Socio-Economic Review (2009) 7, 695–726 doi:10.1093/ser/mwp008 Advance Access publication June 16, 2009

DISCUSSION: A new labour economics? Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021

At the 2008 SASE meeting in San Jose´ , Costa Rica, David Marsden organized a session on the prospects for a renewed institutional labour economics. The debate began with introductory remarks by Paul Osterman, who sketched out an argument that at the time was still in its very early stages. The introduction and the subsequent comments were found by the audience to be highly productive. After the session, the editors of Socio-Economic Review asked the participants to share their views with the readers of the journal. We are grateful to Paul Osterman for taking up the challenge and summarizing the state of his thinking in a brief draft of what has yet to be developed into a formal paper. We also thank the discussants who agreed to write up their comments on the basis of Osterman’s intermediate draft. Keywords: employment, labor economics, labor market institutions, organiz- ations, personnel management JEL classification: J01 labor economics

The contours of institutional labour economics: notes towards a revived discipline

Paul Osterman

Sloan School of Management, Massachusetts Institute of Technology, Cambridge, MA, USA

Correspondence: [email protected]

Labour Economics has increasingly become an extension of standard econ- omic theory. This tendency is exemplified in the so-called New Personnel Economics (NPE), which seeks to explain the personnel practices of firms [their internal labour markets (ILMs)] from an optimization perspective (Lazear and Shaw, 2007). However, while this view has gone largely

# The Author 2009. Published by Oxford University Press and the Society for the Advancement of Socio-Economics. All rights reserved. For Permissions, please email: [email protected] 696 Discussion unchallenged within the profession, there are important reasons to regard it as incomplete in explaining the employment practices of organizations and, in particular, in explaining the current evolution of these practices. An alternative perspective draws from an older tradition of institutional labour economics (ILE) as well as from recent developments in organizational sociology and political science. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 The goal of this paper is to sketch the outlines of this alternative view. In the next section, I describe the main features of the debate and, in subsequent sections, I illustrate the argument with evidence regarding the rise of ILMs, the spread of new forms of work organization, efforts by firms to obtain employee commitment, and the growing diversity of employment practices. I then discuss whether the arguments developed here represent merely transitory deviations from standard theory and conclude with some final observations.

The debate Mainstream labour economics does study institutions, such as unions or the minimum wage or temporary help firms, but the perspective is to view them as external constraints on a firm which is optimizing a well-defined objective function. In contrast, the goal of this paper is to introduce greater reality into our understanding of how firms make decisions regarding their employment rules. The core argument developed here is that a fruitful approach to an institutional analysis is to focus on the importance of groups inside of organizations, the persistence of conflict of interest among these groups and the role played by power and politics in working through these conflicts. It is this group level process that creates the employment prac- tices that we are seeking to explain. In emphasizing these factors, the argument shifts attention away from the optimizing models that NPE uses to explain employment practices while at the same time acknowledging that the impulses central to the NPE models are one of the multiple factors at play and need to be given weight. One possible critique of NPE might be termed the socio-psychological per- spective. The standard model typically sees economic actors as atomistic and selfish utility maximizers whereas social-psychologists emphasize interpersonal comparisons of utility and alternative sources of motivation such as gift exchange. It is easy to show (Pfeffer, 2007; Baron and Kreps, forthcoming) that the behavioural assumptions inherent in NPE models are incomplete and shallow and lead to misunderstandings about how people actually behave at work and about what motivates them. However, as useful as this socio-psychological viewpoint is, it is also important to see that the discussion A new labour economics 697 and examples centre on such individual level behaviours as the role of intrinsic motivation, the importance of reciprocity and gift exchange, status inconsis- tency, the anchoring of expectations and interpersonal comparisons of utility. That is, the entire discussion remains almost entirely at the level of individual actors. The other end of the spectrum from the social-psychological view is the point Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 that the labour market (and other markets) is embedded in larger cultural, legal and institutional systems. This perspective is exemplified in the substantial pol- itical science literature on ‘varieties of capitalism’, which argues that alternative equilibria are possible and that there is no single ‘most efficient’ outcome (Katz and Darbishire, 1999; Hall and Soskice, 2001; Streeck and Yamumara 2001; Thelen, 2004). For example, Japan, Germany and the USA are all successful market systems, but they show considerable variation in how they have responded to economic shocks and the impact technological change has had on the wage distribution. This contrasts not just with standard theory but also to the older institutional view that industrial societies were converging towards a single model (Kerr et al., 1960). One review of the varieties of capitalism literature concludes that the consensus among scholars that there is a persistent diversity across nations in how capitalism is organized is ‘truly striking’ (Thelen, 2004, p. 1). The implication of these national effects is that in order to understand why a given ILM practice is adopted it may be important to pay close attention to the national context. For example, research by Marsden and Belfield (2008) shows that the use of performance pay differs between Britain and France, and the difference can be attributed to the impact of variation in employee protection laws across the two countries. National effects are important because, in addition to having a direct impact upon ILM practices, they also shape the relative power and resources of different groups within firms. Although, in one sense, national effects are a challenge to NPE—because their existence demonstrates that context matters and that ILM practices cannot be explained via timeless and spaceless optimization models— a more sophisticated version of NPE can deal with national effects by simply treating them as an additional constraint on firm decision-making. It is also worth noting that a weakness in the national effects literature is its tendency to view nations as internally homogeneous in their practices while the reality, as I will show below, is that there is substantial diversity within countries in ILM rules. It is this diversity that I seek to better understand.

The alternative perspectives NPE scholars identify a personnel practice that is empirically prevalent and develop a model in which the practice in question emerges as the most efficient 698 Discussion solution to the problem posed in the model. Added to this is the standard economics equilibrium assumption that if organizations err and adopt sub- optimal practices they will lose out in the competitive market. The institutional perspective offers a more complex view of organizational decision-making and behaviour. There have, of course, been a number of rela- tively recent important contributions that have sought to elaborate the Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 meaning of ILE and to create space for this school of thought (Jacoby, 1990; Kaufman, 1994, 2004; Marsden, 1999), and the goal of this paper is to add to this literature. The interpretation of institutional thinking developed here is that within organizations there are competing rationales as well as competing fac- tions with distinct interests. The decisions organizations make about their ILM practices flow from a complex interaction of rationales and interests that includes an important role for factors such as norms and customs, social structure, com- peting interests, search for legitimization and power. Only by understanding this mixture can we explain why a given practice emerged and took the form it did and, more importantly, how it is likely to evolve and adapt to shifts in the environment. The view of organizations as an arena in which different visions and power configurations are played out has a distinguished provenance in sociology and political science. Nearly a century ago John Commons wrote that ‘ ...economic conflicts are not merely conflicts between individuals ...but are conflicts between classifications or even classes of individuals ...’ (quoted in Parsons, 1963). In his description of a French bureaucratic organization Crozier wrote of ‘opposing forms of rationality’ (p. 298) and went on to observe a ‘complex equilibrium affecting the patterns of action, the power relationships, and the basic personality traits characteristic of the cultural and the institutional systems of a given society’ (p. 294). Gouldner took a similar position in his classic book Patterns of Industrial Bureaucracy (Gouldner, 1954). He examined, via a case study of one firm, the origins and implementation of bureaucratic rules regarding issues such as attendance, safety, bidding for jobs and effort. As a result, the firm’s personnel practices, i.e. ILM, reflected a political and social process that resolved these conflicts. Gouldner wore that ‘Rather than assuming that bureaucracy possesses a Gilbrator-like stability, the perspective directs atten- tion into the tensions and problems evoked by bureaucratization’ (p. 11). The NPE use of principal–agent theory does not address these issues because it is about how one designs incentives for individuals. It is not about group be- haviour, competing conceptions of purpose and collective interest in organiz- ations. Furthermore, agency theory implicitly assumes that there is a decision-maker with power to impose an incentive compatible solution, but in fact, decisions in organizations are the outcome of a political bargaining process among the groups as described above. The so-called decision-maker A new labour economics 699 cannot act unilaterally. Once groups are introduced and bargaining among groups, implicit or explicit, becomes the explanation for how ILM rules are estab- lished, we have departed from the standard world of principal–agent theory.1

Illustrating the argument Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 In this section, I will illustrate the argument by first discussing the emergence of ILMs and then discussing the more recent evolution of ILMs. With ILMs, employment within firms is governed by a set of administrative rules that, at least at first glance, are only loosely connected to the external market. The NPE perspective does offer insights into the emergence of ILMs since pressures for ILMs emerged in part from the efforts of firms to address turn- over and incentives concerns. Turnover is costly because it entails a waste of human capital and because in a high turnover organization, provision of and investment in on-the-job training is problematic. To reduce turnover, compensation is back loaded, and this leads to a system in which people spend long periods of time with the same employer. This effect is intensified by the nature of human capital acquisition. The early investments in training need to be rewarded over time, from the perspective of both the employees and the firm, and again this implies long careers. The initial emergence of ILMs does provide some support for this view. At the turn of the twentieth century, American firms, even large ones, did not have well- developed ILMs. Instead, they managed their workforce via the so-called ‘drive system’ in which substantial discretionary (often arbitrary) power was put in the hands of the foreman (Jacoby, 2004). Turnover was high and was problematic during periods of growth and labour shortage, and this led to efforts to stabilize the workforce via the creation of job ladder personnel systems. Adoption of these practices was uneven but does seem to be correlated with NPE-like considerations. With the onset of the Depression, the storyline shifts. In sectors that were unionized, pressures built to establish the formal job ladders and seniority

