The Employment Relationship: Causes and Consequences of Modern Personnel Administration (Springer Studies in Work and Industry)

  • 30 953 9
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up

The Employment Relationship: Causes and Consequences of Modern Personnel Administration (Springer Studies in Work and Industry)

The Employment Relationship Causes and Consequences of Modern Personnel Administration PLENUM STUDIES IN WORK A N D IN

1,645 647 9MB

Pages 247 Page size 198.48 x 318.48 pts Year 2010

Report DMCA / Copyright


Recommend Papers

File loading please wait...
Citation preview

The Employment Relationship Causes and Consequences of Modern Personnel Administration

PLENUM STUDIES IN WORK A N D INDUSTRY Series Editors: Ivar Berg, University of Pennsylvania, Philadelphia, Pennsylvania a n d A r n e L. Kalleberg, University of North Carolina, Chapel Hill, North Carolina

WORK AND INDUSTRY Structures, Markets, and Processes Arne L. Kalleberg and Ivar Berg

Current Volumes in the Series: THE BUREAUCRATIC LABOR MARKET The Case of the Federal Civil Service Thomas A. DiPrete THE EMPLOYMENT RELATIONSHIP Causes and Consequences of Modern Personnel Administration William P. Bridges and Wayne J. Villemez ENRICHING BUSINESS ETHICS Edited by Clarence C. Walton ENSURING MINORITY SUCCESS IN CORPORATE MANAGEMENT Edited by Donna E. Thompson and Nancy DiTomaso INDUSTRIES, FIRMS, AND JOBS Sociological and Economic Approaches Edited by George Farkas and Paula England LABOR AND POLITICS IN THE U.S. POSTAL SERVICE Vern K. Baxter LIFE AND DEATH AT WORK Industrial Accidents as a Case of Socially Produced Error Tom Dwyer MATERNAL EMPLOYMENT AND CHILDREN'S DEVELOPMENT Longitudinal Research Edited by Adele Eskeles Gottfried and Allen W. Gottfried THE STATE AND THE LABOR MARKET Edited by Samuel Rosenberg WHEN STRIKES MAKE SENSE--AND WHY Lessons from Third Republic French Coal Miners Samuel Cohn

A ContinuationOrder Plan is availablefor this series. A continuationorder will bring deliveryof each new volumeimmediatelyupon publication.Volumesare billedonlyupon actual shipment.For further informationplease contact the publisher.

The Employment Relationship Causes and Consequences of Modern Personnel Administration William P. Bridges University of Illinois Chicago, Illinois


Wayne J. Villemez University of Connecticut Storrs, Connecticut

Plenum Press • New York and London

Library of Congress Cataloging in Publication Data Bridges, William P. The employment relationship: causes and consequences of modern personnel administration / William P. Bridges, Wayne J. ViUemez. p. cm.--(Plenum studies in work and industry) Includes bibliograpldcal references and index. ISBN 0-306-44744-4 1. Personnel management--Illinois--Chicago Metropolitan Area--Case studies. 2. Bureaucracy-Illinois--Chicago Metropolitan Area---Case studies. 3. Line and staff organization--Illinois--Chicago Metropolitan Area--Case studies. I. Villemez, Wayne J. II. Title. HI. Series. HF5549.2.USB75 1994 94-34950 658.3'009773'11---dc20 CIP

ISBN 0-306-44744-4

©1994 Plenum Press, New York A Division of Plenum Publishing Corporation 233 Spring Street, New York, N.Y. 10013 All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise, without written permission from the Publisher Printed in the United States of America


In 1979, serious research was just beginning on the connections between stratification outcomes and organizations. Data suitable for investigating these connections were scarce, and the general wisdom was that they would remain scarce--since organizational case studies were seen as the only means of gathering linked individual and organizational data. The case study approach does allow one to link the two types of data, but gathering such data on more than a few organizations is prohibitively expensive and difficult, and having only a few organizations limits generalizability. To help solve this problem, we developed the idea of a survey of a random sample of several thousand employed individuals, followed by a second survey of their several thousand employing organizations. This method, we reasoned, would provide us with a generalizable, simple random sample of individuals, coupled with a weighted random sample of organizations (weighted, of course, by size of organization). An added benefit would be that these valuable data could be gathered by a survey organization for the price of two simple surveys. It was not an easy idea to sell. We developed it into a proposal to the National Science Foundation (NSF), and though the reviewers were otherwise sympathetic, they were almost unanimous in their contention that such a survey would not work because "obviously" the great majority of respondents would refuse to reveal exactly who their employers were. To counter this idea, we fielded a large pilot study (funded by the University of Illinois at Chicago Office of Social Science Research), the results of which demonstrated quite conclusively that not only were over 90% of respondents quite happy to identify their employer, but they just as freely provided the exact street address of that employer. Armed with this fact, w e returned to the NSF and were funded, and the survey finally got into the field in 1981. We have debated at some length the relative merits of a national sample versus one confined to a single large labor market. Given the



uniqueness of the research design and the fact that nothing like it had ever been done before, we had no way of knowing exactly what sorts of methodological problems we might encounter. This concern tipped the balance to our choice of a more manageable single-labor-market study. Many years later, others successfully applied our method to a continuing national study (see, e.g., Spaeth, 1985; Spaeth, O'Rourke, Kalleberg, Knoke, & Marsden, 1993; Kalleberg, Knoke, Marsden, & Spaeth, 1993), and Parcel, Kaufman, and Jolly (1991) have recently concluded in a review article that our methodology is probably the most efficient means of gathering a matched sample of employers and workers. Once we had decided on a single labor market, our next two decisions were easy ones. We were both working in the middle of the Chicago SMSA, a very large labor market with an industrial structure that mirrored that of the United States. For a study of the Chicago labor market, the University of Illinois Survey Research Laboratory was an obvious choice; the other obvious choice, the National Opinion Research Center, was not interested in doing a local study. This book is not in any sense the culmination of this research project, but it is the largest product to date. We have published much from these data elsewhere, as have our students, and though we have both moved on to new projects, we expect other colleagues to continue using these data for many years. Like most large data sets, this one required many skilled hands to bring it in successfully. For their hard work, methodological acumen, and extraordinarily intelligent coping skills, we thank our students (now colleagues) who saw the project through to completion: Jack Beggs, the project straw boss; and Cathy Fladung and Sibylle Allendorf, his talented coworkers. We are also grateful to the University of Illinois's Survey Research Laboratory for its professionalism and expertise, which helped bring an untried methodology to a successful conclusion. James Zuiches, the NSF Sociology Program director who funded the project, was unfailing in his support and assistance throughout the three years of the original and supplementary grants. Our colleagues at the University of Illinois at Chicago and the then department head, John Johnstone, went far out of their way to accommodate our needs as the survey was done. As the manuscript took shape, Mildred Schwartz provided many useful editorial suggestions and gave generously of her time in reading and rereading several sections. Finally, we thank Arne Kalleberg for adamantly insisting that we write this book, and Eliot Werner of Plenum Publishing Corporation, for waiting patiently for it while we completed other things first. This project has been a fully collaborative effort from its beginnings,



and this book is no exception. We have greatly enjoyed working in tandem over the past decade in what seems to us a very fruitful effort-although an effort complicated somewhat by the peripatetic inclinations of one of us. We do not attempt to disentangle our respective contributions to this book and simply list our names alphabetically. WILLIAM P. BRIDGES WAYNE J. VILLEMEZ


Chapter 1

Introduction Examining Bureaucratic E m p l o y m e n t Relations . . . . . . . . . . . . . . . . . S u m m a r y a n d Plan of the Book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 5 25

Chapter 2

Assembling Data on Employment Relations


The Metropolitan E m p l o y e r Worker (MEWS) S u r v e y . . . . . . . . . . . E m p l o y e e Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E m p l o y e r Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . O t h e r Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29 35 41 50

Chapter 3

The Elements of Bureaucratic Personnel Management


Theories of Bureaucratic Personnel M a n a g e m e n t . . . . . . . . . . . . . . . Empirical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

54 60

Chapter 4

The Scope of Bureaucratic Management M e a s u r e m e n t Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary ...................................................

71 77 78 101




Chapter 5

Higher-Level Effects on Bureaucratic Control and Due Process in Organizations: The Influence of State, Industry, Organization, and Job


Previous Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measures a n d Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . S u m m a r y a n d Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

103 107 109 123

Chapter 6

Bureaucratic Structure and Attachment to Firms: Linkages to Potential M o b i l i t y in the Market


Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hypotheses ................................................. Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125 130 136 140 158

Chapter 7

Employment Relations and Earnings


Previous Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis a n d Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

161 164 175

Chapter 8



The The The The

178 180 183 189

Shape of Bureaucratic E m p l o y m e n t Relationships . . . . . . . . . . Scope of Bureaucratic E m p l o y m e n t Relationships . . . . . . . . . . C o n s e q u e n c e s of Bureaucratic E m p l o y m e n t Structures . . . . . . Future of Bureaucratic E m p l o y m e n t Relations . . . . . . . . . . . . . .



A. Screening Questionnaire, S t u d y #410: Jobs a n d Workers in a Metropolitan Labor Market . . . . . . . . . . . . . . . . . . . . . . . . . . .




B. Employer-Identification Q u e s t i o n Sequence . . . . . . . . . . . . . . . . . C. A C o m p a r i s o n of Similar Data G a t h e r e d f r o m Workers a n d E m p l o y e r s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. C o n s t r u c t i o n of M e a s u r e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E. Classifying O c c u p a t i o n s as O r d e r l y External Markets . . . . . . . .

199 201 211 217





Series Publications


1 Introduction

In the fall of 1988, the U.S. Congress passed a bill requiring that employers with more than 100 employees provide 60 days' advance notification to affected workers in the event of a major reduction in the plant's work force. The reaction of employers, of their associations, and, indeed, of a wide spectrum of American conservatives was quite predictable. For them, employment was seen as equivalent to any other kind of economic contract, and the imposition of a state-mandated notification requirement was perceived as one more example of unnecessary government interference in the operation of free markets. In opposing this legislation vigorously, and ultimately successfully, business leaders were carrying on a long ideological tradition of asserting the primacy of the "employment-at-will" doctrine. The essence of this doctrine is that, unlike serfs or indentured servants, workers in a capitalist economy are free to quit, and employers are free to discharge them, whenever and for whatever reason they choose. Of course, if the plant-closing law had gone into effect, employment relations would have changed dramatically for those American workers whose jobs were held on a purely employment-at-will basis. However, what the conservatives' arguments tended to gloss over in this debate is the fact that many employer-employee relationships in American society have already transcended this short-run, here-todaygone-tomorrow set of assumptions. Or as the legal scholar Ian MacNeil (1980) expressed it, "discrete contracting" has given way to "relational contracting" in the establishment of many employment relationships. Although there are many publicly visible examples of this shift, such as professional athletes' no-trade clauses and mulfiyear contracts, executives' "golden parachutes," and the experimentation of some unions and companies with Japanese-style "lifetime employment" systems, more subtle changes have also occurred as codified seniority systems, internal


Chapter 1

labor markets, and bureaucratic promotion practices have filtered through the occupational structure. Without debating the specifics of the 1988 plant-closing bill, it is clear that our evaluation of this, or any other legislation, depends on accurate knowledge of the kinds of social arrangements that employers and employees have fashioned in different industries and different firms. These employment relationships are the subject of this book. Our perspective on them is sociological; that is, we see them as providing a link between employees and their employers in which mutual rights and obligations are established. Furthermore, we assume that there are regularities in these relationships that express the properties of a larger system, and we follow the current sociological convention of referring to these regularities as social structure. Sociological rhetoric and everyday experience coalesce w h e n we recognize that the structure of employment relationships--the rules and norms and the legitimate expectations that develop about t h e m - - o f t e n put limits on the damage that could be done if free rein were given to the personal prejudices and individual quirks that each of us brings to our job. Thus, there will always be questions that our boss cannot ask and requests that he or she cannot expect us to satisfy, and from his or her point of view, there are excuses that we cannot hope will be accepted. Moreover, the basis of these protections is not in many instances our threat to quit or the bosses' willingness to fire, but a set of rules and procedures that have become established as organizational norms. To the extent that this has happened, the employment relationship has become bureaucratized. Our working definition of an employment relationship is the typical set of terms and conditions that regulate the exchange of labor for money between an employer and a given category of employees laboring under his or her auspices. Transcending the specific quantities of work and money (or other material benefits) agreed to by the employee and the employer, this employment relationship most often extends to other matters such as grievance procedures, expectations about promotion chances, and stipulations about procedures for making any change in the relationship that might be desired by either party. Defined in this manner, employment relationships are shown by our research to vary considerably among employers and among workers, but they are most often bureaucratically defined (rather than uniquely negotiated) relationships. Historians, economists, and other sociologists have written extensively about the emergence of institutional mechanisms that regulate employment relationships and that blunt the impact of raw market forces on employment decisions. However, much of their discussion has



been motivated by purely theoretical, or even ideological, concerns. Neo-Marxists have portrayed these mechanisms as yet another system of capitalist labor control. Some sociologists and economists have attempted to explain their adoption according to market structure considerations (monopoly, oligopoly, etc.). Other economists have examined them from the viewpoint of understanding whether these practices enhance economic efficiency, both for individual firms and for the economy as a whole. Organizational theorists have suggested that both creeping bureaucracy and corporate imitation play an important role. Ironically, because these studies have emphasized sorting out institutional and historical causes, attention has been diverted from examining the workers and employers who are the key actors in the establishment of these employment structures. As a complement to these largely historical studies, which rely on aggregated and documentary data sources, we will survey the kinds of arrangements that exist, and who is affected by them: Do blacks and women, for example, have access to structured mobility systems and grievance procedures equal to the access enjoyed by white males? How do the expanding service industries compare to the manufacturing sector in the formalization of employee rights and responsibilities? We will also consider some possible effects of these institutionalized structures: To what extent are employees' pay and the causes of their pay molded by other aspects of their employment relationship? What about their job satisfaction? To what extent do employees who are sheltered from supply and demand forces become dependent on their current employers? Although answers to these and similar questions have direct relevance to the historical and ideological issues mentioned above, they are also of considerable importance in their own right. In answering these questions, we will build on what we believe is the most important insight generated by existing writing: Employment relationships are strongly influenced by organizational considerations. These considerations are relevant because employment ties usually connect individuals to organizations, not simply to other individuals, as in marital, filial, and friendship relationships. One implication is that continuity in employment relationships frequently assumes more importance than continuity in other commercial transactions. When employers are organizations, be they private corporations, government agencies, or nonprofit associations, they share in the characteristics of all groups. Thus, their interactions with workers and potential workers can be interpreted in terms of the concept of membership. Even though employing organizations are highly inconsistent in their treatment of employees as m e m b e r s - - a s the debate over plant-dosing legislation illustrates--it is


Chapter I

an unusual employing organization that never invokes the worker-asmember metaphor. With this ideological construct as a backdrop, it becomes more difficult for employers to act on a day-to-day basis as if their employees are only trading partners in a purely economic exchange. More important, the fact that employers are often organizations opens up the possibility that various aspects of the employment relationship become institutionalized as part of the operating procedure of the organization itself. This institutionalization can be illustrated by the principle of seniority, especially its application to the problem of rationing layoffs during economic cutbacks. Thus, Philip Selznick (1969) noted that seniority (1) has become a prevalent operating procedure in a large number of nonunion contexts and (2) has been "welcomed as a contribution to orderly management" even though it is an "important limitation on [managerial] discretion" (p. 88). He added that a "keen sense of professionalism in personnel relations may be compatible with a 'mechanical' procedure [seniority] w h e n the latter is an integral part of a rationalized system of personnel practice" (p. 89). What we would like to emphasize is that this kind of institutionalization, where one of m a n y possible solutions to what is essentially a market-based problem becomes transformed into a legally and culturally supported norm of procedure, is intimately associated with the advent of organizations--and large organizations--as principal employers. Many other aspects of employment relationships, not just seniority, have become institutionalized in this manner, a notion aptly encapsulated in the title of Sanford Jacoby's book, EmployingBureaucracy(1985). Here we extend his historical inquiry by exploring, in contemporary society, the prevalence of other aspects of bureaucratic employment relations, such as internal labor markets, the codification of employment rules and regulations, and the establishment of rights to due process in the administration of these rules. It is important to know where these relationships are most highly developed, and which types of workers and which types of employers are joined together in them. While "good jobs" are likely to be imbedded in these kinds of bureaucratic employment relationships, there has been little investigation of their consequences for individual workers, particularly in their relationships to external labor markets. Our research leads us to argue that bureaucratic employment relationships, with their emphasis on "internal" opportunity and organizationally grounded fairness, have the potential both for creating dependence and for liberating workers from the arbitrary whims of individual bosses and supervisors. By dependence, we mean a behavioral withdrawal from the outside market so that bureaucratically employed workers lose touch with alternative employ-



ment possibilities in the outside world. Whether or not this dependence is accompanied by feelings of loyalty and psychological commitment to their current employers, workers are just as tied to their current situations. To summarize, employment relationships are a critical element in the experience of life in industrialized societies. Because employment in these societies typically involves the establishment of a social tie between an individual and an organization, the form these relationships take is unavoidably influenced by the organizational context in which they occur. Prototypical bureaucratic employers have bureaucratic personnel systems that include provisions for organizational "careers" (internal labor markets), along with mechanisms for ensuring due process and dispensing some measure of industrial justice. In their fully developed mode, these personnel administration systems have been described, only partly metaphorically, as "internal states" (see Burawoy, 1979). But it is unlikely that these practices are fully developed or implemented ubiquitously. For one thing, there are many small, nonbureaucratic employers w h o are largely immune from most of the forces that have encouraged the development of personnel systems (see Granovetter, 1984). For another, even in large organizational settings, there may be inconsistent adoption of bureaucratic personnel practices with regard not only to the type of practices that are implemented but also to the evenness of implementation across organizational subunits. Thus, we can anticipate that different types of workers will have different amounts of exposure to bureaucratic personnel management.


The central issues of this book are related directly or indirectly to the "stratification question": Who gets what and why? The stratification approach applies, first, because the practices and the protections that we are studying are inherently a "social reward," different in type from but parallel in importance to income and social prestige.1 Other things being equal, it is better to be in a job where one has good promotion prospects, the right to appeal the arbitrary decisions of one's superiors, and clearly stated expectations of rights and responsibilities. Although some studies a As David Stark (1986)insightfully pointed out, there may be more uncertainty in capitalist economies about job security than there is about income (the reverse being true in some socialisteconomies). From this possibility,he deduced that American internal labor markets are structured around the problem of rationing employment stability, whereas socialist ones are oriented around rationing income.


Chapter 1

have implicitly recognized these principles, these studies have often b e e n r a t h e r l i m i t e d i n scope. By e x a m i n i n g a r e p r e s e n t a t i v e s a m p l e of m e t r o p o l i t a n w o r k e r s a n d t h e i r e m p l o y e r s , w e s h a l l p r o v i d e a less bia s e d a n d m o r e c o m p r e h e n s i v e set of a n s w e r s t h a n o t h e r s h a v e offered. 2 By f o c u s i n g o n b u r e a u c r a t i c e m p l o y m e n t i n s t i t u t i o n s , o u r a n a l y s i s will b o t h b r o a d e n a n d d e e p e n t h e s t u d y of o r g a n i z a t i o n a l c o n t r i b u t i o n s to stratification. It will b e d e e p e n e d b e c a u s e , r a t h e r t h a n t a k i n g i n t e r n a l l a b o r m a r k e t s a n d o t h e r b u r e a u c r a t i c p e r s o n n e l practices as " g i v e n s , " w e will b e e x p l a i n i n g t h e i r d i s t r i b u t i o n a n d t h e i r i n t e n s i t y . It will b e b r o a d e n e d b e c a u s e w e will b e c o n s i d e r i n g t h e effects of t h e s e s t r u c t u r e s n o t o n l y o n w a g e s b u t o n o t h e r d i m e n s i o n s of i n d i v i d u a l o u t c o m e . 3 S i n c e t h e p r i m a r y f o c u s of this b o o k is o n t h e i n t e r f a c e b e t w e e n i n d i v i d u a l s a n d o r g a n i z a t i o n a l f o r m s , t h e l i t e r a t u r e o n t h e e m p l o y m e n t rela2 For example, Marcia Freedman's book Labor Markets: Segments and Shelters (1976) is based entirely on aggregate data and does not address consequences. 3 Despite our focus on individuals in their jobs, it is important to note at the outset that this book is not intended to be an argument for, or even an exemplar of, the individualist perspective. We do not choose to join that argument, but if we did, it would be on the structuralist side (see Pfeffer, 1982, pp. 18-23, for a summary of the argument in the organizational literature; see Stinchcombe, 1979, for a summary of the argument in the stratification literature). Although employment relationships are surely socially constructed phenomena, they are structural features that constrain individual behavior on both sides of the relationship (as do Marx's "relations of production"); they are neither emergent nor renegotiated in everyday interaction. Certain job configurations and bureaucratic structures produce, in the aggregate, higher or lower levels of monetary return for the personal investments of the occupants of the job and may be expected to produce higher or lower levels of satisfaction with the job, both of which may be expected to produce, in turn, higher or lower levels of tenure or, its converse, turnover. These simple examples suffice to make the point: Income level, job satisfaction, and quit or stay decisions are individual phenomena. Yet, we would not agree with Collins (1981) that statements regarding organizational effects on these phenomena are simply macrosummaries of microevents. Individuals come and go in the organization and in the job; levels of income and satisfaction and rates of turnover are most often more constant. Some jobs in some organizations are known to have high turnover, others to have high satisfaction, and still others to produce a high return on human capital--all regardless of the individuals in them. These are known characteristics of the jobs, not aggregations of accidental confluences of individuals of certain dispositions. They are structural phenomena, not individual phenomena, and do not depend in the aggregate on the idiosyncratic reactions of individuals. Unlike Mayhew (1980), we do concede that structural statements require certain implicit micro-level assumptions, but then microlevel theories themselves require assumptions regarding the constancy of certain biochemical reactions. We see little difference between the two sets of assumptions. So, statements about organizational effects on, say, job satisfaction levels are structural statements, just as are statements about the effect of "environmental uncertainty" on "organizational decentralization." In both cases, the actions of individuals are ultimately involved; in both cases, such actions are a stochastic given and are thus irrelevant.



tionship is crucial. Although much of the extant literature in this area is historical and comparative, it is nonetheless a useful source of ideas for addressing our basic issues.

Previous Literature The major perspectives on bureaucratic management can be usefully categorized into three groups: the technical-rational perspective, the industrial-sectorial perspective, and the institutional perspective. In a focus on the employment relationship, three issues are critical in each of these perspectives: what they imply about the shape of bureaucratic management, the scope of bureaucratic management, and the consequences of bureaucratic management. By the shape of bureaucratic management, we mean the defining characteristics: the key elements of management isolated by this perspective and the interrelationships among these elements. The scope of bureaucratic management refers to the distribution of various bureaucratic elements of management structure: distribution by industry, by occupation, by firm or establishment characteristics, by level of technology, and by many other aspects. The different perspectives imply different distributions. The consequences of bureaucratic management for employees (at all levels) must be the major focus of any stratification study, and the different organizational perspectives, with their different predictions of shape and scope, lead to different predictions of the effects of management structures on worker attachment, reward patterns, distributional inequalities, worker satisfaction, and other aggregate individual outcomes.

The Technical-Rational Perspective Before what we call the technical-rational perspective, there was the merely "rational" perspective. Taylor's (1911) "scientific management" prescriptions, with their assumption of at least economically rational workers behaving rationally, are the earliest example of the perspective. Perhaps the best single example of this view, though, is Blau's emphasis (1964) on the interaction patterns of individuals as the key to organizational structure. Explaining the phenomenon by which normal "direct" exchange transactions between individuals become "indirect" exchange transactions in organizations, Blau noted: As long as subordinates obey the orders of a superior primarily because they are obligated to him for services he has rendered and favors he has done for them individually, he does not actually exer-

Chapter 1 cise authority over the subordinates, and there is a direct exchange between him and them . . . . The establishment of authority means that the normative constraints that originate among the subordinates themselves affect their compliance with the orders of the superior--and indirect exchanges now take the place of the former direct ones. The individual subordinate offers compliance to the superior in exchange for approval from his colleagues; the collectivity of subordinates enforces compliance with the superior's directives to repay its joint obligations to the superior; and the superior makes contributions to the collectivity in exchange for the selfenforced voluntary compliance of its members on which his authority rests. (p. 329) This exchange perspective clearly assumes that order depends on rational choices being made in the workplace; they are more benign than Hobbesian choices because compliance is being exchanged for approval rather than for life, but the model of humans is the same. Also the same is the conception of an organization as an aggregate of individuals; here, the core of a complex organization lies in the rationality-based interactions among individuals. Another clear, though very different, example of the early rational perspective can be found in the works of those usually characterized collectively as members of the " h u m a n relations" school. The origins of this school are usually traced to the work of Roethlisberger and Dickson (1939) and their classic report on research in the Hawthorne plant in Cicero that seemed to isolate important informal group structures. Others identified with this school include Rensis Likert (1961, 1967), Douglas McGregor (1960, 1967), and Chris Argyris (1960, 1964). The essential argument is that h u m a n relations (rather than just economic relations) are essential to a smoothly functioning organization. As Likert (1967) phrased it: The best performance, lowest costs, and the highest levels of earnings and of employee satisfaction occur when the drive for a sense of personal worth is used to create strong motivational forces to cooperate. (p. 75) Persons in organizations were driven by what Argyris (1960) termed a desire for self-actualization, and the overall point was that an organization can efficiently achieve its own self-actualization only if the individuals in it do so as well. This work has been criticized on both empirical and ideological g r o u n d s - - t h e latter because of its argument that workers who question organizational values need " a d j u s t m e n t " - - b u t its influence remains in the field.



T h e general perspective of individual rational action, indeed, is pervasive in the field, many of its assumptions being more-or-less taken for granted. In an excellent summary and critique of the perspective, Pfeffer (1982, Chapter 2) identified several variants sharing the same basic assumptions. He pointed to "expectancy theory" (Vroom, 1964), which asserts that individuals act expecting (or considering the probability) that their action will lead to a valued outcome; to "goal theory" (Locke, 1968), which argues that individuals act to achieve personally derived goals; to "needs theory" (Alderfer, 1972; Maslow, 1993), which argues that individuals act to satisfy their needs; and to theories like that of his own (Pfeffer, 1981), which argue that people act to achieve power as well as to achieve other outcomes. The problems with all of these similar theories, Pfeffer (1982) noted, include: (1) Their presumption of the preexistence of purpose or intent; (2) their tendency to ignore the effects of context on behavior; (3) their use of individual-level constructs to build theories of collective or macrolevel behavior; (4) their heavy reliance on cognitive, information-processing assumptions about the causes of human activity; (5) their reliance on hypothetical constructs that reside largely in people's heads and thus that are problematic to observe and measure; and (6) their fundamentally tautological nature, which makes them somewhat theoretically suspect. (p. 42) Even more telling, perhaps, is the criticism of this perspective by Perrow (1986), who contended that, in the extensive research conducted by devotees of this school, we have learned "a great deal about psychology and social psychology, but little about organization per se" (p. 114). In part, he argued, this is because organizations cannot be explained by explaining the actions of individuals or groups within them, but more fundamentally, it is because what we think we have learned about psychology and social psychology from these studies is probably itself incorrect:

People in these fields no longer agree that such things as norms, values, and personality really exist or account for much; these concepts may only give a false sense of order to a world that both the academic and the person in the street desperately want to order. Sociology, too, is having some difficulty swallowing the simple, obvious proposition that attitudes predict behavior. (The proposition that morale predicts productivity is just one specification of this.) (Perrow, 1986, p. 115) Instead, Perrow noted, it may be that "organizations run backward," that is, that individuals socially construct a world that they genu-


Chapter 1

inely believe was there all along, a world actually moved by accidents and random events. If such is the case, it is small wonder that attitudes do not predict behavior because behavior precedes them in time. Perrow cited Karl Weick (1979) and March and Olsen (1976) for this counter to the rational model, but the idea far predates them. Perrow was referring to what Pareto called "sentiments," those rational after-the-fact explanations of nonrational behavior. Pareto (and then Parsons) moved from this argument to the necessity of lower-level explanations of behavior; Perrow inferred from the argument the necessity for higher-level explanations.

