Employees and Entrepreneurship: Co-Ordination and Spontaneity in Non-Hierarchial Business Organizations (New Thinking in Political Economy)

Employees and Entrepreneurship

NEW THINKING IN POLITICAL ECONOMY Series Editor: Peter J. Boettke, George Mason University, USA New Thinking in Political Economy aims to encourage scholarship in the intersection of the disciplines of politics, philosophy and economics. It has the ambitious purpose of reinvigorating political economy as a progressive force for understanding social and economic change. The series is an important forum for the publication of new work analysing the social world from a multidisciplinary perspective. With increased specialization (and professionalization) within universities, interdisciplinary work has become increasingly uncommon. Indeed, during the 20th century, the process of disciplinary specialization reduced the intersection between economics, philosophy and politics and impoverished our understanding of society. Modern economics in particular has become increasingly mathematical and largely ignores the role of institutions and the contribution of moral philosophy and politics. New Thinking in Political Economy will stimulate new work that combines technical knowledge provided by the ‘dismal science’ and the wisdom gleaned from the serious study of the ‘worldly philosophy’. The series will reinvigorate our understanding of the social world by encouraging a multidisciplinary approach to the challenges confronting society in the new century. Recent titles in the series include: Constitutions, Markets and Law Recent Experiences in Transition Economies Edited by Stefan Voigt and Hans-Jürgen Wagener Austrian Economics and the Political Economy of Freedom Richard M. Ebeling Anarchy, State and Public Choice Edited by Edward Stringham Humane Economics Essays in Honor of Don Lavoie Edited by Jack High Public Choice and the Challenges of Democracy Edited by José Casas Pardo and Pedro Schwartz Fiscal Sociology and the Theory of Public Finance An Exploratory Essay Richard E. Wagner Institutional Competition Edited by Andreas Bergh and Rolf Höijer Political Failure by Agreement Learning Liberalism and the Welfare State Gerhard Wegner The Neoliberal Revolution in Eastern Europe Economic Ideas in the Transition from Communism Paul Dragos Aligica and Anthony J. Evans Employees and Entrepreneurship Co-ordination and Spontaneity in Non-hierarchical Business Organizations Ivan Pongracic, Jr.

Employees and Entrepreneurship Co-ordination and Spontaneity in Non-hierarchical Business Organizations

Ivan Pongracic, Jr. Associate Professor of Economics and William E. Hibbs/ Ludwig von Mises Chair in Economics, Hillsdale College, USA

NEW THINKING IN POLITICAL ECONOMY

Edward Elgar Cheltenham, UK • Northampton, MA, USA

© Ivan Pongracic, Jr. 2009 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical or photocopying, recording, or otherwise without the prior permission of the publisher. Published by Edward Elgar Publishing Limited The Lypiatts 15 Lansdown Road Cheltenham Glos GL50 2JA UK Edward Elgar Publishing, Inc. William Pratt House 9 Dewey Court Northampton Massachusetts 01060 USA

A catalogue record for this book is available from the British Library Library of Congress Control Number: 2008937418

ISBN 978 1 84720 806 4 Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents Series editor’s foreword by Peter Boettke Foreword by Frederic Sautet Acknowledgments

vii ix xi

1. Introduction Understanding decentralization developments The pros and cons of decentralization Modern economic theory of the firm and decentralization Outline of the book 2. The hierarchical theory of the firm Coase’s theory of the firm Williamson’s continued development of the Coasian research program 3. The knowledge problem in firms The institutional taxonomy The knowledge problem in firms Decentralized decision-making through hierarchical flattening The organizational tradeoff between co-ordination and innovation Foss’s critique of intra-firm decentralization Seeking creativity within firms Conclusion 4. Spontaneous order in decentralized firms What motivates employees? Spontaneous order within decentralized firms Conclusion 5. Employees as creative agents Entrepreneurship and creativity The process of entrepreneurial decision-making and judgment derivation Creativity and imagination Creativity, judgment and delegation Conclusion

1 4 8 15 16 20 21

v

24 42 45 51 59 68 72 85 87 92 93 103 109 114 114 116 119 121 123

vi

Contents

6. Conclusion

127

Bibliography Index

132 139

Series editor’s foreword Ivan Pongracic’s Employees and Entrepreneurship breaks new ground in the study of the organization of economic activities within the business firm from the perspective of the Austrian School of Economics. He emphasizes the role of spontaneous order within firms. Successful firms, he argues, use the decentralized decision-making structure to foster creative responses among employees. He argues decentralized firms are decentralized precisely for this reason: they give decision-making power to employees in order to extract creative action from them. This emphasis on cultivating creativity among employees inside the firm, rather than disciplining them by aligning incentives, is a unique contribution of Pongracic. He marshals this analytical framework to interpret the modern evolution of business practice over the last quarter-century, which has seen both delayering of firm structure and an emphasis on creative work in firms such as Google. Since the mid-1980s several works in the Austrian tradition have addressed the theory of the firm. These studies tend to emphasize the importance of decentralized knowledge and the alignment of incentives within the firm to mobilize dispersed knowledge. A classic work in this tradition is Frederic Sautet’s An Entrepreneurial Theory of the Firm, with its emphasis on the M-form organization and the Hayekian knowledge problem that business enterprises must confront and solve for the successful co-ordination of economic plans within the firm. Pongracic’s work fits nicely into this modern literature. However, Pongracic’s work is the first work in this literature to address in depth the radical decentralization and flattening of managerial structures that have been witnessed over the past quarter-century within capitalist economies. The theory of the firm has an unusual history in the discipline of economics. Classical economists acknowledged the role that the entrepreneur played in a dynamic economy. But as the discipline of economics evolved in a technical direction in the twentieth century, the entrepreneurial element in economic life had to be purged. The reason for the purging was straightforward. Entrepreneurship is about creativity, novelty, and innovation – things that are not easily captured in deterministic models. As the model of perfect competition came to dominate the intellectual agenda in microeconomics, not only entrepreneurship but also management and strategy were vii

viii

Series editor’s foreword

pushed aside. Instead, the theory of the firm that developed mid-century was a theory of production functions. Ronald Coase had offered an alternative theory of the firm focusing on the transaction costs of market exchange, and emphasizing the rationale for organization of economic activities in the firm to minimize transaction costs. But Coase’s theory of the firm never really gained wide acceptance until Armen Alchian and Harold Demsetz focused on the problems of team production, and Oliver Williamson engaged in an act of intellectual arbitrage by using Coasian ideas to make sense of Alfred Chandler’s work on managerial capitalism. The focus of this literature was on the structure of the firm and managerial strategy. This literature mainly focused on the appropriate alignment of incentives within the firm. Parallel to these developments, scholars such as Israel Kirzner worked tirelessly to reintroduce the entrepreneurial element of the market process into economic analysis. With respect to the firm, Kirzner tended to emphasize the selling costs associated with the delivery of products to consumers as well as the production costs. But Kirzner did not really discuss the internal structure of the firm and the managerial strategies adopted to realize profits. The Austrian literature on the firm really took off with Richard Langlois. Langlois reached backward to Edith Penrose, Fritz Machlup, and F.A. Hayek, and synthesized these works with the contemporary theory of the firm found in Alchian, Coase, and Williamson. The literature within the Austrian School that Langlois inspired includes works by Nicolai Foss, Peter Klein, Peter Lewin, and Frederic Sautet. But this literature did not focus on the contemporary empirical trend of demanagement. In fact, Williamson in one of his classic works posed the Coasean question as ‘markets’ or ‘hierarchy’. But in the years since he published Markets and Hierarchies (1975), firms have flattened their hierarchical structures without eliminating the unity of their business enterprises. Enter Pongracic. Here we start from the existing reality of the ‘free agent nation’ and the ‘delayering’ of the business enterprise. A rather stark example of this sort of new freedom in the organization of economic activities within the firm would be Google, where the work environment and structure of labor contracts are far more flexible and creative than could even be imagined in a traditional hierarchical firm. Ivan Pongracic’s Employees and Entrepreneurship is a welcome addition to the Austrian literature on the theory of the firm, and represents a significant step forward by demonstrating the relevance of the ideas of Mises, Hayek, Kirzner and Lachmann for the contemporary world of business and the entrepreneurial market process. Peter Boettke University Professor, Economics Department, George Mason University

Foreword Economists have regarded the market as a decentralized system of resource allocation for a long while. In the market, no one is in charge of allocating resources beyond what every individual has under his or her command. For instance, there is no one centrally deciding whether car makers should produce white cars or whether computer makers should make laptops rather than desktop computers. In such a system, entrepreneurs constantly bid resources away from their current uses in order to reallocate them towards what consumers may want to buy. In the marketplace, resources are bought and sold without anyone deciding what the overall allocation of resources should look like. The opposite is true in business firms – or so it seems. Indeed, the traditional story is to say that firms are islands of planning and hierarchy in the sea of the market. For years, firms’ internal structures were not a subject of inquiry. Economists assumed they were centralized hierarchies because in economic theory firms were nothing more than production functions. Things started to change in the 1960s with, for instance, the work of Alfred Chandler, which looked at the inside of the black box to discover that firms can have some degree of decentralization. Progressively economists started paying more and more attention to the inner structure of firms and the role of hierarchy. More specifically, the idea that hierarchy may be an impediment to firm development has become a subject of inquiry in recent years with the emergence of corporations with flattened structures. Many management articles and books, as well as economics ones, have been written on the subject. It is in this context that Dr Pongracic offers the economics profession remarkable research into the causes of organizational change and the rationales for firms’ structural evolution. It draws on the economics of organizations, entrepreneurship theory, and the insights from, among others, Joseph Schumpeter and Friedrich Hayek. It also draws on other disciplines to explain phenomena within the firm that economists have not been able to fully account for. Dr Pongracic’s book is not intended to dispense advice to managers; it is about studying the causes as to why some firms flatten their hierarchical structure and decentralize while others don’t. And as economists have discovered over the last few decades, firms’ structural changes don’t occur without their share of difficulties, among them opportunistic behavior and agency problems. ix

x

Foreword

Yet, for some economists, an even more important issue exists. The major problem that economics must explain is that of the co-ordination of knowledge and resources. Ludwig von Mises warned that no central planning exercise can solve the co-ordination problem absent market prices and private property. Because firms rely to a great extent on centralization and hierarchies (and not on actual market prices), they face an internal coordination problem, which worsens as they grow. A potential solution is to introduce high-powered incentives similar to what one finds in markets. Flattening the hierarchy and decentralizing decision-making are two main avenues to introduce those market-like incentives. However, absent real prices, internal co-ordination may remain a difficulty. Countless studies have shown how firms can use shadow prices to reveal the opportunity costs of using internal resources and how intrapreneurs can take advantage of them. Still, these studies rarely focus on the existence of other (entrepreneurially driven) mechanisms that would solve the internal co-ordination problem (as well as problems of opportunism and agency). A flattened hierarchy and a more decentralized structure may bring about processes that go a long way towards (paradoxically perhaps) creating a greater internal co-ordination and making some firms more entrepreneurial. This is where Dr Pongracic’s fresh perspective on the subject offers new insights. He skillfully presents an original approach arguing for the presence of non-price co-ordinative mechanisms within firms with a flattened hierarchy. He explains why, under certain circumstances, firms can introduce strong entrepreneurial incentives and marketlike structures without dissolving into chaos. Scholars in the area of management as well as Austrian economists and specialists of the theory of the firm will find important insights in this book. Dr Pongracic doesn’t fall into the trap of arguing that decentralization is the new big thing that every CEO should be thinking about. He simply consistently applies the approach of Austrian economics and other disciplines to explain firms’ structural changes. Dr Pongracic’s book is a major contribution to our understanding of why firms do what they do and it provides a way for concepts developed in the Austrian literature to have a voice in the larger debate. We can only be grateful to him for all that. Frederic Sautet Senior Research Fellow Senior Editor, Mercatus Policy Series Mercatus Center at George Mason University February 2008

Acknowledgments It is no exaggeration to say that this book would not have been possible without the help of several important individuals. I owe them an enormous debt of gratitude. The book began as my doctoral dissertation at George Mason University, so I must first express my deep thanks to my dissertation committee director, Dr Karen Vaughn. In addition to successfully (and with great patience) guiding my dissertation process to completion, Professor Vaughn also had the greatest influence on my learning in graduate school. The years I spent as her research assistant and the two classes I took with her taught me how to be a successful heterodox economist as well as what is important and distinctive about Austrian economics. I would also like to thank the rest of my committee, which consisted of Drs Richard Wagner, Peter Boettke, and John Crockett. I greatly appreciate their guidance and comments. Broader thanks must also go to the George Mason University graduate program, which has provided me with an education that was exactly what I was seeking. My years there have set my intellectual path for the rest of my years, and I am grateful to all of my faculty there, especially to those from whom I learned so much about the amazing insights of Ludwig von Mises and Friedrich Hayek, economists who have provided the foundation for all my economic understanding. My dissertation would never have become a book if it had not been for Dr Peter Boettke. I had not even considered attempting to turn it into a book and publish it until he suggested it to me. He also recommended my work to Edward Elgar Publishing, which led to this publication. Professor Boettke’s influence on my life goes beyond this, though. In addition to owing a great intellectual debt for all I have learned from him through the years as well as having him on my dissertation committee, it is his unceasing energy to promote as well as to do good economics that has been a true inspiration. I am also very grateful for his excellent foreword. My sincere thanks must also go to Dr Richard Ebeling, a former colleague and a friend. His unmatched, encyclopedic knowledge and thoughtprovoking, insightful questions played a very important role in the development of this book. In addition, by pointing me to his own theoretical work on Alfred Schutz and intra-group co-ordination, he enabled me to make much greater sense of decentralization within firms than I would have otherwise. His suggestions led me to important breakthroughs at key xi

xii

Acknowledgments

junctures of my work, and I cannot thank him enough for this. He has also played a large role in my life, making possible many accomplishments that I probably would not have attained without him. In this, thanks must also go to his wife, Dr Anna Ebeling. I am happy and proud to call them both friends. I have known Frederic Sautet since the mid-1990s, when we were both in the midst of our graduate studies. Our intellectual interests and backgrounds were very similar, and I have learned much from our many conversations through the years. His contributions to the theory of the firm, in particular how intra-firm structures affect the entrepreneurial responses by employees (as presented in his 2000 book), have also proved to be invaluable to my work. In addition, his detailed comments were absolutely fundamental in allowing me to convert a dissertation into a book, and tighten up my arguments and rhetoric. Finally, his unceasing enthusiasm for and faith in my work have been an important source of encouragement that allowed me to get through many doubts. As though all of that were not enough, he also wrote a very kind and wonderful foreword for this book. Words cannot express my gratitude for all of this. I owe much of my initial insights into the topic of decentralization of decision-making within firms to my father, Ivan Pongracic, Sr. who provided me with many references and sources from management and business literature dealing with this topic. I have learned a lot from our many conversations on the topics of employee ‘empowerment’ and flattening of hierarchical structures. It is safe to say that my dissertation would have looked very different – if it were even ever to be finished – without his many suggestions, and I am deeply thankful. I must also thank my colleagues in the Hillsdale College economics department, namely Gary Wolfram, Charles Steele, and Nikolai Wenzel, for being so helpful, accommodating, encouraging, enthusiastic, fun, and always interesting. You have made work much more pleasant and rewarding, and I thank you. Thanks must also go to everyone at Hillsdale College, especially the administration, for providing a stimulating, enjoyable, and supportive workplace and helping me in my many endeavors over the years. Special thanks also to my assistant Hannah Mead for her great help with the book index. Finally, I must thank my wonderful wife, Christina Pongracic, as well as my parents, Ivan Sr. and Ana Pongracic, for their unceasing encouragement, bottomless patience, and support. I dedicate this book to you. Ivan Pongracic, Jr. Hillsdale, MI March 2008

1. Introduction We live in a global economy . . . To have a fighting chance, companies need to get every employee with every idea in their heads and every morsel of energy in their bodies, into the game. (Jack Welch, ‘The “But” Economy’, Wall Street Journal, October 30, 2003)

It is commonly accepted that the existence and survival of firms rest on their ability to generate new market knowledge in the form of innovations more effectively than (or at least as effectively as) their competitors. Yet economics has so far not explained very well how firms create and introduce innovations, and why some are more successful at it than others. Joseph Schumpeter defined the function of entrepreneurship as introducing novelty (as in something new) and initiating change, but there has been little economic examination of either the process of change or the introduction of novelty within firms and even fewer systematic inquiries into how entrepreneurship works within a firm.1 In modern economic theory firms either have nothing to do with innovation, or else are but vessels for carrying out the vision of a great, charismatic leader/entrepreneur (for example, Casson 1982; Nelson and Winter 1982). Circumstances where employees act in creative and entrepreneurial ways are not often considered. But in order to explain many of the most vibrant of today’s businesses and industries, we must not only incorporate entrepreneurship into the theory of the firm but do it in a way that will also explain employee creativity. That is the goal of this book. I will examine how the structure of the internal organization of firms affects employee entrepreneurship. I will focus in particular on the phenomenon of decentralization of decision-making within a firm, as I believe this is the clearest manifestation of attempts to create a thoroughly entrepreneurial firm, one where most or even all employees are encouraged to act in creative and innovative ways. I will compile and develop a set of theoretical tools which can be used to better explain the existence of decentralization within firms as well as why decentralized firms are sometimes very successful. There will be three main components of my analysis: 1) a systematic application of the knowledge problem within firms, 2) a development of a theory of spontaneous order within firms, based on a realistic accounting of factors affecting employee motivations, and 3) a development of a framework to explain human creativity. 1

2

Employees and entrepreneurship

It is unfortunately not an exaggeration to claim that economists have so far not fully come to grips with the institutional richness and complexity that are evident upon even a cursory look at real-life firms. Little recognition has been given to institutional heterogeneity of firms. Firms are usually studied as abstract institutions that operate under a hierarchical, command-and-control structure, independent of internal self-organization or spontaneous order. The focus has been on understanding the common aspects of firms. This is a valid endeavor, but it will not answer all the questions related to the firm. In particular, a theory of homogeneous firms cannot explain firm decentralization. Decentralization of managerial control, also sometimes known as ‘decontrolling’, is a phenomenon evident in many business organizations in the world today, ranging from small, privately owned ventures all the way to large conglomerates. Management literature has been grappling with explanations of this development for some time, but economics for a long time remained relatively unaffected by it. The theory of the firm usually assumed a hierarchical structure as one of the fundamental features of a firm despite there being in the real world few firms that are absolutely controlled from the top through a strict hierarchy. All firms rely to some extent on forces of self-organization because they must give their employees at least some decision-making ability in order to operate. Thus, modeling firms by assuming one central decision-maker acting through the hierarchical structure may effectively simplify complex organizations of firms in order to explain some aspects of their operations, but it will not explain why firms are organized as they are. Radically decentralized firms force us to recognize the heterogeneity of firm organization as well as the existence of intra-firm self-organization. Given that decentralization can be carried out in many different ways and can vary in depth (some firms going as far as removing all of the middle management to create a two-level hierarchical structure with the responsibilities formerly held by middle managers now entrusted to regular employees), inevitably we are left with firms that exhibit different internal structures and institutional features, some of which do not adhere closely to the hierarchical theory of the firm, built on assumptions of opportunistic and non-creative employees. There is mounting evidence that decision-making within firms is increasingly decentralized and allocated to lower tiers of the managerial hierarchy. There have been many anecdotal indicators showing this in the business and management literature over the years (a particularly good example being Malone 2004). But now we are starting to see some empirical support for it as well. A recent article by Rajan and Wulf (2006) relied on a detailed database of job descriptions of top managers, reporting relationships, and compensation structures to find ‘that the flattening of the

Introduction

3

senior management hierarchy . . . is widespread in the United States among leading firms in their sectors’ (p. 4) .2 They find that there are more managers reporting directly to the CEO, reflecting fewer positions between the CEO and the division heads.3 The authors examine whether this increasing CEO ‘span’ could be due to the natural growth of firms or mergers. They find that neither factor plays a role. They also examine whether the increasing CEO span could be due to profit center units becoming larger and more important and thus needing to report directly to the CEO. Again, the data does not support this hypothesis, as the average size of profit center divisions has actually been decreasing: ‘[E]ven though the structure of the division has not changed drastically over time, its head has moved nearer the top. The organized hierarchy is indeed becoming flatter’ (p. 18). They find that middle management has been shrinking, just as the anecdotal evidence has suggested. This development on its own does not mean that there is greater decentralization of decision-making – it could simply indicate greater control and decisionmaking by top management rather than middle management. But this does not appear to be the case, for the simple reason that as the hierarchies have flattened the individuals who started out at lower levels of the hierarchy also got paid more: ‘[O]ne strong piece of evidence suggests that these changes are not all form without any function: they seem to be accompanied by systematic changes in pay’ (p. 20). The only explanation for increasing pay is that authority and responsibility have been delegated to the lower rungs of the organization and employees are expected to act in more creative ways. Decision-making has become more decentralized and less subject to oversight by higher rungs of management, and in the process raised the pay of regular employees. The phenomenon is real – decentralization of decision-making accompanied by hierarchical flattening is on the rise. Economists are increasingly recognizing this fact and attempting to explain these phenomena. This area is no longer strictly within the domain of business and management literature. But one area remains insufficiently elucidated – the source of innovations within firms. This area will be the focal point of this book. My primary hypothesis is that, in order to create innovations and increase market responsiveness, some firms engage in an experimental process of altering their institutional structure by introducing various degrees of decentralization of control and decision-making. The decentralization is done in order to spur the creative and entrepreneurial impulses of the employees. In some cases, these decentralized forms are appropriate and firms are successful; in others, they are not and firms fail (or revert to a more hierarchical form). The appropriate institutional structure cannot be known with certainty ex ante, but can only be discovered through the

4

Employees and entrepreneurship

process of institutional experimentation guided by intra-firm entrepreneurial insights. Two key elements of this explanation of decentralization bear emphasis: 1) the scope for entrepreneurial (that is, innovative) action (whether by employees or managers) within firms is strongly influenced by their internal structure; 2) in some cases innovations introduced by a firm are the result of creative or entrepreneurial actions of employees rather than owners or managers (who are usually thought of as the primary entrepreneurial actors within firms).

UNDERSTANDING DECENTRALIZATION DEVELOPMENTS We can observe a general pattern in growth of firms. The great majority of firms start out as mechanisms to carry out the entrepreneurial vision of one or few men. As the firm is successful and its complexity grows, it becomes impossible for a single person to run the entire company. Through a hierarchy, the entrepreneur-promoter can delegate certain tasks to the various management levels while the important decision-making remains centralized among the managers or owners. The upside of this structure of control is assured co-ordination of activities of various sections and divisions of the firm, and subsequent orientation of employees toward achieving the common goal. As I will show in Chapter 2, Ronald Coase believed this to be one of the benefits of the firm mode of production. It allows the resource owners to reduce or eliminate the costs of co-ordinating through simple control by the entrepreneur (or, in his words, ‘entrepreneur-coordinator’, who appears to be more of a managerial rather than an entrepreneurial figure in the usual Austrian economics sense). But there is also a downside: all change within a firm (whether due to the entrepreneur-promoter’s own innovative initiative or as a necessary response to competitive pressures) has to take place through the execution of a central plan, as designed by the entrepreneur and/or the strategic management. Thus, any progress by the firm is necessarily constrained by the knowledge of the entrepreneur and/or the strategic management. Ludwig von Mises presents an alternative to the purely centralized view of the firm in his 1944 book Bureaucracy. He shows that, even in gigantic multinationals, individual managers of different sections are still ultimately responsible to the market as well as guided by the market, even if they are not ‘micromanaged’ from above. A firm is divided into divisions, which are further subdivided into other divisions, which leads to a decentralization of responsibility and decision-making. At the head of each division is a manager. This manager must produce profit, or at least produce partially

Introduction

5

completed goods (to be finished by another section) at a lower cost than an outside company could. By holding each section manager responsible for the profit of his section, Mises claims, a large corporation can ensure that consumer sovereignty controls that manager’s actions. As long as each division manager has access to the tools of economic monetary calculation through the existence of market prices and the standard tools of accounting, ‘the general manager of the whole enterprise can assign to each section’s management a great deal of independence’ (p. 33) and decentralization will not result in any loss of ‘the sovereignty of consumers and the democratic operation of the market’ (p. 36). The bottom line, according to Mises is that ‘[the profit motive] joins together utmost centralization of the whole concern with almost complete autonomy of the parts, it brings into agreement full responsibility of the central management with a high degree of interest and incentive of the subordinate managers of sections, departments and auxiliaries’ (p. 36). It does this by simply leaving ‘most of the details . . . to the head of every department’ (p. 34). When we take a look at the business world today (and for the last few decades) we see many firms experimenting (sometimes explicitly, often implicitly) with ways to increase the amount of authority and responsibility not just of the heads of different sections and departments, but, even more radically, of each employee. Some firms, such as W.L. Gore & Associates and McKinsey Consulting, do so to a practical extreme where hierarchy is all but eliminated. Other businesses experiment with less radical but still substantial reductions of hierarchical control of employees. The bottom line is the same, though: top management gives up at least some control, and firms rely on spontaneous order at least to some extent to co-ordinate the actions of employees. Today’s world looks more like the one described by Mises in the pages of Bureaucracy rather than the one described by Coase, with many firms relying on internal self-organization to a greater extent than ever before. W.L. Gore & Associates The aforementioned W.L. Gore & Associates is worthy of a closer look to illustrate these claims. It is a stunningly successful company, producing a wide variety of products, ranging from sophisticated medical equipment to guitar strings to weatherproof clothing (made with the company’s famous GORE-TEX material). It was started in 1958 by Bill Gore and his wife, Vieve, and initially operated from their basement. The 46-year-old chemical engineer quit his job with Du Pont to start his own company because Du Pont refused to take seriously his suggestions on the potential uses of Teflon. In 1998, only 40 years after the founding of the company, its revenue

6

Employees and entrepreneurship

was over $1.1 billion with estimated annual operating profits of $250 million (information taken from Shipper and Manz 1998). This exceptional performance earned the company consistently high rankings on Forbes magazine’s list of the 500 largest private companies in the United States. W.L. Gore & Associates achieved all this with a radically decentralized internal structure, instituted at the very founding of the firm. There is no hierarchy of which to speak, and Gore associates (they refuse to use the word ‘employee’ to describe the people working for the company) have an immense amount of freedom in deciding which projects to pursue, where in the company to work, and so on. This is why the company’s management style has been often referred to as ‘unmanagement’. That W.L. Gore & Associates is a radically ‘flat’ company can be seen by the fact that there are only three persons with actual titles: ‘Bill Gore was chief executive officer for over 20 years . . . As of 1995 the executive hierarchy consisted of a president (Bob Gore [Bill Gore’s son]) and a secretarytreasurer (Vieve Gore) because the two titles were required by the laws of incorporation’ (Shipper and Manz 1998, p. C-500). It is a company ‘without titles, hierarchical organization charts, or any other conventional structural arrangements typically employed by enterprises with hundreds of millions of dollars in sales revenues and thousands of employees’ (ibid., p. C-496). The founder, Bill Gore, described the company’s structure as a ‘lattice organization’ (in contrast to the orthodox ‘pyramid organization’), intended to allow all associates to communicate and cooperate with each other in a spontaneous manner (that is, as they see necessary) and thus to promote creativity among the associates. The lattice structure depends on ‘interpersonal interactions, self-commitment to group-known responsibilities, natural leadership, and group-imposed discipline’ (ibid., p. C-499). The lattice structure is based around continuously fluctuating crossfunctional teams, consisting of associates who voluntarily choose to become a part of a team. It would seem a miracle that anything gets done under this kind of a system, and even the founder once admitted that the whole process is somewhat of a mystery to him, as we can see in the following passage: ‘When a puzzled interviewer told Bill [Gore, the founder] that he was having trouble understanding how planning and accountability worked at W.L. Gore & Associates, Bill replied with a grin, “So am I. You ask me how it works? Every which way” (ibid.). I cannot imagine a better description of an intra-firm process of spontaneous order! Probably the most radical aspect of W.L. Gore & Associates’ ‘unmanagement’ is that the employees choose what work they will carry out. Nobody assigns a place to any associate, and in fact ‘words such as employees, subordinates, and managers [are] taboo in the Gore culture’ (ibid., p. C500).4 When the associates are hired they are given time alone to familiarize

Introduction

7

themselves with the operations of the company and, once they are ready, select the team where they think they can make the greatest contribution. Nobody has the power to deny their participation on the team they choose. The associates ‘were expected to commit to making a contribution to the company’s success’ (ibid., p. C-502), and their pay depends on the magnitude of their contributions to the company. On the other hand, the company has committed to ‘providing a challenging, opportunity-rich work environment and reasonable job security’ (ibid.). The associates are not managed, but are (voluntarily) guided and advised by leaders who naturally emerge in teams. The power of choice in W.L. Gore & Associates rests with the associates, though from the legal perspective the company is privately held. It may seem quite strange for those ultimately responsible for a business to give up their control over it. There obviously must be something to be gained by it. The benefits of decentralization are indirect: decentralization, if done right, intensifies incentives and creates an environment more conducive to creative action and problem-solving by non-management employees, which in most cases actually leads to more creativity and innovation by non-management employees. This is very clear in the case of W.L. Gore & Associates, where the flat structure was a conscious attempt to encourage innovative behavior by the employees: ‘Bill Gore believed great products alone did not make a great company. He felt hierarchy stifled creativity and that therefore it was important to avoid smothering the company in thick layers of formal management’ (ibid., p. C-503). And at Gore, employees must be creative since all operations are innovationdriven: Like everything else at Gore, research and development activities were unstructured. There was no formal R&D department charged with coming up with new products and more efficient manufacturing processes; Associates were expected to be inventive. Empowering Associates to be innovative in developing new products and ferreting out new market opportunities had worked well over the years . . . All Gore Associates were encouraged to think, experiment, and follow a potentially profitable idea to its conclusion. At the plant in Newark, Delaware, Fred L. Eldreth, an Associate with a third-grade education, designed a machine that could wrap thousands of feet of wire a day. The design was completed over a weekend. The year before he died, Bill Gore claimed that at his company ‘the creativity, the number of patent applications and innovative products is triple [that of Du Pont].’ (Ibid., p. C-507)

Gore’s internal organization is a radical and dramatic example of decentralization, one that probably would not prove to be a good fit for most businesses. But by revealing what is possible when decentralization is carried out in the right way and within the appropriate market circumstances, this

8

Employees and entrepreneurship

extreme can be very useful in clarifying our thinking. It also presents an important challenge: if the theory of the firm is unable to explain an organization such as Gore, it is in some ways incomplete.

THE PROS AND CONS OF DECENTRALIZATION To explain why some firms have been engaging in decentralization of decision-making we must recognize two important epistemological points: First, those who are closest to the product or the customer, obviously most often the non-management workers, can be usually expected to have the best insight into how to make that product more attractive to the customer. Second, it is very difficult to know in advance which innovations will capture the imagination – and pocketbooks – of consumers, which is why business history is littered with skeletons of innovations gone awry. Most innovation is at least to some extent a process of trial and error, companies never knowing for certain what will work and what won’t. Given this, it seems clear that the greater the amount of total innovations generated within an organization, the better the odds of having at least a few of them actually appeal to consumers and promote the business. The question then is, how does a firm promote innovative and creative work by their employees that generates the greatest amount of potentially viable innovations? The central hypothesis of this book is that this can be done through a process of decentralization. Innovations are the benefits of decentralization. But as with everything else, this reduction of hierarchical control does not come without cost. If workers are no longer told exactly what to do with every minute of their time, but are expected to come up with creative solutions to the problems facing the business, then there is no longer any guarantee of perfect co-ordination of activities of all the employees and all the departments of the firm. At the very least, such co-ordination is no longer ex ante but rather ex post (as it is in markets), and therefore in some ways probably wasteful as management finds that activities of certain departments are not compatible with the rest of the firm and might need to be abandoned or modified. The firm then has to bear the costs of the wasted resources. The wasted resources can also come from two other sources within decentralized firms: poor decisions and greater shirking by the employees. So, the main questions of decentralization are: 1) Why engage in this loosening of the hierarchical authority (generically labeled as ‘decentralizing’ though also referred to as ‘decontrolling’) in the first place? 2) Why do different businesses engage in different amounts of decentralization? 3) How is it possible for a company to give up almost all of the central control

Introduction

9

and still exist as a company and survive – or even thrive? 4) What then is the role of top management in decentralized firms? Let me tackle the answers in turn: 1) Firms engage in decentralization because the potential benefits of decontrol (greater flexibility, faster response to competition, greater innovation, faster growth, better workplace characteristics) are in some cases greater than the costs (instability of the firm, potential waste of resources due to incompatibility of projects, fragmentation of the company’s purpose, territorialism and lack of cooperation between employees, greater shirking, poor decision-making), which can ultimately be discovered only through a process of experimentation with the institutional structure. 2) Giving up central managerial control only makes sense in a market which exhibits continuous innovative potential, a market where firms must constantly strive to innovate in order to simply survive. The markets that are based on taking advantage of proven technology in order to continue to satisfy consumers and that do not require (or the consumers are not calling for) further innovation will feature firms with much stronger centralization and hierarchical structure in order to ensure the greatest possible co-ordination between all of the resources of the firm, and thus the greatest amount of cost minimization. However, for certain industries cost minimization is not as important as innovation. Furthermore, it is possible that the structure of certain industries is more amenable to decentralization than others. 3) Companies can give up quite a bit of central control and still do well because there exist forces of ‘spontaneous order’ operating within firms that are akin to the ‘invisible hand’ in the economy. Obviously, they are not the same. Spontaneous order refers to ‘the principle by which a beneficient social order emerged as the unintended consequences of individual human actions’ (Vaughn 1987, p. 168). The actions of the employees within a firm are in fact intended for the firm’s benefit, so at first glance it would appear that the principle of spontaneous order does not apply to the firm. However, there is a way of thinking about the concept of spontaneous order that is of relevance to the study of the firm. Vaughn continues: ‘In [Scottish Enlightenment] literature, a spontaneous order was one that had consequences that were mutually beneficial but were nevertheless unplanned by any one person’ (1994, p. 124). In the case of at least some radically decentralized firms, we observe no ex ante planning of any significance. These firms are relying on the innovative actions of their employees and innovation cannot be planned. In addition, Vaughn goes on to say that: spontaneous order can be seen as a process of systematic, ordered change in either the formal or informal rule structures by which people attempt to achieve

10

Employees and entrepreneurship their purposes. In this sense, the spontaneous order is the unplanned and often unconscious changes in rules and institutions that occur as the by-product of purposive action.’ (Ibid., p. 125)

This is of great relevance to decontrolled firms where continuously evolving informal rules play as important a role in governing behavior as formal rules. 4) Managers still have an important role, though they are no longer expected to engage in detailed planning. Instead, they are expected to understand which tools they have at their disposal to ‘empower’ the workers and facilitate their creativity. They must have a good understanding of the evolution of the informal rules within the firm in order to promote the rules that encourage creative work and eliminate the rules that promote free-riding and routine labor. That may entail sometimes formally adopting the successful rules that have been introduced informally by subdivisions of the company; in other cases it may entail actually introducing – or rejecting – certain informal rules. A decentralized firm can really be thought of as a nexus of rules intended to facilitate innovation (which of course requires successful co-ordination of the activities of many different people). This role of management differs significantly from the traditional one. Managers in decontrolled firms no longer monitor employee performance in order to discourage shirking and penalize the employees who do not meet some planned output goals, but rather aid each worker in being more creative. Management literature often uses the phrase ‘management must be a resource for the employee’ to describe this new managerial role. They cannot simply co-ordinate the activities of the employees; they must in some instances educate, in other instances act as a service to match employees with compatible ideas with each other, and in still other instances simply provide suggestions rather than orders. The following comments by Steve Morton, a former head of an engineering group at Tektronix, illustrates this. If one of the employees that Morton was managing had a problem, Morton did not impose his own solution: ‘I may suggest alternatives, but in the end I let him choose his own way. Then I do whatever has to be done on my end to get that alternative to happen. If we have to go over budget I arrange it’ (Pinchot 1985, p. 145). The same principle applied to purchase orders presented to Morton by employees, which he always signed: ‘I may ask what it is, but in the end, even if I disagree, I’ll sign. Their project is their responsibility and they decide’ (ibid., p. 146). It is worth emphasizing that decentralized firms do not – and should not – randomly give up control over employees, simply hoping that this decentralization will unleash their creative impulses. Rather, they create a struc-

Introduction

11

ture of rules that can reduce (if never entirely eliminate) the free-rider problem, while increasing the incentives for the employees to engage in creative rather than routine work effort. To achieve these goals, firms engage in a process of continuing experimentation by instituting different rules and control structures. The important thing to understand is that many of the rules develop through a trial-and-error, evolutionary process and may only be applicable and effective in those firms under those particular circumstances. It is very difficult to say how many of these rules are general and will apply in different situations and in other firms. It is very likely that many of them are specific to their time and place. Also, very importantly, a decentralized firm is not a ‘co-op,’ or a selfmanaged company. It always has an ultimate decision-maker who decides which projects will be carried through to the market and which ones will not. The interesting thing about this, though, is that more than a few ultimate decision-makers in decentralized firms today see the benefits of ‘not putting all of their eggs in one basket’ by allowing some amount of funds to go to projects in which they do not have full confidence, but which exhibit an intense devotion of an individual employee or a group of employees. The most radical form that this decontrol can take is when the decisionmakers in the firm simply allot particular funds to a team of employees and do not interfere in their development process. In some cases, the top management then becomes more like a venture capital firm rather than the controlling authority of the company. And like true venture capital firms, decontrolled firms may continue to fund projects that are not a sure bet because they see the potential for something great. Intrapreneurship The best-selling 1985 management book, Intrapreneuring by Gifford Pinchot III, refers to the funds devoted to employee creativity as well as the time they can devote to their creative pursuits as ‘corporate slack’. Letting employees pursue their own pet projects has often been very successful for companies for a very simple reason: ‘When all corporate resources are committed to what is planned, nothing is left for trying the unplannable. Yet innovation is inherently unplannable. Companies that successfully innovate empower their employees to use corporate resources in ways that cannot always be predicted or justified’ (Pinchot 1985, p. 211). It is clear that creativity requires time and in many cases funds to keep it going. Many companies have embraced this viewpoint, and have instituted programs that provide this corporate slack. For example, Ore-Ida instituted just such a program with great success:

12

Employees and entrepreneurship Every two years Ore-Ida names five fellows, each of whom is given a $50,000 annual budget to fund other employees in the exploration of new ideas. The results have been impressive: – A new computerized scale system funded by a $15,000 fellows grant has already saved more than $2 million. – A $10,000 fellows grant supported engineers in developing a novel heatrecovery system that has already saved $170,000 in one year. (Ibid.,p. 213–14)

Here is another example of corporate slack from Intrapreneuring: At Texas Instruments, managers have three distinct funding options for new R&D projects. If their proposal is rejected by the centralized Strategic Planning System because it is not expected to yield acceptable economic gains, intrapreneurs can seek a ‘wild hare’ grant . . . [designed] to ensure that good ideas with long-term potential were not systematically turned down. Alternatively, if the project is outside the mainstream of Strategic Planning, managers or engineers can contact one of dozens of individuals who hold ‘IDEA’ grant purse strings and can authorize up to $25,000 for prototype development. The briefness of the one-page application form expresses both a commitment not to become bureaucratically slow and a high level of trust in the people they have hired. It was an IDEA grant that resulted in TI’s highly successful Speak-n-Spell learning aid. (Ibid., pp. 214–15)

The bottom line is that nobody can know ex ante exactly what is going to work in the markets and what isn’t, so some firms have adopted a policy of innovation diversification. Providing time and funds for employees to be creative is one way in which firms are decentralizing the decision-making in their organizations. In such cases no longer is the top management making all the decisions about how to spend the company funds. But many companies are going further than that by experimenting with their institutional structures. They are engaging in ongoing learning about the strengths and weaknesses of their institutional configuration through a trial-and-error process, attempting to discover the optimal one for that time and place – meaning, the arrangement of intra-firm rules that will achieve the proper balance between innovation and co-ordination. In order to extract as much creativity as possible out of their workers, some companies are relinquishing managerial power and letting the employees make their own mistakes and engage in their own trial-and-error processes – in other words, allowing them to learn on their own. Ultimately, decentralized firms are a realworld manifestation of an experimental process of discovery of the optimal organizational form – where the ‘optimal’ means one that is most suited to its environment, one that will have the best mix of various tradeoffs, especially the tradeoff between stability and innovation.