1There is an NPE literature on organizational politics, but it remains at the level of individuals and how they maximize their self-interest rather than at the level of group political contests over rules. For example, Lazear (1989) examines whether wage compression is good or bad for efficiency, given that it may enhance cooperation but perhaps also reduce incentives for the best employees. However, his analysis asks how individuals react with respect to their private self-interest to different wage distributions and he maintains the view that there is a decision maker within the firm who can set the wage structure based on which is the most optimal. In another effort along these lines, Prendergast and Topel (1996) model the behaviour of supervisors who exhibit favouritism towards particular employees. But, again, the modelling is purely at the individual level. Supervisors simply ‘like’ some workers and not others; there is no sense of group action or interest. 700 Discussion systems that characterize industrial ILMs. Non-union firms imitated these systems in order to avoid unionization. The Federal Government, via the War Labor Board, imposed standardized personnel practices on Federal contractors in order to maintain labour peace. In addition, personnel specialists within firms were very active in pushing for the formalization of employment practices, a point of view that was clearly in their self-interest. Indeed, during the Depression Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 and the World War II periods, there was very little correlation between the stan- dard optimizing correlates of ILMs—firms size and turnover rates—and the adop- tion of specific formal personnel practices (Baron et al., 1986). The conclusion to be drawn from this material is that when it comes to under- standing how the prototypical ILM of the post-war period emerged, the story is considerably complex than implied in NPE models, and the complexity is consist- ent with the institutional perspective developed above. There is certainly con- siderable room for the play of market forces and efficiency-maximizing considerations, but to these were joined the actions of two important interest groups within firms, unions and personnel professionals, as well as the actions of an external institution, the government. ILMs can only be understood as resulting from the interaction of these actors as they used their resources to pressure organizations in the directions that they preferred.

The slow diffusion of high-performance work systems New strategies for organizing work, which often involve the use of teams and various forms of quality programs, have been implemented in a range of firms (Osterman, 1994, 2000; Handel and Gittleman, 2004) because there is consider- able evidence that they lead to higher levels of quality and productivity (Kochan and Osterman, 1994; MacDuffie, 1995; Ichniowski et al., 1997; Pfeffer, 2007). Yet despite this evidence, adoption has been slow and the question is why. Institutional explanations for slow diffusion find support in survey-based research. For example, in their survey of the adoption of HPWO practices across a sample of automobile firms, Pil and MacDuffie (1996) find that a set of variables capturing the presence or absence of internal organizational barriers and resistance perform better than simple measures economic performance. In a cross-national survey, Bloom and Van Reenen (2007) show that ownership pat- terns play an important role in explaining differences in the adoption of modern managerial practices, a finding that is consistent with explanations that emphasize the balance of power considerations within firms. In his survey of establishments, Osterman (1994) found that a variable capturing employee- friendly values held by managers was significant in explaining HPWO diffusion. Case studies lead to a similar conclusion. Among the patterns in the literature are that first line supervisors resist because they feel threatened by the role of A new labour economics 701 self-managed teams in taking over some of their functions (Coyle-Shapiro, 1999), unions can be oppositional when their adversary function is threatened by the more cooperative nature of HPWO systems (Walton et al., 1994) and, as Pfeffer (2007) also notes, the growing power of the finance function within the firm puts heavy weight on immediately measurable costs and less weight on more abstract or hard to measure future benefits. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 In short, both surveys and case studies support the view that the adoption of HPWO systems is strongly affected by group pressures and politics within organizations. Whether and how a firm adopts these practices depends on much more than simply an optimizing decision by senior managers.

Effort and commitment Firms today are increasingly finding that their competitive success is due to high levels of employee commitment to the organization. This is because of the channel between employee commitment and product quality and customer sat- isfaction. In NPE, commitment is obtained via efficiency wages and incentive compensation models that are grounded in principal–agent incentive theory. These perspectives are somewhat different in that efficiency wage/shirking models are based on the level of effort, whereas in principal–agent models the worry is that the effort may be in the wrong direction, i.e. for the self-interest of the agent rather than the principal. However, what both have in common is the view that effort is largely decided at the level of the individual and that the key to any solution is to get that individual’s incentives right. The institutional perspective emphasizes group processes, organizational structure and culture. At the core of the institutional perspective is the idea that the level of effort delivered by the workforce is at least partially determined at the group level, i.e. by considerations other than individual incentive schemes. Consider Toyota or Southwest Airlines. These firms are tremendously successful in their industries, and both are widely known for obtaining high levels of effort from their workforce. They are also successful in obtaining organizational com- mitment, an idea that is basically foreign to NPE thinking (Lincoln and Kalleberg, 1990). But, in contrast to NPE expectations, both are quite conventional in their incentive systems (and this is as true of Toyota of America as it is of Toyota of Japan) and in both organizations seniority (frequently seen in the NPE literature as the enemy of meritocracy) is important. These organizations are able to obtain exceptional individual effort and organizational commitment. The explanation lies in a set of factors that institutionalists emphasize: group norms and culture. In thinking about this it is important to recognize that norms and culture are group phenomena and not examples of the kind of individual level social–psychological exchange discussed earlier. Indeed, there is a long tradition 702 Discussion in institutional research, for example in the literature on production norms and a fair day’s work for a fair day’s pay, of recognizing that group culture and norms are important. It is clear to observers that, at the level of the workgroup, each organization has succeeded in obtaining effort via broadly shared norms and not simply via individual incentives or individual gift-exchange. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021

Diversity of employment systems We are increasingly observing considerable variety in ILM systems among simi- larly situated organizations. Yet, if firms that face similar external constraints have different ILM practices, then this diversity poses a problem for NPE models, which would posit a single optimal response. In contrast, the diversity can be understood from an institutional perspective that emphasizes different internal political configurations across firms. A range of evidence demonstrates diversity. At the case study level are examples such as the contrast in employment practices between Wal-Mart and Costco (Greenhouse, 2005), the divergence of Southwest Airlines from the prac- tices of other large carriers, and the HRM practices of the software firm SAS compared with its competitors (Pfeffer, 1998). More systematic work within industries, such as automobiles and telecommunications, also points to the same conclusion. For example, in their study of telecommunications and automobiles, Katz and Darbishire (1999, p. 4) report that ‘within both union and non-union sectors the extent of variation in wages, work practices, and other employment conditions has increased’. A similar finding regarding substantial diversity emerges from research in a very different set of firms: small, young, non-union start-ups in the Silicon Valley (Baron et al., 1999; Burton, 2001). The study followed a sample of start-ups and collected data on their personnel practices with respect to hiring, promotion, compensation and retention. Firms varied in their approach towards motivation and retention, and the researchers identified five HR models, each with a different set of personnel practices. Broader survey research also finds substantial diversity. Osterman (1994, 2000) surveyed roughly 700 establishments and, after holding constant a wide range of characteristics, found considerable variation in the adoption of teams, job rotation and training practices. A similar pattern was observed by Bloom and Van Reenen (2007) in their survey of firms. In short, there is considerable evidence of diversity in ILM practices across similarly situated firms. Of course, the evidence that the firms face exactly the same set of external constraints is not ironclad, and if constraints are different then variation is not a problem for NPE models. The survey-based research does hold constant a substantial number of variables as does the Silicon Valley A new labour economics 703 work that is focused on a narrow and relatively homogenous sector within the same geographic area. The case study work is more subject to this concern. Nonetheless, when the wide range of evidence is considered it does seem reason- able to conclude that real diversity, holding constant constraints, is a fact of organizational life. Because there is not a unified literature on diversity, each of the different Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 studies offers its own set of explanations. The Silicon Valley research focused upon the values of the founders of the firms and how those founders enforced their values and imprinted their organization. Bloom and Van Reenen emphasize the importance of ownership patterns and in particular whether the firm is family controlled. The relative power of different internal constituencies—unions, per- sonnel professionals, the finance function—show up as key variables in much of the literature. Taken as whole, while recognizing that the definitive test has not been undertaken, a fair reading of the literature would be that an important component of the very substantial observed diversity can only be understood as flowing from institutional considerations.