Expectations following from Rational Theories. Rational theories have straightforward implications for bureaucratic management effects. The shape of bureaucratic management should be everywhere the same because universal principles of organization govern efficiency considerations. The scope of bureaucratic management is relatively fixed. It should vary only by the size of the organization, and variation across industries, occupations, and firms or establishments should occur only to the extent that the size distribution of organizations varies across such units. As to the individual effects of bureaucratic management, most rational theories would see its overall effect as neutral. Changes necessitated by increased size that lead to decreased worker satisfaction (e.g., deskilling) would have to be balanced by changes that increase worker satisfaction (e.g., higher wages), or else worker productivity would decline. A search for consequences of bureaucratic management that began from this perspective would be essentially a search for the compensating mechanisms that organizations differing in various dimensions use to achieve optimal exchange relations with workers. As theories become more complex, however, their implications for employment relationships change. The preceding types of "rational" theories consider primarily the rationality of individual actors in organizations. Many other variations emphasize a higher level of analysis and dwell on the rational, goal-seeking behavior of whole organizations (whether treated as aggregations of interlocked individual actions or as actors in their own right). The notion that organizations have goals apart from the individuals in them (which, indeed, can determine the goals of those individuals) and act rationally to pursue those goals has informed the work of many organizational theorists, most of w h o m have accepted Simon's idea (1957) of "bounded rationality," a rationality that results not in choices that are optimal, but in choices that seem optimal given the environment and the political history of the organization. Actors are, in Simon's terms (1957), "intendedly rational, but only limitedly so"



(p. xxiv). In organizational choice, that is, there is no simple m e a n s - e n d scheme; there are multiple ends (objectives), each with its value, and many possible paths to each end. The process of choosing the best path to the "best" objective is one that is circumstantially circumscribed. Theories of such processes require underlying explanations of h o w individually rational actions come to be aggregated to organizationally rational actions. None of the existing theories explain the aggregation procedure; some simply assume it, others argue that those organizations in which it does not occur do not survive, and still others argue that organizations, whatever the true reason, behave "as if" they were rational actors. The first group assumes what is to be demonstrated, the second group ignores noncompetitive markets and the public sector, and the third group casts out explanation, arguing for the value of prediction instead. If, indeed, there are rational organizations rationally pursuing their goals, one can assume that their internal organization is optimally structured to attain these goals, and that variations in goals and in the appropriate means to achieve them result in differences across organizations in job content, bureaucratic structure, and other relevant characteristics. This is not a radical notion, nor is it a very complex idea. Weber saw bureaucratization as an ideal type, and thus as something that would vary with historical and other circumstances. Discovering a conceptual dimension that effectively organizes these differences, however, has proved difficult for organizational theorists. Most early suggestions for an appropriate dimension have been interesting but not terribly useful. They have included sorting by the type of task involved (programmed vs. nonprogrammed) (March & Simon, 1958); sorting by the type of leadership (authoritarian, laissez-faire, or democratic) (Gouldner, 1954); sorting by the type of power used to gain compliance (coercive power, remunerative power, or normative power) (Etzioni, 1975); sorting by w h o benefits from the organization (members, clients, owners, or the public) (Blau & Scott, 1962); and many others.

Contingency Theory. The most promising research was stimulated by the work of Woodward (1958, 1965), w h o argued for technological type as a key organizing dimension, a focus that is one aspect of what has come to be called contingency theory. Hall (1987) summarized her work: Woodward's findings show that the nature of the technology vitally affected the management structures of the firms studied. The number of levels in the management hierarchy, the span of control of first-line supervisors, and the ratio of managers and supervisors to other personnel were all affected by the technology employed. Not only was structure affected, but the success or effectiveness of the


Chapter 1 organizations was related to the "fit" between the technology and structure. The successful firms of each type were those that had the appropriately structured technical systems. (pp. 105-106)

The most immediate implication of this approach is that one could expect the type of technology to help determine the nature of the employment relationship. Woodward's findings have been replicated for the United States (Zwerman, 1970), greatly extended and refined (see Perrow, 1986, pp. 140-146, for an excellent summary), and challenged and refuted (the Aston group found little support for the idea in its crude form; Mohr, 1971, found no empirical support for it in any f o r m - - a t least, in public organizations). The major problem comes in defining technology, a more serious problem than it seems at first glance, because there may be a variety of technological forms within the same organization. Hall (1962), for example, showed that different units of the same organization may have different structural forms. Perrow (1967) offered a still-popular typology based on the nature of the search process, the number of exceptions encountered in doing the work, and the nature of the raw material. The Aston group identified three types of technology-knowledge, materials, and operations technology--and then concentrated on the last in their research (see Hickson, Pugh, & Pheysey, 1969). Scott (1981) provided an extensive review of definitional attempts and constructed a typology of technology definitions by cross-classifying the three types just mentioned with processing stages (input, throughput, and output); all of the resulting nine cells describe a definition of technology, and all nine have work cited within them. Current thinking sees the type of technology as crucial to organizational form, but most likely in interaction with other factors. But it is not only technology that is assumed to have an effect on organizational structure, and thus indirectly on the bureaucratic employment relationship. The other two most cited contextual factors that are presumed to influence efficient organizational structure are the size of the organization and its competitive environment. Variations in these two, along with technology, are seen as the contingencies that determine the mode of organization; the central tenet of this approach is that the appropriate organizational design depends on the organization's context: "There is no one best way to organize" (Galbraith, 1973, p. 2). The idea of size as a determinant of organizational structure has probably generated more research than any other dimension. Blau has been one of the major figures in this research, investigating the relationship between size and differentiation and finding, to oversimplify somewhat,



that with increasing size come increases in the number of levels, departments, and even job titles within an organization (see Blau, 1968, 1970, 1972, 1973, and a host of articles by Blau and various associates). The relationship between size and differentiation was found not to be linear because the rate of increase in differentiation declines with size. The Aston group also found size to be crucial to organizational structure, causing, among many other things, decreased concentration of authority and increased specialization and reliance on written records. Some have called the relationship into question (e.g., Argyris, 1972; Hall, Haas, & Johnson, 1967; Hall & Tittle, 1966), but it seems empirically established beyond doubt (see Kimberly, 1976, for a review of studies). Meyer (1972) found in a longitudinal study that size caused structural changes, a finding that adds weight to the argument of genuine causality. Pfeffer (1982), however, noted: If one assumes reasonably constant spans of control, the relationship between size and differentiation is mathematically true by definition. Thus, there is some concern that the effects of size on differentiation are not of great theoretical interest or importance. (pp. 149-150) Though the theoretical importance of the size component may be moot in organizational theory (and most would not agree that it is), it is clearly of crucial importance in the employment relationship because, of all the factors that affect that relationship, most have been found empirically to covary with size. Size has been found to affect differentiation, mechanisms of control, centralization, formalization, the size of the administrative structure, reliance on written records, job satisfaction, wages, return to education, and many other factors that clearly impinge on the employment relationship (see Kalleberg, Wallace, & Althauser, 1981; Stolzenberg, 1978; ViUemez & Bridges, 1988). The environment of organizations as a crucial variable has been noted on many occasions (e.g., Burns & Stalker, 1961), but it really became a central theoretical focus in 1967, with the publication of three major works that all made essentially the same argument (Lawrence & Lorsch, 1967; Perrow, 1967; Thompson, 1967). Burns and Stalker (1961) found that a more formalized structure was appropriate for organizations in a stable environment, whereas a more flexible, less formalized and centralized structure was optimal for organizations in less predictable environments. Subsequent work has expanded this basic notion, and the earliest focus was on environmental uncertainty. Duncan (1972) expanded the definition of uncertainty to two elements: "complexity" and "variability." The first refers to the number of elements that must be


Chapter 1

dealt with, the s e c o n d to the extent that these e l e m e n t s c h a n g e o v e r time. L a w r e n c e a n d Lorsch (1967) e x p a n d e d the definition d o w n w a r d , a r g u i n g that different units within the s a m e organization faced different e n v i r o n m e n t s . 4 T h e e n v i r o n m e n t a l - u n c e r t a i n t y c o n c e p t h a s b e e n red u c e d to its p r e s u m e d p r e c u r s o r s in a n u m b e r of w a y s , the m o s t w i d e l y accepted b e i n g that of Pfeffer (1982; Pfeffer & Leblebici, 1973; Pfeffer & Salancik, 1978). Pfeffer c o n t e n d e d that the r e l e v a n t d i m e n s i o n s of the organizational e n v i r o n m e n t w e r e the d e g r e e of c o n c e n t r a t i o n of resources, the relative scarcity of resources, a n d the d e g r e e to w h i c h t h o s e in the organizational field are interconnected. 5 O t h e r s h a v e e x p a n d e d a n d contracted the n u m b e r of e n v i r o n m e n t a l d i m e n s i o n s e v e n m o r e , a d d i n g further analytic detail. For e x a m p l e , Aldrich (1979) isolated six m a j o r d i m e n s i o n s : e n v i r o n m e n t a l capacity, h o m o g e n e i t y , stability, concentration, turbulence, a n d m a r k e t c o n s e n s u s . Dess a n d Beard (1984) collapsed these six to three; Hall (1987) cross-classified t h o s e t h r e e w i t h s e v e n o t h e r d i m e n s i o n s . A l t h o u g h c o n c e p t u a l clarity h a s p r o g r e s s e d (or not, d e p e n d i n g o n w h o m o n e reads), the essential p o i n t h a s r e m a i n e d the s a m e : Organizational e n v i r o n m e n t has a strong impact o n organizational structure a n d m u s t be t a k e n into account. If it is n o t s e e n as quite the essential d e t e r m i n a n t of structure it once w a s (in P e r r o w ' s w o r d s , if the w a v e of theorizing h a s lost its h a n d s o m e crest of frothy promise), organizational e n v i r o n m e n t is still r e g a r d e d as a n i m p o r t a n t influence.

Implications for the Employment Relationship T h o s e theories that w e h a v e t e r m e d merely rational i m p l i e d that the s h a p e of bureaucratic e m p l o y m e n t relations w o u l d be e v e r y w h e r e the s a m e , that their s c o p e w o u l d v a r y only b y the size of the organization, a n d that their c o n s e q u e n c e s w o u l d h a v e to be neutral. The technicalrational theories i m p l y m u c h m o r e variation. In these theories, it no longer necessarily follows that universal principles of o r g a n i z a t i o n gov4 That is, they pointed out that the subenvironment of a marketing unit was very different from the subenvironment of an R & D unit, which in turn was very different from the subenvironment of the production division. The most effective organizations, they found, were those that accomplished two things: (a) allowed each department to have its own different internal structure in accord with its subenvironment and (b) effectively coordinated and integrated these different substructures within a broader structure. s He argued that concentration reduces conflict, which in turn reduces uncertainty; that resource scarcity increases uncertainty by increasing conflict; and that interconnectedness increases uncertainty by increasing both interdependence and conflict. The obvious implication of Pfeffer's work is that research should focus on the causal precursors of uncertainty, like concentration, scarcity, and interconnectedness, rather than on the tooglobal concept of uncertainty itself.



ern efficiency considerations. What is efficient is now contextually determined, and the shape of employment relations can be expected to vary across organizations with different technologies and environments. Similarly, the scope of those elements of management structure that define the employment relationship can no longer be expected to be fixed. They will be differentially distributed across the relevant contexts. As to the individual-level consequences that follow from this perspective, they have not been systematically explored, but no one has suggested any effect other than a neutral one. We argue below that a case could be made for individual consequences following from this perspective that are far from neutral--as other perspectives contend. While debate was raging over the issue of how (and whether) environments influenced organizations, there was little apparent interest in understanding how workers were affected by their employers' environments. The key word here is apparent, for the environmental argument was being made in a different way at the same time, though by researchers from a number of different traditions who did not call it an environmental argument and who made little reference to the organizational literature; some of them may have been unaware of the related research. For example, at the core of the segmented-labor-market theorists' arguments is a clearly environmental argument. As Baron (1984) put it: Institutionalists and Marxists suggest that firms' relations to their product markets and their control over industrial and political environments influence how work is organized and rewards distributed (Galbraith 1967, Averitt 1968, Bluestone 1970, O'Connor 1973, Friedman 1977, Edwards 1979, Berger & Piore 1981, Gordon et al. 1982). While the combination of factors stressed depends on the investigator's theoretical perspective, the expectation that good jobs are concentrated in "core" or monopolistic firms and industries is explained by some combination of the following attributes: technical mix; level of union and management interest in employment stability; ability to absorb higher labor costs due to market structure and demand schedules; growth, concentration, and change in organizational forms; differences in the quantity and quality of managerial activity; and economic and political relationships with the state and foreign markets. (p. 48) These are the studies reviewed below as examples of the workercontrol (or industrial-sectorial) perspective, but Baron was quite correct to characterize them as studies of the impact of organizational environments. They do not characterize themselves that way, but they clearly are, because the attributes stressed that differentiate sectors are ele-


Chapter 1

ments of organizational environments. The argument is simply narrowed, and it is assumed that organizations must adapt different means of worker control in response to differing environments. It is also true that at least one variant of the "institutionalist" theory to be discussed below is based on reasoning similar to that found in the organization-environment literature. Thus, despite the lack of explicit attention given to the labor implications of "environment" in this tradition, its premises have not gone unheeded. Furthermore, if these various sidetracks of organizational theory are followed to their own conclusions, one is no longer operating in a world that runs according to either technical or rational principles. For that reason, we defer our discussion of these theories until we consider another recent attempt to explain employment relationships from a more purely technical-rationalist stance. Though the theory to follow falls clearly in the "technicalrational" category, we discuss it separately and extensively because it is a large and sophisticated attempt to deal with employment relationships (though it does not use that terminology).

Transaction Cost Analysis. One of the best-known recent "rational organization" approaches is the transaction cost (or "market failure") analyses of Williamson (1975, 1981, 1985; Williamson & Ouchi, 1981). Seeking to explain w h y organizations exist at all, instead of all their functions simply being handled by the market, Williamson (1985) argued that some transactions can be handled more efficiently by an organization than by the market. He summarizes: Transaction cost economics maintains.., that organizational variety arises primarily in the service of transaction cost economizing. That approach is to be distinguished not merely from the technological approach to economic organization but also from power approaches, which ascribe nonstandard forms of organization to monopoly purposes or class interests . . . . Transactions, which differ in their attributes, are assigned to governance structures, which differ in their organizational costs and competencies. (pp. 387-388) The organization of work and of labor, that is, is governed by efficiency considerations, for labor market transactions, like all other transactions of organizations, have costs that can be minimized by effective organizational arrangements. Technology is important, as is the ownership of assets, but, Williamson maintained, neither separately determines economic organization, nor do they do so in interaction with each other. The transaction is the key unit of analysis. For an example of what he meant, consider the difference between this approach and the neoclassical approach to the most basic employment relationship. A worker



w h o has assets ( h u m a n capital) to sell to an employer would, in the neoclassical model, sell t h e m in the marketplace where "faceless buyers a n d s e l l e r s . . , m e e t . . , for an instant to exchange standardized goods at equilibrium prices" (Ben-Porath, 1980, p. 4, quoted in Williamson, 1985, pp. 55-56). In the transaction cost model, the worker's sale of assets is neither instantaneous nor faceless; it takes place within an organizational structure, for such a transaction can be more cheaply a n d efficiently m a d e within such a structure t h a n in the marketplace. Out of this need for within-organization transactions grow different organizational structures with differing governance procedures. This need is h e i g h t e n e d by w h a t Williamson called asset specificity--in this particular context it could be called job- or firm-specific skills. Williamson (1985) noted: Specialized governance structures are more sensitively attuned to the governance needs of nonstandard transactions than are unspecialized structures, ceteris paribus. But specialized structures come at a great cost, and the question is whether the costs can be justified . . . . The benefits of specialized governance structures are greatest for transactions supported by considerable investment in transaction-specific assets . . . . More generally, the object is not to economize on transaction costs but to economize in both transaction and neoclassical production cost respects . . . . Human capital investments that are transaction-specific.., evolve during contract execution. Specialized training and learning-by-doing economies in production operations are illustrations. Except when such investments are transferable to alternative suppliers at low cost, which is rare, the benefits can be realized only so long as the relationship between the buyer and seller is maintained. (pp. 60-62) Williamson (1981, 1985) delineated four basic types of governance structures, keyed to two central dimensions: h u m a n capital specificity a n d w o r k - r e l a t i o n separability, the latter referring to the ease with which individual work output can be monitored. In the first case, where h u m a n capital is nonspecific a n d the work is easily monitored, there exists no special governance structure to sustain the e m p l o y m e n t relation; on dissatisfaction, it can be terminated by either party without serious costs being incurred (workers can move to other employers without loss of productivity, a n d employers can secure other workers without start-up costs). This e m p l o y m e n t relationship is called a "spot market" by Williamson. The second type, a "primitive team" market, occurs w h e n the worker assets are nonspecific, and individual output is difficult to meter with any precision. Individual team members can be


Chapter 1

replaced without cost on either side, but a supervisory role is required in this arrangement. The third and fourth types of employment relations both involve very specific human capital assets, a situation in which the relationship cannot be terminated without costs, usually to both sides. In the "obligational market," individual performance is easily monitored, so the only concern is protecting the relationship. Procedural safeguards and monetary incentives (such as severance pay) will protect the worker by discouraging dismissals; nonvested retirement and other benefits will protect the employer by discouraging quits. In the "relational team" market, assets are very specific and individual performance is difficult to monitor. Bureaucratic management devices become most complex under these circumstances. Williamson (1985) noted: The firm here will engage in considerable social conditioning to help assure that employees understand and are dedicated to the purposes of the firm, and employees will be provided with considerable job security, which gives them assurance against exploitation . . . . A sense that management and workers are "in this together" furthers all of those purposes. (p. 247) According to this perspective, we should expect to find bureaucratic management forms (both shape and scope) varying by the type of labor transaction required, that is, by the type of employment relationship mandated by the type of work. In a "spot market," neither workers nor employers are constrained to maintain the relationship, and we can expect to find no special bureaucratic management devices in place. In a "primitive-team" market, we can expect to find supervisory positions and should expect to find relatively high rewards associated with supervision, as it is a crucial link. In an "obligational market," there will be bureaucratic structures whose express purpose is to maintain the employment relationship, as well as structures that provide both sides with rewards for maintaining it and penalties for severing it. Worker control will not be at issue in these markets, only worker continuity. Bureaucratic management should come to full flower in the "relational-team" market, adding authority structures and other specific control devices to the maintenance devices. From this perspective, there can be expected differences in the shape of bureaucratic management across occupations and industries, as well as across occupations within the same organization, depending on the occupational labor markets and on the technology of the work. Individual consequences will vary as well. For example, supervisory positions should be best rewarded in "team" markets and should be most available (as a mobility possibility) in those markets where main-



taining the relationship is crucial (only one of which is a team market). Another set of hypotheses that is consistent with this approach relates to the consequences of "mismatches," that is, situations where the governance structure is out of step with the requirements of the "transactional situation." For example, firms employing workers in non-asset-specific jobs (e.g., truck drivers) who nevertheless offer elaborate governance structures might be expected to compensate by paying lower wages. All in all, the effect of bureaucratic management should be largely benign because it will be in place only when needed to protect the interests of worker and employer. The consequences, that is, should be neutral, as the early rational theories implied. Not all scholars agree that the individual consequences of bureaucratic employment relationships will be largely neutral. We turn to the second of our broad perspectives, the industrial-sectorial perspective.

The Industrial-Sectorial Approach The industrial-sectorial (or worker-control) approach to organizational structure begins from a different premise, treating that structure as emerging not from a quest for efficiency in production, but from a quest to increase control over the workers. On the face of it, this may seem a precious, even silly, distinction, for if the efficiency (and thus the profitability) of production can be achieved by a more effective control of the workers (and it is clear that it can), then the seeking of such control should be just one more example of the quest for efficiency, and not a separate school of thought. Indeed, most major studies in the field grant that gains in control led to (and were engendered by the desire for) increased productivity and profitability. The differences between this perspective and that of the technicalrational one outlined above are not differences in substance but differences in focus and most especially in ideological reaction to the consequences of organizational change. To say this is not to trivialize those differences; they are substantial in some ways. To most in this school, worker control is a rational and obsessive concern of employers, so much so, in fact, that the whole of the historical transformation of work and of the employment relationship in the United States can be traced to the employer-worker conflict over work force control. Most who argue from this perspective adopt Marxian assumptions about the true sources of profit and capital accumulation: If such must come from the surplus value created by labor, then effective control of that labor will naturally be seen as crucial to any rational capitalist (e.g., Bluestone, Murphy, & Stevenson, 1973; Doeringer & Piore, 1971; Edwards, 1979; Gordon, 1972;


Chapter 1

Gordon, Edwards, & Reich, 1982; Harrison, 1972; Wachtel & Betsey, 1972). This perspective broadens the definition of efficiency, taking to heart Perrow's advice (1986): Beware of "efficiency" arguments that do not ask, "Efficient for whom, and at what costs to others?" Ask instead about "externalities," the social costs of organized activities that are not included in the price and are borne by those who benefit very indirectly, if at all, from the activity. (p. 278) The m e a n s - e n d efficiency at issue here results from what Useem (1982) aptly termed "classwide rationality," a phrase that captures the essence of this perspective. It argues not only that the goals of an organization are not necessarily those of all of its members, but that those goals can actually be antithetical to the desires of a whole class of organizational members. Drawing from Goldman (1980), Pfeffer (1982, pp. 163-164) summarized the control perspective as consisting of four basic steps. Because employers desire a labor force that can be controlled so that it will work toward the employers' interests and desire a labor force that is relatively inexpensive and powerless, (1) they select means of production that "deskill" workers as much as possible (see Braverman, 1974), and (2) they structure the employment relationship so that the owners' control over labor is as hidden as possible, but control over the workers and the managers is achieved. These two actions result in (3) resistance from workers, manifested as lack of motivation, absenteeism, turnover, and collective action, and (4) a cycle of continuing conflict and change caused by the underlying capital-labor conflict. Examining the same phenomena as others, those in the control school interpret them quite differently. For example, the internal labor market was seen by Williamson simply as a rational response to efficiency pressures. As Baron (1984, characterizing Williamson's thought) put it, "An employer's investment in long-term contracts is viewed as a rational response to turnover and costs, and any short-run disequilibria between marginal productivity and wages should cancel out throughout a worker's career with the firm" (p. 40). Edwards (1975, p. 7), on the other hand, saw the proliferation of the internal labor market in its various forms as a system "contrived and consciously designed" to perpetuate control over the work force. He saw the rise of bureaucratic management structures in the same context, arguing that the resulting legitimization of hierarchical authority structures in large firms made control of the work force less obtrusive and more acceptable (Edwards, 1979, p. 9).



From this perspective, there emerged a subfield of research. Edwards suggested that the control differences between the large firms in the monopoly sector and the smaller ones in the competitive sector would be mirrored by differences in the labor markets serving the two sectors. Following up on this suggestion, a number of researchers examined labor market "dualism" or "segmentation" (e.g., Beck, Horan, & Tolbert, 1978; Bibb & Form, 1977), and found empirical evidence of employment relationship differences across the labor markets. Specifically, they found differences in the stability of employment, in the existence of internal labor markets, in the individual return to human capital, and in the extent of bureaucratic management. There have been many criticisms of these findings (e.g., Baron & Bielby, 1980; Hauser, 1980; Pfeffer & Ross, 1981; Zucker & Rosenstein, 1981), but the main effects persist and at least require an alternative interpretation. According to this approach, bureaucratic management structures should vary by industry and should have control of the work force as their primary function. Workers should be best off in those industries where their need for skill and training give them bargaining power, for the structure of work has been largely determined by the struggle between owners and labor over who gets to specify the employment relationship. Core industries were formed out of the unionization struggle: The emergent decay of the drive system of labor management accelerated in the 1930s, as workers were able to organize industrial unions that dramatically advanced their struggles to resist corporate control over the production process . . . . Corporations responded to this challenge from labor by pursuing the implications of some of their early explorations of new "internal labor market" structures and by encouraging an integration of the industrial unions into a new collective bargaining structure. This corporate initiative... established the foundation upon which the process of labor segmentation was established. (Gordon et al., 1982, p. 236) Bureaucratic management structures can thus be expected to be found primarily in core industries, and their particular structure should reflect the extent of union power and union cooptation within each particular industry. Although it has never been stated clearly, within the core sector one might be expected to find the same balancing of interests predicted by the exchange perspective (or, in the language of contemporary economics, "compensating differentials"; see England, Farkas, Kilbourne, & Dou, 1988; Filer, 1985; Jacobs & Steinberg, 1988). Even here, such effects may be found only among those occupying positions of privilege based on racial, gender, occupational, or union status. According to this perspective, then, both the shape and the scope of bureaucratic management should vary widely across different industrial


Chapter 1

sectors and should vary in accordance with the history of labor demand and unionization.

The Institutional Approach Both the technical-rational and the sectorial and control perspecfives "rest on the shaky ground of rationality" (Hall, 1987, p. 119). An approach that does not rest on this ground, and that argues instead that organizational structure owes as much to coercion and imitation as to rational or efficiency considerations, has come to be called the institutional approach. In its most recent manifestations, this theory argues that organizations act within the confines of a "field" of other organizations and are heavily influenced by this fact. An organizational field consists not only of organizations actually interacting or those competing with each other, but of the whole group of organizations that "in the aggregate, constitute a recognized area of institutional life: key suppliers, resource and product consumers, regulatory agencies, and other organizations that produce similar services or products" (DiMaggio & Powell, 1983, p. 148). Organizations within the same field tend to model themselves on each other, not necessarily because the "field model" is the most efficient one, but because of outside regulatory influences, because of a desire to imitate the most legitimate and successful of their kind, because of normative pressures stemming from professionalization, or because of some combination of these three. Organizations, that is, tend to become institutions; this is an important transmogrification, as rational organizations will come and go or change as efficiency considerations change, but institutions are self-perpetuating and extremely protective of their existing structure. To speak of institutionalized organizations is "to speak of practices or procedures that are continued and transmitted without question, to speak of meanings that become typified and transmitted to newcomers in the organization and shared without thought or evaluation" (Pfeffer, 1982, p. 239). Economic theories predict the greatest success for those organizations that are most efficient and most effectively coordinated, but in institutional theory, as Zucker (1987) noted, efficiency and success do not necessarily covary: "Organizational conformity to the institutional environment simultaneously increases positive evaluation, resource flows, and therefore survival chances, and reduces efficiency" (Meyer & Rowan, 1977, quoted in Zucker, 1987, p. 445). This view of organizations and their adaptation to their environment began formally with Selznick's treatment (1957) of some organizations as institutions with a life that transcends the task for which they



were created. He saw the strain toward institutionalization as resulting from a "quest for organizational immortality" (Pfeffer, 1982, p. 239) a n d as likely to lead to a deemphasis on (or even an a b a n d o n m e n t of) the original goals of the organization a n d the adoption of the goals of survival a n d growth instead. Meyer and Rowan (1977) a n d Zucker (1987) both defined institutionalization as a variable, that is, as something that an organization could have lower or higher levels of, but more significantly, as something which could also vary within an o r g a n i z a t i o n - - s o m e acts a n d structures could be institutionalized while others were based on efficiency considerations. As Pfeffer (1982) summarized: For institutionalized acts or institutionalized structures, there is a stable structure of interaction and definition of the situation which is not necessarily based on rational or instrumental considerations. For acts that are not institutionalized, there are more likely to be incentives or a rational calculus involved in understanding them, and the idea of emergent structure in the context of not yet consensually shared definitions and situational meanings is more relevant and applicable. Thus, the degree of institutionalization determines the extent to which rational versus emotional and nonrational bases of action predominate. (p. 240) The implications of this perspective for the e m p l o y m e n t relationship are plain. Bureaucratic m a n a g e m e n t structures can proliferate even where they are in neither employers' nor workers' interests. From this perspective, one w o u l d argue that all organizations within a specified organizational field should have the same bureaucratic "shape"; that the scope of bureaucratic m a n a g e m e n t forms is limited in extent by the imitative impulse; a n d that the consequences of bureaucratic managem e n t can be extremely negative for the worker (as well as the organization), but that t h e y will persist nonetheless. In some organizations, the imitated structure is efficient; in others, it is not. This leads to the prediction that similar structures m a y have positive consequences in one context a n d negative ones for the same worker in another.