Introduction

13

Oticon The Danish hearing-aid company Oticon A/S is a dramatic example of the process described in the previous paragraph. By the late 1980s, the company which had once been an undisputed leader in its industry had fallen behind the competitors. In the late 1970s they had a 15 per cent share in the world market for hearing aids; by 1987, that market share had shrunk to 7 per cent (Foss 2001c). The main reason for this decline was that Oticon was not keeping up with the new technology introduced by their competition (in particular, the new ‘in-the-ear’ hearing aids and the application of digital technology to hearing aids). The company was in trouble: Oticon management had to realize that in an industry where the race to bring the next technologically sophisticated product to the market was fast becoming the competitive criterion, the competition had leapfrogged Oticon in terms of technological developments and in terms of reducing the time length of product development. (Ibid., p. 8)

The added problem was that the prevalent culture in the company was one of incremental technological change. Oticon preferred to rely on proven products and proven technology. However, the markets were changing and that was no longer an option. In 1989 Lars Kolind became the new CEO of Oticon, replacing the executive team that had been in place for 30 years and had made the company into a huge success in the 1970s. That team voluntarily stepped down when it became clear that they were unable to cope with the new problems of the company. After some introductory cost-cutting measures, Kolind decided that ‘something more radical was needed with respect to the strategic orientation of the firm and the administrative systems’ (Ibid., p. 10). In 1990, Kolind undertook a complete overhaul of the company. At the heart of the overhaul was ‘empowering of employees’ through less managerial intervention. Kolind called the new organizational structure the ‘spaghetti organization’ to contrast it with the ‘pyramid organization’ (notice the similarity to W.L. Gore & Associates’ ‘lattice organization’). As Foss explains, Kolind’s goals were clear: [T]he new organization should be able to change rapidly, yet still possess coherence. The new administrative structure should be explicitly ‘knowledge-based’ . . . It should therefore be capable of combining and re-combining skills in a flexible manner where skills and other resources would move to those (new) uses where they were most highly valued, with only minimal intervention on the part of Kolind and other managers being required to secure this aim. (Ibid.)

14

Employees and entrepreneurship

In order to achieve this, the employees were given an unprecedented amount of decision-making power. Kolind eliminated most functional departments, and the company was now almost completely ‘project-based’. Just as in W.L. Gore & Associates, the firm relied on spontaneous order to match employees with the project-based groups: [R]ather than being assigned tasks from above, employees now had a choice to decide which projects they would join. All projects were to be announced on an electronic board, where employees who would like to join them could sign in. The much noted ‘multi-job’ principle meant, first, that employees were not restricted in the number of projects they could [join], and, second, that employees were actively encouraged (and in the beginning actually required) to develop and include skills outside their skill portfolio. (Ibid., p. 11)

Kolind seemed to implicitly accept that employees have a greater knowledge of their skills than the managers, and if motivated properly will insert themselves in those teams where they can make the greatest contributions. In the project-based teams there existed ‘project managers’ but as Foss explains ‘ “management” [was] understood more in terms of playing the role of facilitator and coordinator than that of a directing principal’ (Ibid.). In addition, the company implemented new information technology greatly facilitating communication among employees to better ‘coordinate plans and actions in such a decentralized company’ (ibid.). The spaghetti organization was a great success. All of the goals set by Kolind were met within two or three years of the implementation of the new administrative structure: Yet another positive outcome of the spaghetti organization was that the development time of new products became half of what it used to be . . . Customer orientation, another explicit aim of the spaghetti, also dramatically improved. In 1993, half of Oticon’s sales stemmed from products introduced in 1993, 1992, and 1991. A total of 15 new products had been introduced since the introduction of the new organization. Moreover, the ambition to broaden the business areas of Oticon was successful; it was characteristic of the new products that they were not just hearing-aid hardware, but complete integrated hearing solutions, many of them drawing upon recent advances in digital signal processing technology and embodying sophisticated software. (Ibid., p. 12)

The company was pulled back from the brink of bankruptcy as a result of the radical decentralization. However, by 1995 the mismanagement of the spaghetti organization had led to widespread demoralization and organizational instability, and the company eventually reintroduced (limited) hierarchical structures. The experiment created a lot of innovative behavior but ultimately at too high a price. It would appear that either 1) radical

Introduction

15

decentralization is the proper structure for W.L. Gore & Associates but was not for Oticon at that point in time, or 2) Oticon made mistakes which undermined the successes of the flat organizational form. I will discuss this aspect of the Oticon case in Chapter 3.

MODERN ECONOMIC THEORY OF THE FIRM AND DECENTRALIZATION Despite the increasing examination of decentralization in modern management and organizational literature, economists have approached this issue very cautiously and even skeptically, to the extent that they have even acknowledged it. For a long time most economists would not even consider the possibility that firms where employees were granted some – or even great – decision-making power and whose remuneration was not directly tied to some objective measurement of their performance could function at all. Foss (2001b) is a revealing example: [T]heoretical arguments suggest that emulating market organization inside firms, for example, by radically decentralizing the firm and allocating farreaching decision rights to employees may be hard to accomplish in a successful manner. Unlike independent agents in markets, corporate employees never possess ultimate decision rights. They are not full owners. This means that those who possess ultimate decision rights can always overrule employees. Thus, there are incentive limits to the extent to which market mechanisms can be applied inside firms, and delegation, while not exactly a rare flower, is certainly a very delicate one. (p. 7)

Foss’s quote effectively presents several misconceptions commonly held by modern economists. First, there is a perception that decentralized firms consciously attempt to copy market forms of organization. But decentralization is not necessarily carried out in order to emulate the market. Instead, it is an alternative organizational form which in comparison to a traditional hierarchically run firm features certain advantages and certain disadvantages. Second, most economists believe that, because employees within firms can never legally claim the full fruits of their creative labor and because those with ‘ultimate decision rights can always overrule employees’, the incentives within a firm are insufficient to get employees to act in truly creative, entrepreneurial ways. In addition to this motivational problem, there is also the problem of intra-firm co-ordination of employees’ potentially complex activities in the absence of detailed planning by the management.

16

Employees and entrepreneurship

But as Foss acknowledges in the paragraph above, and as even the most ardent critics of firm decentralization must concede, firms that are organized along these lines do exist. A skeptical economist will offer some possible explanations for this apparent aberration: 1) the organizational form is a fleeting, impermanent experiment which is doomed to failure (that is, it is not an organizational equilibrium), or 2) the firm is not decentralized in reality, but rather the decentralization claims are a part of a public relations campaign, possibly intended to reduce the unionization pressures, or something similar. These explanations are not convincing to anyone who has followed the business literature over the last ten years. Examples of enduring and successful decentralized firms abound. How can they be explained? I believe that we can explain decentralized firms only by extending and even reconsidering some of the building blocks of the theory of the firm. The New Institutionalist theory of the firm may have led us astray to a certain extent by presenting central planning as the salient characteristic of firms and neglecting to account for the decentralization of decisionmaking that exists at some minimum level in all firms. Examining decentralized firms is an effective way of ‘testing’ the real-world applicability of the modern theory of the firm, and subsequently suggesting the ways in which we should modify it. In particular, I will focus on these neglected or even ignored elements of the theory of the firm: the knowledge problem within firms, spontaneous order within firms and creativity within firms. With better understanding of these three elements we can begin to explain why at least some intra-firm decentralization exists and why some radically decentralized firms can even thrive, contrary to today’s dominant theory.

OUTLINE OF THE BOOK The modern economic theory of hierarchical firms is mostly due to the work of two men: Ronald Coase and Oliver Williamson. I will analyze their view of the firm as the antithesis of the market in Chapter 2, and explain their theories of the hierarchical firm (based on top-down planning). I will show that these theories cannot explain existing, real-life decentralized firms because they do not account for dispersal of knowledge within firms. Thus, we must look beyond the Coase/Williamson theory of the firm to explain decentralization. In Chapter 3 I will look at the recent work on the knowledge problem, that is, dispersal of knowledge, within firms. A growing body of literature has recognized the existence of the knowledge problem within firms, and some of it has actually even dealt with decentralized firms from

Introduction

17

this perspective. The contributions in this area have come basically from three economists: Richard Langlois, Nicolai Foss and Alanson Minkler. Though their work was foundational and essential, none of them have gone far enough to fully explain intra-firm decentralization. In Chapter 4 I will explain the role of spontaneous order within firms with decentralized decision-making. Such firms by definition eliminate at least some consciously directed internal co-ordination, and thus implicitly rely at least somewhat on spontaneous co-ordination. But economists often seem to hold the view that the market system is the only example of spontaneous order. They therefore come to the conclusion that, in the absence of the market institutions (private property and money prices), spontaneous order is impossible, or at least unlikely. However, there are other examples of spontaneous order, which emerge without relying on monetary prices. Languages and money itself are two such spontaneous orders. In fact, as Richard Ebeling has shown using the sociological and philosophical insights of Alfred Schutz (see Chapters 4 and 5), the entire market order itself is dependent not only on the above-mentioned market institutions but also on the sociological connections between individuals established through the existence of structures of intersubjective meanings. I will argue that it is these structures of intersubjective meaning (which are embodied in various organizational rules) that are responsible for spontaneously coordinative properties of decentralized firms. Few issues are more important to decentralized firms than that of employee motivation, which I will also consider in Chapter 4. Unfortunately, economists have not done much to realistically explain the complex factors that motivate employees. The only notable exceptions appear in the work of Herbert Simon and Alanson Minkler, both of whom find that employees are motivated by an intricate set of factors outside of simple incentives. Even more, the actions of most employees at most times are far from being ‘opportunistic with guile’, which was Williamson’s fundamental assumption of employee behavior. It is this fact that allows firms to function at all; otherwise the principal–agent problem and the accompanying monitoring and transaction costs would be absolutely overwhelming. In Chapter 5 I will deal with one element of decentralization that has so far remained mostly ignored – the fostering of creative responses among employees. In this chapter I will discuss how delegation of decision-making in firms can bring about greater active creativity among employees. Though the emerging economic literature dealing with decentralization has so far enumerated many of its benefits and costs, little attention has been given to the role of creativity. Creativity (in the sense of introduction of ‘novelty’) within the firm will be explained by relying on Schumpeter’s conception of entrepreneurship as creativity in action. Schumpeter stated that ‘a study of

18

Employees and entrepreneurship

creative response in business becomes coterminous with a study of entrepreneurship’ (1947, p. 223). Much attention has of course been given to Schumpeter’s theory of entrepreneurship, but it has so far not been applied in a way to help us understand the origins of creativity within firms. An important implication of his definition is that it is not only the owners or head managers in a firm who can act in entrepreneurial ways, but rather employees at all levels of the hierarchy. Decentralized firms are decentralized precisely for this reason: they give decision-making power to employees in order to extract creative action from them. Also, Schumpeter’s conception of entrepreneurship holds that in addition to pecuniary factors (which certainly play a role) entrepreneurs are often motivated by other factors, an important one being an innate need to create. This Schumpeterian conception of entrepreneurship can be used to explain why and how workers in decentralized firms will act in creative ways and how internal structures of decentralized firms are designed to encourage this behavior. Finally, Chapter 6 will present some concluding thoughts.

NOTES 1. A notable exception is Sautet (2000). Sautet incorporates the entrepreneurship theory of Israel Kirzner into the ‘capabilities’ and transaction cost theories of the firm in order to show that the multidivisional form of the firm has emerged to spur greater intra-firm entrepreneurial discovery of new opportunities and more effectively solve the Hayekian knowledge problem present within firms. While Sautet focuses primarily on the M-form of the firm (which he refers to as a ‘decentralized form’) to explain how entrepreneurship fits into a firm, this book will deal with decentralization in a broader, more general way. Nevertheless, to the extent that Sautet did address the connection between the structure of firms and intra-firm entrepreneurship, it is an important predecessor to my own work, with which it shares more than a few affinities. 2. The database consists of over 300 large US firms, spanning multiple industries, tracked over a period of up to 13 years (1986–1998). The database comes from the largest private compensation survey to date, carried out by Hewitt Associates, a leading human resources consulting firm. In addition, the database consists mostly of large, established companies, belying the common perception that hierarchical flattening is mostly seen in start-ups and newer high-tech companies: ‘The firms in the sample are large, U.S. publicly-traded firms that are well-established and profitable with average size of approximately 47,500 employees, age of 85 years since founding, and return on sales of 19%. The typical firm in the sample is thus a large mature stable firm, not one whose organizational structure is likely to be in flux . . . The survey participants are typically the leaders in their sectors and, in fact, more than 75 percent of the firms in the dataset are listed as Fortune 500 firms in at least one year and more than 85 percent are listed as Fortune 1000 firms. These firms represent a significant fraction of the activity of publicly-traded firms in the U.S.’ (Rajan and Wulf 2006, p. 9). 3. ‘Our first finding is that the number of managers reporting to the CEO has increased steadily over time, from an average (median) of 4.4 (4) in 1986 to 8.2 (7) in 1998 . . . Our second finding is that the depth, which is the number of positions between the CEO and the lowest managers with profit center responsibility (division heads), has decreased by more than 25% over the period’ (Rajan and Wulf 2006, p. 4).

Introduction

19

4. We can see this also from the interview held with some employees. One said ‘What attracted me to the company was the opportunity to be my own boss and determine my own commitments . . . I felt a strong need to be sure that I was working on something that had value, something that truly needed to be done . . . I did want to make sure that I was helping the company to make money in some way’ (Shipper and Manz 1998, p. C-504). Another said simply ‘You’re unshackled here. There’s a lot less bureaucracy that allows you to be a lot more productive’ (ibid.).

2. The hierarchical theory of the firm [Klein, Crawford and Alchian (1978)] imply that in distinguishing between transactions carried out within the firm and in the market, I believed that all that existed were these polar and clear-cut cases. This is incorrect. Indeed, I was quite explicit about it. In ‘The Nature of the Firm’ I state that ‘it is not possible to draw a hard and fast line which determines whether there is a firm or not. There may be more or less direction.’ (Coase 1988, p. 27) The study of contractual relations plainly involves more than an examination of discrete markets on the one hand and hierarchical organization on the other. As Llewellyn observed in 1931, the exchange spectrum runs the full gamut from pure market to hierarchy and includes complex ‘future deals’ located between market and hierarchy extremes . . . Suppose that transactions were to be arrayed in terms of the degree to which parties to the trade maintained autonomy. Discrete transactions would thus be located at the one extreme; highly centralized, hierarchical transactions would be at the other; and hybrid transactions (franchising, joint ventures, other forms of nonstandard contracting) would be located in between . . . Whereas I was earlier of the view that transactions of the middle kind were very difficult to organize and hence very unstable . . . I am now persuaded that transactions in the middle range are much more common . . . Whatever the empirical realities, greater attention to transactions of the middle range will help to illuminate an understanding of complex economic organization. If such transactions flee to the extremes, what are the reasons? If such transactions can be stabilized, what are the governance processes? (Williamson 1985, reprinted in Putterman and Kroszner 1996, p. 135)

Prior to the ascendancy of the New Institutional Economics, most industrial organization economists worked within the Structure–Conduct– Performance theoretical framework. SCP theory is most often associated with J.S. Bain (1956, 1968). Internal organization of the firm is insignificant in this theory; instead, what is important is the competitive environment of the industry. Firms are modeled as profit-maximizers, and they strive to achieve this goal given the external constraints of the industrial structure (perfect competition, monopolistic competition, oligopoly or monopoly). It is presumed that firms will organize in the best possible way to achieve the greatest profit, given these constraints. This is a useful field of study on its own, but it does not answer many important questions, such as: how important the entrepreneur (in the Schumpeterian sense of an innovator) is to the operation of firms; how important innovations (in the form of 20

The hierarchical theory of the firm

21

R&D) are for a firm; why one firm gains a larger market share while the other loses it; how particular regulations affect markets by impacting the ability of firms to change and adapt; and so on. These are important questions that theory was unable to answer until economists began to peek inside the black boxes of firms. It is no exaggeration to say that the emergence of the economics of the internal organization of the firm is principally due to one man: Ronald H. Coase. His 1937 article ‘The Nature of the Firm’ was so ahead of its time that its implications went unexplored for more than 30 years. This is the right place to start my analysis of hierarchy in the firm. I will also examine the work of Oliver Williamson, who probably made the most important contributions to the Coasian theory of the firm. Williamson’s examination of firms led him to the conclusion that the hierarchical internal structure is the very source of their institutional advantage. In other words, firms have superior transactional capabilities compared to markets under some circumstances precisely owing to the hierarchies found within them. It is important to explain Williamson’s theory since in Chapter 3 I will go on to show that intra-firm hierarchies are not an unmixed blessing and that in some situations it may make sense to weaken them.

COASE’S THEORY OF THE FIRM Coase (1937) began by posing a question: ‘In view of the fact that it is usually argued that co-ordination [of productive resources] will be done by the price mechanism, why is [conscious] organization necessary? Why are there these “islands of conscious power”?’ (p. 35). In other words, he tried to explain why we see so much of the production in the economy organized through the institution of the firm, which is marked by replacement of the price mechanism with the conscious control by an entrepreneurcoordinator. Coase’s great insight is that ‘[i]t can, I think, be assumed that the distinguishing mark of the firm is the supersession of the price mechanism’ (p. 36) and ‘[a] firm, therefore, consists of the system of relationships which comes into existence when the direction of resources is dependent on an entrepreneur’ (pp. 41–2). Coase thus contrasts two ways of achieving coordination of production: through the price mechanism or the entrepreneur-coordinator. In the markets we see co-ordination through the price mechanism, but in the firm we observe co-ordination by direct control of the entrepreneur-coordinator. Coase wants to ‘explain the basis on which, in practice, this choice between alternatives is effected’ (p. 37). His famous explanation is that it is costly to use the price mechanism, and that therefore it is sometimes less costly to co-ordinate through direct

22

Employees and entrepreneurship

control by the entrepreneur-coordinator. The costs of using the price mechanism, the transaction costs, include discovering the relevant prices as well as negotiating and concluding a separate contract for each exchange transaction that takes place on a market. The contracting cost is especially important to Coase. In the continuing market exchanges, the exchanging parties require a series of contracts, each of which is costly to negotiate, draw up, and execute. But within the firm, that relationship can be reduced down to one open-ended, or incomplete, contract. This incomplete contract: is one whereby a factor, for a certain remuneration agrees to obey the directions of an entrepreneur within certain limits. The essence of the contract is that it should only state the limits to the powers of the entrepreneur. Within these limits, he can therefore direct the other factors of production. (p. 39, italics in the original)

It is clear that to Coase the existence of hierarchical control is the central feature of the firm mode of production. Firms exist because it is cheaper to organize production through direct control by a hierarchical management than through voluntary market transactions. It should be pointed out that Coase has a peculiar understanding of the word ‘entrepreneur’. It certainly has nothing to do with the entrepreneur found in the writings of Schumpeter or Knight (and certainly not of the modern Austrian economists). To Coase the entrepreneur is basically a firm owner who acts like a manager, and directs the activities in the firm. Coase even emphasizes this role by referring to him as ‘entrepreneur-coordinator’ in several places of his paper. In another article, Coase explicitly says that management’s function (as a factor of production) ‘was to coordinate’ (1992, p. 715). There is no reference to entrepreneurial innovation or discovery or even uncertainty-bearing. Nevertheless, Coase’s article was groundbreaking simply for placing transaction costs at the center of his analysis. This allowed him to explain that firms exist because ‘central planning’ within the firm can be done under certain circumstances at a lower cost than by relying on market transactions. As important and influential as these insights have been, they are not perfect. From my perspective an important problem is that Coase never considers the possibility that forces of spontaneous order may have a role to play in the organization of a firm. Instead, there is the unspoken assumption that all firms are internally organized through a fully consciously rational and intentional design, illustrated by Coase’s use of a quote by D.H. Robertson: ‘we find islands of conscious power in this ocean of unconscious cooperation like lumps of butter coagulating in a pail of buttermilk’

The hierarchical theory of the firm

23

(p. 35). There would appear to be an implicit assumption that managers possess all the necessary knowledge to correctly co-ordinate resources and organize production. Coase does allow for limits to the abilities of the entrepreneur-coordinator, most importantly as he discusses the limits to the firm-mode of production. He asks the question: ‘Why, if by organizing one can eliminate certain costs and in fact reduce the cost of production, are there any market transactions at all?’ (pp. 42–3). He offers three explanations: 1) ‘as firms get larger . . . the costs of organizing additional transactions within the firm may rise’ (p. 43); 2) ‘as the transactions which are organized increase, the entrepreneur . . . fails to make the best use of the factors of production’ (ibid.); 3) ‘the supply price of one or more of the factors of production may rise’ (ibid.). He says ‘the first two reasons given most probably correspond to the economists’ phrase of “diminishing returns to management”’ (pp. 43–4). These two are the most interesting here. In the first explanation Coase appears to be referring to some sort of a cost of bureaucracy (Sautet 2000, p. 23). He recognizes that there are limits to the managers’ organizing abilities, and at some point the organization becomes too complex to be controlled (whether by a single manager or a group of managers Coase doesn’t make clear). The second limit to the firm’s growth is based on the imperfect knowledge and fallibility of the manager, which get worse as firms grow. Coase thus concludes that ‘all changes which improve managerial technique will tend to increase the size of the firm’ (p. 46). Despite the limits on managerial capabilities, Coase’s view was that firms are an example of successful central planning on a microeconomic level, as we see in this quote: ‘How did one reconcile the views expressed by economists on the role of the pricing system and the impossibility of successful central economic planning with the existence of management and of these apparently planned societies, firms, operating within our economy?’ (1992, p. 715). Though he says ‘apparently’, Coase never engages in a deeper discussion of whether firms truly qualify as centrally planned ‘societies’. Thus, Coase appears to be saying that within existing firms a ‘division of knowledge’ does not exist or is at least not pertinent: there is no dispersion of relevant pieces of information among many different individuals separate in time and space, but rather all relevant knowledge of the productive task is held by the entrepreneur-coordinator, that is, manager, who can successfully co-ordinate the resources, including the labor resource, through some form of a central plan to achieve his goal. Interestingly enough, Coase in a later article allows that the existence of transaction costs (costs of using the market and the pricing system) ‘implies that methods of coordination alternative to the market, which are themselves costly and in various ways imperfect, may nonetheless be preferable

24

Employees and entrepreneurship

to relying on the pricing mechanism, the only method of coordination normally analyzed by economists’ (1992, p. 715, emphasis added). There is recognition that, first, the intra-firm ‘central planning’ is not perfect and, second, there may be more than two ways of co-ordinating productive activities (the price mechanism and central planning), and that therefore co-ordination within the firm can be a result of something other than internal central planning. Unfortunately, he did not specify what that may be.

WILLIAMSON’S CONTINUED DEVELOPMENT OF THE COASIAN RESEARCH PROGRAM Coase’s work did not inspire many economists initially, and until relatively recently (the mid-1970s) very few had anything to say about the internal organization of firms. The major reason for this state of affairs was that despite Coase’s insights few economists thought about the firm in behavioral ways; rather, they were mostly concerned with technological and production aspects, such as economies of scale and technological non-separability (nonseparation of contribution of effort, as was discussed by Alchian and Demsetz 1972). But one economist followed a different path. Oliver Williamson held that ‘the interesting problems of labor organization involve the study of transactions and contracting and, except in a rather special idiosyncratic sense, do not turn mainly on technology’ (1975, p. 58). Williamson carried on Coase’s work and advanced his insights, even if only after a 30plus-year delay. Williamson has probably contributed more than anyone else in the last four decades to the theory of the firm and in the process has grounded it firmly on Coasian foundations. Williamson analyzed firms as institutions designed to minimize costs of transactions, but he held that Coase had not gone far enough in explaining the nature of these costs. As Coase himself wrote, describing Williamson’s critique of his own work: Williamson has also pointed out that although I was correct in making the choice between organization within the firm or through the market the centerpiece of my analysis, I did not indicate what the factors were that determined the outcome of this choice and thus made it difficult for others to build on what is often described as the ‘fundamental insight’. This also is true. But the interrelationships which govern the mix of market and hierarchy, to use Williamson’s terms, are extremely complex, and in our present state of ignorance it will not be easy to discover what these factors are. What we need is more empirical work. (1992, p. 718)

Williamson has done some of that empirical work, but more importantly he has introduced new theoretical insights that have significantly advanced

The hierarchical theory of the firm

25

our understanding of the firm. He has arguably done more than any other economist to focus our attention on internal organization of firms for its own sake, owing to internal organization playing the central role in how firms are different from markets: I furthermore regard organization form – by which I mean the hierarchical structure of the firm, the way in which internal economic activities are decomposed into operating parts subject to internal controls – to be distinctly interesting and warranting separate attention. (1975, p. 8)

It is in large part due to Williamson’s work that internal organization has become an accepted topic of inquiry for economists, in great contrast to 30 or 40 years ago. As one can glimpse in the quote above, in Williamson’s world the choice of organization of production continues to be quite stark: the market, that is, the invisible hand working through the pricing system, or central planning within the firm. In fact, Williamson uses the terms ‘firm’ and ‘hierarchy’ interchangeably, singling out the method of control within the firm (that is, central planning carried out through a hierarchical management structure) as its most important characteristic. This view has much to recommend it, for undoubtedly the vast majority (if not in fact all) of existing firms are in fact hierarchies. However, just like Coase, Williamson never considers the possibility that firms, even those with apparently hierarchical internal organizations, can sometimes rely on spontaneous processes to achieve a state of co-ordination. Williamson’s model is important and mostly right, but incomplete because of this. In the sections that follow I will examine his analytical framework, as well as his views on the benefits and costs of the hierarchical structures of firms. Williamson’s Basic Model The basic Williamsonian model is very distinct and has been highly influential in the New Institutional Economics/Transaction Cost Economics. He first introduced it in the 1975 book Markets and Hierarchies, and has used it in many of his subsequent articles and books (for example 1986, 1991, 1996, 2000) in a relatively unchanged form. The model is based on several behavioral assumptions that are different from textbook orthodoxy. Williamson grounded his theoretical work on a realistic model of human behavior which recognizes the constraints on rationality, whether through inability to gather all relevant knowledge or through constraints of the cognitive power of our brains. Both themes are outside of the mainstream neoclassical theory, which Williamson clearly

26

Employees and entrepreneurship

recognizes and embraces.1 As Williamson puts it, ‘more self-conscious attention to rudimentary human attributes is essential if we are to accurately characterize and more adequately understand many of the problems of markets and hierarchies’ (1975, p. 2). Williamson was also under the influence of Coase in regard to behavioral assumptions, often referring to the neoclassical view of man’s rationality as ‘hyperrationality’ and then quoting Coase where he calls for a study of ‘man as he is, acting within the constraints imposed by real institutions’ (Coase 1984, p. 231).2 Williamson begins his analysis by assuming that agents act with bounded rationality, defined as behavior that is ‘intendedly rational, but only limitedly so’ (Simon 1961, p. xxiv, italics in the original, quoted from Williamson 1996, p. 56).3 Williamson’s primary influence in this area was Herbert Simon, who explained bounded rationality in the following way: ‘The capacity of the human mind for formulating and solving complex problems is very small compared with the size of the problems whose solution is required for objectively rational behavior in the real world’ (1957, p. 198, italics in original). Williamson uses the assumption of bounded rationality to enlarge the scope of rationality analysis, driven as he is to understand the ‘man as he is’. The second step of Williamson’s model joins the assumption of bounded rationality with ‘opportunism’, which he describes as self-interest seeking with guile. Self-interest is, of course, the fundamental behavioral assumption in economics, but Williamson explicitly adds ‘with guile’ in order to emphasize the fact that human beings, if given a chance, will inflict harm on others to benefit themselves, something that the economic science does not usually emphasize or even recognize. Nevertheless, this assumption becomes the driving force of Williamson’s model. When these two are combined, many potential transactional problems emerge. Most real-world transactions are not spot transactions, but are instead forward-looking (whether taking place in the future, or taking place repeatedly over some period of time). Therefore, they require some sort of a promise or commitment. Given that actors are self-interested with guile, these promises must be made binding, preferably by making them enforceable in the court of law. Written contracts have evolved to carry out this role. However, since all of us are also bounded in our rationality, it is impossible to account for every future contingency in contracts that we write, and therefore all contracts are necessarily incomplete. As Williamson (1996) sums it up, ‘taken together, the lessons of bounded rationality and opportunism lead to the following combined result: organize transactions so as to economize on bounded rationality while simultaneously safeguarding them against the hazards of opportunism’ (p. 254). This is where the firm comes in, which does exactly that, as we will see shortly.

The hierarchical theory of the firm

27

One more factor needs to be added, however. Even in the presence of both bounded rationality and opportunism, economic actors could always simply leave and not engage in further exchange with the party that has done them wrong. Unfortunately, the situation is often complicated by the presence of what Williamson calls asset specificity: Asset specificity has reference to the degree to which an asset can be redeployed to alternative uses and by alternative users without sacrifice of productive value . . . Asset specificity not only elicits complex ex ante incentive responses but, even more important, it gives rise to complex ex post governance structure responses. (1996, p. 59)

It is a necessary precondition for opportunism to become destructive. ‘Asset specificity’ is a term that Williamson only adopted in his later writings. In his earlier writings he referred to this problem as one of ‘smallnumbers exchanges’. Most transactions are initially large-numbers transactions, meaning that there are many competitors all willing to provide a particular asset at the smallest price. However, once a firm settles on one supplier of a particular asset, the transaction undergoes what Williamson terms a ‘Fundamental Transformation’, which in turn will lead to ‘bilateral dependency’. Fundamental Transformation can be thought of as a situation where ‘many transactions that at the outset involve a large number of qualified bidders are transformed in the process of contract execution, so that small-number supply condition effectively obtains at the contract renewal interval’ (1975, pp. 9–10). Competition in supply of an asset is weakened because that asset becomes more differentiated to the buyer, and there are no longer perfect substitutes. There is ‘bilateral dependency’ due to the fact that both the buyer and the asset supplier alter their production processes and thus make their resources more complementary in comparison to the situation that prevailed prior to the Fundamental Transformation. If there were no opportunism, this would not be a problem. But as Williamson points out: [w]hen . . . opportunism is joined with a small-numbers condition, the trading situation is greatly transformed. The transactional dilemma that is posed is this: it is in the interest of each party to seek terms most favorable to him, which encourages opportunistic representations and haggling. The interests of the system, by contrast, are promoted if the parties can be joined in such a way as to avoid both the bargaining costs and the indirect costs (mainly maladaptation costs), which are generated in the process. (1975, p. 27)

Thus, the buyer of an asset and the asset supplier will find themselves in a sticky situation where one party can take advantage of the situation to

28

Employees and entrepreneurship

obtain the greater part of the surplus value that is created through the transaction. The hazards are especially great to the buyers because the asset supplier will likely have cost advantages over the other potential suppliers, leading to the small-numbers condition. There are hazards to the suppliers, too, in the form of the emergence of a possible monopsony situation. Williamson’s view is that the institution of the firm has evolved as a way to solve this dilemma created by the simultaneous presence of bounded rationality, opportunism, and asset specificity. Sautet (2000) provides a cautionary critique of Williamson’s conception of the firm, which should be mentioned here: Williamson’s theory of the firm and integration depends almost exclusively on a behavioral assumption . . . [T]he question here is to know whether or not human nature is opportunistic the way he describes it . . . This implies that, in a totally honest world, firms would not exist because any contract would be selfenforcing. (p. 28)

Though it appears that Williamson’s assumptions are realistic, it is by no means a matter of certainty that human beings will always act as these assumptions would indicate. In fact, there is some evidence that people are at least sometimes not opportunistic in this way, and that they will engage in transactions as though, to use Williamson’s terminology, the largenumbers conditions are prevailing even when they are not. Simon (1991) poses the dilemma in the following way: In most organizations, employees contribute much more to goal achievement than the minimum that could be extracted from them by supervisory enforcement of the (vague) terms of the employment contract. Why do employees not substitute leisure for work more consistently than they do? Why do they often work so vigorously for the welfare of the organization? (p. 32–3)

I will discuss this in more detail in Chapter 4. For now, it is sufficient just to point out this potential problem with Williamson’s model of the firm. Governance It is at this point that we can put the pieces of the firm together. The basic economic ‘environment’ as represented by human characteristics of bounded rationality and opportunism must be joined with the technological state that Williamson calls asset specificity. Alchian and Woodward (1988) explain this Williamsonian situational setting in their own effective way:

The hierarchical theory of the firm

29

The degree to which resources are vulnerable to morally hazardous exploitation depends on what can be called their ‘plasticity’ and on monitoring costs. We call resources or investment ‘plastic’ to indicate that there is a wide range of discretionary, legitimate decisions within which the user may choose. For example, compare a drug research laboratory with a steel manufacturer. There are fewer options for discretionary behavior in steel manufacturing. The technology is largely determined by the nature of the plant. Absentee owners and the debt holders have little cause to worry about the managers turning the resources of the plant into personal consumption, or increasing the riskiness of prospective outcomes. In contrast, a drug research firm could be working on some mundane project with a modest but sure payoff, or on some long shot with a slight chance of a high payoff. Research and development firms are plastic. (p. 69)

Alchian and Woodward then go on to emphasize that ‘plasticity must be combined with high monitoring costs to result in opportunities for moral hazard’ (p. 69). If all these factors prevail, suppliers of particular resources will have to somehow be restrained from taking advantage of the buyers of the resource, and thus a firm is born. A firm is a way to reduce moral hazard problems, and align the incentives of the buyers and sellers of productive resources. Buyers of resources can enter into an open-ended, incomplete contract that gives them the authority to command the owners how the resources will be used. Because the resource owners recognize the mutually beneficial nature of this arrangement, they agree to turn over control to the buyer. Since the greatest moral hazard dangers exist when the resource in question is labor, firms will engage in heavy monitoring of the workers to ensure that the workers are carrying out their orders. Of course, different degrees of monitoring are necessary in different industries, depending on the ‘plasticity’ referred to above. Nevertheless, even in those industries that feature high monitoring costs, monitoring is still cheaper when a transaction is done within a firm than through the market. In Williamson’s theory the hierarchical nature of firms is the logical outcome of the very existence of firms. Firms exist because they have hierarchies that are able to reduce the hazards of opportunism: If . . . it is very costly or impossible to identify future contingencies and specify, ex ante, appropriate adaptations thereto, long-term contracts may be supplanted by internal organization. Recourse to the latter permits adaptations to uncertainty to be accomplished by administrative processes in a sequential fashion. Thus, rather than attempt to anticipate all possible contingencies from the outset, the future is permitted to unfold. Internal organization in this way economizes on the bounded rationality attributes of decision makers in circumstances in which prices are not ‘sufficient statistics’ and uncertainty is substantial. (1975, p. 9)

Williamson thus specifies the fundamental transaction costs that are responsible for production being organized within the institution of the

30

Employees and entrepreneurship

firm, the costs that Coase could only discuss in very general terms. To Williamson firms exist in order to handle the problem of uncertainty inherent in our world. Bounded rationality requires incomplete contracts given that full contract specification of all feasible contingencies is impossible. But incomplete contracts cannot handle the inevitable and necessary adjustments to unexpected and unspecified circumstances. One solution to this problem is to replace the contract with a hierarchical organization which can be trusted to internally adjudicate disputes in a fair and trustworthy way and not take advantage of the vulnerability of the resource suppliers while creating the requisite flexibility that would not exist under state contingent contracts. The reason resource owners can trust firms in this way is because of the reputational effects. If a firm acquires a reputation of acting in an opportunistic manner, it will be unable to secure the future supply of needed resources. Williamson summed up his argument for the existence of firms in the following way: The markets and hierarchies approach attempts to identify a set of environmental factors which together with a related set of human factors explain the circumstances under which complex contingent claims contracts will be costly to write, execute, and enforce. Faced with such difficulties, and considering the risks that simple (or incomplete) contingent claims contracts pose, the firm may decide to bypass the market and resort to hierarchical modes of organization. Transactions that might otherwise be handled in the market are thus performed internally, governed by administrative processes, instead. (1975, p. 9)

Obviously, Williamson is very much in the Coasian spirit here, identifying the marginal conditions necessary to substitute firm control (that is, hierarchy) for market transactions. The bottom line here, very much extending the argument of Coase (1937), is that, as Williamson puts it, ‘markets and firms are alternative instruments for completing a related set of transactions . . . whether a set of transactions ought to be executed across markets or within a firm depends on the relative efficiency of each model’ (1975, p. 8).4 And Williamson has now specified on which basis we can judge this relative efficiency. Williamson has come to use a very handy term for these ‘alternative instruments for completing a related set of transactions’ – he calls them governance structures. Governance basically refers to an organizational construction of a set of transactions. Williamson lists the institutions of governance as consisting of the four following groups: markets, hybrids, hierarchies, bureaus. Transaction Cost Economics describes these alternative modes of governance ‘as syndromes of related attributes, on which account governance structures differ from one another in discrete structural ways’ (1996, p. 7).