Objections A likely NPE response to the foregoing is to acknowledge some divergences from the expectations of the standard model but argue that they are only short-term phenomena and that the market will quickly reward optimal behaviours and punish sub-optimal ones. However, while mainstream economics literature does assume convergence to optimal practice, there is in fact an alternative vein of theory that points away from this view. Nelson and Winter (1982) argue that standard theory ‘cannot address such questions as the duration of the struggle or the durability of mistakes made in the course of it’ (Nelson and Winter, 1982, p. 32) and they go on to describe what is essentially a search model for adoption of practices which is ‘gaping, disorderly, and error ridden’ with ‘general habits and strategic orientations coming from the firm’s past’ (pp. vii, viii). Their explanation is different than that of the current paper in that they focus on the nature of the managerial decision-making process while this paper emphasizes the importance of politics among groups. Nonetheless, the central point of importance here is the rejection of the view that market forces inevitably force out practices that seem sub-optimal from the standard perspective. The literature on increasing returns makes a similar point. Arthur (1989) argues along these lines with the idea being that that with increasing returns, a technology that is adopted gets advantages relative to a competing technology because it improves more quickly. Because of this, a random event or accident can set the adoption on a particular path whereas with a different event or 704 Discussion accident, a different path and different technology would win. As a result he shows that while ‘Under constant and diminishing returns the evolution of the market reflects only a-priori endowments, preferences, and transformation possibilities .... But while this is comforting it reduces history to the status of a mere carrier, the deliverer of the inevitable. Under increasing returns, by contrast, many outcomes are possible’ (p. 127) and ‘competition between Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 economic objects ...takes on an evolutionary character ...with a “founder” effect mechanism ...history becomes important ...’ (p. 128). This argument about ‘lock-in’ has a parallel in the political science literature that examines the adoption of institutions as well as the sociology isomorphism literature. The general point is that the existing institutions take on normative or customary legitimacy and thus are hard to change (DiMaggio and Powell, 1983; Mahoney, 2000; Thelen, 2004). In addition, Stinchcomb long ago argued that institutions are ‘imprinted’ by the environment in which they are founded and are often unable to adapt to new conditions (Stinchcomb, 1965). The population ecology literature in sociology has taken off from this insight to argue that the source of change in response to environmental shifts is less the reconfiguration of the existing institutions but rather the gradual rise of new organizations (Hannan and Freeman, 1984). There is extensive empirical research in the literatures on technology, in popu- lation ecology and political science that documents these arguments concerning persistence and lock-in. The literature in ILE has not been as systematic, but there are a good deal of scattered observations that support it. There is, for example, a long history of empirical work that demonstrates the power of customs and norms in the workplace, and these customs and norms clearly serve to lock past practices into place (Dunlop, 1958). The earlier observation about the slow diffusion of new high-performance work systems is also evidence of how slow change can come. Numerous business school cases focus on firms that were slow to adapt due to the weight of past practice. It is not that the market has not had an impact—it certainly has—but one reasonable conclusion is that the discipline of the market, the pressures to eliminate ‘sub-optimal’ forms, is con- siderably weaker than often assumed. Hence, any effort to understand the current landscape is radically incomplete if viewed simply through an optimizing lens.

Conclusion The argument of this paper is pitched at the level of the organization. It does not address the psychology of individuals (i.e. whether or not they are utility maxi- mizers) nor does it address the operation of external markets. The creation of employment rules within organizations is the terrain on which I think that ILE has the most purchase. It is important in this context to understand that even at A new labour economics 705 the level of the organization, the approach proposed here does not reject market forces. Rather, the goal is to develop a more sophisticated understanding of what determines employment systems. The central proposition, one which as noted has a long tradition in the literature, is that these systems are the result of a political process in which competing objectives and rationalities play out a contest. Market or efficiency considerations are certainly one player in this contest, and in the examples Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 I presented I made an effort to give them their due. But they are not the only player. The arguments developed here are clearly related to the sociological literature on embeddedness and to both the sociological and political science literatures on institutionalization; however, these literatures have thus far not penetrated very deeply into labour economics discussions. At the same time, this paper ascribes a more important role to market forces as a factor in the design of ILMs and also places more emphasis on the idea of competing rationalities and balance of power than is found in those literatures. At the same time, the argument devel- oped here does leave several important questions unanswered. I noted earlier that national institutions can affect the balance of power of different groups within firms and thus impact outcomes, but this paper does not pursue this point. In addition, the emphasis on group processes and politics raises the question of how groups are formed and re-reformed. This, of course, goes to the issue of how people and collectivities come to identify themselves as having commonal- ities and how they frame their self-interest. This is a topic of growing interest but it would take this paper too far afield to enter into it.

Acknowledgements The idea for this paper grew from a conference presentation at University Carlos III in Madrid, and I am grateful to Isabel Calderon Gutierrez for suggesting the idea and for having invited me. I have received useful comments and suggestions from Lucio Baccaro, Frank Dobbin, Sandy Jacoby, Thomas Kochan, Mike Piore, and participants in my session at the summer 2008 SASE meetings.

References Arthur, W. B. (1989) ‘Competing Technologies, Increasing Returns, and Lock-in by His- torical Events’, Economic Journal, 99, 116–131. Baron, J. N., Dobbin, F. and Devereaux Jennings, P. (1986) ‘War and Peace: the Evolution of Modern Personnel Administration in U.S. Industry’, American Journal of Sociology, 92, 350–383. Baron, J., Burton, M. D. and Hannan, M. (1999) ‘Engineering Bureaucracy: The Genesis of Formal Policies, Positions, and Structures in High-Technology Firms’, Journal of Law, Economics, and Organization, 15, 1–41. 706 Discussion

Baron, J. and Kreps, D. (2009, forthcoming) ‘Employment as an Economic and a Social Relationship’. In Gibbons, R. and Roberts, J. (eds) Handbook of Organizational Econ- omics, Princeton, NJ, Princeton University Press. Bloom, N. and Van Reenen, J. (2007) ‘Measuring and Explaining Management Practices Across Firms and Countries’, Quarterly Journal of Economics, CXXII, 1351–1408. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Burton, M. D. (2001) ‘The Company They Keep: Founders’ Models for Organizing New Firms’. In Schoonhoven, C. B. and Romanelli, E. (eds) The Entrepreneurship Dynamic: Origins of Entrepreneurship and the Evolution of Industries, Stanford, CA, Stanford University Press, pp. 13–39. Coyle-Shapiro, J. A.-M. (1999) ‘Employee Participation and Assessment of an Organiz- ational Change Intervention: a Three Wave Study of Total Quality Management’, The Journal of Applied Behavioral Science, 35, 439–456. DiMaggio, P. J. and Powell, W. W. (1983) ‘The Iron Cage Revisited: Institutional Iso- morphism and Collective Rationality in Organizational Field’, American Sociological Review, 48, 147–160. Doeringer, P. B. and Piore, M. J. (1971) Internal Labor Markets and Manpower Analysis, 2nd edn, Lexington, MA, Heath. Dunlop, J. (1958) Industrial Relations Systems, New York, Henry Holt. Gouldner, A. (1954) Patterns of Industrial Bureaucracy, New York, Free Press. Greenhouse, S. (2005) ‘How Costco Became the Anti-Wal-Mart’, New York Times, Section 3, p. 1. Hall, P. and Soskice, D. (eds) (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, New York, Oxford University Press. Handel, M. and Gittleman, M. (2004) ‘Is There a Wage Payoff to Innovative Work Prac- tices?’, Industrial Relations, 43, 67–98. Hannan, M. and Freeman, J. (1984) ‘Structural Inertia and Organizational Change’, American Sociological Review, 49, 149–164. Ichniowski, C., Shaw, K. and Prennushi, G. (1997) ‘The Effects of Human Resource Man- agement Practices On Productivity: A Study of Steel Finishing Lines’, American Econ- omic Review, 86, 291–313. Jacoby, S. (1990) ‘The New Institutionalism: What Can It Learn From The Old?’, Industrial Relations, 29, 316–340. Jacoby, S. (2004) Employing Bureaucracy, Mahwah, NJ, Lawrence Erlbaum. Katz, H. and Darbishire, O. (1999) Converging Divergences, Ithaca, NY, Cornell University Press. Kaufman, B. (1994) ‘The Evolution of Thought on The Competitive Nature of Labor Markets’. In Kerr, C. (ed) Labor Economics and Industrial Relations, Cambridge, MA, Harvard University Press, pp. 145–188. A new labour economics 707