Hybrid Approaches Like a n y classification scheme, the triumvirate of rational, workercontrol, and institutional theories is an imperfect ordering of w h a t various authors think about e m p l o y m e n t relationships. As we have seen, not only does the rational perspective bleed into the institutional a n d control approaches w h e n the concept of environment is recognized, but there are also some theories that adopt elements from each of the underlying perspectives. Two hybrid approaches that we will have occasion to


Chapter 1

refer to in subsequent chapters are an organizational governance model proposed by Selznick (1969) and an "industrial relations systems" model offered by Kochan, Katz, and McKersie (1986). Each, in its o w n way, reflects the concerns of all three basic perspectives. In spirit, Selznick's account of the development of industrial justice mechanisms most strongly resembles the institutional approach of recent organizational theory. That is, in the sociocultural environment of American liberal democracy, large organizations of all kinds (public, private, industrial, and service-providing) adopt quasi-legal mechanisms for ensuring fairness in the treatment of workers across a wide range of issues. What separates Selznick's account (1969) from the institutional approach is the idea that pressure for "procedural justice" may arise almost exclusively from an internal source--the link between bureaucratization and administration through rational rules: Thus the "legality" of bureaucratic authority does not necessarily derive from the public status of the agency or enterprise. It is founded in the internal life and order of the association . . . . The source of these attributes [rational-legal legitimacy and rule-governed decision making] is internal; the dynamic they create calls forth the ideals of legality. (p. 81) Nevertheless, Selznick's approach (1969) also parts company with that of the technical-rationalists. Because of its concern about legality, the bureaucracy necessarily places limits on unbridled rationality: The bureaucrat came upon the scene to enlarge the possibilities of free and rational manipulation of resources, including human resources. Yet his very commitment to legality must fetter managerial decision . . . . Vested rights and policy commitments are expected costs of rational-legal bureaucracy" (p. 94) Thus, in contrast to the transaction cost approach, Selznick's model predicts less within-organization heterogeneity in the adoption of bureaucratic managerial practices. However, in contrast to that of the more recent institutionalists, under his theory one might expect to find more within-industry variation in these matters and, in particular, withinindustry variation that is explained by a concomitant variation in organizational size. Perhaps the most integrated of all perspectives is that offered by Kochan et al. in The Transformation of American Industrial Relations (1986). Theirs is essentially a historicist argument that attempts to explain matters of personnel administration as being the result of a succession of "industrial relations systems." Of particular importance is the "New Deal system," which is currently in the process of being replaced by a



new, more flexible, less unionized alternative. Kochan et al.'s model is based on the idea that managerial practices become institutionalized, but it also recognizes the environment in which institutionalization takes place as a changing one. They see at least two changes as being of fundamental importance since the early 1970s: (1) economic difficulties occasioned by the growth of foreign competition and (2) decreasing tolerance by the business community of labor unions in general. In response, at the level of employment relations, Kochan et al. expect fundamental changes in the face of bureaucratic personnel management where the older system can be displaced: The emphasis on uniformity in the workplace left little room for concern with individual employee differences in motivation and talent. Indeed the specification of collective rules was a reaction to the distrust workers and unions had for the way foremen and other managers had handled their discretionary power. More than anything else, the new nonunion model can be differentiated from the traditional union system by the attention it pays to individual employee concerns and its effort to provide employees with more discretion through the design of broader jobs. (p. 89) Although there is much to applaud in Kochan et al.'s approach, particularly its sensitivity to recent empirical developments, its heuristic value for our purposes is rather limited. Its major shortcoming is exactly the one that plagued an earlier generation of "industrial relations" literature. By default, it seems to assume exactly what we wish to make problematic, that the diffusion of any system--be it "New Deal" or otherwise--is extensive in terms of both the types of employers and the types of workers which are included. To be sure, Kochan et al.'s model is a fruitful source of some alternative hypotheses, particularly with regard to negative findings. Thus, the explanation for a lack of correlation between establishment size and "due process" may lie in the prevalence of alternative "human-resource-management" techniques. Nevertheless, this model does not deal directly with either the scope or the consequence issues at the heart of this book.


Like the larger society, the American workplace has become an arena in which individuals regularly press claims based on relatively abstract notions of "rights" and "opportunities." At work, however, the legitimacy of these claims rests not on principled appeals to a social charter or a constitution, but on the institutionalization of a variety of


Chapter 1

practices by personnel bureaucracies. As a result of the adoption of these practices and of the example that they set, employment relationships have been transformed from webs of personalistic ties into rule-bound transactions that both reflect and frustrate the operation of external markets. At least, this is the picture that would emerge if one distilled a consensus view from the numerous accounts of this process. Despite some recent, and largely historical, work on these developments, there are still unanswered questions. Although we n o w know more about the causes of bureaucratic personnel management (see Baron, Dobbin, & Devereaux-Jennings, 1986; Baron, Devereaux-Jennings, & Dobbin, 1988; Jacoby, 1985), it is still difficult to reconcile competing hypotheses without a further investigation of three key empirical areas. We have labeled these areas shape, scope, and consequences. It is also our belief that a complementary methodology, the analysis of a detailed cross-sectional survey, will go a long w a y toward filling in the gaps in our existing knowledge. Because the data we are analyzing are highly representative of a local labor market, because they offer observations from the vantage point of both workers and employers, and because they have greater depth of information than is usually found in employer-only surveys, they will shed light on many issues that have been hidden from other types of studies. Before we present a brief overview of the contents of each of the following chapters, one point should be stated. First, we have deferred the presentation of specific hypotheses about shape, scope, and consequences to the individual chapters themselves. This choice not only helps to preserve the interplay between theory and data as the empirical story unfolds, but it also prevents confusion between the broader goals of this book and the details of testing specific hypotheses. In Chapter 2, we describe in detail the design and procedures of the study that generated our data on bureaucratic employment practices. Included are discussions of sample size, representativeness, and response rates. Because this study generated a matched sample of workers and employers, special attention is given to evaluating two issues: (1) the comparability of the data obtained from these two divergent sources and (2) the possible selection biases that may have resulted from the fact that the employer data were obtained for a large, but nonexhaustive, subset of those responding to the employee survey. We turn our attention to the "shape" issue in Chapter 3. Each of the theoretical perspectives we have mentioned above contains some implicit hypotheses about the relationships among the due-process, the formalization, and the internal-labor-market mechanisms under scrutiny here. After discussing the implications of these various perspectives, we reveal the results of a



confirmatory-type factor analysis that identifies five separate dimensions of bureaucratic management. Here, our procedures and goals resemble those first used in an article (Baron et al., 1988), and we produce findings that update and extend those obtained in that historical study. Chapters 4 and 5 are devoted to the clarifying issues related to the diffusion and adoption of bureaucratic personnel practices--what we define as the issue of scope. In Chapter 4, scope is considered from an individual point of view. That is, we develop and evaluate hypotheses about the relationship between a variety of individual characteristics and specific bureaucratic personnel practices. For example, there is considerable interest in comparing men and w o m e n with regard to both the quantity and the quality of their participation in internal labor markets (ILMs). Our analysis will make a twofold contribution to this line of inquiry. First, we attack the subject as a problem in "structural attainment." That is, we define ILMs as an objective, exogenously determined attribute of one's position in the workplace and evaluate h o w similar men and w o m e n are in both their chances and their means of entering internal labor markets. Second, because we have information on multiple job positions with each employer, we are able to define ILMs and their consequences on the basis of actual movement as well as structural predispositions. Thus, we can measure and compare the quantity of movement actually experienced by the individual men and w o m e n in this sample. The second scope chapter (Chapter 5) treats the topic from a "structural" point of view. That is, we ask how bureaucratic personnel practices are distributed according to organizational and industrial criteria. As we have already noted, different theories make different predictions about the extent of variation in these practices within and between organizational and industrial units. Closely related to these questions are uncertainties about which characteristics of firms, organizations, and industries predict the adoption of bureaucratic employment relationships. The answers here depend on understanding the role not only of size and of environment (in the general sense, as defined above), but also of job and subunit characteristics. Finally, this chapter addresses h o w demographic and structural distributions interrelate in determining the scope of bureaucratic employment relationships. Our discussion of consequence issues begins in Chapter 6, where we address the normative and attachment consequences of bureaucratic personnel management. Here, we focus both on satisfaction and on the stance that individuals assume toward the external market. The general theme of this analysis is "behavioral" d e p e n d e n c e - - a possible price that may be paid for the comforts of a stable relationship in a protected


Chapter 1

environment. Chapter 7 turns to the relationship between employment practices and individual earnings. A crucial concern in this area is whether there is evidence in favor of a "compensating-differentials" argument. That is, within the same markets, do workers make tradeoffs between higher earnings, better advancement prospects, and greater procedural justice? We realize, of course, that this study, too, is incomplete. It suffers from the limitations that afflict all cross-sectional surveys. It lacks greater depth on some topics than it has on others. But these are limitations that can be properly judged only when they are encountered in the path of our analysis. We begin d o w n that road now with a survey of the methods and materials at hand.

2 Assembling Data on Employment Relations Inherent in the employment relations perspective is the idea that data need to be gathered on structures, arrangements, and processes that are neither purely individual nor purely organizational. The premise of our investigation is that, to uncover underlying patterns in h o w these relationships are organized, one needs to consider not only individuals and their career experiences, but also organizations and their managerial policies. Thus, from the outset of the project, we knew that we wanted to survey both a representative sample of workers and a sample of their employers. It was also apparent that other large-scale structural units (e.g., occupations) were important contexts in which these individualto-firm relationships played themselves out. In this chapter, we describe h o w we gathered data on a sample of workers and their employers in a metropolitan labor market and h o w those data were supplemented with information from other relevant information sources.


As other commentators have recently noted, several strategies may be followed for producing matched samples of workers and employers (Parcel et al., 1991). Our approach was to begin with a probability sample of employed persons, to identify the name and location of their employers, and to conduct a second survey of the employers caught in our initial dragnet. To our knowledge, then and now, this was the first use of the "bottom-up" sampling methodology in a large-scale survey of 29


Chapter 2

workers and work organizations. It has since been adopted by several others (Parcel et al., 1991; Spaeth, 1985; Spaeth et al., 1993). Early in the project, several decisions were made that had longlasting consequences for the success of the research. First, it was decided to contract out the sampling and interviewing phases of the study to the Survey Research Laboratory (SRL) of the University of Illinois. As a wellestablished survey institution, the SRL provided guidance throughout the data-gathering phase of the project. More important, it was willing to adopt what was at the time a rather unorthodox methodology, and it brought its previous experience to bear on finding creative solutions to problems that had rarely been encountered before in survey interviewing. Second, largely for financial reasons, it was decided that all of the employee interviewing would be carried out by telephone and that the same method would be used, as far as possible, in the employer phase of the project. This decision militated heavily in favor of the use of random-digit telephone dialing as the sampling method for choosing individual employee respondents. Although this method does introduce some potential biases into the design (see below, in "Employee Sample"), to choose another technique would have greatly reduced the size of both the employee and the employer samples because of the added expense (Groves & Kahn 1979; Thornberry & Massey 1988). Third, several initial limitations were imposed on the definition of the study population. Because one of the substantive foci of the project was how workers found jobs, we excluded self-employed workers from the study. Another restriction imposed was that respondents had to work at least 20 hours per week in a single job to be eligible for our sample. However, the most important population limitation was that all employee interviewing was limited to the then six-county Chicago Standard Metropolitan Statistical Area (SMSA). (Respondents who lived in the SMSA but who worked outside it were included in the study, however.) The main reason for restricting the geographic coverage of the project was not the expense of long-distance phone interviewing, but our judgment that the second (employer) phase of the project would be much more successful if restricted to the Chicago SMSA. Partly, we guessed--probably correctly--that interviewers identified with the University of Illinois would have better luck in gaining cooperation from area employers than they would from a nationwide sample. Many local businesses recruit management personnel from the university, and it does enjoy a fair measure of prestige among the population at large in the area. A second, and more vital, consideration was our anticipation that the process of finding employers would be more easily carried out in


Assembling Data on Employment Relations

Table 2.1. G e n e r a l E c o n o m i c C h a r a c t e r i s t i c s , C h i c a g o S M S A v e r s u s U n i t e d States A. Price and wage levels A.1. Consumer Price Index, August 1979 All urban areas of U.S. Chi-NW Indiana SCSA (1967 = 100)

221.1 218.1

A.2. Expenditures for family of four at intermediate budget level, A u t u m n 1979 Dollars Index All urban areas, U.S. $20,517 100 All metropolitan areas, U.S. $20,935 102 Chi-NW Indiana SCSA $20,564 100 A.3. Pay indices for selected occupations, Chicago SMSA (December 1978) Category Total Manufacturing Office clerical 105 101 Electronic data processing 102 98 Skilled maintenance 108 105 Unskilled plant 121 101 Base = 100 = Pay level in all 262 SMSAs A.4. Hours and earnings for production workers in manufacturing, Chicago SMSA, 1978 Chicago SMSA Total U.S. Average weekly hours 39.5 39.9 Average weekly earnings $319.16 $311.20 Average hourly earnings $ 8.08 $ 7.80 B. Other characteristics B.1. Annual growth rate in employment, 1970-1980 (in thousands) a Chicago Total U.S. Annual employment, 1970 3,015 70,664 Annual employment, 1980 3,261 90,657 Average growth rate (%) 0.7 2.5 B.2. Percentage of workers involved in work stoppages, 1980 Percent involved Chicago SMSA 1.5 Total U.S. 1.5 aFigures for the Chicago SMSA exclude agricultural workers and are from January of each year. Figures for United States include agricultural workers and are annual averages. Sources: U.S. Department of Labor, Bureau of Labor Statistics (1981). Handbookof Labor Statistics. (Washington, DC: U.S. Government Printing Office); U.S. Department of Labor, Bureau of Labor Statistics (1981). Employmentand Earnings, March 1981. (Washington, DC: U.S. Government Printing Office); U.S. Department of Labor, Bureau of Labor Statistics (1981). Employment and Earnings, June 1981. (Washington, DC: U.S. Government Printing Office); U.S. Department of Labor, Bureau of Labor Statistics (1981). Employment and Earnings, June 1971. (Washington, DC: U.S. Government Printing Office).


Chapter 2

Table 2.2. L a b o r Force C h a r a c t e r i s t i c s , C h i c a g o S M S A v e r s u s U n i t e d S t a t e s A. Sex by race by labor-force status, Chicago SMSA vs. U.S.

Employed (%) Male White Black Spanish Female White Black Spanish

Total (%)

Number of cases

Adult population (%)

76.5 55.7 74.9

4.4 11.0 8.7

19.1 33.3 16.3

100 100 100

1,937,578 430, 306 193,760

36.4 8.1 3.6

51.5 44.4 46.8

2.4 6.9 6.5

46.0 48.7 46.8

100 100 100

2,121,424 535,900 174,701

39.8 10.1 3.3

Number of cases

Adult population (%)

Employed (%) Male White Black Spanish Female White Black Spanish

Chicago SMSA Out of Unemployed labor force (%) (%)

United States (numbers in thousands) Out of Unemployed labor force Total (%) (%) (%)

70.3 57.6 71.0

4.6 8.1 6.6

25.1 34.4 22.4

100 100 100

66, 411 8,122 4,631

38.8 4.7 2.7

48.0 47.1 44.4

2.7 6.0 4.7

49.3 46.9 50.8

100 100 100

77,770 9,888 4,878

45.4 5.8 2.8

B. Occupation of employed persons, Chicago vs. U.S.

Executive, administration and managerial Professional specialty Technical, including health Sales Administrative, support/clerical Service Farm, forestry, fishing Precision products and craft Operatives and laborers Total Number

Chicago SMSA (%) 11.8 12.5 3.0 10.4 20.6 11.6 0.5 11.4 18.3 100.0 (3,239)

Urban U.S. (%) 11.2 13.2 3.3 10.6 18.7 13.3 1.1 11.9 16.7 100.0 (73,749)

Total U.S. (%) 10.4 12.3 3.1 10.0 17.3 12.9 2.9 12.9 18.3 100.0 (97,636)

Sources: U.S. Commerce Department, Bureau of the Census (1982). Census of Population and Housing: General Social and Economic Characteristics-Illinois. (Washington, DC: U.S. Government Printing Office); U.S. Commerce Department, Bureau of the Census (1982). Census of Population and Housing: General Social and EconomicCharacteristics-United States Summary (Washington, DC: U.S. Government Printing Office).


Assembling Data on Employment Relations

a geographic area with which we were familiar and that was at least somewhat limited in size. Again, we believe that this judgment was a sound one. To what extent would our results have differed if this research had been done on a nationally, rather then a locally, representative sample? Without redoing the study, one can never be sure of the extent of this sort of potential bias. Nevertheless, there is something to be learned by comparing the Chicago SMSA in 1980 to the nation as whole with regard to its economic, demographic, and industrial characteristics. The relevant figures are presented in Tables 2.1 through 2.3.

Table 2.3. E s t a b l i s h m e n t a n d I n d u s t r y D i s t r i b u t i o n s , Chicago SMSA versus United States Establishment size distribution, Chicago SMSA vs. U.S.

Number of employees 1-4 5-9 10-19 20-49 50-99 100-249 250-500 500-999 1,000 plus Total Number

Chicago SMSA (%)

U.S. (%)

50.6 19.1 13.4 9.6 3.9 2.3 0.7 0.3 0.2 100 (99,941)

54.3 19.8 12.4 8.3 2.9 1.6 0.4 0.2 0.1 100 (4,543,167)

Distribution of employment by industry group in Chicago SMSA and U.S.

Industry group Mining Construction Manufacturing Transportation Trade Finance, insurance, and real estate Services Government Total Number

Chicago SMSA (%)

U.S. (%)

0.2 3.2 25.3 6.2 23.0 6.8

1.2 4.9 22.2 5.6 22.8 5.8

20.2 15.1 100.0 (3,197)

19.9 17.6 100.0 (91,705)

Sources:U.S. Department of Labor, Bureau of Labor Statistics (1981). Employment and Earnings, June 1981 (Washington, DC: U.S. Government Printing Office); U.S. Commerce Department (1982); County Business Patterns 1980. Volume CPB 80-15 (Washington, DC: U.S. Government Printing Office).


Chapter 2

In its general economic standing, the Chicago SMSA appeared little different from the rest of the United States when our study was conducted. Perhaps the largest discrepancies, which appear in Table 2.1, were in the wage levels of blue-collar workers, especially in the nonmanufacturing sector. While manufacturing wages appeared to be slightly above average, "unskilled plant" workers in the general population enjoyed levels of compensation that were about 20% above average. A second, and probably not related, difference between the Chicago area and the United States generally was the much slower rate of labor force growth in the area during the 1970s. Although it is difficult to be definitive about such things, it is possible that this slower growth may have been associated with some weakness or slack in external labor markets. Whether this was so, and whether it would have affected the prevalence of internal labor markets in the metropolitan area are questions that need further attention. This slower growth did not seem to be associated with any greater docility in the Chicago labor force, as the proportion of workers involved in strikes and work stoppages was nearly identical to the national average. Table 2.2 compares the Chicago SMSA and the United States with regard to the composition of the adult population, the labor force status of various groups, and the occupational distribution of the employed population. In the last column of Table 2.2A, the data shown represent the percentage distribution of the adult population across sex, race, and Spanish-origin categories. (Note that these percentages do not add to 100% since Hispanic adults can be of any race.) Compared to the U.S. adult population, the Chicago area had relatively more minority group members of both sexes and fewer whites. Although it did not mirror the general population exactly, it is probably better to err on the side of studying an area that overrepresents than one that underrepresents minority group members. When looking at labor force status, one notes that whites in Chicago had higher rates of employment than whites elsewhere, while blacks had lower rates. (Hispanics resembled whites more than blacks in this instance.) If these two tendencies are combined, they probably tend to offset one another. That is, the lower employment rate of blacks in Chicago was applied to a relatively larger population base, the result being that the employedpopulation of the Chicago SMSA more closely represented the employed population of the country than that the overall population represented the corresponding national total. The occupation distributions presented in Table 2.2B also reveal a high degree of correspondence between the Chicago SMSA and the rest of the country. There was a slight tendency for Chicago to have relatively more administrative support workers and to have fewer service workers, but these differences are small enough to be ignored.

Assembling Data on EmploymentRelations


Table 2.4. Disposition of All Randomly

Dialed Phone Calls in MEWS Survey Disposition Eligible Interviewed Strong refusal Noncontact, unavailable,etc. Ineligible Screened Other Eligibilityunknown--refusal Nonresidential Disconnected or nonworking Noncontact Other Total Number

Percentage 31.1 3.7 1.5 21.6 .1 6.7 10.3 20.3 3.7 1.1 100.1 (8,717)

Looking at industrial and establishment characteristics shown in Table 2.3, we see that the Chicago area employed slightly fewer workers in mining, construction, and government and slightly more in manufacturing, finance, and services. These differences are so small, however, that it seems safe to accept Chicago as being quite industrially representative. The same similarity is also found with regard to the employment size distribution of the Chicago SMSA. Here, it is true that Chicago was somewhat unrepresentative insofar as it had relatively fewer small establishments and relatively more large ones, but the overall percentages are quite similar. We now turn to a consideration of our specific sampling procedures and how well the sample that they produced represents this particular metropolitan area. EMPLOYEE SAMPLE

As we have noted, we chose to draw a sample of employed persons by using random-digit dialing. All in all, 487 exchanges were selected from the "312" (458 exchanges) and "815" (29 exchanges) area codes (exchanges outside the Chicago SMSA found in the "815" area code were not used in the survey). Based on our pretest experience and existing SRL protocols, random combinations of four-digit numbers were sampled within these exchanges. Ultimately some 8,717 calls were made in April, May, and June 1981. Table 2.4 shows a breakdown of the disposition of these calls.

Chapter 2


Table 2.5. Response Rates of Screened

Eligible Respondents in MEWS Survey Disposition


Interviewed Refused Unavailable Total

85.7 10.1 4.1 (3,164)

Several levels of distinction need to be made in examining this table. First, one needs to eliminate from consideration those calls that would not have resulted in an interview with an eligible respondent under any circumstances. This category includes nonresidential phones as well as disconnected or nonworking numbers. Noncontacts and "Eligibility Unknown--Refusal" are a somewhat different matter. The former includes those instances in which phones rang but were never answered (SRL protocols required seven callbacks in this situation), and the latter includes those instances in which the person answering the phone hung up before the interviewer could complete his or her introduction. Both of these types of outcomes may have been associated with households in which an eligible respondent was present but was unreachable, although there is no reason to think that all of them were. Moving beyond these categories, we encounter calls that resulted in respondents being classified as either "Eligible" or "Ineligible." The vast majority of both groups were respondents with w h o m the interviewers were able to complete the screening portion of the survey instrument. Other ineligibles consisted mostly of respondents w h o answered the phone and did not speak either English or Spanish (interviews were conducted in each language). Table 2.5 presents statistics on the refusal rate among the screened eligible population. By the standard criterion used in survey research, we calculate the response rate to be 85.7% of the screened eligible population. This rate compares quite favorably with the response rates of 67% and 52% achieved, respectively, by Parcel et al. (1991) and Spaeth (1985) w h e n they used a similar methodology, although with nonprofessional interviewers. To further interpret the statistics just presented, a brief examination of the "screening page" of the employee questionnaire is in order (Appendix 1). First, it should be noted that the within-household sampling methodology relied on a quasi-quota procedure and was not based on a complete household enumeration of all eligible workers in the household. In previous studies, the SRL had found that its success with this

Assembling Data on Employment Relations


Table 2.6. Effect of N o n r a n d o m Respondent Selection

within H o u s e h o l d s

Percentage male Mean age (years) Yearly earnings

Actual sample

Hypothetical sample (with random respondent selection within households)

58.4 38.0 $19,800

58.6 38.2 $19,800

p r o c e d u r e , a s k i n g f o r h a r d e r - t o - r e a c h r e s p o n d e n t s first, o f t e n p r o d u c e d r e s u l t s t h a t w e r e c o m p a r a b l e to h o u s e h o l d e n u m e r a t i o n , a n d t h a t c o u l d be produced at a lower cost per completed interview. During the course of the interview, the respondents were asked whether anyone else c o n t r i b u t e d to t h e f a m i l y i n c o m e . A l t h o u g h t h e a n s w e r s w e r e n o t a p e r fect i n d i c a t o r of w h o e l s e w a s e l i g i b l e for t h e s u r v e y i n t h e s e h o u s e h o l d s ( s o m e o t h e r s c o n t r i b u t e d to i n c o m e b y o w n i n g s t o c k , b y selfe m p l o y m e n t , o r b y w o r k i n g for f e w e r t h a n 20 h o u r s p e r w e e k ) , it is a f a i r l y r e a s o n a b l e a p p r o x i m a t i o n . Table 2.6 p r o v i d e s s o m e e s t i m a t e s of h o w o u r s a m p l e w o u l d h a v e c h a n g e d if, i n s t e a d of u s i n g t h e s a m p l i n g procedure outlined above, we had chosen randomly from among those l i s t e d a s i n c o m e c o n t r i b u t o r s i n t h e h o u s e h o l d . ~ W i t h i n t h e l i m i t s of t h e questions asked about other household members, this table shows that the quota procedure used in the MEWS sample did indeed produce a s a m p l e t h a t w a s q u i t e c l o s e to t h e o n e t h a t w o u l d h a v e b e e n o b t a i n e d if 1 Although we did not enumerate all wage earners per se in each household before selecting a respondent, several questions were asked that did allow us to estimate how the sample would have changed if a random selection had been made among various income contributors in the household. In the sequence of questions devoted to ascertaining individual and household income, respondents were asked how many persons besides themselves contributed to household income. The breakdown of responses was as follows: no other contributors, 1,450; one other contributor, 1,080; two other contributors, 110; three others, 34; four others, 14; and five or more others, 9. Working just with those respondents who said that one other person contributed (which would have given us access to about 67% of the total possible additional sample elements), we were able to identify the age and sex of the other contributors, who were wives or parents of the respondents who were actually selected. Eventually, we were able to do this for 954 of our original respondents. A new hypothetical sample was constructed by using all the original respondents from the one-earner and the three-or-more-earner households and a randomly chosen person from the two-earner households. Although this sample is admittedly not a perfect representation of what household enumeration would have obtained, it does provide a sense of how robust the original sampling procedures were.

Chapter 2


Table 2.7. C o m p a r a t i v e U n i v a r i a t e Statistics, 1981 M E W S S a m p l e v e r s u s 1980 C e n s u s 1980 Census Age in years Mean Median Persons in household Mean Median Wage and salary income Mean Median Years of school completed Mean Median Marital status (%) Married Widowed Divorced Separated Never married Sex (%) Male Female Place of birth (%) U.S. Outside U.S. Occupation group (%) Executive, administrative, and managerial Professional Technical Sales Administrative support Private household Service Farming, etc. Precision products, craft, and repair Operatives, laborers, etc. Race or Ethnicity (%) White Non-Hispanic Hispanic Black Non-Hispanic Hispanic

1981 MEWS

38.0 36.0

38.0 35.0

3.3 3.0

3.3 3.0

$15,500 $13,400

$19,400 $17,500

12.7 12.0

13.7 13.0

60.0 3.3 8.2 2.5 26.0

58.3 3.8 10.2 3.2 24.7

57.3 42.7

58.4 41.6

88.0 12.0

91.1 8.9

12.7 12.8 3.3 7.6 22.4 0.2 9.7 0.3 12.0 19.0

17.9 16.1 2.9 8.2 19.9 0.1 7.8 0.4 12.5 14.0

79.0 (75.3) (3.7) 15.0 (14.9) (0.1)




Assembling Data on Employment Relations Table 2.7.