The hierarchical theory of the firm

31

Economic actors rationally choose governance structures depending on the nature of the transaction, that is, depending on their economic context or circumstances: ‘economic agents purportedly align transactions with governance structures to effect economizing outcomes’ (1996, p. 5). They do so because, though bounded in rationality, they are able to engage in learning, ‘foresight’, and problem-solving: Transaction cost economics assumes that economic actors have the capacity to look ahead and recognize contractual hazards and investment opportunities. Often, however, the requisite recognition will come as a product of experience. Whether positive or negative, the basic proposition is that, once the relevant features have been disclosed, the firm will react to such knowledge by taking actions that mitigate future hazards and more fully realize future gains. Learning through experience – by discovering more about the environment and suppliers and rivalry, after which appropriate adaptations are worked out – is more ambitious than merely trial-and-error learning but is less ambitious than the idea of farsighted contracting to which I referred earlier. (2000, p. 47)5

In this way Williamson is able to use the governance structure as the main explanatory apparatus for most questions relating to the firm (its existence, its boundaries, and so on). Once he has established the fundamental rationale for the very existence of the firm, he extends that rationale to every aspect of firm organization. As he explains it, ‘[t]he study of governance is concerned with the identification, explication, and mitigation of all forms of contractual hazards’ (1996, p. 5, italics in the original). This in turn leads to asking the following basic questions: ‘(1) . . . what the dimensions are on which transactions differ that present differential hazards. It further (2) asks what the attributes are on which governance structures differ that have hazard mitigation consequences. And it (3) asks what main purposes are served by economic organizations’ (1996, p. 13). The purpose of the whole enterprise is to provide refutable implications for the institutional study of the firm by allowing the firm’s ‘marginalist logic’ to be tested.6 Williamson’s framework has caused a revolution in economics, and New Institutional Economics would be much weaker without it. And it has certainly enlarged our understanding of the firm, while simultaneously leading economists to start paying much more attention to the black box of the firm. Nevertheless, in some ways it is an overly restrictive model. The fact is that this ‘governance study’ is focused on the hazard mitigation aspect of firms to the extent of ignoring all other potential roles or rationales for the institution of the firm.7 I will explore the role that knowledge creation and co-ordination play in the existence of firms in Chapters 3 and 5.

32

Employees and entrepreneurship

Limits and Strengths of Hierarchies from Williamson’s Perspective Williamson must deal with the fact that there are constraints on the extent of firm organization. The basic question is ‘If firms have such attractive properties for economizing on transaction costs, what prevents the firm from preempting the market quite generally?’ (1975, p. 117). As I discussed previously, Coase grappled with this same issue, and Williamson’s answer is very much in the Coasian spirit: the marginal costs of ‘the hierarchical mode of organization’ become greater than its marginal benefits at some point. This question of the limits to vertical integration is really one of the costs of the hierarchical structure of organization, and Williamson groups these costs into three loose categories: 1) general, non-size-dependent factors; 2) size-related factors; 3) incentive factors (1975, p. 118). The first limit to vertical integration is quite familiar to students of the socialist calculation debates.8 This limit is due to managers being unwilling to act in profit-seeking ways if that would lead to the diminishing of their status, or even loss of their jobs. In other words, managers are above all selfinterested, and they will act first and foremost to improve their own situation. Williamson refers to this as ‘systems rationality versus subgroup rationality’, but today it is usually known as the agency problem. The agency problem can also be manifested in ‘sub-optimal’ decisions when managers are influenced by social connections (such as friendship) within a firm. This situation can be made even worse if there exist expectations of future reciprocity. Also, at times of disputes between different sections of the firm we may observe managers engaging in compromises intended to preserve the size of their departments rather than increase the profitability of the firm. Another problem is that managers may be unwilling to admit failure, instead escalating their commitment to a failing project. This is more likely to be a problem in firms than in markets since failing projects can be funded by cross-subsidization by other (successful) parts of the firm. The last type of a non-size, general limit to vertical integration Williamson calls ‘communication distortion’. By this he means a situation where managers purposely lie or at least distort the situation in order to promote their personal goals. Williamson points out that this distortion can take two forms, defensive (tell the boss what he wants to hear) and assertive (tell the boss what you want him to know). Both are especially likely if there exists a prospect of promotion. It is for all the above reasons that at some point a bureaucracy will become too unwieldy and it will be cheaper to resort to market transactions rather than try to remain inside the firm. Williamson’s conclusion is that:

The hierarchical theory of the firm

33

although internal organization is (potentially) better able than the market to distinguish between random events and meritorious performance, and in this respect is superior to the market as a mechanism for assigning rewards to deeds, the inference difficulties that internal auditing experiences as the firm grows in complexity eventually limit its power in this respect. (1975, p. 125, italics in the original)

Williamson’s view is that, though a hierarchical structure will never eliminate all opportunism, it can significantly reduce it: ‘whenever a market transaction is shifted from the market to the firm, a tradeoff is struck in which it is understood that some irreducible degree of opportunism will continue’ (1975, pp. 125–6). In other words, a hierarchy solves one set of problems (opportunism from asset specificity) by introducing a different set of problems (agency-related incentive misalignment). Given that firms exist, Williamson concludes that the magnitude of the former problem must eclipse the magnitude of the latter in all those cases where firms actually exist. Williamson also considers two other types of limits to vertical integration in addition to those due to managerial self-interest. The second type is size-related factors. The obvious problem is that information has to be communicated through more levels of hierarchies, and therefore even free of attempts at strategic distortions it will inevitably be distorted, in the sense that there will be a loss of meaning of pertinent information.9 More importantly than distortion of information in large hierarchies will be the problems of bureaucracy: ‘bureaucratic insularity varies directly with organizational size’ (1975, p. 127). As firms grow larger, bureaucracies will become more entrenched, more powerful, and larger, the reason being that there are greater opportunities for managerial discretion in larger firms, which often leads to ‘indulging in corporate personal consumption’. Of course, managers can be replaced either by boards or through takeovers, but this still remains a constraint on firm size. Finally, the third type of limit to vertical integration concerns the incentive constraints of the employment relation. Williamson believes that it is impossible (or at least very difficult) for firms to award large bonus payments for entrepreneurial activities because this disrupts ‘strict correspondence between income and hierarchy’ (1975, p. 129), which is the foundation of the firm. Awarding such bonuses could lead to greater demands for quid pro quo payments, which could subsequently greatly increase the transaction costs – and therefore undermine the very purpose of a hierarchy. Williamson points out that it is for this reason that the ‘entrepreneurial types’ are not often to be found in large corporations, and also that: the large firm is frequently at a disadvantage to the small enterprise in supporting early stages of development – because, among other things, of the bureaucratized

34

Employees and entrepreneurship reward structure in the large firm which relies on salary and promotion rather than direct participation in the earnings associated with successful innovation. (1975, pp. 129–30).

I will show in later chapters, though, that in fact many firms today are trying to find ways in which to do exactly this in order to attract the ‘entrepreneurial types’ and become more competitive with smaller firms. The bottom line in Williamson’s own words is that ‘[t]he shift of a transaction or related set of transactions from market to hierarchy is not all gain . . . Flexibility may be sacrificed in the process and other bureaucratic disabilities may arise as well’ (1975, p. 40).10 Selective Interventions Williamson revisits this topic of ‘why firms are superior to markets’ in his subsequent work in the form of the issue of ‘selective intervention’, by which he means ‘that each production stage is directed to perform in the pre-acquisition manner except when misalignments occur and the substitution of authority for autonomy yields net gain’ (1991, p. 100; see also 1985, p. 133). In other words, selective intervention means that the commanding office always has the option of choosing ‘autonomy or authority’ for the controlled stages. Thus, autonomy (or, more precisely, semi-autonomy) of different operating levels can still be achieved under hierarchy, but with the added benefit that self-interested bargaining by the agents can be avoided on those occasions when it threatens to be disruptive or destructive.11 The theory of selective intervention returns us to the basic dilemma: why there isn’t perfect vertical integration across the entire economy, since with selective intervention a large firm could always do at least as well as many small firms. As Coase put it: ‘Why is not all production carried out in one big firm?’ (1937, p. 43). Williamson rejects the non-strategic communication distortion argument this time around. He points out that this argument does not allow for ‘selective intervention’, but rather assumes a fully top-down-managed firm: Internal organization need not, however, adopt this structure. Suppose instead that the parent firm deals with each of its parts by exercising forbearance with respect to those activities where no net gains are in prospect (in which event the parent directs the operating part to replicate small firm behavior) and intervenes wherever coordination yields net gains. (1985, p. 135)12

In this case, Coase’s question remains unanswered – in the presence of selective intervention, one big firm should always be superior to many small ones.

The hierarchical theory of the firm

35

Williamson solves this dilemma by concluding that ‘selective intervention’ is simply not possible. As he himself admits, the reasons for this get a bit involved (1991, p. 102), so I will just summarize the argument (which can be found in full detail in 1985, pp. 135–61). Williamson’s main assumption (and, one could say, a requirement to show that selective intervention doesn’t work) is that the ‘selectively de-intervened’ stage of production gets to keep all of its profits (‘preserves high-powered incentives’). More precisely, the manager of the division is assumed to be the residual claimant – though, obviously, not the owner. According to Williamson, when the manager is not the owner of a vertically integrated division (though he may be a residual claimant), opportunism will kick in: he will skimp on employing labor, instead overusing physical capital in order to maximize short-run profits, and then quit. The checks against these abuses are of two kinds, and are both dismissed by Williamson: 1) an owner could engage in monitoring, but then there will be monitoring costs which were not borne before vertical integration; 2) reputational effects might prevent managers from engaging in such opportunistic actions for fear of difficulty of finding a job in the future, but reputation is at best an imperfect tool to prevent opportunism and often the downsides of a poor reputation are not large enough to overcome high short-term profits. Therefore, there are no effective solutions to the problems of selective interventionism. This is how Williamson arrives at the conclusion that there is no tendency towards a meta-monopoly of the sort that Karl Marx envisioned. Opportunism rears its ugly head again, this time through the bureaucracy necessarily present in the hierarchical structure of the firm rather than through imperfectly specified contracts. In the 1991 article Williamson also appealed to Grossman and Hart (1986), who showed that vertical integration leads to investment distortions compared to market organization. He emphasized the fact that their conclusion ‘makes no appeal to any added costs of bureaucracy’ (1991, p. 101) and thus is not an alternative but rather a complementary explanation to his own. Through these two explanations Williamson basically tried to specify the marginal conditions, first posited by Coase, that decide whether transactions will take place within a firm or a market, as we see here: ‘Thus construed, the basic tradeoff is between comparative bureaucratic costs (where the market enjoys the advantage) and comparative adaptive capacity (where, under bilateral trading, internal organization enjoys the advantage). This tradeoff switches from net negative to positive as the condition of asset specificity deepens’ (Williamson 1991, p. 103). Also: ‘The basic market and hierarchy trade-off that is incurred upon taking transactions out of markets and organizing them internally substitutes one form of inefficiency (bureaucracy) for another (maladaptation)’ (Williamson 1996, p. 237).

36

Employees and entrepreneurship

Hybrid Organizations One conclusion that can be drawn from Williamson’s critique of selective interventionism (and from the two quotes above especially) is that there are only two relevant alternative modes of production: pure hierarchy or pure market transactions. Though Williamson does refer to ‘hybrids’ on occasion, he rarely deals with them in any detail – he basically exhibits a blackand-white view of economic institutions, as demonstrated here: ‘Note, moreover, that the relevant comparative institutional choice . . . is only twoway, market or hierarchy’ (1991, p. 103). His one discussion of hybrids came relatively late, in 1996, and only as a response to criticisms.13 Williamson defines hybrids as ‘various forms of long-term contracting, reciprocal trading, regulation, franchising, and the like’ (1996, p. 104).14 They are organizational forms that are intermediate between markets and hierarchies in almost all ways, and that are in some ways superior to a hierarchy, and in others inferior: The hybrid mode is located between market and hierarchy with respect to incentives, adaptability, and bureaucratic costs. As compared with the market, the hybrid sacrifices incentives in favor of superior coordination among the parts. As compared with the hierarchy, the hybrid sacrifices cooperativeness in favor of greater incentive intensity. (1996, p. 107)

Here is a very clear statement of the institutional tradeoffs that organizations have to weigh when deciding the proper level of decentralization of decision-making. It is worth noting that at the very end of the 1996 book Williamson considers a connection between innovation and hybrids. Williamson claims that some hybrids, such as joint ventures or ‘parallel R&D’, are better at ‘timely entry’ when ‘events are fast-moving or if learning-by-doing is essential’ (1996, p. 118). Williamson concedes that ‘added apparatus is needed to deal with the full set of issues that arise when responsiveness in real time, rather than equilibrium contracting, is the central concern’ (ibid.), but he does not propose what that apparatus may be. His last words on this matter are that some hybrids, like joint ventures, may in fact be transient forms of governance intended to temporarily increase the responsiveness of firms. Thus, they are necessarily examples of institutional disequilibria. Given that his own model is (self-consciously) only concerned with studying equilibrium phenomena, it would seem to be irrelevant to the study of hybrids. For many real-world experiments in hybrid forms of governance the overriding objective is to stimulate innovation, therefore rendering Williamson’s explanatory framework largely inappropriate.

The hierarchical theory of the firm

37

Williamson and Hayek’s Knowledge Problem The issue of hybrids is strongly connected to the issues of knowledge dispersion. As I mentioned earlier, Williamson was influenced by Hayek in this area, especially with regard to the existence of ‘idiosyncratic’ knowledge. The following quote demonstrates the debt he owed to Hayek in this area: The purpose is to better assess the employment relation in circumstances where workers acquire, during the course of their employment, significant job-specific skills and related task-specific knowledge. What Hayek referred to as knowledge of ‘particular circumstances of time and place’ (1945, p. 521) . . . thus play[s] a prominent role in the analysis. (1975, p. 57)15

It is to Williamson’s credit that he recognized the importance of idiosyncratic knowledge, but to him it has different implications than to Hayek. Hayek famously explained that markets respond to the dispersal of knowledge of time and place by assigning the economic decision-making to those that contain that idiosyncratic knowledge, in a fundamental sense decentralizing that decision-making. Furthermore, in markets prices play an important role as guideposts to those dispersed decision-makers by conveying some economically relevant knowledge. In contrast, Williamson was not interested in the issue of idiosyncratic knowledge to explain how markets work, but rather to examine its implications to the existence of the firm, that is, hierarchy, as is evident in the following quote: ‘Given bounded rationality, uncertainty, and idiosyncratic knowledge, I argue that prices often do not qualify as sufficient statistics and that a substitution of internal organization (hierarchy) for market-mediated exchange often occurs on this account’ (1975, p. 5).16 Williamson turns Hayek’s insights upside down. Whereas Hayek’s conclusion is that in the presence of idiosyncratic knowledge decision-making should be decentralized and assigned to the people holding the relevant knowledge, Williams comes to the conclusion that idiosyncratic knowledge can (and most likely will) be used in opportunistic ways, which means that the hierarchical internal organization is necessary to dissipate the risks of ‘hold-up’: [Incumbent employees] are in possession of a valuable resource (knowledge) and can be expected to fully and candidly reveal it only in exchange for value. The way the employment relation is structured turns out to be important in this connection. The danger is that incumbent employees will hoard information to their personal advantage and engage in a series of bilateral monopolistic exchanges with the management – to the detriment of both the firm and other employees as well. (1975, p. 63)

38

Employees and entrepreneurship

Thus, to Williamson the problem presented by the presence of idiosyncratic knowledge is once again in the end a problem of the possibility of opportunistic actions by the employees containing that idiosyncratic knowledge. It is not about how to best align the decision-making with the holder of the idiosyncratic knowledge, as it would be to a Hayekian economist. A Wall Street Journal article from July 1, 2002, ‘Tricks of the Trade: On Factory Floors, the Workers Hide Secrets to Success’, has special significance to Williamson’s explanation of the potential problems of idiosyncratic knowledge. The article describes how factory workers gain idiosyncratic knowledge through sheer accumulated experience. As they gain that edge, they sometimes act opportunistically in two ways: one, by hiding their true maximum productivity in order to be able to enjoy some leisure; two, by refusing to share their ‘tricks’ in order to keep their job security. Both would seem to support Williamson’s explanation of the downsides of idiosyncratic knowledge. However, upon a closer look it turns out there is more to this story. What emerges in the article is that employees often refuse to share their superior knowledge in fear that the company will use that knowledge to replace them. More interestingly, they sometimes refuse to share their knowledge simply because firms were never interested in it in the first place. It is often the case that managers completely ignored the employees’ suggestions on how to improve the production, suggestions stemming from the superior knowledge of the employees. Finally, sometimes it is a matter of knowledge being inarticulable (tacit), and therefore unable to be shared, as one worker explained: ‘It’s knowledge you couldn’t really write down if you wanted to.’ The article goes on: ‘Mr. Bosco, Blackmer’s president, agrees that it’s almost impossible to document all of what he calls “native intelligence” of his workers. There are so many variables that only someone who does a job from day to day can fully comprehend what’s involved.’ But, contra Williamson, when given a chance to contribute to a solution to a particular problem, it appears that many, if not most, workers are happy to share whatever knowledge they can: As they searched for solutions, both the plant’s managers and workers agreed that the root of the problem was that the former management hadn’t given workers any input into the redesign. ‘They ignored our experience because they thought that any monkey could build pumps,’ says Mr. Bancroft, one of the plant’s most seasoned pump builders . . . Mr. Taylor says that he was pleased that the plant implemented the workers’ ideas. ‘The most frustrating part of sharing your ideas is when nobody listens,’ he says.

If the sharing of knowledge does not happen, it is not necessarily because workers are acting opportunistically: ‘Paul Adler, a professor of manage-

The hierarchical theory of the firm

39

ment and organization at the University of Southern California, believes workers have a predisposition to share knowledge but are often stopped by “fairly clear signals that management is going to use this knowledge to hurt them”’. The bottom line here appears to be that Williamson’s behavioral assumptions appear to be at the very least too strong, and that in fact we do not see opportunistic use of idiosyncratic knowledge nearly to the extent that we would expect going by Williamson’s work. In the following chapters I will argue that, though Williamson’s conception of the firm contributed some important insights, it is ultimately incomplete. A more complete theory of the firm has to present a more thorough account of hybrid organizational forms and better account for the dispersal of knowledge within firms. It is my claim that, in order to take advantage of the relevant, but dispersed, knowledge of their employees, firms have begun to experiment with different ways of ‘decontrolling’ the firm, that is, relaxing the hierarchical structure of the firm in order to decentralize the intra-firm decision-making, in the process creating an increasing variety of hybrid forms of governance. The raison d’être of some firms is to act as generators of entrepreneurial insight through better use of idiosyncratic knowledge of creative, imaginative, experienced employees. The owners/managers of such firms strive to shape the institutional environment of their firms in ways that bring about an alignment of incentives between the owners and the incumbent employees with idiosyncratic information. In other words, owners encourage creation and proliferation of idiosyncratic information. I will examine these issues in more detail next.

NOTES 1.

Williamson was influenced by F.A. Hayek’s 1945 article ‘The Use of Knowledge in a Society’, which he quotes in nearly every book and article. His 1975 book, where he first laid out his model, dedicates a separate section to Hayek, entitled ‘Hayek on Information’. Here Williamson enumerates Hayek’s insights that were important to his own work. Interestingly, there is another area where Williamson’s views are quite close to Hayek: comparative institutional analysis: ‘Transaction cost economics eschews hypothetical ideals and insists that the relevant comparisons are with feasible alternatives, all of which are flawed (Coase 1960). The relevant criterion is thus that of remediableness, according to which an outcome for which no superior alternatives can be described and implemented with net gains is presumed to be efficient. (Not surprisingly, public-policy intervention to correct market failures is much more circumspect when the remediableness test is applied)’ (1996, p. 7). Williamson also criticizes the Competence school (represented by Foss, Mahnke, and others) for talking about inefficiencies (such as path dependency) by ‘comparing an actual condition with a hypothetical ideal. That is in the zero transaction cost tradition of Pigou to which Coase (1960) took vigorous exception. A feasible criterion for judging dynamic efficiency is never proposed. Remediableness considerations are never reached’ (2000: 38). This is very much in the spirit of Hayek’s own critiques of neoclassical perfect competition model and welfare economics.

40 2. 3. 4.

5.

6. 7.

8. 9.

10. 11.

12.

13.

Employees and entrepreneurship For example, see 1985, p. 44; and 1996, p. 55. Williamson uses the same quote in Williamson 1991 and Williamson 2000. Williamson has remained consistent on this point even 20 years after initially discussing it: ‘Transaction cost economics describes the firm as a governance structure . . . firms and markets are alternative modes of governance (Coase 1937) and the allocation of activity between firms and markets is not taken as given but is something to be derived’ (1996, p. 7). See also the following quote: ‘Are human agents myopic, in the manner of the behavioral theory of the firm (Cyert & March 1963) or do they have the capacity for foresight, whereupon they look ahead and reposition? . . . Transaction cost economics ascribes foresight rather than myopia to human actors’ (Williamson 2000, p. 24). ‘The discriminating match between transactions and governance structures is the main source of refutable implications’ (1991, pp. 95–6). Williamson is increasingly coming under criticism by a collection of economists loosely operating under the ‘competence’ or ‘capabilities’ perspective of the firm for this very reason (see Foss and Mahnke 2000). The economists operating in this perspective have deliberately positioned themselves as an alternative to the governance perspective by downplaying the hazard-mitigation rationales of the firm, and emphasizing the technological rationales. However, these are not the same technological rationales relied on by the economists of the past. The new competence economists focus their attention on the role of firm routines as the embodiments of the institutional knowledge, and ways in which these routines can increase the value and decrease the cost of production by economizing on knowledge utilization. It should be noted that it is not quite right to speak of ‘the’ competence perspective at the moment, since it is very much in development and thus still quite heterogeneous. However, the economists working within this perspective can point to a body of literature which has influenced their studies: most importantly Nelson and Winter (1982), and previous to them Penrose (1959) and Richardson (1972). A thorough discussion of the competence perspective of the firm is beyond the scope of this book. Williamson acknowledges as much when he explains that the problem of the limits to vertical integration is ‘similar to [issues] which have been examined in the context of central planning versus market economies’ (1975, p. 117). Williamson bases this theory on the work of F.C. Bartlett (1932), who did experiments on oral transmission of stories, and found that there was radical alteration by the end of the chain of transmissions. This has been turned into a kids’ game called ‘broken telephone’. Williamson was very explicit about the comparative institutional nature of his analysis when examining markets versus hierarchies – see 1975, p. 130; 1996, p. 93. Williamson (1991): ‘Whereas contractual misalignments previously elicited self-interested bargaining, which delayed the adjustment and itself was costly, the operating divisions in the post-merger interval simply accept hierarchical realignment decisions as determinative’ (p. 100). Williamson is somewhat arbitrary in his introduction of ‘selective intervention’ into his work. In fact, he never really explores the full implications of selective intervention, which is in some ways quite similar to the concept of ‘decontrolling’. He mostly sticks to considering its implications to vertical integration, especially to limits to vertical integration. Although he occasionally refers to this system as another possible form of internal organization, he does not explore it in much depth. He also fails to examine the real-world examples of selective interventionism, the most famous being of course General Electric. Williamson in the 1996 book admits that, though he had spoken of hybrids in the past, he had failed to specify any of their characteristics: ‘[T]ransaction cost economics has been criticized because it deals with polar forms – markets and hierarchies – to the neglect of intermediate or hybrid forms . . . the abstract attributes that characterize alternative modes of governance have remained obscure. What are the key attributes and how do they vary among forms?’ (1996, pp. 93–4).

The hierarchical theory of the firm 14. 15.

16.

41

Though even in this book the only example of hybrids that he really explores in any detail is franchising. Williamson’s appreciation of idiosyncratic knowledge is even more obvious in this passage: ‘Interestingly, Marshall recognized that idiosyncratic human capital could sometimes accrue during the course of employment (1948, p. 626). Becker (1962), moreover, made express provision for human capital in his examination of labor market incentive schemes. Marshak expressly took exception with the readiness with which economists accept and employ assumptions of fungibility. As he put it, “There exist almost unique, irreplaceable research workers, teachers, administrations, just as there exist unique choice locations for plants and harbors. The problem of unique or imperfectly standardized goods . . . has indeed been neglected in the textbooks” (1968, p. 14). Polanyi’s (1962) remarkable discussion of “personal knowledge” further illustrates the importance of idiosyncratic knowledge and working relations’ (1996, p. 59). Williamson shows here that he, like most of the economics profession, misunderstood Hayek’s argument. Hayek rarely talked about equilibrium prices, which can be summed up in the phrase ‘sufficient statistics’; rather, he attempted to explain the important social role of disequilibrium prices through which economic actors can attempt to learn and discover opportunities to make profit by serving customers. Sautet quotes Kirzner on this point: ‘the nature of the coordinative property of the price system lies in “its ability to communicate information concerning its own faulty information-communication properties” (Kirzner 1985: 196)’ (2000, p. 37). Prices therefore act as guides to economic error avoidance and/or correction, by leading to discovery of heretofore unseen profit opportunities. In addition, Hayek never argued that prices convey all the information that is necessary for free markets to operate successfully. In fact, prices are just a guide, and economic actors rely on many different forms of subjective knowledge, such as the knowledge of the time and place of particular circumstances, as well as their own skills (which can be thought of as tacit or inarticulate knowledge), accumulated over time.

3. The knowledge problem in firms In a free market . . . any advantages that may be derived from ‘central planning’ . . . are purchased at the price of an enhanced knowledge problem. We may expect firms spontaneously to tend to expand to the point where additional advantages of ‘central’ planning are just offset by the incremental knowledge difficulties that stem from dispersed information. On a small scale the latter difficulties may be insignificant enough to be worth absorbing in order to take advantage of explicitly coordinated organization. Knowledge problem dispersed over a small geographical organizational area may mean a Hayekian knowledge problem that, unlike that relevant to large, complex entities, is solvable through deliberate search. Beyond some point, however, the knowledge difficulties will tend to reduce the profitability of firms that are too large. Competition between firms of different sizes and scope will tend, therefore, to reveal the optimal extent of such ‘central planning’. (Kirzner 1992, p. 162)

Coase and Williamson have had a great influence on modern economists, and it is mostly due to them that today we usually think of firms as designed, centrally planned institutions. But a different perspective on the theory of the firm has emerged over the last decade, one explicitly anchored on an Austrian, or more specifically Hayekian, epistemological basis rather than on transaction costs. The fundamental insight of this new Austrian theory of the firm is that much of the economically important knowledge is subjective (that is, knowledge of time and place) and often tacit (or inarticulable), and therefore necessarily dispersed among the multitudes of individuals within a firm just as much as within a society. The implications of dispersed knowledge for economic systems have been discussed at great length over the last 70 years (for a few examples, see Hayek 1944, 1948; Lavoie 1985a, 1985b; Boettke 1993), and it is not necessary to revisit this argument here. However, applying these epistemic conditions within a firm is a relatively new research agenda, and one of great significance to the theory of the firm. There have been several attempts (Cowen and Parker 1997; Gable and Ellig 1993) to import the dispersed knowledge analysis from the macro realm of economic systems to the micro phenomenon of the firm in a direct and rather simplistic way, and to force the policy conclusions relevant to economic systems to apply to the internal organization of the firm.1 This literature aimed to demonstrate that firms and markets are basically the same. Therefore, the authors concluded that firms should attempt to 42

The knowledge problem in firms

43

emulate markets and make themselves less firm-like; in other words, there should be no ‘islands of conscious power in this ocean of unconscious cooperation’.2 This was a rather forced enterprise, and one that ultimately did not yield great insight into internal structure and operations of firms. Though knowledge is certainly dispersed within a firm just as it is within an economy, it should be obvious that most successful firms engage in some straightforward form of command-and-control, hierarchical operations. It is highly unlikely that all these multitudes of entrepreneurs have been doing it wrong all along! Unlike markets, firms actually have one ultimate purpose – to earn profit – and require at least some conscious co-ordination. In fact, a major reason why free markets can be shown to be ‘spontaneously’ orderly is because owners and managers and promoter-entrepreneurs consciously run and command their firms. They use their localized, often tacit, specific knowledge of time and place within their firms, to satisfy their customers more effectively than their competition, and in such a way bring about the proper allocation of scarce resources. And it is exactly their ability to consciously direct production within firms that allows us to be confident of the market’s ability to co-ordinate disparate plans and bring about beneficial outcomes. And yet we are left with a very real problem: knowledge still is dispersed inside firms. Is this entirely irrelevant, as Coase’s and Williamson’s analyses would imply? Or have firms found ways to deal with internal dispersed knowledge and even use it to their advantage? Can this be done while retaining the hierarchical command structure of firms, a structure akin to economic central planning? Is it possible to reduce some hierarchical command layers and still have a co-ordinated firm? In other words, how important are ‘tall’ managerial hierarchies for a firm to be successful? In order to answer these questions, we must be willing to inquire within the ‘black box’ of the firm – to examine firms’ internal structures and managerial methods as they are, rather than as we would wish them to be, and construct a body of theory which can ultimately explain if and how firms attempt to solve the knowledge problem that confronts them. These are exactly the questions (among others) that a small group of Austrian and Austrian-sympathetic scholars has been attempting to answer over the last decade and a half. A growing list of economists, including Richard Langlois, Nicolai Foss, Frederic Sautet and Alanson Minkler, have been emphasizing the Hayekian knowledge problem within firms. In general, they do not reject Coase’s and Williamson’s contributions, but are instead attempting to expand upon them and extend them into new areas. For example, Langlois, one of the most prominent members of this group, has developed a concept of ‘dynamic transaction costs’, defined as ‘coordinating costs of informing, negotiating with, and persuading potential

44

Employees and entrepreneurship

contracting parties who may not share one’s faith in the proposed innovation or even, in a fundamental sense, one’s view of the world’ (1992, p. 177), and turned the concept into the center of his analysis of the firm. In this formulation the Hayekian knowledge problem shares center-stage with Coasian transaction costs.3 Foss in particular has done much to raise the awareness of the issues of dispersed knowledge in firms within the Coase/Williamson modern economics of organization, and has consciously started a combined research program.4 And Sautet has made an important contribution with his 2000 book in which the intra-firm knowledge problem is at the very center of the institutional analysis of the firm.5 I will examine the contributions by all four of these scholars in detail in this chapter. I will first discuss the nature of firms from an institutional perspective, relying on Carl Menger’s and Hayek’s taxonomies. Second, I will examine the significance of dispersed knowledge within firms. Third, I will explain decentralized decision-making within firms as a response to the existence of dispersed knowledge in firms. Finally I will present Foss’s recent critique of firm decentralization from the Austrian perspective, and comment on it. The main argument that I will develop in this chapter, based on the work of Langlois, Minkler, Foss, and Sautet, can be summarized in the following way. Firm managers are aware of the existence of dispersed knowledge in firms and they decide whether or not to take advantage of specific and unique knowledge of the circumstances held by employees by adjusting the organizational structure of the firm: loosening the managerial control and hierarchy – giving greater responsibility and decision-making power to the employees – when they want to access and use more employee-held knowledge, and tightening the managerial control when they need less employeeheld knowledge. The reason why they would sometimes choose to have less access to employee-specific knowledge is that decontrolling comes at a cost – loss of some conscious co-ordination of the firm’s productive efforts, leading to possible discoordination and therefore the possibility of wasted resources. Managers then must decide between the relevant tradeoffs: the value contributed by greater access to specific knowledge of their employees compared to the losses due to lack of perfect co-ordination of efforts and employee pursuit of some ultimately wasteful activities. Finding the proper tradeoff is a process of entrepreneurial judgment as well as trialand-error discovery, and thus cannot be known ahead of time. It should stand to reason then that decontrolling makes the most sense during periods of intense competition, when innovation becomes more important than simple cost-minimizing. Managers are thus confronted with a multitude of organizational possibilities, ranging from a perfectly centrally planned structure in the mold of

The knowledge problem in firms

45

the ‘Taylor firm’ from the early twentieth century, all the way to radically decentralized organizations such as W.L. Gore & Associates. We must therefore recognize the open-ended, non-deterministic nature of the internal structure of firms. Finding the structure that results in the best mix of features is a process of discovery unfolding through time and requiring entrepreneurial judgment on the part of owners/managers.