Kaufman, B. (2004) ‘The Institutional and Neoclassical Schools in Labor Economics’. In Champlin, D. and Knoedler, J. (eds) The Institutionalist Tradition in Labor Economics, New York, Sharp, pp. 13–38. Kerr, C., Dunlop, J. T., Harbison, F. H. and Myers, C. A. (1960) Industrialism and Industrial Man, Cambridge, MA, Harvard University Press. Kochan, T. and Osterman, P. (1994) The Mutual Gains Enterprise, Cambridge, MA, Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Harvard Business School Press. Lazear, E. (1989) ‘Pay Equality and Industrial Politics’, Journal of Political Economy, 97, 561–580. Lazear, E. and Shaw, K. (2007) ‘Personnel Economics: The Economist’s View of Human Resources’, Journal of Economic Perspectives, 21, 91–114. Lincoln, J. and Kalleberg, A. (1990) Control, and Commitment: a Study of Work Organiz- ation and Work Attitudes in The United States and Japan, Cambridge, UK and New York, Cambridge University Press. MacDuffie, J. P. (1995) ‘Human Resource Bundles and Manufacturing Performance: Organizational Logic and Flexible Production Systems in the World Auto Industry’, Industrial and Labor Relations Review, 48, 197–221. Mahoney, J. (2000) ‘Path Dependence in Historical Sociology’, Theory and Society, 29, 507–548. Marsden, D. (1999) A Theory of Employment Systems, London, Oxford University Press. Marsden, D. and Belfield, R. (2009) ‘Institutions and the Management of Human Resources: Incentive Pay Systems in France and Great Britain’, mimeo, London School of Economics. Nelson, R. and Winter, S. (1982) An Evolutionary Theory of Economic Change, Cambridge, MA, Harvard University Press. Osterman, P. (1994) ‘How Common Is Workplace Transformation and How Can We Explain Who Adopts It? Evidence from a National Survey’, Industrial and Labor Relations Review, 47, 173–188. Osterman, P. (2000) ‘Work Organization in an Era of Restructuring: Trends in Diffusion and Impacts on Employee Welfare’, Industrial and Labor Relations Review, 53, 179–196. Parsons, K. H. (1963) ‘The Basis of Commons’ Progressive Approach to Public Policy’. In Somers, G. (ed) Labor Management and Social Policy, Madison, WI, University of Wisconsin Press, pp. 3–24. Pfeffer, J. (1998) The Human Equation: Building Profits by Putting People First, Boston, Harvard Business School Press. Pfeffer, J. (2007) ‘Human Resources from an Organizational Behavior Perspective: Some Paradoxes Explained’, Journal of Economic Perspectives, 21, 115–134. Pierson, F. (ed) (1957) New Concepts in Wage Determination, New York, McGraw Hill. Pil, F. and Macduffie, J.-P. (1996) ‘The Adoption of High Involvement Work Practices’, Industrial Relations, 35, 423–456. 708 Discussion

Prendergast, C. and Topel, R. (1996) ‘Favoritism in Organizations’, Journal of Political Economy, 1, 958–978. Streek, W. and Yamamura, K. (eds) (2001) The Origins of Neoliberal Capitalism; Germany and Japan, Ithaca, NY, Cornell University Press. Thelen, K. (2004) How Institutions Evolve; The Political Economy of Skills in Germany, Britain, the United States, and Japan, Cambridge, Cambridge University Press. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Walton, R. E., Cutcher-Gerschenfeld, J. E. and McKersie, R. B. (1994) Strategic Nego- tiations: A Theory of Change in Labor–Management Relations, Boston, MA, Harvard Business School Press.

Towards a new institutional labour economics?

Peter Auer

ILO, Geneva, Switzerland

Correspondence: [email protected]

Paul Osterman’s paper is important and thought provoking, and it will probably signal the beginning of a longer debate about theoretical assumptions and methods of ILE versus those of (standard) economic theory as applied to organ- izations by NPE. Hopefully, it will also contribute to the redefinition of the dis- cipline and its (re)casting as an important approach to the study of labour markets. It may well be that the present crisis will lead to a more careful and weighted approach towards rational maximization of utility and efficiency; the consideration of possible conflicts between short-term maximization and long- term optimization in organizations; and the rediscovery of the value of smart regulations and institutions.

A matter of diverging assumptions and methods? The crux of Paul Osterman’s critique is that NPE assumes that individuals and firms are efficiency or utility maximizers, and that NPE is void of any concerns for group processes, organizational structure or culture (Ostermann, p. 11) or, as Je´roˆme Gautie´ puts it in his comment, for history and contingency, and A new labour economics 709 thus is more or less void of institutions.2 Indeed, reading, for example, through Lazear and Shaw’s (2007) arguments, power relationships, hierarchies, interest representation, national diversity and so on seem to be absent or, at best, to have a marginal role in the generalization of the rational efficiency/utility max- imizing principle. In addition to being concerned with these assumptions, there is also a (less Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 explicit) concern about methods, in particular the use of formalization tech- niques and the use of general principles rather than empirical facts. Narrow general assumptions about behaviour and strict but reduced models unduly sim- plify the real complexity of the object of study. On a more general level, this debate has antecedents in the debates on the differences or boundaries between sociology and economics. A good account of these debates can be found in Richard Swedberg’s (1990) Economics and Sociology: Redefining their boundaries. Conversations with Economists and Sociol- ogists. He cites, for example, Paul Samuelson, who repeats Pareto’s opinion that economics is about rational behaviour and sociology about irrational behav- iour; he notes Talcott Parsons’ view on ‘Economic Imperialism’ and the straightjacket in which it puts contiguous but more complex sciences like soci- ology; and he cites Knight’s angry reply to Parsons in which he states that ‘Sociology is the science of talk, and there is only one law in sociology. Bad talk drives out good talk’ (Knight as cited by Samuelson, 1983, p. 161, in Swedberg, 1990, p. 15). The economist James Duesenberry (1960, p. 233) once made the comment that ‘economics is all about how people make choices; sociology is all about how they don’t have any choices to make’ referring to the social determination of choices; and Olivier Williamson contested that unfamiliar business practices and organizational forms3 operate in the service of price discrimination, barriers to entry and risk aversion, but also in the service of lowering transaction costs (Williamson, 1996). While these debates on sociology versus economics do not exactly overlap with the debate on the differences between ILE and NEP (which are, in principle, two subfields within economics, although NEP may not see it that way), there are indeed similarities between the former and the latter. Lazear (1993, p. 3) writes

2Lazear and Shaw (2007) seem to acknowledge institutions but argue that while ‘the typical textbook on human resource management would often eschew generalization, arguing that each situation is different’, economists would use ‘specific institutional details to identify the causal sources of the general principles’, a premium being put on the identification of the latter (p. 3). 3Such as hierarchies, which also develop internal labour markets. An internal labour market is ‘an administrative unit, such as a manufacturing plant, within which the pricing and allocation of labor is governed by a set of administrative rules and procedures’ (Doeringer and Piore, 1971). 710 Discussion that ‘Traditional issues in personnel have generally been analysed by industrial psychologists or sociologists. While their work has produced insights, some of the problems are inherently economic’. Thus, he puts into opposition sociology and economics, but he could have mentioned ILE or Industrial and Labour Relations as well. This seems particularly annoying as NEP is concerned with some of the same topics as ILE, Industrial and Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Labour Relations and other more sociologically connected fields: compensation, hiring practices, training and teamwork (Lazear and Shaw, 2007). While Lazear and Shaw (2007, p. 110) stay vague on the boundary between their field of study and other disciplines in the area of the study of work and employment (which includes sociology, human resource management and ‘those who studied organizational behaviour’, but always stressing their non-economic approach), they give an overtly pragmatic explanation of why they think NPE is a positive addition to ‘traditional methods of studying human resources management’: NPE will increase the attractiveness of labour economics for students of business schools because it uses the degree of forma- lization that economists are accustomed to.4 Indeed, as I understand Paul Osterman’s paper, his critique of NEP is also one of economic and formalization imperialism that would make the discipline of labour economics subject to impoverishment and put it in the straightjacket of the assumptions that both firms and workers are rational maximizing agents which seek utility and profit and that ‘efficiency is a central concept of personnel economics’ (Lazear and Shaw, 2007, p. 92). As for the dispute: there is indeed an understandable dislike of the attempt to construct standard economic theory and its methodological tool-kit as the superior discipline for studying labour markets, in general, and those internal to firms, in particular. The main criticism remains that NPE is a ‘reductionist’, apolitical and ahistorical theory that reduces all the complexity of organizations (and the institutions in which they are embedded) to a sort of functionalist econ- omic Darwinian survival of the fittest, the fittest being the most efficient because they have the most rational type of management of workers’ compensation, careers, training and so on, with suboptimal forms of organizations being wiped from the market.5

4‘Most labour economists who have taught in business schools have traditionally encountered a disinterested audience’ and ‘The issues studied by human resource specialists were of interest to economists, but the approach taken by non economists lacked the formal framework to which economists have grown accustomed’ (Lazear and Shaw, 2007, p. 23). 5It would be interesting to apply this to the failure of Lehman Brothers when compared with, say, Citigroup, the one having been let down while the other was bailed out. Is there any A new labour economics 711

Where do we go from here? Now let us turn to what I understand as the objective of Paul Osterman’s essay. One of the underlying aims is to (re)construct ILE/ILM as a discipline that is accepted academically and has a (better) grip on reality both in a positive way and in a normative way. This cannot be done by narrow assumptions and econo- metric methods (see above) but rather by keeping (or introducing) some of the Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 real, observable complexities of the workings of (internal) labour markets such as the importance of groups inside organizations, which contribute to creating employment rules and determining outcomes for individuals (Osterman, 2009, p. 11).6 However, the goal of the endeavour is also to ‘develop a more sophisti- cated understanding of what determines employment systems’ which he sees as the result of a political process among groups in which competing interests and rationalities play out in a contest.7

A new direction for institutional labour economics? While I agree with the ‘reductionist’ view of NEP, which is unduly generalized and universalized, and with the importance of groups and their interactions inside of organizations (as opposed to individuals), I think that this alone is not sufficient to reinstate ILE/ILM as a thriving discipline. Reinstating ILE/ILM would indeed eliminate much of what I see as the advantage of a broader approach in the tra- dition of sociological economy or political economy, which is part of ILE and part of an interdisciplinary and international comparative approach. It is, by the way, the latter that can teach those who generalize the lesson of real diversity out of standard theory. Those who study employment systems must take into account the interaction between external institutions (e.g. unions and collective bargain- ing, social protection systems) and the internal behaviour of groups even if these behaviours or political processes and contests are not completely determined by existing rules and institutions. There is, of course, an individual component as well, the degree of which depends on where people are placed in the hierarchy, their qualifications and in which labour market segments they operate. The insti- tutional embeddedness of firms plays a role, and Osterman himself points out the intra-organizational rational efficiency maximizing explanation, or is the explanation rather superior social connectedness, for example?

