistence wage assumption. It remains his belief that the logic of Ricardo’s argument requires the assumption: ‘I am amazed to be told that “all the evidence” points to Ricardo believing that the wage will fall secularly’; similarly, ‘without this assumption [Ricardo’s] fundamental theorem on distribution (only a rise in wages will lower profits) is not rigorously true’; and ‘his chapter on gross and net revenue and his repeated proposition that only net revenue (which excludes wages) can be taxed or saved are wrong’ (1981, 101). Much depends on the precise assumption made regarding wages; the subsistence wage attribution, for example, has distorted the entire body of ‘Cambridge’ historiographical doctrine relating to classical economics. Stigler is right to focus on this issue. But I believe it can be shown that his version of Ricardianism is invalid. If we are ever to do justice to the historical Ricardo and to the course of nineteenth-century economics we must by all means abandon the fixed-wage attribution. 219

RICARDO - THE NEW VIEW

I. I shall proceed by drawing upon another well-known contribution by Stigler. In his paper on ‘Textual Exegesis as a Scientific Problem’ (1965b), he observes that in seeking an accurate interpretation ‘we should not be so literal-minded as to count the passages in a book to decide an author’s general position because the passages are not of equal importance’ (448). Where there exists a clash of texts the solution is to investigate the dependency of a writer’s main analytical conclusions upon the alternative readings: ‘We increase our confidence in the interpretation of an author by increasing the number of his main theoretical conclusions which we can deduce from (our interpretation of) his analytical system’ (448). This procedure Stigler refers to as ‘scientific exegesis ’, and it is designed to isolate the ‘net scientific contribution’ of the writer. If, however, the historian is concerned with what the author in question ‘really believed’, then the criterion should be which of the various alternative interpretations best suits what is known of the man’s style – ‘ personal exegesis ’. Let us accept that the inverse wage–profit relation and the falling secular profit rate are amongst Ricardo’s ‘main analytical conclusions’. How does the notion of scientific exegesis help us? Were Stigler correct that these results require the constant wage assumption, the answer would be selfevident. But he is mistaken. A model can be devised, incorporating the fundamental theorem and generating the falling return on capital, wherein the wage rate is a variable. The model in question is a genuine growth model in the sense that until stationariness is achieved, the wage rate and the profit rate are both above their respective ‘minima’ encouraging net capital accumulation and population growth; and both will ultimately decline from whatever their ‘present’ values may be to those respective minima in consequence of the pressures increasingly exercised by land scarcity. Only in the stationary state itself is the wage at subsistence and the profit rate at its minimum. Versions of this model were developed independently by inter alia John Hicks and myself (1977), Carlo Casarosa (1978), and Paul Samuelson (1978). Not only is a variable-wage growth model technically meaningful, it is precisely what Ricardo himself specified to be his own. Ricardo explicitly describes the falling commodity wage of the secular path: In the natural advance of society, the wages of labour will have a tendency to fall, as far as they are regulated by supply and demand; for the supply of labourers will continue to increase at the same rate, whilst the demand for them will increase at a slower rate. If, for instance, wages were regulated by a yearly increase of capital, at the rate of 2 per cent., they would fall when it accumulated only at the rate of 1-1/2 per cent. They would fall still lower when it increased only at the rate of 1, or 1/2 per cent., and would continue to do so until the capital became stationary, when wages also would become 220

THE ASSUMPTION REGARDING WAGES

stationary, and be only sufficient to keep up the numbers of the actual population. (Ricardo 1951, I, 101) The complication deriving from the rising prices of wage goods is then allowed for; although the commodity wage falls the money wage will rise: As population increases, these necessaries will be constantly rising in price, because more labour will be necessary to produce them. If, then the money wages of labour should fall, whilst every commodity on which the wages of labour were expended rose, the labourer would be doubly affected, and would be soon totally deprived of subsistence. Instead, therefore, of the money wages of labour falling, they would rise; but they would not rise sufficiently to enable the labourer to purchase as many comforts and necessaries as he did before the rise in the price of those commodities. (Ricardo 1951, I, 101) It is precisely this rise in the money wage that causes the profit rate to fall despite the decline in the commodity wage. The model can also easily incorporate an initial section of rising commodity wages; and Ricardo thought that it should (see my 1979 book, 395ff; also my paper with Hicks, 1977, 365). Stigler is unaware of the declining path of wages, but in 1952 he did allude to Ricardo’s references to rising wages. But these, he says, ‘must simply be recorded as correct views which Ricardo did not know how to incorporate into his theoretical system’ (1965a, 172). Poetic justice indeed! To attribute the fix-wage model to Ricardo is illegitimate in terms of Stigler’s own criterion of scientific exegesis. For what Ricardo insisted upon was a secular fall in the profit rate along with an initial stage of increasing wages and a final stage of declining wages and Stigler’s version simply cannot accommodate this complexity. I do not intend to suggest that Ricardo’s account is faultless. It would be remarkable were that the case. In fact, it is probable that there are two Ricardo models – those characterized by the versions offered by Hicks and myself, and by Samuelson or Casarosa. On the first view, secular variations in real wages are the outcome of differential growth rates of capital and labour – the upward trend a result of capital growth outpacing labour growth at an early stage of development and the downward trend a result of the reverse relationship (see the citation above from Ricardo). This version does not, in short, portray a ‘dynamic equilibrium’ or balanced growth path of wages. Such a path is expounded in Samuelson’s ‘canonical classical model’. It traces out the unique values of the wage, given the general labour and capital supply functions, 221

RICARDO - THE NEW VIEW

which assure balanced factor growth subject always to the constraint imposed by land scarcity; the wage falls although labour supply decelerates in line with capital. There is much evidence to suggest that this version too is a valid attribution to Ricardo (see section IV below), although the clearest exposition, which Ricardo accepted, is by Malthus (1820; see Ricardo 1951, II, 255–6). II. What is the source of Stigler’s misunderstanding? His own notion of personal exegesis provides the key. That Ricardo’s Principles and his Essay are replete with references to the constant wages assumption has never been denied. The growth model in the Essay on Profits is expounded on the basis of a constant wage though not at the subsistence level (see Hollander 1979, 136). But Ricardo himself tells us precisely why he so proceeds; he was no methodological neophyte: ‘We will, however, suppose that no improvements take place in agriculture, and that capital and population advance in the proper proportion, so that the real wages of labour continue uniformly the same; – that we may know what peculiar effects are to be ascribed to the growth of capital, the increase of population, and the extension of cultivation, to the more remote, and less fertile lands’ (Ricardo 1951, IV, 12). All that he intended was a simplifying assumption for the sake of clear exposition – to allow him to focus upon one causal variable at a time playing on profits. There is no reason to believe that any more was at stake when he devised the famous numerical illustration of the Principles (chapter 5 and chapter 6) which incorporates the constant wage assumption. Stigler has, it seems, mistaken Ricardo’s typical ‘first approximations’ for a full growth model – the man’s style for the substance. III. The same conclusion follows in the wage taxation context. As I explain in my book (1979, 383) Ricardo’s taxation theorems can be interpreted as applying to the case where a subsistence wage rules. I provide textual evidence from the chapter ‘On Wages’ and elsewhere to show this. But I then demonstrate that the wage taxation theorems do not stand or fall with the subsistence assumption (386). Stigler is mistaken when he cites passages in Ricardo (1951, I, 215, 219) purportedly indicating adherence to a subsistence wage, for it is a constant wage above subsistence to which Ricardo there alluded. A word of explanation. During the course of his discussion (in the chapter ‘Taxes on Raw Produce’) of the effect upon the money-wage rate induced by the taxation of necessaries, Ricardo introduced the qualification that the ‘rate of progression’ of the economy is throughout taken for granted, clearly implying that the analysis was intended to 222

THE ASSUMPTION REGARDING WAGES

apply whether or not wages are initially at ‘subsistence’: ‘Those who maintain that it is the price of necessaries which regulates the price of labour, always allowing for the particular state of progression in which the society may be, seem to have conceded too readily, that a rise or fall in the price of necessaries will be very slowly succeeded by a rise or fall of wages’ (I, 161; emphasis added). If this statement were the only one of its kind, it might perhaps be dismissed as unrepresentative. But the fact is that the chapter ‘On Taxation of Wages’ itself is formally contingent upon it. The chapter ‘On Taxation of Wages’ unlike that ‘On Wages’ is based upon Adam Smith’s proposition that ‘the demand for labour, according as it happens to be either increasing, stationary, or declining, or to require an increasing, stationary, or declining population, regulates the subsistence of the labourer, and determines to what degree it shall be either liberal, moderate, or scanty’ (I, 215). And the general applicability of the taxation theorems is much emphasized: ‘The price of labour will express, clearly, the wants of the society respecting population’ [Malthus]; it will be just sufficient to support the population, which at that time the state of the funds for the maintenance of labourers, requires. ... Suppose the circumstances of the country to be such, that the lowest labourers are not only called upon to continue their race, but to increase it; their wages would be regulated accordingly. Can they multiply in the degree required, if a tax has taken from them a part of their wages, and reduces them to bare necessaries? (Ricardo 1951, I, 219–20) That Ricardo should have proceeded in the chapter ‘On Wages’ and in other contexts on the assumption of a subsistence wage despite his acceptance of the Smithian position according to which the equilibrium wage is that wage which assures an appropriate positive rather than a zero growth rate of labour supply, is not difficult to appreciate. It is characteristic of his general method to simplify the analysis wherever this can be done without loss. The broader application of the taxation theorems means simply that the mechanism of population adjustment to wages above or below a given ‘subsistence’ rate must now be applied to wage variations about a long-run labour supply curve relating the wage rate to the growth rate of population. Once again Stigler has mistaken a strong-case simplifying assumption for the entire analysis. IV. Ricardo’s full position as formulated in the taxation context – that of Smith and Malthus – has broad implications for growth in general. The rate of capital accumulation acts, in the first instance, as an independent variable which determines what 223

RICARDO - THE NEW VIEW

the commodity wage rate must be to guarantee an equivalent growth rate of population. Any disturbance which raises the price of wage goods – the effects say of increasing land scarcity and not merely taxation – necessitates an appropriate monetary compensation to assure that the growth rate of population is not impeded. The ‘natural’ wage – albeit ‘unnatural’ because a limiting case (1951, II, 227–8) – is then the wage which assures a constant population and will correspond to zero net accumulation, while to each positive growth rate of capital there will correspond a higher real wage rate to assure the equivalent growth rate of labour supply. (A ‘high’ but constant, rate of accumulation according to this analysis entails a high real wage rate; while an increase (decrease) in the rate of accumulation generates an increase (decrease) in the real wage rate.) But the assumption that the rate of capital accumulation is an independent variable – totally unaffected by the reduction in profits corresponding to the increase in money wages – is, however, no more than a first approximation. A reduction in profits would, it is conceded, probably have some effect on accumulation so that the compensatory increase in money wages would not entirely prevent a fall in real wages in consequence of taxation (I, 221–2, 225–6). In effect, the ‘dynamic equilibrium’ path of the system has been altered to entail reduced (net) factor returns and factor growth rates. The full analysis of wage taxation thus allows for a decline in the real wage and flies in the face of a fixed wage interpretation. All the ingredients of the canonical growth model – and as remarked earlier Ricardo approved of that version formulated by Malthus – come into play in the wage taxation context. V. There remains Ricardo’s conception of net revenue to which Stigler also alludes in support of his case. Malthus had made the same erroneous attribution in his Principles (Ricardo 1951, II, 381). His position, Ricardo complained, had been misunderstood by Malthus (and by Say). The inclusion, within net income, of profit and rent alone was merely a simplifying assumption without substantive intent: ‘Mr. Malthus says “the additional two millions of men would some of them unquestionably have a part of their wages disposable”. Then they would have a part of the neat revenue. I do not deny that wages may be such as to give to the labourers a part of the neat revenue – I limited my proposition to the case when wages were too low to afford him any surplus beyond absolute necessaries’ (1951, II, 380– 1). Similarly regarding profits and rent as the source of accumulation and taxation: ‘Perhaps this is expressed too strongly, as more is generally allotted to the labourer under the name of wages, than the absolutely necessary expenses of production. In that case a part of the net produce of the country is received by the labourer, and may be saved or ex224

THE ASSUMPTION REGARDING WAGES

pended by him; or it may enable him to contribute to the defence of the country’ (I, 348 fn). VI. In another of his seminal articles (1976), Stigler makes the point that if we seek to understand the scientific role played by a figure in the history of economic theory then what is relevant is how his work ‘appeared to his contemporaries’, for science ‘consists of the arguments and the evidence that lead other men to accept or reject scientific views’ (60). The ideas they may have intended to express are not relevant from this perspective. This is very true and it is a perspective that casts much light on a number of issues in scientific history. But we must also allow for erroneous interpretations. While the variable-wage model was an open book – the above account is not based on correspondence or other unpublished and not easily accessible items – from the very outset readers have often misunderstood Ricardo. Erroneous readings bedevil the problem of his legacy. REFERENCES Casarosa, Carlo (1978) ‘A New Formulation of the Ricardian System’, Oxford Economic Papers 30,1 (March): 38–63. Hicks, John and Hollander, S. (1977) ‘Mr. Ricardo and the Moderns’, Quarterly Journal of Economics 91,3 (August): 351–69. Hollander, Samuel (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 Vols), Vol. I: Principles of Political Economy (1821), Vol. II: Notes on Malthus’s Principles of Political Economy (1820), Vol. IV: Pamphlets, 1815–1821, Cambridge: Cambridge University Press. Samuelson, Paul A. (1978) ‘The Canonical Classical Model of Political Economy’, Journal of Economic Literature 16,4: 1415–34. Stigler, George J. (1965a) ‘The Ricardian Theory of Value and Distribution’, Journal of Political Economy June 1952; reprinted in Essays in the History of Economics, Chicago: University of Chicago Press, 165–98. ——(1965b) ‘Textual Exegesis as a Scientific Problem’,Economica 32 (November): 447–50. ——(1976) ‘The Scientific Uses of Scientific Biography, with Special Reference to J.S. Mill’, in J.M. Robson and M. Laine (eds) James and John Stuart Mill: Papers on the Centenary Conference, Toronto: University of Toronto Press, 54–66. ——(1981) ‘Review’, Journal of Economic Literature 19 (March): 100–2.

225

14 THE WAGE PATH IN CLASSICAL GROWTH MODELS: RICARDO, MALTHUS AND MILL

One general growth model encapsulates a variety of classical perspectives on the secular path of wages. The common features of the model utilized by Ricardo, Malthus and Mill are land scarcity manifested in diminishing agricultural returns at least beyond a certain labour–land ratio, and a positive functional relation between the return on capital ( r ) and the capital growth rate ( g K ). There are minor differences only between Ricardo and Malthus concerning the relation between the labour growth rate ( g L) and the wage ( w ). For Ricardo g L is irresponsive to wage reductions except at levels close to subsistence, whereas Malthus allowed a regular (positive) g L-w relationship. But this difference is not a substantive one, and in each version the outcome of the growth process is a downward path of wages until the subsistence level (w s) where g L (as well as g K ) has fallen to zero. A substantive difference between the two is that for Ricardo wage movements result from deviations between the factor growth rates, while Malthus posits a declining wage path notwithstanding equality of the factor growth rates – i.e. his constitutes a ‘dynamic equilibrium’ path. Mill’s concern was with the implications of ‘prudential’ behaviour on the part of labour. In principle, ‘prudence’ can be incorporated within the labour-supply function itself in which case it is manifested in a wage floor at some conventional wage (w c) or as exogenous upward shifts in a regular, positively sloping, g L function. The outcome is the same in both cases, namely constancy of the wage on the path to the stationary state despite the pressures deriving from land scarcity. In his Autobiography Mill paid tribute to Malthusianism as a reform doctrine: ‘This great doctrine, originally brought forward as an argument against the indefinite improvability of human affairs, we [the Philosophical Radicals] took up with ardent zeal in the contrary sense, as indicating the sole means of realizing that improvability by securing full employment at high wages to the whole labouring population through a voluntary restriction of the increase of their numbers’ (Mill 1981, 107–8). 226

THE WAGE PATH

Now Malthus himself formulated briefly but precisely the ‘constant wage’ model in the Essay on Population. To do justice to the historiographical record we must refer to the Malthus–Mill Prudential Wage Path, and distinguish between the ‘basic’ Malthusian model and its modification. That Ricardo would also have subscribed to this analysis is clear from his famous pronouncement that ‘the friends of humanity cannot but wish that in all countries the labouring classes should have a taste for comforts and enjoyments. ... There cannot be a better security against a superabundant population’ (Ricardo 1951, I, 100). The wage theories of Malthus, Ricardo and Mill can in fact be set out with the help of the same basic diagram. In this, wages are measured in commodity and not in value terms; and throughout the article we proceed as if wage goods are a single good. Panel A in each diagram portrays the marginal product curve generated by variable labour on given land ( HH’) (see Figure I). HH’ corresponds to the roof in Hicks and Hollander (1977), and it depicts the maximum wage, w max, that is obtainable when the minimum feasible return on capital (the return that reduces g K, the rate of growth of capital, to zero) is itself zero. If this minimum return on capital is labelled r *, w = w max when r = r * = 0, and w max is depicted by HH’ in panel A. In formal terms, the wage per worker, w, plus profits per worker, wr (in circulating capital models where wages are paid one period in advance of the sale of the product), exhausts the marginal product of labour, F’(L), so that:

w + wr = F’ (L) and therefore: F’(L)

w = 1 +r

and if w max is the wage where r = r * = 0, then:

w max = F’(L). In each diagram, panel B portrays the wage on the vertical axis and g K and g L, the rates of growth of capital and labour, on the horizontal axis, and the g K -w relationship it depicts is derived from panel A. For any (absolute) labour supply, the marginal product of labour on HH’ indicates the maximum wage ( w max) that reduces r to r * and therefore g K to zero. A lower wage at that same marginal product of labour will be associated with a higher r and therefore a higher g K : thus at any given marginal product, the lower the wage the higher the rate of profit and the higher the rate of growth of capital, and this generates a negative g K -W relationship, the curve emanating from the y axis at the appropriate level given by the height of HH’. At the same 227

RICARDO - THE NEW VIEW

time, at any given w, a declining marginal product along HH’ implies a decline in r and thus in gK , and this effect of a declining marginal product at a given wage will be reflected by a continual inward shift of the gK-w function in panel B. This diagram will now be used to help to set out the wage theories of Ricardo, Malthus and Mill, so that the differences and the similarities between them can be clearly seen.

HH’ corresponds to the marginal product curve where r* = 0. For each L read off value of MP on HH’. This yields intercept in B such that gK = 0. For each given MP value gK rises as w falls. For each w in B the relevant gK is read off the family of curves, gK falling as MP declines along HH’. Note: The diagrams contain the same information as in Samuelson (1978) with the following formal differences: 1) Samuelson utilizes the gK = g(r) directly whereas we relate gK to w, consistently with 1 gK = g(r) having in mind that r = F’(L) w 2) Samuelson allows r* > 0, so that the ‘roof’ (HH’) then lies within the MP curve. Figure I The derivation of gK -w RICARDO In the growth context Ricardo asserted that ‘the motive for accumulation will diminish with every diminution of profit, and will cease altogether when [capitalists’] profits are so low as not to afford them an adequate compensation for their trouble, and the risk which they must necessarily encounter in employing their capital productively’ (Ricardo 1951, I, 122). Similarly, ‘savings may be so rapid and profits so low in consequence as to diminish the motive for accumulation, and finally to destroy it altogether’ (II, 8). We take it then that gK is related positively to r and negatively to w. 228

THE WAGE PATH

Ricardo postulates a secular increase in the wage rate followed by a secular decline until w = w s when g L = 0. These movements are generated by deviations between constant gL and g K with g K > g L to begin with and g K < g L subsequently. (The presumed constancy of g L is subject to a qualification to be examined presently; it is not required by the logic of the argument.) There may be an intermediate period of ‘steady-state’ equilibrium with constant wages during which g K = g L. Agricultural productivity is said to be constant at early periods, diminishing returns setting in only after a particular labour–land ratio has been passed. The upward wage trend, over the range of constant marginal product is expressed in the following passage: In different stages of society, the accumulation of capital, or of the means of employing labour, is more or less rapid, and must in all cases depend on the productive powers of labour. The productive powers of labour are generally greatest when there is an abundance of fertile land: at such periods accumulation is often so rapid, that labourers cannot be supplied with the same rapidity as capital. It has been calculated, that under favourable circumstances population may be doubled in twenty-five years; but under the same favourable circumstances, the whole capital of a country might possibly be doubled in a shorter period. In that case, wages during the whole period would have a tendency to rise, because the demand for labour would increase still faster than the supply. (Ricardo 1951, I, 98) The marginal (average) product curve is likely to be peculiarly high in new settlements which can rely on advanced foreign technology: ‘In new settlements, where the arts and knowledge of countries far advanced in refinement are introduced, it is probable that capital has a tendency to increase faster than mankind: and if the deficiency of labourers were not supplied by more populous countries, this tendency would very much raise the price of labour’ (98). What can be said of the initial wage ( w )? Assuming constant marginal product 0 and given capital-supply and labour-supply conditions there will be no motive force generating the upward wage movement if, at w , g K =g L.It must be presumed, therefore, 0 that at w , g K >g L.As the wage rises so the deviation between g L and g K narrows until 0 wage constancy is achieved. This process is consistent with ‘static’ or short-run equilibrium such that the labour market is regularly cleared which seems indeed always to be Ricardo’s presumption (Figure II). Once the equality of g L and g K has been achieved, there will be no further motive for a change in the wage, which will proceed at a steady level above subsistence: ‘Notwithstanding the tendency of wages to conform to their natural 229

RICARDO - THE NEW VIEW

Each position on the upward stretch of w is one of shortrun (static) equilibrium; not long-run(dynamic) equilibrium. Each position on the horizontal stretch of w is one both of static and dynamic equilibrium.

The intercept of gK on yaxis is determined by the height of HH’; its slope reflects the inverse w-r relation since gK = g(r) 1. where r = F’(L) w At w , gK > gL generating 0 upward pressure until wm. Thereafter the wage remains steady at wm. gL is assumed vertical although this is not essential for the logic.

The Market Period equilibrium wage: Period by period L shifts rightward by a given percentage amount (i.e.: an increasing absolute amount). Capital increases initially at a greater percentage amount, subsequently at same percentage amount.

Figure II Ricardo: the increasing wage trend: initial gK > gL rate [thesubsistence wage, w , at which gL = 0], their market rate may, in an improving s society, for an indefinite period, be constantly above it; for no sooner may the impulse, which an increased capital gives to a new demand for labour be obeyed, than another increase in capital may produce the same effect; and thus, if the increase in capital be gradual and constant, the demand for labour may give a continued stimulus to an increase of people’ (I, 94–5). It may be remarked – although Ricardo does not refer to the possibility – that an increasing wage trend may be generated even if gK = gL at the initial position w , in 0 the event of increasing returns. For a deviation between the growth rates will be continually generated as population and productivity rise putting continuous upward pressure on gK . In this case too the shortrun labour market is in equilibrium while the factor growth rates differ. Balanced factor growth is only achieved when increasing returns peter out (Figure III). 230

THE WAGE PATH

The upward path is the locus of short-run equilibria and does not represent dynamic equilibrium.

Because of rising HH’, gK shifts continuously rightwards. The wage rises because of excess gK > gL. But as the wage rises so the excess is recreated. gK = gL at w0 by construction and at wm only.

Figure III Ricardo: the increasing wage trend: initial gK = gL We turn next to the downward stretch of wages, Ricardo’s main concern (Figure IV). In the following passages reference is made to the role of diminishing returns in depressing gK the constancy of gL, and the downward trend of wages as gK falls steadily relatively to gL: Although, then, it is probable, that under the most favourable circumstances, the power of production is still greater than that of population, it will not long continue so; for the land being limited in quantity, and differing in quality, with every increased portion of capital employed on it, there will be a decreased rate of production, whilst the power of population continues always the same. (Ricardo 1951, I, 98) In the natural advance of society, the wages of labour will have a tendency to fall, as far as they are regulated by supply and demand; for the supply of labourers will continue to increase at the same rate, whilst the demand for them will increase at a slower rate. If, for instance, wages were regulated by a yearly increase of capital, at the rate of 2 per cent., they would fall when it accumulated only at the rate of 1 ½ per cent. They would fall still lower when it increased only at the rate of 1, or ½ per cent., and would continue to do so 231

RICARDO - THE NEW VIEW

until the capital became stationary, when wages would also become stationary, and be only sufficient to keep up the numbers of the actual population. (Ricardo 1951, I, 101) The notion that wages are ‘regulated by [the] yearly increase of capital’ conflicts with the notion in the same paragraph that they are ‘regulated by supply and demand’ unless the labour growth rate is taken to be a constant, which indeed is Ricardo’s presumption (with one qualification to be taken up presently). The wage falls because workers ‘[find] it more difficult to maintain the market rate of wages above their natural [subsistence] rate’ (102). But as in the upward trend generated by increasing returns (or a steadily shifting marginal product curve) so in the reverse trend the labour market may be cleared in the short run notwithstanding the deviations between gK andgL. Ricardo’s presumption that ‘the supply of labourers will continue to increase at

Steady-state equilibrium applies over the range L1 → L3 where MP is assumed constant; and at E where gL = gK = 0. The downward path is the locus of short-run equilibria and does not represent dynamic equilibrium.

As gK shifts continually inwards, the wage falls. But at no time is gK = gL except at B and C; the gL curve does not trace out long-run equilibria. As the wage falls in response to gK< gL so the divergence is recreated. The entire movement from wm to ws is a response to divergence gK gL. The gL curve can be generalized to avoid the ‘bend’ with no alteration in the basic logic.

The Market Period equiibrium wage: Period by period the rightward shift in L is greater than that in C so that w declines.

Figure IV Ricardo: the decreasing wage trend: initial gK = gL (continuation of Figure II or III) 232

THE WAGE PATH

the same rate’ notwithstanding reduced wages, implies a precipitous fall in gL from some positive constant to zero when the wage reaches subsistence. There is, however, reason to believe that allowance was made for some response in gL to wage reductions from levels above but close to subsistence. For Ricardo (I, 101) seems to have been aware of the fact that otherwise ‘the labourer ... would be soon totally deprived of subsistence’, and saw some deceleration in the decline as realistic, a deceleration that can only be assured by a reduction in gL. A generalized positively sloping gL-w relationship might have been utilized throughout without changing the substance of the general argument. MALTHUS: THE BASIC MODEL A clear statement of a ‘dynamic equilibrium’ or balanced growth wage path – a path satisfying the conditions gK = gL – will be found in Malthus’s Principles (1820). There are four propositions in the relevant passage: (1) that in a system entailing population growth the real wage must exceed w s; (2) that in consequence of land scarcity (diminishing returns) the excess w > w s must fall to zero; (3) that the profit rate declines steadily, as proportional wages rise, to r*; and (4) that assuming gK = gL at all times, this common growth rate must decline: the supposition ... of a constant uniformity in the real wage of labour is not only contrary to the actual state of things, but involves a contradiction. The progress of population is almost exclusively regulated by the quantity of the necessaries of life actually awarded to the labourer; and if from the first he has no more than sufficient to keep up the actual population, the labouring classes could not increase, nor would there be any occasion for the progressive cultivation of poorer land. On the other hand, if the real wages of labour were such as to admit of and encourage an increase of population, and yet were always to remain the same, it would involve the contradiction of a continued increase of population after the accumulation of capital, and the means of supporting such an increase had entirely ceased. We cannot then make the supposition of a natural and constant price of labour, at least if we mean by such a price, an unvarying quantity of the necessaries of life. And if we cannot fix the real price of labour, it must evidently vary with the progress of capital and revenue, and the demand for labour compared with the supply. We may however, if we please, suppose a uniform progress of capital and population by which is not meant in the present case the same rate of progress permanently, which is impossible, but a uniform progress towards the greatest practicable amount, without temporary accelerations or retardations . . . 233

RICARDO - THE NEW VIEW

. . . if poorer lands which required more labour were successively taken into cultivation, it would not be possible for the corn wages of each individual labourer to be diminished in proportion to the diminished produce; a greater proportion of the whole would necessarily go to labour; and the rate of profits would continue regularly falling till the accumulation of capital had ceased. Such would be the necessary course of profits and wages in the progressive accumulation of capital, as applied to the progressive cultivation of new and less fertile land, or the further improvement of what had before been cultivated; and on the supposition here made, the rates both of profits and of real wages would be highest at first, and would regularly and gradually diminish together, till they both came to a stand at the same period, and the demand for an increase of produce ceased to be effective. (Malthus 1951, 255f)

The wage corresponds to g K = g L at all times. There is no competitive mechanism to assure this result so that the wage path is purely a reference path. Note: Samuelson (1978) posits constant K/L ratio technologically. Figure V Malthus: the decreasing wage trend: g K = g L Wage movements due to divergencies of g L and g K are distinctly contrasted with those reflecting balanced growth and treated in a separate section (285f). 234

THE WAGE PATH

Malthus presents the dynamic equilibrium path as a ‘supposition’, implying thereby a hypothetical reference path. And indeed this path cannot (assuming regular capitaland labour-supply growth functions) be achieved in a competitive world where capitalists and labourers act independently. For since r = F’w(L) (1 r moves inversely to proportional not absolute wages) it follows that, in an expanding system, the initial impact of diminishing returns at the going wage is on capitalists alone inducing a fall in g K relative to g L; and assuming a regular decline in the marginal product the excess g L > g K is continually reconstituted notwithstanding the wage decline. This was Ricardo’s insight. The dynamic-equilibrium path can only be, therefore, a construct of the mind derived by asking what the wage path would be if it is assumed that g L = gK under conditions of diminishing returns. Evidently the return to the ‘joint’ factor declines; but the share of the incidence of declining productivity can be calculated by reference to the capital-supply and labour-supply growth functions, namely the g K - r and the g L - w relationships. There is a unique wage rate for each marginal product which is consistent with balanced factor growth as is illustrated by Figure V. THE MALTHUS–MILL PRUDENTIAL WAGE PATH In the Essay on Population Malthus investigated the implications for wages of ‘prudence’. In the following passage we have a contrast between that check to population growth exercised by falling ‘corn’ wages and that check exercised by deliberate constraint designed precisely to avoid the deterioration of living standards. That the population growth rate must in one way or the other decline is a necessary implication of land scarcity. Correspondingly, two categories of ‘stationary state’ are defined – the one entailing a ‘low’ corn wage and the other a ‘high’ corn wage: A diminished power of supporting children is an absolutely unavoidable consequence of the progress of a country towards the utmost limits of its population. If we allow that the power of a given quantity of territory to produce food has some limit, we must allow that as this limit is approached, and the increase of population becomes slower and slower, the power of supporting children will be less and less, till finally, when the increase of produce stops, it becomes only sufficient to maintain, on an average, families of such a size as will not allow of a further addition of numbers. This state of things is generally accompanied by a fall in the corn price of labour; but should this effect be prevented by the prevalence of prudential habits among the lower classes of society, still the result just described must take place; and though, from the powerful operation of the preventive check to increase, the wages of labour estimated even in corn might not be low, yet it is obvious that, in this case, the power of supporting children would rather be 235

RICARDO - THE NEW VIEW

nominal than real; and the moment this power began to be exercised to its apparent extent, it would cease to exist. (Malthus 1890, 420; first introduced in 5th edn, 1817) The two Malthusian versions were effectively utilized in response to a proposal by Arthur Young ‘so to adjust the wages of day-labour as to make them at all times equivalent to the purchase of a peck of wheat’ (Appendix, 1890 [1817] 583). This proposal ‘in its general operation, and supposing no change of habits among the labouring classes’, Malthus objected, ‘would be tantamount to saying that, under all circumstances, whether the affairs of the country were prosperous or adverse; whether its resources in land were still great, or nearly exhausted; the population ought to increase exactly at the same rate – a conclusion which involves an impossibility’. Here we have an application of the ‘standard’ model – the wage path must fall to assure the appropriate deceleration in the population growth rate. But allow for prudence and the picture is transformed, as Malthus proceeds to show: ‘If, however, this adjustment, instead of being enforced by law, were produced by the increasing operation of the prudential check to marriage, the effect would be totally different, and in the highest degree beneficial to society. A gradual change in the habits of the labouring classes would then effect the necessary retardation in the rate of increase, and would proportion the supply of labour to the effective demand, as society continued to advance ... without the pressure of a diminishing quantity of food’ (583). We turn now to J.S. Mill. Mill is explicit that ‘the greater the profit that can be made from capital, the stronger is the motive to its accumulation’ (1965, 161) and that it is ‘an almost infallible consequence of any reduction of profits, to retard the rate of accumulation’ (843). Moreover, when the minimum profit rate has been reached ‘no further increase of capital can for the present take place’ (738). In early states of society marginal product is high and constant and gK proceeds at the same rate as the maximum physiological capacity of population growth gL (and may possibly exceed it): In countries like North America and the Australian colonies, where the knowledge and arts of civilized life, and a high effective desire of accumulation, co-exist with a boundless extent of unoccupied land, the growth of capital easily keeps pace with the utmost possible increase of population, and is chiefly retarded by the impracticability of obtaining labourers enough. All, therefore, who can possibly be born, can find employment, without overstocking the market: every labouring family enjoys in abundance the necessaries, many of the comforts, and some of the luxuries of life ... (Mill 1965, 343–4) 236

THE WAGE PATH

In the absence of this special confluence of circumstances it would be impossible for population to expand at its maximum capacity without downward pressure on the wage, and this because of impediments to gK emanating from land scarcity: ‘The increase of capital is checked, because there is not fresh land to be resorted to, of as good quality as that already occupied.’ All this was ‘standard’ Ricardian analysis. But Mill also allowed for a case where population and capital proceed at the same rate implying a constant commodity wage: ‘We shall suppose them ... to increase with equal rapidity; the test of equality being, that each labourer obtains the same commodities as before, and the same quantity of those commodities’ (723). This is the case of dynamic equilibrium at constant wages already recognized by Malthus. The consequence of such expansion in conditions of land scarcity is necessarily a fall in the profit rate: ‘The labourer obtaining the same amount of necessaries, money wages have risen; and as the rise is common to all branches of production, the capitalist cannot indemnify himself by changing his employment, and the loss must be borne by profits.’ But we know that a fall in profits generates a decline in gK so that it must be presumed that gL too is declining. Ongoing population growth is possible at constant wages even in conditions of land scarcity provided gL declines simultaneously with gK. The downward trend in wages – inevitable if population grows at the maximum physiological rate characterizing the unskilled labouring class of contemporary Britain – could only be avoided by population control, i.e. ‘prudence’. The foregoing constant wage path reflects the adoption of such behaviour (for details see Hollander 1984). In principle, ‘prudence’ can take two forms in the Mill–Malthus model: a) Endogenous: In this case workers deliberately constrain the population growth rate to prevent the wage falling below some designated ‘conventional’ level. In effect, the labour supply function becomes horizontal at w , and this wage traces out the c ‘dynamic equilibrium’ path. For population growth to fall and prevent any wage decline implies ascribing to labour prescience of the steady decline in gK. Since the model was designed to portray the effect of an instillation of responsible habits this strict assumption makes considerable sense at least as a limiting case. If, however, allowance is made for a lag in the response of gL the actual path will sag below the horizontal; and will not be one of ‘dynamic equilibrium’. The notion of a subsistence wage as that wage at which gL = 0 is no longer relevant since the conventional wage is consistent both with gL= 0 andgL> 0. But if subsistence is defined as that wage below which gL< 0 it can be given independent status (ws < wc) (see Figure VI). b) Exogenous: Here population control is reflected in shifts of the positively sloped gL-w function generated by ongoing educational propaganda programmes. Whereas in case a) prudence is embodied in labour’s given reaction pattern, now changes in 237

RICARDO - THE NEW VIEW

habits are required. The outcome is as before – a ‘dynamic equilibrium’ constant wage path – but now the notion of a subsistence wage at which g L= 0 retains its validity for each given labour-supply relationship. Unlike Malthus’s ‘standard’ case where the wage path declines towards the constant subsistence level, the subsistence level rises towards the constant wage path which it meets at E, the ‘stationary state’ (see Figure VII). CONCLUSION An often repeated opinion has it that ‘hard-line’ Malthusianism constitutes the core of the classical theory of distribution: ‘their distribution model only had content to the extent that it could derive this from a “hard” empirical Malthusian proposition’ (Hutchison 1978, 71 and references there cited). This is a questionable view. J.S. Mill, and before him Malthus, based policy conclusions on a growth model which allowed for the effects of ‘prudential’ population control on wages, and was specifically designed to serve an exhortatory purpose, namely to show how wages might be maintained (even increased) despite land scarcity by recognition of the alternatives – uncontrolled population growth and the depression of wages to some ‘subsistence’ minimum, or their maintenance at some ‘cultural’ level. The model yields two categories of Stationary State – one consistent with ‘low’ and the other with ‘high’ wages. The theory of distribution turns out to be a powerful one indeed, capable of dealing with ‘hard’ and ‘soft’ Malthusianism.

238

THE WAGE PATH

From B to E, g K = g L if lagged population growth is assumed away. The dashed line represents the original g Lfunction with [ w ] the s subsistence wage prior to the adoption of prudential habits.

An initial phase of constant MP is allowed generating steady-state wages. The declining stretch L1→L2 reflects the positively sloped section of gL. The Mill–Malthus contribution relates to the constant dynamic equilibrium path L2 → L4.

Figure VI Malthus–Mill: the constant wage trend; ‘prudence’ endogenous

Steady state wage, L0 ? L1 reflects constant MP. Thereafter wage constancy is due to shifting gL in line with g K . It is assumed that the original subsistence wage applies until L1. Unlike Hicks and Hollander where

the wage path declines to subsistence, now the subsistence wage ‘rises’ to E.If gL shifts leftwards more rapidly then g K a rising wage trend will be generated (as Malthus recognized).

Figure VII Malthus–Mill: the constant wage trend: ‘prudence’ exogenous 239

RICARDO - THE NEW VIEW

REFERENCES Hicks, J.R. and Hollander, S. (1977) ‘Mr. Ricardo and the Moderns’, Quarterly Journal of Economics 91, 3 (August): 351–69. Hollander, S. (1984) ‘Dynamic Equilibrium with Constant Wages: J.S. Mill’s Analysis of the Secular Wage Path’, Kyklos 37: 247–65. Hutchison, T.W. (1978) On Revolutions and Progress in Economics, Cambridge: Cambridge University Press. Malthus, T.R. (1890) Essay on the Principle of Population. Reprinted from the Last Edition Revised by the Author (1826), London: Ward, Lock. —— (1951) Principles of Political Economy (1820) in Works and Correspondence of David Ricardo, Vol. II, Cambridge: Cambridge University Press. Mill, J.S. (1965) Principles of Political Economy in Collected Works, vols II, III, Toronto: University of Toronto Press. Autobiography in Collected Works, vol. I, Toronto: University of Toronto Press. —— (1981) Ricardo, David (1951–73) Principles of Political Economy in The Works and Correspondence of David Ricardo, Vol. I, and Notes on Malthus’s Principles of Political Economy in Vol. II,Cambridge: Cambridge University Press. Samuelson, P.A. (1978) ‘The Canonical Classical Model of Political Economy’, Journal of Economic Literature XVI (December): 1415–34.