THE INSTITUTIONAL TAXONOMY A fundamental issue (one ultimately inseparable from the knowledge problem, as I will show below) distinguishing the new Austrian work on the firm from the Coase/Williamson paradigm is one of the institutional nature of the firm. The Coasian perspective is clearly that firms are ‘designed’ and ‘planned’ institutions, while Langlois, Foss, Minkler, and Sautet choose to loosen these characterizations to varying degrees. In order to better understand the Austrian contribution, we must first look to the work of Carl Menger and Friedrich Hayek. Menger introduced a distinction between pragmatic and organic institutions. Pragmatic institutions come about through a ‘common will directed toward establishing them’, while organic institutions are ‘the unintended result of innumerable efforts of economic subjects pursuing individual interests’ ([1883] [1985], p. 146, italics in the original). Thus, the difference between organic and pragmatic institutions is related to their origin: pragmatic institutions are the result of someone’s intent and can therefore be thought of as designed, while organic institutions came into existence without anyone’s intent and can therefore be thought of as undesigned. In addition to the differentiation between organic and pragmatic origins of institutions, we can classify institutions along another dimension. Hayek distinguished between organizations (taxis) and spontaneous orders (cosmos) or simply orders. Organizations are institutions that exhibit explicit, conscious planning to meet or accomplish someone’s (or a group’s) goals; orders are institutions that feature ‘abstract’ rules independent of anyone’s purpose apart from allowing the individuals to achieve their goals (Hayek 1973, p. 38). Thus, Hayek’s distinction between orders and organizations is not about the origins of institutions, but rather about their internal structure, or, more precisely, the nature of the systems of rules within them. The question that Hayek is concerned with is whether an institution exists to serve the specific interests of one person or a small group, or whether it exists to serve the interests of multitudes of people. Given these definitions it is clear that the Coasian view of firms is as ‘pragmatic organizations’: firms are consciously designed with concrete

46

Employees and entrepreneurship

rules intended to serve the specific interests of one person or a small group of people. They are both designed and planned. However, while many, if not most, firms may fit this description, it is by no means clear that all firms do. There is growing recognition that business firms are much more heterogeneous than previously thought. Pragmatic and Organic Institutions In an influential book, Nelson and Winter (1982) showed that firms undergo an evolutionary process which shapes their internal structure and their capabilities as much as, if not more than, the original, intentional designs of the founders. Langlois further developed their work by placing it within an explicitly Austrian (or, more precisely, Mengerian and Hayekian) framework. According to Langlois, Nelson and Winter’s de facto claim is that firms are often (though not always) organic organizations (1992, p. 169), because their rules of operations are not etched in stone but rather evolve through time. Moreover, the evolution of rules within organizations is often driven by ‘the same sorts of unintended consequences and unplanned outcomes . . . that one finds in spontaneous orders’ (ibid.). Langlois’s argument is that Nelson and Winter provided a challenge to the usual assumptions about firms by demonstrating that there exist many undesigned aspects to their structures. Langlois arrives at the contraCoasian conclusion ‘that the internal workings of the firm are far less in the nature of conscious planning than popular accounts would have it’ (1994b, p. 176). Langlois’s explanation is based on the differentiation between the original structure of a firm, which may in fact be consciously designed in detail, and its subsequent structures as the firm grows and changes. He interprets Menger’s definition of a pragmatic institution to apply to a firm when it is fully consciously designed not only initially but also subsequently – not just ex ante, but ex post. In other words, an institution is truly pragmatic only if all of its structural features at every point in time are the result of conscious, planned design. Langlois explains it this way: Any individual firm may be of pragmatic origin: it is consciously designed in the sense that someone consciously filed incorporation papers, acquired financing, hired secretaries, and did all the things one associates with creating a firm. At the same time, however, that firm does not spring full-blown from anyone’s head. Its ultimate size, its economic success, its particular characteristics become known only as the result of an economic process. Moreover, the general organizational concepts that one uses in designing a firm – such as the idea of the M-form studied by Chandler (1962) – certainly develop in an organic way from earlier forms. (1986a, p. 19)

The knowledge problem in firms

47

The fascinating implication of this explanation is that all firms must by necessity possess at least some organic elements. Thus, it is incorrect to refer to firms strictly as pragmatic institutions, as Coase and Williamson would. The evolutionary, unpredictable and non-determinate process of growth and change will lead to a firm structure which is at least to some degree a consequence of human action, though not of human design, that is, the existing structure was not intended by any one person. Rather, it is a result of ‘the diverse intentions of many individuals and groups interacting with one another and with external circumstances over time. For an organization to be genuinely of pragmatic origin, then, unintended consequences must not intrude to alter the intentions of the founders’ (Langlois 1992, p. 169). The ‘origin’ therefore does not refer to just a firm’s initial state, which is often severely disconnected from its present reality, but instead to the origin of the current structure of the firm. And it should be obvious that many firms feature structures that are at least partly influenced by non-designed, unintended factors, whether internal or external. The basic reason most firms are organic organizations is the simple one of survivability: Langlois emphasizes that for firms to survive they must exhibit what he calls ‘a reconfiguration of capabilities’ (1994a, p. 15), which is simply the ability to change the rules and routines of operation as the internal and external circumstances dictate. In other words, firms must be able to change and grow, which is more difficult for pragmatic organizations relative to organic organizations. Firms cannot remain competitive by simply preserving the same set of rules and capabilities they had at the outset. They must change by acquiring some new capabilities and shedding some old irrelevant ones, all of which often takes place without a founder’s or leader’s design, and sometimes explicitly in opposition to a grand plan. Langlois describes this process in the following way: In effect, the founding (or, in some cases, the ‘reengineering’) of the firm is a historical event that keys a path-dependent process of learning through the accretion and improvement of routines. Under some – but by no means all – circumstances, such a process may result in a large, complex organization. And, as was once the case for biological organisms, such coherent organizations may easily be misinterpreted as being the creatures of conscious design. The founding of Apple Computer in 1976 was surely an act of conscious design, even if founders Jobs and Wozniak had quite different visions of what the organization was and could become. But it would be nonsense to say that the corporation called Apple Computer that exists today is the result of conscious design. The corporation grew in what is in effect an organic fashion through the slow accretion of routines and capabilities. A firm thus begins [as a pragmatic organization]. But if it lives very long, it eventually [transforms into an organic organization] as it grows and learns organically. (1994a, p. 16, italics in the original)

48

Employees and entrepreneurship

The bottom line is that many if not most of today’s firms would be better described as organic rather than pragmatic institutions, contrary to Coase and Williamson. Orders vs. Organizations But we can go even further by reexamining the claim that firms must be organizations – that their rules must be ‘concrete’ and directed toward a unitary purpose. It must be conceded that on some level a firm can never be classified as an order by definition, since all for-profit firms do have a bottom line, an overall, unitary purpose to earn the greatest return on investments. But the issue is slightly more complex than that. Just because firms cannot be classified as orders, they should not by implication immediately become classified as organizations. We must consider the possibility that some firms could fall under some intermediate classification between organization and order. In other words, the relevant question is whether firms exhibit any elements of spontaneous order. Langlois again says yes: ‘Hayek’s theory of spontaneous order can in fact include the case of such apparently purposive and extra-market forms as the business firm’ (1994a, p. 2). To explain how, Langlois first summarizes Hayek’s definitions of organizations and orders: An organization is also a system of rules of conduct. In comparison with the rules of an order, the rules of an organization are concrete: rather than facilitating many different purposes, they are focused on achieving certain specific goals. Yet, the rules of an organization are similar to those of a more abstract institution in the sense that we can view them as evolving in much the same way and as having many of the same informational benefits. (1992, p. 174)

So the first point to understand is that internal structures of firms consist of systems of rules, whether explicit or tacit. Second, in firms these systems of rules are supposed to focus the efforts on achieving certain specific goals. But these rules, in addition to sometimes changing and evolving in response to unforeseen and undesigned influences (that is, organically), can also be shown to be abstract to a certain extent. Many firms do not have a goal more specific than the one mentioned above of earning the greatest returns on investments. Beyond that, internal structures of a firm may become quite abstract. The primary purpose of institutional rules within firms is often to give the individuals within the firm a structure through which their actions and knowledge can be co-ordinated. In at least some instances these rules will be of a rather abstract nature in the sense that they are intended to spur people to use their specific knowledge in unanticipated and novel ways rather than just following commands.

The knowledge problem in firms

49

In fact, the organic nature of the firm is to some extent interrelated to the firm exhibiting some elements of order. Harper (1996) describes it in the following way: As omniscient economists, however, we may observe ex post that the actual structure of the venture is not the sole result of conscious design on the part of the entrepreneur: although of pragmatic origin, much of the structure – such as connections of goodwill and other institutions of inter-firm cooperation – will evolve at least in part spontaneously over time. Because of structural uncertainty, the development of an innovating firm will depend upon many actual circumstances and economic changes which the entrepreneur could not have predicted. (p. 154)

Therefore, rules within at least some if not many firms are made to be sufficiently abstract to allow for some spontaneous (that is, not planned by the higher levels of the hierarchy) action and interaction, which leads to the organic evolution of firms. These rules must by definition then contain some elements of order (rather than being pure organizations). Langlois’s definition of spontaneous order supports this reading: ‘[S]pontaneous orders [are] abstract patterns of behavior whose structure emerges organically from the interplay of individual intention rather than from any grand scheme’ (1994a, p. 3). If the system of rules within a firm was fully concrete no such abstract patterns of behavior could ever emerge, and therefore organic change within a firm would be impossible. In fact, Langlois explains that firms that exhibit fully concrete rules would find themselves at great disadvantage during times of rapid change: As Hayek suggests, the more abstract the rules of a system, the better able that system is to coordinate a diversity of concrete purposes. Concrete rules are limited by the purposes they embody, and so are necessarily ill-adapted to circumstances that necessitate a change – especially an unpredictable change – of concrete goal. (1994a, pp. 20–21)

So, we can draw the conclusion that changing and surviving firms must in some cases be classified as organic orders. In order to better understand how the concepts of orders and organizations are applied to firms, we could think of firms as operating along a continuum between two extremes of order and organization. Harper explicitly refers to it in these terms at one point: For any particular economic activity, the entrepreneur’s choice of governance structure is not limited to polar cases of pure markets or hierarchical organization. There are complex hybrid modes of governance located between both ends of the continuum, and entrepreneurial activity in this intermediate range is widespread. (1996, p. 154, italics added)6

50

Employees and entrepreneurship

The continuum would imply the existence of some kind of a tradeoff in organizing and running a firm, which I will explore below. For the moment it is important to recognize that most real-world firms do not fit the pure definition of an organization but rather would seem to reside in this vast institutional space between pure organizations and pure orders, containing elements of both, but combining those elements in unique and often unexpected ways. This interpretation of the institutional structure of firms is obviously opposed to the orthodox view. Foss (1997) also briefly discusses this very issue of the institutional nature of firms. He also questions the usual clean distinction between planned and spontaneous orders (as he refers to organizations and orders, respectively), and argues that firms must allow a certain amount of flexibility in rules governing the behavior of employees in order to bring about localized discovery procedures (or intra-firm learning processes) (p. 192). One particular example of this is incomplete employment contracts: by not specifying exactly the activities of employees in some cases firms create a situation conducive to these localized discovery procedures. I believe this gets to the core of the issue in many ways, but Foss fails to carry this insight any further. However, his contribution is significant for the fact that in that article he provides a justification for the abstract nature of rules within firms on the basis of dispersed knowledge. He explains that firms are unlikely to fit the definition of a pure ‘made order’ or organization owing to the existence of the internal knowledge dispersal problem: Much of Austrian economics, and in fact also standard economics and organizational economics, has been working with an overly restrictive separation between spontaneous and made orders, or between ‘markets and hierarchies’. However, as I have indicated, the distinction is far from watertight. Firms, at least large firms, face a knowledge dispersal problem that may be on a par with the societal knowledge dispersal problem, and all sorts of emergent processes take place inside firms. (1997, p. 190)

This theme was also briefly explored by Langlois, as we can see in the following quote: In the Hayekian picture, firms and markets are both systems of rules of conduct. And both are systems for economizing on knowledge in the face of economic change, albeit quite different kinds of knowledge and change. Moreover, there is a sense in which the firm exists not in order to centralize control over knowledge but – like the market – precisely to decentralize the use of knowledge. (1994a, p. 3)

This is the point that I will explore in the next section. It is due to knowledge dispersal that firms are sometimes better thought of as organic orders

The knowledge problem in firms

51

rather than pragmatic organizations. Of course, there is a great amount of heterogeneity among firms and some will fit the pragmatic organization classification very well, while others will not. But until recently most analyses of firms have implicitly assumed that firms must be pragmatic organizations. But by assuming this we cannot explain firms that have decentralized the decision-making. Firms featuring decentralized decision-making are organic orders to an extent beyond that made absolutely necessary by the complexities of the real world. In other words, decentralized firms are explicitly and intentionally organic orders, usually to a much greater degree than ordinary firms, which themselves, as Langlois explained, inevitably must have some organic and order elements. In ordinary firms where decision-making is hierarchical, the spontaneous elements are usually unintentional and often unrecognized even by the firms’ owners or managers. But decentralized decision-making within firms exists because the owners and/or managers intentionally converted their firms into organic orders. In order to be able to explain why they would do so, we must better understand the issue of knowledge dispersal within firms. One final word: the organic order nature of decentralized firms does not imply that firms with decentralized decision-making must exhibit market features. Though both markets and decentralized-decision-making firms are examples of organic order, it does not logically follow that they must consist of the same kinds of rules and structures. In fact, to get back to the notion of the governance continuum, while markets are at the ‘order’ extreme of the continuum, decentralized-decision-making firms are somewhere along the intermediate points of the order–organization continuum. It stands to reason therefore that they will not be identical. This is worth emphasizing so as to better understand decentralized-decision-making firms. They are a distinct institutional form, neither pure organizations nor orders, and there is a lot more to them than simply incorporating elements of market structures.

THE KNOWLEDGE PROBLEM IN FIRMS It is only in the last ten years that economists have begun to consider the possibility that firm owners/managers face the Hayekian knowledge problem. In a way, that is not too surprising: the mainstream of the profession has in general given little attention to Hayek’s epistemological insights, and the few Austrian and Hayekian economists have mostly devoted their scarce time to topics other than the theory of the firm. Fortunately, the 1990s saw much greater interest and inquiry into the implications of the knowledge problem to the firm organization.

52

Employees and entrepreneurship

Hayek famously defined the knowledge problem as: the fact that the knowledge of the circumstances of which we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess. The economic problem of society is . . . [one] of the utilization of knowledge which is not given to anyone in its totality. (1945, p. 77–8)

In this quote it becomes obvious why the knowledge problem is also often referred to as the problem of dispersed knowledge: as Hayek explains, economically relevant knowledge is often dispersed among (or held by) many different individuals. Almost all productive activities require that this dispersed knowledge somehow be either collected or co-ordinated, frequently not easily done – thus the ‘knowledge problem’. The usual (implicit) assumption about firms has been that the entrepreneur-promoter-coordinator possesses superior knowledge relative to his employees, and therefore that the knowledge problem in the firm is irrelevant. But that assumption has been coming under increased scrutiny recently as more economists ask the fundamental question: does the condition of dispersed knowledge have any relevance to the firm? Dispersed Knowledge in Spontaneously Ordering Firms In the previous section I showed that firms may, in some cases, be classified as ‘organic orders’ because the rule systems within them result from a continuous and often unintended evolution as a consequence of unplanned employee actions and interactions. This fact can only be understood from the epistemological perspective: organic order firms evolve in unexpected and undesigned ways because of the dispersal of knowledge within them. In other words, change happens because employees are allowed to use their specific knowledge in unexpected ways (from the perspective of the owners/managers). A system of rules within a firm may be made very abstract for the express purpose of getting the employees to engage in individual decision-making based on personal knowledge. Sautet quotes a passage by Hayek to make this point: For this reason, the complex firm must ‘rely also on rules and not only on specific commands. The reason here is the same as that which makes it necessary for a spontaneous order to rely solely on rules: namely that by guiding the actions of individuals by rules rather than by specific commands it is possible to make use of knowledge which nobody possesses as a whole’. (Hayek 1973, pp. 48–9) (Sautet 2000, p. 99)

The knowledge problem in firms

53

As long as a firm is run by rules rather than specific commands, it can hope to use the dispersed, scattered knowledge without having to collect it and centralize it. Sautet here implicitly recognizes that a firm can also exhibit elements of a (spontaneous) order, as I argued above; moreover, his point is that the only way to make full use of the dispersed knowledge is by transforming a firm from a fully planned organization to something that looks like a spontaneous order by replacing commands with rules. Merely recognizing the possibility of the existence of the knowledge problem within firms does not go far enough, though. In order to explain how the knowledge problem in some cases plays an important role in firm structure, we must show that it can be significant enough for the firm to be somehow harmed, or at least become less competitive, if it is unable to handle it. Foss has identified two such cases. The first case concerns the issue of the firm size: [F]irms may confront knowledge problems of a magnitude comparable to those that confront a social planner in a socialist economy. Firms, such as AseaBrown-Boveri, with more than 200,000 employees certainly face a knowledge dispersal problem of almost the same caliber as the general societal knowledge dispersal problem. (Foss 1997, p. 186)

So, the larger the firm, the more serious the knowledge dispersal problem. This certainly has great significance for large firms. And if firms on average are becoming larger than in the past, the knowledge problem in firms would be increasingly more relevant. But it is not clear that this is the case. Also, we cannot dismiss the possibility that some managers may in fact possess greater knowledge than all their employees combined. This is where the second case becomes relevant. It is identified by Foss as ‘knowledge work’: [B]ecause of the increased need for firms to be source diverse, specialized knowledge in production . . . is becoming increasingly dispersed . . . [B]ecause of the increased importance of sourcing specialist knowledge, knowledge assets controlled by individual agents (‘knowledge workers’) are becoming increasingly important in production. (Foss 2001a, p. 8)

Given recent industrial developments, problems created by knowledge work would certainly seem to have growing significance. ‘Knowledge firms’ are becoming more numerous in the real world as work becomes more ‘knowledge-based’ and the economy moves further into the realm of the ‘knowledge economy’. In knowledge firms the managers often have less knowledge about particular activities than the employees (knowledge is dispersed), leading to the managers’ inability to fully direct and plan the

54

Employees and entrepreneurship

employees’ actions. The clear implication is that the knowledge problem in firms is becoming more relevant than in the past, and at least some (and probably an increasing number of) firms in the economy will seek creative, innovative ways to handle their internally dispersed knowledge. In other words, knowledge firms must successfully co-ordinate their internally dispersed knowledge in order to survive the rivalrous market process. Modern Economics of Organization and the Knowledge Problem The view of firms as co-ordinating mechanisms is certainly not a novel one. As seen in Chapter 2, Coase and Williamson developed just such a theory, though with two significant differences: in their frameworks, firms do not co-ordinate knowledge but rather incentives, and they co-ordinate them strictly through managerial commands. Foss points out that the Coasian perspective leaves out the possibility of any other sort of co-ordination being needed: One meaning [of co-ordination] that has become prevalent – in fact, almost completely dominant – in the [modern economics of organization] is that of mitigating the effects of incentive-conflicts. Thus, incentive-conflicts are ubiquitous and all-important; in contrast, coordination problems that do not turn on incentive conflicts are implicitly taken to be either empirically rare, theoretically trivial or both. (Foss 1999, p. 460)

Even the asymmetric information models that have become widespread in the last couple of decades do not account for the knowledge problem. Though these models ostensibly appear to be about the differences in knowledge held by the principal and agent, in fact they are just another way to examine the incentive conflicts between the employer and employee. Minkler effectively exposes the implicit rejection of the existence of dispersed knowledge in the asymmetric information models: These models are based upon the strong, and increasingly untenable, supposition that the principal could know what agents should do. These models preclude the possibility that agents could ‘know’ more than the principal. Instead, asymmetric information models presuppose that agents possess information which would be of immediate use to the principal if its transfer wasn’t prevented by costly observation (or some other transaction cost). In contrast, knowledge problems remain even if observation is costless. (1993b, p. 569)

Minkler is here noting a difference between ‘information’ and ‘knowledge’. Knowledge held by some individuals is often not transferable except at a prohibitive cost, whereas information is easily transferable. In addition, the relevant information held by the agent in the asymmetric information

The knowledge problem in firms

55

problem is not of a technical nature,7 but rather of a behavioral nature: the agent has more information than the principal on how the agent is likely to act in different circumstances. It thus becomes clear that the asymmetric information models do not deal with knowledge but rather incentive conflicts, and therefore they are irrelevant to the knowledge dispersal problem. It then would appear that the existence of dispersed knowledge within firms poses a critical problem for the Coase/Williamson theory of the firm. Though their model may be applicable in some, or even many, circumstances where firms fit the pragmatic organization mold, it does not apply to those firms that are better explained as organic orders. Those firms tend to exhibit a significant amount of knowledge dispersal, the best example being the so-called ‘knowledge firms’. Because Transaction Cost Economics is insufficiently general in its work on the firm I believe that we must go beyond their theories in order to explain the phenomenon of decentralization of decision-making. In this same spirit, Foss calls for a development of an alternative, more general, theory of firms that would explain firms as having: advantages relative to the market in the coordination of certain types of knowledge. In this context, real-world firms coordinate distributed knowledge by various means, such as command, management information systems, routines, corporate intrapreneurship, transfer prices, and shared mental constructs. (Foss 1997, p. 191)8

This is a view of the firm that goes far beyond the Coase/Williamson model, and one that I will further develop below. Foss gets the last word on the knowledge problem and the Coasian firm: [T]he [Modern Economics of Organization] may be criticized for assuming that virtually all that is worth discovering has already been discovered. There are a few spanners in the works – all of which are only related to misaligned incentives – but there is no genuine knowledge problem of the sort that Hayek talked about. The bottom line is that, in the MEO, knowledge is not truly dispersed and coordination problems are trivialized. One may press the claim that the MEO has not fully absorbed one of the main messages of the socialist calculation debate: that knowledge is dispersed, subjectively held and tacit, and that there is more to efficient resource allocation (in socialist economies or firms) than providing managers with the right incentives. There is also the problem of the coordination of knowledge, which is not necessarily an incentive issue. (Foss 1999, p. 464)

56

Employees and entrepreneurship

Meaning and Implications of Dispersed Knowledge in Firms Foss defines the ‘truly dispersed knowledge’ epistemic conditions as occurring when ‘agents hold different mental constructs for making sense out of reality’ (1999, p. 466). Under those circumstances, ‘they may be ignorant about what knowledge they are ignorant about (sheer ignorance), and they will, therefore, experience unforeseen contingencies’ (ibid.). So uncertainty is unavoidably bound up with the knowledge problem. In fact, an important aspect of true knowledge (rather than information) is that it is often not communicable, for the basic reason that even the holders of the knowledge may not be consciously aware that they are in possession of it. They are sometimes uncertain about the knowledge they themselves hold. They may be perfectly capable of using it in the right circumstances, but would not be able to articulate it. They may even be in a situation where they may not be able to consistently apply the appropriate knowledge, though they do actually possess it. Owing to these considerations, it is worth taking a closer look at the nature of knowledge and its implications within firms. True knowledge is deeply embedded in the individual, accumulated as a result of years of study and experience, and can only be obtained by others through a similar lengthy process of study and experience-acquisition. All individuals differ in their knowledge for the basic fact that all individuals differ in their studies and experiences. These are the basic facts behind the three ‘dynamic characteristics to knowledge’ from the perspective of the Austrian school, listed by Minkler (1993b). Knowledge is: 1. 2. 3.

empirical – it often consists of knowledge of particular circumstances of time and place, and is irrelevant outside of that time and place tacit – sometimes it cannot be wholly articulated, and therefore cannot be communicated or otherwise transferred to other individuals a source of surprises – present knowledge can lead to discovery of elements (ideas and insights) not anticipated by the searching agent: ‘One could not sit down and intentionally have a specific idea (i.e., one that could be specified initially). And certainly one could not direct another to have a specific idea. Ideas originate spontaneously, and their usefulness cannot be known or planned beforehand’ (1993b, pp. 570–71).

Given its characteristics, it becomes clear why knowledge is dispersed within firms just as it is in societies. The knowledge held by employees cannot be centralized, in the sense of being collected and used by a central manager, because it is empirical, tacit, and a source of surprises. Sautet adds that employees may hold relevant and important knowledge ‘about the internal allocation of resources or about a profit opportunity in the

The knowledge problem in firms

57

marketplace’ (2000, p. 99) – in either case it is knowledge that could potentially benefit the firm and that a central manager would find useful. But the knowledge problem in the firm will prevent it from being used. In fact, often the manager will be completely unaware of the fact that such knowledge could be held by the employees, or certainly which employees would possess exactly which knowledge. Thus, Sautet defines the Hayekian knowledge problem in the firm as a situation where ‘the entrepreneur-promoter can be ignorant of his/her ignorance with respect to the knowledge possessed by some of his/her employees (and this knowledge could be crucial to the firm)’ (2000, p. 9). This intra-firm Hayekian knowledge problem is also recognized by Minkler, who refers to it as the asymmetrical knowledge problem (in contrast to the asymmetrical information problem). Minkler summarizes the organizational significance of the knowledge problem thus: A superior can only direct a subordinate over the activities she knows about. If the subordinate is knowledgeable about activities that the superior is not, the superior could not direct the subordinate to engage in those activities. That is, Alchian and Demsetz and the literature they have spawned presuppose that the principal could direct the agent over all activities, even activities in which the agent but not principal knows about . . . If the agent is more knowledgeable than the principal, then the principal is structurally uncertain9 about the agent’s activities and therefore could not direct her over those activities. (1993b, p. 576)10

We can therefore easily imagine a situation where an agent in a contractual relationship will possess greater knowledge about the production process or the product than the principal. But such a situation cannot exist within the Coasian conception of the firm: managers by definition know at least as much as, and probably much more than, agents.11 This is what allows them to engage in the process of monitoring of the employees, which presumes the principals’ ability to understand what they are observing and reward or sanction the agents appropriately. Minkler emphasizes that ‘asymmetrical knowledge implies structural uncertainty which renders observation irrelevant. In these predicaments, monitoring . . . cannot assure performance even if it was costless’ (Minkler 1993a, p. 20). The foundation of the Coasian firm begins to erode in a situation where the principal has less knowledge than the agent. The traditional hierarchical firm can no longer rely on simple monitoring to solve the standard moral hazard problem.12 When confronted with the asymmetric knowledge problem the only way that a manager could engage in effective monitoring is if he ‘made an investment to learn the employee’s production possibilities’ (Minkler 1993a, p. 24), or, in other words, if he undertook the process of learning all the relevant knowledge held by the employee. Only then

58

Employees and entrepreneurship

would he be in a position to properly evaluate the worker’s claims about the problems and possibilities of the productive process carried out by the employee.13 In situations of knowledge work this would obviously be impossibly costly and inefficient: workers are hired exactly because of their particular unique and costly-to-acquire knowledge. Sautet adds an important insight to this discussion: the knowledge problem in firms exists in addition to the market knowledge problem, where the relevant knowledge of the best uses of scarce resources for the purpose of satisfying consumer wants and earning a profit is dispersed among many individuals. Sautet explains it this way: Therefore, the entrepreneur-promoter may not only be ignorant of his/her ignorance with respect to profit opportunities in the market, but also with respect to what his/her employees know. [Thus, there exists a] double Hayekian Knowledge Problem. (Sautet 2000, p. 99, italics in the original)

The significance of the double Hayekian knowledge problem would appear to depend on the level of hierarchical separation between employees and management and it is possible that in some instances it could lead to serious problems. Employees may be ordered to adhere to a plan and perform tasks that they know are wasteful. Perhaps they may know a better way of doing something but they do not have the ability to use their knowledge. In either case, they may not be able to fully communicate their knowledge to the managers. Situations like these lead to demoralization of the workforce. Firms confronting a significant double knowledge problem that they are unable to overcome may find themselves at a competitive disadvantage and may not be able to survive. The double knowledge problem is of special significance ‘when a firm grows – that is to say, when it tries to exploit new profit opportunities by diversifying into new products and by expanding in new geographical areas’ (Sautet 2000, p. 116). It is during those times that a firm is faced with the greatest amount of ignorance and uncertainty, unable to rely on time-tested routines, and it is during those times that the economic knowledge held by the employees is of the greatest importance. So the fundamental question with regard to the knowledge problem in the firm is: ‘[H]ow can central managers mobilize and utilize the knowledge necessary for the development of the firm when they do not possess it?’ (Sautet 2000, p. 116).14 Or, more precisely, ‘how can the principal capture the benefits of agent’s knowledge?’ (Minkler 1993a, p. 18). This is the question which not only has not been asked by the Coasian theory of the firm, but which ultimately cannot be answered without going beyond the accepted transaction cost paradigm.

The knowledge problem in firms

59

DECENTRALIZED DECISION-MAKING THROUGH HIERARCHICAL FLATTENING An important way in which firms have been handling the problem of internally dispersed knowledge, or the double Hayekian knowledge problem, is by decentralizing decision-making through flattening of organizational hierarchies. This process has been given many different names: decontrolling, demanaging, empowering. They all refer to the same process: removal of some, or all, layers of middle management and giving employees the power to make some, and in some cases many, decisions. Managers and/or owners voluntarily transfer some power to ‘call the shots’ to the employees. In other words, the owner ‘chooses who gets to be the decision-maker in production, that is, whether the organizational structure will be participatory or non-participatory’ (Minkler 1993a, p. 18). Decentralized firms are participatory in the sense that non-top-management workers participate in the decision-making in the firm. In some cases the decision-making is limited to an employee’s own narrow sphere of operations, but in other cases the decision-making granted to employees encompasses even longterm strategy. In addition, the proper amount of decision-making decentralization fitting the particular circumstances often cannot be known in advance, but rather must be discovered by the owners/managers through a process of dynamic interaction between the firm and the markets in which it operates. An interesting example of just such a case can be seen in a 1989 Harvard Business Review article, ‘Managing without Managers’ by Ricardo Semler, president of Semco S/A, Brazil’s largest marine and foodprocessing machinery manufacturer (at least at the time). The article attempted to explain the radically decentralized managerial structure in his company. It was not a very large company (only 800 employees), and it did not deal in work that most people would think of as knowledge work – it was basically a manufacturing company. Semco reduced ‘management levels to three – one corporate level and two operating levels at the manufacturing units’ (p. 78): the top level consisted of five ‘counselors’ including Semler himself; the intermediate level was formed of heads of eight divisions referred to as ‘partners’; the bottom layer consisted of all other employees, ‘associates’ and ‘coordinators’. As Semler put it, ‘Counselors, partners, coordinators and associates. Four titles. Three management layers’ (ibid.). The company combined the flat managerial structure with a system of profit sharing. The reason that the employees were allowed to claim a share of the profits was that, though Semler was the president of the company, he was not the key decisionmaker in his own company:

60

Employees and entrepreneurship We insist on making important decisions collegially, and certain decisions are made by a company-wide vote. Several years ago, for example, we needed a bigger plant for our marine division, which makes pumps, compressors, and ship propellers . . . [T]he workers voted [on which of the three available buildings the company would acquire] – and they chose a plant the counselors didn’t really want . . . But we accepted the employees’ decision, because we believe that in the long run, letting people participate in the decisions that affect their lives will have a positive effect on employee motivation and morale. We bought the building and moved in. The workers designed the layout for a flexible manufacturing system, and they hired one of Brazil’s foremost artists to paint the whole thing, inside and out, including the machinery. That plant really belongs to its employees. I feel like a guest every time I walk in. I don’t mind. The division’s productivity, in dollars per year per employee, has jumped from $14,200 in 1984 – the year we moved – to $37,500, and for 1989 the goal is $50,000. Over the same period, market share went from 54% to 62%. Employees also outvoted me on the acquisition of a company that I’m still sure we should have bought. But they felt we weren’t ready to digest it, and I lost the vote. In a case like that, the credibility of our management system is at stake. (pp. 78–9)

In addition, Semco employees were expected to read and understand the company balance sheets (Semco provided classes to teach them to do so) and to ‘question the management decisions that affect future profits’ (p. 82). It is because of this that Semler can say that: Marketing is everybody’s problem. Everybody knows the price of the product. Everybody knows the cost. Everybody has the monthly balance sheet that says exactly what each of them makes, how much bronze is costing us, how much overtime we paid, all of it. And the employees know that 23% of the after-tax profit is theirs. (p. 84)

Semco is, to be sure, an extreme case, but it successfully illustrates the measures that some companies are – successfully – taking to become more competitive. If Semco did not actually exist, I am certain that most economists would confidently claim that such an intra-firm structure is the stuff of fairy-tales; and yet Semco showed dramatic productivity gains and became the market leader through these practices. The scant existing economic literature on hierarchical decentralization has explained the success of such firms strictly in epistemological terms: ‘[T]he primary advantage of participatory organizations may lie in their utilization of dispersed knowledge’ (Minkler 1993a, p. 18) and ‘The decentralized structure produces knowledge that would be impossible to produce in a different organizational structure’ (Sautet 2000, p. 127). In decentralized structures a greater amount of knowledge is utilized simply by giving the decision-making power to those workers who are in the best position to make the decisions. Oticon is another effective example. In

The knowledge problem in firms

61

Chapter 1, I described Oticon’s experimentation with radical decentralization of decision-making as a response to loss of market share and the threat of bankruptcy. Oticon was in trouble because it did not keep up with competitors’ innovations, which mostly centered around the development of an in-the-ear hearing aid. Once the decentralization was undertaken, it was soon discovered that Oticon employees, working independently of the company commands and unbeknownst to their managers, actually did develop the in-the-ear hearing aid technology, but owing to the inflexibility of the hierarchical organizational structure existing at the time they were unable to bring the product to the manufacturing stage. The management did not think to tap the knowledge of their employees until Kolind restructured the company and made it radically decentralized (he referred to it as the ‘spaghetti organization’), as seen in the following quote: One of the things soon to be realized when the spaghetti organization became a reality was that Oticon actually already had almost fully developed an in-the-ear hearing aid back in 1979 . . . A result of the spaghetti organization was that work on the old in-the-ear hearing aid could be resumed. (Foss 2001c, p. 12)

And: [A] basic problem in the old organization had been that commercially important knowledge simply didn’t reach the relevant decision-makers. A reflection of this is the example . . . of Oticon employees already having invented the in-the-ear hearing aid, that invention basically being shelved and forgotten until the spaghetti organization recovered it. (Ibid., p. 14)

The decentralized organization form is often the result of the managers recognizing that they themselves cannot make a decision that is as good as the decision that the employee will make, because the managers do not possess the proper knowledge to do so. Oticon was decentralized based on this assumption, and CEO Kolind was proven absolutely correct, as the above anecdote shows. This knowledge asymmetry is obviously very relevant in the areas of knowledge work, as discussed above, and, to the extent that more work is becoming knowledge work and the workers are becoming more knowledgeable, decentralized organizational forms are likely to become more common.15 Though decentralization of decision-making has innate logical appeal in the cases of knowledge work, we find it being tried even in manufacturing. The accepted wisdom in manufacturing firms has been that the workers are there simply to carry out the work which has been precisely set out for them through specified orders handed down from above. There was not much consideration given to the possibility of dispersed knowledge in

62

Employees and entrepreneurship

manufacturing settings. However, this status quo has been increasingly challenged. Minkler presented a dramatic example of decentralization in manufacturing: Saturn automobile company, a semi-autonomous division of General Motors, is another example of an American firm that recognizes the importance of the dispersed knowledge. The company has a flat hierarchy and employs teamwork and worker participation. The workers belong to the United Automobile Workers Union and were chosen from GM’s other plants based upon their motivation and desire to work in a cooperative environment. From the very conception of the plant’s design workers were involved in all stages of decision-making. They have continued to exert significant decision-making authority and responsibility. At Saturn any production worker (called member) can stop the production line – but if she does she has the responsibility to correct the problem. The intent is to develop an organization that fosters commitment and that promotes ideas which can be used not only at Saturn, but at GM’s other facilities as well. (1993b, pp. 573–4)

Saturn’s organizational structure would have been unthinkable in the US even 30 years ago. But more recently there have been many examples of manufacturing firms with decentralized decision-making not only surviving but actually thriving, as Minkler explains: Evidence suggests that manufacturing workers possess and can develop significant knowledge and are the source of important ideas and innovations, especially process innovations and especially in conducive environments. Many, if not most, Japanese industrial firms have recognized this and are using organizational innovations like employee participation, team production (cooperative rather than hierarchical or sequential decision-making), just-in-time inventory systems, and employment guarantees. (1993b, p. 572)

As Minkler points out, Japanese firms were the pioneers in the area of decentralization of decision-making, many of them eagerly adopting the managerial techniques introduced by William Deming. But with a few exceptions,16 economists have largely ignored the unique managerial and organizational advances made by Japanese firms. In fact, as I asserted above, the economic profession in general has mostly turned a blind eye to the decentralization developments. Foss also draws attention to this fact: From an Austrian point of view, it is striking that the [modern economics of organization] has focused almost exclusively on the negative aspects of decentralization, that is, to the extent that it has treated decentralization in organizations at all. Large firms may confront knowledge dispersal problems of magnitude comparable to those that confront the social planner in a socialist economy. . . .the Hayekian point that decentralization is an effective response to

The knowledge problem in firms

63

the local emergence of unforeseen contingencies is either neglected or given a static mechanism design interpretation. (Foss 1999, p. 471)

Foss emphasizes the superior ‘response to unforeseen contingencies’ as the main benefit of decentralized decision-making, an issue not studied much in modern economics, mostly owing to the great difficulty of mathematically modeling dynamic factors in the face of uncertainty. In light of this, it shouldn’t be surprising that we find the most fruitful discussion of decentralization of decision-making in management literature, as Foss also recognizes: There is, in fact, an enormous management literature that explicitly addresses how to handle the knowledge dispersal problems (not just the incentive problems) that exist in, typically, multinational firms. The focus is normally on choosing the right degree of decentralization . . . The literature explicitly begins by rejecting the idea that top management in large firms, such as [Asea-BrownBoveri], can simply centralize all the relevant knowledge and issue in a top-down fashion the relevant commands to different business units. This is seen as plainly absurd; and the arguments advanced in favor of this judgment are closely related to Austrian arguments against central planning: the size, complexity, and partially tacit character of the relevant knowledge, in addition to the need for flexibility and local adaptation, makes centralization not only inefficient, but truly impossible. Firms must resort to other means to handle dispersed knowledge. (Foss 1997, p. 187)

Management is a practical discipline, so maybe it is to be expected that on this issue it was ahead of economics:17 management theory follows managerial practices, and the practitioners do not have the luxury of assuming away the very real problem of internally dispersed knowledge. In attempting to handle the knowledge problem, managers discovered this tool of decentralization of decision-making. As Mises explained in Bureaucracy ([1944] 1983), this is the result of even division managers being ultimately guided by the bottom line of profit: managers must in the end contribute to a larger firm profit to justify the operation of their divisions, and in some cases the divisions run by the managers who gave greater decision-making power to their employees showed a consistently better performance, maybe owing to the enhanced response to the unforeseen contingencies mentioned above. Allowing the employees with the greatest knowledge of time and place to make the appropriate decisions will give firms the greatest ‘capability to react quickly to change’ and ‘promote innovation’ (Minkler 1993b, p. 569). This response to unforeseen contingencies, noted by Foss as the central benefit of decentralization, becomes more important as a market becomes more rivalrous, and survival no longer depends simply on cost

64

Employees and entrepreneurship

minimization but rather rapid adjustment to unexpected actions by competitors. Minkler explains the benefits of decentralization in the following way: ‘[A]s global markets demand greater flexibility, on-the-spot decisions become crucial and workers become “knowledge” workers – they come to have decision-making advantages over superiors higher up the hierarchy’ (Minkler 1993b, p. 569). We can see that desire for flexibility in the following example by Langlois: [IBM]’s hierarchical structure, internal sourcing procedures and elaborate system of controls made it too inflexible to respond well to a rapidly changing market. As a result, IBM chose in effect to disintegrate vertically into the production of PCs. The company ‘spun off’ a small group of executives and engineers, exempted them from IBM internal sourcing and other procedures, and treated them as, in effect, a venture-capital investment . . . IBM’s motives for disintegration were . . . the need to gain rapid access to capabilities that would not otherwise have been available in time. (1994b, pp. 177–8, italics in the original)

To the extent that the spin-off was still a part of IBM, the process described by Langlois is better understood as a case of limited decontrolling or decentralizing rather than vertical disintegration, which would require sale of the division. IBM officials understood that there was much locally held knowledge which could not be utilized under their standard hierarchical structure. In other words, the internal structure, while successful in co-ordinating the activities of the giant company, also stood in the way of speedy development of a product necessary to remain competitive in a rapidly changing market. IBM managers removed the constraints of a hierarchical decision-making in order to create the flexibility to respond to a market challenge. Langlois draws a general conclusion that is closely applicable to the above examples: The hypothesis I propose generalizes this idea: the more radical the change – the more radical the deviation from the customary path – the more abstract will be the institutions necessary to change, create, or otherwise redirect concrete capabilities in an effective direction. If this hypothesis is right, then, the best way for an organization to plan for the future, especially an unpredictable future, is to emulate in some degree a spontaneous order. (1994a, p. 21, italics in the original)

In light of Langlois’s hypothesis, I claim that firms with decentralized decision-making are consciously and intentionally endeavoring to become more like organic orders (which feature spontaneously evolving abstract rules). They do so in order to spur greater use of the knowledge dispersed among a firm’s employees and thus to better cope with unexpected changes created through rivalrous market competition.