6The creation of employment rules within organizations is the terrain on which I think institutional labour economics has the most purchase (Osterman, 2009, p. 16). 7Paul Osterman adds that an analysis of the psychology of the worker and the operations of the external (labour?) market, along with the impact of national institutions on the balance of power between different groups within firms is excluded, as this would go beyond the scope of his essay. 712 Discussion importance and attractiveness of this branch of literature, a fact which is also pointed out by Gautie´ (2009). While the comparison of group processes in firms within one country and across diverse firms in diverse sectors is certainly a field of study and a complex endeavour, it seems to be insufficient in terms of reinvigorating a disci- pline. And for the study of (labour market) institutions, an international com- Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 parative aspect seems important. For example, could not one easily accuse NEP of being an American discipline? Indeed, one operates in a less-regulated labour market where individuals might be freer in their decisions, although ato- mized individual decision-makers have always been an illusion, making the nor- mative aspects of NEP that Je´roˆme Gautie´ highlights in his paper a somewhat illusionary goal. One could add that it is also an American discipline because it involves almost exclusively American and American-trained scholars.8 But it is also true that many of Lazear’s examples are based on the higher skilled, higher paid management levels, where one might indeed assume more individual freedom in pay determination. There is less freedom for the average worker, even in the USA. And, given the importance of wage determination in much of NEP, it is particularly strange for a European reader who is so well acquainted with the concept of pay being determined by collective bargaining that such insti- tutional variables are omitted. The real richness of our field of study emerges only when different organiz- ational forms and international variation are taken into account, including the workings of ILMs. It is true that the discipline of ILE must also work on its meth- odological toolkit, and it can (and does) also espouse quantitative analysis. It must work on both its dependent and independent variables and must not shy away from statistical analysis and econometrics. But it should also work on its narratives and in doing so, it should not forget its many roots in the field of labour economics (old and new institutional economics, ILMs in particular and also British, German and French variants of institutionalism—and probably much more—new blends coming from the study of transition economies and developing nations, etc.). The discipline will have to continue to work on the boundaries of economics, sociology and political science (and probably more) and across these boundaries. It is in essence and in reality interdisciplinary. In a global world, pure national models, even if they come from a leading power, will become less interesting,

8‘American economic textbooks typically analyze and celebrate the workings of a highly idealized market economy. Little or nothing is said about the many limitations of a formal analysis of markets, or about the importance of the physical, historical, and social contexts within which markets exist. As the academic arm of economics has become increasingly rigorous and formal and less concerned with application ...mainstream textbooks have reduced the attention given to older concerns of political economy and institutional economics’ (Goodwin et al., 1997, p. 1). A new labour economics 713 and new blends of models will become more attractive.9 So the ‘new’ in New Institutional Labour Economics (NILE)10 could be—apart from the stronger focus on groups, power, hierarchies, institutions and embeddedness—precisely the international and interdisciplinary option we have been looking for. It might not yet be seen in full, but the present crisis will probably contribute to the efforts of Paul Osterman and others to reinstate ILE/ILM as a valid, thriving Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 and challenging discipline which should ending up being called NILE. We had better hurry up and claim the copyright, and we might also need a new textbook!

References Duesenberry, J. (1960) ‘Comment’. In Demographic and Economic Change in Developed Countries, a Conference of the Universities-National Bureau Committee for Economic Research, special conference series, vol. 11, National Bureau of Economic Research, Prin- ceton, NJ, Princeton University Press, pp. 231–234. Gautie´, J. (2009) ‘Institutional Labor Economics: from survival to revival? Comments on Paul Osterman’, Socio-Economic Review (this issue). Goodwin, N. R., Ananyin, O. I., Ackermann, F. and Weisskopf, T. E. (1997) ‘Economics in Context: The Need for a New Textbook’, paper 00-02, Medford, MA, The Global Devel- opment and environment Institute (G-DAE), Tufts University. Hillard, M. and Mcintyre, R. (1994) ‘Is There a New Institutional Consensus in Labor Economics?’, Journal of Economic Issues, 28, 619–630. Lazear, E. P. (1993) ‘The New Economics of Personnel’, Labour. Review of Labour Economics and Industrial Relations, 7, 3–23. Lazear, E. P. and Shaw, K. L. (2007) ‘Personnel Economics: the Economist’s View of Human Resources’, Journal of Economic Perspectives, 21, 91–114. Ostermann, P. (2009) ‘The Contours of Institutional Labor Economics: Notes Towards a Revived Discipline’, Socio-Economic Review (this issue). Rodrik, D. (1999) ‘Institutions for High Quality Growth: What They Are and How to Acquire Them’, paper for the IMF Conference on Second Generation Reforms, Washing- ton, D.C., IMF. Samuelson, P. A. (1983) ‘Frank Knight, 1885–1972’. In Samuelson, P. A. Economics from the Heart: A Samuelson Sampler, New York, Harcourt Brace Jovanovich, pp. 160–161.

9Rodrik (1999, p. 12) observes that ‘an approach that presumes the superiority of a particular model of a capitalist economy is quite restrictive in terms of the range of institutional variation that market economies can (and do) admit’. 10Hillard and Mcintyre (1994) already saw a consensus between neo-classical approaches and institutional labour economics emerging. The last 15 years seem to have reaffirmed rather the differences than the communalities, but this may change. 714 Discussion

Swedberg, R. (1990) Economics and Sociology: Redefining Their Boundaries. Conversations with Economists and Sociologists, Princeton, NJ, Princeton University Press. Williamson, O. E. (1996) The Mechanisms of Governance, Oxford, Oxford University Press. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Institutional labour economics: from survival to revival? Comments on Paul Osterman

Je´ roˆ me Gautie´

University of Paris I Panthe´ on-Sorbonne, Paris, France

Correspondence: [email protected]

Paul Osterman’s article is very stimulating. It not only revitalizes the traditional ‘institutionalist’ arguments against mainstream economics: beyond the confron- tation with NPE, it is an attempt, in a more positive way, to identify some of the specificities of ILE when compared with other non-mainstream approaches, such as the socio-psychological perspective and the political economy perspective. Nevertheless, it is still a long way between the drawing of the ‘contours’of‘insti- tutional thinking’ (or ‘perspective’) and the building of an integrated theory which could compete with standard Labour Economics in general and NPE in particu- lar. Osterman’s contribution is an important stepping stone, but it leaves many crucial questions unanswered.

Groups versus individuals; history and contingency versus functionalism and optimality Osterman’s main statement is that there is a collective dimension in organizations that is not reducible to individual decision-making: organizations are a coalition of groups with not only competing interests, but also competing views and rationalities; they are not just a nexus of contracts between optimizing rational and self-interested individuals. This point had also been put forward almost 25 years ago in Doeringer’s criticism of the efficiency wage theories, in which he advocates ‘a socioeconomic theory of the labour market – one which empha- sizes collective instead of individualistic behavior ...’ (Doeringer, 1986, p. 48, emphasis in the original) and insists on the importance of social relations, A new labour economics 715 collective bargaining power and group cohesion in the workplace. Focusing on this collective dimension makes clear that the difference between ILE and main- stream (Labour) economics does not rest primarily on the divergence concerning the conception of individual motives and behaviours—and this is why ILE goes beyond what Osterman calls the ‘socio-psychological’ criticism of NPE. In that sense, following these remarks, even if it not explicitly mentioned here, ‘behav- Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 ioural economics’—which has experienced a tremendous development in the recent years and which has induced very strong and negative reactions from mainstream economists such as Lazear himself11—misses a crucial point. It has focused, at least until recently, too much on psychology (reference-dependant and time-biased preferences and fairness considerations, for instance) and too little on social relations and structures (culture, networks, etc.). The second crucial point, which is in part a consequence of the first one, is the rejection of the functionalist view according to which institutions, in general, and ILM rules, in particular, can be seen as efficiency optimizing devices, selected by the market process. Apart from the fact that it is often not possible to identify the ‘one best way’, even well-identified sub-optimal practices and institutions can survive for a long time, and this contributes to the explanation of why there is much more variety among organizations and ILMs than the simplistic main- stream economics view can account for. An important implication is that history, and therefore contingency, matters. All these arguments are strongly put forward by Osterman and illustrated with impressive empirical evidence. But they are not specific to ILE: as Osterman also notes in his concluding remarks, ‘they are clearly related to the sociological litera- ture on embeddedness’. Indeed, Granovetter, in his seminal paper, raises the same issues when assessing the standard economic view: he criticizes it for overlooking social relations in which individual behaviours are embedded, for adopting a functionalist perspective on institutions (coupled with the focus on the market selection process argument), and for denying the role of history and contingency (Granovetter, 1985). Beyond Granovetter’s focus on social relations (in particular networks), many other sociologist also take into account other forms of ‘embedd- edness’ and therefore put more emphasis on ‘competing rationalities and balance of power’, just what ILE is supposed to do.12 Overall, ILE may simply appear to be one of the branches of (economic) sociology, with minor specificities, applied to