(Continued) 1980 Census


Non-Hispanic Hispanic Spanish Other Non-Hispanic Hispanic

2.3 (2.3) (0.0) 3.2 0.5 (0.2) (0.3)

1981 MEWS 1.7 5.2 0.3

complete household enumeration had been used to select a respondent within each household. The sample would have changed only slightly in terms of its age and gender composition. But how well did the sample actually represent the population that we had set out to capture? One w a y of answering this query is to compare our sample to the Public Use Samples of the 1980 census of Population and Housing. Of particular interest for this purpose are the 1% B samples from the census, which have the desirable properties of having a large number of observations, of identifying the variables that are needed to reconstruct our eligible population, and of identifying the Chicago SMSA as a separate geographic entity (U.S. Department of Commerce, Bureau of the Census, 1983). After imposing the sampling-frame limitations by excluding selfemployed workers and those employed fewer than 20 hours per week, one obtains a comparison sample of about 26,704 individuals, about 10 times the number in the MEWS sample itself. Table 2.7 presents a comparison of the univariate distributions of some key variables from the two samples. Age and household size had virtually identical distributions in the two samples. Moreover, the differences in sex, black versus nonblack race, and marital status are small enough to be considered negligible. Spanish origin was measured somewhat differently in the MEWS survey than in the decennial census. In the Chicago data, Spanish was provided to respondents as an option in the closed-end question that elicited the other racial responses. In the census, people of Spanish origin were largely identified through a separate "Spanish-origin" question, so that people of all races could claim Spanish roots. Nevertheless, some 3.2% of the census sample did volunteer Spanish as a response to the race item. The proportion of Spanish in our sample did fall between the value of 3.2% (the self-identified Spanish) and 7.3% (the sum of the relative frequencies of the various Spanish-origin groups within each racial category).


Chapter 2

Table 2.8. Comparative Regression Statistics, 1981 MEWS Sample and 1980 Census of Population 1 in 100 Public Use Sample, Chicago SMSA Dependent variable (income in 1979) a Independent variables

1980 Census

1981 MEWS

Race (Black = 1)

-2,180 (190) 9,110 (132) 1,120 (22.3) 225 (4.9) .317 20,585

-2,330 (533) 9,540 (379) 1,240 (71.0) 239 (15.0) .332 2,483

Gender (Male = 1) Years of education Age R Squared (adjusted) N ~Standard errors in parentheses.

Income, occupation, and years of education were substantially higher in the MEWS sample than in the census. Although the income difference appears to be especially large, this was in good measure due to inflation. When the 1981 MEWS data, which asked about 1980 income, are deflated to 1979 dollars, the reference point for the census income questions, one obtains a mean of $17,093. This is much closer to the census value of $15,500, and about half of the difference can be explained by inflation. Nevertheless, there were persistent differences in the two samples with regard to education and occupation that correspond to the remaining income differences. It seems safe to conclude that the MEWS sample had a slight tendency to overrepresent highersocioeconomic-status respondents. This tendency has been noted in previous literature on telephone interviewing. For example, Thornberry and Massey (1988) estimated that, w h e n one takes into account both the degree of noncoverage of households due to the absence of universal access to phones and the lack of response to telephone surveys in general, the expected response rates to telephone surveys in general range from 56% for those with less than high school education to 88% among those with more than high school education. We estimate that our rates of differential response were of roughly the same magnitude as those estimated by Thornberry and Massey. In large part, the consequences of this bias depend on h o w well the MEWS sample matched the census in terms of its multivariate charac-

Assembling Data on Employment Relations


teristics rather than its univariate characteristics. In other words, has the slightly higher proportion of higher-status persons in our sample thrown off our estimates of the key relationship under study? Table 2.8 compares the census and MEWS data w h e n each are used to predict respondents' incomes in a simple regression model. Because the census did not ask about work experience directly, age is used as a predictor variable in its place for both samples. With the exception of the coefficient for education, the MEWS coefficients are about 5% larger than the corresponding coefficients in the census sample (the education coefficient is about 10% larger). It is worth noting that because the signs, significance, and relative sizes of the coefficients are so similar for the two samples, one would probably reach the same substantive conclusions by examining results from either data set. On the basis of these tests we conclude that the MEWS data are quite representative of the population that they were intended to reflect. Where there are some discrepancies in univariate statistics, as in education, we are reasonably confident that the multivariate results are not affected very much. Having resolved the issue of the adequacy of the employee sample, we turn now to a consideration of the employer sample and its characteristics.


In constructing a set of procedures for locating and interviewing the employers of the individuals in our sample, we had to keep the purpose of these interviews clearly in focus. One of our two main goals was to gather information about the characteristics of each establishment and firm more reliably than could be done by simply asking the workers, a group that may not be terribly well informed about such matters. A second, and perhaps more important, aim was to find out information that pertained specifically to jobs within those organizations. We sought information about hiring, recruiting, and screening techniques; about the chances for mobility for each job; about the kinds of work rules and due-process arrangements that applied to each job; and about the training requirements for each position. To get information of the second variety, it was necessary to talk with someone who had detailed knowledge about the specific position in which each of our respondents worked. Thus, where we had a choice, we opted to have the SRL interviewers speak with people at lower levels of the organization w h o would have the most detailed knowledge about the specific positions in question. Often, this was a staff member in the personnel department or

Chapter 2


Table 2.9. Results of Attempts to Locate Employers

Disposition Respondent refused to identify employer Employer identified Located Not located Total



268 2,444 2,400 44 2,712

9.8 90.1 100.0 98.2 1.8 100.0

a first-line supervisor. In any event, the real unit of data collection for the second phase of this project was the "job within the organization." In those scattered instances where we had more than a single respondent from each establishment, we gathered data separately on each job that appeared in our sample. Before anyone could be interviewed, it was necessary to identify the establishment where each individual respondent was employed. The basic method for doing this was to augment the standard series of occupation and industry questions so that the necessary information would be revealed. Appendix B contains the segment of the MEWS questionnaire in which these responses were sought. Because the employee fieldwork extended over three months, arrangements were made with the SRL to immediately photocopy each of these pages from the completed questionnaires so that other project staff could begin the sometimes arduous task of tracking dow n the right establishments. The sources used were fairly standard ones (the appropriate metropolitan p ho n e directories and published directories of industrial and commercial businesses), but occasionally it was necessary to consult more esoteric documents, such as the registered list of Illinois corporations published by the state attorney general's office. Table 2.9 shows our level of success in eliciting and finding the employers of the 2,712 individual MEWS respondents. Overall, we believe that these results reveal considerable willingness of individuals to supply the names of their employers; only about 10% refused to comply with this request. Likewise, the number of employers wh o ultimately proved to be impossible to find was very modest, amounting to less than 2% of our total individual sample. 2 A brief 2 Again, comparisonswith other studies with similar methodologiesare informative. Parcel et al. (1991)reported a dropoffof 9% between their sample of interviewedemployees and a sample of supervisors who could be located. Assuming that this dropoff included attrition that resulted from employees'refusalto provide adequate informationand from

Assembling Data on Employment Relations


digression will illustrate that not all of the problems e n c o u n t e r e d h e r e resulted from lack of goodwill o n the part of our r e s p o n d e n t s . In the course of interviewing, the following colloquy was r e c o r d e d on one of the questionnaires: Q: A: Q: A: Q: A:

W h a t type of business or i n d u s t r y is this; that is, w h a t p r o d u c t is m a d e or w h a t service is given? It's the construction or, more, the destruction business. Is this mainly, manufacturing, wholesale trade, retail trade, a service, or something else? A b u n c h of us w o r k together gutting old a p a r t m e n t s and buildings. W h a t firm or c o m p a n y in the destruction business do y o u w o r k for? There's no real official name, but w e go u n d e r the "Burrito Brothers."

Given the informal nature of some segments of the u r b a n economy, it is not surprising that published p h o n e directories, or e v e n t e l e p h o n e books, are not a w h o l l y sufficient source for locating e m p l o y e r s or developing a sampling frame. In the next section, we address the question of w h e t h e r the unidentified or unlocated e m p l o y e r s a p p e a r e d to be a rand o m subset of the larger p o p u l a t i o n or instead a p p e a r e d to be biased in some fashion. The job a n d e m p l o y e r list c o r r e s p o n d i n g to the 2,400 individual r e s p o n d e n t s was r e d u c e d to 2,386 with the deletion of 14 m e m b e r s of the military w h o were included in the first wave of sampling. Furthermore, it was d e t e r m i n e d that some 66 individuals w e r e e m p l o y e d in identical jobs for the same employer, further reducing the effective sample size of jobs to 2,320. A n example of this sort of duplication was "letter carrier for the Chicago post office" or "elementary-school teacher for the Chicago public schools." Thus, in the section that follows w e n e e d to distinguish w h e t h e r the completion or refusals that w e discuss are being tallied b y the n u m b e r of jobs that were included or lost, or b y other sources, it can be compared to the 13% dropoff that we experienced at this stage. In a national study using a similar methodology, Spaeth et al. (1993) attempted to identify and interview the employers of respondents (and their spouses) sampled in the 1991 General Social Survey (GSS). Their experience was not dissimilar from Parcel's and her collaborators' or from our experience. Refusals to identify employers arose in about 12% of the GSS cases, and in an additional 5% of cases, information sufficient to locate the employers was not obtained.

Chapter 2



2.10. Disposition of Employer Interviews in MEWS Study


Unavailable from employee survey Interviewed in employer survey Not interviewedin employer survey Refusal Noncontact Unavailable Out of business No new listing Other Total

Number 312 1,926 474

Percent of all c a s e s 11.5 71.0 17.4

351 24 10 9 18 62 2,712

Percent of employer cases

80.2 19.8 14.6 1.0 0.4 0.3 0.7 2.5



the n u mb er of persons included or lost. In Table 2.10 values are shown for the n u mb er of individual employees whose employer or job interviews met with various fates. About 71% of all employees w ho were interviewed in the first survey had employers who were interviewed in the second wave. Among those who had employers w ho were not interviewed, a substantial number (312) were in this category because an individual r e s p o n d e n t failed to provide the name of an employer or, in some instances, the name of one wh o could be found. Employer refusals accounted for another sizable share of missing employer data. However, it is worth pointing out that if the identified employers are taken as the point of comparison, less than 20% of these employers refused to be interviewed. Not surprisingly, a small number of employers had either relocated or gone out of business in the interval between the two surveys. If the completion rates shown in Table 2.4 are combined with those shown in Table 2.10, the combined completion rate for both waves of the survey is approximately 61%. Although this percentage is not as large as one might wish for, it is still within an acceptable range and provided an adequate base for proceeding. Finally, we consider the issue of what kinds of bias might have been introduced in the employer phase of our interviewing. That is, to what extent were the 71% of respondents whose employers provided data similar to the 29% whose employers did not. Relevant data for this comparison are provided in Tables 2.11 through 2.13. Considering first the patterns that are shown for different individual characteristics in Table 2.11, we see that male and female respondents divulged the names of their employers at about equal rates. However,

Assembling Data on Employment Relations


Table 2.11 Employer Completion Rates by Worker Demographic Characteristics

Sex Female Male Race Nonblack Black Education HS or less Some college College or more Age group Under 25 25-34 35-44 45-54 55 and over Region of birth U.S. or Canada Northern or Western Europe Southern or Central Europe Latin America Other

Percent of respondents not identifying employer

Total N

Percent of identified employers not interviewed

N of employers

11.6 11.4

1,128 1,583

15.8 22.5

997 1,402

12.1 7.8

2,313 386

19.9 18.5

2,033 356

10.4 10.5 13.6

1,126 714 865

17.9 20.7 21.3

1,009 639 747

6.4 11.6 15.4 16.7 15.5

910 979 423 228 84

19.4 20.0 21.2 18.9 18.3

852 865 358 190 71

10.8 17.2

2,135 128

19.0 23.6

1,905 106





12.9 14.3

93 140

12.3 26.7

81 120

the e m p l o y e r s of female r e s p o n d e n t s were m o r e likely to be contacted a n d interviewed t h a n the e m p l o y e r s of males. In large part, this difference is explained b y s o m e other characteristics of male a n d female employees, such as the industries in w h i c h they w o r k e d , as w e shall see shortly. The results for race, education, a n d age are s o m e w h a t u n e x p e c t e d insofar as individual a n d e m p l o y e r nonresponses t e n d e d to be associated with h i g h e r status r e s p o n d e n t s rather t h a n lower-status ones. That is, y o u n g e r , black, a n d less-educated r e s p o n d e n t s w e r e m o r e likely t h a n older, nonblack, a n d better-educated people in the sample to identify a locatable employer. This same pattern also holds true with r e g a r d to e d u c a t i o n a n d e m p l o y e r refusals. S o m e w h a t ironically, the biases that

Chapter 2


Table 2.12 Employer Completion Rates by Worker Labor Market Characteristics

Percent of respondents not identifying employer Number of previous jobs None 2-3 3-4 5 or more Years of labor force experience 0-5 6-10 11 or more Occupation group Managerial, professional, technical Sales/clerical Skilled blue collar Other blue collar Farm

Total N

Percent of identified employers not interviewed

N of employers

7.8 10.2 11.3 13.9

408 489 942 839

19.9 21.0 18.5 20.2

376 439 836 722

7.2 12.4 16.6

1,185 635 821

19.8 18.0 21.0

1,100 556 685





11.5 10.9 9.8 8.3

758 339 590 12

19.2 21.5 18.2 9.1

671 302 532 11

w e n o t e d for individual-level r e s p o n s e s to the first w a v e of interviewing w e r e r e v e r s e d , at least partially, b y the p a t t e r n s of e m p l o y e r c o m pletion. While there are s o m e systematic differences according to res p o n d e n t s ' region of origin, t h e y do not f o r m a readily interpretable pattern. In Table 2.12, similar data on e m p l o y e r r e s p o n s e s are b r o k e n d o w n b y the labor m a r k e t characteristics of the individual r e s p o n d e n t . H e r e , there are s o m e s t r a i g h t f o r w a r d p a t t e r n s of individuals' willingness a n d ability to identify e m p l o y e r s , but there are no real differences in rates of e m p l o y e r s ' refusals. The regularities that do a p p e a r in the first c o l u m n are quite consistent w i t h those s h o w n in the p r e v i o u s table. T h a t is, r e s p o n d e n t s in higher-status o c c u p a t i o n s w e r e less likely to identify their e m p l o y e r s , as w e r e those w i t h m o r e years of labor force e x p e r i e n c e a n d t h o s e w h o h a d h a d a greater n u m b e r of p r e v i o u s e m p l o y e r s . O f course, w i t h o u t u n d e r t a k i n g a multivariate analysis, it is difficult to k n o w w h i c h of the effects s h o w n here, a n d in the p r e v i o u s table, are i n d e p e n d e n t of o n e another. Finally, w e e x a m i n e in Table 2.13 h o w r e s p o n d e n t s ' industrial a n d e s t a b l i s h m e n t affiliations w e r e associated w i t h the level of e m p l o y e r r e s p o n s e . P e r h a p s the m o s t dramatic result in a n y of these last three

Assembling Data on Employment Relations


Table 2.13 Employer Completion Rates by Worker Industry

and Establishment Characteristics

Industry group Other Construction Manufacturing, utility, commercial Trade Finance, service, public administration Government worker No Yes Establishment size 10 and under 11-50 51-250 251-500 Over 500

Percent of respondents not identifying employer

Total N

Percent of identified employers not interviewed

7.7 22.6 9.2

13 106 1,016

0.0 32.9 18.9

12 82 923

12.1 11.0

462 1,087

25.6 17.4

406 967

13.4 0.0

2,326 386

20.2 17.4

2,014 386

14.0 13.4 10.3 12.0 9.4

243 612 717 299 795

25.8 18.9 18.5 14.4 20.4

209 530 643 263 720

N of employers

tables is the h i g h rate of nondisclosure of e m p l o y e r s b y construction e m p l o y e e s a n d the equally high rate of refusal or n o n c o n t a c t a m o n g construction employers. 3 In part, this is a matter of the social organization of construction w o r k itself, and, in part, it m a y reflect s o m e cultural features of the industry. O n e aspect of construction w o r k with w h i c h o u r m e t h o d s were n o t well aligned is the fact that the e m p l o y e r ' s place of business a n d the actual place of e m p l o y m e n t are frequently separated b y large distances. It is not u n c o m m o n in large cities for the general contractor o n a large w o r k site to be situated in a different m e t r o p o l i t a n region. At the same time, the place of w o r k m a y meet m a n y of the usual criteria of being an o n g o i n g w o r k establishment. It exists in the same place for m o n t h s , or often years, at a time, a n d there is some decision m a k i n g r e g a r d i n g hiring a n d supervision at the location itself. A d d e d to these complexities are the further ambiguities associated with the exis3 In the study by Spaeth et al. (1993) mentioned in Footnote 2, there was also a tendency for workers in the "mining and construction" category to furnish fewer than expected completed employer interviews. While their overall completion rate for all employees was 50.8%, it was 42.7% in this category. (Figures recalculated from Spaeth et al., 1993, Table 4, p. 17.)


Chapter 2

tence of subcontracting relationships. When one asks what firm someone works for, and where she or he works, how the question should be answered is not completely obvious w h e n the employee in question has been sent by the union hiring hall to work for a subcontractor on a larger project u n d er the auspices of a completely different general contractor. Not only do these characteristics of construction work make it difficult for individual respondents to name their "employer," but they also make it hard for survey interviewers to locate and interview appropriate "establishment-level" respondents. Aside from these problems, the other noticeable industrial artifact is that employers of workers in the retail and wholesale trade were somewhat less compliant in furnishing interviews than other employers. This reduced compliance may be attributed not only to the small scale of some retail operations and their more transient existence, but also to the fact that many retail sites lack the luxury of a private enough place and a long en o u g h time for a proper interview to be conducted. 4 While Table 2.13 seems to indicate a high rate of disclosure among government employees, that particular piece of information is a complete artifact of the way in which we measured being a government employee: Our primary method was to rely on the responses to the question that asked for the name of the employer. Nevertheless, there were few differences in the rates of refusals among government and private-sector employers in the second wave of our interviewing. Finally, we may consider how employer size, as measured by the size of the workplace or establishment, affected our success. There was a consistent tendency for the employees of smaller places to be more reluctant to divulge the names of their employers. Perhaps these workers suspected that they would be more likely to "get into trouble" by discussing their jobs than those workers w h o felt more a n o n y m o u s in their larger, and perhaps less personal, work environments. O n the employer side, although it was more difficult to interview managers and owners in the smallest establishments, there were not m any differences among those establishments with over 10 employees. On balance, we believe that these results establish the general representativeness of the sample we gathered, particularly w h e n it is remembered that our goal was to generalize to a population of employed 4 Data from Table2.13 can be recalculatedto show the percentage distribution of employees with completed employer data across these industry categories. The results are as follows: other, .6%; construction, 2.8%; manufacturing, etc., 39.0%; trade, 15.8%; and finance, service,and public administration, 41.7%. With the exceptionof the shortfall in trade and a surplus in manufacturing, these figures compare very favorablywith the SMSA percentages shown in Table2.3.

Assembling Data on Employment Relations


individuals. In dealing with the attrition that did occur between the individual data-gathering phase and the completion of the employer interviews, there are at least two possible strategies of action. One is to use some formal estimation techniques that have been specifically developed for the problem of "sample selection bias" (Berk, 1983). The second is to combine the employee and employer surveys at the level of individual indicators, so that employer data are used when they are available, but w h e n they are not, employee data are used as supplementary indicators so that the larger number of cases can be preserved. This procedure can be illustrated for the measures of establishment and firm size that are used in the subsequent analysis. To understand these procedures, it is also important to know that we acquired additional information on the sampled work establishments through a commercial credit-rating vendor, Dun and Bradstreet Inc. Although this source of data proved to be rather incomplete (of the 2, 712 workers in our sample, only 62% were employed by establishments for which the commercial vendor had listings), it was, nevertheless, a useful supplement to the data obtained in our survey. Our final measure of establishment size was derived somewhat differently, depending on which sources of information were available for each case. The basic principle that underlies these decision paths is that where private-sector employer survey data were available, we regarded them as the best information available on establishment size. If, in the private sector, these employer responses were not available, then a regression formula was used to predict what the employer data would have been had the information been collected. The predictors in the regression equation are the establishment-size figures provided by the workers, and by Dun and Bradstreet if both were available. (The coefficients in the regression equations were estimated from that subset of cases where both employer and supplementary data were available.) For the public sector, employer data were more likely to be missing, so employeereported values were regarded as the best source of information. If they were also missing, then the Dun and Bradstreet information was used in the regression approach just described. Although the details are somewhat different, a similar kind of procedure was used to estimate the total company size of each firm in the sample. Although employee estimates could not be used in this instance, because the employees were not asked to provide this information, combinations of Dun and Bradstreet and employer-survey information were used to supplement one another. The more general point here is that decisions about missing data and resulting sample-selection problems are often best made on a case-by-case--or to be literal, a measure-


Chapter 2

by-measure--basis than by an abstract statistical algorithm. Thus, in the case of establishment size, at least one study (Parcel et al., 1991) showed that management-provided estimates of size are often superior to those provided by workers. Our procedures allowed us to incorporate this finding and to make full use of our complete sample by supplementing employer-provided information where it was necessary to do so. (A more general discussion of h o w our data gathered from employees compared to our parallel data gathered from employers appears in Appendix C).


Although firms and establishments are a very important context in which individuals live out their working lives, it is important to consider "how" one makes his or her living as well as "where" one does so. The convention of incorporating this kind of information is the concept of occupation. Although ordinary discourse tends to blur their meanings, the terms job and occupation are carefully distinguished in most empirical studies. The crux of the distinction is that occupation is taken to refer to those aspects of a work role that are recognized as common or equivalent across different employment settings. In contrast, jobs are regarded as particular, that is, what tasks one carries out at his or her place of work. Although both jobs and occupations were at issue in this study, our theoretical interests dictated that we investigate some aspects of the occupational context of work as well as its job context. Specifically, in Chapter 4, we consider h o w the pattern of withinoccupation mobility across firms affects employment relationships. To accomplish this task, we relied on data about occupational movement that were taken from a very different data source: the Current Population Survey (CPS, U.S. Department of Commerce, Bureau of the Census, 1985). Unlike the MEWS sample, the CPS is a nationwide, monthly survey of households and employed persons in the United States conducted by the Census Bureau for the Department of Labor. In January 1983, as part of its ongoing effort to monitor job training and job mobility, the bureau fielded a supplement to the CPS that asked a series of questions relating to individuals' recent occupational and employer mobility. The answers to these questions provided the basis for identifying occupations with more and less viable mobility prospects through a process that is described in detail in Appendix E. We believe it is safe to conclude that the information available to us about this metropolitan labor market is unique in its scope and depth.

Assembling Data on EmploymentRelations


Although the techniques that were used were somewhat controversial at the outset, they have become widely accepted as a standard tool of e m p lo y men t relations research. In the next chapter, we redirect attention to the theoretical issues raised in the first chapter, but we now investigate them using the data from the sources we have just described.

3 The Elements of Bureaucratic Personnel Management James Q. Wilson (1989, pp. 139-142) relates the tale of the rise and fall of PACE (the Professional and Administrative Career Examination) in the federal bureaucracy. This examination, developed over a five-year period by 24 research psychologists, was designed to identify those with the ability to become managers, so that civil service job ladders could stretch from the bottom of the hierarchy to near the top, and so mobility chains would be consistent and predictable. Federal agencies could either promote to managerial positions directly from within the agency or else choose new managers from the PACE list. In either case, the mobility chain was clear and consistent. In 1979, following a lawsuit brought on behalf of black and Hispanic workers (who tended to have subqualifying scores on PACE more often than white workers), the Carter administration entered into a consent decree that abolished PACE. After that, movement into the managerial ranks happened in several different ways. A few occupations had n e w tests developed, and mobility continued in the old way. But in most cases, either mobility was internal and based on agency head preference without test data, or it was a "Schedule B" appointment from outside, with no particular standards applied. About 10 years later, new clusters of tests were developed, but a college degree allowed exemption from them. There are now many avenues of movement to the managerial level, and the type of avenue varies considerably. Even in a bureaucracy as rigid as that of the federal government, the degree of protection offered to "protected" jobs varies from occupation to occupation and from agency to agency. And h o w it got to be that 53


Chapter 3

way has far more to do with outside events than with some internal rational dynamic driving the employment relationship. In contrast to the haphazardness that reigns in the everyday world of organizational administration, scholarly accounts of modern personnel management often convey a picture of a system of policies that coherently regulate aspects of organizational functioning as diverse as hiring, promotion, transfers, wage setting, supervision, layoffs, punishments, fringe benefits, work specification, and job performance. This portrayal of bureaucratic personnel management as a comprehensive system of organization or control that appears in contemporary writing is likely to be convincing, until one recognizes that the source of the alleged coherence of these policies lies not so much in real-world practices as in the different theories that underlie each author's views of modern personnel administration. In this chapter, we are concerned with these policy elements and their diversity. As a point of reference, we begin with some of the explanatory schemes that have been proposed to see how they have distinguished among the various provisions found in bureaucratic personnel environments. The larger question of whether these elements combine to form a system, or even an emergent system, depends, of course, on the kind of system one is looking for. Therefore, the second goal of this chapter is to analyze the strength, the pattern, and the nature of relationships among a selected set of bureaucratic elements, with the aim of detecting the empirical outlines of a system of bureaucratic personnel management.