The knowledge problem in firms

65

Even Alfred Chandler, often ascribing almost superhuman powers to managers, recognized the benefits of decentralization and demanaging, as can be seen in the following passage: Yet the dominant centralized structure had one basic weakness. A very few men were still entrusted with a great number of complex decisions . . . As long as the enterprise belonged in an industry whose markets, sources of raw materials, and production processes remained relatively unchanged, few entrepreneurial decisions had to be reached. In that situation such a weakness was not critical, but where technology, markets, and sources of supply were changing rapidly, the defects of such a structure became more obvious. (Chandler [1962] 1990, p. 41; in Sautet 2000, p. 114)

These comments imply that a tradeoff of some sort is present as firms change their internal structures.18 Decentralization appears to be most appropriate during times of rapid change. It is then that firms must have the greatest flexibility, greatest use of dispersed subjective knowledge held by their employees, and greatest ability to create new knowledge, embodied in new products and processes. We see this explained by Bill Gore, founder of the radically decentralized W.L. Gore & Associates: I’m told from time to time that a lattice [decentralized] organization can’t meet a crisis well because it takes too long to reach a consensus when there are no bosses. But this isn’t true. Actually, a lattice by its very nature works particularly well in a crisis. A lot of useless effort is avoided because there is no rigid management hierarchy to conquer before you can attack a problem. (Shipper and Manz 1998, p. C-499)

The obvious question in light of this is, given that flexibility and the ability to create and use new knowledge are generally desired by all firms at all times, why is decentralization not always appropriate? I will answer this question in ‘The organizational tradeoff between co-ordination and innovation’ (page 68). Delegation and Decentralization in Recent Economics Literature There has been increasing interest in the issues of decentralization of decision-making in mainstream economics (Jensen and Meckling 1992; Aghion and Tirole 1997; Holmstrom and Roberts 1998; Hart and Moore 1999; Stein 2002; among many others). Most of this work has been focused on incentive and opportunism problems that are created by a decentralized decision-making structure (usually analyzed under the umbrella of agency theory and asymmetric information problems), and thus falls short of thoroughly dealing with the true intra-firm or double knowledge dispersion

66

Employees and entrepreneurship

problem as I explained above. In addition, these articles rely on mathematical models of decentralization of decision-making for analysis and are thus fundamentally constrained by the simplifications necessary to make the models tractable. Nevertheless, it is informative to take a brief look at this literature. The article that began this line of inquiry and set out the fundamental approach is Jensen and Meckling (1992). In this piece the authors consider two different types of knowledge that can exist within a firm: general and specific. Following Hayek (1945), they explain how the ‘market automatically moves decision rights to the agents with the relevant knowledge’ (Jensen and Meckling 1992, p. 252) when specific knowledge is difficult to transfer: Because it is costly to transfer, getting specific knowledge used in decisionmaking requires decentralizing many decision rights in both the economy and the firm. Such a delegation in turn creates two problems: the rights assignment problem (determining who should exercise a decision right), and the control or agency problem (how to ensure that self-interested decision agents exercise their rights in a way that contributes to the organizational objective). (Jensen and Meckling 1992, p. 251)19

This article was among the first anywhere to explicitly set out the tradeoffs of decentralization: on the one hand, costs due to the lack of use of existing relevant information in the case of too little decentralization (recognizing the existence of dispersal of knowledge within firms) and, on the other hand, ‘agency costs’ (principals’ loss of control over agents) in the case of too much decentralization. Jensen and Meckling come to the conclusion that managers must weigh these two carefully in deciding the optimal allocation of decision rights. Another influential article in this emerging field is Aghion and Tirole (1997). This article showed that the primary benefits of delegation consist of greater ‘initiative’ to acquire relevant information as well as more active ‘participation’ in the organization, since employees derive greater utility from work when they are able to make decisions rather than being commanded. Aghion and Tirole think of this primarily in terms of the asymmetric information problem: a principal can overrule the agent, but wouldn’t want to if the agent is better informed. Another benefit of decentralization is the faster response in situations of urgency of decisionmaking: ‘It is sometimes observed that the need to adapt quickly to customer requirements has forced firms to decentralized decision-making’ (p. 24). They do not explain exactly why decentralized decision-making could be expected to bring about a quicker adaptation. Following the precedent of Jensen and Meckling, Aghion and Tirole accept that the downside of decentralization is the standard one of loss of

The knowledge problem in firms

67

control by principals. They hypothesize about one other interesting downside of centralization – worsened communication due to the threat of selective interventions by principals. If principals are likely to overrule the decisions made by the employees, though the decision-making was delegated to them, employees might respond by hiding relevant information in order to prevent selective interventionism. Aghion and Tirole discuss at some length different ways in which the principal can create the trust and/or credible commitment necessary to reduce the potential for selective interventionism, and thus encourage proper communication as well as provide better incentives for the agents to become better informed. One important insight is that principals may intentionally remain somewhat uninformed in order to increase their credible commitment to abstain from selective interventions. Ultimately, Aghion and Tirole conclude that centralization prevails if the principal has the superior information, while decentralization prevails when the principal is not as well informed as the subordinate and thus fears the possibility that forcing her decision on the agent will lead to a worse situation. A more recent article to consider decentralization within firms is Stein (2002), which asks the question ‘what organizational form – decentralization or hierarchy – does the best job of allocating capital to competing investment projects?’ (p. 1916). The main explanatory factor in this article is the existence of ‘soft’ information ‘that cannot be directly verified by anyone other than the agent (the “line manager”) that produces it’, in contrast to ‘hard’ information that is easily verifiable and communicable. Soft information cannot be verified by upper management or the CEO. It is somehow unique to the person who has it, even if that person is located at the lower levels of the hierarchy. (Notice the similarity to Jensen and Meckling’s use of ‘general’ and ‘specific’ information, closely corresponding to ‘hard’ and ‘soft’ information here.) The CEO will want to ensure that this soft information is used for the benefit of the firm (and that the agent engages in further ‘research’ – that he continue to accumulate more information and generate more ideas) so he will allocate sufficient capital to the decision-maker at a lower level in a hierarchy, such as a production line manager, in order for him to ‘lever’ his expertise. The implication is that in the presence of soft information firms should attempt to align authority over capital with expertise – in other words, decentralize the decisionmaking. However, when the information produced by the line managers is ‘hard’, Stein comes to exactly the opposite conclusion: ‘separating authority from expertise actually improves research incentives, as line managers struggle to produce enough information to convince their bosses that they should get more of the firm’s resources’ (p. 1893). In many ways, we can

68

Employees and entrepreneurship

think of Stein (2002) as an extension of Aghion and Tirole (1997), with one key addition: a strong hierarchical structure will weaken incentives for the better-informed agents to properly utilize their information only when the information is ‘soft’. In the presence of soft information, ‘line managers are discouraged when they do not have full authority’ (p. 1894) and will therefore engage in less ‘research’ and information generation. Finally, if the information can be ‘hardened’, the relative merits of a hierarchy will increase. So, to sum up, in this emerging literature we see greater appreciation for the knowledge problem. The differentiation seen in Stein (2002) between ‘hard’ and ‘soft’ information and in Jensen and Meckling (1992) between ‘general’ and ‘specific’ information would seem to adhere at least somewhat to Hayek’s differentiation between tacit and subjective knowledge of time and place on the one hand and more straightforward data-like, quantifiable knowledge on the other. Rajan and Wulf (2006), discussed at the beginning of Chapter 1, also introduce the idea of heterogeneity of information when speculating about the possible reasons why hierarchies are flattening: [G]reater competition may increase the complexity of the decisions that have to be made as well as the variety of data that impinge on the decision . . . Also, information may be hard to convey up a hierarchy with the necessary detail and color, thus reducing managers’ incentive to collect it. (p. 23)

The insights generated by these articles are valuable, though ultimately incomplete, as I will show below. Nevertheless, it is good to see that economists in the mainstream of the profession are beginning to take the issue of intra-firm knowledge heterogeneity seriously, and I certainly hope that this sort of work on delegation and decentralization continues to be further explored.20

THE ORGANIZATIONAL TRADEOFF BETWEEN CO-ORDINATION AND INNOVATION I have shown above that there is an emerging consensus that flexibility, and thus a greater ability to cope with change, is the main benefit of decentralization. This benefit is largely due to the fact that in decentralizeddecision-making firms knowledge and decision-making are aligned. In addition, such firms are better at spurring innovative activities by their employees, which is conceptually distinct from (though not entirely unrelated to) making the best use of individually held knowledge. These benefits have become widely recognized. But the proponents of decentralization

The knowledge problem in firms

69

(especially Cowen and Parker 1997) tend to emphasize these upsides of decentralization without acknowledging any possible downsides. Their logic explicitly follows the following pattern: the greatest amount of decentralization in markets is desirable; decentralized firms emulate markets; ergo, the greatest amount of decentralization in firms is desirable. However, this is simply not so. The first reason is theoretical: decentralized firms do not emulate markets – they are distinct institutional forms that, though they may appear so, are not like markets. It should be clear that firms are fundamentally different from markets owing to the fact that they can never fully be an ‘order’, in the Hayekian sense of an institution featuring a system of rules that do not aim at any one purpose apart from allowing the individuals within the system to achieve their goals. Ultimately all firms have one over-arching final goal – to make profit. In order to achieve that goal they rely on conscious co-ordination to at least some extent. It is improper thus to attempt to draw analogies between markets and decentralized firms. The second reason why Cowen and Parker’s logic is faulty is an empirical one: if there were only benefits to decentralization, we would observe many more real-world firms featuring radically decentralized decision-making than we actually do. Furthermore, we have seen some radically decentralized firms which diminished the extent of their decentralization and did very well (Foss 2001c). If decentralization were always superior to a hierarchy, it would be difficult if not impossible to explain such cases. It should be clear that decentralization entails some costs and that there most definitely are benefits to the hierarchical institutional structure. Those benefits follow the traditional Coase/Williamson/New-Institutionalist theory: hierarchies can reduce transaction costs due to opportunism and bounded rationality. In broader terms, we can say that hierarchies can effectively create institutional stability and incentive co-ordination. Langlois explains it in the following way: A social institution, then, is a mechanism to reduce the entropy of the environment. The presence of such a mechanism means coordination, high payoffs, and, in this context at least, a rigid and predictable pattern of behavior by both agents. (1986b, p. 175)

The nature of the tradeoffs facing firms with decentralized decisionmaking should be clear: they must balance the costs of lesser stability and poorer intra-firm incentive co-ordination with the benefits of greater flexibility, better use of dispersed knowledge, and superior innovative capabilities. Finding the right balance appears to be a difficult task. Firms with successful decentralized decision-making replace explicit control systems with systems of abstract rules that are able to achieve some minimum

70

Employees and entrepreneurship

required amount of incentive co-ordination, but these systems of rules are very difficult to develop. As I will discuss in the next chapter, they must consist of both an organizational culture providing a thoroughly pervasive structure of intersubjective meanings and an effective system of incentives (team-production, profit sharing, and so on). The above tradeoff, ever present in decentralized institutions, was first recognized by Ludwig Lachmann: In our view the central problem of the institutional order hinges on the contrast between coherence and flexibility, between the necessarily durable nature of the institutional order as a whole and the requisite flexibility of the individual institution. (1971, p. 13)

Lachmann was not talking about decentralization within a firm but rather within an extended market order. But this quotation gets right to the heart of the matter as far as firms with decentralized decision-making are concerned, as well. The tradeoff is between the ‘coherence’ of a firm on the one hand and its ‘flexibility’ on the other. As a firm becomes more participatory, a greater number of individuals will be acting according to their own knowledge, interpretations, and desires, leading to an inevitable splintering of a firm’s purpose and capabilities. In other words, the firm can and probably will lose some ‘coherence’: there will inevitably be less co-ordination of the actions of the employees, and therefore firms will be less stable and maybe even less enduring and robust. But a purely hierarchical firm will lack flexibility. An analogy that Hayek drew is of an army where each soldier’s actions must be dictated by a superior. Such an army would be impossibly handicapped. These are the relevant tradeoffs faced by firm managers, and, unfortunately, they are not easily quantifiable, or often even explicitly understood ex ante. What is not well recognized is that the internal structures of firms rarely stay fixed for a long period of time. Managers/owners can and do actively search for the proper level of institutional decentralization of decisionmaking through a process which combines trial and error, experience, and learning, all based on the judgments formed by their own subjective, tacit, empirical knowledge of the institutional characteristics of the firm. The internal structures of firms evolve through time: firms change, sometimes by design, other times spontaneously. They can acquire characteristics of an organic order for some period of time, which can be shed later as firms attempt to change into organic organizations, or even pragmatic organizations. They search for the best organizational structure for that moment in time – or, more precisely, the organizational structure most appropriate to the particular market circumstances that prevail at that moment in time.21

The knowledge problem in firms

71

These tradeoff factors have rarely been accounted for in the normative study of institutional structures, that is, a study of optimal institutional structures, which led Langlois to make a case for a dynamic understanding of institutional efficiency: One implication of [a flexibility–efficiency tradeoff] is, in effect, that efficiency is not an absolute concept: it can’t be defined independently of the organization’s environment. A firm in a very rapidly changing environment may have very bad transaction-cost properties but be far more efficient – far better able to survive – than a relatively less flexible organizational structure with good transaction-cost properties in equilibrium. It’s not clear how important this problem is in practice, although I conjecture that it may be quite significant in situations of rapid technical change. In any event, it’s far from clear that one can’t do comparativeinstitutional analysis in a way that accounts for these dynamic considerations. Most current analyses do seem to assume that the criterion for the organization’s survival is efficiency in the allocation of resources rather than flexibility or something like it. (1986a, pp. 20–21)

In other words, New Institutionalist economists have been examining only the ability of firms to reduce the transaction costs of their operations but neglecting to examine their ability to be flexible in the face of market instability and rapid change. It is clear that firms with decentralized decisionmaking are unlikely to have very good transaction-cost-reducing properties, which explains why economists have been so suspicious of – or even hostile to – decentralized forms of institutional structures of firms. But firms with decentralized decision-making exist because their managers are willing to forgo some of that coherence if it will gain the firm better use of dispersed knowledge and greater innovative capabilities. We can go a bit further and also explain the limits of firm operations from the perspective of these tradeoffs. This is nothing new: management literature has taken the view for decades that firms are constrained in their size by their ability to handle dispersed knowledge, as Foss points out: Implicitly the knowledge dispersal problem is seen in this literature as determining the boundaries of the firm, for there is a point where the ‘loss of control’ is so overwhelming that it more than offsets any gains from, say, integrating one more line of business or making one more foreign direct investment. (Foss 1997, p. 187)

To the extent that this was the question that originally motivated Coase to study firms, fully incorporating the double knowledge (or the knowledge asymmetry) problem into the theory of the firm contributes an important element to the Coasian perspective. All of these elements together form a solid foundation for a more thorough study of decentralized forms of organization. The work done on

72

Employees and entrepreneurship

decentralization by Langlois, Minkler, and Foss was of a peripheral nature: decentralization was not the center of their inquiry. Sautet (2000) was a large step forward in this area, as it explained decentralization as a way to generate new knowledge for the firm. However, Sautet restricts his examination of decentralization to the development and adoption of the multidivisional structure. I am interested in explaining the more radical application of decentralization of decision-making that we see in modern markets. So, though the work of the above scholars is important and a fundamental starting point, I hold that there is much more that remains to be explained about decentralization. I will suggest several new theoretical tools that I believe will be important in that study, tools that have not gotten much if any attention yet. It is the lack of understanding of these elements that recently led several economists working in the Austrian tradition to question the benefits of decentralization. I will discuss this next.

FOSS’S CRITIQUE OF INTRA-FIRM DECENTRALIZATION Seemingly paradoxically, as the mainstream of the economic profession discovers the Hayekian knowledge problem in the process of attempting to explain the decentralization developments, a few of the more recent Austrian works in this field (Foss 2001a; Foss and Klein 2005; Witt 1998) have been emphasizing the importance of centralized control within a firm and de-emphasizing the benefits of a reduction of managerial hierarchies. The logic behind this development is summed up well in the following passage: ‘Application of basic Austrian ideas suggests that the entrepreneur needs the relevant knowledge to organize the activities of his firm. In other words, the more important an individual’s knowledge, the higher he should be in the hierarchy’ (Garrouste 2002, p. 79). The primary question of decentralization of decision-making is who will get the decision rights within a firm, and the Austrian economists listed above increasingly answer that the ‘entrepreneur’ (in the sense of the owner or the CEO) will be the one with the greatest – or at least decisive – information and thus must be given the decision rights. They recognize that the Hayekian dispersion of knowledge within the firm exists and matters, but their claim is that it is dominated by the benefits of conscious control and co-ordination of productive resources (both capital and labor), a self-consciously Coasian perspective. This stands in stark contrast to the earlier contributions by Austrian economists in the field of the theory of the firm which, as I have shown above, tended to be generally supportive of intra-firm decentralization developments.

The knowledge problem in firms

73

The most striking example of this reversal can be seen in the work of Nikolai Foss. In contrast to Foss 1997 and 1999, Foss 2001a, 2001b, and 2001c strongly question the logic of decentralization. In this series of articles Foss attempts to show the dominating benefits of authority (and the corresponding strongly hierarchical organizational structure), and the problems with ‘hybrid’ organizational forms (those combining aspects of markets and firms), in particular their instability (because they are not an organizational equilibrium). Foss is interested in showing that, even when decentralization is applied in ‘knowledge firms’, it does not work nearly as well as the proponents would make it seem. Recall that knowledge firms are the most likely to exhibit an internal situation that Sautet defined as the ‘double knowledge problem’ and Minkler defined as ‘asymmetric knowledge’. Foss explains this situation in the following way: One reason why authority is (allegedly) waning in importance is that it is becoming increasingly more difficult to monitor and direct workers, because of the specialist nature of knowledge work . . . [The principal] may be ignorant about members of the set of possible actions open to the agent, or the agent may be better informed than the principal with respect to how certain tasks should (optimally) be carried out, or both. (2001b, p. 4)

He goes on to say that, ‘even in such a setting, it is possible to provide efficiency explanations of authority’ (ibid.). Foss’s argument is a direct response to Cowen and Parker (1997), as can be seen in the following passage: [I]t is argued that only those firms that emulate markets inside their internal organization to the largest possible extent will survive and prosper in the knowledge economy (Cowen and Parker 1997). The ‘Coasian firm’, characterized by well-defined boundaries, authority, etc. will, in contrast, wither. (2001a, p. 5)

I have already critiqued Cowen and Parker above: firms with decentralized decision-making do not emulate markets. They are alternative institutional forms that have their own characteristics and structures. It is not only overly simplistic to describe them as being ‘market-like’, but profoundly inaccurate: markets are (complex) systems of exchange based on money prices and property rights, neither of which is featured to any great extent within (considerably less complex) internally decentralized firms. The employees are not necessarily the owners of the capital that they utilize, and certainly very few firms with decentralized decision-making have internal exchanges based on ‘prices’.22 It is hard to understand how anybody could describe an institution that lacks both of these characteristics as marketlike! We must come to the conclusion, as I did above, that Cowen and

74

Employees and entrepreneurship

Parker (1997) do not offer a good explanation of decentralization, and therefore any critique of decentralization based on their work should be immediately suspect. And so it is with Foss (2001a), the central theme of which is that ‘it does not follow that firms should emulate markets as far as possible’ (p. 5, italics in the original). Foss presents a rather novel critique, though, pitting Mises against Hayek, in a manner of speaking.23 Foss describes the ‘prodecentralization’ literature as basically Hayekian, in the sense that there is an emphasis on the Hayekian knowledge problem.24 But Foss objects that the Hayekian knowledge problem has been overemphasized in the prodecentralization literature. As he says, ‘the very fact that firms exist is prima facie evidence that they can somehow cope with the problems implied by Hayekian settings’ (p. 14). And, according to Foss, while the knowledge problem has been overemphasized, the role of an entrepreneur-leader who holds ultimate authority has been greatly underemphasized. He backs up this claim by referring to Mises: In his critique of market socialism, Mises (1949) pointed to the folly of ‘playing markets’, and I draw on his overall argument that bringing coordination mechanisms characteristic of market organization into a planned organization is inherently problematic . . . I . . . argue that firms are also systems of complementary elements and that this fact places constraints on the extent to which firms may be made ‘market-like’. In particular, I agree with Mises that ‘[t]he function of the entrepreneur cannot be separated from the direction of the employment of factors of production for the accomplishment of definite tasks. The entrepreneur controls the factors of production’ (Mises 1949: 306). (2001a, p. 5)

We see in this paragraph that Foss derives two ideas out of Mises: 1) just as it is impossible to have intermediate forms of economic systems which combine elements of markets and socialism, so it is with markets and firms – we cannot ‘arbitrarily’ combine markets and firms and expect to create stable institutions; 2) the entrepreneur must direct the factors of production within a firm, a task which cannot be assigned to the employees themselves. I will argue that Foss is wrong (to different degrees) on both points. Decentralized Firms as Combinations of Firms and Markets? First I will discuss Foss’s claim that it is impossible to combine firms and markets. He explains the ‘firms-as-markets’ claims made by prodecentralization writers: One possible interpretation of the recent literature on economic organization in the knowledge economy is that as Hayekian settings become increasingly

The knowledge problem in firms

75

prevalent, traditional authority relations vanish . . . and coordination mechanisms (i.e., authority, norms, teams, prices, etc.) will increasingly be combined in new, innovative ways – resulting in what is often referred to as ‘new organizational forms’, and substituting for traditional relations of authority. This final claim implies that organizational forms do not cluster in a few rigid, discrete forms, but, on the contrary, that coordination mechanisms are highly malleable. In particular, firms may adopt coordination mechanisms that we normally think of as characteristic of the market rather than of planned coordination. (2001a, pp. 8–9)25

Foss is implying here that ‘new organizational forms’ cannot really exist, a claim I believe is rather difficult to stake: the existence of ‘new organizational forms’ would appear to be self-evident. One need only take a look at the last 10 or 20 years of business and management literature to see that this is so. However, though Foss is not very clear on this point, it would appear that his view is not that firms featuring various forms of decentralization are not observed at any point in time, but rather that such firms are unstable in the sense that they are unable to retain their decentralized features for any lengthier period of time. He supports this claim in Foss (2001c) by further examining the Danish firm Oticon A/S. As I explained in Chapter 1, Oticon implemented a radical form of decentralization for a period of time only to ultimately reject it and return to a more-or-less straightforward hierarchical form. Thus, Foss draws the conclusion from this example that it is very difficult – if not impossible – to properly ‘shape’ a ‘new’ and, more importantly, lasting form of organization. In other words, it is unlikely that a decentralized organization can ever be an equilibrium solution to the organizational problem. This raises the question of how long an organizational structure must last to be considered an equilibrium solution. The decentralized structure in Oticon lasted between five and eight years (depending on the interpretation of events), during which time Oticon became a market leader in several innovative products, gained significant market share and became very profitable. If a period greater than five years does not qualify as an equilibrium solution, then we need to rethink our notion of equilibrium. In addition, it should be obvious that one example of failure of radical decentralization does not condemn the entire organizational form, and I have shown examples of radically decentralized and successful companies that have lasted for much more than five years and which still continue, such as W. L Gore & Associates Inc. In the case of Oticon, it would appear that management made key mistakes that unraveled the entire organizational form, as I discuss just below. But it is by no means a matter of certainty that management teams in all firms exhibiting greater degrees of decentralization will always make mistakes that render decentralization impossible.

76

Employees and entrepreneurship

And as I pointed out in Chapter 1, even if the Oticon management failed to create a stable decentralized structure, in many ways their experiment was a success, placing the company back in the position of market leader, in terms both of technical and product innovations and of market share. I believe that Foss goes wrong when he accepts the perspective of decentralization proponents who repeatedly refer to decentralization as a process of emulating the market, which, as I explained above, is not what decentralized firms do. Therefore, it is incorrect to analyze these ‘new organizational forms’ as combining hierarchy and markets. Whatever they may be, they are not markets. Once we recognize that, it very much calls into question the applicability of Mises’ insight that markets and central planning cannot be mixed to derive a stable economic system. Foss explains his use of Mises’ insight in more detail in the following paragraph: [R]ather than being combinable at will, coordination mechanisms, such as authority, delegation, pricing, etc., tend to cluster in predictable ways. This is an application of sorts on the level of the firm of Mises’ demonstration that the various elements that make up the capitalist market economy are complementary ones; one cannot simply take a subset of these away, say, unhampered capital markets, and substitute them with elements that are characteristic of a different system. In a similar manner, concentrated ownership, authority, circumscribed decision rights, and incentives that are less ‘powered’ than those of the marketplace are all complementary elements of a system, namely, the firm, and they will continue to be so, even in the knowledge economy. (2001a, p. 19)

It bears pointing out that Foss is not actually using Mises’ argument directly, but rather only by analogy: Mises was not speaking about intra-firm structures when making his claim. In light of this, we can see that the first problem with Foss’s argument against decentralization is that, if decentralized firms do not actually emulate markets, the analogy breaks down, and Mises’ insight is irrelevant to the question of intra-firm decentralization. But an even more significant problem with Foss’s argument is that Mises’ above objections to ‘playing market’ were aimed at coercive interventions into the spontaneous market order. These coerced market interventions were done (or proposed to be done) in a way that would not allow the consumers to have any say in determining the efficiency and desirability of the new structure – consumers were prevented from ‘voting with their dollars’. That is most certainly not the case with decentralized firms. First of all, decentralized firms are often far more experimental rather than designed in nature: they are generally an outcome of a spontaneous – or semispontaneous – incremental evolutionary process, rather than rationalistically ‘designed’ in the way that market interventions and non-market, governmental institutions are. Second, and more importantly, the decen-

The knowledge problem in firms

77

tralized firm structure is not coerced into the markets; instead, it exists within a wider catallactic order where employees unhappy with the decentralized firm structure (or performance of such a structure) can freely quit and get jobs in more hierarchical firms; and consumers unhappy with the products of decentralized firms can freely choose goods made by more hierarchical firms. And this underlies my ultimate objection to Foss’s analogy between Mises’ rejection of a mixed economic system and Foss’s rejection of decentralized firms: decentralized firms are still very much subject to market forces and must compete with predominantly hierarchically structured firms. If indeed decentralized structures were always inferior, as Foss appears to imply, that would very quickly become discovered through the process of market competition. As it is, even Foss is forced to acknowledge that decentralized firms do in fact exist.26 Thus when Foss argues for the sanctity of the traditional ‘Coasian’ structure of the firm by saying ‘the very fact that firms exist is prima facie evidence that they can somehow cope with the problems implied by Hayekian settings’ (2001a, pp. 13–14), we can just as easily say that the very fact that firms exhibiting decentralized elements exist is prima facie evidence that they have been able to cope with whatever problems that structure may create for them (more on that below). The Role of an Authoritarian Entrepreneur The second critique of decentralization that Foss derived from Mises is that firms must have direction of resources within them by an entrepreneur with ultimate decision-making authority. Foss claims that entrepreneurs holding some ultimate authority in firms are a key component of any firm’s organizational structure, and cannot be dispensed with, as firms with decentralized decision-making appear to wish to do. But as I discussed above, though firms with a great amount of internal decentralization tend to have fewer managerial layers and therefore less direction of employees, they inevitably still feature some ultimate decision-maker, whether an owner or some sort of a manager. Foss defines the ‘principal’ in this relationship to be the entrepreneur in the Misesian definition;27 the agents, of course, are either lower-level managers or sometimes the ordinary employees. In all firms, whether centralized or decentralized, the Misesian entrepreneurs delegate their own decision-making rights to the agents to different extents. Foss discusses Mises’ explanations of this phenomenon: The reason that firms can thrive even though their internal organization exemplifies Hayekian settings is that they have recourse to delegation. As Mises (1949, p. 303) emphasized, ‘entrepreneurs are not omnipresent. They cannot themselves attend to the manifold tasks which are incumbent upon them.’ Mises

78

Employees and entrepreneurship (p. 305) clearly recognized that in many firms decision rights are allocated by the entrepreneur (and the board of directors) to lower levels, presumably in order to better cope with distributed knowledge, an insight that is not present in Coase (1937). He perceptively recognized that delegation leads to agency problems, but argued that the system of double-entry book-keeping and other control measures may partly cope with such problems. Mises also understood that delegation of decision rights is circumscribed in an attempt to cope with the control problem that follows from delegation. (2001a, p. 14)

Foss demonstrates that Mises was aware of inherent tradeoffs associated with organizing a firm: delegation (or decentralization) has the benefits of relieving the entrepreneur from carrying out many tasks that others can do just as well, thus allowing her to focus on just those tasks that she is uniquely qualified to carry out; but it comes at a cost of creating agency problems. Foss reads Mises (accurately, in my opinion) as saying that the costs of delegation become higher than the benefits rather quickly, leading to a situation where relatively few, and mostly unimportant, decisions can be delegated. Foss further extrapolates that firms with some decentralization ‘are prompted by a market-driven pressure to delegate decision rights (for example, to better serve customer preferences) and structure reward schemes in such a way that optimal trade-offs are reached’ (pp. 14–15). Foss adds a proviso, though: decision-making rights ‘are delegated as means to end, their use is monitored, and top-management reserves ultimate decision rights for itself’ (p. 15). The first of these three seems like a rather odd thing to point out: of course decision rights are delegated as a means to an end. As I pointed out, every firm will always possess important elements of an ‘organization’ (in a Hayekian sense) because all firms have a bottom line, an ultimate end, of maximizing profits. Delegation/decentralization is not done for its own sake, whatever that may mean, but rather to improve the performance of a firm, and as soon as it appears unlikely that it will do that over the long run it should be abandoned. As for monitoring the use of decision rights, this is another rather obvious element of decentralization, with an important limitation: as Minkler explained, in an asymmetric knowledge situation the entrepreneur may not be able to understand fully whether the agent is acting responsibly or irresponsibly with the delegated decision-making rights, so monitoring, while to a certain extent absolutely necessary and unavoidable, can sometimes be of limited use. The third point that Foss makes, that the entrepreneur reserves ultimate decision rights for herself, I will discuss in greater detail next. Foss concludes that radically delegated decision rights are unlikely to be used in an efficient manner by employees, and therefore attempts at intrafirm decentralization are most likely doomed:

The knowledge problem in firms

79

An immediate implication of this kind of reasoning is that emulating market organization inside firms, by radically decentralizing the firm and allocating farreaching decision rights to employees may be hard to accomplish in a successful manner . . . [T]here are incentive limits to the extent to which market mechanisms can be applied inside firms, and delegation, while not exactly a rare flower, is certainly a very delicate one. (2001a, pp. 18–19)

Foss questions the logic of delegating decision-making rights for two reasons: 1) there are several reasons why the entrepreneur, that is, the person with ultimate authority, should have and keep her decision-making rights; 2) the entrepreneur can never make a credible commitment to not engage in selective intervention, thus distorting the incentives faced by the agents. I will address each of these in turn. As far as the first reason goes, Foss makes an important contribution with his discussion of the benefits of centralized decision-making. It should be emphasized again that Foss carries out his analysis within an implicit asymmetrical knowledge setting, where employees have greater knowledge than an employer. Foss shows that even in such settings there are significant reasons why the intra-firm decision-making should still be centralized: [I]t is possible to explain the presence of authority in [an asymmetric knowledge] setting, in the sense of it being rational to give one agent decision-making power over another one. I discuss the rationales for this under the headings of ‘the need for urgent coordination’, ‘decisive information’, ‘economies of scale in decisionmaking’ and ‘defining incentive systems’. (2001a, pp. 11–12)

These four rationales have been improperly disregarded or dismissed by proponents of decentralization. The first rationale for centralized or authoritarian decision-making can be summed up as ‘the need for urgent coordination’: Coordinated adaptation or action may be required when actions or activities are complementary, when it is important to make some urgent choice (possibly highly inefficient), because doing nothing is worse. In such cases, it may be better to have somebody pick a strategy and make everybody play this strategy . . . [T]he decentralized solution performs poorly if urgency is important. (2001a, p. 12, italics in original)

The second rationale for centralized decision-making is the possession of ‘decisive information’ by one person: Even under distributed knowledge, where the centralized decisionmaker per definition does not possess (at least some) local information, he may in many cases still hold the information that is decisive. (2001a, p. 12, italics in original)

80

Employees and entrepreneurship

The third rationale for centralized decision-making is ‘economies of scale in decision- making’: [Economies of scale in managing in the form of effort costs of negotiating, learning about potential suppliers, etc.] may relate both to managing the internal relations between agents inside the firm and managing relations to outside agents (customers, suppliers, government agencies). (2001a, p. 13)

The fourth and final rationale for centralized decision-making is ‘defining incentive systems’: [E]ven under hidden knowledge, the principal may be able to form conjectures of the financial results that result from the agent’s activities, and he can check whether these conjectures are actually confirmed using the control system of the firm . . . Hidden knowledge does not imply that subjective performance measurement becomes impossible. (2001a, p. 13)

These four reasons go a long way towards explaining why we do not see more decentralization of decision-making than there already exists. But the most important conclusion we can draw from them is that determining the internal structure of a firm is a very complex process involving many different factors. In addition, the importance and nature of each one of the factors is often very difficult to evaluate ex ante. For example, the question of whether there will actually be economies of scale in decision-making accruing to the entrepreneur or whether by holding on to those decision-making rights the entrepreneur is giving up important contributions by the employees is an extremely difficult one, and likely to vary dramatically through time and circumstance. Ultimately, there must be a combination of experience, judgment, and trial and error to arrive at the correct answer to such a question. The existence of these four factors is not a ‘slam-dunk’ for the ‘Coasian firm’ but rather an important consideration for entrepreneurs in deciding how many of their decision rights to delegate. The second factor leading Foss to question the delegation of decisionmaking rights is that an entrepreneur can never make a credible commitment to not engage in selective intervention, a scenario that will have a distorting effect on the incentives faced by the agents. In other words, it is impossible for principals in a firm to credibly reject their power to make ultimate decisions. Employees will understand this, and will therefore respond by changing their behavior in undesirable ways. I discussed this problem in Chapter 2: Williamson, who referred to it as ‘the impossibility of selective intervention’, was the first to formulate it, and Foss explains it in the following way:

The knowledge problem in firms

81

The main problem is that incentives are diluted. This is because the option to intervene ‘. . . can be exercised both for good cause (to support expected net gains) and for bad (to support the subgoals of the intervenor)’ (Williamson 1996: 150–151). Promises to only intervene for good cause can never be credible, Williamson argues, because they are not enforceable. Although Williamson may be going too far, a main conclusion in this literature is indeed that credible delegation may be very hard to accomplish, since reneging on a promise to delegate will in many cases be very tempting and those to whom rights are delegated anticipate this. An immediate implication of this kind of reasoning is that emulating market organization inside firms, for example, by radically decentralizing the firm and allocating far-reaching decision rights to employees may be hard to accomplish in a successful manner. Unlike independent agents in markets, corporate employees never possess ultimate decision rights. They are not full owners. This means that those who possess ultimate decision rights can always overrule employees. (2001a, pp. 18–19)

As I explained in Chapter 2, this logic leads Williamson to more or less reject the possibility of the existence of any sort of delegation (or, as Williamson refers to it, a hybrid organizational form). Though Foss notes that Williamson goes too far, he does not clarify exactly how he thinks Williamson’s claims should be moderated. Foss instead seems to fully accept Williamson’s critique. I propose that both Williamson and Foss are failing to take into account two important factors, which could sufficiently moderate the ‘non-credibility of rejection of selective intervention’ problem to make decentralized decision-making within firms workable. They are: 1) the wide use of the tit-for-tat strategy by employees, and their subsequent accumulation of information through the reputational effect; 2) absence or successful removal of opportunism on the part of the entrepreneur and/or top managers. First, I will discuss the role that the tit-for-tat strategy and reputational effects play in decentralization of decision-making. Game theorists have shown that the so-called tit-for-tat strategy, where a person cooperates in the first round and in each subsequent round adopts the opponent’s strategy from the previous round, can be very successful in maximizing gains from trade. The most important factor for this to happen is that the game must be repeated many times, with no known end. I believe that this applies in most firm situations. The entrepreneur could always interfere and take away the delegated decision-making rights, but she will know that she will lose the trust of the employees, and therefore any subsequent delegation she attempts to engage in will not be trusted, because she will develop a reputation as a ‘cheater’. Therefore, the entrepreneur would forfeit all possible future benefits of delegation/decentralization by overruling the employees on a decision that was previously delegated to them.