11Behavioural economists are ‘Barbarians at the gates’ according to Lazear (Lazear, 2000, p. 140).

12Beyond Granovetter’s ‘structural embeddedness’,Zukin and DiMaggio (1990), for instance, mention ‘cultural embeddedness’, which refers ‘to the role of shared collective understandings in shaping economic strategies and goals’, as well as ‘political embeddedness’, which refers to the struggle for power between different groups. For a presentation of the firm as a ‘field of struggle’, see, for instance, Bourdieu (2005). 716 Discussion the field of (internal) labour markets—in the same way that NPE is an application of standard micro-economics to the same field. Identifying the differences between ILE and the existing sociological approaches would require more than the drawing of the ‘contours’ of the disci- pline: namely, the building of an integrated theory. From this point of view— as Osterman himself acknowledges in his conclusion—many questions would Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 need to be addressed: how groups are formed, what exactly the meaning and role of ‘culture’13 are, etc. But so far they remain unanswered.

Taking the new personnel economics challenge seriously The lack of an integrated theory is not only a problem of identity and label (econ- omics or sociology?); it also raises the issue of the capacity of ILE to challenge standard economic theory. The traditional criticism from mainstream economics (in a wide sense, including ‘new institutional economics’) over the (old) institu- tionalist approach, assimilated to the field of industrial relations since WWII, is that it emphasizes facts over theory—‘a mass of descriptive material waiting for a theory or a fire’, according to Coase (1984; quoted in Budd, 2005, p. 179). Oster- man’s clarification of ILE’s theoretical foundations (in the line of Doringer and Piore) can be seen as an attempt to overcome this criticism.14 But, as we have seen in the previous section, he turns towards sociology. For many economists, sociology faces the same kind of criticism: it provides a lot of very interesting empirical material, it is ‘broader-thinking’ than (mainstream) economics— which may seem like an oversimplification, but it is much less rigorous and ana- lytic—all in all, ‘it may be better at identifying issues, but worse at providing answers’ (Lazear, 2000, p. 103). Economists ‘place a premium on identifying general principles’ (Lazear and Shaw, 2007, p. 91) which transcend the insti- tutional and historical variations sociologists (and ILE) focus on. Of course, sociologists reject this view. They criticize economists for focusing on ‘clean models’ and avoiding ‘sloppy reality’, while sociologists accept ‘dirty hands’ because it is the price you have to pay in order to comprehend the com- plexity of reality (Hirsch et al., 1990). In other terms, economists are accused of having nice models but weak empirical relevance, especially when it comes to organizations and firms, which are particularly complex. When tracing the history of the relations between personnel/human resource management

13To give an example, Osterman seems to put less emphasis on ‘culture’ than Champlin and Knoedler (2004) when identifying the theoretical premises of the institutionalist approach. 14When he first presented his contribution at the SASE 2008 annual meeting, Osterman insisted more on the attempt to contribute to an ‘integrated theory’, distinguishing ILE (as a theory) from industrial relations (as a ‘multi-disciplinary field of study’). A new labour economics 717

(HRM) and economics, Bruce Kaufman (2000) emphasizes the growing split between the two fields from the 1930s to the 1990s. Up until the 1930s, when institutionalist economists dominated labour economics, there was a strong con- nection between the two fields, and economics served as the principle intellectual discipline around which business curricula were built. ILE can be seen as applied economics—if the term is used in the context of art, practice and problem Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 solving—and this has long been its comparative advantage, when compared with mainstream labour economics, which was left outside business curricula until recently. Whereas, according to Kaufman, ‘the economics of personnel is largely non applied in that it remains an effort first and foremost at science build- ing, and only secondarily intends to inform and advance the practice of person- nel/HRM’ (Kaufman, 2000, p. 249). This may have been true in the past, but the promoters of the NPE would not agree with this statement: of course, they give a priority to science building, but at the same time they intend to provide prescriptions for HRM. As Lazear clearly puts it:

Economic theory’s slow entrance in business was in part a fault of econ- omics itself. Economists, as scholars, have emphasized positive aspects of economics and eschewed the normative [ ...]. Thus much of the contribution of economics to both the study and teaching of business has been a result of translating positive lessons into normative ones. (Lazear, 2000, p. 117)

In this sense, NPE poses new challenges to ILE, when compared with the previous mainstream economic theories which also intended to open the ‘black box’ of the firm. One big difference between this and the efficiency wage theories Peter Doer- inger was reacting against in the mid-1980s, for instance, is that these theories were only interested in finding some mechanisms in ILMs to better understand external labour markets dysfunctions—i.e. wage rigidity and the resulting unem- ployment. They were not really interested in what was really going on in ILMs and even less interested in giving advice to managers about what they should do. Whereas the normative dimension is central to the NPE approach:

Why should pay vary across workers within the firms? [ ...] Should firms pay workers for their performance on the job or for their skills or hours of work? [ ...]Howshould all these human resource management practices, from incentive pay to team work be combined within firms? Personnel economists offer new tools to analyze these questions – and new answers as well. (Lazear and Shaw, 2007, p. 91, emphasis added) 718 Discussion

Following Lazear (2000), we may even suggest that because of the centrality of the concepts of efficiency and optimality it relies on, the normative dimension of NPE (i.e. its capacity to provide prescriptions) is even more important than its positive dimension (i.e. its capacity to account for, via its theoretical models, reality as it is). In fact, Lazear acknowledges that there is still a gap between the NPE models and real HRM practices, but he stresses that the gap is progress- Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 ively shrinking: ‘personnel economics explicitly think in terms of substitution and trade-offs, and firms have begun to do so as well’ (Lazear and Shaw, 2007, p. 105, emphasis added). Why is this so? Simply because managers find interest in adopting behaviours that economists claim are optimal (Lazear, 2000)? It is more than the usual argument (criticized by Osterman, among many others) that market forces select optimal behaviours. The diffusion of economic ideas plays a crucial role here. It relates to the performativity of economic discourse.15 Because of its normative dimension, economic theory shapes the world and becomes real (i.e. empirically relevant) afterwards: in a sense, its ‘normativity’ precedes its ‘positivity’. Whether the performative power of NPE—and, in the first place, its capacity to provide useful tools for HRM—is largely overestimated is an open question. Still, ILE faces a big challenge because it may lose the battle on both grounds: in science building, because of the weakness of its theoretical foundations (see above); and also as an applied discipline, because the insistence on contingency and complexity is not easy to sell to human resource practitioners in search of the clear-cut prescriptions NPE claims to be able to provide.

References Budd, J. (2005) ‘Ideas versus Ideology: the Origins of Modern Labor Economics – Com- ments on Gallaway and Vedder’, Journal of Labor Research, XXVI, 177–180. Bourdieu, P. (2005) ‘Principles of an Economic Anthropology’. In Smelser, N. and Swedberg, R. (eds) The Handbook of Economic Sociology, New York, Russell Sage Foundation and Princeton, NJ, Princeton University Press. Callon, M. (2006) ‘What Does It Mean to Say that Economics Is Performative?’. In MacKenzie, M. F. and Siu, L. (eds) Do Economists Make Markets? On the Performativity of Economics, Princeton, NJ, Princeton University Press. Champlin, D. and Knoedler, J. (eds) (2004) The Institutionalist Tradition in Labor Econ- omics, Armonk, NY, Sharpe. Coase, R. (1984) ‘The New Institutional Economics’, Journal of Institutional and Theoretical Economics, 140, 229–232.

