Interpreting Ricardo

Citation preview

This book proposes a textually-based, non-mathematical reconstruction of the substance and evolution of David Ricardo's thought on the interrelated topics of value, distribution and accumulation. It also provides a detailed summary of, and critical commentary on, the vast secondary literature. Among the conclusions reached, the author rejects Sraffa's influential 'corn model' interpretation of Ricardo's early writings, and he questions the alleged similarity between the work of Ricardo and Sraffa. He reaffirms the 'old' interpretation that Ricardo's primary concern was to demonstrate the link between worsening conditions of production in agriculture and 'permanent' reductions in the general rate of profit, and he opposes the Hollander and Hicks view of Ricardo's treatment of wages. He also addresses the questions of the role and significance of Ricardo's labour theory of value. The neoclassical interpretation of Marshall and Hollander is also rejected. Dr Peach argues that Ricardo's work has been persistently, and sometimes wilfully, misinterpreted, and that this state of affairs can be remedied only through an attempt to understand Ricardo's writings in his terms, taking account of his objectives.

INTERPRETING RICARDO

INTERPRETING RICARDO TERRY PEACH Department of Economics University of Manchester

CAMBRIDGE UNIVERSITY PRESS

CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo, Delhi Cambridge University Press The Edinburgh Building, Cambridge CB2 8RU, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9780521119757 © Cambridge University Press 1993 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1993 This digitally printed version 2009 A catalogue record for this publication is available from the British Library Library of Congress Cataloguing in Publication data Peach, Terry. Interpreting Ricardo / Terry Peach. p. cm. Includes bibliographical references and index. ISBN 0 521 26086 8 (hardback) 1. Ricardo, David, 1772-1823. I. Title. HB103.R5P43 1993 330.15'3-dc20 92-9219 CIP ISBN 978-0-521-26086-2 hardback ISBN 978-0-521-11975-7 paperback

To My Mother

Contents

Preface Acknowledgements

page xi xiv

1

Interpretations of David Ricardo

2

From bullion to corn: the early writings The early monetary writings Ricardo's new 'theory' The prt-Essay correspondence The Essay on Profits The pre-Essay agricultural analysis: further considerations The post-Essay correspondence Conclusion

3

4

39 40 49 55 68 76 81 85

The falling rate of profit, wages and the law of markets

87

Diminishing agricultural returns, the falling rate of profit and the Corn Law debate Wages The law of markets Conclusion

88 103 131 143

The labour theory of value (I) Before the Principles The first edition of the Principles From the first to the second edition of the Principles Conclusion

5

i

The labour theory of value (II) Between the second and third editions of the Principles The third edition of the Principles The final writings Conclusion

ix

145 145 154 172 186

189 189 211 224 238

CONTENTS

6 The appropriation of Ricardo Utility Demand Product pricing: the role of'demand and supply' Ricardo and the 'Say tradition': 'opportunity cost' and 'imputation' Ricardian 'general equilibrium' analysis Ricardo and Sraffa

241 241 250

256 263 277 286

7 Concluding remarks

294

Bibliography Index

3°4

Preface

I have two main objectives in this book. First, to propose a reconstruction of the substance and evolution of Ricardo's central views on value, distribution and accumulation. Secondly, to give a critical commentary on the interpretative literature. The appearance of another study of Ricardo's work may not command widespread approval. Professor D. P. O'Brien has censured (unnamed) historians of economic thought for 'endlessly quibbling about what Ricardo might have said' (O'Brien 1990), which would seem to be his doleful characterisation of recent scholarly debate (in which, curiously, he has been a participant). He may not be alone in holding this view. Yet I will offer no apology for this contribution. If, as I believe, Ricardo's impostors continue to lurk in the pages of learned journals, monographs, dictionaries of economics and textbooks, they must be exposed as such, regardless of the irritation, or inconvenience, that this may bring. The stimulus for much of the recent interpretative controversy was provided by Samuel Hollander's Economics of David Ricardo (1979), on the basis of which he has constructed a sweeping, revisionist account of the history of economic thought, according to which virtually all roads lead to (his version of) 'mainstream' economics. When I reviewed Hollander's book (Peach 1981), I commented that it would have a salutary effect in stimulating a reassessment of orthodox interpretations. My position remains that Professor Hollander has performed a valuable service by challenging received opinion. There is also (as Malthus said of Ricardo's Principles) much collateral material in Hollander's book with which I agree; and in its breadth of coverage, it is a truly impressive work (the scope of my own contribution is modest in comparison). However, on many central issues it is my conviction that Professor Hollander's view of Ricardo is radically flawed. Thus, to take some areas of

Xll

PREFACE

disagreement, I will argue that Ricardo was not primarily concerned to develop a general framework of analysis; that he did attribute importance to his falling rate of profit thesis as an argument against agricultural protection; that he was not a committed 'new view' growth theorist; that he was susceptible to the 'Ricardian Vice'; that the 'Ricardo-Say-Walras' tradition of'factor' pricing is a hoax; that Ricardo's work does not incorporate a fundamentally important 'core' of general equilibrium analysis; and, more generally, that Ricardo's work has little in common with later 'neoclassical' theory. For all my criticisms of Professor Hollander's work, I am not advocating an unqualified return to orthodoxy. I can at least agree with Hollander's conclusion that Piero Sraffa's widely accepted 'corn model' interpretation is textually unsupported (a criticism which I extend to Sraffa's interpretation of Ricardo's invariable standard). I also reject currently popular interpretations of Ricardo's labour theory of value, including the views that the theory appealed to Ricardo merely because of its supposed empirical relevance; that it was no more than a cost of production theory; and that there was no essential difference in the treatment of value over the successive editions of the Principles. Moreover, although I shall argue that the traditional interpretation of Ricardo as a natural wage theorist has more to commend it than competing interpretations, I accept that his writings do contain evidence, often conveniently ignored or distorted, in favour of the 'new view' interpretation. In this last instance, as in others, orthodox teaching invited a response such as Hollander's. Its sharpness of focus pushed Ricardo's qualifications and inconsistencies so far into the background that they became almost invisible. Another aspect of this study which conflicts with recent teaching is a portrayal of Ricardo as a man capable of quite startling inconsistencies and errors. These are evident in his treatment of wages and, even more so, in his treatment of value. Here, it should perhaps be stated that I did not set out to discredit Ricardo, but neither was I prepared to fudge a pleasing consistency when it became clear that none existed. The plan of the book is as follows. Chapter i presents a selective review of the literature relevant to the topics discussed in future chapters, and an outline of my own interpretation. Chapter 2 concentrates on Ricardo's early writings, from 1809 to 1816.

PREFACE

Xlll

Chapter 3 considers elements of Ricardo's treatment of distribution and accumulation, with the emphasis on his later writings. Chapters 4 and 5 trace the evolution of his treatment of the labour theory of value. Chapter 6 considers the similarities and dissimilarities between Ricardo's work, 'neoclassical' economics and 'Sraffian' economics. Chapter 7 contains some concluding remarks. NOTE ON REFERENCES

References of the form (iv, p. 40) are to The Works and Correspondence of David Ricardo, edited by Piero Sraffa (with the collaboration of M. H. Dobb). Capitalised roman numerals indicate the relevant volume.

Acknowledgements

I have contracted so many debts of gratitude over the long gestation period of this book that it is almost invidious to single out the role of only some individuals. Yet there are three people whose contributions were sufficiently important that they cannot be allowed to go unmentioned: Andrew Glyn (undergraduate tutor and doctoral supervisor, 1977-8), Walter Eltis (doctoral supervisor, 1978-82), and Ian Steedman (colleague, 1980 onwards). To them, and to the many others who gave their encouragement and advice, I express my thanks.

xiv

CHAPTER I

Interpretations of David Ricardo

My purposes in this chapter are to review selectively the literature bearing directly on the principal issues discussed in the following chapters, and to give an overview of my own interpretation. In chapter 2, I investigate the evolution of Ricardo's early treatment of value, distribution and accumulation, from the time of his first published writings, when he was embroiled in the 'bullion controversy', to the period immediately following his Essay on Profits (1815), when he was obliged to defend his arguments in favour of a liberated corn trade. The contemporary literature on this period of Ricardo's work is dominated by discussion of Piero Sraffa's 'corn model' interpretation (as I shall call it), presented by him in the Introduction to the first volume of Ricardo's Collected Works (published in 1951). Previously, there had been relatively little attention paid to Ricardo's early writings, with interest having centred on the later Principles of Political Economy and Taxation (here-

after abbreviated to Principles).1 What had been said, of any significance, may be summarised briefly. It was generally agreed that, after his initial preoccupation with monetary issues, Ricardo's concern with distribution as a separate topic had arisen in response to new and highly controversial proposals to restrict the importation of cheaper foreign corn.2 This has remained the popular view ever since. To take Edwin Cannan's version of Ricardo's early argument, 'restrictions on importation raised the price of food . . . the price of food regulated the wages of labour, and . . . cheap labour was necessary for high profits' (Cannan 1893, p. 164). In other words, the Corn Laws would result 1

2

It is pertinent to note that there was far less publicly available material prior to Sraffa's edition of the Collected Works. See, for instance, Cannan (1893, p. 148), Gide and Rist (1948, p. 160), J. Hollander (1904) and Mitchell (1949, p. 136). Cf. Marx ([1862-3] 1969, p. 236).

2

INTERPRETING RIGARDO

in higher money wages and, in consequence, a lower rate of profits. There was, however, a 'vulnerable point' in the early Ricardo's argument, remarked upon by Jacob Hollander as follows: 'It was impossible [for Ricardo] to prove that a rise in wages was the exclusive cause of a fall in profits [Hollander's version of Ricardo's doctrine], if it were true that a rise in wages necessarily occasioned a rise in prices' (Hollander 1904). That is, Ricardo's implied adherence to an 'adding up' treatment of pricing, as it has come to be known, was allegedly an obstacle to the prosecution of his anti-Corn Law case.3 In lectures delivered during the 1930s, Wesley C. Mitchell indirectly suggested a way in which Ricardo might have circumvented his pricing dilemma. Commenting on the Essay on Profits (1815), Mitchell averred that the 'whole discussion' of agriculture was conducted 'on a commodity basis', with output and the entire capital outlay consisting of so many quarters of wheat (Mitchell 1949, pp. 141, 174). Now, one facet of Ricardo's argument in the Essay was that, if corn imports were prohibited, any extra demand for corn would have to be met either from the more intensive cultivation of land already under the plough, or from new land which was comparatively less fertile or less advantageously situated. But, whichever strategy was adopted, the attempt to expand output would eventually run up against diminishing returns. The implication of Jacob Hollander's interpretation is that if Ricardo had attempted a similar analysis in his earlier writings, when reasoning in money terms, the higher corn price (reflecting diminishing returns) would certainly increase money wages, because bread was the staple diet of the labourers, but the impact on agricultural profitability would be indeterminate, perhaps even non-existent, because corn prices would rise again as farmers passed on their higher labour costs. However, if the pre-Essay Ricardo had adopted the assumption of capital-product homogeneity (a point on which Mitchell did not speculate), he would have escaped the indeterminacy. For if agricultural output and capital (including wages) consist of the same commodity, corn, diminishing returns (implying 3

The 'adding up' theory is often attributed to Adam Smith, seemingly for good reason if attention is confined to Book i, chapter vii of the Wealth of Nations. However, elsewhere in his treatise, Smith suggested various interdependencies between outputs, natural prices and the natural rates of wages, profits and rent. Although he did not reach the stage of combining his remarks into a single, polished doctrine, the 'adding up' description of his position is misleading, nevertheless.

Interpretations of David Ricardo

3

more corn capital per quantity of corn output) must result in a lower rate of profit. Mitchell's interpretation made no discernible impression on mainstream opinion. The same certainly cannot be said of Piero Sraffa's later variation on the 'corn model' theme. Prompted by discoveries made in the course of his own theoretical work (see below, p. 289), Sraffa suggested that the pre-Essay Ricardo had indeed formulated a model in which corn was the output and sole input for the agricultural sector of a closed economy: the sole input by virtue of the assumption that production is effected by unassisted labour and seed corn, with corn as the only wage good both in agriculture and in other sectors of the economy. This model had supplied Ricardo with the 'rational foundation' for the 'principle of the determining role of the profits of agriculture', supposedly articulated in a letter of 1814 with the words, 'it is the profits of the farmer that regulate the profits of all other trades'. The logic of such a model was explicated by Sraffa as follows: 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 [if only as the wage-good for their workers]. 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. (1, p. xxxi, emphasis in original) According to this interpretation, when Ricardo wrote that agricultural profits 'regulate' other profits, he had meant that they uniquely determine them in a precise, mathematical sense; he had known the assumptions required in order to validate his pronouncement, and it was the 'primitive "agricultural" form' of his analysis which had prevented him from recognising the problems created by his subscription to 'the generally accepted ['Smithian'] view that a rise in corn prices, through its effect upon wages, would be followed by a rise of all other prices' (1, p. xxxiii). Perhaps surprisingly, given the alleged importance of the 'corn model', Sraffa revealed that it was 'never stated by Ricardo in any of his extant letters and papers', although, on the basis of various items adduced by way of indirect textual support, he claimed that Ricardo 'must have formulated it' either in lost papers or in

4

INTERPRETING RICARDO

conversation (i, p. xxxi). On that point, Sraffa differed from Mitchell, who had suggested that the analysis in the Essay was itself based on 'corn model' reasoning. Sraffa's view of the Essay was rather that its numerical examples 'reflected' the corn model approach: 'both capital and the "neat produce" are expressed in corn, and thus the profit per cent is calculated without need to mention price' (i, p. xxxii). Given his admission that a record of the fully specified 'corn model' is inextant, he presumably meant that the Essay had the appearance of a 'corn model' but that, for some undisclosed reason, the appearance was misleading. Sraffa's interpretation became the received wisdom, although, as so often happens, the subtleties of the original statement were often overlooked; hence the claim, identical to Mitchell's, that the Essay itself contained a 'corn model'. 4 From the early 1970s onwards, however, a powerful challenge to this entrenched orthodoxy has been delivered by Professor Samuel Hollander, who argues that the 'corn model' interpretation must be abandoned (although he allows that there is an extant statement of the model in Ricardo's correspondence). According to Professor Hollander, Ricardo's general model of distribution turned on the interrelationships between money wages, prices and other distributive variables. And, with a further thrust at received opinion, he questions whether Ricardo's interest in distribution had been stimulated by the Corn Law controversy (cf. Tucker 1954). On Hollander's alternative reading, 'the distinctive Ricardian analytical construct arose from dissatisfaction on purely logical grounds with the Smithian analysis of profits' (1979, p. 647, emphasis in original). It was therefore through a detached contemplation of doctrine that Ricardo had developed a general model in which higher money wages, however caused, would result in a lower rate of general profitability. But how had Ricardo overcome the supposed obstacle created by his 'adding up' treatment of pricing, to which attention had been drawn by Hollander's namesake at the beginning of the century? On this, S. Hollander suggests that it was Ricardo's application of quantity-theory reasoning which had yielded the insight that a rise in money wages could not be passed on in higher prices. This is commended (by Professor Hollander) as 'a very attractive reconstruction' of Ricardo's early thought (1979, p. 118). 4

See, for example, Barkai (1965), Eatwell (1975b), Dasgupta (1985, p. 27), Gillman (1956), Groenewegen (1972), Hartwell (1971, p. 18), Hicks (1965, chapter.4) and (1985, p. 312), and Ong (1983).

Interpretations of David Ricardo

5

The attraction of Hollander's interpretation has not been widely celebrated. Bharadwaj (1983a) (1983b), de Vivo (1985) (1987a) (1987b), Eatwell (1975a), Garegnani (1982), Langer (1982), O'Brien (1981) and Roncaglia (1982), to name only the principal contributors, have been unanimous in their rejection of Hollander's case. And, according to one of these critics, 'it would be fair to say' that recent debate over the legitimacy of the 'corn model' reading has actually strengthened its appeal (Garegnani 1990, p. 296). Quite remarkably for a model of which there is no surviving record, its erstwhile existence is beyond reasonable doubt. That, at least, appears to be the message transmitted by several modern writers, particularly those who champion Sraffa-based alternatives to 'neoclassical' economics. I argue, however, that the textual evidence cited in favour of Sraffa's reading is plainly inadequate to sustain his interpretation (cf. Faccarello 1982).5 This is not to say that I endorse Professor Hollander's alternative interpretation, or that I believe the many criticisms of Hollander to have been entirely devoid of merit. On the contrary, the argument put by Garegnani (1982), that Hollander fails to supply a convincing alternative version of Ricardo's early theory, is, I argue, fundamentally correct. Where Professor Hollander (following Tucker) stands on firmer ground is with his claim that the origins of the early theory cannot be traced unambiguously to the debate over agricultural protection. I also develop a constructive interpretation of Ricardo's early work. Rather than presenting his adherence to the 'adding up' treatment of price as a theoretical inconvenience, or spare wheel in his analytical system, I suggest that it was integral to his early 'agricultural' analysis. On my reading, it was because he believed that all prices rise with the corn price that he could confidently 'predict' the decline in agricultural profitability when more (heterogeneous) inputs were required to produce a certain corn output: the resulting price changes would be common to both output and 5

I do not consider the so-called 'further evidence' in support of Sraffa's interpretation, marshalled by Langer (1982), de Vivo (1985) and Prendergast (1986b), among others. The generic argument of these authors seems to be that the use of anything resembling a 'corn model', by anyone acquainted with Ricardo, is evidence that Ricardo used the same model. Porta (1986) has commented that these 'contributions fail to give any conclusive evidence that Ricardo held the corn-ratio theory of profits at any stage; so the contributions are designed to bring consolation to those who already believe in the neo-Ricardian ['corn model'] interpretation of the formative steps of Ricardo's profit theory' (emphasis in original). I agree.

O

INTERPRETING RIGARDO

inputs, irrespective of the physical nature of those inputs. By the time of the Essay, with Ricardo having abandoned the 'adding up' approach, he proceeded in his 'agricultural' analysis by assuming that all prices remain unchanged as more inputs are applied per unit of corn output. These inputs are, quite explicitly, heterogeneous, but it is their unchanging valuation in wheat, not their wheat composition, which could give a misleading 'corn model' appearance to the analysis. The Essay neither contains, nor truly reflects, the elusive 'corn model'. Chapter 3 contains three sections which have, as an overlapping focus, Ricardo's core theory of'permanent' movements in profitability during the accumulation process. The first section deals with the genesis, orientation and, as perceived by Ricardo, the applied relevance, of his 'nature' analysis of distribution-cum-accumulation, the second with his treatment(s) of wages, and the third with the 'law of markets'. I will preview each section in turn. For most nineteenth-century commentators, and the great majority of later ones, it has been common knowledge that Ricardo was particularly concerned to elaborate the dynamic course of wages, profits and rent under the influence of progressively diminishing agricultural returns, that he envisaged a declining general rate of profit in the absence of a free trade in corn, and that the 'prediction' of a falling rate of profit was central to his case against agricultural protection. In recent times, each element of this composite orthodoxy has been disputed. Take, first, the claim that Ricardo's analysis of distribution was located within the dynamic setting of the accumulation process. In the aftermath of Sraffa's editorial introduction to the Collected Works and his later Production of Commodities by

Means of Commodities (i960), some writers have placed greater emphasis on Ricardo's alleged concern with purely 'notional' redistributions of 2L given social product: a static object of analysis, similar to Sraffa's. Indeed, one 'Sraffian' has claimed recently that Ricardo 'was almost exclusively interested in the relations between prices and income distribution and very little in economic growth' (Sylos Labini 1990, p. 3) which, if true, would suggest a radical misunderstanding of Ricardo by legions of previous commentators. Although Samuel Hollander is trenchantly opposed to the 'Sraffian' reading, and especially to its message that Ricardo's work is foreign to the 'neoclassical' tradition, he, too, questions the interpretation that Ricardo's core doctrine incorporated the dynamic prin-

Interpretations of David Ricardo

7

ciple of diminishing agricultural returns. Building on his suggestion that the inspiration for Ricardo's early theory of profitability did not come from the Corn Law debate, Professor Hollander claims that 'Ricardo's primary objective ... was to provide an analytical framework whereby to investigate the various factors that play upon the profit rate' (1979, p. 605, cf. pp. 12, 640-1, 647). On Hollander's reading, Ricardo had sought to develop a general analytical framework within which he could establish his 'fundamental theorem of distribution' (p.7): an inverse relationship between movements in the rate of profit and, for whatever reason, movements in money wages. This framework could be, and was, applied to the case of higher money wages resulting from diminishing agricultural returns, but merely as one of a number of possible applications on which, in Professor Hollander's opinion, undue importance has been placed by most other commentators. Indeed, a 'main theme' of his study of Ricardo is, he proclaims, 'that to single out the theorem relating to a declining rate of return on capital is to exaggerate the import of a particular application of the basic theory' (p. 12). In a further challenge to orthodoxy, Professor Hollander presents the 'fact' that Ricardo's case 'against agricultural protection was not based upon the secular downward trend in the rate of return on capitaP (p.604,

emphasis in original), although the iconoclasm of this statement is diluted when, in an accompanying footnote, he discloses that his interpretation 'is stated too strongly . . . [because] on occasion the case is made in such terms'. However, he is quick to add that such recalcitrant instances were 'an exception to the rule, and argued for public consumption' (p.604 n.13). Apparently, these 'public' utterances are not to be taken seriously. Professor Hollander's interpretation is complemented by his view of Ricardo as an 'optimist', who had 'great faith in British prospects despite agricultural protection' (p.6oo).6 If this was so, then it might seem curious to find Ricardo stressing the falling rate of profit argument in his case against agricultural protection. On the other hand, Professor Hollander apparently realises that there is something curious about his own interpretation. For if Ricardo was an 6

Professor Hollander is not alone on this point. Also see Blaug (1958, pp.31-2), Eltis (1984, p. 216), Hicks (1983b), Kolb (1972), Rogin (1956, p. 117), and Stigler (1952). More writers have favoured a 'pessimistic' interpretation of Ricardo's view of future prospects. They include Dobb (1973, p.88), Gide and Rist (1948, p. 170), Halevy ([1928] 1972, pp.319, 340), Mitchell (1949, pp.165-6), O'Brien (1975, pp.41-3), (1981), Roll (1973, p.186), Schumpeter (1954, pp.571-2), and Whewell (1833).

O

INTERPRETING RICARDO

'optimist', did not base his case against agricultural protection 'on the secular downward trend in the rate of return on capital' (except when he traduced his real meaning for the sake of'public consumption'), and was motivated by the detached ambition to develop a general framework of analysis, why was it that he 'chose to adopt an analytical framework incorporating the principle of diminishing agricultural returns and the stationary state'? Professor Hollander thinks that a 'definite answer . . . cannot be given' (p. 640) and, having dismissed nearly all the traditional answers, his puzzlement is understandable. Without making any claim to novelty, I argue that, if anything deserves to be called Ricardo's 'primary objective', it was to establish that progressively diminishing agricultural returns must ('permanently') depress general profitability; and, following on from this point, that Ricardo's 'core' analysis of distribution was inextricably linked with his analysis of growth.7 This is not to deny that the theoretical framework which Ricardo developed could be, or was, applied to other ends: it clearly was so applied as, for example, the analysis of taxation in the Principles bears witness. But the development of that general framework was part of the same intellectual process by which Ricardo sought to establish his more narrowly focused 'agricultural' thesis and was not, contrary to Hollander's account, a separate objective. I also dispute Professor Hollander's view that the falling rate of profit argument was peripheral to Ricardo's case against the Corn Laws. Ricardo's central argument against protection, both in the early correspondence, and in the Essay, was precisely that it would result in a (gradual) decline in general profitability. While it is true that other arguments were deployed, and emphasised in Ricardo's later work, they were complementary to the projection of falling profitability and were highlighted because of contemporary circumstances (this applies, in particular, to the argument that protection encouraged price instability). In my judgement, Ricardo's continued enunciation of the 'secular trend' argument in his later speeches, notes, correspondence, and in the pamphlet On Protection to Agriculture (1822), at times when external circumstances rendered it somewhat inappropriate, is evidence of the importance 7

Recent commentators who have 'rediscovered' the dynamic contextualisation of Ricardo's analysis of distribution include Caravale and Tosato (1980, pp. 15, 55), Ong (1983), Tosato (1985), Hicks (1985) and Caravale (1985).

Interpretations of David Ricardo

9

he placed upon it, not a sign that he was being untrue to his real convictions. I can agree with Professor Hollander, and others, over Ricardo's 'optimism'. He did not envisage a rapid fall in general profitability as a result of protection. But, although he believed that the declension would be gradual, with the negative effects on accumulation offset by cheaper 'luxuries' and a 'natural' reluctance to export capital, this must be set against his view that the alternative future, without restrictions, was one in which accumulation would proceed for a virtually unlimited period without any 'permanent' reduction in profitability. As far as Ricardo was concerned, a rising corn price posed the only empirical threat of'permanently' falling profitability, and it is therefore scarcely surprising to find him retaining an analytical framework which incorporated the principle of progressively diminishing agricultural returns. Indeed, as I read Ricardo, it was the introduction of that principle into the study of the 'natural course' of wages, profit and rent which he regarded as one of his major contributions to the analysis of distribution. Turning to Ricardo's treatment of wages, the subject matter of the second section in chapter 3, most commentators have favoured variations on a 'subsistence' or 'natural' wage interpretation, according to which Ricardo assigned labour a 'natural wage' of unique significance. This natural wage, a datum of analysis at any point in time, was the commodity wage just adequate to sustain a stationary labouring population, and, if market wages were above (or below) the natural level, a population mechanism would operate, with the labouring population expanding (or contracting), thus reestablishing a parity between the natural and market rates. Opinions have long differed, however, over the speed and regularity with which this parity was supposedly achieved. At one extreme, it has been claimed that any divergence between market and natural rates was treated by Ricardo as a fleeting, unusual occurrence, while other commentators have credited him with a more flexible position, encompassing the possibility of significant disparities in exceptional circumstances.8 But, whichever version is taken, the common 8

Writers who have endorsed the more 'hard-line' interpretation include Barkai (1959), Bohm-Bawerk ([1921] 1959, p. 60), Dmitriev ([1904] 1974, p. 42m), Halevy ([1928] 1972, p. 331), Kaldor (1955-6), Jevons ([1879], 1965, pp. 269-70), Pasinetti ([i960] 1974), Read (1829, pp. 213, 251, 253), Samuelson (1959) (but cf. Samuelson, 1978), Wermel (1939, p. 158), West (1826, pp. 38-9, 74), and Whewell (1833). Those favouring a more 'flexible' account include Haney (1924, p. 267), Hicks (1973, pp. 49, 124), Knight (1935), Ramsay

10

INTERPRETING RICARDO

attribution is that Ricardo treated the natural wage as the normal centre of gravity for market wages in an unrestricted historical domain. A markedly different view was presented by Edwin Cannan, whose account presaged the 'new view' interpretations, as they have come to be known, advanced by some latter-day writers. Basing his reading on passages from the chapter 'On Wages' in the Principles, Cannan's version of Ricardo's doctrine was as follows: market (real) wages remain above the natural level in an advancing economy and correspond to that level in the terminal 'stationary state' alone; the 'tendency' for market and natural wages to coincide was intended only in the sense of an historical trend; and the trend involved a falling 'real' market wage, a rising money wage (owing to a progressively higher corn price), and declining rates of profit and capital accumulation (Cannan 1893, especially pp.247-53, 35°) • Cannan's interpretation made little, if any, impression on mainstream opinion, and even appears to have been abandoned by Cannan himself in his later writings (1929, p.299). But 'old' interpretations have a habit of reappearing if they can lay claim to at least some textual support.9 Such has been the case with Cannan's interpretation, independently resuscitated by Carlo Casarosa and Samuel Hollander (the latter with the authoritative support of Sir John Hicks).10 These protagonists of the new view are agreed, first, that the natural wage is irrelevant to Ricardo's analysis of the growing economy; secondly, that his growth model is one in which ('equilibrium') market wages, in commodity wages, secularly decline towards their natural level; and, thirdly, that the relevant 'equilibrium' wage at any point in time is determined simultaneously by the dynamic interplay between the prices of wage goods, the supply of labour, the rate of profit and (thence) the rate of capital accumulation, which governs the demand for labour.11 The new

9

10 11

([1836], 1974, p. 13511.) and Stigler (1952); Samuelson (1979) might also be included in this category. Since this is undeniably true of the new view interpretation, it may be wondered how the relevant passages were dealt with by other commentators. The answer is that those who did not ignore them completely seem to have thought that their import was fully compatible with the subsistence interpretation. See Ashley (1891), Gide and Rist (1948, pp. 172-3, 177), Gray (1980, pp.169-71), Haney (1924, pp.267-71), Rubin ([1929], 1979, chapter 30), and St. Clair (1965, chapter 5). Also see Rogin (1956, pp. 116-37), Tucker (i960, pp. 113-16) and Levy (1976). Among the more important of the recent 'new view' publications, see Casarosa (1978), (1985), Hicks and Hollander (1977), Hicks (1979), Hollander (1979, especially chapter 7), (1983), (1984), (1987, chapter 8), and (1990). Morishima (1989, especially chapter 5)

Interpretations of David Ricardo

11

view interpretation, with its stress on the simultaneous determination of profits and wages, thus dissolves the old perception of Ricardo as a 'surplus theorist', who calculated profit as a residual: a point made explicitly by Casarosa (1978) and Hicks (1979).12 I argue that the sponsors of the new view do not plausibly account for the presence of the natural wage analysis in the Principles, and that it is only the natural wage analysis which is consistent with Ricardo's doctrine of 'permanent' movements in profitability, and with his characteristic style of analysis. Thus, it is my judgement that the natural wage interpretation is superior to the new view alternative in the sense that it alone is faithful to the core analysis of distribution and accumulation in the mature Ricardian system. However, this is not to deny the textual basis for the new view interpretation, both in the chapter 'On Wages' and elsewhere in the Principles. The presence of this material invites an explanation. It is my contention that the new view analysis was derived from Malthus, who had advanced the idea that commodity wages undergo a secular decline in an attempt to persuade Ricardo that higher corn prices would not result in higher money wages and therefore lower profits. The new view passages in the chapter 'On Wages' were a reformulation of the Malthusian scenario, the objective being to prove that higher corn prices would depress general profitability, even allowing for falling commodity wages (cf. Rosselli 1985). If those were the only such passages, it could be concluded that they were introduced to controvert Malthus's argument and had no further significance. But the evidence suggests that Ricardo may have gone further, on some level adopting the Malthusian-inspired doctrine as his own. Yet he had evidently not appraised its (in)compatibility with a natural wage analysis which was, I argue, an outgrowth from his earlier, dominant train of thought. As to the later evolution of Ricardo's thought (after the Principles), there is some evidence that criticism from Malthus may have

12

might also be listed as a 'new view' exponent, at least to the extent that his Ricardo also limited the relevance of the natural wage to the ultimate stationary state. Hollander (1979) adopts a more qualified position. Recognising that Ricardo often calculated profit as a residual, or surplus, if only as a 'deliberate simplification', Professor Hollander suggests that 'the profit rate is merely a formal residual, there being a mutual dependency of the one [profit] upon the other [wages]' (1979, p. 688, emphasis in original).

12

INTERPRETING RIGARDO

prompted a more favourable disposition to the new view analysis on Ricardo's part (Malthus certainly did not recognise Ricardo as a kindred new view theorist). At the same time, however, Ricardo deliberately reaffirmed his view of the natural wage as the normal centre of gravity for market wages, thus confirming that the new view was never Ricardo's one-and-only 'true view' (contrary to Samuel Hollander's opinion). I also comment on a compromise interpretation offered by Caravale (1985), who expands on ideas previously ventilated in Caravale and Tosato (1980) (cf Tucker i960, pp. 113-16). This interpretation has the 'merit' of liberating Ricardo from the charge of incoherence, first put by Malthus, allegedly implied by the use of a 'stationary population' natural wage in the analysis of a growing economy, while retaining the concept of a given natural wage and, per force, a surplus treatment of profit. Caravale's resolution is achieved by redefining the natural wage as 'that constant level of the wage in correspondence with which the rates of growth of population and of capital coincide', whatever those rates happen to be (Caravale 1985, p. 143). Far from being gratuitous, Professor Caravale claims that his redefinition is supported by the 'considerable amount of flexibility' shown by Ricardo to the definition of the natural wage, and also by Ricardo's own use of the 'dynamic' natural wage in the Essay (pp. 135-43). It is thus commended as something with a Ricardian flavour. I argue that Professor Caravale's redefinition is textually unsupported, and that Ricardo's use of the natural wage in the dynamic context is unobjectionable, providing we grant him his ingrained comparative static (or 'discrete') method of analysis. In any case, surely the first task of interpretation is to understand texts, not to rewrite them. Ricardo's attachment to the 'law of markets', discussed in the third section of chapter 3, also raises some controversial questions. According to Keynes, who regarded the 'law' as a source of great falsehood, the classical economists (including Ricardo) had taught that 'supply creates its own demand in the sense that the aggregate demand price is equal to the aggregate supply price for all levels of output and employment (1936, pp. 21-2). This, probably the best-known characterisation of the doctrine, has been interpreted as implying 'Say's identity': the proposition that there is always zero excess demand for money (and therefore zero excess demand for all other

Interpretations of David Ricardo

13

commodities).13 For Schumpeter, however, the 'classical' doctrine was that the 'aggregate demand price of output as whole is capable of being equal to its aggregate supply price for all volumes of output' (1954, p. 624, emphasis in original); in other words, the 'classics' purportedly subscribed to 'Say's equality', according to which the zero excess demand characteristics are restricted to equilibrium states. Specifically with reference to Ricardo, Mark Blaug (1985b, chapter 5) and Samuel Hollander (1979, chapter 9) have followed Schumpeter in favouring the 'equality' interpretation, while Michio Morishima has claimed that Ricardo subscribed to the 'rather antiquated version' of the 'law' described by Keynes (1989, pp. 52,155, 167, cf. Stigler 1953). The 'identity versus equality' issue overlaps with the disputed question of Ricardo's intellectual debt. All commentators are agreed that the 'law' did not originate with Ricardo and that he borrowed it from either Jean Baptiste Say or James Mill, the latter having developed a more uncompromising version of the doctrine incorporating 'Say's identity'. Dobb (1940, pp. 40-1) and Stigler (1953) cite James Mill as Ricardo's source of inspiration, whereas Marx claimed that the 'law' was 'adopted by Ricardo from the tedious Say' ([1862—3] 1969, p. 493). More cautiously, Samuel Hollander finds it 'tempting to conclude that if Ricardo had any early debt it was probably to Say' (1979, p. 502), although he repeatedly finds that Ricardo's exposition of the 'law' has more in common with Mill's version (1979, pp. 506-10). As well as disagreement over the substance and origin of Ricardo's version of the 'law', there have been differences of opinion over its importance in his work. Ronald Meek (1967) declared that it could not 'plausibly be described as fundamental to Ricardo's system'; similarly, Denis O'Brien contends that 'the attribution to this idea of key importance in relation to Ricardo is rather dubious' (1975, p. 41). But these are minority opinions.14 Indeed, some writers have stressed that the 'law' encapsulated Ricardo's 'vision' of actual economic behaviour and is, in consequence, of immense importance to the understanding of his work.15 13

14

15

In an earlier essay, Keynes explicitly accused Ricardo of dealing 'with the abstraction of a neutral money economy' (1933, pp. 138-9). For the opposing view see, for example, Checkland (1949), Dobb (1940, p. 42), Morishima (1989, p. 136), and Stigler (1953). The 'law' is also one element in Pasinetti's famous 'mathematical reformulation of the Ricardian system' (Pasinetti i960). See, in this vein, Halevy ([1928], 1972, pp. 324-5), Hutchison (1952), (1953) and (1978, pp. 45-51), Keynes (1936, pp. 34, 192) and Marx ([1862-3] l9&% PP- 528-29).

14

INTERPRETING RICARDO

Among those who have propounded the last position, Professor Terence Hutchison has made a point of linking Ricardo's use of the 'law' with his weakness for the so-called 'Ricardian Vice'. It was Schumpeter who coined this pejorative description, the substance of his general criticism being that Ricardo constructed highly simplified models between a few aggregative variables, grossly undervalued the importance of frictional (short-run) behaviour, and applied the results of his models directly to the solution of real-world problems. This account of Ricardo's approach represented a distillation of views that had been circulating for over a century, and have continued to circulate in more recent times.16 The boldest challenge to the interpretative consensus has come, almost needless to say, from Samuel Hollander, who seems to regard Schumpeter's charge as an egregious slander. In Hollander's opinion, expressed with great vigour in the opening chapter of his The Economics of John Stuart Mill (1985), Ricardo was 'preoccupied' with the short-run in both his theoretical and applied work; he was well aware that his models were highly abstract, to such an extent that he reposed no great faith in their 'predictions' or in their relevance for policy making; and, because of his sensitivity towards the operation of counteracting causes, frictions and lags, he displayed a 'responsible' attitude towards problems of 'theory confirmation'. Specifically with regard to Ricardo's use of the 'law of markets', however, Professor Hollander allows that the charge of the 'Ricardian Vice' 'appears better substantiated', although he adds that Ricardo's law-based 'refusal to admit excess aggregate supply in the postNapoleonic period' reflected 'less a methodological predisposition than a horror of inflation' (1985b, p. 4). And he later conjectures that Ricardo's use of 'law of markets' reasoning in depressed times can be explained in terms of'the strength of his objections to those economists who maintained that the post-war crisis reflected characteristics of secular stagnation' (p. 33n.). The message appears to be that Ricardo was not 'really' guilty of 'Vice' and, in this spirit, Hollander also emphasises Ricardo's allowance 'for temporary excess demand for money in consequence of an initial reduction in the money supply' (p. 27); that is, his subscription to the weaker, 16

See, for example, the Preface to Jones (1831), Whewell (1833), Bagehot ([1898] 1973, pp. 197-205), Ingram (1915, p. 120), Haney (1924, pp. 276-7), Rubin ([1929] 1979, pp. 242-3), Checkland (1949), and Stigler (1952, 1953). For a selection of later writings, see Blaug (1956, 1968), Pasinetti ([i960] 1974), Fetter (1969), Sowell (1974, chapter 4) and O'Brien (1975, pp. 3, 42, 55, 69).

Interpretations of David Ricardo

15

'Say's equality' version of the 'law of markets'. Ricardo's newly discovered credentials as a 'responsible' applied economist thus emerge relatively untarnished. I argue that the 'law of markets', apparently known to Ricardo at the time of his early 'monetary' writings, was anything but an insignificant element in his analytical system. As Ricardo developed his theory of distribution, the full force of the 'law' became increasingly apparent to him, eventually providing him with the seemingly watertight argument that general profitability could never 'permanently' alter in the absence of changes in the conditions of producing wage goods. The 'law' was central to his mature theoretical system. It was also much more than a purely abstract doctrine. The 'law', which subsumed the principles of rationality and near-perfect market knowledge, was evidently thought to capture the modus operandi of the real economic system; and, contrary to S. Hollander's account, the nature of the doctrine and, more importantly, the manner in which it was applied by Ricardo, are alone sufficient to convict him on the charge of the 'Ricardian Vice'. 17 This judgement stands regardless of his adherence to the 'equality' version of the doctrine (an interpretation which, I argue, is weakly preferable to the 'identity' alternative). For, although Ricardo did make limited allowance for temporary maladjustments, his emphasis always remained firmly, and characteristically, on 'equilibrium' tendencies. Moreover, when he was confronted by an external world which confounded his 'law'-based expectations, his response was to blame reality, not his model. His responsible attitude to theory confirmation, alleged by Samuel Hollander, was not conspicuously apparent. Finally, on the question of Ricardo's source of inspiration for the doctrine, it seems to me that James Mill should probably take 17

More generally, it seems to me that Schumpeter's characterisation of Ricardo's analytical method - the focus on long-run, 'permanent' outcomes to the comparative neglect of 'temporary' phenomena, the use of simple, highly aggregated models, and the isolation of linear, mono-causal relationships - is essentially correct. But in claiming that Ricardo's theoretical results were applied directly to the solution of practical problems, Schumpeter gave a hostage to fortune. In the areas of agricultural protection, monetary policy and the poor laws, theoretical work supplied Ricardo with his ultimate policy objectives although, for a variety of humanitarian, political and purely economic reasons, he was prepared to dilute his specific policy proposals. (The important example of agricultural protection is discussed in the first section of chapter 3.) Samuel Hollander takes Ricardo's various 'concessions' as evidence that the charge of the 'Ricardian Vice' is totally inapplicable, whereas I infer that it requires qualification, not rejection.

l6

INTERPRETING RIGARDO

the credit (or blame). I should add, however, that this question receives only passing attention. Chapters 4 and 5 chart the evolution of Ricardo's thinking on the labour theory of value from the time, after the Essay's publication, when he first committed himself to the idea that (changes in) exchange relationships are governed by (changes in) the amounts of labour required to produce commodities, to his final writings, when he struggled to resolve a complex of problems which had arisen in connection with the theory. Of all the areas of Ricardo's work, this one has given rise to the greatest diversity of interpretative opinion. I begin my literature review with Karl Marx's controversial, often maligned, and more frequently misunderstood interpretation, based on his reading of the third (1821) edition of Ricardo's Principles.18 Marx's most sustained commentary is contained in his three-part Theories of Surplus Value (written in 1862-3, hereafter abbreviated to TSV). In Part 11 of TSV, Marx paid tribute to Ricardo's 'great historical significance for science': [A]t last Ricardo steps in and calls to science: Halt! The basis, the startingpoint for the physiology of the bourgeois system - for the understanding of its internal organic coherence and life process - is the determination of value by labour-time. (TSFPart 11 [1862-3] I969> p- 166, emphasis in original) Coming from Marx, this was praise indeed.19 But Marx was not bestowing upon Ricardo the ultimate accolade of having anticipated the true, Marxian concept of'value'. Thus he complained bitterly that Ricardo had failed to investigate the 'form' or 'characteristic' of labour which constituted the 'substance' of value (see, for example, pp. 163, 172 and T S F P a r t m [1862-3] 1972, pp. 131, 137).20 And he also censured Ricardo for his 'faulty architechtonics' in not following Marx's own, step-by-step approach to the analysis of'value' (T-STPartu [1862-3] 1969, pp. 166-7, X74)-21 Furthermore, Marx 18

19

20

21

Although I refer to M a r x ' s comments on R i c a r d o as an 'interpretation', it must be borne in mind that in Theories of Surplus Value, especially, M a r x appears to have been partly, if not largely, concerned to develop his own theoretical system, with the result that t h e line between interpretation and translation is often extremely blurred. As Bortkiewicz observed, ' R i c a r d o is practically the only a u t h o r who, in a certain sense, finds any mercy in M a r x ' s eyes at all' ([1907], 1952, p . 30 n. 56). According to the account given in Volume 1 of Capital, the value-substance is abstract labour, conceived (in units of time) as the 'productive expenditure of h u m a n brains, muscles, nerves, hands, e t c ' without regard to the specific form that the productive activity takes ([1890], 1976, p p . 134-7). M a r x believed that it was first necessary to abstract, inter alia, from the establishment of a general rate of profit (as in volume 1 of Capital). O n e of Ricardo's m a n y serious failings,

Interpretations of David Ricardo

17

chided Ricardo for the several uses he had made of the word 'value': a criticism that must be fully grasped for a balanced appreciation of Marx's position. Marx observed that Value' had been used by Ricardo 'very often' in the sense of mere exchangeable or relative value (the ratio in which commodities exchange for each other) (p. 172); and that it had been used to denote (relative) 'cost' or 'natural' prices (pp. 175, 329).22 But he also claimed that Ricardo sometimes attributed a third meaning to 'value': 'value' as determined exclusively by the labour expended on the production of a commodity, occasionally referred to by Ricardo as 'absolute value' or 'real value' (pp. 172—3). According to Marx, this last species of'value' was not clearly differentiated by Ricardo from 'exchange value' (TSV Part in [1862-3] 1972, p. 125), or from 'value' in the sense of cost or natural price (TST Part 11 [1862-3] J 969, P- J99)- But the impression given by Marx is that 'labour value' (as Rubin later termed it [1929] 1979, p. 249), although imperfect in its conception and expression, was both present in, and important to, Ricardo's work. That alleged importance was manifested, so Marx believed, in the objective Ricardo had set for the 'invariable measure of value': a commodity numeraire which should isolate those price variations for other commodities that were 'caused by the rise or fall "in their values", in the labour-time necessary for their production' (p. 202, emphasis in original). Marx saw little merit in this aspect of Ricardo's analysis, although he thought he had grasped its significance: 'The problem of an "invariable measure of value" was simply a spurious name for the quest for the concept, the nature, of value itself (TSV Part m [1862-3] 1972, p. 134, emphasis in original).23 Yet although Ricardo had, on one level, identified this 'nature of value' with labour-time, he had also retained his other 'value' concepts. His problem was in not clearly distinguishing one concept from another.

22

23

according to M a r x , was that he h a d 'presupposed' (or 'smuggled in') the general rate of profit at the beginning of his chapter ' O n V a l u e ' in the Principles. These are the prices which obtain when wage a n d profit rates are respectively uniform, assuming, for simplicity, single uniform rates. R e n t is excluded from Ricardo's cost of production. M a r x diagnosed R i c a r d o ' s ' p r o b l e m ' as being his 'false assumption that money [the 'invariable measure of value'] . . . exchanges as a commodities for commodities' (TSV Part n [1862-3] 1969, p . 200, emphasis in original; cf. TSV Fart m , [1862-3] 1972, p p . 133-5). Per contra, M a r x ' s 'gold', in which his values and 'modified values' ('prices of p r o d u c t i o n ' in

10

INTERPRETING RICARDO

The accusation that Ricardo confused his Values' was made repeatedly by Marx (for example, TSV Part n [1862-3] ! 969 ? pp. 166, 199, 215). Nowhere was this confusion allegedly more evident that in the first chapter of the Principles, 'On Value', where Ricardo had initially postulated a situation of equality between (relative) 'labour-values' and (relative) cost or natural price 'values': a special context in which the 'value' concepts coincided (pp. 164—5, 242; TSV Part in [1862-3] 1972, pp. 70-1 ). 24 Ricardo had then proceeded to investigate the consequences for natural prices of changes in distribution when (in my terms) 'capital structures' differed between production processes: the natural prices of at least some commodities would alter even though nothing had changed with respect to their physical conditions of production. 25 Marx's verdict on the analysis was equivocal. On the one hand, he claimed that Ricardo had ('erroneously') deduced that his 'labour values' were affected by the change in distribution (TSV Part 11 [1862-3] 1972, pp. 175, 191) ,26 On the other hand, he stated that Ricardo had 'virtually demonstrated, the difference between cost and value, cost-prices and [labour] values' (p. 199); and that 'Ricardo has a notion that there is a difference between value and cost-price', a 'notion' which is warmly commended as a 'great contribution' (TSV Part in [1862-3] 1972, p. 71). My reading of Marx's position is that

24

23

Capital Volume in) were expressed, was a commodity from outside the domestic production system. Its 'price of production' did not even exist. Marx also observed (for example, in TSV Fart n [1862-3] ^ p j , p- 215) that Ricardo did the same thing in later chapters. A 'dated labour' framework (with credit due to Whewell, 1833) may serve to capture the spirit of the analysis. For commodity *, let

Px = wLt(l+r) +wL,-I{\+r)2+

26

... + wL,-n(\ +r)n+\

where p stands for natural price, L, represents labour expended in the last (discrete) production period, /,,_, ... Lt-n are labour expenditures in previous production periods, and w and r denote the current period uniform values for, respectively, wages and the rate of profit. Differences in 'capital structure' reduce to differences in the 'time profiles' of labour inputs between commodity production processes (for example, one commodity, with the same total labour input as another, may have had more of that labour expended in a recent period). If such differences exist (with r > o ) : (i) natural price ratios will not correspond to ratios of expended labour (the summation of L terms), and (ii), a rise (or fall) in w, accompanied by a fall (or rise) in r, will result in natural price changes which are independent of variations in labour expenditures. The direction of movements in absolute natural prices will depend on the 'time profile' of labour inputs to the production of the 'measured' commodity relative to the 'time profile' for the commodity-numeraire. This is the version of Marx's interpretation which has been criticised by Ian Steedman, who avers that 'Marx persistently misinterprets Ricardo's use of the term value to be his (Marx's) use and then accuses Ricardo of "mistakenly" identifying value and cost' (Steedman 1982, p. 121). I cannot agree that Marx imputed to Ricardo his (Marx's) concept of'value'.

Interpretations of David Ricardo

19

Ricardo's overloaded use of the term 'value' had prevented him, or rather, it had necessarily prevented him, from clearly stating the essential conceptual difference, thus leading to apparent contradictions. Yet Ricardo had understood, if only vaguely, that there was a difference. There was an element of'insight'. There is no ambiguity surrounding Marx's pronouncement on 'the true course' of Ricardo's investigation. Ricardo had believed that the: variations in the cost-prices of commodities resulting from a rise or fall in wages are insignificant compared with those variations ... brought about by changes in the values of commodities, that is changes in the quantity of labour employed in their production. (Ricardo is far from expressing this truth in these adequate terms). One can therefore, by and large, 'abstract' from this and, accordingly, the law of value remains virtually correct. (TSV Part 11 [1862-3] !969> PP- 193-4) Ricardo had 'abstracted' from troublesome inequalities between (changes in) cost or natural price 'values' and (changes in) 'labour values' on supposed 'empirical' grounds, and had thus obtained his 'law of value' by direct assumption. Marx's interpretation of Ricardo can be located within an interpretative tradition that extends back to Ricardo's own time. Thus, the anonymous author of the pamphlet Observations on Certain Verbal Disputes in Political Economy (1821) remarked that Ricardo had departed from using 'value' in a relative sense, occasionally turning it into an absolute entity dependent on labour expenditure (pp. 14—16). Samuel Bailey was even more explicit in making that allegation (1825, pp. 31, 58m, 162—3). Bailey also claimed, as did Marx, that Ricardo's 'invariable measure of value' was supposed to indicate 'variations in the producing labour of other commodities', described by Ricardo as variations in those commodities' 'values'; and that Ricardo had been perplexed and confused in his conception and handling of 'value' categories (pp. xvii-xviii, 233-9). 27 With an even greater stress on Ricardo's 'absolute' usage of'value', related to labour expenditure, John Stuart Mill, in his Essays on Some Unsettled Questions, made the point Ricardo had only 'occasionally' used 'value' in the relative sense, and had 'more frequently employed it in a sense peculiar to himself, to denote . . . the quantity of labour required to produce [an] article' (1874, p. 96, emphasis in 27

Marx was conversant with both these earlier works.

20

INTERPRETING RICARDO

original).28 But it was Marx's 'labour theory' reading, and specifically the much abridged version in Volume i of Capital, that became the focus of attention by the closing decades of the nineteenth century.29 Unfortunately for Marx's reputation as a critic, Capital contained a distinctly one-sided account of Ricardo's value analysis in comparison with the lengthier treatment in Theories of Surplus Value. Thus, Marx claimed in Capital that Ricardo had treated labour as 'the source of value' ([1890] 1976, p. 174 n. 33) and had assumed that his 'average prices' (natural prices) 'directly coincide with the values of commodities' (p. 269 n. 24). The impression conveyed by these curt remarks is that 'labour value' was Ricardo's sole use of 'value'. That was how Alfred Marshall seems to have understood Marx, albeit imperfectly: 'Karl Marx claim[s] Ricardo's authority for the statement that the natural value of things consists solely of the labour spent on them' ([1920] 1949, p. 672); and he added that 'Marx interpreted Ricardo's doctrine, to mean that interest [profit] does not enter into that cost of production which [partly] governs . . . value' (pp. 672-3n.). With regard to the latter accusation, Marx had not, in fact, offered any such interpretation, even in Capital. But, letting this pass, it was the claim that Ricardo had any 'labour value' concept which Marshall robustly denied, although he seems to have accepted that Ricardo reasoned on the assumption 'that the values of two commodites are to be regarded as .. .proportionate to the amount of labour required for making them' (p. 672, my emphasis). Marx (in TSV) had said much the same with a different emphasis. His Ricardo had assumed the 'law of value' in order to obtain the sought-after correspondence between exchange values and 'labour values'. For Marshall, however, Ricardo's true concept of 'value' was 'cost of production'. And the impression given by Marshall is that the 'proportionality' assumption (relative costs in proportion to labour expenditures) was nothing more than a special assumption of 28 29

F o r Mill's later position, see below, p . 21 n. 30. T h e first (German) edition of Capital 1 was published in 1867; the English translation followed in 1887. M a r x served notice in Capital 1 that a n e x p a n d e d t r e a t m e n t of Ricardo's analysis would a p p e a r in ' t h e third a n d fourth books of this work' ([1890] 1976, p . 173 n. 33). T h e third 'book', Capital V o l u m e m , was published in 1894 with a n English edition in 1909. T h e fourth 'book' was the three-part Theories of Surplus Value, first G e r m a n edition published in 1956.

Interpretations of David Ricardo

21

no great significance. Marshall's Ricardo was, first and foremost, a 'cost of production' theorist.30 Both Marx and Marshall had based their interpretations on the third edition of Ricardo's Principles, but with the later work ofJacob Hollander and Edwin Cannan, attention shifted to the evolution of Ricardo's thought. According to Hollander (1904), Ricardo first adopted a 'labour measure' of exchangeable value in his early 'monetary' writings, explaining relative values in terms of comparative amounts of labour expended on the production of commodities. Even at this early time, Hollander believed that Ricardo harboured reservations towards the 'labour measure': the conviction was 'present to some extent from the first in Ricardo's mind, that the claim of "embodied labour" to serve as a measure of value must be relative rather than absolute excellence'. These reservations had increased by the time Ricardo came to write the first edition of the Principles, in which the exceptions to the 'labour measure', owing to unequal capital structures, were 'clearly recognised'. Subsequently, criticism from Ricardo's contemporaries, Torrens and Malthus, resulted in textual changes to the chapter 'On Value' in the second edition of the Principles which, Hollander believed, were 'highly significant': there was 'an appreciable increase of reserve in the advocacy of "embodied labour" as a universal measure of value'. But, according to Hollander, this 'appreciable increase in reserve' did not placate Ricardo's opponents, especially Malthus, whose further criticism led to quite dramatic textual changes in the third edition of the Principles. The rewritten chapter 'On Value' was 'in content and tendency very different from that in the original edition of 1817 and a conspicuous though logical advance over that in the edition of 1819', opined Hollander. By this stage, it seems that Ricardo did not view the 'labour measure' as a remotely satisfactory approximation. As for 30

This interpretation may have reflected the influence of the later J. S. Mill. In his Principles of Political Economy, Mill presented Ricardo's natural value or natural price as the 'value of a thing which is proportional to its cost of production', without mention of labour quantity (Mill [1871] 1987, pp. 452-3); Ricardo was also said to have never denied that the 'wages of labour have as much to do with value as quantity of labour' (p. 461); and with reference to writers, presumably including Ricardo, who had 'formed a notion of something, under the name of a measure of value', Mill claimed that the measure 'would be more properly termed a measure of cost of production' (p. 566). Ricardo's 'more frequent' use of value to denote 'the quantity of labour required to produce [an] article' alleged by the younger J. S. Mill in his Unsettled Questions (1874, p. 96, emphasis in original) had disappeared from view. The possible reasons for Mill's change in emphasis are discussed briefly in my concluding chapter.

22

INTERPRETING RICARDO

the later use of the measure, this was merely 'the result of deliberate convention', with Ricardo's faith in the labour theory having suffered a massive haemorrhage. Edwin Cannan agreed that criticism had forced Ricardo's 'retreat' from 'the quantity-of-labour-required doctrine' (as Cannan described it, 1929, p. 176) over the successive editions of the Principles. But, unlike Jacob Hollander, who gave the impression that Ricardo had responded to the critical pressure with cheerful resignation, Cannan implied that Ricardo had sustained a beguiling emotional attachment to his 'discredited' theory, preventing him from consigning it to oblivion (or even to 'deliberate convention'): 'His heart clung to the pure labour theory, but his candid brain showed him that it was weak' (Cannan 1929, p. 177). Unfortunately, perhaps, Cannan did not elaborate on his insight to Ricardo's obsession. In his Introduction to the Collected Works (1951), Sraffa directly challenged the 'widely accepted opinion' that Ricardo had 'steadily retreated under pressure of his critics from the theory of value presented in edition 1' (Worksti, p. xxxvii).31 Although Sraffa granted that Ricardo had 'shown signs of weakening' during 1820 (between the second and third editions of the Principles), newly discovered evidence had convinced him that the 'retreat' interpretation was based on a misunderstanding of the textual changes and that 'the theory of edition 3 appears to be the same, in essence and in emphasis, as that of edition 1' (p. xxxviii). The 'signs of weakening' were 'no more than a passing mood' (p. xl). Although Sraffa dismissed the 'retreat' interpretation, he allowed that there had been 'considerable alterations' to Ricardo's specification of an 'invariable measure of value' over the three editions of the Principles (p. xl). These alterations were made in connection with a problem supposedly encountered by Ricardo with his distribution analysis. According to Sraffa, Ricardo was concerned with 'the division of the national product between classes and in the course of that investigation he was troubled by the fact that the size of this product appears to change when the division changes' (p. xlviii). 31

The 'retreat' interpretation had received passing support from Edelberg (1933), and a further variation on the same theme was developed by Cassels (1935), who triumphantly pronounced that the 'real significance of Ricardo's chapter on value lies not in the fact that occasional passages seem to support a labour theory but rather in the evidence it gives us that he had examined this theory critically and was finally led to reject it entirely'. For a later version of the 'retreat' interpretation, see Rogin (1956, pp. 14-5).

Interpretations of David Ricardo

23

This 'troublesome' change reflected unequal 'capital structures' and the resulting distribution-induced natural price changes. Hence: the problem of value which interested Ricardo was how to find a measure of value which would be invariant to changes in the division of the product; for, if a rise or fall of wages by itself brought about a change in the magnitude of the social product, it would be hard to determine accurately the effect on profits, (p. xlviii) This problem had not arisen when (Sraffa's) Ricardo had relied on his 'corn model' analysis, for at least in the agricultural sector, where the general rate of profit was determined, measurement was effected in physical (corn) terms. Nor would it arise if the 'labour theory' held without exceptions. In that event, the magnitude of the social product would be determined uniquely by the total labour required to produce it, not its distribution, with the rate of profit determined 'by the ratio of the total labour of the country to the labour required to produce the necessaries for that labour' (p. xxxii).32 The newly specified third edition standard was, it seems, designed to restore the validity of 'unmodified' labour theory reasoning in a 'modified' context. Thus: In edition 3 ... the standard adopted was money 'produced with such proportions of the two kinds of capital as approach nearest to the average quantity employed in the production of most commodities'; and the relevant passages were accordingly altered to the effect that, with a rise of wages, some commodities would fall and others rise in terms of this standard. (If measured in such a standard, the average price of all commodites, and their aggregate value, would remain unaffected by a rise or fall of wages.) (xliv—xlv) The new standard ensured that the money value of national income would be invariant to changes in distribution between wages and profits.33 Ricardo may have solved his 'problem'. 34 32

33

34

SrafFa may have meant that the rate of profit would be determined by the ratio of total labour minus the labour required to produce necessaries to that latter quantum of labour. The formula implies that all capital is reducible to wage outlays. Cassels (1935) h a d also suggested that R i c a r d o w a n t e d a s t a n d a r d in terms of which 'total national incomes . . . can only vary in proportion to the changes in the total amount of labour expended in producing it'. But Cassels failed to supply a coherent explanation of how Ricardo's third edition standard might have helped to achieve this objective in the context of differing 'capital structures'. The 'price balancing' role for the invariable standard is not, however, the only one mentioned by Sraffa. He also suggests that Ricardo wanted a standard which would measure changes in the 'difficulty of production' (in terms of expended labour) of commodities, and (in Ricardo's later work) a standard that could distinguish between price changes reflecting altered labour expenditures and those that were distribution-induced {Works 1, pp. xlviii-xvlix). But as Eatwell has observed, it must be assumed that 'output, wages a n d social capital [are] all commodity bundles of average composition' (Eatwell 1975b).

24

INTERPRETING RIGARDO

SrafFa's interpretation of the 'price-balancing' function of Ricardo's third edition standard acquires greater significance in relation to the putative Ricardian pedigree of his own theoretical work. This matter is taken up in chapter 6, but I will observe here that many of the commentators who have forthrightly endorsed the interpretation (such as Bharadwaj 1983b, Dobb 1973, pp. 81-4 and Eatwell 1975b) have also proclaimed the notion of a Ricardo-Sraffa lineage. In the case of the 'no retreat' interpretation, however, most commentators seem to have accorded it a favourable reception, irrespective of their position on other matters. 35 But what was it, exactly, to which Ricardo had remained committed? What was Ricardo's 'labour theory'? Sraffa gave no explicit guidance as to whether Ricardo had perceived the labour theory as anything more than a useful analytical tool. But Maurice Dobb, who expedited the publication of the Collected Works (see Pollitt 1988), evidently believed that there was a strong Marxian flavour to the theory. He wrote in private correspondence: Sraffa's edition of Ricardo's works and correspondence ... is, I'm glad to say, nearing completion ... I think we conclusively establish (in opposition to the traditional Hollander-Marshall-Cannan view) that there was no 'weakening' of Ricardo's enunciation of the labour theory as time went on; that in fact he reached at the end of his life a position rather close to that of Marx, so that the true line of descent is certainly from Ricardo to Marx, and not from Ricardo to cost-of-production theory au Mill to Marshall as the bourgeois tradition has it. [Ricardo] was at the last still exercised with a notion of an Absolute Value (= embodied labour) as something distinct from, but underlying, exchange-value. (Letter of 23 December 1950 to Theodor Prager, emphasis in original, quoted by Pollitt 1988) The claim that Ricardo's theory incorporated a notion of labourdetermined value had been a central feature of Marx's and other nineteenth-century writers' interpretations. Many later commentators, unpersuaded by Marshall's contrary interpretation, have also (directly or indirectly) supported elements of the 'labour theory' reading by variously crediting Ricardo with a notion of 'value', 'absolute value', or 'real value', related exclusively to the labour expended on a commodity; the view of labour as the 'substance' or 'source' of value; and the perception of labour as both quantitatively 33

But see Tosato (1985, p. 210) for a note of dissent on this point.

Interpretations of David Ricardo

25

and qualitatively the most important factor determining (relative) value.36 But these views have not prevailed. Marx's reading was subject to scathing criticism by Ladislaus von Bortkiewicz, who targeted his criticism on Marx's claim that Ricardo had 'confused prices with [Marxian] values'. Marx was simply wrong and, in general, his 'discussions of Ricardo . . . redound in unjust, petty, wordsplitting remarks' ([1907] 1952, p. 30 n. 56). Similarly, Mark Blaug described Marx's reading of Ricardo as 'a comedy of errors' (1958, p. 236), while Ian Steedman accuses Marx of'a mere verbal muddle' and 'verbal insensitivity' (1982, pp. 120-1). For the majority of modern commentators, it would seem that Ricardo's 'labour theory', insofar as it even deserves that title, was intended as no more than a reasonable empirical approximation. One of the most uncompromising versions of the 'empirical' interpretation was advanced by Professor George Stigler, who appeared incredulous of the opinion that 'Ricardo attributes more than quantitative importance to labour in determining values'. Ricardo's theory was 'an empirical labour theory of value, that is, a theory that the relative quantities of labour required in production are the dominant determinants of relative values' (Stigler 1958, emphasis in original). But Stigler also asserted, in a way reminiscent of Marshall (and Whewell 1833), that the theory 'is of course a cost of production theory', as if the two descriptions are identical. What can be made of this? An unqualified 'cost of production' theory may be understood as one which generates the proposition that (changes in) relative prices are governed by (changes in) relative costs of production, regardless of whether those relative costs are proportional to relative labour expenditures. In contrast, a 'pure' labour theory (or, following Gordon 1959, a 'labour theory of relative price') might be defined as one which postulates (changes in) relative expenditures of labour time as the determinants of relative prices. If a 'pure' labour theory holds, and prices are decomposable into wages and profits, then relative costs of production (wages plus profits) are equal to 36

See, for example, Ashley (1891), Bell (1907), Gillman (1956), Gonner (1913, pp. 306—7), Gordon (1959), MacDonald (1912), Meek (1967) and (1973, p. 112, but cf. p. iv), Myint (1948, p. 33), Patten (1893), Roll (1973, p. 178), Rubin ([1929] 1979, p. 251) and Whitaker (1904, especially pp. 43-4, 13m., 137).

26

INTERPRETING RICARDO

comparative labour expenditures.37 That being so, it could be argued that the 'pure' labour theory is a 'special case' cost of production theory and, in one sense, it is. However, it would seem that the reason why some commentators have baulked at the cost of production description of Ricardo's work is that they believed it would convey a wholly misleading impression.38 For them, Ricardo's putative emphasis on labour as the determinant of (at least) relative values would be obliterated by the 'cost of production' description.39 This point is well taken.40 Returning to Stigler, his emphasis, like Marshall's, was decidedly on 'cost of production' (cf. Steedman 1982). We are therefore invited to believe that this was Ricardo's primary 'value' concept, with the 'special case' labour theory used only because of its supposed 'empirical' credentials. The latter, 'empirical' facet of his interpretation is the one that has been widely accepted and, if it is combined with the 'no retreat' thesis, we arrive at something approaching the current 'received wisdom'. Samuel Hollander accepts most of that wisdom. Although he allows that 'Ricardo seems to accord labour a special role in value creation - treating it as, in some sense, the source of value', he stresses that the sine qua non for the labour theory's adoption by Ricardo was its supposed 'empirical validity' (1979, p. 263). And he accepts that there was no 'retreat' over the successive editions of the Principles ('The contrary is closer to the truth', p. 217.) His most distinctive contribution is over the role and (lack of) significance of the invariable standard. According to Sraffa, one function of Ricardo's third edition standard was to keep the money value of national income constant following a redistribution of the aggregate product. But for Professor Hollander, although that solution 'is certainly implied by the Ricardian materials' (p. 238) it was not Ricardo's main objective. On Hollander's reading, 'Ricardo's primary concern, for the analysis of which he sought to devise a suitable measure of value, was the 37

38

39

40

T o obtain the link between labour expenditures a n d 'costs', (changes in) relative wage costs per unit of output must be proportional to (changes in) labour expenditures, a n d profit-to-wage ratios must be uniform (a case of identical 'capital structures'). See Gide a n d Rist (1948, p p . 164-5), Schumpeter (1954, p . 594) a n d Whitaker (1904,

p. 13m.). J . Hollander and E. C a n n a n may have used alternative descriptions to 'cost of production' for the same reason. M y adopted definition of a ' p u r e ' labour theory is not intended to exhaust the possible meanings of a 'labour theory of value'.

Interpretations of David Ricardo

27

rate of return on capital' (p. 6). 41 The purpose of the 'invariable measure', or 'standard measuring device' to use Hollander's neologism, was to 'predict . . . the precise profit rate and price structure that will be yielded in the new equilibrium state following a disturbance to wages' (p. 304). Professor Hollander elaborates on this 'predictive' function as follows: Ricardo proceeded . . . by holding constant the price of a commodity in the face of a wage change, implicitly assuming the factor proportions ['capital structure'] involved to be identical with those of the medium of exchange. The new lower profit rate, which (in principle) could be easily calculated for this commodity, was then applied throughout the system to obtain the new cost price and thus the new price structure, (p. 303) The problem was that the new profit rate and price structure would depend on the commodity selected as the 'medium'. The new selection of the 'mean' or 'average' commodity in the third edition of the Principles 'was a formal solution [to this problem] at best' (P- 3O3)Professor Hollander appears confident that Ricardo 'did not consider the alteration to be of great significance' (p. 220). Nor, on his interpretation, would we expect otherwise, for the standard was only 'a most convenient but not an indispensable predictive device' (p. 305) of which Ricardo could (or should) have rid himself by using 'process' (supply and demand) analysis (I return to this aspect of Hollander's interpretation in chapter 6). Furthermore, even the limited significance of the 'standard measure' was constrained to the first chapter of the Principles, since the complications resulting from unequal 'capital structures' were overcome in subsequent chapters 'by the simple expedient of assuming identical factor proportions across the board' (p. 202; cf. pp. 221, 258).42 The (respecified) 'standard' is thus presented as a specialised and rather insignificant tool in Ricardo's analytical kit. I come to an outline of my interpretation. Ricardo seems to have first adopted an unmodified 'labour theory' in early 1816 (contrary 41

42

Professor Hollander's statement connects with yet another interpretative debate, this one being whether R i c a r d o targeted his distribution analysis on the (proportional or absolute) share of profit or on the rate of profit. Since I agree with Hollander that the latter was, indeed, Ricardo's 'primary concern', I will not pursue the matter. Cf. D o b b (1940, p p . 14-15), Eltis (1984, p p . 185-6) a n d Groenewegen (1972). O t h e r writers had noted that R i c a r d o used an unmodified 'labour theory' beyond his first chapter but did not credit him with founding that usage on the equal 'capital structures' assumption. See, for example, Bohm-Bawerk ([1921] 1959, p . 302) and Whitaker (1904, p . 58).

28

INTERPRETING RICARDO

to Jacob Hollander's interpretation, there is no evidence of Ricardo's embrace of the labour theory in the early 'monetary' writings). Even at this time, there are indications that he went beyond the formulation of a 'pure' labour theory of relative value by using 'value' (or sometimes 'natural value') to signify an absolute quantity dependent on labour expenditure, but his enquiries were not motivated by a desire to find the 'nature' or 'essence' of value. He was more straightforwardly concerned to overcome certain logical difficulties with his previous analysis, which Malthus had brought to light after the Essay's publication. The labour theory, which provided a simple 'physical' analogue to the notion of (changes in) 'difficulty or facility of production', long at the centre of his speculations, appeared to serve his purposes admirably (cf. Dobb 1940, p. 20 and Meek 1974). It seems likely that the first chapter of the Principles, 'On Value', together with several other chapters, were drafted before Ricardo became aware of the modifying influence on the labour theory associated with different 'capital structures'. In that state of analytical innocence, the stipulation of a commodity-numeraire ('invariable standard'), always requiring the same labour input, was enough to ensure that changes in natural prices exclusively reflected proportional changes in labour expenditures, as we find in parts of the chapter 'On Value' and in later chapters. We also find the crude notion of'labour value' (as Rubin described it), sometimes referred to by Ricardo as 'real value' or simply 'value', and the usage of 'value' to denote (relative) costs of production. The beauty of the 'pure' labour theory was that all these concepts of value coincided in the sense that (changes in) values qua comparative costs of production were directly associated with (changes in) 'labour values' as measured vicariously by the 'invariable standard'. 43 It is when there are modifications to the 'labour theory' that the association breaks down. The modifications, owing to unequal 'capital structures', were discovered by Ricardo at a comparatively late stage in the writing of the Principles. At first, this discovery was greeted with dismay, but then Ricardo had an inspiration, or so it seemed at the time. The new material could be turned to his advantage in the shape of the 'curious effect': the novel demonstration that the natural prices of 43

This is to assume away the 'heterogeneous labour' problem (see below, pp. 157-8).

Interpretations of David Ricardo

29

commodities would actually fall when wages rose (and the rate of profit fell). This result was obtained by using a commodity-?*umeraire (described as an 'invariable standard') supposedly issuing from the shortest end of the 'capital structure' spectrum (it was produced with the least capital in the shortest possible time). That was all very well, but the interpolation of the 'curious effect' material into the chapter 'On Value' was disastrous. The 'effect', which Ricardo gleefully played up in numerical illustrations, signalled the breakdown of the 'pure' labour theory and, relatedly, the impossibility of finding any commodity-numeraire which could limit all natural price changes to those associated with altered labour expenditures. The latter point would be otiose if, as Samuel Hollander has claimed, Ricardo based his use of the labour theory in subsequent chapters on the assumption of identical 'capital structures'. But there was no such assumption (cf. Bharadwaj 1983b and Blaug 1980), nor did Ricardo offer an 'empirical' justification for using the theory. Indeed, he offered no real justification at all, save for the repeated claim, which he had himself discredited, that the sole criterion of an unchanging labour input to the production of an 'invariable standard' was sufficient to ensure the exclusive relationship between changes in labour expenditures and changes in natural prices. Although it remains a possibility that Ricardo had, at some point during the drafting process, perceived a means of reconciling the 'curious effect' material with his use of an unmodified labour theory, no such 'solution' was incorporated in the finished work. It was not until late 1819 that Mai thus compelled him to realise the seriousness of his oversight. Malthus's role as a critic of Ricardo's 'labour theory' has long been recognised, and in the interpretations of Sraffa and S. Hollander, for example, he is credited with having prompted Ricardo's respecification of the 'invariable standard' in edition 3 of the Principles. But the full force and ingenuity of Malthus's critique, and the significance of the resulting textual changes to the chapter 'On Value' in the third edition of Ricardo's Principles, have been greatly underestimated. By means of some incisive arguments, Malthus effectively raised the question of how any 'invariable standard' could validate the use of a 'pure' labour theory when 'capital structures' differ. Ricardo was noticeably shaken by Malthus's criticisms (perfected, with some relish, in Malthus's own Principles), to the point where he seemed to be on the verge of abandoning his 'labour theory'. But, in agreement with

30

INTERPRETING RIGARDO

Sraffa, this 'weakening' was a 'passing mood'. After further deliberation, he resolved to defend the 'labour theory' and, to that end, two defensive strategies were employed. The first strategy found expression in the new section 'On an invariable measure of value' in the third-edition version of the chapter 'On Value'. It would seem that Ricardo could not rid himself of the idea, deriving from intellectual habits formed at the pre-'curious effect' stage of his deliberations, and reinforced by dubious 'natural science' analogies, that a perfect 'invariable standard' would limit natural price movements to those resulting from altered labour expenditures alone. His first priority thus became to specify such a standard. The stumbling-block was his own ('curious effect') analysis, which implied the non-existence of the 'perfect' standard when 'capital structures' differ. The best Ricardo could devise was a commodity-numeraire produced under [unweighted) average conditions (with respect to 'capital structure') which were allegedly representative of the production conditions of the 'great mass' of commodities. At least for them, changes in relative natural prices and changes in 'labour values' would coincide. It is true, to a point, that the new standard played the part of Samuel Hollander's 'measuring device'. That role was certainly played by the 'extreme' standard in the 'curious effect' analysis of editions i and 2. But this had never been the only function of an invariable standard, nor was it the main function of the edition 3 standard. In the first two editions there were, in effect, two types of standard, one a 'measuring device' pure and simple (in the 'curious effect' analysis), the other a 'labour value' standard. In the third edition these two functions were collapsed into one, but the 'measuring device' function was clearly subordinated to the role of isolating changes in 'labour value'. Primarily, the new standard was designed to salvage the 'labour theory' (cf. Ong 1983 and Tosato 1985). Pace Samuel Hollander, this was scarcely an insignificant development. Professor Hollander is not alone in mistaking Ricardo's purpose. Sraffa's claim that one function of the new standard was to keep the money value of the national product invariant to changes in its distribution is also wide of the mark. There is no evidence that the 'balancing solution' was ever formulated by Ricardo and, indeed, he explicitly suggests that the distribution-induced price movements

Interpretations of David Ricardo

31

will not cancel out in the aggregate.44 Furthermore, in his later writings he made a point of explaining that his analysis of changes in the rate of profit was to be framed at the 'micro' level of the individual farm (or firm), not at the 'macro' level of national income shares.45 The specification of the new standard was not Ricardo's only tactic for defending his labour theory. Disappointed by his inability to find a 'perfect' standard, but determined to retain the labour theory all the same, he also adopted the questionable expedient of removing those numerical examples from 'On Value' which he had used to illustrate the 'considerable' modifications and, ostensibly on empirical grounds, he replaced them with others which gave the impression that the modifications were distinctly trivial. It would seem to be on the basis of the amended text that he has been credited with an 'empirical' labour theory. I argue, however, that Ricardo's empirical assertions should be seen as a manifestation of his strong attachment to the 'labour theory'. Hence, from my perspective, the claim that the 'labour theory' appealed to Ricardo because he thought it was a good empirical approximation is an inversion of the truth. The labour theory had supplied Ricardo with a graphic means of demonstrating that worsened conditions of agricultural production (more labour expenditure per unit of output) would 'permanently' depress general profitability. His appeal for the theory undoubtedly stemmed, in part, from its suitability to his specific requirements in the realms of distribution analysis. But there was, or rather, there became, more to it than that. Provoked by Malthus's criticisms, Ricardo's thoughts turned to the very meaning or 'essence' of value, and it seemed to him with increasing force that 'value' should mean, or be related exclusively to, quantities of labour expenditure; and that (changes in) cost of production 'values' should solely reflect (changes in) 'labour values' (variously referred to at different times 44

45

However, as Whitaker remarked most perceptively (1904, p . 62), the 'balancing solution' can be found in works published after Ricardo's death by J a m e s Mill a n d McCulloch. This might tempt some writers to claim that Ricardo must have formulated the same solution himself, perhaps in lost papers or conversation, in much the same way that the 'corn model' interpretation is currently defended (see above, p . 5 n. 5). I would not doubt that Mill a n d McCulloch were inspired by their (mis) understanding of Ricardo's work, although it is clear from Ricardo's later correspondence on 'value' that the three friends were working on completely different wavelengths. T h e 'balancing solution' belongs to them (and also to M a r x ) , not to Ricardo. Cf. Blaug (1958, p . 11), H a n e y (1924, p . 271) a n d Schumpeter (1954, p p . 652-3).

32

INTERPRETING RICARDO

as 'real values', 'absolute values', 'natural values' or simply 'values'). Hence the search for the 'perfect' standard, 'perfect' in the sense that the 'value' concepts could be mutually reconciled. Ineluctably, given the way the problem was posed, it proved to be a futile exercise. One implication of my study is that Marx's often-ridiculed interpretation is considerably closer to the truth than those advanced by some of his critics. While it is true that Ricardo did use 'value' in the sense of (relative) cost of production (which Marx never denied), that was not his only usage. The concept of 'labour value' can be found in the Principles and, as Ricardo himself admitted in later correspondence, it was fundamentally important to his thinking on 'value'. Marx exaggerated the extent to which Ricardo confused his value categories, but those writers who, following Marshall (and the later J. S. Mill), have claimed that Ricardo held a mere 'cost of production' theory, are responsible for perpetrating a more serious distortion of his work. In chapter 6, I disentangle the main interpretative threads in the ongoing debate over Ricardo's 'neoclassical' or 'Sraffian' credentials; and I evaluate the rival claims which have been made. The question of Ricardo's lineage has preoccupied generations of economic theorists and historians of economic thought. Alfred Marshall argued for Ricardo's acceptance within the 'neoclassical' tradition, but failed to gain many supporters for his cause. By the 1950s, the dominant view appears to have been that Ricardo belonged in a pre-'neoclassical' underworld, having failed to anticipate many, if any, 'correct' ideas. Then came Sraffa's contribution, with his Introduction to Ricardo's Collected Works, and his later Production of Commodities by Means of Commodities (i960), in which he

drew parallels between his own work and Ricardo's. Sraffa's implication, vigorously propounded by his followers, was that Ricardo belonged with him in an intellectual lineage radically distinct from, and superior to, the 'neoclassical' or 'supply and demand' tradition. That, in brief, is the background against which Samuel Hollander has proposed his own rehabilitation of Ricardo, a throwback to the interpretation of Marshall, from which Ricardo emerges as an intellectually virile progenitor of'neoclassical' economics. It is not a thesis which has been widely acclaimed, by Sraffians or, for that matter, anyone else. In the first section of the chapter, I consider whether Ricardo

Interpretations of David Ricardo

33

could be said to have anticipated, or appreciated elements of, a utility-based theory of consumer demand. This has long been an absorbing question for those who have attempted to gauge Ricardo's relationship with 'neoclassical' economics, for understandable reasons. In the case of Jevons, Menger and Walras, the putative founding fathers of 'neoclassical' economics, subjective theories of value were fundamental to their reconstructions of the subject. Needless to say, there were important differences between their approaches, but something approximating to the textbook distinction between total and marginal utility, together with the 'law' of diminishing marginal utility, was common to all three accounts, decidedly so for the contributions of Jevons and Walras. The 'founding fathers' were unanimous that Ricardo had not made significant progress in the subjective direction. Jevons acknowledged Ricardo's awareness that 'utility is absolutely essential to value' ([1879] 1965, p. 161), but apparently saw no trace of the essential distinction between total utility and, in Jevons's terms, (eventually diminishing) final degree of utility. Menger seems to have been equally unimpressed with Ricardo's contribution. Thus he 'protested' with reference to the treatment of rent that Ricardo had 'brought to light merely an isolated factor having to do with differences in the value of land but not a principle explaining the value of the services of land to economizing men' ([ 1871] 1981, p. 167). In short, Ricardo had failed to appreciate the all-pervasive, subjective nature of 'value'. Similarly, Walras castigated Ricardo for allegedly tracing the origin of'value' to labour ([1926] 1954, p. 201). For Walras, it was scarcity, the ability of something to command a positive price because of its positive rarete (marginal utility), which constituted the origin of'value'. By implication, the concept of (diminishing) rarete had eluded Ricardo. Alfred Marshall did his best to challenge those negative perceptions of Ricardo. Following Jevons, he pointed out that Ricardo had viewed 'utility' as being 'absolutely essential' to value. And he also drew attention to Ricardo's 'insistence', as he put it, that market price fluctuations are determined by supplies relative to 'the wants and wishes of mankind' (these last are Ricardo's words) (Marshall [1920] 1949, p. 670). This was all very well, but what of the crucial distinction between total and (diminishing) marginal utility (or something similar)? Applying his principle of 'general interpretation', Marshall believed that Ricardo had been 'feeling his way'

34

INTERPRETING RICARDO

towards it; and that, had it not been for his ignorance of'the terse language of the differential calculus', the distinction might have been stated explicitly (pp. 670-71 ). 46 For most subsequent commentators, however, with the notable exception of Samuel Hollander, this facet of Marshall's attempted rehabilitation of Ricardo was nothing more than wishful thinking.47 To summarise my position, it is plain that Ricardo did recognise utility (subjective esteem) as a precondition for exchange. And there is even a sense in which he can be credited with a crude, undeveloped notion of the distinction between total and marginal utility. But there is no evidence that he appreciated either the 'law' of diminishing marginal utility, or the 'equi-marginal' principle.48 Moreover, his insights to the utility doctrine, such as they were, must be placed firmly in perspective. In his deliberations on 'value', the later Ricardo was evidently concerned to find a monocausal principle which could explain the ratios of natural prices; and he hankered after an objective 'standard in nature' by which he could judge the direction and extent of changes in the 'values' of individual commodities. The utility doctrine, as he understood it, was irrelevant to his project. To say that Ricardo failed to anticipate or develop a subjective theory of demand is not to say that he was entirely silent on demand influences. This point has been latched upon by Samuel Hollander who, perhaps in order to compensate for Ricardo's disappointing performance as a 'subjective' theorist, compliments him for his 'particularly sophisticated' treatment of demand (1979, p. 273). On the basis of the evidence reviewed in the second section of chapter 6, I conclude that the compliment is unearned. Although Ricardo was cognisant of the 'law of demand' (the inverse relationship between price and quantity demanded) as a market phenomenon, and could even be said to have shown awareness of the 'elasticity property' without being an 'originator' of the elasticity concept, contrary to Professor Hollander - his scattered remarks are undistinguished by 46

47

48

The principle of'generous' interpretation was stated thus: 'If... we seek to understand him [Ricardo] rightly, we must interpret him generously . . . When his words are ambiguous, we must give them that interpretation which other passages in his writings indicate that he would have wished us to give them' (Marshall [1920] 1949, p. 670). See, for example, Haney (1924) p. 255 n. 1, Hutchison (1952), Mitchell (1949, pp. 175-7), Robbins (1936), Schumpeter (1954, pp. 600, 911-12), Stigler (1950) and Whitaker (1904, P- 45)The 'principle' being that individuals adjust their purchases so that ratios of marginal utilities are equal to ratios of prices.