82

Employees and entrepreneurship

To better illustrate the important role that reputational effects play in the process of delegation it is useful to take another look at the two examples of radical decentralization of decision-making I have discussed previously, Oticon and Semco. Foss uses the story of Oticon to demonstrate the selective intervention problem. Though the radical decentralization of Oticon between 1990 and 1995 was a great success in terms of innovations and gaining a competitive advantage, Oticon reverted to a much more hierarchical structure in 1996. The main reason according to Foss’s analysis was that the CEO, Lars Kolind, frequently engaged in selective interventions, leading to employees becoming discouraged and disillusioned. Though employees had a large amount of discretion, there continued to exist one all-important hierarchical layer within Oticon that led to the ultimate undoing of the company’s radically decentralized decision-making organization structure. As Foss explains, [P]rojects had to be evaluated by the Products and Projects Committee that was staffed by Kolind, the development manager, the marketing manager, and the support manager. The Project and Products Committee either rejected or approved of the project . . . Although a considerable amount of variety was indeed allowed to evolve, the selection over this variety was very much guided by the visible hand of the Products and Projects Committee. (Foss 2001c, p. 15)

Of course, the existence of the Products and Projects Committee is not necessarily a barrier to successful decentralization of decision-making within a firm. But Foss shows that the Committee abused its gate-keeping power by engaging in frequent selective interventions in those five years of radical decentralization, leading to ‘diluted incentives and a general state of de-motivation’ (ibid., p. 22) among the employees. The Committee created a situation where ‘projects were interrupted in seemingly arbitrary ways’ (ibid.). Since the Committee was unable to credibly commit to a policy of non-intervention, the employees stopped introducing new projects. The management’s view may have been ‘that in important respects and in many situations, they were likely to possess decisive knowledge, and that efficient resource-use dictated intervening in, and sometimes, closing down projects’ (ibid.). The management obviously did not have faith in the radical decentralization of decision-making they had created, and ultimately the employees engaged in a tit-for-tat strategy, bringing about an unraveling of the decentralization. Oticon in 1996 adopted a more structured, hierarchical organizational form (though, according to Foss, ‘Oticon headquarters is still by any reasonable standard an organization characterized by much delegation of decision rights’ (ibid., p. 16)) because once the management had lost the ability to convince the employees

The knowledge problem in firms

83

that they would not give in to the temptation of selective intervention there was nothing else they could do but re-establish some hierarchical controls.28 A rational, long-term-oriented entrepreneur who wished to take advantage of decentralization of decision-making would have to discipline herself not to give in to this type of temptation, understanding that it is necessary to gain employee trust in order to allow delegation of decisionmaking rights to succeed. Ricketts (2002) explains that a ‘sufficiently high probability of repetition thus changes the structure of the exchange game . . . cheating is no longer the dominant strategy’ (p. 21). This explains why Ricardo Semler (see Chapter 3, ‘Decentralized decision-making through hierarchical flattening’, page 59) refused to overrule the employees of his firm even when he disagreed with them: ‘Employees also outvoted me on the acquisition of a company that I’m still sure we should have bought. But they felt we weren’t ready to digest it, and I lost the vote. In a case like that, the credibility of our management system is at stake.’ Semler showed a keen awareness of the value of ‘cooperating’ in this one round (despite the possibility that the value of his company might be harmed in the short run) in order to gain the benefits of continued trust and continued cooperation by the employees. Semler demonstrated that in fact an entrepreneur can be credible in his commitment to non-interventionism. Since all parties understand the potential benefits of a successfully decentralized firm, it is possible that in the first round of the game the entrepreneur and employees will engage in cooperative behavior, which in turn will lead to a self-reinforcing process of continuous cooperation. As Ricketts again explains: ‘In certain conditions, self-interested behavior may result in the widespread adoption of the “tit for tat” strategy in trading games. “Tit for tat” becomes . . . a convention. Once established there are powerful forces of self-interest tending to maintain it’ (2002, p. 22). It would seem that, unlike Oticon between 1990 and 1995, Semco (as well as W.L. Gore & Associates, as I showed in the first chapter) successfully created a convention of managerial non-interventionism, and they reaped the benefits. This process of the development of a convention is more likely to take place within a firm than in a market situation for two important reasons: 1) unlike market participants, employees in a firm are guaranteed to have repeated dealings with one another and with the entrepreneur; and 2) within a firm it is relatively easy to accurately observe whether a participant cooperated or cheated, which is sometimes more difficult to determine in market situations. For these reasons we can confidently expect widespread adoption of the tit-for-tat strategy among employees, thus severely restraining the incentives for the entrepreneur to engage in undesirable (in the long run) selective interventionism.29

84

Employees and entrepreneurship

The second factor indicating that Williamson and Foss are exaggerating the problem of selective interventionism is one of opportunism. A key part of Williamson’s theory is his assumption of opportunism on the part of not only employees but also the entrepreneur. However, if we could show that people within firms are not opportunistic in the ways that Williamson assumes, then the whole credible non-interventionism commitment problem breaks down. I will discuss opportunism and worker motivations in the next chapter. There I will show that several recent studies have come to the conclusion that opportunism is not nearly as widespread as Williamson makes it seem, therefore allowing for the possibility of credible commitment by entrepreneurs and managers to abstain from engaging in selective interventions into delegated decision-making by employees. In fact, contrary to Williamson and Foss, we can make a case that the existence of occasional opportunistic individuals can in fact explain why we have been observing a ‘flattening’ (removal of layers of middle management) of the organizational structures as a common part of the decentralization process. Even though most managers with authority may not be opportunistic, opportunistic behavior by just one can spoil the entire reputational environment. For this reason entrepreneurs (who can be expected to truly want to maximize the long-run benefits of decentralization of decision-making and will therefore have the most to lose from selective intervention, which will undermine the confidence in the delegation process) could remove many of the middle managers as a way to ensure the least amount of selective interventionism. Conclusion Despite my objections to Foss’s arguments, I whole-heartedly agree with him that pro-decentralization literature has been far too eager to proclaim the death of the Coasian firm: Proponents of the knowledge economy notion assert that these settings are becoming increasingly prevalent in today’s business landscape, in the sense that an increasing fraction of firms experiment with decentralizing their internal structures, build relations to external knowledge sources, etc. (2001a, p. 4)30

Whether they are becoming increasingly prevalent or not is an empirical question, which has so far not been conclusively answered.31 But I suspect that the reports of the death of the Coasian firm have been greatly exaggerated. Finding the ‘right’ institutional structure to make a firm properly decentralized indeed is not easy, and it is possible and even likely that many firms will get it wrong – meaning that the costs of decentralization will turn

The knowledge problem in firms

85

out to be greater than its benefits and the decentralization experiments will be abandoned, firms returning to a more hierarchical structure. In addition, external circumstances may change in ways that will make a radically decentralized decision-making structure no longer necessary, and we will observe a return to a hierarchical, controlled structure. Ultimately I argue, as did Langlois above, that, owing to the unavoidable ‘flexibility–efficiency tradeoff’, it is impossible for economists to determine the absolutely more efficient internal structure of the firm – rather, it must be discovered by entrepreneurs and other actors within firms themselves. This is ultimately one of the most important tasks for the figures with authority within firms. Foss is certainly correct to point out that the proponents of decentralization are not right in claiming that all firms would be better off if they engaged in a greater degree of delegation. But he goes too far in the opposite direction to show this. I believe that the proper conclusion to draw from this discussion is that firms must have an appropriate structure for their particular competitive environment. Firms deal with Hayekian or knowledge-work settings by finding the proper amount of decentralization befitting the situation. It may or may not in fact be the case that competitive environments are changing in such ways that more firms will have to become more decentralized in order to survive. This also is not a theoretical issue, but rather an empirical one, and we as economists cannot say how much decentralization is appropriate nor which firms should be decentralized – this must be done by entrepreneurs and employees in the firms themselves.

SEEKING CREATIVITY WITHIN FIRMS In most of the pro-decentralization literature (with the exception of Sautet 2000), the authors largely failed to discuss how intra-firm decontrolling brings about a stimulated creative response from employees. In fact, there has not been much work done exploring the connection and distinction between the use of dispersed knowledge and the creation of new knowledge (whether in the form of new products, new production processes, or new markets). Though these two would clearly seem to be connected, they are not the same. My claim is that the primary benefit of decentralization of decision-making within firms is not necessarily the better co-ordination of dispersed knowledge, but rather the generation of the greatest amount of new knowledge. To the extent that creativity is the necessary antecedent for innovative behavior, firms seeking to foster their innovative capabilities can do so by adopting intra-firm structures that encourage creative problem-solving by the employees. Minkler recognizes the importance of

86

Employees and entrepreneurship

creativity and innovation, as can be seen in the following quote, though ultimately he does not advance this important insight any further: ‘Successful firms are the ones that develop organizational capabilities that allow them to continue to innovate; firms that rest on their past accomplishments decline’ (1993b, pp. 583–4). Sautet (2000) does go further, explaining that the multidivisional form of intra-firm decentralization is primarily instituted as a way to generate new knowledge: [T]he decentralized structure is a locus of discovery and of exploitation of profit opportunities. This is possible, not simply because bounded rationality problems were overcome (even if this is certainly true), but rather because the knowledge necessary for the growth of the firm could be discovered only through a decentralized structure that is conducive to entrepreneurial discovery. The M-form allows for the discovery of knowledge that would not have been discovered otherwise. (Sautet 2000, p. 132)

New knowledge is discovered or generated as there is greater ability for employees to act on their existing knowledge. And the ability for employees to act on their existing knowledge is very much determined by the intrafirm structure or, more precisely, the rules governing the employees. The rules within successful firms characterized by a great amount of decentralized decision-making have been designed (or in some cases have evolved) to ‘extract’ the maximum amount of creativity from employees at different levels of the firm hierarchy. In a world of intense competition, it is often too costly not to take advantage of employees’ creative problemsolving, since firms need to be as responsive to changing market conditions as possible. In fact, many businesses fully expect that their employees will not act as simple automatons, carrying out orders from above, but will rather creatively engage whichever problems may arise in the process of production. In other words, employees are expected to behave as entrepreneurs, and we can thus explain the benefits of decentralized firms as encouraging entrepreneurial – meaning creative, innovative, alert – employee behavior. Firm rules which are capable of harnessing and using these employees’ entrepreneurial judgment, alertness, and imagination may be more likely to survive the competitive evolutionary process, as seen in multitudes of examples of decentralized firms in today’s economies. Such rules must allow for some level of personal decision-making and, by logical implication, restrict the amount of planning from above that can take place.32 Therefore, the benefits of decentralized firms may be explained not only in terms of the Hayekian ‘dispersed-knowledge’ co-ordination but also in terms of Schumpeter’s evolutionary-entrepreneurial stimulus to

The knowledge problem in firms

87

innovation.33 I will explore the connection between decentralization and employee creativity in more detail in Chapter 5.

CONCLUSION In this chapter I have summarized and critiqued the current state of economic knowledge about decentralization within firms, resting mostly on the idea of the knowledge problem within a firm. Sautet refers to this as the double knowledge problem, while Minkler labels it an asymmetrical knowledge problem. Some important foundations have been laid in the past 20 years, but there has also been work that I believe has led us down a few wrong paths. I have tried to separate out the valuable contributions from the false ones. But, as insightful and beneficial as the valuable contributions are, they do not go far enough. Several important questions remain mostly unanswered, and the analysis has not been pushed as far as it should. The unexplained or insufficiently addressed aspects of decentralization of decision-making within firms can be summed up in the following two questions: 1) What motivates workers to work hard if they are not continuously monitored – in other words, how do decentralized firms attenuate the agency problem? 2) How do firms with decentralized decision-making coordinate the actions of the multitudes of employees empowered with their own decision-making rights – in other words, how are they able to avoid absolute chaos and discoordination? I will answer these questions in the next chapter. I will show that there exists a process generating a spontaneous order within firms characterized by a high degree of decentralization that is different than the spontaneously ordering process we observe in the markets, and that we can explain it by relying on Schutzian sociological philosophy. In addition, I will show that firms with properly structured decentralized decision-making can effectively reduce the agency problem, contrary to the established wisdom of transaction cost economists, and that workers who are not acting in opportunistic ways are much more likely to spontaneously co-ordinate their actions.

NOTES 1.

As an example, see Cowen and Parker 1997: ‘We argue that firms should rely on marketbased mechanisms to an increasing degree, and that the problems of command approaches to management resemble the problems faced by all forms of central planning of resource allocation’ (p. 17, emphasis added). Also, see Gable and Ellig 1993: ‘Historical experience shows that market economies, which rely on the dispersed

88

2.

3.

4.

5.

6.

7. 8.

Employees and entrepreneurship knowledge and independent judgment of numerous consumers and producers, consistently provide a dramatically higher quality of life than centrally planned economies. Given that reality, it makes sense to examine how market economies coordinate human activity in order to glean lessons for improving business management. Unfortunately, many analysts in business and academia resist this approach, out of a belief that market concepts apply only “out there”, beyond the boundaries of the firm. In this view, the principles of a free society apply in the external market, but the firm’s internal affairs are the province of brilliant planners making command decisions . . . [W]e believe firms that fail to learn and adapt market principles internally will one day find themselves distant competitors to firms that do’ (pp. 8–9). Again, see Cowen and Parker 1997: ‘[I]t is far from obvious that the border between the firm and the market was ever as distinct as the earlier literature implied . . . Arguably, therefore, the differences between markets and hierarchies are best analyzed in terms of a unified theory of contracts and incentives rather than as fundamentally different institutional structures’ (p. 42, emphasis in the original). Langlois was one of the pioneers in this field, and it is largely due to him that the study of the firm, as he himself put it, is ‘beginning to attract – and to yield to – insights and approaches one could characterize as fundamentally Austrian. Among these insights are the importance of economic process and the tacit and decentralized character of economic knowledge’ (1994b, p. 173). The following quote effectively sums up Foss’s research agenda.’[T]he ambition of the present paper is not to take steps towards an alternative Austrian theory of the firm. Rather, it is the more humble one of suggesting that Austrian economics is a challenge to the [modern economics of organization], that economists of organization may derive inspiration from Austrian economics, and perhaps even that some sort of combined research program may be a worthwhile endeavor’ (Foss 1999, p. 459). Sautet refers to the intra-firm knowledge problem as the ‘double Hayekian knowledge problem’, and explains it in the following way: ‘In the complex firm, individuals are alert, on the one hand, and possess knowledge that can be local, tacit and social in nature, on the other. In that sense, the complex firm cannot function like a simple firm, for the employees’ knowledge cannot be entirely centralized. There will always be knowledge possessed by the employees that will depend on the particular circumstances of time and place and which the promoter cannot know (even if this would be valuable to him/her). This knowledge could be about the internal allocation of resources or about a profit opportunity in the marketplace. In other words, there is a [Hayekian knowledge problem] in the complex firm: the entrepreneur-promoter can be ignorant of his/her ignorance with respect to the knowledge possessed by some of his/her employees (and this knowledge could be crucial to the firm). This is in addition the [Hayekian knowledge problem] in the marketplace’ (2000, pp. 98–9). To a certain extent Harper is restating Williamson’s work here. They both define paradigmatic firms as purely hierarchical organizations that are completely without any element of order. Harper’s list of hybrid modes of governance is almost identical to Williamson’s: ‘venture capital contracts, venture nurturing (including corporate new venture divisions and R&D partnerships), venture spin-offs, joint ventures, franchising, reciprocal trading (including product exchange agreements), share contracts, quasivertical integration and other forms of non-standard contractual arrangements’ (1996, p. 154). However, Harper differs from Williamson by virtue of being more open to the possibility of the existence of multitudes of governance structures, in addition to those listed here, along the governance continuum. Williamson appeared to believe that intermediate forms of organization are very few and he left out any consideration of decentralized-decision-making firms, as I discussed in Chapter 2. Foss puts it this way: ‘[T]he Austrian insight that most economically relevant knowledge is local and tacit is not systematically incorporated into contemporary Coasian theories of the firm, at least with regards to production knowledge’ (1994, p. 56). Foss brings up this same critique in another article: ‘The dispersal of knowledge creates coordination problems that go beyond the incentive coordination problems that are

The knowledge problem in firms

9. 10.

11. 12.

13.

14.

15.

16. 17.

89

treated in the [Modern Economics of Organization], and are consequently sidestepped in this body of theory. A consequence of this neglect is that the MEO tends to neglect the role of management, leadership, routines, capabilities, and shared cognitive categories (e.g. corporate culture) in coordination, except when these can be interpreted as either manifestations of ex ante incentive alignment or ex post governance problems that are non-standard in the context of the MEO’ (Foss 1999, p. 459). Minkler defines structural uncertainty as ‘the type of uncertainty that exists if a decisionmaker cannot ex ante specify all relevant alternatives or outcomes’ (1993b, p. 571). It is ultimately caused by dispersed knowledge. Minkler in a different article phrases this problem in the following way: ‘But if a principal does not know what agents should be doing (maybe because knowledge is dispersed through the organization or because an agent knows about production processes and outcomes the principal does not) the problem confronting the principal is different and cannot be solved by monitoring. In certain predicaments, the principal would not even be able to decipher what was observed. Direction would be pointless’ (Minkler 1993a, p. 18). According to Minkler: ‘Information-based models preclude the possibility that agents could understand more than the principal about certain aspects of production’ (1993b, p. 570). This is one thing not well understood by critics of decentralization: they object that a decentralized firm cannot solve the moral hazard problem while remaining oblivious to the fact that in this asymmetrical knowledge situation the standard firm is just as helpless against it. Minkler gives an interesting example of the futility of monitoring in a knowledge work situation: ‘Even if Millard, an entrepreneur who saw the potential market for personal computers but technically incompetent himself, could have costlessly observed the two engineers who developed the IMSAI 8080 microcomputer, it would have served no purpose. Because he did not have sufficient knowledge to direct the engineers, Millard could not decipher what they were doing, or equivalently if they were shirking, by watching them’ (1993a, p. 20). Sautet’s answer to this question is in many ways similar to mine: ‘I contend that one of the ways for a firm to grow is by becoming multidivisional . . . I see the main function of the M-form as solving the inner [Hayekian Knowledge Problem] (with which the general managers are confronted) and, at the same time, as overcoming the impossibility of planning (completely) the growth of the firm (in a diversified and geographically dispersed way) from the central office’ (2000, p. 131). My answer is more general than Sautet’s, in the sense that I explain decentralization of decision-making as the response to the double knowledge problem, but certainly the multidivisional form is an important way of decentralizing decision-making within a firm. In addition, I am interested in explaining the extreme cases of decentralization, whereas in an M-form divisions can still be centrally planned to a great extent. Nevertheless, Sautet’s work is certainly of great relevance to my own inquiries. Sautet makes the same claim: ‘With an increasing level of human capital, delegation of authority (team-based operations) becomes increasingly possible, and this delegation diminishes the inner Hayekian Knowledge Problem (provided new coordination problems are under control)’ (2000, p. 124). For example, Aoki (1990), Ricketts (2002). At least Schumpeter would not have been surprised by this development. He was aware of the failure of economics to realistically deal with business institutions, and called for more study by economists of business literature: ‘I would commend to economic historians – and, for that matter, to economic theorists, if they will interest themselves in the problem – that they examine the already available secondary literature for data upon entrepreneurial characteristics and phenomena. A miscellany of such writings – from general economic histories to biographies of businessmen, and from local histories to studies of technological change – all hold information, which sifted and arranged with definite hypotheses in mind will carry us a goodly distance toward our goal. New facts

90

18.

19.

20. 21.

22.

23.

24.

25.

Employees and entrepreneurship will doubtless be needed in the end, but already we have a multitude that have as yet not been digested’ (Schumpeter 1949, p. 271). Let us also not forget that Coase wrote his groundbreaking 1937 article only in response to spending months visiting different factories and interviewing their managers. Ultimately, economists must examine the meanings that managers and employees put on their own business structures, and the best way to do so at this point would be to better familiarize ourselves with business and management literature. We can also see a vague reference to a tradeoff existing between centralization and decentralization in the following passage by Langlois: ‘Indeed, decentralization is very much the imperative for any organization once it becomes successful and established . . . Once the innovation of mass production of parts became assimilated and disseminated, centralization becomes more costly and less beneficial’ (1994a, p. 19). It is informative to contrast Jensen and Meckling with a passage from the recent book The Science of Success by Charles Koch, founder of Koch Industries, Inc.: ‘We should also expect decision rights to change over time, as our businesses and our comparative advantages change and we make good or bad decisions. This is a dynamic process meant to ensure that those with the best combination of values, knowledge, motivation, demonstrated capability and opportunity cost are making the decisions’ (Koch 2007, p. 128). Also: ‘Those with local knowledge are often in a better position to solve the problem at hand. The ideas and creative energy of all employees should be leveraged . . . Decisions should be made by those with the best knowledge, taking comparative advantage into account’ (Koch 2007, p. 133). See also Rajan and Zingales (2001), Sliwka (2001), Zabojnik (2002), Dessein (2002), Rivkin and Siggelkow (2003), Colombo and Delmastro (2004), Mookherjee (2006), and De Paola and Scoppa (2006). This constant flux to which institutional structures are prone calls into question the applicability of the equilibrium concept to institutional structures. Foss defines an equilibrium state for institutions in the following way: ‘An organizational equilibrium obtains where decision rights are delegated in such a way that the benefits of delegation in terms of better utilizing local knowledge are balanced against the costs of delegation in terms of agency losses’ (Foss 2001b, p. 5). But in a world of constant change and imperfect knowledge, it would seem that such a notion of equilibrium is more or less irrelevant, especially given the fact that the ‘benefits of delegation in terms of better utilizing of local knowledge’ are just as impossible to quantify as are the ‘costs of delegation in terms of agency losses’. Foss would agree with this perspective on decentralized firms, as he admits that ‘playing market . . . may [be] broadly interpret[ed] as the introduction of pricing in the context of hierarchy’ (2001a, p. 17). Most firms featuring decentralized decision-making categorically do not do this. Foss (2001a) explains it in the following way: ‘Misesian arguments are used to criticize arguments derived from Hayekian insights that firms should emulate markets to the largest possible extent. In a sense, Misesian arguments resurrect the Austrian and Coasian notion that markets and hierarchies are indeed different mechanisms for resource allocation’ (p. 5). In Foss (2001a) he explains it in the following way: ‘[M]any of those who have addressed economic organization in the knowledge economy have explicitly drawn upon Austrian – more precisely, Hayekian – ideas on the need for decentralization fostered by the presence of dispersed knowledge. They have used such Austrian ideas to argue that hierarchy and planning methods are as problematic inside firms as they have proved to be outside firms, that firms need to harness the ability of markets to utilize, exchange and build information rapidly in response to changing contingencies, and that extensive delegation of decision rights and the use of high-powered incentives to support this are imperative’ (p. 4–5). Another version of this same statement can be found in Foss (2001a): ‘[B]ecause authority declines in importance as knowledge becomes distributed and knowledge inputs increase in importance, resort to other coordination mechanisms is necessary. Thus,

The knowledge problem in firms

26.

27.

28.

29. 30. 31. 32.

33.

91

firms increasingly rely on high-powered incentives, implement employee stockownership programs, invest in building “corporate cultures”, try to price corporate resources to the largest possible extent, and so on. An outcome of this is the emergence of “new organizational forms”. The theoretical implication is that various mechanisms for coordinating resources are combinable to a much larger extent than hitherto assumed in, for example, organizational economics, where economic activities are normally assumed to be organized across three discrete governance structures, firms, markets and hybrids (e.g. Williamson 1996)’ (p. 6). For example, see Foss (2001a): ‘economic organization in settings where rapid changes in the external environment necessitate a high degree of organizational decentralization and “empowerment” of employees, where relations to outside knowledge sources (other firms, universities, and so on) are paramount, and where “knowledge assets” account for a large (and increasing) part of value-added in production. While such settings have no doubt existed in some industries for a long time, they are not exactly the dominant mode of production that characterizes, say, American business history for a great part of the 20th century’ (p. 4, italics added). Foss defines the Misesian entrepreneur as one that engages in speculation, that is ultimately in charge of a business venture, that hires the managers and employees, makes the expansion and contraction decisions, and has ownership over a firm’s key assets. A case could be made that, when one defines the entrepreneur in this way, the argument that a firm needs an entrepreneur, in a sense of a person with ultimate authority, is somewhat tautological. Interestingly enough, Foss, though critical of radical decentralization in this article, admits that the spaghetti organization in Oticon ‘may indeed have caused a degree of innovativeness that might not have been obtainable in its absence’ (Foss 2001c, p. 4). He also says that, though the spaghetti organization may be an ‘inherently unstable administrative system, this system was in fact necessary to realize the benefits of increased innovativeness’ (ibid.). This is a very interesting quote, since it implies that even temporary decentralization, as long as it is at least partially successful, may be worth any instabilities that it will introduce. The instabilities can be fixed later, once the innovations have been created and are being exploited by the company. This perspective brings into question the necessity of the decentralized institutional structure being some sort of an ‘institutional equilibrium’. Of course there may be circumstances where it makes absolute sense for an entrepreneur to intervene in the delegated decision-making, especially if she determines that delegation and decentralization have outlived their usefulness. See also the following quote: ‘Misesian arguments help to demonstrate the continued viability of the “Coasian firm”, as against those critics who have argued that it will wither under the impact of the increasing prevalence of Hayekian settings’ (2001a, p. 19). Though I did show at the beginning of Chapter 1 that recent empirical studies have been providing support to the notion that decision-making within firms is increasingly decentralized and allocated to lower tiers of the managerial hierarchy. A striking example that demonstrates creative problem-solving which does not emanate from the top (that is, is designed) is given by Harper (1996): ‘The concept of moving production lines was not in fact solely [Henry] Ford’s brainchild. “It is clear that the impression given in Ford’s My Life and Work that the key ideas of mass production percolated from the top of the factory downward is erroneous; rather, seminal ideas moved from the bottom upwards” (Nevins & Hill 1954, p. 474). In fact, they credit Clarence Avery, a recent university graduate, as having played the largest single role in introducing the new production technique into Ford’ (p. 205). The benefits of decentralization can also be explained in terms of the discovery of opportunities for profit and new knowledge, which is what Sautet, building on the work of Israel Kirzner, does.

4. Spontaneous order in decentralized firms When business profits leap ahead, the entrepreneur will eventually even be ready to go beyond the yield attributable to labor and to let the workers share to some extent in the entrepreneurial profits, adding such share to their wages. A remarkable fact! It may be viewed as a symptom of the fact that between employers and employees, much as they may be at odds with one another, there yet exists fundamentally a farreaching community of interests. In today’s combative mood the existence of this community is not publicly admitted, but it nevertheless is at work, if tacitly. The whole circle of people engaged by the enterprise, from the top managers to the lowliest workers, is bound together by their common stake in the success of the business, and in the struggle with the customers and competitors it feels as a unit as a companionship of fate. (Wieser 1926, p. 354)1

The most important hitherto unexplained aspect of decentralization within firms is how the actions of multitudes of employees can ever be co-ordinated in the absence of a conscious, explicit command-and-control system. There are several questions to be answered in regard to this point: Can an intra-firm decentralization actually be an example of spontaneous order? What is the mechanism by which this order comes about? Is it possible to achieve spontaneous order if people are always acting in opportunistic ways? Even if people are willing to be cooperative, moral, fair, and selfless (to a degree, anyway), is it not necessary to have managers who will consciously co-ordinate the employees’ actions? In the absence of such coordinative force, won’t chaos rule? When it comes to answering these questions, it would appear that we are (almost) all Coasians now. Coase established a bold differentiation between firms and markets: a firm was always characterized by conscious co-ordination of resources because of the transaction-cost advantages of intrafirm organization compared to relying on markets. In markets, on the other hand, we observe spontaneous order, which is entirely absent within firms. I argued in Chapter 2 that this view is flawed, and even Coase himself stated decades after writing his 1937 article that he wished he had not stated his argument in such black-and-white terms. And, indeed, we have been seeing much evidence that spontaneous order can exist within firms, and there is today growing recognition of this fact by organization and management scholars, if not yet entirely by economists. 92

Spontaneous order in decentralized firms

93

In this chapter I will explain how spontaneous order can be generated within decentralized firms. I will first focus on explanations of employee motivational factors which are more realistic than the simplistic assumption of widespread opportunistic behavior on which much of modern Transaction Cost Economics is founded. It is important to see that employees have a variety of motivations that make co-ordination within firms easier to attain. So having a good explanation of employee motivations is important not only to explain how firms with decentralized decision-making can mitigate the problem of incentive alignment without relying on close monitoring by managers, but also to explain why and how employees voluntarily co-ordinate their actions with other employees to bring about intra-firm order. In the second part of this chapter, I will apply the theory of spontaneous order derived from Alfred Schutz and Richard Ebeling to the firm. When combined with an examination of employee motivations, this theory can successfully explain how and why decentralized decision-making within firms works.

WHAT MOTIVATES EMPLOYEES? Most economists accept that the principal–agent problem within firms can best be solved by employee monitoring. Because employees face lowpowered incentives within firms the most effective way to ensure their full effort is to establish a monitor (usually a residual claimant of some form) vested with the power to determine the employee’s wage or even their very employment based on the observed effort and results of the employee’s labor (Alchian and Demsetz 1972). But how realistic is this explanation? In the real world no employees are watched for the duration of their working hours – such a thing would be unimaginable, as well as unimaginably costly: it would require one monitor for each worker, and then of course monitors for the monitors. Fortunately, in most modern workplaces successful outcomes are achieved without having to resort to anything approaching such extreme measures. Economists must start with the right motivational premises if we hope to explain this reality. As we will see, those include not only traditionally considered positive and negative incentives, but also factors such as peer pressure, work ethic, morals, identification with the organization, and satisfaction from doing a good job. Simon on Motivational Factors The most significant work on employee motivation was done by Simon (1991). He asks the following questions: ‘[How are] the employees of

94

Employees and entrepreneurship

firms . . . motivated to work for the maximization of the firm’s profit[?] . . . Why do employees often work hard?’ (1991, p. 26). The usual answer is that the hierarchical firm governance structures provide the proper incentive alignment, wherein the owners and managers of the firm retain the power to give commands and therefore get the employees to act in ways desired by their superiors. However, that leads Simon to ask two further questions: ‘how are employees induced to work more than minimally, and perhaps even with initiative and enthusiasm? Why should employees attempt to maximize the profits of their firms when making decisions that are delegated to them?’ (1991, p. 26). In other words, how do firms avoid the principal–agent problem which can lead to widespread shirking? The theory of the firm along transaction cost lines holds that monitoring of employees plays the key role here. It is through monitoring of employee performance that employers can decide on the appropriate rewards or punishment or the appropriate place in the hierarchical structure of the firm. Shirking can be restrained only through a well-designed system of negative and positive incentives, enforced through managerial monitoring. The primary role for managers is to monitor employees and report on their work effort and productivity. Simon does not accept this explanation, accusing it of presenting ‘an unrealistic utility function’ of employees which ‘does not provide a basis for understanding real organizations’ (1991, p. 30).2 He thus sets out to derive a more detailed, more realistic, and empirically richer utility function that is built upon the factors that ‘motivate real people in real organizations’, and he relies on organizational literature to do so. More specifically, he focuses on three major motivational factors within firms: rewards, authority, and identification. Rewards consist of positive incentives such as bonuses, wage or salary increase, promotion, or even recognition for doing good work. Simon does not deny that rewards can motivate employees, but he points out that rewards will only be effective when the employees’ contribution to the overall firm performance can be accurately measured and assigned. Without this ability, either because ‘the indices used to measure outcomes are inappropriate, [or] because they do not measure the right variables, or because they do not properly identify individual contributions’ (1991, p. 33), positive incentives can have a distorting or even counterproductive effect as free-riding becomes rampant or as workers become discouraged or even resentful. But in firms almost all activities are greatly interdependent (this interdependence is at the very heart of the advantageousness of firm organization) and therefore it is very difficult if not impossible to measure the separate contributions of each worker. Thus, it becomes equally difficult or impossible to assign proper rewards to deserving

Spontaneous order in decentralized firms

95

employees.3 Simon’s bottom line on rewards is that, although they ‘play an important part in securing adherence to organizational goals and management authority, they are limited in their effectiveness. Organizations would be far less effective systems than they actually are if such rewards were the only means, or even the principal means, of motivation available’ (1991, p. 34). The second motivational factor Simon discusses is authority, which refers to giving power (authority) to employees to make decisions on how best to achieve some end goal. It is common for a command simply to present the end goal, rather than specific actions necessary to achieve it.4 It is up to the employee to figure out how best to reach the goal. As Simon puts it: employees, especially but not exclusively at managerial and executive levels, are responsible not only for evaluating alternatives and choosing among them but also for recognizing the need for decisions, putting them on the agenda, and seeing to the generation of possible actions. Doing the job well is not mainly a matter of responding to commands, but is much more a matter of taking initiative to advance organizational objectives. (1991, p. 32)

Simon, without explicitly identifying it as such, points to the benefits of delegation on the basis of the knowledge problem: if authority is used to spur the employees to engage in decision-making and problem-solving, rather than simply carrying out detailed orders, then ‘many different experts can contribute their knowledge to a single decision’ (ibid.). This contribution can most effectively happen if people are given authority to make decisions, even if that authority will often be within certain guidelines only. So authority can be used to motivate employees to the extent that it’s actually given up by the owners or the managers and delegated to the employees themselves. Having authority will make the employees more likely to assume responsibility and Simon notes that ‘willingness of employees at all levels to assume responsibility for producing results – not simply “following the rules” – is generally believed to be a major determinant of organizational success’ (1991, p. 37). Ultimately employers want their employees to ‘take initiative and apply all their skill and knowledge to advance the achievement of the organization’s objectives’ (1991, p. 32). But we still observe a puzzle: In most organizations, employees contribute much more to goal achievement than the minimum that could be extracted from them by supervisory enforcement of the (vague) terms of the employment contract. Why do employees not substitute leisure for work more consistently than they do? Why do they often work so vigorously for the welfare of the organization? (Ibid.)