240

15 RICARDIAN GROWTH THEORY: A RESOLUTION OF SOME PROBLEMS IN TEXTUAL INTERPRETATION*

My concern in this paper is to confirm the validity of the attribution to Ricardo of a growth model entailing secularly declining real wages. In this model both the realwage and profit rates are above their respective minima during the course of expansion; and both decline to those minima in consequence of the pressures exerted by increasing land scarcity (e.g. Tucker 1960, 110–16; Casarosa 1985; Hollander 1987). This interpretation – the so-called ‘New View’ – has been rejected by various commentators in favour of a subsistence-wage growth model (cf. Pasinetti 1960; Stigler 1981; Peach 1988). There is a danger that the protagonists will speak past each other, and fail to bring the issue to a conclusion. This would be unfortunate, for there is strong evidence – not yet all taken into account – pointing to the variablewage interpretation. That interpretation is preferable (1) because, unlike the alternative, it satisfies certain minimum, common sense, requirements of textual exegesis, and (2) because the alternative is subject to serious analytical error. I. ON TEXTUAL EXEGESIS A word first regarding the celebrated rule of ‘scientific exegesis’ recommended in Stigler’s paper ‘Textual Exegesis as a Scientific Problem’: We should not be so literal-minded as to count the passages in a book to decide an author’s general position because the passages are not of equal importance. We increase our confidence in the interpretation of an author by increasing the number of his main theoretical conclusions which we can deduce from (our interpretation of) his analytical system. The test of an interpretation is its consistency with the main analytical 241

RICARDO - THE NEW VIEW

conclusions of the system of thought under consideration. If the main conclusions of a man’s thought do not survive under one interpretation, but do under another, the latter interpretation must be preferred. (The analogy to maximum likelihood is evident.) (Stigler 1982 [1965], 69) Here Stigler adds that the interpretation must reveal ‘the main concepts in the man’s work, and the major functional relationships between them’, but ‘need not account for careless writing or unintegrated knowledge’. For Stigler (1981), Ricardo’s argument regarding the growth process requires the assumption of constant secular wages, since ‘without this assumption the fundamental theorem on distribution (only a rise in wages will lower profits) is not rigorously true’. Stigler recognizes ‘a considerable number of passages ... in which the assumption, of constant long-run wages is qualified or ignored’, but he treats them as ‘inconsistent views’. Stigler’s approach – its insistence on ‘rigour’ – is potentially dangerous at least from a purely historiographical perspective. For it implies that we are not strictly concerned with what Ricardo himself maintained. This implication of Stigler’s rule of scientific exegesis is confirmed by a contrast drawn with ‘personal exegesis’: This rule of interpretation is designed to maximize the value of a theory to the science. The man’s central theoretical position is isolated and stated in a strong form capable of contradiction by the facts. The net scientific contribution, if any, of the man’s work is thus identified, amended if necessary, and rendered capable of evaluation and possible acceptance. This rule of consistency with the main conclusions may be called the principle of scientific exegesis. Of course men make logical errors or slip into tautologies and otherwise blemish their work. One may seek to determine what the man really believed, although this search has no direct relevance to scientific progress. One will then invoke a different criterion to choose between conflicting passages: that interpretation which fits best the style of the man’s thought becomes decisive. This may be called the principle of personal exegesis. (Stigler 1982 [1965], 69; emphasis added) Evidently, ‘scientific exegesis’ is not necessarily concerned with what the subject ‘really believed’, but in a theoretical statement formulated ‘in a strong form’ and ‘amended if necessary’. Now it is surely conceivable that an author may have deliberately adopted a ‘weakened’ theorem. In the case that now concerns us – realwage movements in a progressive economy – the impact on profits of increasing land scarcity is lessened if the real wage declines, although the declining profit rate cannot be prevented as I shall explain presently. But in fact, in this case, much more is involved: the constant242

RICARDIAN GROWTH THEORY

wage growth model makes no analytical sense except under special conditions relating to ‘prudential’ population control, conditions which Ricardo did not specify; without these conditions a declining real wage is a necessary feature of the growth process in the presence of increasing land scarcity. As for the inverse wage– profit relation (the ‘wage’ in this context is the ‘gold’ or ‘proportional’ wage), it is continuously at play and indeed accounts for the fall in the profit rate: the fundamental theorem is rigorously true despite falling real wages. The ‘scientific rule of exegesis’ is, however, acceptable provided it is limited to a test of interpretation understood simply as consistency with the main analytical conclusions. We must isolate the central theoretical position from the texts without amendment – even if logical error should be involved, and a fortiori if logical error is not involved. Where we are faced by apparently conflicting passages, we select as the author’s ‘true’ position that formulation which reflects him ‘at his best’, and set it up as standard for the evaluation of other of his statements whether earlier, later, or contemporaneous relating to the same issue. It is then possible, by reference to the author’s own best considered terms, to recognize partial formulations and special cases and also error, backslidings and other infractions. The hallmark of a good interpretation is its convincing rationalization of all such ‘residual’ passages. Stigler is, of course, right that ‘passages are not of equal importance’, but that cannot justify jettisoning as ‘unintegrated knowledge’ those that do not conform with the main analytical conclusions and yet cannot be classed as obvious error. For those who maintain the variable-wage interpretation the indubitable existence at various junctures of the constant-wage assumption constitutes the ‘residual’ that we must account for convincingly. Here what Stigler calls ‘personal exegesis’ helps: ‘The simplifying assumption was Ricardo’s trademark’ (1982, 70). So indeed it was. But this fact does not conflict with the equally factual circumstance that his general model of growth involves wage variability. The simplifying constant-wage assumption is adopted for various specific reasons depending on context (which we outline below) but not when wage variability is a central and logically necessary feature of the argument. That Ricardo’s style involves ‘the simplifying assumption’ is no excuse for reading out substantial parts of his main text. ‘Scientific’ and ‘personal’ exegesis must complement each other. Dr Terry Peach, in a recent ‘Review of Some Interpretative Issues’ relating to Ricardo, also concludes that ‘the new view cannot stand as a representation of Ricardo’s true position’ (1988, 114). Like Stigler, he recognizes that texts pointing to the New View interpretation do exist in the Principles : ‘there are undoubtedly passages in the chapter “On Wages” which lend it superficial credibility’; in that chapter Ricardo ‘maps out a trajectory for wages that points directly in the new-view direction’ (111– 12). But he suggests ‘that the principal new-view passages may have been an attempt to show, on Malthus ’ premises, that a declining real wage need not compensate for 243

RICARDO - THE NEW VIEW

rising corn prices’ (114), implying that the concept of a declining real wage itself is not Ricardo’s at all. It is also allowed, again as ‘a possibility’, that Ricardo ‘had at one time [in pre- Principles correspondence, see below, p. 257] accepted Malthus’s vision of the accumulation process although, to borrow an expression from Stigler (1965), it could only ever have been “unintegrated knowledge”, irreconcilable with the natural wage doctrine’. So the falling-wage version, we are asked to believe, though formulated by Ricardo is not Ricardo’s, while to the extent he may ‘at one time’ have accepted it he was unable to deal with it analytically. As noted, Peach concludes unambiguously that ‘the new view cannot stand as a representation of Ricardo’s true position’. His substantive case turns on the view that Ricardo was concerned ‘with “permanent” movements in profitability’ reflecting the higher real cost of producing food (107), where the ‘natural wage’ is treated ‘as an active center of gravity with an unrestricted domain of influence’ rather than an influence exerted solely in stationary state conditions (112). ‘The picture is one of market wages oscillating around the natural level’ though not necessarily or always with near-instantaneous adjustment of population. The main point is that ‘permanent movements in profitability’ cannot be caused by real-wage variation; ‘the impact of real-wage variations on profitability is ... relegated to a temporary status’ (113). And in summary: There is nothing baffling about this as long as we stick with the traditional interpretation, suitably qualified to allow for cases of tardy population adjustment. However, if the natural wage is not a potent center of gravity at all times – Ricardo’s position according to the new view – his general thesis, that permanent movements in profitability are given exclusively by changed conditions of producing wage-goods, loses the significance it was evidently thought to possess: temporary influences on profitability from real-wage variations would be of equal importance with permanent ones. We must turn then to the texts used by Peach in favour of the ‘old view’ formulated rigorously by Pasinetti (1960). I shall show that his treatment (like Stigler’s) fails to satisfy the primary requisite of exegesis which is to get the model right on the author’s own terms. His error involves the identification of the falling wage trend in Ricardian theory with wage fluctuations about the trend as far as concerns profit-rate determination. That Ricardo, interested as he was with secular movements, often set aside fluctuations – or insisted on the speedy adjustment of population to disturbances – in no way implies subscription to a subsistence wage (or even a constant wage) growth model. Similarly, the New View has no difficulty at all in making sense of unambiguous declarations that the profit rate depends solely on the cost of producing wage goods. 244

RICARDIAN GROWTH THEORY

II. THE PRINCIPLES OF POLITICAL ECONOMY We begin with the celebrated statement in the Principles that ‘However much the market price of labour may deviate from its natural price, it has, like commodities, a tendency to conform to it’ (Ricardo 1951, I, 94; emphasis added). Dr Peach takes the natural wage presented here to be ‘the active center of gravity with an unrestricted domain of influence’ (1988, 112); oscillations of the real wage occur around the ‘natural’ (in the sense of the subsistence) level. The New View by contrast has the real wage drawn down to the subsistence wage in the stationary state but remaining above it during the course of growth. Doubtless in some cases the term ‘tendency’ should be understood in Peach’s sense. In establishing categories and definitions and in undertaking expository exercises the subsistence wage is taken (on a par with the natural price of goods) as a convenient point of departure. For it is the simplest. We have such categories in the following statement which relate to alternative initial states, one where the real wage exceeds subsistence and one when it falls short of subsistence – the correction occurring by way of absolute population increase and decrease respectively: It is when the market price of labour exceeds its natural price, that the condition of the labourer is flourishing and happy, that he has it in his power to command a greater proportion of the necessaries and enjoyments of life, and therefore to rear a healthy and numerous family. When, however, by the encouragement which high wages give to the increase of population, the number of labourers is increased, wages again fall to their natural price, and indeed from a re-action sometimes fall below it. When the market price of labour is below its natural price, the condition of the labourers is most wretched: then poverty deprives them of those comforts which custom renders absolute necessaries. It is only after their privations have reduced their number, or the demand for labour has increased, that the market price of labour will rise to its natural price, and that the labourer will have the moderate comforts which the natural rate of wages will afford. (Ricardo 1951, I, 94) 1 Similarly, the materials given on I, 95–6 which describe the value of the given ‘natural price of labour’ rising, remaining stationary, or falling depending on what occurs upon expansion to the real cost of producing wage goods are a matter of classification. Classification and expository exercises involving static comparisons however, do not constitute a general account of growth. It is the existence of a carefully specified, genuine growth model – i.e. one accounting for on-going capital accumulation and 245

RICARDO - THE NEW VIEW

population expansion – that permits us to recognize, by way of contrast, the limited status of the foregoing formulations. The argument specified in the analysis of ‘an improving country’, commences thus: ‘Notwithstanding the tendency of wages to conform to their natural rate, their market rate may, in an improving society, for an indefinite period, be constantly above it; for no sooner may the impulse, which an increased capital gives to a new demand for labour be obeyed, than another increase of capital may produce the same effect; and thus, if the increase of capital be gradual and constant, the demand for labour may give a continued stimulus to an increase of people’ (I, 94–5). The real wage might thus remain indefinitely above subsistence. But once deceleration of capital growth sets in (with increasing land scarcity) the real wage declines from above subsistence levels to subsistence – the level attained in the stationary state when the profit rate too will have fallen to its minimum with net accumulation reduced to zero: In the natural advance of society, the wages of labour will have a tendency to fall, as far as they are regulated by supply and demand; for the supply of labourers will continue to increase at the same rate, whilst the demand for them will increase at a slower rate. If, for instance, wages were regulated by a yearly increase of capital, at the rate of 2 per cent., they would fall when it accumulated only at the rate of 1 ½ per cent. They would fall still lower when it increased only at the rate of 1, or ½ per cent., and would continue to do so until the capital became stationary, when wages also would become stationary, and be only sufficient to keep up the numbers of the actual population. (Ricardo 1951, I, 101) Thus far Ricardo has been, provisionally, assuming that the decline in the real wage is accompanied by a decline in the ‘money’ wage. But this cannot be so; the downward trend in the real wage is necessarily accompanied by a money-wage increase and therefore a profit–rate fall – the inverse wage– profit relation is at play. We see very clearly that the secular fall in the real wage is not of that order to which can be ascribed a rise in the profit rate: I say that under these circumstances, [money] wages would fall, if they were regulated only by the supply and demand of labourers; but we must not forget, that wages are also regulated by the prices of the commodities on which they are expended. As population increases, these necessaries will be constantly rising in price, because more labour will be necessary to produce them. If, then, the money wages of labour should fall, whilst every commodity on which the wages of labour were expended rose, the labourer would be doubly affected, and would be soon totally deprived of subsistence. Instead, therefore, of the money wages of 246

RICARDIAN GROWTH THEORY

labour falling, they would rise, but they would not rise sufficiently to enable the labourer to purchase as many comforts and necessaries as he did before the rise in the price of those commodities ... he would ... receive an addition to his money wages, though with that addition he would be unable to furnish himself with the same quantity of corn and other commodities, which he had before consumed in his family. Notwithstanding, then, that the labourer would be really worse paid, yet this increase in his wages would necessarily diminish the profits of the manufacturer; for his goods would sell at no higher price, and yet the expense of producing them would be increased. This, however, will be considered in our examination into the principles which regulate profits. ... The fate of the labourer will be less happy [than that of the landlord]; he will receive more money wages, it is true, but his corn wages will be reduced; and not only his command of corn, but his general condition will be deteriorated, by his finding it more difficult to maintain the market rate of wages above their natural rate. (Ricardo 1951, I, 101–2) It is a process involving labour market pressures that dictates the rate of decline of the real wage and profit rates. Those pressures assure that money wages rise sufficiently to prevent the real wage falling to subsistence prematurely. The statement that ‘the supply of labourers will continue to increase at the same rate’ is evidently a first approximation since population growth is responsive to declining real wages; it is precisely such responsiveness that assures that the real wage does not collapse prematurely to subsistence. Ricardo’s full analysis of growth is an impressive achievement. His object was to demonstrate that with real wages initially above the stationary state level and with scope for a decline, the profit rate must still fall with increasing land scarcity. For since the corn wage initially falls short of the marginal product, but ultimately absorbs the marginal product entirely – subject possibly to a positive profit-rate minimum – the proportionate share of wages in the marginal product necessarily rises, and on this share depends the profit rate. Thus the burden of diminishing returns cannot fall entirely on labourers. Moreover, a rise in the proportionate share of wages is reflected in a higher ‘money’ wage provided money satisfies certain Specific conditions. Diminishing returns is necessarily reflected in rising money wages (and thus a falling profit rate) despite a declining corn wage. 2 Despite his unambiguous statement of the falling wage path, when Ricardo proceeds to devise a set of data to illustrate his propositions we find him assuming a constant wage basket – money wages are always such as ‘to enable [the worker] to live just as well, and no better, than before’ (I, 103). His corn wage falls with the rise in the corn price but his full (fixedproportions) basket is constant: ‘In proportion as corn 247

RICARDO - THE NEW VIEW

became dear, he would receive less corn wages, but his money wages would always increase, whilst his enjoyments ... would be precisely the same’ (I, 103–4). Are we justified in describing Ricardo as ‘inconsistent’ because of this procedure? I think not. He has chosen to provide a limited set of data to demonstrate the rise of the money wage with the rise of corn – the simplest he could easily devise. Doubtless a little more effort to make up a full set entailing a falling basket would have been desirable since some readers tend to be more struck by figures than words; but the fact remains that nothing substantive would have been gained (apart from an opportunity to avoid misunderstanding). For the data were designed to show that money wages rise with the corn price. But we know his position, stated a paragraph before, that this is so – this must be so – notwithstanding a fall in the commodity wage. Readers should be able to make up their own data to illustrate the general case. The same set of simplified data is reproduced in the next chapter ‘On Profits’ to demonstrate that the higher money wage is paid out of a marginal product of constant value, necessarily raising labour’s proportion of that constant value and lowering the profit rate. Again Ricardo would have done well to make the effort to devise a more complex set of data to allow for a falling commodity wage, but there is no reason to believe that more than deliberate simplification (or at the worse ‘laziness’) was involved. For in this chapter Ricardo again alludes to the falling realwage trend established earlier: ‘We have shewn that in early stages of society, both the landlord’s and the labourer’s share of the value of the produce of the earth, would be but small; and that it would increase in proportion to the progress of wealth, and the difficulty of procuring food. We have shewn, too, that although the value of the labourer’s portion will be increased by the high value of food, his real share will be diminished ; whilst that of the landlord will not only be raised in value, but will also be increased in quantity’ (I, 112; second italics added). There is one allusion in the chapter to falling real wages that requires special attention: It may be said that I have taken it for granted, that money wages would rise with a rise in the price of raw produce, but that this is by no means a necessary consequence, as the labourer may be contented with fewer enjoyments. It is true that the wages of labour may previously have been at a high level, and that they may bear some reduction. If so, the fall of profits will be checked; but it is impossible to conceive that the money price of wages should fall, or remain stationary with a gradually increasing price of necessaries; and therefore it may be taken for granted that, under ordinary circumstances, no permanent rise takes place in the price of necessaries, without occasioning, or having been preceded by a rise in wages. (Ricardo 1951, I, 118) 248

RICARDIAN GROWTH THEORY

This latter statement refers to the possibility that real wages fall in the event that they happen to be ‘previously at a high level’. This formulation is unnecessarily conciliatory to hypothetical critics if contrasted with those statements which refer to the certainty of a declining real wage. Still, the main analytical result is insisted upon as always: the money wage rises notwithstanding a declining commodity wage so the profit rate necessarily falls. One notes too the remark that the falling commodity wage ‘checks’ the fall in the profit rate; the profit rate will always be higher than it otherwise would be were the commodity wage held constant. Still the profit-rate trend is downward. Again it emerges that the falling commodity wage of the growth model cannot be classified with wage reductions which reflect exogenous changes or cyclical movements or random movements. Such wage reductions might, if severe enough, actually raise (temporarily) the profit rate. To exclude variations of the latter kind, as Ricardo usually does, is not to exclude the secular decline of the real wage reflecting the impact of diminishing returns; that decline cannot prevent the falling profit rate. When Ricardo states his objective at the outset of the chapter ‘On Profits’ to be ‘the cause of the permanent variations in the rate of profit’ (I, 110), his object is to exclude fluctuations in the profit rate (which would evidently reflect fluctuations in the money wage for whatever reason), but not to exclude the falling wage trend. After citing Ricardo’s concern in the chapter ‘On Profits’ with ‘permanent variations in the rate of profit’ Dr Peach proceeds to another quote to prove that such variations could not be caused by real-wage variation: ‘From the account which has been given of the profits of stock, it will appear that no accumulation of capital will permanently lower profits, unless there be some permanent cause for the rise of wages. ... If the necessaries of the workman could be constantly increased with the same facility, there would be no permanent alteration in the rate of profits or wages, to whatever amount capital might be accumulated ’ (I, 289, italicized by Peach 1988, 113). Peach concludes that here and on other occasions in the Principles (cf. I, 132, 296) ‘the impact of realwage variations on profitability is implicitly relegated to a temporary status’. And he elsewhere challenges advocates of the New View to ‘confront’ this fact (1986, 117). The challenge falls flat. The foregoing statement by Ricardo asserts that in the absence of diminishing agricultural returns, there are no forces at play depressing the profit rate. What specifically concerned Ricardo in this context was the exclusion of downward pressure on the profit rate due (1) to rising real wages reflecting labour scarcity, and (2) to inadequate aggregate demand. This is precisely what is captured by the New View. The falling real wage reflects the impact of diminishing agricultural returns; with constant returns there will be steady real-wage and profit rates. To repeat: the falling commodity wage, in the case that real corn costs are rising, is not 249

RICARDO - THE NEW VIEW

to be identified with temporary or short-run fluctuations in the wage; the secular fall is itself a ‘permanent’ variation generated by the impact of rising agricultural costs on the rate of accumulation. Nothing that is said by adherents of the New View points away from, or minimizes the significance of, agricultural conditions in profit-rate determination. Dr Peach (1988, 111–12) admits to the evidence given above in favour of the New View in the Principles, I, 94–5 (and also 101–2) though he asserts that the texts offer ‘superficial credibility’ only. But those same pages are said also to contain material suggesting the subsistence wage to be a true centre of gravity with the secular path following that wage (above, p. 244). If this were so, we would have to ascribe to Ricardo two conflicting growth models within the passage of only a few paragraphs. I suggest the explanation I have given rings truer. We must distinguish the statement of categories or classificatory matter from the full general growth model; and in the case of the texts on pages 101f, distinguish the simplified hypothetical data from the full model. I shall proceed with the exegesis. Dr Peach (1988, 112) cites the following passage from ‘Taxes on Raw Produce’ as evidence of rapid wage adjustment to the subsistence level, and therefore favouring the Pasinetti interpretation: A tax ... on raw produce, and on the necessaries of the labourer would ... raise wages. From the effect of the principle of population on the increase of mankind, wages of the lowest kind never continue much above that rate which nature and habit demand for the support of the labourers. This class is never able to bear any considerable proportion of taxation; and, consequently, if they had to pay 8s. per quarter in addition for wheat and in some smaller proportion for other necessaries, they would not be able to subsist on the same wages as before, and to keep up the race of labourers. (Ricardo 1951, I, 159) This view is unconvincing. In the first place, only two pages later we find Ricardo generalizing, indicating that the subsistence wage of the page 159 passage is a special case. The constant real wage is not necessarily the subsistence wage: ‘Those who maintain that it is the price of necessaries which regulates the price of labour, always allowing for the particular state of progression in which the society may be, seem to have conceded too readily, that a rise or fall in the price of necessaries will be very slowly succeeded by a rise or fall of wages’ (I, 161; emphasis added). 3 Moreover, in the context of the growth process – involving a ‘gradually increasing demand [for produce], which may be ultimately attended with an increased cost of production’ 4 – Ricardo also introduces the notion of real-wage fluctuations about what is termed the ‘natural level’, in consequence of excessive labour-supply adjustments: ‘It generally happens 250

RICARDIAN GROWTH THEORY

indeed, that when a stimulus has been given to population, an effort is produced beyond what the case requires; the population may be, and generally is so much increased as, notwithstanding the increased demand for labour, to bear a greater proportion to the funds for maintaining labourers than before the increase of capital. In this case a re-action will take place, wages will be below their natural level, and will continue so, till the usual proportion between the supply and demand has been restored’ (I, 164). On this passage Peach writes that the ‘natural wage is treated as an active center of gravity with an unrestricted domain of influence’ (1988, 112), though he accepts the passage on page 161 as supporting the New View (his note 12). Again we are asked to envisage the chaotic adoption of two distinct models with Ricardo switching from Pasinetti’s view to the New View and back again within the space of five pages. There is no need to take this unlikely route. There is a more convincing explanation. The passage on page 161 assumes constant real wages assured by a ‘particular state of progression’; the capital growth rate (positive, negative, or zero) is thus assumed to be given. Ricardo here neglects the deceleration of accumulation outlined earlier in the Principles (see above, p. 246) which generates the secular decline in the real wage. Conceivably he had simply forgotten that complexity; alternatively, he had introduced a deliberate simplification. And this latter possibility would not be surprising. Assume the case of positive growth. Here the real wage exceeds subsistence. The wage variation described on page 164 then occurs around this level, with an initial speeding up of accumulation rather than an absolute increase in capital. This is sufficiently complex. But add to it the complexity of a positive but decelerating growth rate and one can appreciate why it might have been set aside. It is for the reader to make the necessary amplification as is always the case with any ‘textbook’. Whatever the reason for the neglect of decelerating accumulation, such neglect indicates that the analysis of taxation is not undertaken within the confines of a full-fledged growth model. This episode provides a further lesson: it warns us to avoid word-mindedness. The ‘natural rate’ often refers to the subsistence wage. But positively not here (on p. 164); here it is the rate which assures population growth in line with the specified capital growth rate. The fluctuations are not about ‘subsistence’. I turn now to the chapter ‘Taxes on Wages’ to consider a passage also admitted by Dr Peach (his note 12) to support the New View: ‘The demand for labour, according as it happens to be either increasing, stationary, or declining, or to require an increasing, stationary, or declining population, regulates the subsistence of the labourer, and determines in what degree it shall be either liberal, moderate, or scanty’ (1951, I, 215). Here Ricardo cites Adam Smith, and as on page 161 takes three discrete categories of cases – positive, zero, or negative accumulation – thus avoiding 251

RICARDO - THE NEW VIEW

the full growth model with its deceleration of growth. Once again he simplifies the analysis of wage taxation to avoid the complexity of a falling real wage. It is for the reader to do a little work of their own. Certain passages are said by Dr Peach (1988, 112) to support the traditional or Pasinetti view by their emphasis on speedy adjustment of the wage to the ‘natural’ (read: subsistence) wage. Ricardo 1951, I, 159 (above, p. 250) is one such. In fact that passage says nothing about the speed of adjustment. But consider a reference to the text on page 125: ‘Whilst the land yields abundantly, wages may temporarily rise, and the producers may consume more than their accustomed proportion; but the stimulus which will thus be given to population, will speedily reduce the labourers to their usual consumption.’ Certainly there is reference here to speedy adjustment. But it is not to the subsistence wage, but rather to ‘the usual’ or ‘accustomed’ level. One notes that the case assumes a state of society not subject yet to land scarcity, reinforcing the impression of a real wage initially above subsistence. Ricardo then elaborates for the general case – with the onset of diminishing returns: But when poor lands are taken into cultivation, or when more capital and labour are expended on the old land, with a less return of produce, the effect must be permanent. A greater proportion of that part of the produce which remains to be divided, after paying rent, between the owners of stock and the labourers, will be apportioned to the latter. Each man may, and probably will, have a less absolute quantity ; but as more labourers are employed in proportion to the whole produce retained by the farmer, the value of a greater proportion of the whole produce will be absorbed by wages, and consequently the value of a smaller proportion will be devoted to profits. This will necessarily be rendered permanent by the laws of nature, which have limited the productive powers of the land. (Ricardo 1951, I, 125–6; emphasis added) Peach’s exegesis takes the passage from page 125 as evidence for the old view because of its allusions to speedy adjustment. But on page 126 we have an explicit statement of falling real wages (and a falling profit rate) due to diminishing returns. The New View is supported directly; and it is clear that the fluctuations issue – whether speedy or slow – is irrelevant for the trend of the wage path. Finally, we turn to Ricardo I, 16 and an allusion to a reduction in the real costs of producing wage goods, taken by Dr Peach (1988, 112) as proof of rapid adjustment and the subsistence wage path: ‘it is probable his wages would in no long time be adjusted by the effects of competition, and the stimulus to population, to the new value of the necessaries on which they were expended. If these improvements extended 252

RICARDIAN GROWTH THEORY

to all the objects of the labourer’s consumption, we should find him probably at the end of a very few years in possession of only a small, if any, addition to his enjoyments.’ Dr Peach is mistaken. The rapid adjustment relates to some constant real wage but not necessarily at the subsistence level. But again the matter of fluctuations has nothing to do with the trend path. Much has been made by commentators of a statement by Ricardo to Malthus regarding their respective methods. It is important for us to have it in mind in considering the Principles : It appears to me that one great cause of our difference in opinion, on the subjects we have so often discussed, is that you have always in mind the immediate and temporary effects of particular changes – whereas I put these immediate and temporary effects quite aside, and fix my whole attention on the permanent state of things which will result from them. Perhaps you estimate these temporary effects too highly, whilst I am too much disposed to undervalue them. To manage the subject quite right they should be carefully distinguished and mentioned, and the due effects ascribed to each. (24 January 1817; Ricardo 1951, VII, 120) Dr Peach illustrates Ricardo’s representation of the ideal procedure by his allowance for cases where ‘wages are above or below the natural level, with oscillations and time-consuming population adjustments’ (1988, 114). But this is unconvincing. Ricardo is alluding rather to the short-run and the long-run impacts of given changes in conditions (say new technology), his own preference being for the investigation of the latter, though conceding that ideally he should deal with both categories of effects. While it is not clear to me that he had in mind the need to deal with ‘oscillations’, that is how Malthus understood him, apparently confusing causes with effects; at least he took the opportunity to state his own preferences in terms of an interest in ‘irregular movements’: ‘I really think that the progress of society consists of irregular movements, and that to omit the consideration of causes which for eight or ten years will give a great stimulus to production and population, or a great check to them, is to omit the causes of the wealth and poverty of nations – the grand object of all enquiries in Political Economy’ (26 January 1817; Malthus, in Ricardo 1951, VII, 122). And he alluded by way of illustration to Ricardo’s fixed-wage assumption in the Essay on Profits : A writer may, to be sure, make any hypothesis he pleases; but if he supposes what is not at all true practically, he precludes himself from drawing any practical inference from his hypothesis. In your essay on profits you suppose 253

RICARDO - THE NEW VIEW

the real wages of labour constant; but as they vary with every alteration in the price of commodities, (while they remain nominally the same) and in reality are as variable as profits, there is no chance of your inferences being just as applied to the actual state of things. We see in all the countries around us, and in our own particularly, periods of greater or less prosperity, and sometimes of adversity, but never the uniform progress which you seem alone to contemplate. (26 January 1817; Malthus, in Ricardo 1951, VII, 122) This criticism of Ricardo does not make a case for a falling secular real wage in contrast to Ricardo’s fixed wage, but only for concern with fluctuations in the real wage (brought about by fluctuations in the corn price relative to a stable money wage). In the Principles Ricardo continues in his own fashion to ignore such considerations in his analysis of profits: ‘The reader is aware, that we are leaving out of our consideration the accidental variations arising from bad and good seasons, or from the demand increasing or diminishing by any sudden effect on the state of population. We are speaking of the natural and constant, not of the accidental and fluctuating price of corn’ (I, 115n). That, in effect, was Ricardo’s reply to Malthus. By choice, his concern remained with the upward secular pressures on the price of corn, and the implications thereof, pre-eminently the falling paths of the real-wage and profit rates. He made no ‘concessions’ to Malthus. III. ‘ON GROSS AND NET REVENUE’ Stigler’s case for a fixed-wage interpretation (above, pp. 241–2) includes the argument that without that assumption Ricardo’s chapter ‘On Gross and Net Revenue’ and his proposition that the disposable surplus excludes wages are ‘wrong’ (1981, 101). Surely ‘wrong’ is not the right word. Ricardo indeed sets out in this chapter with the resounding theme that ‘It is from the two last portions only [profits and rent], that any deductions can be made for taxes, or for savings’ (Ricardo 1951, I, 347–8). But we must consider the context. His object was to controvert what he read to be Adam Smith’s inclusion within disposable surplus of the entire national income. He pointed out the need to correct for ‘necessary expenses’. And following his predilection for strong cases (here ‘personal exegesis’ is certainly in order) he did so in the most forcible fashion. But nothing was lost – except perhaps effect – by generalizing, as he himself pointed out in the Principles, to cases where wages contained part of the disposable surplus: ‘Perhaps this is expressed too strongly, as more is generally allotted to the labourer under the name of wages, than the absolutely necessary expenses of production. In that case a part of the net produce of the country is received by the 254

RICARDIAN GROWTH THEORY

labourer, and may be saved or expended by him; or it may enable him to contribute to the defence of the country’ (I, 348n). Similarly, in refutation of Malthus’s misunderstanding: ‘I do not deny that wages may be such as to give to the labourers a part of the net revenue – I limited my proposition to the case when wages were too low to afford him [sic] any surplus beyond absolute necessaries’ (II, 381). Allowing for wages in excess of the subsistence wage rules out the identification of the disposable surplus with the returns to property owners. But then, so what? The collapse of a ‘rigorous’ notion of surplus does not, as Stigler apparently believes, preclude the growth model involving abovesubsistence, though declining, wages. IV. THE EARLY CORRESPONDENCE Malthus allowed early on for falling wages during secular growth. He refers for example, to ‘the constant tendency to a fall in the wages of labour’ (23 November 1814, Ricardo 1951, VI, 152); but he does so in the context of a denial that the state of cultivation ‘ regulates ’ the profit rate. In reply Ricardo gave his own position. He is explicit that his falling profit rate presumes falling real wages per capita : Accumulation of capital has a tendency to lower profits. Why? because every accumulation is attended with increased difficulty in obtaining food, unless it is accompanied with improvements in agriculture; in which case it has no tendency to diminish profits. If there were no increased difficulty, profits would never fall, because there are no other limits to the profitable production of manufactures but the rise of [money] wages. If with every accumulation of capital we could tack a piece of fresh fertile land to our Island, profits would never fall. ... A diminution of the proportion of produce, in consequence of the accumulation of capital, does not fall wholly on the owner of stock, but is shared with him by the labourers. The whole amount of wages paid will be greater, but the portion paid to each man, will in all probability, be somewhat diminished. (18 December 1814; Ricardo 1951, VI, 162–3; emphasis added) Here then in late 1814 we already have what was to be the full argument of the Principles : a falling profit rate due to rising money wages (the inverse wage–profit relation) accompanied by declining real wages – the burden of diminishing returns is shared between capital and labour. I turn now to an earlier passage from the correspondence cited by Dr Peach in support of his rejection of the New View. ‘Nothing’, wrote Ricardo on 8 March 1814, ‘can increase the profits permanently on trade ... but a really cheaper mode of obtaining food’ (VI, 104; cited by Peach 1986, 107). On Peach’s reading this statement 255

RICARDO - THE NEW VIEW

indicates that ‘permanent movements must be defined relative to the subsistence wage’. This is going too far: by its neglect of real-wage variability the statement may imply a constant real wage, though a real wage above subsistence might be intended. But even if we assume that early in 1814 Ricardo had adopted the falling real-wage secular trend recognized in December that year, it would not be surprising to find him neglecting that matter in dealing with the profit rate. For, as explained above, that secular trend does not act to raise the profit rate, but reflects the same force by which profits are forced down – diminishing returns. Furthermore, neglect of shortrun wage fluctuations in this specific context is also not surprising, for the position Ricardo contested was the Smithian case held by Malthus regarding the positive impact on profits of new markets, the passage cited continuing thus: ‘A cheaper mode of obtaining food will undoubtedly increase profits says Mr. Malthus but there are other circumstances which may also increase profits with an increase of capital. The discovery of a new market where there will be a great demand for our manufactures is one’ (VI, 104–5). Why complicate the case by introducing extraneous matter? One is obliged to read the texts in context. However, we cannot be certain that Ricardo had adopted his falling realwage trend before the explicit statement of 18 December. That letter – coinciding as it does with the full-fledged statement of the Principles – provides us with a standard by which to evaluate Ricardo’s intellectual development in the earlier letter of 1814. Taking the standard as reference point, we encounter various ‘concessions’ by Ricardo to Malthus that go too far. For example: ‘You infer that my doctrine [‘facility of obtaining food’ as determinant of the profit rate] is not correct because improvements may take place in agriculture or manufactures, because new leases may not be granted at the time of the rise in the price of raw produce, and because the price of labour may not rise without delay in the same proportion’ (VI, 145). Ricardo then conceded that a failure of the money wage to rise in proportion will temporarily raise the profit rate. Now this is somewhat misleading in terms of the later formulation whereby even in the long run there is no proportional adjustment and yet the profit rate declines. A more significant concession by Ricardo to Malthus implies that a decline in the secular wage path might prevent a falling profit rate: You observe [p. 141] that in rich countries [large absolute population and capital] profits are often much higher, and in poor countries much lower than according to my theory, to which I reply that profits are very much reduced in the poor country by enormous wages; – the wages themselves may be considered as part of the profits of stock, – and are frequently the foundation of new capital. In rich countries wages are low, too low for the 256

RICARDIAN GROWTH THEORY

comforts of the labourers; – too large a portion of the gross produce is retained by the owner of stock, and is reckoned as profit. (23 October 1814; Ricardo 1951, VI, 147) Here we encounter a notion of falling secular wages with the transition from a low- to a high-density population, thwarting any clear trend to profits. But if despite falling corn wages the profit rate declines (the mature view), this concession is uncalled for, though it will be noted that Ricardo is emphasizing a situation where the real wage is initially enormously high, which might suggest a special case; in the event of a lesser initial value the profit-rate decline would not perhaps be thwarted. Even so, the passage provides an instance where ultimately Ricardo was to refine his position to indicate that, with increasing land scarcity, the profit rate falls along with falling real wages no matter what the initial value of the latter might be. The increasing incidence of diminishing returns, he was to show, is necessarily shared by both capitalists and labourers. I turn next to a passage from a letter by Ricardo of 10 January 1816 that causes Dr Peach some difficulty, pointing as it does to the New View. It is after citing this particular passage that he draws his strikingly original inference that ‘the principal new-view passages may have been an attempt to show, on Malthus ’ premises, that a declining real wage need not compensate for rising corn prices’ (above, pp. 243–4). I cannot think it inconsistent to suppose that the money price of labour may rise when it is necessary to cultivate poorer land, whilst the real price may at the same time fall. Two opposite causes are influencing the price of labour[:] one the enhanced price of some of the things on which wages are expended – the other the few enjoyments which the labourer will have the power to command, – you think they may balance each then, or rather that the latter will prevail. I on the contrary think the former the more powerful in its effects. I must write a book to convince you. (Ricardo 1951, VII, 10; cited in Peach 1988, 114) 5 Dr Peach’s reading of the text seems to be invalid. Apart from the fact that both in this letter of 10 January 1816 and in the Principles he ultimately did write, Ricardo argued not that ‘a declining real wage need not compensate for rising corn prices’ but that it cannot compensate for them (rising corn prices are ‘the more powerful’), there is no suggestion in the passage that Ricardo was merely making a logical case for argument’s sake, i.e. on Malthus’s terms. And why should there be, since he had in late 1814 stated as his own position that the profit rate falls with higher corn prices despite a reduced wage basket. 257

RICARDO - THE NEW VIEW

Dr Peach’s reading also flies in the face of the Ricardo–Malthus correspondence immediately preceding the letter of 10 January 1816. On 22 December 1815 Malthus had laid out the issues, following a personal encounter, thus: I think the principal difference between us relating to the subjects we last discussed, is, your opinion that General Profits never fall from a general fall of prices compared with labour, but from a general rise of labour compared with prices. I own, I can see no reason for this opinion. ... I believe I assented in par[ting] to the proposition that in the successive cultivation of poorer land, the price of labour would rise as well as the price of corn. When however the rise in the price of corn is occasioned solely and exclusively by the necessity of cultivating poorer land I am now convinced that the rise must be very small, and as the real price of labour must fall, I see no reason why the nominal price of labour should rise. How can such a country purchase the precious metals so as to occasion a general rise in the price of labour? (Malthus; in Ricardo 1951, VI, 341–2; see editorial comment p. 342n) In his reply Ricardo insisted that general prices are constant, corn and money wages only rising. He protested at Malthus’s representation of his (Ricardo’s) position in terms suggesting that money wages rise relative to general prices, a formulation which allows for the latter rising to some degree: As for the difference between us on Profits of which you speak in your letter, – you have not I think, stated it correctly. You say that my opinion is ‘that General Profits never fall from a general fall of prices compared with labour, but from a general rise of labour compared with prices’. I will not acknowledge this to be my proposition. I think that corn and labour are the variable commodities, and that other things neither rise nor fall but from difficulty or facility of production, or from some cause particularly affecting the value of money, – and that no alteration of price proceeding from these causes affect general profits; – allowing always some effect for cheapness of the raw material. (2 January 1816; Ricardo 1951, VII, 3) In replying to this letter Malthus insisted: ‘Can you give me a good reason why the money price of labour should rise because it is necessary to cultivate poorer land, and the real price of labour must fall’ (8 January 1816; VII, 8). Clearly Malthus here takes for granted that Ricardo accepted that the real wage declines. He is not asking for a rationale on his own terms, but challenging Ricardo to 258

RICARDIAN GROWTH THEORY

justify Ricardo’s own position that despite a fall in the real wage there necessarily occurs a rise in the money wage with rising corn prices. Malthus himself maintained that the money wage is more or less unchanged: ‘In the progress of cultivation ... the money price of labour [will] remain stationary nearly, and the money price of manufactured good[s] fall from the fall of profits’ (VII, 9). And precisely at this point we have Ricardo’s answer that it is quite consistent ‘to suppose that the money price of labour may rise when it is necessary to cultivate poorer land, whilst the real price may at the same time fall’ (above, p. 257); and his aside that to prove fully that this is so – that Malthus was in error for supposing that the real wage decline can ‘balance’ the impact on the money wage of higher corn prices – would require a book. There is, I conclude, not a shadow of a doubt that the falling wage path reflects Ricardo’s own considered position from at least December 1814, and was taken up in the Principles as a leading feature of the general theory of growth. V. THE ‘ESSAY ON PROFITS’ I turn now to the Essay on Profits (1815). This pamphlet poses interesting interpretative problems, providing an important instance where Ricardo fell into error on his own terms. Let us have at hand a statement in the essay of the inverse wage–profit relation: ‘The sole effect then of the progress of wealth on prices, independently of all improvements, either in agriculture or manufactures, appears to be to raise the price of raw produce and of labour, leaving all other commodities at their original prices, and to lower general profits in consequence of the general rise of wages’ (Ricardo 1951, IV, 20). Here the money wage upon which general profits depend is said to be governed by the price on corn. Similarly: ‘If by foreign commerce, or the discovery of machinery, the commodities consumed by the labourer should become much cheaper, wages would fall; and this, as we have before observed, would raise the profits of the farmer, and therefore, all other profits’ (IV, 26n). With this relation in mind we can better appreciate an important summary passage regarding forces that might raise the profit rate: If then, the principles here stated as governing rent and profit be correct, general profits on capital can only be raised by a fall in the exchangeable value of food, and which fall can only arise from three causes: 1st. The fall of the real wages of labour, which shall enable the farmer to bring a greater excess of produce to market. 2d. Improvements in agriculture, or in the implements of husbandry, which shall also increase the excess of produce. 259

RICARDO - THE NEW VIEW

3dly. The discovery of new markets, from whence corn may be imported at a cheaper price than it can be grown for at home. (Ricardo 1951, IV, 22) In one central respect the line taken here differs from that ultimately adopted in the Principles. The main relationships above are, first, general profits will fall should money wages rise in consequence of higher corn prices; conversely, they would rise in the event of reduced corn prices. Secondly, the reasons for reduced corn prices include a fall in the real wage. It has this latter effect by allowing a ‘greater excess of produce’ to reach the market, i.e. by generating an increase in supply; similarly in the case of new agricultural technology. But in the Principles a fall in the real wage of either a cyclical or a casual kind – a wage fluctuation – acts on profits directly via the money wage not through the intermediary of the corn price; a real-wage fall leaves the price of corn unchanged, raising the profit rate by depressing the money wage. On the other hand, a decline in the real wage of the secular sort, since it is an outcome of the growth process and incorporates the impact of land scarcity is not (according to the Principles ) accompanied by a falling money wage but by a rising money wage, and does not raise the profit rate at all. But the question now is: what is the nature of the secular real-wage path according to the Essay? Our passage continues thus regarding forces acting on (raising) the profit rate: ‘The first of these causes [the fall in the real wage] is more or less permanent, according as the price from which wages fall, is more or less near the remuneration for labour which is necessary to the actual subsistence of the labourer’ (IV, 22). Taking the word ‘permanent’ literally, we should read Ricardo as stating that the further the initial real wage is from subsistence, the greater the likelihood a real-wage reduction will be subsequently reversed. This would seem to imply a downward trend in the real wage – at least a statistical trend reflecting the circumstance that wage reductions are compensated for to an increasingly small degree by reverse movements. Dr Peach, however, reads Ricardo here as defining permanent movements in the real wage ‘relative to the subsistence wage’ (1988, 107), and takes it as proof that the secular real wage follows a subsistence path. I do not see that this is justified by the text. At the same time what follows denies a declining wage trend, though certainly without implying a subsistence-wage path: The rise or fall of wages is common to all states of society, whether it be the stationary, the advancing, or the retrograde state. In the stationary state, it is regulated wholly by the increase or falling off of the population. In the advancing state, it depends on whether the capital or the population advance, at the more rapid course. In the retrograde state, it depends on whether population or capital decrease with the greater rapidity. 260

RICARDIAN GROWTH THEORY

As experience demonstrates that capital and population alternately take the lead, and wages in consequence are liberal or scanty, nothing can be positively laid down, respecting profits, as far as wages are concerned. But I think it may be most satisfactorily proved, that in every society advancing in wealth and population, independently of the effect produced by liberal or scanty wages, general profits must fall, unless there be improvements in agriculture, or corn can be imported at a cheaper price. (Ricardo 1951, IV, 22–3) Having implied a downward trend of the real wage – a statistical if not an analytical trend – Ricardo now seems to deny any trend pattern at all, or at least he abandons any attempts to define a secular pattern. This, of course, contrasts with the Principles. In point of fact Ricardo earlier in his Essay had justified the assumption of constant secular wages following an allowance that profits might increase ‘because the population increasing, at a more rapid rate than capital, wages might fall’ or because of new technology: ‘We will, however, suppose that no improvements take place in agriculture, and that capital and population advance in the proper proportion, so that the real wages of labour, continue uniformly the same; – that we may know what peculiar effects are to be ascribed to the growth of capital, the increase of population, and the extension of cultivation, to the more remote, and less fertile land’ (IV, 12). Quite clearly Ricardo has failed to take account of the fact – upon which he himself insisted before the Essay in the 1814 correspondence and later in the Principles – that the secular process involving diminishing returns entails falling real wages. To remove wage fluctuations would not, on his own terms before and after the Essay, leave a constant secular wage path. A characteristic of the Essay is thus its partial treatment of the secular problem. Ricardo describes the stationary and advancing states; presumably wages are at or above the subsistence level respectively, and so must necessarily fall in the approach to a stationary state. Yet this latter implication, rationalized at length in the Principles, is neglected. The argument in the Essay constitutes an incompletely formulated analysis. 6 VI. THE ‘NOTES ON MALTHUS’ There is one further substantive matter which ought alone to be decisive. In his Principles (1820) Malthus expressly adopts the view – Ricardo’s view we have shown from 1814 – that the profit and real-wage rates decline simultaneously. He provides a brilliant account which sets aside the complexity of fluctuations and explains why secular wage constancy is an impossibility:

261

RICARDO - THE NEW VIEW

the supposition ... of a constant uniformity in the real wage of labour is not only contrary to the actual state of things, but involves a contradiction. The progress of population is almost exclusively regulated by the quantity of the necessaries of life actually awarded to the labourer; and if from the first he had no more than sufficient to keep up the actual population, the labouring classes could not increase, nor would there be any occasion for the progressive cultivation of poorer land. On the other hand, if the real wages of labour were such as to admit of and encourage an increase of population, and yet were always to remain the same, it would involve the contradiction of a continued increase of population after the accumulation of capital, and the means of supporting such an increase had entirely ceased. We cannot then make the supposition of a natural and constant price of labour, at least if we mean by such a price, an unvarying quantity of the necessaries of life. And if we cannot fix the real price of labour, it must evidently vary with the progress of capital and revenue, and the demand for labour compared with the supply. We may however, if we please, suppose a uniform progress of capital and population by which is not meant in the present case the same rate of progress permanently, which is impossible, but a uniform progress towards the greatest practicable amount without temporary accelerations or retardations ... ... [If] poorer lands which required more labour were successively taken into cultivation, it would not be possible for the corn wages of each individual labourer to be diminished in proportion to the diminished produce; a greater proportion of the whole would necessarily go to labour; and the rate of profits would continue regularly falling till the accumulation of capital had ceased. Such would be the necessary course of profits and wages in the progressive accumulation of capital, as applied to the progressive cultivation of new and less fertile land, or the further improvement of what had before been cultivated; and on the supposition here made, the rates both of profits and of real wages would be highest at first, and would regularly and gradually diminish together, till they both came to a stand at the same period, and the demand for an increase of produce ceased to be effective. (Malthus 1820; in Ricardo 1951, II, 255f) 7 Now Ricardo stated explicitly in his Notes his agreement with this formulation: ‘I agree throughout this section with Mr. Malthus in principle, we only differ in our ideas of what constitutes a real measure of value’ (II, 258). Elsewhere in the Notes Ricardo responds to Malthus’s criticism of his rent doctrine thus: ‘I do not see that Mr. Malthus’s language is very different from my own; in page 152 [ Principles, 1820] 262

RICARDIAN GROWTH THEORY

he says “The fall of profits and wages which practically takes place, undoubtedly transfers a portion of the produce to the landlord”’ (II, 119); and again he applauds Malthus’s reference to ‘the transfer from profits and wages’. These posthumously published statements alone should suffice to settle the debate as to Ricardo’s ‘true’ view. VII. SUMMARY AND CONCLUSION To return to the main issue. No one questions that the permanent decline in the profit-rate is given exclusively by changed conditions of producing wage goods. This fact in no way points away from the New View. The assertion that the ‘traditional interpretation’ of Ricardo (the fixed-wage interpretation) has the merit of ‘leav[ing] intact Ricardo’s thesis on permanent movement in profitability’, so that ‘its distortion factor is markedly lower than that of its main rival’ (Peach 1988, 114) is without merit. The New View captures the permanent movement in profitability. The impact of diminishing agricultural returns is on both the real wage and the profit rate. That the declining real-wage trend is not to be identified with real-wage fluctuations is scarcely surprising. Apart from diminishing returns acting on the price of wage goods the only other cause which, for Ricardo, could account for the falling profit rate is a rising, not a falling, real wage. Moreover, to exclude wage fluctuations in the account of secular profit decline makes good analytic sense. It amounts to the exclusion of wage increases due to fortuitous movements away from trend in the relative growth rates of capital and population. We must avoid identifying the impact on profits of fluctuations in the real wage with that of the downward trend – they have quite different analytical statuses. NOTES *A version of this paper was read at the History of Economics Society annual meeting, Richmond, Va., June 1989. I thank A.M.C. Waterman, Sue Golding, and D. Pokorny for their helpful comments. 1. See also the discussion of accumulation in the face of labour scarcity (Ricardo 1951, I, 165). For this exercise, the ‘natural wage’ is taken as starting point of the analysis. 2. Without a full appreciation of the characteristics of the growth model many of Ricardo’s observations lose their meaningfulness. For one might wonder why so much effort was put into proving that proportionate wages rise with diminishing returns if the real wage is constant; it is then self-evidently the case. The issue only becomes truly meaningful should the real wage be falling. Ricardo’s proposition is then that the marginal product declines faster than the real wage, raising labour’s proportionate share and lowering the profit rate. 3. Peach himself accepts, in his note 12, that the qualification in question supports the New View. 263

RICARDO - THE NEW VIEW

4. In the growth process, a high corn price is the effect of an increasing demand [and] always preceded by an increase of wages’ (Ricardo 1951, I, 163; emphasis added), so that there is no question of a slow adjustment of the wage. 5. The passage is cited by Peach only from ‘Two opposite causes ...’. 6. For these reasons it is unlikely that Ricardo was deliberately describing a ‘dynamic equilibrium’ path with wages above subsistence at a constant level. This would require deceleration of population growth, without the inducement of a wage decline, along with (at the same rate as) that of capital (Caravale 1985). 7. The entire argument assumes an absence of ‘prudential’ population control, that would allow a constant real wage (above subsistence) on the path to the stationary state.