The three views of personnel management that see it as a unified development are the capitalist control model of Edwards, the organizational-governance model of Selznick, and the institutionalist model of Baron et al. A fourth account of bureaucratic personnel policies, Oliver Williamson's transaction cost model, recognizes the significance of the these same organizational practices. However, his theory regards these practices not as elements of a system, but as alternative means of "contractual governance," which arise in response to needs for temporal continuity in the relationship between employers and employees. Not surprisingly, each model incorporates a similar, although not completely overlapping, list of bureaucratic personnel practices. However, each model contains significantly different emphases; different elements are seen as reflecting the core institutions of bureaucratic management, and with different processes and logics are seen as giving rise

Bureaucratic Personnel Management


to its empirical consolidation as a social form. This is to be expected, as the four models follow different traditions, emerging, respectively, from what we called in the first chapter the industrial-sectorial approach (Edwards), the institutional approach (Selznick), the technical-rational approach (Williamson), and a hybrid approach (Baron and, to some extent, Selznick). In the capitalist control model, bureaucratic personnel systems ("bureaucratic control" in Edwards's lexicon) are viewed as the culmination of a series of control forms evolved since the beginning of the industrial revolution to deal with the problem of industrial management. Of critical importance in this model, and the related model of Burawoy, is the fact that industrial management in the West has invariably been situated within a capitalist context marked by structured conflict between workers and managers over the distribution of surplus value and over the nature of the employment relationship under which surplus value is realized. In the 20th century, what is crucial for any method of exercising capitalist power is that it address both the problem of immediate control (getting enough work out at the right cost) and the problem of long-run control (deflecting potential challenges to the capitalist system itself). With the increasing scale and concentration of capitalist enterprises, Marxist theorists see solutions to these problems as necessarily involving strategies of "hierarchicalization," "stratification," and "subdivision": creating more layers, more status divisions, and more little boxes within slightly larger boxes. Whether or not these organizational tactics create efficient systems of production, these theorists see them as reliable means of redirecting workers" attention from large, collective, and threatening issues to smaller, individual, and more manageable ones: problems of being in the right job classification, at the proper salary step, and in the correct evaluation category. And as an added feature, successful control strategies are frequently described as those that render managerial control as an impersonal, abstract force rather than as a willful and deliberate regime imposed and maintained by other people. Modern enterprise strives for a government by rules, not people. Thus, bureaucratic management, with one foot planted on the bedrock of impersonal rules and the other on the steady ground of specialized jurisdictions, becomes the organizational partner of corporate capitalism. Seen in this light, the central institutions of bureaucratic personnel management are classification schemes and rules that order production, mobility, rewards, and penalties according to impersonal criteria. In his account, Edwards attributes primary importance to wage administration schemes, job classifications systems (and their attendant formalized job


Chapter 3

descriptions and job instructions), and routinized methods of performance appraisal and evaluation. Although recognizing the empirical prevalence of due-process mechanisms--grievance procedures and seniority provisions--Marxist theories typically ascribe to them an ancillary, almost incidental role: The positive incentives, the relief from capricious supervision, the right to appeal grievances and bid for better jobs, the additional job security of seniority--all these make the day-to-day worklife of Polaroid's employees more liveable. They function as an elaborate system of bribes, and like all successful bribes, they are attractive. But they are also corrupting. (Edwards, 1979, p. 145) To describe these practices as "bribes" and as "corrupting" is revealing, for a bribe is indelibly tinged with personalism, and corruption is as abstract as a wink and a satchel of unmarked bills. In short, the Marxist view of these bureaucratic elements is that they represent a slightly disreputable side of the modern corporation rather than its public face of rationality, impersonality, and efficiency. Likewise, internal labor markets and systematic mobility, to the extent they are considered in some Marxist treatments of modern personnel management, are viewed as another mechanism of cooptation whose significance lies in their potential to distract workers rather than in their tendency to enhance workers' job security and improve their bargaining positions. (Burawoy, 1979, however, does recognize the potential that internal labor markets and job transfer schemes have for diluting supervisors' full exercise of authority.) A second perspective on personnel management is represented by what we are terming the bureaucratic-governance model of industrial relations. (Although it is equally appropriate to refer to it as the institutionalist model, doing so risks confusing this perspective with the next one we will discuss.) This viewpoint is expressed perhaps most fully in Philip Selznick's Law, Society, and Industrial Justice (1969). Here, the dynamic force that gives rise to a different style of personnel administration in contemporary organizations is the growth of bureaucracy as a means of administration. To simplify the argument somewhat, this model adopts the Weberian position that increases in the scope and scale of organizations of all types are accompanied by a steady trend toward the adoption of a bureaucratic form of administration. This is true primarily for reasons of technical efficiency, but also as a reflection of the "rationalization of all spheres of social life." More important to our purposes is what this model has to say about the effects of bureaucratic organization on employment relations in

Bureaucratic Personnel Management


m o d e r n enterprises. If "control" is the central theme of Marxist approaches to this topic, "governance" is the d o m i n a n t motif of the bureaucratic model. W h a t this term means is that there is an inevitable tendency for bureaucracies to establish themselves internally as private governments, with accompanying notions of "legality, .... citizenship," and, in the ideal case, "pluralistic participation.'1 True to its roots in the growth of bureaucracy, industrial governance is marked by its propensity to manage and dominate t h r o u g h the mediu m of formal a n d written rules. In contrast to the Edwards model, which emphasizes the conflict-reducing a n d conflict-dispelling functions of impersonal rules, the governance model emphasizes the limitations that written policies place on the unfettered exercise of managerial discretion. To the extent that such limits are real, they give substance to the construction of the employee's relationship with his or her employer as one based on a citizenship status as well as one of being a party to an atomistic contract. Like the Marxist model, the governance model explicitly recognizes that capitalist e m p l o y m e n t relationships are asymmetrical a n d subordinating. Precisely because e m p l o y m e n t relations are e m b e d d e d in a context of worker subordination, the portrayal of bureaucratic m a n a g e m e n t as a "quasi-legal" governance system highlights its tendencies to act as a countervailing force to unlimited managerial autonomy. The basis of these limitations is the worker's "membership right in a going concern" (Selznick, 1969, p. 67). By emphasizing that bureaucratic m a n a g e m e n t of necessity is an "authority'-based relationship, this model calls attention to the limits that the exercise of that authority places on the ruler as well as on the ruled (cf. Bendix, 1974). Therefore despite its emphasis on the bureaucratic roots of this form of industrial relations, the governance model puts something other t h a n detailed classification schemes at the center of its analytic universe. W h a t forms the axis of this system are two related pivot points: practices that create membership rights a n d statuses for workers a n d practices that restrict employers' discretion, especially with regard to worker membership and status. Thus, the core elements of bureaucratic personnel systems are formalized seniority rights, protection from arbitrary dismissal, a n d written rules pertaining to work behavior, discipline, a n d 1 Selznick (1969), although clearly ambivalent about the extent to which industrial justice requires industrial democracy,did assert that employee rights must be positive as well as negative: "In the special-purpose organization, the chances that a viable group life will emerge, sustained by 'natural' interests, are much more limited than in the more comprehensive community. . . . Thus pluralism has a limited worth as a guarantor of even minimal freedom in the special-purposeorganization. . . . This does not mean that internal pluralism can be entirely discounted" (p. 41).


Chapter 3

employee grievances. Internal-labor-market practices are of lesser importance in this model and matter largely insofar as they reinforce the idea that an enterprise's workers enjoy special membership rights not shared with outsiders. In other words, the governance model suggests that highly regimented job ladders are less essential than mechanisms that ensure preferential access to positions for insiders. Although structured mobility and pseudocareers are not necessarily seen as foreign by this approach, they are viewed as contributions to the overall system only to the extent that they further solidify the position of "workers-asmembers." Yet another view of bureaucratic personnel management is offered by the "neoinstitutionalists." This group adheres to the belief that many aspects of organizational structure arise less as solutions to purely technical problems and more as responses to a normative environment that prescribes certain practices and structures as "standard operating procedures." Although recognizing the press of many concrete historical circumstances, this theory locates the operant motivations in organizational leaders" tendency to behave ritually, ceremonially, and imitatively (DiMaggio & Powell, 1983; Meyer & Rowan, 1977). Baron, Dobbin et al. (1986) directly applied this model to the problem of the diffusion of bureaucratic personnel systems from the early 1940s onward. In solving problems of organizational administration and labor force control, the neoinstitutionalists stress that solutions are not chosen haphazardly but represent the influence of two groups that exert what may, for lack of a better term, be described as cultural authority. These groups are the federal government, which during World War II also had the legal authority to mandate specific employment outcomes, and the personnel administration profession, which had been an emergent force in American industry since the beginning of the century (see Jacoby, 1985). For our purposes, the institutionalists see the core of personnel administration as consisting not of one set of practices or another (control or governance), but of the institutional group themselves (i.e., the personnel department). In fact, Baron et al.'s study seems virtually indifferent to the choice of any one of a number of personnel practices as indicators in building its impressive empirical case: Centralized hiring, time-andmotion studies, service emblems, seniority provisions, and job evaluation are only many manifestations of a common underlying force--the adoption of the "system." In contrast to these three systems models of bureaucratic personnel practices, the transaction cost approach (Williamson, 1975, 1985) emphasizes the instrumental quality of the adoption of these mechanisms. Furthermore, the instrumentalities are seen as operating at a very imme-

Bureaucratic Personnel Management


diate level: protecting transaction-specific investments in h u m a n capital in particular jobs. Compared to the Marxist approach outlined above, failures to achieve appropriate regulation of employment relationships result not in the resurgence of a smoldering class struggle, but in declines of economic efficiency and squandered resources. This avowedly "microanalytic" approach also leaves room for considerable within-firm and within-establishment variability in personnel practice. In this model, the "word-processing" department of a steel mill is likely to confront a very different set of labor-market transaction problems than the "continuous-casting" department is, with the predicted result being that each unit will develop a unique package of institutional practices to deal with its unique problems. Although the contingent character of this approach makes any specification of relationships among parts of the personnel system a risky endeavor, several tentative observations can be suggested. First, the emphasis placed on relational continuity under the condition of a firm's specific investment in h u m a n capital means that internal-labor-market mechanisms play a central role in these situations because of their mutually obligational quality. Firms and workers both consent to enter a relationship in which organizational rules reinforce their joint condition of technically based dependence. Second, the model also stipulates that the ease of measuring and preplanning job performance also makes a difference in how employment relationships work themselves out. Even in the case where skills are firm-specific, if work performances are easily monitored at the individual level, employers will have less need of costly internal labor markets--the primary advantage of which is their capacity for building normative commitments--and can instead rely on detailed bureaucratic control supplemented by minimal due-process guarantees. What this reasoning suggests is the likelihood of a negative statistical relationship between bureaucratic work-regulation mechanisms and internal-labor-market mechanisms--especially mobility guarantees. In the next section, we examine our survey data with the aim of describing the overall distribution of these personnel practices among jobs and employers, and of deciphering, to a limited extent, the associations among the various dimensions of bureaucratic work administration. Of course, the issues raised in the preceding paragraph go considerably beyond these distributional questions, and we will defer the investigation of causes and consequences to subsequent chapters. There are two parts to the following analysis, First, we examine the weighted and unweighted frequencies of occurrence of various practices. That examination is followed by a two-level confirmatory analysis that evalu-


Chapter 3

ates the presence of various dimensions underlying our observed measures, and that, in turn, investigates whether and in what way these subdimensions can be accounted for by a single common latent factor that accounts for the association among them. Examination of this second-order structure should clarify several of the questions we have posed. The three "systems" models of bureaucratic personnel practice all suggest that a single, second-order factor should exist. What differentiates them is the strength of the relationship between this single higher-order factor and the various first-order factors. Thus, the Marxist model implies that the central institution of bureaucratic control is the structure of formalized controls over work activities. The institutional governance model, in contrast, stresses that due-process rights are crucial. Although we emphasize that these findings are not absolutely decisive with regard to these theories, the results begin to suggest which of them are more and less plausible.

EMPIRICAL RESULTS The Scope of Bureaucratic Personnel Management Information about various bureaucratic employment practices was collected in both the employer and the employee surveys described in Chapter 2. Appendix D provides the necessary details concerning the translation of specific questionnaire responses into the variables reported in Tables 3.1 and 3.2. In some instances, the recording of questionnaire answers produced variables having more than two simple " y e s " "no" response categories (see Appendix D). When this occurred, Table 3.1 shows percentages based on unambiguous "yes" responses. However, where more detailed statistical procedures were employed, as in Table 3.2, the full range of response categories are preserved. Table 3.1 displays the relative frequencies of a wide variety of bureaucratic personnel practices. Although the data we are examining do not provide estimates of two widespread features of these "systems" (the presence of wage classification plans and of seniority-based layoff mechanisms), they do provide reasonable coverage of the basic dimensions elaborated in the preceding discussion. The categories used to group these items were derived from the various theories discussed above. There are, of course, alternative sets of categories that might have been used to organize these items. In particular, two dimensions that were considered but rejected are (1) practices that create continuity in employer-employee relationships and (2) practices dependent on

Bureaucratic Personnel Management


T a b l e 3.1. F r e q u e n c y of Various B u r e a u c r a t i c P e r s o n n e l Practices i n C h i c a g o

Percent of workers Practices r e l a t e d . . . mostly to due process: 1. Employer must give reason for firing. 2. Reason for firing must be written. 3. Employer must give advance notice or severance pay. 4. Employee has grievance procedure. both to due process and bureaucratic work control: 5. All or most rules about employee conduct are written, a 6. Employees are given company handbooks, a 7. Employer has set penalties for rule violations, a mostly to bureaucratic work control: 8. Employee is given written performance evaluation. ~ 9. Employee is given written job description, a 10. Employee is given written job instructions. ~ to formalized personnel practices: 11. Hiring for job is centralized in personnel office, a 12. Firm has personnel department.a 13. Employer uses formal evaluation procedures in hiring. to exposure to labor market competition (ILM I): 14. Employee has had more than one job in company. 15. Inside hiring for current position (employee-reported). 16. Inside hiring for current position (employer-reported). a

Percent of establishments





















































Chapter 3



Table 3.1.

Percent of workers Practices r e l a t e d . . . to institutionalized mobility opportunities (ILM II): 17. Job is a stepping stone to a higher-level position in organization (employee-reported). 18. Chances of promotion to higher level position are 75% or greater (employer-reported). a 19. Job is part of a promotion ladder (employer-reported). a

Percent of establishments













abased on responses from employer questionnaire.

written rules and procedures. In this scheme, the first group would include Items 1, 3, 4, and 14-19 (the internal-labor-market indicators), and the second would include Items 2, 5, 6, and 8-10. This alternative conceptualization is perhaps a better representation of the transaction cost approach with its emphasis on mechanisms that create relational continuity, but there are several reasons that tip the scales against it. For one, it assumes that internal-labor-market mechanisms and due-process guarantees are realizations of the same latent attribute of organizational policy, an assumption that overlooks other important causes of market sheltering, such as employee-interest-group politics. For another, the categories it provides are even less exhaustive than those presented in Table 3.1. Where, for example, do centralized hiring or "set penalties for rule violations" belong, not to mention the various indicators of formalized personnel practices? Finally, using the methods of confirmatory factor analysis, we have been able to verify statistically the particular scheme presented in Table 3.1, lending this approach the added weight of consistency with empirical observation. When we turn to the items themselves, Table 3.1 indicates that internal labor markets, mechanisms often regarded as central to the establishment of systematic internal governance, are among the least diffused of all aspects of personnel management. Thus, less than a third of the positions in this sample were regularly filled through inside hiring (according to both employer and employee reports), and structured promotion opportunities also seem to have been extended to less than a majority of all workers. Although the percentage of workers who en-

Bureaucratic Personnel Management


joyed either of these two internal-labor-market features would undoubtedly have been somewhat higher, a comparison with the presence of "due-process" features is instructive. On average, due-process rights seem to have been much more prevalent than internal-labor-market mechanisms. If employment-at-willis defined as the ability of employers to sunder employment relationships without having to account for their actions in any way, only about 20% of the workers were vulnerable to this form of day-to-day uncertainty. Although even this proportion may seem unacceptably large to m a n y readers, it is dearly an exceptional form of employment relationship in today's labor market rather than the typical one, which, according to Sanford Jacoby (1985), prevailed under the "drive system" at the beginning of the 20th century. Nevertheless, deeper levels of protection of job rights seem to have faded off rather quickly. About one third of the workers were unprotected from instantaneous dismissal or the application of unwritten rules of performance and conduct, and about half lacked access to grievance procedures and the right to written explanations of involuntary terminations. Thus, jobs in this late-industrial setting contained important variability in both the presence and the extent of their protections of "rights of membership in a going concern," and explaining this variation is an important task of subsequent sections of this book. Comparing the prevalence of due-process rights among all workers and just those in nonsupervisory jobs (Columns I and 2) helps to establish one important generalization about their uneven spread in the economy: Due-process guarantees were not to be explained as a privilege enjoyed only by those higher up in an administrative hierarchy. In fact, on four out of five of these measures (or six of seven if the two mixed indicators are included), nonsupervisory workers were more likely to enjoy this protection than those one or more levels up in their firms. In part, this is a matter of style rather than substance--note that advance warning and severance pay were less likely to be provided to the rank and file--but it also suggests that where due process did exist, it was available to the humble as well as the exalted. Thus, although the contrast is not great, it is interesting that internal-labor-market practices were less prevalent among lower-level workers than among those higher in the chain of command, reversing the pattern observed for the dueprocess items. As the literature makes clear, bureaucratic personnel management has a control face, as well as a "protection" face. Table 3.1 indicates that formalized mechanisms of control are also very widely diffused in the contemporary employment scene. Written job descriptions and perfor-


Chapter 3

mance evaluations were provided to about three fourths of all workers and were used in nearly one half of all establishments. These frequencies are large relative to most other indicators of bureaucratic personnel management and suggest the possibility that these practices had diffused quite widely and had become institutionalized as culturally appropriate practices for managing relationships with employees in a variety of contexts other than large bureaucracies. Not surprisingly, written job instructions were distributed less frequently than job descriptions. Judged on the basis of the number of affected employees, personnel departments were also a staple feature of organizational life for most of the workers. In fact, because of their high frequency of occurrence, it is tempting to jump to the conclusion that the existence of these departments was the underlying cause of, or at least a necessary condition for, the other aspects studied here. However, compared to the other attributes, personnel departments were less frequently found on an establishment basis and appeared in only 27% of the cases. What this indicates is a stronger relationship between establishment size and the existence of personnel departments than was true for due-process rights or bureaucratic work control. Taken by itself, this finding does not disprove the claims of the institutionalists, who argue for the centrality of the personnel professionals, because the consistent use of the other managerial techniques, as opposed to their sporadic use, may depend on the presence of these formal personnel units within organizations. In the next section, we address this possibility through the use of a factoranalytic model of all of these indicators. Finally, we see in Table 3.1 that even where personnel departments did exist, they were not necessarily the dominant force that some accounts would have us believe. In fact, personnel departments did not play a substantial role in either screening or making the final hiring decision for a majority of jobs in this sample. Although the data are not available in this study, one would strongly suspect that the role of personnel departments in terminations is even less salient.

The Shape of Bureaucratic Personnel Management Although there were some interesting differences among these types of personnel practices in their apparent frequency of occurrence, these disparities may have been less meaningful than they seem at first glance. Imagine, for example, that every job that provided a worker with good internal mobility opportunities (Variables 17, 18, and 19 in Table 3.1) also provided significant due-process guarantees, sheltered its incum-

Bureaucratic PersonnelManagement


bents from external-labor-market competition, regulated them through bureaucratic control mechanisms, and placed them under the auspices of a well-established personnel department. Furthermore, suppose that this sort of pattern was generalized, so that every job marked by a less common bureaucratic personnel practice was of necessity also characterized by the more common practices. What this would mean is that the five dimensions suggested in Table 3.1 could really be subsumed under a single dimension of bureaucratic control, with the more narrowly dispersed practices simply representing a higher level of adoption than the more widely dispersed ones. Or to put the matter in statistical terms, these five dimensions might be regarded as a (Guttman) scale type of bureaucratic personnel management. Although the various theories of personnel management are not very explicit about this point, this sort of logic seems to underlie at least the Marxist approaches to the topic. In their view, bureaucratic control is indeed a "system," and the empirical prevalence of "bureaucratic work control" compared to "structured mobility chances" could be easily interpreted as managerial attempts to "get by" with the least expensive and least constricting forms of labor force control in situations where those techniques were sufficient to the task. These considerations imply that we need to look more closely not at the frequency of occurrence of these practices, but at the relationships among them. Two main questions can be addressed empirically: (1) To what extent do the five dimensions posited in Table 3.1 capture the variability among these items? (2) Assuming these five dimensions are sufficient, can they themselves be explained as realizations of some underlying single dimension of bureaucratic personnel management (perhaps of the Guttman scale variety)? To explore these questions with our data, we undertook a confirmatory factor analysis of the items in Table 3.1 using the LISREL technique for analyzing covariance structures. This analysis was carried out in two stages. In the first stage, we attempted to reproduce the tentative dimensional structure shown in Table 3.1. 2 Achieving successful results at this stage, we then attempted to estimate a "second-order" model in 2 Unfortunately,even the preliminaryresults could not be achievedwithout eliminating two of the items in Table 3.1. The items dropped were Number 7, " Employerhas set penalties for rule violation,"and Item 13, "Employeruses formalevaluationprocedures in hiring." The latter created problems because of its extremelyweak relationship to the other indicators. The highest zero-order correlationof this item with any other was .176 (with "written performanceevaluation"),and a majorityof these correlationswere less than .10.


Table 3 . 2 . E s t i m a t e d

Chapter 3



Due process Employer must g i v e r e a s o n for firing R e a s o n for f i r i n g must be written Employer must give advance n o t i c e or severance pay Employee has grievance procedure A l l or m o s t r u l e s about employee conduct are written Employees are given company handbooks E m p l o y e e is g i v e n w r i t t e n performance evaluation E m p l o y e e is g i v e n written job description E m p l o y e e is g i v e n w r i t t e n j o b instructions Firm has person . nel dept. H i r i n g for j o b is centralized in p e r s o n n e l office I n s i d e h i r i n g for current position ( e m p l o y e e reported) Employee has had more than one job in company

from Confirmatory

Bureaucratic work control

Market sheltering

Factor Analysis a

Mobility assurance

.275 (10.9)





.491 (14.6) .521 (8.37)














Personnel formalization

.586 (6.49)

.431 (2.20)




.109 (4.45)

.849 (10.0)










.981 (16.6)









(11.2) .









.235 (21.8)








1.00 (13.5)




Bureaucratic Personnel Management



Table 3.2. Due process Inside hiring for current position (employer reported) Job is stepping stone to higher level position in organization (employee reported) Chances of promotion to higher level position are 75% or greater (employer reported) Job is part of a promotion ladder (employer reported)


Bureaucratic work control



Market sheltering

Mobility assurance

1.19 (8.74)




Personnel formalization -


1.23 (8.80)


1.43 (11.5)

aNumber of cases = 1738; T values in parentheses.

which the five latent variables derived in the first stage were themselves taken as indicators of a single underlying second-order latent variable. The key substantive results are presented in Tables 3.2 and 3.3. The factor loadings shown in Table 3.2 offer strong support for the model that groups these practices into five dimensions. As can be seen in the first line of Table 3.3, this model also provides an acceptable fit T a b l e 3.3. S u m m a r y of C o n f i r m a t o r y Factor M o d e l s of B u r e a u c r a t i c P e r s o n n e l I t e m s



Chi sq.


Ajusted goodness of fit

1. 1st-order factors = 5 No 2nd-order factors 2. 1st-order factors = 5 2nd-order factors = 1 3. 1st-order factors = 5 2nd-order factors --- 2














Chapter 3

with the observed data. As expected, the distribution of company handbooks to employees and the implementation of written rules of conduct were each statistically significant indicators of the presence both of "due-process" rights and of "bureaucratic work control." Although it may not be immediately obvious that company handbooks should be related to the protection of employee rights, the courts, in several instances, have held that the policy statements proffered in these documents constitute legally enforceable contract claims. As for the dual character of written rules, this finding is consistent with the "industrial governance" model discussed above. Employer discretion is limited, although hardly eliminated, when rules and regulations are written and, in well-developed cases, codified. What may be more surprising in these results is the presence of two internal-labor-market dimensions rather than one. We have argued elsewhere (Villemez & Bridges, 1988) that there are persuasive theoretical grounds for analytically separating the "protection" (from market competition, that is) and the "mobility" dimensions of internal labor markets. Our findings here reinforce that earlier claim by demonstrating that this duality persists even when a broader range of related practices are included in a statistical model. And as we shall demonstrate in the pages below, these two internal-labor-market facets can also be differentiated in terms of their heterogeneous causes and consequences. In sum, our results allow us to state, in answer to the first question, that the five posited dimensions do indeed organize these variables rather well. In Table 3.4, the estimated correlations among these five latent dimensions are shown. The establishment and use of personnel departments quite clearly has the strongest links to the other dimensions identified here, and the degree of market sheltering has the weakest links. However, our most immediate concern is whether some single unitary "metadimension" might be uncovered that organizes these five latent

Table 3.4. Correlations among Latent Variables Estimated under Model 1

Due Bureaucratic Market Mobility Personnel process workcontrol sheltering assurance formalization Due process Bureaucratic work control Labor-market sheltering Mobilityassurances Personnel formalization

1.00 .53 .25 .16 .70

1.00 .30 .60 .59

1.00 .60 .42

1.00 .62


Bureaucratic Personnel Management


properties. Line 2 of Table 3.3 s h o w s model-fitting results for an alternative s e c o n d - o r d e r factor m o d e l with a single s e c o n d - o r d e r factor. 3 It does not fit the data v e r y well. H o w e v e r , inspection of the correlation matrix in Table 3.4 suggests that there are p e r h a p s two u n d e r l y i n g aspects of these systems: one related to internal labor markets that w o u l d be a cause of "labor-market sheltering" a n d "mobility assurances" a n d a s e c o n d d i m e n s i o n related to " p e r s o n n e l formalization," "bureaucratic w o r k control," a n d " d u e process." Nevertheless, w h e n this structure is i m p o s e d o n the model, there is not m u c h i m p r o v e m e n t . In fact, the m o d e l p r e s e n t e d in Line 3 d e m o n s t r a t e s that the estimated correlation p a t t e r n is such that e v e n a two-"metavariable" m o d e l fails to capture the diversity of these latent traits. In short, we are forced to conclude that the practice of bureaucratic p e r s o n n e l relations is i n h e r e n t l y multifaceted. To the extent that either a s y s t e m of control or a s y s t e m of rights exists, it is a loosely coupled one. Before turning to an examination of the causes a n d c o n s e q u e n c e s of these various dimensions, w e m u s t briefly c o m m e n t o n one m i n o r technical matter: the appropriate w a y to convert these g r o u p s of items into useful m e a s u r e m e n t scales. The options are combining the various items in each d i m e n s i o n b y using relatively equal weights or combining t h e m b y using the factor scores that e m e r g e as a b y - p r o d u c t of the confirmatory factor analysis. For reasons of simplicity, w e chose to do the former. N e v e r t h e less, w e did evaluate the consequences of this decision b y constructing the m e a s u r e s b o t h w a y s a n d c o m p a r i n g the resulting scales. The u p s h o t of these comparisons is that it probably makes little difference w h i c h set of scales is u s e d in the s u b s e q u e n t analysis. O n the three m o s t i m p o r t a n t d i m e n s i o n s (due process, mobility assurance, and m a r k e t sheltering), the correlations b e t w e e n the w e i g h t e d a n d u n w e i g h t e d scales for the corres p o n d i n g concepts r a n g e d from .83 to .96. Likewise, the w e i g h t e d a n d u n w e i g h t e d m e a s u r e s had nearly identical correlations with establishm e n t size. Thus, our decision to use the simple, equally w e i g h t e d , additive scales is not likely to have colored our findings v e r y much. 3 These models differ from the previous model (that described in Line 1) only in the way in which they represent the intercorrelations of the five "first-order" latent variables. In the model with no second-order factors (i.e., Line 1), these relationships are represented by the full correlation matrix of the latent variable errors. In LISREL terms, this is a PSI matrix with 10 estimated parameters. In Line 2, the model posits a single second-order factor that "causes" the first-order factor. Thus, the PSI matrix is replaced by five causal coefficients (one of which is fixed by design) and by a single variance term for the secondorder factor--a total of five estimated parameters. This accounts for the difference in degrees of freedom between the two models. Parallel modifications were made to arrive at the two-factor model summarized in Line 3.


Chapter 3

With these preliminaries behind us, we n o w investigate h o w various demographically defined groups compared in their involvement with bureaucratic personnel structures. This endeavor will also shed additional light on which of the foregoing theoretical accounts best explains these rather heterogeneous elements of bureaucratic personnel management.