96

Employees and entrepreneurship

This brings us to the very heart of Simon’s explanation of employee motivation: employee loyalty. He makes a case that loyalty plays a more significant role in employee motivation than generally acknowledged by economists. It is no great insight to recognize that free-riding is common, but the more interesting puzzle is, why is there anything besides free-riding? It is obvious that many if not most workers exert more (and in some cases much more) than the minimal effort they could get away with (in order to simply retain their employment). We all know people who work very hard in the absence of any monitoring or direct rewards for such behavior. Strict Transaction Cost Economics would have to proclaim such behavior irrational, though the future of many businesses depends on their employees behaving in such seemingly irrational ways. So what explains this ‘irrational’ behavior? Simon claims that it is driven by identification of employees with organizational goals. Employees often exhibit pride in their work, a phenomenon evident among skilled and managerial employees more than employees engaged in routine work. One could say that this can be explained by noting the connection between the success of the firm and the personal benefits that accrue to the employees as a result. But this neglects to take into account the very real fact that each employee contributes only a small amount to the overall success of the firm, and therefore free-riding would still seem to be the dominant strategy for all employees. The best explanation of the phenomenon of organizational loyalty can be derived by recognizing that humans often exhibit what we will call, following Adam Smith’s precedent, empathy. Under proper conditions they are able to identify with other people, and consider (to various degrees) their well-being when making decisions. This quality allows people to exhibit some, or even substantive, responsiveness to social influence, creating a situation where people are more likely to engage in behavior considered desirable by a society. If the relevant society is a firm, we may find employees who exhibit organizational pride and loyalty. They will be able to gain satisfaction from the firm success that transcends the pecuniary benefits accruing to them directly. So, as Simon states, ‘organizational identification becomes a motivation for employees to work actively for organizational goals. Of course, identification is not an exclusive source of motivation; it exists side by side with material rewards and enforcement mechanisms that are part of the employment contract’ (1991, p. 36). I should also add that organizational identification does not happen easily and automatically. Employees will only develop it if they feel that their employer – as well as the other employees – is fair and honest, that the work the firm is engaged in is important, and that they are given work that is

Spontaneous order in decentralized firms

97

interesting and stimulating. These conditions cannot be met easily, but they are met at least sometimes if not often. So now we are seeing a more complete picture. Unlike a simplified and unrealistic picture painted by Transaction Cost Economics, the question of employee motivation is a complex one of not only incentives, but also the existence of the authority for employees to make their own decisions as well as organizational loyalty. The latter point is the one emphasized by Simon himself as a way to solve all sorts of potential problems in firms – for example, organizational loyalty can eliminate employee attachment to their own sub-goals and it can decrease the likelihood of employees producing substandard products. We can say that, in general, organizational identification is responsible for mitigating the problems of moral hazard and opportunism. One could easily accuse Simon of being a bit too optimistic about the positive influence exerted by organizational loyalty. But Simon concedes that firms never fully rely on it, nor should they attempt to. Rather, all firms employ a wide variety of mechanisms to properly motivate their employees. Simon should be recognized for bringing to our attention factors that we ignore at our own peril. Any explanation of the internal structure of a firm is at best highly incomplete, and at worst highly inaccurate, if it neglects to account for organizational identification and authority as important motivational factors. Simon was himself motivated by a desire to show that large, hierarchical enterprises are not necessarily stiflingly bureaucratic, but can be very productive.5 However, I think that we can draw another conclusion from his arguments: it is possible to give employees quite a bit of independence without hopelessly handicapping firm operations. Firms may actually run even more successfully than if they relied on close monitoring and control. This is an important conclusion. Minkler on Motivational Factors In a 2002 article, Minkler examined the employee motivational factors in firms in more detail. He set up an extensive national survey to better understand what actually motivates workers to work hard. His findings were broadly consistent with Simon’s insights. Minkler found employees in the great majority of the cases worked hard, even when it was impossible for their employers to monitor them. Minkler came to the conclusion that an assortment of factors influences worker motivations (listed in the order of their empirical importance): moral factors, ‘intrinsic’ factors, ‘peer-pressure’ factors, and finally ‘positive’ incentives (in the form of wage increases and promotions).

98

Employees and entrepreneurship

As mentioned several times already, the theory of the firm, as it has evolved over the last 30 years, has been mostly predicated upon one fundamental assumption: workers will shirk in the absence of sufficient incentives. The fundamental question for this transaction cost theory of the firm was ‘would workers shirk if the firm did not know how hard each was working’ (2002, p. 2). As Minkler found that the answer was overwhelmingly ‘no’, he issued a challenge to Transaction Cost Economics which is all about the firm creating a proper alignment of incentives. Like Simon, Minkler certainly had no wish to dismiss the impact of negative and positive incentives: ‘While it would be foolish to suggest that incentives do not matter, the point is that after a consensual agreement has been struck, other motivations kick in that may be even more important’ (2002, p. 19). So, once an employee is in a working situation, the question of remuneration becomes of secondary importance to his effort. Minkler even suggests that aligning incentives may be a ‘second-order problem’. Minkler’s results are largely supported by the experimental economics literature,6 where fairness and morality rather than basic instrumental rationality played an important role to get people to work. This literature also finds that simple communication between employees very frequently results in cooperation, and, more surprisingly, that unenforceable promises do matter. Minkler examined several different motivational factors in his study, some of which echo Simon’s, but most of which were not considered by Simon: The first factor is the existence of straightforward incentives, whether negative (monitoring and the associated threat of firing, demotion, or a cut in pay) or positive (wage increases and promotions). The second factor is what he calls intrinsic motivation, which can be thought of as an innate work ethic. A work ethic leads people to engage in hard work for its own sake. Minkler suggests, based on the work of Frey (1997), that intrinsic work motivation may result from work that is ‘interesting, involves the trust and loyalty of personal (as opposed to anonymous) relationships, and is participatory’ (2002, p. 5). As a counterpoint, intrinsic motivation may be reduced by the existence of monitoring or even different positive and negative incentive schemes. Minkler reports Frey’s finding that: [c]ommands are most controlling in the sense that they wrest self-determination from the agent, whereas rewards might still permit autonomy of action. Promotions believed to be an acknowledgement of general competence may even increase intrinsic motivation. But if the reward is closely linked to the performance set by the principal and contingent on specific performance, it can be seen as controlling with a resulting decrease in intrinsic motivation. (2002, p. 6)

Spontaneous order in decentralized firms

99

Obviously, this is a complex issue that seems to be very much tied up with an employee’s personal interpretation of the employer’s valuation of and trust in that employee. It would appear that employees are more likely to have an intrinsic drive to work hard in a situation where they feel their contribution is respected, and they are trusted to make the right decisions. We can see some connections here to Simon’s discussions of authority and delegation of decision-making power. The third factor is peer-pressure, which Minkler defines as a situation where an employee is concerned with how he is perceived by other workers. This is most important in interdependent work situations, especially where rewards are contingent on a team’s success: in those cases the presence of other employees can effectively lead to prevention of shirking through the employee’s feelings of guilt (due to the existence of some sort of loyalty or team spirit) or the imposition of active social sanctions, such as shame. In fact, firms often attempt to create and strengthen social connections between employees through fostering activities such as company sport teams, picnics, Christmas parties, and so on. These connections will create stronger feelings of empathy between employees, and guilt may play a more important part in preventing shirking. This factor would seem to loosely correspond to Simon’s organizational identification, though Minkler’s peer-pressure factor would appear to be more directly connected to employees’ self-interest than organizational identification. The fourth factor is moral motivation or personal ethics, which can result from either religious feelings or feelings of some higher sense of duty, and which are fostered through the institutions of family, education, and religion. Often moral considerations lead people to act in ways that are contrary to their narrow self-interest. Minkler follows the recent research in this area, which explains that moral actions are undertaken in order to preserve one’s ‘self-identity’. Shirking is prevented in those cases by simply relying on people’s innate moral principles. Finally, the fifth factor is fairness, based on reciprocity. It is a version of a tit-for-tat game, where the proper response to another actor is predicated upon that actor’s behavior. So, within the context of a firm, employees will act honestly if they perceive the employer to be honest. They will not shirk if they think their employer has been acting in an honorable way towards them. The survey found the following:7 1) almost 95 per cent of the respondents are likely to work hard even if it is impossible for their employer to check up on them; 2) they are strongly motivated by morals, followed by intrinsic factors, peer-pressure, and positive incentives. The results are supportive of Simon’s broad explanations of worker motivation. Minkler sums up the findings in the following way:

100

Employees and entrepreneurship

Most fundamentally, incentive mechanisms may not be necessary to assure the provision of effort. According to the survey instrument used in this study, moral and intrinsic motivations are most important to workers. In order for firms to benefit from both motives, they could promise to provide interesting work and a fair employment package and then keep those promises. If firms cannot organize work in such a way that it is interesting, they might still be able to elicit effort by acting honestly and by promoting mutual commitment. If, instead, mutual commitment is not possible because of changing demographics, high turnover, or demand uncertainty, firms might still elicit effort by providing interesting work. (2002, p. 19)

Other economists working in the field of the theory of the firm have observed behavior along the lines examined by Simon and Minkler. For example, Langlois noted that: in many situations, people follow norms of behavior – like honesty – even in endgame situations. One often tells the truth when lying would be costless and privately beneficial. The reason is that norms of this sort are often internalized to form a part of culture. They are, in effect, instances of the tacit rule following . . . the norm is itself an enforcement mechanism. (1992, p. 173)8

Langlois’s insights provide support for Minkler’s notions of a work ethic and moral ethic. There is an additional motivational factor which neither Simon nor Minkler mention: the actual enjoyment of work. If people are given interesting jobs that are a good fit for their personality and skills, they may in fact work hard and creatively for the simple pleasure of doing so, rather than because of peer-pressure, a work ethic or some higher moral ethic. Employees can in effect sometimes partly become consumers of the process of production. We certainly observe this type of behavior in different walks of life. It is often quite obvious in the domain of the arts, where an author or a painter or a musician works for the pure creative joy of the process. We also can observe the same situation with software writers, teachers, political party volunteers, and so on. There are many cases where people engage in particular work outside of their real career (hobbies), often for nominal (if any) money, deriving great pleasure from this work.9 This could explain the near-obsession among firms with a great amount of decentralized decision-making with finding the right job fit for their employees. If the firm can provide the employees with interesting work they are much more likely to work hard and in a creative manner.10 An effective example of such a fit was Art Fry at 3M. Fry almost singlehandedly created the omnipresent Post-it notes. Though his formal job title at the company was product designer, he not only came up with the idea for Post-it notes, but also worked relentlessly to find the proper adhesive for the

Spontaneous order in decentralized firms

101

notes, design the production process, persuade the management of the potential profitability of the project, and finally create the right marketing campaign. These were all jobs far outside of his formal duties, but 3M’s internal structure was designed to be decentralized enough for Fry to have plenty of scope to act on his strong intrinsic motivation. 3M’s chairman, Lew Lehr, explained it this way in 1983: For many years the corporate structure [at 3M] has been designed specifically to encourage young entrepreneurs to take an idea and run with it. If they succeed, they can and do find themselves running their own business under the 3M umbrella. The entrepreneurial approach is not a sideline at 3M. It is the heart of our design for growth. (Pinchot 1985, p. 5)

This organizational structure allowed Fry to create one of the most successful 3M products ever. Fry stumbled upon the idea for adhesive notes outside of work, but decided to try to turn the idea into reality at 3M. He took advantage of ‘a 3M policy that gives technical people 15 percent of their time to work quietly on ideas of their own’ (ibid., p. 138), otherwise known as ‘corporate slack’. Fry made use of the adhesive that was independently developed by another part of the company, which had distributed the samples of the adhesive ‘in the hope that someone would find a use for it’ (ibid.). Once he had developed the rough product, he presented it to manufacturing. Their response was that it would be impossible to produce. Most people in most companies would have given up at this point. Fry instead began using the machines during their off-hours to develop the manufacturing process. He sometimes went to extremes: ‘When he got official permission to use a pilot plant, he worked five consecutive eighthour shifts, or forty hours, without stopping’ (ibid., p. 139). Finally, when he had designed the machine that would produce the notes, the manufacturing engineer told him that it would ‘take six months to build and cost a small fortune’ (ibid.). Fry’s response was typical: ‘The next morning when people came to work they found Art’s new process up and running. He had built a crude version of the machine overnight in his basement, brought it to work, and installed it. It was working’ (ibid.). Fry continued this level of dedication to the product until it was a giant success for 3M. The truly surprising element in this story is that Fry, a person clearly showing great entrepreneurial abilities,11 would choose to stay employed with 3M rather than start his own company. But, though he did not capture the full profits, there were obvious benefits to innovating within 3M: ‘Even with a good dose of venture capital, Art would not have been able to create Post-it notes outside of the large multifaceted corporate environment’ (ibid., p. 137). The choice was not between keeping the profit for himself, working within his own firm, and sharing the profit with others,

102

Employees and entrepreneurship

less deserving, by working within 3M; rather, it was between being able and not being able to create something like Post-it notes at all. It should be emphasized that I do not mean to suggest that people will labor for no remuneration. They must receive some pecuniary incentives. But these other motivational factors must be considered in addition to the pecuniary incentives, which have traditionally been the only factor considered by economists. These are important conclusions for firms featuring decentralized decision-making, since, as I pointed out above, many economists are dismissive of such firms on the grounds that they will be unable to solve the rampant employee free-rider problem once employees are given so much freedom on the job. But here we see that employees are much less likely to engage in free-riding behavior than traditional theory would predict. In addition, many firms with decentralized decision-making rely on some form of teamwork to organize production. This is consistent with the above findings, since the teamwork organization will provide two benefits: 1) create significant empathy among employees, leading to strong feelings of guilt preventing shirking; 2) create a system of informal monitoring which will diminish the likelihood of shirking through the imposition of shame on shirkers. The fact that some real-life firms do not adhere to their decentralization experiments for very long periods of time could be explained on motivational grounds. Decentralized firms that rely too much on honesty (that is, an individual work and moral ethic) and not enough on the provision of positive and negative incentives, peer-pressure, and organizational loyalty to avoid free-riding behavior have not properly designed their internal institutional structure. Such a firm may be able to operate (even successfully) for some period of time, but, in the face of a few setbacks or discouraging factors, employees may revert to narrowly selfinterested behavior. However, even in such cases, a decentralized structure can bring about many beneficial consequences while it lasts (in other words, as long as individuals are not acting in narrowly self-interested ways). Firms with decentralized decision-making can produce a very high level of experimentation and accordingly a high level of innovation, whether in the form of new products or new processes. Thus, many owners or managers may view decentralization as a matter of a tradeoff between lesser organizational stability and greater innovative capabilities. They may understand that introduction of different forms of decentralization may make the firm more unstable and they will very likely eventually have to revert to a more controlling structure. But these organizational adjustments may be worth the gains in innovations in the short run, if not in the long run.

Spontaneous order in decentralized firms

103

SPONTANEOUS ORDER WITHIN DECENTRALIZED FIRMS If we can accept that employees do not usually act in the ways presumed by Transaction Cost Economics and that they are often willing to engage in something approaching consummate cooperation, then next we must explain how cooperation can come about within firms that have decentralized decision-making. All firms must have some – if not great – internal coordinative forces. Within regular, hierarchical firms co-ordination of course is achieved through conscious command-and-control methods. But within firms with decentralized decision-making we observe co-ordination which is not fully, or even mostly, a result of conscious command-and-control, and which can only be explained by recourse to the theory of spontaneous order. But that certainly does not mean that we thus need to rely on the market theory to explain co-ordination with firms exhibiting some decentralization. As I pointed out previously, firms with decentralized decisionmaking do not emulate markets, and thus it is incorrect to attempt to explain their internal co-ordinative forces by relying on the institutions that generate spontaneous order within market institutions, most fundamentally consisting of property rights and monetary prices. We need to take a more subtle view of the forces shaping spontaneous orders in general. I believe those forces can be explained by the groundbreaking work of sociologist and philosopher Alfred Schutz. Schutz was greatly interested in the issue of co-ordination of human actions. Ebeling (1995a)12 places Schutz’s fundamental social problem in the context of the economic world: ‘How can a multitude of individuals participating in a complex and world-encompassing division of labor successfully coordinate and adjust their actions and minimize frustration and disappointment, when each participant possesses different and everchanging knowledge and expectations about the possibilities of the future?’ (p. 81). A simpler way to ask this question is: ‘How is a society possible?’ An equivalent question for the purposes of this book would be: How is a firm with a great amount of decentralized decision-making possible? 13 At the very bottom of this problem is the existence of a great amount of interdependencies among the people in any relevant society, whether an economy or a firm. When other people’s actions have a great amount of bearing on one’s own chances of success, there must be a method by which those actions are somehow co-ordinated. This co-ordination problem is even more relevant in a firm with decentralized decision-making, since interdependencies and externalities (both positive and negative) have greater importance here than in a large society, in the sense that they will have a larger effect on the chances of success of each individual. How can

104

Employees and entrepreneurship

employees correctly anticipate the actions of their colleagues that are relevant to their own decision-making, as well as signal to others what they themselves will be doing? That is the crux of the problem. To Schutz, ‘social action . . . was human action in which the actors incorporated the possible actions of others into their own plans and modes of conduct; it consisted of intricate webs of “mutual orientation” (Ebeling 1995a, p. 84). That mutual orientation is possible through the existence of a shared set of meanings that social actors put on their own actions as well as the actions of other individuals. They are shared in the sense that they are not private meanings, unique to a particular individual. In a wider social order it is our parents, friends, religious institutions, media, or own experience that teach us ‘to interpret the meanings embedded in the multitude of social relationships’ (ibid.). These meanings become ‘ideal types’14 which are subject to constant reconsideration and flux. But, as impermanent as they may be, ‘typifications emerge out of social interaction and become formalized into structures of intersubjective meaning’ (ibid., italics in the original), or meanings which we all share in common. These structures of intersubjective meaning allow us to understand other people’s gestures, their inflections of words, purposes of different products, and uses of resources. In turn, they bring about greater social coordination. The structures themselves are not objective because they are neither permanent nor uniform in meaning from person to person; rather, they constantly change, and we all may have slightly different interpretations of their meaning. In addition, they are usually learned through experience rather than any formal method. Nevertheless, we each individually act as though they are objective to us in the here and now. They are tacit or informal institutions of the web of social relationships. As Ebeling eloquently explains it, ‘for the human actor in the social arena of everyday life, it is the structure of intersubjective meanings, as captured in ideal typifications, that incorporates and envelops the “reality” of mundane action’ (1995a, p. 87). The structures of intersubjective meaning evolve as actors assign new meanings to actions and objects, owing to new thoughts, ideas, and experiences. Since the structures are necessarily informal, they are continuously modified by different individuals. Some modifications are rejected by the society (that is, others do not adopt the same meanings, and therefore they remain unique rather than shared), but those that are accepted are institutionalized through repeated actions based on that meaning. It is an evolutionary process of meaning selection that Ebeling sums up in the following way: ‘Thus a dynamic process of mutual dependency emerges in which the “given” meaning structures serve as points of individual and social orientation for “understanding,” while being themselves modifiable through

Spontaneous order in decentralized firms

105

their use for the expression of new meanings by individuals’ (1999, p. 128). The meaning structures are flexible and do not act as constraints, but rather as guides for co-ordinating our actions with those of others. Expectations and Co-ordination The structures of intersubjective meaning allow us to build a relatively reliable model of expectations of the actions and reactions of others, one which is applicable in a wide variety of circumstances, even relatively unique ones. This is done through the typification process, which can reveal ‘patterns of regularity or “types” of response for prediction of unique cases . . . No matter how imperfect, it introduces an additional source of knowledge for coordination of plans in the complex social setting of the market’ (Ebeling 1994, pp. 92–3). The structures of intersubjective meaning will also lead us to expect that others will have the same ability to interpret and understand our own actions. This co-ordination of expectations will allow us to form plans with a great deal of confidence that we will be able to successfully fulfill them, even when these plans are heavily dependent upon the actions of others. We can expect that others will be able to interpret our intentions and requests in ways that we desire. As Ebeling again explained it, ‘the routinization of behavior along typical patterns introduces ranges of knowability about the possible future conduct and motivations of others. It is what makes society and economies possible in lieu of a “perfect knowledge” of each separate individual and his or her unique eccentricities and differences’ (Ebeling 1995b, p. 146, italics in the original). The co-ordination process is therefore predicated upon those shared meaning structures. When our expectations of the actions of others are largely met, it is because we have the same intersubjective understanding of our roles in a particular situation or relationship. As Ebeling points out, these structures of intersubjective meaning seem to be exactly what Ludwig Lachmann (1971) referred to as the informal ‘nodal points’ of mutual orientation and plan co-ordination that accrue ‘among the interstices’ of the structure of formal rules (laws) in society. It is a powerful but sadly neglected account of the forces of spontaneous order in a society. Structures of Intersubjective Meaning within Firms Though neither Schutz nor Ebeling applied this construct to firms, I believe that its import is natural.15 It explains very effectively how individuals within firms are able to co-ordinate their complex activities and get things done in the absence of explicit commands. It is important to recognize that even the most closely controlled firms rely somewhat upon the ‘common

106

Employees and entrepreneurship

sense’ of their employees rather than the spelling out of every detail of the productive process, but that common sense is ultimately always predicated upon a structure of intersubjective meaning present in that firm. There is one major difference between a society and a firm: in a society many of our relationships are by necessity anonymous; in contrast, within firms most relationships are better described as being ‘face to face’, ‘in which the participating actors form typifications not of all men or some group of men, but of the particular other with whom one is interacting’ (Ebeling 1995a, p. 85), ‘on the basis of his typical attitudes, motivations, and responses as they have come to be seen and analyzed in direct interpersonal contact’ (Ebeling 1995b, p. 146). It is for this reason that structures of intersubjective meaning within firms can have a more powerful co-ordinating effect than in a large society. Employees can engage in a more concrete typification process based on close, repeated contact with others with whom they are co-ordinating their activities. This is an added benefit of organizing most work within firms as teamwork. The benefits of teamwork do not consist only of influencing the motivations of employees, but also of allowing all team members to get to know each other better – and therefore form more accurate typifications of each other. In addition, the existence of a particular ‘corporate culture’ can make the typification process even more effective by emphasizing what sort of actions and attitudes can be expected by employees and managers. In other words, the corporate culture can greatly facilitate mutual comprehension of meanings.16 When firms are viewed from this perspective, the role of managers is seen very differently than in the transaction cost literature. Managers become facilitators of co-ordination, rather than monitors or planners. Their primary task may be the establishment or simple fostering of a particular culture within a firm. Herbert Simon pointed out this aspect as well: ‘seeing that commands are obeyed is not simply a matter of observing behavior, but of affecting the thought processes and the decision premises of employees’17 (1991, p. 32). This can be effectively done through the establishment of an organizational culture, imparting a structure of intersubjective meaning on the firm. 3M is a famously innovative company, and it has come up with just such a culture, consisting of a compelling narrative which promotes the entrepreneurial self-image among the employees. In the process 3M clearly presented the kind of approach towards work that it expects of its employees: The legends of 3M almost all speak of the dogged dedication of the intrapreneur. When his bosses told intrapreneur Phil Palmquist to stop working on reflective coatings because that wasn’t his job, he continued four nights a week from 7:00 P.M. to 11:00 P.M. Soon he had a product 100 times brighter than white paint. Among other things, it now lights up roadway signs at night when

Spontaneous order in decentralized firms

107

your headlights shine on them. In a more extreme case, George Swenson, another 3M intrapreneur, was fired when he wouldn’t stop working on a new roofing material. He continued working on the project despite the fact that he was no longer employed. Once he had it working, the company relented and rehired him. By treasuring such stories, 3M encourages others to try to innovate despite opposition. (Pinchot 1985, pp. 46–7)

Managers in decentralized firms must act as co-ordinating conduits by facilitating a strong organizational culture.18 Another example of a manager attempting to establish just such an organizational culture can be seen in the development of Pontiac Fiero in the late 1970s. This project was headed by Hulki Aldikacti. He was convinced that a sporty two-seater car would be very successful for GM, but he had a hard time selling the concept to the headquarters. So he forced a form of decentralization through to GM: as he put it, he decided to ‘go outside and set up shop and run it like a small business’ (Pinchot 1985, p. 78). Though he didn’t have the full support of the GM headquarters, the heads of the Pontiac division were behind him. They also understood that ‘the Pontiac organization as it then existed was too ponderous for the job’ (ibid.), so they let him bend the rules while they protected him in his unofficial intra-firm horizontal disintegration. Aldikacti rented a space some ten miles away from the Pontiac office to get away from the bureaucracy. Also, he was aware that they had a very limited time to complete the project before the headquarters put a stop to it. Moving away from the main office allowed his cross-functional team to easily communicate and in the process develop a common culture, greatly speeding up the development: One of the great barriers to speed in innovation is slow communication. To avoid that problem, Hulki [Aldikacti] brought all of his people together in a small building away from the noise of bureaucratic interference. Formality slows communication too, so Hulki insisted on direct communication. Only with a small team can members of the group know and trust each other well enough to dispense with formality. (ibid., p. 82)

In this case, the manager very clearly acted as a facilitator of the development of an informal culture of communication that established a structure of shared meanings. The result was that the team was able to set a new record in the length of the car development period: ‘The Fiero project cut almost two years and many dollars out of the normal timetable for new car development. GM top management considered this system so significant that they are trying to recreate it’ (Ibid., p. 80). Another important (and related) role for managers is to judge human character: managers must be able to determine whether a particular person has the right characteristics (innate or formed through experience) to fit

108

Employees and entrepreneurship

into the organizational structure. Once the employees with the ‘correct’ attitudes and ambitions are selected, firms can rely on organizational identification, job satisfaction, morals, peer pressure, and so on, to get employees to exert maximum effort. Thus the employee selection process is of a greater importance in firms with decentralized decision-making than within orthodox hierarchical firms. We see this employee selection process at play in W.L. Gore & Associates, a company with an extremely flat organizational structure and a great degree of decision-making decentralization, as described in Chapter 1. At Gore there are no managers, and the hiring process is carried out mostly by the employees themselves. But the hiring process is very methodical and detail-oriented in order to identify those employees who hold the appropriate outlook on work and fit within the company’s relatively unique organizational culture: Job applicants at Gore were initially screened by personnel specialists, who contacted as many as 10 references for each applicant. Each candidate who passed this screening was then interviewed by Associates working in the area of the company where the candidate was being considered for a position. According to those who had gone through them, the interviews were rigorous. Before a candidate was hired, an Associate had to agree to be his or her sponsor. The sponsor’s role was to take a personal interest in the new Associate’s contributions, problems, and goals, acting as both a coach and an advocate. (Shipper and Manz 1998, p. C-503)

Ultimately not all employees will be cut out for this way of working. Some employees may find a decentralized corporate culture stifling, despite others finding it free or even nurturing. In order to create an effective organizational culture the structures of intersubjective meanings must be voluntarily shared by all, and it is very important that only those employees who are willing and able to do so are employed. We see this dynamic in action at Gore: Not all Gore Associates functioned well in Gore’s unstructured work environment, especially initially. Those who had worked at other companies and become accustomed to a more structured work environment usually encountered adjustment problems. As [founder] Bill Gore said, ‘All our lives most of us have been told what to do, and some people don’t know how to respond when asked to do something – and have the very real option of saying no – on their job. It’s the new Associate’s responsibility to find out what he or she can do for the good of the operation.’ A few Associates concluded that Gore’s flexible, unstructured workplace was not for them and soon left the company . . . However, the vast majority of new Associates, after some initial floundering, adapted quickly. Overall W.L. Gore’s lattice organization proved to be good for the company’s bottom line. The year before he died, Bill Gore estimated that the company’s profit per Associate was double Du Pont’s profit per employee. (Ibid., p. C-500)

Spontaneous order in decentralized firms

109

Another example of the importance of the employee selection process can be seen in the following segment of an interview with Eric Schmidt, CEO of Google Inc.: Wall Street Journal: How do you choose people to work at your company? Mr Schmidt: The principle that Google operates under is to hire very, very strongwilled, sort of driven persons. We have relatively little management and the management is very, very thoroughly vetted. They both have the intelligence and the history of working in high-tech and they want to work, they want to change the world. We always talk at Google about how brilliant the engineering teams are, which is indeed true. It’s just as important to have corresponding managers or leaders who have the strategic understanding of what we’re trying to do because it changes every nine months. (Interview by Mylene Mangalindan, Wall Street Journal, March 29, 2004, p. B1)

Several interesting points emerge here: first, Google can operate with little management because they hire people with high intrinsic motivation and great love of their work; second, they hire people who have worked in this industry and most likely already have similar structures of shared meanings, facilitating intra-firm co-ordination; finally, we also see that having few management layers creates a situation where all managers can have a fuller strategic picture which allows great flexibility, necessary when the plans, as Mr Schmidt put it, are changing every nine months. So, to sum up, an organizational culture should accomplish two purposes: enable greater co-ordination through structures of shared meaning and encourage creative, hard work by creating a feeling of a common cause among employees. I have already talked about the decentralization developments in the Danish firm Oticon, and we can interpret CEO Lars Kolind’s instituting of a new organizational culture befitting a radically decentralized firm as aiming to accomplish both of these, as Foss notes: ‘[Kolind’s] attempt to infuse the organization with a strong set of shared values may also be seen as an attempt to assist the coordination of multiple efforts in a decentralized setting while simultaneously keeping agency problems at bay’ (Foss 2001c, p. 16).19

CONCLUSION There are forces at work that even in the absence of conscious co-ordination by managers can bring about a great amount of co-ordination within firms. This explains why we have been seeing a major push towards eliminating middle-management positions in many firms (the so-called process of ‘delayering’). It is a result of greater decentralization of decision-making powers within firms, which inevitably required a redefinition of the managerial roles.

110

Employees and entrepreneurship

Many firms (especially in the high-tech sector) have created an internal organization where the managers are expected to encourage the creativity of their employees while the employees are expected to freely cooperate with others within an open job framework (one where their duties are continuously changing, often on their own initiative). Though sometimes seeming chaotic, such firms appear to be able to effectively co-ordinate the actions of multitudes of individuals within them under at least some circumstances. I believe that they do so because they have been able to create structures of intersubjective meaning manifested as an organizational culture in which, first, their employees do not act in opportunistic ways, but are rather motivated by a variety of factors, such as morals, peer-pressure, intrinsic factors (that is, a work ethic), pride in a job well done, and positive incentives in the form of profit-sharing or stock options, and second, these structures create sufficient compatibility of actions among employees. In addition, it is worth noting that, though the intra-firm behavioral rules may be few in number in firms with a great deal of decentralized decision-making, the employees are also governed by wider social rules, such as, for example, courtesy and honesty, which also contribute to making order and cooperation possible. If at this point stable firms with decentralized decision-making are not very common, a possible reason for it may be that the business world and the society know little about the institutional structures that will properly establish the right motivational factors and create an effective structure of shared meanings. This might also be the reason for the seemingly transient nature of some firms with great degrees of decentralized decision-making. My conjecture is that we are in the early part of the evolutionary development of decentralized decision-making, and much more needs to be discovered and understood to fully take advantage of this institutional form. However, I certainly do not mean to give the impression that decentralization is some sort of an organizational panacea. There are many circumstances where it would be inappropriate and its ultimate effect nothing but institutional destabilization and a great amount of internal discoordination. As Ebeling pointed out above, the structures of intersubjective meanings are not perfect as co-ordinating forces, and it is possible that they will be insufficient to achieve necessary co-ordination in some or even many situations. In fact, I believe it is safe to say that relying on structures of intersubjective meaning will never achieve the level of co-ordination that can be achieved through conscious co-ordination. Also, despite the existence of all the other motivational factors, employees may nevertheless decide to act in narrowly opportunistic ways. Yet we see – and will continue to see – firms still engaging in decentralization of decision-making because sometimes the potential benefits of decentralization, such as the greater employee

Spontaneous order in decentralized firms

111

creativity, innovation, and entrepreneurship, are simply too attractive and trump the potentially considerable downsides.

NOTES 1.

Contrast Wieser’s quote with a remarkably similar quote by Alfred Marshall, made only months prior: ‘Thus there is de facto some sort of profit-and-loss sharing between almost every business and its employees; and perhaps this is in its very highest form when, without being embodied in a definite contract, the solidarity of interests between those who work together in the same business is recognized with cordial generosity as the result of true brotherly feeling. But such cases are not very common; and as a rule the relations between employers and employed are raised to a higher plane both economically and morally by the adoption of the system of profit sharing’ (1925, p. 627). 2. In fact, Simon presents his arguments in this article as a direct critique of New Institutionalist Economics: ‘The attempts of the new institutional economics to explain organizational behavior solely in terms of agency, asymmetric information, transaction costs, opportunism, and other concepts drawn from neoclassical economics ignore key organizational mechanisms like authority, identification, and coordination, and hence are seriously incomplete’ (1991, p. 42). 3. Simon here is, without acknowledging it, adopting a part of the argument for the existence of the firm first made by Alchian and Demsetz (1972). However, Simon rejects the rest of the Alchian and Demsetz story: they explain that, owing to this inseparability of individual employee contributions to the firm’s bottom line, managers will simply pay employees a fixed wage, assuming the residual claimant position for themselves. They can then monitor employees to determine whether they are exerting the proper effort. 4. Simon makes an interesting point on operations according to simple commands: ‘For the organization to work well, it is not enough for employees to accept commands literally. In fact, obeying operating rules literally is a favorite method of work slowdown during labor–management disputes, as visitors to airports when controllers are unhappy can attest’ (1991, p. 32). 5. Simon in particular intended to restore the reputation of government: ‘Large organizations, especially governmental ones, are often caricatured as “bureaucracies,” but they are often highly effective systems, despite the fact that the profit motive can penetrate these vast structures only by indirect means’ (1991, p. 43). His insights would explain why we sometimes observe government agencies that work relatively efficiently, despite having a very ‘wrong’ incentive structure. 6. See Kahneman, Knetsch, and Thaler (1986); Frey and Bohnet (1995); Konow (1996). 7. Minkler includes a section in the paper on the methodology of the survey. He acknowledges that many economists are skeptical of surveys for the basic reason that they measure attitudes rather than actions. He examines the literature on the survey methodology, which in general finds that attitudes are a good indicator of future behavior. He also discusses two sources of potential biases in surveys, including a social desirability bias. It refers to a situation where ‘respondents tailor their responses to draw a favorable picture of themselves at the cost of providing truthful answers’ (2002, p. 16). Minkler dismisses the concern by noting that the ‘respondents subject to social desirability bias in their survey responses are also those most likely to be subject to its effect in firms. If the norm is to work hard in their firm, those most likely to overstate their propensity to work hard because of social desirability bias would also be those most likely to conform to the norm’ (ibid.). For further discussion, see pp. 15–19. 8. Langlois continues: ‘This is not to say that a norm must always emerge or that the mechanism of repeated play must always solve the prisoner’s dilemma in a happy fashion. There are far too many examples of social situations in which norms have collapsed or

112

9.

10.

11.

12.

13.

14.

15.

Employees and entrepreneurship failed to emerge and in which the dilemma of this game is all too real. It is a major task of research in this area to understand the circumstances under which efficiency-enhancing norms will in fact emerge’ (1992, p. 173). Though commonly seen with hobbies, we also observe this phenomenon with very successful and wealthy individuals who often continue to work very hard for the pure satisfaction that it brings. Individuals such as Donald Trump or Paul McCartney come to mind. We can put another twist on this: a job is ‘interpreted’ by the worker according to his own valuations and circumstances and context. Some jobs will be very enjoyable for some people according to the meaning that the actors assign to that job. A teacher of economics may be willing to work for a relatively low pay because the meaning that he puts on the job is that he is imparting an important worldview to others who are likely to act, politically or otherwise, according to that worldview. If his own ideology is important enough to him, this contextual, subjective meaning of the job can make the job worthwhile to him even at a relatively low pay. One of his superiors described Fry in the following way: ‘Art is not just an inventor . . . He is an innovator. He has a good feel for economics, practicality, and a strong profit motive. He proves things out in his own mind, but he has a strong sense of the end user’ (Pinchot 1985, p. 138). In this section I will be mostly relying on Richard Ebeling’s interpretation and extension of Schutz’s work into the economic world for the basic reason that, as Ebeling explained, Schutz’s ‘published writings contain few concrete applications of his ideas to economic and market processes’ (1999, p. 129). Ebeling (1999) actually brings up Coase to support his application of Schutzian theories to economics: ‘Coase does not deny that the “economic way of thinking” may have useful and indeed valuable applications in other surrounding disciplines. But he reminded his fellow economists that economics has potentially as many interesting things to learn from those other social sciences as it has to contribute to them’ (p. 134). Ebeling does not extend any of his work on Schutz to the theory of the firm, though. Ebeling (1994) presents a definition of an ‘ideal type’, based on the work of Max Weber: ‘An ideal type is meant to be a stylized reconstruction, a selection of typical traits or characteristics conceived to represent for purposes of the analysis at hand those qualities in an individual, social institution or order, or historical period that enable an interpretive understanding of that individual, institution or order, or historical period . . . it represents an accentuation of certain qualities or characteristics and thus an idealization of the various attributes the individuals or objects possess or have in common’ (p. 86). He went on to point out that ‘[t]he ideal type is a composite image of an individual or group of individuals created in the mind of a person wishing to either understand their actions in the past or anticipate their actions or reactions to various circumstances in the future. The complexity and difficulty for the individual attempting to construct and apply ideal types for the purposes of forming expectations about the possible actions of others arises from the fact that two problems confront him: first, it is necessary to enumerate the various idealized characteristics and attributes believed to be relevant for understanding what makes particular human beings “tick”; and, second, and often much more difficult, it is necessary to evaluate the relative importance (the “weight”) of each of these behavioral characteristics or qualities in alternative and changing circumstances. Only success in both – an understanding of the relevant behavioral characteristics and their relative importance in an individual’s actual conduct in a specific setting – enables correct expectations to be formed’ (p. 89, italics in the original). Interestingly enough, Foss (1997) criticizes the existing Austrian work on economic organizations for underestimating ‘the role of shared mental constructs – theories, norms, ideologies, culture, and so forth – in coordinating a complex division of labor’ (pp. 188– 9), and points to Ebeling’s early work on typifications (1986) as one exception (p. 189, footnote 9). Though Foss 1997 and 1999 contain some discussion of these issues (see below), he does not go far enough in explaining how these shared mental constructs can effectively bring about order in a firm.

Spontaneous order in decentralized firms 16.

17.

18.

19.