REFERENCES Caravale, G.A. (1985) ‘Diminishing Returns and Accumulation in Ricardo’, in G.A. Caravale (ed.) The Legacy of Ricardo, Oxford: Basil Blackwell, 127–88. Casarosa, C. (1985) ‘The “New View” of the Ricardian Theory of Distribution and Economic Growth’, in G.A. Caravale (ed.) The Legacy of Ricardo, Oxford: Basil Blackwell, 45–58. Hollander, S. (1987) Classical Economics, Oxford: Basil Blackwell. Pasinetti, L.L. (1960) ‘A Mathematical Formulation of the Ricardian System’, Review of Economic Studies 27: 78–98. Peach, T. (1986) ‘David Ricardo’s Treatment of Wages’, in R.D.C. Black (ed.) Ideas in Economics, London: Macmillan, 104–28. ——(1988) ‘David Ricardo: A Review of Some Interpretative Issues’, in W.O. Thweatt (ed.) Classical Political Economy: A Survey of Recent Literature, Boston: Kluwer Academic Publishers, 103–31. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Stigler, G.J. (1981) ‘Review’, Journal of Economic Literature 19 (March): 100–2. The Economist as Preacher and Other Essays, Chicago: University of Chicago Press. ——(1982) Tucker, G.S.L. (1960) Progress and Profits in British Economic Thought, Cambridge: Cambridge University Press.

264

16 A REPLY TO PROFESSOR STIGLER AND DR PEACH*

I am happy to let readers work through the textual evidence as presented in this paper [above, Chapter 15] and Dr Peach’s reaction and decide for themselves as to ‘the timing and extent’ of Ricardo’s adoption of the falling wage growth model (Peach 1990, 764). Since I allow that between Ricardo’s first statement of that model in late 1814 and the Principles in 1817 there are instances of erroneous or partial formulation (as in the Essay on Profits ) there is less of a difference between Peach and myself than Peach suggests. As for the Principles, perhaps the main substantive issue is simply this: was Ricardo inherently inconsistent, as Peach believes; or did he adhere to the falling wage growth model and make use of the ‘discrete’ analysis when attempting to expound specific parts of the full model, as I believe? Readers can also come to their own conclusion regarding the specific feature of the ‘continuous’ model I maintain is so essential to an understanding of Ricardo’s position – that it accounts simultaneously for both a falling wage and a falling profit rate in terms of diminishing agricultural returns. I shall, therefore, devote what small space is available to me to Professor Stigler’s challenging observations. My critic believes that I am interested ‘only in Ricardo’s inner convictions’ (Stigler 1990, 765). In fact he is mistaken. I pretend to know nothing of Ricardo’s inner convictions; only the extant written word is available to me – at least in this world – and this I read in an attempt to discern what message Ricardo apparently wished to convey to his readers. It is also unclear to me why Professor Stigler should think that the only question worth investigation is that of interpretations by contemporaries. Certainly that is an important question. But another question is: were the contemporary interpretations accurate or were they distortions? To approach that question we are obliged to have some conception of the apparent intentions – not, I repeat, in the sense of ‘inner convictions’ – of the original author. And that was the sole purpose of my exercise. If an interpretation turns out to be inaccurate then a further question 265

RICARDO - THE NEW VIEW

arises relating to the source of error. This I should think is a most worthwhile issue, but one excluded from consideration by Stigler’s recommendations. On numerous occasions Ricardo protested at Malthus’s misreadings; his Notes and correspondence are filled with protestations. Thus, for example, Malthus maintained that ‘in his very ingenious chapter on profits’, Ricardo had dwelt on no other cause of falling secular profits ‘than the increasing difficulty of procuring food’, precluding an impact of fluctuating wages (in Ricardo 1951, II, 263–4). Ricardo reacted with feeling and at length against this misinterpretation (II, 264–8). I am not sure why Professor Stigler provides the citations from James Mill and McCulloch since there is no express attribution in either of them to Ricardo. More important, Mill had a penchant for using ‘strong cases’ for political ends, and as Winch has put it: ‘Unlike Smith and Ricardo, Mill makes no allowance for the possibility that real wages may remain above the psychological or physical minimum for what amounts to a lengthy short run’ (1966, 193). Why saddle Ricardo with James Mill’s extremism? As for J.S. Mill, I have myself noted elsewhere that he attributed a constant wage model to Ricardo (1987, 235–6). ‘It is regrettable’, I write there, ‘that Mill failed to record Ricardo’s full argument which recognized conspicuously the falling-wage possibility, and he must be held partly responsible for thereby misleading commentators ever after.’ But there is one contemporary – Mountifort Longfield – who did understand the orthodox growth model as entailing a necessary excess of the real wage over its final equilibrium level in a growing economy, although he could not appreciate why in that case labour could not bear the entire brunt of diminishing returns: Before I proceed to give what I consider a more accurate system of profits ... I shall briefly point out the fallacies in the two arguments of which I have just given you a sketch. In the argument used to prove that the decreasing fertility of the soil is the great and necessary cause of a decline of profits, it is, I conceive, unwarrantably assumed, that the effect cannot be entirely borne by the labourer, and that therefore of necessity some part of it must fall upon capital. This necessity I cannot perceive. As population was advancing, the wages of labour must have been more than what would be necessary to the subsistence of the labourers, with such families as would keep up an unvarying population; they may sustain some reduction, and why not the entire amount of the reduction that has taken place in the returns made to labour and capital? (Longfield 1971 [1834], 184–5) This latter criticism happens to be technically unwarranted. Nevertheless, here is a case where a correct reading coupled with technical error led to analytical ‘progress’, 266

A REPLY TO PROFESSOR STIGLER AND DR PEACH

for Longfield was stimulated to seek an alternative approach to profit-rate determination. I am not sure what Stigler would make of this case. But neither of us should be allowed to get away with selective citation. In his second section Professor Stigler raises an irrelevant issue for the purposes of my paper – the quality of Ricardo’s growth theorizing. (In my opinion, the falling wage model happens to be more impressive than the alternative which entails, as Stigler admits, only ‘a routine exercise in comparative statics’ (1990, 768).) Some of his comments apparently reflect misunderstandings: ‘This “theory” does not tell us when (and under what conditions) an (unexplained) act of saving will launch a period of wages exceeding equilibrium’ (767). But the theory does not involve ‘unexplained acts’ or (elsewhere in his paper) ‘periodic bursts’ of saving, or ‘initial displacements’, or an ‘ ad hoc cause’; the initial conditions are such as to generate continuous, albeit decelerating, growth rates of capital and population. As for the link with Smith, certainly there was some development of a growth model of the Ricardian kind in the Wealth of Nations – as it happens including invocation of diminishing returns. (This is precisely the point insisted on by Paul Samuelson 1978.) All the more reason not to be so taken aback to find the model re-emerging and strengthened at some of its weak points by Ricardo. It was in the air when Ricardo came on to the scene. NOTE * My thanks to Elizabeth Allgoewer for helpful comments.

REFERENCES Hollander, S. (1987) Classical Economics, Oxford: Basil Blackwell. Longfield, Mountifort (1971) [1834] ‘Lectures on Political Economy’, in The Economic Writings of Mountifort Longfield, New York: Augustus M. Kelley. Peach, T. (1990) ‘Samuel Hollander’s “Ricardian Growth Theory”: A Critique’, Oxford Economic Papers 42: 751–64. Ricardo, David (1951–73), The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 Vols), Cambridge: Cambridge University Press. Samuelson, P.A. (1978) ‘The Canonical Classical Model of Political Economy’, Journal of Economic Literature 16: 1415–34. Stigler, G.J. (1990) ‘Ricardo or Hollander?’, Oxford Economic Papers 42: 765–68. Winch, D. (ed.) (1966) James Mill: Selected Economic Writings, Edinburgh: Oliver and Boyd.

267

17 ON THE TEXTUAL INTERPRETATION OF RICARDIAN GROWTH THEORY: THE ‘NEW VIEW’ CONFIRMED (AGAIN)

I. INTRODUCTION In Chapter V of his Principles (1951, I, 103) Ricardo devised an arithmetic example demonstrating the impact on the profit rate of expansion under conditions of diminishing agricultural returns assuming the wage basket to be constant. Each worker receives initially £24 annually permitting expenditure of £12 on manufactured wage goods and £12 on food amounting to 3 quarters of corn with the corn price at £4 per quarter. When the corn price rises to £4.4.8 (a 5.8 per cent rise) the money wage is assumed to rise sufficiently to assure unchanged purchasing power over the original basket, on which assumption the corn wage falls from

In various publications over the past two decades, I have argued that these data represent a simplified case devised for illustrative purposes since in full-fledged accounts of the course of growth, before and after composition of the Principles and in the Principles itself, we find Ricardo has the real wage declining, in the sense of a declining basket, under pressure of decelerating capital accumulation. Keith Gibbard denies that there is any validity to this interpretation, and takes one vocal critic to task for conceding some small merit to my interpretation of Ricardo’s early writings and indeed of the Principles itself. He maintains that Ricardo throughout operated with a constant basket, any decline of the corn wage emerging only in the arithmetical sense outlined above (the ‘price effect’). From the 268

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

correspondence of 1814 to the Notes on Malthus (1820) the point at issue is ‘the valuation of a heterogeneous wage basket’; the capital–population mechanism, and its corollary of a secularly declining real wage under conditions of diminishing agricultural returns, ‘played no central role in Ricardo’s work’ (Gibbard 1994, 483). Gibbard refers to a ‘certain irony’ in that the New View critics of Sraffa’s reading of Ricardo ‘have explicitly based their interpretation upon the assumption of a homogeneous or corn-only wage’ and thereby ‘placed Ricardo’s work in an unnatural and misleading light’ (483). This is not the case. The New View of Ricardo claims that where the wage basket consists of corn alone, the corn wage declines under pressure of decelerating labour demand because of diminishing returns; and where the model is complicated by allowance for a mixed wage basket the commodity basket falls under the same pressure. There is nothing ironic in sometimes working with a simple and sometimes with a complex model and to convey essentials by the former if nothing for the purpose at hand is lost thereby. Adherents to the Old View – if I may use the term for convenience – used to insist on a constant corn wage in making the ‘corn-profit’ interpretation of the early Ricardo on profits. Their goal was to preclude labour-market pressures generated by decelerating accumulation from playing on the trend path of the real wage. A fall in the corn wage is apparently no longer disputed; but Gibbard seeks to achieve the same objective by representing any fall in the corn wage as nothing but an arithmetic reflection of a constant fixed-proportions mixed wage basket due to the increase in the price of the food item. This latest rescue mission is no more successful than earlier efforts. II. THE EARLY CORRESPONDENCE Gibbard sets out to prove that Ricardo used the so-called ‘price effect’ as early as 25 July 1814. He asserts first that Ricardo allowed for a heterogeneous wage basket in an earlier letter of 26 June and ‘was thus aware of the question of relative valuation at an early stage’, having in mind Ricardo’s statement that ‘the rise in the price of corn . . . without any augmentation of capital must necessarily diminish the demand for other things even if the prices of those commodities did not rise with the price of corn, which they would (tho’ slowly) certainly do’ (1994, 467–8). Gibbard gives a false impression by asserting that Ricardo’s concern in the June letter was with ‘relative valuation’, for this term suggests the ‘price effects’ defined earlier, a reading subsequently used to reinforce the fixed real-wage interpretation (below, p. 272). In fact, this passage contains at most an implicit reference to a mixed wage basket but not a word about ‘price effects’, the point at issue being unrelated to the wage path. 1 In any event, Gibbard himself allows that the mechanism he perceives could only be at play temporarily – Ricardo ‘possibly regards it as a secondary issue of timing’ – 269

RICARDO - THE NEW VIEW

considering the ultimate increase of manufacturing prices and the fact that (as Gibbard sees the matter) ‘price effects’ require the manufacturing price to be constant; 2 and he passes on quickly to the letter of 25 July, in which (following Garegnani) he finds a fall in the corn wage ‘due entirely to the presence in the wage basket of non-wheat elements, unchanging in price, resulting in ... a “price effect”, with the real wage (basket) constant’ (468–9). This reading of the July letter is unsatisfactory. The text used in support is the second sentence in the following passage – the first is omitted: The capitalist ‘who may find it necessary to employ a hundred days labour instead of fifty in order to produce a certain quantity of corn’ [Malthus; in Ricardo 1951, VI, 111] cannot retain the same share for himself unless the labourers who are employed for a hundred days will be satisfied with the same quantity of corn for their subsistence that the labourers employed for fifty had before. If you suppose the price of corn doubled, the capital to be employed estimated in money will probably be also nearly doubled, – or at any rate will be greatly augmented and if his monied income is to rise from the sale of the corn which remains to him after defraying the charges of production how is it possible to conceive that the rate of his profits will not be diminished. (Ricardo 1951, VI, 115, emphasis added) The reference to a capital ‘nearly doubled’ alludes, according to Gibbard – here basing himself on Garegnani – to ‘the unchanged price of the non-wheat elements in the farmers’ costs’ (1994, 468). Even were this so his case would not follow, since the matter in dispute between us entails manufactured wage goods not other non-corn items in the farmer’s costs. But assuming Gibbard intended to write ‘the unchanged prices of manufactured wage-goods’, his interpretation would be no stronger than that of the June letter; if Ricardo assumed not only a mixed wage basket but constant prices of manufactured wage goods, he must have abandoned – some time between 26 June and 25 July – his Smithian position whereby a rise in the corn price acts on manufactures, whereas the letter of 25 July is explicit that ‘the prices of all commodities must increase if the price of corn be increased’ (cited, note 1). For my part, I find that the passage, especially when the first sentence is kept in mind, suggests a corn-only wage basket; there is no need to bring in non-corn capital to understand it. 3 At this juncture, incidentally, Ricardo does not positively insist upon a fall in the corn wage, but merely observes that any such fall would have to be at least as great as the fall in productivity to prevent a profit-rate decline; nor does he explain the mechanism at play dictating whether or how far the corn wage would fall. On the other hand, he does insist on working in money terms as is clear from the sentence Gibbard himself cites. Gibbard asks in his note 4 why if the real wage consists only of corn and if the money wage does not rise sufficiently to 270

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

compensate for the increased corn price thus depressing the corn wage, Ricardo was so confident that the profit rate falls. The answer is obvious. The reduction in the corn wage would have to be of unlikely magnitude (halved in fact) to insulate the profit rate. We come now to the reduction in the corn wage alluded to in the letter of 18 December which, according to Gibbard, ‘does not, as Hollander suggests, imply a reduced wage basket but results from the impact of the price effect on the non-corn wage goods’ (469). In that letter Ricardo concedes to Malthus, who had criticized his emphasis on the agricultural sector as ‘regulator’ of the profit rate, that the profit rate would be affected by changes in the prices of non-agricultural wage goods: I have been endeavoring to get you to admit that the profits on stock employed in Manufactures and commerce are seldom permanently lowered or raised by any other cause than by the cheapness or dearness of necessaries, or of those objects on which the wages of labour are expended. Accumulation of capital has a tendency to lower profits. Why? because every accumulation is attended with increased difficulty in obtaining food, unless it is accompanied with improvement in agriculture; in which case it has no tendency to diminish profits. If there were no increased difficulty, profits would never fall, because there are no other limits to the profitable production of manufactures but the rise of wages. If with every accumulation of capital we could tack a piece of fresh fertile land to our Island, profits would never fall. I admit at the same time that commerce, or machinery, may produce an abundance and cheapness of commodities, and if they affect the prices of those commodities on which the wages of labour are expended they will so far raise profits; but then it will be true that less capital will be employed on the land, for the wages paid for labour form a part of that capital. (Ricardo 1951, VI, 162) So Ricardo does not insist on a corn wage basket, his point being that changes in the prices of wage goods of any kind will, by playing on the money wage, affect the profit rate. He then returns to his main agricultural theme: ‘A diminution of the proportion of produce, in consequence of the accumulation of capital does not fall wholly on the owner of stock, but is shared with him by the labourers. The whole amount of wages paid will be greater, but the portion paid to each man, will in all probability, be somewhat diminished’ (VI, 162–3). I have always read this as referring to the corn wage. And – notwithstanding Gibbard’s doubts (1994, 469) – I find it perfectly conceivable that Ricardo here does indeed switch his assumptions in the course of one letter, the preceding reference to manufactured wage goods being in 271

RICARDO - THE NEW VIEW

the nature of an aside (his ‘I admit at the same time’ suggests just this). Even so, this would not necessarily mean that in the event of a mixed wage basket he denied a general real wage fall, merely that he was not focusing upon it. Alternatively, it cannot be excluded that his reference to a falling wage does in fact indicate the general basket. Gibbard recalls ‘the argument pressed upon [Ricardo] by Malthus and acknowledged in his own letters of 26 June and 25 July’, and insists that since Ricardo could not have forgotten them in December he must be using the ‘price effect’ argument to assure his falling corn wage but a constant basket. This is unconvincing. Ricardo had acknowledged nothing in June and July regarding the ‘price effect’ case. In any event, what ‘price effects’ could Gibbard – who defines the ‘price effect’ in terms of constant manufacturing prices – be attributing to Ricardo? Ricardo at this time did not yet hold the prices of manufactures to be constant when the corn price rises. And finally, there is an evident response to Gibbard’s complaint: ‘What Hollander does not explain is how a rise in the money wage relative to the price of manufactures, led Ricardo to conclude by 18 December 1814 that the real commodity wage would fall’ (470). The answer again is self-evident. Consider the extreme case where the prices of manufactures are in fact constant, having in mind for convenience Ricardo’s data in the Principles (above, p. 268). The money wage rises with the corn price, the extent depending on labour market pressures. Should the rise in the money wage be less than 14s. the worker could no longer buy the original basket. He might choose to maintain his manufacturing purchases and his corn consumption or take the same amount of corn and fewer manufactures, though the simplest case would be a reduction of the full range of his purchases. In all cases one would say that the real basket is reduced. ‘A rise in the money wage relative to the price of manufactures’ does not therefore, preclude a fall in the basket, a lesson that can be applied to the case at hand. Gibbard has taken us not a step further to show that Ricardo in the early correspondence used the ‘price effect’ argument. III. THE ‘ESSAY ON PROFITS’ AND ITS AFTERMATH In his next Section Gibbard observes that ‘Had Ricardo indeed adopted New View reasoning by December 1814, as Hollander maintains, we might have expected it to permeate the argument of the Essay’ (471). He points out that I adduce just one passage in support of a downward trend in the real wage. Gibbard might do me the service of reading what I actually wrote. I did not claim in my account that the passage in question implies a downward trend in the real wage commensurate with the ‘continuous’ model (1990, 747–8 [above, Chapter 15, pp. 261–2]). I said to the contrary (a) that there is a possible implication of a downward statistical trend – not of an analytical trend, and (b) that even this is followed immediately by a denial of 272

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

any trend variation, in line with Ricardo’s assumption at the outset of the Essay that he was going to ‘suppose ... that capital and population advance in the proper proportion, so that the real wages of labour continue uniformly the same’. I emphasized that Ricardo ‘had failed to take account of the fact ... that the secular process involving diminishing returns entails falling real wages’. How clear could one be? But why not turn Gibbard’s own question on himself thus: ‘Had Ricardo indeed adopted the “price effect” argument in the 1814 correspondence, as Gibbard maintains, we might have expected it to permeate the argument of the Essay.’ Yet Gibbard is explicit that the price effect is applied in the Essay ‘not to the question of wages but against the landlord’ (emphasis added), with ‘the impact on the labourer ... relegated almost to the status of an afterthought towards the end of the essay’ (I would point out in a reference to what ‘Mr. Malthus thinks’). And after all those confident assurances that the ‘price effect’ was applied by Ricardo in the 1814 correspondence it is difficult not to be taken aback to read that ‘It was not until the post- Essay debate with Malthus in the Spring of 1815 that Ricardo was to realize its full significance’ (Gibbard 1994, 471). As Gibbard points out regarding those post- Essay materials (474), in my Economics of David Ricardo I myself discuss an illustration by Ricardo which assumes ‘the wage basket to be composed of a fixed-proportions mix of corn and (manufactured) necessaries, the commodity wage to be constant, and per capita corn wages to fall in the course of development as the relative price of food rises’ (1979, 157). He asks: ‘Are we then to believe, that there were in fact two “thoroughly Ricardian foundations”4 for wage movements at this time?’ My answer is yes. I take the illustration based on the assumptions just mentioned to be just that, an illustration to get across a particular point but not to describe a full growth process; and it is ‘thoroughly Ricardian’ since this is what appears in the Principles (above, p. 268). As for the growth process, Gibbard has somehow managed to miss Ricardo’s letter to Malthus of 10 January 1816 which confirms all that I have been maintaining, namely a fall in the real wage understood as the basket – not the corn wage due to ‘price effects’ – under pressure of diminishing returns: I cannot think it inconsistent to suppose that the money price of labour may rise when it is necessary to cultivate poorer land, whilst the real price may at the same time fall. Two opposite causes are influencing the price of labour[:] one the enhanced price of some of the things on which wages are expended, – the other the fewer enjoyments which the labourer will have the power to command, – you think they may balance each then, or rather that the latter will prevail. I on the contrary think the former the more powerful in its effects. I must write a book to convince you. (Ricardo 1951, VII, 10) 273

RICARDO - THE NEW VIEW

Gibbard and other opponents of the New View would do well to mull over this passage carefully – including its closing sentence. IV. THE ‘PRINCIPLES OF POLITICAL ECONOMY’ Gibbard (1994, 476) takes up my ‘challenge’ to confront the passage in the Principles on the downward impact on wages of decelerating capital accumulation (I, 101–2 [above, pp. 246–7]). He concedes that it is ‘apparently a convincing passage in support of the New View. ... Supply and demand it seems have become an integral part of Ricardo’s distribution mechanism in the Principles ’ (1994, 477); but he immediately withdraws his concession (478): ‘It is clearly something of a special case’, supporting this by reference to a statement in the passage indicating that ‘every commodity on which the wages of labour were expanded rose’, whereas we know Ricardo at this time held prices of manufactures constant, a position actually restated a few lines later, namely that the manufactures ‘sell at no higher price’. Gibbard does well to bring this complexity to our attention, but he avoids the obvious conclusion. First, we have Gibbard’s suggested rationalization for Ricardo’s assumption that ‘every commodity on which the wages of labour were expended rose’ – the special case: ‘Possibly Ricardo chose this simplifying assumption to avoid the complexities of analyzing the impact of a falling rate of capital accumulation on a wage basket where relative values were changing’ (478). Now a couple of paragraphs later he asks rhetorically in the manner of an earlier reaction: ‘Had Ricardo truly been wedded to the idea of a secularly declining real wage since December 1814, as Hollander insists, we might have expected him to insist on the prevailing strength of labour market pressures’ (479). Does he not see that one can reverse the question: ‘Had Ricardo truly been wedded to the idea of “price effects” since December 1814 (actually since June/July), as Gibbard insists, one might have expected him to insist on them in dealing with the impact of a falling rate of capital accumulation’ – unless, that is, we are to take it that Gibbard and other critics deny deceleration of accumulation to be a principal Ricardian concern. To offer Gibbard’s rationale is simply to admit that Ricardo does not use ‘price effects’ in the analysis of that standard problem. Gibbard then asserts that Ricardo’s assumption that all wage-goods’ prices rise is in effect ‘to treat the wage as though it consisted only of corn’ (478). This implies a proportionate increase of all wage goods’ prices – a matter I take up shortly; but it should have been obvious to Gibbard that he has destroyed his own case by allowing that in this case ‘with capital accumulation lagging behind population growth ... the corn wage and real wage must both fall’ – since this recognizes the market mechanism acting to depress the real wage which I insist upon, and undermines the notion of a constant real wage during the course of growth characterized by decelerating accumulation. 274

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

Gibbard’s presumption that where Ricardo allows the prices of manufactures to rise it is a proportionate rise that was intended is unjustified. Any increase in manufacturing prices reflects specifically an increase in the real costs of producing raw materials as in the passage cited by Gibbard himself (479, from Ricardo 1951, I, 104) and elsewhere in the chapter ‘On Profits’: ‘There are few commodities which are not more or less affected in their price by the rise of raw produce, because some raw material from the land enters into the composition of most commodities. Cotton goods, linen, and cloth will all rise in price with the rise of wheat; but they rise on account of the greater quantity of labour expended on the raw material from which they are made, and not because more was paid by the manufacturer to the labourers whom he employed on those commodities’ (I, 117–18). But there can surely be no doubt that Ricardo never relinquished the notion that the corn price rises relative to the prices of manufactures during the course of growth. Since that is so, nothing need have prevented him from incorporating the ‘price effect’ formally to derive a falling corn wage and a constant wage basket in the growth analysis had that been his intention. Notice that he does incorporate it informally when he reverts to his illustrative data (above, p. 268) leaving it to his readers to provide a numerical example: ‘In all these calculations I have been obvious only to elucidate the principle. ... My object has been to simplify the subject, and I have therefore made no allowance for the increasing prices of the other necessaries, besides food, of the labourer; an increase which would be the consequence of the increased value of the raw materials from which they were made, and would of course further increase wages, and lower profits’ (I, 121–2). Assume, therefore, that while the corn price rises by 5.8 per cent, that of manufactured wage goods also rises but by half as much, namely by 2.9 per cent; in this case the money wage must rise from £24.00 to £25.05 (rather than to £24.70) to assure a constant wage basket, the corn wage falling to 2.36 (rather than to 5.83) quarters. So the ‘price effect’ emerges when manufacturing prices rise as far as the illustration is concerned. Yet Ricardo insists on a falling commodity wage when dealing with genuine growth, i.e. with decelerating accumulation, maintaining that the labourers ‘will receive more money wages but his corn wage will be reduced; and not only his command of corn, but his general condition will be deteriorated, by his finding it more difficult to maintain the market rate of wages above their natural rate’ (I, 102; emphasis added). It is also necessary to correct an analytical error on Gibbard’s part. He asserts that ‘Ricardo’s theoretical dilemma is acute’, since should the ‘prices of all wage goods, including manufactures ... rise, then the clear-cut conclusion that manufacturing profits will fall with the onset of diminishing agricultural returns, is lost’ (Gibbard 1994, 478). This, of course, is not so since the rise in manufacturing prices reflects specifically and solely the increased real cost of producing raw materials not the increase 275

RICARDO - THE NEW VIEW

in the (money) wage as the passages I cite above make clear; the inverse wage– profit relation holds good, and there is no ‘acute dilemma’ in accounting for the fall in the manufacturing profit rate. And since Ricardo’s general conclusion regarding a falling real wage (general basket) holds good whether or not manufacturing prices rise, we may understand better his rather cavalier switching between assumptions on this matter. V. RICARDO’S ‘NOTES ON MALTHUS’ Gibbard’s remarks on the Notes on Malthus are largely irrelevant. The quote given from Ricardo 1951, II, 98 does not assert a constant basket; it is beyond me on what he bases the assertion that this passage indicates that ‘As capital and population advance, the rise in the price of corn, consequent upon increasing difficulty of production, induces a rise in the money wage sufficient to sustain the real wage at its “natural” level’ (1994, 481). Gibbard seems to think that a mixed wage basket, a rise in the corn price relative to the price of manufactures and a rise in the money wage relative to manufactures necessarily imply a constant basket. They do not (above, p. 272). Ricardo did certainly sometimes revert implicitly in his Notes on Malthus to the illustrative data of the Principles, as for example II, 78 where the reason given for the money wage rising less than the corn price is that ‘corn is not the only thing consumed by the labourer’ rather than because of market pressures, but this is not indicative of the full growth process entailing deceleration of accumulation. He resorted to simplification as usual when his concern – in this case the measurement of value – did not require a full account. Even so, one finds instances of unambiguous decline in the basket in the measurement context as, for example, in a letter to McCulloch (written at the time of composing the Notes) where Ricardo adopts for argument’s sake Malthus’s special silver measure: If produced by labour only – if half an ounce of silver could be picked up on the sea shore by a day’s labour, the natural price of labour would be always half an ounce of silver, it could neither rise nor fall. Corn might however be produced with more difficulty, and by the rise of its price, the wages of the labourer would be less adequate to procure him comforts and conveniences. In this case I should say wages would rise, because I always measure the rise of every thing by the quantity of labour necessary to produce it, and the wages though less in quantity would require more labour to produce them. (2 May 1820; Ricardo 1951, VIII, 179; emphasis added) As for the full account, that is intimated in Ricardo’s acceptance of Malthus’s statement of a falling wage trend involving market pressure. Gibbard admits all I would ask for: ‘It is, it seems, a highly significant concession, providing clear evidence 276

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

that, by 1820, Ricardo had come to accept Malthus’s position on this central question’ (1994, 482). But Gibbard will still have none of it, pointing to revisions made by Malthus in manuscript notes to his own Principles (first published in 1989) involving the substitution of corn wages for real wages in the extract in question. Am I expected to take this seriously? These are Malthus’s revisions not Ricardo’s and of no relevance for the issue of Ricardo’s position. In any event, Malthus’s modification does not affect his insistence on labour market pressures on the corn wage due to decelerating accumulation; all that is retained. Gibbard’s claim to the contrary results from the erroneous belief that a fall in the corn wage can only reflect ‘price effects’. VI. CONCLUSION Why the contortions on the part of critics of the New View to avoid facing the evidence that Ricardo has market processes at play in his growth model and the insistence that growth proceeds in his model at a constant real wage despite all the illogicalities created by that constraint? And why the panic when a critic slackens in his opposition? I suggest that any answer should take into account the following consideration. Gibbard apparently accepts that Ricardo addressed the issue of a decelerating rate of accumulation under pressure of land scarcity (above, p. 274). That is the cause of downward pressure on the real wage, though for Gibbard this is a ‘special case’ which reduces to corn only. He is wrong on this as I have shown, but taking him at his word he still has not been able to remove the market pressures in question. He cannot make them disappear at a wish. And there is the rest of the elaborate Ricardo text involving market pressures in the reverse direction forcing real wages up over the long period: In different stages of society, the accumulation of capital, or of the means of employing labour, is more or less rapid, and must in all cases depend on the productive powers of labour. The productive powers of labour are generally greatest when there is an abundance of fertile land: at such periods accumulation is often so rapid, that labourers cannot be supplied with the same rapidity as capital. It has been calculated, that under favourable circumstances population may be doubled in twenty-five years; but under the same favourable circumstances, the whole capital of a country might possibly be doubled in a shorter period. In that case, wages during the whole period would have a tendency to rise, because the demand for labour would increase still faster than the supply. (Ricardo 1951, I, 98) All this must be meaningless to Gibbard. In fact, only if the growth problem is erased from the agenda is it possible to argue for a constant real wage. I would not 277

RICARDO - THE NEW VIEW

be surprised were the source of much of the hostility to the New View to be found just here, since secular variability of the wage basket undermines the entire concept of surplus. NOTES 1. Ricardo was opposing Malthus’s position that agricultural protection reduces aggregate output more than aggregate demand thereby raising the return on capital, since this position neglected the negative impact of a higher corn price on the demand for manufactured goods. How would such an impact work? Since it operates even with manufacturing prices unchanged Ricardo must be assuming some kind of income effect due to the corn-price increase. A mixed wage basket is then implied unless this economizing behaviour refers specifically to non-wage earners assumed to purchase food as well as manufactures, as in the treatment of corn taxation in the Principles : ‘it is not necessary that my demand for corn should diminish as I may prefer to pay £100 per annum more for my corn, and to the same amount abate in my demand for wine, furniture, or any other luxury’ (1951, I, 237). There is a fortiori a contraction in the demand for manufactures should their prices rise with the price of corn, which they will do we are told. (This is clearer still on July 25: ‘It appears to me that the difficulty and expense of growing corn will necessarily regulate the demand for the products of capital, for the demand [our quantity demanded] must essentially depend on the price at which they can be afforded, and the prices of all commodities must increase if the price of corn be increased’ (VI, 114).) This latter would certainly include demand for manufactures by non-wage earners. 2. In fact, the ‘price effect’ might operate even when manufacturing prices rise provided the relative corn price rises [see below, p. 275]. A belated correction in this respect will not save the paper considering its remaining deficiencies. 3. I once believed the passage contained an error [above, p. 38, n.13]. It does not as the following example shows. I apologize to Ricardo; but it becomes even clearer now, that he probably had in mind in this instance a capital consisting only of corn wage goods. The first line indicates the initial situation; the second, the new situation assuming that the corn wage per man falls in proportion to the productivity decline; the third, the new situation with a decline of only 25 per cent; and the fourth, the new situation with a yet smaller decline of 10 per cent such that ‘the capital to be employed estimated in money will probably be also nearly doubled’. This formulation is correct if line 4 is compared with line 2 rather than line 1.

1. 2. 3. 4.

1

2

3

Corn output (qs)

Labour (men)

Corn wage per man

100 100 100 100

50 100 100 100

1.0 0.5 0.75 0.9

4 Total corn wages

5 Corn price $

50 50 75 90

5 10 10 10

6

7

8

9

‘Capital Gross Net Profit estimated sales sales rate in money Rev. $ Rev. $ (8/6) (4 x 5) (1 x 5) (7 - 6) 250 500 750 900

500 1000 1000 1000

250 500 250 100

100% 100% 33% 11%

4. This from Hollander 1975, 197, in my account there of the illustration [above, p. 53]. 278

THE ‘NEW VIEW’ COMFIRMED (AGAIN)

REFERENCES Gibbard, K. (1994) ‘Hollander’s Textual Interpretation of Ricardian Growth Theory: A Reconsideration’, History of Political Economy 26: 464–85. Hollander, S. (1975) ‘Ricardo and the Corn Profit Model: Reply to Eatwell’, Economica 42: 188–202. —— (1979) The Economics of David Ricardo, Toronto: University of Toronto Press. —— (1990) ‘Ricardian Growth Theory: A Resolution of Some Problems in Textual Interpretation’, Oxford Economic Papers 42: 730–50. Ricardo, D. (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press.