Interpretations of David Ricardo

35

theoretical innovation or stunning insight. There is little here to detain the archaeologist of'neoclassical' doctrine. In the third section of the chapter, I turn to the vexed question of the role accorded by Ricardo to 'supply and demand' in price determination. According to Alfred Marshall, Ricardo was fully aware 'that the conditions of demand played as important a part as those of supply in determining value'. However, Marshall admitted that the coordinate role of demand and supply in price determination was not entirely obvious to all but 'the most careful readers' of Ricardo's Principles ([1920] 1949, p. 71). This was because Ricardo had assumed, without stating it explicitly, that all commodities 'obeyed the law of constant return' (p. 671), which made it seem as though conditions of supply exercised the sole, determining influence on price. But this was an appearance which belied Ricardo's 'true' position. Marshall's interpretation suggests that Ricardo had been thinking (with greater or less precision) in terms of the determination of price by the intersection of demand and supply schedules which, if true, might be taken as a movement in the direction of'neoclassical' economics.49 Without necessarily drawing the latter inference, a minority of commentators, recently joined enthusiastically by Samuel Hollander, have at least supported the gist of Marshall's interpretation. 50 But more have dissented.51 Within this latter stream of opinion, some writers allow that Ricardo introduced a 'supply and demand' explanation, of sorts, for the prices of nonreproducible commodities when these had departed from their natural levels. And some grant that Ricardo had given a 'supply and demand' rationale for the process by which market prices tend to equate with natural prices. But there is unanimity that Ricardo eschewed 'supply and demand', in any form, when it came to the determination of the natural price levels themselves. 49

50 51

O f course, it d e p e n d s on w h a t is m e a n t by 'neoclassical' economics. A c c o r d i n g to W a l r a s , w i t h o u t the concept oirarete (or s o m e t h i n g similar), a n d a n u n d e r s t a n d i n g of its relationship to price, 'it is impossible either to formulate or to d e m o n s t r a t e the law of supply a n d d e m a n d ' ([1926] 1954, p . 181). T h u s , if W a l r a s is o u r a u t h o r i t y on 'neoclassical' economics, a R i c a r d i a n theory of supply a n d d e m a n d without a subjective t r e a t m e n t of value is not a p r o p e r theory at all (cf. S c h u m p e t e r 1954, p . 613). H o w e v e r , it will b e c o m e clear t h a t there is n o need to agonise over w h e t h e r a m o r e ' v a c u o u s ' theory of supply a n d d e m a n d is, or is not, 'neoclassical'. A p a r t from H o l l a n d e r , also see Cassels (1935), V i n e r (1954) a n d R a n k i n (1980). See, for e x a m p l e , Blaug (1987, p p . 4 4 2 - 3 ) , D m i t r i e v ([1904] 1974, p p . 8 6 - 7 ) , D o b b (1940, p p . 9 - 1 0 , 1973, p p . 118-19, 1975), Eatwell (1975b, 1977), M e e k (1974), M y i n t (1948, p p . 6 2 - 3 ) , R o n c a g l i a (1982), Shove (1942) a n d Sowell (1974, p . 105).

36

INTERPRETING RICARDO

I argue that the majority opinion is essentially correct. On the central issue, which I take to be the determination of natural price levels. Ricardo was methodologically opposed to any form of'supply and demand' analysis. For, although he accepted that the process by which market price equates with natural price must involve output variations, he was adamant that (changes in) the natural price level itself must be explained solely in terms of (changes in) conditions of production. And he effectively denied the notion of a functional relationship between those conditions of production and supply, even for the case of corn. The determination of natural price by the intersection of a demand schedule with a supply schedule (horizontal or otherwise) was entirely foreign to Ricardo's analytical approach. The fourth section of the chapter is devoted to an examination of Samuel Hollander's thesis that Ricardo followed J. B. Say in according a 'key role . . . to opportunity cost and derived factor demand' (1985a, p. 18), with the implication that Ricardo 'belongs' within a Say-Walras tradition of analysis: 'Ricardo's line is that of Lausanne: equality of wage rates and of profit rates maximise the return to the factors "capital" and "labour"' (p. 24). In his defence, Professor Hollander attests that Walras 'certainly believed Say to have been on the right road' in his formulation of the 'mutual interdependence between product and factor markets incorporating the principles of opportunity cost and of imputing the values of factors from the values of their products' (pp. 19-20). But, Hollander exclaims, 'so did Ricardo' believe this of Say (p. 20, emphasis in original). Indeed, Ricardo is said to have both 'subscribed to' and 'applauded' Say's position (pp. 20, 27). Thus, a Walras-Ricardo link, which has managed to escape the attention of most other commentators, and Walras himself, is vicariously established.52 I argue that there is no substance to Professor Hollander's interpretation. Although it is possible to place an 'opportunity cost' construction on Ricardo's writings, the 'neoclassical' version imposed by Hollander is entirely inappropriate. Ricardo did not distinguish rigorously between 'capital' and 'labour' as factors of production, nor did he conceive of profit and wages as returns to the separate factors, somehow related to their productive contributions. As for the 'agreement' with Say, this was an attempt to convince the 52

Walras explicitly criticised Ricardo for claiming that cost of production determines selling price when, according to the Walrasian analysis of'imputation', it is selling price which (in part) determined the prices of productive services ([1926] 1954, Lesson 38, pp. 398-403).

Interpretations of David Ricardo

37

Frenchman that some of his pronouncements were consistent with Ricardo's doctrine. There was no acceptance of Say's 'imputational' perspective. Professor Hollander's work also figures prominently in the fifth section of the chapter, in which I consider whether Ricardo's economics has much, if anything, in common with (forms of) general equilibrium analysis. Precedents for an affirmative answer to this question were set by Edelberg (1933) and Barkai (1965), but it is the recent work of Professors Hollander and Morishima, conducted independently, which contain the most far-reaching and imaginative statements of the case. I focus on their contributions. According to Professor Hollander, there is a 'fundamentally important core of general-equilibrium economics' in Ricardo's work, which implies a 'strong continuity of doctrine' between Ricardian and later 'neoclassical' analysis (1987, pp. 6-7). From this perspective, the views that Ricardo attempted to divorce the analysis of distribution from that of general pricing, and that his work has more in common with the 'modern "Cambridge" school of economists . . . [which] champions . . . an approach involving the treatment of prices, production levels, and distribution by means of separate models with an eye upon the isolation of "one-waydirection" relationships or the "causal ordering" of variables' (1979, p. 689), are serious misconceptions if not, as Professor Hollander sometimes intimates, repugnant distortions of the 'historical record'. Professor Morishima is even bolder. Ricardo's economics is 'in fact' a general equilibrium theory (1989, p. 43), and 'Ricardo is the forerunner, if not the founder, of the general equilibrium school' (p. 149). By implication, Walras had, once again, inexplicably failed to recognise his doctrinal roots.53 On my interpretation, Walras's 'oversight' was entirely justified. Neither Hollander nor Morishima is able to cite a single instance of a fully developed 'general equilibrium' (style) analysis in Ricardo's writings. In an attempt to fill the void, both adopt the expedient of constructing versions of 'Ricardian' analysis from fragments of his work, reassembled in a pattern suited to their objectives. This would be sufficiently dubious in itself- it is enough to dispel the idea that Ricardo himself produced a form of general equilibrium analysis 53

Walras plainly believed that Ricardo had not treated the relevant economic variables as 'unknowns within the same problem'; unknowns which 'must always be determined together and not independently of one another' ([1926], 1954, p. 418).

38

INTERPRETING RIGARDO

but some essential elements of their reconstructions (such as essentially variable, endogenously determined commodity wages) are not even authentically Ricardian. 54 I conclude that the 'historical Ricardo' should not be confused with his 'general equilibrium' namesake. It would be possible to quibble endlessly over what is, or is not, a truly 'neoclassical' analysis. But if the above outline of my interpretation is correct, Ricardo's work fails to satisfy so many of the usual defining criteria that it is needless to pursue the matter. According to any known usage of the term, Ricardo was not a 'neoclassical'. The 'Sraffians' would celebrate that judgement. According to Pierangelo Garegnani's canonical version of 'Sraffian' historiography, Ricardo and Sraffa are located within a 'social surplus' tradition of economic analysis, variously distinguished by, first, the treatment of all outputs, techniques of production, and one distributive variable, as given data; secondly, the determination of the unknown distributive variable(s) as a residual; thirdly, the rigorous exclusion of 'supply and demand' analysis; and fourthly, a prime concern to analyse price changes following notional redistribution of the given (social) product. 55 It is my contention that all but the second of the above characteristics are either totally inapplicable to Ricardo, or applicable only in a qualified way. The basic problem with Sraffa's interpretation would seem to stem from a tendency to read more of himself into Ricardo than was warranted. Thus, the 'corn model' interpretation and the 'price-balancing' interpretation of Ricardo's third edition standard, two of the supporting pillars for the 'Sraffian' reading, are both instances of Sraffa having reconstituted Ricardo in his own image. No wonder that the 'Sraffians' find so much to admire in 'their' Ricardo. 54

55

Even Marshall found himself unable to say anything 'in defence of Ricardo's positive doctrine of wages' (letter t o j . B. Clark, 2 J u l y 1900, reprinted in Pigou, ed, 1925, p . 413). G a r e g n a n i (1983, 1984, 1987, 1990). Also see B h a r a d w a j (1983a, 1983b, 1990), d e Vivo (1987a), D o b b (1973, 1975) a n d Eatwell (1975a, 1975b, 1977).

CHAPTER 2

From bullion to corn: the early writings

In this chapter I propose a reconstruction of Ricardo's treatment of value, distribution and accumulation, from the time of his earliest writings to the period following the publication of his Essay on Profits Ricardo's contribution to the 'Bullion Controversy' is discussed in the first section, with a view to establishing his early position on matters relevant to this and to subsequent chapters (Ricardo's monetary analysis, as a topic in its own right, is not especially my concern). In the second section I consider the possible origins of Ricardo's new 'theory', if indeed it deserves such a grand description, first intimated by him in August 1813, when he suggested an important role for agricultural conditions of production in governing the course of general profitability. I also speculate on the content of this 'theory'. Then, in the third section, I turn to Ricardo's pre-Essay correspondence. This is the longest section in the chapter, and also the most difficult, which reflects the confusing, not to say confused, nature of Ricardo's writings during a period in which he was struggling to articulate new ideas within an old framework. The analytical content of the Essay on Profits is discussed in the fourth section, while in the fifth section I return to the pre-Essay writings and offer a reconstruction of Ricardo's more formal analysis at that time. All the evidence cited in favour of this reconstruction, which is merely supplementary to the interpretation developed in the third section of the chapter, is avowedly indirect but, then again, so too was the evidence cited by Sraffa in support of his 'corn model' interpretation (which I reject). The fifth section, while focusing on the post-Essay debate between Ricardo and Malthus over Ricardo's pricing assumptions, also briefly documents their other areas of disagreement. A short conclusion follows. 39

40

INTERPRETING RICARDO THE EARLY MONETARY WRITINGS

Ricardo first ventured into print in 1809, with an anonymous letter to the Morning Chronicle newspaper. At this time his preoccupation was with monetary issues as he sought to expose the inflationary consequences of the 1797 Bank Restriction Act, which had suspended the free convertibility of paper currency into bullion. His message was promulgated in The High Price of Bullion (four editions, 1810-11), the Reply to Bosanquet (1811) and the Morning Chronicle

(1809-10). As well as those published contributions, Ricardo also expressed his views in various sets of notes and in private correspondence, much of the latter with T. R. Mai thus, who was to remain his principal adversary on matters of political economy. Ricardo's central contention was that the Bank of England (and, following suit, the country banks) had overissued paper currency which, in the absence of convertibility, had led to an increase in prices of approaching 20 per cent, and a rise in the market over the mint price of gold, the latter being 'the only sure test by which we may know that [the paper issue] is . . . excessive' [Reply to Bosanquet in, p. 244).l To support his diagnosis, it was necessary for him to explain how the monetary system would operate with convertibility. His reasoning was as follows. Suppose that all countries operate a pure specie system with gold as the basic currency. Fundamental to Ricardo's account of the modus operandi of this and any other system with a metallic currency was the oft-repeated claim that the metal is subject to exactly the same 'laws' as all other commodities with regard to the regulation of its 'value' and the incentives for its exportation or importation. 2 In this context, and throughout the 'monetary' writings, 'value' was used mostly in a purely relative sense to denote the ratio(s) of exchange. Hence the value of gold would be ascertained by its command (in exchange) over other commodities and vice versa: 'Commodities measure the value of money in the same manner as money measures the value of commodities' [High Price of Bullion in, p. 104).3 If, then, a quantity of gold commands less of commodity X 1

2

3

The mint price of gold was legally fixed at £ 3 17s. ioV^d. per ounce. The market price, which was considerably higher, was the price in terms of inconvertible paper currency. See, for example, The High Price of Bullion in, pp. 54-5, 60, 103-4, an( ^ t n e letter to Mai thus of 18 June 1811, vi, pp. 23-8. Cf. Adam Smith, Wealth of Nations [1776] 1981, pp. 49-50. In one place, Ricardo did allude to the 'intrinsic value' of precious metals (and other commodities), which was said to depend on 'their scarcity, the quantity of labour bestowed

From bullion to corn: the early writings in country A than in country B (i.e., the money price of X is higher in A), gold would be 'cheaper' in A than in B (it would take less X in country A to purchase a unit of gold). Ricardo was emphatic that the general price level in any country was governed by strict 'quantity theory' considerations: '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 . . . [This is] in all cases the regulator of prices' (Notes on Jeremy Bentham's Sur Les Prix in, p. 311). Consequently, the effect of an increase in the quantity of gold {ceteris paribus) would be to raise the general price level or, equivalently, to depress the 'value' of gold (in Ricardo's terminology, gold would have become 'redundant' or 'superfluous'). But this is not the end of the story, because 'it is highly beneficial to us to exchange that commodity which is superfluous, for others which may be made productive' {High Price of Bullion in, p. 55). Consequently, in an unrestricted system, gold would be exported in exchange for foreign commodities. Then, because of the monetary efflux, prices return to a lower 'equilibrium' level such that the (gold) prices of exportable commodities in the trading countries are approximately the same. (The argument reverses if'the mass of commodities on one side of the line' increases: prices fall and, as a result of the cheapness of commodities, they are exported in exchange for gold: 'it is only when gold is high compared with goods, or in other words that goods are cheap, that any temptation is offered for its importation'. Letter to the Morning Chronicle, 29 August 1809, in, p. 16). The introduction of convertible paper money would not interfere with the 'quantity theory' mechanism. Banks 'could never issue more notes than the value of the coin which would have circulated had there been no bank' {High Price of Bullion m, p. 57), for the simple reason that, if they tried, paper would be converted into gold and exported.4 However, under a paper system without convertibility, there would be no such mechanism to offset the inflationary effects of (unwarranted) monetary expansion (p. 78). The suspension of convertibility was thus responsible for higher prices and the

4

in procuring them, and the value of the capital employed in the mines which produce them' (m, p. 52). This species of'value', which superficially clashes with the purely relative notion, received no further attention. See, for example, letter to the Morning Chronicle, 29 August 1809, in, p. 16; High Price of Bullion in, pp. 57, 60-1; Notes on Bentham's Sur Les Prix m, pp. 269-70.

41

42

INTERPRETING RICARDO

sustained rise of the market price of gold (in terms of paper) over the mint price. There are various aspects of Ricardo's position which merit closer attention. First, it is striking that his economic actors were presumed to respond with great alacrity to any monetary-induced disturbance. To take some examples: on the assumption that a new mine was discovered in a country, he claimed that gold and silver 'would immediately become articles of exportation' {High Price of Bullion in, p. 54); if, with a convertible currency, the banks were responsible for an unwarranted increase of paper money, 'the excess would be immediately returned to them for specie' (p. 57); and a 'void' in convertible paper 'would speedily be filled by importations of bullion' (p. 61). These three cases all involve intermediate links: a rise in domestic prices for the first two and a fall in prices for the third. Yet it was typical of Ricardo's natural style of reasoning that when he thought something should happen, he presumed it would happen, virtually without delay. This formed the basis for the (in)famous Ricardian distinction between 'temporary' and 'permanent' phenomena, first applied in the monetary context. So, for example, with convertibility, he considered that an excess of paper 'would produce little permanent effect on the value of the currency, because nearly an equal quantity of the coin would be withdrawn from circulation and exported' (p. 90). His overriding interest was in these 'permanent' effects rather than the intermediate, 'temporary' ones.5 Ricardo's confidence that economic actors would respond as he predicted was grounded in the belief that it 'is self-interest which 5

As R. S. Sayers remarked with reference to the monetary writings: 'To say that Ricardo ignored everything but the long-run [permanent] forces would . . . be a great injustice, but it is the fact that his real interest was confined to these forces, to the exclusion of short-period [temporary] disturbances' (Sayers 1953). However, it should be added that Ricardo was prepared to invoke 'temporary' phenomena when it suited him. One conspicuous instance is the following, from the Appendix to the High Price of Bullion: 'the operations of an increased currency are not instantaneous, but require some interval of time to produce their full effect' (111, p. 118). The context was an attempt to reconcile 'the principal maintained by me with the actual facts of the case' (p. 118). Similarly, when in 1819 he was asked by a Commons Select Committee to explain why the reduced supply of Bank of England notes between December 1817 and December 1818 had not produced any effect 'upon the exchange and on the price of gold' (v, pp. 371-72), contrary to Ricardo's theory, he replied that 'countervailing causes' were merely postponing an inevitable outcome (v, p. 377). In these instances, and in others (m, pp. 114, 118, 195), 'temporary' and 'countervailing' influences were used much as dei ex machina, to 'explain' any observed disparity between Ricardo's theoretical predictions and empirical reality, while the (unobserved) truth of the predictions continued to be asserted.

From bullion to corn: the early writings regulates all the speculations of trade' {High Price of Bullion m, p. 102). Thus, in the case of international monetary flows, gold is used to pay for imports only if it is relatively 'superfluous' in the importing country, for only then is it profitable to pay in gold (rather than in other commodities). As Ricardo explained to Mai thus: I wish to prove that if nations truly understood their own interest they would never export money from one country to another but on account of comparative redundancy. I assume indeed that nations ... are so alive to their advantage and profit . .. that in point of fact money never does move but when it is advantageous both to the country which sends and the country that receives that it should do so. The first point to be considered is, what is the interest of countries in the case supposed? The second what is their practise? Now it is obvious that I need not be greatly solicitous about this latter point; it is sufficient for my purpose if I can clearly demonstrate that the interest of the public is as I have stated it. (22 October 1811, vi, pp. 63-4) Ricardo seemed to be in no doubt that 'the interest of the public' was as he had stated it. Nor did he doubt that the bullion merchants did, in fact, behave in accordance with their interests. His theory took for granted: that whenever enormous profits can be made in any particular trade, a sufficient number of capitalists will be induced to engage in it, who will, by their competition, reduce the profits to the general rate of mercantile gains. It assumes that in the trade of exchange does this principle more especially operate; it not being confined to English merchants alone; but being perfectly understood, and profitably followed, by the exchange and bullion merchants of Holland, France, and Hamburgh; and competition in this trade being well known to be carried to its greatest height. Does Mr. Bosanquet suppose that a theory which rests on so firm a basis of experience as this can be shaken by one or two solitary facts not perfectly known to us? {High Price of Bullion m, p. 165)

Self-interested behaviour, the possession of accurate market knowledge (implicitly credited to the bullion merchants in the above passage), and impressively brisk responses to market signals were, and remained, the behavioural characteristics of Ricardo's 'economic man'. The forces regulating general profitability were also discussed in the 'monetary' writings, although on this subject Ricardo's views were later to change profoundly. Obliged to answer the argument that the behaviour of the interest rate was the litmus test of an

43

44

INTERPRETING RICARDO

'excessive' supply of money, Ricardo curtly instructed his opponent to consult Adam Smith's Wealth of Nations: 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. (Letter to the Morning Chronicle, 18 September 1810, 111, p. 153) Ricardo's approval for the 'competition of capitals' argument, variously described as 'incontrovertible' and 'manifest', was affirmed on numerous occasions.6 But what was it with which he so effusively agreed? Turning to The Wealth of Nations•, as Ricardo instructed, we find the following; As capitals increase in any country, the profits which can be made by employing them necessarily diminish ... There arises ... a competition between different capitals ... But upon most occasions he [the owner of one capital] 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 ... 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 ... [This] competition raises the wages of labour, and sinks the profits of stock ... [T]he profits which can be made ... are in this manner diminished, as it were, at both ends. (Smith [1776] 1981, pp. 352-3) It is possible to distinguish two influence on profits. The first, operating in the output market, may be taken to imply the interdependency of outputs, prices and profits: as new capitals enter a market, the increased aggregate output is associated with lower prices and, with a given money wage, lower profits. But also, the entry of a new capital implies an increased demand for labour which, with a given labour supply, results in higher money wages and, for a given product price, lower profits. Hence the second influence.7 Ricardo was also called upon to answer the argument that the increase in bank notes was the concomitant, perhaps even the cause, of Britain's recent economic growth. He did not deny that 'this 6

7

See, for example, letter to the Morning Chronicle, 20 September 1809, in, pp. 25-6; High Price of Bullion in, pp. 88-9, 92; letter to the Morning Chronicle, 24 September 1810, in, p. 150; and Reply to Bosanquet in, p. 19411. The two influences crop up elsewhere in the Wealth of Nations. Contrasting events in a 'thriving town' with those in 'remote parts of the country', Smith mentioned only the 'wage effect' of competition ([1776] 1981, p. 107). However, both influences, operating in tandem, were invoked subsequently (pp. 110-11).

From bullion to corn: the early writings

45

country has since 1797 greatly increased in wealth and prosperity' {Morning Chronicle, 18 September 1810, in, p. 141) and, to that extent, some increase in the circulation was justified. However, insofar as the market price of bullion had risen above the mint price, the increase of paper money must have been excessive {Reply to Bosanquet m, p. 244). As for the argument that the growth in the money supply had been responsible for the prosperity. Ricardo commented: There is but one way in which an increase of money no matter how it be introduced into the society, can augment riches, viz at the expence of the wages of labour; till the wages of labour have found their level with the increased prices which the commodities will have experienced, there will be so much additional revenue to the manufacturer and farmer [because] they will obtain an increased price for their commodities, and can whilst wages do not increase employ an additional number of hands, so that the real riches of the country will be somewhat augmented. A productive labourer will produce something more than before relatively to his consumption, but this can be only of momentary duration. (Notes on Bentham's Sur Les Prix HI, p p . 318—19,

my emphasis)8

In this passage we meet again the distinction between 'temporary' and 'permanent' effects, even though the words are not actually used. Clearly, the real effect of an increased money supply is constrained to a temporary interval: an interval so temporary as to be 'of momentary duration'. It is also assumed by Ricardo that there is a certain real (commodity) wage which prevails before and after the monetary disturbance. Some pages later in his Notes on Bentham's Sur Les Prix, the given real wage assumption is made explicitly: Labour is paid not by money but by money's worth[;] therefore if prices rise it will not occasion any increased production because more money must be given to the labourer to enable him to obtain the same amount of commodities, (m, p. 329) The presumption now is that an increased money supply will not result in any increased production: 'temporary' phenomena have been ignored completely. 8

Cf. Ricardo's letter to Mill of 1 January 1811: 'There appears to me only one way in which any addition would be made to the Capital of a country in consequence of an addition of money; it would be this. Till the wages of labour had found their new level with the altered value of money, - the situation of the labourer would be relatively worse; he would produce more relatively to that which he consumed, or rather he would be obliged to consume less. The manufacturer would be enabled to employ more labourers as he would receive an additional price for his commodities; he might therefore add to his real capital till the rise in

46

INTERPRETING RIGARDO

The assumption of a given real wage was to be a prominent feature of Ricardo's later writings. So, too, was the denial of'general gluts' (an implication of the 'law of markets'), an argument which also made its debut appearance in the early monetary writings. In an Edinburgh Review article of February 1811, Mai thus contended, in opposition to Ricardo, that a 'comparative redundancy or deficiency of currency' was not the sole reason for the exportation or importation of bullion. According to Mai thus, 'the varying demand for different sorts of produce arising from the varying desires and necessities of... nations' could be, and often was, an 'original cause' of bullion flows unconnected with 'redundancy' (Malthus 1811, pp. 342—3). To illustrate this argument he took the case of a poor harvest when, by supposition, money was not 'redundant'. Corn would be imported and 'if the debt [to the exporting country] . . . be considerable' a large part of it would be paid for in bullion rather than in (other) domestic commodities. Some commodities might be exported, but 'the prices of commodities are liable to great depression from a glut in the [foreign] market' and this would give a disincentive to their exportation. Hence the payment in bullion, the cause of which 'is to be found in the wants and desires of one of the two nations, and not in any original redundancy or deficiency of currency in either of them' (p. 345). Ricardo could not accept Malthus's argument. Although he granted that there might be a glut in particular markets in the corn-exporting country, the idea that there could be a general glut was roundly dismissed as fallacious. Thus: Whilst the desire of accumulation is not extinguished in the breast of man, he will be desirous to realise the excess of his productions, above his own consumption, into the form of capital. This he can only do by employing, himself, or by loans to others, enabling them to employ, an additional number of labourers, as it is by labour only that revenue is realized into capital. If his revenue be corn, he will be disposed to exchange it for fuel, meat, butter, cheese and other commodities in which the wages of labour are usually expended, or, what is the same thing, he will sell his corn for money, pay the wages of his labourers in money, and thereby create a demand for those commodities which may be obtained from other countries in exchange for the superfluous corn ... No mistake can be greater than to suppose that a nation can ever be without wants for commodities of some sort. It may possess too much of one or more the wages of labour placed him in his proper sphere. In this interval some trifling addition would have been made to the Capital of the community' (vi, pp. 16-17).

From bullion to corn: the early writings

47

commodities for which it may not find a market at home . . . but no country ever possessed a general glut of all commodities. It is evidently impossible. If a country possesses every thing necessary for the maintenance and comfort of man, and these articles be divided in the proportions in which they are usually consumed, they are sure, however abundant, to find a market to take them off. It follows therefore, that whilst a country is in possession of a commodity for which there is no demand at home, it will be desirous of exchanging it for other commodities in the proportion in which they are consumed. {High Price of Bullion 111, pp. 107—8, emphasis in original)

According to the first part of this argument, Malthus had failed to recognise that net savings will be channelled into capital accumulation in the corn-exporting country, thus generating a demand for (at least) extra wage-goods from the corn-importing country.9 The second paragraph then asserts the more general proposition that, in any case, there can never be a 'general glut' of all commodities. This blanket denial of general gluts was not entirely convincing, since Ricardo appeared to overlook the point stressed by Malthus, that increased supplies might be expected to depress commodity prices. All that matters, so it seems, are the proportions in which commodities exchange. In that respect, the argument is similar to the version of the 'law of markets' developed by James Mill in his Commerce Defended (1808), from which it possibly derived.10 I turn to the connections envisaged by Ricardo between corn prices, money wages and general prices. In accepting that some price increases could be attributed to 'the scarcity of corn, and the expences which have attended its importation' [Reply to Bosanquet in, p. 240), Ricardo intimated his approval for Adam Smith's thesis that, via money wages, general prices move in the same direction as The underlying 'no savings' principle was expressed more strongly, also during the early 'monetary' period, in a note written on Jeremy Bentham's Sur Les Prix: 'revenue is in all cases spent, but in one case the objects on which it is expended are consumed . . . in the other those objects form a new capital' (in, p. 299). The roots of this principle extend back to Adam Smith, who had written: 'In all countries where there is tolerable security, every man of common understanding will endeavour to employ whatever stock he can command in procuring either present enjoyment or future profit . . . A man must be perfectly crazy who, where there is tolerable security, does not employ all the stock which he commands . . . in some one or other of those . . . ways' ([1776] 1991, pp. 284-5). Mill asserted that 'a nation may easily have more than enough of any one commodity, though she can never have more than enough of commodities in general' providing that the commodities were supplied in their 'due proportion[s]' (Mill [1808] 1966, p. 137). John Stuart Mill noted that his father had greatly 'prized' his Commerce Defended for 'having been his first introduction to the friendship of David Ricardo' ([1871 ] 1987, p. 563).

48

INTERPRETING RICARDO

the price of corn.11 A similar impression is conveyed by his remark that a bread-tax would raise general prices 'as there is no commodity to the production of which the labour of man is not necessary' (Notes on Bentham's Sur Les Prix in, p. 270).12 This view of price interrelationships continued to be held by Ricardo until shortly before the publication of his Essay on Profits (1815), and I shall argue that it was integral to his agricultural 'theory' of profits. In the years to come, Ricardo's main preoccupation was with corn price increases wrought by diminishing returns on the land, not by taxation or 'scarcity' (reduced supply). Interestingly, in 1810 an argument due to Coutts Trotter (a banker) came to Ricardo's attention which contained some of the elements of the one later espoused by Ricardo himself. Trotter had written: 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 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. (111, pp. 388-9) In December 1810, or soon after, when, according to Sraffa, Ricardo's comments on Trotter were probably written (m, p. 380), the above scenario was tersely dismissed on 'competition of capitals' grounds as applied, it would seem, to the output sphere: 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, (m, p. 389) However, in his notes on Bentham's Sur Les Prix, we find Ricardo expressing his qualified approval for the 'diminishing returns' hypothesis: 'the decreasing power of the land to produce in proportion to the labour and capital employed on it' is a 'sentiment [which] is in the main undoubtedly true, - but I think it requires some quali11

12

Smith had written: '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' ([1776] 1981, p. 510). Cf. 'If a tax were laid upon bread, and, in consequence, the wages of labour were raised, the tax would eventually fall on all those who consumed the produce of the labour of man.' (Reply to Bosanquet in, p. 243).

From bullion to corn: the early writings

49

fication. — I see this is admitted in the next paragraph' (111, p. 287).13 It would be odd if this remark had pre-dated Ricardo's notes on Trotter and, indeed, it is possible that the annotation of Trotter's pamphlet had precedence.14 Be that as it may, Ricardo's early awareness of the 'diminishing returns' hypothesis is beyond dispute. I have drawn attention to several features of Ricardo's writings during the 'bullion' period: his claim that, for analytical purposes, money should be treated as a commodity like all others, his relative concept of 'value', his inclination to assume speedy responses to market disturbances, his dominant concern with 'permanent' (postdisturbance) rather than 'temporary' (transitional) phenomena, his boldly stated assumption of self-interested behaviour on the part of economic agents, his implicit assumption that agents have nearperfect knowledge, his pledges of allegiance to the 'competition of capitals' thesis, his use of 'fixwage' analysis, his denial of general gluts, his opinion that an increased price of corn, via money wage, leads to higher prices generally, and his awareness and qualified acceptance of the diminishing agricultural returns hypothesis. Of these, 'value' later acquires an additional, 'absolute' meaning, and the 'competition of capitals' thesis is eventually abandoned along with the view that higher money wages necessarily lead to higher prices. But before either of those changes came about, there was the development of Ricardo's agriculturally based 'theory' of profits. The possible origins and content of the 'theory' are considered next. RIGARDO'S NEW 'THEORY'

Monetary issues continued to dominate Ricardo's discussions with Mai thus until the summer of 1813, when there is evidence of a partial change in emphasis: the forces governing movements in the rate of profit became, for the first time in their correspondence, a disputed matter. Frustratingly, this development is preceded by a gap of over four months in the extant correspondence, added to which there is only a one-sided record of the exchanges of August 1813, consisting of two letters by Ricardo in which he first intimated an important role for agricultural productivity in the regulation of 13

14

The qualification 'admitted' by Bentham was that new land might be more productive by virtue of its natural fertility, climate, or 'some other circumstance', unspecified, which could make the same quantity of labour more effective (in, p. 288). Sraffa reports that the Bentham notes were written between 25 December 1810 and 11 January 1811 (m, p. 261).

50

INTERPRETING RICARDO

profits. Needless to say, any attempted reconstruction of the intellectual developments in this period must be highly speculative. From the first letter of 10 August 1813, it would appear that Ricardo and Malthus had been discussing 'the effect of opening new markets or extending the old', with Malthus arguing that the extension of foreign trade had, in reality, led to an increase in 'not only the nominal but the real value of our exports and imports' and 'a general increase of profits . . . [together with a] material growth of prosperity' (vi, p. 93). While accepting that the real value of trade had increased, Ricardo doubted the significance of this development: England may have extended its carrying trade with the Capital of other countries. Instead of exporting sugar and coffee direct from Guadaloupe and Martinique to the continent of Europe the planters in those colonies may first export them to England, and from England to the continent. In this case the list of our exports and imports will be swelled without any increase of British Capital. The taste for some foreign commodity may have increased in England at the expence of the consumption of some home commodity. This again would swell the value of our exports and imports but does not prove a general increase of profits nor any material growth of prosperity, (vi, p. 93) The apparent increase in foreign trade proved absolutely nothing. Missing from Ricardo's letter was an acknowledgement that there had been either a 'material growth in prosperity', or an increase in profits. Judging from his next letter to Ricardo, of 17 August 1813, Malthus must have quizzed him on those matters. Ricardo rejoined: 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 ... 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 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, (vi, pp. 94-5) The 'question' between Ricardo and Malthus had turned on the regulation of (general) profitability. Leaving aside, for the moment, the possible content of Ricardo's 'theory', what might have stimulated the change in the focus of his debate with Malthus?

From bullion to corn: the early writings In his masterful article on this question, G. S. L. Tucker drew attention to a number of possible stimuli which, either singly or in combination, could furnish an answer. The first was the ongoing debate with Malthus over currency issues. As Tucker observed, each of Ricardo's letters contained a paragraph immediately following those discussing 'wealth and prosperity' in which monetary issues were the subject of debate. In the letter of 10 August, the paragraph reads: 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, (vi, pp. 93-4) And in the letter of 17 August: I quite agree, that an increased value of particular commodities occasioned by demand has a tendency to occasion an increased circulation, but always in consequence of the cheapness of some other commodities. It is therefore their cheapness which is the immediate cause of the introduction of additional money, (vi, p. 95) Not unreasonably, Tucker suggested that the monetary issues apparently raised by Malthus in his missing letters, to which Ricardo was replying, may have been 'successive steps in a single line of reasoning' (Tucker 1954). Malthus's argument, as Tucker reconstructed it, could have been that the extension of foreign trade had resulted in a general rise in profitability and an associated rise in commodity prices; and that 'this had necessitated an increase of the means of payment; that although notes had certainly been issued in excess . . . some more moderate increase in the supply of money would have been justified' (Tucker 1954). We then find Ricardo rehearsing his argument that only a 'redundancy' of commodities (their cheapness in money) can facilitate the introduction of new money.15 The second possibility, perhaps the most tenuous, is that the question of 'profits' had arisen in connection with the chancellor of 15

In support of his suggestion, Tucker drew attention to passages from Malthus's later Principles which contain a very similar argument to his reconstructed one. He need not have looked so far ahead. The argument can be found in Malthus's An Inquiry into the Nature and Progress of Rent ([1815] 1970, p. 223).

51

52

INTERPRETING RIGARDO

the exchequer's 'New Plan of Finance', announced in March 1813. In connection with his proposals, the chancellor, Vansittart, had 'asserted that the rapid redemption of public debts would lead to a glut of capital seeking investment, and a disastrous fall of profits and interest' (Tucker 1954). Six years later, Ricardo wrote an article for the Encyclopaedia Britannica in which he specifically criticised Vansittart's assertion using 'law of markets' reasoning (iv, pp. 179-80). From this, Tucker inferred that Vansittart's proposals 'may have set Ricardo thinking about the effect of the accumulation of capital on the rate of profit', although he added that there 'is no direct evidence in favour of this supposition' (Tucker 1954). The third and most widely accepted possibility is that the new interest in 'profits' had been prompted by the 'corn law' controversy. In May 1813, the Select Committee on the Corn Trade published a report which was presented to Parliament on 15 June. The Committee recommended the setting of a high price before foreign corn could be imported into Britain, and the discontinuation of bounties on corn exports. It was claimed that these measures would increase the domestic corn supply (so reducing the dependency on foreign grain), provide for low and stable prices and, generally, contribute to the 'wealth and improvement of the country' (Smart 1910, p. 281). As Tucker observed, the issues raised by the Committee's report were 'calculated to arouse the interest of an economist' and, to that extent, this long-favoured explanation of Ricardo's initial concern with 'profits' retains a measure of plausibility.16 A fourth possibility, overlapping with Tucker's 'currency' suggestion, may be added to the list. In 1813, as the war with Napoleon entered its final phase, overseas markets were beginning to open up to British trade (see Smart 1910, p. 352). It is therefore no surprise at all to find Ricardo and Mai thus debating 'the effects of opening new markets or extending the old'; and the 'profits' question could have arisen quite straightforwardly from this. But to give Tucker the last, judicious word: 'perhaps the only conclusion that should be drawn ... is that a number of economic problems arose in the first half of 16

Samuel Hollander takes a different view: 'Any stimulus provided by the corn law debate for Ricardo's new theory is not a self-evident one' (Hollander 1979, p. 122, emphasis in original). His argument is that contemporary discussions did not explicitly link general profitability with agricultural productivity. But that is only to say that a relationship central to Ricardo's perspective was not articulated by others. The relevant question is whether 'external' developments may have stimulated Ricardo to make his own connections.

From bullion to corn: the early writings the year 1813, any one of which would have been sufficient to evoke Ricardo's interest in a theory of profits' (Tucker 1954). There remains the question of the content of Ricardo's 'theory'. According to Garegnani (1982), Ricardo had 'reached by 1813 [a] theory of profits which contrasted . . . sharply with the generally accepted one, running in terms of the "competition of capitals'", Garegnani's presumption being, it would seem, that Ricardo had arrived at the 'corn model' theory, attributed to him by Sraffa. Samuel Hollander agrees that Ricardo had 'abandoned' the competition of capitals thesis, although his own version of Ricardo's position, which I consider presently, differs markedly from the Sraffa—Garegnani interpretation. Ronald Meek, in contrast to both Garegnani and Hollander, perceived a 'marriage' between the Smithian doctrine and 'the current idea, then very prevalent in business circles, that a decrease in the price of corn . . . means higher profits' (Meek 1973, pp. 89-90): the competition of capital thesis had not been rejected comprehensively by Ricardo at this time. Ricardo's position was, to recall, 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' (vi, p. 95). This sentence is not unambiguous. Garegnani and Hollander take it to mean that the conditions of producing food govern profitability to the exclusion of competition of capitals influences. But it \s possible that 'new facilities in the production of food' were regarded as a counteracting influence on the 'Smithian' effects of'a rapid increase of Capital': the reading seemingly favoured by Meek. The evidence is too slender to support an unequivocal judgement, but regardless of whether Ricardo had broken from the Smithian position as a matter of logic, the fact is that he continued to believe, for some time, that his position was compatible with 'orthodoxy'. 17 There was no 'sharp contrast' in his opinion. As to the underlying reasoning for Ricardo's 'theory', there is no direct evidence whatever. But how might Ricardo have supported his (ambiguous) position. This question is addressed by Samuel Hollander, who proposes the following 'attractive reconstruction', as he calls it: the principle of diminishing returns was known to Ricardo at least as early as 1810... [I]n the same year Ricardo firmly rejected Trotter's position 17

This is substantiated in the course of the present chapter.

53

54

INTERPRETING RICARDO

that accumulation implies rising food prices .. . and accordingly rising money wages and general prices . . . Ricardo's objection to Trotter . . . turned upon the Smithian principle of 'competition of capitals.' But by 1813, when he himself had abandoned the principle on monetary grounds, the objection to Trotter had fallen away and Ricardo would have been left with the basic relationship linking diminishing returns to the money-wage rate . . . but with general prices held constant (on quantity theory grounds), the burden of the wage increase falling on profits. (Hollander 1979, pp. 117-18)

The problem is that Ricardo continued to argue, well into 1814, that a higher corn price would result in higher prices generally. This is acknowledged by Professor Hollander, who comments that 'the relevance for [Ricardo's] own approach to profits of the monetary argument used in 1813 to counter the principle of "competition of capitals" was apparently not self-evident' (Hollander 1979, p. 118, emphasis in original). Professor Hollander thus invites us to believe that Ricardo's 'theory' had originated in the application of a quantity-theory argument which was then forgotten. In agreement with Garegnani (1982), this is not particularly convincing. Professor Garegnani's alternative explanation, that Ricardo had adopted a 'corn model' treatment of profitability, at least has the merit of explaining how Ricardo could have arrived at a theoretical relationship between agricultural conditions of production and general profitability. Reasoning on the assumption that corn is both the output of, and sole input to, agricultural production, Ricardo might have inferred that agricultural conditions of production, reflected in the agricultural profit rate, uniquely determine the general rate of profit. But there is no evidence on which to credit Ricardo with such a 'refined' theoretical position. From the letters of 1813, we can infer only that Ricardo had come to regard agricultural conditions of production as an important influence on general profitability. At the expense of stating the obvious, an important influence is not the same as a unique determinant. An alternative reconstruction, based on Ricardo's recorded beliefs, may be offered as follows. We know from the monetary writings that Ricardo endorsed the Smithian view of all prices moving in the same direction as the corn price; and we also know that he expressed his qualified approval of the 'diminishing agricultural returns' hypothesis. Now, diminishing returns imply a reduction of produce 'in proportion to the labour and capital employed' (m, p. 287). But with the consequent price changes being

From bullion to corn: the early writings

55

common to both produce and 'the labour and capital employed', regardless of the physical composition of capital, it might have seemed evident to Ricardo, once his thought turned in this direction, first, that agricultural profitability must fall in such circumstances, and, secondly, that the establishment of a uniform rate of profit would involve a reduction in profitability elsewhere in the economy to the immutably lower agricultural rate. With 'decided improvements in agriculture', this downward influence on profits would be neutralised, even reversed, thus allowing 'a rapid increase of Capital' at the same, or at a higher, rate of profits. In this way, the facts would be 'perfectly reconcilable to my theory'. The letters of 1813 have raised more questions than answers. Let us move on to 1814. THE PRE-ESSAT CORRESPONDENCE

Following the letter to Mai thus of 17 August 1813, there is a break in the extant correspondence of over six months before the discussion of 'profits' is resumed.18 During this period Ricardo had drafted some 'papers on the profits of Capital' which he sent first to Malthus and then to Hutches Trower, both of whom gave their written comments.19 All these items are missing. However, we do have a letter of 8 March 1814 to a reportedly confused Trower ('To one not aware of the whole difference between Mr. Malthus and me, the papers you read were not clear, and I think you have not entirely made out the subject in dispute') in which Ricardo attempted to clarify his position. Malthus, it seems, had persevered with his argument of 1813: He thinks ... [that] if new markets are discovered, in which we can obtain a greater quantity of foreign commodities in exchange for our commodities ... profits will increase and interest will rise, (vi, p. 104) Ricardo disagreed, but he did not deny all common ground. Thus: When Capital increases in a country, and the means of employing Capital already exists, or increases, in the same proportion, the rate of interest and of profits will not fall. 18

19

Only three letters survive from Ricardo's correspondence during this time, each written by him, the last dated i January 1814. See Trower's letter to Ricardo of 2 March 1814 (vi, p. 102). Hutches Trower was, like Ricardo, a wealthy stockbroker and landowner, whose friendship with Ricardo had been formed at the beginning of the century (see Sraffa, vi, pp. xxiii-xxv). He was to make some astute comments on Ricardo's later treatment of value.

56

INTERPRETING RICARDO

Interest rises only when the means of employment for Capital bears a greater proportion than before to the Capital itself, and falls when the Capital bears a greater proportion to the arena, as Mr. Malthus has called it, for its employment. On these points I believe we are all agreed ... (vi, P- I0 3) The 'agreement' was that a higher (or lower) rate of profit is associated with an increased (or decreased) demand for capital relative to the supply of capital, just as the 'competition of capitals' doctrine required.20 In fact, this 'agreement' had little substance. Future correspondence will establish that Ricardo and Malthus differed {inter alia) over both the meaning of 'demand', and the causal mechanism by which profitability is 'permanently' altered. However, it is an arresting feature of the period under consideration that Ricardo believed he was working within the 'orthodox' framework, so contributing to the dense fog of confusion which shrouded his exchanges with Malthus. Ricardo continued his letter to Trower: 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, (vi, pp. 103-4) And in the next paragraph, introducing the concept of 'permanence': 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. (vi, p. 104) What can be deduced from these passages? Very little. It is taken for granted that agricultural profits must fall if more capital is employed on the land (ceteris paribus), that all other profits will fall in consequence, and that this outcome is compatible with 'competition of capitals' reasoning in the sense that 20

Cf. Ricardo's phrasing of the 'competition of capitals' thesis in his Reply to Bosanquet: 'the rate of interest is governed wholly by the relation of the amount of capital with the means of employing it' (HI, p.

From bullion to corn: the early writings

57

'the Capital bears a greater proportion to the arena . . . for its employment' (although note that the demand for capital is apparently deprived of any causal role). Precisely how 'the profits of the farmer ... regulate the profits of all other trades' is left unstated. Of course, we can always impose a precise meaning by translating 'regulate' as 'uniquely determine', in which case it is true, as opined by Garegnani (1982), that Ricardo's 'proposition can find a "rational foundation" only in the ["corn model"] argument described by Sraffa' (cf. Bharadwaj 1983b). But to take this as evidence in favour of the 'corn model' reading it must be supposed that the translation is justified, and that there was a logically sound 'rational foundation'. The letter to Trower offers no justification for either supposition. Following that letter, there is another break in the extant discussion of'profits', this time of over three months. It was during this period (in May) that parliament had resumed the debate on the proposed alteration to the Corn Law and, from this point onwards, the 'protection' issue dominated Ricardo's correspondence with Malthus. It can be inferred from Ricardo's letter of 26 June 1814 that Malthus had expressed doubts 'respecting the effects of restrictions on the importation of corn, in tending to lower the rate of interest' (vi p. 108). They were not doubts shared by Ricardo: T never was more convinced of any proposition in Polit: Economy than that restrictions on importation of corn, in an importing country have a tendency to lower profits' (vi, p. 109). In defence of his conviction he wrote: The rise of the price or rather the value 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. With the same Capital there would be less production, and less demand. Demand has no other limits but the want of power of paying for the commodities demanded. Every thing which tends to diminish production tends to diminish this power. The rate ofprofits and of interest must depend on the proportion of production to the consumption necessary to such

production, - this again 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, p. 108, my emphasis) According to Sraffa, the italicised line, described as a 'striking passage', is the 'nearest that Ricardo comes to an explicit statement on these [corn model] lines' (1, p. xxxii). That this reading should

58

INTERPRETING RICARDO

have seemed plausible is, perhaps, indicative of the extra-textual origins of Sraffa's interpretation (see below, p. 289). First, it is contextually obvious that Ricardo's expression did not refer specifically to events in the agricultural sector and, indeed, his main preoccupation seems to have been with 'other things' (non-agricultural commodities). Secondly, whereas Sraffa's interpretation suggests that Ricardo was alluding to a 'formula' for the agricultural profit rate, what he was stressing, with no particular reference to agriculture, was the dependency of profit rates on the magnitude of the surplus after defraying the cost of the direct labour input, with this cost 'essentially' depending on the 'cheapness of provisions'. Finally, exactly the same 'striking expression' had been used by Ricardo in his earlier writings, explicitly within a monetary context, and without specific reference to agriculture.21 The 'striking passage', in itself, offers no support for the 'corn model' interpretation. 22 Ricardo's ambition was to explain how an extension of cultivation, brought about by restrictions on corn importation, would impinge on other sectors of the economy. Taking for granted that agricultural profitability would fall, his aim was apparently to convince Malthus that the effect of restrictions would harmonise with 'competition of capitals' reasoning to the extent that an economywide reduction in demand must ensue. This would be so, Ricardo claimed, even on the counterfactual assumption (in terms of his current thinking) that the general price level does not rise with the corn price, for in that case profitability outside agriculture must fall if money wages increase (reflecting the higher corn price). This, in turn, supposedly leads to less production and therefore less demand within the manufacturing sector, reinforcing the reduction in demand from the agricultural sector. But what happens if prices do increase? Ricardo's phrasing (demand falls 'even if prices do not rise) suggests that this would reinforce the 'demand-effect', but it would also tend to subvert the inverse relationship between higher money wages and lower profits. The latter point was not lost on Malthus. In his reply to Ricardo of 6 July 1814, Malthus first questioned Ricardo's assumption that the amount of capital would be unchanged with restricted corn importation. Surely there would be 21 22

See above, p. 45, cf. p. 45, n. 8. But Sraffa is not alone in having thought otherwise. A p a r t from Sraffa's followers, the later S. H o l l a n d e r has also come to accept Sraffa's reading. See H o l l a n d e r (1986, p . 1093), a n d c o m p a r e with H o l l a n d e r (1979, p . 125).

From bullion to corn: the early writings 'much less capital\ he argued, and the 'usual effects' would follow (vi, p. 1 io, emphasis in original). This was a 'competition of capitals' argument. The 'usual effects' were a reduction in real wages and, via the output market, a rise in prices, combining to produce a higher level of profitability. The moral drawn by Malthus was 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' (vi, p. i n ) . As we have seen, Ricardo gave the impression that he accepted the traditional 'demand and supply' explanation of profits, but it was dawning on Malthus that they might mean different things by 'demand': 'The demand for capital depends, not upon the abundance of present produce, but upon the demand for the future products of capital' (vi, pp. 111-12, emphasis in original). There was indeed a difference. Ricardo's 'demand' was the amount actually demanded which depended solely on production. The implications of these different perspectives were to become gradually more apparent to both men. In an attempt to persuade Ricardo that the 'demand and supply' principle (as Malthus understood it) applied to the agricultural sector, as well as to other sectors, Malthus gave the following example: If it is necessary to employ a hundred days labour instead of fifty, in order to produce a certain quantity of corn, there seems to be no reason whatever that the person who possesses an accumulation sufficient to make the necessary advances should have a less remuneration for his capital. The effects of a great difficulty in procuring corn would in my opinion be, a diminution of capital, a diminution of produce, and a diminution in the real wages of labour, or their price in corn; but not a diminution of profits ... [unless there happened to be] a great abundance of capital. In short all will in my opinion depend upon the state of capital compared with the demand for it. This will be the prime mover, and it is this which will determine the profits which a capital employed in agriculture shall yield, whether the land be naturally rich or naturally poor, much worked or little worked, (vi, p. i n , emphasis in original) Here we do encounter a 'corn calculation', but with Malthus as its author. However, the illustration was poorly suited to Malthus's purpose, which was in part to establish that demand could sustain a high level of agricultural profitability regardless of the 'physical'

59

60

INTERPRETING RIGARDO

conditions of production. If agricultural production is undertaken by unassisted labour (as Mai thus appears to suggest), with real wages consisting of corn, then changes in the corn price would affect capital and product to the same extent: the output effect of the 'competition of capitals' would not be operating in agriculture. Malthus was to realise his blunder. Ricardo replied in his letter of 25 July 1814. He claimed that the effects of restrictions on the corn trade, and those of reductions in capital, were 'two operating causes' on profits which lacked a 'necessary connection'. Yet he conceded that a 'reduction of capital independently of restrictions on importation of corn will have a tendency to raise profits and interest', thus revealing his continuing attachment to 'orthodoxy'. The attachment is even more apparent in his attempt to explain why 'the difficulty and expence of procuring corn will necessarily regulate the demand for the products of capital'. Unlike the letter of 26 June, general price increases were now held to be directly responsible: 'the demand must essentially depend on the price at which [commodities] can be afforded, and the prices of all commodities must increase if the price of corn be increased' (vi, p. 114). The 'regulatory' mechanism was of a traditional demand and supply variety. On the 'realistic' assumption that all prices rise with the corn price, demand falls and therefore profitability declines. Ricardo then turned to Malthus's 'corn calculation': The capitalist 'who may find it necessary to employ a hundred days labour instead offiftyin 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 subistence that the labourers employed forfiftyhad 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? (vi, pp. 114-15) This passage is notoriously ambiguous. Take the interpretation favoured by Samuel Hollander (1979, p. 128, 1983, 1986), that the 'capital to be employed estimated in money' actually consists of corn. We would then be faced with a true 'corn calculation', with Ricardo making the point I anticipated above, that capital-product homogeneity rules out a separate 'price effect' on profits. However,

From bullion to corn: the early writings this would tell us nothing about the 'rational foundation' for Ricardo's own treatment of profits, since the 'corn calculation' had derived from Malthus. But is it a 'corn calculation'? Or rather, did Ricardo understand it as such? If capital comprises of corn wages, and assuming a slightly reduced per capita corn wage, then doubling the direct labour input would nearly quadruple money capital, not double it as Ricardo claimed. Perhaps Ricardo intended to write 'nearly quadrupled', but it is also possible that the 'nearly doubled' capital excluded the increased charges of production (cf. the Principles i, p. 117), in which case the reasoning could have been as follows. Following the doubled corn price, the 'original capital' (as Ricardo described it in the Principles 1, p. 117) is not itself doubled in money value because of the existence of non-corn inputs. However, Ricardo believed that all prices rise with the corn price, hence his conviction that the money capital 'is nearly doubled'. Finally, providing that the corn surplus is diminished after deducting the (corn valued) expenses of production, as it will be 'unless the labourers who are employed for a hundred days [are] satisfied with the same quantity of corn . . . [as] the labourers employed for fifty were before', the rate of profit must fall. Malthus returned to the 'corn calculation' in his letter of 5 August 1814: 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? (vi, p. 117) His understanding was that capital does consist of corn. He continued: 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, (vi, p. 117) According to Sraffa, this passage is 'no doubt an echo of Ricardo's own ['corn model'] formulation' (1, p. xxxi). To accept this reading, it is necessary to believe that Malthus's observation was not directly prompted by the internal dynamic of a close and unbroken correspondence which had extended over a six-week period, beginning

61

62

INTERPRETING RICARDO

with Ricardo's letter of 26 June 1814. Malthus, it would seem, abruptly saw fit to recall and criticise, for the first time, a formulation made by Ricardo in the lost 'papers on the profits of Capital' or in erstwhile correspondence (there is no evidence that they met during the six-week period). I can offer a more straightforward explanation. The passage in question immediately follows Malthus's ironic comment that profits must be increased if capital is only doubled, suggesting that it was intended as a related point. Malthus had realised that the capital—product homogeneity assumption, introduced by him in the letter of 6 July 1814, could backfire: it allowed no scope for the influence of 'demand, and of the abundance or scarcity of capital'. In counselling that 'we' (both of them) could not 'properly refer to a material rate of produce', he was therefore abandoning his own formulation which he believed, rightly or wrongly, Ricardo had borrowed. He was not making an obscure reference to past conversation or lost papers. My interpretation receives support from Ricardo's reply to Malthus in his letter of 11 August 1814 (which Sraffa passed over): 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, (vi, p. 121, emphasis in original)

All Ricardo defends, for 'monetary' reasons, is a 'material' conceptualisation of the amount of profit. Neither the rate of profit nor the assumption of capital—product homogeneity is mentioned, which, from Sraffa's Ricardo, would be a remarkable silence. Why should he miss such a fitting opportunity to defend, or even refer to, the putative 'rational foundation' for his theory? I return to Malthus's letter of 5 August, in which he reaffirmed that import restrictions 'must necessarily be attended with a diminution of capital, and therefore . . . they must tend to raise rather than lower profits' (vi, p. 116, emphasis in original). He agreed with Ricardo that the effect of higher money wages would be to reduce output in the manufacturing sector, and therefore raise prices. Where he parted company was in claiming that prices would rise to such an extent that profitability would increase. This difference again turned on the meaning of'demand': 'The whole amount of demand

From bullion to corn: the early writings

63

will from advanced prices diminish of course', as Ricardo had maintained, 'but the proportion of demand to the supply which is always the main point in question, as determining prices and profits, may continue to increase as it does in all countries the capital of which is retrograde' (vi, p. 117, emphasis in original). Malthus's conception of the meaning and role of demand was not Ricardo's. In his letter of 11 August 1814, Ricardo effectively denied that changes in the amount of capital could influence profits in any other way than through the conditions of agricultural production. Now allowing that a diminution of capital is a 'probable effect' following import restrictions, he wrote: I should say that a real diminution of capital will diminish the work to be done, and consequently will affect the wages of labour, and the demand for food. In the case supposed, restrictions on importation of corn, encouragement is given to the further cultivation of our own land - but if accompanied by a diminution of capital a discouragement is also given ... and whether profits rise or fall must in my opinion depend on the degree of these contrary operating causes, (vi, p. 119, emphasis in original) Changes in profits depend solely on the state of agricultural cultivation, without the independent causal role for 'demand' which Ricardo had seemingly allowed for in his letter of 25 July. He then endeavoured to explain why 'demand' could not counteract the agricultural influence: It is true [as Malthus had contended] 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 ... in money at the increased value to which all goods would rise in consequence of the rise of the wages of labour. - In such case I should say that the effective demand had diminished, because the same quantity of commodities could not be annually consumed. If [they were] ... then it must be evident that it would be at the expence of capital, (vi, p. 120) For Ricardo, the 'effective demand' falls if less commodities are consumed and, without a depletion of capital, less commodities are consumed if less are produced. But less commodities are produced only if the rate of profit falls, which implies that 'the rise of... goods

64

INTERPRETING RICARDO

will not be in the same proportion as the rise of labour'. By smuggling in his own concept of'demand' [alias supply) he had assumed the very thing that he set out to prove. A little over two weeks later, it was becoming more obvious to Ricardo where his difference with Malthus lay. 'I sometimes suspect that we do not attach the same meaning to the word demand', was his classic understatement. He continued: 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 larger quantity. (30 August 1814, vi, p. 129) Ricardo's 'demand' was equal to the quantity supplied and, consequently, 'demand and supply', as applied to output, could not explain differences in general levels of profitability: There may be two countries in one of which from bad government and the consequent insecurity of property, — or from the little disposition to saving in the people, profits may be permanently high . . . whilst in the other where these causes do not operate, profits may be permanently low ... It would surely be incorrect to say that the cause of the high profits was the greater proportion of demand for produce, when in both countries, the supply would be, or might be, precisely equal to the demand, and no more, (vi, pp. 128-9) However, he was prepared to accept that 'demand' could exercise a 'temporary' influence on profitability. After suggesting to Malthus that 'our principal difference is about the permanence of... effects', he added that 'the scarcity of a commodity, or the increasing demand for it will for a time increase profits' (vi, p. 128, my emphasis) . He did not elaborate, but it may be presumed that he believed the output of the commodity would be expanded, possibly by new market entrants, thus reducing profitability to the rate prevailing elsewhere in the economy.23 Malthus believed that he and Ricardo were agreed on broad principles, which is symptomatic of the continuing confusion in their debate, but he again perceived that their differences turned on the meaning of'demand': 23

The converse of Ricardo's proposition that the increasing demand for a commodity can only 'temporarily' increase profits is that falling demand can only 'temporarily' depress them, so undermining his earlier claim that the reduction in manufacturing profitability could be viewed as the result of lower demand.

From bullion to corn: the early writings

65

I think you would allow that when capital is scanty compared with the means of employing it . . . profits will be high not only temporarily but permanently. And I cannot help being of opinion that these high profits always indicate a comparative excess of demand above supply, even though the demand and supply should appear to be precisely equal . . . I by no means think that the power to purchase necessarily involves a proportionate will to purchase; and I cannot agree with Mr. Mill in an ingenious position which he lays down in his answer to Mr. Spence [in Mill's Commerce Defended], that in reference to a nation, supply can never exceed demand. A nation must certainly have the power of purchasing all that it produces, but I can easily conceive it not to have the will . . . You have never I think taken sufficiently into consideration the wants and tastes of mankind. It is not merely the proportion of commodities to each other but their proportion to the wants and tastes of mankind that determines prices. (11 September 1814, vi, pp. 131-2, emphasis in original)

In his reply, Ricardo misleadingly 'agreed' with Malthus on the point about 'scanty capital', although he added that it was 'very important to ascertain what the causes are which make capital scanty . . . and how far . . . they may be considered temporary or permanent'. As far as he was concerned, 'the state of the cultivation of the land is almost the only great permanent cause' (16 September 1814, vi, p. 133) by which he apparently meant, as before, that worsened conditions of agricultural production result in lower profitability and therefore less production—cum—demand: the fall in demand is not the cause of reduced profitability. This led into a spirited defence of 'Mr. Mill's idea, that in reference to a nation, supply can never exceed demand' (vi, p. 134) — the so-called 'law of markets' - with Ricardo arguing that the equality of supply and demand held regardless of the distribution of produce between wages and profits; and that 'the wants and tastes of mankind [are] unlimited', with the amount actually demanded only depending on the supply.24 The same and related points were made repeatedly by Ricardo in the letters that followed.25 For his part, Malthus 24

25

Samuel Hollander has suggested that 'the a p p a r e n t use m a d e of Mill's Commerce Defended for the first time in 1814 . . . a n d this at Malthus's instigation' lends support to the view that it was J e a n Baptiste Say's version of the 'law of markets', not Mill's, which h a d influenced R i c a r d o (Hollander 1979, p. 502, cf. Meek 1967). I d o u b t whether Professor Hollander's inference is justified. While it is true that Malthus first referred to Mill's work, he presumably did so because he h a d been reminded of it by Ricardo. Furthermore, it is clear that R i c a r d o was well prepared to defend ' M r . Mill's idea' in his reply to Malthus. T h u s : 'the wants a n d tastes of m a n k i n d . . . frequently occasion large profits on particular commodities for short periods . . . b u t they d o not . . . operate on general profits because they d o not often influence the g r o w t h of raw p r o d u c e ' (23 O c t o b e r 1814, vi, p p . 147-8);

66

INTERPRETING RICARDO

continued to insist that Ricardo underrated 'the wants and tastes of mankind in affecting prices' (9 October 1814, vi, p. 141); and that demand was causally responsible for higher prices and profits (p. 142). Their disagreement on these matters was total. There is an implication of the 'law of markets' that deserves our attention. But first, I consider an exchange of letters between Mai thus and Ricardo which took place towards the close of 1814. Malthus corresponded: I have never ... doubted or denied the general tendency of the accumulation of capital upon the land to diminish profits. But the acknowledgement of this obvious truth appears to me to be very different from the general position that the state of the land regulates profits ... 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 caeteris paribus be on a level. But the question is whether agriculture always takes the lead in the determination? and I should certainly say that it did not. (23 November 1814, vi, pp. 152-3, emphasis in original) Replying, Ricardo expressed his 'proposition' or 'principle' as being that 'the rate of profits can never permanently rise unless capital be withdrawn from the land' (18 December 1814, vi, p. 163). But this did not imply a 'determining' role for agriculture in the 'corn model' sense: 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; 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, vi, p. 162) This argument is similar to the one advanced in the letter of 11 August (see above, p. 63) in the sense that the agricultural influence on general profitability is supposedly dominant regardless of the sectoral origin of any perturbation. But here, explicitly, 'capital employed on the land' includes non-agricultural output (in the form of manufactured wage-goods) and this, according to Ricardo, makes no difference to his 'principle'. The 'lead role', as Malthus had 'Give men but the means of purchasing and their wants are insatiable' (p. 148); 'those commodities only will be produced which will be suited to the wants and tastes of mankind, because none other will be demanded' (p. 148); 'demand is regulated by production. Demand is always an exchange of one commodity for another.' (pp. 163-4).

From bullion to corn: the early writings described it, did not depend on capital-product homogeneity. But, of course, Ricardo had never claimed otherwise. I return to the 'law of markets' which, except for 'temporary' variations in particular rates of profit resulting from either the production of the 'wrong' commodities (those not suited to 'the wants and tastes of mankind') or, in the case of agriculture, from 'one or more bad seasons' (23 October 1814, vi, p. 146), implies that a reduction in general profitability cannot be explained in terms of an oversupply of commodities. The quantity supplied (in 'equilibrium') is the quantity demanded. In the early letters, however, a 'demand constraint' was apparently invoked by Ricardo to explain why manufacturing profitability falls if corn rises in price. But, as I have suggested, Ricardo in effect presumed the very fall in 'demand' (production) that he set out to explain, by taking for granted that general profitability must fall to the lower level established in agriculture. The question thus arises of how could he be so sanguine that agricultural profitability declines when corn is more difficult to produce? On Sraffa's interpretation, his confidence is easily accounted for: the use of more corn input relative to corn output implies that agricultural profitability declines and that manufacturing profitability falls commensurately. But there is no need to fabricate this rationale in order to understand Ricardo's confidence. We know he maintained that all prices rise with the corn price, and this implies that corn price changes affect product and capital in the same direction, if not to quite the same extent, regardless of the physical composition of capital. With diminishing agricultural returns involving a reduction in corn output for a given input of capital 'things' (including wage-goods), it may indeed have seemed evident that agricultural profitability must fall whatever the resulting corn price happens to be. To put this another way, a property of the 'corn model' is that price changes affect agricultural product and capital to precisely the same extent, so making it certain that agricultural profitability declines with the onset of diminishing returns. My suggestion is that Ricardo would have thought that price changes affect product and capital to roughly the same extent, not because of an assumed physical homogeneity of capital and product, but because of his continuing belief that general prices vary with the corn price. I have argued that two pieces of indirect evidence adduced by

67

68

INTERPRETING RIGARDO

SrafFa in support of his 'corn model' interpretation do not withstand scrutiny and that, in any case, the 'corn model' is not required in order to understand Ricardo's confidence that agricultural profitability falls as conditions of production deteriorate. My interpretation is not one which has Ricardo constructing, or relying upon, a logically consistent model, but I do not regard this as a weakness: there is no evidence that there was such a model. In the words Keynes used to describe his own intellectual development, there was rather 'a struggle of escape from habitual modes of thought and expression' (Preface to the General Theory 1936, p. vii), with Ricardo initially articulating his 'theory' using his old, familiar discourse. The agricultural 'theory', heralded in 1813, was constructed from existing materials, including the 'competition' of capitals' thesis, the distinction between 'permanent' and 'temporary' effects, 'orthodox' pricing assumptions, the denial of general gluts (supported by the 'law of markets'), and diminishing agricultural returns. ('Fixwage' analysis should be added to the list, although, to avoid further complicating the discussion, I postpone detailed consideration of Ricardo's treatment of wages until the next chapter.) Not all the old doctrines were well suited to Ricardo's new analytical direction, most conspicuously the 'competition of capitals'. And there was also the inconsistency, as Ricardo was soon to perceive it, between the 'commodity' treatment of money and the claim that all prices rise with the corn price. His 'struggle of escape' was just beginning. THE

ESSAY ON PROFITS

Ricardo's Essay was published on 14 February 1815, soon after Malthus's An Inquiry into the Nature and Progress of Rent and The Grounds of an Opinion on the Policy of Restricting the Importation of Foreign

Corn. It was only a few days after reading Malthus's pamphlets that Ricardo presented his own, incorporating the Malthusian treatment of (differential) rent into his argument against import restrictions on corn.26 As Ronald Meek pointed out (1973, p. 93), Ricardo developed his 26

Malthus's Inquiry was published on 3 February and his Grounds of an Opinion on 10 February (iv, p. 4). It was perceptively observed by Patten (1893) t n a t ' t n e n e w theory theory [of rent] fitted admirably into Ricardo's system, and made the proof of his theory of the relation of the price of food to profits more perfect. The law of rent came into Ricardo's system, not as a basis, but as a better proof of a theory already developed.'

From bullion to corn: the early writings analysis in two stages, with the first concentrating on agriculture. The agricultural analysis begins with Ricardo accepting Malthus's definition of rent: the rent of land [is] that portion of the value of the whole produce which remains to the owner, after all the outgoings belonging to its cultivation, of whatever kind, have been paid, including the profits of the capital employed, estimated according to the usual and ordinary rate of the profits of agricultural stock at the time being, (iv, p. 10)

We are then asked to consider 'the first settling of a country rich in fertile land . . . which may be had by any one who chooses to take it . . . without any deduction whatever for rent' (iv, p. 10). In this situation: if the capital employed . . . on such land were of the value of two hundred quarters of wheat, of which half consisted of fixed capital, such as buildings, implements, &c. and the other half of circulating capital, - if, after replacing the fixed and circulating capital, the value of the remaining produce were one hundred quarters of wheat, or of equal value with one hundred quarters of wheat, the neat profit to the owner of capital would be fifty per cent, or one hundred profit on two hundred capital, (iv, p. 10)

There is evidently capital on this land - 'buildings, implements, &c' - which is not of the same physical nature as output. Total capital is calculated as being 'of the value of two hundred quarters of wheat', presumably on the basis of a tacitly assumed set of wheat 'prices'. Wheat [alias corn) is not the sole agricultural input, explicitly so for items of fixed capital. Still with reference to the 'first settling', Ricardo continued: Profits might . . . increase, because the population increasing, at a more rapid rate than capital, wages might fall; and instead of the value of one hundred quarters of wheat being necessary for the circulating capital, ninety only might be required, (iv, p. 11)

To write that circulating capital is 'of the value of one hundred quarters of wheat' does not necessarily imply that commodity wages consist only of wheat. Nor is there an explicit assumption anywhere in the Essay that wheat is the sole wage-good.27 Hence, it may have been that Ricardo was adopting the same practice for manufactured wage goods as he had for the non-wheat elements of fixed capital, tacitly calculating their wheat value on the basis of 'given' relative 27

Samuel Hollander confidently refers to Ricardo's 'assumption in the Essay that corn alone enters the wage basket' (1979, p. 146). For a critique of his interpretation, see Rankin (1984).

69

JO

INTERPRETING RIGARDO

prices. Be this as it may, Ricardo had not assumed product-capital homogeneity in agriculture. Ricardo next drew attention to the possible influence on profits of variations in the wheat valued per capita wage, and of'improvements . . . in agriculture, or in the implements of husbandry (my emphasis) again acknowledging the heterogeneity of agricultural inputs. After remarking that these 'are circumstances which are more or less at all times in operation' (iv, p. 11), he assumed 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, p. 12) The farmer's rate of profits in the 'first settling' is, by assumption, 50 per cent, which is also taken to be the rate prevailing outside agriculture: '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, p. 12). As Malthus had remarked, this type of statement 'is merely saying what every body knows, that all profits must caeterisparibus be on a level' (cited above, p. 66). It is not a statement of'corn model' determination (cf. S.Hollander 1979, P-I39)- 28 Animating his 'model', Ricardo assumed that capital and population increase, so generating a rise in the demand for food which must be met 'from land not so advantageously situated'. The story unfolded as follows: [T]he necessity of employing more labourers, horses, &c. to carry the produce from the place where it was grown, to the place where it was to be consumed, 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. Suppose this addition to be of the value often quarters of wheat, the whole capital employed on the new land would be two hundred and ten, to obtain the same return as on the old; and, consequently the profits of stock would fall from fifty to forty-three per cent, or ninety on two hundred and ten. (iv, p. 13) 28

It is curious that 'trade' is mentioned at all, both here and elsewhere in the main 'agricultural' analysis, where Ricardo baldly asserts that general profits are 'regulated' by movements in the agricultural rate (pp. 13-14). The problem is that the (new) 'regulatory' mechanism is only detailed at a later analytical stage. These references may reflect the Essay's slipshod composition.

From bullion to corn: the early writings But what happens to the rate of profit on the first land cultivated? Ricardo wrote: [T]he return would be the same as before ... but, the general profits of stock being regulated by the profits made on the least profitable employment of capital on agriculture, a division of the one hundred quarters would take place, forty-three per cent, or eighty-six quarters would constitute the profits of stock, and seven per cent, or fourteen quarters, would constitute rent, (iv, p. 13) The wheat value of the capital on the first land is thus unchanged even though conditions of production on new land are inferior: this is the only way in which the return 'would be the same as before'. Since the total capital was not assumed to comprise wholly of agricultural output, the implication is that the exchangeable value of wheat relative to other commodities was treated as a constant. It was not allowed to change with altered conditions of production, and profitability, on the land 'regulating' general agricultural profits. In the continuation of the analysis, with 'land of worse quality, or less favourably situated' successively brought into cultivation, or land already in cultivation worked more intensively, the extra capital employed is consistently referred to as being 'of the value of so many quarters of wheat, and the relative value of wheat continues to be treated as a constant. The numerical analysis is reproduced in the Essay's table, with the first column heading again reinforcing the point that capital is only 'estimated in quarters of wheat' (iv, p. 17, my emphasis). Having explicitly allowed for heterogeneous inputs to agriculture, and on the tacit assumption that the wheat value of those inputs remains constant, Ricardo had illustrated the inverse relationship between rent and profits. If more wheat valued inputs are required to obtain a certain wheat output on new land, or on a portion of existing land, the rate of profits on that (portion of) land falls, being given by wheat output minus wheat measured inputs over wheat measured capital. But that rate of profit is applied to all (portions of) land. Ceteris paribus, less of the wheat surplus on those lands is required to yield the new, lower rate of profit, with the residual appropriated by the landlord as a rent which increases as the rate of profit declines. It remained for Ricardo to explain how the lower agricultural rate of profit would be communicated to other sectors in the

71

72

INTERPRETING RICARDO

economy. It was at this juncture that he announced his new 'rudimentary theory of exchange value' (Meek 1973, p. 93): The exchangeable value of all commodities, rises as the difficulties of 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. On the contrary, facilities in the production of corn, or of any other commodity . . . which shall afford the same produce with less labour, will lower its exchangeable value . . . Wherever competition can have its full effect, and the production of the commodity be not limited by nature, as in the case with some wines, the difficulty or facility of their production will ultimately regulate their exchangeable value, (iv, p. 19-20)

This 'theory', associating 'ultimate' changes in exchangeable value with 'difficulty or facility of production', sounded the rejection of Ricardo's earlier view that all prices must rise following a rise in the corn price. But what did Ricardo mean by 'difficulty or facility of production'? It might seem that by identifying conditions of production with the labour input, he had formulated a 'pure' labour theory of value, according to which (changes) in exchangeable value are governed by (changes) in the quantity of labour expended on the production of commodities.29 However, in claiming that corn would vary in price and exchangeable value simply because of a change in real (commodity) wages (iv, pp. ign., 22), Ricardo belied such an analytical perspective; and his subsequent correspondence confirms that the adoption of the labour theory came much later. Even so, Ricardo evidently believed that he had sufficient grounds to reject his previous, Smithian-inspired doctrine, which he did with the argument that money is a commodity like any other. The necessity to treat money as another commodity had been espoused in the early writings, but the long overlooked inference was now drawn that with a stable, corn-measured relative cost of producing the money and non-corn commodities, prices (except for corn) would be unchanged (iv, p. 2 in.). Having, at last, emancipated himself in this way from the idea of a positive relationship between movements in the corn price and general prices, Ricardo put forward the following mechanism for the regulation of general profitability: 29

An early interpretation along precisely these lines was advanced by West (1826, p. vii).

From bullion to corn: the early writings

73

The sole effect ... 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, (iv, p. 20)30 Compared , with the pre-Essay arguments, which were always questionable when it came to explaining the fall in manufacturing profitability, this was blissfully straightforward. I return to the 'agricultural' analysis, conducted on the assumption that the wheat value of non-wheat inputs to agriculture is constant even though conditions of production deteriorate as cultivation is extended (or intensified). According to the new treatment of exchangeable value, the wheat value of those inputs would fall as wheat became more difficult to produce (assuming that there is no change in their own conditions of production). By ignoring this change in relative prices, Ricardo had, in effect, assumed a constant price of wheat (and other commodities) throughout his 'agricultural' analysis. (Ronald Meek was, to my knowledge, the first to recognise this: Meek 1973, p. 93; also see Napoleoni 1975, pp. 64-5.) Hence the following passage which, far from being an 'aside' (S. Hollander 1979, p. 140), may be read as an informal conclusion to that reasoning: 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 have increased, whilst the value of the produce continued the same, (iv, p. 18, emphasis in original) Ricardo apparently believed he had loaded the argument against himself with the 'constant prices' assumption. This is hinted at in the above passage and emerges more clearly in the following extract: Not only is the situation of the landlord improved ... by obtaining an increased quantity of the produce of the land, but also by the increased exchangeable value of that quantity. If his rent be increased from fourteen to 30

This was amended soon after the Essay's publication: 'I would indeed rather modify what I have said concerning the stationary state of the prices of commodities, under all variations of the price of corn . . . I made no allowance for the altered value of the raw material in all manufactured goods; they would I think be subject to a variation in price not on account of increased or diminished wages, but on account of the rise or fall in the price of the raw produce which enters into their composition' (9 March 1815, vi, p. 179). Ricardo attributed his error to haste.

74

INTERPRETING RICARDO

twenty-eight quarters, it would be more than doubled, because he would be able to command more than double the quantity of commodities, in exchange for the twenty-eight quarters, (iv, p. 20, my emphasis)

As Malthus was to argue in the post-Essay correspondence, the 'price' effect (the application of this terminology in the present context is due to Faccarello, 1982) actually works against Ricardo because the farmer, like the landlord, also benefits from its favourable influence on profitability: if the price of wheat rises then the wheat value of non-wheat agricultural capital must fall. That this should have been overlooked may reflect the Essay's hasty composition. Ricardo had attempted to demonstrate how 'the general profits of stock . . . [are] regulated by the profits made on the least profitable employment of capital on agriculture' (iv, p. 13).31 He then attempted to explain how other influences conformed with this 'principle'. The behaviour of wages, and their effect on profits, was considered first. I postpone discussion of this part of his argument until the following chapter. Ricardo then confronted the position of'those who ascribe to the extension of commerce, and discovery of new markets . . . the progress of profits . . . [and who argue] that profits on agriculture no more regulate the profits of commerce, than that the profits of commerce regulate the profits on agriculture' (iv, p. 23). It was presumably Malthus, in particular, who Ricardo had in mind. His reply followed similar lines to those rehearsed in their correspondence. He allowed that 'the first discoverer of a new and better market may, for a time, before competition operates, obtain unusual profits'; but, he continued, 'those profits will themselves sink to the ordinary level' (iv, pp. 24—5). The underlying reasoning was that doughty profit-seeking capitalists would swarm into the 'better market', thus erasing the initial advantage. Only a 'temporary' variation in profitability was involved. (The same argument was 31

This 'principle' appears in different guises throughout the Essay, with Ricardo switching between the various formulations as if they were identical. Thus: 'profits are regulated by the difficulty or facility of procuring food' (iv, p. i3n.); 'general profits of stock depend wholly on the profits of the last portion of capital employed on the land' (p. 21); 'general profits on capital, can only be raised by a fall in the exchangeable value of food' (p. 22); 'the increase of the general rate of profits . . . can never take place but in consequence of cheap food' (p. 25); and, profits 'depend on the price, or rather on the value of food' (p. 26).

From bullion to corn: the early writings applied to the case of 'improved machinery at home' in particular markets: iv, p. 25) A further counter-argument ran as follows: 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, (iv, p. 24)

This argument hinges on the supposition of a given agricultural capital, not on the composition of that capital. Ricardo could accept that events outside agriculture, such as cheaper manufactured wage goods (iv, p. 26n.), and improvements 'in the implements of husbandry' (iv, p. 22), would affect agricultural profitability; but then, as he had claimed in 1814, cit will be true that less capital will be employed on the land' (cited above, p. 66). In terms of the agricultural analysis in the Essay, there would be less wheat measured capital. But assuming that the value of capital is given and fixed, and that the demand for wheat is also given (by 'the same population'), agricultural profitability cannot vary. Here, I suggest, we find the substance of Ricardo's position, turned into an 'argument' about other rates of profit with a flourishing reference to 'all the principles of political economy': 'all profits must caeteris paribus be on a level', as Malthus had said. The Essay contains much familiar material, although there are some novelties, the most important of which is the new treatment of exchange relationships. It also contains a far more detailed analysis of agriculture than anything to be found in the extant pre-Essay writings. What it does not contain is any support for the 'corn model' interpretation. Sraffa believed otherwise, claiming that the 'numerical examples reflect' the 'corn model' approach, 'particularly in the well-known Table which shows the effects of an increase of capital . . . and [where] profit per cent is calculated without need to mention price' (1, p. xxxii). But although the analysis did not 'mention price', that was only because of a tacit assumption of constant prices. It had nothing to do with capital-product homogeneity. Ricardo's analysis, including the 'regulatory' agency of farmers' profits, did not rest, and was evidently not thought by him to depend, on 'corn model' assumptions.