4 The Scope of Bureaucratic Management In this chapter, we evaluate the spread of bureaucratic employment structures across different segments of the population defined with regard to sociodemographic categories, namely, age, race, and gender. The results of this investigation are interesting not only from a descriptive or "social accounting" viewpoint, but also for the insight they provide into h o w these structures affect employees' labor market behavior. That is, if we accept the basic premise that the five dimensions of bureaucratic employment structure are important because they establish more durable links between workers and firms than exist in an unstructured market, the prevalence of these links across different groups will depend on two main factors. The first factor is the relative ability of each group to effect such a connection or, to use terms prevalent in recent social stratification theory, the level of social resources held by each group. In general, we would predict that groups with better resources--more education, more work experience, and better representation in the current decision-making structure--would disproportionately populate the ranks of internal labor markets and rule-ordered personnel systems. However, an equally important factor is the motivation of each group to forge a stronger bond with particular employers. Because this concept is less familiar and, we would argue, often overlooked, some elaboration is in order. The degree of motivation for strong employment ties is a matter that depends upon each group's assessment of its alternative social and economic prospects. Groups that perceive, whether accurately or not, that they face discrimination in the broader labor market are likely to prefer arrangements that stabilize their relationships with particular employ71


Chapter 4

ers. On the other hand, groups with strong commitments to alternative bases of affiliation are less likely to prefer these arrangements. In fact, the most t h o r o u g h g o i n g version of this a r g u m e n t postulates that these alternative commitments m a y be not only of the e m p l o y m e n t variety, such as professional membership, but also of the "extraemployment" variety, such as family a n d h o u s e h o l d commitments. The nature of alternative e m p l o y m e n t opportunities is critical, however. As a starting point for our analysis, we draw on transaction cost analysis, which explicitly posits bureaucratic e m p l o y m e n t arrangements ("govemance structures") and market mechanisms as alternative m e a n s of regulating the terms of employment. The key i n d e p e n d e n t variable in that theory, however, is skill specificity--a factor w h o s e causes are not explained u n a m b i g u o u s l y in the theory but that appears to result in large measure from the nature of the technology a d o p t e d by particular firms for particular jobs: "Continuity of the e m p l o y m e n t relation is valu e d by both employer and employee for tasks that involve the acquisition of significant transaction-specific skills, while tasks for which skill acquisition is insubstantial a n d / o r general purpose do not create the same continuity of interests." (Williamson, 1985, p. 255). By implication, where this " h u m a n asset specificity" is missing, external markets are readily available as a means of adjusting worker and employer contributions a n d interests. 1 Our view is based on a s o m e w h a t different model, in which allowance is m a d e for the fact that nonskill considerations m a y influence the " g o v e r n a n c e ' - m a r k e t tradeoff. The most important of these are occupational institutions that arise not only in response to the degree of "firm specificity" of the work performed by a typical incumbent, but also in response to a variety of other social forces. A m o n g these forces are the political activities of the occupation, which result in licensure requirements a n d other restrictive practices, and, second, the occupation's efforts to build occupational social networks and other infrastructures that promote and facilitate intrafirm mobility a m o n g its members. In other words, if the degree of skill specificity is held constant a n d two kinds of 1 Actually,Williamsonalso introduced a second dimension in the classificationof types of employment relationships: the degree to which individual productivity is easily monitored or "metered." Thus, his analysis subdivides work arrangements where asset specificityis present into two subtypes, depending on whether individual productivity can or cannot be easily reckoned. The situation of high observability,he argued, leads to "obligational markets," and the situation of low observability leads to "relational teams." From the present point of view, this distinction is not particularlyimportant because both obligational markets and relational teams include mechanisms for encouraging the continuity of employment relationships between firms and workers.


Scope of Bureaucratic Management


work differ in the extent to which external opportunities (mobility chances) are institutionalized according to well-defined occupational channels, the less "occupationalized" job will be more likely to be part of a governance structure than the more occupationalized one. The logic of this proposition is derived from Hirschman's familiar "exit-voice" dic h o t o m y (1971) and, in contrast to the single-factor transaction cost model, rests on the assumption that the efficacy of labor markets, at least from the point of view of workers, is itself a contingent variable that cannot be a s s u m e d to be constant. 2 The identification of occupation-specific external markets as an alternative source of stability in e m p l o y m e n t transactions was anticipated by Althauser a n d Kalleberg (1981) in their five-element labor market typology (see also Stolzenberg, 1975). The three types of market arrangements in their scheme entail frequent recourse to between-employer movement: occupational "internal" labor markets (OILMs), occupational labor markets (OLMs), and secondary labor markets (SLAMs). Interfirm m o v e m e n t in the latter category (SLAMs) is largely r a n d o m and unstructured a n d differs markedly from that f o u n d in the first two categories: "Turnover is typical, expected by employer a n d employee alike . . . . Job training is simple and b r i e f . . , there is no inherent potential for orderly m o v e m e n t into other jobs" (p. 136). On the other hand, both occupational "internal" labor markets a n d occupational labor markets per se enable between-firm mobility that is purposive, normatively sanctioned, a n d initiated largely at the discretion of employees. In the Althauser a n d Kalleberg model, these two types are further differentiated by w h e t h e r external job m o v e m e n t necessarily involves a progressive e n h a n c e m e n t of workers' skills and responsibilities (in the "occupational internal" mode, it does). For our purposes, there is no necessity to distinguish b e t w e e n these two occupation-specific types of markets. In this sense, our usage is very reminiscent of Stinchcombe's identification (1979) of a "craft a n d professional" mobility type, in which "people keep the same occupation, but move a m o n g employers. Presumably the most c o m m o n cause of this p h e n o m e n o n is that t h e y are more valuable in the occupation in which t h e y have experience than they are on the open market" 2 There is the possibility, however, that occupational institutions have no direct causal influence on the governance-market tradeoff and that all of their effect is mediated by the degree of skill specificity.This view places occupation forces further back in the causal chain and locates them as a factor parallel to technology as a source of skill specificity. This channel of influence is consistent with a diverse literature that documents the role of occupational and professional groups in striving for work autonomy within larger organizations (Bucher & Strauss, 1961; Freidson, 1984). Although these effects undoubtedly exist, we believe they are not the whole story.


Chapter 4

(p. 239). In Stinchcombe's scheme and ours, it is crucial, however, that occupation-specific mobility not be confused with the disorderly movement endemic among jobs in the "secondary" or "open" type of external market. Where movement is haphazard and uncertain, little protection is offered to workers through the "exit" option, and pressure for internal "governance" solutions is not averted. Under these conditions, whether these bureaucratic arrangements are in fact achieved will turn on the resource considerations mentioned above. The transaction cost model also adds a second consideration to the explanation of the mobility component of structured employment relations: the difficulty of making short-run measurements of individual employees' productive contribution. The basic idea here is that, if information about an employee's worth is unreliable, either because the nature of the work inhibits easy observation, because the employee manipulates his or her presentation of himself or herself, or because the employer lacks appropriate measuring skills, it is advantageous for the employer to defer rewards through a sequential mobility process rather than deliver all of the benefits "up front." We might suspect that, if employers, a generally white, middle-class group, have particular difficulty in assessing the performance of minority group workers, a common strategy would be to assign them more frequently to job ladders-although ones at a fairly low level--than they would comparable white employees. This may be particularly true for minority individuals w h o are at the low end of the usual scales on which future productivity is judged, that is, education and experience. On this basis, therefore, we might predict a negative association between the conventional "human capital" factors and participation in job ladders for minority group workers. On the other hand, it is possible that a diametrically opposed set of forces is operating. If minority group members place greater value on job security and procedural fairness in their dealings with their current employers, it is likely that blacks with the most social resources would be those with the most success in obtaining bureaucratically structured jobs. Perhaps the most realistic expectation, therefore, is that education and experience for minorities are more positively related to due process and bureaucratic work control than they are to participation on job ladders. This hypothesis will be examined empirically when we look at the "within-group" results below. To return to our main argument, to evaluate the access of w o m e n and racial minorities to internal labor markets and other bureaucratic job structures, two major considerations must be addressed. The first is the issue of social resources. Are members of disadvantaged minority

The Scope of BureaucraticManagement


groups handicapped by a lack of education and continuous labor force participation in escaping from dead-end jobs? To what extent are they deprived of opportunities to acquire firm-specific skills that, if obtained, would induce employers to include them in schemes for enhancing mobility and firm commitment? The second is the matter of access to alternative employment prospects. To what degree do w o m e n and minorities work in occupational contexts in which orderly between-firm movement is difficult? To what extent does such difficulty impel them to seek a bureaucratic or organizational solution to the problem of employment stability? A somewhat different set of issues is posed in analyzing the relationship between age and participation in various internal-labor-market structures. By their very nature, internal promotion sequences and other governance mechanisms are implemented as career-organizing devices and, of necessity, are age-related. This patterning formed the basis of Kaufmann and Spilerman's innovative analysis (1982) of the relationship between age distributions and institutional forces across occupational groups. However, because their study regards the age structure of an occupation as the dependent variable, it is of only indirect relevance to our purpose of explaining how workers' ages affect their levels of participation in bureaucratic governance structures. We begin this endeavor by focusing on the two variables most strongly related to internal labor markets in our scheme: protection from outside hiring and mobility chances. The total effect of an individual's age on these variables may be usefully considered as the combination of two components. The first of these is the probability that an individual of a given age is part of a job ladder at any level or "rung." Because recruitment to job ladders typically takes place at the beginning of careers, most age-related change in participation rates on job ladders results from exits, not entrances. For example after age 30, very few workers are selected into "ports of entry," and as any cohort ages, the numbers of people exiting job ladders can be expected to increase--perhaps gradually at first, but then with increasing force. Thus, if this were the only operant consideration, we would expect that individuals' mobility chances and protection from outside hiring would decrease throughout the life cycle at an increasing rate. However, the second component we must consider is the conditional probability that an individual has either hiring protection or job mobility opportunities given that he or she is an internal-labor-market participant. The age patterning of these conditional probabilities is probably different for the two variables. For individuals in an internal labor market, mobility chances are probably high initially, level off, and then


Chapter 4

decline as workers reach the end of their productive careers. By contrast, protection from outside hiring is probably minimal at the "entry-port" career phase and gradually increases, although at a decreasing rate, as one's career in an internal labor market (ILM) progresses. In this model, the overall probability of being in a job with structured mobility opportunities or protection from outside hiring would be the product of the probability of being part of an ILM and the conditional probability of having structured mobility chances given membership in the ILM. Based on the discussion above, we would expect the age patterning of both mobility chances and hiring protection to have an inverted and somewhat asymmetric U shape. For mobility chances, the "hump" in the U should occur relatively early in the life cycle and should be followed by consistent declines in chances thereafter. For job protection, the peak levels ought to occur later in the career, with a somewhat steep drop off in subsequent years, as significant proportions of individuals leave ILMs. The other "governance" factors investigated in this s t u d y - - d u e process arrangements, bureaucratic work control, and personnel departments--have less significance as "career-organizing" mechanisms and are probably less patterned by age than the internal-labormarket variables. Consequently, the relationships between these variables and age is probably governed to a much larger extent by the factor of work experience than is true of mobility chances and job protection. That is, age as a sociological variable is necessarily related to a "socialresource" component (work experience) and a life-cycle component (career stage). For the ILM variables, the life-cycle component can be expected to exert a strong independent influence; for the other variables, this is less true, and after work experience is taken into account, any age effects should be slight. The social resources normally associated with the term human capital can also be expected to influence individuals' exposure to a structured employment relationship. In the institutional-bureaucratic theory, one would expect education to be strongly related to almost all aspects of these personnel structures. With their emphasis on written rules and procedures and with their proclivity to recruit workers on the basis of technical competence, bureaucracies tend to have more educated work forces and more structured personnel relationships. Furthermore, to the extent that education raises individuals' bargaining power with employers, they should enjoy better mobility prospects, more protection from outside competition, and more protection through due-process guarantees. The role of work experience is somewhat less clear-cut, but it is likely that it, too, functions as a resource that employees can bring to

The Scope of Bureaucratic Management


bear to enhance the institutionalization of their job security and their employment rights. With these considerations in mind, we are n o w in a position to state our expectations about the relative access that various groups have to structured employment relationships. White males should have consistently better mobility chances and protection from outside hiring than white w o m e n and blacks of either sex. These advantages in internallabor-market participation should, however, decline after education and work experience are taken into account. Similarly, we expect that white males have better opportunities to obtain postschool training, and since we expect a positive relationship between postschool training and mobility chances, white males' internal-labor-market advantages should be further reduced after training opportunities are statistically adjusted. On the other hand, white males are probably better positioned with regard to external mobility opportunities in occupational labor markets than are the other groups. Since we expect external occupation labor markets to be negatively related to internal markets and other bureaucratic employment mechanisms, controlling for this factor should increase our estimates of white males' position vis-a-vis the other groups. Predictions about how white males compare to other race and gender groups on access to due process, exposure to bureaucratic work control, and coverage by personnel departments would have to be tempered by the recognition that these work aspects, especially bureaucratic work control, are not themselves unmitigated benefits to all individuals. Although, in general, having one's work life subject to lawlike rules and procedures may seem preferable to serving at the whim of an arbitrary master, bureaucratic control comes at the price of a sacrifice of autonomy, flexibility, and, perhaps, creativity. For this reason in particular, we might expect that these aspects of structured employment relationships are more appealing to minority group members than to majority group members and that controls for social resource variables would not diminish these patterns.


Appendix E contains a full description of the procedures used to produce one measure of employees' access to a well-defined occupational labor market. This measure is constructed entirely from a data source that is external to the MEWS study and its questionnaires: the January 1983 Current Population Survey (CPS). Through an analysis of job movement patterns of individuals classified according to the detailed occupa-


Chapter 4

tion scheme (at the three-digit level), a factor score was derived from the CPS that measures the extent of between-firm occupational mobility that is characteristic of each occupation. Individuals in the MEWS sample were then assigned the "occupational labor market" score that corresponded to their current occupation. As a supplementary measure, we included an additional variable that was based on individuals' responses in the MEWS survey. That variable is a dichotomy on which positive responses were accorded to those who said that they had had a distinct occupational preference w h e n they were searching for their current job and who reported that their first occupation for their current employer was the type of work they had been seeking. The reasoning behind this measure is that a successful within-occupation search depends to a considerable extent on the presence of a well-defined occupational labor market, particularly after education and work experience are taken into account, as they were in the subsequent analysis. In the analysis that follows, the CPS-based measure is referred to as "high-mobility occupation," and the individual job-search-based measure is called "successful occupational search."


Because of the complexity introduced by the age variable, we will defer those results until we have completed a basic orientation to the differences between the race and sex groups. Table 4.1 contains a summary of the univariate distributions of the variables used in our models. The pattern of differences on the social resource variables--education and work experience--is close to expectations. White males had the highest mean years of schooling (nearly 14) and the longest mean number of years of work activity (about 18). Black males were, on average, the least-educated group, and white females had the shortest length of working life. In general, the work experience differences appear to be more dramatic than the educational ones. The training variables present something of an ambiguous message. White males reported by far the longest length of training--almost 21/2 y e a r s - - a n d black males reported the least, a figure close to 1 year on average. The levels of participation in white-collar jobs also follow an orderly pattern: Women were more likely to work in nonmanual jobs than men, and whites were more likely to be white-collar workers than blacks. The combined effect of these trends is such that white women were nearly twice as likely to work in white-collar jobs than were black men. This same pattern reoccurs with regard to external occupational-labor-market

The Scope of Bureaucratic Management


Table 4.1. Means of Independent and Dependent Variables

by Sex and Race Categories

Years of education Years of work experience Training by coworkers Length of training period (weeks) White-collar job Successful occupational search Occupation mobility score Age category (%) Under 25 25-34 35-44 45-54 55 and over Mobility Hiring protection Due process Bureaucratic control Personnel department Number of cases


White female

White male

Black female

Black male

13.8 16.3 0.157 89.4 0.653 0.231 -0.120

13.7 13.2 0.167 47.0 0.792 0.273 0.060

14.0 18.1 0.144 124.5 0.594 0.220 -0.236

13.5 15.8 0.141 61.2 0.673 0.234 0.025

12.7 17.3 0.240 55.9 0.373 0.093 -0.303

14 35 22 16 13 1.70 1.47 2.60 2.80 3.21 2,487

17 34 21 16 12 1.54 1.25 2.60 2.85 3.13 825

12 36 23 17 13 1.76 1.60 2.50 2.67 3.20 1,307

12 34 23 18 13 1.86 1.38 3.04 3.22 3.50 205

12 41 21 12 14 1.84 1.74 2.98 2.96 3.29 150

participation, a l t h o u g h the size of the differences a p p e a r s to be d i m i n ished. N e v e r t h e l e s s , white w o m e n w e r e the g r o u p m o s t likely to be in a high-external-mobility o c c u p a t i o n a n d to h a v e f o u n d a job in a n occupation t h e y w e r e searching for, a n d black m a l e s w e r e the g r o u p at the o t h e r e n d of each of these continua. Because of the strong similarity a m o n g these patterns, it is t e m p t i n g to speculate that i n c u m b e n c y in a white-collar job is a factor that explains the possibilities for orderly external m o v e m e n t . As w e shall see shortly, there is only a w e a k association b e t w e e n these characteristics, a n d their effects on participation in internal-labor-market structures are largely i n d e p e n d e n t . Tables 4.2 t h r o u g h 4.6 p r e s e n t the differences b e t w e e n w h i t e males, black males, w h i t e females, a n d black females o n o u r five g o v e r n a n c e structure variables before a n d after the introduction of a variety of statistical a d j u s t m e n t s . (The race-by-sex g r o u p coefficients are b a s e d o n the "effect c o d i n g " m o d e l , in w h i c h the coefficient for a particular g r o u p r e p r e s e n t s the difference b e t w e e n that g r o u p a n d the overall m e a n . In this coding s c h e m e , w h i t e females w e r e the reference g r o u p that w a s excluded f r o m the equation. The coefficients for w h i t e females s h o w n in these tables w e r e derived b y calculating the a p p r o p r i a t e linear c o m b i n a -










,.O "0



' I

~ I




A r e t h e r e s e t p e n a l t i e s for ru le violations?


9 = Missing 1 = Yes--all 2 = Yes--some 0 = DNA 8 = Don't know 3=No

Dropped ~ 4 ---> 3 ---* 2 ---> 2 "--~ 1

2 = No

aDNA = does not apply.

test, we divided the questionnaire items into two groups, those from the employer questionnaire and those from the employee questionnaire. The employee responses used were DPROC2, DPROC3, DPROC4, and DPROC5; the employer responses used were DPROC6 and DPROC7. The items were coded according to the scheme given in Table D.1. We then estimated the canonical correlation between these sets of items, including within each set of items dummy variables representing nonstandard responses. For example, the employer variables were


Construction of Measures

Table D.2. Relationships between Selected Due-Process Items Must get advance notice of firing: % saying "yes"

Have formal appeal process if dispute: % saying yes


59 30 21

62 26 22

2,132 126 452

70 46 37

76 48 37

1,319 264 542

56 53 49 43 41

70 54 31 45 32

926 356 59 279 242

55 55 55 49 42

69 55 55 31 39

976 299 11 59 505

DPROC2. Employer must give reason if person is fired Yes Don't know No DPROC3. Reason for firing must be in writing Yes Don't know No DPROC6. How much of rules are in writing? All Most No rules Some None DPROC7. Are there set penalties for rule violation? Yes, for all Yes, for some Don't know No rules No

DPROC6, DPROC7, a d u m m y variable for the " n o rules" r e s p o n s e on the first variable, a n d a d u m m y variable for the " d o n ' t k n o w " r e s p o n s e on the s e c o n d variable. If the n o n s t a n d a r d r e s p o n s e s were severely miscoded, these d u m m y variables w o u l d have exhibited large stand a r d i z e d canonical coefficients. H o w e v e r , as expected, these coefficients were v e r y small (less t h a n .08) in every instance, a n d it is unlikely that our final scales are c o m p r o m i s e d b y our treatment of these responses. I N T E R N A L LABOR M A R K E T S As w e have explicitly a r g u e d in the text, internal labor markets n e e d to be r e g a r d e d as combinations of t w o discrete elements: the practice of inside hiring for jobs (the protection c o m p o n e n t ) a n d the a t t a c h m e n t of mobility assurances to specific jobs (the mobility c o m p o n e n t ) . Consequently, w e derived t w o additive scales describing each job o n these


Appendix D

Table D . 3 . Internal Labor Market Items

Variable PROTQ46

Item(s) A r e m o s t p e o p l e in y o u r job h i r e d from o u t s i d e

Reference Q46

or inside the company?

Initial codes a 9 8 4 2

= = = =

Missing Don't know No experiences Outside

3 = Half/half I = Inside PROTQ42A

Total number of jobs with current employer


99,98 = M i s s i n g

1 = One 2 = Two 3-24

Resulting codes Dropped --* 0 ~ 0 ---* 0 --* 3 ---* 4 Dropped --~ 0 ---* 2 --* 4


Are vacancies for this job usually filled from outside thiscompany?


1 = Outside 3 = Both 2 = Inside


Is this job a steppingstone to another job in this company?


9 = Missing 8 = Don't know 0 = DNA 2= No 3 = Depends 1 =Yes

Dropped --~ 0 ---* 0 --*0 ~ 3 ---~4


If person is still with


999 = M i s s i n g 0 = DNA 0%-10% 11%-74% 998 = D o n ' t k n o w 74%-100%

Dropped --~ 0 ~ 0 ~ 2 ---* 2 ~ 4


9 = Missing 8 = Don't know 2=No I = Yes

Dropped --~ 2 --*0 ---*4

company in 10 years, what is chance of promotion?


Is job p a r t of a p r o m o tion l a d d e r ?

~ 0 ---* 2 --~ 4

aDNA = does not apply.

dimensions. Each scale consists of three items, and procedures similar to those discussed above were used w h e n missing values were encountered on any of the constituent items. Table D.3 presents an overview of the questionnaire items used for the internal-labor-market variables. In this case, three items were taken from the employer interview and three from the employee interview. Again, certain items were answered with nonstandard responses. In particular, w h e n employees were asked about the usual source of recruits for their o w n jobs (PROTQ46), some were reluctant to choose one of the extreme categories and answered "half and half"; others professed to have too little experience in the company to be able to say.

Construction of Measures


Table D.4. Relationship between Selected Internal-Labor-Market Items

Item Are most people in your job hired from inside or outside the company? Inside Half and half Don't know Outside If person is still with the company in 10 years, what is the chance of promotion? 75%-100% 11%-74% Don't know 0%-10% Does not apply

Number of jobs in company % with 3 or more

Stepping-stone to another job? % saying yes


41 24 8 8

42 39 28 26

588 222 72 934

27 22 23 12 9

41 36 34 22 22

926 356 59 505 65

Similarly, w h e n e m p l o y e r s were asked about the p r o m o t i o n chances (MOBIl3) of individuals i n jobs (provided individuals stayed with the c o m p a n y at least 10 years in those sorts of jobs), some e m p l o y e r s w e r e unable to assess w h a t the p r o m o t i o n chances might be. In addition, because this question was p r e c e d e d by a question asking w h a t the chances w e r e that e m p l o y e e s w o u l d stay with the c o m p a n y for 10 years after t h e y e n t e r e d the job in question, the chances-for-promotion question was not asked of those e m p l o y e r s w h o said there was no chance of individuals in the job staying 10 years. O u r p r o c e d u r e for dealing with these ambiguities was similar to that u s e d for the due-process items. Table D.4 displays the tabular relationships b e t w e e n the items in question a n d some criterion items t a k e n f r o m the same c o n t e n t area. In the case of the employee-hiring variable, the " d o n ' t k n o w " r e s p o n s e s m a p p e d onto the criterion variables m o r e like the "outside hiring" r e s p o n s e than like a neutral response, a n d our coding of this item reflects that p a t t e m . For the e m p l o y e r mobility variable (MOBIl3), the " d o n ' t k n o w " r e s p o n s e s b e h a v e d after the fashion w e have n o t e d b e f o r e - - t h a t is, as neutral or middling r e s p o n s e s - - a n d w e r e c o d e d accordingly. Again, these decisions w e r e verified b y the canonical correlation p r o c e d u r e described above, the results indicating that the coding decisions were generally correct (i.e., the s t a n d a r d i z e d canonical coefficients for the n o n s t a n d a r d r e s p o n s e d u m m y variables w e r e quite small, b o t h relatively a n d absolutely; all w e r e . 15 or smaller).

E Classifying Occupations as Orderly External Markets Despite the obvious importance of the concept, little attention has been devoted to examining the degree to which occupations provide a context for orderly between-firm movement. Although other studies have examined aggregate transition rates among occupations and other labor force statuses (Hayward & Grady, 1986), studies of occupational patterns of movement among employers are quite scarce (Stinchcombe, 1979, is one exception). Even in cases where the desirability of data on interfirm mobility is recognized, researchers have often resorted to using the very poor proxy measure of interindustry mobility (Kaufman & Spilerman, 1982). The classification scheme developed here is based on an aggregate analysis of data from the January 1983 Current Population Survey, in which individual respondents were asked a variety of questions related to their occupational and firm mobility. Respondents who were working both at the time of the survey and one year earlier were asked, "You told me that [you] are now working as [occupation]. Were [you] doing the same kind of work a year ago, in January 1982?" For respondents w h o had changed occupations, a series of additional questions were posed. These included the name of the employer one year earlier and the detailed occupation, industry, and class-of-worker items used regularly in census data collections. All employed respondents were also asked h o w many years they had been working in their current occupation category and h o w many months or years they had been working for their current employers. From these basic items, a number of occupation-level indicators were created. Two of these were based on a measure that we call the occupationfirm tenure ratio, or hereafter OFTR. This measure is simply the ratio of 217


Appendix E

the number of months the individual had worked in his or her current occupation divided by the number of months worked for his or her current employer. Values below 1 on this ratio indicated people whose current occupation had changed since coming to work for their current employer. Higher values indicate those whose involvement with their occupation predated their involvement with their current employers. The assumption behind these calculations is that occupations in which workers changed firms while maintaining their occupational identities were occupations characterized by orderly and institutionalized occupational labor markets. All individuals in the CPS sample were then identified as to their status on two dichotomies: whether they were low-OFTR workers with a ratio value of less than one or whether they were high-OFTR workers with a ratio value greater than 3.5. These measures were then aggregated to the three-digit occupation level, and each occupation was characterized according to its proportion of high- and low-OFTR workers. For the occupations included in this analysis, the mean occupation had about 13% low-OFTR workers and about 14% high-OFTR workers. More important was the distribution of these proportions across occupations. Thus, occupations ranged from a minimum value of zero on the proportion of low-OFTR workers to a maximum value of .43 on low-OFTR workers. The distribution of the proportion of high-OFTR workers was similar: from a minimum value of zero to a maximum value of .48. (It should be noted that, at the occupation level, there is theoretically some built-in negative correlation between these ratios, but that, empirically, the conditions are not met for there to be much artifactual relationship here. Thus, occupations with 100% low-OFTR workers would have to have 0% high-tenure workers. However, as we have just seen, there were no 100% low- or high-OFTR occupations. Moreover, occupations with 0% low-OFTR workers would not have to have any high-OFTR workers, as all individuals could be in the 1 - 3 . 5 0 F T R range, and the same is true of occupations with 0% high-OFTR workers.) These two proportions then make up a positive indicator of an occupation offering orderly between-firm mobility and a negative indicator of it. The other two occupation indicators are based on responses to the questions about each individual's occupational status one year before the interview and their length of employment with their current employer. Each individual was classified as having experienced one of four types of occupational movement: (1) none, or same employer and same occupation as last year; (2) occupation "changer" and employer "stayer" in the previous year; (3) occupation "stayer" but employer "changer" in the previous year; and (4) both occupation and employer mobility in the

Classifying Occupations as Orderly External Markets


Table E.1. Correlation Matrix of Occupational Mobility Measures

1 1. Retrospective proportion of workers making within-occupation moves 2. Proportion of low Occupation-Firm Tenure Ratio (OFTR) workers 3. Proportion of high Occupation-Firm Tenure Ratio (OFTR) workers 4. Prospective proportion of employer changers keeping occupation




1.00 -.522






1.00 .264


p r e v i o u s year. T h e n , in the a g g r e g a t i o n process, each o c c u p a t i o n w a s characterized b y the p r o p o r t i o n of all of its current i n c u m b e n t s w h o w e r e "mobility t y p e 3%," that is, firm c h a n g e r s but o c c u p a t i o n stayers. At the o c c u p a t i o n level, values for this variable r a n g e d f r o m a l o w of 0% to a h i g h of 48%, w h i c h is quite h i g h considering the t i m e p e r i o d of interest w a s only one year. The final m e a s u r e w a s created b y a similar logic, except p r o s p e c tively rather t h a n retrospectively. Each individual w a s identified accord-

Table E.2. Factor Analysis of Occupational Mobility Measures

Eigenvalues before iteration Eigenvalue Difference Proportion Cumulative


Factor 2

Factor 3


2.095 1.348 0.524 0.524

0.746 0.068 0.187 0.710

0.679 0.198 0.170 0.880


Factor loadings 1. Retrospective proportion of workers making within-occupation moves 2. Proportion of low OFTR workers 3. Proportion of high OFTR workers 4. Prospective proportion of employer changers keeping occupation

0.120 1.000











Appendix E

ing to the occupation he or she worked in during January 1982. For all individuals who changed employers in the next year, some changed occupations and some did not. For each 1982 occupation, a proportion was constructed of the number of employer changers who were also occupation stayers. Again, occupations where prospective employer movement was associated with occupational stability were regarded as "orderly-movement occupations." The range of values across occupations for this measure was from 0% to 100%. That is, in some occupations, all employer moves were also occupation moves, and in others, all employer moves were made within occupational boundaries. To create a summary measure of orderly occupational movement, these four indicators were analyzed with common factor analysis. The correlation matrix of the input variables is shown in Table E.1. As is readily apparent, all correlations are in the expected direction, and all are substantial. The results of the factor analysis are shown in Table E.2. The one-factor solution is revealed to be adequate, as the next largest eigenvalue falls considerably below the conventional cutoff criterion of 1.00. The factor pattern indicates that the best indicators of orderly movement are the retrospective proportion of all workers making between-firm within-occupation moves and the negative indicator of the proportion of the occupation, which is low-OFTR workers. Based on this analysis, factor scores were created for the set of occupations included in the analysis. Before inspecting the results of this scoring on an occupation-byoccupation basis, it is necessary to explain briefly the steps that were taken to deal with small cell sizes for some occupations. Although the CPS mobility questions were responded to by 69,637 workers, they were not evenly distributed across the 496 detailed census occupation categories. For the purposes of this analysis, we defined the minimally acceptable cell size as 20 workers. By this criterion, there were 336 occupations that qualified for inclusion. To avoid missing values on the remainder of the occupation sample, we collapsed some adjacent and/or related occupation categories. For example, 24 categories of post-secondary-school teachers that had sample sizes too small to be included on their o w n account were merged with the category "post-secondary-school teachers not elsewhere classified" and were entered in the analysis as a single observation. Or as a second example, the detailed category of "ushers" was merged with the detailed category of "guides." Thus, the factor analysis reported above was carried out on a sample of 345 partially grouped occupations that accounted for nearly all of the occupations in the original detailed list. (A residual group of 24 occupations still contained 20 or fewer observations and were dropped from the analysis.)