279

Part V FURTHER INTELLECTUAL LINKAGES

18 THE RECEPTION OF RICARDIAN ECONOMICS*

I am concerned in this paper with two central themes of the modern literature on classical economics: the supposed early demise of ‘Ricardianism’, and the related notion of a ‘dual’ development of nineteenth-century theory entailing, on the one hand, typically Ricardian procedures and on the other embryonic neo-classical procedures. My investigation reveals an impressive resilience of the fundamental Ricardian theorem on distribution, and its derivation in terms of the standard measure of value, amongst the so-called ‘dissenters’ as well as those traditionally classified as members of Ricardo’s school; and it casts doubt on the historical accuracy of a welldefined duality of analytical development. An introductory word is in order regarding the secondary interpretations, and the precise scope of the present article. The key features of the Ricardian theory of distribution, as it is now generally expounded, are its conception of real wages as a social or institutional datum, and that of profits as a form of residual. (In the more complex models attributed to Ricardo – in contrast to the simple singlesector models – distribution is said to have logical priority over exchange values in the sense that the price structure is known only after the prior and independent determination of the wage and profit rates.) 1 Ricardian procedure is sharply contrasted with that of the neo-classical (general equilibrium) economists either to its advantage or to its disadvantage. Thus from one perspective it is said that scientific economics requires the treatment of productive organization within a general-equilibrium context wherein the returns to the productive factors are envisaged as the competitively determined prices of services; Ricardian procedure thus appears to constitute the arbitrary reduction of the number of variables in the system leaving one variable (profits) to be determined residually (Schumpeter 1954, 568f; Knight 1956, 37–88). But from another perspective it is the general-equilibrium approach that is defective; the function of the economist is seen to consist precisely in the specifications of ‘causal’ or one-way relations where appropriate (Pasinetti 1974, 44), and great merit is attached in particular to the 283

RICARDO - THE NEW VIEW

specification of the real wage in cultural or institutional terms and the treatment of profits as residual. The notion of a dual development of nineteenth-century analysis is shared by modern commentators from whichever of the diametrically opposed perspectives they may view the early literature. One stream of nineteenth-century doctrine – the nonRicardian stream – is said to originate in Adam Smith’s ‘adding-up-components’ theory whereby competition, through the operation of supply and demand, assures that market prices gravitate towards ‘natural’ prices defined as the sum of unit wage, profit and rent costs; in this scheme the factors themselves are paid at their natural rates determined, in their turn, by the general conditions of demand and supply in factor markets. The embryonic neo-classical approach of Smith is identified in the work of Turgot, J.B. Say, Lauderdale and Malthus, and in the later writings of the Longfield–Senior group of ‘dissenters’ from Ricardianism, J.S. Mill, Jevons and Marshall (Schumpeter 1954, 465, 474, 568, 673n; Dobb 1973, 44f, 112f). The second or Ricardian line, runs the argument, was short-lived although revitalized decades later by Marx. According to the reading of the record by Schumpeter the Ricardian system ‘failed from the start to gain the assent of the majority of English economists’, and by the early 1830s ‘was no longer a living force’ (1954, 478). 2 This position is much the same as that of Marx, although for Schumpeter Ricardo’s poor fortune was a welcome sign while for Marx the (supposed) early demise was a symptom of the degeneration of economic science; 1830 was for Marx the dividing line between ‘scientific’ and ‘apologetic’ economics. 3 It was in fact Schumpeter’s belief that J.S. Mill must be excluded from that group which constitutes Ricardo’s ‘school’. 4 This evaluation is equally characteristic of Marxist interpreters. Marx himself speaks of Mill’s work as an example of the ‘eclectic, syncretistic compendia’ which characterized the period after the collapse of ‘scientific’ political economy in 1830 (1973, 883). Along these lines Maurice Dobb has observed of Mill: ‘when looking back on him from a distance one can see quite clearly that in major respects his own work was much nearer to Marshall than it was to Ricardo; and that so far as his theory of value was concerned, on the contrary to continuing and improving on Ricardo, in essentials he took his stand on the position of Smith where Ricardo had been opposing him’ (1973, 122). 5 In his volume on J.R. McCulloch, Professor O’Brien has added his authority to the view that the central Ricardian model suffered a serious decline soon after Ricardo’s death. For it is the general theme of this work that while McCulloch ‘did much to popularize economics ... it was not Ricardo’s economics that he was popularizing’ (1970, 402–3).6 McCulloch, runs the argument, must be considered as full square in the Smithian tradition. A similar revisionist interpretation has recently been put forward regarding De Quincey (Groenewegen 1974, 193). 284

THE RECEPTION OF RICARDIAN ECONOMICS

We shall be concerned with the two issues raised in the preceding account: the supposed early demise of Ricardianism and the related notion of a dual development in nineteenth-century economics. Needless to say it is impossible to approach these matters without a rather precise specification of the Ricardian paradigm, extending beyond the general questions of doctrinal position and archetypal method. Investigation not only of the content but also of the origins of Ricardo’s Principles – particularly the process whereby Ricardo, in Spring 1813, commenced to discern what he considered to be a number of logical errors in the Smithian position – suggests that what is characteristically ‘Ricardian’ is the use of a special theory of value involving an absolute standard in the derivation of the inverse relationship between wages and profits – the famous fundamental theorem on distribution. 7 In the Ricardian structure, an increase in the proportionate share of wages appears as an increase in wages, expressed in terms of the measure of value – a commodity produced with constant labour input and thus constituting a labour-embodied unit. An increase in per capita ‘gold’ wages implies a rise in the labourer’s share in per capita output which is necessarily of constant ‘value’. The entire Ricardian scheme is thus designed to relate the rate of return on capital to the ‘value’ of per capita wages (Ricardian ‘real’ wages) – which in effect amounts simply to the proportion of the work day devoted to the production of wages – and variations in the rate of return to (inverse) variations in the ‘real’ wage rate. It follows from the basic analysis that – assuming unchanged input coefficients in the production of the monetary metal, or ruling out nominal changes in money values – wage-rate increases are non-inflationary and at most generate an alteration in relative prices within limited bounds; capitalists are unable to pass on increased wage costs in the form of generally higher prices. This fundamental conception, it must be emphasized, was initially formulated as a direct challenge to received doctrine based upon Adam Smith’s analysis whereby wage-rate increases are passed on by capitalists in the form of higher prices in manufacturing industries and lower rents in agriculture. Here lies the essence of the Ricardian contribution. The significance – indeed the objective – of Ricardo’s work cannot be accurately evaluated unless placed in this historical context. 8 It is my primary conclusion that the Ricardian theorem on distribution – the inverse wage–profit relationship – left a firm and positive impression on the work of a number of authors normally regarded as ‘dissenters’ par excellence – including Malthus, Bailey, Torrens and Longfield – and this despite their frequent formal criticisms of Ricardo and his followers and their declared objective to break new ground or at least to refute the merit of Ricardo’s divergencies from the Wealth of Nations. 9 Although I shall only briefly consider the status of McCulloch, J.S. Mill and 285

RICARDO - THE NEW VIEW

Thomas De Quincey, it is my belief that to play down their adherence to Ricardianism and place them in Smith’s camp as far as concerns the theory of value and distribution is unjustified. 10 A demonstration of the resilience in the works of the dissenters of the Ricardian distribution theorem raises some serious questions regarding the legitimacy of the notion of a dual development in the course of nineteenth-century economic analysis. These doubts are reinforced by two further circumstances. My investigation reveals a very considerable degree of disagreement within the group of dissenters. The fact is that there occurred no concerted attack on Ricardo. Of particular significance for an accurate perspective is the weight of objection directed by various ‘dissenting’ writers against Bailey, frequently regarded as the head of the line of Ricardo critics. Secondly, it becomes clear that in many cases the criticisms constitute a misunderstanding of Ricardo’s position, while in others the contributions of the dissenters would not have been considered objectionable by Ricardo. I have in mind particularly in this latter category the emphasis upon the relativity dimension of exchange value (Bailey); the scarcity theory of rent (Thompson, Senior); the insistence that the cost determination of price operates by way of demand–supply variation (Malthus); the principle of diminishing marginal utility (Lloyd); and the abstinence approach towards interest (Longfield, Scrope, Read, Senior). But of outstanding significance is the application of market demand–supply analysis to long-run wage determination (Malthus, Longfield, Torrens, Read, Scrope, Senior). Ricardo I believe stood squarely in the Smithian tradition regarding wage theory. The notion of a subsistence wage as central feature of his system, the key theme in modern representations of Ricardian economics, has been seriously exaggerated. 11 I. THE DISSENTING CRITICS I shall give some indication in this section of the extent of adherence to the inverse profit–wage relationship amongst several of the main writers usually classified as dissenters. As already indicated, the acceptance of Ricardo’s position implies – and this implication was usually well understood by nineteenth-century economists – the rejection of key features of Smithian analysis. I wish first to emphasize that Malthus himself, who is commonly taken for granted to be a severe critic of the Ricardian theory in fact accepted much of its substance. It is a serious error to neglect the extent of his accord with the inverse theorem. Thus in his famous Quarterly Review article for 1824 treating McCulloch’s Encyclopaedia Britannica contribution of the previous year Malthus wrote: ‘Of all the truths which Mr Ricardo has established, one of the most useful and important is, that profits are determined by the proportion of the whole produce which goes to labour. It is, indeed, a direct corollary from the proposition, that the value of commodities is 286

THE RECEPTION OF RICARDIAN ECONOMICS

resolvable into wages and profits; but its simplicity and apparent obviousness do not detract from its utility’ (1963, 189). Similarly: ‘We fully agree with the author of the present treatise, that when it is said that profits depend on wages, they must not be understood to depend on wages estimated in money, in corn, or in any other commodity, but on proportional wages, that is, on the share of the commodities produced by the labourer, or of their value, which is given to the labourer’ (199). His only substantive criticism is of the unsatisfactory manner in which the division is actually determined by Ricardo. The inverse profit-wage relationship, he insisted, is ‘only one important step in the theory of profits, which of course cannot be complete till we have ascertained the cause, which under all circumstances, regulates this proportion of the whole produce which goes to labour immediately and accumulated’ (189). Malthus also believed that his own doctrine relating to the role of aggregate demand stood up empirically better than the Ricardian emphasis upon the difficulty of production on the land in accounting for the (supposed) historical decline in the profit rate, but at the same time insisted that the evidence accorded ‘most perfectly with the more general proposition of Mr Ricardo respecting profits, namely that they are determined by the proportion of the whole produce which goes to labour’ (198). 12 That Malthus accepted the substance of Ricardo’s position is also confirmed by a note of fundamental import attached to his Measure of Value (1823). Malthus had observed in the text that ‘it may be laid down ... as a general proposition, liable to no exception, that when the value of any produce can be resolved into labour and profits, then as the proportion of such produce which goes to labour increases, the proportion which goes to profits must decrease in the same degree, and as the proportion which goes to labour decreases, the proportion which goes to profits must increase in the same degree’ (1957, 28–9). 13 To this he added: This proposition is essentially the same as that which is very clearly and ably expressed by Mr Ricardo in his chapter On Profits in the following terms: ‘in all countries and at all times profits depend on the quantity of labour requisite to provide necessaries for the labourers on that land, or with that capital which yields no rent’; a proposition which though incomplete in reference to the ultimate causes of the variations of profits, contains a most important truth. From this truth the legitimate deduction appears to me to be, the constant value of labour; but Mr Ricardo has formed his system on a deduction exactly opposite to it. He has, however, in my opinion, amply compensated for the errors into which he may have fallen, by furnishing us, at the same time, not only with the means of their refutation, but the means of improving the science of Political Economy. (Malthus 1957, 29) 287

RICARDO - THE NEW VIEW

The criticisms alluded to briefly in the foregoing passage relate to Ricardo’s (supposed) failure to deal satisfactorily with real (basket) wage-rate determination and to the tortuous issue of the measure of value. Communications virtually broke down between the correspondents on the latter issue. But from our present perspective one outstanding fact remains; namely, Malthus’s insistence that by the relationship which he envisaged between the constancy of the value of a given command over labour and the proportionate shares, he was attempting to convey precisely the same concept as Ricardo himself. Schumpeter gave the central place to Samuel Bailey amongst those who reacted against orthodoxy: ‘Bailey’, he maintained, ‘attacked the Ricardo– [James] Mill– McCulloch analysis on a broad front and with complete success. His Dissertation, which said, as far as fundamentals are concerned, practically all that can be said, must rank among the masterpieces of criticism in our field, and it should suffice to secure to its author a place in or near the front rank in the history of scientific economics’ (Schumpeter 1954, 486). More specifically, he showed convincingly ‘the defects of the concept of real value and of the Ricardian theory of profit’ (599). Let us consider this evaluation. While Bailey maintained that the notion of an invariable standard ‘amidst universal fluctuations’ involves an inherent contradiction, it is quite clear that this follows from his own definition of value as price, or exchange value. His case turns out to be a purely terminological one: The specific error of Mr Ricardo on the subject of invariable value consists ... in supposing, that if the causes of value affecting one commodity remained the same, the value of that commodity could not vary, overlooking the circumstance that value denotes a relation between two objects, which must necessarily alter with an alteration in the causes affecting either of them. He incessantly identifies constancy in the quantity of producing labour with constancy of value. Hence he maintains, that if we could find any commodity invariable in the circumstances of its production, it would be in the first place invariable in value; and, 2ndly, it would indicate, or would enable us to ascertain, the variations in value of other commodities. (Bailey 1825, 121) In point of fact a commodity produced with a constant labour input, Bailey observed, ‘would enable us to ascertain, not the fluctuations in value between two or more commodities (for these are facts to be gathered from appropriate evidence), but the fluctuations in the quantity of labour which produced them’ (124). Alternatively expressed, we would be enabled to establish in which commodity ‘those [observed] fluctuations had originated’ (121). Now this renders precisely the purpose of Ricardo’s measure, and the critique amounts to no more than an unwillingness to use the term ‘value’ in 288

THE RECEPTION OF RICARDIAN ECONOMICS

the Ricardian fashion. Bailey himself seems to concede as much: ‘Many of the strictures, which have been made on Mr Ricardo’s writings ... would be in some degree obviated if two things were conceded, namely, if we assumed that he was constantly speaking of real value, and if we were to grant him the absurdity which we have shown this expression to imply; or, in other words, if we were to consider it as importing [ sic] cost of production, without relation to the power of commanding in exchange’ (253). 14 The terminological penchant of the book is also evident from Bailey’s comments upon Ricardo’s identification of a change in the ‘value’ of wages – wages in terms of the invariable measure – with a change in the proportion of wages in aggregate output, that is the representation of labour as rising or falling in value ‘only when a larger or smaller proportion of the commodity produced goes to the labourer’. For Bailey simply altered the sense of the term ‘value’ to refer to exchange-value; from which it followed that a change in the proportion of wages in total output was only one among several possible causes of variation in the ‘value of labour’, that is Bailey’s term for commodity wages (46). On this terminological usage, when allowance is made for changes in aggregate output generated by increased labour productivity, and neglecting rent, wages and profit may both vary in the same direction. 15 As Bailey pointed out towards the close of his work his contention was simply ‘that in cases ofi mproved productive power, the product might be so divided, that the rate of profits [identified with the profit share in output] should be increased while the value of labour [the commodity wage] was enhanced’ (241). This conclusion, however, is totally in accord with Ricardian analysis. Bailey’s criticisms of the inverse profit–wage relation were evidently not directed at Ricardo’s position as Ricardo understood it. When Bailey actually considered this matter he conceded frankly enough the legitimacy of Ricardo’s case: Mr Ricardo’s inference is a legitimate deduction from his premises, if we concede certain postulates. Grant him the kind of value called real, which has no relation to the quantity of commodities commanded, but solely to the quantity of producing labour, and it inevitably follows, that there could be no alteration in the real value of labour, but from an alteration in the proportion of the product which went to the labourer. Neither, if money were always produced by a uniform quantity of labour, could there be any alteration in the money-value of labour. (Bailey 1825, 58n) Indeed, Bailey’s recognition that the profit rate falls when commodity wages rise (given labour productivity) runs along strictly Ricardian lines, involving as it does the proposition, which conflicts with that of Adam Smith, that capitalists cannot 289

RICARDO - THE NEW VIEW

pass on the higher wage costs to consumers in the form of higher prices, since the prices, that is the exchange values, of all commodities cannot logically be raised simultaneously: If labour rises while [its] productive powers remain the same, profits will inevitably fall. This may be easily proved from the principles already advanced; for if labour rises in value, whoever purchases labour must give a greater quantity of other things for it, and as the capitalist purchases labour, he must pay more for it. It will be said, perhaps, that he may raise the value of his goods, that is, he may require a greater quantity of other commodities than before, in exchange for his own. But the capitalist who produces these other commodities is in the same predicament, and they cannot both raise their goods ... it is a contradiction to maintain, that a universal rise in the value of labour can increase the value of commodities. (Bailey 1825, 64–7) Bailey’s acceptance of this basic Ricardian objection to Smith cannot be ignored. It follows that Schumpeter’s high praise for Bailey on the grounds that he showed convincingly the ‘defects of the concept of real value and of the Ricardian theory of profit’ is scarcely justified. Bailey failed on the whole to analyse Ricardo’s objectives in devising his measure, concentrating as he did upon terminology, but he nonetheless accepted the accuracy of Ricardo’s position regarding the profit–wage relationship on Ricardo’s use of terms and rejected Smith’s view of the relation between wages, profits and prices. The entire Bailey episode has been overdone as an indication of revolt against Ricardo. It is also pertinent that Bailey himself presented no alternative model, 16 although he did show a preference for far greater ‘generality’ than was apparent among the Ricardians, and in his conclusion he emphasized the oversimplification characteristic of Ricardian method. Even in this regard he made allowances, conceding that Ricardo himself had recognized increasing costs and therefore elements of ‘scarcity’ or of ‘partial monopoly’ – utilizing the early nineteenth-century term for less than infinitely elastic supply – at least in the case of corn and gold. He took issue with Ricardo’s formal position that ‘commodities can be classified according to the “source” of their exchangeable value’ – namely ‘scarcity’ and the quantity of labour embodied – and attention focused upon the latter since this group is quantitatively the most significant; and in dealing with price formation he talked of the role of ‘considerations acting on the human mind’ (227–8). Yet at the same time the ‘main considerations’ in this context are said by Bailey himself to be production costs (199). In fact, insofar as concerns matters involving the so-called ‘causes of value’ – that is the theory of price – there is no divergence of substance between Ricardo and Bailey. 290

THE RECEPTION OF RICARDIAN ECONOMICS

Bailey generalized the conception of differential rent to apply to labour but also adopted the position that rent is not a ‘cause’ of value (197), while his insistence that rent may be generated even in the absence of differential land qualities, by dint of land scarcity, was conceded also by Ricardo. Bailey forcefully rejected statements of the labour theory by James Mill, De Quincey and McCulloch (207–8), but was less harsh with respect to Ricardo himself who, he recognized, qualified the theory (213– 14).17 C.F. Cotterill’s famous comment that there are ‘some Ricardians still remaining’ led Professor Schumpeter to conclude that ‘the decay of the Ricardian school must have became patent’ shortly after Bailey’s contribution (1954, 478). It is not, however, difficult to show that it is unjustified to draw such broad implications from what in any event is a most ambiguous statement, for the fact is that Cotterill also refers to variations of exchange value proportionate to alterations in relative labour input as a proposition which ‘most economists maintain’ – in contrast to Smith’s position relating increases in prices ‘to the rise in wages’ (1831, 107). Space limitation precludes a full demonstration of the fact that much of Cotterill’s criticism of Ricardo was misplaced. We wish, however, to emphasize Cotterill’s recognition of the need for an ‘invariable standard’, and his belief in the possibility of its construction. Amongst the most important features of his pamphlet Cotterill himself counted the statement of ‘the conditions necessary to a standard of value’ and the refutation of ‘the erroneous doctrine that the supposition of such a standard involves contradictory conditions’ (v). And it was Bailey (as well as Torrens) whom he largely had in mind by this latter statement (99). 18 His main objection as far as concerned Ricardo himself was to the specific character of Ricardian ‘money’ as an invariable standard – in his estimate it was impossible to talk meaningfully of exchange value in terms of such a medium – but not to the principle of the matter. In fact Cotterill himself provided illustrations of the ‘correct’ procedure designed to guarantee that aggregate money value remains unchanged in the face of an alteration in distribution (76). And he candidly recognized Ricardo’s contribution to the proposed solution: ‘The conditions, however, essential to an invariable standard, do appear to me so very obvious, especially upon Mr Ricardo’s own principles in opposition to Adam Smith’ (75n). H. Merivale’s single adverse comment on Ricardian rent theory – ‘a lapsis offensionis, startling and offending many’ – cannot safely be regarded as indicative of a generally hostile position as is sometimes maintained (Gordon 1969, 378). The review in question considered in its entirety is not hostile. On the contrary, Merivale supported Ricardian method which he talked of as the method of the ‘modern English school’ (1837, 77, 79–80), and was very favourably disposed to McCulloch’s edition of the 291

RICARDO - THE NEW VIEW

Wealth of Nations since it indicated ‘the principal notions of an erroneous character which pervades the whole of the great work of Dr Smith’. By contrast, ‘Mr Ricardo investigated the fundamental truths of the science with singular profoundness; his theories, while they have led many followers astray, have nevertheless penetrated thoroughly into all subsequent lucubrations on the subject and he may be regarded, more justly than any other, as the real founder of the school which at present exists in England’ (1840, 429). We turn now to Longfield. Mountifort Longfield probably represents for historians of economic thought the most original of that group of writers known variously as ‘the Dissenters’, ‘the Men who Wrote above Their Time’, and ‘Voices in the Wilderness’. Professor Seligman referred to him in his celebrated paper ‘On Some Neglected British Economists’ as ‘in some respects the most remarkable of all’ (1903, 525), and Professor Schumpeter wrote of Longfield that ‘he overhauled the whole of economic theory and produced a system that would have stood up well in 1890’ (1954, 466). Professor Blaug has maintained that ‘of all the treatises on political economy which appeared in the 1830’s the most original was Longfield’s Lectures ’ (1958, 159), and Professor Bowley awards Longfield ‘the title of the first of the Victorian economists, or perhaps the first of the neo-classics’ (1973, 217). We may also refer to the following evaluation by Professor Black in his Introduction to the recently published Economic Writings of Mountifort Longfield : ‘It would certainly be wrong to represent Longfield’s analysis as a complete anticipation of neoclassical pricing and distribution theory. Yet whatever limitations it may have, whatever classical elements may be left, incompletely integrated, within it, the fact remains that it is a fundamentally and systematically different analysis from the Ricardian one .... Only Longfield, amongst those who wrote in English at this period, offered his readers a theory of value which was complete – and completely original’ (1971, 15–16). In our view the flavour of much of the discussion of the Lectures is unmistakably ‘Ricardian’. We have in mind the emphasis upon long-run cost price, the choice of a labour measure of value, the rationale given for such choice in terms of the overwhelming role played in price determination by alterations in labour input, and the exclusion of alterations in relative wages or relative profits as a source of disturbance. It has been observed of some of these Ricardian features that they appear in the early lectures given at Dublin in the Trinity term of 1833, while all Longfield’s ‘original contributions’ occur in the sixth and later lectures given in the Michaelmas term. His conception, runs this argument, altered as the year progressed and only by late 1833 was he in a position to formulate the subjective theory of value and the productivity theory of distribution. It was during the period between terms in mid-1833 that Longfield ‘seems to have decided to break away from classical 292

THE RECEPTION OF RICARDIAN ECONOMICS

distribution theory and from Lecture VI on his analysis proceeds with impressive originality’ (13). If, however, we examine this hypothesis closely it will be found wanting in several respects and it is our view that, on the whole, Longfield retained a Ricardian structure throughout. At the very commencement of Lecture VI – where the fullest treatment of demand is undertaken – the main Ricardo-like propositions regarding value are re-stated. Thus the labour measure is again justified as a suitable (though not unique) measure: ‘as most of the commodities in which the wealth of the country consists are produced by labour, political-economists make use of it as a measure of value’ (Longfield 1834, 109). And it is in Lecture VIII where we read that ‘all the commodities which men consume, and which can be made the subject of exchanges, owe their existence and their value to labour. The exceptions to this are very trifling, and are of such a nature that they do not vitiate any of the conclusions drawn from it’ (164). Moreover, the assumption of a constant wage and profit structure is maintained throughout the book. The secondary importance of demand theory as far as concerns the main issue of distribution is also quite apparent in the sixth lecture, which is largely devoted to the theory of rent. The differential rent theory served for Longfield precisely the same function as for Ricardo, namely to permit the focus of attention upon the wage–profit relationship: ‘This analysis I shall enter upon, merely for the purpose of disengaging the cost of production from this element of complexity, and thus of rendering questions concerning wages and profits more simple, by freeing them from a source of confusion and vicious reasoning, in circles to which they are particularly liable’ (116). (The same theme is reiterated in the seventh lecture; 132.) 19 We must here take into consideration also Longfield’s rejection of Malthus’s insistence – evidently derived from Smith – upon the multifold uses of land which renders rent a cost in any single use: ‘I cannot agree with Mr Malthus, in supposing that the rent which the worst land could pay, as pasture land, should be deemed part of the cost of production of corn’ (148). Thus the standard rent theory presented in Lecture VI was not only accepted by Longfield but defended competently along lines which, for the most part, would have been favoured by Ricardo. The only matter with which Longfield took issue related to some formal data in the Principles which imply a continually increasing proportion of rent in total produce (153f). Yet we know that Ricardo himself did not stand by this relationship and altered his text at various points in the third edition to avoid any such implication. Longfield defined profits as a ‘discount which the labourer pays for prompt payment’, or as the ‘sacrifice of the present to the future’ by the possessor of wealth ‘who uses it as capital’. But he recognized that these expressions merely constitute a definition of ‘what profits are’ (179). He also attempted to show ‘how their amount 293

RICARDO - THE NEW VIEW

is calculated’, and in the course of this exercise he was led to define – as he phrased the issue in the Preface – ‘the meaning of proportional wages as they regulate profits’ (xi), and to defend the Ricardian conception against criticisms by Torrens and McCulloch. We are faced with a quite extraordinary reversal of roles. Assuming that all advances are made for one year, and also that all advances constitute wages, the rate of profit can be defined as ‘the proportion in which the value of any commodity is divided between the labourer and the capitalist’ (171). 20 ‘This proposition’, he conceded, ‘may be considered useless or untrue, as depending upon false suppositions. It is, however, true in those cases in which it does apply, and all other cases may, with a little care, be reduced to them.’ And now it becomes clear for what ultimate purpose the measure of value was devised. ‘Such reduction’, he continued, ‘must be made, whenever we resort to labour as a common measure for comparing the values of commodities. Whatever advances are not made in labour must be reduced to the measure of labour. If a capitalist expends £50 in raw materials, it must be considered as so much advanced on account of labour. In order to make use of labour as a measure of value, we, as it were, reduce every thing else of value to that denomination.’ The object of all this was to confute Torrens’s assertion that the doctrine according to which profits depend on wages ‘is equally untenable, whether the terms alteration of wages, alteration of profits, are employed with a reference to proportions, or whether they are used in relation to quantities’ (1826, xv–xvi). Torrens based his objection upon an arithmetical illustration purporting to show that proportional wages may remain unchanged while the rate of profits increases. The error of the calculation lay, Longfield insisted, in a failure to recognize that all inputs other than labour may be reduced to labour so that the only payments to be taken into consideration are wages and profits (1834, 172f). Longfield thus accepted the Ricardian proposition that the profit rate ‘depends upon the proportion of the shares of the final value received by the labourer and the capitalist’. Indeed in an Appendix he formulated, and justified, Ricardo’s isolation of the conditions required to assure that changes in money wages will reflect changes in proportionate wages: Mr Ricardo frequently asserts that wages and profits together are always of the same value, and that nothing but a rise in one can produce a fall of the other. He uses the term ‘wages’, sometimes to signify the absolute wages which the labourer receives, and sometimes as the proportional wages. This did not arise from an abuse of words or a confusion of ideas, it was the natural and necessary consequence of a hypothesis, which for simplicity of illustration, he laid down at the commencement of his work. He assumes that any given quantity of gold, or the metal of which money is made, is always produced by the same quantity 294

THE RECEPTION OF RICARDIAN ECONOMICS

of labour, with the same interval of time before its production. On this hypothesis, if gold is raised from mines within the kingdom, absolute money wages will be identical with, and be measured by, proportional wages, and a rise of wages, that is of money wages, will always be accompanied by a fall of profits. (Longfield 1834, 266–7; emphasis added) 21 The entire discussion thus far has merely provided a definition of profits and a statement of how the rate of profits is to be calculated. The ninth lecture is devoted ‘to the investigation of the laws which determine their actual amount’ (179), wherein Longfield finally set out his objections to his conception of orthodox theory and advanced his alternative. And although he fails to state precisely how his own theory was related to the preceding defence of the Ricardian inverse profit–wage relationship, it seems clear enough that he envisaged the latter as a valid framework for a satisfactory theory built around the efficiency of capital (envisaged as ‘machinery’) in its least productive application within a demand-supply context (187f). Longfield’s market-price analysis, involving the notion of marginal demand price, had rather limited direct influence upon subsequent subjectivist economic thought even in Ireland. And those writers, such as Butt and Lawson, who drew upon elements of Longfield’s distribution analysis did not appreciate its relationship with an archetypal supply–demand theory (Moss 1974, 427f). 22 Nonetheless, the positive influence of Longfield was profound for his lectures elicited a formal retraction from Torrens of his severe critique of Ricardian distribution theory as we shall now see. In his third edition (1826) of the Essay on the External Corn Trade Torrens had described the Ricardian inverse wage–profit relationship as ‘equally untenable, whether the terms alteration of wages, alteration of profits, are employed with a reference to proportions, or whether they are used in relation to quantities’ (above, p. 294). Yet the fact is that there probably exists no clearer statement in the entire literature of the period of the pure Ricardian theory of distribution involving the proportionsmeasuring ‘gold’ than that presented, favourably, by Torrens in his Colonization of Southern Australia (1835, 22–35). Similarly, in 1844 Torrens candidly withdrew his earlier objections: ‘Some of the commentators on the doctrines of Ricardo appear to have fallen into the misconception, that, in altering his nomenclature and in modifying his principles as varying circumstances required, they refuted his theory of profits. In this censure I include myself.’ He had in mind precisely the Ricardian principle that ‘profits rise or fall only as wages fall or rise’ (1844, xxxvi), and provided a defence of the proportions measuring money against a number of strictures by Senior and Jones (xxii f). The change in position was attributed by Torrens himself 295

RICARDO - THE NEW VIEW

to Longfield’s lectures which had ‘succeeded in removing the main objection to the reception of Ricardo’s theory of profit, by showing ... that the cost of production is to be measured by the single standard obtained by reducing previous and proximate labour to a common denomination’ (li–lii). 23 The Ricardian frame of reference is less marked in the works of Samuel Read – which Schumpeter noted ‘bears witness to the influence of Bailey’ (1954, 488) – and G. Poullet Scrope than in that of the other ‘dissenters’ we have considered so far. But even in their cases the evidence is mixed: Read insisted upon an inverse relationship between the wages of regular labour and the wages of management, 24 and Scrope’s objections relate not so much to the inverse relationship as such but to the precise differential effects exerted by wage changes – a characteristic Ricardian problem (Scrope 1831, 29). To take this position is in effect to concede a very great deal of the Ricardian position. And the fact is that Scrope formally granted the validity of the inverse relationship on Ricardo’s own use of terms albeit as an uninteresting truism. 25 As far as concerns increases in wages in ‘the correct and ordinary sense of the word’, Scrope insisted, the proposition that a wage increase will leave unaffected the price level is ‘directly false’ (3). Unfortunately Scrope does not justify this assertion. What he presumably meant by the ‘ordinary’ sense of the term is ‘the money-price of labour’, and even Ricardo agreed that a money-wage increase, if merely nominal, will be accompanied by a general price rise. 26 It is commonly asserted that R. Whately’s celebrated suggestion to rename political economy ‘Catallactics’, or the science of exchanges, implies a significant indication of a deflection of classical economics from its primary preoccupations to a narrowly constrained emphasis upon the ‘sphere of circulation’. But this interpretation does not accurately reflect Whately’s intentions. His purpose in proposing the change of nomenclature was to assure that a study of the wealth of nations should be limited to exchangeable commodities and exclude, for example, the analysis of problems relating to a ‘Robinson Crusoe’ type economy (Whately 1832, 8). 27 That the emphasis was not on the phenomenon of exchange as such is clear from his rejection of the term ‘philosophy of commerce’ as a designation of the subject matter (9). Whately’s suggested redefinition cannot be interpreted as a sophisticated attempt to redefine the scope and nature of political economy. In the final resort he remained concerned with ‘the nature, production, and distribution of wealth’ (26), and regarded the study of what wealth ‘consist[s] in’ and what are ‘the fundamental laws that regulate its distribution’ as essential for an understanding of topical questions (223). 28 It will be clear that the extent of substantive hostility towards Ricardo is considerably 296

THE RECEPTION OF RICARDIAN ECONOMICS

less than is generally maintained. In concluding this section it may be appropriate to refer to the history of thought written by Travers Twiss, who, as Professor of Political Economy at Oxford, held the same chair as Senior, Whately and Lloyd. Twiss in 1847 accepted Ricardo’s position that rent is not an element in price (169–70). He adopted the view, in sharp contrast to Smith, that ‘variations in the rate of wages and profit, where the proportion of fixed capital is unchanged, affect all commodities in the same degree, so that their relations, in respect of such variations, would remain undisturbed, and their exchangeable value not be altered’ (170–1). And he accepted Ricardo’s distinction between circulating and fixed capital in terms of durability, observing that ‘variations in the rate of wages or the rate of profits, will ... affect the cost of production very differently, according as the use of perishable or durable capital preponderates in any branch of manufacture’ (188). There was certainly no general Oxford revolt against Ricardo. II. SOME REACTIONS BY ‘DISSENTERS’ TO BAILEY We return now to consider Bailey’s place in the dissenting literature. That Bailey cannot legitimately be regarded, as is so common, as standing at the head of a line of anti-Ricardians is clear if we consider a little more closely some reactions to his work by other so-called ‘dissenters’. 29 Professor Schumpeter commented of Samuel Read’s work that it ‘bears witness to the influence of Bailey whom Read followed in his Ricardo criticism’ (1954, 488). This comment is quite inaccurate. On the contrary, Read explicitly took Bailey to task in very strong terms for neglecting to consider the value of commodities in terms of ‘their relation to mankind and to human labour’: Notwithstanding the very high respect I entertain for this author, it will be seen in the following pages, that I find occasion to differ from him very widely in his main positions in the ‘Critical Dissertation’. It appears to me that the fundamental error in that work, and that from which all the others to be found in it flow, consists in his treating of value as if it were a mere relation of commodities between themselves ; whereas it appears to me that the idea of value in commodities cannot even be conceived without being mingled with the idea of their relation to mankind and to human labour, of which some portion must always be employed in producing or procuring them originally. (Read 1829, viii) 30 Professor Meek in his study of the ‘decline of Ricardian economics in England’ refers to this same passage relating value to ‘mankind and human labour’ but mistakenly understands it as a position rejected by Read who attributed it to 297

RICARDO - THE NEW VIEW

Ricardo, rather than one maintained by Read himself, and observes that ‘it was this vital concept which virtually vanished from English political economy after Ricardo’s death’ (1967, 67). The evidence to the contrary speaks for itself at least as far as concerns Read. Criticism of Bailey was in fact widespread amongst the dissenters. It is not apparently common knowledge that Cotterill was dissatisfied with significant parts of Bailey’s performance regarding the ‘cause’ of value: ‘the causes of value, as stated by the author of the “Critical Dissertations &c” are extremely unsatisfactory: notwithstanding many parts of that work cannot but be admired by everyone friendly to the science of political economy’ (1831, 53). 31 Similarly, he did not think that Bailey’s work constituted a performance likely to set the controversy at rest; on the contrary, I consider the Author’s use of the term value, on many occasions, inconsistent with its obvious meaning, and the chapter on the causes of value, in my opinion, leads to nothing less than complete scepticism. So that, however certain I am, that the work has done much to evolve the difficulties in which the subject was acknowledged to be enveloped, and however dexterously I conceive the author detected the double meaning in which Ricardo used the term I cannot but regret the appearance of these blemishes. (Cotterill 1831, 40–1) 32 He was equally critical of Bailey’s position on the measure of value as we have seen (above, p. 291). Lloyd is quite specific as to the doctrine to which he objected. His criticisms are not directed against Ricardo at all, although of course Ricardo’s definition of ‘value’ differed from his own. The purpose of his pamphlet was to demonstrate the legitimacy of saying of a commodity that it varies in value ‘without any reference to other objects’ (1834, 28). He rejected the assertion (which in fact was by J.S. Mill) that ‘value is a relative term: if it is not this, it is nothing if any one talks about absolute value, or any other kind of value than exchangeable value, we know not what he means’ (34). 33 He took J.B. Say to task for his analogy between motion and value in his argument that ‘all value is relative’. And he found fault with the pamphlet entitled Observations on Some Verbal Disputes in Political Economy (1821), which, as we shall see, bears a close resemblance in many respects to Bailey’s Critical Dissertation. 34 III. THE RICARDIANS We devote this section to a brief consideration of the ‘Ricardians’ themselves, specifically McCulloch, J.S. Mill and De Quincey, and their relationship with Ricardo. 298

THE RECEPTION OF RICARDIAN ECONOMICS

Recent commentaries on McCulloch perform a ‘disappearing trick’, as it were, in his case. He is either dismissed, as by Schumpeter, as a ‘henchman’ of Ricardo and therefore apparently not worthy of much attention, or he is classified as by O’Brien as a ‘Smithian’. Neither position seems adequate. In particular the argument of Professor O’Brien according to which McCulloch was not essentially Ricardian but directly in the Smithian tradition is overstated. The treatment of the invariable measure is taken – with considerable justification – to be the hallmark of Ricardianism; it is the statement that the invariable measure ‘never interested McCulloch at all’ (1970, 46) which I believe involves a misreading of McCulloch’s position. 35 McCulloch’s formal denial of the existence of an invariable measure of value must be seen in proper perspective. In all editions of his Principles the existence of such a measure is denied on the empirical grounds that all commodities are subject to ‘perpetual variations’ in the quantities of labour required to produce them. 36 Yet the fact remains that money was assumed to have the mean factor proportions so that at least one of the Ricardian prerequisites for a suitable measure is presumed to be actually satisfied in practice. Moreover, McCulloch distinguished conceptually between the case of a nominal wage change, and that of a ‘real’ wage change reflected in an altered wage in terms of money of ‘invariable value’, and it is the latter which concerned him. Equally significant is McCulloch’s insistence that while it could be demonstrated that a rise in wages would leave unchanged the level of prices, yet even in the event that the price level did rise the real value (in the sense of purchasing power) of profits must fall. And, finally, and I believe conclusively, since Ricardo also had forcefully insisted that his ‘fundamental theorem of distribution’ (to the effect that the burden of a wage increase falls on capitalists) held good whether or not the conditions required for a suitable measure of value were satisfied in practice, I conclude that McCulloch’s position in all essentials is identical with that of Ricardo. Schumpeter shares with Marxist interpreters the view that J.S. Mill must be excluded from that group which constitutes Ricardo’s ‘school’ namely James Mill, McCulloch and De Quincey (1954, 476). It is contended in particular that Mill rejected on Bailey’s grounds the conception of a measure of absolute value (589, 603).37 We have considered this view elsewhere and found it to be untenable: Mill accepted the Ricardian theory in the strict sense outlined earlier, specifically the derivation of the theorem about profits and wages from the concept of the invariable measure.38 It may not appear necessary to demonstrate De Quincey’s loyal adherence to Ricardo. But the tide is so strong in favour of the view that the Ricardian influence was short-lived that even he is being pushed into the opposition. 39 We must, however, keep in mind the opening declaration in the first part of the three-part essay for Blackwood’s in 1842 (1970b, 113) that ‘David Ricardo made the first and last effort that can be made to revolutionize that science which, for nations, professes to lay 299

RICARDO - THE NEW VIEW

bare the grounds of their prosperity, and for individuals, as distributed by nature into three great orders of proprietors, the grounds of their expectations’. There can be no doubt of his continued allegiance to Ricardo as will be made clear in the discussion which follows of The Logic of Political Economy. 40 De Quincey’s dissatisfaction with the treatment in the literature of market prices diverging from long-run costs is quite apparent and Ricardo is indeed included amongst the culprits; Ricardo’s fourth chapter is termed ‘a very short chapter and a very bad one’ (cf. 1970c [1844], 200). But it is Adam Smith who throughout bears the brunt of the criticism for his statement of the ‘paradox of value’ (144–5, 188–90), as well as those who ‘fancy that the relation of Supply and Demand could by possibility, and that in fact it often does, determine separately per se the selling price of an article’. 41 Furthermore, De Quincey spoke very warmly of Ricardo’s chapter ‘On Value and Riches’, writing of it as ‘essentially novel’ and worthy of ‘special admiration’ (127). More significantly, when he turned to consider the case of longrun price – and without question this constituted his main concern in value theory – Ricardo is allowed to come fully into his own. Thus the chapter ‘On Value’, upon which is based ‘a total revolution of political economy’, is described as ‘Ricardo’s great reform’ (181), his ‘great inaugural chapter’ (200), after which ‘the powerful hand of Ricardo, will be felt in every turn and movement of economy’ (179). 42 It is in fact the labour theory which De Quincey had in mind modified by the complication created by different proportions of fixed and circulating capital. Moreover, De Quincey’s distribution theory is strictly Ricardian, as indeed is the relationship between distribution and value. Ricardo’s illustrative data on wages and rent are fully reproduced (224–5, 246). The emphasis is upon the inverse profit– wage relation which in Ricardian fashion is set squarely upon the labour theory: ‘Even the novice is now aware that a rise in wages would leave prices undisturbed’ in contrast to the position of ‘the superannuated economic systems smashed by Ricardo’ which envisaged wage increases passed on in the form of higher prices (248n). 43 More specifically: In one brief formula, it might be said of profits that they are the leavings of wages : so much will the profit be upon any act of production, whether agricultural or manufacturing, as the wages upon that act permit to be left achieved. ... But do not the wages and profits as a whole, themselves, on the contrary, predetermine the price? No; that is the old superannuated doctrine. But the new economy has shown that all price is governed by proportional quantity of the producing labour, and by that only. Being itself once settled, then, ipso facto, price settles the fund at which both wages and profits must draw their separate dividends. ... 300

THE RECEPTION OF RICARDIAN ECONOMICS

But if that is true, then it follows that wages and profits vary inversely. ... Any other rise in profits, such as would leave wages virtually undiminished, could be only an apparent rise through some depreciation in the currency; and that depreciation, changing any one thing nominally, must change all other things: affecting all apparently, really it would affect none. (De Quincey 1970c [1844], 257–8) 44 Finally we recall De Quincey’s adherence to Ricardian rent theory and more specifically Ricardo’s analysis of its consequences for profits, wages, and prices, which placed economists ‘irredeemably’ in his debt. 45 There can surely be no question of De Quincey’s continued adherence to strict Ricardianism. IV. SOME RICARDIAN REACTIONS TO THEIR CRITICS In many important instances the post-Ricardian critics simply misinterpreted Ricardo. Malthus, for example, believed quite erroneously that Ricardo maintained his cost theory of exchange value as an alternative to demand–supply theory (1963 [1824], 181–2). Both Malthus and Longfield (1971 [1834], 184f) asserted without justice that in Ricardo’s system rising capital with population unchanged leaves the profit rate unaffected – that the only cause of falling profits was resort to inferior land. Bailey implied that Ricardo failed to appreciate the relativity dimension of exchange value. Senior’s objection to Ricardo – adopted also by Bailey and T.P. Thompson – that to say ‘it is the price of [the] last portion of corn, which governs that of the remainder, is to mistake the effect for the cause’ and his adoption, as an alternative, of a demand– supply or ‘monopoly’ explanation falls into the same category. 46 The fact is that the Ricardians – and to a considerable degree Ricardo himself confirms the point – accepted much of the substantive argument of the ‘critics’. We consider first some Ricardian reactions to Bailey. A Westminster reviewer, possibly James Mill, made the point that Ricardo would not have objected to the principal theme of the first chapter of the Critical Dissertation relating to the nature of value: ‘This chapter is logamarchy, simply and purely. It makes profession, or rather ostentation and parade, of being a controversy with Mr Ricardo. But it contains not an assertion, to which, as far as ideas politico-economical are concerned, Mr Ricardo would not have assented; it contains, not indeed, as far as such ideas are concerned, an assertion which is not implied in the propositions which Mr Ricardo has put forth. It is a criticism of some of Mr Ricardo’s forms of expression.’ 47 And that this would have been Ricardo’s view is clear from his reaction to the anonymous Verbal Disputes of 1821, which in its essentials is so similar to the Critical Dissertation that Bailey 301

RICARDO - THE NEW VIEW

feared he might be accused of plagiarism. Ricardo observed:

48

Writing to Trower in August 1821

With respect to our difference of opinion on the subject of exchangeable value it is more an apparent difference than a real one. In speaking of exchangeable value you have not any idea of real value in your mind – I invariably have. Your criticisms on passages in my book are, I have little doubt, correct, because they are also the criticisms of others on the same passages. A pamphlet has appeared ‘On Certain Verbal Disputes in Polit. Econ.’ where the same ground of objection is taken as you take; the fault lies not in the doctrine itself, but in my faulty manner of explaining it. (22 August 1821; Ricardo 1951, IX, 38) J.S. Mill, in a Note attached to McCulloch’s edition of the Wealth of Nations (1828) alluded to Senior’s objections to Ricardo’s rent theory when he wrote of those ‘who affect to suppose that Sir Edward West, Mr Malthus, and Mr Ricardo, considered the cultivation of inferior land as the cause of a high price of corn’. The argument that ‘the cultivation of inferior soils is not the cause but the effect of high price, itself the effect of demand’ was, he insisted, a doctrine ‘explicitly laid down by the distinguished authors previously referred to, and particularly by Mr Ricardo’ (1967, IV, 174). Much the same argument would apply to the strictures of T.P. Thompson.49 Mill’s reaction was perfectly justified. Ricardo himself said (to Trower) of Senior’s article: ‘I am glad to have got so good an ally, for what I think the correct principles, and you must partake of the pleasure which I feel in observing that they are every day making way’ (11 December 1821: 1951, IX, 122). The fact is that in the Principles Ricardo had shown a thorough awareness that differential rent is but a special case of a more general phenomenon. In his discussion of early settlements, he observed, ‘no one would pay for the use of land, when there was an abundant quantity not yet appropriated, and, therefore, at the disposal of whosoever might choose to cultivate it’. 50 The further extension of this statement runs explicitly in terms of demand and supply (and refers to the authority of J.B. Say): ‘On the common principles of supply and demand, no rent could be paid for such land, for the reason stated why nothing is given for the use of air and water, or for any other of the gifts of nature which exist in boundless quantity ...; as the supply is boundless, they bear no price. If all land had the same properties, if it were unlimited in quantity and uniform in quality, no charge could be made for its uses unless where it possessed peculiar advantages of situation’ (I, 69–70). Indeed Ricardo takes Smith to task for failing in his occasional emphases upon the physical contribution of the land factor to appreciate the dependency of the value dimension of rent upon scarcity: ‘Dr Smith does not reflect that rent is the effect of high price. ... It is ... from the price at which the 302

THE RECEPTION OF RICARDIAN ECONOMICS

produce is sold, that rent is derived; and this price is got not because nature assists in the production, but because it is the price which suits the consumption to the supply’ (I, 71n). It should be added that it is perfectly consistent with the Ricardian position that rent may be generated even in the absence of differentials. We have in mind not merely differentials reflecting differences in the ‘qualities’ of successive plots, for it is diminishing returns at the intensive margin which constitutes a more general case (I, 71, 72) but the increase in rent due to a constraint upon output expansion of any kind – whether at the extensive or intensive margins – in the face of rising demand. 51 We turn next to the Ricardian response to Malthus. In his Encyclopaedia Britannica contribution of 1823, commenting upon Malthus’s Principles of 1820, McCulloch observed that while ‘some of the subordinate doctrines respecting value advanced by Mr Ricardo in the first and second editions of his Principles of Political Economy and Taxation were opposed by Mr Malthus in his recent publication’, yet ‘Mr Malthus does not attempt to invalidate the leading principles established by Mr Ricardo, and the alterations and corrections which the latter has made in the third edition of his work have gone far to remove the objections of Mr Malthus’. 52 And that this is indeed so is clearly revealed by Malthus’s remarks, as we have already noted, which suggest that he adopted much of the Ricardian inverse profit– wage relationship (above, pp. 286–8). How did Ricardo himself view Malthus’s position on distribution published in 1820? The treatment of the Profits of Capital in Malthus’s Principles met with Ricardo’s approval. Ricardo observed in his Notes that ‘I maintain no other doctrine than that which has been well explained by Mr Malthus in the 2 first sections of 5th chapter. His own statements are sometimes at variance with it, mine I believe never’ (1951, II, 288). More specifically, Ricardo commented favourably on Malthus’s analysis of the adverse effects upon the profit rate generated by increased difficulty of producing wage goods, and by increased commodity wages and concluded: ‘These 2 causes may both be classed under the name of high or low wages. Profits in fact depend on high or low wages, and on nothing else. ... In all this Mr Malthus and I appear to concur. Whenever the difficulty of production on the land is such that a greater proportion of the value of the whole produce is employed in supporting labour, I call wages high, for I measure value by these proportions ; and from Mr Malthus’s language here, everybody would think he agreed with me’ (II, 252). Ricardo’s great complaint was that Malthus unjustly insisted that his analysis failed to allow for the effect of real (basket) wage variation upon the profit rate: ‘I have invariably insisted that high or low profits depended on low and high wages, how then can it be justly said of me that the only cause which I have recognized of high or low profits is the facility or difficulty of providing food for the labourer. 303

RICARDO - THE NEW VIEW

I contend that I have also recognized the other cause, the relative amount of population to capital, which is another of the great regulators of wages’ (264). 53 Ricardo also saw eye to eye with Malthus in the Principles regarding the effects of changes in distribution upon prices. On reading of Malthus’s acceptance of the proposition that a fall in profits will generate a decline in the prices of capitalintensive goods, Ricardo observed: Now I confess that I feared Mr Malthus himself would have found the proposition paradoxical, because in some of his works he has maintained that a rise in the price of corn will be followed by an equal rise in the price of labour, and an equal rise in the price of all commodities; and it was only after further consideration that he thought it fit to reduce the proportion in which commodities would vary when corn varied, and to fix it at 25 or 20 p.c., when corn varied 33-1/3 – that is to say when corn varies 100 p.c. commodities are to vary 75 to 60 p.c. (Ricardo 1951, II, 60–1)54 Actually Ricardo saw his position as closer to that of Malthus than to that of McCulloch in the light of the extreme position sometimes adopted by the latter regarding the labour theory. 55 It has recently been observed, quite correctly we believe, that while Ricardo was on good terms with both McCulloch and Malthus ‘he did not value their opinions equally ... Malthus seems to be the person whom Ricardo took more seriously than he did anyone else’ (Grampp 1974, 284). As for Malthus’s contention in the Quarterly Review of 1824 that the New School – unlike Adam Smith whose position ‘strongly savours of the effects of demand and supply’ (1963, 189) 56 – rejected demand and supply in accounting for competitive long-run price, Mill insisted that in fact the very opposite was the case. All influences upon price operated by way of demand and supply, and the cost theory was dependent thereupon (1967, IV, 33–4). Mill’s insistence that there is no conflict between cost and demand– supply analyses reflects precisely Ricardo’s position: ‘Mr Malthus mistakes the question’, Ricardo protested of the formulation in Malthus’s Principles, ‘I do not say that the value of a commodity will always conform to its natural value without an additional supply, but I say that the cost of production regulates the supply and therefore regulates the price’ (1951, II, 48–9). J.S. Mill’s reaction to De Quincey is also pertinent. In his review of De Quincey’s Logic for the Westminster of June 1845, he observed that ‘the larger half of the volume is occupied with the theory of Value; which he rightly esteems the master key to the principal difficulties of the science’. But Mill denied that De Quincey’s observations on the relation between value-in-use and value-in-exchange had ‘all the originality which he ascribed to them’, for he had merely brought ‘into full theoretical explicitness 304

THE RECEPTION OF RICARDIAN ECONOMICS

what was known to all clear thinkers’ (1967, IV, 395–6). He also believed that De Quincey had not given due credit to received doctrine relating to prices diverging from cost: ‘Have not all political economists distinguished between articles which can be multiplied to an indefinite extent by labour, and articles naturally or artificially limited to a quantity short of demand, and have they not all, from Ricardo downwards, affirmed that in the former, and more common case, the value conforms on an average to the cost of production, while in the latter there are no limits to the value except the necessities or desires of the purchaser?’ (1967, IV, 398).57 But considering the work as a whole, Mill found little to object to and his recognition of De Quincey’s allegiance to Ricardo is clear, as indeed is his own position: ‘One of [De Quincey’s] merits is his early and consistent appreciation of Ricardo, the true founder of the abstract science of political economy, and whose writings are still, after all that has been written, its purest source’ (394). On matters of distribution in particular ‘De Quincey thoroughly understands his master, and is therefore able to supply new developments and illustrations of his master’s doctrines’ (401). This general evaluation by Mill seems quite accurate, as we have seen above. V. THE CONSISTENCY OF RICARDIAN AND ‘DISSENTING’ PROPOSITIONS That the Ricardians were able to see eye to eye with so much of the apparently critical work on value is not surprising. As already remarked Ricardo did not envisage his cost of production theory as an alternative to supply–demand analysis. 58 Conversely, it should be emphasized that the majority of ‘dissenters’ continued to accentuate the cost determination of price. This is true of Bailey himself and Longfield, both of whom spoke of production costs as the main consideration in price determination. It is true also of Cotterill who included in costs only wages and profits and like Bailey excluded rent. Read adopted a cost theory although he was inconsistent in emphasizing in some circumstances Smith’s conception, while elsewhere he accepted the opinion that ‘rent is the consequence, not the cause, of high price of raw produce’. Scrope, following Read, also adopted a theory of cost price which was an amalgam of that of Smith and Ricardo, and rejected the ‘vulgar opinion’ that rent affects price. In Longfield’s analysis of changes in relative prices the emphasis, as with Ricardo, was upon variations of the labour input, and here too was seen to lie the justification of a labour measure. 59 Much weight is sometimes placed on Whately’s criticism of the labour theory; 60 but to our mind, it would be unjustified to exaggerate its import, for his criticism explicitly limits the discussion to the case of given supplies. No labour theorist would have quarrelled with Whately’s observations since they do not relate to long-run equilibrium values in the case of infinitely elastic supplies, and are thus not inconsistent with a labour or cost explanation of 305

RICARDO - THE NEW VIEW

long-run value. Indeed Whately himself conceded that ‘valuable articles are, in almost all instances, obtained by Labour’. What, however, of Lloyd’s contribution to marginal utility or Longfield’s conception of intensity of demand? In this context the recent researches of Dr Marian Bowley are particularly pertinent. As she puts the matter ‘no revolutionary significance’ was attached to discussions of the law of diminishing marginal utility and related conceptions. Moreover, ‘these contributions did not affect the main classical conclusions as to the nature of market and natural prices and their determination’ (1972, 27). Her remarks are made with reference to attempts ‘to clear up particular problems’ generated in the Wealth of Nations, but in our view Ricardo would not have objected to the developments. While, of course, his main interests lay in longrun price determination, his economics required and, implicitly at least, hinged upon the operation of the competitive mechanism involving demand–supply analysis. His rejection of demand–supply theory did not apply to the particular version elaborated by Longfield, and Longfield himself appreciated Ricardo’s objections to the ‘indefinite’ and ‘vague’ expression ‘proportion between the demand and supply’ as unhelpful in the prediction of market price (1971 [1834], 247). It may also be remarked that Lloyd’s famous analysis of marginal utility is not inconsistent with a cost or even a labour theory, and was not so envisaged by Lloyd himself : ‘if labour becomes more effective, so that commodities of all kinds shall be produced in a degree of abundance greater in proportion to the wants of mankind, all sorts of commodities, though exchangeable in the same proportions as before for each other, could be said to have become less valuable’ (Lloyd 1834, 28). This statement is quite consistent with a cost or labour theory of exchange value. 61 To what extent may the conception of interest as a return to ‘abstinence’ developed by Scrope, Read and Longfield be interpreted as a sharp break with Ricardian procedures? To what extent would Ricardo have objected to an analysis of the precise nature of the savings supply function? The conception in Ricardo’s work of profits as residual is we believe little more than a formal consequence of the implicit presumption that the only contractual payment is that made to labour. There can be no doubt that Ricardo recognized the necessity of interest in the limiting case. More importantly, he took into account the effect of a declining profit rate on accumulation. It is true that he gave no name to the effect but it is by no means certain that he would have objected to the investigation of the time preference notion which the so-called ‘dissenters’ insisted upon. J.S. Mill, of course, found no difficulty in ascribing at one and the same time to the inverse wage–profit relationship and to the abstinence conception. 306

THE RECEPTION OF RICARDIAN ECONOMICS

What finally of the widespread application of market demand–supply analysis to long-run wage determination, as for example by Malthus, Longfield, Torrens, Read, Scrope and Senior? Here too there occurred no breakaway. The story would be a different one were it the case, as is apparently quite generally believed, that the subsistence wage played a key role in Ricardo’s work, not only in the context of his growth model but also in basic applications such as wage taxation. But this is far from an accurate perspective. Ricardo’s model was a growth model in the true sense with wages and profits above their respective minima which become relevant only in the stationary state; while the taxation theorems were applied by Ricardo to the case of above-subsistence wages. It is of great historical significance that Longfield did not actually attribute to Ricardo a subsistence theory of wages. On the contrary, he recognized Ricardo’s ascription to a model entailing secularly falling wages. 62 There can be no doubt that Longfield hit upon one of the most complex issues of Ricardian growth theory when he objected that, in principle, there is no reason why wages cannot bear the full brunt of secularly rising food prices since, in a context of ongoing population expansion, real (commodity) wages must initially exceed the ‘subsistence’ level (1971 [1834], 182f). Yet Ricardo did address himself to this issue and attempted to justify the secular decline in the rate of profit in circumstances of declining commodity wages, and Longfield failed to consider Ricardo’s case which turns upon the effect of a declining wage upon the population growth rate, and the presumed operation of a market process which assures that the population growth rate is kept in line with a secularly falling capital growth rate. 63 VI. THE REVIEW LITERATURE: A SECOND VIEW A widespread hostility to the Ricardian theory of value and distribution has been discerned by a recent investigator of the review literature of our period (Gordon 1969). This position, we believe, cannot be substantiated as far as concerns the four leading journals. In the first place twelve (supposedly) ‘hostile’ writers – of whom only seven (De Quincey, Ellis, Malthus, Merivale, Scrope, Senior and T.P. Thompson) can strictly speaking be described as economists – writing no more than sixteen articles over almost three decades constitutes too small a sample to permit any strong generalizations. In any complete evaluation it would certainly be necessary to allow for the contributions by such ‘pro-Ricardians’ as James and J.S. Mill and also the articles by McCulloch in the Edinburgh Review – and his influence on editorial policy – which ‘by the standards of the time, and also by the judgements of later economists’ had the reputation of being the leading economic journal (Fetter 1965, 427). Professor Fetter, writing of the Edinburgh Review during the period 1818–37, has 307

RICARDO - THE NEW VIEW

described McCulloch as ‘preaching the Ricardian doctrine with almost fanatical discipleship’ – a phenomenon which did much to bring about ‘the victory of Ricardian dogmatism’; over this period, ‘McCulloch’s virtual monopoly of economic articles enabled him to throw the influence of the Edinburgh back of the doctrines of the Ricardian system and also to keep out of the Review articles critical of the Ricardian system’ (Fetter 1953, 234, 238). Its readers would have had scarcely a hint of anti-Ricardian sentiment. 64 But even this takes for granted that we can accept the seven economists named above as hostile to Ricardian theory as far as concerns its key propositions. We have shown in this study that this would be a questionable position to take. Merivale and De Quincey positively supported Ricardo. The positions of Senior and Thompson on rent do not conflict with Ricardian principles – indeed Ricardo welcomed Senior’s article of 1821. And Malthus recognized the value of the inverse profit–wage relationship in his Quarterly Review contribution of 1824, a position consistent with various other statements. VII. THE POLITICAL ECONOMY CLUB: A SECOND VIEW Our investigation leads us to question the accuracy of J.L. Mallet’s famous account of the position of leading members of the Political Economy Club as conclusive evidence of a decline in Ricardo’s influence. 65 When we find that so much of the fundamental Ricardian position relating to the specific matters of the measure of value and the inverse profit–wage relationship – and not merely to general matters of method – left a strong positive impression on Longfield, and brought about a concession of error on the part of Torrens, and that even Bailey and several other ‘dissenters’ writing during the early 1830s recognized the logical validity of Ricardo’s basic theoretical structure, provided his use of terms is accepted, we are obliged to conclude that Mallet’s report fails to reflect the state of opinion during the postRicardian period as a whole even if it does provide an accurate account of the mood of particular Club meetings early in 1831. But even when we limit our attention to early 1831, and set aside the change in opinion on Torrens’ part under the influence of Longfield, there are reasons not to exaggerate the import of Mallet’s report. The account is ambiguous. If Torrens took a negative view of Ricardo’s stature, Mallet makes it clear that both Tooke and McCulloch considered that Rent was in point of fact the effect of differences in the productiveness of soils, because inferior soils were not brought into cultivation until the demand for food and the increased price enabled the cultivator to bring those soils into culture with a fair profit. McCulloch [unlike Tooke] stood up vigorously for Value as well as Rent, and paid very 308

THE RECEPTION OF RICARDIAN ECONOMICS

high compliments to Ricardo, whom he still considered as right in most points, and at all events as having done the greatest service to the science, his methodical and scientific way of treating it, so that even where he was mistaken, his errors could be detected by a subsequent and more accurate analysis. He was not an inventor, and no more was Bacon; and their merits in McCulloch’s opinion had very much the same character. 66 Now according to Mallet ‘This seemed to be generally assented to’. Similarly, at the second of the two meetings in question Mallet noted: ‘It was generally admitted that Ricardo is a bad and obscure writer, using the same terms in different senses; but that his principles are in the main right. Neither his Theories of Value nor his Theories of Rent and profits are correct, according to the very terms of his propositions; but they are right in principle.’ 67 Torrens (according to Mallet) referred both to Bailey and to Perronet Thompson as conclusively refuting Ricardo on Value and Rent respectively. Now we have already suggested that such an evaluation of Bailey’s work as far as concerns value is an exaggeration. As for the matter of rent, it is clearly the case that several of the socalled ‘dissenters’, in addition to Tooke and McCulloch who spoke favourably at the Club meetings, supported Ricardo’s position. We have in mind Cotterill and Longfield and also Malthus who, in correspondence during 1829 and 1831, rejected T.P. Thompson’s position that contemporary rent must be accounted for in terms of a monopoly theory rather than the principle of diminishing returns (de Marchi and Sturges 1973, 385, 386). Indeed it is of some interest that Malthus expressed displeasure at Torrens’ particular formulation of the question debated by the Political Economy Club: ‘I was hardly prepared to expect that in so short a time as has since elapsed [since his debates with Ricardo], one of the questions in the Political Economy Club should be “Whether any of the principles first advanced in Mr Ricardo’s work are now acknowledged to be correct?” My apprehension at present is that the tide is setting too strong against him.’ 68

VIII. CONCLUSION: A ‘DUAL’ DEVELOPMENT? Acceptance of the Ricardian theorem on distribution – the inverse profit– wage relationship – was far more widespread during the post-Ricardian period than is generally believed. But it is important to recognize that Ricardo obtained his results, paradoxical as it may appear, by a more consistent and rigorous use of the conception of relative value than appears in the Wealth of Nations or for that matter in Bailey’s work. For the demonstration of Ricardo’s main theme obliged him to pay considerable attention to inter-industry linkages. It is in fact impossible to do 309

RICARDO - THE NEW VIEW

justice to Ricardian economics without keeping to the fore the micro-economic or allocative foundations thereof; indeed it is helpful to envisage Ricardo as engaged in an effort to distinguish between a disturbance (such as a tax) affecting a single industry which will leave the general profit rate unaffected – this conclusion stands even if the industry in question happens to be agriculture – and a disturbance (such as a general wage increase) affecting all industries which will reduce the general profit rate.69 Here we note that even those sections on the Wealth of Nations most consciously preoccupied with resource allocation are severely constrained from this very perspective. For the average returns to the factors are not themselves determined in the adjustment process set in motion, for example, by demand disturbances; the movement of factors between industries acts upon product prices and the prices of factors in particular industries but the average returns remain unaffected. His procedures in formal analysis – as distinct from applied studies – imply identical factor proportions between sectors. Furthermore, Smith neglected the matter of variable proportions in any productive unit. In much of this Ricardo followed Smith. But he was much more conscious of the analytical problems created in the pricing process by the existence of differential factor proportions and was preoccupied by the relative dimension of price to a very marked degree. Similarly, he was in a large part of his work engaged in the rejection of Smith’s isolation of the pricing of corn for special treatment, insisting that corn must be treated in the same manner as every other product of the system. Ricardo thus corrected a variety of deficiencies in Smithian allocation economics along what were to be the ‘neoclassical’ lines. As far as concerns the matter of abstinence or recognition of a supply price for capital, it would appear that Ricardo as well as J.S. Mill, Senior, Read, Scrope and Longfield abandoned the standard eighteenth-century position according to which savings are institutionally determined. In this regard too Ricardo – unlike Smith – stands on the neoclassical (Marshallian) road. Most striking of all is Ricardo’s adherence to a supply–demand analysis of longrun wages. In this respect, and also in his analysis of wage taxation Ricardo was wholly dependent upon Smith. From all these perspectives – and bearing in mind the relationship between the dissenters and Ricardianism outlined in this paper – it appears unhelpful to think of a ‘dual development’ of economic analysis during the nineteenth century, involving on the one hand Smith, the dissenters, Mill and Marshall, and on the other Ricardo and Marx. We do not intend, of course, to suggest an identity of objective or of procedure between Ricardo and Marshall but the picture is far too complex to permit a neat and clearcut categorization such as that outlined at the outset. In conclusion we wish to consider a recent remark by Professor Frank Fetter to 310

THE RECEPTION OF RICARDIAN ECONOMICS

the effect that ‘if, as was the general impression by the 1830s, Ricardo’s theoretical framework had been largely rejected, there is a mystery as to why the name of Ricardo was so closely associated with the economic theory of John Stuart Mill and English neoclassical economics’ (1969, 80). A number of explanations for the phenomenon alluded to are offered, amongst which are the respect of Alfred Marshall for tradition, and Mill’s respect for the memory of Ricardo which left the impression that ‘even when the analysis was quite contrary to Ricardo’s ... the Ricardian framework remains almost unaltered. No one would suspect, from reading Mill’s Principles, the criticisms to which Ricardian ideas on rent, value, wages, and profits had been subjected in the previous twenty years’ (81). 70 But in the light of our own discussion it is not at all clear that there is a problem to resolve. For J.S. Mill did not ignore the criticisms of the preceding quarter century; the point to keep in mind is that he did not believe these criticisms to be destructive of the main Ricardian theoretical structure. His position is very clearly formulated in his review of De Quincey’s Logic in which he comments on De Quincey’s belief that there had occurred little advance since the 1817 ‘revolution’: We dissent from the opinion that political economy does not advance. We think it is in a state of most rapid progression. But ... the superstructure seems to be overgrowing the foundation. The science is growing at the extremities, without a proportional and suitable enlargement of the main trunk. Many important new views – new, at least, in having been previously overlooked – have dawned upon political economists during the last twenty years. But for want of sufficiently careful habits of systematic thought, these new views have been too frequently promulgated as contradictions of the doctrines previously received as fundamental; instead of being, as they almost always are, developments of them; corollaries flowing from these fundamental principles, certain conditions of fact being supposed ... What has been added to the science since Ricardo, does not need to be substituted for his doctrines, but to be incorporated with them. They do not require alteration or correction, as much as fuller exposition and comment. (Mill 1967, IV, 394) Our analysis of the monograph and periodical literature suggests that J.S. Mill was quite justified in taking this view. Many of the criticisms of Ricardo turn out to be quite in accord with Ricardian theory and Ricardo himself, to the extent that he was familiar with the critical literature, took a view quite consistent with that of Mill.

311

RICARDO - THE NEW VIEW

NOTES *I am indebted to Walter Elter and Lawrence Moss for contructive criticism. 1. Dobb 1973, 36; ‘income-distribution (e.g. the profit–wage ratio) was a precondition of the formation of relative prices’. (Also Dobb 1973, 169, 261, 266.) And see p. 116: Given the wage rate ‘the conditions of production in the industry or industries producing necessities for wage-earners played a key role in determining the ratio of profits or surplus to wages, and hence (given necessary labour-expenditures in various lines of production) relative exchange values’. 2. See also Hutchison 1952, 428–9: ‘nearly all the great economists of the second quarter of the nineteenth century, especially those regarded as significant and original today, abandoned the main Ricardian questions and the main Ricardian answers (with the very doubtful exception of J.S. Mill). The obvious examples are Bailey, Longfield, the Oxford group of Whately, Senior and Lloyd and also Richard Jones ... to take only the main English names.’ 3. Marx 1965, 14–15 (Afterword to the second German edition, 1873, of Capital, vol. 1); Marx’s rough draft notes (written 1857) Grundrisse (1973, 883). (See also Meek 1967, 54.) T.W. Hutchison (1974, 14) has observed that by adopting this position Marx placed all his ‘bourgeois’ contemporaries ‘beyond the pale of serious scientific discussion’ – thus representing the development of economics ‘as a process uniquely culminating in, and consummated by, the writings of Marx himself’. 4. ‘From Marshall’s Principles, Ricardianism can be removed without being missed at all. From Mill’s Principles it could be dropped without being missed very greatly’ (Schumpeter 1954, 629). 5. But see the position of Pedro Schwartz (1972, 16–17) which places Mill more firmly in the Ricardian tradition at least as far as concerns analysis: ‘The sway of Ricardian doctrine over British economic thought was in some way restored with the publication of John Stuart Mill’s Principles in 1848. From an analytical point of view, Mill’s treatise followed the pattern of the master, while the doctrinal changes that Mill introduced made it easier to accept the Ricardian heritage.’ 6. The treatment of the invariable measure of value, which is said to be ‘central to Ricardo’s system’, we are told, ‘never interested McCulloch at all’ (O’Brien 1970, 146). 7. Insofar as concerns the theory of value the essential point to keep in mind is its function as a necessary preliminary for the analysis of distribution, particularly the analysis of profit-rate determination and the effect of wage-rate changes upon the rate of profit. Ricardo was preoccupied with the problem of defining the minimum conditions required of a medium of exchange – at least in principle – to assure approximate constancy in the value of the output to be shared between the recipients of income net of rent in the face of changes in distribution. Cf. editor’s introduction to the Principles, Ricardo 1951, I, xlviii. For accounts of the Ricardian procedure see also Stigler 1965, 156–97; and Blaug 1958, esp. 23–5. 8. The weakness of the Smithian position came to the fore partly in consequence of preoccupation with the great monetary issues during the early years of the century. And it should be remarked that, despite the attention paid to the problem of constructing a suitable measure of value, the conclusion that wagerate increases are non-inflationary was maintained quite generally and applied to the ‘real world’ where the conditions required of the theoretical measure are not fulfilled . 312

THE RECEPTION OF RICARDIAN ECONOMICS

9. Professor Jacob Viner observed many years ago that any investigation of the extent and nature of the dissension requires that attention be paid to the replies of the Ricardians to their critics, for ‘some of the criticisms directed against the Ricardian analysis were either based on misinterpretations of it or would have been accepted by the Ricardians’ (1958, 419–20). 10. Mark Blaug has also argued the case for the resilience of ‘Ricardian’ economics using the term, however, in this context in a sense different from ours. Blaug has in mind ‘the proposition that the yield of wheat per acre of land governs the general return on invested capital as well as the secular changes in the distributive shares’ (1958, 3). See also p. 61: ‘The rate of capital formation was still held to be governed by returns in agriculture, and the core of the Ricardian system, the law of diminishing returns, continued to dominate the body of economic thought.’ 11. Ricardo’s emphasis upon the role of demand–supply analysis in both product and factor markets is substantiated more fully in my Economics of David Ricardo [1979]. See also Hicks and Hollander 1977. 12. See also Malthus 1964 [1836], 260–1. 13. See also Malthus 1964 [1836], 88. 14. However, Bailey continues rather less generously: But then, although some inconsistencies would by this means be obviated or explained away, we should obtain in their place a number of others equally irreconcilable, and also a series of unmeaning and identical propositions. For instance, the proposition that a million of men always produced the same value, but not the same riches, would be reduced to this, that what a million of men produced always cost the labour of a million men: a = a. The truth appears to be, that the idea of real value was seldom distinctly present to his mind, although there was almost constantly an obscure reference to it.

15. It should be carefully noted that Bailey retains the Ricardian usage of profits as a share in aggregate output while at the same time he utilizes the conception of commodity wages, concluding, not surprisingly, that in the case of constant output an increase in the commodity wages paid to a given work force must involve a decline in the profit share (Bailey’s ‘rate of profit’) – while in the case of an increasing output due to technical progress an increase in the commodity wages need not have this effect (Bailey 1825, 64f). 16. He asserted (Bailey 1825, xii) that ‘the science cannot yet be exhibited in a regular and perfect structure’. 17. His concession was, however, made rather grudgingly. In this context, Bailey praised De Quincey’s analysis of the severe problems relating to qualitative differences in labour and the use of the wage scale to obtain units of standard labour (Bailey 1825, 214f). Bailey recognized Ricardo’s concession that a variation in wages will influence relative prices in the event of differing factor proportions but asked: ‘Why persist in calling quantity of labour the sole determining principle of value?’ (217). 18. He might have included Say, in principle, who thought of the invariable standard as a ‘chimera’. 19. Here Longfield insisted that differential land fertility was not per se the ‘cause’ of rent, since ‘whatever is useful, and is limited in quantity is capable of possessing value, if it can be made the subject of exchange’ (1971 [1834], 134). Ricardo would have accepted this proposition. 20. Longfield uses the term ‘depends upon’ but he obviously intended an identity relation. 21. Despite his adherence to the main propositions of the Ricardian position, Longfield simultaneously ascribed in the same lecture (VIII) to Smith’s view that an increase in 313

RICARDO - THE NEW VIEW

22.

23.

24. 25.

26.

27. 28. 29.

30.

wages will generate a general rise in the level of prices. The latter argument was not, however, soundly based upon a specific theory of value. And since he ends the lecture with a defence of Ricardo’s inverse profit–wage relationship we conclude that in adopting the Smithian position he fell into serious error from the perspective of his own argument. Butt’s combination of the utility teaching of Say and Senior with aspects of Longfield’s theory of distribution to arrive at an early statement of the marginal utility theory of imputation seems to be one of the most impressive outcomes but it is only in this century that Butt’s work is becoming generally known. See Moss 1973. For a full discussion of this matter see Lionel Robbins 1958, 53f. This evidence has largely been overlooked; cf. Black 1971, 26; Blaug 1958, 127, 159. Lord Robbins conjectures that Ricardo would probably not have welcomed the Longfield–Torrens defence insofar as it was based upon a theory of value depending on quantity of capital, or accumulated labour, which Ricardo had turned down in earlier debate with Torrens (1958, 57). Read 1829, 249: ‘The more the inferior labourer gets as his wages, the less will remain to the superior labourer as his’. ‘Having persuaded themselves of the truth of these false conclusions, our economists go on, in the most self-satisfied way, to draw from them several corollaries, such as, “It is abundantly certain, therefore, that no rise of wages will ever occasion a general rise of prices, and no fall of wages a general fall of prices” ( McCulloch, ...). Why certainly, when wages are defined to be merely the labourer’s share, as compared to that of the capitalist, their fall or rise can have no influence on the value of the joint return. ... We need no professor of political economy to announce, as a recondite proposition, what is identical with their own postulate’ (Scrope 1831, 3). It is an increase in wages expressed in money of unchanged general purchasing power – either because of constant real costs of production of the metal or because of a constant supply of money – which cannot be accompanied by a general price rise. Whately does not, however, limit the investigation to material objects. It is certain that opponents of Ricardianism writing from an inductivist or historical point of view, would have read Whately’s work as a defence of Ricardian procedures. Despite his assertion that ‘the decay of the Ricardian School must have become patent’ soon after 1826 Schumpeter (1954, 478) hedged his account with multifold qualifications: the influence of Bailey ‘was much greater than appears on the surface’; ‘though but few contemporaries did justice to him, it became clear in time that he had turned the tide and dealt a fatal blow’ (599). Similarly, while Bailey’s critique did not ‘pass unnoticed’, and while ‘it is safe to presume that his influence extended beyond the range of explicit recognition’, and although ‘a poll of writers on value from 1826 to 1845 would produce a considerable majority for Bailey’, nevertheless there occurred no ‘spectacular victory’, Bailey’s ‘defeat’ (our emphasis) being accounted for on two grounds, that his challenge was ‘premature’ and that Bailey failed to show how Ricardo’s system might be replaced. Thus Bailey and those who followed his example helped ‘undermine’ the system ‘but they did so by a slow process of attrition’ (487). The same kind of criticism is directed by Read against Lauderdale’s Inquiry of 1804. Any attempt ‘to reduce the idea of value to a mere relation of commodities between themselves, without any connexion with mankind, with labour, or with cost of production’ is fatal, since ‘the connexion of the exchangeable value of commodities with labour and cost of 314

THE RECEPTION OF RICARDIAN ECONOMICS

31.

32.

33.

34. 35.

36. 37. 38. 39.

40.

41. 42.

production is indeed the only circumstance which confers any importance on the connexion of the exchangeable value of commodities between themselves’ (221). In his chapter dealing with the causes of value Bailey (1825, 185) had distinguished between ‘commodities which are monopolized, or protected from competition by natural or adventitious circumstances’; those ‘in the production of which some persons possess greater facilities than the rest of the community, and which therefore the competition of the latter cannot increase, except at a greater cost’; and those ‘in the production of which competition operates without restraint’. The general notion is that ‘their respective causes of value cannot be the same’. Thus if Cotterill was critical of Ricardo in the context of the cause of value, he was equally dissatisfied with the treatment of Bailey whose explanation of longrun exchange value in terms of cost of production was ‘defective’: ‘it does not include profit, and therefore never can determine value, because, though the labour or capital should not vary, the profit may vary, and consequently the value of any commodities’. On similar grounds Cotterill objected also to Torrens (1831, 46). The citation is from J.S. Mill’s (unsigned) review of Malthus’s contribution of 1824, ‘Periodical Literature – Quarterly Review’. Cf. Westminster Review III, 5, January 1825. Mill by his assertion intended to deny Malthus’s conception of aggregate demand, not, as is often maintained, to challenge the Ricardian position regarding absolute value. Lloyd, however, took the statement out of context and applied it generally. Bailey himself (1826, 37) applauded the sentiment of the article but expressed surprise at finding an insistence upon value as a relative conception in the Westminster Review, which in 1826 had published an attack – which we know to be by James Mill – upon his own Critical Dissertation. If the respective contexts are borne in mind the apparent inconsistency of position between the journal volumes creates no difficulty. See below, pp. 301–2, regarding this pamphlet. Cf. Schwartz 1972, 15: ‘After reading Dr O’Brien, one might even say that in a short time he [James Mill] was the only one left in the Ricardian succession – together with John Stuart Mill.’ Principles of Political Economy, edn 1, 1825 (London, 1872), 119; edn 5, 1864 (New York, 1965), 243. See also Blaug 1958, 172–3. On Mill see my ‘Ricardianism, J.S. Mill and the Neo-Classical Challenge’ (1976). Groenewegen 1973, 193, refers to De Quincey’s remarks on the mutual determination of exchange value by ‘intrinsic utility’ and ‘difficulty of attainment’ as a ‘change of heart in favour of the Smith tradition in value theory’. De Quincey wrote a three-part essay for Blackwood’s in 1842 entitled ‘What is the Radical Difference between Ricardo and Adam Smith? With an occasional Note of Ricardo’s Oversights’. This essay was recast and expanded as The Logic of Political Economy (1844). A view attributed to ‘almost every journal in the land’ which he knew (De Quincey 1970c [1844], 121). De Quincey also referred to his early work Dialogues of Three Templars on Political Economy, which originally had appeared in three successive numbers of the London Magazine in 1824 (1970a, 37) and which had been considered with some respect by Bailey in his Critical Dissertation. The object of this early work, De Quincey observed (1970c, 119), had been ‘to draw into much stronger relief than Ricardo himself had done that one radical 315

RICARDO - THE NEW VIEW

43.

44.

45.

46.

47.

48.

49.

doctrine as to value by which he had given a new birth to Political Economy’. And while he has been pleased with the attention given to the work by Bailey, nevertheless ‘with all his ability, that writer failed to shake any of my opinions. I continue to hold my original ideas on the various aspects of this embarrassing doctrine.’ In the Logic reference is also made to an earlier demonstration in the ‘Templars’ Dialogues’ that the concept of a stable measure of value was ‘demonstrably a chimera, an ens rationis, which never could be realized’ (De Quincey 1970c [1844], 151–2). In the ‘Dialogues’ Ricardo himself is cited to the same effect, the third edition of his chapter on value ‘having for its direct object to expose the impossibility of any true measure of value’ (93). Yet Ricardo used the notion of a constant measure as a conceptual device, and De Quincey followed the same procedure: Ricardo’s practice was to ‘neglect the variations in the value of money ... he made it understood, that ... he would always assume money as ranging at its stationary natural value’ (201–2). De Quincey emphasized the fundamentally important feature of Ricardian economics that ‘a change cannot commence in profits; that function of industry is not liable to any original affection of change’ (1970c [1844], 277); see also 258– 9: ‘Any change that can disturb the existing relations between wages and profits must originate in wages: Whatever change may silently take place in profits always we must view as recording and measuring a previous change in wages ... Ricardo’s chapter upon profits is substantially no more than a reiteration of his two chapters upon wages and rent.’ But De Quincey at the same time criticized Ricardo for playing down ‘that eternal counter-movement which tends, by an equivalent agency, to redress the disturbing balance’ (1970c [1844], 249). Accordingly, he denied that one could speak of an ‘immovable law of declension’ of the profit rate (294). Yet oddly enough he also conceded that technical progress ‘may be altogether torpid’ while diminishing returns is a phenomenon which ‘is sometimes for a century together proceeding with activity’ (259). But it is doubtful whether De Quincey’s charge can be sustained; see Hollander 1977. Nassau W. Senior, ‘Report on the State of Agriculture’, Quarterly Review XXV, 50, July 1821; T.P. Thompson (1832) The True Theory of Rent in Opposition to Mr Ricardo and Others, 9th edn, London. ‘On the Nature, Measures, and Causes of Value’, Westminster Review V, 9, January 1826: 157. (Reprinted with the London, 1967 (Cass) edition of the Critical Dissertation. ) For the attribution of the review of James Mill, see Fetter 1962, 584. The authorship of Observations on Certain Verbal Disputes in Political Economy (1821) has not yet been ascertained. The Cass edition of Bailey’s works (London, 1967) includes the pamphlet, but see the note by Ken Dennis (1973, 17–18) on ‘The Bailey Notebooks and Authorship of “Verbal Disputes” ’. Marx (1971, 125) commented on the similarity between the works in question. Thompson himself (1832, 12) raised the question whether there was, in fact, any substantive difference between his approach and that of the Ricardians, and gave a rather unconvincing answer to the affirmative : If it is urged that both the present writer and Mr. Ricardo from whom his theory is derived, have distinctly represented the cultivation of the inferior soils as flowing from the appropriation of land and the limitation of its quantity, and that to affirm that rent proceeds from the secondary cause is tantamount to affirming that it proceeds from the first; – the answer is, that it is one thing to affirm that the competition is the simple cause of rent in all cases, and would equally cause it if no difference in the qualities of soils and

316

THE RECEPTION OF RICARDIAN ECONOMICS no knowledge of the art of forcing crops were in existence, – and another thing to affirm, that it causes rent through the intervention of these circumstances, and would not cause rent without them.

50.

51.

52. 53. 54. 55.

56.

57.

58.

59.

Thompson, it may be added, commended Ricardo for his proposition that ‘rent will be paid because corn sells high; and not corn sell high because rent is paid’, which ordered the causal sequence correctly (8). Ricardo 1951, I, 69. Conversely, ‘[if] air, water, the elasticity of steam; and the pressure of the atmosphere, were of various qualities; if they could be appropriated, and each quality existed only in moderate abundance, they, as well as the land, would afford a rent, as the successive qualities were brought into use’ (75). See J.S. Mill 1965, II, 428. Ricardo on various occasions explained in great detail the process of output expansion in response to rising demand. Within his structure, permanent constraint on such expansion must imply an increase in rents to prevent the profit rate in agriculture remaining permanently in excess of that in manufactures. See McVickar 1966 [1825], 135. See also Ricardo 1951, II, 285, 446. The reference is to the Grounds of an Opinion, 1815. Malthus’s emphasis upon the disturbance to the price structure of a wage-rate is reiterated, in Ricardo 1951, II, 285–6. Thus he wrote to McCulloch (9 October 1822; IX, 279): ‘You go a little further than I go in estimating the value of commodities by the quantity of labour required to produce them: you appear to admit of no exceptions or qualification whatever, whereas I am always willing to allow that some of the variations in the relative value of commodities may be referred to causes distinct from the quantity of labour necessary to produce them.’ Here he had in mind the effects of changing wages: ‘To this second cause I do not attach near so much importance as Mr Malthus and others but I cannot wholly shut my eyes to it.’ To Malthus himself, he insisted that their difference was only one of degree, that is to say, that the effects of a variation in wages were limited: ‘no great effects can follow because profits themselves constitute but a small portion of price, and no great addition, or deduction can be made on their account’ (9 October 1820; VIII, 279). Malthus had in mind Smith’s notion of the ‘natural prices’ of commodities determined by the natural rate of wages and profits or ‘the ordinary or average rate which is found in every society or neighbourhood, and which is regulated ... by the general circumstances of the society, their riches or poverty, their advancing, stationary, or declining conditions’. De Quincey actually recognized the distinction in question but believed that the literature treated the matter at a superficial level (1970c, 155). De Quincey was much concerned with bilateral monopoly while J.S. Mill operated in terms of a competitive structure and wished to fit De Quincey’s case into demand–supply analysis (1967, IV, 399f). Extraordinary as it may seem Read went so far as to charge the ‘Ricardo School’ for following Lauderdale’s lead in emphasizing utility and scarcity or supply– demand rather than cost analysis (1829, 220–1). Several of the dissenters accorded labour a special role in their definitions of wealth. Thus Read defined wealth as ‘those external material objects, necessary, useful, or agreeable to mankind which it costs some considerable exertion of human labour or industry to produce or acquire originally, and which once acquired can be transferred from one to another – appropriated or alienated’. Further, he concluded that ‘as the existence of all value in exchange is wholly dependent on the necessity of some portion of human labour 317

RICARDO - THE NEW VIEW

60. 61.

62.

63.

64.