75

76

INTERPRETING RICARDO

THE PRE-^\SMr AGRICULTURAL ANALYSIS! FURTHER CONSIDERATIONS

The speed with which Ricardo was able to produce his Essay has been remarked upon. If, as Sraffa suggested, the first half was 'the revised version of a text prepared before the appearance of Malthus's pamphlets' (iv, p. 4 n.3), the swiftness of Ricardo's composition becomes credible. Let us then suppose that Ricardo did have a manuscript prior to February 1815. It may be conjectured, again following Sraffa (iv, p. 102 n.2), that this was a revised version of the 'papers on the profits of Capital', referred to in March 1814. What might this suggest for the pre-Essay treatment of agriculture? The agricultural analysis in the Essay proceeded with given and unchanging wheat values for heterogeneous inputs, reflecting the initial working assumption that all prices are given and constant. Before the Essay, however, Ricardo believed that all prices would rise with the price of corn. Had he at that time assumed, again as an initial working assumption, that prices varied pari passu with the price of corn, he could have produced an identical agricultural analysis to the Essay's, save for the underlying assumption about pricing. Later, when he came to revise his document, he would have reproduced the existing analysis with a different rationalisation. Given his new treatment of exchangeable value, it was no longer possible to assume that all prices rise with the corn price, and so he assumed that they were all unchanging, thinking, in his haste, that this actually strengthened his argument. Those revisions which he needed to make to the manuscript would have comprised the annotated references to Mai thus, and the incorporation of the theory of differential rent. And to this newly revised portion of the document would have been appended the remainder, containing a 'theory' of exchange relationships which played no active role in the agricultural analysis. Indirect evidence that Ricardo may have assumed equi-proportional price increases comes, first, from Malthus's letter of 9 October 1814: You 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 profits of stock ... But unless it could be shewn that no improvements were ever to take place either in agriculture or manufactures, and that upon a rise in the price of raw produce . . . the price of labour and of

From bullion to corn: the early writings

77

every other commodity both foreign and domestic would rise without delay exactly in

proportion, the doctrine is evidently not correct in practice. And as these contemporaneous effects are in my opinion not only improbable but impossible, it would be quite useless to lay much stress upon it even as a theoretical groundwork, (vi, pp. 139-40, my emphasis) Malthus continued: The Profits of stock ... 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 ... 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, (vi, pp. 140-1, emphasis in original) Malthus's perception of Ricardo's 'theoretical framework' was that it relied on contemporaneous and proportional price increases for all commodities following a 'rise of raw produce'. Moreover, if, as in the Essay, Ricardo's earlier 'agricultural' analysis had represented 'difficulty of production' as a rise in wheat measured inputs relative to wheat output, but at that time on the 'proportionality' assumption, it would be understandable for Malthus to interpret his 'proposition' as being that 'the quantity of produce compared with the expence of production . . . determines profits'. In his reply to Malthus of 23 October 1814, Ricardo seems to confirm that he had assumed proportional and contemporaneous changes between the prices of labour and corn (vi, p. 145) but he is silent on the question of price changes for commodities generally. However, it is possible that the particular and general cases were mutually implied. Here it is relevant to consider an argument put forward by Malthus in his Observations on the Effects of the Corn Laws.

Malthus was insistent that the 'expenditure of the labouring classes of society . . . by no means consists wholly in food, and still less, of course, in mere bread or grain'. The household budget is broken down as follows: 40% on 'meal or bread'; 20% on 'meat, milk, butter, cheese, and potataoes' which are 'slowly affected by the price of corn'; and 40% on 'house-rent, bricks, stone, timber, fuel, soap, candles, and clothing' which change 'still more slowly' following variations in the price of corn: indeed, 'as far as some of

78

INTERPRETING RIGARDO

them depend, in part or in the whole, upon foreign materials (as is the case with leather, linen, cottons, soap, and candles), they may be considered as independent of it; like the two remaining articles of tea and sugar, which are by no means unimportant in their amount'. Hence Malthus's conclusion that 'the whole of the wages of labour can never rise and fall in proportion to the variations in the price of grain' ([1814] 1970, pp. 98-9). Two things should be clear. First, that at this stage of their correspondence, Malthus could have been expected to tackle Ricardo for assuming corn as the sole wage good. And, secondly, that, with a mixed wage basket, the two assumptions of equi-proportional changes between corn and labour prices, and between corn and commodities (foreign and domestic), are inseparable. The fact that the corn-wage assumption was not included in Malthus's wideranging critique, taken in conjunction with Ricardo's implicit acknowledgement that he had assumed equi-proportional price changes between the prices of corn and labour, suggests that a more general application of the 'proportionality' assumption could have been quite acceptable to him. The letter of 23 October also dealt with Malthus's interpretation of Ricardo's 'proposition': 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 profits 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', (vi, p. 144, emphasis in original) At stake here may have been Ricardo's perception of the corn output to corn measured expenses formulation. Thus, when Ricardo 'concluded' the agricultural analysis in the Essay, in which the formulation had figured prominently, he characterised his result as being that 'the cost of production would have increased, whilst the value of the produce continued the same' (iv, p. 18, my emphasis). So, although the numerator in the rate of profit expression had been presented in physical terms, it was given a 'value' interpretation. At the same time, the Essay contains statements of precisely the sort apparently singled out by Malthus. Thus, the general rate of profit was said to depend on 'the produce compared with the cost of production on the land'; and 'every thing which shall augment the

From bullion to corn: the early writings

79

cost of production without augmenting the quantity of food, will . . . lower the general rate of profits' (iv, p. 26). Had an earlier exposition run along similar lines, both Malthus's 'misunderstanding' and Ricardo's 'defence' become explicable. In Ricardo's opinion, it would not have been that he was failing to allow for price considerations, rather that he was making the strong 'proportionality' assumption about price behaviour. Further indirect support for the suggestion that Ricardo may have assumed equi-proportional price movements comes from his correspondence with Malthus after the Essay's publication. Malthus correctly pointed out that if, as Ricardo now maintained, corn rises in price when it becomes more difficult to produce, with all other prices unchanged, then the wheat value of non-wheat capital on the land must fall. If this was taken into consideration, Malthus claimed, agricultural profitability might actually increase (vi, p. 185). I return to this exchange in the following section. Of relevance now is Ricardo's argument that if all commodities rose in price along with corn — a view he attributed to Malthus and Torrens — the criticism would be nullified: 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; - but your theory requires that there should be no rise in the price of those commodities on which the wages of labour are expended, for if they rose in the same proportion as corn, there could be no fall in the corn wages of labour. Is it not however very

improbable that all manufactures should rise at home and yet that those on which [the wages of] labour are expended should not rise? Is not the price of soap, candles &ca, though foreign commodities, necessarily affected by the rise in the price of those home goods which are given in exchange for them[?] (17 April 1815, vi, pp. 212—13, my emphasis) And further: I will... 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 [when additional corn is more difficult to produce] it will still bear the same proportion to the money value of the capital employed, (vi, p. 213 my

emphasis)

For there to be an unchanged relationship between the money value of produce 'on the old land' and the money value of capital when the price of corn increases, prices of agricultural inputs other than corn

80

INTERPRETING RIGARDO

(including non-corn wage goods) would have to rise proportionately: something explicitly stated for wage goods. This 'proportionality5 assumption was derived neither from Malthus nor Torrens. Malthus complained to Ricardo: 'Surely I have always maintained that when corn rises, though other commodities would rise they would not rise in proportion' (23 April 1815, vi, p. 222). He was right.32 Similarly, numerical examples in Torrens's Essay on the External Corn Trade reveal that he had not made the 'proportionality' assumption either.33 It is therefore conceivable that Ricardo was making use of the same analytically convenient assumption adopted by him before the Essay, at the time when he endorsed the positive relationship between corn and general price movements.34 To put this suggestion in perspective, it is by no means necessary to credit Ricardo with the 'proportionality' assumption in order to appreciate his confidence that agricultural profitability must decline when conditions of production deteriorate. Reinforcing this point, consider the following passage from Ricardo's post-Essay letter to Malthus of 8 May 1815: I have an account before me of the Capital actually employed on a farm ... It amounts to £3433 ... 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 &ca &ca. 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, (vi, p. 226, emphasis in original) In his defence of Sraffa's 'corn model' interpretation, John Eatwell seized on the observation that 'the fall in money will be common to both capital and produce' which he read as an echo of Ricardo's 32

33 34

See his Enquiry into the Nature and Progress of Rent (1815, p p . 24, 49); a n d also the a r g u m e n t reproduced from his Observations on the Corn Laws (1814): above, p p . 7 7 - 8 . Torrens (1815, pp. 83-4, 190). T h e assumption could have derived from a n acquaintance with J a m e s Mill's An Essay on the Impolicy of a Bounty on the Exportation of Grain (1804). According to Mill, 'the price both o f . . . raw material, a n d the wages of labour in all manufactures [Mill included overseas manufactures], are raised in exact proportion to the price of corn' ([1804] 1966, p. 64). As noted by Donald Winch, by the third edition of the Wealth of Nations, A d a m Smith h a d exempted foreign commodities from rising with the rise in the domestic corn price (p. 63n.).

From bullion to corn: the early writings

81

'corn model' reasoning (Eatwell, 1975a). But consider Ricardo's words: the agricultural profit rate must fall even though one third of capital is not subject to 'the fall in money value'. How much more confident he would have been if, as he previously believed, all prices moved in the same direction as the corn price. THE POST-£55/ir CORRESPONDENCE

Following the Essay's publication, Ricardo waited in some trepidation for a response from his peer group. He wrote to Malthus on 9 March 1815: 'My acquaintance lies so little amongst political economists that I have very few opportunities of knowing whether, what you consider as my peculiar opinions, have any supporters, or indeed whether they are read or attended to' (vi, pp. 178-9). He added: 'It is a matter of mortification to me that my execution has been so faulty, - I was too much in a hurry, and have not made my meaning intelligible even to those who are familiar with such subjects, much less to those who skim over these matters' (vi, pp. 178-9). Malthus responded: 'Considering the short time you were employed about it, the essay has great merit; but it might certainly have been improved by more time and attention' (10 March 1815, vi, p. 182). The element of praise was backhanded, since at least one 'improvement' would, in Malthus's view, overturn Ricardo's entire argument. In this section, I briefly review the debate sparked by Malthus's criticisms, which exposed the illconsidered nature of Ricardo's new treatment of value; and, for future consideration, I note the other main areas of controversy in the post-Essay correspondence. Malthus had noticed the dissonance in the Essay between Ricardo's agricultural analysis, with all prices held constant, and the new treatment of value which implied that corn-measured capital on the land would decline as conditions of production deteriorated. Not only would the landlord benefit from the higher exchangeable value of a given quantity of corn, as Ricardo had allowed, so too would the farmer, thus tending to raise the agricultural profit rate. This point was expressed to Ricardo orally soon after the Essay's publication.35 It was so perceptive, Malthus evidently believed, that it deserved another airing: 35

In his letter of 10 March 1815 Malthus wrote: 'I confess I think that the kind of calculation which I mentioned to you in Town [London], shews in what manner profits on land may

82

INTERPRETING RICARDO

Pray think once more on 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 offixedcapital, and of the circulating capital of tea sugar cloaths &c: for the labourers; and consequently more clear surplus will remain in the shape of rent and profits together, (no matter which) for home demand. Pray tell me whether any objection to this strikes you. (12 March 1815, vi, p. 185)36 In his reply to Malthus, Ricardo could not 'hesitate in agreeing . . . that if from a rise in the relative value of corn less is paid 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' (14 March 1815, vi, p. 189). However, something less than 'self-evident' was the impact of the 'price effect' for the farmer on the last (portion of) land cultivated. Here we encounter a stunning piece of sophistry. Ricardo's 'opinion' was that 'corn can only permanently rise in its exchangeable value when the real expences of its production increase. If 5000 quarters of gross produce cost 2500 quarters for the expences of wages &ca, and . . . 10000 quarters cost 5500 quarters . . . then the price would rise 10 pc* because such would be the amount of the increased expences' (14 March 1815, vi, p. 189). The problem is that the rise in the corn-measured expenses of production is incalculable without knowledge of the new corn price and does not, therefore, 'explain' that price. In agreement with Samuel Hollander, 'a more perfect case of circular reasoning would be difficult to imagine' (1979, p. 159). Ricardo continued to favour his 'circular' arithmetic, declaring its superiority over an alternative basis for calculation: 'the price of corn would not I think rise in proportion to the greater number of men employed but to the greater amount of wages paid' (letter to

36

rise decidedly, from the alteration in the relative value of corn'; and in the same letter Malthus refers to a conversation he held about the published Essay 'just before I left Town' (vi, p. 182). Malthus's visit to London probably began a day or two after the Essays publication, taking in Saturday 25 February which would have been the occasion for the monthly meeting of his club. In Peach (1986) I mistakenly suggested that 4 March 1815 was the meeting-day. The argument is elaborated in Malthus's letter to Horner of 14 March 1815 (vi, pp. 186-8). Malthus's self-satisfaction was manifest: 'The view I have taken of the subject would greatly alter his [Ricardo's] conclusions.'

From bullion to corn: the early writings Mai thus, 17 March 1815, vi, p. 193).37 The rejected approach was the one he adopted just over a year later. As part of his post-Essay defence, Ricardo repeated that it was the commodity-status of money which prevented all commodities from necessarily rising in price following a rise in corn: I have observed in the bullion pamphlet that many who say they consider money only as a commodity, and subject to the same laws of variation in value from demand and supply as other commodities, seldom proceed far in their reasoning about money without shewing that they really consider money as something peculiar, — varying from causes totally different from those which affect other commodities ... If money be a commodity does not corn and labour enter into its price or value? and if they do, why should not money vary as compared with corn and labour by the same law as all other commodities do? (Letter to Malthus of 27 March 1815, vi, p. 203)38 This is ironic. The first edition of Ricardo's High Price of Bullion was published in January 1810, and it had therefore taken him five years to perceive an implication of the 'commodity money' argument which he now paraded as almost self-evident. The episode does, however, provide a (further) warning not to presume consistency on Ricardo's part. According to Garegnani (1982) the post-Essay correspondence shows 'how unprepared Ricardo was as yet for any reasoning based on the relative value of commodities . . . as opposed to the argument based on physical quantities of corn'. This is partially correct. Ricardo's treatment of relative value was seriously flawed, as he came to realise, and there is no evidence that he, or anyone else, had subjected it to searching consideration before the Essay's publication. Relatedly, there is scant evidence to support Samuel Hollander's suggestion that Ricardo's post-i£«Yy> defence had been formulated before the Essay's publication (Hollander 1979, p. 162, 1986, cf. Prendergast, 1986a). On the contrary, all the indications are that the Essay was hastily assembled, that the new 'theory' of value - only the germ of an idea - was poorly considered, and that the unsatisfactory defence was in response to a criticism which was only articulated after the Essay's publication: there is nothing to suggest that it had previously occurred to Ricardo. Where Professor Garegnani goes astray, however, is in thinking that these considerations 37 38

Cf. Ricardo's letter to Malthus of 21 March 1815, vi, p. 197. Cf. Ricardo's letters to Malthus of 4 April 1815, vi, pp. 210-11, and 17 April 1815, vi, pp. 212-13.

83

84

INTERPRETING RICARDO

somehow point in a 'corn model' direction. Ricardo may have come unstuck with his new 'theory', but this proves nothing about the nature of a previous analysis, developed when subscribing to an 'old' theory. I turn to other areas of post-Essay controversy. Mai thus valiantly persisted with his argument that profitability is 'determined by the general supply of stock compared with the means of employing it, and not merely by the stock employed on the land' (10 March 1815, vi, p. 182).39 For over a year Ricardo continued to believe that he, too, accepted the 'competition of capitals' thesis, but in a form compatible with his own position. Thus he wrote in his letter to Mai thus of 8 May 1815: It is true that the rate of profits depends upon the scanty or abundant supply of capital compared with the means of employing it profitably, and these means will as you say upon the common principles of supply and demand be increased either by a diminution of capital or by an extension of the market for it. Our enquiry is in fact what the causes are of an extension of the market and I hold that the most powerful and the only one which operates permanently, is a reduction in the relative value of food, (vi, pp. 228-9) This was a similar style of accomodation with 'received' opinion that had characterised Ricardo's earlier correspondence and, as before, the autonomous influence of demand on profitability is only 'temporary'. The prime mover for 'permanent' variations in profitability is altered conditions of agricultural production. 40 A related line of attack from Mai thus was the criticism of Ricardo for 'greatly underrating] the effect of the wants and tastes of mankind, on which, after all every exertion of human industry depends' (8 September 1816, vn, p. 70).41 Both this and the former criticism were essentially directed at one facet of the emerging Ricardian system: the law of markets. Ricardo's attachment to this doctrine is discussed in the following chapter. Yet another line of attack from Malthus, also considered in the following chapter, comprised the twin theses that real wage vari39

40

41

Cf. M a l t h u s ' s letters to R i c a r d o of 18 April 1815, vi, p . 217; 23 April 1815, vi, p . 223; 5 M a y 1815, v i , p p . 2 2 4 - 5 ; 1 O c t o b e r 1815, vi, p p . 289-90; 11 O c t o b e r 1815, v i , p . 296; 30 O c t o b e r 1815, vi, p p . 3 1 9 - 2 0 ; 13 N o v e m b e r 1815, vi, p . 323; 28 April 1816, v n , p . 30; a n d 6 August 1816, VII, p . 52. See R i c a r d o ' s letters to M a l t h u s of 17 O c t o b e r 1815, vi, p . 305; a n d 9 A u g u s t 1816, VII, PP-57-8). Cf. Malthus's letter of 26 January 1817, VII, pp. 122-3.

From bullion to corn: the early writings

85

ations undermined Ricardo's 'agricultural' theory of profit regulation; and that a supposed downward secular tendency in real wages would counteract any influence on profitability from worsened conditions of producing food. Suffice it to say, at this juncture, that these arguments were also rebutted. Where Malthus 'succeeded', however, was in making Ricardo belatedly recognise the distance he had travelled from Smithian orthodoxy. In a letter of 9 October 1816, Malthus gave a succinct statement of his own position: Will it not be true in all cases that rent will depend upon the demand compared with the supply of good land, wages on the demand compared with the supply of labour, and profits on the demand compared with the supply of capital[?] (vn, p. 77) Here, unambiguously, he had separated the labour and output market dimensions to the 'competition of capitals' thesis. In his reply, Ricardo tentatively agreed ('perhaps') that wages and rent depended on demand and supply considerations. He continued: 'I do not quite understand the expression that profits depend on the demand compared with the supply of capital' (11 October 1816, vn, p. 78). Never again would he attempt to reconcile his treatment of profits with the 'competition of capitals' doctrine. CONCLUSION

The corn model interpretation has been found wanting. The 'striking passage' from Ricardo's letter to Trower of 8 March 1814 which, according to Sraffa, was the 'nearest that Ricardo comes to an explicit statement on these [corn model] lines', offers no particular support for Sraffa's reading. Malthus's objection that 'we can never properly refer to a material rate of produce', interpreted by Sraffa as 'no doubt an echo of Ricardo's own [corn model] formulation', is explicable in terms of the extant correspondence, with the 'corn model' formulation actually deriving from Malthus. And the analysis in the Essay, said by Sraffa to 'reflect' the 'corn model' approach, was in fact based on the corn valuation of heterogeneous inputs on the assumption that all prices remain constant: any 'reflection' of a 'corn model' analysis is wholly superficial. It is perfectly true that there are various 'strong' statements from Ricardo, including the pithy expression that 'it is the profits of the farmer which regulate the profits of all other trades', which, as

86

INTERPRETING RICARDO

Professor Blaug has remarked, are 'perfectly rationalized by the corn model construction' (Blaug 1985a, p. 711.). But there is no evidence that this was how they were rationalised by Ricardo himself. Indeed, a 'lead role' for farmers' profits was claimed (as in the Essay) despite the clearly acknowledged existence of heterogeneous inputs to agriculture. Even so, the 'strong statements' do suggest that there was thought to be something 'special' about agriculture, since it was the agricultural rate to which other rates supposedly conformed. I have offered an explanation for this that runs in terms of Ricardo's known beliefs and, specifically, his preEssay opinion that all prices move in the same direction as the corn price. With price changes being (roughly) common to both agricultural product and capital, irrespective of the composition of capital, it may have seemed natural to suppose that agricultural profitability must fall with the onset of diminishing physical returns, and hence, for a uniform rate of profit, that other profit rates must adjust to the lower rate in agriculture [mutatis mutandis for the case of 'improvements' in agriculture). Having thus arrived at the view that other rates of profit 'track' the agricultural rate, Ricardo then endeavoured to make everything conform to this same, mono-causal principle, including the case of cheaper manufactured wage goods which, supposedly, resulted in less capital employed on the land and therefore a sequential rise in agricultural and thence general profits. I have granted that my suggestion is not one that has Ricardo heroically creating a logically consistent model, fully vindicating a 'determining' role for agricultural profits in the 'corn model' sense. But, as I view this phase of Ricardo's work, it was marked not by the construction of neat, consistent models, but by the structured chaos of intellectual emancipation and development. As I have stated elsewhere (Peach 1987), the early writings contain no more than the overstated beginnings of a theory, not the fossils of one which had been perfectly formed.

CHAPTER 3

The falling rate ofprofit, wages and the law of markets

In this chapter I consider various aspects of Ricardo's mature analysis of distribution and accumulation. In the first section, I show that Ricardo's later theoretical objective (after the Essay) remained one of demonstrating a falling rate of profit in consequence of progressively diminishing agricultural returns; I argue that there is no mystery surrounding his retention of the (developed) 'agricultural' model; and I investigate the connection between the falling rate of profit analysis and Ricardo's case against agricultural protection. In the second section, devoted to Ricardo's treatment of wages, I establish that the Principles contains evidence of two conflicting doctrines, namely the natural wage analysis and the 'new view' analysis; I argue that it is only the natural wage analysis which is consistent with Ricardo's central thesis on 'permanent' changes in profitability, and with his characteristically 'discrete' style of analysis; I speculate on the reasons for the inclusion of the 'new view' passages in the Principles', and I consider whether Ricardo's position changed in later years. The 'law of markets', commonly interpreted as the doctrine that 'supply creates its own demand', provides the subject matter for the third section. It was this powerful doctrine, incorporating Ricardo's vision of real economic behaviour, which supported his 'prediction' of virtually unlimited economic growth, without any 'permanent' fall in profitability, in a British economy open to the free importation of corn. I recount the story of Ricardo's quite remarkable attachment to the 'law'; I comment on the implications for his reputation as an 'applied' economist; and I tender an explanation as to why the doctrine came to hold such a powerful attraction for him. A brief conclusion follows.

bb

INTERPRETING RICARDO

DIMINISHING AGRICULTURAL RETURNS, THE FALLING RATE OF PROFIT AND THE CORN LAW DEBATE

In the letter to Trower of 8 March 1814, which contained the aphorism that 'the profits of the farmer . . . regulate the profits of all other trades', Ricardo delivered his opinion that the profits of the farmer, and hence general profitability, 'must necessarily decrease with every augmentation of Capital employed on the land' in the absence of freely imported corn and improvements in husbandry (vi, p. 104). On the basis of the material reviewed in the last chapter, there cannot be any doubt that Ricardo's overriding objective in later correspondence, in the Essay, and in the immediate post-Essay correspondence, was to establish his 'theory' (Ricardo's word) of 'a diminishing rate of profit, in consequence of being obliged to cultivate poorer lands' (letter to Malthus, 4 April 1815, vi, p. 209). The first question addressed in this section is whether his objective subsequently changed. Writing to Jean-Baptiste Say in August 1815, Ricardo mentioned that: 'Mr. Mill wishes me to write it [the Essay on Profits] over again more at large. I fear the undertaking exceeds my powers.' Ricardo's fears at this time, which continued to plague him, mainly concerned his 'want of talent for composition' (vi, p. 249). But it is the visualisation of the future work as an expanded Essay that is significant, for in many respects the theoretical 'core' of the Principles indeed reflects this early conception. Ricardo made slack progress with the Mill-inspired project but, writing to Trower on 29 October 1815, he was in no doubt over the intended substance of his book: Mr. Malthus and I continue to differ in our views of the principles of Rent, Profit and Wages. These principles are so linked and connected with every thing belonging to the science of Political Economy that I consider the just view of them as of thefirstimportance. It is on this subject... that I should wish to concentrate all the talent I possess, not only for the purpose of establishing what I think correct principles but of drawing important deductions from them, (vi, pp. 315-16).l Ricardo's differences with Malthus had continued to centre on the role of the agricultural conditions of production in generating ('permanent') movements in profitability. So, for example, we find 1

Ricardo's authorship difficulties are discussed at greater length in the next chapter.

Falling rate ofprofit, wages and the law of markets Ricardo defending his 'opinion' that 'in the progress of society, independently of all improvements in skill and machinery, the produce of industry constantly diminishes as far as the land is concerned, and consequently capital becomes less productive [i.e., the rate of profit falls]' (7 October 1815, vi, p. 294).2 This was nothing more than a rewording of his position in 1814. Malthus continued to be a vigorous critic of Ricardo's 'agricultural' thesis, one of his arguments being that an alleged secular tendency for real wages to decline would offset the effect of higher corn prices on money wages and profits. A full consideration of this argument, and of its ramifications, is postponed to the following section of this chapter, but Ricardo's immediate response is of importance in the present context. He wrote: 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 these may balance each other, or rather that the latter will prevail, I on the contrary think the former the most powerful in its effects. I must write a book to convince you. (10 January 1816, vn, p. 10)

Even allowing for an element of hyperbole, Ricardo's concern to establish the connection between worsening conditions of agricultural production and general productivity is clearly evident. It must be stressed, in view of some contemporary misunderstanding of the point, that the 'principles' of distribution which Ricardo was so keen to explicate were not conceived in static, abstract terms. He later wrote to Malthus: I wish much to see a regular and connected statement of your opinions on what I deem the most difficult, and perhaps the most important topic of Political Economy, namely the progress of a country in wealth and the laws by which the increasing produce is distributed. (23 February 1815, vn, p. 24)

2

Cf. 'I contend that there are no other causes which will for any length of time make capital less in demand [implying a lower rate of profit], however abundant it may become, but a comparatively high price of food and labour' (17 October 1815, vi, p. 301). At this stage Ricardo was still expressing himself in 'competition of capitals' language.

89

go

INTERPRETING RICARDO

The study of distribution was to be located within the dynamic setting of the accumulation process.3 As the Principles testifies, Ricardo was himself wrestling with 'perhaps the most important topic of Political Economy'. Thus we read in the book's Preface that to 'determine the laws which regulate ... distribution' in 'different stages of society' is 'the principal problem in Political Economy' (i, p. 5). This project was a direct outgrowth from the Essay, itself the result of deliberations extending back to 1813. Even though there are important differences between the Essay and the Principles, it is possible to discern a central theme which encapsulates Ricardo's 'core' doctrine and attests to the essential continuity of his thought. Before substantiating this claim, brief reference must be made to Ricardo's treatment of value in the Principles, considered at length in the following two chapters. The analysis beyond the first chapter of the Principles is conducted using a 'pure' labour theory of value. Assuming that commodities are selling at their 'natural prices', those ruling when wage and profit rates are respectively uniform, (changes in) natural price ratios are taken to depend exclusively on (changes in) comparative amounts of labour expended in production, with the 'labour expended' covering both the direct, or 'living' labour-time spent producing something, and the labour-time expended on used-up tools, machines and raw material (more generally, anything required in the production process other than the direct labour). The analysis further assumes that one commodity, gold, is always produced with the same amount of labour-time. Using this commodity, the 'invariable standard', as numeraire, movements in the natural prices of other commodities reflect corresponding changes in the labour expended on their production. If, say, the natural price of a commodity rises by 20 per cent, this indicates a 20 per cent increase in the labour input per unit of its output. The second chapter of the Principles, 'On Rent', opens with Ricardo asking 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. (1, p. 67) 3

Cf. 'I shall be glad to see in a connected form your matured opinions on the progress of rent, profits, and wages, and in what manner they are affected by the increasing difficulty of procuring food, by the increase of capital, and the improvements in machinery. I fear we shall not agree on these subjects, and I should be very glad if we could fairly submit our different views to the public' (Ricardo to Malthus, 5 October 1816, vn, p. 71).

Falling rate of profit, wages and the law of markets

91

It will not: The exchangeable value of all commodities ... is always regulated, not by the less quantity of labour that will suffice for their production under circumstances highly favorable ... but by the greater quantity of labour necessarily bestowed ... by those who have no such facilities. (1, p. 73) Translating the theory of differential rent adumbrated in the Essay into the 'pure' labour theory discourse, the exchangeable value of agricultural produce is regulated by the greatest quantity of labour bestowed on output from what is, ipso facto, the least productive (portion of) land. And since rent is not paid on that (portion of) land, it cannot undermine the 'pure' labour theory.4 True to Ricardo's stated objectives in the Preface to the Principles, the chapter 'On Rent' also contained a disquisition on 'the effects of the natural progress of wealth and population on rent, in a country in which the land is of variously productive powers' (1, p. 78). The analysis is similar to the one contained in the Essay. Population is presumed to increase and, with the demand for corn determined by the size of population ('With the same population, and no more, there can be no demand for any additional quantity of corn', 1, p. 79) less fertile lands are taken into cultivation, or existing lands are worked more intensively. Hence, with the successive employment on the land 'of an additional quantity of labour with a proportionally less return', rent increases (1, p. 72). This is initially illustrated in terms of a declining corn surplus on new land relative to the employment of a given 'capital and labour': the procedure adopted in the Essay (indeed, the figures chosen correspond to the first three entries in the Essay's table: 1, pp. 70—1, cf. iv, p. 10.) Also reminiscent of the Essay, explicit consideration of the exchangevalue of corn is deferred to a later stage in the analysis (1, p. 72). The fifth chapter of the Principles, 'On Wages', poses some thorny interpretative problems, considered in the following section. But Ricardo's preoccupation with his 'agricultural' thesis is luminous, both at the beginning of the chapter, when he writes of a 'tendency' 4

As Ricardo later emphasised, in opposition to Mai thus and Say, the analysis of differential rent did not depend on there being whole tracts of land which pay no rent: 'they neither of them advert to the . . . principle which cannot be touched, of capital being employed on land, already in cultivation, which pays no rent' (letter to Mill, 22 December 1818, vn, p. 372); i.e., the so-called 'intensive' case. For similar protestations, see Ricardo's letters to McCulloch, 3 January 1819, vm, p. 4; and to Say, 11 January 1820, vm, pp. 149-50. Also see On Protection to Agriculture (iv, p. 240). The principle of differential rent was also applied to manufactured produce (1, p. 73).

92

INTERPRETING RICARDO

for (money) wages to rise in consequence of the 'tendency to a rise in the price of [agricultural] necessaries' (i, p. 93), and in the so-called 'new view' passages, where he attempts to show, in opposition to Malthus, that progressively diminishing agricultural returns will depress general profitability, even allowing for a secular fall in commodity wages (1, pp. 101—2). The sixth chapter, 'On Profits', reveals the same preoccupation. It 'remains for us to consider what is the cause of the permanent variations in the rate of profit' (1, p. 110). So opens the chapter 'On Profits', somewhat misleadingly. 'Variations' can be upward or downward, whereas the chapter is orientated to a detailed explication of the ('permanently') falling rate of profit thesis.5 This was the 'proof that the previous chapters had been building up to and, fundamentally, it amounted to a restatement of the central theoretical argument in the Essay, using the 'pure' labour theory of value, and now taking account of a change in the price of corn within the 'agricultural' analysis.6 The analysis proceeded as follows. It is initially assumed that a diminishing corn output is produced by the labour often workers on (portions of) land successively taken into cultivation. Since the same number of labourers are always employed on the rentless land, regardless of its physical yield, the monetary value of the output from that land is constant (according to 'pure' labour theory reasoning). However, with the physical output declining as new land is taken into cultivation, the unit price, which applies to all corn, rises in proportion to the presumed physical decrease. As this process continues, the monetary value of the greater physical output from the 'better' land increases, but these gains pass to the landlord in rent, leaving the farmers on all lands with an unchanging money revenue out of which to pay their labourers. If, as Ricardo assumed, corn products enter the (given) commodity wage basket, money wages must rise with the corn price, and therefore each farmer's residual profit will be reduced. Similarly for the manufacturers, if 5

6

The Principles does contain the ingredients for a more general theory of 'permanent' movements in profitability, which may be expressed as follows. Abstracting from the effects of taxation, 'permanent' movements in profitability are associated exclusively with altered conditions of producing the 'natural' wage bundle. The proponents of the 'new view' interpretation might be expected to dissent from this statement of Ricardo's doctrine, since it includes the natural wage as an integral element. This aspect of my reading is defended in the following section. Ricardo's failure to allow for a revaluation of corn in his analysis of agriculture had been the particular feature of the Essay which attracted Malthus's criticism (see above, pp. 81-2).

Falling rate ofprofit, wages and the law of markets the difficulty of producing their commodities is unchanged, so too will be the natural prices of those commodities and also the monetary value of any given output. Higher money wages therefore erode the profits of the manufacturers (1, pp. 113-15).7 Ricardo next turned to the rate of profit, first on the assumption that there is an 'original capital' for the farmer which, in money terms, does not change as new land is brought into cultivation; and then allowing for a rise in capital value on the grounds that capital 'consists in a great measure of raw produce, such as . . . corn and hay-ricks . . . unthreshed wheat and barley . . . horses and cows, which would all rise in price in consequence of the rise of produce'. 8 The effect of making the latter allowance is that the farmer's rate of profit falls to an even greater extent. The manufacturer's capital is assumed to be of the same monetary value as the farmer's, 'and therefore his [rate of] profits would conform to the altered rate of those of the farmer' following a rise in the corn price and money wages. This corresponds to the first case of an unchanged 'original value' of capital. Taking the second case, Ricardo commented that 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' (1, p. 117). Although he does not state it explicitly, the reasoning seems to be that the monetary capital of the manufacturer would also rise, with conformity between the farmer's and manufacturer's profitability again achieved.9 The implications of a declining general rate of profit for capital accumulation were then set out: no one accumulates but with a view to make his accumulation productive ... Without a motive there could be no accumulation ... The farmer and manufacturer can no more live without profit, than the labourer without wages. Their motive for accumulation will diminish with every diminution of profit, and will cease altogether when their profits are so low as not to afford them an adequate compensation for their trouble, and the risk which 7

8

9

Ricardo confusingly refers to the rate of profit in the course of this analysis, but the explicit focus is on the shares of wages and profits in an output of constant value. The calculations imply that 23.33 P e r c e n t of capital 'consists . . . of raw produce'; later in the chapter, in another calculation, this rises to 50 per cent (1, p. 122). As Samuel Hollander has observed, the analysis seems to imply that wages are not included in the 'original capital' (Hollander 1979, p. 206). Ricardo was careful to point out that if commodities 'rise in price with the rise of wheat... 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' (1, pp. 117-18).

93

94

INTERPRETING RIGARDO

they must necessarily encounter in employing their capital productively, (i, p. 122)10

Ricardo had newly explicated his thesis that the 'natural tendency of profits . . . is to fall; for, in the progress of society and wealth, the additional quantity of food required is obtained by the sacrifice of more and more labour' (1, p. 120). The content of the chapters on rent, wages and profits, together with Ricardo's earlier correspondence, amply confirm that, if anything deserves to be called his 'primary objective', it was to establish his thesis of a falling rate of profit in consequence of diminishing agricultural returns. Moreover, Ricardo's summary expression of his doctrine clearly betrays his 'agricultural' focus. Thus he claimed that 'in all countries, and 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' (1, p. 126). But profits could equally be said to 'depend' on the labour required to produce manufactured necessaries, given the conditions of producing corn (a mixed commodity wage is assumed explicitly throughout the chapter). Ricardo's preference for the one-sided 'agricultural' expression reflects his long-standing analytical preoccupation.11 It may still be wondered why, in Samuel Hollander's words, 'Ricardo chose to adopt an analytical framework incorporating the principle of diminishing agricultural returns and the stationary 10

11

In the words of Samuel Hollander, this statement implies 'a regular positive dependency of savings upon the rate of return on capital' (1979, p. 318). But on the basis of his exhaustive review of the evidence, Professor Hollander concludes that this cannot be said to represent Ricardo's 'true' position (see, especially, Hollander 1979, pp. 310-26), since the idea of a diminishing relationship between capital supply and the rate of profit was more often denied than affirmed by Ricardo. According to Piero Sraffa, 'while the theory that the profits of the farmer determine all other profits disappears in the Principles, the more general proposition that the productivity of labour on land which pays no rent is fundamental in determining general profits continues to occupy a central position' (1, p. xxxiii). On this view, there is continuity between the earlier and later writings with respect to 'the more general proposition', but not with regard to the more specific 'theory' that agricultural profits determine general profits (in the mathematically precise, 'corn model' sense). On my interpretation, the early writings provide no evidence that Ricardo had formulated a 'corn model' analysis, so all that disappears is a phrase, not a tightly woven theory. Moreover, the similarity between Ricardo's earlier and later treatment of profits is far greater than Sraffa implied with his notion of a transition between structurally dissimilar models. Thus, Ricardo continued with his sequential approach of first establishing a fall in agricultural profitability in consequence of diminishing agricultural returns, and then explaining why manufacturing profitability must conform to the reduced agricultural rate. The 'lead role' for agricultural profits, as Malthus had once described it, continued to be a feature of the later analysis.

Falling rate of profit, wages and the law of markets

95

state'? (1979, p. 640).12 On the interpretation of Ricardo as a 'pessimist', who believed that progressively diminishing agricultural returns would rapidly propel a 'closed' economy into the growthless stationary state, the question would not even arise. But as Samuel Hollander, and others, have argued persuasively, the view of Ricardo as a 'pessimist' is unfounded.13 Even with restrictions, Ricardo predicted that the stationary state 'is yet far distant' [Funding System 1820, iv, p. 179); and, in the same vein, he wrote in earlier correspondence that 'we are happily yet in the progressive state, and may look forward with confidence to a long course of prosperity' (to Trower, 4 February 1816, vn, p. 17); similarly, he wrote in later years to his more alarmist ally, James McCulloch, that 'we may ... have a more limited measure of prosperity notwithstanding the continued operation of our corn laws' (letter to McCulloch, 23 March 1821, vm, p. 358). Ricardo was not a prophet of economic gloom.14 His position on the likely course of 12

13 14

Although Professor Hollander thinks that 'a definite answer' cannot be given to the question, he does speculate on a number of possible answers. He claims, first, that the form of Ricardo's growth model, and particularly its incorporation of diminishing agricultural returns, was 'dictated' by Ricardo's ambition to combat the 'competition of capitals' thesis (1979, p. 640). However, the argument in the chapter 'Effects of Accumulation on Profits and Interest' in the Principles, which was specifically designed to refute the 'competition of capitals' thesis, stands independently of the presence, or absence, of diminishing agricultural returns. Secondly, he suggests that it is 'probable that the notion of ultimate stationarity . . . was "taken for granted" by Ricardo' (1979, p. 640). This is also unconvincing. The main thrust of Ricardo's argument was that 'ultimate stationarity' could be avoided by the liberation of the corn trade. Thirdly, he claims that the 'principle of diminishing agricultural returns had for Ricardo a theoretical significance quite apart from its utilization in the analysis of secular profit rate movements. I have in mind the corollary of the principle, namely that at the various margins rent is absent, thus permitting the easier treatment of the distribution problem' (1979, p. 641). This is no more successful, since, according to the theory of differential rent, rent is absent even if all agricultural output is produced under identical conditions. Fourthly, Professor Hollander suggests that the 'conception of diminishing returns was also of practical significance in the sense that, despite recognition of technical change in agriculture, progress would be yet more rapid in manufacturing. There would in this case be a secular rise in the relative price of corn and at least some of the empirical propositions derived from theoretical reasoning on the basis of unchanged technology would still hold good' (1979, p. 642, emphasis in original). However, the 'propositions' that would not 'hold good' include rising money wages and a 'permanently' falling general rate of profit, which happen to have been among the main corollaries of diminishing agricultural returns, at least for Ricardo. More promisingly, Professor Hollander suggests in a later work that 'since technical progress takes the form of random shocks it was to be accorded secondary status compared to diminishing returns' (1985b, p. 36). This is undoubtedly true, but it does not amount to a full recognition of the central importance that Ricardo attached to his 'agricultural' model. For references to those who have held the opposing views, see above, p. 7 n. 6. This point receives vivid confirmation in the 'law of markets' section of the present chapter.

0,6

INTERPRETING RIGARDO

profitability was, it seems, that although diminishing returns had dominated over agricultural improvements in reality, and would continue to do so without the free importation of corn, the resulting upward pressure on money wages was insufficient to produce more than a very gradual fall in general profitability, especially when account was taken of the falling prices of manufactured wage goods. Moreover, cheaper luxuries, together with the 'natural' disinclination to export capital, would combine to offset the negative effect of lower profitability on accumulation.15 It is this relatively 'optimistic' outlook that may lend a degree of pertinence to Hollander's question. However, in order to appreciate the significance of the 'agricultural' model, we need only consider Ricardo's view of the likely course of profitability without restrictions. In December 1814, he wrote to Malthus: If there were no increased difficulty [in obtaining food], 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. (18 December 1814, vi, p. 162) This argument, supported by the law of markets, was reproduced in the Essay (iv, p. 18), in subsequent correspondence with Malthus (17 October 1815, vi, p. 301), and in the Principles.16 Later, in his notes on Malthus's Principles, Ricardo pronounced that with free importation of corn 'there will be no transfer neither from profits or wages to rent' in the progress of society (11, p. 222). On the same theme he wrote to Trower: I contend for free trade in corn on the ground that while trade is free, and corn cheap, profits will not fall however great be the accumulation of capital. (21 July 1821, vm, p. 208) Similar views were also expressed in Ricardo's article on the Funding System (iv, p. 179) and in parliamentary speeches (v, pp. 55, 188). In the absence of restrictions on the free importation of corn, it was therefore Ricardo's belief that the British economy could look forward to a golden future in which profits would never ('per15

16

My position, thus far, is sufficiently close to Samuel Hollander's (1979, chapter n ) that there would be little gain from reviewing all the supporting textual evidence. See the passages quoted below, pp. 114-15.

Falling rate of profit, wages and the law of markets

97

manently') fall and the spectre of the stationary state was forever banished. Set against that future, the prospect of a protectioninduced decline in profitability, however gradual, and an ultimate slide into stagnation, however distant, assumed a greatly magnified importance. A rising corn price posed the sole, ongoing threat of 'permanently5 falling profitability. No wonder, then, that the principle of diminishing agricultural returns should have been incorporated in Ricardo's study of the 'natural course' of wages, profit and rent, his 'optimism' notwithstanding. There is also the associated matter of the Corn Law connection. If, as many commentators have argued, the falling rate of profit argument was central to Ricardo's case against agricultural protection, this would make his retention of the 'agricultural' model even less surprising. For Professor Hollander, however, Ricardo's case 'against agricultural protection was not based upon the secular downward trend

in the rate of return on capital' (1979, p. 604, emphasis in original). Or rather, to take the footnoted version of his contention, it was based on the 'secular' argument, but only 'as an exception to the rule, and argued for public consumption' (p. 609 n. 13). The remainder of this section will be spent evaluating Professor Hollander's interpretation. In June 1814, Ricardo wrote in private correspondence to Malthus that he 'never was more convinced of any proposition in Polit: Economy than that restrictions on importation of corn in an importing country have a tendency to lower profits' (26 June 1814, vi, p. 109). This boldly stated conviction, which cannot be dismissed as an aside for 'public consumption', is reflected in the content of the Essay on Profits, published the following year. In the Introduction to the Essay, Ricardo explained that consideration of the principles of rent, 'together with those [principles] which regulate the profit of stock, have convinced me of the policy of leaving the importation of corn unrestricted by law' (iv, p. 9). These 'principles' were explicated in the first half of the Essay, with Ricardo showing that a closed economy would experience a secularly falling rate of profit [etcetera) owing to diminishing agricultural returns (ceteris paribus). His case against the Corn Laws in the Essay was directly centred on a 'secular trend' argument (cf. iv, p. i6n). There is no reason to suppose that this was a disingenuous tactic, somehow unrepresentative of Ricardo's 'true' position. Professor Hollander chooses to emphasise a passage from the Essay

98

INTERPRETING RICARDO

in which Ricardo claimed, following Malthus's lead, that a great increase in 'wealth and prosperity' was possible even with restrictions, albeit at a slower pace (iv, p. 34). Professor Hollander comments: 'His major point is simply that the rate of growth is likely to be higher in a free system than in a protected system? (1979, p. 609, emphasis in

original). The reason why growth would be higher is that the price of corn (and therefore money wages) would be lower, so increasing profitability which is the spur for, and means of financing, the accumulation of capital. However, the reason why the domestic corn price had risen above the overseas price was, according to Ricardo, that diminishing returns had outweighed the counteracting effects of agricultural improvements.17 And he apparently believed that this net tendency for a rising corn price was only to be expected in a restricted system.18 His claim that growth would be higher in an unrestricted system was therefore complementary to his projection of a downward trend in profitability (owing to rising food prices), not divorced from it as Professor Hollander appears to suggest (1979, pp. 600-1). Nor, in the Essay was the differential growth rate argument his 'major point' against the Corn Laws: the prominence he gave to the demonstration of falling profitability is evidence of that. Rising corn prices, falling profitability, and the prospect of higher growth in an unfettered economy, were interrelated propositions. The Essay was written to demonstrate the 'inexpediency of restrictions on importation' (those words form part of its full title) whereas the Principles bore the imprint of a more general treatise on 'the laws of profits and wages, and on the operation of taxes' (Preface, 1, p. 6). Yet the contrast between the two works is not so great as this may suggest, since both were concerned with the 'principles' of distribution and, more specifically, with the analysis of the 'natural course' of wages, profit and rent under the influence of diminishing 17

18

Ricardo did not 'attempt to deny' that 'great improvements have been made in agriculture', although he added that 'with all those improvements, we have not overcome the natural impediments resulting from our increasing wealth and prosperity, which obliges us to cultivate at a disadvantage our poor lands, if the importation of corn is restricted or prohibited' (iv, p. 32). War-time restrictions had forced the cultivation of domestic land, with the result that the overseas corn price was 'not much above half that at which we can ourselves produce it on some of our poorer lands' (iv, p. 30). Thus, 'the price of corn . . . has been invariably observed to rise as a nation became wealthy, and was obliged to have recourse to poorer lands for the production of part of its food; and very little consideration will convince us, that such is the effect which would naturally be expected to take place under such circumstances' (iv, p. 19).

Falling rate of profit, wages and the law of markets

99

agricultural returns. Moreover, although the main chapters on distribution in the Principles lacked the same policy orientation of the Essay, they still contained direct and indirect allusions to the deleterious affects of a restricted corn trade, a refrain that is also heard in subsequent chapters. 19 The application of the falling rate of profit analysis to the 'protection' issue may not have been so bold, or sustained, as it had been in the Essay, but the relevant connections were made. 20 I turn to Ricardo's later writings (from 1819) on which Professor Hollander relies heavily (as he must) to support his contention(s). It was during this period that Ricardo had taken (or rather, purchased) his seat in Parliament and, with others, was exercised by the vexed question of the 'agricultural distress'. Events had unfolded as follows.21 The post-war battle for a liberated corn trade had been lost. A new Corn Law was passed in March 1815 which prohibited the importation of foreign corn until the domestic price had been at, or above, 80 shillings a quarter for six weeks, when the ports could be opened to unlimited duty-free importation for three months. The harvest in 1815 was excellent and was followed by a falling corn 19

20

21

'Wealth increases most rapidly in those countries where the disposable land is most fertile, where importation is least restricted, a n d where through agricultural improvements, productions can be multiplied without any increase in the proportional quantity of labour, and where consequently the progress of rent is slow' ('On Rent', i, p . 77); 'With the progress of society the natural price of labour has always a tendency to rise, because one of the principal commodities by which its natural price is regulated, has a tendency to become dearer, from the greater difficulty of producing it. As, however, the improvements in agriculture, the discovery of new markets, whence provisions may be imported, m a y for a time counteract the tendency to a rise in the price of necessaries, a n d m a y even occasion their natural price to fall, so will the same causes produce the correspondent effects on the natural price of labour' ('On Wages', 1, p. 93); 'However extensive a country may be where the land is of a poor quality, a n d where the importation of food is prohibited, the most moderate accumulations of capital will be attended with great reductions in the rate of profit, a n d a rapid rise in rent; a n d on the contrary a small b u t fertile country, particularly if it freely permits the importation of food, may accumulate a large stock of capital without any great diminution in the rate of profits' ('On Profits', 1, p . 126); ' N o point in political economy can be better established, than that a rich country is prevented from increasing in population, in the same ratio as a poor country, by the progressive difficulty of providing food. T h a t difficulty must necessarily raise the relative price of food, a n d give encouragement to its importation' ('On the Comparative Value of Gold, Corn, a n d Labour, in Rich and Poor Countries' 1, p . 373); 'prohibitions of the importation of corn increase the demand, a n d drive us to the cultivation of poorer lands' ('Bounties on Exportation, a n d Prohibitions on Importation', 1, p p . 313-14). Also see Ricardo's letter to Trower of 21 J u l y 1820, in which he referred to the chapter ' O n Profits' in support of his argument in favour of a liberated corn trade (vm, p . 208). M y brief account is based on Smart (1910, chapters 24-36), and Smart (1964, chapters 1-10).

IOO

INTERPRETING

RICARDO

price towards the end of the year, with the fall continuing into 1816, much to the despair of the agriculturalists. In 1816, however, crops were deficient with the per quarter price of grain rising to over 103 shillings in January 1817, peaking at 111 shillings in June. Free importation was triggered, as it was in 1818 (when imports were the greatest on record) and 1819. Due to the flood of imports, combined with good domestic harvests, prices eventually dropped, falling to approximately 69 shillings in June 1819, 67 shillings in December, 65 shillings during the first half of 1820, and 55 shillings during the second half. But, for the landed interest, worse was to come. The harvests of 1819 and 1820 were good, while that of 1821 was superb. Prices plummeted. From 54 shillings a quarter in January 1821, corn fell to 51 shillings in May and, after rising briefly, closed the year at 47 shillings. The year 1822 witnessed the lowest average corn price since 1792, with a fall to 34 shillings during November. The seven years following the triumphant passing of the 1815 Corn Law had been disastrous for the agriculturalists. They clamoured for a revision of the 1815 Act. From 1819 to 1821, 1,200 petitions were presented to Parliament, 'the greatest number of petitions ever presented to Parliament on one subject' (Smart 1964, p. 3). Parliament's response was to set up a Committee to investigate the causes of, and possible remedies for, the 'distress', with Ricardo as one of its members. As he wrote to McCulloch in 1822, T have no hope of good measures being adopted, the landlords are too powerful in the House of Commons to give us any hope that they will relinquish the tax which they have in fact contrived to impose on the rest of the community' (8 February 1822, ix, p. 158). Ideally, he envisaged a totally free trade in corn, although he granted that in view of the widespread and severe distress, and taking into account the influence of landlords, it would be prudent to adopt interim protectionist measures which could be phased out gradually. He placed his proposals before the public in the pamphlet On Protection to Agriculture, published on 18 April 1822, and aptly described as a 'minority report of the [Agricultural] Committee' (J. Hollander [1910] 1968, p. 54). Briefly, he reluctantly condoned a 'monopoly of the home-market to the British grower till corn reaches seventy shillings per quarter' (iv, p. 263). Then, importation should be allowed but with a duty of 20 shillings per quarter, which would be reduced by 1 shilling every year until it reached 10 shillings. Growers would also

Falling rate of profit, wages and the law of markets

101

be allowed a drawback (a type of subsidy) of 7 shillings a quarter which, like the 10 shillings duty, would supposedly compensate them for the peculiar taxes they incurred. The advocacy of this package, described as 'a substantially free trade in corn' (iv, p. 266), was very much an exercise in realpolitik?2 But the compromise did not go far enough for Parliament and, once again, Ricardo found himself on the losing side. The rationale for Ricardo's proposals, concessionary motives aside, was that they would dampen fluctuations in the price of corn. As he stated in On Protection to Agriculture (1822), the effect of the 1815 Act had been 'to make the price of corn in this country habitually and considerably above the price in other countries; and therefore, on occasion of abundant crops, it must fall below the price of those other countries, before any relief can be afforded to the grower by exportation'. But also, as soon as free importation was triggered by high prices, the home market was swamped with overseas corn, with a powerfully depressing effect on the corn price (iv, p. 242, cf. p. 263). This argument against the 1815 Corn Law, that it promoted price instability, was emphasised repeatedly in the later writings.23 According to Samuel Hollander, Ricardo's corn price stability argument does 'not relate at all to the rate of return' (1979, p. 629). This is an untenable claim. Instability was in part a reflection of diminishing agricultural returns, encountered because of restrictions on free importation, which had elevated the price of corn above the continental price.24 And a rising corn price had been the 22

I n a speech delivered on 29 April 1822, R i c a r d o described the 20 shilling duty as something he 'considered . . . as forced upon him, a n d he consented reluctantly to this duty, on account of the distress which now existed, a n d only on that account' (v, p . 157). Even with the lowest duty of 10 shillings, the proposal was protectionist. According to Ricardo's own calculations, the average corn price in continental Europe was 'somewhat about 40 [shillings] per quarter' (iv, p . 242) which, allowing for his 'liberal allowance' of 10 shillings as a countervailing duty (iv, p . 264), a n d taking into account transportation expenses incurred by overseas growers (estimated by McCulloch, 1822, at 10 shillings a q u a r t e r ) , would still leave the home-grower with a n advantage. R i c a r d o consoled himself with the thought that a total repeal would 'ultimately' come about (letter to McCulloch, 7 M a y

23

See, for example, Ricardo's Parliamentary Speeches of 8 February 1821, v, p p . 73-4; 18 February 1822, v, p . 133; a n d 3 April 1822, v, p . 151. Also see the following items of correspondence: t o T r o w e r , 2 M a r c h 1821, v m , p . 350; a n d to McCulloch, 23 M a r c h 1821, VIII > PP- 356-7See, on this theme, Ricardo's speech of 7 M a y 1822, v, p p . 167-8. Also consider one of Ricardo's notes on Malthus's Principles (written during the summer of 1820). M a i thus h a d begun a sentence as follows: ' I f restrictions u p o n importation necessarily increased the

1822,

24

ix, p.

192).

quantity of labour and capital required to obtain corn ... ' At that point Ricardo

102

INTERPRETING RICARDO

reason invoked previously by Ricardo for rising money wages and falling profitability. There is, in other words, at least a logical relationship between the 'unstable prices' and the 'falling profitability' arguments, irrespective of whether the latter argument was actually put by Ricardo in his later writings.25 But the fact is that the argument was put, several times, during the later period. It may be found in speeches, in correspondence, and in notes.26 It is also found in the pamphlet On Protection to Agriculture: With a permanently high price of corn, caused by increased labour on the land, wages would be high; and . . . profits would necessarily fall . .. There is no other way of keeping profits up but by keeping wages down. In this view of the law of profits, it will at once be seen how important it is that so essential a necessary as corn, which so powerfully affects wages, should be at a low price; and how injurious it must be to the community generally, that, by prohibitions against importation, we should be driven to the cultivation of our poorer lands to feed our augmenting population, (iv, P- 237) interposed the comment: 'It is only because they do so, that they are attacked. Can any man doubt of their having this effect?' (11, pp. 204-5). Professor Hollander also claims that a so-called ' "welfare" case in favour of corn law repeal . . . was made quite independently of that relating to the effects of repeal upon the profit rate' (1979, p. 634). The 'welfare' argument was that free importation would be attended by a reallocation of productive resources away from agriculture and a resulting increase in aggregate output and consumption (see, for example, the Essay iv, p. 32; and the Principles 1, pp. 314, 316-18). The connection with the thesis of a rising corn price in a closed economy was made explicit by Ricardo in On Protection to Agriculture: 'in the progress of society, when no importation takes place, we are obliged constantly to have recourse to worse soils to feed an augmenting population, and with every step of our progress the price of corn must rise . . . A high price, if the effect of a high cost, is an evil . . . the price is high, because a great deal of labour is bestowed in obtaining the corn. If only a little labour was bestowed upon it, more of the labour of the country, which constitutes its only real source of wealth, would have been at its disposal to procure other enjoyments which are desirable' (iv, pp. 212-13). The argument was only superficially 'independent' of'that relating to the effects of repeal upon the profit rate', since repeal would imply, as related consequences, a 'better' allocation of labour and a higher profit rate (reflecting the lower corn price). 'We had passed corn laws, that made the price of that necessary of life, grain, higher than in other and neighbouring countries, and thus interfered with the article which was considered the chief regulator of wages. Where grain was dear, wages must be high, and the effect of high wages was necessarily to make the profits on capital low' (speech of 16 December, v, p. 33); the Corn Law 'had tended to raise the price of sustenance, and that had raised the price of labour, which of course diminished the profit on capital' (speech of 24 December 1819, v, p. 38); 'If you confine yourself to the resources of your own soil, I say, rent will in time absorb the greatest part of that produce which remains after paying wages, and consequently profits will be low. Not only individual profits but the aggregate amount of profits will be diminished' (letter to Trower, 21 July 1820, vm, p. 208); 'On my plan [of free corn importation] there will be no transfer neither from profits or wages to rent, prohibit importation and there will be' (note on Malthus's Principles 11, p. 222).

Falling rate ofprofit, wages and the law of markets

103

Admittedly, less prominence was given to the 'secular trend' argument in the later writings, but this was only to be expected in view of contemporary circumstances. As a dedicated opponent of the corn laws, it was natural that Ricardo should have focused on their short-term consequences (fluctuating prices) at a time when these were arousing so much public disquiet. But he never lost sight of the long-term 'secular' effects of protection on general profitability, even in times offalling corn prices (which is quite remarkable), nor were the various arguments unrelated. The post-Essay correspondence, and the content of the core theoretical chapters in the Principles, fully support the long-standing view that Ricardo's primary concern was to establish his theory of a falling rate of profit in consequence of progressively diminishing agricultural returns. Nor is there any mystery surrounding his retention of the 'agricultural' model, despite his undoubted 'optimism' towards Britain's economic prospects. As for the Corn Law connection, the falling rate of profit argument was Ricardo's main case against agricultural protection in the early writings and in the Essay; its later overshadowing by the (logically related) priceinstability argument is perfectly explicable in terms of changed economic circumstances; and it cannot be dismissed as an argument produced merely for the sake of public consumption. In these areas, the traditional view of Ricardo emerges relatively unscathed. WAGES

The plan of this section is as follows. First, I show that the Principles contains evidence of two conflicting treatments of wages: the dominant 'natural wage' analysis, and the 'new view' analysis. Secondly, I argue that only the natural wage analysis is compatible with Ricardo's core doctrine of 'permanent' movements in profitability. Thirdly, I investigate the influences which may have led to the inclusion in the Principles of both the natural wage analysis and the new view alternative. Fourthly, I consider whether Ricardo's position changed in later years (after the publication of the Principles). Fifthly, I take up the question of the relationship between the natural wage doctrine and Ricardo's 'discrete' method of analysis. A brief conclusion then rounds off the section.

104

INTERPRETING RIGARDO

The 'Principles': conflicting doctrines

In this sub-section I will establish that the Principles contains evidence of two broad, conflicting treatments of wages, one of them singled out in 'traditional' interpretations, the other highlighted in the 'new view' literature. According to the first treatment, the natural wage is an operative centre of gravity for market wages without a stationary state restriction: Ricardo's 'natural wage doctrine'. However, it is necessary to distinguish between a qualified version of the doctrine and various strong, unqualified statements which may suggest that the population mechanism responsible for the gravitational tendency achieves an almost instantaneous equality between market and natural wages following any divergence between them. In contrast to both these variations on the natural wage theme, the new view passages suggest that the natural wage becomes a centre of gravity for market wages only in the far distant 'stationary state'. Before that time, the market wage (in commodity terms) initially rises, and then falls over a long historical stretch. I begin with the natural wage doctrine, comprehensively set out by Ricardo in the opening pages of the chapter 'On Wages'. The chapter begins: Labour, like all other things which are purchased and sold, and which may be increased or diminished in quantity, has its natural and its market price. The natural price of labour is that price which is necessary to enable the labourers, one with another, to subsist and to perpetuate their race, without either increase or diminution, (i, p. 93) In order to 'perpetuate their race', labourers must be enabled to purchase those commodities which, from 'habits and customs', they have come to regard as essential to their needs. There are two qualifications. First, the 'natural' bundle of commodities is not 'absolutely fixed and constant. It varies at different times in the same country, and very materially differs in different countries' (1, pp. 96—7). Secondly, rather than there being just one natural wage, Ricardo's allusion in the chapter 'On Value' to a 'scale' of wages for labour of 'different qualities' implies a whole spectrum of natural wages (1, pp. 20-2). Both these qualifications are, for the most part, swept aside. The 'market price of labour' is 'the price which is really paid . . . from the natural operation of the proportion of the supply to the demand' (1, p. 94). 'However much the market price . . . may

Falling rate of profit, wages and the law of markets

105

deviate from [the] natural price', Ricardo added, 'it has, like commodities, a tendency to conform to it' (1, p. 94). The rationale for this tendency is elaborated as follows: It is when the market price . . . exceeds [the] natural price, that the condition of the labourer is flourishing and happy, that he has it in his power . . . 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 . . . poverty deprives [labourers] 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. (1, p. 94) There is no mention of a 'stationary state' restriction on the natural wage. If the tendency to conformity between market and natural wages is analogous to that between the market and natural prices of commodities, as Ricardo has asserted, then it must be regarded as the normal case.27 But there are, as he points out, certain differences between labour and most 'other' commodities. The most important difference is implicit in the last quotation. If the equalising tendency is to operate following a rise in the market wage, there must be an increase of population which, to swell the number of labourers, could take many years. Ricardo expatiated on the process involved for a single male labourer in the chapter 'Taxes on Raw Produce': An accumulation of capital naturally produces an increased competition among the employers of labour, and a consequent rise in its [labour's] price. The increased wages are not ['always', added in edn. 3] 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 . . . (1, p. 163) Marriage is assumed to result in a new supply of future labourers. What 'generally happens' afterwards is that: an effect 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 27

Ricardo presumed that there was normally a strong tendency for the market prices of commodities to conform to their natural levels (see below, pp. 135-6).

106

INTERPRETING RIGARDO

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, p. 164) Both here and in the passage from 'On Wages', conformity between market and natural wage rates is always tending to occur but, as Ricardo wrote elsewhere in the Principles with reference to labour, houses and gold, the 'effect cannot, under some circumstances, be speedily produced' (1, p. 196, cf. pp. 165-6). And there may be oscillations of the market wage around the natural wage as a result of're-action'. These qualifications granted, the presumption is that the natural wage constitutes the attraction point for market wages at all times. From the evidence reviewed, the equalising tendency is forever operating. A further qualification is introduced when Ricardo considers the impact of accumulation on the value of 'capital' ('food, clothing, tools, raw materials, machinery, &c. necessary to give effect to labour', 1, p. 95). Taking two cases, the first when a burst of capital accumulation is accompanied by a rise in the value of wage-goods, the second when items of capital remain at an unchanged or even lower value, he reasoned as follows: In the first case, the natural price of labour ... will rise; in the second, it will remain stationary, or fall; but in both cases the market rate of wages will rise ... above its natural price; and in both cases it will have a tendency to conform to its natural price, but in the first case this agreement will be most speedily effected. The situation of the labourer will be improved, but not much improved; for the increased price of food and necessaries will absorb a large portion of his increased wages; consequently a small supply of labour, or a trifling increase in the population, will soon reduce the market price to the then increased natural price of labour. (1, p. 95-6) In the second case, it 'will not be till after a great addition has been made to the population, that the market price of labour will again sink to its then low and reduced natural price' (1, p. 96). It might take some time, but Ricardo is explicit that it will happen. And again, a stationary state restriction is conspicuously absent. Thus far, the natural wage has been presented by Ricardo as the normal 'centre of gravity' for market wages in an unrestricted historical domain. The speed with which equality is achieved between market and natural wage rates depends on population (supply) lags, and possible changes in the value of wage-goods. The

Falling rate ofprofit, wages and the law of markets

107

totality of these views may be described as the qualified natural wage doctrine. There is also more limited textual support for an unqualified version of the same doctrine. In the chapter 'Taxes on Raw Produce', Ricardo wrote: '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' (1, p. 159). This was one of the few places where he distinguished between different 'qualities of labour'. With 'the rate which nature and habit demand' coming unmistakably close to the phrase used in the description of the natural wage (see above, p. 104) it would seem that, for the 'lowest kind' of labour, he was claiming a powerful tendency for market wages to conform to the appropriate natural wage. There ar^e two other passages in which changes in population are explicitly presumed to occur with great speed, although the natural wage is not mentioned in either. The passage considered here is from the chapter 'On Value' (the other, from the chapter 'On Profits', is quoted below, p. 112): If the shoes and clothing of the labourer, could, by improvements in machinery, be produced by one fourth of the labour now necessary to their production, they would probably fall 75 per cent; but ... 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 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. (1, p. 16) This analysis contrasts with the one given in 'On Wages', where Ricardo stated that it would take 'a great addition . . . to the population' and, presumably, more than 'a very few years', for the same {natural) commodity wage to rule again (see above, p. 106). But it does provide limited support for a 'strong' version of the 'traditional' account. It is also indicative of an unevenness in Ricardo's treatment, but of a minor order when set against the implications of the new view passages. The most compelling new view evidence is found in the chapter 'On Wages', beginning with the following passage: 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

IO8

INTERPRETING RICARDO

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 Vi per cent. They would fall still lower when it increased only at the rate of 1, or Vi 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. I say that, under these circumstances, wages would fall, if they were regulated only by the supply and demand of labourers . . . (1, p. 101)

With (commodity) wages regulated by 'supply and demand' alone, they supposedly fall in the 'natural advance of society' and, given the definition of the natural wage, only reach that level in the stationary state. However, the analysis is incomplete, since no reason has been given for the reduction in the demand for labour. After commenting that 'wages are also regulated by the prices of the commodities on which they are expended', Ricardo produced a fuller account: 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 . . . were expended rose, the labourer would be doubly affected, and would be soon totally deprived of subsistence. Instead, therefore, of the money wages . . . falling, they would rise; but they would not rise sufficiently to enable the labourer to purchase as many [commodities] . . . as he did before [their rise in price] . . . 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 . . . (1, pp. 101-2)

This supplies the missing explanation for the declining rate of capital accumulation (and hence for the falling labour demand). In the 'natural advance of society', market wages in commodity terms are falling, although in money terms they persistently rise, partially offsetting the constant rise in the price of'necessaries', but reducing general profitability and hence the pace of accumulation. Only in the eventual stationary state, when the labouring population is static, will market and natural wages coincide. Other passages point in various new view directions. In the chapter 'Taxes on Raw Produce' there is a reference to 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

Falling rate of profit, wages and the law of markets

109

which the society may be' (1, p. 161). There is also the chapter 'Taxes on Wages', which S. Hollander finds pleasingly supportive of his new view interpretation (Hollander 1979, pp. 386-95.) In the second paragraph of the chapter, Ricardo quoted Adam Smith's claim that the 'subsistence of the labourer' depends on the 'demand for labour, according as it happens to be either increasing, stationary, or declining, or to require an increasing, stationary, or declining population' (1, p. 215). Taking the reading most favourable to Professor Hollander's case, that Ricardo's attitude to Smith's position was approbatory, his new view credentials are apparently strengthened. But there is a complication. In the quoted passage, Smith had adverted to his treatment of wages in the first book of the Wealth of Nations. What he had said there, among other things, was that the natural rate of wages is the 'ordinary or average' rate to which market rates tend at a given 'time and place'; this natural rate would therefore change accordingly as the society happened to be in an 'advancing, stationary, or declining condition' (Smith [1776] 1981, p. 72). So, depending how far we are prepared to go in attributing Smith's position to Ricardo, there could be two new views: the first based on the chapter 'On Wages', with market wages always above but asymptotically approaching the natural level, as defined for a stationary population; and the second with an entirely different, Smithian concept of the natural wage, secularly declining over time. The same interpretative dilemma is met later in the chapter when Ricardo cites an 'able passage from Mr. Malthus's work on population' which repeats the Smithian natural wage doctrine (1, pp. 218-19). Shortly afterwards he wrote: 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. (1, p. 220) Whether he was referring to the regulation of natural or market wages is wholly unclear, although support for (some version of) a new view interpretation might seem to be given. However, rather like the chapter 'Taxes on Raw Produce', when siding with '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', but more directly in 'Taxes on Wages', it was obviously convenient for Ricardo to 'borrow' the Smithian-Malthusian analysis. The extract from the Wealth of

IIO

INTERPRETING RICARDO

Nations served as a necessary prelude to a refutation of James Buchanan's criticism of Smith; and the 'able passage' from Malthus's Essay on Population was also introduced to controvert Buchanan (with a nice touch of irony, since the passage had been quoted by Buchanan himself). Even so, Ricardo may seem to have given his approval to the new view treatment(s) of wages, if only vicariously. Finally, there is one rather curious passage that must be considered, from the chapter 'On Wages': 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 (i, pp. 94-5). This passage is situated between others in which the natural wage is presented as the normal centre of gravity for market wages, perhaps suggesting that the 'indefinite excess' was regarded as no more than an exceptional possibility (cf. St Clair 1965, p. 109).28 If, on the other hand, Ricardo was outlining a probable course for wages, his remarks would signal a fairly radical departure from, rather than a qualification to, the natural wage analysis. This would then raise the question of whether the 'indefinite excess' was a detached view, separate from the natural wage doctrine yet unintegrated with the new view analysis (an interpretation seemingly favoured by Stigler 1952 and Rowthorn 1980), or whether it was part of the new view analysis (Cannan 1893, pp. 247ff., Hicks and Hollander 1977 and Casarosa 1978), in which case it more properly belonged at a later point in the chapter. I leave this as an open question. It has emerged that there are two broadly conflicting accounts of wages in the Principles. First, there is the natural wage doctrine, mostly qualified, which depicts that wage as an active centre of gravity without a 'stationary state' limitation. And secondly, there is the new view account, either with market wages above the natural wage (as explicitly defined by Ricardo) until the stationary state, or with Smithian 'natural wages' secularly falling over time. I stress that these accounts are conflicting, pace Samuel Hollander who 28

The passage appears to have been a strong rewording of the following, from the Essay: 'greater profits would lead to further accumulation; and thus would a stimulus be given to population by really high wages, which could not fail for a long time to ameliorate the condition of the labouring classes' (iv, p. 35).

Falling rate of profit, wages and the law of markets

111

glosses over the many 'natural wage' utterances by presenting them as mere 'simplifications', unrepresentative of, yet (somehow) fully compatible with, Ricardo's 'true' (new view) position.29 I can see little point in a 'simplification' which repeatedly presents the natural wage as something which it is not. There is a puzzle thrown up by the presence of incompatible wage doctrines in the Principles which invites exploration. But before embarking on that project, I consider the connections between Ricardo's treatment(s) of wages and his central doctrine on 'permanent' movements in profitability during the accumulation process. Wages, profitability and accumulation in the 'Principles'

The main purpose of the chapter 'On Profits' was to establish the logical truth of the proposition that progressively diminishing agricultural returns must result in a 'permanent' fall in general profitability (ceteris paribus). What, if anything, does the chapter reveal about Ricardo's treatment of wages? It indicates, as a considerable understatement, that his position was not straightforward: the chapter contains material that can be cited in support of both the natural wage and the new view interpretations. Towards the beginning of the chapter, he wrote: 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, p. 112, emphasis in original) This is presumably a reference to the new view analysis in the chapter 'On Wages'. Six pages later, the following passage is encountered: 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 29

Hollander (1979, pp. 327, 387, 396, 646, 687). Also see Hollander (1990), in which the natural wage doctrine becomes a 'convenient point of departure', a matter of 'classification', an 'expository exercise'; but never, of course, a doctrine reflecting Ricardo's 'true' (new view) position.

112

INTERPRETING RICARDO

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, (i, p. 118) Samuel Hollander enlists this passage in support of his own interpretation (1979, p. 399), but more would be expected from a committed new view Ricardo than the mere possibility of a fall in commodity wages following a corn price increase; he would be expected to state that commodity wages are (secularly) reduced. In terms of the natural wage doctrine the passage is less mysterious: 'high wages' are simply those ('temporarily') above the natural level.30 Later in the chapter, Ricardo invited a natural wage interpretation by explicitly assuming a given and constant commodity wage. Following a rise in the price of corn, the farmer 'would be obliged to pay his labourers . . . [enough] to enable them to consume the same quantity of necessaries as before, and no more' (1, pp. 124—5). Yet within the space of a few lines he claimed that: each labourer would receive more money wages; but the condition of the labourer, as we have already shewn, would be worse, inasmuch as he would be able to command a less quantity of the produce of the country. (1, P- 125)

The last passage takes us in the direction of the new view, but the very next paragraph is more suggestive of the natural wage analysis: 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. (1, p. 125)

This not only evokes the natural wage doctrine, with the reference to the 'usual consumption' of labourers, but the 'speedy' population response suggests the more hardline, unqualified version. However, to further complicate matters, Ricardo continued: In his later work, Professor Hollander concedes that this passage is not quite so supportive of his new view interpretation. Ricardo was being 'unnecessarily conciliatory to hypothetical critics' (Hollander 1990).

Falling rate of profit, wages and the law of markets

113

But when poor lands are taken into cultivation .. . the effect must be permanent. A greater proportion of that part of the produce which remains to be divided . . . 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. (1, pp. 125-6)

This might be read as a restatement of the new view contention that commodity wages secularly decline. On the other hand, Ricardo was considering the distribution of the material net surplus in agriculture and, ceteris paribus, the corn measured wage per labourer will fall, even with wages given and constant. In that sense, each man does have a 'less quantity'. The material in the chapter 'On Profits' fails to establish Ricardo's unambiguous subscription to any particular treatment of wages, or a link between one or other treatment of wages and the doctrine of'permanent' movements in profitability. It is necessary to look elsewhere in the Principles. Consider, first, the following paragraphs, from the chapter 'On Foreign Trade: 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. If, therefore, by the extension of foreign trade, or by improvements in machinery, the food and necessaries of the labourer can be brought to market at a reduced price, profits will rise . . . but if the commodities obtained at a cheaper rate . . . be exclusively the commodities consumed by the rich, no alteration will take place in the rate of profits. The rate of wages would not be affected, although wine, velvets, silks, and other expensive commodities should fall 50 per cent., and consequently profits would continue unaltered. Foreign trade, then, though highly beneficial to a country, as it increases the amount and variety of the objects on which revenue may be expended, and affords, by the abundance and cheapness of commodities, incentives to saving, and to the accumulation of capital, 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. (1, pp. 132-3)

Ricardo was associating a 'permanent' reduction in wages (hence a 'permanent' rise in profits) exclusively with cheaper wage goods. At a stretch, this might seem to be compatible with a new view style of reasoning. Riding roughshod over logical niceties, one could argue for the case of the downward secular tendency that a rise in the price of corn is the 'cause' of both a 'permanent' rise in money wages and

114

INTERPRETING RIGARDO

a 'permanent' fall in profits, these effects being 'permanent' because they are irreversible, ceteris paribus (disregarding oscillations around the 'dynamic' trend). 31 Applying similar dynamic reasoning to the case in hand, cheaper wage goods would lead to 'permanently' lower wages and 'permanently' higher profits in the sense that the secular trend line for money wages is downwardly displaced. But this interpretation is doomed to failure. Ricardo has also stated that cheaper luxuries (the commodities consumed by the rich) will stimulate an additional accumulation of capital; and that there is no consequential (permanent) alteration in the rate of profits (or wages). According to dynamic new view reasoning, in contrast, increased saving and accumulation by capitalists will 'permanently' raise money wages and lower profits (the trend line shifts upward). Ricardo was not expressing himself in those terms. The sense of the passage is rather that 'permanent' movements in profits are inversely related to changes in the value of a given wage bundle of enduring relevance. The only wage bundle mentioned by Ricardo in the Principles that fits this description is the natural wage. The same is true for the wage bundle implied by the opening paragraph to the chapter 'Effects of Accumulation on Profits and Interest': 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 funds for the maintenance of labour were doubled, trebled, or quadrupled, there would not long be any difficulty in procuring the requisite number of hands, to be employed by those funds ... 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, (i, p. 289) According to new view reasoning, the last sentence in this extract is simply false. Even if necessaries 'could be constantly increased with the same facility', the dynamic interrelationships which are central to the new view imply that the (steady state) rates of wages and profits will depend on capitalists' accumulation. There would be a 'permanent alteration in the rate of profit and wages' if the rate of accumulation changed. Ricardo, in contrast, was taking for granted that the same wage-bundle prevails 'to whatever amount capital might be accumulated'. And his position was that 'permanent' 31

This interpretation is in the spirit of Hollander (1990).

Falling rate ofprofit, wages and the law of markets

115

movements in money wages (and therefore profits) were uniquely given by changes in the value of this bundle. If we now ask which wage (bundle) has been isolated by Ricardo as having the required, potentially lasting relevance, the answer must be as before: the natural wage. It is only the natural wage doctrine which is compatible with Ricardo's pronouncements on 'permanent' variations in wages and profits. The chapter on 'Accumulation' contains other statements carrying the same implication. I will confine myself to presenting just one further passage but, first, a little on its background. In November 1818 Ricardo wrote to Mill: I hear from various quarters that my book is selling very fast, and that a new edition will soon be required. It will perhaps be necessary that I should carefully go over the whole subject again . . . as I have not looked at it now for more than a twelvemonth, (vn, p. 327)

This 'careful' review took all of a 'few days' (vn, p. 337) and mainly served to confirm Ricardo 'in the truth of the general doctrines' (vn, p. 334). But among the 'few very trifling alterations' where Ricardo thought he could make his ideas 'more clear' (vn, p. 331) was the following interpolation: however abundant capital may become, there is no other adequate reason for a fall of profit but a rise of wages, and further it may be added, that the only adequate and permanent cause for the rise of wages is the increasing difficulty of providing food and necessaries for the increasing number of workmen. (1, p. 296)

This is presumably one of the 'general doctrines' that Ricardo was so convinced of. Like the other passages that have been cited, it is consistent only with the natural wage analysis. It is the 'traditional' interpretation of Ricardo as a natural wage theorist, and that interpretation alone, which preserves intact his core thesis on 'permanent' movements in profitability. The new view interpretation inevitably detracts from the 'permanence' thesis, as well as rendering the natural wage doctrine completely unintelligible. Yet the fact is that the new view can also claim textual support. A rather prosaic deduction is that Ricardo's treatment of wages in the Principles was eclectic, to say the least. More interestingly, what were the influences which had led to Ricardo's espousal of the contradictory doctrines, and did his

I I6

INTERPRETING RICARDO

position change over time? These questions are addressed in the following sub-sections. Wages: the background influences

The assumption of given and constant real wages, a characteristic of Ricardo's 'monetary' writings, was also a feature of his early work on profits.32 For instance, in claiming that 'the cheapness of provisions . . . is after all, whatever intervals we may be willing to allow, the great regulator of the wages of labour' (26 June 1814, vi, p. 108), Ricardo was implicitly demoting real wage variability to the status of a secondary, relatively unimportant influence on money wages. This prompted yet another dispute with Malthus, who took a critical view of the assumption that, in Malthus's words, money wages 'rise without delay exactly in proportion' following 'a rise in the price of raw produce' (9 October 1814, vi, p. 140).33 For Malthus, real wages necessarily fall if corn becomes more difficult to produce, thus tending to offset the effect of the higher corn price on money wages.34 His criticisms made little impression on Ricardo, who at first denied any necessary connection between the effects of a higher corn price (higher money wages and lower profits) and those of a reduction in capital (lower real wages, according to Malthus) (25 July 1814, vi, p. 114). Subsequently he conceded that a reduction in capital was a 'probable effect' of restrictions on corn importation, but in making any consequent change in profits depend entirely on 'the cultivation of the land', he effectively denied that real wage variations played a significant, independent role in governing money wages (11 August 1814, vi, p. 119). Then, in the Essay, he supposed 'that capital and population advance in the proper proportion, so that the real wages of labour, continue uniformly the same' (iv, p. 12): an application of 'fixwage' analysis to which Malthus objected strongly (see below, pp. 119—20). This was an assumption which could be made regardless of a theoretical position on the enduring relevance of any particular real wage (such as the natural wage in the Principles), but there is some evidence in the 32 33

34

F o r t h e earlier use o f ' f i x w a g e ' analysis, see a b o v e , p . 4 5 . Ricardo had no objection to this characterisation of his position. His reply suggests that he regarded the impact on profitability from a decline in real wages (implied by non-'proportionality'), as a 'temporary' occurrence. 23 October 1814, vi, p. 145. See his letters t o R i c a r d o of 6 J u l y 1814, v i , p . 11155 A u g u s t 1814, v i , p p . 1 1 6 - 1 8 ; a n d 2 3 November 1814, vi, p. 155.

Falling rate ofprofit, wages and the law of markets

117

Essay suggesting that Ricardo may have singled out the 'subsistence' wage - the conceptual precursor to the 'natural' wage - as the real wage of special importance. After remarking that general profits could be raised by a 'fall of the real wages of labour', Ricardo added that such a fall 'is more or less permanent, according as the price from which wages fall, is more or less near that remuneration for labour, which is necessary to the actual subsistence of the labourer' (iv, p. 22). His sentence is ambiguous, but it may be read as implying the definition of 'permanent' variations in profit with reference to the 'subsistence' wage (cf. Cannan 1893, p. 243).35 If 'natural wage' is substituted for 'subsistence' wage, this would give us one of the central doctrines of the Principles.

The fact is, however, that the developed category of the 'natural wage' does not appear in Ricardo's writings until the Principles itself.36 What seems to have happened is that, rather like his 'borrowing' of Malthus's theory of differential rent, Ricardo was indebted to Torrens's Essay on the External Corn Trade (1815) for supplying a mode of analysis which fitted well with his own train of thought. 37 Torrens claimed that the 'proper way of regarding labour, is, as a commodity on the market' which, like other commodities, has both a market and a natural price. The market price is regulated by demand and supply conditions, but the natural price, which is 'governed by other laws', consists: in such a quantity of the necessaries, and comforts of life, as, from the nature of the climate, and the habits of the country, are necessary to support the labourer, and to enable him to rear such a family as may 35

36

37

Samuel Hollander presents a different interpretation. H e claims that Ricardo's words 'seem to imply a d o w n w a r d trend in the real wage - at least a statistical trend reflecting the circumstance that wage reductions a r e compensated for to a n increasingly small degree by reverse movements' (emphasis in original). But, having advanced this interpretation, Professor Hollander is forced to a d m i t that 'what follows [in the Essay] denies a declining wage trend' (emphasis in original). His resolution of this paradox is that R i c a r d o had fallen into error 'on his o w n terms' (Hollander 1990). M y resolution is that those terms a r e Professor Hollander's, not Ricardo's. This is overlooked by Professor Caravale, who misleadingly refers to the 'precise dynamic notion of [the] natural wage . . . [which] is in fact given by R i c a r d o in the Essay' (1985, p. 135, cf. p p . 137, 143). Professor Caravale compounds his error by identifying Ricardo's 'subsistence' wage with t h e constant wage assumption m a d e at the outset of the detailed agricultural analysis. N o such identification was m a d e by Ricardo. Ricardo read Torrens's pamphlet soon after its publication (vi, p. 188).