Classifying Occupations as Orderly External Markets


Table E.3. Factor Scores for Illustrative Group

of Census Occupations Occupation Lowest 10 occupations Locomotive operators General office supervisors Separating, filtering, and clarifying machine operators Production coordinators Purchasing managers Administrators, protective services Production testers Lathe and turning-machine operators Supervisors, production occupations Supervisors, cleaning and building-service workers Middle 10 occupations Eligibility clerks, social welfare Mechanical engineers Weighers, measurers, and checkers Religious workers, not elsewhere classified Therapists, not elsewhere classified Teachers, elementary school Authors Announcers Data-processing-equipment repairers Insurance sales occupations Top 10 occupations News vendors Animal caretakers Dental hygienists Shoe repairers Construction helpers Protective service occupations, not elsewhere classified Athletes Hoist and winch operators Drywall installers Short-order cooks

Score -2.56 -2.14 -2.11 -2.05 - 1.94 -1.93 -1.86 -1.84 -1.80 -1.79

-.030 -.026 -.024 -.019 .001 .003 .018 .020 .024 .030 1.54 1.59 1.65 1.70 1.77 1.80 1.85 1.98 1.98 2.21

A l t h o u g h this g r o u p i n g p r o c e d u r e is o p e n to criticism o n t h e g r o u n d s that some occupations with divergent mobility patterns may have been l u m p e d t o g e t h e r in t h e p r o c e s s , it is v e r y u n l i k e l y t h a t this a g g r e g a t i o n i n t r o d u c e d m u c h bias into t h e analysis. First, o c c u p a t i o n s w i t h f e w inc u m b e n t s n a t i o n a l l y w e r e also likely to h a v e relatively f e w i n c u m b e n t s


Appendix E

in the Chicago area and were thus less likely to have materialized in the MEWS study. Second, in most cases, the observations aggregated were quite similar, as in the case of college teachers cited above. Third, the results of the correlation and factor analysis were nearly identical for both the analysis of 339 ungrouped occupations and the analysis of 345 grouped observations. Table E.3 presents the aforementioned factor scores for three groups of occupations: those with the least orderly occupational mobility; those with middling levels (i.e., the 10 occupations with factor scores close to 0); and those with the highest factor scores. The last phase in the construction of these measures was to assign to each individual in the MEWS sample the orderly-mobility score for his or her detailed occupation.


Abraham, Katherine G. 1990. "Restructuring the Employment Relationship: The Growth of Market Mediated Work Arrangements." In Katherine Abraham & Robert McKersie (eds.), New Developments in the Labor Market: Toward a New Institutional Paradigm. Cambridge, MA: MIT Press. Alderfer, Clayton P. 1972. Human Needs in Organizational Settings. New York: Free Press. Aldrich, Howard E. 1979. Organizations and Environments. Englewood Cliffs, NJ: Prentice-Hall. Althauser, Robert P. 1989. "Internal Labor Markets." In W. Richard Scott & Judith Blake (eds.), Annual Review of Sociology, Vol. 15, pp. 143-161. Palo Alto, CA: Annual Reviews. Althauser, Robert, and Ame L. Kalleberg. 1981. "Firms, Occupations, and the Structure of Labor Markets: A Conceptual Analysis." In Ivar Berg (ed.), Sociological Perspectives on Labor Markets. New York: Academic Press. Argyris, Chris. 1960. Understanding Organizational Behavior. Homewood, IL: Dorsey. Argyris, Chris. 1964. Integrating the Individual and the Organization. New York: Wiley. Argyris, Chris. 1972. The Applicability of Organizational Sociology. London: Cambridge University Press. Baron, James N. 1984. "Organizational Perspectives on Inequality." Annual Review of Sociology 10: 37-69. Baron, James N., and William Bielby. 1980. "Bringing the Firms Back In: Stratification, Segmentation and the Organization of Work." American Sociological Review 45: 737-765. Baron, James N., Alison Davis-Blake, and William T. Bielby. 1986. "The Structure of Opportunity: How Promotion Ladders Vary within and among Organizations." Administrative Science Quarterly 31:248-273. Baron, James N., P. Devereaux-Jennings, and Frank R. Dobbin. 1988. "Mission Control? The Development of Personnel Systems in U.S. Industry." American Sociological Review 53:497-514. Baron, James N., Frank R. Dobbin, and P. Devereaux-Jennings. 1986. "War and Peace: The Evolution of Modem Personnel Administration in U.S. Industry." American Journal of Sociology 92:350-383. Beck, E. M., Patrick M. Horan, and Charles M. Tolbert, II. 1978. "Stratification in a Dual Economy: A Sectoral Model of Earnings Determination." American Sociological Review 43:704-720. 223



Bendix, Reinhart. 1974. Work and Authority in Industry. Berkeley: University of California Press. Ben-Porath, Yoram. 1980. "The F-connection: Families, Friends, and Firms and the Organization of Exchange." Population and Development Review 6:1-30. Berk, R. 1983. "An Introduction to Sample Selection Bias in Sociological Data." American Sociological Review 48:386-398. Bibb, Robert, and William H. Form. 1977. "The Effects of Industrial, Occupational, and Sex Stratification on Wages in Blue-Collar Markets." Social Forces 55:974-996. Blau, Francine D., and Lawrence M. Kahn. 1981. "Race and Sex Differences in Quits by Young Workers." Industrial and Labor Relations Review 34:563-577. Blau, Peter M. 1964. Exchange and Power in Social Life. New York: Wiley. Blau, Peter M. 1968. "The Hierarchy of Authority in Organizations." American Journal of Sociology 73:453-467. Blau, Peter M. 1970. "A Formal Theory of Differentiation in Organizations." American Sociological Review 35:201-218. Blau, Peter M. 1972. "Interdependence and Hierarchy in Organizations." Social Science Research 1:1-24. Blau, Peter M. 1973. The Organization of Academic Work. New York: Wiley. Blau, Peter, and Otis Dudley Duncan. 1967. The American Occupational Structure. New York: Wiley. Blau, Peter M., and Richard A. Schoenherr. 1971. The Structure of Organizations. New York: Basic Books. Blau, Peter M., and W. Richard Scott. 1962. Formal Organizations. San Francisco: Chandler. Bluestone, Barry, William M. Murphy, and Mary Stevenson. 1973. Low Wagesand the Working Poor. Ann Arbor, MI: Institute of Labor and Industrial Relations. Braverman, Harry. 1974. Labor and Monopoly Capital. New York: Monthly Review Press. Bridges, William P., and Wayne J. Villemez. 1986. "Informal Hiring and Income in the Labor Market." American Sociological Review 51:574-582. Bridges, William P., and Wayne J. Villemez. 1991. "Employment Relations and the Labor Market: Integrating Institutional and Market Perspectives." American Sociological Review 56:748-764. Bucher, Rue, and Anselm Strauss. 1961. "Professions in Process." American Journal of Sociology 66:325-344. Burawoy, Michael. 1979. Manufacturing Consent. Chicago: University of Chicago Press. Bums, Tom, and G. M. Stalker. 1961. The Management of Innovation. London: Tavistock. Child, John. 1973. "Predicting and Understanding Organization Structure." Administrative Science Quarterly 19:168-185. Clegg, Stewart, and David Dunkerly. 1980. Organization, Class and Control. London: Routledge & Kegan Paul. Cole, Robert. 1979. Work, Mobility, and Participation: A Comparative Study of American and Japanese Industry. Berkeley: University of California Press.



Collins, Randall. 1981. "On the Microfoundations of Macrosociology." American Journal of Sociology 86:984-1014. Dess, Gregory G., and Donald W. Beard. 1984. "Dimensions of Organizational Task Environment." Administrative Science Quarterly 29:52-73. DiMaggio, Paul, and Walter W. Powell. 1983. "The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields." American Sociological Review 48:147-160. DiPrete, Thomas A. 1989. The Bureaucratic Labor Market: The Case of the Federal Civil Service. New York: Plenum Press. Dobbin, Frank R., Lauren Edelman, John W. Meyer, W. Richard Scott, and Ann Swidler. 1988. "The Expansion of Due Process in Organizations." In Lynne G. Zucker (ed.), Institutional Patterns and Organizations: Culture and Environment. Cambridge, MA: Ballinger. Doeringer, Peter B., and Michael J. Piore. 1971. Internal Labor Markets and Manpower Analysis. Lexington, MA: D. C. Heath. Duncan, Greg J. 1986. "Do Women 'Deserve' to Earn as Much as Men?" In Greg J. Duncan (ed.), Years of Poverty, Years of Plenty. New York: Basic Books. Duncan, Robert B. 1972. "Characteristics of Organizational Environments and Perceived Environmental Uncertainty." Administrative Science Quarterly 17:313-327. Edelman, Lauren B. 1990. "Legal Environments and Organizational Governance: The Expansion of Due Process in the American Workplace." American Journal of Sociology 95:1401-1440. Edwards, Richard C. 1975. "The Social Relations of Production in the Firm and Labor Market Structure." In Richard C. Edwards, Michael Reich, and David M. Gordon (eds.), Labor Market Segmentation. Lexington, MA: D. C. Heath. Edwards, Richard C. 1979. Contested Terrain. New York: Basic Books. Emerson, Richard M. 1962. "Power-Dependence Relations." American Sociological Review 27:31-41. England, Paula, G. Farkas, B. S. Kilbourne, and T. Dou. 1988. "Explaining Occupational Sex Segregation and Wages: Findings From a Model With Fixed Effects." American Sociological Review 53:544-558. Etzioni, Amitai. 1975. A Comparative Analysis of Complex Organizations, rev. ed. New York: Free Press. Evans, Sara M., and Barbara J. Nelson. 1989. Wage Justice: Comparable Worth and the Paradox of Technocratic Reform. Chicago: University of Chicago Press. Felmlee, Diane H. 1982. "Women's Job Mobility Processes." American Sociological Review 47:142-150. Filer, Randall. 1985. "Male-Female Wage Differences: The Importance of Compensating Differentials." Industrial and Labor Relations Review 38:426-437. Freedman, Marcia. 1976. Labor Markets: Segments and Shelters. Montclair, NJ: Allanheld, Osmun. Freeman, Richard, and James Medoff. 1979. "New Estimates of Private Sector Unionism in the United States." Industrial and Labor Relations Review 32:143-174. Freeman, Richard B., and James L. Medoff. 1984. What Do Unions Do? New York: Basic Books.



Freidson, Eliot. 1984. "The Changing Nature of Professional Control." Annual Review of Sociology 10:1-20. Galbraith, Jay. 1973. Designing Complex Organizations. Reading, MA: AddisonWesley. Galbraith, John K. 1967. The New Industrial State. New York: Houghton Mifflin. Glenn, Norval D., and Charles Weaver. 1982. "Further Evidence on Education and Job Satisfaction." Social Forces 61:46-55. Goldman, Paul. 1980. "The Marxist Analysis of Organizations: Values, Theories, and Research." Unpublished manuscript, Department of Sociology, University of Oregon. Gordon, David M. 1972. Theories of Poverty and Underemployment. Lexington, MA: D. C. Heath. Gordon, David M., Richard Edwards, and Michael Reich. 1982. Segmented Work, Divided Workers. New York: Cambridge University Press. Gouldner, Alvin. 1954. Patterns of Industrial Bureaucracy. New York: Free Press. Granovetter, Mark. 1973. "The Strength of Weak Ties." American Journal of Sociology 78:1360-1380. Granovetter, Mark. 1984. "Small is Bountiful: Labor Markets and Establishment Size." American Sociological Review 49:323-334. Groves, Robert, and Robert L. Kahl~ 1979. Surveys by Telephone: A National Comparison with Personal Interviews. New York: Academic Press. Gruenberg, Barry. 1980. "The Happy Worker: Determinants of Job Satisfaction." American Journal of Sociology 86:247-271. Halaby, Charles N., and David L. Weakliem. 1989. "Worker Control and Attachment to the Firm." American Journal of Sociology 95:549-591. Hall, Richard H. 1962. "Intraorganizational Structural Variation: Application of the Bureaucratic Model." Administrative Science Quarterly 7:295-308. Hall, Richard H. 1987. Organizations: Structures, Processes, and Outcomes, 4th ed. Englewood Cliffs, NJ: Prentice-Hall. Hall, Richard H., J. Eugene Haas, and Norman Johnson. 1967. "An Examination of Blau-Scott and Etzioni Typologies." Administrative Science Quarterly 12:118-139. Hall, Richard H., and Charles R. Tittle. 1966. "Bureaucracy and Its Correlates." American Journal of Sociology 72:267-272. Harrison, Bennett. 1972. Education, Training, and the Urban Ghetto. Baltimore: Johns Hopkins Press. Hauser, Robert M. 1980. "On 'Stratification in a Dual Economy.'" American Sociological Review 45:702-712. Hayward, Mark, and William R. Grady. 1986. "The Occupational Retention and Recruitment of Older Men: The Influence of Structural Characteristics of Work." Social Forces 64:644-666. Herzberg, Fredrick. 1966. Work and the Nature of Man. New York: T. Y. Crowell. Hickson, David J., Derek S. Pugh, and Diana C. Pheysey. 1969. "Operational Technology and Organizational Structure: An Empirical Reappraisal." Administrative Science Quarterly 14:378-397. Hirschman, Albert. 1971. Exit, Voice, and Loyalty. Cambridge, MA: Harvard University Press.



Hodson, Randy. 1984. "Companies, Earnings, and the Measure of Economic Segmentation." American Sociological Review 49:335-348. Jacobs, Jerry, and Ronnie Steinberg. August, 1988. "Compensating Differentials and the Male-Female Wage Gap: Evidencefrom the New York State Comparable Worth Study." Paper presented at the annual meetings of the American Sociological Association. Jacoby, Sanford M. 1985. Employing Bureaucracy: Managers, Unions and the Transformation of Work in American Industry, 1900-1945. New York: Columbia University Press. Janson, Philip, and Jack K. Martin. 1982. "Job Satisfaction and Age." Social Forces 60:1089-1102. Kalleberg, Arne L., Michael Wallace, and Robert P. Althauser. 1981. "Economic Segmentation, Worker Power and Income Inequality." American Journal of Sociology 87:651-683. Kalleberg, Arne L., David Knoke, Peter V. Marsden, and Joe L. Spaeth. 1993. "The National Organizations Survey: A Descriptive Overview." Mimeo. Katz, Donald. 1987. The Big Store: Inside the Crisis and Revolution at Sears. New York: Viking Press. Kaufman, Robert L., and S. Spilerman. 1982. "The Age Structures of Occupations and Jobs." American Journal of Sociology 87:827-851. Kaufman, Robert L., Randy Hodson, and Nell D. Fligstein. 1981. "Defrocking Dualism: A New Approach to Defining Industrial Sectors." Social Science Research 10:1-31. Kimberly, John R. 1976. "Organizational Size and the Structuralist Perspective: A Review, Critique, and Proposal." Administrative Science Quarterly 21:577-597. Kochan, Thomas A., Harry C. Katz, and Robert B. McKersie. 1986. The Transformation of American Industrial Relations. New York: Basic Books. Lawrence, Paul R., and Jay W. Lorsch. 1967. Organization and Environment. Boston: Graduate School of Business Administration, Harvard University. Leicht, Kevin T., and Michael Wallace. 1988. "The Property Rights of Jobs: Job Entitlement Versus Managerial Prerogative in Manufacturing Firms." Research in Social Stratification and Mobility 7:189-221. Likert, Rensis. 1961. New Patterns of Management. New York: McGraw-Hill. Likert, Rensis. 1967. The Human Organization. New York: McGraw-Hill. Lincoln, James R., and Arne Kalleberg. 1985. "Work Organization and Workforce Commitment: A Study of Plants and Employees in the U.S. and Japan." American Sociological Review 50:738-760. Locke, E. A. 1968. "Toward a Theory of Task Motivations and Incentives." Organizational Behavior and Human Performance 3:157-189. MacNeil, Ian R. 1980. The New Social Contract: An Inquiry into Modern Contractual Relations. New Haven, CT: Yale University Press. March, James G., and Johan P. Olsen. 1976. Ambiguity and Choice in Organizations. Bergen, Norway: Universitetsforlaget. March, James G., and Herbert A. Simon. 1958. Organizations. New York: Wiley. Markey, James P., and William Parks, II. 1989. "Occupational Change: Pursuing a Different Kind of Work." Monthly Labor Review 112:3-12.



Maslow, Abraham H. 1993. "A Theory of Human Motivation." Psychological Review 50:370-396. Mastekaasa, Arne. 1983. "Comment on Japan's and Martin's 'Job Satisfaction and Age': A Test of Two Views." Social Forces 62: 252-254. Mayhew, Bruce H. 1980. "Structuralism versus Individualism: 1. Shadowboxing in the Dark." Social Forces 59:335-375. Mayhew, Bruce H. 1981. "Structuralism versus Individualism: 2. Ideological and Other Obfuscations." Social Forces 59:627-648. McGregor, Douglas. 1960. The Human Side of Enterprise. New York: McGraw-Hill. McGregor, Douglas. 1967. The Professional Manager. New York: McGraw-HiU. Meyer, John W., and Brian Rowan. 1977. "Institutionalized Organizations: Formal Structure as Myth and Ceremony." American Journal of Sociology 83:440463. Meyer, John W., and W. Richard Scott. 1983. Organizational Environments. Beverly Hills, CA: Sage. Meyer, Marshall W. 1972. "Size and Structure of Organizations: A Causal Analysis." American Sociological Review 37:434-440. Miller, Delbert C., and William H. Form. 1964. Industrial Sociology: The Sociology of Work Organizations. New York: Harper & Row. Mohr, Lawrence B. 1971. "Organizational Technology and Organizational Structure." Administrative Science Quarterly 16:444-451. Montgomery, David. 1979. Workers' Control in America. Cambridge, UK: Cambridge University Press. Mottaz, Clifford. 1985. "The Relative Importance of Intrinsic and Extrinsic Rewards as Determinants of Work Satisfaction." Sociological Quarterly 26:365386. Parcel, Toby, Robert L. Kaufman, and Leeann Jolly. 1991. "Going Up the Ladder: Multiplicity Sampling to Create Linked Macro-to-Micro Organizational Samples." SociologicalMethodology 21:43-79. Perrow, Charles. 1967. "A Framework for the Comparative Analysis of Organizations." American Sociological Review 32:194-208. Perrow, Charles. 1986. Complex Organizations: A Critical Essay, 3rd ed. New York: Random House. Pfeffer, Jeffrey. 1981. Power in Organizations. Marshfield, MA: Pitman. Pfeffer, Jeffrey. 1982. Organizationsand Organization Theory. Marshfield, MA: Pitman. Pfeffer, Jeffrey, and James N. Baron. 1988. "Taking the Workers Back Out: Recent Trends in the Structuring of EmploymenL" In Barry Staw and L. L. Cummings (eds.), Research in Organizational Behavior. Greenwich, CT: JAI Press. Pfeffer, Jeffrey, and Yinon Cohen. 1984. "Determinants of Internal Labor Markets in Organizations." Administrative Science Quarterly 29:550-572. Pfeffer, Jeffrey, and Huseyin Leblebici. 1973. "The Effect of Competition on Some Dimensions of Organizational Structure." Social Forces 52:268-279. Pfeffer, Jeffrey, and Jerry Ross. 1981. "Market Power and Sectoral Variation in Wageand Status Attainment." Unpublished manuscript, Graduate School of Business, Stanford University.



Pfeffer, Jeffrey, and Gerald R. Salancik. 1978. The External Control of Organizations: A Resource Dependence Perspective. New York: Harper & Row. Pugh, D. S., D. J. Hickson, C. R. Hinings, and C. Turner. 1969. "The Context of Organization Structures." Administrative Science Quarterly 13:91-114. Roethlisberger, F. J., and W. J. Dickson. 1939. Management and the Worker. Cambridge: Harvard University Press. Scott, W. Richard. 1981. Organizations: Rational, natural, and open systems. Englewood Cliffs, NJ: Prentice-Hall. Selznick, Philip. 1957. Leadership in Administration: A Sociological Interpretation. Evanston, IL: Row, Peterson. Selznick, Philip. 1969. Law, Society, and Industrial Justice. New York: Sage. Simon, Herbert. 1957. Models of Man. New York: Wiley. Spaeth, Joe L. 1985. "Job Power and Earnings." American Sociological Review 50:603-617. Spaeth, Joe L., Diane O'Rourke, Arne L. Kalleberg, David Knoke, and Peter V. Marsden. 1993. Design and Implementation of a National Probability Sample of Work Establishments. Mimeo. Stark, David. 1986. "Rethinking Internal Labor Markets: New Insights from a Comparative Perspective." American Sociological Review 51:492-504. Stevens, Gillian, and Joo Hyun Cho. 1985. "Socioeconomic Indexes and the New 1980 Census Occupational Classification Scheme." Social Science Research 14:142-168. Stinchcombe, Arthur L. 1959. "Bureaucratic and Craft Administration of Production: A Comparative Study." Administrative Science Quarterly 4:168-187. Stinchcombe, Arthur. 1979. "Social Mobility in Industrial Labor Markets." Acta Sociologica 22:217-245. Stolzenberg, Ross M. 1975. "Occupations, Labor Markets, and the Process of Wage Attainment." American Sociological Review 40:645-665. Stolzenberg, Ross M. 1978. "Bringing the Boss Back In: Employer Size, Employee Schooling, and Socioeconomic Achievement." American Sociological Review 43:813-828. Taylor, Frederick W. 1911. The Principles of Scientific Management. New York: Harper. Thompson, James D. 1967. Organizations in Action. New York: McGraw-HiU. Thornberry, Owen T., and James T. Massey. 1988. "Trends in United States Telephone Coverage Across Time and Subgroups." In Robert Groves, P. Biemer, L. Lyberg, J. Massey, W. Nicholls III, and J. Waksberg (eds.), Telephone Survey Methodology. New York: Wiley. Tolbert, Pamela S., and Lynne G. Zucker. 1983. "Institutional Sources of Change in the Formal Structure of Organizations: The Diffusion of Civil Service Reform, 1880-1935." Administrative Science Quarterly 28:22-39. U.S. Department of Commerce, Bureau of the Census. 1983. 1980 Public Use Microdata Sample (B Sample). [Computer File]. U.S. Department of Commerce, Bureau of the Census. 1985. Current Population Survey, January 1983. [Computer File]. Useem, Michael. 1982. "Classwide Rationality in the Politics of Managers and



Directors of Large Corporations in the United States and Great Britain." Administrative Science Quarterly 27:199-226. Villemez, Wayne J., and William P. Bridges. 1988. "When Bigger is Better: Differences in the Individual-Level Effect of Firm Size and Establishment Size." American Sociological Review 53:237-255. Vroom, Victor H. 1964. Work and Motivation. New York: Wiley. Wachtel, Howard, and Charles Betsey. 1972. "Employment at Low Wages." Review of Economics and Statistics 54:121-129. Weber, Max. 1947. "Bureaucracy." In H. H. Gerth and C. W. Mills (eds.), From Max Weber:Essays in Sociology. New York: Oxford University Press. (Original work published 1922.) Weick, Karl. 1979. The Social Psychology of Organizing, 2nd ed. Reading, MA: Addison-Wesley. Williamson, Oliver E. 1975. Markets and Hierarchies:Analysis and Antitrust Implications. New York: Free Press. Williamson, Oliver E. 1981. "The Economics of Organizations: The Transaction Cost Approach." American Journal of Sociology 87:548-577. Williamson, Oliver E. 1985. The Economic Institutions of Capitalism. New York: Free Press. Williamson, Oliver E., and William G. Ouchi. 1981. "The Markets and Hierarchies Program of Research: Origins, Implications, Prospects." In Andrew H. Van de Ven and William E. Joyce (eds.), Perspectives on Organizational Design and Behavior. New York: Wiley-Interscience. Wilson, James Q. 1989. Bureaucracy: What Government Agencies Do and Why They Do It. New York: Basic Books. Woodward, Joan. 1958. Management and Technology. London: Her Majesty's Stationery office. Woodward, Joan. 1965. Industrial Organizations: Theory and Practice. London: Oxford University Press. Wright, James D., and Richard Hamilton. 1978. "Work Satisfaction and Age." Social Forces 56:1140-1158. Zucker, Lynne G. 1987. "Institutional Theories of Organization." Annual Review of Sociology 13:443-464. Zucker, Lynne G., and Carolyn Rosenstein. 1981. "Taxonomies of Institutional Structure: Dual Economy Reconsidered." American Sociological Review 46:869-884. Zwerman, William L. 1970. New Perspectiveson Organization Theory. Westport, CT: Greenwood Press.