65.

being employed in the production of those commodities in which value in exchange forms an attribute, so labour is the only certain measure of that value’. Longfield, for his part, observed that ‘as most of the commodities in which the wealth of the country consists are produced by labour, political-economists make use of it as a measure of value’, and also that ‘all the commodities that men consume, and which can be made the subject of exchanges, owe their existence and their value to labour. The exceptions to this are very trifling, and are of such a nature that they do not vitiate any of the conclusions drawn from it.’ Whately, for his part, asserted that ‘valuable articles are in almost all instances, obtained by labour’. And while Scrope spoke harshly against the notion of labour as the ‘only source of wealth’, he nonetheless chose to describe capital as the ‘result of previous labour’. Whately 1832, 252–3: ‘It is not that pearls fetch a high price because men have dived for them; but on the contrary, men dive for them because they fetch a high price.’ In a recent study of Lloyd, Romano (1977) observes that Lloyd adopted an approach to rent as residual which grows in the course of secular expansion; however, he ascribed to the ‘scarcity’ theory of rent developed by Thompson and Senior according to which differential productivity is not a necessary condition. He made only scattered references to profits but suggests that profit rates had fallen due to increased ‘competition of capitals’ – the Smithian position. But the charge of inconsistency on Ricardo’s part – having maintained simultaneously the conception of declining secular wages and the proposition that custom and habit assured constant wages – is quite unjustified (Longfield 1971 [1834], 205). There is nothing in the orthodox position which requires secular wages to be a constant in a growing economy. Torrens, quite independently, formulated in his fifth (1829) edition of the Essay on the External Corn Trade (London: Longman, Rees, Orme, Brown, and Green, 1829), 468–9, a statement of the theory of development with special reference to the course of wages and profits, which reflects, we believe, Ricardo’s own position on the matter. See also On Wages and Combination (London: Longman, Rees, Orme, Brown, Green, and Longman, 1834), 21–2 for the same passage. Actually very little appeared on the specific issue of value and distribution. But one such article by Empson (‘Life, Writings, and Character of Mr. Malthus’, Edinburgh Review 130, January 1837) was very well disposed towards Ricardo. Empson refers to Ricardo’s ‘masterly criticism’, in his last letters, of Malthus’s pamphlet of 1823; and observed (correctly) that the influence of Ricardo remained strong with Malthus (1837, 470–2). The proceedings of the Political Economy Club, particularly those of 13 January 1831, are frequently cited as evidence of widespread hostility amongst economists towards Ricardian economics. The debates in question turned on the question put by Torrens: ‘What improvements have been effected in the science of Political Economy since the publication of Mr Ricardo’s great work; and are any of the principles first advanced in that work now acknowledged to be correct?’ ( Centenary Volume of the Political Economy Club (London: Political Economy Club, 1921, 35). According to J.L. Mallet’s account it was Torrens’s opinion that all the great principles of Ricardo’s work had been successively abandoned, and that his theories of Value, Rent and Profits were now generally acknowledged to have been erroneous. As to value the dissertation on the Measure of value published in 1825 by Mr.

318

THE RECEPTION OF RICARDIAN ECONOMICS Baillie [Bailey] of Leeds has settled that question. As Thompson had shown that Rent was not the effect of differences in the relative productiveness of soils, but the effect of demand and price, and as to profits, it is clear that the part that goes to replacing the capital employed, which Mr Ricardo had omitted to take into the account, was decisive of the unsoundness of his views (223–4).

66. Political Economy Club Centenary Volume, 224. (Cf. also McCulloch’s ‘boldly standing by Ricardo’s doctrine, that equal quantities of labour are equal in value all over the world’ – in opposition to Torrens and Malthus.) However, both McCulloch and Tooke are reported as accepting the criticism that Ricardo’s theory of profit was vitiated by the failure to allow for the replacement of fixed capital; while Tooke took exception to Ricardian value theory. 67. Political Economy Club Centenary Volume, 225 (emphasis added). Mallet himself took a more reserved view. 68. Letter to Whewell, 31 May 1831, in de Marchi and Sturges 1973, 391. 69. Despite Schumpeter’s strong ‘indictment’ of Ricardo on the grounds of a failure to appreciate the production–distribution process as a ‘web of exchanges’, on occasion he was less severe. (See Schumpeter 1954, 568, 690.) 70. See also Schwartz 1972, 236–7.

REFERENCES Bailey, S. (1825) A Critical Dissertation on the Nature, Measure and Causes of Value, London: R. Hunter ——(1826) A Letter to a Political Economist on the Subject of Value, London: R. Hunter. Black, R.D.C. (1971) ‘Introduction’, The Economic Writings of Mountifort Longfield, New York: Augustus M. Kelley. Blaug, Mark (1958) Ricardian Economics, New Haven, Conn.: Yale University Press. Bowley, Marian (1972) ‘The Predecessors of Jevons: The Revolution that Wasn’t’, The Manchester School of Economic and Social Studies XL (March): 9–29. Studies in the History of Economic Theory before 1870, London: Macmillan. ——(1973) Cotterill, C.F. (1831) An Examination of the Doctrines of Value, London: Simpkin Marshall. de Marchi, Neil and Sturges, R.P. (1973) ‘Malthus and Ricardo’s Inductivist Critics’, Economica XL (November): 379–93. Dennis, Ken (1973) ‘The Bailey Notebooks and Authorship of “Verbal Disputes”, History of Economic Thought Newsletter No.11 (Autumn): 17–18. De Quincey, Thomas (1970a) [1824] Dialogues of Three Templars on Political Economy, in D. Masson (ed.) Political Economy and Politics, Vol. X, Collected Works of Thomas de Quincey, [London, 1897] New York: Augustus M. Kelley, 37–112. ——(1970b) [1842] ‘Ricardo and Adam Smith’,Blackwood’s Edinburgh Magazine LII, 323 (September), in D. Masson (ed.) Political Economy and Politics, 113–17. ——(1970c) [1844] The Logic of Political Economy, in D. Masson (ed.) Political Economy and Politics 118–294. Dobb, Maurice (1973) Theories of Value and Distribution since Adam Smith, Cambridge: Cambridge University Press. 319

RICARDO - THE NEW VIEW

Fetter, Frank W. (1953) ‘The Authorship of Economic Articles in the Edinburgh Review, 1802–1847’, Journal of Political Economy LXI (June): 232–59. ——(1962) ‘Economic Articles in theWestminster Review and their Authors, 1824– 51’, Journal of Political Economy LXX (December): 570–96. ——(1965) ‘Economic Controversy in the British Reviews, 1802–50’,Economica XXXII: 424–37. ——(1969) ‘The Rise and Decline of Ricardian Economics’,History of Political Economy I (Spring): 67–84. Gordon, Barry (1969) ‘Criticism of Ricardian Views on Value and Distribution in the British Periodicals, 1820–1850’, History of Political Economy I (Fall): 370–87. Grampp, W.D. (1974) ‘Malthus and his Contemporaries’, History of Political Economy VI (Fall): 278–304. Groenewegen, P.W. (1973) ‘Review of DobbTheories of Value...’, Economic Journal LXXXIV (March): 192–3. Hicks, John and S. Hollander (1977) ‘Mr. Ricardo and the Moderns’, Quarterly Journal of Economics 91,3: 351–69. Hollander, S. (1976) ‘Ricardianism, J.S. Mill, and the Neo-classical Challenge’ in J.M. Robson and M. Laine (eds) James and John Stuart Mill: Papers on the Centenary Conference, Toronto: University of Toronto Press. ——(1977) ‘Ricardo and the Corn Laws: A Revision’,History of Political Economy IX (Spring): 1–47. Hutchison, T.W. (1952) ‘Some Questions about Ricardo’, Economica XIX (November): 415–32. ——(1974) ‘The Cambridge Version of the History of Economics’,University of Birmingham Occasional Paper No. 19 (January). Knight, F.H. (1956) On the History and Method of Economics, Chicago: University of Chicago Press. Lloyd, W.F. (1834) A Lecture on the Notion of Value, London: Roake and Varby. Longfield, Mountifort (1971) [1834] Lectures on Political Economy (Dublin), in The Economic Writings of Mountifort Longfield, New York: Augustus M. Kelley. McCulloch, J.R. (1872) [1825] The Principles of Political Economy, 1st edn, London: A Murray. ——(1965) [1864]The Principles of Political Economy, 5th edn, New York: Augustus M. Kelley. McVickar, John (1966) [1825] Outlines of Political Economy, New York: Augustus M. Kelley. Malthus, T.R. (1957) [1823] The Measure of Value Stated and Illustrated, New York: Kelley and Millman. ——(1963) [1824] ‘Political Economy’,The Quarterly Review 30,60: 297–334, in Bernard Semmel (ed.) Occasional Papers of T.R. Malthus, New York: Burt Franklin, 71–104. Principles of Political Economy, 2nd (posthumous) edn, New York: Augustus ——(1964) [1836] M. Kelley. Marx, Karl (1965) [1863] ‘Afterword’ to the 2nd German edition (1873) of Capital, I, Moscow: Progress Publishers. 320

THE RECEPTION OF RICARDIAN ECONOMICS

——(1971) Theories of Surplus Value, III, Moscow: Progress Publishers. Grundrisse: Foundations of the Critique of Political Economy, London: Penguin. ——(1973) [1857] Meek, R.L. (1967) Economics and Ideology and Other Essays, London: Chapman and Hall. Merivale, H. (1837) ‘Senior on Political Economy’, Edinburgh Review 133 (October). ——(1840) ‘McCulloch’s Edition of The Wealth of Nations’, Edinburgh Review 142 (January). Mill, J.S. (1965) Principles of Political Economy, Collected Works, II, Toronto: University of Toronto Press. ——(1967) ‘The Nature, Origin, and Progress of Rent’,Collected Works, IV, Toronto: University of Toronto Press, 161–80. Moss, Lawrence (1973) ‘Isaac Butt and the Early Development of the Marginal Utility Theory of Imputation’, History of Political Economy V (Fall): 317–38. ——(1974) ‘Mountifort Longfield’s Supply-and-Demand Theory’,History of Political Economy VI (Winter): 405–34. O’Brien, D.P. (1970) J.R. McCulloch: A Study in Classical Economics, London: George Allen and Unwin. Pasinetti, L.L. (1974) Growth and Income Distribution, Cambridge: Cambridge University Press. Read, Samuel (1829) Political Economy: An Inquiry into the Natural Grounds of Right to Vendible Property or Wealth, Edinburgh: Oliver and Boyd. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Vol. I: Principles of Political Economy, Cambridge: Cambridge University Press. London: Robbins, Lionel (1958) Robert Torrens and the Evolution of Classical Economics, Macmillan. Romano, R.M. (1977) ‘William Forster Lloyd – A Non-Ricardian?’ History of Political Economy 9: 412–41. Schumpeter, J.A. (1954) History of Economic Analysis, New York: Oxford University Press. Schwartz, Pedro (1972) The New Political Economy of John Stuart Mill, London: London School of Economics. Scrope, G. Poulett (1831) ‘The Political Economists’, Quarterly Review XLIV, 87 (January): 1–52. Seligman, E.R.A. (1903) ‘On Some Neglected British Economists, II’, Economic Journal XIII (December): 511–35. Senior, Nassau W. (1821) ‘Report on the State of Agriculture’, Quarterly Review XXV, 50 (July). Stigler, G.J. (1965) Essays in the History of Economics, Chicago: University of Chicago Press. 9th Thompson, T.P. (1832) The True Theory of Rent in Opposition to Mr Ricardo and Others, edn, London: R. Heward. Torrens, Robert (1826) An Essay on the External Corn Trade, 3rd edn, London: Rees, Orme and Brown. Torrens, Robert (1835) Colonization of Southern Australia, London: Longman, Rees, Orme, Brown, Green, and Longman.

321

RICARDO - THE NEW VIEW

——(1844) The Budget: On Commercial and Colonial Policy, London: Smith, Elder. Twiss, Travers (1847) View of the Progress of Political Economy in Europe since the Sixteenth Century, London: Longman, Brown, Green and Longman. Viner, Jacob (1958) The Long View and The Short, Glencoe, Ill.: Free Press. Whately, R. (1832) Introductory Lectures on Political Economy, 2nd edn, London: B. Fellowes .

322

19 THE ROLE OF BENTHAM IN THE EARLY DEVELOPMENT OF RICARDIAN THEORY: A SPECULATIVE ESSAY

This paper constitutes a speculative discussion of Ricardo’s possible ‘indebtedness’ to Jeremy Bentham during the early course of development of his theoretical system. A precondition for good scholarship is, of course, the phrasing of worthwhile questions. There are, in fact, two ‘mysteries’ which set the stage for our investigation. The first relates to Ricardo’s breakaway from Smithian principles – the abandonment of the famous conception ‘competition of capitals’ in the context of profit-rate theory, and more particularly the introduction of agricultural productivity as a key variable in that context. The second turns upon a lost Bentham manuscript, apparently once in Ricardo’s possession, dealing with the effects on profits of cultivating successive qualities of land. A little elucidation is required regarding these matters. In his pamphlet of 1810–11 Ricardo appears as a full-fledged Smithian on most theoretical matters. Thus, for example, in the High Price of Bullion he insisted that ‘profits can only be lowered by a competition of capitals not consisting of circulating medium’ (Ricardo 1951, III, 92). But a concern with the general profit rate is not conspicuous at this stage and the matter is scarcely raised between Ricardo and Malthus in their correspondence of 1811. Indeed a preoccupation with currency and the exchanges characterizes the extant Ricardo correspondence until March 1813. The first indications of a specific concern with the theory of profits, and one which implies a breakaway from Smith (partial only at this stage) are contained in two letters to Malthus dated 10 and 17 August 1813. 1 In the first of the two letters in question Ricardo objected to Malthus’s contention (made in correspondence and perhaps also in conversation) that contemporary extensions of foreign trade could be taken as evidence of a rising rate of profit: ‘On further reflection I am confirmed in the opinion which I gave with regard to the effect of opening new markets or extending the old. I most readily allow that since 323

RICARDO - THE NEW VIEW

the war, not only the nominal but the real value of our exports and imports has increased – but I do not see how this admission will favour the view which you take of this subject ... [Extension of trade] does not prove a general increase of profits nor any material growth of prosperity’ (VI, 93). 2 In the second letter Ricardo explicitly stated the position that attention should be directed at agricultural productivity to understand trends in the rate of profit: That we have experienced a great increase of wealth and prosperity since the commencement of the war, I am amongst the foremost to believe; but it is not certain that such increase must have been attended by increased profits, or rather an increased rate of profits, for that is the question between us. I have little doubt however that for a long period, during the interval you mention [1793–1813], there has been an increased rate of profits, but it has been accompanied with such decided improvements of agriculture both here and abroad, – for the French revolution was exceedingly favorable to the increased production of food, that it is perfectly reconcileable to my theory. My conclusion is that there has been a rapid increase of Capital which has been prevented from shewing itself in a low rate of interest by new facilities in the production of food. (Ricardo 1951, VI, 94–5; emphasis added) Quite clearly, Ricardo presumed that but for improved agricultural technology during the Revolutionary and Napoleonic periods the profit rate would have declined with accumulation. Each of the extracts given above is followed by statements insisting that an increase in general prices occurs only in consequence of a rising money supply (or velocity), while only a falling level of prices can induce a monetary inflow. 3 Although we cannot be positive about the matter it is most likely that the statements alluding to the profit rate and those relating to the monetary mechanism are connected rather than independent. If this is so, Ricardo appears to be objecting to a formulation by Malthus very similar to that which was to appear later in the latter’s Principles of 1820 – namely that an increase in the value of aggregate output from given resources may be ‘occasioned by commerce’ and will be associated with a rise in profits of those engaged in the new or expanded trade which extends to the general rate of return. 4 It would seem to be precisely this body of doctrine, which attributes the rising value of national product since 1793 to the opening of new markets for British goods, with which Ricardo took issue in 1813, apparently on the grounds that the expansion of the means of finance required to assure a higher level of prices and profits, of which Malthus was very confident, would not in fact be forthcoming. Ricardo’s rejection of Malthus’s position implies the rejection of the standard Smithian view according to which ‘the acquisition of new territory, or new branches 324

THE ROLE OF BENTHAM

of trade, may sometimes raise the profits of stock ... even in a country fast advancing in the acquisition of riches’, by way of the reduction in the pressure exerted by ‘competition of capitals’ (Smith 1937, 93). Early in 1811 Ricardo had been maintaining the Smithian position that the price level will vary with the general profit rate (analysed in terms of ‘competition of capitals’; 1951, III, 92), although at the same time (as we shall see) he also insisted (in a different context) that a monetary expansion was essential to assure a rise in general prices. He was bound sooner or later formally to take issue with the Wealth of Nations. Malthus’s analysis of the period 1793–1813 in terms of Smithian principles may have merely served as catalyst in this intellectual process. Still, dissatisfaction with Smith along the above lines does not explain adoption of the alternative analysis relating the profit rate to agricultural productivity. The precise nature of this transition remains an uncertain matter. Whence this positive aspect of Ricardo’s new position? Here a potentially complicating feature of the record requires attention. What conclusions can be drawn regarding Ricardo’s intellectual indebtedness from the recognition by Smith himself of the phenomenon of increasing land scarcity and its implications for distribution? For at one juncture in his chapter ‘Of the Profits of Stock’ Smith related the secular decline of the return on capital to the necessity of extending cultivation to increasingly inferior land: ‘As the colony increases, the profits of stock gradually diminish. When the most fertile and best situated lands have been all occupied, less profit can be made by the cultivation of what is inferior both in soil and situation.’ The (real) wage is likely to rise, we are told, at least for a period, but since the ultimate state of stationarity is characterized by minimum wages (‘subsistence’ wages in the technical sense) the effects of increasing land scarcity must ultimately include a falling trend of average wages: ‘In a country fully peopled in proportion to what either its territory could maintain or its stock employ, the competition for employment would necessarily be so great as to reduce the wages of labour to what was barely sufficient to keep up the number of labourers’ (Smith 1937, 92f). (This is not the only discussion in Smith’s work of this kind of conception. The implications of changing factor ratios for resource allocation between agriculture and manufacturing and different categories of manufacturing – plentiful and cheap land and scarce and dear labour at an early stage of development, and increasing scarcity of land and abundance of labour as growth proceeds – are developed quite extensively in Book III of the Wealth of Nations in the chapter ‘Of the Natural Progress of Opulence’ and in Book IV in the chapter ‘Of Colonies’.) 5 Despite the presence in Smith’s work of these allusions to the implications of increasing land scarcity, it would, I believe, be an error to give much weight to any dependency by Ricardo upon his predecessor as far as this matter is concerned. On the contrary, the evidence suggests that he was quite unaware of Smith’s discussions. 325

RICARDO - THE NEW VIEW

But if Ricardo in fact misunderstood Smith’s intentions, Smith himself must bear a large part of the responsibility, for his notion of ‘competition of capitals’ occurs repeatedly in the analysis of the profit rate and is presented as the general theory into which the consequences of diminishing returns may be fitted as a special case. It is not surprising then that what appear to be the obvious implications for Ricardo’s own argument of what Smith had to say about secular changes in factor ratios in general and land scarcity in particular should have escaped him. I conclude therefore that Ricardo’s new theory of profit – that aspect involving a linkage between agricultural productivity and the profit rate – constituted an intellectual break with Smith. While for some purposes it may be legitimate to talk of a ‘canonical’ classical theory of growth and distribution, 6 it would be inappropriate to do so in the context of our present problem. Where does Bentham fit into the story? In The Works and Correspondence of David Ricardo, there appears the following editorial note: ‘In the Mill–Ricardo case [of the Ricardo Papers now deposited at Cambridge] there is also a paper probably in Bentham’s hand-writing on the effects on profits of cultivating successive qualities of land’ (1951, X, 388n). I have made a thorough search of the Ricardo Papers and no such document is to be found. It is perhaps fair to assume that the editor felt that the item in question lacked any connection with the origins of Ricardo’s theory, or was not in any other way interesting. 7 Yet to leave the matter there is not satisfying. The existence of such a manuscript whets the appetite. It opens up the possibility at least that Ricardo may have learned something from Bentham regarding the matter at hand. For this reason I wish to explore further some of the potentialities of the connection. I shall consider first, within an appropriate general framework, Bentham’s own position on the question of land scarcity and its implications for distribution, before turning to what we know of Ricardo’s familiarity therewith. I. LAND SCARCITY AND THE FALLING RATE OF PROFIT Throughout the entire range of Bentham’s writings allusions will be found to a downward trend in the rate of profit during the course of secular expansion – a trend formally related to increasing ‘competition of capitals’. There can be no doubt of the Smithian pedigree for the principle, on the basis of which Bentham rationalized the supposed fact that ‘Holland excepted, there is no nation that has so much capital as England, and consequently none which invests it with so little profit’. 8 But there is an ambiguity to be reckoned with. When Smith referred to ‘competition of capitals’ he had in mind pressure in both commodity and labour markets. It is not 326

THE ROLE OF BENTHAM

always clear which of these phenomena – the first clashing with the law of markets and the second quite consistent with it – Bentham had in mind in his various statements or whether both were intended. The falling profit rate is sometimes discussed without reference to pressure from rising real wages: ‘It is the tendency and property of what may be called naturally formed capital, to lower prices in a variety of ways, in proportion to its encrease: to lower the cost of production in respect of labour, by division of labour, introduction of machinery, and so forth, as per Adam Smith; to lower the rate of profit on stock, to wit by competition, as again per Adam Smith’ (1954f [1800–1], II, 450). It appears that Bentham, here at least, had in mind downward pressure on final prices, an impression confirmed in the lengthy tract on Circulating Annuities in a passage touching upon some consequences of a rise in the rate of accumulation: By adding to the number of capitalists of all sorts, [i.e.] proprietors of stock employed in the several branches of productive industry in the production of the several vendible articles of national wealth, and thereby, by encreasing the competition amongst them, in their capacity of vendors of such articles, reducing the rate of profit upon stock (including the rate of interest on money borrowed to be employed as stock), reducing the real as well as money prices of such articles, in as far as the profit upon stock constitutes a component part in the composition of such prices, and thus encreasing the quantity of vendible articles of each given kind brought to market by the employment of a given sum of money. (Bentham 1954e [1800], II, 303) It cannot however be ruled out that, even when not explicit about the matter, Bentham may also have intended pressure on profits by way of rising real wages. And there are instances where this certainly appears to be his intention. Thus in the early work In Defence of Usury : ‘when capital is plenty, interest will be low, and real price of labour high’ and when capital is scarce, interest is high, and real price of labour low. More generally, ‘the relative quantity of capital will encrease, and consequently the rate of interest fall, where thesaurisation goes on faster than population’ (1954a [1787], I, 206–7). From this perspective profits are treated as a ‘residual’, 9 varying inversely with the wage rate. Rising real wages appear in Bentham’s vision of the recent past – the last three or four decades of the eighteenth century – given in the paper of 1795 entitled Supply Without Burden. Here Bentham observed that ‘Children in greater and greater numbers are every year produced, children which for a certain number of years in the earlyer part of their lives consume something without producing anything. But wealth in still greater quantities than is necessary for the maintenance of the children during 327

RICARDO - THE NEW VIEW

that expensive period of their lives is at the same time every year laid up’ (1954d [1795], I, 361). 10 It must be noted that a rapid rate of accumulation (and consequently rising wages) is not necessarily inconsistent with ‘diminishing returns’. But on the whole the impression seems to be that Bentham here took for granted increasing returns. Reference is frequently made to increasing ‘relative wealth’, or ‘wealth in proportion to the numbers of those who share it’ (366), which strictly interpreted includes non-wage earners as well as wage earners. If increasing real average wages are in fact a feature accounting for the falling rate of profit it must logically be presumed that their upward movement exceeds that of productivity. In the later Institute of Political Economy further reference is made to growth rates during the last three or four decades of the eighteenth century, implying that ‘wealth has gone on encreasing faster than population’ (1954j [1801–4], III, 376). At the same time Bentham also observed that ‘the half, or thereabout, of the aggregate wealth will be that which is shared among individuals of the poorest class: and in the case of that class, the wealth of an average individual appears within the period in question to have rather diminished than encreased’. The reference to falling real wages is specifically to the poorest class and it does not preclude rising wages to labour as a whole.11 Whatever may have been the experience of the past with respect to productivity and general wages, prospects for the future were less promising. I come at last to the matter of land scarcity and its implications for distribution. In some of his earlier writings Bentham made a case against colonies on the grounds that domestic investment opportunities – with particular reference to agriculture – were unlimited; colonial activity implied merely the transfer of capital from one use to another with no net benefit: ‘It is impossible ... [that] you can ever have too much agriculture. It is impossible that, while there is ground untilled, or ground that might be better tilled than it is, any detriment should ensue to the community from the withholding or withdrawing capital from any other branch of industry and employing it in agriculture. It is impossible ... that the loss of any branch of trade can be productive of any detriment to the community’ (Bentham 1954b [c.1790], I, 217) The assertion hinges on the proposition, that ‘by promoting an encrease of the productions of agriculture ... you create a demand for the industry of those who afford a demand for the productions of agriculture’ (216). It is unlikely that it would have been so formulated in the event that land scarcity constituted an issue. But, somewhat later, a case was made out for colonization in the Institute of Political Economy, as a means of counteracting the prospective downward trend of real wages attributed to growing land scarcity: ‘taking futurity into the scale, the well-being of mankind appears to have been promoted upon the whole by the establishment of colonies. Taking Britain for example, at the rate at which population 328

THE ROLE OF BENTHAM

has been encreasing for this last century, long before the conclusion of the present century, the population would have extended beyond the utmost number for which the soil would be capable of affording sustenance: long before which period [a] great diminution of relative opulence, a severe sense of general poverty and distress, would necessarily have taken place’ (1954j [1801–4], III, 355). 12 The True Alarm similarly spells out the consequences of land scarcity as an even more immediate prospect: ‘If we consider further the rapid encrease of population such as it has been even during the war, if we observe that it would soon, by its natural course, reach the point where it exceeds the means of subsistence which the two isles could produce, it will be recognized that the emigration of men and capital is a real good in the present state of Great Britain’ (1954g [1801], III, 68). 13 And, in the Defence of a Maximum, ‘exportation of capital and emigration’ are cited as the sole valid solutions, apart from the setting up of state magazines, for the consequences of the increasing scarcity of corn: The scarcity has for its ulterior cause prosperity in all its shapes: an exuberant population, exuberant not with reference to wealth taken in all its shapes, for that too is an exuberance, but with respect to the capacity of raising within the local precincts of the chief seat of empire the quantity of food necessary for the sustenance of its inhabitants .... In regard to scarcity, two remedies commonly relied on as sufficient are essentially inadequate: cultivation of wastes, and importation with or without bounties: ... two others commonly shrunk from, but the only ones upon which any safe reliance can be placed: magazines on public account, and facility afforded, allowance declaredly and liberally given, to exportation of capital and emigration. By inadequate, in speaking of cu[l]ture of wastes, I certainly do not mean undesirable: but where is the resource when all shall have been brought into culture? a state of things which many now living may perhaps live to see. The arrival of the period is an event worth calculation, but [this] is not a fit place [for it]. ... In the mean time encrease of mouths is going on, as fast perhaps as the encrease of land in a state to feed them. ... It is time to cast off antipathies and panic, and look difficulties in the face. Subsistence must remain for ever precarious, or magazines must be established. Wheat with the inferior grains rather than none – rice from Hindostan would stand clearest of objection. The objections that have been urged against magazines are strong, perhaps conclusive. But they all turn upon a state of things out of which we have emerged, and in which nothing but some unexampled calamity can replace us. They turn upon an habitual sufficiency, either actual or possible, of the average stock of grain for the 329

RICARDO - THE NEW VIEW

subsistence of [all] inhabitants. With us, barring calamity as above, or emigration to an unexampled and improbable amount, the very possibility of such a sufficiency is gone for ever. Population has already outstripped culture. Population having no limit, so long as food is to be had from abroad in exchange for wealth, that it should ever be overtaken by culture seems altogether improbable, that it should long continue so to do is, unless contiguous land were to arise out of the sea, impossible. (Bentham 1954i [1801], III, 293–6) In the same work will be found pertinent references to the limits imposed upon the application of capital to agriculture (flowing in part from the small size of holdings) (299f). And a criticism of the expenses entailed by colonies concludes in the following terms: Thus stands the account, so long as the land suffices for its inhabitants in prospect as well as in existence, and so long as emigration, whether of hands or capital, is a loss. But when efflux in both ways is become a relief – efflux of hands and mouths by mitigating scarcity, efflux of capital by mitigating the income tax imposed by capitalists upon capitalists as capital accumulates, and the rate of interest, the income obtainable for the use of it, is borne down – in this already impending, if yet scarcely so much as imagined, state of things, colonies, though still a drain, are notwithstanding, and even because they are a drain, a relief. (Bentham 1954i [1801], III, 301–2) The outcome of this seems clear enough – the consequences of land scarcity are a falling rate of real wages and a falling rate of profit. 14 The former trend is explicitly noted and also suggested by the allusions to emigration, while the latter is implied by the references to capital export also within the context of the implications of land scarcity. It also follows that the falling profit rate does not necessarily hinge upon rising commodity wages; it may, in conditions of falling productivity, occur despite a downward trend in the real wage rate. This constitutes what was to be standard Ricardian Can more be said doctrine and Bentham preceded Ricardo in the general formulation. regarding the logic for the case? I turn to this matter in the next section. II. AN IMPLICIT LOGICAL BASIS FOR BENTHAM’S POSITION It is possible to construct a logical basis for Bentham’s position from his writings along the following lines. The reasoning amounts in substance to that subsequently utilized by Ricardo. 330

THE ROLE OF BENTHAM

Land scarcity is manifested in the rising prices of cereals. Thus in Paper Mischief the contemporary price movements are explained partly by an ‘occasional’ cause – poor harvests – and partly by an ‘habitual’ cause, namely ‘a deficiency in the quantity of land employed in the production of the article in question, regard being had to the continually encreasing numbers of the mouths that call for it’ (1954f, II, 435). 15 Moreover, Bentham adhered to Smith’s approach towards both real-wage determination, running in terms of the relative growth rates of capital and population, and money-wage determination involving the real wage in conjunction with the prices of wage goods. According to the Smithian analysis a change in the price of cereals will imply subsequently an appropriate change in the money wage to assure against any reduction in the commodity wage. (It must be emphasized that nothing in the argument hinges upon the presumption that ‘subsistence’ wages prevail.) The matter is stated repeatedly but nowhere clearer than in the Manual of Political Economy in the context of a hypothetical fall in the price of oats in consequence of a subsidy: Oats would be sold to the poor for less money to the amount of the bounty, but they would get less money to buy oats with. What the poorest class of the poor has to live upon, is the wages of labour: that is the quantity of wealth given to them in exchange for their labour. But the rate of wages depends upon, and is necessarily governed solely and exclusively by, the degree of opulence in the country at the time: that is by the proportion of the quantity of wealth in readiness to be employed in the shape of capital in the purchase of labour to the number of persons for whose labour there is a demand: and this degree of opulence is supposed ... to be the same. (Bentham 1954c [1793–5], I, 247–8)16 For the next step in the argument we must refer back to Bentham’s adherence to the Smithian ‘adding-up’ cost theory of price. This has been briefly documented earlier from various statements relating reduced general prices to a lower return on capital in the context of the discussion of ‘competition of capitals’ (above, p. 327). In one of the relevant discussions – that in Circulating Annuities – there is a considerable elaboration of the consequences for prices of upward pressure on wages. A distinction is there drawn between cases of initial unemployment and initial full employment; money supply increase in the latter case will generate upward pressure on general prices in consequence of higher wage costs per unit: If it were possible that such offer should confine itself to hands whose time was already fully employed, and employed to as much advantage (in respect of the amount of their contribution to the general mass of national wealth) as they would be were the supposed fresh offer to be accepted by them, it could not, although it were accepted, be productive of any the smallest 331

RICARDO - THE NEW VIEW

addition whatever to the general mass of national wealth: it could have no other effect than that of raising the price of labour, and thence the prices of goods, in respect of such part of the price of goods as is composed of the price of the labour employed in the production of those goods. (Bentham 1954e [1800], II, 304–5)17 It appears from this passage read in context that the adding-upcost approach, according to which a rise in a cost element (such as wages) implies a consequential increase in general prices, was strictly conditional upon a concomitant expansion of the money supply. Moreover, detailed discussions of the effects on general prices of increased money supply acting by way of the higher expenditure by labourers on consumer goods will also be found in The True Alarm (1954g, III, 114f). Bentham’s adherence to Smithian doctrine was evidently much qualified. Before exploring the implications of this qualification it will be fruitful to examine Bentham’s highly critical comments on Smith regarding the general price level. In The True Alarm Bentham took Smith to task for failing to deal with two questions relating to the monetary mechanism: ‘In which way and by what causes does cash receive an encrease?’ and ‘Why is the Bank of England the only channel, or almost the only channel, through which gold in bars passes to the Mint and enters into circulation?’ (131n). 18 In a subsequent passage, apparently part of the same manuscript (sometimes referred to as ‘Of the Balance of Trade’) the first criticism is applied to Smith’s analysis of contemporary price movements, according to which – at least on Bentham’s reading – increased general prices, over an extended period of time, were accounted for in terms of a series of poor harvests, without allowance for the necessary increase in money supply : Speaking of his own time, [Smith] denies the facts of an encrease of prices in so far as it depends on an encrease of money. He does admit an encrease of prices, and even for the period of the ten or twelve years ending in 1775; but he attributes it to the bad seasons, and not to money. He prefers the assumption of ten or twelve bad seasons to any other: not considering that, though the dearth of corn may encrease the relative price of this commodity for a time, yet it cannot encrease the aggregate of prices for any considerable length of time: because the encrease in the aggregate of prices implies by definition an encrease in the aggregate of money, both quantity and [velocity of] circulation being taken into account. (Bentham 1954h [1801], III, 237) 19 Bentham frequently restated his position – without necessarily referring to the Wealth of Nations – that an increased volume of money (or higher velocity) is 332

THE ROLE OF BENTHAM

essential to maintain a higher price level, except in the case of short-term grain shortages and war: These two causes suffice to bring about, each in its sphere of action, a rise of prices, and even a considerable rise, without any addition having been made to the mass of money. These two cases excepted, prices can only rise by reason of a pecuniary augmentation in the mass of the national revenue, that is to say, in the total of individual revenues: and that augmentation can only take place by the addition of a new quantity of money, or by a greater rapidity in its circulation. (Bentham 1954g [1801], III, 114–15) It follows logically from all this that a rise in money wages during the course of a regular process of capital accumulation – assuming relatively slow population growth, and an unchanged supply of money – cannot generate a general rise of prices. 20 It also follows logically, and this is the key point for our purposes in this paper, that the rising price of corn in the course of secular expansion due to land scarcity, induces a compensatory rise in the money wage rate to assure unchanged real wages which cannot be passed on in higher general prices (in the event of a constant money supply) and which therefore exerts pressure on the return on capital. 21 In the event that the burden of land scarcity is, so to speak, shared between labourers and capitalists – as Bentham implies by his allusions to a falling real-wage rate as well as a falling profit rate in his discussions of secular trends – a slight modification is required: The money-wage increase in consequence of the higher corn price will be insufficient to prevent some decline in the commodity wage in which case the fall in the profit rate is somewhat checked though not prevented. 22 III. ON RICARDO’S POSSIBLE ‘DEBT’ TO BENTHAM The inverse relationship between wages and profits constitutes what came to be known as the fundamental Ricardian theorem on distribution. It is my belief, based on the foregoing evidence, that the theorem itself and the logic upon which Ricardo constructed his case for a falling profit rate in consequence of land scarcity – which turns strategically upon the demonstration of constant general prices in the face of a rising price of corn and money wage rate 23 – are to be found in Bentham’s writings at the turn of the century. Yet we must proceed cautiously. Even if it were certain that the Bentham documents examined above had been available to Ricardo we could not be sure of the impression conveyed thereby. It must be remembered that while Bentham related the secular decline of the profit rate to land scarcity the logic for the relationship traced out in 333

RICARDO - THE NEW VIEW

our reconstruction was by no means formulated by Bentham himself in a deliberate and concerted fashion; in particular, the squeeze on profits exerted by the constancy of prices in the face of rising money wages is only an implicit property of his argument, but not explicitly stated. Nor did he formally disclaim the principles of ‘competition of capitals’ and the ‘adding-up-cost’ approach to pricing. 24 What we in fact do know of Ricardo’s reaction to Bentham and other contemporaries during the period of his transition from Smithian theory, and immediately before, reinforces the need for prudence in the drawing of intellectual linkages. This matter requires our attention. A French translation, entitled ‘Sur les Prix’, by Etienne Dumont of Bentham’s The True Alarm – made during the period 1802–10 – found its way into Ricardo’s hands through James Mill, and Ricardo made informal notes on the manuscript. It is not clear that he saw the entire document reproduced by Stark and the specific criticism therein of Smith. 25 Moreover, the document sometimes referred to as ‘Of the Balance of Trade’, which was originally probably part of The True Alarm, may not have been seen by Ricardo (Stark 1954, III, 27). But Ricardo did comment upon Bentham’s allowances regarding general price increase (given the money supply) in the case of short-term grain shortages: ‘If any rise in the price of commodities is caused in the way here supposed it must be by diminishing the amount of commodities, which will make the money which circulates them more relatively abundant. If the commodities remained the same and their price was increased, more money would be absolutely necessary to circulate them. But if it is the mass of prices of which the author speaks, he is mistaken because what [ sic] one commodity rose in price another would fall’ (Ricardo 1951, III, 300). The passage is somewhat unclear for it would seem logical to suppose that in the case of a poor harvest the volume of transactions does decline. However, the following note suggests that Ricardo was insisting upon an unchanged price level in final equilibrium alone, conceding a temporary price increase as part of the adjustment mechanism: ‘[Bentham’s] arguments are all founded on the supposition of the country to which they are applied being insulated from all others. If not it is evident that the rapidity of the circulation would cause an exportation of money, and would not therefore raise prices at home’ (III, 300). 26 An increase in the money supply was thus said to be essential for a (permanent) rise of general prices. 27 Moreover, at least in the ‘sound state’ of the currency, such an increase in the money supply occurred only in response to an initially reduced price level; no other means of raising the money supply was conceded (III, 325). Indeed, Ricardo firmly rejected Bentham’s various allusions to real output effects of monetary expansion in terms making this very clear: ‘money cannot call forth goods’, he insisted, ‘but goods can call forth money’ (III, 301). 334

THE ROLE OF BENTHAM

Ricardo’s insistence, in line with that of Bentham, upon an increase in the money supply to assure a higher level of prices is inconsistent with subscription to the Smithian proposition that a It is precisely this money wage increase or an increase in profits will be passed on to consumers. direction that the subsequent Ricardian argument was to take. But there is nothing to suggest at this stage that Ricardo recognized these implications. We must now address the matter of land scarcity and its relevance for distribution. Ricardo explicitly commented on a passage in Dumont’s translation involving the consequences of land scarcity for the returns to labour and capital, 28 but denied any advantage to capital outflows and insisted that stationariness was a far distant prospect: ‘It can never be allowed that the emigration of Capital can be beneficial to a state. A loss of capital may immediately change an increasing state to a Stationary or retrograde state. A nation is only advancing whilst it accumulates capital. Great Britain is far distant from the point where capital can no longer be advantageously accumulated. I do not mean to deny that individual capitalists will be benefited by emigration in many cases, – but England even if she received the revenues from the Capital employed in other countries would be a real sufferer’ (III, 274). By itself this might appear to constitute an index of Ricardo’s appreciation of the key theoretical issue at hand – his difference with Bentham amounting solely to the empirical question of proximity to the stationary state. There exists, moreover, an independent allusion in Ricardo’s notes on Bentham to the matter of diminishing returns: ‘It appears to me that the possession of new Land would add to our own sum of riches without additional labour, because the same labour employed on double the quantity of equally good land now in cultivation in England would produce a greater return. This opinion is founded on the decreasing power of the land to produce in proportion to the labour and capital employed on it’ (III, 287). 29 But the implications of these references should not be exaggerated. This becomes clear when we carry the story further. I turn to an obscure work of 1810 by Coutts Trotter which came to Ricardo’s attention. 30 Trotter had denied the existence of a contemporary excess paper issue: ‘the currency of the country is in no degree depreciated by the use of paper’ (in Ricardo 1951, III, 388). That the currency had lost some part of its value is conceded but the fact is ascribed to gold itself (throughout Europe) falling in value relative to commodities; to the burden of taxation; and to the increase in population. It is the last which interests us here. For Trotter gave a clear statement of diminishing returns, and attempted to relate the phenomenon to the level of general prices by way of the price of corn and the wage rate: In a country insulated as ours now is, by political as well as natural circumstances, every increase of population must make an increase in the 335

RICARDO - THE NEW VIEW

demand for all the articles which land and industry produce. To raise the former, worse soils and more unfavourable situations must be taken into cultivation; and the produce therefore will be obtained, and must be sold, at an increased expense. To create the latter, men must be paid at a higher rate of wages, because in every state of society, and especially in one progressive, as that of England is, men must receive somewhat above what is necessary for their support; and the expense of that support will be regulated principally by the cheapness or dearness of food. (Trotter, in Ricardo 1951, III, 388–9) Trotter thus linked the Smithian relationship between the corn price and general prices to the principle of diminishing returns. Ricardo’s response is of the very first importance. It is that ‘every increase of population must arise from an increase of capital, and has a tendency to lower the prices of commodities and therefore the wages of labour, not to raise them’ (III, 389). The notion of diminishing returns, and implications thereof for the rising real costs of food production, would appear from this observation to have been totally foreign to Ricardo – despite the allusions to the phenomenon in the Notes on Bentham. Ricardo had his eye solely upon the depressing effects upon prices of Smith’s ‘increased competition’ between capitalists: ‘An increase of Capital never raises the prices of commodities. ... Can there be an increase of population without an increase of Capital having preceded it?, yet ... we are told that an increase of population will occasion a rise in the price of commodities, and in the wages of labour. ... A competition of Capitalists keeps down prices’ (III, 389–90). It is evident then that, at least until the early months of 1811, Ricardo fell squarely in the Smithian camp. His attachment to the principles of the Wealth of Nations extended to the effect on the price level of changes in the price of corn by way of wage-rate variation; the ‘adding-up’ cost theory of price; and the explanation of profit-rate determination in terms of the ‘competition of capital’. At the same time, building-blocks for a reconstruction were already at hand in Trotter’s explicit formulation of the principle of diminishing returns applied to explain rising corn prices and money wages – if not in Bentham’s less formal statements – coupled with Bentham’s insistence – to which Ricardo subscribed – that any general price increase cannot be sustained without an increased volume of money. Ricardo’s total unawareness of this potent mixture is clear not only from the fact that he rejected Trotter’s case based upon diminishing returns, but that he did so on the grounds of inconsistency between this argument and the Smithian notion of ‘competition of capitals’ whereby capital accumulation was supposed to lower general prices, rather than the ‘monetary’ grounds of the non-sustainability of higher general prices (along which lines the breakaway was ultimately to occur). 336

THE ROLE OF BENTHAM

It seems fair to conclude from this episode, quite generally, that the extant documents which came to Ricardo’s attention in 1810–11 – and these include some of the Bentham items we have discussed – left no impression upon him (no conscious impression at least) regarding the main issue at hand. There remains then only the lost Bentham manuscript on the effects on profits of cultivating successive qualities of land, referred to by Piero Sraffa. Unless and until this comes to light I cannot see that much more can be said on the potential influence of Bentham upon Ricardo. In conclusion, I would add only that since Bentham had written most of his economics by 1813 it is probable that the lost document was of no later date; it therefore remains an open possibility that Ricardo’s new doctrine, formulated in the first half of that year, may have owed something to Bentham. This possibility is reinforced by the presence in the extant Bentham manuscripts of arguments, outlined in this paper, which are quite consistent with those subsequently used by Ricardo in making his case. NOTES 1. There are no extant letters between those of late March and early August 1813, while those up until 24 March make no mention of profit-rate determination. As it is most unlikely that what Ricardo called ‘my theory’ in August would have been under consideration without appearing in the correspondence, the transition from the Smithian conception, it seems safe to conclude, probably occurred at some time during the fourmonth period from April to July 1813. 2. Malthus’s letter, to which Ricardo’s of 10 August is a reply, and Malthus’s answer are both wanting; indeed we do not have any Malthus letters from March 1812 until June 1814. The two met frequently, however, during 1813 and the issue of profit-rate determination was doubtless discussed. Letters by other correspondents to Ricardo between January 1812 and November 1813 are missing. 3. Thus immediately after the passage from the letter of 10 August Ricardo formalized the proposition in the following terms: ‘I am of opinion that the increased value of commodities is always the effect of an increase either in the quantity of the circulating medium or in its power, by the improvements in economy in its use, – and is never the cause. It is the diminished value, I mean nominal value, of commodities which is the great cause of the increased production of the mines, – but the increased nominal value of commodities can never call money into circulation. It is certainly an effect and not a cause.’ Similarly, immediately following the passage from the second letter Ricardo asserted of commodities in general that it is ‘their cheapness which is the immediate cause of the introduction of additional money’ (1951, VI, 93–4, 95). 4. This interpretation was first suggested by G.S.L Tucker, ‘The Origins of Ricardo’s Theory of Profits’ (1954, 323–4). 5. The discussion of Smith’s Books III and IV is analysed in my Economics of Adam Smith (1973), chap. 10, dealing with investment priorities. 6. Great weight is placed on the ‘Ricardian’ implications of Smith’s discussion of growth and distribution in Samuelson 1977 and Samuelson 1978. 7. Professor Burns, who heads the University College ‘Bentham Project’, has written to the 337

RICARDO - THE NEW VIEW

8. 9. 10. 11.