I I8

INTERPRETING RICARDO

preserve, in the market, an undiminished supply of labour. (Torrens 1815, p. 62) This natural price 'may, in any given time and place, be regarded as very nearly stationary' (p. 65), whereas the market price 'may often be considerably more, and often considerably less' than the natural price. However: the natural, and the market price of labour, have a mutual influence on each other, and cannot long be separated. When the market price of labour falls below the other . .. deaths are increased ... births are diminished; and thus, by a double operation, the level between the natural, and the market price of labour, is restored. On the other hand, if the market price should, at any time, be raised above the natural, the increased comforts enjoyed by the labourer and his family, would diminish deaths, and, by giving encouragement to marriage, increase births, until, by a double operation, the supply of labour was augmented, and its market price brought back to that natural level, from which it can never permanently recede, (pp. 65-6) Elsewhere, Torrens insisted that market wages cannot diverge from the given natural level 'for any length of time' (pp. 75—7). Torrens's analysis is very similar to the natural wage doctrine espoused by Ricardo in his Principles,38 This was not lost on Torrens, who reportedly felt aggrieved by Rieardo's failure even to mention his name. 39 In an act of conciliation, Ricardo 'mentioned Torrens twice with approbation' in the second edition of the Principles (1, pp. o,6-7n., 27m.), although his private opinion was that 'none of his [Torrens's] doctrines appeared to me strikingly new' (vn, p. 180) partly, it may be suspected, because they were so close to ideas of his own. This is true in a double sense. From 1814 onwards Ricardo had, implicitly and explicitly, treated variations in commodity wages as an influence on profits of subordinate (or 'temporary') importance when set against the influence of changes in the difficulty of producing wage-goods, especially corn. His comparative lack of interest in real-wage variability, of which Malthus complained incessantly, could be justified if it was possible to specify a real wage of 'permanent', enduring significance. I have suggested that the 'subsis38

39

Cannan argued that, although Rieardo's account was influenced by Torrens's work, there was a major difference because Ricardo defined the natural wage for a stationary population, whereas Torrens's natural wages were only 'ordinary' or 'average' rates (1893, p. 247). It seems to me that Torrens's usage of his natural wage doctrine could certainly bear a 'stationary population' interpretation, whatever he may have intended. See Rieardo's letter to T r o w e r of 23 August 1817, v n , p p . 179-80.

Falling rate of profit, wages and the law of markets

119

tence' wage in the Essay may have played that role. If so, Torrens's more rounded conception of the 'natural wage' would have been an embellishment of an existing category. Secondly, Torrens's presumption of only transient departures of market from natural wages must have been euphonic to Ricardo who, as he frankly admitted, was disposed to ignore 'temporary' phenomena (as he had done in his early 'monetary' writings). Thus he wrote to Malthus: It appears to me that one great cause of our difference in opinion, on the subjects which we have so often discussed, is that you have always in your 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. (24 January 1817, vn, p. 120) We know, however, that the natural wage doctrine in the Principles was mostly hedged with qualifications on the speed with which the 'gravitational' pull would restore equality between market and natural wage rates. Introducing such qualifications went against the grain, but Ricardo thought he should make the effort. Again to Malthus: 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, (vn, p. 120) There seems little doubt that this applied, at least in part, to the treatment of wages. Indeed, in Malthus's scathing reply, 'wages' were mentioned as a specific bone of contention: I agree with you that one cause of our difference in opinion is that which you mention. I certainly am disposed to refer frequently to things as they are, as the only way of making one's writings practically useful to society ... 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 inferences from his hypothesis. In your essay on profits you suppose the real wages of labour constant; but as they vary with every alteration in the prices of commodities, (while they remain nominally the same) and are in reality 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 and less prosperity, and sometimes of adversity, but never the uniform

120

INTERPRETING RICARDO

progress which you seem alone to contemplate. (26 January 1817, vn, pp. 121-2, emphasis in original)

We will see in the following sub-section that Ricardo truly believed he had insured himself against Malthus's criticism with his attempt to 'manage the subject quite right'. We will also find that Mai thus was distinctly unimpressed. From the evidence reviewed, Ricardo's adoption of the natural wage doctrine may be seen as an evolutionary step from the more informal treatment of wages which had pervaded his earlier writings. The allowances which he self-consciously made for 'temporary' disparities between market and natural wage rates also become explicable in terms of his reaction to Malthus's criticism. But what of the origins of the scattered new view passages? Ricardo had first been confronted with Malthus's thesis of a secularly declining commodity wage in November 1814: as the state of the land is by no means the sole cause which determines profits, but as they are powerfully influenced by the varying demands for produce . . . and the constant tendency to a fall in the wages of labour, it neither accords with theory or experience to call the state of the land the regulator of profits. (Malthus to Ricardo, 23 November 1814, vi, p. 152, emphasis in original) 40 In his reply, Ricardo granted that real wages would fall 'in consequence of the accumulation of capital': 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, vi, pp. 162-3)

On the basis of this passage alone, it is unclear whether Ricardo was endorsing Malthus's 'falling wages' thesis or, alternatively, whether he was merely allowing for a 'temporary' reduction in real wages that would be reversed by a subsequent fall in the supply of labour. If the former, then we might expect to find the falling real wage scenario in the Essay on Profits, published a couple of months later. We would search in vain. Indeed: The rise or fall of [real] wages is common to all states of society, whether it be the stationary, the advancing, or the retrograde state . . . As experience demonstrates that capital and population alternately take the lead, and 40

This thesis was vigorously propounded in Malthus's An Inquiry into the Nature and Progress of Rent: 'There is nothing so absolutely unavoidable in the progress of society as the fall of wages . . . ' ([1815] 1970, p. 193 ff.).

Falling rate ofprofit, wages and the law of markets

121

[real] wages in consequence are liberal or scanty, nothing can be positively laid down, respecting profits, as far as [real] wages are concerned, (iv, pp. 22-3) There is no mention of a secular downward trend in real wages either here or anywhere else in the Essay, 'liberal or scanty5 wages are common to all states of society, which is compatible with the suggestion that the 'subsistence' wage had come to be viewed as the wage of 'permanent' significance ('liberal or scanty wages' being wages above or below the 'subsistence' level).41 Following the Essay's publication, Malthus persisted with his general argument that real wage variability undermined Ricardo's view that 'profits . . . depend upon facility of production' (5 May 1815, vi, p. 225). He made little headway.42 Malthus was undeterred and, in his letter to Ricardo of 16 October 1815, he again promoted his secular tendency: I agree with you ... in thinking that in the progress of cultivation there is a regular tendency to a diminution in the productiveness of industry on the land, in the same manner as there is a regular tendency to a diminution in the real wages of labour; but it is quite erroneous I think to infer that the profits of stock may not, as will the wages of labour rise for periods of some duration during the course of this progress. In fact, profits on the land will always rise, when population increases faster than capital, independently of improvement &c. (vi, pp. 303—4) Ricardo's only comment was the following: That profits may rise on the land if population increases faster than capital, I am not disposed to deny, but this will be a partial rise of profits on a particular trade, for a limited time, and is very different from a general rise of profits on trade in general. (17 October 1815, vi, p. 305) 41

42

Cf. a footnote in the Essay in which, on the hypothetical assumption that agricultural improvements counterbalance diminishing returns, R i c a r d o claimed that corn would remain subject to a n 'accidental variation of price' arising (inter alia) 'from greater or less real wages of labour' (iv, p . io,n.). T h e variations in real wages were, by implication, themselves 'accidental'. This again contradicts the notion that R i c a r d o may have adopted the Malthusian scenario of secularly (and necessarily) falling real wages at this time. It is, on the other h a n d , compatible with t h e view that market wages 'accidentally' fluctuate around some given real wage of lasting relevance. See, particularly, Ricardo's letter of 17 October 1815, in which he expressed his view that profits will not 'necessarily fall, 'however a b u n d a n t [capital] may become', providing 'there be no difficulty on the score of food a n d r a w produce' (vi, p . 301, emphasis in original). T h e reasoning should be familiar. As with similar passages in the Principles, the a r g u m e n t implies that the same wage bundle tends to prevail over time, whatever happens in the sphere of accumulation. Real wage movements are of no ' p e r m a n e n t ' consequence.

122

INTERPRETING RICARDO

All Ricardo concedes is the possibility of a temporary variation in commodity wages (and hence profits) during the accumulation process. On the 'secular decline', he remained silent. Soon afterwards, writing to Hutches Trower, Ricardo (for good reason) included 'wages' among the topics 'of the first importance' over which he and Malthus disagreed (29 October 1815, vi, pp. 315—16). There followed a two day visit to Malthus's home, which provided the opportunity for an 'abundance of discussion' (vi, p. 325). Presumably with reference to this meeting, Malthus subsequently wrote: I believe I assented in parting 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. (22 December 1815, vi, p. 342) Perhaps it was the now withdrawn agreement which Ricardo had in mind when he reported to James Mill that he and Malthus had 'differed less in opinion' in their discussions (vi, p. 325). Malthus had at least conceded that money wages move in the same direction as the corn price. Ricardo initially 'responded' to Malthus's retrenchment by ignoring it (2 January 1816, vn, pp. 1-3). But Malthus pressed for an answer: 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, vn, p. 8, emphasis in original) Ricardo could not decline this invitation. He responded: 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 these may balance each other, or rather that the latter will prevail, I on the contrary think the former the most powerful in its effects. I must write a book to convince you. (10 January 1816, vn, p. 10) The new view passages in the chapter 'On Wages' were the ones in which he pursued this objctive.

Falling rate of profit, wages and the law of markets

123

If those were the only such passages, it might be concluded that they were introduced to controvert Malthus's argument and had little or no further significance. Ricardo was attempting to demonstrate that, even on Malthus's assumptions, progressively diminishing agricultural returns must depress general profitability. Yet we know that various new view refrains are also heard (intermittently) in the chapters 'On Profits', 'Taxes on Raw Produce' and 'Taxes on Wages'. It is therefore possible that Ricardo had, on some level, internalised the Smith-Malthus style of analysis. But this is not to suggest that the new view analysis was anything approaching Ricardo's one-and-only 'true view'. I repeat, the new view passages cannot be reconciled meaningfully with the (dominant) natural wage doctrine, which had evolved from Ricardo's early 'fixwage' analysis. And it is only the natural wage analysis which is compatible with Ricardo's 'core' thesis on 'permanent' movements in profitability. Whether his position changed in later years is considered next. The 'Notes on Malthus3 and after

Malthus's Principles, published in April 1820, contained a frontal assault on the fundamental tenets of Ricardo's work. 'It has been my wish', wrote Malthus in his Introduction, 'to avoid giving to my work a controversial air' (11, p. 11). Yet the temptation to try and puncture Ricardo's authority, and thus reestablish his own, had evidently proved irresistible. Ricardo's treatment of wages and profits received a drubbing from Malthus. I begin by considering the first two sections of Malthus's fifth chapter, 'Of the Profits of Capital', and Ricardo's comments on them.43 According to Malthus, movements in the general rate of profit depended on both altered conditions of producing wage-goods, especially corn, and changes in the commodity wage: 43

Ricardo read Malthus's book immediately on publication and, three months later during the Parliamentary recess, shut himself away in Brighton to study it in detail (letter to Mill, 27 July 1820, VIII, p. 212). It was during this time that he began writing his notes, which were never published. Explaining why, Ricardo wrote to Trower: 'Your opinion, I perceive, is in favor of publishing them as an appendix to a new edition of my "Principles . . . " That was the form in which I at first had an idea of giving them to the public, but I was strongly dissuaded from it by Mill, who thought I ought by all means to avoid giving too controversial a character to my book' (14 January 1821, VIII, p. 333).

124

INTERPRETING RIGARDO

It is the combination of the two, and of others in addition to them, sometimes acting in conjunction and sometimes in opposition, which occasions in the progress of society those varied phenomena which it is not always easy to explain, (n, p. 254) He then assumed, as a debating ploy, that commodity wages are given and constant. In that case, ceteris paribus, 'the rate of profits must regularly and without any interruption fall, as the society advanced', because of the resort to inferior land. However: a moment's consideration will shew us, that the supposition here made of a constant uniformity in the real wages of labour is not only contrary to the actual state of things, but involves a contradiction. (11, p. 255) His two-stage argument was, first, that if the given and constant wage was set at 'subsistence' (corresponding to Ricardo's definition of the natural wage), 'the labouring classes could not increase, nor would there be any occasion for the progressive cultivation of poorer land' (11, p. 255); and, secondly, that if the given wage was above subsistence, but still constant, the analysis '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' (11, p. 255). Neither wage assumption would do. For the sake of 'curiosity', Malthus proceeded to assume that capital and population advance at the same rate, whatever that rate may be (11, pp. 255—6). Again ceterisparibus, diminishing agricultural returns imply that both commodity wages and the general rate of profit 'continue regularly falling till the accumulation of capital had ceased' (11, p. 256). This virtually corresponds to the new view analysis in Ricardo's chapter 'On Wages'.44 In the next section of his chapter Malthus switched his attention to the 'second main cause' of variations in general profitability: changes in commodity wages. After singling out various possibilities to illustrate the effects of this 'cause', and asserting their practical relevance, he commented that periods 'in which capital and population do not keep pace with each other, are evidently of sufficient extent to produce the most important results on the rate of profits, The difference lies in Malthus's concept of the natural wage. According to his preferred definition (see below, p. 126) it is the natural wage which secularly declines, not the market wage as in Ricardo's analysis.

Falling rate of profit, wages and the law of markets

125

and to affect in the most essential manner the progress of national wealth' (11, p. 263). He then trained his sights on Ricardo: the natural and necessary tendency of profits to fall in the progress of society, owing to the increasing difficulty of procuring food, is a proposition which few will be disposed to controvert; but to attempt to estimate the rate of profits in any country by a reference to this cause alone, for ten, twenty, or even fifty years together ... 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. (11, pp. 263—4) Mai thus, for one, did not recongise Ricardo as a new view theorist: a salient point in itself, since according to the new view interpretation, Ricardo and Malthus shared the same analysis.*5 Ricardo's reaction to Malthus's criticisms is of great importance, especially so in the light of Samuel Hollander's declaration that Ricardo's statements 'alone should settle the debate as to [his] "true" view' (in Professor Hollander's favour, of course) (Hollander 1990). It is true that Ricardo was fulsome in his praise for Malthus: I have read the 1 and 2d sections of Chap. 5 of Mr. Malthus work with great pleasure [;] they express with great clearness and ability the doctrines which appear to me to be true respecting profits - I had, very imperfectly indeed, attempted to explain the same principles myself, in my work. (11, P- 265) He noted later that he maintained 'no other doctrine than that which has been well explained by Mr. Malthus' (11, p. 288); and he further expressed his agreement 'throughout . . . with Mr. Malthus in principle' (11, p. 258, cf. pp. 125-6, 138). Professor Hollander might seem to be on firm ground. But wait. Even supposing, purely hypothetically it will transpire, that the above remarks constituted the only germane evidence from Ricardo's notes, they would not establish his 'true' position at the time he wrote his Principles, some four years earlier. It may be conceivable that Ricardo had, for the first time, fully converted to Malthus's new view treatment of wages. But he had certainly not 45

Samuel Hollander's conviction that this was indeed so may account for his faulty attribution of a letter written by Malthus (16 October 1815) to the pen of Ricardo (Hollander 1979, p. 403).

126

INTERPRETING RICARDO

done so before publishing his major work. We are thus faced with the possibility of a significant development in his thought, not the confirmation of his earlier doctrinal preference. From Ricardo's response to Malthus's criticism of his (Ricardo's) natural wage doctrine it is clear that any 'conversion' was far from total. After quoting Ricardo's definition of the natural price of labour, Malthus had inveighed: This price I should really be disposed to call a most unnatural price; because in a natural state of things ... such a price could not generally occur for hundreds of years. But if this price be really rare, and, in an ordinary state of things, at so great a distance in point of time, it must evidently lead to great errors to consider the market-prices of labour as only temporary deviations above and below that fixed price to which they will very soon return, (n, pp. 227-8) This criticism is highly pertinent. So too is Malthus's preferred ('Smithian') definition of the natural wage as ' t h a t . . . which, in the actual circumstances of the society, is necessary to occasion an average supply of labourers, sufficient to meet the average demand' (11, p. 228). Ricardo had been presented with a perfect opportunity to clarify his position. In terms of Malthus's definition, the new view scenario which had attracted Ricardo's 'warm and repeated praise' (Hollander 1979, p. 399) was one with natural wages gradually declining over time. If Ricardo was now a true convert to this reasoning, we could therefore expect him to accept Malthus's criticism and, perhaps, to pledge his allegiance to Malthus's definition. The following note might seem to clinch the matter: With every demand for an increased quantity it [the market price of labour] will rise above this [natural] price[;] therefore if capital and population regularly increase the market price may for years exceed its natural price. I am however very little solicitous to retain my definition of the natural price of labour - Mr. Malthus's will do nearly as well for my purpose. (11, p. 228n.) Ricardo's overriding 'purpose' was to prove that worsening conditions of production on the land must raise money wages and depress profitability. And as he had shown with his new view passages in the chapter 'On Wages', and Malthus had demonstrated with his own analysis, this objective could be achieved even allowing for declining commodity wages. To that extent, Malthus's definition

Falling rate ofprofit, wages and the law of markets

127

of the natural wage would do 'nearly as well'.46 But further thought must have convinced Ricardo that Malthus's definition would not do 'nearly as well5. The former note was deleted and replaced with this one: I have done so [given his definition of the natural wage] that we may have one common language to apply to all cases which are similar. By natural price I do not mean the usual price, but such a price as is necessary to supply constantly a given demand. The natural price of corn is the price at which it can be supplied affording the usual profits. With every demand for an increased quantity the market price of corn will rise above this price and probably is never at the natural price but either above or below it, — the same may be said of the natural price of labour. (11, pp. 227-8) As Ricardo had stated in the Principles, his natural price of labour had the same significance as the natural prices of 'other' commodities; as a centre of gravity, market prices could be 'above or below it' and, to that extent, the 'cases' were 'similar'. It follows that Ricardo's praise for Malthus's chapter 'On Profits' cannot be taken as evidence of his 'true' or unambiguous subscription to new view reasoning in 1820 (let alone earlier). Understandably, he must have been delighted to find Malthus articulating the very same argument which had been rejected in 1816 (see above, p. 122). He had convinced Malthus of the point that progressively diminishing agricultural returns will depress general profitability. And, on one (detached) level, he was favourably disposed to the Malthusianstyle analysis. But for all that, he could not bring himself to relinquish his natural wage doctrine. The evidence provided by Ricardo's second note thus delivers a serious blow to commentators who, in Professor Hollander's case, wish to present the new view passages as quintessentially 'Ricardian', or, in the case of Professor Caravale, claim textual licence to redefine the natural wage in ways that are more fitting to their restatements of Ricardo's system.47 46

47

Whether the above note betokens full acceptance of Malthus's position is unclear. T h e possibility of market wages exceeding the natural rate 'for years' is far weaker than 'hundreds of years' predicted by Malthus or, for that matter, the 'indefinite' period Ricardo h a d allowed in one passage from the chapter ' O n Wages' (see above, p . 110). T h e y d o their best to minimise t h e d a m a g e . C o m m e n t i n g on the second note, Professor H o l l a n d e r writes: ' R i c a r d o ' s reply merely conceded that . . . the market price is likely to fluctuate'; then, in the next breath, ' b u t in a deleted Note ... R i c a r d o was even more conciliatory . . . ; a n d immediately following, 'All the evidence thus points to the conclusion that average commodity wages, in R i c a r d o ' s view, tend to decline in t h e course of secular expansion' (1979, p p . 4 0 2 - 3 ) . Professor Caravale, after mis-quoting part of the deleted note in his m a i n text (he omits 'nearly' in 'I a m however very little solicitous to retain my definition of the n a t u r a l price of labour. M r . Malthus's will d o nearly as well for m y

128

INTERPRETING RICARDO

However, it seems clear that, despite his refusal to recant his natural wage doctrine, Ricardo was shaken by Malthus's rebukes. Some of them, he thought, were unfair, evidently believing, with some justice, that he had allowed for ('temporary') supply and demand influences on commodity wages: how then can it be justly said of me that the only cause which I have recognised of high or low profits is the facility or difficulty of providing food for the labourer)?] I contend that I have also recognised the other cause, the relative amount of population to capital, which is another of the great regulators of wages, (n, pp. 264-5) But there was something defensive, even vulnerable, in his tone. Thus: Profits in all countries must mainly depend upon the quantity of labour given for corn ... I say mainly depend, because I think wages mainly depend on the price of corn. After the observations of Mr. Malthus on the other causes which may affect labour, I must guard myself against being supposed to deny the effect of those other causes on wages. (11, p. 291) The 'other causes' had never been so important to Ricardo as they were to Malthus, and it had taken a conscious effort to 'guard himself against the charge of overlooking them entirely. Now, stung by Malthus's criticisms, he went out of his way to protect himself from possible censure. Both in the Notes and in later writings, this defensiveness went so far that his 'profits' doctrine was stated in such a way as to be superficially compatible with almost any theory of real wages, always providing that some real wage could be taken as given. The following note is a case in point: I do not say that the labourers earnings will always be the same, but whatever they may be, profits will depend on the proportion which their value bears, to the whole value produced on the last land, (n, p. 266) He was guarding his flanks. Whatever the given commodity wage happened to be, and however it was determined, profits depended on (changes in) its value. This strategic formulation still allowed him to claim, so he believed, that (changes in) wages, hence profits, depended 'mainly . . . on the price of corn' (11, p. 291); and, as he put it subsequently, that wages depended 'in a great measure', 'to a purpose'), adds in a footnote that this 'sentence is extremely illuminating, although it belongs to a paragraph that Ricardo, in order to avoid repetitions, replaced with another one' (1985, pp. 137, 13711.). Professor Caravale does not present the second of Ricardo's notes.

Falling rate of profit, wages and the law of markets

129

great degree', 'particularly', 'mainly' and 'powerfully' on the price of corn (respectively, v, p. 130 and iv, pp. 236, 347, 237). But in hedging his bets — refusing either to have the courage of his own 'natural wage' convictions, or to jump decisively in favour of Malthus's doctrine — he left himself open to the charge of vacuous theorising. Why not reverse the 'causality' and state that given the conditions of producing wage-goods, profits depend on 'the labourers earnings'? Ricardo, in his attempt to straddle different theoretical positions, had regressed to an empty style of'fixwage' analysis. Perhaps, with further thought, he would have committed himself to something more substantial. We will never know. The fact is that his attention was increasingly captured by problems with his (labour) theory of value, not by those with his treatment of wages. The 'wages' file was never closed. Wages and Ricardo's analytical method: a note

Ricardo's central object of analysis in the Principles - his investigation of the natural course of wages, profit and rent — was set in a dynamic context, and incorporated the principle of progressively diminishing agricultural returns. For the contemporary economist, versed in mathematical techniques, some kind of exponential growth analysis might seem to be appropriate to such a theoretical investigation. With the new view scenario, a literary formulation is encountered which lends itself to such a translation. Although the scenario cannot be represented by a 'steady state' model (growth rates are not constant) it is nevertheless possible to postulate a set of dynamically interrelated variables and describe the trajectory of the system over continuous time. The new view has quite a modern feel to it. A characteristic of Ricardo's analytical approach is that he did not spontaneously express himself in 'continuous' terms. The method which seems to have come as second nature to him might be more aptly described as 'discrete' analysis or, as Hicks put it before his conversion to Samuel Hollander's new view interpretation, the analysis of the single 'Impulse' (Hicks 1972). That word is used by Ricardo when referring to 'the impulse, which an increased capital gives to a new demand for labour' (1, p. 95). Why he could claim, as he so often did, that the same

13O

INTERPRETING RICARDO

commodity wage would ultimately prevail 'notwithstanding the accumulation of capital', was precisely because the accumulation was conceived as coming in discrete impulses, each one being allowed to work itself through the system. Initially, the system is tranquil, with wages given at their subsistence or 'natural' level. Then comes an impulse of capital accumulation, creating a new demand for labour and hence a rise in commodity wages, which induces an expansion in the labouring population. This, in turn, depresses commodity wages back to their original level, although whether this corresponds to the original money wage depends on the price of wage-goods. If, say, food has risen because of diminishing returns, the reestablished commodity wage is of a higher monetary value, and the (general) rate of profit is 'permanently' reduced. Such was the style of analysis that appealed to Ricardo, even more so when the 'intermediate' stages (the 'temporary' phenomena) were omitted altogether. Evidence of Ricardo's ingrained 'discrete' style of reasoning has been found in his earliest writings (for example, with his analysis of the effects of an increased money supply on the 'real' economy), in his later correspondence with Malthus, and also in the Principles. However, with the new view analysis (and the articulated possibility of market wages exceeding the natural level for an 'indefinite period' in an improving society) Ricardo shifted to a more 'continuous' approach: the 'impulses' come so rapidly that their individual effects have no time to dissipate. But this way of viewing economic activity was the exception rather than the rule. Ricardo was not a 'dynamic' theorist in the modern sense. Conclusion

There is much to be said in favour of the traditional interpretation of Ricardo as a natural wage theorist. The natural wage analysis, well documented in the Principles, was a developed form of the 'fixwage' analysis that is found in Ricardo's early monetary writings, and in his early writings on profits; it was integral to the 'core' thesis on 'permanent' movements in profitability, presented in the Principles', and it coheres with his ingrained, 'discrete' style of analysis. However, extreme and common forms of the natural wage interpretation have erred in their treatment of the new view passages, which have been either explained away by silence or bizarrely incorpo-

Falling rate ofprofit, wages and the law of markets

131

rated as part of the natural wage doctrine. 48 Following in the footsteps of Edwin Cannan, later writers, including Samuel Hollander, have performed a valuable service by drawing attention to the neglected, or distorted, textual material. But in claiming that the new view was Ricardo's 'true view', that he was not a natural wage theorist at all, in any substantive sense, and that both the natural wage analysis and the 'permanence' thesis are fully compatible with the new view analysis, they, and Professor Hollander in particular, have rebounded to a yet more untenable extreme. Ricardo, unlike Malthus, was at no discernible time a fully committed new view theorist. And, even in his later writings, when he showed the greatest signs of wavering, his clear preference was for the treatment of real (commodity) wages as an analytical datum and not, contrary to the spirit of the new view analysis, as an endogenous variable. It would seem that Ricardo had partially adopted the new view without appraising its (in) compatibility with his dominant natural wage analysis, thus suggesting, once again, that his textbook reputation as a master of overall theoretical consistency stands in need of correction. But to put this in perspective, the problems with his treatment of wages pale into insignificance when set against those with his treatment of'value', as we shall see. THE LAW OF MARKETS

The 'law of markets' must be one of the most powerful doctrines in the history of economic thought. In this section I review the evolution of Ricardo's position on the 'law', conceived by him as a doctrine resting on real-world tendencies; I comment on the implications for his reputation as an 'applied' economist; and I speculate on the reasons for his abiding faith in the doctrine. Ricardo's early argument against the possibility of general gluts, found in the 'monetary' writings, rested on two propositions: that aggregate net profits, after allowing for personal consumption, will be turned 'into the form of capital' (net savings are zero); and that commodities 'are sure, however abundant, to find a market to take them off if they are 'divided in the proportions in which they are usually consumed' (Appendix to the fourth edition of the High Price of Bullion 1811, in, pp. 107-8). These propositions were central to 48

See the references cited above, p. i o n . 9.

132

INTERPRETING RIGARDO

James Mill's version of 'law', and their articulation by Ricardo is indicative of his acquaintance with and, to some extent, his acceptance of, that doctrine. I introduce the qualification because it is unclear how Ricardo thought he could reconcile the 'law' with the 'output effect' of Adam Smith's 'competition of capitals' thesis (see above, p. 44). If aggregate supply (in the 'due proportions') creates its own demand, it is far from obvious why an increased national output (associated with a growth in domestic capital) should be accompanied by lower profitability. Ricardo's apparent lack of attention to this matter may be explicable in terms of his 'monetary' preoccupations (it is worth recalling that the early Ricardo also affirmed that money has a 'commodity' status and that higher money wages lead to higher prices generally which, as he subsequently argued, were irreconcilable propositions). Later, however, as he developed his 'agricultural' theory of profitability, 'law'-related propositions were deployed in support of his contention that 'profits would never fall' without a rise in (money) wages (letter to Malthus, 18 December 1814, vi, p. 162). If total demand expenditure is equal to the revenue from production, and 'those commodities only [are] produced which [are] suited to the wants and tastes of mankind' (letter to Malthus, 23 October 1814, vi, p. 148), then (aggregate) demand and supply factors cannot affect general profitability, leaving money wages as the sole governing influence.49 That was to assume (among other things) that the 'right' commodities were produced. As Ricardo explained to Malthus using a simple 'real barter' illustration (a pedagogical device also favoured by James Mill), he regarded any other market outcome as nothing more than an incautious error on the part of the producers: demand is regulated by production. Demand is always an exchange of one commodity for another. The shoemaker when he exchanges his shoes for bread has an effective demand for bread, as well as the baker has an effective demand for shoes, - and although ... the shoemaker's demand for bread must be limited by his wants, yet whilst he has shoes to offer in exchange he will have an effective demand for other things, — and if his shoes are not in demand it shews that he has not been governed by the just 49

This is not to say that Ricardo made use of the law of markets 'only because it provided a means of establishing that the mechanism of his model which produced declining profits was in fact the only mechanism which could produce declining profits' (O'Brien 1975, p. 41, emphasis in original). The fact that Ricardo used the 'law' in his early 'monetary' writings is sufficient to falsify Professor O'Brien's statement. Subsequently, Ricardo came to perceive the implications of the doctrine for his emerging theory of profitability, with which it fitted admirably.

Falling rate of profit, wages and the law of markets

133

principles of trade, and that he has not used his capital and his labour in the manufacture of the commodity required by the society, - more caution will enable him to correct his error in his future production. (18 December 1814, vi, pp. 163-4) Ricardo's producers were invariably 'governed by the just principles of trade', swiftly and perceptively adapting their production to suit the 'wants' of the (other) demanders. However, when it came to writing the Principles, he was forced to explain the disparity between this vision, according to which only 'temporary' excess supplies would occur in particular markets, and the external reality of an economy in turmoil. Some of his explanations or, perhaps more accurately, his ex post rationalisations, were rehearsed during the troublesome year of 1816. In the evocative language of the period, 1816 was a year of widespread industrial (and agricultural) 'distress' (see Smart 1910, chapter 25). By Ricardo, however, the severe economic difficulties were nothing more than a 'temporary' set-back on the road to prosperity. Writing to Trower on 15 July 1816, he pronounced: Surely the disastrous effects which always attend an important change in the employments of capital cannot much long continue and we shall soon witness a renovation of commercial activity and credit, (vn, p. 49) Napoleon had been defeated at Waterloo over one year earlier, and it seemed to Ricardo that capitalists had been given ample time to overcome the dislocating effects of the transition from war to peace. They ought to be producing the 'right' commodities in the 'proper proportions'. Two months later he was becoming noticeably impatient with their reluctance to do so. It 'appears to me that a sufficient time has elapsed to make that new distribution of employments which our altered circumstances have made necessary', he opined to James Mill (8 September 1816, vn, p. 67). But why were the capitalists being so obdurate? Ricardo tendered an explanation which took the form of an analogy with the effects of a reduction in the money supply: The duration of the intervals between marked changes are often much longer than is generally supposed. It proceeds from the opposition which is naturally given to such change. Thus 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

134

INTERPRETING RICARDO

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 any thing like accurate calculation. (8 September 1816, vn, p. 67) By implication, capitalists were engaged in the production of the 'wrong' things and, by refusing to sell their commodities at reduced prices in the vain hope that markets would revive, they were needlessly postponing 'the new distribution of employments'. This futile behaviour could only be explained (away) in terms of irrationality ('prejudices') and inaccurate information. What should be done about such human failings? Writing to Mill on 17 November 1816, Ricardo was 'sorry that the distress still continue[s]', adding that providing there is 'the most rigid economy in every branch of the public expenditure . . . nothing further can be done for us' (vn, p. 90). Presumably he believed that the capitalists would soon come to their senses, accept their losses, and transfer to new lines of production. There was certainly no cause for government intervention, as he made perfectly clear during January 1817: I am not one of those who think that the raising of funds for the purpose of employing the poor is a very effacicious mode of relief, as it diverts those funds from other employments which would be equally if not more productive to the community. That part of the capital which employs the poor on the roads for example cannot fail to employ men somewhere and I believe every interference is prejudicial. (To Malthus, 3 January 1817, vn, p. 116) 50

This was to assume that the 'diverted' funds were being, or would be, employed 'productively' by profit-seeking capitalists. The fact that this was not actually happening seems to have been regarded as almost immaterial. The 'law of markets' received its formal blessing from Ricardo in the chapter 'Effects of Accumulation on Profits and Interest' in the Principles, which opened with the claim that 'no accumulation of capital will permanently lower profits, unless there be some per50

Cf. his subsequent letter to Malthus in which Ricardo complained of'taking it [capital] out of the hands of those who know best how to employ it, to encourage industry of a different kind and under the superintendence of those who know nothing of the wants and demands of mankind and blindly produce cloth or stockings of which we have already too much, or improve roads which nobody wishes to travel' (24 January 1817, vn, p. 121).

Falling rate of profit, wages and the law of markets

135

manent cause for the rise of wages' (1, p. 289).51 In defence of this position, entailing a rejection of the once 'incontrovertible' competition of capitals doctrine, Ricardo devoted most of the chapter to an elaboration of the 'law of markets' argument that 'there is no amount of capital which may not be employed in a country, because demand is only limited by production'. The steps in his reasoning should be familiar: commodities are supplied only with a view to demanding other commodities; each producer is not 'for any length of time . . . ill-informed of the commodities which he can most advantageously produce' (1, p. 290); the desire to consume something is 'implanted in every man's breast; nothing is required but the means, and nothing can afford the means, but an increase of production' (1, p. 292); and all money income is spent, either by the direct recipients, or by those to whom they lend their unspent money balances, thus implying that aggregate net savings are zero (1, p. 291).52 The condition for (minimally) 'advantageous' production was that commodities should sell at natural prices: those prices that yield the going rate of profit, as determined by the conditions of producing the (given) real wage.53 According to the account presented by Ricardo in the chapter 'On Natural and Market Price', the 'accidental and temporary deviations of the actual or market price of commodities . . . from their primary and natural price' (1, p. 88) is erased through the expansion or contraction of output in the different spheres of production. The Titans in Ricardo's drama of economic life were the capitalists who, with their 'restless desire . . . to quit a less profitable for a more advantageous business' (behaviour which, he informs us, is 'more active than is generally 51

52

53

On the basis of the material discussed in the last section, 'permanent' wage increases should be understood as those reflecting increases in the natural prices of commodities entering the natural wage bundle. T h e 'unlimited desires' principle applies only to commodities other than food: ' N a t u r e . . . has necessarily limited the a m o u n t of capital which can at a n y one time be profitably engaged in agriculture, b u t she has placed no limits to the a m o u n t of capital that m a y be employed in procuring " t h e conveniences a n d o r n a m e n t s " of life' (i, p. 293). Ricardo's quotation is from The Wealth of Nations. Smith h a d remarked that the 'desire of food is limited in every m a n by the narrow capacity of the h u m a n stomach; b u t the desire of the conveniences a n d ornaments of building, dress, equipage, a n d household furniture, seems to have no limit or certain b o u n d a r y ' ([1776] 1981, p . 181). I n the chapter ' O n Machinery', added to the third edition of the Principles, R i c a r d o stated that he 'was, a n d am, deeply impressed with the truth of [Adam's Smith's] observation' (1, p . 387). A further qualification is that the general rate of profit must afford capitalists 'an adequate compensation for their trouble, and the risk which they must necessarily encounter in employing their capital productively' (1, p. 122).

136

INTERPRETING RICARDO

supposed') will ensure that capital is directed 'to each trade in the precise amount that is required' (1, pp. 88-90). Hence market prices will be prevented from diverging 'for any length of time' from their natural levels (1, p. 91). Nor is there any likelihood that capitalists will be unable to find a profitable outlet for capital. If 'there is no limit to demand' (1, p. 296), then something can be produced which will find a profitable market. In the passages just quoted, and elsewhere in the Principles, the impression is given of a hitchless economic system, ever hastening towards a no-saving 'equilibrium' in which all commodities exchange at natural prices.54 But there were some qualifications. Ricardo granted that 'the strength of that disinclination, which most men feel to abandon that employment of their capital to which they have long been accustomed' (1, p. 265) could dampen their ardour to redirect production in unfavourable circumstances. He also drew attention to the specificity of fixed capital, which might create a similar obstacle (1, p. 266, cf. pp. 191, 196). In more general terms: When the market prices of goods fall from an abundant supply, from a diminished demand, or from a rise in the value of money, a manufacturer naturally accumulates an unusual quantity offinishedgoods, being unwilling to sell them at very depressed prices. To meet his ordinary payments, for which he used to depend on the sale of his goods, he now endeavours to borrow on credit, and is often obliged to give an increased rate of interest. This, however, is but of temporary duration; for either the manufacturers expectations are well grounded, and the market price of his commodities rises, or he discovers that there is a permanently diminished demand, and he no longer resists the course of affairs; prices fall, and money and interest regain their real value. (1, pp. 297-8) According to this passage, it is 'natural' for producers to stockpile their goods when market prices fall and, it should be noted, to demand money directly without a prior sale, thus implying that money can be more than a mere lubricant in the exchange process (contrary to an earlier assertion that 'money is only the medium by which . . . exchange is effected', 1, p. 292). Yet Ricardo's habitual inclination to minimise the significance of'frictions' also comes to the fore when he writes that the producers' behaviour 'is but of 54

Cf. 'the profits of the favoured [foreign] trade will speedily subside to the general level' (i, p. 129); and 'If the profits of capital employed in Yorkshire, should exceed those of capital employed in London, capital would speedily move from London to Yorkshire' (i, p. 134). I return to the capital transfer mechanism in chapter 6.

Falling rate ofprofit, wages and the law of markets

137

temporary duration'. Some recognition of 'temporary' maladjustments had been necessitated by Britain's post-war 'distress', of which Ricardo's readers would have been acutely aware, but his emphasis remained firmly, and characteristically, on 'equilibrium' forces. In the second half of 1817, as the economy revived, it seemed to Ricardo that those forces were, not before time, beginning to assert themselves. In his letter to Mill of 7 August he looked forward to 'a long course of prosperity' (vn, p. 170); writing to Mai thus, on 4 September, he doubted 'whether we have even during the late distress ceased to advance as a nation in wealth, but at present I think no one can doubt that we are again making forward strides in prosperity' (vn, pp. 185-6); and again to Mai thus, on 10 October, he corresponded: We shall now I hope for some years sail before the wind. You and I have always agreed in our opinions of the power and wealth of the country, - we were not in a state of despair at the discouraging circumstances with which we were lately surrounded. We looked forward to the revival which has taken place, (vn, p. 192) Alas, the 'revival' turned out to be short-lived. By March 1819 a pall of gloom was enveloping the economy with reports of glutted markets and stagnation (see Smart 1910, chapter 31). Was Ricardo aware of the 'considerable stagnation in trade, and . . . [the] great reduction of prices in consequence?', he was asked by the Commons Committee on the Resumption of Cash Payments (4 March 1819, v, p. 384). He had 'heard so' but, he added, with disarming if not disconcerting honesty, 'I am not engaged in trade, and it does not come much within my own knowledge' (v, p. 385). This confessed lack of direct knowledge did not prevent Ricardo from knowing how the economic system should operate. Thus, in the 'Funding System' article, written during September 1819 for the Encyclopaedia Britannica, he recited the 'law of markets' in the following, uncompromising terms: demand is only limited by production; whoever can produce has a right to consume, and he will exercise his privilege to the greatest extent .. . [T]he demand for particular commodities is limited, and therefore ... there may be a glut of such commodities, but in a great and civilized country, wants, either for objects of necessity or of luxury, are unlimited, and the employment of capital is of equal extent with our ability of supplying food and necessaries for the increasing population, (iv, pp. 178-9)

138

INTERPRETATING RIGARDO

The lugubrious external circumstances had not shaken Ricardo's faith in one of his most cherished doctrines. With the publication of Malthus's Principles in May 1820, the doctrine became a major source of controversy. Malthus's position was unequivocal: the 'law' 'appears to me to be utterly unfounded, and completely to contradict the great principles which regulate supply and demand' (11, pp. 306-7); and he added that Say, Mill and Ricardo, all supporters of (versions of) the 'law', 'have fallen into some fundamental errors in the view which they have taken of this subject' (11, p. 309). Ricardo was equally unimpressed with Malthus's arguments. Writing to James Mill on 27 July 1820, he fulminated: 'The attack on Say's and your doctrine of accumulation, is supported by the weakest arguments . . . and [is] so palpably fallacious that one's wonder is he could have deliberately written it' (vm p. 212). There was to be no meeting of minds. Malthus's principal contention was that the labourers' demand for commodities is insufficient to purchase the aggregate output without a profit-destroying fall in prices. Consequently, if capitalists, landlords and 'other rich persons' desist from immediate consumption, there must be a general glut of commodities and a cessation of accumulation (see, for example, 11, pp. 300-2). This made no sense to Ricardo. In response to one of many restatements of Malthus's argument, he wrote, 'this is only saying that labour is so high that it absorbs all that fund which ought to belong to profits, and therefore the capitalist will have no interest in continuing to accumulate' (11, p. 308). This was not, in fact, what Malthus was saying, but rather a translation of his argument into Ricardo's own theoretical discourse.55 According to Ricardo's reasoning, the assumed reduction in 'luxury' consumption would increase the fund for 'productive' consumption (investment) which would in turn lead to a higher demand for labour and, before population increased, higher real (commodity) wages and therefore a fall in profitability. Higher wages and not, as Malthus contended, a general oversupply (glut) of commodities, were the cause of lower profits. The 'power to consume' was merely transferred to 'another set of consumers' (11, p. 309) - the labourers - and provided the capitalists produced commodities 'suited to the wants of those who can command them' 53

Cf. Samuel Hollander's remark that 'each used logic relevant to his own vision of the accumulation process to counter the arguments of the other' (1979, p. 527, emphasis in original).

Falling rate ofprofit, wages and the law of markets

139

then commodities 'cannot exist in such abundance as not to find a market' (11, p. 304). Ricardo evidently believed he had won the day: What I wish to impress on the readers mind is that it is at all times the bad adaptation of the commodities produced to the wants of mankind which is the specific evil, and not the abundance of commodities. (11, p. 306) Nor, as far as he was concerned, was it at all likely that people in 'developed' countries, 'with tastes for every enjoyment that nature, art or science will procure' (11, p. 341), would refrain from 'luxury' consumption. As he wrote to his doctrinal ally, McCulloch, 'Mr. Mai thus speaks of an indisposition to consume being very common I say it never exists any where, not even in South America to which he has so triumphantly alluded' (2 August 1820, vm, p. 216). The full-blown 'law of markets', complete with the 'unlimited desires' principle, continued to be preached as 'the true faith' (the apt religious description is Ricardo's, vm, p. 227). Only 'errors', 'mistakes', 'accidents' and 'miscalculations' by capitalists in particular markets, for brief periods of time, were allowed as minor exceptions to the general rule of 'rational' production.56 Malthus, for his part, maintained his implacable opposition to Ricardo's vision. A well-known exchange of letters in the autumn of 1820, with the economic 'distress' still continuing, vividly captures their respective positions. Malthus wrote: The present state of things ... in England America Holland and Hamburg, still more than in France does appear in the most marked manner to contradict both his [Say's], and your theory. The fall in the interest of money and the difficulty offindingemployment for capital are universally acknowledged, and this fact, none of your friends have ever accounted for in any tolerably satisfactory manner; but what confidence can be placed in a theory, as the foundation of future measures which is absolutely inconsistent with the past and the present state of things[?] (25 September 1820, vm, p. 260) To which Ricardo replied: The difficulty of finding employment for Capital in the countries you mention proceeds from the prejudices and obstinacy with which men persevere in their old employments, - they expect daily a change for the 56

See the following letters: to Malthus, 4 September 1820, vm, pp. 227-8; to Trower, 15 September 1820, vm, pp. 236-7; to Trower, 26 September 1820, vm, pp. 255-9; t o Trower, 3 October 1820, vm, pp. 272-4; to Malthus, 9 July 1821, ix, p. 16; and to Mill, 18 December 1821, ix, p. 132.

140

INTERPRETING RICARDO

better, and therefore continue to produce commodities for which there is no adequate demand. With abundance of capital and a low price of labour there cannot fail to be some employments which would yield good profits, and if a superior genius had the arrangement of the capital of the country under his controul, he might, in a very little time, make trade as active as ever. Men err in their productions, there is no deficiency of demand. If I wanted cloth, and you cotton goods, it would be great folly in us both with a view to an exchange between us, for one of us to produce velvets and the other wine, — we are guilty of some such folly now, and I can scarcely account for the length of time that this delusion continues. (9 October 1820, VIII, p.

277)

There was nothing wrong with the 'law of markets'. The fault lay with the economic actors who had inexplicably, and inexcusably, forgotten their lines. It should be clear that the 'law of markets' was not, for Ricardo, a purely abstract doctrine. Nor should there be any doubt as to the significance of a doctrine which generated the most important policy imperative of all: laissezfaire?1 What does Ricardo's attachment to this doctrine reveal about his standing as an 'applied' political economist? More precisely, was this an area in which Ricardo was culpable of the 'Ricardian Vice'? Samuel Hollander has claimed that Ricardo's 'refusal to admit excess aggregate supply in the post-Napoleonic period . . . reflects less a methodological predisposition than a horror of inflation' (1985b, p. 4). And he later pleads that Ricardo used 'law of markets' reasoning in depressed times 'because of the strength of his objections to those economists who maintained that the post-war crisis reflected characteristics of secular stagnation' (p. 33n.). But in making these submissions, which doubtless contain elements of truth, Professor Hollander conveniently overlooks the point that Ricardo evidently believed, passionately, in the practical relevance of his doctrine, to such an extent that he blamed reality for not conforming to the 'model', not vice versa. If that does not indicate a 'methodological predisposition', one wonders what would. More promisingly, Professor Hollander also draws attention to Ricardo's allowances 'for temporary excess demand for money in 57

In his Proposalsfor an Economical and Secure Currency, Ricardo did allow for 'a few, and a very few exceptions' to 'the liberty of trade' when 'the purchasers are not supposed to have, or be able to acquire sufficient knowledge to guard them against deception'; as examples of legitimate interference he gave examination-setting for medical practitioners, the validation of drug purity, and the hallmarking of plate and money (iv, pp. 71-2).

Falling rate of profit, wages and the law of markets

141

consequence of an initial reduction in the money supply' (p. 27). This links up with the question of whether Ricardo subscribed to a 'Say's equality' version of the 'law' (implying that there is always zero excess demand for money, and therefore zero excess demand for other commodities), or a 'Say's identity' version (with the zero excess demand characteristics restricted to equilibrium states). I judge the 'equality' interpretation to be preferable, although it must be said that the distinction between 'equality' and 'identity' statements is extremely blurred in Ricardo's writings. His emphasis on 'equilibrium' tendencies was so pronounced that it can easily appear as if properties which hold in 'equilibrium' alone were regarded as necessarily true at all times. Yet he did allow for the possibility of excess demand for money (and unsold commodities), if only for periods of'temporary duration' (1, p. 298).58 More generally, the various 'frictions' in the system which he rather grudgingly allowed for do point in the 'equality' direction. But all such phenomena were, in the end, regarded as ephemeral departures from the dominant course of events. I submit that this is an area of Ricardo's work which displays many 'Vice'-like characteristics. The 'law of markets' rested on a highly simplified 'model' of economic behaviour, so simple that it could be reduced to barter relationships between two producers. It was taken to imply an unambiguous policy (laissez faire). And, although it did not totally ignore 'temporary effects', their implications were treated as nugatory. One final question remains for consideration. Since Ricardo's confidence in the 'law of markets' did not derive from his detailed acquaintance with trade (as he admitted), from whence did it come? It has been perceptively stated that Ricardo assumed 'a stockbroker in every man' (Fetter 1969).59 This might be put more strongly, for Ricardo seems to have assumed that there was a stockbroker similar to himself in every man. The story of his success on the financial markets is well known, but it may be worth recalling Mallet's sketch of his market behaviour: 38

59

The full passage is quoted above, p. 136. Also see the passage from Ricardo's letter to Malthus of 8 September 1816, above pp. 133-4. Alfred Marshall had made a similar point. He observed that 'Ricardo and his followers' had 'regarded man as . . . a constant quantity, and gave themselves little trouble to study his variations. The people whom they knew were chiefly city men; and they took it for granted tacitly that other Englishmen were very much like those they knew in the City' (Marshall [1885] I925» PP- !54-5)-

142

INTERPRETING RIGARDO

He is said to have possessed an extraordinary quickness in perceiving in the turns of the market any accidental difference which might arise between the relative price of different stocks, and to have availed himself of this advantage . .. He is also said never to have carried his stock transactions to any speculative extent; but to have always, or generally, sold out on the turn of the market, (x, p. 73) Ricardo apparently presumed that capitalists (like the bullion merchants in the early 'monetary' writings) had the same propensities as himself. They were economically self-interested, rational and, as something of an understatement, highly attuned to actual and potential market possibilities. It was therefore inconceivable that they would languish in relatively unprofitable spheres of production when, as Ricardo believed, they could produce something which would bring them a better rate of return. This may seem an innocuous proposition to modern readers, habituated to the assumption of perfect knowledge, but the difference is (perhaps) that Ricardo thought he was describing the behaviour of real producers. It was, then, a deep conviction that capitalists behave as Ricardo thought they ought to behave, as he thought he would behave in their position, that may help to explain his faith in the 'law of markets'. Additionally, it was necessary for him to believe that demand (including the demand for 'power' through capital accumulation) was, in principle, unlimited, and that people would not choose to hoard money. Having convinced himself of the truth of the former proposition, and it does not seem that Ricardo needed much convincing, the latter proposition followed as a rational consequence. Why hoard money when it could be spent on immediate consumption or profitably invested? Ricardo could not think of a reason, so he concluded that none existed. And again, he took for granted that real people would behave accordingly. The final ingredient was methodological: Ricardo's disposition to put 'immediate and temporary effects quite aside, and fix my whole attention on the permanent state of things which will result from them' (letter to Malthus, 24 January 1817, vn, p. 120). Implicit in this statement was the belief that there is a 'permanent state of things' (loosely speaking, an 'equilibrium' state) to which 'temporary' ('disequilibrium') paths must lead. And since Ricardo 'knew' the characteristics of this 'permanent' state — commodities exchanging at natural prices, with all revenue spent - that clinched the matter. The result was a vision of free-market capitalism which

Falling rate of profit, wages and the law of markets

143

functioned with almost clockwork precision: an economic system that would continue to prosper and expand without foreseeable limit, on the sole condition that legislators take the trouble to learn the few 'simple principles' of Political Economy and desist from unwise legislative interference (letter to Trower, 12 November 1819, VIII, pp. 132—3). The connection between the 'law' and this vision of economic behaviour was captured brilliantly by Keynes: The celebrated optimism of traditional economic theory, which has led to economists being looked upon by Candides, who, having left this world for the cultivation of their gardens, teach that all is for the best in the best of all possible worlds provided we will let well alone, is also to be traced ... to their having neglected to take account of the drag on prosperity which can be exercised by an insufficiency of effective demand. For there would obviously be a natural tendency towards the optimum employment of resources in a Society which was functioning after the manner of the classical postulates. It may well be that the classical theory represents the way in which we should like our Economy to behave. But to assume that it actually does so is to assume our difficulties away. (Keynes 1936, pp. 33-4) How well this applies to Ricardo. CONCLUSION

Writing to his friend Hutches Trower in January 1818, Ricardo claimed to have 'a very consistent theory in my own mind' (vn, p. 246). In comparison with his earlier work, he had indeed produced a more consistent theory by the time he wrote his Principles, but not to the extent that he imagined. What appears to have happened is that he developed a basic 'core' of doctrine which was then overlaid with various qualifications and supplemented by material, notably the new view treatment(s) of wages, which was strikingly incongruous. The theoretical core included the following: the 'pure' labour theory of value (yet to be considered in detail), the natural wage analysis, progressively diminishing agricultural returns and the related theory of differential rent, and the 'law of markets'. Using the framework constructed from these materials, Ricardo's prime ambition, reflected in the content of the main theoretical chapters in the Principles, was to trace the 'permanent' effects of accumulation on wages, profits and rent; and in doing so he mostly used the simple, and (in)famous, long-run comparative static (or 'discrete')

144

INTERPRETING RIGARDO

method of analysis. More particularly, he sought to prove that accumulation must result in a 'permanently' declining rate of profit as a consequence of diminishing agricultural returns. Since, as far as he was concerned, that was the only reason why general profitability would 'permanently' fall, the message was clear: the Corn Laws should be scrapped. The above sketch of Ricardo's work is a familiar one. Some commentators would have reservations over the inclusion of the 'pure' labour theory in the core, others might dispute the presence of the 'law of markets', but I believe it is a sketch which most would find broadly acceptable. It corresponds to the 'mature Ricardo' of textbook fame. The fact is, however, that the picture is incomplete. It leaves out certain things that simply do not fit, such as the new view treatment of wages, and it suggests that Ricardo was almost entirely concerned with the single issue of the Corn Laws. These and other weaknesses have been fastened upon by Samuel Hollander who has, in recent times, led the challenge to received opinion. I hope to have shown that Professor Hollander's contrary view of Ricardo is even more seriously flawed, most conspicuously so with his interpretation of Ricardo's treatment of wages, but also with his attempt to diminish the significance of the 'agricultural' model, and to portray Ricardo as almost completely innocent of the 'Ricardian Vice'. On all these issues, my conclusion is that, although received opinion may stand in need of qualification, it does not deserve wholesale rejection.

CHAPTER 4

The labour theory of value (I)

In this and the following chapter, I trace the evolution of Ricardo's treatment of value from 1815 to 1823. The present chapter proceeds as follows. In the first section I reconstruct the stages in Ricardo's thinking on 'value' — encompassing his adoption of the 'pure' labour theory and his subsequent development of the 'curious effect' — as he struggled to compose the Principles. Then, in the second section, I critically evaluate the published outcome of his deliberations. In the third section I consider, in turn, Ricardo's correspondence on 'value' in the interval between the first two editions of the Principles, his debate with Robert Torrens, which ultimately led to the adoption of a 'dated labour' analysis, and the significance of the revisions to the chapter 'On Value' in the second edition of his major work. An interim conclusion completes the chapter. BEFORE THE PRINCIPLES

In August 1815, James Mill genially threatened Ricardo with 'no rest, till you are plunged over head and ears in political economy' (vi, p. 252). Two years later the Principles was published but, for Ricardo, authorship was to be an uphill struggle. Some of his difficulties were immediately communicated to Mill. There was his 'disinclination to write', the 'visits of friends [which] make great encroachments on my time', and 'the temptation of being out in the air in fine weather' (30 August 1815, vi, pp. 262-3).* Mill was 1

Cf. the following, from Ricardo's letter to Malthus of 10 September 1815: 'It is the general remark in our part of the country that a finer season was never remembered . . . The temptation to enjoy it has been so great that I have been incessantly out with someone or other of my friends, who have been staying with me . . . which makes such inroads on my time that I find I have much less leisure here for reading and study than I have in London' (vi, p. 267).

H5

146

INTERPRETING RICARDO

unimpressed. In October, he requested 'some account of the progress you have been making in your book' (10 October 1815, vi, p. 309). This met with a discouraging response: 'Here at Gatcombe we go on much as usual, - we are always full of visitors, some of which make great demands on my time . . . At the present moment we are full to overflowing. Last night we had a dance and sat down to supper a party of no less than 49 which did not disperse before 4 oClock this morning'; and then, 'you will not be surprised to hear that I have done very little in the way of studying or writing' (24 October 1815, vi, p. 312). Mill responded to Ricardo's indolence by 'commanding' him 'to begin to the first of the three heads of your proposed work, rent, profit, wages — viz. rent, without an hours delay. If you entrust the inspection of it to me, depend upon it I shall compell you to make it all right, before you have done with it' (9 November 1815, vi, p. 321, emphasis in original).2 And a few weeks later: This is what you must do with the opus magnum. In thefirstwriting, be not very solicitous about any thing, but about getting out all the ideas which appear to you to bear upon the subject, and to be conducive to its elucidation. When this is done, we shall have no great difficulty in marshaling them, and putting each in the place in which it will receive most light from others and shed most light upon them. (1 December 1815, vi, p. 330, emphasis in original)3 It was understandable that Mill should have recommended this style of writing in view of Ricardo's prevarication and his neurotically self-proclaimed difficulties in the art of composition. But when, eventually, Ricardo followed Mill's advice, the final results were disastrous. Three unrequited weeks after his last letter, Mill's new gambit was to set Ricardo 'a school exercise' to prove that 'improvements in agriculture . . . raise the profits of stock, and produce immediately no other effects' (22 December 1815, vi, p. 339). (The 'exercise' was supplemented with the helpful advice that Ricardo 'should resist 2 3

This order of topics for the 'proposed work' was the same as in the Essay on Profits. Cf. the following, from Mill's letter to Ricardo of 14 August 1816: 'I solemnly command and ordain that you proceed, without loss of time, on the plan which you have already sketched out, till you have gone over the whole field of Political Economy, from the beginning to the end, thinking nothing of order, thinking nothing of repetitions, thinking nothing of stile regarding nothing, in short, but to get all the thoughts blurred upon paper some how or another. We shall see what is to be done with it after that' (vn, p. 60).

The labour theory of value (I)

147

some of the most frivolous of social calls', vi, p. 340.) His errant pupil was not quite ready to oblige: As soon as I get rid of it [the pamphlet, Proposals for an Economical and Secure

Currency, published in February 1816] I will begin on the proposition which you require to be proved. I know I shall be soon stopped by the word price, and then I must apply to you for advice and assistance. Before my readers can understand the proof I mean to offer, they must understand the theory of currency and of price. They 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. (30 December 1815, vi, pp. 348-9)

It was indeed Ricardo's understanding of the 'theory of currency and price' that underwent the greatest change before the Principles, without any substantive 'advice and assistance' from Mill. Here we must retrace our steps. Before the Essay on Profits (1815), Ricardo had espoused the Smithian view that all prices rise following an increase in the price of corn. Then, in the Essay, the Smithian doctrine was repudiated on the grounds that money is a commodity like any other. This 'commodity' view of money had been articulated by Ricardo in his earliest writings, but it was only by the time of the Essay that he perceived the implication that if, measured in corn, the relative expenses of producing the money and other commodities remain unchanged when corn measured costs of agricultural production increase, the other commodities would not rise in price. So far so good. But in stating and elaborating the principles which do govern changes in exchangeable value, Ricardo came unstuck. In March and April 1815, he expressly rejected Malthus's basis for calculating price variations — prices change in proportion to changes in the number of labourers employed for a given output — in favour of the circular one which had corn rising in price in proportion to changes in the corn measured expenses of production relative to the corn

148

INTERPRETING RICARDO

output. 'Difficulty or facility of production', which supposedly regulated changes in 'value' [exchangeable or relative value), was itself conceived as a 'value' relationship between output and inputs. It was not an independent determinant of'value'. There was a related problem. In referring to the value of a commodity, Ricardo's occasional practice was implicitly to identify an 'absolute' value, determined by circumstances peculiar to an individual commodity, regardless of its general exchangeable value. But if calculation of the 'absolute value' requires information of exchangeable value — of the cost of producing the commodity in terms of output - then it cannot be conceived apart from exchangeable value. Moreover, there was the confusion surrounding the use of'value' in both the relative and the 'absolute' senses (this problem was to remain with Ricardo, and has often befuddled commentaries on his work). It may well have been the gradual recognition of these various contradictions, and especially the 'circulatory' problem, which impelled Ricardo to recast his value analysis. To substantiate the above points consider, first, the letter to Mill of 30 December 1815 (quoted more fully above, p. 147). If there can be an 'alteration in the value of the particular commodity . . . which affects the value of no other thing', 'value' is not being used in the sense of exchangeable or relative value; for in the latter sense of 'value', a change in the (relative) value of one commodity necessarily implies an inverse change in the (relative) value of other commodities. Similarly, the 'invariability of the value of the precious metals' also implies a concept of 'absolute' value. Money, supposedly, can be of an 'invariable value' even though it exchanges for more or less of other things. The usage of 'value' in such contexts thus departed from mere exchangeable value. But concurrently, it was only the relative concept of'value' that received formal enunciation, as in the pamphlet Proposals for an Economical and Secure Currency, which pro-

vides a snapshot of Ricardo's position in late 1815 on the 'theory of currency and price'. The 'value of a commodity', he declared in the pamphlet, is 'estimated by the quantity of other things generally for which it will exchange', (iv, p. 60). This concept of value is unambiguously one of exchangeable, or relative value. Yet, within the space of a couple of paragraphs, he wrote: A hat may exchange for less of tea, sugar, and coffee, than before, but, at the same time, it may exchange for more of hardware, shoes, stockings, &c,

The labour theory of value (I)

149

and the difference of the comparative value of these commodities may either arise from a stationary value of one, and a rise, though in different degrees, of the other two; or a stationary value in one, and a fall in the value of the other two; or they may have all have varied at the same time, (iv, pp. 60-1) In this passage, the possible source of changes in relative value is traced to changes in the 'value' of individual commodities in an 'absolute' sense. But how would a change in this latter concept of 'value' be discovered? Only by reference to an 'invariable measure of value', a commodity invariable in its own value (iv, p. 60). This species of exchangeable value (relative to the standard) would thus indicate a change in the 'absolute' value of the measured commodities: a change in their prices (the exponent of their values relative to the standard) would uniquely reflect changes in their own values. In practice, Ricardo argued, there is no such thing as an 'invariable measure of value' (iv, p. 60). But could there be one in principle? This question is not directly confronted, although we do find the following, pertaining to the source of variations in 'value': Some commodities are rising in value, from the effects of taxation, from the scarcity of the raw material of which they are made, or from any other cause which increases the difficulty of production. Others again are falling, from improvements in machinery, from the better division of labour, and the improved skill of the workman: from the greater abundance of the raw material, and generally from greater facility of production, (iv, pp. 59-60) The inference may be drawn that 'absolute value', as I have termed it, was related to 'difficulty or facility of production'. But how was 'difficulty or facility of production' conceived? Had Ricardo escaped from its circular computation in terms of the exchangeable value of produce relative to costs of production? The passage from the Proposals does not yield an answer. 'Difficulty of production' covered taxation, 'scarcity' or 'abundance' of raw material, improvements in machinery, and the worker's skill, with no guidance as to how these various elements are commensurated. Ricardo's treatment of'value' at this time left abundant scope for clarification. He came to realise that there were problems, writing to Malthus on 7 February 1816: I have not thought much on our old subject, - my difficulty is in so presenting it to the minds of others as to make them fall into the same chain of thinking as myself. - If I could overcome the obstacles in the way of

150

INTERPRETING RICARDO

giving a clear insight into the origin and law of relative or exchangeable value I should have gained half the battle, (vn, p. 20)

The 'obstacles' proved difficult for him to surmount. Two months later, Ricardo described them as 'almost invincible', adding that he found 'the greatest difficulty to avoid confusion in the most simple of my statements' (24 April 1816, vn, p. 28). Malthus could barely disguise his schadenfreude: I cannot help thinking that the reason why with your clear head, you find a difficulty in your progress is that you are got a little into a wrong track. On the subject of determining all prices by labour, and excluding capital from the operation of the great principle of supply and demand, I think you must have swerved a little from the right course. But on this point of course you differ from me. (28 April 1816, vn, p. 30)

It would seem that Ricardo had come to associate (changes in) exchangeable value with (changes in) comparative quantities of labour required to produce commodities, thus supplying a concept of 'difficulty or facility of production' which could be conceived independently of exchange value.4 By August 1816, Ricardo was at last triumphing over his recurring bouts of indolence and defeatism, thanks in part to an atrocious English summer which had thwarted his outdoor pursuits. A 'pretty mass of papers' had been assembled, 'written first and last upon the subject', which Mill 'commanded' Ricardo to send for inspection.5 In the same month, Ricardo wrote to Malthus: • This raises an interpretative conundrum. It was Malthus who had favoured the use of a 'labour quantity' method of calculating price changes in the post-Essay debate with Ricardo. Why, then, should he implicitly criticise Ricardo for 'determining all prices by labour'? I suggest that his criticism should be read as a composite: 'determining all prices by labour, and excluding capital from the great principles of supply and demand'. Malthus seemingly identified (at least) two influences on prices: facility of production, and the level of profits as regulated by 'the demand compared with the supply of capital' (vn, p. 77). Admittedly, he had himself calculated price changes as a sole function of changes in 'facility of production', but this may have been a (somewhat disguised) form of ceteris paribus reasoning. Perhaps his complaint against Ricardo was that he appeared to deny the existence of any influence on (changes in) exchange value other than labour-quantity conditions of production. ' The 'pretty mass of papers' quotation is from Mill's letter to Ricardo of 14 August, vn, p. 60. Ricardo confirmed the existence of the papers in his reply of 8 September, vn, pp. 65-6. For his outbursts of authorial despair, see his letter to Malthus, 28 May 1816, vn, p. 36, and to Mill, 8 August 1818, VII, pp. 53-4. For his reports on the weather, see his letter to Trower, 15 July 1815, VII, p. 48, his letter to Mill, 8 August 1816, VII, pp. 53-4, and his letter to Malthus, 5 October 1816, VII, pp. 70-1.

The labour theory of value (I)

151

there are two causes which raise the wages of labour, one the demand for labourers being great in proportion to the supply - the other that the food and necessaries of the labourer are difficult of production, or require a great deal of labour to produce them. (9 August 1816, vn, p. 57) And also: if the demand for cloth or gold be either great [er] or less than the supply they may rise or fall in their exchangeable value. That is to say their market value might rise or fall but their natural value would probably undergo little variation, and therefore after a time they would exchange at their usual rates, (vm, p. 58) In the first of these extracts, 'difficulty of production' is conceived in terms of the labour required to produce commodities, as would be expected from a 'pure' labour theorist. (It was only later that Ricardo discovered the exceptions to the labour theory, out of which he fashioned the 'curious effect', discussed below.) Linking this with the second extract, there is the new category of 'natural value' which, presumably, was a function of labour expenditure, and played the role of the shadowy concept in the earlier writing which has been dubbed 'absolute value'. Now assume, as an extension of this reasoning, the existence of an 'invariable measure of value' - the 'sheet anchor on which all my propositions are built' defined as a commodity always requiring the same quantity of labour in its production and hence of an unvarying 'natural value'. It would follow that changes in the exchangeable value of other commodities relative to the standard — excluding those resulting from changes in 'market value', to use Ricardo's 1816 terminology — uniquely reflect changes in their 'natural value', irrespective of their exchangeable value with other commodities generally. It may be conjectured that the 'pretty mass of papers' contained similar reasoning. Mere development of a 'pure' labour theory was not, however, an intellectual resting place for Ricardo, who came to realise that 'capital structures' may differ between production processes, with modifying consequences for the labour theory. The detailed reasoning belongs in the following section but, briefly, 'capital structure' (which is not Ricardo's term) covers both the ratio of fixed to circulating capital, and the durability of fixed capital; 'fixed capital' refers to (expenditure on) long-lasting means of production, such as machinery (the longer they last, the greater their 'durability'); and

152

INTERPRETING RICARDO

'circulating capital' comprises of (expenditure on) items which, in the limit, are completely used up in a production period (for instance, wage payments to 'direct' labourers). If capital structures differ, commodities will not exchange in proportion to labour expenditures, and a change in distribution will itself result in a new set of exchange relationships, without any change in labour expenditures.6 The first intimation of Ricardo's new 'discovery', initially regarded as a stumbling block, comes in a letter to Malthus of 5 October 1816: I have been very much impeded by the question of price and value, my former ideas on those points not being correct. My present view may be equally faulty, for it leads to conclusions at variance with all my preconceived opinions, (vn, pp. 71-2) Nine days later, Ricardo dispatched his 'papers' for Mill's inspection, the draft manuscript containing material that later took shape as the first seven chapters of the Principles.1 His accompanying letter included the following: I have been reading [the papers] over in the Post Chaise, and they are really so little connected, so imperfect, and altogether so very bad in their present state, that I am not doing myself common justice to expose them even to your friendly eye. They are worse than they otherwise would be in consequence of my becoming better acquainted with the subject as I have proceeded. Much of what is said in the beginning should be left out or altered to agree with what I think the more correct views which I afterwards adopted. You will see the curious effect which theriseof wages produces on the prices of those commodities which are chiefly obtained by the aid of machinery and fixed capital. (14 October 1816, vn, p. 82) He continued: I have been beyond measure puzzled to find out the law of price. I found on a reference to figures that my former opinion could not be correct and I was full a fortnight pondering on my difficulty before I knew how to solve it. During that time I could not proceed or I should have made greater progress, (vn, pp. 83-4) 6 7

See above, p. i 8 n . 25, for a preview of the underlying logic. When sending the papers Ricardo wrote, 'I shall now consider the subject of taxation' (vn, p. 84). The first chapter on taxation in the Principles is the eighth.

The labour theory of value (I)

153

The 'curious effect' was the demonstration that a rise in money wages and an associated reduction in general profitability will induce a fall in the prices of all commodities produced with the aid of fixed capital, where price is expressed in terms of a commodity produced without any fixed capital at all. The 'effect' is therefore predicated on unequal capital structures and was, I suggest, perceived on one level as a 'solution' to the problems which Ricardo's prior recognition of such inequalities had caused him for the 'full fortnight'. As we find in the Principles', it was as if he thought that he had turned the influence of divergent capital structures to his own advantage, with the 'curious effect' not merely overturning, but actually inverting, the Smithian prediction of price increases following a rise in money wages. This was all very well, but on another level the allowance for differences in capital structure seriously undermined a reliance on 'pure' labour theory reasoning which was, I have suggested, a characteristic of the 'pretty mass of papers'. 8 If Ricardo still intended to use the 'pure' labour theory, he would surely have to square this with the 'curious effect' and its underpinnings. Perhaps he was aware of these problems, or at least some of them, hence his remark to Mill that '[m]uch of what is said in the beginning should be left out or altered to agree with what I think the more correct views which I afterwards adopted', where 'the beginning' presumably refers to the first part of the manuscript, in which he had outlined the 'unmodified' labour theory, and the 'more correct views which I afterwards adopted' were the implications of unequal capital structures and the 'curious effect'.9 But as we shall see, the revisions that he made were far from satisfactory. And James Mill was no help at all. It will be recalled that Mill had promised to 'make it all right' by polishing and rearranging Ricardo's text. Perhaps he was so delighted to receive anything from Ricardo that his critical faculty went into suspension. At any rate, his reaction to the manuscript bordered on ecstasy, especially with regard to the treatment of 'value': 8

9

It cannot be established whether the earlier papers covered so much ground as the ones eventually sent to Mill, but Mill's description of them as having been written 'from first and last upon the subject' (vn, p. 60) makes it plausible to suppose that their scope was similar. Oddly, Sraffa apparently thought that 'the beginning' referred to a much later part of the draft manuscript: 1, pp. xvi-xviii.

154

INTERPRETING RIGARDO

I think you have made out all your points. There is not a single proposition the proof of which I think is not irresistible. With the curious result pointed out in your letter . . . I was very much struck; but have no doubt whatsoever as to the validity of your conclusions, the proof of which I think is incontrovertible. Your explanation of the general principle that quantity of labour is the cause and measure of exchangeable value, excepting in the cases which you except, is both satisfactory, and clear. Your exposition and argumentation to shew, in opposition to A. Smith, that profits of stock do not disturb that law, are luminous. So are the exposition and argumentation to shew that rent also operates no such disturbance. . . . The argument thus far is not only convincing, but clear, and easily understood. (18 November 1816, vn, p. 98)

This came as a tonic to Ricardo. 'How very encouraging your letter is!' he exclaimed, adding that T shall pursue my work with increased energy, and I hope with more confidence than has hitherto accompanied my exertions' (2 December 1816, vn, p. 100). Four years later, this newly found confidence would be severely dented. THE FIRST EDITION OF THE PRINCIPLES Value and the invariable standard

The opening chapter of the Principles, 'On Value', is replete with unanswered questions, misplaced material, and internal contradictions. It provides vivid evidence of the evolution of Ricardo's thought on 'value' and, when considered together with the value analysis in subsequent chapters, of his failure to reconcile the 'pure' labour theory with the 'curious effect' material. 10 In this subsection, I present a detailed commentary on Ricardo's first chapter and draw attention to its uneasy relationship with subsequent chapters. More detailed issues are taken up in the following sub-section. The chapter 'On Value' begins with a quotation from Adam Smith's Wealth of Nations on the distinction between 'value in use' and 'value in exchange', with Ricardo drawing the moral that utility, by which he means the power of something to 'contribute to 10

Cassels' observation that Ricardo had failed to 'rewrite the text completely as his thinking on the subject progressed' was extremely perceptive, although the same cannot be said of his major thesis that Ricardo was neither 'an adherent or an exponent' of the labour theory. Cassels (1935).

The labour theory of value (I)

155

our gratification', is 'not the measure of exchangeable value, although it is absolutely essential to it': If a commodity were in no way useful... it would be destitute of exchangeable value, however scarce it might be, or whatever quantity of labour might be necessary to procure it. (1, p. 11)11 All commodities that exchange possess utility, but Ricardo's theory of value applies only to a 'mass of commodities daily exchanged' all of which 'can be increased in quantity by the exertion of human industry, and on the production of which competition operates without restraint' (1, p. 12). The force of assuming 'competition' and 'reproducibility' is that these conditions together ensure a tendency for commodities to sell at their natural prices: those prevailing when wage and profit rates are respectively uniform. As far as Ricardo was concerned, his labour theory only (ever) applied to commodities selling at their natural prices. 'In the early stages of society', Ricardo continued, the exchangeable value of the selected 'mass' of commodities 'depends solely on the comparative quantity of labour expended on each' (1, p. n ) . Following more quotations from the Wealth of Nations, apparently introduced to gain Smithian authority for his position, Ricardo wrote: That this [comparative quantities of labour expended] is really the foundation of the exchangeable value of all things, excepting those which cannot be increased by human industry, is a doctrine of the utmost importance in political economy. (1, p. 13) And then, in a remark which might justly be applied to himself: for from no source do so many errors, and so much difference of opinion in that science proceed, as from the vague ideas which are attached to the word value. (1, p. 13) Except for Ricardo's brief discussion of'use value', 'value' has so far been used in the sense of exchangeable value, with the focus on the determination of static exchange ratios. But the 'labour expended' principle is also extended to cover changes in such ratios: If the quantity of labour realized in commodities, regulate their exchangeable value, every increase of the quantity of labour must augment the value of that commodity on which it is exercised, as every diminution must lower it. (1, p. 13) 11

I postpone discussion of Ricardo's treatment of 'utility', 'scarcity' and other related matters to chapter 6.

I56

INTERPRETING RICARDO

We have the outline of a simple labour theory: (changes in) exchange relationships are related exclusively to (changes in) comparative amounts of labour expended on the production of commodities. Ricardo next directed some critical fire against Adam Smith, hitherto portrayed as a venerable ally, for professedly reneging on 'labour expended' as the 'original source of exchangeable value' and conflating the measures of 'labour expended' and 'labour commanded' (1, p. 14).12 His remarks are most illuminating for the light they cast on his own preoccupations. It is here that Ricardo first mentions an 'invariable standard', which should be capable of 'indicating correctly the variations of other things' (1, p. 14). The 'other things' are (bundles of) freely reproducible, competitively produced commodities, and, in context, the 'variations' which should be indicated are those arising from changes in the 'quantity of labour necessary to . . . production' (1, p. 15, emphasis in original). Now suppose, following Ricardo's argument, that an approximately constant wage bundle prevails over a period of time in which the quantity of labour required to produce wage-goods has changed (1, pp. 15-16). According to his reasoning, 'the value of labour' has changed (1, p. 15), whereas a 'labour commanded' standard would imply that the two wage bundles are of (approximately) equal value. The sum of money corresponding to the first bundle commands, say, one day's labouring activity, but so does the second. That made little sense to Ricardo. His concern with changes in the difficulty of producing commodities, especially wagegoods and more particularly corn, was such that it evidently seemed absurd to claim that the value of the wage bundle is unchanged regardless of changes in its circumstances of production. Understandably, if erroneously, he thought he could prove that the contrary view to his own was flawed logically. On the criterion for identifying an 'invariable standard', Ricardo seemed to be in no doubt: If any one commodity could be found, which now and at all times required precisely the same quantity of labour to produce it, that commodity would be of an unvarying value, and would be eminently useful as a standard by which the variations of other things might be measured. (1, p. 17, n. 3) 12

According to the 'labour commanded' measure of real exchangeable value which, Ricardo's protestations notwithstanding, appears to have been Smith's choice for all stages of society, the 'real value' of commodities is ascertained by dividing their nominal value

The labour theory of value (I)

157

In practice, he conceded that of 'such a commodity we have no knowledge', but he added that it is 'of considerable use towards attaining a correct theory, to ascertain what the essential qualities of a standard are, that we may know the causes of the variation in the relative value of commodities' (1, p. 17 n. 3). We would know whether other commodities required more or less labour to produce them. Observe, further, that the description of the standard as being 'of an unvarying value' implies an 'absolute' concept of value. The standard might exchange for more or less of other things - there could be changes in its exchangeable value - but its own ('absolute') 'value' is always the same.13 The next phase of the discussion turned on (some of) the implications of 'different qualities of labour' (1, p. 20) for the labour theory. Conveniently leaning on Adam Smith, Ricardo claimed: The estimation in which different qualities of labour are held, comes soon to be adjusted in the market with sufficient precision for all practical purposes, and depends much on the comparative skill of the labourer, and intensity of the labour performed. (1, p. 20)14 The existence of such a 'scale' immediately creates a problem for the labour theory, since commodities will not exchange in proportion to physically calculated ratios of'labour expended'. Account must also be taken of the different qualities of labour, as indicated by wage relativities.15 But neither here, nor anywhere else in his writings, did Ricardo take the problem seriously. Having assured his readers that the scale of wages is 'liable to little variation' (1, p. 20), he continued: If a piece of cloth be now of the value of two pieces of linen, and if, in ten years hence, the ordinary value of a piece of cloth should be four pieces of linen, we may safely conclude, that either more labour is required to make

13

14 15

(money prices) by the ruling nominal price of labour (money wage). It is not a 'determinant' of exchange relationships between commodities, contrary to Ricardo's insinuation. This was pounced upon by Samuel Bailey: 'If value, however, denotes merely a relation, this proposition cannot be true' ([Bailey] 1825, P- 9)Cf. Smith ([1776] 1981, pp. 48-9). 'If the differences of skill in different employments are so little variable . . . it proves only that they are circumstances which permanently affect value, and that it must be altogether incorrect to designate quantity of labour the sole cause, when quality of labour is so steady in its effects' ([Bailey] 1825, PP- 2 I 2 ~ I 3> emphasis in original). This has remained a favourite line of attack for critics of labour theories. See, for example, Cannan (1929, p. 175), and Schumpeter (1954, p. 594). Here it is worth noting that an analogous problem will arise with differential profit rates, a phenomenon alluded to by Ricardo in a later chapter, but without reference to the theory of value (1, p. 90).

I58

INTERPRETING RICARDO

the cloth, or less to make the linen, or that both causes have operated. (1, p.2l)16

Moreover, the analytical focus was not, in any case, on 'long-period' value comparisons: As the inquiry to which I wish to draw the reader's attention, relates to the effect of the variations in the relative value of commodities, and not in their absolute value, it will be of little importance to examine into the comparative degree of estimation in which the different kinds of human labour are held ... [It] can have little effect, for short periods, on the relative value of commodities. (1, pp. 21-2, my emphasis) The 'inquiry' indeed related to 'the effect of variations in the relative value of commodities', especially the effect of a rise in the relative value of corn. But, as in the chapter 'On Profits', this rise in the relative value of corn is attributed to a rise in its own value as dependent on changes in the quantity of labour required to produce it. A change in 'absolute value', in the above context, appears to mean something more akin to changes in the labour expended on commodities over long periods of time.17 The consequence of allowing for differential wage (and profit) rates is, to repeat, the destruction of the 'pure' labour theory if conceived as a means for determining relative natural prices from knowledge of labour expenditures alone. But if differentials are stable, and the same types of labour are always required to make the same commodities, in the same proportions (and also remaining at the current level of abstraction), then, as Ricardo stated, changes in exchange ratios are still uniquely referable to changes in labour expenditures. However, Ricardo did not confine himself to such a limited interpretation of the labour theory. Having mentioned the 'heterogeneous labour' problem, he proceeded as if it did not exist. He may therefore be read 'as if the assumption of homogeneous labour had been made. Having 'disposed' of the different 'qualities of labour' problem, with Adam Smith's assistance, Ricardo wrote: He should have added that either more or less labour might be required to make both the cloth and the linen, to differing degrees. Ricardo's somewhat opaque usage of'absolute value' in this context was noted by Samuel Bailey, who confessed that he could not pin down its meaning (1825, p. 224). In contrast, Marx, commenting on the same passage in the third edition of the Principles, claimed that 'absolute value' was being used to denote the labour time expended on a commodity ([1862-3] 1969, p. 172). It does not seem to me that the context justifies Marx's certitude.