15‘A discourse is performative [ ...] if it contributes to the construction of the reality that it describes’ (Callon, 2006, p. 7). A new labour economics 719

Doeringer, P. (1986) ‘Internal Labor Markets and Noncompeting Groups’, American Economic Review, 76, 48–52. Granovetter, M. (1985) ‘Economic Action and Social Structure: the Problem of Embeded- ness’, American Journal of Sociology, 91, 481–510. Hirsch, P., Michael, S. and Friedman, R. (1990) ‘Clean Models vs. Dirty Hands: Why Economics is Different from Sociology’. In Zukin, S. and DiMaggio, P. (eds) Structures Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 of Capital: the Social Organization of the Economy, Cambridge, Cambridge University Press. Kaufman, B. (2000) ‘Personnel/Human Resource Management: Its Roots as Applied Econ- omics’, History of Political Economy, 32, 227–256. Lazear, E. (2000) ‘Economic Imperialism’, Quarterly Journal of Economics, CXV, 99–146. Lazear, E. and Shaw, K. (2007) ‘Personnel Economics: the Economist’s View of Human Resources’, Journal of Economic Perspectives, 21, 91–114. Zukin, S. and DiMaggio, P. (1990) Structures of Capital: the Social Organization of the Economy, Cambridge, Cambridge University Press.

The revival of institutional labour economics: a societal dimension

David Marsden

London School of Economics, London, UK

Correspondence: [email protected]

The proposal to take the groups within the firm, and the relationships among them, as one of the key intuitions of New Institutional Labour Economics (NILE) is surely very welcome. Paul Osterman has argued that promoting effective coordination between groups and managing their potential conflicts of interest within the firm’s ILM is a more fruitful starting point than optimizing individual relationships within a principal–agent framework as proposed by Lazear and Shaw (2007). Osterman has focused primarily on North American examples. I should like to argue that the value of his proposed approach can be greatly strengthened if applied to the findings of international comparative research on management of the workplace. The time is ripe because one achievement of the New Economics of Personnel (NEP), which Lazear’s work has done much to promote, has been to revive interest among economists in what goes on inside the firm. I should like to focus on Osterman’s four examples to show that his argument is greatly reinforced if we consider the findings of a generation of international 720 Discussion comparative research on workplace management. In doing so, I should like to extend his argument beyond firm ILMs and propose that managers often face an important strategic choice: between individual solutions to workplace pro- blems, and those which require collective action, sometimes with other firms, and sometimes through joint institutions. Within labour economics as within industrial relations, for too long we have conceived of collective bargaining as Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 a vehicle for workers’ collective action. This has overshadowed the important element of collective action among employers, not just to counter the power of organized labour, but also to deal with issues that are problems for employers and which can benefit from a collective rather than an individual approach. Osterman proposes four areas as examples: the emergence of firm ILMs, the slow diffusion of high performance work systems, the commitment and incen- tives, and finally, the diversity of employment systems. ILMs are more than administrative units in which management decisions replace market pricing mechanisms. Whether employers considered them the right solution to their labour management problems has been greatly influenced by the institutional context in which they operate. Lorenz’s (1987) historical com- parison of British and French shipyards showed that whereas British employers had an alternative source of skilled labour available from a set of occupational labour markets, their French counterparts lacked this option, and had to build ILMs in order to develop their own supply of skilled labour, as a ‘quasi-fixed factor’ to use Oi’s (1962) term. The presence of a well-organized system of occu- pational labour markets in Germany has similarly enlarged the options facing employers there regarding whether to set up their own ILMs or to participate in occupational markets (Sengenberger, 1987). In recent years, many large employers have built hybrid forms using their involvement in training for occu- pational skills, as a way of recruiting able young workers, and then offering them supplementary training and careers that build on occupational foundations (Franz and Soskice, 1995). Many smaller firms, on the other hand, have opted for the use of Germany’s well-developed occupational markets (Drexel, Steedman). Thus, a firm’s decision whether to build an ILM depends partly on how their local labour markets are organized. This choice between internal and occupational market strategies also affects the structure, and capabilities, of the different occupational groups within the workplace. A good example is provided by the comparison between engineers in similar French and Japanese firms (Maurice et al., 1991). In both countries, employment is strongly shaped by firm ILMs, but the strength of occupational groups within them differs greatly. In the French firms, the position of engineers is stratified by their initial training, with top management positions taken by those from the e´lite engineering schools, and a hierarchy of different technical diplomas structuring the career ladders for other groups of engineers. In contrast, A new labour economics 721 in the Japanese firms, graduate engineers enter by a common portal alongside other graduate white collar recruits, and slowly emerge into senior positions. Both countries’ firms have highly competent engineers, but the group identities and loyalties in the French firms are much more strongly anchored in distinct qualification groups. From the work of Maurice and his Japanese and French col- leagues, one can further trace out the implications of how the different group Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 identities play out in workplace cooperation. The role of collective action among employers for the formation of transfer- able skills and the occupational groups they underpin is well illustrated by Johan- sen’s (2000) study of employer-provided training in Norway. Johansen used Olson’s (1971) theory to examine whether employers’ capacity for collective action in training depends on industrial structure, whether there are many small firms, or a few large ones, and whether there is an overarching coordinating body. He identified four sectors to illustrate each of these possible situations, and found that employer-funded training for transferable skills worked best where the collective action problems could be best resolved. Thus, the Norwegian engineer- ing industry, characterized by many small firms and no coordinating body, experienced the greatest difficulties, whereas in the insurance industry, leadership was provided by a nucleus of large firms which set standards that other, smaller, firms followed. Municipal schools and regional hospitals also came out quite well, the former having a coordinating body, and the latter benefiting from both this and a nucleus of large organizations. Johansen focused mainly on employer pro- vision; nevertheless, in many countries, state education and training provision is closely articulated with employer action, and so will be influenced by employers’ capacity for collective action. The diffusion of high-performance work systems likewise displays strong international differences and the importance of employer collective action. Although Womack et al. (1990) forecast that lean production would sweep aside the diversity of national experiments in the design of mass production systems, in fact, this has not happened. Flexible work teams may have spread across the industrial world, as observed by Katz and Darbishire (2000), but the way they have been enacted differs greatly, and to a considerable extent has been shaped by the forms of regulation that grew with mass production. A good example is provided by Lorenz and Valeyre’s (2005) study of ‘lean’ and ‘learning’ forms of team organization across the European Union. Lean teams have relatively little discretion and are closely controlled by management. Learn- ing teams give workers more discretion, and more control over their time so that they can learn from the multiple, unanticipated problems that arise day to day. Lean teams tend to be more common in countries such as Britain, Ireland, and to a lesser extent France, whereas the learning teams are located predominantly in Scandinavia and Germany. Koike and Inoki’s (1990) account of work 722 Discussion organization in Japan and South-East Asian countries suggests a similar pattern, with the learning model more developed in Japan, and the lean model in some of its neighbouring countries. How much autonomy management allows work teams is partly an issue of efficiency and learning, but also a deeply political issue in the workplace. Work group autonomy can be a great source of efficiency because management has Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 limited access to knowledge of micro-level organizational improvements, and it can be a valuable source of peer group pressure against free-riding (Freeman et al., 2008). However, group autonomy can also be a source of work group power, as illustrated by the classic studies of job control (Brown, 1973; Burawoy, 1979), and a potent source of opposition to management policies. Thus, a critical issue is how far management and work groups see it as in their joint interest to cooperate as opposed to what Katz (1985) described, in an earlier study of the automobile industry, as trench warfare. Again the collective action choice emerges. A key issue touched on in Turner and Auer’s (1994) study of approaches to team working lies in the independence of workers’ skills vis-a`-vis the firm, and the degree to which they can trust management so that the group power, which could be used for conflict, is harnessed instead for cooperation. In many respects, building trust is like convincing one’s partner that ‘tit-for-tat’ is a viable strategy. One has to convince the other that one’s long- term intentions are for cooperation, even though the opportunities to prove this may occur only infrequently. Shirai argued that Japanese employers’ commitment to long-term employment was not tested until the first oil shock when they showed how much they were prepared to sacrifice in order to sustain this prac- tice. The other factor in ‘tit-for-tat’ is the capacity to play the punishment strategy if the other side has reneged; in other words, it is easier to engage trust if both parties have the sufficient power to punish the other. In the case of work groups, this means having organizational resources, such as those of trade unions or works councils, or solidarity from other groups if they feel the employer has reneged on a promise; likewise, on the other side, if the employer knows that fellow employers will be supportive. Diffusion of beneficial but risky ‘high transaction cost’ policies, such as those associated with ‘high performance work systems’ (HPWS), faces familiar ‘first mover’ disadvantages. These were well described in the US by Appelbaum and Batt (1994) in ‘The New American Workplace’.In the language of personnel econ- omics, such policies incur high transaction costs because of the related invest- ments in training and workplace procedures required in order to make them work effectively. Thus, firms that invest heavily in training to make HPWS effective are exposed to a number of risks. Rival employers may poach their skilled labour; during the early years of high investment they be could easily be undercut by firms adopting policies with lower transaction costs; unions and A new labour economics 723 workers, with experience of how peer firms behaved, could be deeply suspicious of the firm’s motives; and the promise of more secure jobs could lead to problems of adverse selection as they might attract less efficient workers who would have been dismissed by the non-innovator firms. In the US environment, these and other authors proposed legislative action largely because the institutions to support collective action were weak or absent. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Turning to the ‘efficiency wage’ practices, as Osterman points out, these have been urged by Personnel Economics as a means to generate employee commit- ment and high levels of effort and quality. One element of commitment behav- iour, which has assumed considerable importance in recent debates, is employee willingness to engage in peer monitoring. There can be financial motives for peer monitoring as Freeman et al. (2008) show. ‘Fair share capitalism’ can boost performance by encouraging peer monitoring. Their results are also con- sistent with the emphasis on groups which is at the centre of Osterman’s programme for NILE. It is widely accepted that collective incentives are more effective for small groups than for large groups because they deal better with potential free-rider pro- blems. Thus, for any manager designing a group incentive pay system, an essential question must be: ‘who are the relevant peers?’ Thinking within an individualistic framework, one could ask whether a share in profits or a modest team bonus would be enough to encourage fellow workers to inform on their colleagues or to pressure them if they do not perform. However, as the classic work group studies remind us, such pressures can cut both ways: work groups can sanction both free-riders and ‘rate busters’.Thus once more, the evidence on group behaviour in different environments lies at the centre of a revived NILE. Compared with the pure world of choice within the NEP, the recognition of group conflicts within the firm makes the world more idiosyncratic, and increases the importance of contextual knowledge if managers are to operate incentive systems successfully. ‘Merit pay’, for example, is a high transaction cost incentive system because of the need for procedural justice mechanisms, such as perform- ance appraisal and appeals procedures. If employees do not believe it is operated fairly, they are unlikely to be motivated by it. Idiosyncrasy means that general solutions have to be adapted to local contexts and puts a premium on the infor- mation needed for this adaptation. Again a collective action choice arises: firms can seek individual solutions, or they can work with other firms which operate in similar contexts, pool their experiences, and learn how to adapt and apply incentive systems in their own environments. Such joint action appears to lie behind the greater diffusion of merit pay and other kinds of incentive pay among similar types of firms in France as compared with Britain. As Osterman points out, the initial impulse for developing incentive pay in France may have arisen because of strong job protection that made dismissal for poor performance more difficult in France than in Britain (Marsden and Belfield, 2009). However, 724 Discussion given this initial impulse, it appears that French firms then made extensive use of their employer organizations and especially of local employer networks to improve and implement their incentive pay systems. The workplace of the NEP is peculiarly transparent because work groups and the politics of their inter- relationships are absent. Knowledge of the latter is an essential part of the success- ful operation of incentive systems. As the contours of these groups are often Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 similar across workplaces in the same region, firms face a real choice between acting individually or collectively. As French firms had built a greater capacity for acting collectively, they had progressed further with the development of incentive pay systems than their British counterparts. International diversity of employment systems even within the same sectors was vividly emphasized in Katz and Darbishire’s sectoral studies of employment relations in different countries. They emphasized what Dunlop called the ‘power context’, that firms adjusted to globalization and the spread of team-working and other HPW practices, according to the local institutional arrangements. Such influences are an important part of the case for a revived NILE, but they are sometimes dismissed as causing temporary although non-essential deviations from the competitive norm. Osterman counters this argument with Thelen’s (2004) evidence on the long-term presence of such factors, over many decades, for example, in shaping skill training systems in different countries. Likewise, the research report, which heralded the ‘Social foundations of industrial power’ (Maurice et al., 1986), contained a number of historical accounts of the genesis of aspects of the distinctive employment systems in France and Germany that support Thelen’s view. There is another perhaps more fundamental aspect of employment systems which shapes the different architectures that we can observe today. If we consider the decentralized decisions by workers and firms, the same actors as those used in Personnel Economics, it can be shown that these architectures correspond to four different solutions to the basic problem of the employment relationship: the bureaucratic, craft, J-form and the professional. They balance the flexibility of an incomplete contract for the employer with sufficient protections for the worker (Marsden, 1999). Without those protections, workers would not agree to be opt for an employment relationship as opposed to some other form of service contract. Firms may opt for these different architectures on a purely indi- vidualistic basis, but the advantage that they can gain from one or the other model is affected by how widely it has been adopted by other employers. Although at first sight, the reference to firms and workers might seem retrograde with regard to NILE, there are also many externalities, like those observed by Appelbaum and Batt (1994), which affect the profitability of different models for the firm. These open up scope for collective action solutions. This applies par- ticularly to the craft and professional architectures where transferable skills are A new labour economics 725 involved, but it is also relevant for the more organization centred bureaucratic and J-form architectures which function more flexibly when there is a high level of trust. Osterman’s agenda for a revived NILE is an important one. His emphasis on the importance of groups within the firm not only opens the door to closer collaboration with sociologists and political scientists, but it also clears the way Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 for integrating the findings of many recent comparative studies of the workplace. The latter also highlight the need to look beyond the boundaries of firm ILMs and to look at collective action, and the decision facing firms whether to opt for individual or collective solutions.