12.

13.

14.

15.

16.

17.

present writer: ‘Nothing that I am aware of here, in the British Museum or among the Dumont papers in Geneva seems to correspond to the paper Sraffa described.’ Jeremy Bentham, Of the Balance of Trade [1801], in W. Stark (ed.) Jeremy Bentham’s Economic Writings (London, 1954) III, 222. I rely throughout on this collection. See, for example, 1954a [1787], I, 116–17, where Bentham writes of profits as the revenue remaining to the employer after deducting wages. It may be noted that the falling profit rate is alluded to in this context, p. 359. Cf. also some rough notes by Bentham cited by Stark 1954, III, 482, referring to ‘the impossibility of raising the wages of ordinary labour beyond mere subsistence’, but also asserting that ‘the higher the wages of labour the better consistent with national security’ which implies a distinction between different groups of labourers. (See also 352–3.) There is a similar (but not identical) reference recorded by Bowring in his edition of the Manual of Political Economy : Let us take England, for example. According to the progress which population has made during the last century, it may be supposed that it would soon have attained its extreme limits – that is to say, that it would have exceeded the ordinary means of subsistence, if the superabundance had not found means of discharging itself in these new countries. But a long time before population has reached these limits, there will be a great diminution of relative opulence, a painful feeling of general poverty and distress, a superabundance of men of all the labouring classes, and a mischievous rivalry in offering their labour at the lowest price. (John Bowring, The Works of Jeremy Bentham, London 1833–43, III, 53) According to Stark (1954, I, 49), the ‘Manual’ as published by Bowring is in fact largely from a manuscript developed after the turn of the century, and designed for the Institute of Political Economy. In certain materials apparently prepared for the Manual of Political Economy (1793–5), as classified by Stark, there is explicit reference to the stationary state and subsistence wages, reminiscent of the Wealth of Nations ; cf. 1954, III, 539. Also in the Manual will be found the celebrated passages in Latin alluding to birth control as a means of preventing population increasing more rapidly than capital (1954c, I, 272–3). On this matter see J.R. Poynter, Society and Pauperism; English Ideas on Poor Relief, 1795–1834 (London and Toronto, 1969), 122f. It is not clear to me that the case for capital export is necessarily connected to Bentham’s allowances of a possible divergence between savings and investment; but see Winch 1965, 33. The falling profit rate for Bentham (as for Smith and the later classicals) was a ‘contingent prediction’, conditional upon appropriate ceteris paribus conditions. See, for example, 1954d, I, 359–60; 1954h, III, 222, 287, 352, 469. Reference is made in The True Alarm (1954g, III, 135) to a relative rise in the prices of agricultural compared to manufactured produce – since 1760 the former had increased by more than 50 per cent and the latter by at most 12 ½ per cent. (Cf. also Circulating Annuities, 1954e, II, 331.) See also 1954c, I, 236n, 272–3, regarding real-wage determination. On the variability of the money wage with the price of provisions see II, 447–8; III, 91–2, 127, 246, 503. For further discussion of Bentham on real-wage rate determination see Poynter 1969, 120–1. (The same principles are applied to rent, 1954e, II, 303.) By contrast: ‘If it were possible 338

THE ROLE OF BENTHAM

18. 19.

20. 21.

22.

23.

24.

25. 26.

27. 28. 29.

30.

that such offer could confine itself to hands as yet altogether without employ, it would, in as far as it were thus accepted, be productive of a real addition to the mass of wealth, without any corresponding addition, or any addition at all, to the price of labour, and therefore without any addition at all to such part of the prices of goods as is composed of the price of the labour employed in the production of these goods’ (304). Cf. also 1954e, II, 339, 342n. Bentham did not, however specifically attribute to Smith (in the present context) a process whereby an increase in the price of corn acts upon general prices by way of the money-wage rate. See a possible allusion to this result 1954g, III, 120–1. The argument is foolproof, however, only if it is certain that the burden of a wage increase does not fall on rent. While Bentham does not develop the differential rent principle it must be borne in mind that Ricardo too formulated the inverse profit-wage relation before realizing the significance of the principle for his case; differential rent was subsequently fitted easily into a well prepared structure. Much later Bentham apparently criticized Ricardo’s formulation in the [ Essay on Profits ]. See Bowring, The Works of Jeremy Bentham, X, 498: ‘In Ricardo’s book on Rent’, runs the report, ‘there is a want of logic. I wanted him to correct it in these particulars; but he was not conscious of it, and Mill was not desirous. He confounded cost with value. ’ In a later manuscript (dated 1821) Bentham asserted that the corn laws entailed a high corn price and ruin to the labourers (1954k, III, 410). This assertion also suggests that the rising corn price is not entirely compensated for. It may be noted that Ricardo did not formulate a complete case in early 1813. And even in the following year he still maintained a positive (if qualified) relationship between the corn price and general prices. In short, the relevance for his own approach to profits of the monetary argument used to counter the principle of ‘competition of capitals’ was apparently not self-evident. The absence of a well thought out case may be illustrated by the apparent failure by Bentham to apply his monetary argument against Smith’s analysis of the corn export bounty according to which the rise in the price of corn would generate, via the money wage rate, a general rise in the level of prices. (For Bentham’s own case against the bounty, see 1954c, I, 268–7; 1954e, II, 295–6.) Above, pp. 332–3. In particular, there are no comments by Ricardo on the questions raised by Bentham, 1954g, III, 131n. See also Ricardo’s response (1951, III, 311): ‘Is not the mass of prices the same after [corn] scarcity as before. May we not as before put the mass of commodities of all sorts on one side of the line, – and the amount of money multiplied by the rapidity of its circulation on the other. Is not this in all cases the regulator of prices?’ This too is presumably a statement relating to long-run equilibrium states allowing therefore for a temporary increase in the general level of prices. Except where taxation was responsible for higher prices; Ricardo 1951, III, 270, 307, 328, 341. See above, p. 329 for corresponding English text given by Stark (1954g, III, 68). Ricardo actually took Bentham to task in this context for failing to appreciate the phenomenon of diminishing returns. Immediately thereafter (1951, III, 288) he conceded that Bentham possessed the principle. The Principles of Currency and Exchanges applied to the Report from the Select Committee of the of Gold Bullion, 1st edn (London, House of Commons Appointed to Inquire into the High Price 339

RICARDO - THE NEW VIEW

1810). See for the full text J.H. Hollander, Minor Papers on the Currency Question (Baltimore: The Johns Hopkins Press, 1932). (Ricardo’s comments were written in December 1810, or thereabouts.)

APPENDIX While searching the Ricardo Papers I came across a manuscript (ADD. 7510. IV. B24) written in French, in Ricardo’s hand, entitled ‘The advantages and disadvantages of colonies are ably explained by Mr Bentham’. (The item bears no watermark.) Here we find Bentham in a ‘conservative’ mood; it is not generally known that Ricardo, at any period, took a similar position, approving of the notions that it were better for the colonies to remain with the mother country; that the United States would have been better served by remaining longer under Britain; and that Egypt would benefit by coming under British rule. Internal evidence suggests, according to Professor J.H. Burns, that the original from which Ricardo copied was a Bentham manuscript in French (rather than a Dumont translation) and that such a manuscript could have been communicated to Ricardo by James Mill during the period 1810–11. Circumstantial evidence also suggests that the manuscript dates to the early years of the new century; see Winch 1965, 34: ‘In the early years of the nineteenth century, possibly as a result of the turn of events in France, Bentham retreated from the radicalism on questions of government expressed in some of his writings in the 1790’s. Prior to his meeting with James Mill in 1808, Bentham’s Toryism seems to have reasserted itself; and this change of heart is reflected in his views on colonies at this time.’ There is a resemblance (although not an identity) between the manuscript in question, and passages in the Etienne Dumont collection Théorie des peines et des recompenses, 3rd edn (Paris, 1826), II, Livre Quatrième, chap. XII (‘Des Colonies’), 336–55, and in the materials reproduced by W. Stark as ‘The Institute of Political Economy’. The manuscript lacks the discussions of the ultimate limits to population growth due to land scarcity, and of the relatively slow growth of population in oldsettled countries contained in these works. (See above, pp. 328–9 and note 12.) It cannot, however, be positively excluded that the manuscript from which Ricardo copied contained these or similar passages.

REFERENCES Bentham, Jeremy (1954a) [1787] Defence of Usury, in Jeremy Bentham’s Economic Writings, W. Stark, London: George Allen and Unwin, 121–207. ——(1954b) [c. 1790] ‘Colonies and Navy’, I, 209–18. ——(1954c) [1793–5] Manual of Political Economy, I, 219–73. ——(1954d) [1795]Supply without Burden, I, 279–367. ——(1954e) [1800]Circulating Annuities, II, 201–423. ——(1954f) [1800–1] Paper Mischief (Exposed), II, 425–58. ——(1954g) [1801]The True Alarm, III, 61–216. ——(1954h) [1801]Of the Balance of Trade, III, 217–246. ——(1954i) [1801]Defence of a Maximum, III, 247–302. ——(1954j) [1801–4] Institute of Political Economy, III, 303–80. 340

I, ed.

THE ROLE OF BENTHAM

——(1954k) [1821] Observations on the Restrictive and Prohibitory Commercial System, III, 383–418. Bowring, John (ed.) (1833–43) The Works of Jeremy Bentham, Edinburgh: Wilham Tait. Hollander, Samuel (1973) The Economics of Adam Smith, Studies in Classical Political Economy Vol. I, Toronto and Buffalo: University of Toronto Press. Poynter, J.R. (1969) Society and Pauperism; English Ideas on Poor Relief, 1795–1834, Toronto: University of Toronto Press. Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 vols), Cambridge: Cambridge University Press. Samuelson, Paul A. (1977) ‘A Modern Theorist’s Vindication of Adam Smith’, American Economic Review LXVII, 1 (February): 42–9. ——(1978) ‘The Canonical Classical Model of Political Economy’,Journal of Economic Literature XVI,4 (December): 1415–34. Smith, Adam (1937) [1776] An Inquiry into the nature and causes of the Wealth of Nations, New York: Modern Library. Stark, W. (ed.) (1954) Jeremy Bentham’s Economic Writings, London: George Allen and Unwin. Trotter, Coutts (1932) [1810] The Principles of Currency and Exchanges, in J.H. Hollander (ed.) Minor Papers on the Currency Question, Baltimore: The Johns Hopkins Press. Tucker, G.S.L. (1954) ‘The Origins of Ricardo’s Theory of Profits’, Economica XXI (November): 320–33. Winch, Donald (1965) Classical Political Economy and Colonies, London: London School of Economics.

341

20 ON PROFESSOR SAMUELSON’S CANONICAL CLASSICAL MODEL OF POLITICAL ECONOMY

It is probable that Professor Samuelson’s statement of the ‘canonical classical model of political economy’ will become the locus classicus for the next generation of textbook writers; teachers of the history of thought would be advised to familiarize themselves with the ingenious diagrams in particular. But the author, who himself warns of the dangers of myth-making in undertaking such reconstructions (1978, 1415), particularly (I would add) such Gestalt reconstructions – has provided no means whatsoever whereby the reader is enabled to evaluate the historical accuracy of the formulation. 1 It is to repair this deficiency, in part at least, that I devote the present note, attending largely to Adam Smith and David Ricardo. By adopting a historical approach, some quite tricky problems of interpretation raised in the article are avoided. Professor Samuelson alludes to a number of modern authorities who take issue with his view that the principle of diminishing returns features prominently in the Wealth of Nations (Samuelson 1978, 1432n). This is surely a matter that can be answered by careful textual analysis. There is in fact much that can be said in a formal sense in favour of Professor Samuelson’s reading of Adam Smith; it is indeed possible to build up from the texts a model wherein the profit rate tends to decline secularly in consequence of extensions to increasingly inferior land. But this would be ‘irrelevant’ 2 if contemporary and near-contemporary economists were totally unaware of Smith’s position. And such indeed seems to have been the case. The outstanding historical fact, which is in danger of being overlooked, is the failure of late eighteenth- and early nineteenth-century writers to recognize the presence in Smith’s work of diminishing returns and its implications for distribution. The significance of this fact, which certainly requires explanation, cannot be overstated. It colours the reading of the early nineteenth-century literature; in particular, an entire range of difficult questions arises relating to the nature and timing of what Ricardo considered, with reason, to be his break-away from Smith 342

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

regarding profit rate determination and the rationalization of the profit-rate trend. For despite the presence in Smith’s work of allusions to the implications of increasing land scarcity, the evidence reveals that Ricardo was totally unaware of Smith’s discussions. The significance to Ricardo of the principle of diminishing returns lay less in any resultant ‘prediction’ of a declining rate of profit and more in the negative proposition that in its absence there would be no other conceptual reason for a decline; in short, he regarded his argument as a counterfoil to Smith’s analysis in terms of ‘competition of capitals’, which had to be ruled out of court because it flew in the face of the law of markets – to which Smith himself subscribed. If Ricardo misunderstood Smith’s intentions, Smith himself must bear a large part of the responsibility, for his notion of ‘competition of capitals’ occurs repeatedly in the analysis of the profit rate and is presented as the general theory into which diminishing returns may be fitted as a special case. Smith’s own responsibility in this regard is further compounded by the presence in his work of propositions (conspicuously formulated) relating to corn pricing that actually imply a denial of the principle of diminishing returns. Professor Samuelson has asserted at one point of his argument that the differences between the classical authorities were matters of semantics alone (1430, 1432); and more specifically that there are ‘no thought experiments proposed by Ricardo to which he has given a different substantive answer than would Smith’s system’ (1432n). It will become clear in what follows that this is simply not the case. Neither is it true that the contradictions in the Wealth of Nations can be regarded as ‘insignificant’ (1415); they were significant enough to have caught the attention of contemporary commentators. Before we proceed to substantive matters, it should be emphasized that Professor Samuelson’s short-circuited version of the ‘canonical’ classical model, attributed to Ricardo, is a serious and regrettable case of myth-making; so too is his (related) assertion that Ricardo failed to appreciate as clearly as did Smith the demand–supply determination of wages (1416–17; also Samuelson 1977). It is unnecessary to provide verse and line references for all of this. The case has been fully documented recently by David Levy (1976), Sir John Hicks and Hollander (1977), and Carlo Casarosa (1978). In what follows I outline the textual evidence suggesting the presence of the ‘canonical’ model in the Wealth of Nations. In Section II, I attempt to account for its neglect by Smith’s contemporaries and near-contemporaries. Section III discusses the Ricardian theory of profits envisaged as a challenge to Smithian doctrine; reference is made also to the position of Sir Edward West from the same perspective. Further objections to Smith in terms of the conflict between the principle of ‘competition of capitals’ and the law of markets are taken up in Section IV. The final section discusses Ricardo’s fundamental innovation on distribution–the inverse profit–wage relationship. 343

RICARDO - THE NEW VIEW

I. THE CANONICAL MODEL IN THE ‘WEALTH OF NATIONS’ The celebrated notion ‘competition of capitals’ occupies centre stage in the profitrate analysis of Smith’s first Book and appears as the general theory into which the implications of diminishing returns are fitted as a special case. This section is devoted to an exposition of the textual evidence. The ‘ordinary’ rate of profit, in Smith’s account, is ‘every where regulated by the quantity of stock to be employed in proportion to the quantity of the employment, or the business which must be done by it’ (1937 [1776], 799). Accordingly, the rate tends to fall with accumulation in consequence of ‘increased competition’ between capitalists: ‘The increase of stock, which raises wages, tends to lower profit. When the stocks of many rich merchants are turned into the same trade, their mutual competition naturally tends to lower its profit; and when there is a like increase of stock in all the different trades carried on in the same society, the same competition must produce the same effect in them all’ (87). The phenomenon is related in Smith’s Book II to what would now be called an increasing paucity of investment opportunities; it will be noted that the pressure on profits manifests itself in the forms of both rising wages and falling final prices: As the quantity of stock to be lent at interest increases, the interest, or the price which must be paid for the use of that stock, necessarily diminishes, not only from those general causes which make the market price of things commonly diminish as their quantity increases, but from other causes which are peculiar to this particular case. As capitals increase in any country, the profits which can be made by employing them necessarily diminish. It becomes gradually more and more difficult to find within the country a profitable method of employing any new capital. There arises in consequence a competition between different capitals, the owner of one endeavouring to get possession of that employment which is occupied by another. But upon most occasions he can hope to justle that other out of this employment, by no other means but by dealing upon more reasonable terms. He must not only sell what he deals in somewhat cheaper, but in order to get it to sell, he must sometimes too buy it dearer. The demand for productive labour, by the increase of the funds which are destined for maintaining it, grows every day greater and greater. Labourers easily find employment, but the owners of capitals find it difficult to get labourers to employ. Their competition raises the wages of labour, and sinks the profits of stock. But when the profits which can be made by the use of a capital are in this manner diminished, as it were, at both ends, the price which can be paid for the use of it, that is, the rate of interest, must necessarily be diminished with them. (Smith 1937 [1776], 336; emphasis added) 344

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

The only specific instance of such increasing paucity of investment opportunities is an allusion in the basic chapter on profits to diminishing agricultural returns. The passage in question, which deals specifically with ‘new colonies’, is important for the reference to both high wage and profit rates at an early stage of development, and a rising wage rate but falling profit rate during the course of subsequent progress:

High wages of labour and high profits of stock, however, are things, perhaps, which scarce ever go together, except in the peculiar circumstances of new colonies. A new colony must always for some time be more under-stocked in proportion to the extent of its territory, and more under-peopled in proportion to the extent of its stock, than the greater part of other countries. They have more land than they have stock to cultivate. What they have, therefore, is applied to the cultivation only of what is most fertile and most favourably situated, the land near the sea shore, and along the banks of navigable rivers. Such land too is frequently purchased at a price below the value even of its natural produce. Stock employed in the purchase and improvement of such lands must yield a very large profit. ... Its rapid accumulation in so profitable an employment enables the planter to increase the number of his hands faster than he can find them in a new settlement. Those whom he can find, therefore, are very liberally rewarded. As the colony increases, the profits of stock gradually diminish. When the most fertile and best situated lands have been all occupied, less profit can be made by the cultivation of what is inferior both in soil and situation. ... [But] the The demand for labour wages of labour do not sink with the profits of stock. increases with the increase of stock whatever be its profits; and after these are diminished, stock may not only continue to increase, but to increase much faster than before. (Smith 1937 [1776], 92–3; emphasis added) In the chapter specifically devoted to colonies in Book IV the high wages and profits characterizing new establishments are discussed in the following terms: Every colonist gets more land than he can possibly cultivate. He has no rent, and scarce any taxes to pay. No landlord shares with him in its produce, and the share of the sovereign is commonly but a trifle. He has every motive to render as great as possible a produce, which is thus to be almost entirely his own. But his land is commonly so extensive, that with all his own industry, and with all the industry of other people whom he can get to employ, he can seldom make it produce the tenth part of what it is capable of producing. He is eager, therefore, to collect labourers from all quarters, and to reward them with the most liberal wages. But those liberal wages, joined to the 345

RICARDO - THE NEW VIEW

plenty and cheapness of land, soon make those labourers leave him, in order to become landlords themselves, and to reward, with equal liberality, other labourers, who soon leave them for the same reason that they left their first master ... In other countries, rent and profit eat up wages, and the two superior orders of people oppress the inferior one. But in new colonies, the interest of the two superior orders obliges them to treat the inferior one with more generosity and humanity; at least, where that inferior one is not in a state of slavery. Waste lands of the greatest natural fertility, are to be had for a trifle. The increase of revenue which the proprietor, who is always the undertaker, expects from their improvement constitutes his profit; which in these circumstances is commonly very great. But this great profit cannot be made without employing the labour of other people in clearing and cultivating the land; and the disproportion between the great extent of the land and the small number of the people, which commonly takes place in new colonies, makes it difficult for him to get this labour. He does not, therefore, dispute about wages, but is willing to employ labour at any price. (Smith 1937 [1776], 532–3) These passages must, however, be read in conjunction with Smith’s allusions to a ‘stationary state’ marked by wage and profit rates at their respective minima: In a country which had acquired that full complement of riches which the nature of its soil and climate, and its situation with respect to other countries, allowed it to acquire; which could, therefore, advance no further, and which was not going backwards, both the wages of labour and the profits of stock would probably be very low. In a country fully peopled in proportion to what either its territory could maintain or its stock employ, the competition for employment would necessarily be so great as to reduce the wages of labour to what was barely sufficient to keep up the number of labourers, and, the country being already fully peopled, that number could never be augmented. In a country fully stocked in proportion to all the business it had to transact, as great a quantity of stock would be employed in every particular branch as the nature and extent of the trade would admit. The competition, therefore, would every-where be as great, and consequently the ordinary profit as low as possible. (Smith 1937 [1776], 94–5) Reference has just been made to a rising trend of real wages and an ultimate state of stationariness when wages are at their minimum or subsistence level. Evidently a 346

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

stage of falling per capita wages must be envisaged along the route. Smith did not spell but there are out the logic behind this downward trend in his analysis of the falling profit rate, pertinent materials from which to draw elsewhere and these perhaps allow us to fill in the picture. The characteristic feature of Smith’s analysis of wages is the role accorded to the rate of capital accumulation as an ‘independent’ variable governing the demand for labour, upon which depends the (long-run or secular) real wage rate and the growth rate of population; to each growth rate of labour demand there corresponds a long-run real wage rate, which assures an equivalent rate of growth of population, and therefore of the work force: ‘It is not ... in the richest countries, but in the most thriving, or in those which are growing rich the fastest, that the wages of labour are highest’ (69). 3 According to the Smithian position, an increase in the (long-run) wage rate will occur in consequence of a change in the rate of capital accumulation; a steady rate of increase of capital would not alter (longrun) per capita wages. By extension it may, therefore, be fair to attribute to Smith the view that the implied falling trend of real wages, alluded to earlier, is attributable to a declining rate of capital accumulation, itself a reaction to the falling profit rate. This interpretation, however, begs many key questions as will shortly become clear. The foregoing textual analysis brings together much of the primary evidence, of which I am aware, in support of Professor Samuelson’s argument that the ‘canonical’ classical model appears in the Wealth of Nations. It is indeed quite an impressive case. 4 Yet from the point of view of the intellectual historian the disturbing fact must be faced that On the Smith left no impression whatsoever on his contemporaries regarding his formulation. contrary, as we shall see, commentators focused on aspects of Smith’s work that point conspicuously away from the phenomenon of diminishing returns. II. SOME CHARACTERISTICS OF THE SMITHIAN PRESENTATION AND THE NEGLECT OF SMITH’S CANONICAL MODEL I shall attempt in what follows to explain the neglect of Smith’s discussion of land scarcity and its implications for distribution by subsequent commentators. Some allowance should perhaps be made for matters of exposition. The discussion is not concentrated in one location; it is scattered and, most significant, much of the evidence given above derives from the analysis of colonies where institutional arrangements differed in some important respects from those in the competitive exchange system, which primarily concerned contemporaries. These formal matters may well have played a part in disguising Smith’s argument. 347

RICARDO - THE NEW VIEW

More important is the fact that Smith, unlike Ricardo, did not establish land scarcity as the sine qua non for a secularly falling profit rate. If we take into account disturbances that check or reverse the regular downward trend of the profit rate, it becomes apparent that increasing land scarcity – diminishing agricultural returns – is not a necessary condition for the supposed trend. Thus the trend would be checked not only by the ‘acquisition of new territories’ but also by the ‘acquisition of new branches of trade’, which reduces the pressure of ‘competition’ in commodity markets, thus permitting final prices and thus profits to rise: The acquisition of new territory, or of new branches of trade, may sometimes raise the profits of stock, and with them the interest of money, even in a country which is fast advancing in the acquisition of riches. The stock of the country not being sufficient for the whole accession of business, which such acquisitions present to the different people among whom it is divided, is applied to those particular branches only which afford the greatest profit. Part of what had before been employed in other trades, is necessarily withdrawn from them, and turned into some of the new and more profitable ones. In all those old trades, therefore, the competition comes to be less than before. The market comes to be less fully supplied with many different sorts of goods. Their price necessarily rises more or less, and yields a greater profit to those who deal in them, who can, therefore, afford to borrow at a higher interest. (Smith 1937 [1776], 93; emphasis added) The analysis is, incidentally, applied to explain an upward movement in interest rates in London after the peace of 1763. ‘New business’ was provided by the British ‘acquisitions’ in North America and the West Indies, and capital was diverted away from the Mediterranean and European branches of trade: ‘So great an accession of new business to be carried on by the old stock, must necessarily have diminished the quantity employed in a great number of particular branches, in which the competition being less, the profits must have been greater’ (93–4). I conclude that Smith’s rationalization of the falling profit rate included considerations apart from land scarcity. The combining of such distinct phenomena could not have helped the case. In any event the argument is vitiated by severe ambiguities. It is not even clear whether the pressure exerted upon the profit rate by increased ‘competition of capitals’ takes the form of rising wages in addition to falling prices. Smith is sometimes circumspect on this matter; as we have already seen ‘the increase of stock, which raises wages, tends to lower profit’ – a fall which is then apparently related formally to competition in commodity markets. 5 We must here keep in mind that capitalists are frequently accorded the ability to escape the burden of rising wages by way of increased 348

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

manufacturing prices and reduced rents (see below, p. 350). It is difficult in any event to appreciate why the wage rate tends ultimately to fall ; to explain it in terms of a declining rate of accumulation as provisionally suggested above does not really solve the problem, for Smith explicitly asserts that ‘the demand for labour increases with the increase of stock whatever be its profits; and after these are diminished, stock may not only continue to increase, but to increase much faster than before’ (93). Finally, and most important, it is not made clear why, if indeed there is scope for a decline in the real wage (and setting aside pressure on final prices), there is any necessary reduction on the profit rate at all. It was Ricardo who ultimately developed a model where all these issues are resolved. These considerations suggest that the main case was not well thought through or clearly argued in the Wealth of Nations. But the failure to leave any impression on the main issues that concern us can be better appreciated if we also keep in mind the common belief that Smith had actually ruled out diminishing returns. This may be illustrated from An Inquiry into the Nature and Progress of Rent (1815) where Malthus set out the principle of differential rent, which he contrasted with the notion of rent envisaged as a ‘monopoly’ return by Smith, Say, the Physiocrats, Sismondi and Buchanan. 6 What is particularly striking from our present perspective is the interpretation and critical assessment of the Smithian approach to corn-price determination in a secular context: Adam Smith has very clearly explained in what manner the progress of wealth and improvement tends to raise the price of cattle, poultry, the materials of clothing and lodging, the most useful minerals, &c. &c. compared with corn; but he has not entered into the explanation of the natural causes which tend to determine the price of corn. He has led the reader, indeed, to conclude that he considers the price of corn as determined only by the state of the mines which at the time supply the circulating medium of the commercial world. But this is a cause obviously inadequate to account for the actual differences in the price of grain, observable in countries at no great distance from each other, and at nearly the same distance from the mines. (Malthus 1970b [1815], 209) 7 For his own part Malthus accounted for a ‘high comparative money price of corn’ in terms of its ‘high comparative real price, or the greater quantity of capital and labour which must be employed to produce it’ (210). Ricardo made precisely the same point. It is appropriate to keep in mind the strength of his objections to Smith’s position on corn pricing whereby ‘if you except corn and such other vegetables as are raised altogether by human industry, ... all other sorts of rude produce, cattle, poultry, game of all kinds, the useful fossils and minerals of the earth, &c. naturally grow dearer as the society advances in 349

RICARDO - THE NEW VIEW

wealth’ (1937 [1776] 217). ‘Why should corn and vegetables alone be excepted?’, he protested. ‘Dr Smith’s error throughout his whole work, lies in supposing that the value of corn is constant; that though the value of all other things may, the value of corn never can be raised’ (1951 [1817], I, 374). There is a closely related aspect of Smithian theory that next requires attention. I have referred to Smith’s apparent denial of a secular increase in the corn price due to the pressure of growing land scarcity. But more generally it was his position that a rise in the price of corn in consequence of government intervention in the corn trade, for example by the corn export bounty, can have no allocative implications; the argument presumes that an increase in the price of corn will be translated by way of the money wage rate to the prices of manufactured goods, leaving unchanged the relative profitability of agricultural activity. The model is laid out in the chapter ‘Of Bounties’, although the various constituent elements are established in the theoretical chapters of Book I: the money price of corn regulates that of all other home-made commodities. It regulates the money price of labour, which must always be such as to enable the labourer to purchase a quantity of corn sufficient to maintain him and his family either in the liberal, moderate, or scanty manner in which the advancing, stationary or declining circumstances of the society oblige his employers to maintain him. It regulates the money price of all the other parts of the rude produce of land, which, in every period of improvement, must bear a certain proportion to that of corn, though this proportion is different in different periods. It regulates, for example, the money price of grass and hay, of butcher’s meat, of horses, and the maintenance of horses, of land carriage consequently, or of the greater part of the inland commerce of the country. By regulating the money price of all the other parts of the rude produce of land, it regulates that of the materials of almost all manufactures. By regulating the money price of labour, it regulates that of manufacturing art and industry. And by regulating both, it regulates that of the complete manufacture. The money price of labour, and of every thing that is the produce either of land or labour, must necessarily either rise or fall in proportion to the money price of corn. (Smith 1937 [1776], 476–7) The outstanding theoretical conclusion is that an increase in the wage rate can be passed on by employers in the form of higher final prices in the manufacturing sector, and lower rents in agriculture (e.g. 816, 824). As for matters of policy, Smith’s fundamental conclusion is that government intervention designed to raise agricultural profits by means of a corn export bounty must fail in its objective because of the 350

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

effects of a higher corn price upon money wages and therefore manufacturing prices in general: ‘The real effect of the bounty is not so much to raise the real value of corn, as to degrade the real value of silver; or to make an equal quantity of it exchange for a smaller quantity, not only of corn, but of all other home-made commodities’ (476). Similarly, a few pages later, in the first two editions, we find an equally strong statement: ‘[Our country gentlemen] did not perhaps attend to the great and essential difference which nature has established between corn and almost every sort of goods. ... The nature of things has stamped upon corn a real value which no human institution can alter. No bounty upon exportation, no monopoly of the home market, can raise that value’ (482). This feature of Smith’s work was remarked upon repeatedly by contemporary and near-contemporary commentators. Critical reference to it is the most striking feature of the literature during the period 1776–1815. It seems scarcely surprising that late eighteenth- and early nineteenth-century critics were unaware of what Adam Smith wrote about land scarcity and its consequences given his treatment of corn pricing and its apparent denial of the possibility of real changes in the price of corn. I wish to elaborate upon this feature of the historical record. 8 In his Observations on ... National Industry, James Anderson developed a case – precluded as we have just seen by Smith – for government intervention to encourage domestic agriculture by means of an export bounty. In the course of an attempt to fix upon the precise magnitude of the subsidy required, Anderson utilized the principle of diminishing returns – from which he deduced the conception of rent as an intra-marginal surplus – on the grounds that it is the level of marginal costs that is relevant for the calculation. The exposition of the principle is striking (1779 [1777], 207–9). Anderson did not formally emphasize that his rent conception differed from that of Smith, and he did not base a full-fledged ‘Ricardian’ system upon it; but the principle of diminishing returns from which the rent concept derives is used nonetheless to counter Smith’s position on the bounty. In effect, the entire body of Smithian theory, according to which variations in the price of corn can be only nominal variations in consequence of the general effect upon prices as a whole, is turned down. The same case was presented by Anderson in his An Enquiry into the Nature of the Corn-Laws. Here he observed, in criticism of Smith’s general analysis of price in Book I, chapter 7 of the Wealth of Nations that ‘it is not ... the rent of land that determines the price of its produce but it is the price of that produce that determines the rent of the land’ (1777, 45n). 9 Accordingly, to fix a low corn price would mean not merely a fall of rent but a reduction in agricultural output despite the ‘popular objection ... that the price to the farmer is so high only on account of the high rents 351

RICARDO - THE NEW VIEW

and avaricious extortions of proprietors. Lower (say they) your rents, and the farmer will be able to afford his grain cheaper to the consumer.’ The notion of no-rent land at the (extensive) margin of cultivation is expounded with most impressive clarity in an account that has much in common with that regarded as ‘revolutionary’ nearly forty years later. On Anderson’s Enquiry, let us note, J.R. McCulloch later wrote that its publication ‘marks an important era in the history of economic science, from its containing the earliest explanation that is anywhere to be met with of the real nature and origin of rent’ (1845, 68). Malthus provides a conspicuous illustration of the present theme. In the second edition of his Essay on Population he added a chapter ‘Of Bounties on the Exportation of Corn’ – a subject of the ‘highest importance’ – which took strong exception both to various aspects of Smith’s analytical treatment of the issue, and to his policy conclusions (1803). Here he insisted inter alia that an export subsidy would in fact raise the agricultural profit rate and thus encourage expansion, denying the existence of any ‘great and essential difference’ between corn and other goods such as that emphasized by Smith: When Dr Smith says that the nature of things has stamped upon corn a real value, which cannot be altered by merely altering the money price; and that no bounty upon exportation, no monopoly of the home market, can raise that value; nor the freest competition lower it; it is evident, that he changes the question from the profits of the growers of corn in any particular country, to the physical and absolute value of corn in itself ... I certainly do not mean to say, that the bounty alters the physical value of corn, and makes a bushel of it support a greater number of labourers for a day, than it did before: but I certainly do mean to say that the bounty to the British cultivator does, in the actual state of things, really increase his profits on this commodity; and by thus making the growth of corn answer to him, encourages him to sow more than he otherwise would do, and enables him in consequence to employ more bushels of corn in the maintenance of a greater number of labourers. (Malthus 1803, 461–2) In the later Observations on the Effects of the Corn Laws (1814), formally designed as an impartial evaluation of the merits of protection and of free trade, the discussion again turned on Smith’s treatment of the corn-export bounty and in particular the special treatment accorded corn pricing: ‘I have always thought, and still think, that this peculiar argument of Dr Smith is fundamentally erroneous’ (1970a [1814], 967). The issue at stake, as in the Essay on Population, was Smith’s assertion that ‘corn is of so peculiar a nature, that its real price cannot be raised by an increase of its money-price; and that, as it is clearly an increase of real price alone, which can 352

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

encourage its production, the rise of money-price, occasioned by the bounty, can have no such effect’ (98). Here too Malthus insisted that the money wage rate does not rise proportionately with the corn price, and this because of the existence in the wage basket of home and imported goods other than corn (99). Moreover, the reaction of money wages to an increase in the price of corn occurs in consequence of a slow response of labour supply so that ‘corn and labour rarely keep an even pace together; but must often be separated at a sufficient distance and for a sufficient time, to change the direction of capital’ (100). In brief, ‘the price of corn does not immediately and generally regulate the prices of labour and all other commodities; and ... the real price of corn is capable of varying for periods of sufficient length to give a decided stimulus or discouragement to agriculture’ (106). It is worth adding that Malthus considered Smith’s theoretical position to fly in the face of the general theory of resource allocation : ‘it cannot be maintained without violating the great principles of supply and demand, and contradicting the general spirit and scope of the reasoning, which pervade the “Wealth of Nations” ’ (97). The Smithian approach excluded corn from those normal supply responses which flowed from changes in the relative demands for commodities and constituted an unjustifiable exception to ‘the operation of that principle, so beautifully explained and illustrated by Dr Smith, by which capital flows from one employment to another, according to the various and necessarily fluctuating wants of society’ (101). Francis Horner, in his well-known article for the Edinburgh Review of 1804 adopted (though with qualifications) the Smithian position in his analysis of the corn export legislation of that year, asserting that ‘the real price of corn will be maintained the same, notwithstanding a nominal variation’, and adding that ‘without a just apprehension of this fundamental truth, it is impossible to reason with accuracy upon the subject’ (1957 [1804], 106). Horner’s article was subsequently described by Ricardo, in his chapter dealing with ‘Bounties on Exportation’, as subject to the ‘common error which has misled Dr Smith, and, I believe, most other writers on this subject’, namely that ‘because the price of corn ultimately regulates wages, ... it will regulate the price of all other commodities’ (1951 [1817], I, 302). Indeed, Ricardo maintained that ‘perhaps in no part of Adam Smith’s justly celebrated work, are his conclusions more liable to objection, than in the chapter on bounties’ (I, 304). It is precisely in the context of corn legislation – it must be remembered that what is said of a corn export subsidy had clear implications for corn import duties as well – that we can perceive most clearly the differences between the Smithian and the Ricardian theoretical models. The point in question is reiterated in one form or another throughout the entire range of Ricardo’s writings. It is worth recalling, for example, Ricardo’s repeated denial, in debate with Malthus, that it is possible to identify a change in the corn price with a change in money value as Smith, we have seen, did: ‘Money I think only 353

RICARDO - THE NEW VIEW

falls in value, when it will exchange for less of all things; not when it will exchange for less of one thing, or of two things, or of a dozen things’ ([1820], II, 144). While Ricardo believed that corn was ‘the most important commodity of all others’ (II, 180), he nonetheless insisted that it must be treated merely as one among the others. Similarly, in the Essay on Profits (1815): ‘It has been thought that the price of corn regulates the price of all other things. This appears to me to be a mistake. If the price of corn is affected by the rise or fall of the value of the precious metals themselves, then indeed will the price of commodities be also affected, but they vary, because the value of money varies, not because the value of corn is altered’ (IV, 21n). The turning point in the transition between the Essay and the Principles came at the end of 1815 when Ricardo, having commenced the expansion of the pamphlet, wrote to James Mill: ‘I know I shall be soon stopped by the word price’. Ricardo then expanded his argument in terms relevant for our case: [My readers] must know that the prices of commodities are affected two ways one by the alteration in the relative value of money, which affects all commodities nearly at the same time, – the other by an alteration in the value of the particular commodity, and which affects the value of no other thing, excepting it enter into its composition. – This invariability of the value of the precious metals, but from particular causes relating to themselves only, such as supply and demand, is the sheet anchor on which all my propositions are built; for those who maintain that an alteration in the value of corn will alter the value of all other things, independently of its effects on the value of the raw material of which they are made, do in fact deny this doctrine of the cause of the variation in the value of gold and silver. (Ricardo 1951 [1815], VI, 348–9) Ricardo reacted quickly enough to what he regarded as the Smithian error in J.R. McCulloch’s essay on the national debt: ‘Your system proceeds upon the supposition that the price of corn regulates the price of all other things, and that when corn rises or falls, commodities also rise or fall, – but this I hold to be an erroneous system, although you have great authorities in your favour, no less than Adam Smith, Mr Malthus, and M. Say’ ([1816], VII, 105). And there can be no question that one of his main objectives in the Principles itself was to demonstrate the error of Smith’s position (common ‘to all the writers who have followed him’), according to which an increase in wages generates an increase in the prices of all commodities: ‘I hope I have succeeded in showing, that there are no grounds for such an opinion, and that only those commodities would rise which had less fixed capital employed upon them than the medium in which price was estimated, and that all those which had more, would positively fall in price when wages rose’ ([1817], I, 46. Cf. 302, 307, 315). 354

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

‘These results’, Ricardo insisted, ‘are of such importance to the science of political economy, yet accord so little with some of its received doctrines, which maintain that every rise in wages is necessarily transferred to the price of commodities’ (I, 61). III. THE RICARDIAN BREAK-AWAY ON THE THEORY OF PROFITS It is scarcely surprising that an economist who denied the possibility of a relative increase in the price of corn during the course of secular expansion should have been understood as neglecting the implications of increasing land scarcity – whatever he might have written about the matter in the context of new colonies. And it is also scarcely surprising that Ricardo should have referred to his theory of profit-rate determination in the first extant Ricardo documents touching on the ‘classical’ approach to the problem. I allude here to the famous letter to Malthus of 10 August 1813: ‘On further reflection I am confirmed in the opinion which I gave with regard to the effect of opening new markets or extending the old. I most readily allow that since the war, not only the nominal but the real value of our exports and imports has increased, – but I do not see how this admission will favour the view which you take of this subject. ... [Extension of trade] does not prove a general increase of profits nor any material growth of prosperity’ (VI, 93). In a second letter Ricardo explicitly stated the position that attention should be directed at agricultural productivity to understand trends in the rate of profit: That we have experienced a great increase of wealth and prosperity since the commencement of the war, I am amongst the foremost to believe; but it is not certain that such increase must have been attended by increased profits, or rather an increased rate of profits, for that is the question between us. I have little doubt however that for a long period, during the interval you mention [1793–1813], there has been an increased rate of profits, but it has been accompanied with such decided improvements of agriculture both here and abroad, – for the French revolution was exceedingly favorable to the increased production of food, that it is perfectly reconcileable to my theory. My conclusion is that there has been a rapid increase of Capital which has been prevented from shewing itself in a low rate of interest by new facilities in the production of food. (Ricardo 1951, VI, 94–5) This formulation, it must be emphasized, was regarded by Ricardo as an alternative to the Smithian view of profit-rate determination and rationalization of the profitrate trend, running in terms of ‘competition of capitals’, which he himself had 355