The labour theory of value (I)

159

though Adam Smith fully recognized the principle, that the proportion between the quantities of labour necessary for acquiring different objects, is the only circumstance which can afford any rule for our exchanging them for one another, yet he limits its application to 'that early and rude state of society, which precedes both the accumulation of stock and the appropriation of land;' as if, when profits and rent were to be paid, they would have some influence on the relative value of commodities, independent of the mere quantity of labour that was necessary to their production. Adam Smith, however, has no where analyzed the effects of the accumulation of capital, and the appropriation of land, on relative value. It is of importance, therefore, to determine how far the effects which are avowedly produced on the exchangeable value of commodities, by the comparative quantity of labour bestowed on their production, are modified or altered by the accumulation of capital and the payment of rent. (1, pp. 22~3n.) All this criticism amounts to is that Smith had shown no interest in developing a labour expended theory of exchangeable value, which is true. But at least Ricardo had set his own agenda. The question of the 'payment of rent' is deferred by Ricardo to his second chapter, where it emerges that rent cannot occasion any modification to the 'pure' labour theory because it is not a component of price (see above, pp. 90-1). With 'capital', however, the outcome can be very different, as Ricardo proceeded to demonstrate. He first located this new phase of his analysis in the 'early stages' of society, where beaver and deer hunters both use 'capital' in the shape of weapons. In these circumstances, the value of beaver and deer 'would be regulated, not solely by the time and labour necessary to their destruction, but also by the time and labour necessary for providing the hunter's capital, the weapon, by the aid of which their destruction was effected' (1, p. 23). (As he stated elsewhere, less capital 'is the same thing as less labour' 1, p. 82.) However, it is not the total amount of labour expended on capital that counts. The relevant quantity depends on the durability of the equipment - its longevity - and, in the case of a durable implement, 'only a small portion of its value would be transfered to the commodity' (1, p. 23). Only one page later, the 'early stages' have been left far behind, and we find ourselves transported to 'a state of society in which . . . improvements have been made, and in which arts and commerce flourish' (1, p. 24). Ricardo pronounced: we shall still find that commodities vary in value conformably with this principle: in estimating the exchangeable value of stockings, for example, we shall find that their value, comparatively with other things, depends on

l6o

INTERPRETING RICARDO

the total quantity of labour necessary to manufacture them, and bring them to market. First, there is the labour necessary to cultivate the land on which the raw cotton is grown; secondly, the labour of conveying the cotton ... which includes a portion of the labour bestowed in building the ship in which it is conveyed ... thirdly, the labour of the spinner and weaver ... and of many others, whom it is unnecessary further to particularize. The aggregate sum of these various kinds of labour, determines the quantity of other things for which these stockings will exchange, while the same consideration of the various quantities of labour which have been bestowed on those other things, will equally govern the portion of them which will be given for the stockings, (i, p. 24—5) The proposition is that static exchange relationships are determined by comparative amounts of 'labour expended' at all vertical and horizontal stages of production processes. 'To convince ourselves that this is the real foundation of exchangeable value', the argument is then extended to changes in value resulting from 'abridging labour in any one of the various processes' in a commodity's production (1, p. 25). Thus, taking 'labour expended' to cover both 'direct' and 'indirect' labour (labour expended on capital), we have a reaffirmation of the 'pure' labour theory, now applied to an 'improved' society: (changes in) exchange relationships are still determined by (changes in) comparative amounts of labour expended. According to the analysis which follows, however, we do not find the 'pure' labour theory holding in such a society, perhaps suggesting that the above passage was a relic from the 'pretty mass of papers', written before the discovery of the 'curious effect'. Indeed, if the analysis is taken literally, the theory may not even apply to the 'early stages', to which Ricardo initially returned. The main conceptual distinction for the analysis is between fixed and circulating capital: According as capital is rapidly perishable, and requires to be frequently reproduced, or is of slow consumption, it is classed under the heads of circulating, or of fixed capital. (1, p. 52)18 Taking the cases of hunting and fishing in the 'early stages', it is assumed that: 'the bows and arrows of the hunter [are] of equal value, and of equal durability, with the canoe and implements of the fisherman, both being the produce of the same quantity of labour' (1, 18

Later in the Principles he added: 'It is difficult to define strictly, where the distinction between circulating and fixed capital begins; for there are almost infinite degrees in the durability of capital' (1, p. 150).

The labour theory of value (I)

161

p. 53); hunting and fishing require the same number of ('direct') labourers; and the 'production' cycle is the same length in both occupations. In short, an assumption of identical capital structures. Three propositions are claimed to follow. First, 'profits would be high or low, exactly in proportion as wages were low or high' (1, pp. 53—4). Secondly, changes in distribution 'could not in the least affect the relative value offish and game, as wages would be high or low at the same time in both occupations' (1, p. 54). And, thirdly, if 'there were any other commodity which was invariable in its value, requiring at all times, and under all circumstances, precisely the same quantity of labour to obtain it', it would be possible to isolate changes in the amount of labour required to obtain either fish or game, providing that their 'catch' or 'output' is the sole variable (1, p. 54): using the 'invariable' commodity as numeraire, changes in the prices of fish and game would reflect corresponding changes in labour expenditures per unit of output. The demonstration of the first two propositions was as much dependent on an 'invariable standard' as the third, but the specification of the standard as a commodity produced by an unchanging quantity of labour was insufficient for Ricardo's purposes. The standard's production process must itself exhibit the same capital structure as in hunting and fishing, for only then will the total monetary output of fish and game remain invariant to changes in distribution (as in Ricardo's calculations). Ricardo's awareness of the stronger requirement is confirmed by the following: If we had then an invariable standard, by which we could measure the variation in other commodities, we should find that the utmost limit to which they could permanently rise [in price], was proportioned to the additional quantity of labour required for their production; and that unless more labour were required for their production, they could not rise in any degree whatever. A rise of wages would not raise them in money value, nor relatively to any other commodities, the production of which required no additional quantity of labour, which employed the same proportion of fixed and circulating capital, and fixed capital of the same durability. (1, P-56)19 Then came the 'curious effect'; a product of divergent capital structures. Its first illustration rested on the following assumptions: the total monetary value of the capitals employed in hunting and 19

Cf. the paragraph beginning 'If the gold mine from which money was obtained were in the same country . . . ' (1, p. 55).

162

INTERPRETING RICARDO

fishing is the same; the hunter uses more fixed capital than the fisherman; the fixed capital is of equal durability in both occupations; and physical outputs are given and constant. Assuming a rise in money wages of 6 per cent, and a corresponding reduction in the profit rate from 10 to 4 per cent, it is shown that the monetary value of both game and fish 'output' will fall, implying lower unit prices, with game falling to the greater extent because of the larger fixed capital requirement in its 'production'. 20 These price reductions occur without any change in the amounts of labour expended. The calculations beg a question which remains unanswered until some six pages later in the text. Having concluded his extended analysis of the 'curious effect', Ricardo belatedly informed his readers: It should ... be carefully remembered, that in this whole argument I am supposing money to be of an invariable value; in other words, to be always the produce of the same quantity of unassisted labour. (1, p. 63, my emphasis) It is with reference to this 'money' commodity that Ricardo had calculated the lower rate of profit, used to obtain the new prices.21 Having considered the implications of a change in distribution when commodities are produced by unequal proportions of fixed and circulating capital, ceteris paribus, Ricardo turned to the case of differing durabilities of fixed capital, still using the 'unassisted labour' commodity as numeraire. Commodities are assumed to be made by a machine, unassisted by human labour, which is first deemed to last for one hundred years, then for ten years. Again, for a rise in wages and fall in the rate of profit, the commodities fall in price, to an extent depending on the size of the reduction in the profit rate and the durability of the machine. 22 The general conclusion was as follows: 20

21

22

T h e new profit rate is applied to calculate the monetary 'contribution' of fixed capital to output using the standard annuity formula. T h e a n n u a l 'contribution', A, is given from A = x [r(i + r)"/(i + r)" — i ] , where r denotes the rate of profit, n stands for the n u m b e r of years that fixed capital lasts, and x is the present monetary value of the fixed capital. For the 10 per cent rise in wages, it can be shown that the rate of profit on any commodity also produced by unassisted labour will fall to 3.77 per cent, which R i c a r d o rounded u p . Because such a commodity has the same capital structure as the standard's, the rise in wages cannot be passed on in a higher price. T o 'make sense' of Ricardo's calculations, it must also be supposed that both fixed capital, and wage-goods, are produced by unassisted labour, otherwise they, too, would be liable to a price change. These complications seem to have been ignored by Ricardo. U s i n g the a n n u i t y formula (above, n . 20) these results follow from the g r e a t e r ' c o m p o u n d i n g effect' of a given r e d u c t i o n in profitability w h e n fixed capital is m o r e d u r a b l e .

The labour theory of value (I)

163

It appears, then, that in proportion to the quantity and the durability of thefixedcapital employed in any kind of production, the relative prices of those commodities on which such capital is employed, will vary inversely as wages; they will fall as wages rise. It appears too that no commodities whatever are raised in absolute price, merely because wages rise; that they never rise unless additional labour be bestowed on them; but that all commodities in the production of which fixed capital enters, not only do not rise with a rise of wages, but absolutely fall; fall too as much as 68 per cent., with a rise of seven per cent, in wages, iffixedcapital be exclusively employed, and be of the duration of 100 years. (1, pp. 62-3)23 This is the 'curious effect'. Ricardo was evidently delighted with the 'effect', trumpeting it as being 'of such importance to the science of political economy, yet accord [ing] so little with some of its received doctrines, which maintain that every rise of wages is necessarily transferred to the price of commodities' (1, p. 61). It also implies that a 'pure' labour theory cannot hold when capital structures differ. But how significant were these differences and their effects supposed to be? Ricardo gives an answer at the end of the chapter: It appears then that the accumulation of capital, by occasioning different proportions of fixed and circulating capital to be employed in different trades, and by giving different degrees of durability to such fixed capital, introduces a considerable modification to the rule, which is of universal application in the early stages of society. (1, p. 66) And, seemingly to reinforce the point, he continued: Commodities, though they continue to rise and fall, in proportion as more or less labour is necessary to their production, are also affected in their relative value by a rise or fall of profits, since equal profits may be derived from goods which sell for 2,000/. and from those which sell for 10,000/; and consequently the variations of those profits, independently of any increased or diminished quantity of labour required for the goods in question, must affect their prices in different proportions. (1, p. 66) It is difficult, if not impossible, to reconcile these pronouncements with those earlier ones which suggested that (changes in) exchangeable value can be explained solely by reference to (changes in) comparative amounts of labour expended, even in a 'state of society 23

The statement that commodities 'never rise unless additional labour be bestowed on them' is patently false, since the effect of a fall in wages and rise in the profit rate would be to increase prices. The correct position is expressed by Ricardo in the last sentence of the chapter (i, p. 66).

164

INTERPRETING RIGARDO

. . . in which arts and commerce flourish' (see above, pp. 159-60). But, comparatively speaking, this is a minor problem. The major problem begins to surface towards the end of the chapter 'On Value', in a sequence of (misplaced) passages located between the detailed 'curious effect' analysis and the concluding paragraphs cited immediately above. (The odd positioning of the concluding paragraphs suggests that they may have been hasty and belated interpolations to the pre-'curious effect' draft.) Ricardo's objective was to 'distinguish between those variations [in price] which belong to the commodity itself, and those which are occasioned by a variation in the medium in which value is estimated, or price expressed', particularly as this bears on questions of distribution (1, p. 64). In the case of a general alteration in the exchangeable value of money, he observed that there is 'a general effect on price, and for that reason . . . no real effect whatever on profits'. He continued: On the contrary, a rise of wages, from the circumstance of the labourer being more liberally rewarded, or from a difficulty of procuring the necessaries on which wages are expended, does not produce the effect of raising price, but has a great effect in lowering profits. In the one case, no greater proportion of the annual labour of the country is devoted to the support of the labourers, in the other case, a larger portion is so devoted. (1, p.64).24 And then: It is according to the division of the whole produce of the land and labour of the country, between the three classes of landlords, capitalists, and labourers, that we are to judge of rent, profit, and wages, and not according to the value at which that produce may be estimated in a medium which is confessedly variable. (1, p. 64) In the first passage, changes in aggregate profits are related inversely to changes in the proportion of labour expended on the production of aggregate wages. In the second, it appears to be the material division of produce that supplies the criterion for judging (changes in) aggregate class incomes. But these are not, in fact, separate criteria. 24

Sraffa cited this passage in support of his interpretation that Ricardo determined the general rate of profit with references to social aggregates (i, p. xxxii). But, as Ricardo was to make clear, the general rate of profits is calculated with reference to the distribution of produce obtained by a given amount of labour on the last (portion of) land cultivated: a 'micro' determination. This point is substantiated in the following chapter.

The labour theory of value (I)

165

From the 'material' angle, the critical assumption is that the total material output is the produce of an unchanging amount oflabour, in which case a change in its proportional allocation signifies corresponding movements in the amount of labour expended on the production of the aggregate shares.25 The 'labour expended' criterion is the dominant one: It is not by the absolute quantity of produce obtained by either class, that we can correctly judge of the rate of profit, rent, and wages, but by the quantity oflabour required to obtain that produce. (1, p. 64)26 And further: 'Wages are to be estimated by their real value, viz. by the quantity oflabour and capital employed in producing them' (1, p. 65).27 But how is this quantity oflabour - the determinant of 'real value', an equivalent term to 'natural value' in the pre-Principles writings - to be ascertained? The answer given is by measuring (changes in) produce with an 'invariable standard' (1, p. 64). But which invariable standard? In the 'curious effect' analysis, something called an 'invariable standard' was produced by an unchanging amount of unassisted labour, and was apparently selected in order to maximise the purely distribution-induced price reductions (which, to emphasise, occur independently of changes in the amounts oflabour expended). So, if capital structures differ, as Ricardo has assured us they do, this standard - the only one specified in such a context - would not give a correct indication of changes in 'real value'. With a new wage-profit couplet there would be changes in shares, and the size of the total product, bearing no relationship to labour expenditures. To put this another way, the analysis of 'shares' is conducted with a 'pure' labour theory and yet, within the same chapter, Ricardo seems to have gone out of his way to underline that theory's shortcomings. This paradox is magnified when we turn to subsequent chapters, particularly those up to, and including, the seventh chapter 'On 23

26

27

T h e assumption is mentioned explicitly when R i c a r d o writes: ' W e might find . . . that though the absolute quantity of commodities h a d been doubled, they were the produce of precisely the former quantity o f l a b o u r ' (i, p. 64). T w o further assumptions underlying his numerical explication a r e that a change in labour productivity affects all commodities equally; a n d that all classes consume the same commodities in fixed proportions. H e can then discuss the changing distribution of 'every h u n d r e d hats, coats a n d quarters of corn produced' (1, p. 64). T h e reference to 'the rate of profit, rent, a n d wages' is misleading, because at this point Ricardo is theorising about (movements in) aggregate shares. Only subsequently does he turn his attention to the 'rate of profits' (1, p. 65, emphasis in original). Recall that 'capital' is also reduced to a quantity oflabour expenditure.

l66

INTERPRETING RIGARDO

Foreign Trade'. For the arresting feature is that the effect on (changes in) natural prices of anything other than altered labour expenditures is entirely absent from these chapters and is, indeed, explicitly denied: 'every diminution in the wages of labour raises profits, but produces no effect on the price of commodities' (i, p. 133). In view of what had gone before, one might think that this use of the 'pure' labour theory called for some justification. One possibility, according with George Stigler's interpretation, is that Ricardo attributed only trifling empirical significance to the distribution-induced price movements, and therefore thought it reasonable to abstract from them. Some years later, he indeed professed a long-standing opinion 'that in the relative variation of commodities, any other case, but that of the quantity of labour required for production, was comparatively of very slight effect' (11, p. 59). Yet the fact is that an explicit 'empirical' claim on behalf of the labour theory cannot be found in the first edition of the Principles. Not only that, but the 'curious effect' analysis, and the closing paragraphs to 'On Value', suggest a truly significant 'modification' to the labour theory. If an 'empirical'justification had been floating in Ricardo's mind, it was not communicated to the readers of the first (or second) edition of his Principles. A second, logically separable possibility, is that he had assumed identical capital structures throughout the commodity-producing system. S. Hollander has claimed that such an assumption was indeed made and, in his favour, the suggestion does agree with the logic of Ricardo's position: if a 'pure' labour theory is to hold then, as he had shown, capital structures must be everywhere the same. But this interpretation, too, is starved of explicit textual support.28 Ricardo had made no visible attempt to reconcile his use of the 'pure' labour theory with his own analysis of divergent capital structures. His only explicit 'justification' for claiming an exclusive relationship between changes in natural prices and labour expenditures (a relationship which implies a 'pure' labour theory) was, in 28

Professor Hollander cites two pieces of indirect supporting 'evidence'. The first is a passage from Ricardo's letter to Mill of 14 October 1816: 'I have been beyond measure puzzled to find out the law of price. I found on a reference to figures that my former opinion could not be correct and I was full a fortnight pondering on my difficulty before I knew how to solve it' (vn, pp. 73-4). The 'solution', according to Professor Hollander, was the assumption of identical capital structures (1979, p. 202 n. 30). But, if this is true, why should Ricardo have hidden such a critical assumption from his readers? Professor Hollander's second item, a passage from chapter 'On the Rent of Mines' (1, pp. 85-6), is remarkable for saying nothing whatever about capital structure.

The labour theory of value (I)

167

fact, the assumption of an 'invariable standard', defined without any reference to capital structure. Thus: Having acknowledged the imperfections to which money made of gold and silver is liable as a measure of value, from the greater or less quantity of labour which may, under varying circumstances, be necessary for the production of those metals, we may be permitted to make the supposition that all these imperfections were removed, and that equal quantities of labour could at all times obtain .. . equal quantities of gold. Gold would then be an invariable measure of value. (1, p. 87 n. 1) And having assumed this invariable measure, Ricardo continued: In speaking therefore of varying price, the variation will be always considered as being in the commodity, and never in the medium in which it is estimated (1, p. 87), where the Variation' is wrought exclusively by altered labour expenditures (see 1, pp. 85-6). The same position is expressed at the beginning of the chapter 'On Profits': 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, and therefore every variation of price to be referable to an alteration in the value of commodity. (1, p. 1 ion.) The analysis which follows confirms the unique relationship between 'alterations in value' and changes in labour expenditure. Turning to the ninth and following chapters, there are a couple of references to the 'curious effect' analysis (1, pp. 207, 239). But for the most part, except where it is explicitly vitiated by foreign trade or taxation, a 'pure' labour theory is presumed to hold with the 'invariable standard' again defined without reference to capital structure: That commodity is alone invariable, which at all times requires the same sacrifice of toil and labour to produce it. Of such a commodity we have no knowledge, but we may hypothetically argue and speak about it, as if we had; and may improve our knowledge of the science, by shewing distinctly the absolute inapplicability of all the standards which have been hitherto adopted. (1, p. 275) Once more, the implication is that, with this hypothetical commodity as the standard, changes in the (natural) prices of other commodities exclusively reflect changes in the labour expended on their production.

l68

INTERPRETATING RICARDO

The material reviewed in the last section may provide an explanation for what had happened. Prior to considering the implications of divergent capital structures, Ricardo had embraced the 'pure' labour theory in terms of which the assumption of an unchanging labour input for the standard is suflficient to yield the comprehensive link between movements in natural prices and labour expenditures. It was on this level of abstraction, or ignorance, that he drafted his first seven chapters (or chapters roughly comparable in scope). Subsequently, following his development of the categories 'fixed' and 'circulating' capital, and his allowance for unequal 'capital structures' (which initially brought him grief), he fashioned the 'curious effect' analysis, which was interpolated in the first draft chapter. At this stage he apparently realised, if only with reference to the first chapter, that further revisions were necessary to reconcile the 'curious effect' material with his earlier analysis (see above, pp. 152-3). But, on Mill's advice, he decided to press ahead with the rest of the book, leaving it until the very last to put his manuscript 'in some tolerable order' (letter to Malthus, 3 January I 8 I 7 , V I I , p. 115). That could have been his greatest mistake. For if he had perceived a solution to the problem of reconciling his use of the 'pure' labour theory with the 'curious effect' material, and on this there is no clear, substantive evidence, it did not find its way into the published manuscript. The final revision process, doubtless undertaken with Mill's 'advice', did nothing to resolve the serious textual inconsistencies. There is another level on which Ricardo sowed confusion with his 'value' analysis, stemming from his erratic and unclarified use of terms. This is considered in the following sub-section, together with the question of the labour theory's 'status'. Value, price and cost of production: further issues

Attention has been drawn to several instances of Ricardo using 'value' in a way signifying an 'absolute' value: something attributable to a commodity regardless of its relative or exchangeable value, such that it is possible to speak of the commodity as being 'of an unvarying value' (1, p. 17 n. 3) even though it may exchange for more or less of other things. This type of'value', once referred to as 'natural value' in correspondence before the Principles, was associated with the amount of labour expended on a commodity and, in the chapter 'On Value', it was rechristened real value (1, p. 65).

The labour theory of value (I)

169

That term, together with cognate formulations, is also found elsewhere in the Principles. In the chapter 'On Foreign Trade', there is a reference to a rise in 'the real or labour price of commodities', generated by 'the additional quantity of labour required to convey them to the various markets where they were to be sold' (1, p. 136); in the chapter 'On Bounties on Production', Ricardo wrote of corn being 'lowered in price by an alteration in its labour price' and, a few lines later, of the 'real fall in the value of corn, arising from less labour being required to produce [it]' (1, p. 323); and in the chapter 'Mr. Malthus's Opinions on Rent', he agreed with Malthus that if the 'real price of commodities, have any meaning . . . it must be measured by the proportionate quantity of capital and labour necessary to produce them' (1, pp. 416-17). Finally, without using 'real value' or equivalent terms, Ricardo wrote in the chapter 'Value and Riches, Their Distinctive Properties' that the 'labour of a million of men in manufactures, will always produce the same value' (1, p. 273), again suggesting an 'absolute' concept of'value', dependent on productive labour expenditure. As Ricardo's contemporaries fairly complained, it was confusing to invest 'value' with a double meaning, sometimes using it to denote exchangeable value, at others 'real value', and the confusion was compounded by the use of additional but contextually synonymous terms. Nor does the possible confusion end here, as may be seen by posing the question, 'What regulates (changes in) natural prices?', and considering Ricardo's answers. One set of his responses postulates the relationship between (changes in) relative natural prices and (changes in) labour expenditures or, equivalently, (changes in) 'facility or difficulty of production' (1, pp. 193, 217, 273, 303-4, 309, 417). But there is another prominent set according to which (changes in) natural prices are regulated by (changes in) costs of production (1, pp. 73n., 156, 164, 364, 382). Can these positions be reconciled? At one point Ricardo explained that cost of production includes 'the usual and ordinary profits of stock' as well as wages (1, pp.405—6). With his utterances made against the backdrop of a 'pure' labour theory, complete with an 'invariable standard', his reasoning may be reconstructed as follows. If there are single natural rates for wages and profit, which Ricardo mostly took for granted, (changes in) costs of production per unit of output, measured by the standard, will be uniformly and exclusively proportional to (changes

170

INTERPRETING RIGARDO

in) wage outlays, as reflecting (changes in) labour expenditures: if the labour expended on a commodity increases so, too, does the wage outlay per unit of output. The proposition that (changes in) natural prices are regulated by (changes in) costs of production is therefore equivalent to one with (changes in) quantities of labour expenditure (or substitute terms) as the 'regulatory' influence. But this equivalence is only true under 'pure' labour theory conditions. It is worth teasing out a feature of Ricardo's reasoning. 'Quantity of labour' is measured in terms of labour time — so many labouring hours or weeks - but comparative labour times of production will be reflected in relative wage bills, providing that a 'pure' labour theory holds. However, although relative wage bills reflect relative labour expenditures, they are not, strictly speaking, the same as those expenditures. Yet in the later writings, particularly, there are many instances of Ricardo identifying, or conflating, quantities of labour with wages, at the expense of some horrendous confusion. Now consider the following statement: 'Corn can be permanently at an advanced price, only because additional labour is necessary to produce it; because its cost of production is increased' (1, p. 335). This may be understood as follows. Measured by the standard, the cost of producing a unit of corn can increase only if more labour is expended and, reflexively, if wage costs per unit of output increase. Profits, in contrast, cannot influence (changes in) natural price, for within a 'pure' labour theory setting, any change in the natural profit rate affects all commodities equally, including the standard, and therefore the exchange relationship between the standard and each commodity is unchanged. Also consider this: By allowing the free importation of corn, or by improvements in agriculture, raw produce would fall; but the price of no other commodity would be affected, except in proportion to the fall in the real value, or cost of production, of the raw produce, which entered into its composition. (1, The reasoning is again explicable. The 'real value' of the raw produce is dependent on the labour expended on its production. But, if this falls, so too does the cost of producing corn as measured by the standard. Indeed, the cost of production falls if and only if 'real value' is reduced. However, the passage does lend itself to a

The labour theory of value (I)

171

misinterpretation which has more widespread roots in Ricardo's shifting locutions, for it may seem as though he has used 'real value' and 'cost of production' as alternative descriptions of the same thing; and this impression is apparently confirmed by his incessant switching between (changes in) 'cost of production' and (changes in) 'quantity of labour' [etcetera) as the determinants of (changes in) natural prices. This left him open to the charge, later pressed by Malthus, that he excluded profit from 'cost of production' (see below, p. 205). But, given a 'pure' labour theory, the accusation is groundless: 'changes in natural prices are determined by changes in labour expenditures' and 'changes in natural prices are determined by changes in costs of production' are interchangeable statements in a 'pure' labour theory context, but this does not imply that 'cost of production' is the same as 'labour expenditure', only that they are related to each other. Finally, I turn to the question of the labour theory's status. On the basis of the material reviewed in the last section, it seems safe to conclude that Ricardo's initial embrace and development of the theory was uninformed by purely speculative concern with the 'substance' or 'essence' of value. Indeed, his attitude towards the theory was rather narrowly instrumental. For him, a 'satisfactory' theory was, I suggest, one which illuminated the connection between agricultural conditions of production and 'permanent' movements in profitbility in a way that was (or rather, seemed to be) logically coherent. In the Principles, however, there are references to labour as the 'foundation of all value' (1, pp. 20, 13, 88), which may sound rather profound. But Ricardo apparently meant little more than the purely analytical proposition that (changes in) exchange relationships are governed by (changes in) labour expenditures. The fact that he sometimes used 'value' in an 'absolute' sense was not, at this time, worked up into a disquisition on the essential meaning of 'value'. As to a possible 'empirical' gloss on the labour theory, he certainly failed to make a strong claim along such lines. Indeed, thoughtlessly perhaps, he made quite a strong case against the labour theory's empirical relevance. What can be said is that the theory clearly possessed analytical appeal for Ricardo, but beyond this we cannot presently go.

172

INTERPRETING RIGARDO FROM THE FIRST TO THE SECOND EDITION OF THE PRINCIPLES

Correspondence

By March, 1817, the Principles was in press. Ricardo confided in Malthus: In their printed form they [the manuscript papers] appear worse, in my eyes, than before, and I need all the encouragement of my partial correctors to keep alive a spark of hope respecting their reception ... As yet I have no misgivings about the doctrines themselves, all my fears are for the language and arrangement and above all that I may not have succeeded in clearly shewing what the opinions are which I am desirous of submitting to fair investigation. (9 March 1817, vn, p. 140) A fortnight later his fears had intensified: 'as the time approaches that I am to appear in print, I seem to become more dissatisfied with my work, and less capable to give any proposition contained in it a patient investigation' (letter to Malthus, 22 March 1817, vn, p. 143). Four months afterwards, refreshed by a six week tour of continental Europe, he received Malthus's verdict: I have read over your book again with much gratification. There is much collateral matter in which I quite agree with you. I also quite agree with you that the difficulty of procuring subsistence is the necessarily limiting cause with regard to profits, but I still cannot agree with you that labour alone in the sense you understand it is either in theory or fact the best measure of exchangeable value ... Pray do you allow that in different countries, where profits are different, your theory of value does not hold good. I dont feel quite sure. (17 August 1817, vn, p. 176) Ricardo replied: You flatter me very much by your second perusal of my book, and I am happy to find that there are but a very few important points on which we materially differ. I certainly allow that my theory of value does not hold good in different countries when profits are different. (4 September 1817, VII, p. 186) With the 'few very important points' including the labour theory of value, the very foundation for Ricardo's 'mature' analytical system, a less self-satisfied response might have been expected.29 On the 29

Malthus had also expressed disagreement with the position 'that the state of the land practically determines the existing rate of profits in different countries' (vn, p. 176). This was passed over by Ricardo.

The labour theory of value (I)

173

other hand, Ricardo's sensitivity to his perceived defects may have led him to expect the worst from his critics, to the point where even heavily qualified praise came as something of a relief. Two months later, a scathing assessment of the Principles appeared in the November 1817 edition of the British Review?0 The recurrent criticism was one of over abstraction, and the sharpness of the review is well captured by the anonymous writer's judgement that the Principles contained 'much more of extravagant paradox and learned absurdity than we have encountered in any similar publication since the commencement of our critical career'. Ricardo could take this criticism in his stride. He wrote to his friend Trower: In the last number of the British there is ... a review of my book, in which in every page I am charged with ignorance and absurdity. Yet it is not done in an ungentlemanly way, and I have the pleasure to have my friend Malthus associated with me in the censure of having made the subject of rent ... obscure and unintelligible. The writer does not in fact see the important part of the subject - he has read but not studied it. He has kindly left unattacked those points which were most assailable, and has fastened on those which are incontrovertible. My style and arrangement are fortunately for me not mentioned. (10 December 1817, vn, p. 219)31 Equally 'fortunate', the problems with the substance of his work, especially those relating to 'value', had passed unnoticed. If only glancingly, however, the reviewer had struck a blow. Ricardo longed for 'an able pen on my side to put my opinions in a clear light, and to divest them of that appearance of paradox which they now wear' (p. 222). And on the 'style and arrangement' theme, he continued to lament his inadequacies.32 But these self-criticisms were only directed against the presentation of his ideas, for, as he wrote to Trower, he believed he had 'a very consistent theory in my own mind' (26 January 1818, vn, p. 246). All he wanted was an 'able pen' and, in James McCulloch's, he thought he had found it. McCulloch had published (anonymously) two short pieces on the Principles in the Scotsman newspaper, but it was his article in the prestigious Edinburgh Review that gave Ricardo the most pleasure: T 30

31

32

Article X V , 'Political Economy a n d T a x a t i o n ' , p p . 309-33. T h e article reviewed both Ricardo's Principles a n d the fourth edition of J . B. Say's Traite d'Economie Politique. H e commented to Malthus that the reviewer h a d evidently not 'understood what I meant to say' (16 December 1817, v n , p. 222). See the letters to Trower, 26 January 1818, vn, p. 246; to Mill, 29 September 1818, vn, p. 305; to Murray, 18 November 1818, vn, pp. 328-9; and to Mill, 23 November 1818, vn, P-333-

174

INTERPRETING RICARDO

cannot but feel highly gratified at [McCulloch's] praise', he wrote to Malthus (20 August 1818, vn, p. 282). And to McCulloch, he sent this bouquet: 'I know not whether I ought to thank you, but I have been exceedingly gratified. My own doctrines appear doubly convincing as explained by your able pen' (22 August 1818, vn, p. 286). Perhaps Ricardo had been blinded by praise, for on the exposition of 'value', at least, the hand of an 'able pen' was not conspicuously apparent in McCulloch's review. McCulloch declared, 'It is the cost of production, which is the permanent and ultimate regulator of the exchangeable value of every commodity' (1818, p. 61, emphasis in original), cost of production being described as a commodity's 'real price' (pp.61, 80). But he also stated that it is 'quantity of labour . . . that determines the real price or the worth in exchange of every commodity' (p. 66, cf. p. 68). From the last sub-section we know that Ricardo's use of terms was such that he could switch between invoking 'cost of production' and 'quantity of labour' (or 'real value', or 'real price') as determinants of (changes in) exchange value, providing that a 'pure' labour theory holds. This (admittedly implicit) proviso seems to have been lost on McCulloch. As a prelude to his summary of the 'curious effect' analysis, he wrote: Although the exchangeable value of a commodity, or its real price, is in no case whatever increased by an increase in the rate of wages, it may, and in very many cases actually is, thereby reduced in its real price, (p. 70, emphasis in original) But if'real price' is determined by quantity of labour, how can it be reduced if labour expenditures are unchanged? Ricardo initially failed to notice any problems with McCulloch's exposition, claiming that his laudatory reviewer 'has so completely understood me' (letter to Sharp, 27 August 1818, vn, p. 291). Certainly, this was not a period remarkable for intellectual endeavour on Ricardo's part, at least with the finer points of theory. In the time that elapsed between the first and second editions of Principles, he was mostly content to bask in the praise bestowed on his work and rejoice in his 'converts'. 33 Hutches Trower, in contrast, had read 33

The 'converts' proudly referred to in correspondence were Robert Torrens (vn, p. 180), Francis Place (vn, p. 189), and Lord Grenville (vn, pp.220, 259). James Mill and McCulloch should also be included, as well as Ricardo's loyal but independently minded friend, Hutches Trower, to whom he wrote: 'The reward which I have received for my labours has far exceeded my expectations' (2 November 1818, vn, p. 319).

The labour theory of value (I)

175

McCulloch's review (thinking, incidentally, that it had been written by Robert Torrens) with 'great attention, and still greater interest'. He ventured the following observation: [The writer] appears to me, in one part of his Review, rather to puzzle himself in endeavoring to explain the difficult question of exchangeable value, and price . . . [He] says, if the labor of production should be encreased, equally on all articles, 'their exchangeable value would remain unaltered, while their real price would however be augmented,' and lower down,

in the same page, he says - 'In such circumstances although the prices of commodities would remain stationary, the wealth and comfort of the whole

society would be diminished.' This surely is a contradiction. In the next page, in endeavoring to shew, that a rise in wages would not affect the prices of commodities ['real prices', according to McCulloch], I think he shows only, that it would not affect their relative values. (23 August 1818, vn, p. 288, emphasis in original) 34 The problems with McCulloch's review were by no means confined to a few pages, but at least Trower had raised some important terminological points. In his reply to Trower, Ricardo admitted that he had not previously noticed 'the inaccuracy of the reviewer'. He continued: In . . . the first quotation you make . . . [McCulloch] used the word price instead of the word value; substitute the latter word and the whole is consistent, though perhaps not quite satisfactory, for it supposes my definition of value to be correct, which may by many be disputed. In the next page he again speaks of real price as synonymous with real value, but his meaning is obvious. The word price I think should be confined wholly to the value of commodities estimated in money, and money only. If so confined, a commodity may rise in real value without rising in price. If more labour should be required than before to work the mines, and to manufacture shoes, it is possible that shoes may continue unaltered in price, but both the shoes, and gold (or money) will have risen in value. (18 September 1818, vn, p. 297, emphasis in original)

Ricardo's liberal use of emphasis is uncharacteristic and, in the light of a previous exchange of letters with Trower, he was probably aspiring, unsuccessfully, to a more precise use of language. 35 It is 34

Trower's quotations, with the emphasis supplied by h i m , are from McCulloch (1818),

35

T r o w e r h a d written to R i c a r d o : ' I f there b e a n y t h i n g , which I should wish d i f f e r e n t . . . [in the Principles]^ it would b e t h e m a k i n g a greater point of defining accurately a n d rigidly, t h e terms employed — T h e Equivocality of l a n g u a g e is t h e p r e g n a n t source of the endless contests that arise upon all scientific and abstruse subjects; and has been especially so in all the discussions upon political economy . . . I would begin with a copious chapter of clear and concise

pp. 68-9.

176

INTERPRETING RIGARDO

fairly clear that he was again using 'real value' in an absolute sense, to denote quantity of labour expenditure, as he had done in the Principles. But, misleadingly, 'real value' and the abbreviated 'value' are also used as synonyms. If 'value' had been used in the sense of exchangeable value, shoes and gold would be of an unchanged 'value', at least with respect to each other. He had cast both light and shade.36 The main correspondence on the theory of value between the first two editions of the Principles is remarkable for the lack of direct attention paid by Ricardo (and others) to the major unresolved problems, detailed in my last section. I next consider the extant fragments of a more substantive debate on 'value' with Robert Torrens, also held between the first two editions, which resulted in Ricardo's acknowledgement of a third basis for differences in capital structure in the second edition of the Principles and, later, to his adoption of a 'dated labour' analysis. The debate with Torrens

In his letter to Mai thus of 24 November 1818, Ricardo alluded to 'a long conversation' with Robert Torrens on 'value' (vn, p. 338). As Sraffa noted (p. 338 n. 2), the conversation probably took place in February 1818, and their debate was also conducted in writing, fragments of which remain.37 The assumptions underlying Torrens's argument were as foll-

36

37

definitions' (28 February 1818, VII, pp. 256-7, Trower's emphasis). This was astute advice, to which R i c a r d o replied: 'Your suggestion of a copious chapter of clear a n d concise definitions would be of great use, b u t it requires a degree of precision a n d accuracy beyond what I could furnish' (22 M a r c h 1818, v n , p . 259). I t also required the time-consuming intellectual commitment which R i c a r d o seemed u n p r e p a r e d to make. It m a y also be noted that R i c a r d o h a d himself used 'price' in contexts where 'real value' would have been more appropriate (according to this latest statement) (1, p p . 136, 323, 416). Mai thus, in reply to a letter from Ricardo which has not been found, referred to Torrens's 'peculiar objections to your [Ricardo's] measure of value' in his letter of 24 February 1818. Malthus continued: 'For myself, . . . I am quite satisfied with your own concessions; and if as you yourself acknowledge, taxation, foreign materials, and the different quantities of fixed and circulating capital employed, all prevent the exchangeable value of commodities from being determined by the labour which they have cost in production, I should say, it followed that your theory was only true caeteris paribus, which might be equally said of the cost of the materials. Given the wages of labour, the profits of stock, and the taxes, the exchangeable value of commodities will vary with the cost of the materials' (vn, p. 253). It was only later that Malthus directly sought to expose the contradictions in Ricardo's Principles.

The labour theory of value (I)

177

ows.38 There are two producers, one manufacturing cambric and the other linen. Each manufacturing process requires the same raw material, flax, and direct labour, although in different proportions. The flax is zto/^ produced by unassisted labour. The rate of profit is set at 100% per discrete production period. And the subsistence of one labourer for one day serves as numeraire. Each production process receives a detailed specification. The cambric manufacturer uses flax of the value often days' subsistence and employs ninety day-labourers, while the linen manufacturer uses flax of the value of eighty days' subsistence and employs twenty day-labourers. The 'value' of both their capital outlays is therefore 100 days' worth of subsistence. Consequently, if these capitals are marked up by 100 per cent, the value of cambric and linen output will also be equal. However, these outputs are not the result of the same quantities of labour expenditure. In the case of cambric, ninety labouring days are spent working up the flax, but that raw material was also produced. To calculate the labour time expended on its production, it is only necessary to halve its 'labour commanded' value: ten days' worth therefore required the effort of five daylabourers. (This follows from the assumptions of a 100 per cent profit rate and the production of flax by unassisted labour.) Totalling, cambric is the product of ninety-five day-labourers whereas linen is the product of sixty (twenty 'direct' labouring-days plus forty for the flax).39 This analysis draws attention to a species of modification to the 'pure' labour theory which had not been explicitly dealt with in the first edition of the Principles, although its ingredients were present. Ricardo had mentioned the production of raw material (or, the same thing, the expenditure of labour on its cultivation and transport), and also the possibility of differences in the proportions of raw material to 'direct' labour between production processes (respectively, 1, pp. 24—5, 171). Torrens's case was a combination of these circumstances: one of unequal combinations of raw material and 'direct' labour, ceteris paribus. Ricardo evidently thought he must translate this case into his categories of fixed and circulating capital. His first suggestion was that 'the raw material in both manufac38

39

In this and the following paragraph, I give Ricardo's summary of Torrens's case. Both versions are effectively the same. In calculating the total labour expenditures, Torrens wanted to include the labour required to produce wage-goods. Ricardo objected to this and, for the sake of argument, Torrens conceded the point (iv, pp. 314-15).

178

INTERPRETING RIGARDO

tures is really fixed capital of equal duration, but of unequal value. That employed in the production of linen being 8 times more valuable, than that employed in the production of the cambric' (iv, p. 311). He could therefore 'absorb' Torrens's 'modification' as a case of different proportions of fixed to circulating capital. Then he wrote: It appears then that every thing is fixed capital which is employed on production except that which resolves itself into wages, (iv, p. 312) But this paragraph was deleted, not surprisingly in view of what followed. Given the manner in which the distinction between fixed and circulating capital had been applied in the Principles, Torrens's, raw material would actually count as circulating capital because, in Ricardo's various illustrations, 'fixed capital' always appears as items of equipment which remain after the production of the final output (for example, 1, pp. 52—3, 59). Raw material would not fall into this category. Ricardo may have recognised this. Departing from Torrens's case, he postulated and contrasted two production processes. One has five labourers employed for one year making iron which is then sold with a profit. All the proceeds are reinvested in production on an expanded scale, and this annual scenario is repeated for twenty years. The other production process also has five men employed for a year, this time in planting acorns. No more labour is employed but, at the end of twenty years, the wood from the oak trees will sell for exactly the same as the iron produced in the twentieth year. (In both cases, output in the twentieth year, expressed in some numeraire, is L.w.{\ + r)2°.) Ricardo commented: Here then are two commodities of equal value one of which [iron] is the production of more than 5 times the quantity of labour employed on the production of the other, (iv, p. 312) This 'strong case' (iv, p. 312) formed the background to the next stage of the analysis. Two more manufacturers entered the analytical frame: the cutler, whose raw material is iron, and the upholsterer, who uses wood. The assumption is that they each spend the same amount on these materials and on 'all other parts of [their] circulating capital' (now suggesting that the produced raw material does count as circulating capital) (iv, p. 313). It follows that:

The labour theory of value (I)

179

the steel made by the cutler in one year, and the furniture made by the upholsterer in the same time would sell for the same money, but as the iron employed would contain much more labour than the wood, the finished commodity in iron would ... represent more labour than the finished commodity in wood, (iv, p. 313) And then, with the aid of a carpenter, Ricardo returned to Torrens's puzzle: In the same way it might be proved that a carpenter and upholsterer employing the same amount of capitals but an unequal value of wood as their raw material, though they would sell their finished goods for the same money, yet these goods would represent different quantities of labour. This last is Major Torrens case and is in my opinion completely answered by shewing that in the production of the raw material different quantities of fixed capital were used, (iv, p. 313) Here, it is no longer the materials themselves which are fixed capital, but rather the 'inchoate' wood viewed from within its own production process: it was partly 'produced' using itself as an input. However, from the standpoint of carpentry or upholstery, the final stage of wood-using productive activity, wood is comparatively a circulating capital in the sense that it is all used up in the final manufacturing stage. To that extent, the classification depends on a temporal perspective. For Ricardo's theory of value, however, the more relevant orientation is the 'backward looking' one, which Ricardo eventually realised himself when, between the second and third editions of the Principles, he formulated his version of a 'dated labour profile' analysis, discussed in the following chapter. One feature of Ricardo's response to Torrens may be emphasised by way of a conclusion to this episode. As with his demonstration of the 'curious effect', Ricardo's formulation of the 'acorns into wood' example is remarkable for undermining the 'pure' labour theory. This very example would return to haunt Ricardo in years to come but, at this time, he remained oblivious to the difficulties he had created for himself. The second edition

When Ricardo came to prepare the Principles for a second edition he had 'not looked at [it] . . . since it was published' (to Murray, 18

l8o

INTERPRETING RICARDO

November 1818, vn, p. 328).40 He read it twice, within the space of a Tew days', and declared himself well pleased with what he found (to Mill, 23 November 1818, vn, p. 333-4). 'With such an unskilful hand' (vn, p. 333), he thought it would be unwise to make major changes, but, in the hope of clarifying the argument, he did effect various alterations to his chapter 'On Value'. They did little to resolve the internal contradictions. One hurried innovation, with Mill's blessing, was the division of the chapter into sections.41 The first section heading read: The value of a commodity, or the quantity of any other commodity for which it will exchange, depends on the relative quantity of labour which is necessary for its production, and not on the greater or less compensation which is paidfor that labour. (1,

p. 11, emphasis in original) The essentially unchanged text for the section was concluded with the discussion of'different qualities of labour'. The second section, headed ' The accumulation of capital makes no difference in the principle stated in the last section' (1, p. 22, emphasis in

original), may be viewed as containing three phases, the first wherein Ricardo explained the necessity to calculate 'labour expended' at all stages of a commodity's production and marketing. It also included the paragraphs suggesting that the 'pure' labour theory applies in the 'improved' society (see above, pp. 159-60). The second phase commenced with the distinction between fixed and circulating capital, to which a footnote was now attached: 'A division not essential, and in which the line of demarcation cannot be accurately drawn' (1, p. 52n.). The point about 'demarcation' had been made in the previous edition some seven chapters later, and it was undoubtedly felicitous to bring it forward. Soon after the footnote, the following passage was inserted: It is also to be observed that the circulating capital may circulate, or be returned to its employer, in very unequal times. The wheat bought by a farmer to sow is comparatively afixedcapital to the wheat purchased by a 40

41

Ricardo presumably meant that he h a d not read it from cover to cover, since h e h a d certainly referred to various passages in correspondence. See his letters to Malthus (4 September 1817, v n , p . 186) a n d to Barton (20 M a y 1820, v n , p. 155-9). R i c a r d o was so lacking in self-confidence that he instructed his publisher, J o h n M u r r a y , to send a courier to Mill's house with the chapter headings: 'Your messenger c a n stop while [Mill] reads them and, he will return them to him with any alteration he may s u g g e s t . . . or he will advise their suppression altogether' (23 November 1818, v n , p. 331). I n the event, Mill approved of the headings, although he m a d e 'a slight alteration, in two (I think) of the expressions' (4 December 1818, v n , p p . 348-9).

The labour theory of value (I)

181

baker to make it into loaves. One leaves it in the ground, and can obtain no return for a year; the other can get it ground intoflour,sell it as bread to his customers, and have his capital free to renew the same, or commence any other employment in a week, (i, p. 53 n. i) 42 But how can circulating capital be 'comparatively a fixed capital'? The answer, deriving from Ricardo's reaction to Torrens's puzzle, is that the categorisation depends on temporal orientation. On a forward-looking criterion, both the farmer's and the baker's wheat is circulating capital because wheat is completely used up at the end of the production periods. But on the backward-looking criterion, more relevant for valuation considerations, the farmer's wheat has been locked into the production process for longer and, in that sense, it is 'comparatively a fixed capital'. The third phase of a second section must now be read as abstracting from possible differences in the 'durability of circulating capital'. It contained the 'fish and game' analysis, conducted on the assumption of identical capital structures, and was newly concluded thus: It appears ... that notwithstanding the accumulation of capital, commodities would not necessarily vary in relative value from a rise in wages, unless it was accompanied by increased facility or difficulty in the production of one or more of them. (1, p. 56 n. 1) Ricardo might well have emphasised 'necessarily' in this passage, since he was only making the logical point that the existence of 'capital' does not affect the labour theory if 'capital structures' are identical. With the above in mind, the third section heading is less jarring: ' The principle stated in the foregoing section considerably modified by the employment of machinery as fixed capital' (1, p. 56 n. 1, emphasis in origi-

nal). In fact, machinery as such does not feature at all in this section, which encompasses the elaboration of the 'fish and game' analysis on the assumption of differences in the ratios of fixed capital (bows, arrows, etcetera) to circulating capital. Given the new divisions, the underlying assumption of a standard produced by unassisted labour 42

This paragraph replaced the following: 'Besides the alteration in the relative value of commodities, occasioned by more or less labour being required to produce them, they are also subject to fluctuations from a rise of wages, and consequent fall of profits, if the fixed capitals employed be either of unequal value, or of unequal duration' (i, p. 53). This paragraph would have been out of place in a section proclaiming that the accumulation of capital makes 'no difference' to the 'pure' labour theory.

l82

INTERPRETING RIGARDO

is now to be found in the fifth section.43 The section ends with a new passage, partly incorporating the first of the three misplaced paragraphs which concluded the first edition chapter: It appears then that the division of capital into different proportions of fixed and circulating capital, employed in different trades, introduces a considerable modification to the rule, which is of universal application in the early stages of society, namely, that commodities never vary in value, unless a greater or less quantity of labour be bestowed on their production, it being shewn in this section that without any variation in the quantity of labour, the rise of its value merely will occasion a fall in the exchangeable value of those goods, in the production of which fixed capital is employed; the larger the amount of fixed capital, the greater will be the fall, (i, p. 58 n. 1) If anything, this basis for a 'considerable modification' to the 'pure' labour theory is articulated with even greater clarity than it had been in the first edition. Section four is headed: The principle that value does not vary with the rise or fall of wages, modified also by the unequal durability of capital, and by the unequal rapidity with which it is returned

to its employer. (1, p. 58 n. 1, emphasis in original) This section includes the 'unassisted machinery' analysis, now footnoted with a thrust at the writer of the hostile British Review article. 44 And, also newly introduced, Ricardo added: The same result [the 'curious effect'] will take place if the circulating capitals be of unequal durability. If from the nature of two different trades, in which equal capitals are employed, one manufacturer could not bring the commodity he produced to market in less than one year, while the other could bring his there in three months, the commodity of the first would fall in relative value to the second with every rise of wages and fall of profits. It must be unnecessary to go into further calculations to prove this to be true, as it rests precisely on the same principle as the case already considered, namely, the different degree of durability of two equal capitals. (1, p. 61 n. 1) 43

44

A further a n o m a l y is created in this section by R i c a r d o ' s insertion of the passage b e g i n n i n g 'Besides the alteration . . . ' , w h i c h referred to alterations in relative value associated with fixed capitals of differing d u r a b i l i t y (quoted above, p . 181 n. 42.). T h i s source o f ' m o d i f i cation' is not dealt with until the fourth section. ' T o p u t the principle in a strong point of view, I h a v e supposed a m a c h i n e to d o work w i t h o u t any assistance from h u m a n l a b o u r , which is evidently impossible. A writer in the British Review has a b s u r d l y a r g u e d as if this supposition was essential to the t r u t h of the principle. But it is obvious that similar results, t h o u g h not equal in degree, will take place w h e n b o t h m a n u f a c t u r e r s employ l a b o u r , a n d m a c h i n e r y or o t h e r capital, if the latter be of u n e q u a l d u r a b i l i t y . ' (1, p . 60 n. 1).

The labour theory of value (I)

183

Ricardo's intention to introduce the above passage had been communicated to McCulloch: I understand that Major Torrens has written an article in the Edinburgh Magazine on my observations on value. I have not yet seen it, but expect to have it here in a few days. Major Torrens and I had a long conversation on this question, without convincing each other. I have distinctly stated in my book, that value is not regulated solely by quantity oflabour, when capitals are employed in production which are not equally durable. I mean to insert ... the following observation to be printed in the next edition and which I think more fully answers Major Torrens' objection. 'The same result will take place ...' (24 November 1818, vn, p. 338) According to Sraffa, the inserted passage was an attempt to counter a separate objection by Torrens from that of the two manufacturers employing different ratios of direct labour to produced raw materials - the case posed in the 'Fragments' - the objection having been stated 'in a Note which has not been found' (iv, pp. 305-6). I find this suggestion implausible. Supposing that Torrens had raised the 'separate objection', one would surely expect to find it mentioned in his article for the Edinburgh Magazine.*5 Yet the only detailed theoretical objections he saw fit to make against the 'pure' labour theory were the following: it broke down if equal money capitals were expended differently on produced raw material or fuel, it could not accommodate different proportional expenditures on durable machinery, and it was vitiated by different wage rates (all ceteris paribus). Of these objections, only the first had not been 'covered' explicitly by Ricardo in the Principles, and it is only the first which is emphasised by Torrens in the 'fragments'. On the other hand, Torrens did mention the possibility of unequal periods of production in his later Essay on the Production of Wealth (1821), but only as a qualification to his own argument, not as a separate criticism of Ricardo's 'pure' labour theory ([1821] 1965, p. 40). I submit that there is no reason to suppose the existence of a separate objection. It is nevertheless true that Torrens's objection in the 'Fragments' was not predicated on differences in the periods of production, seemingly conceived by Ricardo with respect to a single activity (as opposed to a backwardly integrated process, covering all previous stages of productive activity). Yet we know that in his 'answer' to 45

Torrens's piece was titled 'Strictures on Mr. Ricardo's Doctrine respecting Exchangeable Value', Edinburgh Magazine, 1818, pp. 335-8. Together with Ricardo's response, it is further considered in the following chapter.

184

INTERPRETING RICARDO

Torrens, Ricardo had formulated a 'strong case' of differences in production periods in the guise of his 'iron versus wood' example, serving as a prelude to his eventual conclusion that 'in the production of the raw material different quantities of fixed capital [may be] used'. It is therefore possible that in drafting the new passage, Ricardo thought he could answer Torrens with the antecedent stage in his reasoning although, in truth, it was not a direct answer. The fact that he did not explicitly mention the objection actually put by Torrens offers support for his reading. Returning to the textual changes, the next was the insertion of the following paragraph, partly incorporating the second of the concluding paragraphs from the first edition chapter: It will be seen, then, that in the early stages of society, before much machinery or durable fixed capital is used, the commodities produced by equal capitals will be nearly of equal value, and will rise or fall relatively to each other only on account of more or less labour being required for their production; but after the introduction of these expensive instruments, the commodities produced by the employment of equal capitals will be of very unequal value; and although they will still be liable to rise or fall relatively to each other, as more or less labour becomes necessary to their production, they will be subject to variation also from the rise or fall of wages and profits. Since goods which sell for 2,000/. may be the produce of a capital equal in amount to that from which are produced other goods which sell for 10,000/. the profits on their manufacture will be the same; but those profits would be unequal, if the prices of the goods did not vary with a rise or fall in the rate of profits. (1, p. 62 n. 1) As with the relocation and incorporation of the first of the previously concluding paragraphs (see above, p. 182) this alteration might also be read as providing a more expansive testimony to the 'pure' labour theory's shortcomings. The last alteration in the new fourth section involved the excision of the erroneous claim that commodities 'never rise [in price] unless additional labour be bestowed on them' (1, p. 63), and of the statement that 'all commodities in the production of which fixed capital enters, not only do not rise with a rise of wages, but absolutely fall; fall too as much as 68 per cent., with a rise of seven per cent, in wages, if fixed capital be exclusively employed, and be of the duration of 100 years'. The following was substituted: no commodities whatever are raised in exchangeable value merely because wages rise; they are only so raised when more labour is bestowed on their production, when wages fall, or when the medium in which they are estimated falls in value. (1, p. 63 n. 3, emphasis in original)

The labour theory of value (I)

185

The 'correct' point that commodities will rise in exchangeable value when wages fall (relative to the unassisted labour commodity, presumably), was previously made in the now deleted last paragraph in the first edition chapter. As to the omission of the '68 per cent' sentence, this may be seen as an attempt by Ricardo to 'divest' his views 'of that appearance of paradox which they now wear' (vn, p. 222), an 'appearance' which had been lampooned by his British Review critic. Immediately following the amended passage, the fifth section opens under the heading: Different effects from the alteration in the value of money, the medium in which PRICE is always expressed, or from the alteration in the value of the commodities which money

purchases. (1, p. 63, n. 3, emphasis and capitals in original) This was an unfortunate juncture for the new section, coming just before the long overdue explanation that the 'curious effect' analysis rested on the assumption of an 'unassisted labour' standard. Relatedly, it drew attention to the major anomaly in the first chapter and, indeed, the whole book. The fifth section contained the disquisition on the 'real value' of (aggregate) distributive shares, with measurement supposedly effected by an 'invariable standard' which would indicate (changes in) the quantity of labour required to produce (bundles of) commodities. But the only commodity explicitly specified as being 'of an invariable value' is the one that yielded price changes without altered labour expenditures. All the (relatively) good work that had gone into relocating the 'misplaced' paragraphs at the end of the chapter had been undone. (The remainder of the section, and thus of the chapter, was essentially unchanged.) What can be said about these textual changes? Regarding the section headings, Edwin Cannan rather gleefully pointed out that the third and fourth 'flatly contradict' the second which, prima facie, is so (Cannan 1929, p. 176). However, imperfections were only to be expected in view of the hasty, almost uninterested drafting of the section headings. And it is equally unsurprising that their interpolation in a largely unreconstructed chapter should have created fresh problems. More substantially, other writers have contended that the new chapter contained 'a more accurately qualified statement of the practicability of "embodied labour" as a measure of value' (Hollander 1904); and, on the same theme, that there was more emphasis on

l86

INTERPRETING RICARDO

the 'time' dimension to productive activity and, hence, on the 'modifications' to the 'pure' labour theory (Edelberg 1933). These readings are not entirely fanciful. I have drawn attention to a few redrafted passages which may lend credence to such interpretations, but their most compelling support would seem to come from the addition of a third reason for unequal capital structures: differences in the durability of circulating capital. To place this in perspective, however, the 'considerable modifications' to the 'pure' labour theory were generously admitted in the first edition chapter, and it is by no means clear that Ricardo wished to suggest an even greater quantitative modification. Overall, the second edition of the Principles is more remarkable for the revisions that were not made. Ricardo had failed to address the major problems with his treatment of value: how can the pervasive use of the 'pure' labour theory be justified if it is so 'considerably modified' by unequal capital structures? And how can a blandly stipulated 'invariable standard' guarantee the exhaustive link between natural price movements and changes in labour expenditures? Ricardo was soon to be awakened to these questions. CONCLUSION

The 'pure' labour theory, adopted by Ricardo in April 1816 or soon after, provided him with a quantity-of-labour conception of 'difficulty or facility of production' which could be determined independently of general exchange relationships. The notion that ('permanent') changes in exchange relationships were traceable to altered 'difficulty or facility' in the production of commodities had been ventilated in the Essay but, despite any appearance to the contrary, Ricardo had not developed a labour theory at that time. It was only when, as he put it, he had rid himself of his pamphlet Proposals for an Economical and Secure Currency•, in February 1816, that

the 'breakthrough' was made. It was not an altogether surprising development. Ever since he became directly concerned to isolate the forces governing 'permanent' changes in profitability, Ricardo had focused on the role played by altered conditions of production, especially the conditions of producing food. The vague 'theory' in the Essay, which highlighted the importance of'difficulty or facility of production', was a product of that analytical preoccupation, while the identification of

The labour theory of value (I)

187

'difficulty or facility' with labour expenditure may be seen as a conceptual development, enabling Ricardo to overcome the logical inconsistencies which had ensnared him in his post-Essay debate with Malthus. Having adopted the 'pure' labour theory, Ricardo drafted the 'core' chapters of the Principles, those spanning 'On Value' up to, and including, 'On Foreign Trade'. But then, in late September or early October 1816, he turned back to 'value', only to realise that the 'pure' labour theory was 'considerably modified' by unequal capital structures. This discovery was initially perceived as a problem (and no wonder) but, I have argued, Ricardo had the inspiration to make it work in his favour, at least on one level, in the form of the 'curious effect': an inversion of the 'orthodox' prediction that a rise in money wages leads to higher prices. At the same time, he apparently realised that his draft manuscript would require revision in the light of the new material (although he did not spell out his precise intentions in the extant correspondence). But, with Mill's prodding, he decided to press on with the book, leaving the revision process until several months later. The problems with the treatment of 'value' in the published first edition of the Principles were manifold. On the one hand, we find the enunciation of a 'pure' labour theory and, within that theoretical discourse, the stipulation of an 'invariable measure of value', defined by the sole criterion of an unchanging labour input, which would indicate changes in the 'real value' of other commodities ('real value' being an exclusive function of labour expenditure). On the other hand, we also find Ricardo emphasising the significance of the 'considerable modifications' to the labour theory which result from unequal capital structures, with no attempt to justify his use of the 'pure' theory in the light of those modifications. The only condition explicitly stated by him, in order to obtain the unique correspondence between changes in natural prices and changes in labour expenditures, was that of a fixed labour input for the 'standard', without any reference to capital structure. Perhaps, at the time of its discovery, Ricardo had some awareness of the more substantive revision to his manuscript that the inclusion of the 'curious effect' material would necessitate. If so, such thoughts must have been wiped from his mind by the time he came to put his manuscript 'in some tolerable order'. Such an oversight would be entirely consistent with Ricardo's mental habits. He once wrote to

l88

INTERPRETING RIGARDO

Hutches Trower, 'Reading or writing, when one has an object in view, should be followed systematically, and at no distant intervals, for after a time our thoughts are turned into new channels and we cannot easily recall the ideas which were only beginning to be indistinctly formed in our minds' (29 October 1815, vi, pp. 314-15). The 'interval', in this case, would have been three to four months in length, during which time Ricardo had ploughed on from one topic to the next. When it is considered that the Principles appears to have been written in little more than six or seven months, that James Mill was evidently happier to advise on style than on substance, and that Ricardo was struggling with some inherently difficult problems, then the existence of severe textual inconsistencies may not seem quite so shocking. The revisions made to the chapter 'On Value' for the second edition did nothing to resolve the inconsistencies. If anything, Ricardo succeeded in accentuating them. But as I have argued, one should not read too deeply into the textual changes. Certainly, there is no evidence that Ricardo was consciously 'retreating' in his advocacy of the labour theory. What was needed, to detonate Ricardo's awareness of the problems he had created for himself, was a well-considered critique from a worthy adversary. Malthus was to fit the bill.

CHAPTER 5

The labour theory of value (II)

This chapter completes the chequered history of Ricardo's 'labour theory' involvement. The first section begins with a review of the relevant correspondence between 1818 and 1820, during which time Malthus launched his devastating salvo on Ricardo's treatment of value in the form of the half-ounce-of-silver riddle; I next consider the published version of Malthus's critique, contained in his own Principles, together with Ricardo's responses; and then, to close the section, I intrude on Ricardo's private grief as he set about rewriting his chapter 'On Value' in the light of Malthus's criticisms. The substantive changes to the chapter 'On Value' in Ricardo's third edition are documented and discussed in the second section. The final section, divided into four sub-sections, focuses on various aspects of Ricardo's later writings. In the first sub-section, I highlight the 'scientific' input to Ricardo's conception of a 'perfect' measure of value; in the second, I report his increasing tendency to associate (one concept of) 'value' with the physical expenditure of labour time; in the third, I show that he also retained a more conventional concept of 'value' in the form of (relative) cost of production; and in the fourth, I endeavour to explain the functions of the illusory 'perfect' measure of value. A conclusion follows. BETWEEN THE SECOND AND THIRD EDITIONS OF THE PRINCIPLES Correspondence

Soon after revising the Principles for its second edition, Ricardo read Torrens's article in the Edinburgh Magazine ('Strictures on Mr. Ricardo's Doctrine respecting Exchangeable Value'). Torrens opened his paper by citing Adam Smith's limitation of the 'labour 189

igO

INTERPRETING RIGARDO

expended' theory of exchangeable value to the 'first and rudest period of society'. He commented: This limitation of the principle is represented as a great and fundamental error by Mr. Ricardo, who contends that in the most advanced periods of society, as well as in that rude and simple state . .. the labour bestowed upon production is the only foundation of exchangeable value. (Torrens i8i8,p. 335) Then, by means of the arguments reported in the last chapter (above, p. 183), Torrens prosecuted his theme that Smith's lack of perseverance with the labour theory was well justified. Ricardo responded to Torrens informally.1 In a notebook, he juxtaposed Torrens's criticisms with various passages from the chapter 'On Value' which referred to the 'modifications' to the 'pure' labour theory (iv, pp. 315-18), the sub-text being that Torrens traduced his position by claiming that he considered 'the labour bestowed on production [as] the only foundation of exchangeable value'. He also wrote to Mill: I have perhaps said too much on my agreement with Dr. Smith in the passage that I have quoted from Torrens. The fact is that Torrens does not represent Smith's opinion fairly[;] he makes it appear that Smith says that after capital accumulates and industrious people are set to work the quantity of labour employed is not the only circumstance that determines the value of commodities, and that I oppose this opinion. Now I want to shew that I do not oppose this opinion in the way that he represents me to do so, but Adam Smith thought, that as in the early stages of society, all the produce of labour belonged to the labourer, and as after stock was accumulated, a part went to profits, that accumulation, necessarily, without any regard to the different degrees of durability of capital, or any other circumstance whatever, raised the prices or exchangeable value of commodities, and consequently that their value was no longer regulated by the quantity of labour necessary to their production. In opposition to him, I maintain that it is not because of this division into profits and wages, - it is not because capital accumulates, that exchangeable value varies, but it is in all stages of society, owing only to 2 causes: one the more or less quantity Dissatisfied with McCulloch's attempt to answer Torrens, James Mill had hoped that Ricardo would publish his own defence (18 December 1818, vn, p. 364). But Ricardo had no such intention (22 December 1818, vn, p. 372, and 28 December, vn, p. 377). McCulloch's effort, paraded as a 'vindication' of Ricardo's theory of exchangeable value, was printed in the November 1818 edition of the Edinburgh Magazine. He merely asserted that, notwithstanding differences in capital structures, 'quantity of labour' remained 'the only circumstance which determines . . . exchangeable value'. McCulloch seemed to think that this was also Ricardo's position.

The labour theory of value (II)

191

of labour required, the other the greater or less durability of capital. (28 December 1818, vn, p. 377) This betokened a departure from Ricardo's stated position in the first two editions of the Principles, that exchangeable value in the early stages of society 'depends solely' on the labour expended in production: a claim which had coexisted uneasily with the analysis of differing capital structures in fishing and hunting (see above, pp. 155, i6off). Ricardo proceeded to assert, with reference to the two 'causes', that 'the former is never superseded by the latter, but is only modified by it'. That is, changes in 'labour expended' are allegedly of greater importance in accounting for ('permanent') variations in exchangeable value than changes in distribution operating through divergent capital structures. But how much more important? If, as Ricardo had stated in the Principles, and illustrated with his numerical examples, the second cause occasions a 'considerable modification', then the dominance of the first cause might be relatively weak. This, at any rate, would be a reasonable inference from the published text. There is no evidence that Ricardo was greatly perturbed by Torrens's criticisms. Presumably, he thought he had successfully 'answered' the detailed arguments with his new allowance for differences in the durability of circulating capital (see above, pp. 182-4). Beyond that, he would later allow explicitly for 'modifications' in the 'early stages', and drop the passage criticising Smith for limiting the domain of the labour theory. Neither change, effected in the third edition of the Principles, was of great significance. The most penetrating criticisms of Ricardo's treatment of value were presented in Malthus's Principles, published in April 1820. But in condensed form, they were put to Ricardo in earlier correspondence, which I consider here. Malthus had devised the following conundrum: If we suppose half an ounce of silver on an average to be picked up by a days search on the sea shore, money would then always retain most completely the same value. It would always on an average both cost, and command the same quantity of labour. The money price of labour could never permanently either rise or fall; and the accumulation of capital in all cases where capital was used and the same quantity of labour employed, would shew itself in a fall of prices owing to the diminished rate of profits.

IO,2

INTERPRETING RIGARDO

Corn alone would rise in money price on account of the increased quantity of labour required; but the rise would be inconsiderable, and strictly limited by the diminution of corn wages which the labourer could bear. And then, triumphantly: Under these circumstances I should like to know from you how the profits of stock would be regulated. They could not evidently be regulated by the rise in money wages of labour, because labour would not alter in money value. (10 September 1819, vm, pp. 64-5, emphasis in original) Malthus had shown great ingenuity.2 The only standard Ricardo had specified when allowing for differences in capital structure was a commodity produced by unassisted labour in a year, evidently considered to be an 'extreme' in terms of capital structure. But, according to Malthus, there was conceivably an even more 'extreme' commodity, such as the half ounce of silver which, applying the logic to Ricardo's own presentation, should serve as numeraire. Then, linking up with the central Ricardian thesis on distribution, adumbrated in the chapter 'On Profits', Malthus was inviting Ricardo to reconstruct that analysis with silver as the standard and allowingfor differences in capital structure.

But as he remarked, the truly extreme standard 'freezes' the money wage. With homogeneous labour, the (natural) money wage is always half an ounce of silver per day throughout the economy. And if, as Malthus apparently did, we assume a constant corn-only wage basket, it also follows that corn cannot rise in price when it becomes more difficult to produce. If capital structures differ, Malthus was effectively saying, this should have been recognised in the main body of the Principles, especially in the 'core' chapter 'On Profits'. This was a legitimate point given Ricardo's failure to offer an adequate justification for using the 'pure' labour theory beyond, and sometimes within, his first chapter. Ricardo's hasty response was as follows: 1 have not been able to give a proper degree of attention to the subject of your letter. The supposition you make of half an ounce of silver being picked up on the sea shore by a day's labour, is you will confess an 2

He may have been inspired by the passage from The Wealth of Nations in which Smith had written, 'In some parts of Scotland a few poor people make a trade of gathering, along the sea-shore, those little variegated stones commonly known by the name of Scotch Pebbles. The price which is paid to them . . . is altogether the wages of their labour; neither rent nor profit make any part of it' ([1776] 1981, p. 69). The example of Scotch Pebbles was later given by Malthus, in his anonymously published review of McCulloch's Essay on Political Economy ([Malthus] 1824, p. 312).

The labour theory of value (II)

193

extravagant one. Under such circumstances silver could not as you say rise or fall, neither could labour, but corn could or rather might. Profits I think would still depend on the proportions of produce allotted to the capitalist and the labourer. — The whole produce would be less, which would cause its price to rise, but of the quantity produced the labourer would get a larger proportion than before. This larger proportion would nevertheless be a less quantity than before, and would be of the same money value. (21 September 1819, vm, p. 73) This reply embodies a confusion which may have derived from Malthus's formulation, but may also reflect Ricardo's commitment to 'pure' labour theory reasoning. In terms of silver prices, the reduction in corn output does not 'cause' a rise in price, in any straightforward sense, although it does bring about a change in relative prices. This may be seen most clearly if, in addition to the use of silver as numeraire, it is supposed that 'corn model' assumptions hold, with a constant corn wage. When corn output falls, agricultural profitability declines but the corn price remains unchanged; then, for manufacturing profitability to come into line, the prices of manufactures must fall. Not only is the corn price unrelated to changes in physical output (by assumption), but a rise in its price is irrelevant to the reduction in general profitability. The fact that Ricardo apparently took for granted that the corn price would rise with a reduced produce suggests that he could not rid himself of the 'labour theory' presumption that corn must rise in price when it becomes more difficult to produce (and, incidentally, that he did not think through the puzzle using 'corn model' reasoning). Otherwise, his answer exhibits a recurrent feature of his writings between the second and third editions: mere assertion of his position, loosely couched in terms of the proportional allocation of material produce. Uncharacteristically, Ricardo ended his reply to Malthus with an 'empirical' objection to the half-ounce-of-silver assumption: 'It is difficult to conceive that in a great, and civilized country any commodity of importance could be produced . . . without the employment of capital' (vm, pp. 72-3). Malthus retorted: You observe that my supposition is an extravagant one. It is so. But perhaps it is safer to reject capital and profits entirely, than to apply them in any given way, under the certainty that scarcely any other commodity can reasonably be supposed to have required in its production exactly the same quantity of fixed and circulating capital employed for exactly the

194

INTERPRETING RICARDO

same time. On any supposition you can make respecting the capital employed in the production of the precious metals, it is scarcely possible that all your calculations should not be necessarily and fundamentally erroneous. (14 October 1819, vm, p. 108) He had turned the tables. If his riddle was to be criticised on the grounds of 'realism', he could level the same criticism against Ricardo, so cutting off any possible escape route via the respecification of the standard. Ricardo was forced to backtrack. He rejoined: With respect to my calculations, I have only this to say in defence of them, that I never brought them forward for any practical use, but merely to elucidate a principle. It is no answer to my theory to say that 'it is scarcely possible that all my calculations should not be necessarily and fundamentally erroneous', for that I do not deny, but still it is true that the proportion of produce in agriculture or manufactures, retained by the capitalist who sets the labourers to work, will depend on the quantity of labour necessary to provide for the maintenance and support of the labourers. (9 November 1819, vm, p. 130)3 The criterion of empirical realism is dropped and, again, Ricardo is found asserting that the proportional distribution of produce depends upon the labour expended on the labourers' wages. This was the 'opinion' which he had sought to 'elucidate', and had done so in the chapter 'On Profits' using a 'pure' labour theory. That discourse was now in jeopardy. The problems stemmed from differences in capital structure and, fundamentally, it was the legitimacy of using the labour theory that Ricardo was now forced to defend. A little over a month after his skirmish with Malthus, he wrote to his ally, McCulloch: I am more convinced than ever that the great regulator of value is the quantity of labour required to produce the commodity valued. There are many modifications which must be admitted into this doctrine, from the circumstances of the unequal times that commodities require to be brought to market, but this does not invalidate the doctrine itself. I am not satisfied with the explanation which I have given of the principles which regulate value. I wish a more able pen would undertake it - the fault is not in the inadequacy of the doctrine to account for all difficulties, but in the Cf. the following, from an earlier letter to Mill: 'I contend it is of essential use to determine what the causes are which regulate exchangeable value, although they may be so complicated, and intricate, that practically, the knowledge may be very little useful' (28 December 1818, vii, p. 378).

The labour theory of value (II)

195

inadequacy of him who has attempted to explain it. (18 December 1819, vm, p. 142) As so often with Ricardo, serious theoretical difficulties were attributed to his pen. And difficulties there most certainly were. Compelled to defend his 'pure' labour theory, he had convinced himself of its merits, but the fact he was doing his best to externalise was that the 'many modifications' do 'invalidate the doctrine'. He had still to formulate a considered response. With the publication of Malthus's Principles in April 1820, the necessity to do so could no longer be ignored. Ricardo knew that he was vulnerable even after reading Malthus's work 'rather in haste' (vm, p. 183): 'From the very complicated nature of the subject of value', he confided in McCulloch, 'Mr. Malthus has, I think, more chance of discovering some flaw in my argument in the chapter which treats of it, that in any other part of the book.' He continued: and this chance is increased by my supposition of a medium which shall itself be invariable. This medium may be supposed to be produced under a variety of circumstances. It may either be the result of the employment of labour only, as supposed by Mr. Malthus, when he supposes it to be picked up on the sea shore, in which the advance of only one days sustenance is required; or it may be produced under all the variety of different portions offixedcapital, and employed for different portions of time. If produced by labour only ... 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 theriseof 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, vm, p. 179) Ricardo had again presumed that corn must rise in price when 'produced with more difficulty', but now he was more revealing that this was the product of'pure' labour theory logic: 'I always measure the rise of every thing by the quantity of labour necessary to produce it.' He apparently failed to see that (assuming a corn-only wage) a rise in the silver price of corn in proportion to the greater 'difficulty of production' would reduce the 'real wage' (in terms of its labour expended content) pan passu, leaving profitability unchanged. 4 But 4

With a constant mixed wage, corn can rise in price but only to the extent that manufactured wage-goods fall. It seems fairly clear that Ricardo was not relying on such reasoning to explain the higher corn price.

196

INTERPRETING RIGARDO

he did notice the implications of specifying a different standard when capital structures are unequal: if much fixed capital was employed in the mines, or if a considerable time must elapse before the circulating capital is returned by the sale of the silver, labour might be exceedingly variable in such a medium. Every commodity measured in such a medium would rise when labour rose, if it were the produce of labour only, or of a less portion of fixed capital than was employed in the production of money, and on the contrary from the same cause those commodities which were produced by larger proportions of fixed capital, or required more time to finish, would fall in price. This is all implied in my book, but I have not been sufficiently explicit, for I ought to have said that if the medium is produced under certain circumstances, there are many commodities which may rise in consequence of a rise in labour, altho' there are many other which would fall, while a numerous portion would vary very little, (vm, pp. 179-80) This assertion, for such it was, that a 'numerous portion' of commodities would Vary very little', came to assume increasing importance in Ricardo's attempt to rescue his 'pure' labour theory. The final item of correspondence considered here is Ricardo's densely packed letter to McCulloch of 13 June 1820. The first issue addressed by Ricardo was the resolution of all differences in capital structure into differences in the 'time of completing a commodity' (vm, p. 193) which, as I anticipated (above, p. 179), was the belated outcome of his debate with Torrens. Ricardo wrote: [I]f I employ valuable machinery from which I have no return for two years, at the end of the two years, my machinery and my goods together, must be of the value of all the labour employed in producing them, besides the accumulated profit on the capital which yielded me no return for that time. But the same result would take place if I employed circulating capital only and could not bring my commodity to market for two years - at the end of the two years, the commodity will not be worth only all the labour bestowed on it, but also all the accumulated profits for the time that my capital was so employed. Strictly speaking then the relative quantities of labour bestowed on commodities regulates their relative value, when nothing but labour is bestowed upon them, and that for equal times, (vm, PP- 192-3) The treatment of machinery (or other 'leftover' means of production) as a joint product with commodities had been adopted by Torrens in his 1819 Edinburgh Magazine article, whence Ricardo had doubtless drawn his inspiration. However, the salient point is not so much whether there is a fixed capital 'residual', but rather

The labour theory of value (II)

197

the 'time profile' of past labour inputs. As Ricardo says, there can be 'accumulated profits' without the employment of any fixed capital, so that even in a circulating capital model, a 'pure' labour theory may not apply.5 Disregarding the trivial case of zero profitability, the 'pure' labour theory obtains only if time profiles are identical. How significant were differences in these 'labour profiles' (as I shall call them)? Ricardo continued: 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 ... (vm, p. 193) But he also claimed that 'there are such a variety of cases in which the time of completing a commodity may differ'. And further: I sometimes think that if I were to write the chapter on value again ... I should acknowledge that the relative value of commodities was regulated by two causes instead of by one, namely, by the relative quantity of labour necessary to produce the commodities in question, and by the rate of profit for the time that the capital remained dormant, and until the commodities were brought to market, (vm, p. 194) Taking these statements together, it would seem that although he drew back from attributing co-ordinate importance to labour and 'time', he was admitting quite a hefty 'modification' to the 'pure' labour theory. In this regard, however, the last passage is puzzling. After all, Ricardo had allowed for a 'considerable modification' to the 'pure' labour theory in the published versions of his first chapter and, to that extent, the 'two causes' had been acknowledged already. I can only suggest that he thought he should be even more explicit in terms of the newly adopted 'labour profile' framework, although this suggestion is avowedly tentative.6 Ricardo also considered the implications of differing 'labour 5 6

'Fixed' and 'circulating' capital are here defined on the forward-looking criterion. Commenting on the same passage, Samuel Hollander suggests that Ricardo may have been proposing an allowance for a new 'cause of alteration in exchange value . . . namely the effect upon relative prices of exogenous changes in the period of investment' (1979, pp. 222—3, ernphasis in original). However, Professor Hollander then casts doubt on this suggestion by remarking that 'exogenous changes in the period of production . . . were taken up neither in the third edition, nor in subsequent correspondence and the final paper on Absolute Value and Exchangeable Value' (p. 223). But Hollander's suggestion may be more plausible than he himself imagines (see below, pp. 233-4), although the evidence does not permit an unequivocal judgement on the passage in question.

ig8

INTERPRETING RIGARDO

profiles' for the choice of standard. Given the 'variety' of'the times of completing a commodity': it is difficult to fix on any one commodity which may properly be chosen as a general measure of value, even if we could get over the difficulty of not having one which always requires the same quantity of labour to produce it. The two extremes appear to be these: one, where the commodity is produced without delay, and by labour only, without the intervention of capital; the other where it is [the] result of a great quantity of fixed capital, contains very little labour, and is not produced without considerable delay. The medium between these two is perhaps the best adapted to the general mass of commodities; those commodities on one side of this medium, would rise in comparative value with it, with a rise in the price of labour, and a fall in the rate of profits; and those on the other side might fall from the same cause, (vm, p. 193) The 'advantage' of the hypothetical 'medium' commodity as standard was that, for an alleged 'general mass of commodities', their prices would be largely invariant to changes in distribution. This consideration had certainly not been germane to the choice of the unassisted labour standard in editions 1 and 2 of the Principles, which was the only standard rigorously specified in the context of divergent capital structures. On the contrary, that standard had been selected so as to maximise the purely distribution-induced price movements (reductions). But now, the choice of standard was motivated by one ambition: to salvage the 'pure' labour theory. This is underscored when, immediately following the last extract cited, Ricardo wrote: Mr. Mai thus has taken advantage of this defect in my measure of value ... and has not failed, as he justly might do, to make the most of it ... It must be confessed that this subject of value is encompassed with difficulties - I shall be very glad if you succeed in unravelling them, and establish for us a measure of value which shall not be liable to the objections which have been brought against all those hitherto proposed, (vm, PP- J93-4) All distribution-induced price changes were now considered to be indicative of a 'defect' in the standard. The futile quest had begun for an 'unobjectionable' measure which, bizarre as it may seem, would 'validate' the 'pure' labour theory even though 'labour profiles' differed. Ricardo then turned to the implications of varying 'labour profiles' for his distribution analysis, giving the impression that they were nugatory:

The labour theory of value (II)

199

After all[,] the great questions of Rent, Wages, and Profits must be explained by the proportions in which the whole produce is divided between landlords, capitalists, and labourers, and which are not essentially connected with the doctrine of value. By getting rid of rent . . . the distribution between capitalist and labourer becomes a much more simple consideration. The greater the portion of the result of labour that is given to the labourer, the smaller must be the rate of profits, and vice versa. Now this portion must essentially depend on the facility of producing the necessaries of the labourer - if the facility be great, a small proportion of any commodity, the result of capital and labour, will be sufficient to furnish the labourer with necessaries, and consequently profits will be high. The truth of this doctrine I deem to be absolutely demonstrable, yet I think that Mr. Malthus does not fully admit it. (vm, pp. 194-5)7

As in other, similar passages, the claim that the proportional distribution of material outputs is 'not essentially connected with the doctrine of value' was nothing more than sheer assertion. All Ricardo conveyed was his belief that (changes in) labour conditions of producing wage-goods must govern (changes in) the prevailing rate of profits, whatever pattern of natural price movements might ensue. It was all very well to 'deem' this to be 'absolutely demonstrable'. The problem was to demonstrate it.8 There can be no doubt that Ricardo had been jolted into mental activity by Malthus. Just as he did soon after writing to McCulloch, I now turn to a detailed consideration of the case put against him in Malthus's Principles. I also consider his reactions. Malthus3s 'Principles3 and Ricardo3s Notes

In the fourth section of his second chapter, 'Of the Labour which a Commodity has Cost considered as a Measure of Exchangeable Value', Malthus savaged Ricardo's 'pure' labour theory. In 'any stage of society', Malthus asserted, profits are likely to be a component in 'cost of production', and given the 'varying quickness of returns' (unequal 'durability of circulating capital', in Ricardo's 7

8

The claim that distribution 'must be explained by the proportions in which the whole produce is divided' should be taken with a pinch of salt. See below, pp. 202-3. Piero Sraffa's reading of the sentence beginning 'After all, the great questions of Rent, Wages, and Profits . . . ', was that it represented 'an echo of the old corn-ratio theory' (1, p. xxxiii). It was natural for Sraffa to make this connection. But when, in the same letter, Ricardo came to explain how distribution was 'not essentially connected with the doctrine of value', he rather shot himself in the foot, since the 'proportion of any commodity' required to 'furnish the labourer with necessaries' evidently does depend on exchangeable value considerations.