Age discrimination, 177 Age factors in attachment, 135, 185 in bureaucratic control, 95, .96, 97 in bureaucratic employment relationships, 181-182 in due process access, 95, 96, 97, 100 in firm dependence, 140, 141 in hiring protection, 75-76, 95, 96, 97, 98-100 in job ladder participation, 75 in job mobility, 75-76, 95, 96, 97 in job satisfaction, 136, 149, 150, 159, 185 in personnel department access, 95, 96, 97, 181, 182 work experience and, 96-98 in work experience/job mobility relationship, 93, 95-98 Asset specificity, 17, 18 Attachment, 125-160 age factors in, 135, 185 authority and, 128 behavioral linkages in, 126-128, 184 definition of, 131 bureaucratic employment structures and, 136-137, 157, 158 in core industries, 132 decentralization and, 129 educational factors in, 133, 140, 141142, 185 external labor market participation and, 131, 158 firm dependence and, 131, 137-138, 140-149, 155, 156, 157

Attachment (Cont.) firm dependence and (Cont.) age factors in, 140, 141 bureaucratic employment structures and, 142-143 career hiatus and, 148-149, 159-160 in core industries, 140, 141, 145-146 definition of, 184 educational factors in, 140, 141-142 firm size and, 145-146, 154 income and, 145-146 job training length and, 140, 141, 145-146 race factors in, 134, 135, 140, 141 sex factors in, 134-135, 140, 141, 146149 social status and, 140, 141-142, 144 supervisory responsibility and, 140, 141-142 unionization and, 140, 141 work control and, 140, 141-142 firm size and, 132 in high-mobility versus low-mobility occupations, 147-149 instrumental linkages in, 126, 127-128, 183-184 definition of, 130-131 internal labor markets and, 125-126, 128-129, 131 job clemency and, 139, 157-158 job control and, 133-134, 139, 185 job satisfaction and, 131, 135-136, 138, 149-154, 155, 156, 157 age factors in, 136, 149, 150, 159, 185 231


Attachment (Cont.) job satisfaction and (Cont.) bureaucratic employment structures and, 149-150 due process access and, 151 educational factors in, 149, 150, 151 firm size and, 150, 151 in high-mobility occupations, 154 income and, 150, 151, 152-153 job clemency and, 153 job control and, 185 measure of, 138 race factors in, 136, 149, 150, 153, 159, 185 sex factors in, 136, 152-153, 154, 160, 185 unionization and, 149, 150, 151-152, 154 work control and, 153 working conditions and, 153, 159, 185 market opportunity in, 126-128 measure of, 138-139 normative linkages in, 126, 128-129, 184-185 in public sector, 132 race factors in, 134, 135, 140, 141, 160 sex factors in, 134-135, 140, 141, 146149, 160 Socio-Economic Index scores of, 134 supervisory responsibility and, 133-134 unionization and, 132-133, 140, 141, 154, 185 unionization index of, 139-140 working conditions measure, 139 Attitudes, relationship to behavior, 9-10 Authority attachment and, 128 cultural, 58 bureaucratic control and, 109, 117-118, 119 due process access and, 109, 117-118, 119 hierarchical structures of, 20 legality of, 24 in superior-subordinate relationships, 7-8 Banking industry, bureaucratic control institutionalization in, 186 Behavior, relationship to attitudes, 9-10


Behavioral linkages, in attachment, 126128, 184 definition of, 131 Black men due process access by, 100 educational levels of, 78, 79 external labor market participation by, 85-86 hiring protection for, 84-85, 99-100 internal labor market participation by, 84-86 job mobility of, 80, 84, 85 job training length of, 78, 79 personnel department access by, 89, 90 white-collar job participation by, 78-79 Black women due process access by, 100 hiring protection for, 84, 85 internal labor market participation by, 82, 84-85 job mobility of, 80, 81, 82, 84, 101 job training length of, 78, 79 personnel department access by, 89, 90 Black workers internal labor market participation by, 181 job mobility of, 98 market opportunity perceptions of, 134, 135, 141, 142, 143, 144, 160, 186 promotion of, 84 See also Black men; Black women Blue-collar workers Chicago-U.S. comparison of, 31, 34 earnings of, 169, 171-172 Bureaucracies, as government, 57, 179 Bureaucratic control age factors in, 95 96, 97 earnings and, 167-168, 169, 170, 171, 176, 187-188, 189 government intervention in, 105 higher-level effects on, 103-124, 109 authority variable, 109, 117-118, 119 competition measure, 109 efficiency theories of, 104, 112, 113114, 115, 124 government intervention and, 108, 111-114, 116, 117, 123 harm potential variable, 109, 117-118, 119 historical theories of, 104


Bureaucratic control (Cont.) higher-level effects on (Cont.) industrial measures, 107-108, 116, 117, 123-124 industry-specific variations in, 109111, 186-187 institutional theories of, 104-105, 113 internal labor market indicators of, 119-123 job-specific measures, 109, 117-119, 122, 123 learning time variable, 109, 118-119 literacy variable, 109, 118-119, 123 organizational measures, 108-109 per company measures, 108 political theories of, 104 unionization measures, 108-109, 111, 112-113, 114-115, 116, 117, 191 in white-collar employment, 94, 171, 173, 174, 175 working conditions variable, 109, 118, 119, 123 Marxist theory of, 65 multiple causality of, 105, 124 as organizational formalization, 104 personnel department access and, 112 unionization and, 191 workers' earnings and, 167-168, 169, 170, 171, 176, 187-188, 189 Bureaucratic employment relationships. See Employment relationships Bureaucratic employment structures, consequences of, 183-189 Bureaucratic-governance model, of personnel management, 56-58 Bureaucratic management consequences of, 7, 26, 27-28 industry-specific variations of, 21-22 scope of, 7, 18, 26, 27 shape of, 7, 18, 26-27 Bush Administration, 193 Capitalist control model, 54, 55-56 Career interruptions, of women, 148-149, 159-160, 185 Carter Administration, 53 Chicago Standard Metropolitan Statistical Area (SMSA), 30 economic characteristics, 31, 34


Chicago Standard Metropolitan Statistical Area (SMSA) (Cont.) establishment distribution characteristics, 33, 35 industry distribution characteristics, 33, 35 labor force characteristics, 32, 34 Civil rights movement, 106 Civil service PACE examination for, 53 See also Public sector Clinton Administration, 193 Company handbooks, 66, 68 Competition due process access and, 109 hiring protection and, 192 Conduct, rules of, 66, 68 Conflict, employer-employee Marxist theories of, 104 surplus values distribution issue in, 55 in worker control, 19-20 Construction industry employer survey data regarding, 46-48 personnel departments in, 110 Contingency theory, 11-14, 189-192 Core industries attachment in, 132, 140, 141, 145-146 bureaucratic management structure of, 21 firm dependence in, 140, 141, 145-146 market opportunity in, 145-146 personnel departments in, 110, 122 unionization of, 21 Current Population Survey, 50, 77-78 occupational mobility data of, 217, 218, 220 Decentralization, attachment and, 129 Dependence, in employment relationships, 4, 184 as behavioral linkage mechanism, 126 in high-mobility occupations, 128 See also Firm dependence Discrimination age-related, 177 bureaucratic employment relationships and, 71-72, 180 due process access and, 90 Drive system, of employment, 63 Dual labor market, 21, 122


Due process age factors in, 95, 96, 97, 100 black men's access to, 86, 90, 92, 94 black women's access to, 86, 90 company handbook distribution and, 66, 68 earnings and, 164, 165-166, 167, 168169, 171-172, 176, 188-189 frequency of use of, 61, 63 government dominance and, 182, 191 higher-level effects on, 103-124, 109 authority variable, 109, 117-118, 119 competition measure, 109 efficiency theories of, 104, 112, 113114, 115, 124 government intervention and, 108, 111-114, 116, 117, 123 harm potential variable, 109, 117-118, 119 historical theories of, 104 industrial measures, 107-108, 116, 117, 123-124 industry-specific variations in, 109111 institutional theories of, 104-105, 113 internal labor market indicators of, 119-123 job-specific measures, 109, 117-119, 122, 123 learning time variable, 109, 118-119 literacy variable, 109, 118-119, 123 organizational measures, 108-109 per company measures, 108 political theories of, 104 unionization measures, 108-109, 111, 112-113, 114-115, 116, 117, 191 working conditions variable, 109, 118, 119, 123 Marxist theories of, 56 of nonsupervisory workers, 63 personnel department access and, 112 in public sector, 182 questionnaire measures of, 211-213 as social pressure response, 105106 statistical indicators of, 68 white women's access to, 86, 87 women's access to, 86, 87, 181 work-rule indicators of, 211-213 Dun and Bradstreet, Inc., 49


Earnings, and employment relationships, 161-176, 187-189 bureaucratic control and, 167-168, 169, 170, 171, 176, 187-188, 189 due process access and, f64, 165-166, 167, 168-169, 171-172, 176, 188-189 educational factors in, 161-162, 166, 187 external labor markets and, 188 firm dependence and, 166, 167, 169, 170, 172, 173, 174, 176, 189 firm size and, 162-163 effect of formalization on, 161-162, 166, 167, 169, 170, 176 hiring protection and, 163, 164, 165166, 167, 168, 170, 175-176, 187, 189 internal labor markets and, 188 job mobility and, 163, 164, 165-166, 167, 168, 170-171, 176, 187 in private sector, 166, 167, 168, 169, 170-171, 187-188, 189-190 in public sector, 166, 167, 168, 169, 170171, 187-188, 189-190 race factors in, 173, 174-175 sex factors in, 172-174 work experience and, 166 Education attachment and, 133, 140, 141-142, 185 black men's level of, 78, 79 earnings and, 161-162, 166, 187 firm dependence and, 140, 141-142 job mobility and, 90, 91, 94 job satisfaction and, 149, 150, 151 personnel structures and, 76 white men's level of, 78, 79 Efficiency, means-ends, 20 Efficiency theories, of bureaucratic control and due process, 104, 112, 113-114, 115, 124 Employer-Identification Question Sequence, 199-200 EmployingBureaucracy(Jacoby), 4 Employment, as economic contract, 1 Employment-at-will, 1 frequency of use of, 63 Employment opportunities, alternative, 72 Employment relationships, 4-28 age factors in, 181-182 bureaucratization of, 2 components of, 186-187


Employment relationships (Cont.) contingency theory of, 11-14, 189-192 definition of, 2 discrete versus relational contracting in, 1-2 economic theories of, 3 employees' motivations for, 71-72 future of, 189-193 higher-level effects on, 182-183 hybrid approach to, 23-25 implications of, 14-19 individual-level effects on, 180-182 industrial-sectorial perspective on, 1516, 19-22 capital control model of, 54, 55 institutional approach to, 22-23 institutionalization of, 4, 178 long-term, 126-127 membership concept of, 3-4, 57-58, 63, 177 neo-classicial theories of, 16-17 neo-Marxist theories of, 3 organizational factors in, 177 organizational theories of, 3-4 scope of, 180-183 shape of, 178-179 social structure of, 2 technical-rational approach to, 7-19, 54, 55 transaction cost analysis of, 16-19, 54, 58-59, 72, 74, 90 Environment, organizational, 12, 13-14 Exchange transactions, 7-8 Exit-voice, 86 Expectancy theory, 9 External labor market. See Labor market, external Federal bureaucracy/government cultural authority of, 58 job mobility in, 53-54 PACE examination for, 53 public attitudes toward, 192-193 See also Public sector Firm dependence, 127 age factors in, 140, 141 attachment and, 131, 137-138, 140-149, 155, 156, 157 in core industries, 140, 141, 145-146 definition of, 184


Firm dependence (Cont.) earnings and, 166, 167, 169, 170, 172, 173, 174, 176, 189 educational factors in, 140, 141-142 firm size and, 145-146 job training length and, 140, 141, 145146 race factors in, 140, 141 sex factors in, 140, 141-142, 144, 146149 social status and, 140, 141-142, 144 supervisory responsibility and, 140, 141-142 unionization and, 140-141 work control and, 140, 141-142 Firm size attachment and, 132, 145-146, 154 comparison of employer-employee estimations of, 202-204 due process and, 108, 110, 111, 112-114 earnings and, 162-163, 187 job satisfaction and, 150, 151 market opportunity and, 145-146 Firm-specific skills/jobs. See Skill specificity Formalization attachment and, 129 earnings and, 161-162, 166, 167, 169, 170, 176 literacy requirements and, 118, 123 Goals, organizational, 10-11 classwide rationality of, 20 Goal theory, 9 Governance structures basic types of, 17-19 mismatches of, 19 model of, 24 Governance theory, of bureaucracies, 179 Government domination bureaucratic control and, 182, 191 due process access and, 182, 191 job mobility and, 120-121 normative legitimacy of, 182-183 personnel department establishment and, 123 Government intervention due process and, 108, 111-114, 116, 117, 123 internal labor markets and, 105


Grievance procedures unions' association with, 133 See also Due process Hierarchicalization, 55 High-mobility occupations attachment in, 147-149 dependence in, 128 external labor market access in, 102 job satisfaction in, 154 Hiring protection additive scales for, 213-215 age factors in, 75-76, 95, 96, 97, 98-100 attachment and, 125 for black men, 84-85, 99-100 for black women, 84, 85 competition and, 121 decline of, 192 earnings and, 163, 164, 165-166, 167, 168, 170, 175-176, 187, 189 firm size and, 121 industry concentration and, 121 job training and, 91, 93 unionization and, 121 for white men, 77, 181 Historical theories, of bureaucratic governance and due process, 104 Human capital, 17, 18 social resources assessment and, 76-77 Human relations theory, 8 Hybrid approach, in personnel management, 54, 55 Income attachment and, 145-146 job satisfaction and, 150, 151, 152-153 See also Earnings, and employment relationships Industrial governance model, written rules of conduct and, 68 Industrial relations bureaucratic-governance model of, 56-58 system model of, 24-25 Industrial-sectorial approach, to employment relationships, 15-16, 19-22 capital control model of, 54, 55 Industry concentration, job mobility and, 121, 122


Institutional approach to bureaucracies, 179 to due process, 5, 113 organizational-governancemodel of, 54, 55 to organizational structure, 22-23 Institutionalization, of employment relationships, 4, 178 Instrumental linkages, in attachment, 126, 127-128, 183-184 definition of, 130-131 Insurance industry, bureaucratic control institutionalizationin, 186 Job, definition of, 50 Job clemency attachment and, 139, 157-158 job satisfaction and, 153 Job control attachment and, 133-134, 185 job satisfaction and, 185 See also Worker control Job descriptions, written, 61, 63-64 Job discrimination age-related, 177 effect on bureaucratic employment relationships, 180 Job instructions, frequency of use of, 61, 64 Job ladders age factors in access to, 75 attachment and, 125 in public sector, 168, 171, 188 Job mobility additive scales for, 213-215 age factors in, 75-76, 95, 96, 97 of black men, 80, 84, 85 of black women, 80, 81, 84, 101 Current Population Survey data regarding, 50 earnings and, 163, 164, 165-166, 167, 168, 170-171, 176, 187 educational factors in, 90, 91 government domination and, 120-121 industry concentration and, 121, 122 job training and, 91, 93 occupation-specific, 73-74 personnel department access and, 121, 122, 123 of white women, 101


Job mobility (Cont.) of women, 80, 81, 83-84 work experience and, 90, 91, 93 Job satisfaction, 149-154, 155-156 attachment and, 129, 131, 135-136, 138, 149-154, 155, 156, 157 age factors in, 136, 149, 150, 159, 185 bureaucratic employment structures and, 149-150 due process access and, 151 educational factors in, 149, 150, 151 firm size and, 150, 151 in high-mobility occupations, 154 income and, 150, 151, 152-153 job clemency and, 153 job control and, 185 measure of, 138 race factors in, 136, 149, 150, 153, 159, 185 sex factors in, 136, 152-153, 154, 160, 185 unionization and, 149, 150, 151-152, 154 worker control and, 153 working conditions and, 153, 159, 185 bureaucratic employment structures and, 149-140 due process access and, 151 Job search, market opportunity in, 126128 Job-specific skills. See Skill specificity Job training coworker, sex factors in, 94 as instrumental linkage, 183 hiring protection and, 91, 93 internal labor markets and, 91, 93 job mobility and, 91, 93 length of for black men, 78, 79 for black women, 78, 79 firm dependence and, 140, 141, 145146 for white men, 78, 79, 93 for white women, 78, 79, 94 Job turnover union participation and, 132-133 of women, 134 Justice, industrial, 57 unions' association with, 133 workers" security and, 127


Labor market dualism of, 21, 122 efficacy of, 73 segmentation of, 15, 21 Labor market, external attachment and, 131, 158 black men's participation in, 85-86, 94 black workers' participation in, 90 bureaucratic employment relationships and, 180 classification of, 217-222 due process access and, 163-164 employer dependence and, 128 high-mobility occupation membership and, 102 hiring protection and, 163-164 "internal," 73 job mobility and, 163-164 market opportunity perception and, 185-186 occupation-specific, 72-74 relationship to internal labor markets, 95, 102, 181, 183, 184 secondary, 73 white women's participation in, 83 women's attitudes toward, 159 women's participation in, 83-84, 134 in high-mobility versus low-mobility markets, 147-149 workers' awareness of, 140, 141-142, 143 Labor market, internal, 1-2, 82 additive scales for, 213-215 age-related access to, 181-182 age-related job mobility in, 75-76 attachment and, 125-126, 128-129, 131 black men's participation in, 84-86 black women's participation in, 82, 84-85 contingent labor markets versus, 191192 declining need for, 190 due process access and, 119-123 as efficiency pressures response, 20 earnings and, 188 elements of, 213 employment stability rationing in, 5 government intervention in, 105 as instrumental linkage, 183-184 job training and, 91, 93



Labor market, internal (Cont.) Marxist theory of, 56 mechanisms of, frequency of use of, 62-63 occupational, 180 protection versus mobility dimensions of, 68-69 relationship to external labor markets, 95, 102, 181, 183, 184 transaction cost theory of, 59 in white-collar employment, 91, 93 white women's participation in, 101, 102, 181 white men's participation in, 77, 181 women's participation in, 80, 81-84, 101 workers' security and, 127 Literacy due process access and, 109, 118-119, 123 formalization and, 118, 123 Manufacturing industry employer survey data regarding, 47 personnel departments in, 110 Market opportunity, 184 attachment and, 126-128 black workers' perception of, 134, 135, 160 in core industries, 145-146 external labor markets and, 185-186 firm dependence and, 131 firm size and, 145-146 in occupationalized labor markets, 127128 women's access to, 159 women's perception of, 134-135, 141, 144, 146, 185-186 workers' perception of, 127, 140, 141142, 143, 159 Marxist theories of due process, 56 of employer-employee conflict, 104 of internal labor markets, 56 of personnel management, 60, 65 of worker control, 55 Mass-production industries, bureaucratic control institutionalization in, 186187 Membership concept, of employment relations, 3-4, 57-58, 63, 177

Men job satisfaction of, 152-153, 154 See also Black men; White men Metropolitan Employer Worker Survey (MEWS), 29-51 "bottom-up" method of, 29-30 Chicago Standard Metropolitan Statistical Area data of, 30 economic characteristics, 31, 34 establishment distribution characteristics, 33, 35 industry distribution characteristics, 33, 35 labor force characteristics, 32, 34 comparison of employer-employee data, 201-209 employee sample of, 35-41 comparison with 1980 U.S. Census data, 38-41 response rate, 36, 40 employer sample of, 41-50 sources of bias, 44-51 limitations on study population, 30 random-digit dialing method of, 30, 35-36 Microanalytic theory, of bureaucracies, 179 Minority-group workers bureaucratic employment relations of, 180 as contingent workforce, 190-191, 192 due process access of, 181 employers' performance assessment of, 74 See also Black men; Black women Monopoly effect, in worker turnover, 132133 Needs theory, 9 Neo-institutionalist model/theories, 54, 179 New Deal System, 24-25, 189-190 Normative linkages, in attachment, 126, 128-129, 184-185 Obligational market, 18 Occupation, definition of, 50 Occupational mobility measures of, 217-222 See also Job mobility


Occupation-firm tenure ratio (OFTR), 217-218, 219, 220 Occupations, classified as orderly external markets, 217-222 Off-premises workers, 190 Organizational environment, effect on employment relationships, 177 Organizational factors, in employment relationships, 3-4 Organizational field, 22, 23 Organizational-governance model, 54, 56-58 Organizational immortality, 23 Organizational size. See Firm size Organizational structure environmental context, 12, 13-16 firm size and, 12-13 technology's effect on, 11-12 Outside hiring comparison of employer-employee estimations of, 204-206 protection from. See Hiring protection PACE (Professional and Administrative Career Examination), 53 Part-time workers, 190, 191 Performance evaluation, frequency of use of, 61, 63-64 Personnel administration, cultural authority of, 58 Personnel department, 68 access to age factors in, 95, 96, 97, 181, 182 by black men, 89, 90 by black women, 89, 90 bureaucratic control and, 112, 115, 117 due process access and, 112, 115, 117 job mobility and, 121, 122, 123 by white men, 89 by white women, 86, 89 in core industries, 110, 122 prevalence of, 61, 64, 179 firm size and, 110, 111 industry-specific variations in, 109111 role of, 64 Personnel management bureaucratic-governance model of, 56-58


Personnel management (Cont.) capitalist control model of, 54, 55-56, 179 central institutions of, 55 control mechanisms of, 63-64 elements of, 53-70 institutionalist model of, 54 multifaceted nature of, 68 neo-institutionalist model of, 58 organizational control of, 179-180 organizational-governance model of, 54, 56-58 scope of, 60-64 shape of, 66-70 transaction cost model of, 54, 58-59, 62 unified-development theories of, 179 Petroleum industry, bureaucratic control institutionalization in, 187 Primitive team market, 17-18 Private sector earnings in, 166, 167, 168, 169, 170-171 bureaucratic control and, 187-188 due process access and, 169, 189-190 hiring protection and, 175 employer survey data regarding, 47, 49 Productivity, age-related decrease of, 177 Professional and Administrative Career Examination (PACE), 53 Promotion of black workers, 84 bureaucratic practices in, 1-2 as deferred reward, 126 as instrumental linkage mechanism, 126 as satisfaction source, 128-129 of white women, 85 Public sector attachment in, 185 due process access in, 182 earnings in, 166, 167 bureaucratic control and, 187-188 due process access and, 168-169, 189-190 firm dependence and, 169 job ladder access and, 168, 171 job mobility and, 170-171 employer survey data regarding, 47, 48, 49 government rational-legal legitimacy in, 133 job ladders in, 168, 171, 188


Race factors in attachment, 134, 135, 140, 141, 160 in earnings, 173, 174-175 in firm dependence, 140, 141 in job satisfaction, 136, 149, 150, 153, 159, 185 See also Black men; Black women; Minority-group workers; White men; White women Random-digit dialing, 30, 35-36 Rational action perspective, of employment relationships, 7-14 Rationality, bounded, 10-11 Reagan Administration, 193 Relational team market, 18 Retail trade industry, employer survey data, 47, 48 Sample, employee, 35-41 Sample, employer, 41-50 Sample selection bias, 49-50 Sears, Roebuck, and Company, 125 Segmentation, of labor market, 15, 21 Self-actualization, 8 Seniority, in workforce reduction, 4 Seniority systems, codified, 1-2 Sentiments, 10 Service industry employer survey data regarding, 47 personnel departments in, 110 Sex factors in attachment, 134-135, 140, 141, 146149 in earnings, 172-174 in firm dependence, 140, 141-142, 146149 in job satisfaction, 136, 152-153, 154, 160, 185 See also Black men; Black women; Men; White men; White women; Women Skill specificity, 17, 72-73, 83 comparison of employer-employee estimations of, 207-208 of minority-group workers, 75 training costs for, 126 Social resources, 180-181 human capital and, 76-77 of minority-group workers, 74-75 See also Education; Work experience Sodal rewards, in employment, 5-6


Social status, firm dependence and, 140, 141-142, 144 Socio-Economic Index, 134 Spot market, 17, 18 Stratification, 5-6, 55 Subdivision, 55 Supervisory responsibility attachment and, 133-134 firm dependence and, 140, 141-142 in team markets, 18-19 Surplus values distribution, 55 Team market, 18-19 Technical-rational approach, to employment relationships, 7-19, 54, 55 Technology, effect on organizational structure, 11-12 Temporary workers, 190, 191 Tenure, of unionized workers, 133 Training. See Job training Transaction cost analysis, 16-19, 54, 58-59 skill specifidty factor of, 72 employees' productive contribution assessment in, 74 of internal labor market opportunities distribution, 90 Transformed-industrial-relationstheory, 190, 191 Transportation industry, personnel departments in, 10 Uncertainty, environmental, 13-14 Unionization attachment and, 132-133, 140, 141, 154, 185 bureaucratic control and, 114-115, 191 bureaucratic management structures and, 21-22 comparison of employer-employee estimations of, 206-207 in core industries, 21 decline of, 190, 191 due process access and, 108-109, 111, 112-113, 114, 115, 116, 117, 191 due process procedure formalization and, 106 firm dependence and, 140-141 firm size and, 121 hiring protection and, 121


Unionization (Cont.) job satisfaction and, 149, 150, 151-152, 154 tenure and, 133 University of Illinois Survey Research Laboratory, 30, 36-37, 41, 42 Utilities industry bureaucratic control institutionalization in, 187 employer survey data regarding, 47 White-collar employment bureaucratic control in, 94, 171, 173, 174, 175 internal labor market access in, 91, 93 earnings in bureaucratic control and, 171, 173, 174, 175 due process access and, 169, 171-172, 175 firm dependence and, 173, 174 hiring protection and, 173, 174 job mobility and, 171, 173, 174 job training length and, 78, 79, 94 women's access to, 78, 79, 91, 93 White men bureaucratic employment relationships of, 180-181 educational levels of, 78, 79 external labor market participation by, 77 hiring protection for, 77, 181 internal labor market participation by, 77, 181 job mobility of, 77 age factors in, 98 job training length for, 78, 79, 93 personnel department access by, 89 postschool training opportunities of, 77 White women hiring protection for, 99 internal labor market participation by, 101, 102, 181 job mobility of, 80, 81, 82, 83 personnel department access by, 86, 89 promotion of, 85 white-collar employment participation by, 78-79


Wholesale trade industry, employer survey data, 47, 48 Women bureaucratic employment relationships of, 180 career interruptions of, 148-149, 159160, 185 as contingent workforce, 190-191, 192 due process access of, 86, 87, 181 earnings advantages of, 162 firm dependence of, 146-149 internal labor market partidpation by, 80, 81-84 job mobility of, 80, 81, 83-84 job satisfaction of, 152-153, 154 job turnover oL 134 market opportunity perceptions of, 134135, 141, 144, 146, 185-186 white-collar employment participation by, 78-79, 91, 93 See also Black women; White women Worker control employer-employee conflict in, 19-20 firm dependence and, 140, 141-142 job satisfaction and, 153, 159, 185 See also Bureaucratic control Work experience age factors in, 96-98 earnings and, 166 job mobility and, 90, 91, 93 Marxist theory of, 55 personnel structures and, 76-77 as social resource, I81 of white women, 78 Workforce reduction government-mandated notification of, 1 seniority in, 4 Working conditions attachment and, 139 bureaucratic control and, 109, 118, 119, 123 due process and, 109, 118, 119, 123 job satisfaction and, 153, 159, 185 Work-relation separability, 17 Work rules, as due process indicators, 211-213




Series Editors: Ivar Berg, University of Pennsylvania, Philadelphia, Pennsylvania a n d A r n e L. Kalleberg, University of North Carolina, Chapel Hill, North Carolina

WORK AND INDUSTRY Structures, Markets, and Processes Arne L. Kalleberg and Ivar Berg WORKERS, MANAGERS, AND TECHNOLOGICAL CHANGE Emerging Patterns of Labor Relations Edited by Daniel B. Cornfield INDUSTRIES, FIRMS, AND JOBS Sociological and Economic Approaches Edited by George Farkas and Paula England MATERNAL EMPLOYMENT AND CHILDREN'S DEVELOPMENT Longitudinal Research Edited by Adele Eskeles Gottfried and Allen W. Gottfried ENSURING MINORITY SUCCESS IN CORPORATE MANAGEMENT Edited by Donna E. Thompson and Nancy DiTomaso THE STATE AND THE LABOR MARKET Edited by Samuel Rosenberg THE BUREAUCRATIC LABOR MARKET The Case of the Federal Civil Service Thomas A. DiPrete ENRICHING BUSINESS ETHICS Edited by Clarence C. Walton LIFE AND DEATH AT WORK Industrial Accidents as a Case of Socially Produced Error Tom Dwyer WHEN STRIKES MAKE SENSE--AND WHY Lessons from Third Republic French Coal Miners Samuel Cohn THE EMPLOYMENT RELATIONSHIP Causes and Consequences of Modern Personnel Administration William P. Bridges and Wayne J. Villemez LABOR A N D POLITICS IN THE U.S. POSTAL SERVICE Vern K. Baxter