RICARDO - THE NEW VIEW

once held. Each of the extracts is followed by a statement insisting that an increase in general prices occurs only in consequence of a rising money supply (or velocity), while only a falling level of prices can induce a monetary inflow. It is most likely that the statements alluding to the profit rate and those relating to the monetary mechanism are connected. If this is so, Ricardo appears to be objecting to a formulation by Malthus very similar to that which was to appear later in the latter’s Principles of 1820 – namely that an increase in the value of aggregate output from given resources may be ‘occasioned by commerce’ and will be associated with a rise in profits of those engaged in the new or expanded trade, which extends to the general rate of return. It would seem to be this body of doctrine (which attributes the rising value of national product since 1793 and a correspondingly upward trend in the profit rate to the opening of new markets for British goods) with which Ricardo took issue in 1813, on the grounds that the expansion of the means of finance required to assure a higher level of prices and profits, of which Malthus was very confident, would not in fact be forthcoming. In short, the argument was seen to be inconsistent with Hume’s specie-flow mechanism. But Ricardo’s rejection of Malthus’s position implies the rejection of the standard Smithian view referred to above according to which ‘the acquisition of ... new branches of trade, may sometimes raise the profits of stock ... even in a country fast advancing in the acquisition of riches’, in consequence of the reduction in the pressure exerted by ‘competition of capitals’. It is in this light that one may best understand the objective of the Essay on Profits itself: ‘Profits of stock fall because land equally fertile cannot be obtained, and through the whole progress of society, profits are regulated by the difficulty or facility of procuring food. This is a principle of great importance, and has been almost over-looked in the writings of Political Economists. They appear to think that profits of stock can be raised by commercial causes, independently of the supply of food ’ (IV, 13n; emphasis added). And the central theme of the famous chapter on Foreign Trade in the Principles is important from the same standpoint: ‘extension of foreign trade’ in itself alters neither the ‘value’ of the national product nor the general return on capital ([1817], I, 128). Smith’s contention that such extensions – by attracting funds from alternative (domestic) projects thereby reducing the pressure of ‘competition between capitals’ – raise the general return on capital is firmly rejected: ‘I am of opinion, that the profits of the favoured trade will speedily subside to the general level.’ His concern was to provide an ‘independent’ refutation of Smith’s position rather than merely to assert the validity of his own theory of profit rate determination, according to which foreign trade can play a part only in the special case that real wage-goods prices are affected. Nonetheless, the latter proposition is strongly formulated: ‘It has been my endeavour to shew throughout this work, that the rate of profits can never be increased but by a fall in wages, and that there can be no permanent fall of wages but in consequence 356

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

of a fall of the necessaries on which wages are expended. ... Foreign trade, then ... has no tendency to raise the profits of stock, unless the commodities imported be of that description on which the wages of labour are expended’ (I, 132–3). The extent of the break-away from Smith may be gauged also by reference to Sir Edward West’s pamphlet of February 1815, famous for its formulation of the principle of diminishing returns (allowing both for the intensive and extensive margins) and for the use of the principle in the analysis of the declining secular return on capital. What concerns us most here is West’s rejection of the Smithian position, which runs in terms of increasing ‘competition of capitals’ on the grounds of fallacy of composition : Smith therefore attributes this decline in the rate of profit to increased competition. But the slightest consideration will detect the fallacy of this opinion. If the capital employed in one branch of trade alone be increased, doubtless the increased competition of the dealers in that branch will lower the price of their articles, and consequently the profits of those dealers. But why is the price lowered except because that article is now more abundant than others, and could not be sold without such diminution of price. But if the capital in all the different branches of trade, and consequently the quantity of all the articles of those respective trades be increased in the same degree, the same ratio between each and all the rest remains, and each article must sell for the same real price as it fetched before. If the competition be increased in any one article, it for the same reason is increased in all; and as it exists in the same degree in each, it cannot alter the real price of any one. It is only the relative alteration of the demand and supply which can increase the price, and here there is no such alteration (West 1815, 21). West’s implicit distinction in this passage between relative and absolute prices is of the first importance. It is conceded that there would be a temporary decline in general money prices in the case at hand which would, however, be corrected by a monetary inflow: ‘The money price of all articles would no doubt be diminished, and therefore the money-profits of stock; but this would not lower the real price of those articles, nor the real profits; even the money price would soon be raised to a level with the real price, by a favourable balance of trade and the consequent introduction of bullion’ (21). West’s argument bears a close resemblance to that of Ricardo, who also insisted that a disturbance affecting all commodities can have no real consequence, while a change in the general level of prices (in any single country) will be temporary only in the light of money flows. 357

RICARDO - THE NEW VIEW

IV. THE LAW OF MARKETS Ricardo’s objections to the Smithian analysis of the profit rate extended further. For, on his reading, the principle of ‘competition of capitals’ conflicted also with the law of markets to which, of course, Smith subscribed. This aspect of the criticism is made very clear in the chapter of the Principles specifically devoted to the effects of accumulation on the profit rate. The chapter is designed to demonstrate that the only cause of a falling profit rate is an upward trend in ‘wages’: ‘If the necessaries of the workman could be constantly increased with the same facility, there could be no permanent alteration in the rate of profits or wages, to whatever amount capital might be accumulated.’ Similarly, ‘whether these increased productions and the consequent demand which they occasion’ – in the course of accumulation – ‘shall or shall not lower profits, depends solely on the rise of wages’. Adam Smith’s ascription of the declining secular return on capital to ‘increased competition’ between capitalists is rejected as in conflict with the law of markets, for ‘at the same time that capital is increased, the work to be effected by capital, is increased in the same proportion’ (1951 [1817], I, 289, 292, 289–90). Smith’s contentions regarding the effects on profits flowing from ‘competition’ for labour are accepted by Ricardo but only in so far as concerns the short run. The weight of emphasis is upon ‘competition’ in product markets, and Smith’s approach is rejected entirely. Ricardo clearly conceived his chapter as a demonstration of the untenability of the notion of secular stagnation – or an approach to stationariness – independently of the considerations given pride of place in his own model. Accordingly, the only conceivable (permanent) constraint on growth is that deriving from a rising real cost of producing wages, which reduces profits to their minimum: ‘there cannot, then, be accumulated in a country any amount of capital which cannot be employed productively, until wages rise so high in consequence of the rise of necessaries, and so little consequently remains for the profits of stock, that the motive for accumulation ceases’ (I, 290). But provided the profit rate yields an adequate motive to accumulate, expansion of capacity will occur without check, since demand expands pari passu. In support of his general position in the chapter, Ricardo drew upon Say’s authority. Specifically, in support of the case against Smith’s doctrine relating to ‘competition of capitals’, Ricardo observed that ‘M. Say has ... most satisfactorily shewn, that there is no amount of capital which may not be employed in a country, because demand is only limited by production’. And regarding Smith’s statement to the effect that ‘the more disposable capitals are abundant in proportion to the extent of employment for them, the more will the rate of interest on loans of capital fall’, Ricardo, in the light of ‘M. Say’s principle’, asked rhetorically, ‘If capital to any extent can be employed by a country, how can it be said to be abundant, compared with the extent of employment for it?’ 358

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

V. THE INVERSE PROFIT–WAGE RELATIONSHIP The implications of Smith’s ‘error’ were profound. For if, as Smith believed, the impact of a rise in the price of corn and accordingly in the money wage rate were upon prices (in the manufacturing sector) and rents (in agriculture), there need be no fall in the profit rate during the course of secular expansion; capitalists could ‘pass on’ the burden imposed by land scarcity. The appearance of Robert Torrens’s famous Essay on the External Corn Trade in February 1815 provided an ideal opportunity for Ricardo to state this case, for Torrens repeated Adam Smith’s belief that increases in the price of corn are passed on to consumers in consequence of rising money wages. Ricardo simply insisted that Torrens ‘on this part of the subject appears to me defective, as I think that the price of commodities will be very slightly affected either by a rise or fall in the price of corn. If so every rise in the price of corn must affect profits on manufactures, and it is impossible that agricultural profits can materially deviate from them’ (VI, 213). A year later, however, he was pleased to find that Robert Torrens had apparently accepted the ‘Ricardian’ position on the main issues at stake. ‘Have you seen Torrens’ Letter to Lord Liverpool? – He appears to me to have adopted all my views respecting profits and rent; and in some conversation which I had with him a few days ago, he unequivocally avowed that he was now of my opinion, that the price of labour, arising from a difficulty of procuring food, did not affect the prices of commodities. He confessed that his former view on that subject was erroneous’ (Ricardo VII, 24). And a little later he repeated that Torrens had become ‘quite a convert to all what you have called my peculiar opinion on profits, rent, &c.’ (VII, 36). The general argument may be briefly summarized: Wages expressed in terms of ‘money’ of constant ‘value’ depend upon both the commodity wage and the price of wage goods. But in consequence of the rising real cost of producing wage goods – a reflection of diminishing returns – money wages will rise secularly. The secular increase in money wages – despite a reduced commodity wage – in turn has the effect of reducing the profit rate in manufacturing, since manufacturing prices are not raised by the wage increase: ‘Notwithstanding, then, that the labourer would be really worse paid, yet this increase in his [money] wages would necessarily diminish the profits of the manufacturer; for his goods would sell at no higher price, and yet the expense of producing them would be increased ’ ([1817], I, 102; my emphasis). The same conclusion is repeated in the chapter ‘On Profits’: ‘Supposing corn and manufactured goods always to sell at the same price, profits would be high or low in proportion as wages were low or high. But suppose corn to rise in price because more labour is necessary to produce it; that cause will not raise the price of manufactured goods in the production of which no additional quantity of labour is required. If, then, wages continued the same, the profits [of manufacturers] would remain the same; 359

RICARDO - THE NEW VIEW

but if, as is absolutely certain, wages should rise with the rise of corn, then [their] profits would necessarily fall’ (I, 110–11). All of Ricardo’s theoretical efforts were in fact devoted to demonstrating that a general increase in wages implies an inverse fall in the profit rate leaving general prices unaffected. (In the reconstruction, it would seem, the differential rent theory played a key but nonetheless supplementary role. The point here is Ricardo’s insistence that the existence of rent in no way disturbed the inverse wage–profit relationship; the importance of the rent doctrine should be evaluated within the context of the fundamental theorem on distribution.) It was Ricardo’s great achievement to demonstrate that what is at stake in profitrate analysis is proportionate wages; to this end he sought an appropriate measuring device in terms of which a rise in ‘money’ wages would imply a rise in the proportionate share of wages in the net product to be distributed between labourers and capitalists. On these terms it became quite clear how even a falling commodity wage – which characterizes the trend path of an economy subject to diminishing returns – might be consistent with a rising ‘money’ (Ricardian ‘real’ or ‘proportionate’) wage and thus entail an inverse profit-rate change. 10 VI. CONCLUSION To conclude: The ‘canonical’ model can be said to be present in the Wealth of Nations only on the basis of a very selective choice of texts. Even this selection is largely constrained to the special context of colonies and leaves much to be desired in terms both of internal consistency and consistency with central Smithian propositions relating to corn pricing. As for ‘influence’, there was none. On a balance of considerations I suggest that the model was a Ricardian construction insofar as concerns both pure logic and intellectual indebtedness. NOTES 1. Apart from an extraordinary appeal to authority, including his own (Samuelson 1978, 1430). The same deficiency characterizes Samuelson 1977. 2. ‘Irrelevant’ in the special sense suggested by George J. Stigler (1976). Stigler makes the point that if we seek to understand the scientific role played by a figure in the history of economic theory, what is relevant is how his work ‘appeared to his contemporaries’, for ‘science consists of the arguments and the evidence that lead other men to accept or reject scientific views’ (1976, 60). The ideas they may have intended to express are not relevant from this perspective. 3. See Hollander 1973, 157. 4. There is no obvious relationship between the Smithian analysis of ‘investment priorities’ and his ‘forecast’ of a secular downward trend in the average rate of profits in a growing economy. I have argued (1973, 281–3) that the relatively high profit rate available in the agricultural sector frequently emphasized in the Wealth of Nations reflects the extreme 360

ON SAMUELSON’S CANONICAL CLASSICAL MODEL

5. 6.

7.

8. 9.

10.

cheapness of land, which is regarded as typical in an economy at an early stage of development. Economic growth tends to alter the factor endowments such that the There agricultural profit rate falls relatively to that potentially available in manufacturing. does not seem to be any necessary implication for the profit rate trend since, in principle, increasing labour supplies might counterbalance the effect on profitability of increasing land shortage. See above, p. 344. Malthus also took Smith to task for the view ‘that all land which yields food must necessarily yield rent’, but recognized that Smith sometimes ‘contemplates rent quite in its true light’ (1970b [1815], 180). Cf. Smith’s discussion in the ‘digression on silver’ (particularly 1937 [1776], 191) regarding the determination of the corn price (also 36); and 217 regarding the relative prices of agricultural products other than corn to that of corn (also 174–6). I draw freely from Hollander 1979, ch. 1. Cf, also David Hume (1932 [1776], 311): ‘I cannot think’, he wrote to Smith, ‘that the Rent of Farms makes any part of the Price of the Produce but that the Price is determined altogether by the Quantity and the Demand’. See Hicks and Hollander 1977.

REFERENCES Anderson, James (1777) An Enquiry into the Nature of the Corn-Laws, Edinburgh: T. Cadell and C. Elliot. Observations on the Means of exciting a spirit of National Industry, Vol. II, Dublin: ——(1779) [1777] S. Price. Casarosa, Carlo (1978) ‘A New Formulation of the Ricardian System’, Oxford Economic Papers 30,1 (March): 38–63. Hicks, [Sir] John and Hollander, S. (1977) ‘Mr Ricardo and the Moderns’, Quarterly Journal of Economics 91, 3 (August): 351–69. Hollander, S. (1973) The Economics of Adam Smith, Studies in Classical Political Economy, Vol. 1, Toronto and Buffalo: University of Toronto Press. ——(1979)The Economics of David Ricardo, Studies in Classical Political Economy, Vol. 2, Toronto and Buffalo: University of Toronto Press. Horner, Francis (1957) [1804] ‘Observations on the Bounty upon Exported Corn’, Edinburgh Review, 5 (9) (Oct. 1804), Reprinted in The Economic Writings of Francis Horner in the Edinburgh Review 1802–6, ed. F.W. Fetter, New York: Kelley and Millman, 96–114. Hume, David (1932) The Letters of David Hume, Vol. II, ed. J.Y.T. Greig, Oxford: Oxford University Press, Clarendon Press. Levy, D. (1976) ‘Ricardo and the Iron Law: A Correction of the Record’, History of Political Economy 8,2 (Summer): 235–51. McCulloch, John R. (1845) The Literature of Political Economy, London: Longman, Brown, Green, and Longmans. Malthus, T.R. (1803) An Essay on the Principle of Population, 2nd edn, London: Joseph Johnson. ——(1970a) [1814] Observations on the Effects of the Corn Laws, 2nd edn, London: J. Johnson and Co; reprinted in Pamphlets of Thomas Robert Malthus, New York: Kelley, 93–131.

361

RICARDO - THE NEW VIEW

——(1970b) [1815]An Inquiry into the Nature and Progress of Rent, London: John Murray; reprinted in Pamphlets of Thomas Robert Malthus, New York: Kelley, 175–225. Principles of Political Economy, London: John Murray. ——(1820) Ricardo, David (1951–73) The Works and Correspondence of David Ricardo, ed. P. Sraffa (11 Vols), Vol. I: On the Principles of Political Economy and Taxation [1817], Vol. IV: Essay on the influence of a low price of corn on the profits of stock [1815], Vol Vl: Letters 1810–1815, Vol. VII: Letters 1816–1818, Vol. II: Notes on Malthus’s ‘Principles’ [1820], Cambridge: Cambridge University Press. Samuelson, Paul A. (1977) ‘A Modern Theorist’s Vindication of Adam Smith’, American Economic Review 67,1 (February): 42–9. ——(1978) ‘The Canonical Classical Model of Political Economy’,Journal of Economic Literature 16, (4) (December): 1415–34. Smith, Adam (1937) [1776] An Inquiry into the Nature and Causes of the Wealth of Nations, Modern Library edition, New York: Modern Library. Stigler, George J. (1976) ‘The Scientific Uses of Scientific Biography, with special reference to J.S. Mill’, in J.M. Robson and M. Laine (eds), James and John Stuart Mill: Papers of the Centenary Conference, Toronto: University of Toronto Press, 55–66. Torrens, Robert (1815) An Essay on the External Corn Trade, London: J. Hatchard. West, Sir Edward (1815) Essay on the application of Capital to Land, London: T. Underwood .

362

INDEX

abstinence, theory of 186–7, 306, 310 ‘adding-up’ cost approach, Smith’s 72, 101, 124, 161, 331–2, 334, 336 adjustment mechanism: Marshallian 155; Ricardian 140; wage 252–3; Walrasian 155–7 agricultural productivity 19–20, 24, 27–8, 62, 63–4, 65, 324 agricultural profits 19–21, 26–8, 34, 44–5, 60–4, 65, 130–2 allocation theory: in neo-classical economics 138, 204; Ricardian 4–5, 9, 139, 160, 161, 167–8, 182–3 Anderson, J. 103, 351–6 Arrow, K. 202–4, 213 Bailey, S. 285, 286, 296, 297–8, 301–2, 308–9; cost determination of price 305; measure of value 288–91, 298, 299; rent theory 109, 290–1 basket see wage basket Bentham, J. 323–37; and colonization 328–30; and falling rate of profit 326–30, 333–4; and falling rate of real wages 330, 333; his missing paper on profits 326, 337; as influence on Ricardo 323, 330, 333–7; and land scarcity 328–30, 331–3; and Smith 326–7, 331–3 Bharadwaj, D. 139 Black, R.D.C. 292 Blaug, M. 1–2, 5, 6, 8, 77, 292 bounty see subsidy (bounty) on corn Bowley, M. 292, 306

Brewer, A. 195 Bronfenbrenner, M. 2–4, 5–6, 11 Buchanan, J. 147, 349 Bullionism 96, 104–5 Cambridge school 91–2, 137–8, 142, 213, 219 ‘canonical classical model’ 221–2, 224, 342–60 capital accumulation: decelerating 268–9, 271, 274–8; Malthusian theory of 21–2, 25–6, 47, 118, 244; and profit rates 20, 22, 25–6, 27–8; and saving 118–19; Smithian 347; and wage rates 223–4 capital, Marshall’s theory of 157–9 Casarosa, C. 94, 117, 220, 221 ‘Catallactics’ 296 Churchman, N. 10 circular flow process 121–2, 161 class conflict 112–14 classical economics 135–40, 159–60, 168–9, 197; ‘canonical classical model’ 221–2, 224, 342–60 colonization 328–30, 345–6, 348 ‘competition of capitals’, Smith’s 101, 323, 326–7, 336, 344–7, 348–9, 356–7; conflicting with Law of Markets 122, 343, 358 constant real wages: and Malthus 227, 261–3; in Ricardo’s growth theory 96–8, 222, 242–3, 247–8, 253–4 corn: as both input and output 19, 21, 34–5, 44–5; effect of subsidies 100–1, 102, 103–4, 206, 350–1, 352-3;

363

INDEX corn ( cont’d ) restrictions on importation 21–3, 24, 26–7, 323–4; taxation 207; used as numéraire by Ricardo 29, 56, 72 Corn Laws 9–10, 93, 99, 132, 213 corn prices: and corn wages 48–55; Hume on 33; Malthus on 33, 103–4, 352–3, 354; Smith’s theory of 32–3, 83–4, 101–6, 129, 349–51, 353–4 corn wages 48–55 ‘corn–profit model’ 23–4, 44–56, 72–7, 79–86; ascribed to Torrens 79–82; Eatwell on 44–56; lost papers? 23, 60, 61, 63, 69; Moss’s belief in 127, 132; O’Brien’s view of 92–4; Sraffa’s interpretation of Ricardian profit theory as 44–6, 47–8, 56, 60–9, 79–86; use in prediction 93, 99; see also profit rate theory correspondence (Note: only the most frequently cited letters are indexed individually): letter from Ricardo to Malthus: 26 June 1814 21, 23, 61, 64, 66, 269; 25 July 1814 22–3, 61, 64, 66, 73, 74, 269–70; 11 August 1814 24, 66, 72, 73; 23 October 1814 25, 63, 256– 7; 18 December 1814 67, 255, 256, 271–2; 27 March 1815 48, 52–3, 54 Malthus to Horner 48, 54 Ricardo and Malthus: on demand– supply analysis 7, 145, 178, 212, 286; on methodology 97, 253–4; on profit rate theory 19–28, 48–55, 60–4, 72–4, 75–6, 323–4, 355–6; on scarcity 176; on Torrens 83–4; on wages 253–4, 255–9 Ricardo to McCulloch 99, 102, 276; Ricardo to Mill 98, 102, 354–5; Ricardo to Trower 20, 60, 65, 302 cost: rent distinct from 145–7, 180, 183–4, 207, 297; Smith’s ‘adding-up’ approach 72, 101, 124, 161, 331–2, 334, 336; see also opportunity cost; production cost cost approach to value: classical 136–7, 168; Ricardian 28–9, 139, 140, 141–5, 178–80, 301, 305; Walrasian 155–9 Cotterill, C.F. 291, 298, 305, 309 Cournot, A. 172, 173 De Quincey, T. 285–6, 291, 299–301, 304–5, 307–8, 311

de Vivo, G. 82–5, 170 demand: classical omission of 137, 169; derived 120–1, 137, 140, 148–9, 155, 195–9; as determinant of margin 204–11; and distribution 195–9; elasticity of 119–20, 175–6, 207; Malthus’s theory of 22, 24, 62, 287; for manufactured goods 21–2, 23, 66; Ricardian 119–22, 139–40, 142–3, 172–8, 195–9, 202–11; see also demand–supply analysis demand–supply analysis: and long-run wage determination 310; as received doctrine 178–9; Ricardo and Malthus correspondence on 7, 145, 178, 212, 286; Ricardo’s use of 107, 108, 119–22, 123, 124, 144–5, 178–80; Smith’s 120, 310; Torrens’s 178–9, 286; see also demand; supply derived demand 120–1, 137, 140, 148–9, 155, 195–9 differential rent: and land scarcity 109, 145–7, 183–4, 302; Malthusian theory of 28, 293, 349; see also rent theory diminishing returns: present in Smith 326, 342–3, 344–5, 349, 351; and profit rates 20, 27, 82, 85, 249–50, 256, 263; Ricardian 20, 27, 35–6, 51, 132, 219; Trotter’s 335–6; and wages 263, 266, 273 ‘Dissenters’ 283, 285–98 distribution: and demand 195–9; neo-classical interdependence with pricing 91–2, 117, 123–4, 187; relationship with expenditure 187, 195–8 distribution theory: classical 137; De Quincey’s 300–1; influence of Bentham 330, 333–7; Longfield’s 292–5; neo-classical 136; Ricardian 144, 184–5, 283–311; see also inverse wage–profit relation Dobb, M. 170, 195, 199, 284 Dumont, E. 334, 335 Dupuit, J. 173 Eatwell, J. 44–56 elasticity of demand 119–20, 175–6, 207 Essay on the Profits of Stock, Ricardo 28–33, 44–55, 259–61, 272–4; Table 45, 46, 48–55, 74–7, 113–14 expenditure, and distribution 187, 195–8

364

INDEX factor mobility 142–4, 159–60 factor and product market independence 139, 141, 180; Say 140–1, 179–80; Walrasian 140 factor substitution 180–1 falling rate of profit see profit rates falling rate of real wages see wages Fetter, F.W. 307–8, 310–11 food prices, as determinant of profits 21, 28–9 food production, and population growth 174–5, 208–10 French tradition of economics 136, 140–1, 168 Frisch, R. 137, 169 Garegnani, P. 64–9, 270 general equilibrium: in classical economics 167; in neo-classical economics 135, 137–8; O’Brien’s distaste for 92; in Ricardian economics 4, 139–40, 168, 179–80, 185; Say’s 180; Walrasian 140, 180 Gibbard, K. 268–78 Gossen, H.H. 173 Groenewegen, P.W. 79 growth theory: classical 226–39; and constant wages 96–8, 222, 227, 247–8, 253–4; and land scarcity 4, 235–9; Malthus’s 226; Mill’s 226; Ricardian 96–8, 125, 160, 185–7, 208– 11, 241–63; wage rates and population growth 117–18, 233–4, 235–9, 246–7 Hicks, Sir J. 208 Hicks, Sir J. and Hollander, S. 94, 117, 220, 221 High Price of Bullion, Ricardo 112, 323 Horner, F. 353 Hume, D. 33, 110–11, 112, 356 Hutchison, T.W. 10, 11 income effect 175 inflation 102, 105, 324 ‘invariable measure’ of value 107, 288–9, 291, 299 inverse wage–profit relation 4, 21, 127–32, 255–61, 286–97, 333–7; dismissed by Moss 127–9; influence of Bentham 330, 333–7; and land scarcity 106, 210–11; and Longfield 130, 285, 308; and Malthus 105, 130, 255–9, 285, 286–8,

303, 308; O’Brien on 93, 94–5, 104–6, 107–8; and Roncaglia 125; and Torrens 130, 285, 294, 295–6, 308, 359; see also profit rate theory; wages theory Italo-Cambridge school 168, 169–70, 185 Jevons, H.S. 174 Jevons, W.S. 171–4, 180–1, 284; chain of causation 171–2, 174; on Mill 136, 148–9, 155, 171; and pricing theory 136, 140, 168, 169, 173–4, 187 Kaldor, N. 92 Knight, F.H. 137, 169 labour, demand for see derived demand labour theory of value 51–2, 181–3, 202–4, 285, 291, 305–6 land: multi-use 147–8, 154, 160; single-use 204 land scarcity: Bentham on 328–30, 331–3, 335; and differential rent 109, 145–7, 183–4, 302; and growth theory 4, 235–9; Mill on 153–4; and Ricardo’s rent theory 108–9, 121–2, 145–7, 183–4, 205–6; Say on 146; Smith on 325–6, 347–8; as special case of inverse wage–profit relation 106, 210–11; see also rent theory; scarcity principle Langer, G.F. 79–82 Lauderdale, J.M. 284 law of markets 185; conflict with Smith’s competition of capitals 122, 343, 358; Mill’s 15–6, 22, 25–6, 82; Say’s 25, 82, 358 letters see correspondence Levy, D. 94 Lloyd, W.F. 286, 298, 306 long-run cost price determination 49, 123, 150–1 long-run wage determination 242, 254–5, 307, 310 Longfield, M. 4, 266–7, 292–6, 310; as ‘dissenter’ 284, 285, 286, 292–6, 301, 305–7; rent theory 293, 309; and Ricardo’s inverse wage–profit relation 130, 285, 308 McCulloch, J.R. 105, 291, 292, 353; and Ricardo 99, 105, 284–6, 299, 304, 307–9

365

INDEX Mallet, J.L. 308–9 Malthus, T.R. 6, 85–6, 303–4; aggregate demand theory 22, 24, 62, 287; capital accumulation theory 21–2, 25–6, 47, 62–3, 65, 118, 244; on corn prices 33, 103–4, 352–3, 354; correspondence on demand–supply analysis 7, 145, 178, 212, 286; correspondence on profit rate theory 19–28, 48–55, 60–4, 72–4, 75–6, 323–4, 355–6; correspondence on wages 253–4, 255–9; differential rent theory 28, 293, 349; misunderstanding of Ricardo 6–8, 62–4, 141–2, 224–5, 266; population theory 24, 25–6, 175, 208–9; and Ricardo’s concept of net revenue 224–5, 255, 301, 303–4; and Ricardo’s inverse wage–profit relation 105, 130, 255–9, 285, 286–8, 303, 308; and Ricardo’s theory of value 288, 303, 307; and Smith 103–4, 284, 356; and Torrens 85–6; wages theory 226–8, 233–9, 261–3, 276–7, 288, 307 manufactured goods 21–3, 27–8, 55–6, 82–3, 131–2; cost of production 142–3; demand for 21–2, 23, 66; prices 22–3, 45, 53, 73–4, 142–5 margin: endogeneity of 212; Ricardo’s demand determined 204–11 marginal productivity 180–1 marginal utility 159, 172–3, 176–8; concept missing in Smith and Say 176, 178, 188; as determinant of value 171–4, 177–8; Lloyd’s concept of 286, 306; non-paradigmatic nature of 136, 159, 188; Ricardo’s dawning recognition of 120, 172–3, 177–8 marginalist economics 167–8, 188; ‘marginalist revolution’ 135–40, 159–60, 169; see also neo-classical economics Marshall, A. 9, 167–88, 284; and marginal utility 120, 136, 159, 177–8, 188; neo-classical approach 9, 284; and Ricardo 107, 118–19, 120, 136, 139–40, 167–88, 213; theory of capital 157–9 Marx, K. 170, 185–7, 284 Meek, R.L. 297–8 Menger, C. 169, 173, 181 Merivale, H. 291–2, 307, 308

methodology: comparison of Smith and Ricardo 93; Ricardo’s simplification 116, 222, 243, 247–8, 250, 268, 276; Ricardo’s use of ‘strong’ cases 8, 30–2, 97, 112, 223, 254–5 Mill, J.S. 139–40, 148–55, 161, 284–6; and cost approach to value 137, 149–53, 160, 169, 307; and De Quincey 304–5; and demand for labour 148– 9, 155, 199; Jevons on 136, 171; relationship with Ricardo 106, 139–40, 149–55, 299, 307, 310–11; and scarcity of land 153–4; and surplus as source of profits 185–6, 306; wages theory 226–8, 236–9, 266 Mill, James 98, 105–6, 266, 291, 299, 307; Mill’s Law of Markets 15–6, 22, 25–6, 82 model building: Ricardo’s practice 92, 94–5, 98–100; see also ‘corn–profit model’ monetary theory 96, 110–12 money, quantity theory of value of 107 money supply 324, 332–3, 334–5, 336, 355–6, 357 money wages: and prices 25–6, 66–7; and profit rates 21, 23–4, 61–5, 67–9, 246–9; and real wages 246–9; and Ricardo’s theory of distribution 95; see also wages Morishima, M. 6 Moss, L.S. 127–32 ‘natural price’ 124, 142, 178 natural wage: classical approach 137, 168, 224; subsistence wage as 245, 251–2 neo-classical economics: continuity of Ricardo with 9, 120–4, 135–61; criticism of classical economics 135–40, 168–9; general equilibrium analysis 135, 137–8; interdependence of distribution and pricing 91–2, 117, 123–4, 187; see also marginalist economics neo-Ricardianism 169–70, 197, 202, 213 New View 4, 8, 9, 11, 241; Gibbard on 268–78; Peach on 9, 243–4, 245, 249–53, 255–8, 260, 263 Notes on Malthus, Ricardo 197–8, 261–3, 276–7

366

INDEX O’Brien, D.P. 91–114, 284, 299; on Ricardo’s inverse wage–profit relation 93, 94–5, 104–6, 107–8; and Ricardo’s model building 92, 94–5, 98–100; on Ricardo’s Smithian legacy 100–6 opportunity cost: neo-classical 136; Ricardian 139, 140, 141, 143, 181–3; Say’s 141, 148, 179–80 Owen, R. 82 Pasinetti, L.L. 8, 244, 250, 251, 252 Peach, T. 9, 10–11, 71–7, 243–61, 263, 265 Physiocrats 349 Political Economy Club 308–9 population growth: and food production 174–5, 208–10; Malthusian theory 24, 25–6, 175, 208–9; and wage rates 117–18, 233–4, 235–9, 246–7 prediction 98–100; use of ‘corn model’ for 93, 99 Prendergast, R. 85–6 price effect 268, 269–70, 273, 274–5, 277 prices: cost determination of 305; long run cost determination of 49, 123, 150–1; of manufactured goods 22–3, 45, 53, 73–4, 142–5; and money supply 324, 332–3, 334–5, 336; and money wages 25–6, 66–7 pricing: average-cost or marginalcost 51–3, 56; classical model 137, 168; neo-classical model 91–2, 117, 123–4; Ricardo’s cost approach to 136, 140, 141–5, 178–80, 301, 305; Smith’s ‘adding-up’ approach 72, 101, 124, 161, 331–2, 334, 336 Principles of Political Economy and Taxation, Ricardo 245–54, 274–6; ‘On Gross and Net Revenue’ 254–5; ‘On Machinery’ 112–13, 123, 149, 195; ‘On Profits’ 249, 275, 287, 359; ‘On Taxation on Wages’ 110, 118, 223, 251–2; ‘On Value and Riches’ 172–3, 300; ‘On Wages’ 109, 110, 222, 243–4; ‘Taxes on Raw Produce’ 109, 222–3, 250 production cost 28–9, 150–1 productivity: and corn prices 25; marginal 180–1; and profit rates 19–20, 24, 27–8, 62, 63–4, 65, 324; and surplus labour time 186–7

profit rates: and abstinence 186–7, 306, 310; and agricultural productivity 19–20, 24, 27–8, 62, 63–4, 65, 324; and capital accumulation 20, 22, 25–6, 27–8; and discovery of new markets 21, 256, 324–5, 348, 356–7; equalization of 26–8, 30–2, 107–8, 120, 181–2, 187; falling 8, 19–28, 29–36, 220–1, 246–9; ‘material rate’ 60–9; and money wages 21, 23–4, 61–5, 67–9, 246–9; and principle of diminishing returns 20, 27, 82, 85, 249–50, 256, 263; relationship between agricultural and general 19–21, 26–8, 34–5, 44–6, 60–4, 65, 130–2; and trade 21, 26–7; and waiting 118–19; see also inverse wage–profit relation profits: agricultural 19–21, 26–8, 34, 44–5, 60–4, 65, 130–2; food prices as determinants of 21, 28–9; maximization of 140, 142–3, 145, 167, 169; as residual 306, 327; Ricardo’s concept of 118–19; surplus as source of 119, 185–6 profits theory: Bentham’s falling 326–30, 333–4; Malthus’s 21–2, 25–6, 47, 62–3, 65, 118, 244; Marshall on Ricardo’s 118–19; Marx’s 185–7; Ricardian 19–36, 44–56, 60–9, 71–7, 118–19, 158–9; Ricardo and Malthus correspondence on 19–28, 48–55, 60–4, 72–4, 75–6, 323–4, 355–6; Smith’s 101–2, 323, 324–6, 342–3, 358; Sraffa on Ricardo’s 44–6, 47–8, 56, 60–9, 79–86; Stigler on Ricardo’s 35–6; Torrens’s 79–86; Walrasian 157–9; see also ‘corn–profit model’; inverse wage–profit relation proportionate wages 105, 360 ‘rational foundation’ of profit theory 44–5, 46, 47–8, 69, 79–86 Read, S. 286, 296, 297–8, 305, 306, 307, 310 real wages: assumption of constant 96–8, 222, 242–3, 247–8, 253–4; and diminishing returns 263, 273; and money wages 246–9; secularly falling 8, 229, 231–3, 244, 246–9, 255–9; see also inverse wage–profit relation rent: differential 28, 109, 145–7, 183–4, 293, 349; distinct from costs 145–7, 180, 183–4, 207, 297

367

INDEX rent theory: Anderson’s 351–2; Bailey’s 109, 290–1; Longfield’s 293, 309; Malthus’s differential 28, 293, 349; Mill’s 153–4; Ricardian 28–9, 108–9, 121–2, 145–7, 183–4, 203–7, 302–3; Senior on 109, 302, 308; Smith’s 108, 302–3, 349; see also differential rent; land scarcity revenue, Ricardian 219, 224–5, 254–5, 301, 303–4 Ricardian economics: canonical classical model’ 221–2, 224, 342–60; as detour 114, 116; and dual development of economic theory 283–4, 286, 309–11; interpretation of 95–8, 161, 171, 220–2, 225, 241–4, 265–7; misinterpretation of 219, 222–3, 301; neo-classical? 9, 91–2, 117, 120–1, 122–4, 135–61; revitalized by Marx 284; and the Ricardians 298–301; Smithian continuity 109–12, 114; supposed demise 283–5, 311; ‘Walrasian’ claims for 102, 109, 117, 124 Robinson, J. 137–8 Roncaglia, A. 116–25, 138 Samuelson, P.A. 161, 202–4, 220, 221–2, 267, 342–60; ‘canonical classical model’ 221–2, 342–60; income distribution and demand 197–8, 213 saving 267, 306, 310 Say, J.B. 120–2, 140–9, 176–8, 179–80, 298; circular flow 121–2, 161; and marginal utility 141, 176–8, 188; opportunity cost 141, 148, 179–80; and Ricardo 25, 120–2, 140–9, 161, 179–80; Say’s Law 25, 82, 122, 358; and Smith 120, 177, 284; and Walras 121, 137, 140–1, 168 scarcity 123, 124–5, 147, 176–7, 184, 205–6; see also land scarcity Schumpeter, J.A. 167, 169, 284, 288, 290–1, 292, 299 Scrope, G.P. 286, 296, 305, 306, 307, 310 Seligman, E.R.A. 292 Senior, N.W. 286, 301, 302, 307; abstinence 186, 310; and Ricardo’s rent theory 109, 302, 308 Shove, G. 136, 139 Sismondi, J.C.L. S. de 349

Smith, A. 93, 100–6, 323–7, 342–60; ‘adding-up’ cost approach 72, 101, 124, 161, 331–2, 334, 336; and Bentham, J. 326–7, 331–3; and colonization 345–6, 348; corn prices theory 32–3, 83–4, 101–6, 129, 349–51, 353–4; influence on Ricardo 100–6, 109–12, 114, 120, 129, 176–7, 336; lacked concept of marginal utility 176; and land scarcity 325–6, 347–8; law of markets not followed 122, 343, 358; and Malthus 103–4, 284, 356; ‘natural price’ 142; O’Brien on Ricardo’s reaction to 100–6; passing on of wage-rate increases 22–3, 100, 285, 289–90, 335, 350–1, 359; and principle of diminishing returns 326, 342–3, 344–5, 349, 351; profits theory 101–2, 323, 324–6, 342–3, 358; rent theory 108, 302–3, 349; resource allocation 310; and supply and demand analysis 120, 310; on taxation 223; viewed by contemporaries 102–4, 347, 351–5, 357; see also ‘competition of capitals’, Smith’s Sraffa, P. 44–6, 60–9, 79–86, 202–4, 213; Bentham’s missing paper on profits 326, 337; and corn as both input and output 19, 21, 34–5, 72, 86; inference of lost ‘corn model’ papers 23, 60, 61, 63, 69; ‘one-legged’ pricing models 202–4, 213; and Ricardian profit theory as ‘corn profit model’ 44–6, 47–8, 56, 60–9, 79–86 Stigler, G.J. 35–6, 219–25, 241–4, 254–5, 265–7; on textual exegesis 6, 98, 220–2, 225, 241–4 subsidy (bounty) on corn 100–1, 102, 103–4, 206, 350–1 subsistence wage 250–5; classical 136, 137, 168; as natural wage 245–6, 251–2; in Ricardian distribution theory 8, 35–6, 219, 222–3, 286; and taxation 110, 222–3, 250; use in ‘prudential wage’ theory 239 substitution effect 175 supply: and long-run cost price determination 150–1; Ricardo’s emphasis on primacy of 7–8, 145, 178–9, 211–12, 213; see also demand–supply analysis

368

INDEX surplus: concept of 185–7; as source of profits 119, 185–6, 306; and subsistence wage 254–5 taxation 109–10, 148, 184, 207, 222–4 Thompson, T.P. 286, 301, 302, 307, 308, 309 Tooke, T. 308–9 Torrens, R. 79–86, 295–6, 307, 308–9; demand–supply analysis 178–9, 286; and inverse wage–profit relation 130, 285, 294, 295–6, 308, 359 trade: effect of restrictions/ extensions 21–3, 24, 26–7, 323–4; Ricardo’s comparative cost approach 185; and Ricardo’s cost price theory 144, 148 Trefusis, D. 1 Trotter, C. 335–6 Trower, H. 20, 60, 65, 209 Turgot, A.R.J. 9, 284 Twiss, T. 297 utility: final 173; marginal see marginal utility; total 120, 172–3, 177–8 value: Bailey’s measure of 288–91, 298, 299; classical theory of 137, 149–50, 168, 169; cost approach to 136, 140, 141–5, 178–80, 301, 305–6; De Quincey’s theory of 300, 304–5; demand and supply approach to 178–9, 301, 305; ‘invariable measure’ of 107, 288–9, 291, 299; labour theory of 51–2, 181–3, 202–4, 285, 291, 305–6; Longfield theory of 292–4; and marginal utility 171–4, 177–8; measuring 33, 36, 129, 276, 288, 303; paradox of 176–7, 300; quantity theory of value of money 107; Ricardian theory of 171–4, 202–4, 285–97, 300, 305–7; Say’s theory of 140–1, 179–80; Smith’s theory of corn 100–1, 353; Torrens on 83, 294 Viner, J. 8 wage basket 26–8, 67–9, 268–72, 274–6; constant 268, 269–70, 272, 274–6;

mixed 51, 53–4, 67–9, 131–2, 269–72, 273; Sraffian assumption of corn only 46, 69, 268–9 wage rates: and capital accumulation 223–4; demand–supply determination of 343; as endogenous variable 117–18; and population growth 117–18, 233–4, 235–9, 246–7 wages: adjustment mechanism 252–3; Bentham’s falling 330, 333; corn 48–55; determined by food prices 21; and diminishing returns 263, 266, 273; long-run 242, 254–5, 307 310; Malthus-Mill prudential falling 235–9; money 21, 23–4, 25–6, 61–5, 66–9, 95, 246–9; proportionate 105, 360; Ricardian constant 96–8, 222, 242–3, 247–8, 253–4; Ricardian falling real 8, 229, 231–3, 244, 246–50, 255–9, 265–7, 330; Ricardian rising 23–4, 229–30; Ricardian variable 117–18, 185–7, 220–2, 225, 244; Ricardo and Malthus correspondence on 253–4, 255–9; Smith’s passing on of increases 22–3, 100, 285, 289–90, 335, 350–1, 359; Stigler on 242, 254–5; and taxation 109–10, 222–4; see also money wages; real wages wages fund 136, 148–9, 155, 168, 199 wages theory: classical 226–39; of J.S. Mill 226–8, 236–9, 266; of Malthus 226–8, 233–9, 261–3, 276–7, 288, 307; prudential wage path 235–9; Ricardian 226–33; Ricardo’s subsistence 8, 110, 219, 222–3; see also inverse wage–profit relation waiting 118–19 Walras, L. 155–9, 160, 187; admiration for Say 121, 137, 140–1, 168, 180; and classical pricing theory 137, 140, 155–9, 160, 168, 187; and subjective pricing theory 169, 173 West, Sir E. 302, 343, 357 Whateley, R. 296, 297, 305–6 Wicksell, K. 202–3, 205, 210, 212 Winch, D. 266

369