2OO

I N T E R P R E T I N G RICARDO

phraseology), commodities will not exchange according to the comparative amounts of labour expended on their production (n, pp. 56—8). In 'countries advanced in civilization', he continued, profits are likely to be lower than they were in earlier times, 'and consequently neither the varying proportions of the fixed capitals, nor the slowness or quickness of the returns will produce the same proportionate difference on prices'. However, 'to make up for this, the difference in the quantity of fixed capital employed is prodigious, and scarcely the same in any two commodities' (11, pp. 59-60). A 'pure' labour theory is correspondingly flawed. To make his position clear, and with a nice touch of irony, Malthus then repackaged Ricardo's 'curious effect' analysis, retaining a commodity produced by unassisted labour for one year as the numeraire, but allowing for a 'large class of commodities' produced by unassisted labour in less than a year, the presumed existence of which had been signalled in the 'silver' riddle. Following an assumed rise in the money wage and fall in general profitability, this 'large class' rises in absolute price, and another large class' falls. Some commodities would not experience any price variation, but: from the very nature of the proposition, this class must theoretically form little more than a line ... [and] wherever the line may be placed, it can embrace but a very small class of objects. (11, pp. 64-5) For Malthus, there was no 'general mass' of commodities produced with roughly similar capital structures (or 'labour profiles'). Immediately following, Malthus posed the rhetorical question, 'What then becomes of the doctrine, that the exchangeable value of commodities is proportioned to the labour which has been employed upon them?' (11, p. 65). So ended the first stage of his critique. The second stage, encountered in the section 'Of Money, when uniform in its cost, considered as a Measure of Value', may be viewed as a pre-emptive strike against any attempt by Ricardo to respecify a standard qua measure of'labour expended'. Three possible standards are discussed, the first, silver, produced with 'a certain quantity of fixed and circulating capital employed for a certain time', which always requires the same quantity of labour expenditure (11, p. 80). '[E]ven from the admissions of Mr. Ricardo', Malthus observed, 'none of the commodities which would exchange for a given quantity of silver, would contain the same quantity of labour as that silver, except those which had been produced, not only by the same quantity

The labour theory of value (II)

201

of labour, but by the same quantities of the two kinds of capital employed for the same time and in the same proportions'. And given a change in distribution, all commodities would either rise or fall in price, 'except those very few' produced under identical conditions to silver (11, p. 81), independently of changed labour requirements. The second standard is Ricardo's 'unassisted labour for a year' commodity. The same points are made. Finally, Malthus presented as a possible standard the half ounce of silver, obtained in one day by unassisted labour on the sea-shore. 'With regard to the effects of a rise in the price of labour', he remarked, 'they cannot be the subject of our consideration, as it is evident that no rise in the price of labour could take place on the present supposition'. But if the general rate of profits declined, 'every other commodity would fall compared with money' (11, p. 81), again independently of altered labour expenditures. Hence, when capital structures ('labour profiles') differ, there is no 'standard measure of value' which 'would be an accurate measure of the quantities of labour employed upon each commodity' (11, p. 80). The coup de grace was the application of the foregoing points to Ricardo's treatment of distribution: According to Mr. Ricardo, profits are regulated by wages, and wages by the quality of the last land taken into cultivation. This theory of profits depends entirely upon the circumstance of the mass of commodities remaining at the same price, while money continues of the same value, whatever may be the variations in the price of labour. This uniformity in the value of wages and profits taken together is indeed assumed by Mr. Ricardo in all his calculations, from one end of his work to the other; and if it were true, we should certainly have an accurate rule which would determine the rate of profits upon any given rise or fall of money wages. But if it be not true, the whole theory falls to the ground. We can infer nothing respecting the rate of profits from a rise of money wages, if commodities, instead of remaining of the same price, are very variously affected, some rising, some falling, and a very small number indeed remaining stationary. But it was shewn .. . that this must necessarily take place upon arisein the price of labour. Consequently the money wages of labour cannot regulate the rate of profits. (11, pp. 285-6) And then, to press this home, the 'half ounce of silver' argument was dusted down once again. Malthus's two most piercing two lines of attack had been the following. First, if Ricardo wanted an 'extreme' standard, he should select one in terms of which money wages are constant. And

202

INTERPRETING RIGARDO

secondly, with due allowance for differences in capital structure ('labour profiles'), there is no standard which can 'indicate correctly' the labour expended on (bundles of) commodities, or furnish the exclusive link between changes in prices and labour expenditures. The latter criticism vividly exposed Ricardo's 'blind-spot', both in the chapter 'On Value' itself, and in the main body of his Principles. Ricardo was under siege. We have seen from his correspondence that one line of 'defence' was to assert gamely, in terms of the proportional allocation of material produce, that his distribution analysis stood regardless of price behaviour. But when, in the Notes, he came to spell out his position, he exposed his dependency on 'pure' labour theory reasoning. 'It is very probable', he candidly observed, 'that my language about proportions may not have been as clear as it ought to have been. I will endeavor now to explain it.' He continued: Suppose the last land now in cultivation yields 180 qrs. of corn with the employment of a given quantity of labour, and in consequence of the rise of the price of corn a still inferior quantity of land shall be cultivated next year which shall yield only 170 qrs. If this year the labourer shall have one third of the 180 quarters, and next year he shall have one third of the 170 quarters, I say his wages will be of the same value next year, as this, because the whole 170 quarters next year will be of the same value as the 180 quarters are this year . . . When I speak of this division by proportions I always apply it, or ought to apply it, (and if I have done otherwise, it has been from inadvertence), to the produce obtained with the last capital employed on the land, and for which no rent is paid. Now in fact the labourer will get a larger proportion of the 170 qrs., than he got of the 180 qrs., he will get a larger proportion of this equal value, and therefore it is that I say his wages have risen. Whatever may be the quantity of the corn obtained by the last capital employed on the land, it will be of the same value, because it is the produce of the same quantity of labour. A larger proportion of this equal value must itself be a larger value. (11, p. 196)

This is 'pure' labour theory discourse ('My measure of value is quantity of labour', he declared, 11, p. 197). The argument takes for granted that the corn measured total wage bill rises when conditions of agricultural production deteriorate. It follows that the value of the reduced net corn output, total farmer's profit, has fallen, and that 'real wages' have risen, all in terms of a labour expended 'content' which, in the 'pure' labour theory universe, is reflected in money values. Ceteris paribus, the proportion of gross physical output

The labour theory of value (II)

203

'allocated5 to labour by all other capitalists - the value equivalent of wage bills in terms of output - will also rise. This had been Ricardo's repeated assertion, but it is now perfectly coherent within a 'pure' labour theory context. The next stage in the argument, although not detailed in the Notes on Malthus, would be to draw the implications for the rate of profit. Finally, observe that the analysis is conducted at the 'micro' level of'the last capital employed on the land'. It was not framed in terms of the proportional distribution of aggregate (social) output although, needless to say, a changed social distribution is implied (cf. 11, pp. 276, 336). It was helpful of Ricardo to explain his 'language about proportions'. But given his implicit admission that distribution is 'essentially connected with the doctrine of value' - the 'pure' labour theory of value - it was all the more necessary for him to answer Malthus's criticisms. On the extent of the 'modifications' to the 'pure' labour theory, he wrote: I have myself stated that... the general principle of the value of commodities being regulated by the quantity of labour necessary to their production, was modified; but I was of opinion, and still am of opinion, that in the relative variation of commodities, any other cause, but that of the quantity of labour required for production, was comparatively of very slight effect. (11, pp. 58-c))9 Perhaps this really had been Ricardo's private opinion all along, although no such claim for the labour theory had been made in the first two editions of the Principles. Far from it.10 He added: 'Mr. Malthus remark that this [modifying] cause operates in every stage of society is most just' (n, p. 59). He had first 'admitted' this following his brush with Torrens: see above, pp. 190-1. His position in the Notes is itself somewhat contradictory. Although he reasserted the unimportance of the 'modifications' (11, pp. 66, 100-2), he also forgot his newly proclaimed, long-standing conviction when criticising Malthus, rather than defending himself. Malthus had written: 'If we were determined to use only one term [for the determinant of exchange relationships], it would certainly be more correct to refer to capital rather than to labour . . . The natural or necessary prices of commodities depend upon the amount of capital which has been employed upon them, together with the profits of such capital' (11, p. 75). To which Ricardo replied: 'If equal capitals yielded commodities of nearly equal value, there might be some grounds for this argument; but as from a capital employed in valuable machinery, and steam engines, a commodity of a very different value is obtained than from a capital, of the same value, employed chiefly in the support of labour, it is at once obvious that the one term, thought to be the more correct by Mr. Malthus, would be the most incorrect that could be imagined' (11, pp. 74-5).

2O4

INTERPRETING RICARDO

I turn to an associated matter: Ricardo's response to Malthus's visualisation of the pattern of distribution-induced price movements. In one note Ricardo mostly accepted Malthus's position. 'In all cases where the rise in the price of corn, is followed by a rise in the money price of wages, and a fall in profits', it was 'correct' to say that a 'large class' of commodities fall in price and another 'large class' rises, although Ricardo added that the 'last class will rise only in a trifling degree' because 'they [the commodities within the class] will rise on account of the rise of the price of labour, [but] they will fall on account of the fall of profits. The fall from the latter cause, will, in a great measure, balance the rise from the former' (n, pp. 62—3). Missing from this is the counter-assertion to Malthus, that a 'general mass' of commodities would remain unchanged in price. However, in a subsequent note Ricardo did claim that the 'general average' of commodities 'will not be much affected' by a change in distribution (11, p. 288) and, presumably, this class was 'average' in respect of capital structure ('labour profile'). Taking the two notes together, it follows that unless there was assumed to be a more numerous class of rising than falling commodities, the total value of output will not remain invariant to (notional) changes in distribution. There is no evidence in the Notes that Ricardo made any such assumption. Ricardo also tackled the question of a 'measure of value', making his objective, at least, luminously clear. 'What we want', he ventured, 'is a standard measure of value which shall be itself invariable, and therefore shall accurately measure the variations of other things' (11, p. 29). As in the pre-'curious effect' passages in the chapter 'On Value', the 'variations' are those wrought by altered 'facility or difficulty of production', conceived in terms of labour expenditure (11, pp. 31-4). On the criterion for 'invariability', Ricardo was unambiguous: The increase in the price of each individual quarter [of corn], in an unvarying medium, must be owing ... to an increased cost of production[.] ... [C]orn can only rise in an unvarying medium on account of an increased cost of production, more labour must be bestowed to obtain the same quantity. (11,

p. 243, my emphasis) A truly invariable measure of value only yields (natural) price changes when commodities are produced with more or less labour. This was all very well, but if'labour profiles' differ, no commodity is 'invariable' in Ricardo's sense, as Malthus had demonstrated with clinical precision. Ricardo commented:

The labour theory of value (II)

205

Mr. Malthus has clearly shewn that no medium that can be chosen is or can under any circumstance be even supposed to be an accurate measure of value. I not only admit this but have myself pointed it out. (11, p. 288) To this, Sraffa footnoted a reference to the chapter 'On Value' which, presumably, was where he thought Ricardo had 'pointed it out'. However, in the passage cited, Ricardo only stated that if a commodity always required the same quantity of labour expenditure, it would be 'of an unvarying value' (1, p. 17 n. 3, quoted above, p. 156). The import of the passage, and of others beyond the first chapter, is that circumstances could be supposed in which a commodity would be 'an accurate measure'. One might argue that Ricardo's analysis of differing capital structures implied the nonexistence of an 'accurate measure', but it was disingenuous for him to claim that he had 'pointed it out'. He had done no such thing. Underlying Ricardo's attempts to rationalise Malthus's criticisms was an unmistakable attraction for the 'pure' labour theory. Having adopted the theory, which served him well in his demonstration of the link between altered conditions of agricultural production and 'permanent' movements in general profitability, it increasingly made perfect sense to him. What better could express the idea of greater ease or difficulty of production than changes in the amount of labour expended to obtain a certain output? Among others, these attractions were so strong that he would never relinquish the theory. Nor would he escape from a web of conceptual confusion. This brings me to his use of 'value' terminology in the Notes. Here I must mention another of Malthus's criticisms. He wrote: We have the power indeed arbitrarily to call the labour which has been employed upon a commodity its real value; but in so doing we use words in a different sense from that in which they are customarily used; we confound at once the very important distinction between cost and value. (11, pp. 30-1, emphasis in original) The textual basis for this (unfair) criticism probably derived from Ricardo's vacillation between designating either altered 'cost of production' or altered labour expenditure (or 'real value', etcetera) as the cause of changes in natural price (see above, pp. 168—71). Ricardo replied to Malthus: Mr. Malthus accuses me of confounding the very important distinction between cost and value. If by cost, Mr. Malthus means the wages paid for labour, I do not confound cost and value, because I do not say that a commodity the labour on which cost a £1,000, will therefore sell for

206

INTERPRETING RICARDO

£1,000; ^ may sell for £1,100, £1,200, or £1,500, -but I say it will sell for the same as another commodity the labour on which also cost £1,000; that is to say, that commodities will be valuable in proportion to the quantity of labour expended on them. If by cost Mr. Malthus means cost of production, he must include profits, as well as labour; he must mean what Adam Smith calls natural price, which is synonymous with value. (11, pp. 34-5) Quantity of labour is here identified with the cost of labour, that is, the wages paid to labour.11 But is quantity of labour, as measured by its cost to the employer, the same as 'real value'? According to one note, apparently not: 'The real value of a commodity I think means the same thing as its cost of production' (11, p. 35). What is to be made of this? If'real value' is 'the same thing' as cost of production, it is not a unique function of labour expenditure: with unequal 'labour profiles' and a change in distribution, some 'real values' change when measured by any numeraire, even though their labour 'contents' do not. This is not the sense in which 'real value' seems to have been used previously, nor the sense in which it (mostly) would be used subsequently. However, although 'real value' had apparently been redefined in the Notes, 'natural value' comes from the past tofillthe void: If poorer land were cultivated the quantity of produce on that land would not bear the same proportion to the labour employed as before ... [and] I should say that the natural value of corn had risen, at whatever value in money it might be rated. (11, p. 289)12 1

' Cf: 'I do not say a portion of its [a commodity's] cost measures its exchangeable value - but I say its whole value will be in proportion to a portion of its cost... If I had said that the value of commodities was the same thing as the value of the labour expended on them, the [Malthus's] remark would have been well founded, but 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' (11, pp. 100-2, emphasis in original). The same identification of 'quantity of labour' with wages is conspicuous in Ricardo's letter to McCulloch of 13 June 1820. Thus: 'If it be two years before he [a capitalist] can receive the return from the work done by the steam engine, he must have £ 100 for the first years profit, and £i10 for the second, and this is totally independent of the quantity of labour [i.e., wages] actually accumulated in the commodity brought to market'; 'my machinery and my goods together, must be of the value of all the labour employed in producing them [i.e., wages], besides the accumulated profit'; 'the commodity will not be worth only all the labour bestowed on it [i.e., wages], but also all the accumulated profits'; 'When the times are unequal, the relative quantity of labour bestowed on them [commodities] is still the main ingredient which regulates their relative value, but . . . besides compensating for that labour [in the form of wages], the price of the commodity, must also compensate for the length of time that must elapse before it can be brought to market [in the form of profits]' (vm, pp. 192-3, my emphasis). 12 For the earlier usage of'natural value' see above, p. 151.

The labour theory of value (II)

207

'Natural value' is thus presented as the unique correspondent to 'quantity of labour' and this type of'value' is allegedly separate from money 'value', i.e., cost of production measured by some numeraire.1^ Here we have the basis for Ricardo's future dilemma. Ricardo had two primary concepts of 'value', one which was related to 'quantity of labour', the other of a more conventional, (relative) 'cost of production' nature. Where they collided was with the association of labour quantity with labour cost; that is, with the absorption of 'labour quantity' value within the cost of production discourse. This was relatively unproblematical if a 'pure' labour theory held, for in those circumstances the two concepts come into single focus: given a standard, changes in both 'values' are governed exclusively by altered labour expenditures. But when 'labour profiles' differ, a rift opens up: one type of'value' can change independently of the other. From this angle, the purpose of the sought-after 'perfect' standard was to reconcile the two 'value' concepts. It was an impossible requirement. Preparing for the third edition

We know from Ricardo's letters to McCulloch of December 1819 and May and June 1820 that he had been shaken by Malthus's criticisms (see above, pp. 194—9). But when writing to Malthus he was evidently unprepared to reveal his full discomfort, perhaps not wishing to give Malthus the satisfaction that this might bring. Almost as an aside, he informed his adversary: I have been looking over my first chapter, with a view to making a few alterations in it before the work goes to another edition. I find my task very difficult, but I hope I shall make my opinions more clear and intelligible. (4 September 1820, vm, p. 229) Far from just 'a few alterations', the first chapter was to be substantially rewritten, directly as a consequence of Malthus's intervention. Malthus responded waspishly to Ricardo's news: You assert that with few exceptions the quantity of labour employed on commodities determines the rate at which they will exchange for each other. This is a proposition; and one that is not well founded, so that I should doubt whether you will be able to alter your first chapter 13

To complicate matters, Ricardo elsewhere used 'natural value' in the sense of cost of production (n, p. 35).

208

INTERPRETING RIGARDO

satisfactorily to yourself. (25 September 1820, vm, p. 261, emphasis in original) To which Ricardo retorted: You say that my proposition 'that with few exceptions the quantity of labour employed on commodities determines the rate at which they will exchange for each other, is not well founded'. I acknowledge that it is not rigidly true, but I say that it is the nearest approximation to truth, as a rule for measuring relative value, of any I have ever heard ... My commodity and your commodity are both worth £1000 - they will therefore probably have the same quantity of labour realized in each. But the doctrine is less liable to objections when employed not to measure the whole absolute value of the commodities compared, but the variations which from time to time take place in relative value. To what causes, - I mean permanent causes, can these variations be attributed? to two, and to two only; one insignificant in its effect - a rise or fall of wages or what I think the same thing a fall or rise of profits - the other, of immense importance, the greater or less quantity of labour that may be required to produce the commodities. From the first cause 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. To the other cause no very confined limit can be assigned, for the quantity of labour required to produce commodities may vary to double or treble. (9 October 1820, vm, p. 279) The claim that two commodities, of the same money value, 'probably have the same quantity of labour realized in each', was mere assertion. As for the 'argument' about changes in value, this may have been derived from Malthus's published views. It was he who had made the point about profits being a less significant component in cost of production in the 'advanced society' (see above, p. 200). But, conveniently ignored by Ricardo, Malthus had also claimed that this was more than compensated for by differing durabilities of capital. Without an assumption as to the magnitude of such differences, nothing can be deduced about the price effects following a change in distribution. Ricardo signed off his letter defiantly: The subject is difficult, and I am but a poor master of language, and therefore I shall fail to express what I mean. My first chapter will not be materially altered - in principle I think it will not be altered at all. (vm, p. 280) Since the chapter was to be altered quite considerably, in fundamental respects, this should be taken as a signal that Ricardo had no intention of abandoning the labour theory in the light of Malthus's criticisms.

The labour theory of value (II)

209

Malthus pursued his quarry relentlessly. In his next letter he posed the following question: 'Do fifty oak trees valued at 2o£ each contain as much labour as a stone wall in Gloucestershire which has cost iooo^[?]' (26 October 1820, vin, p. 286). The presumption was that the stone wall would have been constructed over a relatively brief time span. As for the oak trees, exactly the same example had been developed by Ricardo himself in his exchange with Torrens (see above, p. 178). At that time, he had assumed a twentyyear 'production cycle', with the implication that the 'pure' labour theory breaks down spectacularly if called on to account for the exchangeable value of wood vis-a-vis more 'normal' commodities. Perhaps he felt irritated by being reminded of this, hence the tartness of his reply: 'Fifty oak trees valued at £20 each do not contain as much labour as a stone wall in Gloucestershire which costs £1000.' I have answered your question let me ask you one. Did you ever believe that I thought fifty oak trees would cost as much labour as the stone wall? I really do not want such propositions to be granted in order to support my system. (24 November 1820, vm, p. 302)

Two years earlier, oak trees were a 'strong case' of a more general phenomenon, described in the second edition of the Principles as 'differences in the durability of circulating capital'. The uncomfortable truth was that this, and the other sources of inequalities in 'labour profiles', did undermine his 'system'. Malthus knew this very well. So did Ricardo. In correspondence with McCulloch, his sturdy ally, he could be more frank. To conclude this sub-section, I consider an exchange of letters between them which took place soon after Ricardo had revised his first chapter. The exchange throws light on Ricardo's more unguarded position at the time. Ricardo had sent McCulloch his notes on Malthus's Principles. They elicited a bewildered reaction: I do n o t . . . think that you are either so perspicuous or so successful in what you have said about value - This in my humble opinion is the least valuable part of your notes - You say that Mr. Malthus 'is quite right in asserting that many commodities in which labour chiefly enters, and which can be quickly brought to market will rise with a rise in the value of labour,' meaning I presume with a rise of real wages . . . - I confess I was not prepared for this proposition . . . You do not I am sure mean to say that a rise of wages can raise the real value of any class of commodities - It can

210

INTERPRETING RICARDO

only raise their relative value, and it does this not in consequence of their rising in absolute value, but of others falling in a still greater ratio - Suppose that the durability of the different capitals employed in production are as i, 2, . . . , 10, &c, and that i is the least and io the most durable — When wages rise they are all affected in the same way but in different degrees . . . [T]hey must... all sink in relative value except the first... and if the standard had been produced by capital whose durability was equal to i they would almost all have fallen as compared with this standard while it is plain none could have risen. (22 January 1821, vm, p. 339, emphasis in original) McCulloch had few credentials for accusing anyone of confusion in the treatment of'value' (see above, pp. 174, 190 n. 1). But he was a faithful 'disciple' whose efforts on Ricardo's behalf, even when they were not, in truth, terribly helpful, deserved a few kind words. T h a t is how I interpret the sub-text of Ricardo's reply. It began: I have made some alterations in the first chapter 'on value' which I fear from the remarks in your letter will not.meet with your approbation . . . I agree in every thing you say respecting the variations which would take place in the relative value of commodities on the supposition that they were produced with capitals of degress of durability, as 1, 2, . . . 10 . . . These are my opinions expressed only in language ten times more clear than I could have expressed them in. (25 January 1821, vm, pp. 342-3) This would appear to be a mixture of flattery and humbug. O n any previous usage of terms, Ricardo's opinion had not been that commodities must fall in 'absolute' value, when capital structures differ, regardless of the adopted numeraire. In the remainder of the letter, Ricardo gently tutored McCulloch on the complexity of the 'value' question. Immediately following the above passage, he wrote: But here is I think the difficulty. You say 'if the standard had been produced by capital whose durability was equal to 1 they would almost all have fallen as compared with this standard, while it is plain that none could have risen'[;] true, if the standard were so produced, but Mr. Malthus and our adversaries say that the standard shall be produced with labour without any capital at all, or at most the capital only that is necessary to support a man in a single day. In this standard your No. 1 would fall with a rise of food and necessaries, and labour could never rise at all . . . If we could take our stand at No. 1 we should do very well but we are driven from it. (vm, PP- 343-4) The 'difficulty' was that with the truly extreme standard, money wages could not change to reflect altered conditions of producing wage-goods, especially corn, which was wholly unacceptable to

The labour theory of value (II)

211

Ricardo. But this was only one of Malthus's scoring points. Ricardo continued: Are you prepared to adopt this standard of daily labour? It may possibly be the correct one, but the circumstances under which it is produced, agree so little with the circumstances under which most other commodities are produced, that by adopting it we introduce a cause of variation of price, which we avoid if we chuse a standard produced under the ordinary circumstances that other commodities are produced, (vm, p. 344)

Malthus had forced Ricardo to confront the question of how any standard could achieve the comprehensive link between changes in prices and changes in labour expenditures. Having done so, it was the minimisation of all distribution-induced price changes which had become Ricardo's priority in the choice of standard. Sadly for him, the perfect solution was beyond his grasp: I am not satisfied, as I have often told you, with the account I have given of value, because I do not know exactly where tofixmy standard. I am fully persuaded that in fixing on the quantity of labour realised in commodities as the rule which governs their relative value we are in the right course, but when I want tofixa standard of absolute value I am undetermined whether to chuse labour for a year, a month, a week, or a day . . . I have reflected so much upon it that I despair of becoming more enlightened upon it by my own unassisted efforts, (vm, p. 344) His faith in the labour theory is again proclaimed. So, too, is his bewilderment. Malthus had inflicted a stunning blow. THE THIRD EDITION OF THE PRINCIPLES

The revised chapter 'On Value7

The changes to Ricardo's treatment of value were heralded in the 'Advertisement' to the third edition of the Principles: 'In this Edition I have endeavoured to explain more fully than in the last, my opinion on the difficult subject of VALUE, and for that purpose have made a few additions to the first chapter' (1, p. 8, capitals in original). CA few additions' was an understatement. As Sraffa remarked, the 'alterations were certainly extensive; little more than half of the final version (edition 3) of the chapter On Value being found in the same form in edition 1' (1, p. xxxviii). Ricardo's first alteration was to substitute 'almost exclusively' for 'solely' in the sentence: 'In the early stages of society . . .

212

INTERPRETING RIGARDO

exchangeable value . . . depends solely on the comparative quantity of labour expended' (i, p. 12).14 The explicit allowance for some differences in capital structure (or 'labour profiles') in the 'early stages', which prompted the change, had been made by Ricardo in December 1818 (see above, pp. 190-1), and is of no great significance. But at least he was removing a minor contradiction which had resulted from claiming labour expenditures as the 'sole' regulator, only to permit differences in capital structure in 'early stages' occupations at a later point in the chapter (see above, p. i6off.).15 The next set of alterations was of great importance. Out went the passage claiming that a commodity with an unchanging labour requirement 'would be of an unvarying value' and could measure the 'variations in other things' arising solely from changes in labour expenditures (1, p. 17 n. 3). This omission had been forced on Ricardo by Malthus, who had played on the point that a constant labour requirement is insufficient to achieve the desired result because of differing capital structures. Four paragraphs were substituted, the first reading: Two commodities vary in relative value, and we wish to know in which the variation has really taken place. If we compare the present value of one, with ... all other commodities, we find that it will exchange for precisely the same quantity of all these things as before. If we compare the other with the same commodities, we find it has varied with respect to them all: we may then with great probability infer that the variation has been in this commodity, and not in the commodities with which we have compared it. If on examining still more particularly into all the circumstances connected with the production of these various commodities, we find that precisely the same quantity of labour and capital are necessary to [their] production ... but that the same quantity as before is not necessary to produce the single commodity whose relative value is altered, probability is changed into certainty, and we are sure that the variation is in the single commodity: we then discover also the cause of its variation. (1, pp. 17-18) 14 15

Cf. similar changes on pp. 20 and 30. This interpretation differs from Sraffa's, according to which 'the "almost exclusively" reflects the change in the choice of standard . . . the new standard permitting a rise of price, as a result of a rise of wages, in the case of commodities produced without fixed capital' (1, p. xxxix). Ricardo had explicitly acknowledged that the 'pure' labour theory was 'modified' in the 'early stages of society' some eight months before Malthus's 'silver' riddle, which was the impulse for the respecification of the standard. Ronald Meek, who thought that Sraffa's interpretation was, in this respect, 'a shade too ingenious' (1973, p. 106 n. 3), and Moore (1966), put forward roughly similar interpretations to the one advanced here, although both traced the stimulus for Ricardo's change in wording to criticisms in Malthus's Principles. On my interpretation, they both post-dated the stimulus, while Moore considerably overestimated the significance of the resulting textual change.

The labour theory of value (II)

213

This passage implies a concept of ('absolute') value applicable to particular commodities. It also suggests that the determinant of this 'value' is the labour expended on production, which receives clear confirmation in the second and third of the newly inserted paragraphs.16 But the tedious exercise which Ricardo had detailed merely sidesteps the difficult questions by implicitly presuming that a 'pure' labour theory holds. Suppose, on the other hand, 'we find' that all but one of the commodities exchange in the same proportions, and that the 'altered' commodity still requires the same labour and capital (accumulated labour): the outcome of a change in distribution when all but the one commodity has the same 'labour profile'. Would we say then that the variation was 'in the single commodity?17 The four new paragraphs had again exposed Ricardo's 'pure' labour theory dependency, since he evidently wanted to say that commodities rise or fall in 'value' in proportion to changes in their individual labour 'contents'. 18 But beyond this, the paragraphs neither directly confront, nor attempt to resolve, the problems he had encountered following Malthus's criticisms. The next major alteration was the deletion of the poorly expressed heading for section 2 (' The accumulation of capital makes no difference . . . ' )

and its replacement with: Not only the labour applied immediately to commodities affect their value, but the labour also which is bestowed on the implements, tools, and buildings, with which such

labour is assisted. (1, p. 22, emphasis in original)19

Immediately following, Ricardo excised the passage censuring Smith for limiting the 'pure' labour theory to the 'early and rude state of society'. These two changes were related and had been 16

17

18

19

For example: 'I find that precisely . . . the cause of the variation between corn and other things, is the smaller quantity of labour necessary to produce it, and therefore, by all just reasoning, I am bound to call the variation of corn . . . a fall in [its] value, and not a rise in the value of the things with which [it is] compared' (11, p. 19). The reference to 'labour and capital' in the first paragraph might suggest something other than 'pure' labour theory reasoning, but I think it is plain, in context, that capital was conceived as accumulated labour (a point Ricardo goes on to make in the third section of the new chapter). It is only much later in the chapter that 'capital' emerges as a more complex entity. The implication from the later section, 'On an invariable measure of value', discussed below, is that the question must be answered negatively. Cf. Marx's reading of this portion of Ricardo's text: 'The "variation" which we want to know about is: whether the "necessary labour-time''' has altered' ([1862-3] 1969, p. 170, emphasis in original). This is now the heading for the third section, Ricardo having created a new second section under the heading: 'Labour of different qualities differently rewarded. This no cause of variation in

214

INTERPRETING RICARDO

anticipated in Ricardo's letter to Mill of 28 December 1818 (see above, pp. 190-1). As Sraffa observed (1, p. xxxix), they were executed to avoid misunderstanding (for which Ricardo was undoubtedly culpable, especially with the old section heading). Nothing more need be read into them. The text is then largely unaltered until the 'fish-and-game' analysis, previously conducted on the assumption of identical 'capital structures' in hunting, fishing, and in the production of an 'invariable standard' (see above, pp. 160-1). But now, although fundamentally the same points are made, the analysis is simplified: discussion of the distinction between fixed and circulating capital is deferred until the next section, and the annuity calculations are purged from the text.20 The fourth section heading in the revised chapter was more general than the corresponding one in the second edition (see above, p. 181). It reads: The principle that the quantity of labour bestowed on the production of commodities regulates their relative value, considerably modified by the employment of machinery and other fixed and durable capital. (1, p. 30, emphasis in original)

It is towards the beginning of this section that the distinction between fixed and circulating capital is now to be found, together with the elaboration of ways in which 'capital structures' may differ: unequal ratios of fixed to circulating capital, differences in the durability of fixed capital, and differences in the durability of circulating capital. It is the first of these sources of difference which Ricardo initially explicated. He did so within the newly adopted 'labour profile' framework, previously explained in his letter to McCulloch of 13 June 1820 (see above, pp. 196—7). Three capitalists are each assumed to employ one hundred labourers for a year, two of them in the construction of machines, the third to cultivate corn. In the following year, the capitalist farmer replicates the first year's activities while the others

20

the relative value of commodities" (i, p. 20, emphasis in original). This second section encompassed the discussion of the relative wages scale: see above, pp. 157-8. A problem with the annuity method was that the monetary present value of fixed capital was treated as invariant to changes in distribution. But, as Ricardo had evidently come to realise, items of fixed capital are themselves susceptible to price changes, reflecting their own 'labour profiles'. As he later wrote to McCulloch: 'capitals are not the same in kind what will employ one set of workmen, is not precisely the same as will employ another set, and if they themselves are produced in unequal times they are subject to the same fluctuations as other commodities' (21 August 1823, ix, p. 360).

The labour theory of value (II)

215

utilise their machines, each assisted by a further hundred labourers, to produce their separate outputs of cloth and cotton goods. Ricardo takes up the analysis: During the second year they will all have employed the same quantity of labour, but the goods and machine together of the clothier, and also of the cotton manufacturer, will be the result ... of the labour of one hundred men for two years; whereas the corn will be produced by the labour of one hundred men for one year, consequently if the corn be of the value of 400/. the machine and cloth of the clothier together, ought to be of the value of 1000/. and the machine and cotton goods ... ought to be also of twice the value of the corn. (1, p. 33) The (produceplus machine) outputs 'ought' to be 'twice the value of the corn5 only if a 'pure' labour theory is to hold. But this is not the case: 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. On account then of the different degrees of durability of their capitals, or, which is the same thing, on account of the time which must elapse before one set of commodities can be brought to market, they will be valuable, not exactly in proportion to the quantity of labour bestowed on them, - they will not be as two to one, but something more, to compensate for the greater length of time which must elapse before the most valuable can be brought to market. (1, pp. 33-4) This is so because profit is calculated on the clothier's and cotton manufacturer's capital outlays (on the construction of machines) at the end of the first year, and is calculated again on the capital—plusprofit sum at the end of the second year. The first year's profit is compounded. Having numerically illustrated the effect of compounding on static exchange relationships, Ricardo turned to the consequences for relative and absolute prices of a 'rise in the value of labour', and associated reduction in the profit rate (1, p. 34). Implicitly assuming a numeraire commodity produced under the same circumstances as corn, a 1 per cent reduction in the rate of profit is shown to lower the absolute prices of cloth and cotton goods by a little over 1 per cent.21 More generally: 21

Ricardo calculated the selling price of the cloth and cotton goods alone according to the (implicit) formula: px = L.2w (i +r) + Lxwr (i +r), where: /,,, L2, stand for the number of labourers employed in the first and second years respectively, and w and r are the second year's wage and profit rates. The labour expended on the machine does not 'count' towards that expended on the cloth or cotton goods. It is as if the machine lasts forever.

2l6

INTERPRETING RICARDO

All commodities which are produced by very valuable machinery, or in very valuable buildings, or which require a great length of time before they can be brought to market, would fall in relative value, while all those which were chiefly produced by labour, or which would be speedily brought to market would rise in relative value, (i, p. 35) He presumably meant that the commodities fall relative to the (assumed invariable) value of the standard, with the 'rising' commodities including, for instance, Malthus's half ounce of silver. A commodity produced by unassisted labour in a year was not an 'extreme' case, as he had been made uncomfortably aware. Ricardo had again shown that when capital structures (or 'labour profiles') differ, prices variations are not confined to those reflecting changed labour expenditures. But now, having extirpated passages which might give the contrary impression (noted below, pp. 218-19) he hastened to add: The reader ... should remark, that this cause of the variation of commodities is comparatively slight in its effects. With such a rise of wages as should occasion a fall of one per cent, in profits, goods produced under the circumstances I have supposed, vary in relative value only one per cent... The greatest effects which could be produced on the relative prices of these goods from a rise of wages, could not exceed 6 or 7 per cent.; for profits could not, probably, under any circumstances, admit of a greater general and permanent depression than to that amount. (1, p. 36) He continued: Not so with the other great cause of the variation in the value of commodities, namely, the increase or diminution in the quantity of labour necessary to produce them ... An alteration in the permanent rate of profits, to any great amount, is the effect of causes which do not operate but in the course of years; whereas alterations in the quantity of labour necessary to produce commodities, are of daily occurrence ... In estimating, then, the causes of the variations in the value of commodities, although it would be wrong wholly to omit the consideration of the effect produced by a rise or fall of labour, it would be equally incorrect to attach much importance to it; and consequently, in the subsequent part of this work, though I shall occasionally refer to this cause of variation, I shall consider all the great variations which take place in the relative value of commodities to be produced by the greater or less quantity of labour which may be required from time to time to produce them. (1, pp. 36-7) Here, for the first time in any edition of the Principles, there is a nominally 'empirical' defence for using the labour theory.22 22

According to the obliterated numerical examples in the first and second editions, which appeared at a corresponding place in the text, an assumed 6 per cent change in profitability

The labour theory of value (II)

217

Whether it was a convincing defence is another matter entirely. Even supposing that 'permanent' variations in profitability are empirically constrained within a band of six or seven percentage points, it follows that relative price movements are similarly limited only because of an implicit assumption about the magnitude of differences in 'labour profiles'. Not a shred of empirical evidence was adduced in support of that assumption.23 Moreover, the putative fact that large permanent variations in profitability take many years is neither here nor there when one considers that a major function of the labour theory was to facilitate the demonstration of how such variations might arise (as in the chapter 'On Profits'). This was not Ricardo at his best. Having newly considered the case of differences in the ratio of fixed to circulating capital within the 'labour profile' framework, Ricardo turned to commodities which 'cannot be brought to market in the same time' (1, p. 37).24 One capitalist is assumed to employ twenty labourers for two consecutive years in the production of commodities, while another employs forty men for one year. Because of accumulated profits, the monetary value of their outputs will differ, even though labour expenditures are the same. Ricardo added: This case appears to differ from the last, but is, in fact, the same. In both cases the superior price of one commodity is owing to the greater length of time which must elapse before it can be brought to market. In the former case the machinery and cloth were more than double the value of the corn, although only double the quantity of labour was bestowed on them. In the second case, one commodity is more valuable than the other, although no more labour was employed on its production. The difference in value arises in both cases from the profits being accumulated as capital, and is only a just compensation for the time that the profits were withheld. (1, p. 37) The case is indeed the same. Given Ricardo's valuation of the commodityjfr/zzj machine 'output', the implicit price formulae for the two cases are identical.25

23

24

25

gave a maximum 7.29 per cent price variation (1, pp. 57-8). At first blush, the contrast with the third edition examples may not appear so great. But the previous examples applied to the 'early stage' occupations of hunting and fishing. Differences in capital structure allegedly became more pronounced in the later stages of society (1, pp. 62 n. 1, 66). According to Barkai (1967), b u t disputed by Wilson a n d Pate (1968) R i c a r d o did have a n empirical w a r r a n t for claiming that profit rates were unlikely to ' p e r m a n e n t l y ' vary by more than 6 or 7 p e r cent. But they a r e u n a n i m o u s that R i c a r d o h a d n o such justification for his implicit position on the extent of differences in 'labour profiles'. This corresponds to the case of 'differences in the durability of circulating capital', introduced in the second edition: see above, p p . 182-3. T h e y a r e of t h e form p = L,2w(i +r) + L1w(i + r ) 2 for t h e 'longer' processes, as against p = L.2w{ 1 + r) for the 'shorter' processes. T h e n o t a t i o n was explained above (p. 215 n. 21).

2l8

INTERPRETING RIGARDO

I come to the fifth section of the rewritten chapter, in which Ricardo recast his analysis of differences in fixed capital durability and, more generally, did his best to minimise the significance of 'modifications' to the labour theory.26 The 'unassisted machinery' analysis (see above, p. 162-3) w a s a prime casualty. Partly, it was no longer acceptable because of its use of annuity formulae, but, more importantly, it was quite out of place given the chapter's new orientation. Far from wishing to lavish attention on, and magnify the effect of, distribution-induced price movements, for which the previous analysis had been a conduit, Ricardo's new tactic was apparently to cover the analytical terrain as swiftly and unceremoniously as he could decently manage. The analytical treatment of fixed capital was newly applied thus: Iffixedcapital be not of a durable nature, it will require a great quantity of labour annually to keep it in its original state of efficiency; but the labour so bestowed may be considered as really expended on the commodity manufactured ... If I had a machine worth 20,000/. ... and if the wear and tear of such machine were of trifling amount, and the general rate of profit 10 per cent., I should not require much more than 2000/. to be added to the price of the goods, on account of the employment of my machine; but if the wear and tear of the machine were great, if the quantity of labour requisite to keep it in an efficient state were that of fifty men annually, I should require an additional price for my goods, equal to that which would be obtained by any other manufacturer who employed fifty men in the production of other goods, and who used no machinery at all. (1, p. 39) Whether Ricardo consciously intended to bamboozle his readers cannot be known, but the amount that is 'added to the price of the goods, on account of the employment of my machine' depends entirely on the length of time that the machinery has been in use: if for five years, that sum would be nearly ^3,ooo. 27 And of course, the longer some machines last, ceteris paribus, the greater the 'modification' to the 'pure' labour theory. The dedicated attempt to detract from the significance of the 'modifications' is again conspicuous at the close of the section. After the introduction of machinery, Ricardo claimed, 'the commodities 26

27

T h e section heading, previously the fourth, was allowed to stand from the second edition: see above, p. 182. According to the logic of Ricardo's earlier analysis, the amount 'to be added to the price of the goods, on account of the employment of my machine' is L,w (1 +r) + Xr{\ +r)'~ ', where L, denotes the labour expended on the upkeep of the machine in period / (with /= 1 being the machine's first period of use) and, as shorthand, X stands for the machine's present value: this is itself dependent on the machine's 'labour profile'.

The labour theory of value (II)

219

produced by the employment of equal capitals will be of very unequal value' (1, p. 42). So far, this corresponds to the wording in edition two (1, p. 62 n. 1; cited above, p. 184). But now he continued, with the insertion of the words italicised by me, 'and although they will still be liable to rise or fall relatively to each other, as more or less labour becomes necessary to their production, they will be subject to another, though a minor variation, also, from the rise or fall of wages and profits' (1, p. 42). And then, with a few deft strokes of his pen, 'goods which sell for 5,000/ [2,000/ in editions one and two] may be the produce of a capital equal in amount to that from which are produced other goods which sell for 10,000/.' The 'commodities produced by the employment of equal capitals' are not so 'very unequal' in their value as they used to be. I now come to the important new sixth section of the chapter, 'On an invariable measure ofvalue\ which opened: When commodities varied in relative value, it would be desirable to have the means of ascertaining which of them fell and which rose in real value, and this could be effected only by comparing them one after another with some invariable standard measure of value, which should itself be subject to none of the fluctuations to which other commodities are exposed. (1, P. 43) This is the same problem which Ricardo had discussed earlier in the rewritten chapter, when he set the exercise: 'Two commodities vary in relative value, and we wish to know in which the variation has really taken place' (1, p. 17). The introduction of the 'invariable standard' did not, however, provide a failsafe means of ascertaining changes in 'real value'. Even supposing that a commodity (gold) always required the same labour expenditure in its production, 'still gold would not be a perfect measure of value' because of differences in capital structures ('labour profiles') (1, pp. 44-5). He continued: [Gold] would be a perfect measure of value for all things produced under the same circumstances precisely as itself, but for no others. If, for example, it were produced under the same circumstances as we have supposed necessary to produce cloth and cotton goods, it would be a perfect measure of value for those things, but not so for corn, for coals, and other commodities produced with [different 'labour profiles'] ... because ... every alteration in the permanent rate of profits would have some effect on the relative value of all these goods, independently of any alteration in the quantity of labour employed on their production. (1, p. 45)

220

INTERPRETING RIGARDO

By implication, changes in 'real value' are the same as 'variations [which have] really taken place': changes wrought by altered productive labour expenditures (cf. Marx [1862-3] l9^9> P- 202). And measured by a 'perfect' or 'accurate' standard, commodities should change in price only as a result of their altered 'labour contents'. Furthermore, as in the correspondence, distribution-induced price movements were manifestations of a defective standard. They did not correspond to changes in 'real value', as associated with labour expenditures. We know from his correspondence that Ricardo 'despaired' at his inability to fix a 'perfect' standard (see above, p. 211). But he had also alleged that a 'medium' commodity would achieve the desired result for a 'general mass' of commodities. This now received ex cathedra enunciation: May not gold be considered as a commodity produced with such proportions of the two kinds of capital as approach nearest to the average quantity employed in the production of most commodities? May not these proportions be so nearly equally distant from the two extremes, the one where little fixed capital is used, the other where little labour is employed, as to form a just mean between them? (1, pp. 45-6) The 'advantage' of making this supposition was, he continued, 'that I shall be enabled to speak of the variations of other things, without embarrassing myself on every occasion with the consideration of the possible alteration in the value of the medium' (1, p. 46), again implying that with distribution-induced price changes, it is the medium which has failed to reveal the 'true picture'. If, on the other hand, the medium is invariable, 'all alterations in price [are] occasioned by some alteration in the value of the commodity of which I may be speaking' (1, p. 46). They are all 'occasioned' by variations in productive labour expenditures. Unlike his position in the first edition, Ricardo was far from desiring a standard which would maximise price variations (reductions, in the earlier editions) following a change in distribution. Such an 'extreme' standard had been part and parcel of the 'curious effect', itself commended as being 'of such importance to the science of political economy' (1, p. 61). But now, as Jacob Hollander astutely remarked, 'the compatibility of higher wages and lower prices was relegated to an incidental statement' (Hollander 1904). This was effected in the new section's penultimate paragraph: Before I quit this subject, it may be proper to observe, that Adam Smith, and all the writers who have followed him, have ... maintained that a rise

The labour theory of value (II)

221

in the price of labour would be uniformly followed by a rise in the price 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. (1, p. 46) But what he had really hoped to succeed in showing was that there would be scarcely any such price movements and, with a 'perfect' standard, none at all. This was in marked contrast to the 'unassisted labour' standard in the first and second editions. In the final paragraph of his newly written section, Ricardo addressed the Malthusian charge that he had confounded 'cost of production' with value (see above, p. 205). He explained: I have not said, because one commodity has so much labour bestowed upon it as will cost 1000/. and another so much as will cost 2000/. that therefore one would be of the value of 1000/. and the other of the value of 2000/. but I have said that their value will be to each other as two to one . . . It is of no importance to the truth of this doctrine, whether one of these commodities sells for 1,100/. and the other for 2,200/., or one for 1,500/. and the other for 3000/. . . . I affirm only, that their relative values will be governed by the relative quantities of labour bestowed on their production. (1, pp. 46-7) He added in an accompanying footnote: Mr. Malthus appears to think that it is a part of my doctrine, that the cost and value of a thing should be the same; - it is, if he means by cost, 'cost of production' including profits. . . . [But] this is what he does not mean, and therefore he has not clearly understood me. (1, p. 47n.) In the first of these passages, Ricardo was again switching between relative quantities of labour expenditures and relative wage bills as the determinant of exchange ratios: an excusable practice if, and only if, a 'pure' labour theory holds. The same proviso applies to the footnote. It was unproblematical for Ricardo to accept the identity of cost with value only within the assumed context provided by a 'pure' labour theory. The final noteworthy alteration to the chapter was the substitution of a 'micro' method of treating distribution for the one previously framed at the level of social aggregates. The amended passage read: It is according to the division of the whole produce of the land of any particular farm [replacing 'of the whole produce of the land and labour of the country'], between the three classes of landlord, capitalist and

222

INTERPRETING RICARDO

labourer, that we are to judge of the rise or fall of rent, profit, and wages (i, P- 49).28 This change had been anticipated in Ricardo's Notes on Malthus (see above, pp. 202—3) and was mirrored by changes elsewhere in the new edition (i, pp. 402-3). The revised chapter: a discussion

The changes to 'On Value' in the third edition were evidently motivated by the ambition to justify the 'pure' labour theory in the light of the discussion of differences in capital structure: a preoccupation which cannot be read even between the lines of previous versions of the chapter. Two converging strategies had been newly adopted. One was the specification of the 'medium' commodity as an invariable standard which would restrict price changes, or as many as possible, to those reflecting altered labour expenditures, specifically as a way of sanctioning the use of the 'pure' labour theory in later chapters. Ideally, however, there would be no distributioninduced price changes, regardless of differences in capital structures. But the discovery of the 'perfect' standard, which would achieve this result, had eluded Ricardo, not surprisingly. There is no counterpart to the new standard in earlier editions of the Principles, simply because Ricardo had not directly posed the question of whether there could be a 'perfect' standard when due allowance is made for differences in capital structure. He had, of course, mentioned 'invariable standards' which would 'indicate correctly' changes in the labour expended on (bundles of) commodities, but always against the background of a 'pure' labour theory in which 'capital structures' did not explicitly figure at all. What he now wanted was the same result when they did. Such appears to have been the thinking behind Ricardo's new section on the invariable measure. As for Sraffa's hint that Ricardo may have felt satisfied with the 'medium' standard to the extent that, when adopted as numeraire, 'the average price of all commodities, and their aggregate value, would remain unaffected by a rise or 28

The new heading for the section containing this passage was: 'Different effects from the alteration in the value of money, the medium in which PRICE is always expressed, or from the alteration in the value of the commodities which money purchases' (i, p . 47, emphasis a n d capitals in original).

The labour theory of value (II)

223

fall of wages' (1, p. xlv), I cannot agree. First, all distributioninduced price changes were taken to be indicative of a 'failure5 in the standard. Secondly, Ricardo had made clear, both in the rewritten chapter and elsewhere, that his task was to explain (changes in) distribution at the level of the individual farm and not, as a prime analytical objective, at the 'macro' level of social aggregates. And thirdly, there is no evidence of Ricardo having formulated a necessary condition for the 'balancing solution' to work, namely that fewer commodities 'fall' than 'rise' following a reduction in profitability. The 'medium' standard was, then, an imperfect means of justifying the 'pure' labour theory (cf. Ong 1983 and Tosato 1985). But its benefits, such as they were, derived from the 'empirical' assertion that the production processes for most commodities had roughly similar 'labour profiles'. This brings me to Ricardo's other line of 'defence': the allegations of empirical relevance. The various 'empirical' claims that were made by Ricardo on behalf of the labour theory amounted to little more than plucking 'appropriate' figures from the air. But their significance, as with the debut for the 'medium' standard, lies in their very introduction, reinforcing the suspicion that Ricardo was, for the first time, fully conscious of the need to reconcile his own analysis of divergent capital structures with his heavy reliance on 'pure' labour theory reasoning. But why, having made such bold 'empirical' claims, did he proceed to put forward a second 'defence' of the 'pure' labour theory in the guise of the new standard? The answer may be that the 'standard' defence was preferred by Ricardo but, unable to find the chimerically 'perfect' candidate, he was panicked into making the wild 'empirical' assertions.29 He had, as it were, doubly indemnified the labour theory. In short, the textual changes were profoundly significant. I therefore cannot agree with Sraffa's view that 'the theory [of value] of edition 3 appears to be the same, in essence and in emphasis, as that of edition 1' (1, p. xxxviii); and even less so with the 'old' argument, an inversion of the truth, that Ricardo was 'retreating' from the labour theory. The stimulus for the textual changes had come from Malthus. But contrary to Sraffa (1, pp. xliii—xliv, and Hollander 1979, 29

In the later writings, when the dust had settled, Ricardo was prepared to make a far greater allowance for differences in capital structure (see below, p. 234).

224

INTERPRETING RICARDO

pp. 219-20), the force of Malthus's critique greatly exceeded the single 'correction' that unless the standard was truly 'extreme', there would be price increases as well as reductions following a fall in profitability. It was the resounding dissonance between Ricardo's analysis of divergent capital structures (incorporating the 'curious effect') and his use of the 'pure' labour theory, sometimes predicated on a blandly stipulated 'invariable standard', that Mai thus had exposed. He had forced Ricardo to attempt an intellectual synthesis which was long overdue. And when that happened, Ricardo knew his priority: to 'save' the labour theory. Why was the theory so important to him? Was it, perhaps, because he regarded it as a sound empirical approximation? Such claims were made, but the direction of causality seems to have run from Ricardo's labour theory commitment to the assertions of'relevance', not vice versa. As to the commitment itself, I have suggested that Ricardo was attracted, first, by the suitability of the labour theory to his analytical requirements - particularly to his demonstration of 'permanently' falling profitability in consequence of diminishing agricultural returns - and, secondly, as a related point, to the theory's 'meaningful' conception of 'difficulty or facility of production' in terms of human labour expenditure. In the final writings, which I consider next, there developed an even more powerful attraction, for it seemed to Ricardo with increasing force that 'value' should mean 'expenditure of labour'. This might be termed a 'philosophical' attraction. THE FINAL WRITINGS

In his letter to McCulloch of 25 January 1821 (discussed above, pp. 210-11), Ricardo had expressed despair at not knowing where to 'fix my standard'. Two years later, in his unfinished papers on Absolute Value and Exchangeable Value (AVEV),

he was resigned to

defending his 'medium' commodity as the least imperfect standard attainable. To that extent, his final position was unchanged from the one expressed in the third edition of the Principles, but this should not detract from the importance of the later writings. Stimulated by criticism from the anonymous author of Observations on Certain Verbal Disputes in Political Economy (1821) (hereafter abbreviated to Observa-

tions), from his friend Hutches Trower, and from Mai thus; and dissatisfied with the dubious help offered to him by James Mill and

The labour theory of value (II)

225

McCulloch, Ricardo attempted to think through 'the most difficult question in Political Economy5, namely the meaning and measurement of'value 5 (letter to Malthus, 3 August 1823, IX> P- 325)- This period of his recorded intellectual development sheds much light on his multifaceted treatment of'value 5 . The later writings are, however, extremely difficult to follow, partly in view of the complexity of the issues raised, partly because of repetition. I therefore adopt a thematic arrangement, beginning with the concept of measurement. The perfect measure: a scientific analogy

The template for discussing the notion of a measure of value had been set by Adam Smith: But as a measure of quantity, such as the natural foot, fathom, or handful, which is continually varying in its own quantity, can never be an accurate measure of the quantity of other things; so a commodity which is itself continually varying in its own value, can never be an accurate measure of the value of other commodities. ([1776] 1981, p. 50) The use of this type of language in Ricardo5s later writings is exemplified by the opening paragraph to the first draft of AVEV: The only qualities necessary to make a measure of value a perfect one are, that it should itself have value, and that that value should be itself invariable, in the same manner as in a perfect measure of length the measure should have length and that length should be neither liable to be increased or diminished; or in a measure of weight that it should have weight and that such weight should be constant, (iv, p. 361)30 The analogy with scientific measurement is revealing. The question of a perfect measure of length (from which other measures, such as weight and volume, would be derived) was of topical interest during Ricardo5s lifetime, with articles on the subject appearing in journals with which he was conversant (the Edinburgh Review and British Review). I will give a summary of an article rather prosaically entitled 'Weights and Measures5, published in the February 1817 edition of the British Review, which 30

The same discourse had been used in Ricardo's letter to Broadly of 14 June 1816 (vn, pp. 42-3), but only with reference to the practical obstacles in the way of specifying a 'measure'. From 1820 onwards, the enquiry turned on theoretical possibilities.

226

INTERPRETING RICARDO

provides both an overview of the points seemingly at issue and a taste of the language used. The anonymous writer began the article by detailing the diversities and anomalies in standards of measurement adopted from Greek and Roman times to his own, his aim being to 'consider by what means these and other such anomalies may best be corrected'. It has 'been long agreed by theorists', he reported, 'that the basis [of measurement] should be the standard of length* (emphasis in original). The 'grand question', therefore, was to determine an 'invariable' unit measure of length 'that shall be founded upon nature'. Or, in short, to determine an 'invariable standard'. The two main contenders for this 'invariable standard' were 'the length of a degree of a meridian in a given latitude' and 'the length of a pendulum that shall vibrate in a given interval, in a given latitude'. Many had been 'seduced' by the former, the writer opined, rather giving away his own predilection. The postrevolution French National Assembly had proposed the 'meridian' approach which, having been endorsed by members of the Paris Academy of Science (including Lagrange, Laplace, Lalande and Borda), was adopted in principle. The astronomers Machain and Delambre were then appointed to 'conduct the grand geodesic operations' and, their work completed, the metric system was introduced in August 1793. Our writer was unimpressed. The 'deduction of a system of measures from the pendulum is much more simple and natural' than the 'arc of a meridian' method for 'determining an invariable standard', he declared, and it should be given preference by the British government (which had set up a Select Committee on the matter in 1814, but without reaching a positive outcome). The foundation of the 'invariable standard' in nature was thought to be essential. As expressed in another article (Kater 1818), published in the Edinburgh Review, 'without reference to some natural object that continues always of the same dimensions', systems of weights and measures give rise to 'perplexity', 'ambiguous language', 'endless frauds', and, more generally, a 'perishable character . . . [to] all our knowledge concerning the magnitude and weight of bodies, and the impossibility, by a description in words, of giving to posterity any precise information on these subjects'. With a natural 'invariable standard', however, the problems of providing intra- and inter-temporal measurements are solved, for in principle the standard measure is always (re)discoverable.

The labour theory of value (II)

227

Ricardo's acquaintance with this (or similar) material seems indisputable (which, apart from anything else, is explicable in terms of his long-standing interest in scientific subjects: see ix, pp. 34-5). Again from AVEV: There can be no unerring measure either of length, of weight, [or] of time ... unless there be some object in nature to which the standard itself can be referred and by which we are enabled to ascertain whether it preserves its character of invariability, for it is evident on the slightest consideration that nothing can be a measure which is not itself invariable. If we have any doubts respecting the uniformity of our measure of length, the foot, for example, we can refer it to the portion of the arc of the meridian, or to the vibrations of the Pendulum under given circumstances, (iv, p. 401) The same should be true for the invariable standard of value which, perforce, would provide a benchmark for both intra- and intertemporal value comparisons.31 It may be inferred that the form in which Ricardo posed the question of the 'invariable standard' was influenced by the prevailing scientific discourse, which may have predisposed him to ask certain questions in certain ways, or reinforced a belief that the questions he was asking were the correct ones. His problem was in answering them.32 'Value' and labouring activity

Soon after the publication of the third edition of the Principles, Ricardo's use of 'value' attracted similar criticism from Hutches Trower and the anonymous author of the Observations. According to Trower, 'the only proper use of the term value is in exchange', whereas Ricardo had used 'value' to 'signify the quantity of labor necessary to acquire or produce a commodity' (24 June 1821, vm, p. 394, emphasis in original). Likewise, the author of the Observations charged that Ricardo had sometimes 'departed from his original use of the term value [qua exchangeable value]' and had 'made of it 31 32

Set AVEV, p p . 3 8 0 - 1 , 397, 4 0 1 - 2 . Samuel Bailey complained bitterly of the falseness of the 'scientific' analogy; 'All that is practicable [in measuring value] appears to be simply this: if I know the value of A in relation to C, I c a n tell the value of A a n d C in relation to each other, a n d consequently their comparative power in purchasing all other commodities. This is an operation obviously bearing no resemblance at all to the process of measuring length' ([Bailey] 1825, p. 96). Granting Bailey his purely relative concept of 'value', the point he makes is evidently correct.

228

INTERPRETING RIGARDO

something absolute' as a function of the labour expended on individual commodities (1821, p. 15). 33 Both critics had a textual basis for their complaints. Ricardo's first replied to Trower thus: I do not, I think, say that the labour expended on a commodity is a measure of its exchangeable value, but of its positive value. I then add that exchangeable value is regulated by positive value, and therefore is regulated by the quantity of labour expended. (4 July 1821, ix, pp. 1-2)34 Seven weeks later, now substituting 'real value' for 'positive value', he endeavoured to explain himself more fully: 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 . . . 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. The exchangeable value of a commodity cannot alter, I say, unless either its real value, or the real value of the things it is exchanged for alter. This cannot be disputed. If a coat would purchase 4 hats and will afterwards purchase 5, I admit that both the coat and the hats have varied in exchangeable value, but they have done so in consequence of one or other of them varying in real value, and therefore if I use the word value without prefixing the word exchangeable to it, it will be correct for me to say that the coat has risen in value whilst hats have not varied, or that hats have fallen in value while coats have remained stationary. (22 August 1821, ix, p. 38) It is of the first importance to note that Ricardo fully accepted the criticisms of his treatment of 'value' in the Principles. He had, sometimes at least, used 'value' in an absolute sense to signify the labour expended on a commodity. However, the proposition that exchangeable value only varies with an alteration in the 'real value' of a commodity obviously can be, and had been, disputed. Ricardo had disproved it himself in the very first edition of his Principles. Furthermore, to leave it for his readers to divine when he was using 'value' in the sense of 'real value' was not especially helpful. These shortcomings aside, Ricardo had unambiguously disclosed his deep33 34

T h i s w a s ' w r o n g ' b e c a u s e , i n t h e w r i t e r ' s o p i n i o n , all v a l u e is r e l a t i v e ( 1 8 2 1 , p . 10). He went on to claim that commodities would possess 'positive value' even if they were not the objects of exchange (ix, p. 2).

The labour theory of value (II)

229

seated preoccupation with 'real value' in the sense of productive labour expenditure. It was a preoccupation not shared by Malthus. In September 1821 he posed a 'knotty point' (ix, p. 63) which rested on his earlier 'half ounce of silver' reasoning. He wrote to Ricardo: No other supposition can render money of the same value at all times than that of supposing it be obtained always by the same quantity of labour without any capital. (25 September 1821, ix, p. 79, emphasis in original) But if such a commodity is adopted as numeraire, if 'labour profiles' differ, and if the rate of profit declines (for Malthus, because of the competition of capitals in the output sphere), all commodities produced with the aid of capital will fall in price. In Malthus's opinion, this meant that no 'commodity can with propriety be considered as of a fixed value . . . which while it continues to require the same advances of labour and capital for its production, is obtained at different periods, at a very different rate of profits' (13 September 1821, ix, p. 65). This was the 'knotty point'. Ricardo had been tangled before. He snapped: Nothing is to me so little important as the fall and rise of commodities in money, the great enquiries on which to fix our attention are the rise or fall of corn, labour, and commodities in real value, that is to say the increase or diminution of the quantity of labour necessary to raise corn, and to manufacture commodities. It may be curious to develop the effect of an alteration of real value on money price, but mankind are only really interested in making labour productive, in the enjoyment of abundance, and in a good distribution of the produce obtained by capital and industry. I cannot help thinking that in your speculations you suppose these much too closely connected with money price. (28 September 1821, ix, p. 83)35 This passage alone should dispel any possible doubt over the centrality of 'real value', alias quantity of labour expenditure, in Ricardo's conceptual schema. Indeed, 'real value' was now so fundamental as to be split off and privileged over the 'independent' behaviour of money prices. One might add, however, that the curiosity which Ricardo was denigrating had once captured his own imagination with the proudly announced 'curious effect'. How far he had travelled since then. In 1823 Malthus published his Measure of Value, advocating a 35

Cf. Ricardo's next letter to Malthus: 'I say again that too much importance is attached to money - facility of production is the great and interesting point' (11 October 1821, ix, p.

100).

23O

INTERPRETING RIGARDO

'labour commanded' measure of value equivalent to using as numeraire a commodity produced by labour alone in the shortest possible time (such as the half ounce of silver).36 The pamphlet was the subject of some animated discussion. Ricardo reported to Trower: [McCulloch] attended the last meeting of our Political Economy Club and the result of our discussions on that day convinced him, as we all had been long before convinced, that the progress of the science is very much impeded by the contrary ideas which men attach to the word value. . . . I mean . . . to turn my thoughts to the subject. (24 July 1823, ix, pp. 312-13) Soon afterwards he wrote to Malthus: We are enquiring about the meaning which should be attached to the words 'increase of value' 'diminution of value'. You tell me that increase of value means an increased power of commanding labour: I deny that this definition is a correct one, because I deny the invariability of the standard measure you have chosen. (3 August 1823, ix, p. 324)

But what he really denied was Malthus's use of'value'. Ricardo's difference with Malthus over the 'proper' meaning of 'value' is evident in a subsequent letter. Ricardo pronounced: I estimate value by the quantity of labour worked up in a commodity . . . I say [a commodity] is only dear when a great quantity [of labour] has been bestowed on its production. (15 August 1823, IX? P-

The very essence of (unprefixed) 'value' is conceived in terms of productive labour expenditure. This is 'value', so to say that something has more or less 'value' - to say that a commodity is dearer (or cheaper) - is to say that it has more (or less) labour expended on its production. From this standpoint, it was nonsensical to deem the value of labour (wages) to be invariable regardless of changes in the labour required to produce necessaries. Evidence of the same train of thought may be found in AVEV: I may be asked what I mean by the word value, and by what criterion I would judge whether a commodity had or had not changed its value. I answer, I know no other criterion of a thing being dear or cheap but by the sacrifices of labour made to obtain it. (iv, p. 397) 36

To reiterate, half an ounce of silver would always represent the money wage for, say, one day's labouring activity. Consequently, the silver prices of other goods will directly express their 'labour commanded'. For example, a commodity selling for two ounces can 'command' four days of labouring activity.

The labour theory of value (II)

231

Again, the essential content of the word 'value' - what it means to say a commodity is more or less valuable - is traced to productive labour expenditure.37 Although, in the passages cited, Ricardo's identification of'value5 and cognate terms with the quantity of expended labour was increasingly made regardless of exchange relationships, there seems little doubt that his ideal was a one-to-one relationship between (changes in) 'labour values' and (changes in) exchangeable values. Consider, in this light, the following extract from AVEV\ Every thing is originally purchased by labour - nothing that has value can be produced without it ... That the greater or less quantity of labour worked up in commodities can be the only cause of their alteration in value is completely made out as soon as we are agreed that all commodities are the produce of labour and would have no value but for the labour expended upon them, (iv, p. 397)38 In this passage, the 'labour value' concept is fused with exchangeable value, with the 'labour worked up in commodities' presented as the condition both for 'value' itself, in an absolute sense, and, unless Ricardo was uttering a tautology, for changes in the exchange relationships between commodities ('the cause of their alteration in value'). Now suppose, following Ricardo's wishful line of reasoning, that an invariable measure of value is required. His problems would be at an end. Thus: The average strength of a thousand or often thousand men it is asserted is always nearly the same, why then not make the labour of man the unit or standard measure of value? If we are in possession of any commodity which requires always the same quantity of labour to produce it that commodity must be of uniform value, and is eminently well qualified to measure the value of all other things ... If this test were adopted it has been said every commodity would be valuable according to the quantity of labour required to produce them, (iv, p. 401-2) 37

38

Cf. the following, also from AVEV: 'To me it appears a contradiction to say a thing has increased in natural value while it continues to be produced under precisely the same circumstances as before [i.e., with the same labour expenditure]' (iv, p. 375). Cf. t h e following passage, i n t r o d u c e d in t h e third e d i t i o n of t h e Principles: ' A franc is n o t a measure of value for any thing, but for a quantity of the same metal of which francs are made, unless francs, and the thing to be measured, can be referred to some other measure which is common to both. This, I think, they can be, for they are both the result of labour; and, therefore, labour is a common measure, by which their real as well as their relative value may be estimated' (1, p. 284). Marx commented on this, correctly in my opinion, that by 'real value, Ricardo . . . understands the commodity as the embodiment of a definite amount of labour-time' (Marx [1862-3] l912i P- 139, emphasis in original).

232

INTERPRETING RICARDO

We would have an invariable 'standard in nature to which we may refer for the correction of errors and deviations' (iv, p. 401), and thus the problem of specifying an invariable standard is resolved according to the accepted scientific procedures of the day. For Ricardo, however, there was another concept of'value', cost of production, and with it a host of problems. These are considered next. 'Value', cost of production and the 'modifications'

In November 1821, James Mill sent Ricardo the manuscript for his Elements of Political Economy. On the topic of 'value', as interpreted by Ricardo, Mill had laid down 'a general and positive rule with respect to quantity of labour realised in commodities] being the rule and measure of their exchangeable value'. Ricardo commented: The exceptions will be opposed to you as they have been to me. In page 76 there is a passage ending with these words 'without in the least affecting the truth of the previous proposition &c &c.['] If a watch and a common Jack altered in relative value without any more or less labour being required for the production of either of them, could we say that the proposition 'that quantity of labour determines exchangeable value' was universally true? What I call exceptions and modifications of the general rule you appear to me to say come under the general rule itself. (18 December 1821, ix, p. 127)

Mill's was not a solution that Ricardo could accept. Neither could he condone McCulloch's strategy for absorbing the modifications. Commenting on McCulloch's lecture notes, he wrote: You go a little farther than I go in estimating the value of commodities by the quantity of labour required to produce them: you appear to admit of no exception 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 . . . 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. (19 March 1822, ix, p. 178)

In saying that he was always willing to allow for the 'exceptions and qualifications', Ricardo was not being entirely accurate. In letters written during the spring and summer of 1821, for instance, he had not mentioned them at all; and they were entirely absent from his

The labour theory of value (II)

233

pamphlet On Protection to Agriculture, published in April 1822.39 But now, perhaps because he recoiled at the unsatisfactory attempts by his friends to explain away the difficulties, there was to be a far greater willingness on Ricardo's part to confront the exceptions more directly. As he later granted to Malthus, the conditions of producing commodities, their 'labour profiles', 'admit of infinite variety' (13 July 1823, IX> P- 3°4)- He continued, tellingly: I should indeed be wanting in candour if I refused to admit that my money measure would not measure the quantity of labour worked up in commodities ... [although it] will measure the quantity of labour and profits together of which commodities are composed. (13 July 1823, IX> P- 3°4) 'Quantity of labour' is identified with the wage component in cost of production which, in other correspondence, and in AVEV, is itself (sometimes) identified with the 'value' of commodities.40 This 'value' did not bear the desired equi-proportional relationship to 'real value'. The exception that now perplexed Ricardo the most derived from the passage of time alone. Having gently reprimanded McCulloch for not doing 'justice to the argument of our opponents', he confessed: I cannot get over the difficulty of the wine which is kept in a cellar for 3 or 4 years, or that of the oak tree, which perhaps originally had not 2 [shillings] expended on it in the way of labour, and yet comes to be worth £100. (8 August 1823, ix, PP- 33°-0 41 39

40

41

See the letters to Say, 8 May 1821, (vm, p. 380); and to Trower, 22 August 1821 (quoted above, p. 228). In Protection to Agriculture he wrote, for example, 'In all cases, the rise of wages, when general, diminishes profits, and does not raise the prices of commodities' (iv, p. 215); according to his presentation, neither does the rise in wages reduce any prices. See the following letters: to Malthus, 3 August 1823 ('The value of almost all commodities is made up of labour [wages] and profits', ix, p. 320); to Malthus, 15 August 1823 ('the value of most commodities is made up of labour [wages] and profits', ix, p. 350); to McCulloch, 21 August 1823 ('value is compounded of two elements wages and profit mixed up in all imaginable proportions', ix, p. 361); and to Mill, 5 September 1823 ('the value of commodities depends upon the value of wages and profits', ix, p. 387). For similar 'cost of production' uses of'value' in AVEV set iv, pp. 365, 372-3, 384, 389, 392, 402-3. Cf. 'We . . . acknowledge . . . that 2 [shillings], with its annual accumulations, at a compound rate of profit, will at last yield £100 . . . but I question the propriety of [you] calling these accumulated profits by the name of labour, and of saying that the commodity so worth a hundred pounds was valuable in proportion to the quantity of labour bestowed upon it. The tree which originally cost 2 [shillings] for labour and becomes in after-time of the value of £100, has never strictly more than 2 [shillings] worth of labour employed on it' (letter to McCulloch, 21 August 1821, ix, pp. 358-9).

234

INTERPRETING RICARDO

The case of the oak tree had apparently originated with Ricardo himself in his debate with Torrens, and had been thrown back at him by Malthus in 1820.42 At the later time, Ricardo had feigned an imperious disdain for this example, but now, after further reflection, the associated problems could not be so easily discounted. The difficulty he perceived was that, even without a change in distribution, the passage of time resulted in a cumulative divergence between cost of production 'value' and 'real value' (in the guise of wages paid). Substituting shrimps for wood, he later wrote to Malthus: Shrimps are worth £10 in my money - it becomes necessary, we will suppose, in order to improve the shrimps to keep them one year, when profits are 10 pc*, shrimps at the end of that time will be worth £11. They have gained a value of £1. (31 August 1823, IX> P- 3^0 This increase in 'value' is unrelated to an alteration in 'real value'. Turning to AVEV, the modifying influence of time on the cost of production 'value' of commodities such as oak trees and shrimps is given full and frank recognition (see, for example, iv, pp. 376—9). It is also striking that when discussing the likely magnitude of distribution-induced changes in cost of production 'values', Ricardo made a threefold greater allowance than he had in his third edition: a 1 per cent rise in wages (and corresponding fall in the profit rate) can bring about a fall in price of 3 per cent for some commodities and a one per cent rise in others (iv, p. 373).43 Assuming Ricardo still believed that profit rates were unlikely to vary 'permanently' by more than six or seven percentage points, we would have (following Stigler) a '58% labour theory'. The implication was that a 'measure of value' could never limit changes in cost of production 'value' to those reflecting changes in 'real value', yet this had been Ricardo's now forlorn aspiration. The specification of the standard, which I consider next, was an exercise in damage limitation. 42 43

See above, p . 209. These absolute price movements were expressed in terms of his ' m e d i u m ' standard, discussed in the following sub-section. Note that, yet again, R i c a r d o did not claim the existence of a larger class of'rising' t h a n 'falling' commodities, thus reinforcing the point that he was not aiming for the 'balancing solution' attributed to h i m by Srafta.

The labour theory of value (II)

235

The 'imperfect' standard

Four months having passed since the publication of Malthus's Measure of Value, all Ricardo had done was 'to convince myself more and more of the extreme difficulty of finding an unobjectionable measure of value'; 'tho' not without fault', he continued, he believed his third edition choice 'is the best' (to Malthus, 3 August 1823; IX> p. 325). And he wrote to Mill: It would be a real service to the science of Political Economy to shew what the real difficulties are in the way offindingany accurate measure of value. We can at last only come to an approximation of a correct measure of value, and that must be the best which is itself the least liable to variation. I believe we have hit upon the best, but something more should be said to prove it such. I wish I could perform that task, but a more able pen must undertake it. (7 August 1823, IX? P- 329) The 'more able pen' had turned out to be no less elusive than the perfect standard, and hence the task of showing the impossibility of 'perfection', and of proving that the 'medium' standard was the best on offer, fell to Ricardo himself. His arguments were rehearsed tirelessly in correspondence and repeated in AVEV. As he stated time and again, if all commodities were produced under identical circumstances with respect to their 'labour profiles', and if one commodity always required the same expenditure of labour on its production, that commodity would be a perfect standard. 44 Conversely, differences in 'labour profiles' meant that any commodity selected and assumed to have an unchanging labour input would be an accurate measure for those, and only those, commodities with the same 'labour profile' as itself.45 But what, exactly, did he mean by 'perfection' or 'accuracy'? There is no mystery about that. For example: If all commodities required a year's time before they were in a state to be brought to market, and required the continued labour of men to produce them during that time, then ... they would be valuable according to the number of men employed on their production ... [and] any commodity which continued uniformly to require the same quantity of labour would 44

45

See, for example, R i c a r d o ' s letters to M a l t h u s of 28 M a y 1823 (IX> P- 2 97~8) a n d 13 J u l y 1823 (ix, p p . 3 0 3 - 4 ) ; a n d passages occurring on t h e following pages of AVEV: iv, p p . 3 6 3 - 7 , 382, 386, 3 9 2 - 3 , 4 0 3 - 4 . See the following correspondence: to Malthus, 13 July 1823 (ix, pp. 303-4); to Malthus, 15 August 1823 (IX» P- 346); and to McCulloch, 21 August 1823 (IX> P- 361); and also the following pages from AVEV: iv, pp. 369, 372-3.

236

INTERPRETING RIGARDO

be an accurate measure of value. It would in fact be invariable, and the variations in the relative value of this commodity to others . . . would be owing to some alteration in the facility or difficulty of producing those other commodities, by which a less quantity of labour was employed on them. (AVEV, iv, p. 364, emphasis in original)

Measured by a perfect standard, the prices of commodities change if, and only if, there are changes in labour expenditures. However, with differences in 'labour profiles', universal perfection is ruled out: The difficulty ... under which we labour in finding a measure of value applicable to all commodities proceeds from the variety of circumstances under which commodities are actually produced. Some, such as shrimps and a few others, are the result of a few hours labour without any advances ... others such as cloth is the result not of labour only, but of advances made probably for a year before the finished commodity is brought to market. Others again such as wine ... is the result also of labour and advances but advances made for a much greater length of time than in the case of the cloth. (AVEV, iv, pp. 368-9) And following an assumed rise in wages and fall in the rate of profit: If we constitute a cloth the measure of value wine would fall in such a measure and shrimps would rise, and if we choose wine as the measure both cloth and shrimps would rise but in unequal degrees, (iv, p. 370) Hence, the 'circumstances of time for which advances are made are so various that it is impossible to find any one commodity which will be an unexceptionable measure' when distribution alone changes (iv, p. 370); for with an 'unexceptionable measure', there should be no price movements at all. Enter, once again, the 'medium' commodity, 'which may fairly be considered ... as agreeing more nearly with the circumstances under which the greater number of commodities are produced than any other which can be proposed' (iv, p. 372). Such a commodity, Ricardo added, would not be a 'perfectly correct measure' because it does not agree with the circumstances of producing all other commodities; consequently, there will be some distribution-induced price movements which, if 'the measure was perfect', would not occur (iv, p. 373). But as he had come to accept, this imperfection was inescapable. Allegedly, however, there were specific advantages in using the 'medium' standard to which Ricardo drew attention in AVEV. 'I confess I prefer a measure', he wrote, 'which will give some idea

The labour theory of value (II)

237

whether when labour varies as compared with commodities it is the value of labour which has undergone a change, or whether it is the commodity which rises or falls' (iv, pp. 371-2). With Malthus's labour commanded measure, he complained, 'labour' (the value of wages) is by definition invariable regardless of (changes in) both the circumstances of producing wage-goods, and the composition of the commodity wage bundle. But with the 'medium' measure: if labour became abundant or scarce, it [the value of wages] would, like all other things, rise or fall. If it became difficult, from the appropriation of land to agriculture, to obtain an additional supply of corn without the expenditure of more capital for each quarter obtained, corn would rise [in price], and nearly in proportion to the increased difficulty, (iv, p. 372) The approximate correspondence between changes in the price of corn and the amount of labour required to produce it was regarded as critical: Let us suppose money to be produced in precisely the same time as corn is produced, that would be the measure proposed by me ... The circumstance of this ['medium'] measure being produced in the same length of time as corn and most other vegetable food which forms by far the most valuable article of daily consumption would decide me in giving it a preference, (iv, pp. 405-6) Ricardo had not forsaken his concern with agricultural conditions of production. With corn and most other commodities allegedly characterised by roughly the same 'labour profile' as the standard's, Ricardo evidently thought he could justify using a 'pure' labour theory in his demonstration of the connections between changes in agricultural conditions of production, money wages, and 'permanent' movements in profitability. As in the chapter 'On Profits' from the first edition of the Principles onwards, the tactic was first to announce a standard, and then claim the consequent association between changes in labour expenditure per unit of output and changes in prices.46 However, the labour theory had become much more than a useful analytical device for adumbrating a central theorem on distribution. 'Value', together with cognate terms such as 'dearness' 46

This solution was extremely restrictive. Apart from the assumption that all wage-goods have the same labour profile as the standard's (allowing for a mixed commodity wage), it must also be assumed that the outputs of the measured commodities are the sole variables. Given Ricardo's focus - 'permanent' movements in profitability - this was tantamount to analysing a long-period phenomenon within a very short-period framework.

238

INTERPRETING RIGARDO

and 'cheapness', had on one level come to be identified with the expenditure of labour. But concurrently, Ricardo retained another concept of 'value' (and relative 'values') in the sense of (relative) production costs: 'labour' (wages) plus profits. From this angle, the job of the standard was to bring the two sets of'value' concepts into harmony, if only for their changes. Measured by the perfect standard, changes in cost of production 'values' would be wrought exclusively by changes in labour expenditure 'values' (sometimes 'real' or 'natural' or 'positive' values). The perfect measure of 'value' was therefore a measure of (changes in) labour 'value' alone, the problem being that the 'value' actually measured by the standard was of the cost of production variety. Thus, when Ricardo wrote in AVEV, with reference to the mass of commodities, that 'every thing to which . .. improvements [savings in labour expenditure] were applied would fall in value, and price and value would be synonymous' (iv, p. 373), he meant 'labour quantity' value. But when he wrote that there might be three commodities, with the same quantity of labour expended on each, selling for different 'values' (iv, p. 384) it was (relative) cost of production 'value' which he must have had in mind. Had the latter been his only concept of 'value', one commodity would be just as serviceable a 'measure' as any other. That was evidently not the case. CONCLUSION

Malthus is often depicted as Ricardo's inferior on matters of 'pure' theory, yet the force and perspicacity of Malthus's criticisms of his contemporary's treatment of value must surely lead us to reconsider that verdict. Malthus alone had recognised the faults in Ricardo's exposition, and he exposed them with consummate skill. If there was ever a time when Ricardo might have abandoned his labour theory, with good cause, it was it the wake of Malthus's criticisms. But a 'retreat' was not to be. The 'pure' labour theory had served Ricardo well, particularly in his demonstration of 'permanently' falling profitability as a result of diminishing agricultural returns, the very problematique out of which the theory had been born, and it was not to be abandoned lightly. Indeed, it was not to be abandoned at all, as the revisions to the chapter 'On Value' demonstrate. Ideally, Ricardo sought to defend the theory via the 'perfect' measure of value, in terms of which all natural price movements are

The labour theory of value (II)

239

restricted to those reflecting altered labour expenditures, but such a solution had escaped him. Hence the 'medium' standard, which was designed not to achieve Sraffa's 'balancing solution', of which there is no evidence in Ricardo's writings, but rather to sanction a constrained labour theory, applicable to an alleged general mass of commodities which, critically, included wage-goods. To further indemnify the labour theory, Ricardo also introduced the independently advanced 'empirical' assertions which, it must be said, reflect little credit on him. It is on the basis of those assertions that he has been attributed a merely 'empirical' labour theory: an interpretation which has been dismissed as superficial. The Marxian reading of Ricardo has fared somewhat better. Marx claimed that Ricardo's treatment of value in the third edition of the Principles was marred by serious confusion: in some places Ricardo used 'value', 'absolute value' or 'real value' to denote 'necessary labour time'; but in others 'value' signified (relative) cost of production. Furthermore, it was Marx's judgement that Ricardo occasionally and falsely identified his 'value' concepts. However, in the opinion of later authorities, including Marshall, Stigler and Steedman, this reading is wholly untenable: Ricardo's 'value' concept was unambiguously one of cost of production. It is certainly true, and acknowledged by all, that Ricardo had a 'value' concept in the sense of (relative) cost of production. But to claim that this was his only 'value' concept has been to perpetrate a one-sided distortion. Even before his embrace of the labour theory, Ricardo had used 'value' in a manner suggesting an 'absolute' entity associated with the facility or difficulty of producing an individual commodity. Once he had made the labour theory his own, that association was with the quantity of labour expended in production, at various times denoted by such terms as 'natural value', 'real value', 'positive value' and even, at the expense of great confusion, 'value' itself. This 'labour quantity' species of'value' both existed in Ricardo's schema and came to be invested with unique significance, especially in the later writings. Perhaps it is appropriate to remind ourselves that soon after the publication of the third edition of the Principles, on which Marx was commentating, Ricardo had written to Trower: 'In speaking of exchangeable value you have not any idea of real value in your mind, I invariably have.' Much needlessly spilt ink might have been saved if Ricardo had made a similar announcement in his Principles.