References Appelbaum, E. and Batt, R. (1994) The New American Workplace: Transforming Work Systems in the United States, Ithaca, NY, Cornell University Press. Brown, W. E. (1973) Piecework Bargaining, London, Heinemann. Burawoy, M. (1979) Manufacturing Consent: Changes in the Labor Process under Monopoly Capitalism, Chicago, University of Chicago Press. Franz, W. and Soskice, D. (1995) ‘The German Apprenticeship System’.In Buttler, F., Franz, W., Schettkat, R. and Soskice, D. (eds) Institutional Frameworks and Labour Market Per- formance: Comparative Views on the US and German Economies, London, Routledge. Freeman, R., Kruse, D. and Blasi, J. (2008) ‘Worker Responses to Shirking Under Shared Capitalism’, NBER Working Paper No. 14227, August, Cambridge, MA, National Bureau of Economic Research. Johansen, L.-H. (2000) Transferable Training and the Collective Action Problem for Employers: an Analysis of Further Education and Training in Four Norwegian indus- tries, PhD Thesis, University of London. Katz, H. C. (1985) Shifting Gears: Changing Labor Relations in the US Automobile Industry, Cambridge, MA, MIT Press. Katz, H. and Darbishire, O. (2000) Converging Divergencies: Worldwide Changes in Employ- ment Systems, Ithaca, NY, ILR Press and Cornell University Press. Koike, K. and Inoki, T. (eds) (1990) Skill Formation in Japan and Southeast Asia, Tokyo, University of Tokyo Press. Lazear, E. P. and Shaw, K. L. (2007) ‘Personnel Economics: The Economist’s View of Human Resources’, Journal of Economic Perspectives, 21, 91–114. Lorenz, E. (1987) ‘Les chantiers navals en France et en Grande-Bretagne 1890–1970’, Le Mouvement Social, 138, 21–44. Lorenz, E. and Valeyre, A. (2005) ‘Organisational Innovation, Human Resource Management and Labour Market Structure: a Comparison of the EU-15’, Journal of Industrial Relations, 47, 424–442. 726 Discussion

Marsden, D. W. (1999) A Theory of Employment Systems: Micro-foundations of Societal Diversity, Oxford, Oxford University Press. Marsden, D. W. and Belfield, R. A. (2009) ‘Institutions and the Management of Human Resources: Incentive Pay Systems in France and Great Britain’, Centre for Economic Per- formance Discussion Paper, London, Centre for Economic Performance, London School of Economics. Downloaded from https://academic.oup.com/ser/article/7/4/695/1606158 by guest on 29 September 2021 Maurice, M., Mannari, H., Takeoka, Y. and Inoki, T. (1988) Des entreprises franc¸aises et japonaises face a´ la me´catronique: acteurs et organisation de la dynamique industrielle, Laboratoire d’Economie et de Sociologie du Travail (CNRS), Aix-en-Provence. Maurice, M., Sellier, F. and Silvestre, J. J. (1986) The Social Foundations of Industrial Power: a Comparison of France and Germany, Cambridge, MA, MIT Press. Oi, W. (1962) ‘Labor as a Quasi-Fixed Factor’, Journal of Political Economy, 70, 538–555. Olson, M. (1971) The Logic of Collective Action: Public Goods and the Theory of Groups, Cambridge, MA, Harvard University Press. Sengenberger, W. (1987) Struktur und Funktionsweise von Arbeitsma¨rkten: die Bundesrepu- blik Deutschland im internationalen Vergleich, Frankfurt, Campus Verlag. Thelen, K. (2004) How Institutions Evolve: the Political Economy of Skills in Germany, Britain, the United States, and Japan, Cambridge, Cambridge University Press. Turner, L. and Auer, P. (1994) ‘A Diversity of New Work Organisation: Human-Centred, Lean, and In-between’, Industrielle Beziehungen, 1, 38–60. Womack, J., Jones, D. T. and Roos, D. (1990) The Machine that Changed the World, New York, Rawson Associates.