24O

INTERPRETING RICARDO

To an extent, therefore, Marx's interpretation is superior to those of his critics: Ricardo sometimes did use 'value' in a manner associated with quantity of labour expenditure (as previous commentors, including Trower, the author of the Observations, Samuel Bailey and, in his Unsettled Questions, John Stuart Mill, had observed). But where Marx went astray was in claiming that 'cost of production' was at times identified with labour expenditure. As we have seen, Ricardo's practice was actually to 'measure' quantity of labour by wages alone, whereas 'cost of production' covered both wages and profits. Ricardo did not 'falsely identif[y] . . . cost-price with real value' (Marx [1862-3] 1969, p. 239). However, if one wished to be unfashionably charitable to Marx, one might add that Ricardo's tendency to associate changes in the two 'values' could, and did to Malthus, suggest that 'cost of production' was (somehow) reduced to labour expenditure. This would be an erroneous reading, but not a totally absurd one. Marx was also miskaken when he suggested that Ricardo thought his 'real' or 'labour values' were affected by a change in distribution (although Marx himself cast doubt on this particular allegation). And there are several further instances where Marx's reading of particular passages is suspect. (See, for example, Marx's interpretation of passages occurring in 1, pp. 385 and 85, discussed by Marx [1862-3] 1969 on pp. 215 and 329 respectively.) I am therefore far from claiming that Marx's interpretation is faultless. But it does not deserve the scorn that some eminent commentators have heaped upon it.

CHAPTER 6

The appropriation of Ricardo

Many historians of economic analysis have been preoccupied with locating Ricardo's work within an intellectual tradition. In recent times, the debate surrounding Ricardo's 'placement' has centred on the question of whether he belongs with the 'neoclassical' or, alternatively, with the 'Sraffians'. This chapter focuses on the main interpretative issues involved in this ongoing debate. The sections deal, in turn, with the question of whether Ricardo could be said to have anticipated significant elements of a 'subjective' theory of consumer demand; his position on the influence of demand, more broadly conceived; the role accorded by him to the forces of 'supplyand-demand' in price determination; his view of factor pricing and, more specifically, the question of whether he belongs within a Say-Walras tradition of analysis; the affinity of his work to forms of general equilibrium analysis; and, finally, the question of his alleged 'Sraffian' connections. My concluding remarks are reserved to the final chapter. UTILITY

The familiar exercise of'going behind' demand schedules to a model of consumer choice no longer necessitates any mention of 'utility' but, however performed, the object is to construct a framework within which an individual's consumption tastes may be formally represented, subject to certain imposed criteria of logical consistency. Having developed the model, the next step is usually to consider the properties of a consumption equilibrium, chief of which is, in the normal two good case (disregarding 'corner solutions'), the equality between the (given) relative price ratio and the individual's (subjective) marginal rate of substitution between the goods. That statement of the equilibrium property is an evolved form of a 241

242

INTERPRETING RIGARDO

result once obtained using (cardinal) utility analysis. Since it has been Alfred Marshall and, with Marshallian inspiration, Samuel Hollander, who have advanced the boldest claims that Ricardo at least came close to appreciating the utility doctrine, a quick dusting down of the Marshallian apparatus is in order. According to Marshall, utility is the 'correlative to Desire or Want', and 'in those cases with which economics is chiefly concerned the measure [of utility] is found in the price which a person is willing to pay for the fulfilment or satisfaction of his desire' ([1920] 1949, p. 78). Marginal utility is the utility obtained from the 'marginal purchase' and is outwardly 'measured' by the 'marginal demand price': the price which the consumer pays for the 'marginal purchase'. The relationship between marginal utility and quantity consumed was expressed by the 'law' of satiable wants or of diminishing utility: 'The total utility of a thing to anyone . . . increases with every increase in his stock of it, but not as fast as his stock increases' (pp. 78-9, emphasis in original). If we follow Marshall in assuming that an individual's marginal utility of money remains constant, the 'law' may be equivalently expressed in terms of a diminishing marginal demand price as quantity increases (p. 80). In equilibrium for the two good case, purchases will be so adjusted that 'the prices which he [the consumer] is just willing to pay for two commodities are to one another in the same ratio as the [marginal] utility of those two commodities' (the 'equi-marginal' principle) ( P . 80). Turning to his interpretation of Ricardo, Marshall believed that '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' (p. 670). The reference is to a passage in the Principles which I come to in due course. But to anticipate my main conclusion, I cannot find any basis for Marshall's reading if, as it did, it implies a recognition by Ricardo of the distinction between total and diminishing marginal utility. Ricardo's first pronouncement on 'utility' was delivered in notes on Bentham's Sur Les Prix: I like the distinction which Adam Smith makes between value in use and value in exchange. According to that opinion utility is not the measure of value. (1810/11, in, p. 284)

The appropriation of Ricardo

243

The distinction to which he referred was this: The word VALUE ... has two different meanings, and sometimes expresses the utility of some particular object, and sometimes the power of purchasing other goods which the possession of that object conveys. The one may be called Value in use'; the other, 'value in exchange.' The things which have the greatest value in use have frequently little or no value in exchange; and, on the contrary, those which have the greatest value in exchange have frequently little or no value in use. Nothing is more useful than water; but it will purchase scarce any thing ... A diamond, on the contrary, has scarce any value in use; but a very great quantity of other goods may frequently be had in exchange for it. (Smith [1776] 1981, pp. 44-5, capitals in original) 'Utility', or 'value in use', was not being used by Smith in the sense of subjective 'esteem': a commodity having a great value in exchange might have 'little or no' value in use (utility) even though it must evidently be a source of pleasure to its purchasers. As Mark Blaug has observed, it would seem that Smith's 'utility' meant something akin to the 'power to satisfy a generalized biological or social need' (Blaug 1985b, p. 40). Whether this was also Ricardo's meaning in the quoted extract, or whether he had even given the matter serious thought, are debatable points. After all, his remark was only an aside. However, with the qualification I come to presently, he persisted in the view that 'utility' (however conceived) 'is not the measure of value', meaning that ratios of exchange are not proportional to 'utilities'. This came nearer to clarification in a letter to Say which also mentioned a positive role for 'utility': Utility is certainly the foundation of value, but the degree of utility can never be the measure by which to estimate value. A commodity difficult of production will always be more valuable than one which is easily produced although all men should agree that the latter is more useful than the former. A commodity must be useful to have value but the difficulty of its production is the true measure of its value. For this reason Iron though more useful is of less value than gold. (18 August 1815, vi, pp. 247-8) If 'utility' is the necessary condition for a commodity to have exchange value, this may signal a movement away from the Smithian conception towards a more 'subjective' notion. On the other hand, with the attention on produced commodities, it was Ricardo's view that exchange ratios reflect the relative 'difficulty of production' rather than the relative 'utilities' of the commodities, where 'utility' appears to have more in common with objective usefulness (as

244

INTERPRETING RIGARDO

suggested by the example of iron). We must await further clarification. Proceeding chronologically, the pamphlet Economical and Secure Currency contained the following passage: If we say that value should be measured by the enjoyments which the exchange of the commodity can procure for its owner, we are still as much at a loss as ever to estimate value, because two persons may derive very different degrees of enjoyment from the possession of the same commodity, (iv, p. 61)1

This is certainly compatible with a 'subjective' conception of 'utility' or 'value in use', providing we grant the equivalence of those terms with the 'degree of enjoyment from the possession' of a commodity. The following curt statement, from the chapter 'On Value' in the Principles, lends support for so doing: Utility . . . is not the measure of exchangeable value, although it is absolutely essential to it. If a commodity were in no way useful, - in other words, if it could in no way contribute to our gratification, — it would be destitute of exchangeable value however scarce it might be, or whatever quantity of labour might be necessary to procure it. (1, p. 11)

Here, 'utility' (or 'usefulness') is expressed in 'subjective' terms: the power of something to somehow 'contribute to our gratification'.2 This 'utility' was a necessary condition for exchangeable value, but not its 'measure', Ricardo's argument being (subject to the modifications and exceptions) that the true 'measure' is provided by relative quantities of labour expenditure. Commodities exchange in proportion to 'labour quantities', not 'utilities'. The 'subjective' use of 'utility' is compatible with Marshallian 'utility' as the 'correlative of Desire or Want'; and the same usage is suggested by other passages in the Principles which, like the one from Economical and Secure Currency, deny the 'objectivity' or interpersonal comparability of 'utility' measurements.3 On the other hand, pas1

2

3

This was written before Ricardo's embrace of the 'pure' labour theory, which did supply a means of estimating 'value'. In context, Ricardo apparently believed, wrongly, that he was adopting Adam Smith's usage of'utility'. In the chapter, 'Taxes on Wages' Ricardo wrote: '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' (1, p. 241). And in 'Mr. Malthus's opinions on Rent': 'One set of necessaries and conveniences admits of no comparison with another set; value in use cannot be measured by any known standard; it is differently estimated by different persons' (1, p. 429). Taking these together, the 'value of enjoyments' is evidently the same as 'value in use' which, in other contexts, is a synonym for 'utility'.

The appropriation of Ricardo

245

sages from Ricardo's later correspondence may intimate the more 'objective' notion, seemingly present in the earlier letter to Say (quoted above, p. 243). Can these uses be reconciled? Possibly they can. In the letter to Say, Ricardo presumed that 'all men should agree' in their estimates of the superior 'utility' of iron relative to gold. Similarly, in later correspondence to Malthus he asserted that gold 'probably is by all mankind considered as the less useful metal' vis-a-vis iron (9 October 1820, vm, p. 277). And in other places, the iron and gold example was rehearsed on the assumption that the 'utility' from iron is always greater (1, p. 283,11, pp. 24-5). So, he may have thought that 'iron versus gold' was a special case in the sense that 'subjective' estimates of comparative 'utility' were remarkably uniform between individuals. However, his confidence for this evidently rested on the 'objective' judgement that iron could be used for a greater variety of concrete, useful functions. The 'subjective' and 'objective' angles had collided. Granted that Ricardo had some concept of the 'utility' derived from or associated with commodities, what of the distinction between total and marginal utility and the 'law' of diminishing marginal utility? Were they present in his work? There is a sense in which Ricardo can be credited with a rough notion of'marginal utility', although it would seem that he took this 'utility' as a given, fixed magnitude which was independent of the total amount consumed. Thus: 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. (Principles, 1, p. 28on, emphasis in original) And further: If two sacks [of corn] be of the value that one was of before, he [the purchaser] ... gets, indeed, double the quantity of riches - double the quantity of utility - ... but not double the quantity of value. (1, p. 281)4 The same message is indirectly conveyed by other passages, but before they are reviewed I will dissect the evidence adduced by Marshall (and Samuel Hollander) in favour of their readings. Their show-cased passage is from the chapter 'Value and Riches, their Distinctive Properties' in the Principles: 4

In these passages, 'value' depends on the quantity of labour expended in producing commodities.

246

INTERPRETING RICARDO

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 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. (1, pp. 275—6) In Marshall's opinion, Ricardo was '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' ([1920] 1949, p. 671). This interpretation, which may seem fanciful, derives, firstly, from the identification of'riches', the 'necessaries and luxuries', with total utility. In fact, this is quite legitimate; as Ricardo wrote to Say, 'I . . . estimate riches . . . by the whole quantity of utility which the commodities which constitute riches possess' (5 March 1822, ix, p. 169). So, total utility is certainly depleted by a reduction in 'the amount of commodities', but what of marginal utility? To understand Marshall one must remember that his 'marginal demand price' - the price paid by the consumer for the 'marginal purchase' - is, for him, an indirect measure of 'marginal utility'. Now, in the case Ricardo was attempting to controvert, he had supposed that aggregate commodity supply is, for some reason, suddenly depleted with the result that the level of (market) prices or, as he put it, 'the value of commodities', is raised. These higher prices are 'marginal demand prices' in the Marshallian sense, and within that discourse, the fact that people make their purchases at such prices carries the implication that the utilities from their 'marginal purchas' - marginal utilities have risen. However, to claim that this was Ricardo's understanding, or that it was the meaning towards which he was 'feeling his way', it is incumbent to show that he had a similar understanding of the relationship between 'marginal demand price' and 'marginal utility' involving, as it did, appreciation of'the law of diminishing marginal utility'. No such evidence was forthcoming from Marshall, presumably for the excellent reason that none could be found. The 'utility' derived from commodites was, for Ricardo, defined indepen-

The appropriation of Ricardo

247

dently of price.5 Marshall had simply imposed his own meaning on Ricardo's words. As for Samuel Hollander, he avers that the 'distinction between total and marginal utility and the conception of diminishing marginal utility could easily have been accommodated' by Ricardo, as if this excuses the absence of a developed utility analysis (1979, pp. 278-9). Leading up to his conjecture, which I return to shortly, Professor Hollander writes: 'The implications of the absence of a conception of diminishing marginal utility should not . . . be exaggerated. For Ricardo . . . was thoroughly conscious of the fundamental role of scarcity in value formation . . . and there can be no doubt at all that he was aware of the solution to the "paradox of value" ... taking for granted that if a hitherto free good (such as water) becomes scarce it will come to have exchange value' (p. 278). However, the 'water' example of a commodity coming to have a 'scarcity value' (1, p. 278) cannot be considered a 'solution to the "paradox of value'" which, for a 'neoclassical', would require the distinction between (diminishing) marginal utility and total utility. Ricardo was merely saying that if the supply of water is reduced (or appropriated), it would have a positive price. But a 'utility paradox' still remains: gold and water might both 'exchange for a great quantity of other goods' even though gold is 'of little use compared with . . . water' (1, p. 11). Further light is shed on Ricardo's actual position, as opposed to the one he 'might have' adopted, by the following passage, which is interesting for setting out the special circumstances in which he thought commodities would exchange in proportion to the relative 'esteem' that they are held by demanders. He commented on Malthus's Principles: In all that Mr. M. has yet said about exchangeable value, it appears to depend a great deal on the wants of mankind, and the relative estimation in which they hold commodities. This would be true if men from various countries were to meet in a fair, with a variety of productions, and each with a separate commodity, undisturbed by the competition of any other 5

In the passage cited from 'Value and Riches', 'wealth' and 'riches' are used as interchangeable terms. Hence the following sentence, from the same chapter, again reveals Ricardo's unawareness of the 'law of diminishing marginal utility': 'There will be double the quantity of commodities . . . and therefore the wealth of the country will be doubled' (i, p. 277). Cf. Ricardo's letter to Say, 5 March 1822: 'Double the production [and consumption] of [commodities] by economical processes and . . . [a]ccording to my view they [consumers] would be singly and collectively doubly rich' (ix, p. 171).

248

INTERPRETING RIGARDO

seller. Commodities, under such circumstances, would be bought and sold according to the relative wants of those attending the fair. (11, p. 24) As Ricardo stated elsewhere, this corresponds to a situation in which each 'commodity is valuable in proportion to its utility' (letter to Malthus, 9 October 1820, vm, p. 276). But this, he insisted, was an exceptional circumstance: when the wants of society are well known, when there are hundreds of competitors who are willing to satisfy those wants, on the condition only that they shall have the known and usual profits, there can be no such rule for regulating the value of commodities. In such a fair as I have supposed, a man might be willing to give a pound of gold, for a pound of iron, knowing the uses of the latter metal; but when competition freely operated, he could not give that value for iron, and why? because iron would infallibly sink to its cost of production, (n, PP- 24-5) For competitively produced, freely reproducible commodities, exchange at natural prices contravened the 'utility principle' as Ricardo understood it, which applied only in 'fairground' contexts. Had he been in possession of or, as the teleologist might say, 'groping towards', the Marshallian apparatus, he would have realised (vaguely) that the two 'rules' are always compatible. Even at natural prices, each individual adjusts purchases so that ratios of marginal utilities to natural prices are the same in each line of expenditure. He showed no such awareness. As far as he was concerned, price could serve as an indicator of 'utility' only in exceptional circumstances, although the 'utility' derived from a commodity would never be less than that outwardly indicated by the price paid. Thus, in a wine-drinking example (quoted more fully below, pp. 250—1) the price at which an individual discontinues consumption as price rises depends on the 'standard in his own mind by which he estimates the value of his enjoyments' (1, p. 241), suggesting that the 'utility' derived from a bottle had previously exceeded the financial 'sacrifice'. 'Utility' and price might converge but, in general, they did not. For the sake of argument, however, let us imbue the spirit of Samuel Hollander's interpretation and suppose that Ricardo had presciently 'discovered' Marshallian-style utility analysis. It seems improbable that he would have regarded it as particularly useful, for his evident requirement was for a single, supply-side determinant of

The appropriation of Ricardo

249

natural price - a 'rule for regulating the value of commodities' which, barring the exceptions, he thought he had found in the expended labour principle. A comparable monocausal function cannot be performed by 'neoclassical' utility analysis, nor could 'utility' provide the objective 'standard in nature' by which Ricardo sought to judge whether commodities had altered in their individual values (see above, p. 232). Professor Hollander's conjecture that the distinction between total and diminishing marginal utility 'could easily have been accommodated' by Ricardo (1979, pp. 278-9) is therefore meaningful, if at all, in the most trivial of senses: it would have been an 'easy accommodation' of no relevance to Ricardo's main analytical work. This is not to say that Ricardo ignored the influence of consumer 'tastes' on the demand (and supplies) of commodities. In fact, he seems to have regarded their influence as the single most important factor - more so than price - in governing consumer demands, often repeating that (shifts in) demands for, hence outputs of, reproducible manufactured commodities and demands for commodities in fixed supply were variously determined by (changes in) people's 'wants, wishes, tastes, inclinations, caprices and prejudices' (and, to make these demands effective, income).6 But these observations were not tantamount to 'going behind the demand schedule' in the 'neoclassical' sense. They were generally accepted, 'common sense' principles. I conclude that the attempt to absorb Ricardo within a 'utility' tradition of analysis must be judged a failure. 'Utility', at times used by Ricardo in a 'subjective' sense, at others in an 'objective' sense, and sometimes in both at once, was a prerequisite for exchange value. But, on his understanding, commodities did not generally exchange in proportion to their 'utilities' (so that price did not, under most circumstances, outwardly reveal 'utility'), nor did 'utility' per unit of commodity diminish with consumption. All this has long been accepted and, had it not been for Samuel Hollander's resuscitation of Marshall's idiosyncratic reading, it would have been needless to reopen the case. 6

See, for instance, i, pp. 12, 88, 194, 240, 249-50. The 'demands for the produce of agriculture are uniform, they are not under the influence of fashion, prejudice, or caprice' (1, p. 263); they are governed by 'want'.

25O

INTERPRETING RICARDO DEMAND

With all commentators agreed that Ricardo did not actually draw or explicitly refer to demand schedules, the issues I investigate in this section are whether he could be said to have grasped their 'essence'; and whether his remarks on demand-price relationships point to the 'particularly sophisticated' treatment of demand for which he has been commended by Samuel Hollander (1979, p. 273). One difference between Ricardo's approach and that of the 'modern' economic theorist is immediately apparent. According to the 'modern' treatment of demand, prices are not distinguished by any special categories, whereas Ricardian discourse makes a distinction between natural and market prices, where the latter have departed from their natural level.7 In principle, however, contemporary teaching would lead one to expect the same relationships between natural prices and quantities demanded as between market prices and demands, providing that ceteris paribus holds. Even so, Ricardo's discussions fall into two groups, depending on whether it is natural prices or non-natural market prices that are under consideration. I begin with the former cases. The behaviour of individual consumers in response to changes in natural price received only passing attention. With reference to commodities other than corn, there are only a few relevant passages in the Principles, and they are remarkably bland. Here is one: If wine were much raised in price in consequence of taxation, it is probable that a man would rather forego the enjoyments of wine, than make any important encroachments on his capital, to be enabled to purchase it. (1, p. 241)

Ricardo elaborated thus: 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 its limits, and experience every day demonstrates than an increase in the nominal amount of taxation, often diminishes the produce. One man will continue to drink the same quantity of wine, though the price of every bottle should be raised three shillings, who would yet relinquish the use of wine rather than pay four. Another will be content to pay four, yet refuse to pay five shillings . . . It is not because they cannot 7

To recall, natural prices are those obtaining when wages and profit rates are respectively uniform. In the present context, the 'natural' cost of production may also include an element of taxation.

The appropriation of Ricardo

251

pay more, that they give up the use of wine ... but because they will not pay more. (1, p. 241) It may not be violently unjust to interpret these remarks in terms of ceteris paribus reasoning: Ricardo had 'assumed', if only by default, that income, tastes and other prices are constants. We thus appear to have the presumption that the individual wine drinker imbibes a certain fixed quantity until price attains some critical level, at which point consumption is discontinued. This might be described by a schedule which is inelastic to the critical price and jumps to zero thereafter. But in aggregate, because of differences in the 'cut off points between individuals, quantity demanded is inversely related to price. At least for non-agricultural commodities, other remarks confirm that Ricardo considered this to be the likely aggregate case.8 In the case of corn, Ricardo's view was that the demand meaning here the amount that people want to consume - is for a fixed quantity. This 'demand' is therefore limited by the size of population, as he repeated on several occasions, in different contexts.9 But the dependence of the quantity actually demanded on natural price was seldom considered on the explicit assumption of given incomes. One rare instance is the following: An increase in the cost of production of a commodity, if it be an article of the first necessity [such as corn], will 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. (1, pp. 343-4) Again translating into the language of demand schedules, the market schedule is likely to be relatively price-inelastic, but when allowance is made for compensatory changes in the (money) natural wage rate following movements in the corn price, the demand is fully inelastic. The total quantity taken becomes dependent solely on the given requirements of the population ('If the market value of corn be increased one tenth by taxation, or by difficulty of production, it is doubtful whether any effect whatever would be produced on the quantity consumed, because every man's want is for a definite quantity, and, therefore, if he has the means of purchasing, 8 9

See, for example, i, pp. 191, 196, 201. See, for example, 1, pp. 79, 384-5; vin, p. 272; v, p. 108; iv, pp. 219-20.

252

INTERPRETING RIGARDO

he will continue to consume as before', 1, p. 193).10 Needless to say, the schedule that might be sketched to represent this viewpoint would violate one of the ceteris paribus conditions — that of given money incomes - on which 'modern' schedules are constructed. So much for Ricardo's observations on the relationship between natural prices and quantities demanded. He was more expansive, and repetitive, on the relationship between quantities demanded and market price variations, especially for corn: an 'agricultural' focus which reflected his engagement with contemporary events. His basic position was expressed in his letter to Malthus of 23 October 1814: A rise of raw produce may proceed from one or more bad seasons, which would undoubtedly increase profits, because the price of produce would rise considerably more than in the proportion of the deficient quantity, (vi, p. 146) In the Essay on Profits, the super-proportional relationship between quantity supplied and movements in (market) price was extended to the case of an 'abundant' supply (iv, pp. 28-9 ff.). It implies a negative relationship between the aggregate demand (consumption) of corn and market price. Part of the rationale was spelt out in the Principles'. A bad harvest will produce a high price of provisions, and the high price is the only means by which the consumption is compelled to conform to the state of the supply. If all the purchasers of corn were rich, the price might rise to any degree, but the result would remain unaltered; the price would at last be so high, that the least rich would be obliged to forego the use of a part of the quantity which they usually consumed, as by diminished consumption alone the demand could be brought down to the limits of the supply. (1, p. 162; cf. 1, p. 217)11 In On Protection to Agriculture, the social implications of bad harvests were stated more clearly. The 'rich would continue to consume precisely the same number of loaves, although the price was tripled or quadrupled' (iv, p. 221); translating again, their demand schedules are vertical for all 'likely' prices. As I read Ricardo, the 10

11

Barkai (1965) and Rankin (1980) reject the view that Ricardo treated the demand (and supply) for corn as a determinate function of population, irrespective of price. But in the passages they adduce to support their position (iv, pp. 219-20, 1, p. 162) Ricardo was discussing 'temporary' phenomena: the consequences of a deficient harvest. However, the quantity-population relationship applied, strictly speaking, to the 'natural course of things', i.e., on the assumption that both corn and wages are at their natural levels. Cf. Adam Smith, ([1776] 1981, pp. 524-5).

The appropriation of Ricardo

253

('implicit') individual demand schedules of the poor are, if not vertical, then highly inelastic, terminating at near income-exhausting prices.12 So it is the labourers who are priced out of the bread market. In aggregate, then, the market price of corn has risen and the quantity demanded (consumed) falls to conform with the reduced supply, as it must. If, on a diagram, the dots are connected between the price-quantity pairs at the natural price and at the higher market price, we obtain, rather obviously, a negatively sloped line: the upper section of the demand schedule for corn. The mirror-image rationale for the lower section of the schedule was frequently detailed in the post-Principles period and was integral to Ricardo's analysis of the prevailing 'agricultural distress'. The case was pithily stated for the benefit of his parliamentary colleagues: The demand [for corn] was limited, because no man could eat more than a certain quantity of bread. If there was more than the usual and ordinary quantity in the market, the price must of necessity fall. (9 April 1821, v, p. 108) In a later speech he expatiated with a numerical illustration: He would put a case to the House, to show how a superabundant supply of an article would produce a sinking of its aggregate value much greater than in proportion to the surplus supply. He would suppose, that in a particular country a very rare commodity was introduced for the first time — superfine cloth for instance. If 10,000 yards of this cloth were imported under such circumstances, many persons would be desirous of purchasing it, and the price consequently would be enormously high. Supposing this quantity of cloth to be doubled, he was of opinion that the aggregate value of the 20,000 yards would be much more considerable than the aggregate value of 12

Ricardo paid very little attention to the opportunities for food substitution. In the third edition of the Principles he did add that 'adopting the most useful substitutes' would relieve 'the distress of the labourer' following a deficient crop, although he was silent on the potency of such a remedy (i, p. 162). He was more forthcoming in a letter to Maria Edgeworth: 'When the crop of grain fails, . . . [some] say the people can have recourse to cheaper substitutes, such as potatoes. They can never make dear food a substitute for a cheap one, but they may make cheap food a substitute for a dear one. This argument would be just if at all times a supply of the cheap food could be obtained, but in a country where wheat constitutes the chief food of the people, no supply of potatoes ever is grown which can be adequate to feed the people if the crop of wheat fails. No more potatoes are grown than what are usually required in addition to the average crop of wheat. How then can potatoes be substituted for wheat? From whence are they to come? There is not limit to the rise in the price of potatoes which would take place under the circumstances supposed. In fact we should not substitute a cheap for a dear food, for this food which was ordinarily cheap would become as dear as wheat' (11 January 1823, IX> P- 2 5^)-

254

INTERPRETING RICARDO

the 10,000 yards, for the article would still be scarce, and therefore in great demand. If the quantity of cloth were to be again doubled, the effect would still be the same; for although each particular yard of the 40,000 would fall in price, the value of the whole would be greater than that of the 20,000. But, 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. This argument he applied to corn. Corn was an article which was necessarily limited in its consumption: and if you went on increasing it in quantity, its aggregate value would be diminished beyond that of a smaller quantity. (7 May 1822, v, p. 171)

Translating once again, the market demand schedule for superfine cloth is relatively elastic over an initial range but, as we move down the schedule, it eventually becomes steeper and, after a 'critical' point, is highly inelastic. Corn and, it would seem, all other commodities^ are similar to the extent that their demand schedules also possess a highly inelastic range: 'No principle can be better established, than that a small excess of quantity operates very powerfully on price. This is true of all commodities' {On Protection to Agriculture, iv, p. 220). 13

Ricardo's two sets of obiter dicta on natural price—demand relationships and on relationships between market prices and demand are mutually consistent in the sense that both imply negatively inclined demand schedules at the market level. In other words, he clearly appreciated the operation of the so-called 'law of demand' when applied to aggregate market behaviour. There is also consistency in his views of individual behaviour, although these may appear rather curious. His presumption was, it seems, that people demand a certain amount of a commodity which, in the case of a 'luxury', continues to be demanded until price reaches an 'unacceptable' level when consumption is forsaken entirely (suggesting, incidentally, that consumption was not conceived in flow terms). And for a 'necessary', all individuals will continue to consume the same amount providing income permits. It is the different 'termination points' of the individual schedules which 'underlie' the negatively inclined market schedules. It would be gilding the lily to describe this as a 'particularly sophisticated' treatment of demand. 'Demand' was not, for Ricardo, 13

Ricardo added that 'no general rule can be laid down for the variations of price in proportion to quantity' (iv, p. 220), by which he apparently meant that it was impossible to formulate a precise numerical relationship of general relevance.

The appropriation of Ricardo

255

a serious object of analysis in its own right. And although it is possible to 'translate' his scattered remarks into the language of demand schedules, as I have done, it must be borne in mind that this is an exercise in translation. As consciously formulated abstractions, demand schedules had no existence at all in his work. His analysis might appear rather more 'sophisticated' if, as Samuel Hollander has claimed, 'Ricardo must be considered as one of the "originators" of the elasticity concept' (1979, p. 277 n. 18). However, Ricardo only made remarks which, translated into a more precise theoretical discourse, imply (changing) elasticity values for some commodities. To the extent that others made similar remarks, they were all '"originators" of the elasticity concept', in Professor Hollander's peculiarly anaemic sense. And, indeed, the views Ricardo expressed, particularly on the effects of changes in the supply of corn on market prices (supposedly an instance of market behaviour generally) were neither particularly original, nor were claimed as such by him. Torrens, for example, had stated exactly the same case in his Essay on the External Corn Trade (1815):

When the supply of any article, particularly if it be one of the first necessity, is diminished below the [effectual] demand [i.e. the quantity demanded and consumed at the 'normal' price], its value rises, not merely in the ratio of this diminution, but in a ratio considerably higher. (1815, PP- H~lb)

And in an 'abundant' year, particularly if there are restrictions on corn exportation, superfluity will be created, but will find no vent until prices have sustained an extraordinary fall. (p. 46) As far as Ricardo was concerned, the same position was held by all serious commentators: 'The principle [that 'a small excess of quantity operates very powerfully on price'], I believe, has never been denied by those who have turned their attention to this subject' {On Protection to Agriculture, iv, p. 220). He, for one, was not claiming doctrinal priority; he was merely stating and sometimes illustrating an established 'truth'. There was no theoretical 'discovery'. In conclusion, statements to the effect that Ricardo 'did not possess a demand theory' or that there was a 'glaring omission of demand from the classical [including the 'Ricardian'] model' are certainly open to misinterpretation, perhaps suggesting, wrongly,

256

INTERPRETING RIGARDO

that 'demand' was something Ricardo passed over in total silence.14 But in claiming that 'Ricardo's treatment of demand turns out to be particularly sophisticated', Professor Hollander had given an equally spurious impression. Ricardo's 'treatment of demand' reduces to the intermittent articulation of a set of commonplaces, undistinguished by theoretical innovation or refinement of technique (not that Ricardo claimed otherwise). One is tempted to wonder what, for Professor Hollander, would count as an unsophisticated treatment of demand. PRODUCT P R I C I N G : THE ROLE OF 'DEMAND AND S U P P L Y '

In this section I enquire into the meaning and role accorded by Ricardo to 'demand and supply' explanations of product—price determination, the main point at issue being whether he can be classified as a 'demand-supply' theorist in the Marshallian tradition. Ricardo allowed scope for a crude 'demand and supply' mechanism when the absolute quantity supplied of a competitively produced, freely reproducible commodity is unequal to the absolute quantity that consumers wish to purchase at the natural price (Smith's 'effectual demand'); and if, for whatever reason, the supply of a commodity is fixed and non-augmentable. These cases may be taken together. Any divergence between the quantity supplied and the 'effectual demand' for a competitively produced, freely reproducible commodity was considered to be a 'temporary' or 'accidental' phenomenon which might arise from circumstances on the 'demand side', such as a change in fashion, caprice, income, and tastes; or, on the 'supply side', from developments over which producers have no control — most importantly changes in weather conditions for agricultural produce, and simple 'errors and miscalculations' in producers' forecasts of aggregate market output viz-a-viz the 'effectual demand' - or from changes in natural price should this alter the 'effectual demand' prior to any variation in market supply. The resulting 14

The first quotation is from Pasinetti ([i960], 1974, p. 12), cited by Hollander (1979, p. 273); the second is from Arrow and Starrett (1973). Pasinetti went on to say: 'The economic theory of demand had not yet been developed at that time, and there is no question of substitution among wage-goods in the Ricardian model' ([i960] 1974, p. 18). However, judging from Ricardo's letter to Maria Edgeworth, quoted above, p. 253 n. 12, there was no substitution because, in his opinion, there were no substitutes.

The appropriation of Ricardo

257

departure of market prices from their natural levels would depend on the extent of the excess or deficiency of market output relative to the 'effectual demand'; and on (prospective) purchasers' incomes and on their 'need' or sheer 'want' to possess the commodity (in the case of a deficient supply). But Ricardo's general rule of thumb was that any shortfall or surplus would lead to a 'powerful' (superproportional) movement of market price away from its natural level. This might be termed a 'demand and supply' explanation of market price in the sense that such prices, more especially their movements, are determined by the 'competition of buyers' relative to the available supplies. The application of the same reasoning to commodities in permanently fixed supply may be found at the beginning of the chapter 'On Value' in the Principles (1, p. 12, cf. pp. 249—50). As for the determination of the natural price levels of competitively produced, freely reproducible commodities, Ricardo was convinced that 'demand and supply' influences, as he understood them, were irrelevant. Indeed, believing that the contrary opinion had 'been the source of much error', he devoted a complete, albeit brief chapter in the Principles, to a defence of his viewpoint. It began: It is the cost of production which must ultimately regulate the price of commodities, and not, as has been often said, the proportion between the supply and demand: the proportion between supply and demand may, indeed, for a time, affect the market value of a commodity, until it is supplied in greater or less abundance, according as the demand may have increased or diminished; but this effect will be only of temporary duration. (1, P- 382) When market price is equal to natural price then, by definition, 'the proportion between the supply and demand', conceived as actual quantities bought and sold, and with 'demand' interpreted in the sense of Smith's 'effectual demand', is one of equality, whatever the natural price happens to be. 'Demand and supply', when couched in terms of proportionality, cannot account for the level of natural price. Of course, no-one would claim otherwise. The chapter contained a more substantial criticism of 'demand and supply' explanations of natural price, with Ricardo arguing that natural prices are determined by conditions of production (as reflected in costs of production), not by the 'equilibrium' levels of output at which absolute quantities supplied are equal to effectual demands. Thus:

258

INTERPRETING RIGARDO

If the demand for hats should be doubled, the [market] price would immediately rise, but that rise would be only temporary, unless the cost of production of hats, or their natural price, were raised. If the natural price of bread should fall 50 per cent, from some great discovery in the science of agriculture, the demand would not greatly increase ... and as the demand would not increase, neither would the supply ... Here, then, we have a case where the supply and demand have scarcely varied ... and yet the price of bread will have fallen 50 per cent. (1, pp. 384-5) The examples given in this passage are, respectively, of an increase in the 'equilibrium' amounts supplied and demanded without, necessarily, any variation in natural price; and of a reduction in natural price with little, if any, affect on the quantities supplied and demanded. The moral drawn by Ricardo was that the level of all natural prices 'will ultimately depend, not on the state of demand and supply, but on the increased or diminished cost of their production' (1, p. 385): a position tantamount to denying any functional relationship between output at natural prices (the amount supplied and effectually demanded) and the conditions of producing that output. The same position was expressed elsewhere in the Principles: If the natural price of cloth were 20s. per yard, a great increase in the foreign demand might raise the price to 25s, or more, but the profits which would then be made by the clothier would not fail to attract capital in that direction, and although the demand should be doubled, trebled, or quadrupled, the supply would ultimately be obtained, and cloth would fall to its natural price of 20s. So, in the supply of corn, although we should export 2, 3, or 800,000 quarters annually, it would ultimately be produced at its natural price, which never varies, unless a different quantity of labour becomes necessary to production. (1, p. 304) Even in the case of corn, Ricardo refused to acknowledge that changes in the amount demanded, and hence supplied, were in any way causally responsible for natural price movements. He vigorously defended his position to Malthus soon after the publication of the first edition of the Principles: Lord King, Mr. Whishaw and you have done me a great deal of honour in making my work the subject of your discussions, but I confess itfillsme with astonishment to find that you think . . . that the measure of value is not what I have represented it to be; but that natural price as well as market price, is determined by the demand and supply, - the only difference being that the former is governed by the average and permanent demand and supply, the latter by the accidental and temporary. — In saying this do you mean

The appropriation of Ricardo

259

to deny that facility of production will lower natural price and difficulty of production raise it? Will not these effects be produced; after a very short interval, although the absolute demand and supply, or the proportion of one to the other, should remain permanently the same? At any rate then demand and supply are not the sole regulators of price ... However abundant the demand it can never permanently raise the price of a commodity above the expence of its production, including in that expense the profits of the producers. It seems natural therefore to seek for the cause of the variation of permanent price in the expenses of production. Diminish these and the commodity must finally fall, increase them and it must as certainly rise. What has this to do with demand? ... I cannot help viewing this question as a truth which admits of demonstration and I am full of wonder that it should admit of a doubt. (30 January 1818, vn, pp. 250-1, emphasis in original) A 'causal' theory of changes in the level of natural price should postulate conditions of production as the determining factor, not the amounts demanded and supplied, or the proportions between those amounts. Malthus was unconvinced, avowing in his own Principles that 'the great principle of demand and supply is called into action to determine what Adam Smith calls natural prices as well as market prices' (11, pp. 45—6). His supporting argument merits a brief review. Having supposed that the cost of producing a manufactured commodity increases, he wrote: But what is it which specifically forces up the price? It ... is a contingent failure of supply. Remove these contingencies, that is, let the extent of the supply remain exactly the same, without contingent failure or excess, whether the price [cost] of production rises of falls, and there is not the slightest ground for supposing that any variation of price would take place. (11, p. 46) With an unchanging market supply and original effectual demand, market price would continue to equal the old natural price even though the new natural price, reflecting the new cost of production, has risen (or fallen). It 'follows', Malthus declared, 'that the relation of the supply to the demand . . . is the dominant principle in the determination of prices whether market or natural' (11, p. 47). However, this was not to deny: that labour and the costs of production have not a most powerful effect upon prices. But the true way of considering these costs is, as the necessary condition of the supply of the objects wanted. (11, p. 49)

260

INTERPRETING RIGARDO

Market price would in all probability rise (or fall) to the new level of costs via output contraction (or expansion); and unless these costs were covered, the commodity would not be produced. But the mere fact that the commodity was selling at its new natural price was not, for Malthus, evidence that cost of production determined this price, contrary to his (correct) interpretation of Ricardo's position. Hutches Trower could not see what all the fuss was about. He wrote to Ricardo: I have made but little progress in his [Malthus's] Book yet; but, hitherto, I cannot think he has succeeded in overturning any of your positions. He labors hard to prove, that the prices of commodities are not regulated by the cost of productions, and, yet, I think, he admits the point, even by his mode of reasoning the subject; for, he allows, that the costs of production have a more powerful effect upon prices. 'But,' he adds, 'the true way of considering these costs is as the necessary condition of the supply of the objects wanted.' Now, if the supply of the objects wanted depends upon the price, covering the costs of production; and that price depends upon the relation between the supply and demand, then, must that price be governed by these costs of production, because if that price be not adequate, the relation between supply and demand is altered, and the price is necessarily affected. After all, I confess, it appears to me little more than a dispute about terms. (5 July 1820, vm, p. 201, emphasis in original) Trower's letter elicited a favourable response: I am pleased . . . with the observations you make on what he [Malthus] has said respecting my doctrine, of price, being ultimately regulated by cost of production. By the very definition of natural price, it is wholly dependent on cost of production, and has nothing to do with demand and supply. (21 July 1820, vm, p. 207, my emphasis)

This restatement of his position was uncompromising. Subsequently, Ricardo attempted to demonstrate that Malthus's way of viewing and expressing the forces governing natural price could be reduced to his own (see 11, pp. 48-9, 224-5), but in doing so two distinct propositions became entangled. The first proposition was that the level of natural price depends on 'cost of production', while the second was that the 'demand and supply' process by which market price tends to equate with natural price involves supply variations: market supply contracts or expands as market price is below or above the natural level and hence the (natural) costs of production. In the sense of market 'processes', the equation of market with natural prices does depend on 'demand and supply' or,

The appropriation of Ricardo

261

more precisely, on supply accommodating itself to the 'effectual demand', which Ricardo had never denied. But he remained adamant that neither market processes, nor the proportions between the amounts supplied and demanded, nor the amounts themselves, could account for the determination of natural price levels and their movements. In view of these enduring convictions, it is unsurprising to find him rejecting the opposing Malthusian position that 'supply and demand5 forces determined natural prices (see 11, pp. 46-77). In that fundamental respect he never wavered, even when discussing the case of corn.15 Does it all amount to an analysis of price determination which is fully compatible with the Marshallian tradition? The crux of the matter turns on the (dis) similarities in the treatments of price determination for competitively produced, freely reproducible commodities. We can all agree (I trust) that Ricardo accepted 'supply and demand' explanations, of a rudimentary sort, for other commodities. Marshall apparently believed that there was an essential analytical continuity between his own work and Ricardo's. Let us remind ourselves of his own position on the determination of value, the essence of which is captured in the following celebrated passage: We might as reasonably dispute whether it is the upper or the under blade of a pair of scissors that cuts a piece of paper, as whether value is governed by utility or cost of production. It is true that when one blade is held still, and the cutting is effected by moving the other, we may say with careless brevity that the cutting is done by the second; but the statement is not strictly accurate, and is to be excused only so long as it claims to be merely a popular and not a strictly scientific account of what happens. . . . [W]e find some commodities which conform pretty closely to the law of constant return; that is to say, their average cost of production will be very nearly the same whether they are produced in small quantities or in large. In such a case the normal level about which the market price fluctuates will be this definite and fixed (money) cost of production. If the demand happens to be great, the market price will rise for a time above the level; but as a result production will increase and the market price will fall: and conversely, if the demand falls for a time below its ordinary level. In such a case, if a person chooses to neglect market fluctuations, and to take it for granted that there will anyhow be enough demand for the 15

So, for example, we find him stating in On Protection to Agriculture, with specific reference to corn: 'the degree in which the price [of corn] must rise to compensate the producer for the charges which he has to pay, does not depend on the quantity produced, nor on the quantity consumed, but on the cost of its production' (iv, p. 211).

262

INTERPRETING RICARDO

commodity to insure that some of it, more or less, will find purchasers at a price equal to this cost of production, then he may be excused for ignoring the influence of demand, and speaking of (normal) price as governed by cost of production — provided only he does not claim scientific accuracy for the wording of his doctrine, and explains the influence of demand in its right place. ([1920] 1949, pp. 290-1)

Ricardo was certainly one of those who spoke of natural price as 'governed by cost of production' which, according to Marshall, is excusable only if commodities 'conform pretty closely to the law of constant return'. By a happy coincidence, he thought 'constant returns' was Ricardo's general assumption: [T]hough he was aware that commodities fall into three classes according as they obey the law of diminishing, of constant, or of increasing return; yet he thought it best to ignore this distinction in a theory of value applicable to all kinds of commodities. A commodity chosen at random was just as likely to obey one as the other of the two laws ... and therefore he thought himselfjustified in assuming provisionally that they all obeyed the law of constant return, (p. 671) The statements 'into which Ricardo often glided with careless brevity, as to the dependence of value on cost of production' could therefore be excused, especially because Ricardo only regarded them as 'part of a larger doctrine, the rest of which he had tried to explain' (p. 673). That 'larger doctrine' was perfectly consistent with Marshall's own. The flaw in this interpretation is that Ricardo's 'careless statements' were not predicated on the assumption of constant returns in the Marshallian sense. They were statements of unrestricted scope, which applied regardless of whether, as a matter of fact or of assumption, changes in conditions of production accompanied changes in output. Unlike the Marshallian world, Ricardo's is therefore one in which demand and supply factors are never assigned co-ordinate importance in natural price determination. In this respect, the two analytical 'traditions' diverge fundamentally. But that is not to deny all points of similarity. Thus, it is indisputable that Ricardo relied on 'demand and supply' analysis, in the form of supply variations, when explaining the process by which market and natural prices converge. Samuel Hollander is quite right in taking Schumpeter to task for claiming otherwise (Hollander 1979, p. 280). But what was the analytical significance of the 'process analysis'? If, as Professor Hollander does, we lavish atten-

The appropriation of Ricardo

263

tion on the material in Ricardo's taxation chapters, we might be tempted to conclude that it was of paramount importance generally. For, in his attempt to demonstrate that taxes must raise commodity prices so that producers obtain the going rate of profit, Ricardo repeatedly insisted that this is effected by way of supply variations, with the 'demand and supply' relationship ultimately elevating price to its new, higher level.16 However, by invariably taking the point of market-natural price convergence as something already given, Ricardo was committing what is, judged by Marshallian or later 'neoclassical' standards, a cardinal analytical sin. He was rigorously separating the determination of natural price - knowledge of the price that must prevail - from the demand-supply processes which ensure that it will prevail. With regard to product pricing, the conclusion must be that Ricardo was not a 'supply and demand' theorist in any substantive sense. RICARDO AND THE

SAY TRADITION .

OPPORTUNITY COST

AND 'IMPUTATION'

Samuel Hollander's claim that Ricardo accorded a 'key role' to 'opportunity cost and derived factor demand' is one facet of his more general argument that there exists an essential compatibility between 'Ricardian' and 'neoclassical' economics. In examining his interpretation I will separate two issues. First, I consider the sense in which Ricardo did employ 'opportunity cost' reasoning. Secondly, I take up the question of his attitude towards Say's doctrine. Professor Hollander's version of Ricardo's 'opportunity cost' analysis is that the 'factors of production' - 'labour' and 'capital' change employment if their 'returns' are higher elsewhere. 'Costs of production' - wages and profits - may therefore be considered as 'opportunity costs' which, in equilibrium, 'maximise the return to the factors "capital" and "labour"' (Hollander 1985a, p. 24). Although Ricardo did fashion an analysis which can be given an 'opportunity cost' interpretation, of sorts, the robustly 'neoclassical' version advanced by Professor Hollander is wholly inappropriate. To be meaningful, it must rest on the supposition that there was, in Ricardo's thought, an analytical distinction between 'capital' and 16

See, for example, i, pp. 156, 181, 205, 209, 243, 257. Ricardo also pointed out that exactly the same mechanism must come into play if physical conditions of production deteriorate (1, p. 191).

264

INTERPRETING RIGARDO

'labour' qua 'factors of production' and that 'profit' and 'wages' were conceived as the 'returns' to these separate 'factors', somehow related to their productive contributions.17 Any such reading of Ricardo is quite untenable. Taking the Principles as our source, let us first consider Ricardo's concept, or rather concepts, of'capital'. There is, first, a 'materialist' conception, exemplified by the following definition: Capital is that part of the wealth of a country which is employed in production, and consists of food, clothing, tools, raw materials, machinery, &c. necessary to give effect to labour. (1, p. 95)18 'Capital' therefore comprises of things which, it should be noted, include the labourers' 'necessaries'. On more numerous occasions, 'capital' is presented as the monetary 'fund' which is expended on the 'things'. For instance: if there be no increased production or diminished unproductive consumption on the part of the people, ... taxes will necessarily fall on capital, that is to say, they will impair the fund allotted to productive consumption. (1, pp. 150-1) The idea of this 'capital fund' being expended on the physical constituents of capital, including the means of 'supporting' labour (subsumed under 'circulating capital'), is conveyed by the following: In rich and powerful countries, where large capitals are invested in machinery, more distress will be experienced from a revulsion in trade, than in poorer countries where there is proportionally a much smaller amount of fixed, and a much larger amount of circulating capital, and where consequently more work is done by the labour of men. (1, p. 266) There are many other passages which bear the same interpretation.19 To confuse matters, 'capital' was occasionally used in a manner perhaps suggesting a 'fund' allotted to the 'support' of labour alone. Thus: 17

18 19

It must also be supposed that the 'returns themselves are, in principle, variables governed by the relative scarcity of the factors' (1985a, p. 24). In other words, Hollander's case rests, in part, on the validity of his 'new view' interpretation of Ricardo's treatment of wages. Since that interpretation has been found wanting, it follows that Hollander's 'opportunity cost' reading is flawed from the very outset. However, the critique of Hollander that I develop here is largely independent of the 'wages' issue. Cf. 'The capital of a country consists of its commodities' (1, p. 270). 1, pp. 132, 291, 324, 361-2, 388. Other passages will be cited below.

The appropriation of Ricardo

265

If the tax had been laid at once on the people of capital, their fund for the maintenance of labour would have been diminished. (1, p. 221) And: The distress which proceeds from a revulsion of trade, is often mistaken for that which accompanies a diminution of the national capital ... When, however, such distress immediately accompanies a change from war to peace, our knowledge of the existence of such a cause will make it reasonable to believe, that the funds for the maintenance of labour have rather been diverted from their usual channel, than materially impaired. (1, P. 265) It would be rash to accord much significance to such apparent usages. It was, for instance, soon after giving his 'materialist' definition of capital, which extended to things other than wage-goods, that Ricardo made reference to 'the accumulation of capital, or the means of employing labour' (1, p. 98). I infer that 'capital' does 'employ labour' but that such expressions were not intended to imply an exhaustive description of the material items of capital. Ricardo has supplied two primary conceptions of'capital': it is a collection of heterogeneous material objects, including wage goods; and it is a 'fund' which is expended on those material objects. Both conceptions were invoked in Ricardo's discussions of 'capital transfer'. The 'restless desire on the part of all the employers of stock, to quit a less profitable for a more advantageous business, has', says Ricardo, 'a strong tendency to equalize the rate of profits of all' (1, p. 88). This is achieved by the 'movement' of 'capital' between employments, the 'real-world' process being described in these words: It is perhaps very difficult to trace the steps by which this change is effected: it is probably effected, by a manufacturer not absolutely changing his employment, but only lessening the quantity of capital he has in that employment. In all rich countries, there is a number of men forming what is called the monied class; these men are engaged in no trade, but live on the interest of their money, which is employed in discounting bills, or in loans to the more industrious part of the community. The bankers too employ a large capital on the same objects. The capital so employed forms a circulating capital of a large amount, and is employed, in larger or smaller proportions, by all the different trades of a country. There is perhaps no manufacturer, however rich, who limits his business to the extent that his own funds alone will allow: he has always some portion of

266

INTERPRETING RIGARDO

thisfloatingcapital, increasing or diminishing according to the activity of the demand for his commodities. When the demand for silks increases, and that for cloth diminishes, the clothier does not remove with his capital to the silk trade, but he dismisses some of his workmen, he discontinues his demand for the loan from bankers and monied men; while the case of the silk manufacturer is the reverse: he wishes to employ more workmen, and thus his motive for borrowing is increased: he borrows more, and thus capital is transferred from one employment to another, (i, p. 89) The capital initially 'transferred' is the 'fund' (in the very next paragraph Ricardo mentions a capitalist 'seeking profitable employments for his funds', 1, p. 90, cf. p. 91) which is then expended on the material elements of capital, including the means of directly 'supporting' wage labourers. Hence the movement of the capital fund is followed by the 'transfer' of capital items. Later in the Principles, Ricardo again endeavoured to explain the 'capital transfer' mechanism, but this time more was involved than discontinuing or increasing demands for loans from 'monied men': It has ... been said, that capital cannot be withdrawn from the land; that it takes the form of expenses, which cannot be recovered, such as manuring, fencing, draining, &c, which are necessarily inseparable from the land. This is in some degree true; but that capital which consists of cattle, sheep, hay and corn ricks, carts, &c. may be withdrawn; and it always becomes a matter of calculation, whether these shall continue to be employed on the land ... or whether they shall be sold, and their value transferred to another employment. (1, pp. 268-9) Once more, the 'capital' withdrawn and transferred is the capital fund which in this case is realised by selling the material constituents of capital. The next step would be to expend the reconstituted fund on all those items of capital required in the new 'employment'.20 To summarise, 'capital transfer' is in the first place a transfer of the capital fund which is then expended on the items of capital, including the means of directly 'supporting' labourers. This is how Ricardo should be interpreted when, in numerous places, he simply refers to the transfer of'capital'. 21 Yet in other places he writes of the transfer, or allocation, of capital and labour.22 Is this significant? I think not. The capital fund includes expenditure on wage-goods - the cost of employing labour - but it does not include labour itself: 20 21 00 22

Also see 1, p. 119. F o r e x a m p l e , 1, p p . 134, 165, 166, 191, 209, 259, 29111., 3 0 3 , 306. See, for e x a m p l e , 1, p p . 133, 238, 263.

The appropriation of Ricardo

267

the 'services' of labour are not part of capital although the 'cost' or means of employing labour is. So, 'capital transfer' may, indeed does, imply a subsequent 'transfer' or redeployment of labour, but they are not one and the same thing. To speak of a transfer of 'capital and labour' is implicitly to distinguish between two volitional acts on the part of economic agents which, although related, are nevertheless separable. The capitalist transfers 'capital' and the labourers change their place of work, but without the redeployment of'capital', labourers would stay put. To speak only of a transfer of 'capital' was therefore to single out the prime mover in the transfer mechanism, taking it as read that 'labour only goes where capital leads the way' (Cannan 1929, p. 185). And so to 'opportunity cost'. The following passage provides a compact statement of Ricardo's position: It is . . . always a matter of choice in what way a capital shall be employed, and therefore there can never, for any length of time, be a surplus of any commodity; for if there were, it would fall below its natural price, and capital would be removed to some more profitable employment. (1, p. 29m.)

The person responsible for making the choice is the capitalist who must decide where to allocate the disposable capital fund. If engaged in a trade in which profits are below the natural level, there is an 'opportunity cost' if the natural rate, or better, can be obtained elsewhere (that it can be was Ricardo's 'law of markets' presumption). The 'primary' opportunity cost is therefore the potential loss of return on the capitalist's total outlay from investing it in one line of business rather than another; if this cost is positive, the capital fund will 'flow' from one occupation to another. But when that happens there is a resulting 'opportunity cost' to the labourers. The redeployment of the capital fund generates an additional demand for labour in the expanding sphere which may be expected to have the 'temporary' effect of raising wages; but labourers also make choices, and they would be expected to shift employment in consequence. To stress, however, the primary opportunity cost which serves as the motivating force, without which everything remains static, is that borne by the capitalist. This 'opportunity cost' reading of Ricardo does not correspond to Professor Hollander's version. Profit is not a return to a distinct 'factor of production' called 'capital'. However, it is easy to see how an unsuspecting reader could be seduced by Professor Hollander's

268

INTERPRETING RICARDO

'factor mobility' interpretation. He quotes the following paragraphs: 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 for woollens; their natural price, the quantity of labour necessary to 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 maufacturer 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 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. // 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 for the labour 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, (i, pp. 90-1; quoted by Hollander, 1985a pp. 21—2, his emphasis)

Without adequate explanation it may well seem that 'capital' and 'labour' are distinct productive 'things', with 'profit' as a return on the 'capital' alone. Indeed, this impression is fostered by Hollander's sentence immediately after the quotation: 'In circumstances of differential factor ratios the same assumption of factor mobility dictates a divergence of (relative) cost prices from (relative) labour inputs ...' (1985a, p. 22). Translated back into Ricardo's vocabulary, 'differential factor ratios' are (connected with) differences in the proportions of fixed to circulating capital which, if we take 'circulating capital' as expenditure on labour alone, implies a 'real' proportion between direct labour inputs and other means of production such as machinery. The latter ratio, if it could somehow be given quantitative meaning, is one between the 'factors of production', 'labour' and 'capital', in something approaching a Walrasian sense. Implicitly, we are therefore invited to impose the same distinction on the preceding paragraphs. But this will not do. The

The appropriation of Ricardo

269

'transference of capital' refers, I submit, to the staggered 'fund' 'real capital' transfer mechanism, with the 'capital' including (expenditure on) wage-goods.23 Moreover, the 'overplus': capital ratio must surely be interpreted as the ratio of net output value, or profit, to the value of all the capital items necessary to replicate the last period's productive activity, including the value of wage-goods.24 'Profit' is therefore a return 'on' or 'for' the (advance of the) entire capital stock. It is not a return to a distinct 'factor of production' called 'capital'. To conclude this phase of the discussion, although it is possible to give an 'opportunity cost' reading of Ricardo's analysis of the process by which market prices come to reach their natural levels, this cannot be couched in terms of foregone opportunities to 'factors of production' in the manner attempted by Professor Hollander, who merely imposes a set of meanings which lack textual resonance. It remains to consider Ricardo's attitude towards Say's doctrine of 'productive services', the latter set out in the following passage from the fourth edition of Say's Traite d'economie 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. This passage, reproduced by Ricardo in the third edition of his Principles (1, pp. 282-3) is said by Professor Hollander to incorporate 'the principles of opportunity cost and of imputing the values of 23

24

The 'fundist' interpretation is evidently confirmed in the second batch of Professor Hollander's italicised lines. It may be suspected that the italics were used not so much to draw attention to this particular point, which he passes over, but rather to emphasise the 'centrality' of'allocation mechanisms' in Ricardo's treatment of natural price. This was Ricardo's considered practice (i, pp. 34-5, 388-9), although he did occasionally omit wage outlays from the capital stock. See above, p. 93 n. 8.

27O

INTERPRETING RIGARDO

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' (1985a, p. 19). One might pause to ask whether there can be a 'principle of imputation' without a 'marginal conception', but I let this pass. However, one would certainly want to know what it is, 'in broad terms only', that is being 'imputed'. The answer provided by Say, it would seem, is 'utility': 'Price is the measure of the value of things, and their value is the measure of their utility' (Say, Traite d'economie politique, quoted by Ricardo, 1, p. 282). Very much in retrospect, this might sound vaguely 'Austrian'. According to Professor Hollander's more recent work, Ricardo believed Say was 'on the right road' with his analysis (1985a, p. 20); indeed, Ricardo is said to have 'applauded' Say's 'doctrine of productive services' (p. 2j).25 As evidence for this he first quotes the following remark from the Principles: 'M. Say maintains with scarcely any variation, the doctrine which I hold concerning value' (1, p. 283), adding that Ricardo's '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 ... whereby rent is excluded from (marginal) cost' (1985a, p. 20). Then, in a footnote, he extracts two sentences from Ricardo's correspondence, the first addressed to Malthus, the second to Say: if he [Say] would give up rent, he and I should not differ very materially on that subject ['productive services']. (9 October 1820, vm, p. 277) And: 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. (8 May 1821, vm, p. 379) On this evidence, Professor Hollander's reading might appear quite plausible. On the basis of my interpretation, however, it would be decidedly odd, if not totally inexplicable, to find Ricardo accepting Say's 25

Cf. Hollander, 1987, p. 139. In his Economics of David Ricardo, Professor Hollander was more cautious: 'It is possible that the acceptance of [Say's] rather ill-defined doctrine of services was a formality, the acceptance of an alternative exposition of the cost-price analysis. But we cannot rule out the possibility that Ricardo intended much more' (1979, p. 670). The later Hollander is evidently in no doubt that Ricardo did 'intend much more'.

The appropriation of Ricardo

271

doctrine. First, he would have had to believe that all commodities are valuable in proportion to their 'utility', which he expressly denied for competitively produced, freely reproducible commodities selling at natural prices. And secondly, he would have had to accept a 'factor of production' analysis of distribution which, I have just argued, he did not. However, since we are dealing 'in broad terms only', it may have been that there was some other sense in which he accepted the 'imputational principle'. Fortunately, it will not be necessary to speculate on what this might have been. For, as I shall now establish, the substance of the 'agreement' was an attempt to convince Say that some of his pronouncements were consistent with Ricardo's view that commodities are valuable in proportion to labour expenditures, not their 'utility'. There was no substantive acceptance of Say's framework. Ricardo enjoyed a cordial relationship with Say, regarding him as a friend and, in important respects, an ally. In his first letter to Say, Ricardo complimented him for 'ably disclosing]' the 'evils' of'all those laws which so materially interfere with the prosperity of nations', which was undoubtedly a reference to Say's advocacy of laissez faire (24 December 1814, vi, p. 166).26 And in his early writings, Ricardo took delight in Say's support for 'the doctrine that demand is regulated by production': the 'law of markets' (letter to Mai thus, 18 December 1814, vi, pp. 163-4). But he was, from the first, implacably opposed to Say's 'utility' doctrine, agreeing that utility is 'certainly the foundation of value' but emphatically denying that commodities generally exchange in proportion to their 'utility' - 'difficulty of . . . production is the true measure of . . . value' - or that 'value' and 'riches' are the same thing (letter to Say, 18 August 1815, vi, pp. 247-8). 27 In this critical respect he was unerring. When it came to writing the Principles, Ricardo's reservations towards Say's work, and his desire to criticise its perceived defects, evidently came into conflict with his overall feelings of friendship. He wrote to Mill: 26

27

Ricardo repeated the compliment in a later letter: 'You, I know, always exert yourself in the good cause' which was 'the establishment of the principles of a liberal policy in trade' (8 May 1821, VIII, p. 381). Also see the reference to Say in Ricardo's Proposals for an Economical and Secure Currency, iv, p. 71, and the passage from the Preface to the Principles, quoted below, p. 272. Cf. remarks in a subsequent letter to Mill, 30 August 1815, vi, pp. 264-5.

272

INTERPRETING RIGARDO

After reading Smith I mean to read Say again, - but as he is a living author, and a friend, I should feel some delicacy in making my objections publicly and strongly to his opinions, which I should not feel with Adam Smith. (17 November 1816, vn, p. 89) The choice of critical tone obviously bothered him to the point where he subsequently asked for Mill's advice (2 December 1816, VII, pp. 100-1). Mill responded: 'I should not fail to point out his errors; though with that respectfulness of manner to which he is entitled' (16 December 1816, vn, p. 108). In fact, Ricardo's handling of Say, in print and in their correspondence, though not in letters to his British friends, always remained both 'delicate' and 'respectful'. Thus we find him paying fulsome tribute to 'the excellent works of M. Say' in the Preface to the Principles] Say, he continued, not only was the first, or among the first, of continental writers, who justly appreciated and applied the principles of Smith, and who has done more than all other continental writers taken together, to recommend the principles of that enlightened and beneficial system to the nations of Europe; but who has succeeded in placing the science in a more logical, and more instructive order; and has enriched it by several discussions, original, accurate, and profound. (1, pp. 6-7) But having sung the Frenchman's praises, he added: The respect, however, which the author entertains for the writings of this gentleman, has not prevented him from commentating with that freedom which he thinks the interests of science require, on such passages of the 'Economie Politique', as appeared at variance with his own ideas. (1, p. 7) This he mainly did in the chapter 'Value and Riches, their Distinctive Properties'. The thrust of his criticism in the first and second editions was that Say 'has confounded two things which ought always to be kept separate, and which are called by Adam Smith, value in use and value in exchange', his argument against Say being that utility is 'not the measure of exchangeable value' (1, p. 28on.). But it was only prior to Ricardo's third edition that Say announced his 'productive services' doctrine. This was in the fourth edition of the Traite, published in late 1819. Soon after receiving his complimentary copy, Ricardo wrote to Say: Your chapter on value is, I think, greatly improved; but I cannot yet subscribe to all your doctrines on that most difficult part of the science of Political Economy. (11 January 1820, vm, p. 149)

The appropriation of Ricardo

273

To Malthus, eight months later, his verdict on Say's new edition was decidedly astringent: I am not convinced by any thing Say says of me - he does not understand me, and is frequently at variance with himself when value is the subject he treats of. (4 September 1820, vm, p. 228) The alleged internal contradictions in Say's treatment were detailed in a subsequent letter to Malthus. Ricardo first repeated that Say 'certainly has not a correct notion of what is meant by value, when he contends that a commodity is valuable in proportion to its utility' (9 October 1820, vm, p. 276). Then came a reference to Say's 'productive services': I think more may be said in defence of his doctrine of 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, (vm, p. 277) This is one of the passages cited by Professor Hollander, who wisely refrains from giving the next line: Tn what he says of services he is quite inconsistent with his other doctrine about utility' (vm, p. 277). If the 'regulation of value' by 'productive services' is 'quite inconsistent' with the notion of value in proportion to 'utility', then, as I have anticipated, the total value of the productive services (the cost of production) cannot be the same as the 'utility' attributed to, or derived from, the product; Ricardo cannot have believed that the 'imputed value' is utility which, in his opinion, remained the same notwithstanding changes in the 'value' of the commodity. But secondly, his language belies an 'imputation' perspective. He thinks that 'productive services' are the 'regulators of value' whereas an 'imputationalist' would presumably reverse the sequence to read: the values of productive services are 'regulated' by the value of the product. In a subsequent letter to McCulloch, the very partial nature of Ricardo's 'agreement' with Say can be read between the lines: I do not know whether you are well acquainted with his [Say's] work on Political Economy; - I have looked over carefully all the new matter in his fourth edition without discovering any thing to induce me to alter the opinion which I have given of the confusion of his ideas respecting value. Utility, riches, value, according to him are all the same thing. A commodity is more valuable because it is more useful. A man is rich in proportion as he is possessed of value - of utility, and it makes no difference whether commodities are of a low value or of a high value. Erroneous as I think

274

INTERPRETING RIGARDO

these views are he has not the merit of uniformly adhering to them, for he often acknowledges that commodities will fall in value if their cost of production be diminished, altho' they preserve the same utility. (23 November 1820, vm, p. 299) These remarks, which were far from 'respectful' towards Say, implicitly reveal that Ricardo took Say's doctrine of'productive services' as an 'acknowledgement' that 'value' is governed by 'cost of production': a doctrine which, if Say 'would give up rent', was tantamount to Ricardo's position.28 This did not constitute an endorsement of Say's imputational perspective (such as it was). Ricardo's later correspondence with Say in 1821 and 1822 confirms my interpretation. But before presenting this material, I consider Ricardo's attitude towards Say's formulation (quoted above, pp. 269-70) as it may be gleaned from the amended chapter on 'Value and Riches' in the third edition of the Principles. Despite 'the corrections [Say] has made in the fourth and last edition of his work', Ricardo saw no reason to retract his opinion that Say had been 'singularly unfortunate in his definition of riches and value' by 'considering value, riches, and utility to be synonymous' (1, pp. 279-81). He added: Indeed, there are many parts of M. Say's work to which I can confidently refer in support of the doctrine which I maintain, respecting the essential difference between value and riches, although it must be confessed that there are also various other passages in which a contrary doctrine is maintained. These passages I cannot reconcile, and I point them out . . . that M. Say may ... give such explanations of his views as may remove the difficulty, which many others, as well as myself, feel in our endeavours to expound them. (1, p. 281) The passages from the Traite, including that on 'productive services' (referred to by Ricardo as 'No. 4'), were therefore set out to expose the internal contradictions in Say's account and his occasional 'support' for Ricardo's doctrine. The first mention of 'No. 4' was in the following passage: When I give 2000 times more cloth for a pound of gold than I give for a pound of iron, does it prove that I attach 2000 times more utility to gold than I do to iron? certainly not; it proves only as admitted by M. Say, {see 4) 28

Ricardo's critical view of Say's performance is underscored by the following remark to Malthus: 'In Say's works, generally, there is a great mixture of profound thinking, and of egregious blundering. What can induce him to persevere in representing utility and value as the same thing?' (24 November 1820, vm, p. 302).

The appropriation of Ricardo

275

that the cost of production of gold is 2000 times greater than the cost of production of iron . . . It is the competition of the producers 'who are perpetually employed in comparing the cost of production with the value of the thing produced,' (see 4) which regulates the value of different commodities. If, then, I give one shilling for a loaf, and 21 shillings for a guinea, it is no proof that this in my estimation is the comparative measure of their utility. (1, p. 283) Then came the following, cited by Professor Hollander: In No. 4, M. Say maintains with scarcely any variation, the doctrine which I hold concerning value. In his productive services, he includes the services rendered by land, capital, and labour; in mine I include only capital and labour, and wholly exclude land. (1, pp. 283-4) From this last passage alone, one might imagine that Ricardo was extending general approval to Say's formulation, subject to the 'exclusion of land5. However, taken with the previous passage, it is fairly clear that he took the substance of Say's doctrine as being equivalent to his own: values are governed by relative costs of production which, in a 'pure' labour context, correspond to ratios of labour expenditures.29 That was the extent of the 'agreement'. The same position emerges in the following passage, from the letter to Say which accompanied a copy of Ricardo's new edition: 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. I have two loaves of bread before me, one raised on the very best land in the country, for which there is probably paid £ 3 or £ 4 p r acre for rent; the other raised on land for which there is not paid p r acre as many shillings for rent, and yet both loaves are precisely of the same value, and are equally good. You would say that in one the productive service of land was highly paid, while comparatively little was paid for the productive services of capital and labour; while in the other much was paid for the productive services of capital and labour, and little for that of the land. This is no doubt true, but the information is not useful and can lead to no inference whatever that can guide our future practice. What we wish to know is what the general law is that regulates the value of bread, as compared with the value of other things, and I think we find that one description of bread, namely, that for the raising of which little or no rent is paid, regulates the value of all bread; and that its value in relation to other things depends on the comparative quantity of labour bestowed on its 29

'To conclude', Ricardo restated his view as being that 'labour is a common measure, by which their [commodities'] real as well as their relative value may be estimated.' (i, p. 284).

276

INTERPRETING RIGARDO

production, and the quantity of labour bestowed on the production of those other things. (8 May 1821, vm, pp. 379-80, emphasis in original) Ricardo's tactic was first to 'translate' the bread example into Say's terminology, stripped of the 'utility' gloss. The resulting information, while 'no doubt true', was useless and inconsequential. The really meaningful doctrine was his own, here expressed in 'labour quantity' terms. He was not following or endorsing Say. Rather, he was inviting Say to follow him. One year later he thought he had succeeded. He wrote to Say: I am happy to observe that the difference between us is much less than I had hitherto considered it. You speak of two different utilities which commodities possess, one, which they derive from nature, without any of the labour of man, the other, which they derive exclusively from his labour. You say that for the first of these, which you call natural utility, nothing valuable can be obtained in exchange, and it is only for that portion of utility which is given to a commodity by labour or industry, for which any thing valuable can be obtained . . . You explain on these principles the case I had put to you of a pound of iron and a pound of gold, which I had supposed had exactly the same utility, though the gold was 2000 times more valuable. If we give 2000 times more for the gold than the iron, you say, it is because that peculiar utility of which only Political Economy treats, namely that given by labour, is 2000 times greater than that given to iron, and you add that the iron has 1999 portions of natural utility for which nothing is given; of which the gold has none. Although I cannot quite approve of the terms used to explain this truth, yet I do now, and always have substantially agreed in the reasoning which proves it, for I have always contended that commodities are valuable in proportion to the quantity of labour bestowed upon them, and when you say that they are valuable in proportion as they are useful, and they are useful in proportion to the quantity of labour or industry bestowed upon them, you are in fact expressing the same opinion in other words. (5 March 1822, ix, pp. 168-9) This was not a capitulation to the 'utility' doctrine: iron and gold do not exchange in the ratio of their 'utilities'. However, Ricardo could happily accept Say's weird reformulation of the 'utility' analysis which, given the way the terms were now being used, reduced to the proposition that commodities are valuable 'in proportion to the quantity of labour or industry bestowed on them'. This was the 'agreement' he had been trying to extract from Say all along. There is no convincing evidence to support Professor Hollander's judgement that Ricardo meaningfully 'subscribed to' or 'applauded'

The appropriation of Ricardo

277

Say's position, where the latter is interpreted as a doctrine of imputation. The substance of Say's argument, as Ricardo understood and 'modified' it by excluding rent from costs of production, was simply that commodities exchange in ratios given by comparative wage bills which, in a 'pure' labour theory context, was equivalent to exchange in proportion to comparative labour expenditures. The hints given by Say of 'imputation' were nothing more than verbiage for Ricardo. They were 'different words' which could be discarded without loss. Certainly, his tone was 'friendly' to Say, at least in print and in their correspondence, but this reflected both their personal relationship and Ricardo's ambition to bring Say 'on side'. It must be concluded that the Ricardo-Say-Walras tradition is purely ficticious. RICARDIAN

GENERAL EQUILIBRIUM

ANALYSIS

In this section I examine two recent attempts, by Professors Hollander and Morishima, to credit Ricardo with a (form of) general equilibrium analysis. I begin with Professor Hollander's interpretation. One possible source of confusion must be addressed at the outset. For a contemporary economist, 'general equilibrium theory' may conjure up the idea of an analysis which attempts the simultaneous determination of the (relative) prices of final products, 'factor returns' and all outputs, taking as data consumer 'tastes' (preference functions), technology, resources and the pattern of their ownership. But Professor Hollander seems not to be claiming that Ricardo's 'general equilibrium reasoning' was similarly located within the framework of simultaneous determination. Rather, his argument suggests that Ricardo's analysis of adjustment processes implies the same kind of economic interdependence that the modern theory exhibits but reduces to timeless, static outcomes. Ricardo's putative analysis bears a loose 'family resemblance' to the modern version, but they are not identical. Professor Hollander has spelt out the alleged affinity between 'Ricardian' and 'neoclassical' theory in the following terms: as in the ... neo-classical system, distribution and pricing are interdependent [in 'Ricardian' economics]. Above all, changes in the pattern of final demand can affect the wage rate, and changes in wages can affect the structure of prices. Demand-supply analysis for Ricardo - as for ... the

278

INTERPRETING RICARDO

neo-classicists - was the vehicle of determination in his general system. (1979, p. io) 30

Let us consider how, according to Professor Hollander, this 'interdependence' shows itself in the 'Ricardian' system. He presents the following 'Ricardian' analysis of a change in the pattern of final demand: If . . . we assume differential capital (machinery)—labour ratios, given aggregate capital (machinery and working capital combined) and aggregate labour, an increase in the demand for a capital-intensive good at the expense of that for a labour-intensive good will disturb the profit-rate and the price structures. There will follow in consequence an expansion of the first category of commodities and a contraction of the second. But as a result the fixed capital or machinery component of total capital will rise at the expense of wage capital as resources are transferred between industries. The wage rate (given the total labour force) must be lower and the general profit rate correspondingly higher in the new compared with the original equilibrium. (1979, p. 300)

And, one may add, the relative natural price of a 'capital-intensive' commodity in terms of one which is 'labour-intensive' will also have risen as a consequence of the change in 'factor returns' (which might, in the spirit of Hollander's account, generate further consequences if commodities are price-elastic in demand). Hence the interdependence of distribution and pricing. A further source of interdependence presents itself if allowance is made for alterations in the 'overall circulating-fixed capital ratio' in consequence of a change in wage and profit rates (p. 680). In the reverse case to that detailed above, machinery would be substituted for labour as wages rose, thus opening up another channel of interdependence between 'factor' and 'product' markets. But in all cases, the crucial point is that the 'equilibrium' values for wages, profits, commodity prices and outputs only emerge as the outcome of allocative processes. This, for Professor Hollander, is Ricardian 'general equilibrium' analysis. If, however, we were to search for examples of such developed reasoning in Ricardo's writings, we would do so in vain: something acknowledged by Professor Hollander himself (see, for instance, Hollander, 1979, p. 679). He is therefore presenting a 'fundamentally important core' of general equilibrium analysis in 30

In the ellipses, Hollander brackets Smithian economics with the other two 'systems'.

The appropriation of Ricardo

279

'Ricardian' economics which Ricardo himself did not rely upon in his own work. But in what sense does it exist at all? Professor Hollander candidly acknowledges that 'a very tricky problem' is involved, defending himself thus: Now were the assumptions of uniform factor ratios and of constant wages used by Ricardo over and again without significant exception, the implication would be that they represent features of his 'basic model' - in which case it would be unconvincing to argue that Ricardo 'could' easily have opened his model in these respects. This objection would be compounded in the event that the techniques of resource allocation were as sparse in his work as is commonly believed. The position adopted here is, however, based upon the twofold demonstration, first, that Ricardo, on matters of fundamental import and not merely casually, himself released the two simplifying assumptions and, secondly, that he himself applied the principles of allocation — demand—supply analysis, profit rate equalization — to a wide variety of issues in a quite sophisticated way. Needless to say, he did not consider all the possible situations where a relaxation of the two key assumptions have profound consequences, or all those that require treatment in terms of allocation theory. But for us to relax the assumptions and to apply the theory of resource allocation to a broader range of issues is to follow along a route laid out by Ricardo himself, using tools of analysis provided by Ricardo. (1979, pp. 680-1, emphasis in original) The 'general equilibrium analysis' is thus obtained by treating wages as essentially variable, assuming differential (and sometimes endogenously variable) 'factor ratios', and applying 'the theory of resource allocation to a broader range of issues' than those to which Ricardo applied it. The various ingredients were all present in Ricardo's work and that, for Professor Hollander, is decisive. Even if we accepted all other aspects of this interpretation, it would be odd to describe a body of reasoning as 'fundamentally important' if it is obtained only by assembling the analytical pieces in ways which Ricardo himself did not do. It would be more accurate, and certainly less misleading, to say that a (form of) 'general equilibrium analysis' can be created out of 'Ricardian economics' rather than locating its existence within that discourse. Even on his own terms, Professor Hollander has overstated his case. But are those terms acceptable? I think not. The 'new view' interpretation of Ricardo's treatment of wages, which allows Professor Hollander to credit Ricardo with 'essentially variable' wages, rests on the elevation of 'unintegrated knowledge' (represented by

280

INTERPRETING RIGARDO

the new view passages) to the status of Ricardo's 'true doctrine'. Once the 'natural wage' is restored to its central location in Ricardo's work, the 'essential variability' of wages reduces to 'temporary' departures of market from natural wages, analogous to discrepancies between the market and natural prices of produced commodities. And these 'temporary' variations were merely incidental to Ricardo's treatment of 'the laws which regulate natural prices, natural wages and natural profits' (i, p. 92) which was, of course, his prime and enduring theoretical objective. Even in his later writings, when Ricardo showed signs of weakening over the natural wage analysis, his commitment to a 'fixwage' analysis of distribution was unswerving. Similarly, although it is undoubtedly true that Ricardo engaged in 'process' or 'resource allocation' analysis - and analysis of'temporary' phenomena, constrained to the analytical 'interval' - this was not his dominant preoccupation. As he wrote to Malthus, 'I acknowledge the intervals on which you so exclusively dwell, but still they are only intervals' (24 November 1820, vm, p. 302). Far from being the central objects of investigation, 'temporary' phenomena were dwarfed by 'the great enquiries on which to fix our attention', namely, 'the rise or fall of corn, labour and commodities in real value, that is to say the increase or diminution of the quantity of labour necessary to raise corn, and to manufacture commodities' (letter to Malthus, 28 September 1821, ix, p. 83). In these, 'great enquiries' the 'temporary' influence of supply and demand - the 'process analysis' - was either marginalised or removed from the analytical frame. To insist on its reintroduction as the 'vehicle of determination' is merely to impose a view of what Ricardo should have done. And this, masquerading as 'interpretation', somewhat inevitably gives rise to absurdities. Such is the case when Professor Hollander attempts to 're-write' Ricardo's treatment of distribution in terms of process analysis, with variable 'factor returns' and differential 'factor proportions', the necessity for the last assumption being that it highlights the soughtafter 'interdependence' between distribution and pricing. This exercise is critical for Hollander's case; it would be scarcely credible if Ricardo, as an overlooked pioneer of 'general equilibrium' economics, had divorced the analyses of distribution and general pricing. Now, we know that the assumption of unequal 'factor proportions' was adopted by Ricardo in the chapter 'On Value', but

The appropriation of Ricardo

281

the 'process analysis' was not centralised. Hollander describes Ricardo's actual approach in the following terms: It was Ricardo's frequent practice in analysing the inverse profit-wage relationship to derive a new (lower) general profit rate in consequence of a change (rise) in wages by use of the standard measuring procedure and then to apply the new rate in the calculation of the price structure. (1979, p. 11)

The 'standard measuring procedure' rests on the (tacit) assumption that a commodity is produced with the same 'capital structure' as that selected for the 'invariable standard' in terms of which all prices are expressed. If, as Ricardo did, it is assumed that money wages rise by a certain amount, a new profit rate may be calculated on the production of the 'measured' commodity since, by assumption, its total cost of production is unchanged.31 Then, with given conditions of producing other commodities, the new wage-and-profit couplet may be imposed to arrive at the new set of prices. But this procedure is precisely the one which Professor Hollander does not want to emphasise, for it implies, first, that distribution is 'solved' at a stage analytically prior to the determination of commodity prices generally; and secondly, that 'process analysis' need not be introduced to determine those prices. With given methods of production, different 'capital structures' and known rates for wages and profits, natural prices can be determined without knowledge of output levels. To that extent the 'demand and supply' mechanism is otiose.32 How does Professor Hollander cope with Ricardo's 'recalcitrance'? He writes: But this does not constitute an account of process analysis entailing causal linkages; it is merely a device designed to yield appropriate predictions regarding the new equilibrium values of key variables following a disturbance to the wage rate. Allocation of resources and thus price movements are in fact governed by profit rate differentials between industries created by the initial disturbance to wages; and a new equilibrium general rate of profit is yielded together with the new equilibrium price structure as a consequence of the reallocation. Distribution is not in any causal sense 'prior' to pricing. (1979, p. 11, emphasis in original) 31

32

T o do this rigorously more stringent conditions may need to be imposed. But, for the sake of this discussion, it is unnecessary to pursue the logic further. But it is not entirely absent from the c h a p t e r ' O n V a l u e ' (see i, p p . 4 0 - 2 ) . However, in t h e papers On Absolute Value and Exchangeable Value, where the same issues a r e discussed at length, the 'process analysis' is not mentioned at all.

282

INTERPRETING RIGARDO

In other words, he substitutes a method of analysis which Ricardo did not make central for the one which he did. More recently, this inversion of reality has become quite astonishing. Still with reference to the 'standard measuring device', Professor Hollander has claimed: But here, as elsewhere, Ricardo was groping his way to a desirable exposition. In the full analysis of this mechanism of adjustment, the increase in wages affects the whole structure of outputs in the system, with the new cost prices and the new general level of profits emerging together as a consequence of a reallocation of resources. And . . . there is nothing in Ricardian logic to preclude a playback from the pattern of activity upon distribution, so that any 'forecast' of the final outcome requires knowledge of the entire set of final demand curves. (1987, pp. 360-1)

It is deeply perplexing how Ricardo could be described as 'groping towards' a 'general equilibrium' analysis which Professor Hollander evidently regards as 'desirable', but which Ricardo did not even attempt to develop for use in the relevant context. The 'groping' is Hollander's wishful thinking, with the unreality of his vision further evidenced by 'the entire set of final demand curves' which he thinks he can find in 'Ricardian logic'. It is a long haul from Ricardo's acceptance of the 'law of demand' as a market phenomenon to a comprehensive set of purely formal analytical constructs. I suggest that we take Ricardo's own analysis of distribution rather more seriously. We should indeed 'follow along a route laid out by Ricardo himself, using the tools of analysis provided by Ricardo' and not divert him to a path of our choosing, diminishing the importance of his 'tools of analysis' in the process. Proceeding in this fashion, the conclusion must be that his analysis of distribution in the context of differing capital structures does not conform to the 'general equilibrium' mould. To 'extend' the analysis he did not use to incorporate variable capital structures, which he left unintegrated, might then prove an absorbing pastime, but we would not thereby be misled into thinking that we are depicting 'the historical Ricardo', to borrow Professor Hollander's felicitous expression (1979, p. 3). It should also be recalled that Ricardo's most developed analytical framework, within which he sought to explicate the central links between changed agricultural conditions of production and 'permanent' movements in profitability, was provided by the 'pure' labour theory of value. And in terms of that framework, applied to that theoretical objective, all the interdependencies which Professor

The appropriation of Ricardo

283

Hollander regards as being so important are nowhere to be found. 'Permanent' movements in profitability are defined with reference to the given (natural) wage; there is no 'interplay' between product and 'factor' markets (to use Hollander's favoured 'neoclassical' terminology); and, by taking the production conditions of commodities as given, we do not have to know output levels (or even the values of uniform wage and profit rates) to determine the natural prices of commodities other than corn. We may not applaud such an analysis, but neither can we wish it away. There is a further peculiarity in Hollander's interpretation. He stresses repeatedly that Ricardo typically used models which rested on 'simplifying assumptions', relaxed in some contexts but not in others (the 'simplifying assumption was Ricardo's "trademark"', 1979, p. 644). I would not disagree, although there is at least one distinction to be made between assumptions which simplify but are doctrinally important (such as a given natural wage) and those that merely simplify (for example, that there is only one natural wage). Be that as it may, even on Professor Hollander's account it would have been quite uncharacteristic for Ricardo to relax all those assumptions in order to use the more complex 'general model' (which, to repeat, he never did). The more valid inference, on this basis alone, is that the economics of the 'historical Ricardo' was methodologically opposed to 'general equilibrium' analysis and in this respect has, pace Hollander, more in common with the '"Cambridge" school'. Wherever we turn, that 'fundamentally important core of general equilibrium economics' is nothing more than Professor Hollander's shadow. Professor Morishima's variation on the 'general equilibrium' theme is no more successful. But at least he is clear on what he means by general equilibrium: 'The purpose of the general equilibrium theory . . . is to construct a system of conditions which are necessary and sufficient to determine all the variables contained in the system'. It is with reference to such a conception that he describes Ricardo as 'the fore-runner, if not the founder, of the general equilibrium school' (1989, p. 149). 'Ricardo's economies', he writes elsewhere in his book, 'is, in fact, a general equilibrium theory, which determines' land intensities and rents 'together with' wages, prices, output and the general rate of profit (p. 43). It is purportedly, a theory of the simultaneous determination of all these economic variables.

284

INTERPRETING RIGARDO

On closer inspection of Professor Morishima's text, it soon emerges that despite Ricardo's pathbreaking work in the field, he never himself produced an explicit general equilibrium analysis. Thus we are variously told that the general equilibrium system was 'concealed' (p. 3) in Ricardo's work, that it was merely 'implicit' (p. 8), that it is only obtained by 'excavating its parts . . . and putting them together properly' (p. 18), that Ricardo did not 'mathematically spell out what conditions were necessary and sufficient to determine the values of the variables with which he was concerned' (p. 149), and that Ricardo 'failed to put the equilibrium supply-demand conditions in the exact form of simultaneous equations' (p. 182). Professor Morishima merely presents yet another reassembly of theoretical bits and pieces which he then combines to form a general equilibrium system. His cardinal error, much like Professor Hollander's, is in thinking that his creation belongs to Ricardo. There are other similarities. In particular, Professor Morishima's reconstruction is, like Hollander's, based on the notion that the wage is an endogenous variable which finds its 'natural' level in the stationary state alone. And he also follows Hollander in taking unequal 'capital structures' as a basic premise of the model.33 I will not repeat my objections to making these attributions in furtherance of the 'general equilibrium' case, although the 'value' aspect of Morishima's interpretation is, perhaps, deserving of a few comments. The 'general equilibrium theory' which Professor Morishima 'finds' in Ricardo is allegedly the 'core' theory which he also 'finds' in Walras (and Marx) (1989, p. viii). On turning to Walras's Elements, however, numerous statements are encountered to the effect that a subjective, utility- (or rarete-) based theory of value is an essential theoretical ingredient. For example: 'the theory of exchange based on the proportionality of prices to intensities of the last wants satisfied [raretes] . . . constitutes the very foundation of the whole edifice of economics' ([1926] 1954, p. 44), emphasis in original, cf. p. 181). To bracket Walras's (extant) general equilibrium system with Ricardo's (inextant) version might therefore seem an even more impossible task, given Ricardo's 'failure' to develop a utility-based theory of value in the Walrasian direction. Yet having convinced himself of the (other) profound similarities between Walrasian and 33

On 'wages', however, he differs from Hollander in advancing a rigid 'wage fund' interpretation. The dubious nature of his interpretation is well exposed by Hollander (i979>PP- 326-39).

The appropriation of Ricardo

285

Ricardian economic analysis, Professor Morishima avers that 'the dissimilarity usually emphasized by historians of economic thought - that is the labour theory of value supported by Ricardo . . . versus the scarcity theory of value by Walras - is of minor or secondary importance' (1989, p. 4). The difference is 'of minor or secondary importance' only because Professor Morishima asserts it to be so. Yet only two pages earlier, he makes the point, when censuring William Jaffe's interpretation of Walras, that if 'an arbitrary selection deletion and inclusion - of an economist's works is permitted, anything can be said of him' (p. 2). Quite so. Not content with having asserted the unimportance of the 'labour versus scarcity' difference between Ricardo and Walras on 'value', Professor Morishima endeavours to detract from the significance of the labour theory in Ricardo's writings. With 'no idea of marginal utilities', he opines, Ricardo 'used the labour theory of value' (1989, p. 8), his (unstated) implication being clear. However, although Ricardo did have a rough notion of marginal utility (but not of diminishing marginal utility), he was steadfast in regarding the utility doctrine, as he understood it, as irrelevant to his 'value' requirements. There is no basis for Morishima's insinuation. Morishima also claims that the labour theory was adopted by Ricardo to compensate for his mathematical ignorance (1989, pp. 9-10). If only Ricardo had possessed the ability to derive 'laws of prices directly from price-cost equations, he would have agreed to discard the labour theory of value' (1989, p. 10). The problem with this form of statement is that we simply do not know, and cannot know, what a different Ricardo might have agreed to. Even Professor Morishima, with his self-proclaimed superiority over historians of economic thought, has no special gifts in this regard. 34 However, we can make interpretative judgements on extant texts, and this route is also followed by Professor Morishima in support of his 'value' argument. Thus he claims, solely on the basis of the third edition of the Principles, that Ricardo 'insisted' on the general deviation between comparative labour expenditures and natural price ratios; for Ricardo, the pure labour theory only held 'in a highly Professor Morishima apparently takes a dim view of mere 'historians of economic thought': 'I have never been a historian of economic thought but have been an economic theorist throughout my life. With such a speciality, I believe, I am allowed to concentrate solely on . . . main works; and by making this constraint I am able to read these works more deeply and more rigorously than specialists in the history of economic thought' (1989, p. 3).

286

INTERPRETING RICARDO

exceptional case' (1989, p. 51). These assertions are without foundation. The main thrust of the rewritten chapter 'On Value' in the third edition of the Principles was precisely to defend the pure labour theory; and this defence incorporated the claim that, for the general mass of commodities, a 'pure' labour theory does apply. The extant reality is the direct opposite of that portrayed by Professor Morishima. A common feature of the interpretations of Hollander and Morishima is that both rely, primarily, on what Ricardo might have done had he followed a line of reasoning supplied by them. It is a telling coincidence. RICARDO AND SRAFFA

In assessing Sraffa's 'Ricardian' pedigree, the subject of this section, I first report various salient features of Sraffa's own work. I begin with his analytical approach, announced in the Preface to his Production of Commodities by Means of Commodities: No changes in output and (at any rate in Parts 1 and 11) no changes in the proportions in which different means of production are used by an industry are considered, so that no question arises as to the variation or constancy of returns. The investigation is concerned exclusively with such properties of an economic system as do not depend on changes in the scale of production or in the proportions of'factors'. This standpoint, which is that of the old classical economists from Adam Smith to Ricardo, has been submerged and forgotten since the advent of the 'marginal' method. (Sraffa i960, p. v)

The assumption of given and unchanging outputs rules out the operation of any supply and demand mechanism following a disturbance to (relative) prices. And indeed, Sraffa's Production of Commodities contains neither a (subjective) demand analysis per se, nor an analysis of market clearing mechanisms. 'Values' or relative prices are defined solely by a uniform rate of profit condition. The determination of relative prices and, more particularly, the investigation of price movements following assumed changes in distribution, are central objects of investigation for Sraffa. For the purposes of his investigation, total outputs are given, one of the distributive variables (wages and profits) is also given, the other variable is calculated as a residual, and prices are determined at a

The appropriation of Ricardo

287

secondary stage in the analysis {after the determination of the unknown distributive variable). It is within this analysis that Sraffa's 'Standard system' and 'Standard commodity' are encountered, both of interest for the present study. Sraffa's model of an economy is described by a set of physical input-output equations, with quantities of labour and commodities on one side of the equations and outputs on the other. Those commodities which are direct or indirect inputs to the production of all commodities are termed 'basic commodities' (as contradistinguished from 'non-basics') and it is only their conditions of production which are relevant to the determination of the general rate of profit. At the expense of simplification, the 'Standard system' is obtained by first segregating the equations for 'basics' and then applying scalars to each equation such that the proportions between 'basic' inputs are made equal, in aggregate, to the proportions between 'basic' outputs; finally, the resulting equations are uniformly scaled up so that the total labour requirement in the 'Standard system' is equal to that in the original set of equations for basic and non-basic commodities. The 'Standard commodity' is defined as the net product of the 'Standard system'. Having obtained these analytical devices, Sraffa demonstrates an inverse, linear relationship between the share of wages in the Standard net product (i.e., as a proportion of the Standard commodity) and the rate of profit, the latter defined as the ratio of the residual net product to the aggregate means of production in the Standard system. This exercise is conducted in purely physical terms, since the net output and input of the Standard system are the same (composite) commodity. Sraffa further shows that the rate of profit, so calculated, must apply in the 'actual' economic system from which the Standard system is obtained. Then, with wages given in terms of the Standard commodity and the (uniform) rate of profit determined as a residual (or, alternatively, the rate of profit given and wages determined as the residual), prices, both of basics and non-basics, are also determined using the Standard commodity as numeraire. Price movements are observed by redistributing the Standard net product and calculating the resulting prices in terms of the Standard commodity. Only in the event that a commodity is produced with the same input proportions as the Standard commodity will its price remain invariant to these 'notional' redistributions. However, this is a most unlikely outcome in Sraffa's models, which are partly

288

INTERPRETING RIGARDO

designed to highlight the 'complicated patterns of price movement' following changes in distribution, thus undermining the 'notion of capital as a measurable quantity independent of distribution and prices' (Sraffa i960, pp. 37-8). Similarities between Sraffa's analysis and Ricardo's are immediately apparent; or, more precisely, they are if Sraffa's interpretation of Ricardo is accepted. So, in the case of Ricardo's putative 'corn model' analysis (or 'corn ratio' theory of profits), we are confronted by a simple Standard system, with corn as the only 'basic', the rate of profit determined as a residual in purely physical terms given the corn wage, and prices determined at a secondary stage in the analysis on the uniform rate of profit principle. There are also striking similarities between the Standard analytical constructs and (Sraffa's interpretation of) Ricardo's 'medium' standard in the third edition of the Principles. The 'problem of value which interested Ricardo' was, according to Sraffa, 'how to find a measure of value which would be invariant to changes in the division of the product; for, if a rise or fall of wages by itself brought about a change in the magnitude of the social product, it would be hard to determine accurately the effect on profits' (1, p. xlviii). Hence the merit of the 'medium' standard: 'If measured in such a standard, the average price of all commodities, and their aggregate value, would remain unaffected by a rise or fall of wages' (1, pp. xliv-xlv). Ricardo's 'problem', as diagnosed by Sraffa, concerned the notional redistribution of a given social product; and through the selection of his third edition standard, Ricardo was allegedly seeking to recapture the advantages of his 'corn model' at the 'macro' level of analysis. These were to count among Sraffa's achievements. His Standard system functions as an elaborate 'corn model' on a 'macro' level (remembering that the Standard system employs the total labour requirement of the original system); and, to underline the similarity, Sraffa himself tells us that his Standard commodity had 'evolved' from Ricardo's conception of 'a medium between two extremes' (Sraffa, i960, p. 94). It is, perhaps, little wonder that some of Sraffa's followers should have claimed that Ricardo's search for a 'perfect' measure of value had at last found a solution within the covers of Sraffa's book (although Sraffa held back from making such a claim explicitly).35 35

See Dobb (1973), pp. 263-6, and Dasgupta (1985), pp. 54-9.

The appropriation of Ricardo

289

The implication of my study of Ricardo is that many of the above similarities are misleading, if not illusory. The jewel in the crown of the 'Sraffian' Ricardo is undoubtedly the 'corn model', hence the vigour with which this central element in Sraffa's interpretation has been defended by those who are most vociferous in championing a Ricardo-Sraffa lineage.36 But, as I have argued, Ricardo's writings are incapable of supporting the weight of the interpretation and, indeed, some of the evidence adduced by Sraffa is demonstrably weak. Sraffa himself provides the clue to what may have happened. Thus, in Production of Commodities, he disclosed that 'it was only when the Standard system and the distinction between basics and nonbasics had emerged in the course of the present investigation that the ['corn model'] interpretation of Ricardo's theory suggested itself as a natural consequence' (Sraffa i960, p. 93). Here, I suspect, we find the clue to Sraffa's interpretative error: 'sense' had been projected onto Ricardo's texts, not inferred from them. The question of the relationship between Ricardo's standard (s) and Sraffa's Standard commodity and system is rather more complicated, although the evidence again suggests a large degree of projection on Sraffa's part. For Sraffa, the investigation of price movements following notional changes in distribution, using the Standard commodity (or one of its derivatives) as numeraire, was a prime theoretical objective. The nearest Ricardo came to a similar analysis was with the 'curious effect' in the first two editions of the Principles, where he, like Sraffa, seemed keen to emphasise the significance of purely distribution-induced price movements. Moreover, in the context of his analysis Ricardo proceeded by assuming changing distributions between wages and profits of a given value of output, produced by labour alone. The profit rate, calculated as the ratio of profit to wages, together with the postulated money wage rate, were then used to compute the prices of commodities produced with the aid of fixed capital of differing degrees of durability. This analysis shares with Sraffa's the characteristic of distribution being 'solved' at a stage logically prior to the determination of general prices. By the third edition, however, Ricardo's emphasis had shifted dramatically. Although the rewritten chapter 'On Value' also contained an analysis of price movements following changes in distribution, what Ricardo desperately wanted, with increasing clarity of 36

See, in particular, Eatwell (1975a), Garegnani (1982), and de Vivo (1987b).

INTERPRETING RIGARDO

purpose if not success, was a 'perfect' standard in terms of which (natural) price fluctuations would be limited to those reflecting changes in 'real value' (altered labour expenditures), even though capital structures differed. The 'medium' standard was a secondbest solution, designed to achieve 'pure' labour theory results for the 'general mass' of commodities (if only by fiat), while all the remaining distribution-induced price movements were seen as indicative of a defect in the standard. Ricardo's search for a 'perfect' measure of value, and his choice of the 'medium' standard in the third edition of the Principles, was inextricably linked with his attempt to rescue the 'pure' labour theory. This consideration alone is sufficient to open a breach with Sraffa, who had no such objective, and thus to falsify the claim that Sraffa may have consummated Ricardo's search for a 'perfect' measure. But the differences between Sraffa and Ricardo go deeper. Sraffa's Standard commodity and system are defined for a given and unchanging matrix of inputs and outputs, whereas a prime function of Ricardo's standard was to sanction the use of the labour theory in the demonstration of his 'agricultural' thesis: the proof that 'permanent' reductions in general profitability must result when an increasing corn output requires proportionally greater inputs of labour. Sraffa's analytical devices would have no relevance in this Ricardian context.37 As for Sraffa's suggestion that Ricardo's 'medium' standard was designed (if only in part) to ensure the aggregate price-invariance of national output to changes in distribution, I have found no evidence whatever. Ricardo made no mention of the 'price-balancing' role for the 'medium' standard, and his own numerical examples suggest a quite different outcome. Two further considerations are also inimical to Sraffa's suggestion. His interpretation presumes that Ricardo's analysis of distribution was framed at the 'macro' level of social aggregates (hence the insistence by 'Sraffians' that Ricardo was a social surplus theorist). However, Ricardo made increasingly clear that the detailed analysis of distribution was to be conducted in 'micro' terms; consider, for example, the amended sentence in the chapter 'On Value' in the third edition of the Principles'. 'It is according to the division of the whole produce of the land of any particular farm [replacing 'of the 37

For similar criticisms of the claim that Sraffa solved Ricardo's problem, see Blaug (1985a), (1985b, pp. 142-3), (1987), Ong (1983), Roncaglia (1978, p. 78), and Tosato (1985).

The appropriation of Ricardo

291

whole produce of the land and labour of the country'] . . . that we are to judge of the rise or fall of rent, profit, and wages' (1, p. 49, my emphasis). This is not to deny his interest in the consequent movements of aggregate shares, which were evidently considered to be of great importance (aggregate profit was, after all, regarded as the main fund for capital accumulation). But it is to deny that the principal distribution analysis was conducted in terms of social aggregates.38 The interpretation also presumes that Ricardo's concern was with purely notional changes in distribution. It is true that in the chapter c On Value' in the Principles, and in the papers On Absolute Value and Exchangeable Value, where a similar analysis was undertaken, Ricardo refers to variations in money wages resulting from either altered supply and demand conditions in the labour market or from changes in the conditions of producing wage-goods. Given his various and changing objectives in these contexts (such as exposing the inadequacies of a labour-commanded measure of value, demonstrating the 'curious effect', illustrating properties of the 'pure' labour theory, and reluctantly conceding the limitations of that theory), the source of the wage-variations was apparently considered to be a matter of indifference. But this is not the case when we turn to his treatment of distribution as a serious object of investigation in its own right (as in the chapter 'On Profits'), where the principal wage variations are those resulting from altered conditions of producing wage-goods, not from 'notional' redistributions of a given product. The 'price-balancing' standard addresses a problem which was peripheral to Ricardo's main concerns. The above considerations also bear on the question of Ricardo's alleged method of holding constant all outputs and techniques of production (see above, p. 286) and the related question of his use (or lack of it) of supply and demand analysis. There are places, notably the chapter 'On Value' and the papers On Absolute Value and Exchangeable Value, where Ricardo at least came close to exhibiting the approach adopted by Sraffa. But again, the similarities with Sraffa's work evaporate when we turn to 'the principal problem in Political Economy' — the analysis of 'the natural course of rent, profit, and wages' (1, p. 5) - in which the output and conditions of producing corn, in particular, were certainly not treated as given 38

Cf. Cannan (1893, pp. 340-1), and, more recently, Tosato (1985).

INTERPRETING RICARDO

constants. To that extent, the issue of returns to scale could be said to arise at the very heart of the Ricardian system, in spite of Ricardo's own refusal to acknowledge functional relationships between physical conditions of production and outputs. The differences between Sraffa and Ricardo would be even greater if, as Samuel Hollander has claimed on behalf of Ricardian analysis, 're-establishment of a system of relative prices following . . . a variation in wages occurs by way of changes in output allowing for the condition of equality between quantities demanded and supplied in commodity markets' (1979, pp. 684—5). But m saying this Professor Hollander appears to have confused his own Ricardo creation with the (truly) 'historical Ricardo'. As I have argued, there are places, grudgingly acknowledged by Professor Hollander when he declaims against Ricardo's use of the 'standard measuring device', where Ricardo removed the 'process analysis' to the sidelines, took at least some conditions of production as given and unchanging, assumed different values for the wage rate and, like Sraffa, determined natural prices on the uniform profit rate principle without explicit regard to conditions of market equilibrium. Professor Hollander's assertion that Ricardo 'nowhere turned his back on ... process analysis' (1979, p. 685, my emphasis) is simply false, even on Hollander's own interpretation. However, although Ricardo at times abstracted from the 'process analysis' in places of great analytical importance, at others he did not. In this respect, there are dissimilarities with Sraffa's theoretical system: dissimilarities which, in some 'Sraffian' commentaries, are masked by the presentation of a 'Ricardian' approach which has little or no supply and demand ingredients (see de Vivo 1987a, for a recent example). The fact is that Ricardo, unlike Sraffa, did provide a supply and demand rationalisation for the convergence of market prices to natural prices (but not for the level of natural prices). And in some places, such as the 'taxation' chapters in the Principles, supply and demand mechanisms figure prominently. Ricardo did not introduce supply and demand reasoning when discussing certain key issues, but a role for supply and demand (involving changes in outputs) was not comprehensively excluded from his work, as it was from Sraffa's. A more pronounced similarity exists over the general treatment of distribution, at least to the extent that both Ricardo and Sraffa took either wages or profits as given (wages, in the case of Ricardo) and

The appropriation of Ricardo

293

determined the other as a residual. Mention should also be made of their shared tendency to isolate dominant causal relationships and, additionally, of certain 'negative' similarities, such as the absence of a 'neoclassical' demand analysis and an account of production and distribution couched in terms of'factors of production'. In these and the other respects that I have indicated, there are broad-brush resemblances. It is when we look more closely that the differences appear.

CHAPTER 7

Concluding remarks

Ricardo's early writings on profitability, discussed in chapter 2, have come to be identified with the 'corn model' analysis, attributed to Ricardo by Piero Sraffa. The reasons for the widespread popularity of Sraffa's interpretation are not difficult to fathom. For some time past it had been generally thought that Ricardo was preoccupied with the Corn Law issue, and that he had a fondness for heroically simplified models, directed to the solution of real-world problems. The 'corn model' interpretation fitted the bill on both counts. Additionally, it had the advantage of pedagogical simplicity, it neatly rationalised the statements of a 'regulatory' role for farmers' profits, it answered the question, raised at the beginning of the nineteenth century, of how Ricardo could sustain his position while subscribing to the 'adding up' treatment of pricing, and it gave order to a confusing body of primary literature. Admittedly, there was no extant record of the model in Ricardo's writings, but various passages could be construed in its reflection; and there was, after all, the Essay on Profits, which had earlier tempted Wesley C. Mitchell to claim a surviving 'corn model' analysis. It is no wonder that the textbooks were rewritten. The implication of my study is that they were rewritten in the service of mythology. Ricardo's writings do not yield any decisive evidence in favour of Sraffa's interpretation: an interpretation which, by imposing sense, obscures the chaotic nature of Ricardo's early writings, as he struggled to articulate new ideas within an inherited discourse. Thus I reject Sraffa's implication that Ricardo had broken decisively with the 'competition of capitals' doctrine in 1813-14, his further implication that the 'adding up' treatment of pricing was thenceforth regulated to a limbo status until it was abandoned in 1815, and his suggestion that Ricardo had penetrated the fog of analytical confusion with a neat, logically consistent 294

Concluding remarks

295

model. Ricardo was passing through a stage marked by the absence of logically consistent models, although this is not to deny that the 'models' he did develop were highly simplified, or that they were (or at least, came to be) focused on the Corn Law issue. In those last two respects, received opinion is essentially correct. The position I have reached is unlikely to find favour with the Sraffians. For them, Ricardo's alleged 'corn model' analysis is an 'analytical success' (Eatwell 1975a), a 'brilliant simplification' (Walsh and Gram 1980, p. 89), which testifies to Ricardo's 'originality and profundity' (Garegnani 1982). The great attraction seems to be that the interpretation transforms Ricardo into the brilliantly precocious child who reaches full maturity in his Sraffian reincarnation. With Ricardo, the Sraffians claim to find their doctrinal roots, from which they evidently derive the pride and fortitude much needed in their confrontation with a hostile 'neoclassical' world. There is nothing new in this self-serving adoption of a prominent historical figure. Nor, more specifically, is there anything new in Ricardo's forced enlistment within a retrospectively conceived lineage: witness Alfred Marshall's interpretation, which drew from Ashley (1891) and Patten (1893) comments to the effect that the new discipline should stand on its own merits, not cling to an artificially reconstructed past for support. Whether the Sraffians will, or could, heed the same advice, are both moot points. The distorting influence of the Sraffians was also noted in relation to the material discussed in the first section of chapter 3, with the attempt to saddle Ricardo with a static analysis of distribution. But this aspect of their interpretation has failed to achieve widespread acceptance, rightly so. In the Principles, as in the Essay, Ricardo's main analytical preoccupation was with the dynamic course of wages, profit and rent, under the influence of progressively diminishing agricultural returns. Unfortunately, the message sometimes transmitted has been that this was Ricardo's sole objective and, relatedly, that the falling rate of profit argument was his only case against agricultural protection. Both these textbook verities stand in need of correction. The associated view, that Ricardo envisaged a catastrophic fall in general profitability if the Corn Laws were not repealed, does supply an easily digested explanation for his use of the falling profitability argument, yet it must be rejected altogether. The significance of the argument for Ricardo was in the com-

296

INTERPRETING RIGARDO

parative sense that although the fall in profitability would be glacially slow, it would not occur at all with an unrestricted corn trade. Samuel Hollander claims that it is virtually all aspects of the received wisdom which must be jettisoned. For him, Ricardo's primary objective was to provide a general framework within which to analyse the inverse relationship between wages and the rate of profit; the demonstration of the falling rate of profit was nothing more than a 'particular application' of the basic model; and, with reference to the Corn Laws, Ricardo espoused the falling profitability argument merely for the sake of'public consumption'. I will not repeat my objections to these claims, but I will speculate on the source of the confusion. Professor Hollander has convinced himself that Ricardo's work is in the Marshallian tradition, and that it incorporates a fundamentally important core of general equilibrium analysis (of sorts). This interpretation could be achieved only by sifting Ricardo's writings through an interpretative sieve, patterned in the image of (later) purely theoretical analysis. One danger of such an exercise is that the objects of Ricardo's work will be sieved away, or at any rate separated from the theoretical material, thus leaving an autonomous, and artificially created, sphere of theoretical discourse, and a residue of 'applied' writings. Thus interpreted, Ricardo the 'pure' theorist would (or should) have focused on the general principle of the wage-profit relationship rather than an 'application', hence the privileging of the general principle as his 'primary objective'. Yet the uncomfortable truth remained that it is the 'application' which pervades Ricardo's work, most inconveniently of all (for Hollander) in the central chapters on distribution in the Principles. Hollander's resolution of this paradox - the dominance of the 'application' over the 'general principle' - is that the 'application' was itself divorced from the external reality. It was an 'application' without application. In this way, Ricardo qua 'pure theorist' is preserved intact, although the reasons for the characteristic shape of his 'theory' are thereby rendered virtually unintelligible. Hence Professor Hollander's uncharacteristic puzzlement when he asks why Ricardo built models incorporating the principle of diminishing agricultural returns. Why indeed, on his interpretation. The difficulties that Professor Hollander creates for himself are magnified when it is considered that the new view analysis, incorporating a rising corn price, rising money wages, and falling

Concluding remarks

297

profitability, is, for him, 'Ricardo's main concern in devising his growth model' (1987, p. 235): a 'growth model' which Hollander is on the one hand forced to split off from both Ricardo's 'distribution theory' proper and his case against the Corn Laws, and on the other hand is determined to emphasise, since it is fundamental to his 'neoclassical' reconstruction. This, in turn, traps him in a new web of difficulties surrounding Ricardo's treatment of wages, the topic discussed in the second section of chapter 3. The presence of new view material in Ricardo's writings is indisputable, but it must be set against the rich textual support in favour of the traditional natural wage interpretation. Professor Hollander's response has been to seek a unifying coherence (most conspicuously so in his 1990 paper), based around the new view, which is achieved by demoting the natural wage analysis to the low-level status of a 'simplification', or 'working hypothesis'; and by concocting a reconciliation between the new view analysis and Ricardo's central doctrine on the causes of'permanent' movements in profitability. There may be something quite appealing about this style of interpretation: I doubt whether many historians of economic thought would disagree with Hollander's advice that 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' (Hollander 1982). And in the case of Ricardo, who has acquired the reputation of being 'a very powerful theorist' (O'Brien 1981), often as his only redeeming feature in the eyes of 'mainstream' economists, the presumption may seem irresistible. However, it is also possible that the 'discovered' coherence is merely imposed, and this has been the case with Hollander's new view interpretation: an interpretation which, apart from straining the meaning of a 'simplifying device' [etcetera) to include the repeated affirmation of something which Ricardo 'really believed' to be false (that the natural wage is the active centre of gravity for market wages in all stages of society), imposes a foreign meaning on his stated doctrine of 'permanent' movements in profitability. The evidence of inconsistency certainly disappears on Hollander's interpretation, as if by magic. The opposite error of interpretative judgement is made by commentators who have either ignored the scattered new view passages, or have invoked them as part of the natural wage analysis. Here again, Ricardo's famed consistency has been purchased in a

298

INTERPRETING RICARDO

debased currency, but for more understandable reasons: the natural wage analysis, the thesis on 'permanent' movements in profitability, and Ricardo's 'discrete' method of analysis, were an organic whole, at the heart of his mature theoretical system. Yet he was neither immune from critical pressure, nor a paragon of consistency, and his sails were trimmed to the persistent wind of Malthusian argument. Hence the new view material and, more particularly, its intrusion into the chapter 'On Profits'. With the 'law of markets', discussed in the third section of chapter 3, we encounter a doctrine or, better still, a creed, to which Ricardo was unerringly committed. My decision to incorporate the section was taken with reservation, since I did not believe I could improve on Professor Terence Hutchison's work in this area. That remains my judgement, but the 'law of markets' material is so important to the understanding of Ricardo's project that I decided it could not be omitted, even from a selective study such as this. Ricardo's faith in the applicability of the 'law' reflected his vision of unregulated capitalism as the best (most prosperous, efficient and rational) of all conceivable economic worlds. The 'law' came to be the very quintessence of that vision, and cannot be dismissed as just another 'model', even less so when it is considered that it was Ricardo's advocacy of laissezfaire, more than anything else, which anchored his work in its historical context, and supplied its dominant thread. For Professor Hollander, however, the goal seems to be one of explaining away the use of the 'law' with a view to establishing Ricardo's innocence of the 'Ricardian Vice'. It is a failed effort. Ricardo may not have been guilty of the 'Vice' on the scale implied by Schumpeter, but guilty he most certainly was, as this area of his writings amply testifies. With the evolution of Ricardo's thought on the labour theory of value, discussed in chapters 4 and 5, we do meet a more truly 'theoretical' Ricardo, especially in his later years, although one who never lost sight of the original function of the theory to provide the framework within which he could demonstrate his thesis of 'permanently' falling profitability in consequence of diminishing agricultural returns. We also meet a bewildering array of interpretations, including the views of Ricardo as a 'labour quantity' theorist (which he was, but not straightforwardly); as a mere cost of production theorist (which he was not); as an 'empirical' labour theorist (a view which only scratches the surface); as a 'price-

Concluding remarks

299

balancing' theorist in the Sraffian mode (which he was not); and as a retreating, steadfast, or (correctly) an increasingly committed labour theorist. How can this diversity of opinion be explained? Part of the answer lies in the complexity and confusion of Ricardo's writings, which has made it difficult to form an overall view of his work, yet easy to pluck individual passages in support of particular readings. It has also made it treacherous to focus on a single text (commonly the Principles) in isolation from Ricardo's correspondence and notes. Thus, the presence of passages in which Ricardo did use 'value' in the sense of cost of production is beyond dispute, but I have argued that this should not detract from his absolute, 'labour quantity' usage of'value', nor from his futile attempt to reconcile his two 'value' concepts by means of the perfect standard. Similarly, it is true that Ricardo advanced empirical claims on behalf of the labour theory in the third edition of the Principles, but in jumping to the conclusion that the theory was merely empirical, and that it would not have been used save for its empirical relevance, commentators have tended to ignore its later, more 'philosophical' appeal to Ricardo; and, more importantly, they have failed to ask why the strong empirical claims were missing from the first two editions of the Principles. It is in the light of such an investigation that I have proposed an inversion of orthodox teaching: the alleged empirical relevance of the labour theory should be taken to reflect Ricardo's attachment to the theory, not vice versa. Sraffa's 'price-balancing' interpretation may also lay claim to textual support, if only superficially. But when one considers that the 'price-balancing' standard is directed towards a problem of limited, if any, genuine interest to Ricardo - the redistribution of a given social product - the suspicion should be raised that Sraffa had again confused Ricardo's objectives with his own. This suspicion is confirmed by Ricardo's later writings, in which he explained that his analysis of distribution was to be conducted at the level of the individual farm, or firm; and it is further confirmed by the absence of a considered 'price-balancing' analysis in his work (indeed, his numerical examples suggest that distribution-induced changes in price will not cancel out in the aggregate). As to the evolution of Ricardo's thought on 'value', it is likewise true that passages may be selected in support of contrasting readings, including the old 'retreat' interpretation. It was Sraffa who succeeded in laying the 'retreat' interpretation to rest (not entirely

300

INTERPRETING RIGARDO

for the right reasons), but in claiming that the treatment of'value' was essentially the same over the successive editions of the Principles, he deflected attention from the significance of the changes to the third edition. With the introduction of the 'empirical' claims on behalf of the labour theory, and the new section on an invariable measure of value, Ricardo was, for the first time, justifying his use of the labour theory in the light of the 'curious effect' material. His previous oversight (which I have attributed to the speed with which the Principles was written, the late discovery of the 'curious effect' material and its interpolation in a 'pure' labour theory manuscript, and the influence of James Mill, whose forte was style rather than substance), had been brilliantly exposed by Mai thus, who alone recognised the true weaknesses in Ricardo's exposition. Ricardo's published reaction, together with his later writings, confirm that his commitment to the labour theory was, if anything, strengthened in the aftermath of Malthus's criticisms. At the same time, however, he was clearly on the defensive, knowing that Malthus had out-smarted him. It is this defensiveness, somewhat neglected in recent years, which forms the real kernel of truth in the 'retreat' interpretation. There are, then, varying degrees of textual support for the competing interpretations of Ricardo's writings on 'value', which may help to explain their genesis and, in some cases, their survival. But it is difficult to resist the conclusion that there have been other forces at work. Thus, to take the case of the 'labour quantity' versus cost of production debate, John Stuart Mill's shift in emphasis from a pronounced 'labour quantity' interpretation of Ricardo (in the Unsettled Questions) to a cost of production rendering (in Mill's later Principles), might be explicable in terms of a simple change of mind. But it would be naive to ignore the alternative possibility that Mill intentionally played down the 'labour quantity' dimension with a view to rehabilitating Ricardo as 'the greatest political economist' ([1871] 1987, p. 397) in the history of Mill's reconstructed subject, whose 'real meaning' was not antagonistic to popular sentiment. This was the view of Ricardo that Mill bequeathed to Alfred Marshall, who once described his own work as a translation of 'Mill's version of Ricardo's or Smith's doctrines into mathematics' (letter to J.B. Clark, 24 March 1908, in Pigou, ed. 1925, p. 416). Having been under Mill's influence from an early stage, it was perhaps natural that Marshall should have proposed a cost of

Concluding remarks

301

production interpretation of Ricardo. But it was also essential for him to do so if he was to counter the perception that Ricardo belonged with Marx in a 'labour quantity5 tradition, radically opposed to the fledgling 'neoclassical' school. Marshall was not about to have Ricardo, his youthful hero (Pigou, ed., 1925, pp. 99-100), suffer that ignominious fate. Hence his insistence that the (Marxian) 'labour quantity' reading was utterly without foundation: an insistence born of Marshall's acquaintance with Mill's Ricardo, and reinforced by the conviction that Ricardo belonged within the history of his reconstructed subject. The cost of production interpretation had been established, and enmeshed with the question of Ricardo's 'place' in the history of economic thought, while the opposing 'labour quantity' interpretation was stigmatised through its association with Marx. So it has remained. The 'labour quantity' interpretation was not, of course, Marx's creation, although modern readers may be forgiven for thinking otherwise. Nor is it necessarily the case that the 'labour quantity' reading of his work consigns him to the Marxian heritage. Ricardo's problem with his later value analysis was that he was struggling to reconcile his 'labour quantity' concept of value with the orthodox cost of production concept, the root difficulty being that (changes in) costs of production were not proportional to (changes in) quantities of expended labour. Given the way in which the problem was posed by Ricardo, with labour quantity identified with the wages advanced to labour, it was necessarily insoluble. With Marx, however, we have the transformation of his 'values' {quanta of abstract labour) into prices of production, the upshot (or perhaps the assumption) being that profit is 'value', regardless of whether this profit is proportional to the labour actually expended on the production of an individual commodity. This has the appearance of a reformulation of Ricardo's problem in a manner capable of'solution'. But as Marx pointed out tirelessly, it required a fundamental concept - value in the sense of abstract labour, covering both paid and unpaid labour — which Ricardo did not possess. Nor, it should be added, did Ricardo even come close to identifying profit with unpaid labour (it is well known and, in this instance, correct, that the question of the 'origin of profit' was never posed by him). So, although I would not deny that Marx saw himself as improving on Ricardo's work, and may have learnt much from it, I am neither claiming that Ricardo was an embryonic Marxian, nor that Marx

302

INTERPRETING RICARDO

was a mature Ricardian. There are enough bogus lineages incorporating Ricardo without adding to the list. The question of lineages was well to the fore in chapter 6, which differed from previous chapters in moving away from Ricardo's own preoccupations to those of his interpreters. The first five sections were spent evaluating various claims which have been made for Ricardo's affinity with 'neoclassical' doctrine, whence it emerged that they ranged from gross overstatement to pure fiction. The real question of interest here is how commentators could be so badly misled. Alfred Marshall's fatal mistake was in coming to believe that he understood 'what he [Ricardo] really meant' ([1920] 1949, p. 670), where Ricardo's 'real meaning' was invariably compatible with Marshall's own theoretical perspective. So, for example, he 'generously' credited Ricardo with the view that demand and supply play a coordinate role in the determination of natural price, he supplied the missing assumption of constant returns in order to excuse Ricardo's many statements that (changes in) natural prices depend on (changes in) supply conditions alone, and in claiming that Ricardo was 'feeling his way' towards the distinction between total and diminishing marginal utility, he implicitly attributed to Ricardo a crude understanding of his own doctrine of marginal demand price. In all these cases, he had made sense of Ricardo in his own terms, not Ricardo's. It was, perhaps, understandable (though not excusable) for Marshall to have identified himself too closely with Ricardo, especially in the light of J. S. Mill's influence on his intellectual development. But what of subsequent generations of commentators? Most of them sensibly concluded that Marshall's generosity had led him astray, but a minority have continued to propound variations on the theme that Ricardo had 'anticipated', 'discovered', or been 'feeling his way towards', elements of 'neoclassical' doctrine. The periodic appearance of such readings has an air of grim inevitability. It is the most natural thing in the world to render the past as a sepia reflection of the present: a present which, for most professional historians of economic thought, who have been drilled in modern economic analysis, is enshrined in 'neoclassical' discourse. While it is entirely possible, a priori, that charming anticipations of 'neoclassical' economics are to be found in classical texts, it is equally possible that they are merely read into those texts; and, even if vague

Concluding remarks

303

similarities do exist, that they are artificially sharpened, separated from context, and repackaged as 'contributions' to a doctrine of which the original author was sublimely ignorant. So it has been with Ricardo and so, on current trends, will it continue to be. Matters have not been helped by the intervention of the Sraffians, whose efforts to claim Ricardo for their reconstructed past have created yet another set of illusions, the brilliance of which derives from Piero Sraffa's genius combined with elements of truth. Here, I should state that I make a sharp distinction between Sraffa's interpretation (honestly given, if confused with his own theoretical preoccupations) and the work of his self-proclaimed followers, some of whose writings tread, unsteadily, the thin line between genuine historical enquiry and doctrinaire tub-thumping. Perhaps it was that disturbing noise which helped push Samuel Hollander to his own interpretative extreme. Be this as it may, the resulting debate has taken on all the characteristics of a bitterly contested paternity suit, with one side battling for Ricardo's custody within the 'neoclassical' home, the other for his adoption within the 'Sraffian' home. My contribution to this debate is simple. Leave him within his own surroundings.

Bibliography

[Anon.] (1817). 'Weights and Measures'. British Review, February, PP- 159-^4(1821). Observations on Certain Verbal Disputes in Political Economy, Particularly Relating to Value, and to Demand and Supply. London: R. Hunter. Arrow, K. J. and Starrett, D. A. (1973). 'Cost- and Demand-theoretical Approaches to the Theory of Price Determination'. In Hicks, J. R. and Weber, W. (eds), Carl Menger and the Austrian School of Economics. Oxford University Press. Ashley, W. J. (1891). 'The Rehabilitation of Ricardo'. Economic Journal, Volume 1, pp. 474-89. Bagehot, W. ([1898] 1973). Economic Studies. New Jersey: Augustus M. Kelley. [Bailey, S.] (1825). A Critical Dissertation on the Nature, Measure, and Causes of Value. London: R. Hunter. Barkai, H. (1959). 'Ricardo on Factor Prices and Income Distribution in a Growing Economy'. Economica, Volume 26, pp. 240-50. (1965). 'Ricardo's Static Equilibrium'. Economica, Volume 32, PP- I5-3I(1967). 'The Empirical Assumptions of Ricardo's 93 Per Cent Labour Theory of Value'. Economica, Volume 34, pp. 418-23. Bell, S. (1907). 'Ricardo and Marx'. Journal of Political Economy, Volume 15, pp. 112-17. Bharadwaj, K. (1983a). 'Ricardian Theory and Ricardianism'. Contributions to Political Economy, Volume 2, pp. 49-77. (1983b). 'On a controversy over Ricardo's theory of distribution'. Cambridge Journal of Economics, Volume 7, pp. 11—36. (1990). 'Sraffa's Return to Classical Theory'. In Bharadwaj, K. and Schefold, B. (eds.) (1990). Bharadwaj, K. and Schefold, B. (eds) (1990). Essays on Piero Sraffa. Critical Perspectives on the Revival of Classical Theory. London: Unwin Hyman. Blaug, M. (1956). 'The Empirical Content of Ricardian Economies'. Journal of Political Economy, Volume 64, pp. 41-58. (1958). Ricardian Economics. A Historical Study. New Haven: Yale University Press. 3°4

BIBLIOGRAPHY

305

(1968). 'David Ricardo'. International Encyclopaedia of the Social Sciences.

Volume 13, pp. 507-12. (1980). 'Marx's Bourgeois Mentor'. Times Literary Supplement, April 11, p. 421. (1985a). 'What Ricardo Said and What Ricardo Meant'. In Caravale, G. A. (ed) (1985). (1985b). Economic Theory in Retrospect. Fourth edition. Cambridge University Press. (1987). 'Classical Economies'. In Eatwell, J., Milgate, M. and Newman, P. (eds) (1987), Volume 1, pp. 434-45. Bohm-Bawerk, E. von. ([1921] 1959). Capital and Interest, Volume 1: History and Critique of Interest Theories. Illinois: Libertarian Press. Bortkiewicz, L. von. ([1907] 1952). 'Value and Price in the Marxian System'. In International Economic Papers, Volume 2, ed. A. Peacock. Cannan, E. (1893). A History of the Theories of Production and Distribution in English Political Economy from IJJ6 to 184.8. London: Percival & Co. (1929). A Review of Economic Theory. London: P. S. King & Son. Caravale, G. A. (ed.) (1985). The Legacy of Ricardo. Oxford: Basil Blackwell. (1985). 'Diminishing Returns and Accumulation in Ricardo'. In Caravale, G. A. (ed) (1985). Caravale, G. A. and Tosato, D. A. (1980). Ricardo and the Theory of Value, Distribution and Growth. London: Routledge & Kegan Paul. Casarosa, C. (1978). 'A New Formulation of the Ricardian System'. Oxford Economic Papers, Volume 1, pp. 38-63. (1985). 'The "New View" of the Ricardian Theory of Distribution and Economic Growth'. In Caravale, G. A. (ed) (1985). Cassels,J. M. (1935). 'A Re-Interpretation of Ricardo on Value'. Quarterly Journal of Economics, Volume 49, pp. 518-32. Checkland, S. G. (1949). 'The Propagation of Ricardian Economics in England'. Economica, Volume 16, pp. 40-52. Dasgupta, A. K. (1985). Epochs of Economic Theory. Oxford: Basil Blackwell. De Vivo, G. (1985). 'Robert Torrens and Ricardo's "corn ratio" theory of profits'. Cambridge Journal of Economics, Volume 9, pp. 89-92. (1987a). 'Ricardo, David'. In Eatwell, J., Milgate, M. and Newman, P. (eds) (1987), Volume 4, pp. 183-98. (1987b). 'Corn Model'. In Eatwell, J., Milgate, M. and Newman, P. (eds) (1987), Volume 1, p. 671. Dmitriev, V. K. ([1904] 1974). Economic Essays on Value, Competition and

Utility. Cambridge University Press. Dobb, M. (1940). Political Economy and Capitalism. Second edition. London: Routledge & Kegan Paul. (1973). Theories of Value and Distribution since Adam Smith. Cambridge University Press. (1975). 'Ricardo and Adam Smith'. In Skinner, A. S. and Wilson, T. (eds.) Essays on Adam Smith. Oxford: Clarendon Press.

306

BIBLIOGRAPHY

Eatwell, J. L. (1975a). 'The Interpretation of Ricardo's Essay on Profits'. Economica, Volume 42, pp. 182-7. (1975b). 'Contributions to the Theory of Surplus Value: Old and New'. Science and Society, Volume 33, pp. 281-303. (1977). 'The Irrelevance of Returns to Scale in Sraffa's Analysis'. Journal of Economic Literature, Volume 15, pp. 61-8. Eatwell, J., Milgate, M. and Newman, P. (eds.) (1987). The New Palgrave Dictionary of Economics. Volumes 1—4, London: Macmillan. Edelberg, V. (1933). 'The Ricardian Theory of Profits'. Economica, Volume i3>PP- 51-74Eltis, W. (1984). The Classical Theory of Economic Growth. London: Macmillan. Faccarello, G. (1982). 'Sraffa versus Ricardo: the Historical Irrelevance of the "Corn Profit" Model'. Economy and Society, Volume 11, pp. 122-37. Fetter, F. W. (1969). 'The Rise and Decline of Ricardian Economies'. History of Political Economy, Volume 1, pp. 67-84. Garegnani, P. (1982). 'On Hollander's interpretation of Ricardo's early theory of profits'. Cambridge Journal of Economics, Volume 6, pp. 65-77. (1983). 'The Classical Theory of Wages and the Role of Demand Schedules in the Determination of Relative Prices'. American Economic Review (P&P), Volume 73, pp. 309-13. (1984). 'Value and Distribution in the Classical Economists and Marx'. Oxford Economic Papers, Volume 36, pp. 291-325. (1987). 'Surplus Approach to Value and Distribution'. In Eatwell, J., Milgate, M. and Newman P. (eds.) (1987), Volume 4, pp. 560-74. (1990). 'Sraffa: Classical versus Marginalist Analysis'. In Bharadwaj, K. andSchefold, B. (eds.). Gide, C. and Rist, C. (1948). A History of Economic Doctrines. Second edition. London: George Harrap. Gillman, J. M. (1956). 'Ricardo's Development as an Economist'. Science and Society, Volume 20, pp. 193-226. Gonner, E. C. K. (1913). 'Ricardo, David (1772-1823)'. In R. H. I. Palgrave (ed.) Dictionary of Political Economy. London: Macmillan, pp. 304-9. Gordon, D. F. (1959). 'What was the Labour Theory of Value?'. American Economic Review (P&P), Volume 71, pp. 462-72. Gray, A. (1980). The Development of Economic Doctrine. An Introductory Survey.

Second edition (with A. E. Thompson). London: Longman. Groenewegen, P. (1972). 'Three Notes on Ricardo's Theory of Value and Distribution'. Australian Economic Papers, Volume 9, pp. 53-64. Halevy, E. ([1928] 1972). The Growth of Philosophical Radicalism. London: Faber & Faber. Haney, L. H. (1924). History of Economic Thought. New York: Macmillan. Hartwell, R. M. (1971). Introduction to Ricardo's Principles of Political Economy and Taxation. Harmondsworth: Penguin.

BIBLIOGRAPHY

307

Hicks, J. (1965). Capital and Growth. Oxford University Press. (1972). 'Ricardo's Theory of Distribution'. In Hicks (1983a). (1973). Capital and Time. Oxford: Clarendon Press. (1979). 'The Ricardian System: A Comment'. Oxford Economic Papers, Volume 31, pp. 131-4. (1983a). Classics and Moderns. Oxford: Basil Blackwell. (1983b). 'From Classical to Post-Classical: The Work of J. S. Mill.' In Hicks (1983a). (1985). 'Sraffa and Ricardo: A Critical View'. In Caravale, G. A. (ed) (1985)Hicks, J. and Hollander, S. (1977). 'Mr. Ricardo and the Moderns'. Quarterly Journal of Economics, Volume 91, pp. 351-69. Hollander, J. H. (1904). 'The Development of Ricardo's Theory of Value'. Quarterly Journal of Economics, Volume 18, pp. 455-91. Hollander, J. H. ([1910] 1968). David Ricardo: A Centenary Estimate. New York: Augustus M. Kelley. Hollander, S. (1979). The Economics of David Ricardo. London: Heinemann. (1982).' The Economics of David Ricardo: A Response to Professor O'Brien'. Oxford Economic Papers, Volume 34, pp. 224—46. (1983). 'On the Interpretation of Ricardian Economics: the Assumption Regarding Wages'. American Economic Review (P&P), Volume 73, pp. 314-18. (1984). 'The Wage Path in Classical Growth Models: Ricardo, Mai thus and Mill'. Oxford Economic Papers, Volume 36, pp. 200-12. (1985a). 'On the Substantive Identity of the Ricardian and Neoclassical Conceptions of Economic Organization: the French Connection in British Classicism'. In Caravale, G. A. (ed.) (1985). (1985b). The Economics of John Stuart Mill. Oxford: Basil Blackwell. (1986). 'On a "New Interpretation" of Ricardo's Early Theory of Profitability'. Economic Journal, Volume 96, pp. 1091-7. (1987). Classical Economics. Oxford: Basil Blackwell. (1990). 'Ricardian Growth Theory: A Resolution of Some Problems in Textual Interpretation'. Oxford Economic Papers, Volume 42, PP- 730-5°Hutchison, T. W. (1952). 'Some Questions about Ricardo'. Economica, Volume 19, pp. 415-32. (1953). 'Ricardo's Correspondence'. Economica, Volume 20, pp. 263-73. (1978). On Revolutions and Progress in Economic Knowledge. Cambridge University Press. Ingram, J. K. (1915).^ History of Political Economy. London: A & C Black. Jevons, W. S. ([1879] 1965). The Theory of Political Economy. Fifth edition. New York: Augustus M. Kelley. Jones, R. (1831). An Essay on the Distribution of Wealth. London: John Murray. Kaldor, N. (1955-6). 'Alternative Theories of Distribution'. Review of Economic Studies, Volume 23, pp. 83-100.

308

BIBLIOGRAPHY

Kater, H. (1818). 'An account of Experiments for Determining the Length of the Pendulum Vibrating Seconds in the Latitude of London'. Edinburgh Review, Volume 30, September 1818, pp. 407-24. Keynes, J. M. (1933). Essays in Biography. London: Macmillan. (1936). The General Theory of Employment, Interest and Money. London:

Macmillan. Knight, F. H. (1935). 'The Ricardian Theory of Production and Distribution'. Canadian Journal of Economics and Political Science, Volume 1,

pp. 3-25 and 171-96. Kolb, F. R. (1972). 'The Stationary State of Ricardo and Malthus: Neither Pessimistic nor Prophetic'. Intermountain Economic Review, Volume 3, pp. 17-30. Langer, G. F. (1982). 'Further evidence for Sraffa's interpretation of Ricardo'. Cambridge Journal of Economics, Volume 6, pp. 397—400. Levy, D. (1976). 'Ricardo and the Iron Law: A Correction of the Record'. History of Political Economy, Volume 8, p. 235-51.

[McCulloch, J. R.] (1818). Review of Ricardo's Principles of Political Economy and Taxation (first edition). Edinburgh Review, Volume 32,

PP- 59-87(1822). Review of the Parliamentary Report on the 'Agricultural Distress'. Edinburgh Review, Volume 36, pp. 452-82. MacDonald, R. A. (1912). 'Ricardo's Criticisms of Adam Smith'. Quarterly Journal of Economics, Volume 26, pp. 549-92. [Malthus, T. R.] (1811). Review of various pamphlets connected with the 'Bullion Controversy'. Edinburgh Review, Volume 25, pp. 339-72. (1824). Review of J. R. McCulloch's 'Essay on Political Economy'. Quarterly Review, January, pp. 297—334. Malthus, T. R. ([ 1814] 1970). Observations on the Effects of the Corn Laws.

Second edition. New York: Augustus M. Kelley. ([1815] 1970). An Inquiry into the Nature and Progress of Rent. New York:

Augustus M. Kelley. Marshall, A. ([1885] 1925). 'The Present Position of Economies'. In Pigou, A. C. (ed.) (1925). ([1920] 1949). Principles of Economics. Eighth edition. London: Macmillan. Marx, K. ([1862-1863] 1969). Theories of Surplus Value. Part 11. London: Lawrence and Wishart. ([1862-1863] 1972). Theories of Surplus Value. Part in. London: Lawrence and Wishart. ([1890] 1976). Capital. Volume 1. Harmondsworth: Penguin. Meek, R. L. (1967). 'The Decline of Ricardian Economics in England'. In Economics and Ideology and Other Essays: Studies in the Development of Economic Thought. London: C h a p m a n and Hall. (1973). Studies in the Labour Theory of Value. Second edition. London:

Lawrence and Wishart.

BIBLIOGRAPHY

309

(1974). 'Value in the History of Economic Thought'. History of Political Economy, Volume 6, pp. 246-60. Menger, C. ([1871] 1981). Principles of Economics. Fourth edition. New York: New York University Press. Mill, James ([1804] 1966). An Essay on the Impolicy of a Bounty on the

Exportation of Grain. In Winch, D. (ed) (1966). ([1808] 1966). Commerce Defended. In Winch, D. (ed.) (1966). Mill, J . S. (1874). Essays on Some Unsettled Questions of Political Economy.

Second edition. London: Longman. ([1871] 1987). Principles of Political Economy. Seventh edition. New York: Augustus M. Kelley. Mitchell, W. C. (1949). Lecture Notes on Types of Economic Theory. New York:

Augustus M. Kelley. Moore, S. (1966). 'Ricardo and the State of Nature'. Scottish Journal of Political Economy, Volume 13, pp. 317-31. Morishima, M. (1989). Ricardo's Economics: A General Equilibrium Theory of

Distribution and Growth. Cambridge University Press. Myint, H. (1948). Theories of Welfare Economics. London: Longman. Napoleoni, C. (1975). Smith, Ricardo, Marx. Oxford: Basil Blackwell. O'Brien, D. P. (1975). The Classical Economists. Oxford: Clarendon Press. (1981). 'Ricardian Economics and the Economics of David Ricardo'. Oxford Economic Papers, Volume 33, pp. 352-86. (1990). Review of International Monetary Economics i8jo—ig6o: Between the classical and the new classical, by M . J u n e Flanders. Economic Journal,

Volume 100, pp. 1356-8. Ong, N-P. (1983). 'Ricardo's Invariable Measure of Value and Sraffa's "Standard Commodity"'. History of Political Economy, Volume 15, pp. 207-7. Pasinetti, L. L. ([i960] 1974). 'A Mathematical Formulation of the Ricardian System'. In Growth and Income Distribution: Essays in Economic

Theory. Cambridge University Press. Patten, S. N. (1893). 'The Interpretation of Ricardo'. Quarterly Journal of Economics, Volume 7, pp. 322-52. Peach, T. (1981). Review of The Economics of David Ricardo, by S. Hollander. Economic Journal, Volume 91, pp. 244-7. (1986). 'Ricardo's Early Treatment of Profitability: Reply to Hollander and Prendergast'. Economic Journal, Volume 96, pp. 1105—12. (1987). 'David Ricardo: A Review of Some Interpretative Issues'. In W. O. Thweatt (ed.), Classical Political Economy. Boston: Kluwer. Pigou, A. C. (ed.) (1925). Memorials of Alfred Marshall. London: Macmillan. Pollitt, B. H. (1988). 'The Collaboration of Maurice Dobb in Sraffa's Edition of Ricardo'. Cambridge Journal of Economics, Volume 12, pp. 55-65. Porta, P. L. (1986). 'Understanding the Significance of Piero Sraffa's Standard Commodity: A Rejoinder'. History of Political Economy, Volume 18, pp. 479-84.

3io

BIBLIOGRAPHY

Prendergast, R. (1986a). 'A Comment on "David Ricardo's Early Treatment of Profitability: A New Interpretation'". Economic Journal, Volume 96, pp. 1098-104. (1986b). 'Malthus's Discussion of the Corn Ratio Theory of Profits'. Cambridge Journal of Economics, Volume 10, pp. 187-9. Ramsay, G. ([1836] 1974). An Essay on the Distribution of Wealth. New Jersey: Augustus M. Kelley. Rankin, S. (1980). 'Supply and Demand in Ricardian Price Theory: A Re-Interpretation'. Oxford Economic Papers, Volume 32, pp. 241-62. (1984). 'The Wage Basket in Ricardo's Essay on Profits'. Cambridge Journal of Economics, Volume 8, pp. 83-6. Read, S. (1829). Political Economy. Edinburgh: Oliver & Boyd. Robbins, L. (1936). 'The Place of Jevons in the History of Economic Thought'. Manchester School, Volume 7, pp. 1 —17. Rogin, L. (1956). The Meaning and Validity of Economic Theory. New York: Harper. Roll, E. (1973). A History of Economic Thought. Fourth edition. London: Faber & Faber. Roncaglia, A. (1978). Sraffa and the Theory of Prices. New York: J. Wiley. (1982). 'Hollander's Ricardo'. Journal of Post-Keynesian Economics, Volume 4, pp. 339-59. Rosselli, A. (1985). 'The Theory of the Natural Wage'. In Caravale, G. A. (ed.) (1985). Rowthorn, R. (1980). 'Marx's Theory of Wages'. In Capitalism, Conflict and Wages. London: Lawrence and Wishart. Rubin, I. I. ([1929] 1979). A History of Economic Thought. London: Ink Links. St. Clair, O. (1965). A Key to Ricardo. New York: Augustus M. Kelley. Samuelson, P. (1959). 'A Modern Treatment of the Ricardian Economy'. Quarterly Journal of Economics, Volume 73, pp. 1-35 and Volume 79, pp. 217-31. (1978). 'The Canonical Classical Model of Political Economy'. Journal of Economic Literature, Volume 16, pp. 1415-34. Sayers, R. S. (1953). 'Ricardo's Views on Monetary Questions'. Quarterly Journal of Economics, Volume 67, pp. 30—49. Schumpeter, J. A. (1954). History of Economic Analysis. London: George Allen & Unwin. Shove, G. F. (1942). 'The Place of Marshall's Principles in the Development of Economic Theory'. Economic Journal, Volume 52, pp. 294-329. Smart, W. (1910). Economic Annals of the Nineteenth Century. London: Macmillan. (1964). Economic Annals of the Nineteenth Century. Volume 11. New York: Augustus M. Kelley. Smith, A. ([1776] 1981). An Inquiry into the Nature and Causes of the Wealth of Nations. Indianapolis: Liberty Press.

BIBLIOGRAPHY

3I I

Sowell, T. (1974). Classical Economics Reconsidered. Princeton: Princeton University Press. Sraffa, P. (ed., with the collaboration of M. H. Dobb) (1951-1955). The Works and Correspondence of David Ricardo. Volumes i-x. Sraffa, P. (i960). Production of Commodities by Means of Commodities. Prelude to a Critique of Economic Theory. Cambridge University Press.

Steedman, I. W. (1982) 'Marx on Ricardo'. In Bradley, I. and Howard, M. (eds.), Classical and Marxian Political Economy. London: Macmillan.

Stigler, G. J. (1950). 'The Development of Utility Theory'. Journal of Political Economy, Volume 58, pp. 307-27 and 373-96. (1952). 'The Ricardian Theory of Value and Distribution'. Journal of Political Economy, Volume 60, pp. 187-207. (1953). 'Sraffa's Ricardo'. American Economic Review, Volume 43, pp. 586-99. (1958). 'Ricardo and the 93% Labour Theory of Value'. American Economic Review, Volume 48, pp. 357-67. Sylos Labini, P. (1990). 'Sraffa's Critique of the Marshallian Theory of Prices'. In Bharadwaj, K. and Schefold, B. (eds.) (1990). Torrens, R.

(1815). An Essay on the External Corn Trade.

London:

J. Hatchard. [Torrens, R.] (1818). 'Strictures on Mr. Ricardo's Doctrine Respecting Exchangeable Value'. Edinburgh Magazine, October, pp. 335-8. Torrens, R. ([1821] 1965). Essay on the Production of Wealth. New York: Augustus M. Kelley. Tosato, D. A. (1985). 'A Reconsideration of Sraffa's Interpretation of Ricardo on Value and Distribution'. In Caravale, G. A. (ed.) (1985). Tucker, G. S. L. (1954). 'The Origins of Ricardo's Theory of Profits'. Economica, Volume 21, pp. 320-33. (i960). Progress and Profits in British Economic Thought. Cambridge University Press. Viner, J . (1954). 'Schumpeter's History of Economic Analysis: A Review Article'. American Economic Review, Volume 44, pp. 894—910. Walras, L. ([1926] 1954). Elements of Pure Economics. London: George Allen & Unwin. Walsh, V. and Gram, H. (1980). Classical and Marxian Theories of General

Equilibrium. Oxford University Press. Wermel, W. T. (1939). The Evolution of the Classical Wage Theory. New York:

Columbia University Press. West, E. (1826). Price of Corn and Wages of Labour, with Observations upon Dr. Smith's, Mr. Ricardo's and Mr. Malthus's Doctrines upon these Subjects.

London: J. Hatchard. Whewell, W. (1833). 'Mathematical Exposition of some of the Leading Doctrines in Mr. Ricardo's "Political Economy and Taxation'". Transactions

PP- 155-99-

of

the

Cambridge Philosophical

Society.

Volume

4,

312

BIBLIOGRAPHY

Whitaker, A. C. (1904). History and Criticism of the Labour Theory of Value in English Classical Political Economy. New York: Columbia University Press. Wilson, G. W. and Pate, J. L. (1968). 'Ricardo's 93 Per Cent Labour Theory of Value: A Final Comment'. Journal of Political Economy, Volume 76, pp. 128-36. Winch, D. (ed.) (1966). James Mill: Selected Economic Writings. Edinburgh: Oliver & Boyd.

Name Index

Note: References to Ricardo and Malthus are omitted from this index Arrow, K. J. and Starrett, D. A. 256n. Ashley, W. J. ion., 25n., 295

Gray, A. ion. Groenewegen, P. 411., 27n.

Bagehot, W. i4.n. Bailey, S. 19, i57n., i58n., 227n., 240 Barkai, H. 4J1., gn., 37, 2i7n., 252n. Bell, S. 25n. Bharadawaj, K. 5, 24, 29, 38n., 57 Blaug, M. 7n., 13, i4n., 25, 29, 3m., 3511., 86, 243, 29on. Bohm-Bawerk, E. von. 9n., 27n. Bortkiewicz, L. von. i6n., 25

Halevy, E. 7n., gn., i^n. Haney, L. H. gn., ion., i4n., 3111., 34n. Hartwell, R. M. 4n. Hicks, J. 4n., 7n., 8n., gn., 10, 11, 129 Hicks, J. and Hollander, S. ion., n o Hollander, J. H. in., 2, 21-2, 24, 26n., 27-8, 100, 185, 220 Hollander, S. xi-xii, 4-5, 6-8, 9, 10, 1 in., 12, 13, 14, 15. 26-7, 2 g, 3°> 3 2 , 34~5> 36-8, 52n., 53-4, 58n., 60, 6 5 ^ , 6gn., 70, 73, 82, 83, 9311., 94-5, 96, 97-9,

Cannan, E. 1-2, 10, 21, 22, 24, 26n., 110, 117, ii8n., 131, i57n., 185, 267, 29m. Carvale, G. A. 8n., 12, 1 i7n., 127 Caravale, G. A. and Tosato, D. A. 8n., 12 Casarosa, C. 10, 11, 110 Cassels, J. M. 22n., 23n., 35n., i54n. Checkland, S. G. i3n., i4n. Clark, J. B. 38n., 300

101, iO2n., 109-10, 111, 112, ii4n., i i 7 n . , 125-7, 129, 131, i38n., 140-1,

144, 166, ig7n., 223, 242, 245, 247, 248-9, 250, 254-6, 262-3, 264^, 267-70, 273, 275-83, 284n., 286, 292, 296-7, 298, 303 Hutchison, T. W. i3n., 14, 34n., 298 Ingram, J. K. i4n.

Dasgupta, A. K. 4n., 288n. De Vivo G. 5, 38n., 289n., 292 Dmitriev, V. K. gn., 35n. Dobb, M. H. 7n., 13, 24, 27n., 28, 35n., 38n., 288n.

JafTe, W. 285 Jevons, W. S. gn., 33 Jones, R. i4n., Kaldor, N. gn. Kater, H. 226 Keynes, J. M. 12-13,68, 143 Knight, F. H. gn. Kolb, F. R. 7 ^

Eatwell, J. L. 4n., 5, 23n., 24, 35n., 38n., 80-1, 28gn., 295 Edelberg, V. 22n., 37, 186 Eltis, W. 7n., 27n. Faccarello, G. 5, 74 Fetter, F. W. i4n., 141

Langer, G. F. 5, 511. Levy, D. ion.

Garegnani, P. 5, 38, 53-4, 57, 83-4, 28gn., 295 Gide, C. and Rist, C. in., 7n., ion., 26n. Gillman, J. M. 4n., 25m Gonner, E. C. K. 25n. Gordon, D. F. 25, 25n.

McCulloch, J. R. 3m., g5, 100, ioin., 139, 173—5, ^ 3 , igon., ig2n., ig4~6, igg, 2o6n., 207, 2og-io, 214, 224, 225, 232-3, 235n., 273

3*4

NAME I N D E X MacDonald, R. A. 2511. Marshall, A. 20-1, 24, 25, 26, 32, 33-5, 3811., 14m., 239, 242, 245, 246-7, 248-9, 261-3, 295, 300-2 Marx, K. in., 13, 16-21, 24, 25, 3111., 32, 15811., 21311., 220, 23m., 239-40, 284, 301 Meek, R. L. 13, 2511., 28, 3511., 53, 6511., 68, 72, 73, 21211. Menger, C. 33 Mill, James 13, 15, 3m., 4511., 47, 65, 8on., 122, 12311., 132, 133, 134, 137-8, 13911., 145-6, 147, 150, 152-4, i66n., 168, 17311., 17411., 180, 187, 188, 190, 19411., 214, 224, 232, 23311., 235, 271-2, 300 Mill, John Stuart 19-20, 2in., 24, 32, 47n., 240, 300, 301, 302 Mitchell, W. C. in., 2, 3, 4, 7n., 34n., 294 Moore, S. 2i2n. Morishima, M. ion., 13, 37-8, 277, 283-6 Myint, H. 25n., 35n. Napoleoni, C. 73 O'Brien, D. P. xi, 5, 7n., 13, 14m, i32n., 297 Ong, N-P 4n., 8n., 30, 223, 2gon. Pasinetti, L. L. gn., i3n., i4n., 256n. Patten, S. N. 25n., 68n., 295 Peach, T. xi, 8in., 86 Pigou, A. C. 38n., 300, 301 Place, F. i74n., Pollitt, B. H. 24 Porta, P. L. 5n. Prager, T. 24n. Prendergast, R. 5n., 83 Ramsay, G. gn. Rankin, S. 3511., 69n., 252n. Read, S. gn. Robbins, L. 34m Rogin, L. 7n., ion., 22n. Roll, E. 7n., 25n. Roncaglia, A. 5, 35n., 2gon. Rosselli, A. 11 Rowthorn, R. 110 Rubin, I. I. ion., i4n., 17, 25n., 28

3*5

St Clair, O ion., 110 Samuelson, P. A. gn. Say, J. B. 13, 36-7, 65n., 88, gin., 138, , 243, 245,, 247n., 263, Sayers, R. S. 42n. Schumpeter, J. A. 7n., 13, 14, i5n., 26n., 3m., 34n., 35n., i57n., 262, 2g8 Shove, G. F. 35n. Smart, W. 52, ggn., 100, 133, 137 Smith, A. 2n., 4on., 44, 47, 48n., 53, 8on., 109-10, 123, 126, 132, 13511., 153, 154, 155, 156, 157, 158, i5g, i8g-gi, ig2n., 206, 213, 220, 225, 242-3, 244n., 246, 252n., 256, 257, 25g, 272, 278n., 286, 300 Sowell, T. i4n., 35n. Sraffa, P. xii, 1, 3-5, 6, 22-4, 26, 2g, 30, 31, 32, 38, 39, 48, 4gn., 53, 57-8, 61, 62, 67, 68, 75, 76, 80-1, 85, g4n., i53n., i64n., 176, 183, iggn., 205, 211, 2i2n., 214, 222-3, 234? 2 3g, 286-go, 2gi-2, 2g3, 2g4-5, 2gg~3oo, 303 Steedman, I. W. i8n., 25, 26, 23g Stigler, G. L. 7n., gn., 13, i4n., 25-6, 34n., 110, 166, 234, 23g Sylos Labini, P. 6 Torrens, R. 21, 7g-8o, 117-ig, 145, i74n., 176-g, 181, 183-4, J 8g-gi, ig6, 2O3n., 2og, 234, 255 Tosato, D. A. 8n., 24n., 30, 223, 2gon., 2gin. Trotter, C. 48-g Trower, H. 55, 57, g5, g6, ggn., 10m., io2n., ii8n., 122, i23n., 133, i3gn., 143, i5on., 173, i74n., 175, 188, 224, 227, 228, 230, 233n., 23g, 240, 260 Tucker, G. S. L. 4, 5, ion., 12, 51-3

Viner,J. 35n. Walras, L. 33, 35n., 36, 37, 277, 284-5 Walsh, V. and Gram. S. 2g5 Wermel, W. T. gn. West, E. gn., 72n. Whewell, W. 7n., gn., i4n., i8n., 25 Whitaker, A. C. 25n., 26n., 27n., 3111., 34n. Wilson, G. W. and Pate, J. L. 2i7n. Winch, D. 8on.

Subject Index

Accumulation, 87, 90, 106, 130, 143; and wages, 10, 114, 122; and profit, 9, 93-4, 96, 98, 108. See also law of markets Agricultural protection, i5n., 99-101. See further under distribution analysis Analytical approach; in early monetary writings, 42-3, 49; and S. Hollander, 14-15, 283; and Ricardian Vice, 14-15, 140-3; and Sraffa, 291-3, 294; and wages, 11-12, 129-30, 298 Annuity formula, i62n., 214, 218, 219 Basic versus non-basic commodities, 287-9 Capital, in correspondence (1814), 60-1, 66-7; in Essay on Profits, 69-71, 75; in Principles, 93, 106, 159, i65n., 2i3n., 2i4n., 263-9. See also capital structure, capital transfer Capital structure, 26n., 160-1, 176, 180-1, 214, 284; and 'curious effect', 28-9, 153, 161-3, 165, 168, 187, 200, 224; and S.Hollander, 27, 29, 166, i97n., 281-2; and invariable measure of value, 22-3, 29, 30, 161-7 (passim), 185-7, I 9 I ~6, 198, 200-6, 210-12, 219-20, 222, 224, 233, 290; and Malthus, 29, 191-5, 199-202, 208-9, 212, 223-4; a n d labour theory of value, 18, 21, 28, 151-3, 163, 166-9, J 77~9? 181, 183, 186, 187, 191, 194-7, 2OO> 203, 208, 209, 212-17, 220, 235-7; a n d (dated) labour profile analysis, i8n.; and Torrens, 176-9, 181, 183-4, 189-91, 196 Capital transfer, 135-6, 265-7 Competition of capitals doctrine; and S. Hollander, 53, 95n.; and law of markets, 132, 135; and Malthus, 59, 60, 85, 229; Ricardo's adherence to (1810-16), 44, 48-9, 53, 56-8, 68, 84, 89n., and abandonment of, 85; and Adam Smith, 44, 132; and Sraffa, 294

Corn laws. See agricultural protection Corn model interpretation, 1, 86, 193; and Essay on Profits, 6, 70, 71, 75; 'further evidence' for, 5n., 3111.; and S. Hollander, 4, 58n., 70; misunderstanding of, 4n.; W. C. Mitchell's anticipation of, 2-4, 294; and Principles, 9411.; and Ricardo's writings of 1813, 53-4; and Ricardo's writings of 1814, 57-8, 61-2, 66-8; and Sraffa, 3-5, 23, 38, 39, 57-8, 61-2, 67-8, 75, 80-1, 85, 94n., iggn., 288-9, 294~5 Corn ratio theory. See corn model Curious effect, 164, 204; discovery of, 28-9, 151-4, 168, 187; in edition 1 of Principles, 161-5 (passim); in edition 2 of Principles, 182; in edition 3 of Principles, 220; and invariable measure of value, 29, 30, 161-2, 165, 181-2, 185, 220; and labour theory of value, 29, 152-3, 154, 160, 163, 168, 174, 179, 187, 224, 300; and Malthus, 200; and Sraffa, 289 Dated labour. See capital structure Demand, 250-6; for corn, 91, 24gn., 251-4; elasticity of, 34, 251, 255; in prt-Essay writings, 56-67 (passim); and S. Hollander, 34, 250, 254-6, 282; law of, 34, 254, 282; limited only by production, 65n., 135, 136, 137, 142; and Sraffa, 286, 293; and utility analysis, 33-4 Demand and supply, 32, 35, 36, 256-63, 28 m., in pre-Essay writings, 59-60, 63-4; and S. Hollander, 27, 35, 262-3, 281-3; and Malthus, 258-61; and Marshall, 35, 261-3, 302; and natural price determination, 35-6, 257-63; and Sraffa, 38, 286, 291-2 Derived factor demand. See imputed value Distribution analysis, and agricultural protection, 1-9 (passim), 52, 57, 58, 88,

316

SUBJECT INDEX 95-103, 294-7 (passim); and diminishing agricultural returns, 2-3, 6-9, 48, 54-5, 67, 88-97, IO3> r23> 126-7, 129, 294-6; dynamic nature of, 6-8, 89-90, 129, 295; in early monetary writings, 43-4; in Essay on Profits, 2, 69-75; and J. Hollander, in., 2, 27-8; and S. Hollander, 4, 6-8, 1 in., 27n., 37, 52n., 53-4, 94-9, 101-3, 280-3, 29^-7; micro versus macro focus of, 31, 38, i64n., I99n., 202-3, 221-3, 290-1, 299; and notional changes, 6, 38, 204, 287-9, 291; in Principles, 90-4; and Sraffa, 3-4, 6, 22-3, 26, 31, 38, 53, 57, i64n., iggn., 234n., 286-91; 'theory' of (1813), 49-55. See also, competition of capitals, corn model, labour theory of value, profit, rent, wages. Economic prospects, 95-6, 103, 133-4, 137; and S. Hollander, 7-9, 95, 96n. Factors of production, 36, 264, 267-9, 27r> 277-9, 280, 283, 293 General equilibrium analysis, 37-8, 277-86; and S. Hollander, 37-8, 277-83, 286, 296; and Morishima, 37-8, 277, 283-6 General gluts. See law of markets. Imputed value, 269-77; and S. Hollander, 36-7, 269-70, 273, 275-7 See also, opportunity cost. Invariable measure of value, 90, 156-7, 169-70, 214-16 (passim); pre-'curious effect', 28, 149, 151, 168; and S. Hollander, 26-7, 29-30, 223-4, 281-2, 292; and Malthus, 29, 192, 195, 200-2, 210-11, 212, 216, 224, 229, 230; and Marx, 17; a n d J . S . Mill, 2in.; 'perfect' conception of, 30-2 (passim), 198, 207, 211, 219-22 (passim), 235-6, 238, 299; and 'pure' labour theory of value, 29, 30-1, 166-8, 186, 187, 198, 222-4, 2375 a n d scientific measurement, 30, 222-3, 225-7, 231-2; and Adam Smith, 225; and SrafFa, 22-4, 26, 29, 30-1, 38, 2i2n., 222-3, 234n., 239, 288-91, 298-9. See also, capital structure, curious effect Invariable standard. See invariable measure of value. Labour commanded, 156, 230, 291 Labour profile, See capital structure

317

Labour theory of value, 16-32, 145-240, 298-302; and 'absolute' usage of 'value', 17, 19-20, 24, 28, 151, 157, 158, 168-9, lll-> r 76, 213, 227-9, 231, 239, 299; adoption of, 150-1, 168, 186-7; and Bailey, 19, i57n., i58n., 227n., 240; between editions 1 and 2 of Principles, 172-9; between editions 2 and 3 of Principles, 189-211; and British Review, 173, 182, 185; and Cannan, 21, 22, 26n., i57n., 185; and cost of production, 20-1, 25-6, 28, 31-2, 169-71, 174, 205-7, 221, 232, 234, 238-40, 298-9, 300-1; and differential profit rates, 157; and distribution analysis, 22-4, 31, 156, 161-2, 164-5, 171, 186-7, r 92-5, 201-3, 205, 217, 224, 237-8, 282-3, 290, 298; and early stages of society, 155, 159, 160-1, 184, 190-1, 211-12, 213, 21611., 237n.; in edition 1 of Principles, 154-71; in edition 2 of Principles, 179-86; in edition 3 of Principles, 211-24; empirical status of 19, 25-6, 29,31, 166, 171, 203, 216-17, 223-4, 239, 298-9; and Essay on Profits, 72, 186; and essence or source of value, 25-6, 28, 31-2, 171, 224, 230-1, 237-8; in final writings, 224-38; and heterogeneous labour; 29n., 157-8; and J. Hollander, 21-2, 26n., 185; and identification of labour quantity with wages, 170, 206-7, 221, 233, 237, 240; justification for, 29-31, 153, 166-8, 186-7, 198, 216-17, 222-3, 238-9, 290, 299-300; and McCulloch, 173-5, i9on., 209-10, 232; and Malthus, 21, 29, 31, 150, 172-3, i76n., 187, 189, 191-6, 198-211, 213, 221, 223-4, 230, 234, 238, 240; and Marshall, 20-1, 25, 26, 32, 239, 300-1; and Marx, 16-21, 24-5, 32, I58n., 2i3n., 220, 23m., 239-40, 301; and J. S. Mill, 19-20, 2in., 32, 240, 300; and Morishima, 285-6; and Observations on Certain Verbal Disputes in Political Economy, 19, 224,

227-8, 240; and rent, 90-1, 159; and Sraffa, 22-4, 30, 183, 205, 211, 2i2n., 214, 222-3, 298-300; and Torrens, 21, 179, 189-91, 2O3n., 209, 234; and Trower, 175, 227-8, 239, 240. See also, capital structure, curious effect, invariable measure of value Laissez-faire, 140, 141-3, 271, 298 Law of markets, 12-16, 52, 65-7, 68, 84, 96, 131-43, 267, 271, 298; in early monetary writings, 46-7, 49, 131-2;

3

i8

SUBJECT INDEX

Law of markets (cont.) and S. Hollander, 13, 14, 65n., i38n., 298; and Keynes, 12-13, 143; and Malthus, 65-6, 84, 138-40; and James Mill, 13, 15, 47, 65, 132; in Principles, 134-7; and Ricardian Vice, 14-15, 298; andJ.B. Say, 13, 65n. Money, as a commodity, 40, 49, 68, 70, 72, 83, 132, 147. See also, quantity theory New view interpretation. See under wages Neoclassical economics, 6, 32-3, 35n., 36-8, 241, 247, 249, 263, 277, 283, 293, 2 95, 297, 302-3 Opportunity cost, 36, 263-4, 267-9

Standard system, 287-90 Subsistence (or natural) wage interpretation. See under wages Supply and demand. See demand and supply Utility, 33-4, 154-5, 241-9, 284-5; anc * demand analysis, 33-4, 241-9; and S. Hollander, 34, 242, 245, 247-9; a n ^ Marshall, 33-4, 242, 245-9, 302; and Say's doctrine of productive services, 269-77 Value, 'absolute' usage of (1815), 148-9; relative usage, in early monetary writings, 40-1, 49. See also demand, demand and supply, imputed value, labour theory of value, pricing assumptions, utility

Pricing assumptions, before Essay on Profits (1813-15), 2-6, 54-5, 61, 67, 76-81, Wages, 9—12, 103-31, 279, 283-4; a n d 147; in early monetary writings, 47-9; Cannan, 10, 110, 117, 1 i8n., 131; and in Essay on Profits, 6, 71-3, 73-5, 76, 81, Caravale, 12, ii7n., 127; in 85, 147; following the Essay on Profits correspondence (1814-15), 116, 118, (1815), 81-4, 147-8 120; in correspondence (1815-16), Process analysis. See demand and supply 120—2; in early monetary writings, Profit, doctrine of'permanent' movements 45-6, 49, 116; in Essay on Profits, in, 6, 9, 11, 15, 65, 66, 89, 92, 96-7, 116-17, 120-1; and S. Hollander, 10, 103, i n , 113-15, 117, 123, 130, 217; 12, 69n., 109, 110-11, 112, 1 i4n., , origin of, 301; as a residual or surplus, 1 i7n., 125-7, I29? 13li 296-7, 284n.; 11, 12, 287-8, 292-3 and Malthus, 11-12, 116-29; anc ^ Marshall, 38n.; and new view Quantity theory, 4, 41, 54 interpretation, 10-12, 92n., 107-15, 120, 123, 125-7, I29~3I> 279-80, Rent, I7n., 33, 68-9, 71, 76, 90-1, 92, 95n., 296-8; in Notes on Malthus, 125-9; a n d 97, 117, 146, 159 'permanent' movements in profitbility, Ricardian vice, 14-15, 140-1; and 11, 111-17, 123, 130-1, 298; in S.Hollander, 14-15, 140-1, 298; and Principles, 94, 105-16; and Adam Schumpeter, 14, i5n., 298 Smith, 109-10; and subsistence (natural wage) interpretation, 9-10, Say's identity versus Say's equality, 12-13, 11-12, 104-7, no-20, 123, 130-1, 14-15, 141 See also, law of markets 297-8; and Torrens, 1 i7n., 118-19. See Standard measuring device. See invariable also accumulation, analytical approach measure of value, and S. Hollander Standard commodity, 287-90