113

Foss 1997 and 1999 contain a cursory discussion of these issues (all discussion in Foss 1997 can be found on p. 192, and in Foss 1999 on pp. 475–7). Foss (1997) discusses the importance of corporate cultures, which he defines as ‘shared mental constructs’. Foss makes no mention of Schutz but rather bases this discussion on the work of Thomas Schelling and Harald Malmgren. He points out that these firm-specific mental constructs often have spontaneous origins, and ‘help coordinate distributed knowledge by infusing employees with firm-specific knowledge’ (p. 192). He also rightly emphasizes that they change over time, and may be influenced, though not predictably designed, by management. In a short and general discussion, Foss emphasizes that a corporate culture may act as a way to organize a localized discovery procedure, which may in turn require incomplete contracts. He does not discuss its implications to creative action of the employees, nor to decentralization of decision-making with in firms. The same is true of Foss (1999), where Foss considers the Schutz-influenced (through Lachmann) contributions by Langlois (1986b) and O’Driscoll and Rizzo (1985). Foss discusses the importance of typification processes in firms to explain how co-ordination can be achieved by managers who are limited in their planning and ordering abilities. But, though Foss does hypothesize that such a system would lead to decentralization pressures within firms, he does not consider its implications to the actions of the employees in any detail. The rest of this quote is also relevant and bears reproducing here: ‘If authority is used to transmit premises for making decisions rather than commands for specific behaviors, then many different experts can contribute their knowledge to a single decision. Information and policy rules can flow through the organization along many channels, serving as inputs – decision premises – for many organizational behaviors’ (Simon 1991, p. 32). Managers can aid the process of knowledge co-ordination by establishing a proper culture or constructive rules, in other words. This is in keeping with Kreps (1990), who introduced a basic theory of corporate culture. Foss sums up his argument in the following way: ‘Kreps argues that firms may develop implicit contracts that align incentives by signaling to employees that management will not opportunistically take advantage of them in the case of unforeseen events, although nothing specific is being said (or can be said) about the event’ (1999, p. 476). Foss builds on Kreps’s insight to conclude that ‘an important aspect of what a firm’s leaders can do is to influence and steer the development of schemes of typification that are flexible enough to accommodate unforeseen events, and that help agents coordinate their interdependent activities’ (ibid.). Kolind ultimately failed in this, as noted in Chapter 3. Oticon was unable to sustain its radically decentralized organization. Though this was most likely due to the inability of Oticon’s management to abstain from selective interventions, again as documented in Chapter 3, it appears that the lack of an organizational culture binding the employees together accounted for some part of the failure. In a meeting that led to the unraveling of the radical decentralization in Oticon in 1995, ‘employees dramatically expressed their concerns about the gap between the Oticon value base, and the way the company was actually run’ (Foss 2001c, p. 22). Without that organizational culture creating the structures of intersubjective meaning and organizational identification, radical decentralization of decision-making will be unstable and will not be able to survive.

5. Employees as creative agents When people make their own decisions about how to do their work and allocate their time, they often put more energy, effort, and creativity into their jobs. Studies of R&D projects, for instance, have found that when the members of project teams feel more freedom and control over their work they become more innovative. That sense of autonomy is probably part of entrepreneurial motivation, too: Not only do you keep the economic rewards of your own work, but you also can make your own decisions and feel like an owner. When people feel tightly controlled, by contrast, they are often less motivated and less creative. Albert Einstein put it well when he remembered the militaristic school he attended as a child: ‘This coercion had such a deterring effect upon me that, after I had passed the final examination, I found the consideration of any scientific problems distasteful for an entire year.’ (Malone 2004, pp. 34–5)

ENTREPRENEURSHIP AND CREATIVITY We have a much better understanding today of the benefits and costs of delegation and decentralization, and the relevant tradeoffs between the two. But there is one benefit of decentralization that has not been discussed in great detail so far: the creative response within firms. Discussion of elements of creativity within firms remains mostly neglected, even by Austrian economists.1 Creativity in action is best understood as innovative behavior that results in the introduction of new knowledge. That new knowledge can take the form of new products, new production processes, or new markets. Co-ordination and use of dispersed knowledge and creation of new knowledge are not the same thing, though they are certainly related, as I discussed at the end of Chapter 3. Firms with decentralized decision-making seek to generate the greatest amount of creativity and creative problem-solving in addition to the best coordination of dispersed knowledge. In a world of intense competition, it is simply too costly not to take advantage of people’s creative problem-solving, since firms need to be as responsive to changing market conditions as possible. Successful firms with decentralization of decision-making are able to ‘extract’ the greatest creativity from their employees.2 In other words, human capital has been taking center-stage in many modern businesses. Rajan and Wulf (2006) speculate that current organizational flattening is due to human capital becoming more important relative to physical capital: 114

Employees as creative agents

115

[A]s the development of financial markets has increased access to physical capital, and as human capital becomes more important to a firm’s comparative advantage, tall hierarchies may lead to top management losing the residual rights of control . . . If physical assets become relatively unimportant, ownership becomes less effective as a means of organization control. Tall hierarchies become less viable. Instead, [top management builds up control] by establishing direct contact with lower level managers (i.e. flattening the firm) and getting them to make human-capital specific investments vis a vis top management. Thus the human-capital-intensive firm is held together by a web of human-capital-specific investments, which are made possible by the flatter hierarchy. (p. 24)3

As workers accumulate more business-specific human capital their decision-making abilities increase, becoming more creative or innovative. In fact, their decisions become more entrepreneurial according to Joseph Schumpeter’s famous conception of entrepreneurship.4 Schumpeter said that ‘when we speak of the entrepreneur we do not mean so much a physical person as we do a function’ (1949, p. 268). That entrepreneurial function is simply the introduction of ‘novelty’. Novelty in the economic world consists of the ‘carrying out of new combinations’ – innovations: ‘Seen in this light, the entrepreneur and his functions are not difficult to conceptualize: the defining characteristic is simply the doing of new things or the doing of things that are already being done in a new way (innovation)’ (1947, p. 223). This definition seems of special significance when it comes to explaining creativity within firms: entrepreneurship is not restricted to the heads of companies or the directors of productive resources, but can potentially take place at all levels of the organizational hierarchy. Schumpeter himself would probably have accepted such a claim, as the following quote would indicate: we call entrepreneurs not only those ‘independent’ businessmen in an exchange economy who are usually so designated, but all those who actually fulfill the function by which we define the concept, even if they are, as is becoming the rule, ‘dependent’ employees of a company, like managers, members of boards of directors, and so forth . . . [The concept of entrepreneurship] does not include all heads of firms or managers or industrialists who merely may operate an established business, but only those who actually perform that function. (1934, pp. 74–5)

Schumpeter’s ‘functional’ theory of entrepreneurship asserts that no person is always an entrepreneur; instead, a person, any person, can only act entrepreneurially: ‘everyone is an entrepreneur only when he actually “carries out new combinations”, and loses that character as soon as he [ceases to carry out new combinations]’ (1934, p. 78). Schumpeter’s functional definition of entrepreneurship leads him to an important insight into the operation of large corporations. As we see in this

116

Employees and entrepreneurship

insightful passage, Schumpeter understood that entrepreneurship within large corporations is usually not found in the figureheads of the corporation, and he thus in some ways anticipated the decentralization developments: the entrepreneurial function may be and often is filled cooperatively. With the development of the largest-scale corporations this has evidently become of major importance: aptitudes that no single individual combines can thus be built into a corporate personality . . . In many cases, therefore, it is difficult or even impossible to name an individual that acts as ‘the entrepreneur’ in a concern. The leading people in particular, those who carry the titles of president or chairman of the board, may be mere co-ordinators or even figure-heads. (1949, p. 261)

Schumpeter here differentiates between the official leaders of a company and those who actually carry out the entrepreneurial function. In many cases, the real entrepreneurs within a company, those who introduce creative change, remain anonymous and are not the well-known persons. According to Schumpeter, this is especially the case in large corporations, where there is no single person that we could label the entrepreneur, though there will almost always be one person who fulfills the function of the director of factors of production. For the purpose of economic change and progress, the introduction of novelty through creativity in action is often accomplished by those other than the figurehead in a company, and may in fact, as Schumpeter himself pointed out, be carried out cooperatively. An important element in Schumpeter’s conception of entrepreneurship is that the introduction of novelty can only come about as a result of creative action. Creativity is necessary for the introduction of novelty, but not sufficient. Just imagining something new is not entrepreneurship. Something new must be actually done5 to be considered entrepreneurial. The action component in ‘creative action’ is absolutely key. This relates to Schumpeter’s well-known differentiation between invention and innovation: new combinations must be carried out within a productive process in order for the action to be innovative rather than just inventive. Invention is not entrepreneurship – only innovation is: ‘The inventor produces ideas, the entrepreneur “gets things done . . .” ’ (1947, p. 224). It is not enough to sit on the sidelines and think up brilliant insights. Entrepreneurship can ultimately be summed up as creativity in action.6

THE PROCESS OF ENTREPRENEURIAL DECISIONMAKING AND JUDGMENT DERIVATION An owner or CEO of a firm obviously must expect to benefit when her employees engage in entrepreneurial behavior (in the above Schumpeterian

Employees as creative agents

117

sense). The potential benefits should be clear: greater innovation, greater flexibility, and better ability to cope with market change.7 There are two important, related questions here: what is behind creative action by employees, and why can employees in some circumstances act more creatively than the head of the company? The answers to these questions lie in the same series of articles by Richard Ebeling on the Austrian theory of expectations (1994, 1995a, 1995b, 1995c, 1999) which I referenced in the previous chapter, explaining the intra-firm process of spontaneous order. As part of his discussion of structures of intersubjective meanings, Ebeling also deals with differences in subjective interpretations of events or actions. At the same time that employees can share a core structure of intersubjective meanings (which permeate all social entities), they all also differ in their perceptions of events and actions, assigning other meanings to them than their co-workers. The cause of this is the simple fact of the differences in individual experiences. As people experience different events, they learn different facts which they apply to their total meaning structure, resulting in multitudes of at least slightly differentiated individuals. It is the actions that people take (which are based upon the subjective meanings that they ascribe to events and others’ actions) that are the ultimate source of ‘modifications in the structures of intersubjective meaning themselves’ (Ebeling 1995c, p. 49). People ‘test’ their interpretations of events and actions of others in the markets, and modify their interpretive schemas accordingly, which ultimately results in modifications to and the evolution of the structures of intersubjective meanings.8 This can help us understand why different people faced with the apparently exact same objective circumstances will make different decisions – though all members of a society or a firm may share certain structures of intersubjective meanings, they will also possess some structures of purely subjective meanings. An instructive example of this phenomenon within the economic realm is a price. Economists often explain the price of a good as an example of market-generated objective information. But that does not account for the fact that different people may derive very different information from that same price. The numerical value of the price may be objective, but the information that a businessman derives from the price is anything but. Ebeling explains it in the following way: A seller finds himself with unsold inventory of a product in excess of desired levels at a particular price. But what exactly is the market telling him at this price? That he needs to relocate his store? That he has failed to advertise the existence or availability of the product sufficiently? That the price is ‘right’ but the quality or characteristics of the product is ‘wrong’? Or that the quality and characteristics are ‘right’ but the price is ‘wrong’? What the price has conveyed is information that something is wrong, that the seller’s plans and expectations are

118

Employees and entrepreneurship

inconsistent with those of others. It has not unambiguously told him in which direction the error lies. The price’s information, in other words, needs interpretation as to its meaning concerning the preferences and plans of others. (1995b, p. 143, italics in original)

That interpretation can partly be generated by the structures of shared meaning within which we are submerged, by applying the right ideal types to derive the common conclusion about the information contained in that price. But each one of us will also use the purely subjective elements of our interpretive schemas, the elements that result from the basic Hayekian division of knowledge. As a result of this division of knowledge, we all have ‘accumulated a stock and structure of knowledge of specific “ideal types” . . . that other people in the division of labor do not possess’ (Ebeling 1995b, p. 148). In other words, some people possess ideal types about situations, actions, and objects (which Ebeling defines as persons’ ‘typifications’) that are more specific compared to other people; moreover, people can hold several sets of typifications (based on different interpretive schemas) without knowing with certainty which one is the correct one. Ebeling introduces an important element to this explanation of differences of human choices when he defines the individual choosing between alternate typifications (attempting to select the ‘correct’ one), as ‘the entrepreneurial element’ of the decision-making process:9 ‘each trader will have a stock of experience-generated specific typifications which he will draw upon to decide the possible meanings the price change or excess supply might have in the particular circumstance’ (1995b, p. 149). So the application of an individual’s differentiated structure of purely subjective interpretation is the source of creativity and thus of the entrepreneurial insights. This, I believe, is one way to gain a deeper understanding of entrepreneurial decision-making, a phenomenon that is ultimately indeterminate. If we accept Ebeling’s explanation of the ‘entrepreneurial element of the decision-making process’, then we must consider the possibility that a collection of people engaging in an ‘interpretive process’ will have a greater chance of arriving at the winning solution.10 In other words, verstehen may come about as a result of a collaborative effort in interpretation. If the role of interpreting the meaning of a price change (or any other market event) is not exclusively held by a head of a company or its upper management but also by some part of the workforce of a company, the chances of coming up with the ‘right’ interpretation may be much greater. Along similar lines, management meetings can be understood as a way to consider several different interpretive schemas of the same situation, and then choose the seemingly most correct one. CEOs seeking advice from their subordinates

Employees as creative agents

119

is another example of this phenomenon. In a situation of an extended division of knowledge, employees may sometimes be in the best position to judge the meaning of some market event. Having a group of employees engage in individual interpretation of such an event by approaching it with their own different ideal types and differentiated structures of subjective meanings may bring about a more successful outcome than having only a few managers do this. Through decentralization of decision-making, firms can tap into the specific ideal types of all the actors at their disposal – in effect, obtaining a large pool of non-shared specific knowledge – while still being able to achieve co-ordination partially through the existence of predominant structures of intersubjective meanings and partially through some remaining conscious co-ordination by a directing figure.11

CREATIVITY AND IMAGINATION Ebeling’s explanation of entrepreneurial and creative decision-making is entirely in harmony with Schumpeter’s understanding of this phenomenon (though Schumpeter does not discuss it in nearly as much detail). To Schumpeter innovations are created through imagination, which is simply a creative ‘vision’ of a better way to use the existing resources or information.12 De Vecchi effectively summarizes Schumpeter’s views on the role of imagination in entrepreneurship: Moreover, because he tries something new, the entrepreneur adds something to the facts, he gives them new forms and contexts and uses his ‘imagination,’ i.e. he detaches himself from the present, opens up to the world of possibilities, identifies ‘a real possibility’ in the new production combinations and sees the future he wants to create, as clearly as the present . . . The entrepreneur is a creator, but his creativity is neither instinctive nor irrational, but founded on knowledge of the present situation and of feasible future situations. (1995c, pp. 18–19)

The creative component of entrepreneurship consists of taking the ‘knowledge of the present situation’, detaching oneself from the ‘objective’ facts of the present, and trying to imagine an alternative situation. The process of entrepreneurial imagination is thus akin to the process of interpretation of the present situation. Ebeling’s phenomenological conception of entrepreneurial creativity is strikingly similar to and compatible with Schumpeter’s: The entrepreneur, however, stands apart from other men and their routines and existing patterns of ‘facts.’ The entrepreneur holds the facts of daily experience

120

Employees and entrepreneurship

and the events of yesterday and turns them in his hands. He sees possibilities, meanings, and new ‘essences,’ that is, new principles of order and arrangement and direction that others around him have failed to possess the intuition to see. The entrepreneur stands as the philosopher of the market place searching for eidetic relationships among the facts of everyday life. Where others experience only the ‘facts’ of plan frustration, the entrepreneur sees patterns of coordination. When the new relationships, patterns, and structures take shape under the guidance of the entrepreneur, those with less imagination frequently say, ‘Why did I not see that?,’ in the same ‘facts’ of the situation. (Ebeling 1995c, p. 47)13

It should be clear that Ebeling and Schumpeter are describing the same process.14 Entrepreneurs provide the creative spark for the introduction of ‘new essences’ and ‘new principles of order and arrangement’, or, phrasing it in Schumpeter’s way, they carry out new combinations. By joining Ebeling’s phenomenological perspective with Schumpeter’s conception of entrepreneurship we get a deeper understanding of entrepreneurship. The mystical ‘creative moment’, while unlikely to ever be fully understood, or certainly predicted, can at least partially be explained as a person’s interpretation of the current facts in different, more appropriately contextual and meaningful ways. This unique interpretation is the result of one’s own different knowledge and experiences, as explained above, and can never be known ahead of time, as Ebeling explains in the following passage: creativity, innovation, and change ultimately cannot be predicted. Creativity, innovation and change arise out of modifications in people’s conduct, i.e., out of new and different choices they make and the actions they undertake. And these choices, as we have seen, emerge out of the fantasizing processes of the human mind; but those human minds do not know ahead of time what their choices and actions will be until they have run through their own projecting processes of imagined possibilities and alternatives. (Ebeling 1999, p. 126)

In light of the above discussion, entrepreneurship can be defined as action within some institutional structure (for example, a market or a firm), consisting of introducing something new into that structure. This ‘novelty’ is born of the intra-personal creative process which is due to the particular individual’s greater specific knowledge and/or more appropriate experiences. In other words, creativity is due to the individual’s possession of the more appropriate typifications and being able to make better sense of those typifications.15 This is the Schumpeter/Ebeling vision of the ‘entrepreneurial element’. Entrepreneurial creativity according to this explanation is a basic human characteristic, and it can be exercised by anyone. Novelty can emerge both in terms of the products and services offered by the firm and also in the way the firm operates, usually consisting of changes in the routines (informal rules) and sometimes even the formal rules of the organization.

Employees as creative agents

121

CREATIVITY, JUDGMENT AND DELEGATION Foss and Klein (2005) and Foss, Foss and Klein (2006) are an important recent contribution in the field of entrepreneurship within the firm. In Foss and Klein (2005) the authors follow Knight’s (1921) lead in thinking of entrepreneurship as an exercise of judgment. Judgment is defined as ‘business decision-making when the range of possible future outcomes, let alone the likelihood of individual outcomes, is generally unknown’ (p. 8). In other words, judgment is action in the face of uncertainty (similar to Mises 1949, where entrepreneurship was defined as action in the face of uncertainty). Judgment primarily involves making decisions about employment of firm resources, and is continuous and often mundane, but absolutely necessary for a firm to operate. But, facing uncertainty, entrepreneurs can never know all possible attributes of the available physical assets. Thus, ‘an important function of entrepreneurship is to create or discover [new productive attributes of physical assets available to the firm]’ (Foss and Klein 2005, p. 14). There is an important proviso, though: physical asset ownership by entrepreneurs is necessary for exercise of judgment. Since, they claim, the facility of judgment can’t be contracted for or bought on the market, entrepreneurs need a firm to fulfill their function. Ownership will also provide high-powered incentives to create or discover new asset attributes. Within this basic context, the authors see benefits from delegation, somewhat contrary to Foss (2001a). They recognize that recent evidence shows that granting employees at least some decision-making rights 1) motivates them to be creative, and 2) generates direct benefits for them, as most employees value having at least some control over their jobs. The firm and the entrepreneur will also benefit because employees will be given ‘opportunities to exercise their own, often far reaching, judgments’ (p. 20), or, in other words, to engage in ‘productive’ entrepreneurship. Ownership will give employers the ultimate authority to ensure that delegation is at a level where productive entrepreneurship will be maximized relative to ‘destructive’ entrepreneurship – which they define as consisting of various opportunistic methods used by employees to benefit themselves at the expense of the firm. Foss, Foss and Klein (2006) extend this discussion of delegation by differentiating between ‘original’ judgment and ‘derived’ judgment. Original judgment is exercised by the firm/resource owners, even in the cases of delegation, since they must make a judgment about how much delegation should be granted and to whom. Derived judgment on the other hand is exercised by the employees who hold the delegated decision rights. Thus, all exercise of employee judgment is derived from the original

122

Employees and entrepreneurship

judgment of the owners. As Foss, Foss and Klein explain, ‘employees are expected not to carry out routine instructions in a mechanical, passive way, but to apply their own judgment to new circumstances or situations that may be unknown to the employer’ (p. 3). This will allow the entrepreneurs to take advantage of employees’ specific knowledge, though they do not elaborate whether this would result in greater knowledge co-ordination or actual creation of new knowledge. One way or another, employees are given more control over the firm capital and face reduced constraints with the ultimate purpose of encouraging them to ‘create or discover new attributes of productive assets’ (p. 10). Thus, employees act entrepreneurially when exercising derived judgment under conditions of uncertainty. The above explanation makes a valuable contribution to explaining delegation and, more broadly speaking, decentralization of decision-making, but it is missing one essential factor: it fails to explain why employees can be expected to hold knowledge which will in at least some cases be superior to the knowledge of the employer, and therefore why derived judgment can sometimes trump original judgment. It is certainly possible to explain this with Hayek’s conception of dispersed knowledge, as I did in Chapter 3, but it is important to also consider the role of employee creativity, and we can do that by using Ebeling’s framework that I explained above. The employee-held knowledge could take the form of different typifications which will allow a different interpretation of the current market conditions – or, within the Foss, Foss and Klein context, a different interpretation of the possible physical asset attributes. Thus, derived judgment may be superior simply because the different typifications held by employees will bring about a creative response by them. Though Foss, Foss and Klein speak of judgment rather than creativity, the difference between the two is rather nebulous and it is probably impossible to separate the two. Most judgment involves, to paraphrase the earlier quote on p. 119 by Ebeling (1995c), adding something to the facts and giving them new forms and contexts. The ultimate purpose of granting the freedom to form judgment to employees is to tap into their unique typifications and interpretative schemas based on differentiated structures of subjective meanings, and thus bring about a creation of something new – an innovation.16 Foss, Foss and Klein show some awareness of this when they explain that derived judgment by employees can involve ‘imagination, creativity, leadership and related factors’ (p. 5). Granted, some if not many judgments may in fact be relatively unimportant, or even mundane, but that should not detract from their ‘novel’ nature, as Schumpeter explained when discussing entrepreneurship: It is but natural, and in fact it is an advantage, that such a definition [of entrepreneurship] does not draw any sharp line between what is and what is not

Employees as creative agents

123

enterprise. For actual life itself knows no such sharp division, though it shows up the type well enough. It should be observed at once that the ‘new thing’ need not be spectacular or of historic importance. It need not be Bessemer steel or the explosion motor. It can be the Deerfoot sausage. To see the phenomenon even in the humblest levels of the business world is quite essential though it may be difficult to find the humblest entrepreneurs historically. (Schumpeter 1947, p. 223)

Combining the Foss, Foss and Klein conception of derived judgment with the above explanations of entrepreneurship and creativity by Schumpeter and Ebeling gives us a deeper, more thorough understanding of the benefits of decentralization and delegation of control over capital goods to employees. It also may at least partially explain why profit-sharing, stock options, and various bonus and reward systems have been instituted in many modern firms. There is greater recognition that the entrepreneurial function may be carried out by employees at lower levels of the hierarchy, and entrepreneurial contributions must be fostered – and rewarded.

CONCLUSION Over the last 15 years a more complete list of benefits and costs of decentralization in firms has emerged. The benefits are increased knowledge co-ordination and innovation. They result from better alignment of decision-making power with the employees holding the most appropriate knowledge and typifications. On the other hand, the costs arise from increased opportunism, incentive misalignment and institutional instability. In other words, the tradeoff is between ‘coherence’ of the firm on one hand and ‘flexibility’ of the firm on the other, as discussed in Chapter 3. As a firm becomes more participatory, a greater number of individuals will be acting according to their own thoughts, interpretations, and desires, leading to a potential splintering of a firm’s purpose and capabilities. In other words, it may lose ‘coherence’: there might be less co-ordination of the actions of the employees, and therefore firms might be less stable and ultimately less durable. But greater access to dispersed knowledge and, maybe more importantly, dispersed innovation could potentially overcome the loss of coherence. Economists have traditionally focused on transaction-costsreducing properties of firms at the cost of neglecting to examine firms’ need for flexibility to rapid market change. That seems to be changing today with a greater emphasis on the potential benefits of delegation and decentralization. It appears that we might be arriving at the point where we finally have a more dynamic understanding of institutional efficiency, as Langlois called for over 20 years ago (see Chapter 3). Firms must have an

124

Employees and entrepreneurship

appropriate structure for their particular competitive environment. And it may be the case that today’s competitive environments are changing in such ways that more firms will have to become more decentralized in order to survive. Ultimately the proper extent of decentralization in a firm is an issue that cannot be decided by economists but rather only by firms themselves.

NOTES 1.

2.

3.

4.

5. 6.

7.

The one exception to this is again Sautet (2000), the theme of which is, as I pointed out already, how an M-form encourages creation of new knowledge in a form of Kirznerian discovery of opportunities on the part of the employees. However, I will rely on Schumpeter’s conception of entrepreneurship rather than Kirzner’s to explain how new knowledge consists of recognizing not only new opportunities but also new products, production processes, and even organizational forms. I mentioned Charles Koch and the company he founded – the world’s largest privately held, Koch Industries – in Chapter 3. Koch credits the company’s success to Market Based Management, a management method that he created and has practiced since the 1960s, which is largely based on decentralization of decision-making throughout the company. Mr Koch champions the benefits of employees’ pursuit of creativity and innovation, or what he calls ‘discovery’, as these quotes demonstrate: ‘To build a culture of discovery, we must encourage, not discourage, the passionate pursuit of our own and others’ hunches’ (Koch 2007, p. 42); ‘Leaving the particulars to the person doing the work encourages discovery. It also enhances adaptation to changing conditions’ (Koch 2007, p. 78); ‘Expectations must also be open-ended and challenging enough to expand an employee’s vision of what can be contributed to an activity. This encourages experimentation and innovation’ (Koch 2007, p. 130); ‘Since the future is unknown and unknowable, those contributing to innovation must be given every possible encouragement and a latitude consistent with their performance and capabilities’ (Koch 2007, p. 166). We observe increasing human capital investment in most sectors of the economy, much of it either subsidized or even fully funded by employers. It is becoming relatively common for firms to pay for employees, especially lower managers, to acquire an MBA degree. One way to explain this is that firms are attempting to establish a credible commitment to decentralized decision-making by investing in their greater decision-making ability. Much analysis has been devoted to Schumpeter’s theory of entrepreneurship, and I will not address this large body of literature in this book. It should be noted that the apparent tensions in Schumpeter’s theories could be problematic to the theme of this chapter, in particular his controversial claims about the eventual obsolescence of the entrepreneur and the routinization of innovation in large corporations (Schumpeter 1942). In addition, Schumpeter sometimes spoke about entrepreneurs in heavily ‘heroic’ terms that are at odds with some of his other statements. I will be self-consciously selective in the use of Schumpeter’s concepts and will not attempt to reconcile Schumpeter’s seemingly contradictory statements. Schumpeter: ‘Now, it is this “doing the thing,” without which possibilities are dead, of which the leader’s function consists’ (1934, p. 88). It bears mentioning that this understanding of entrepreneurship is mostly consistent with Mises’ own definition of entrepreneurship as action in the face of uncertainty. It is only when we are acting in creative, new ways that we are fully confronted with uncertainty. Otherwise, we are acting according to the routine, the consequences of which are mostly (though never absolutely) certain. I dealt with the costs of potential downsides of employee entrepreneurship in Chapter 3.

Employees as creative agents 8.

9.

10.

11.

12.

13.

125

Harper (1996) presents a similar explanation. Harper discusses entrepreneurial learning, and explains it as a sequence of continuous cycles of conjecture and refutation, with the market ultimately subjecting ‘the entrepreneur’s creative imagination . . . to critical control’ (p. 6). Compare this to Ebeling (1995b): ‘The market test inherent in rivalrous competition is, therefore, a competition between interpretive schemas about the message conveyed by the market price signals’ (p. 149). It is a short step then to define rivalrous market competition as actually ‘a competition between interpretive schemas about the message conveyed by the market price signals’ (Ebeling 1995b, p. 149), or more generally ‘a clash of different interpretations of meanings in men’s actions’ (Ebeling 1999, p. 133). Alternatively, a firm could allow several teams to compete against one another by testing their interpretive schemas in the market; this could effectively reduce the risk of failure by ‘not putting all of one’s eggs in one basket’. This has become a relatively common practice among decentralized firms. It is important to emphasize that there is no contradiction between encouraging the widespread development of structures of intersubjective meanings on one hand and allowing individuals to act on their subjectively held knowledge. A firm may successfully impart an organizational culture encouraging creative and individual problemsolving among its employees only by actually relying on the employee application of subjective elements of their knowledge. That is what creative and individual problemsolving requires. If there is a conflict between intersubjective and subjective structures of meanings (for example, a situation where the organizational culture establishes that a team must agree upon or approve a particular creative solution derived from some subjective typification, but the team shoots down the solution), it is the task of managers to find ways to arrive at an acceptable balance or compromise between them. That may entail allowing the creative individual to work outside of his team, or form his own team. The work of G.L.S. Shackle also bears some resemblance to Schumpeter’s theory of entrepreneurship, as seen in the following quote: ‘[E]nterprise is the daring, and desirable, pursuit of imagination, not merely or mainly reason; that enterprise is imagination in action’ (1988, p. 161, italics in the original). To Shackle entrepreneurs imagine possible opportunities, and then work to create them ex nihilo. According to his radically subjectivist perspective, available resources are never objectively given, but rather depend on our ability to imagine their alternate uses. Imagining new possibilities requires entrepreneurs to form subjective projections of the future. As Ricketts (2002) explains it, ‘Shackle’s entrepreneur has perhaps a greater affinity with that of Schumpeter than with that of Kirzner. The emphasis on innovation on the one hand (Schumpeter) and the creative imagination on the other (Shackle) are closely related’ (p. 70). Ebeling adopts the term ‘eidetic’ from the philosopher Edmund Husserl, who defined ‘eidetic reduction’ as ‘an attempt to uncover the principles in concrete, actual experiences in the world. One takes an example or a set of examples drawn from experiences and through one’s mind tries to discern what are the invariant, or generic, qualities or properties exemplified in each. What it is, in other words, that binds them together as forms of a common phenomena and once discovered will be found to be essential and ever present in each and for each to take on the particular concrete configurations they display. It is this unearthing of general principles that Husserl viewed as discovery of the “essence” of a phenomena’ (Ebeling 1995c, p. 44–5, italics in the original). ‘Eidetic reduction’ allows a person to put his imagination into a context where action is possible, as can be seen in this passage by Kohak: ‘one person may experience the world as a realm of endless possibility, while another person may experience the same or even better equipped world as a meaningless context into which he feels, in Heidegger’s term, “contingently thrown.” Ordinary usage distinguishes between the two reactions by speaking of a person either having or lacking imagination. But that is not strictly correct. Both individuals have imagination – what the second person lacks is eidetic imagination. He cannot . . . imagine possibilities, or more exactly, he can imagine a possible fact but not a possible eidos, capable of giving meaning, direction and the dimension of possibility

126

14. 15. 16.

Employees and entrepreneurship to the facts at hand . . . Imagination grows far more with seeing the same thing in many perspectives and contexts’ (1978,p. 21–2, quoted by Ebeling (1995c, pp. 46–7)). In fact, Ebeling recognizes it himself: ‘Herein lies the secret of the Austrian emphasis on the entrepreneur. And in this context it is only fair to include Schumpeter with Herbert Davenport, Mises and Israel Kirzner’ (1995c, p. 47). In other words, an individual’s greater eidetic imagination. See Koch (2007) again: ‘Polanyi believed that discoveries best occur in a system of spontaneous order, of mutually adjusting individual initiatives . . . To begin creating such a system of discovery based on spontaneous order, employees’ roles must include making and encouraging innovations. This requires an environment in which people don’t blindly follow marching orders; rather, ideas are encouraged and challenged but not destructively criticized’ (pp. 164–5). Also: ‘The principle that should guide the implementation of all financial incentives is that they encourage the innovation and creative destruction necessary to maximize long-term sustainable profitability’ (p. 147).

6. Conclusion Thus organizational innovations like worker participation, teamwork and just-intime inventory systems can be seen as attempts by firms to expand their capabilities by placing decision-making authority and responsibility in the hands of those most able to meet the firm’s objectives. Combined with guaranteed employment, these innovations foster commitment to the firm and reduce the need for high degrees of hierarchy and monitoring. And the role of managers changes from one of director of production to one of coordinator and facilitator. Managers now incorporate ideas from all quarters, from marketing, engineering and production, into initial planning decisions and then facilitate needed changes once production is underway. The base for ideas is expanded and everyone becomes responsible for more than just narrowly defined tasks. These kind of organizational changes have enabled firms to enhance their product quality and tailor their output according to specific market segments. Simply put, by recognizing that ideas come from places other than R&D departments and that not only top level managers are capable of idea implementation, firms can now do things they could not do before. This lesson has been, and still is, a difficult one for firms to learn because the entire organizational culture has to change. But by focusing on the capabilities of firms, economists may now learn the same lesson. (Minkler 1993b, pp. 584–5)

There has been an increasing amount of recognition of the importance of decentralization of decision-making in modern firms over the last decade or two. Though radically decentralized institutional structures are unlikely to ever dominate the business world, it is quite possible that greater degrees of firm decentralization in general will become much more common. Decentralization brings about a great deal of flexibility, and it is most certainly not necessary to implement radical flattening of the internal administrative structure to gain some benefits of employee innovation. Many firms have been successful with moderate degrees of decentralization of decision-making. And in the cases where the costs of decentralization become too great, it is easy enough to revert to a more hierarchical structure. As owners and managers accumulate experience in how to make ‘demanaged’ firms work, as these structures become more familiar and acceptable in the business world, as we get a better understanding of the tradeoffs involved, and as more work becomes ‘knowledge work’ we can expect to see a greater amount of decentralization of decision-making. The time has come for economists to devote more resources to studying this phenomenon. Also, understanding firms with decentralized decisionmaking will in the end give us a much better general theory of the firm, 127

128

Employees and entrepreneurship

since no firms are perfectly centrally planned and inevitably feature some elements of decentralization. I have shown how we can explain the existence and operation of decentralization of decision-making within firms. First, I argued that the transaction cost theory of the firm is unable to explain decentralization of decision-making because it in general does not recognize intra-firm division of knowledge, also known as the double knowledge problem or the asymmetric knowledge problem, which can be summarized as a situation where the owners and/or managers may not have sufficient knowledge both of market opportunities and of personal knowledge and ideas held by their employees. In other words, employees may hold knowledge that is both superior to that of their managers and necessary for the success of the firm. In addition, transaction cost theories of the firm are based on an overly strong assumption of employee opportunism which is likely to prevent understanding of decentralized decision-making within firms. Second, I have pulled together recent theoretical work applicable to firms with decentralized decision-making. There has been an increasing level of inquiry into the relevance of the division of knowledge inside firms. Though only some of that work has much to say about decentralization within firms, I have shown that there are important connections between the two: decentralization of decision-making is a way to solve the double knowledge problem that faces managers in firms. In many real-world cases, the double knowledge problem may not be very significant. But when the market turns very competitive, innovation often becomes key to the survival of the firm. At those times owners and managers may wish to capture as much of the personal, idiosyncratic knowledge held by their employees as possible in order to maximize the innovative capability of their firm. This can be done by reducing or eliminating the hierarchical managerial structure within firms, and ‘empowering’ employees with a great deal of decision-making power and choice. Enabling employees to make decisions will allow a firm to use some or all of the knowledge dispersed among the employees of the firm. However, there are also costs to decentralization: as the actions of a firm’s employees are no longer directly and completely guided by managers, there is a serious risk of internal discoordination. Thus, there is an institutional tradeoff between the flexibility of full utilization of internally dispersed knowledge on the one hand, and loss of internal coherence on the other. The exercise of managerial judgment is necessary to determine the proper amount of this tradeoff for each firm. This is something that cannot be known ex ante, but rather must be discovered through a process of trial and error. This is why we sometimes see firms that change their internal structures frequently over a short period of time.

Conclusion

129

Third, I attempted to answer one key question relating to decentralization of decision-making within firms: how can there be any coherence or co-ordination in the absence of direct control by managers? Though some level of discoordination will be the inevitable result of decentralization, we do observe in at least some cases sufficient co-ordination to allow firms with decentralized decision-making to survive and be successful. How is this possible? My claim is that there exist forces of spontaneous order operating within firms, and I tried to provide the theoretical foundations to explain these forces. There are two aspects to this theory: first, we must adopt a more realistic account of employee motivations, one that rejects simple opportunism as the primary employee motivator. We must include motivating factors such as intrinsic motivation, morals, loyalty, peer-pressure, perception of fairness, and of course positive incentives. Second, I applied the Schutz/Ebeling theory of the typification process to the intra-firm analysis. Schutz/Ebeling developed an explanation of social co-ordination predicated on people forming ideal types of other members of a society. These typifications allow individuals greater success as they act in an uncertain world. When a society of people forms mutually dependent typifications (which are more or less harmonized), structures of shared or intersubjective meanings emerge to effectively guide the actions of the members of that society. I have shown that this explanation of social coordination holds with even greater force when it comes to intra-firm coordination, since unlike the situation in the wider society most intra-firm relationships are face to face. When these structures of shared meaning are combined with an organizational culture that encourages creativity and cooperation, decentralized firms are able to achieve a great deal of internal co-ordination. In other words, spontaneous order can emerge within firms with decentralized decision-making. Finally, I explained how it is possible for employees to be more relevantly creative than their superiors. Ebeling showed that individual creativity is driven by a different interpretation of the current facts, which in turn is due to differentiated typifications schemas built on different knowledge and experiences. This allows derived employee judgment to sometimes trump the original judgment exercised by managers. It also introduces an interesting conundrum: co-ordination within firms with decentralized decisionmaking depends on structures of shared meanings while the creativity of employees depends on the uniqueness of their subjective structures of meanings. Some decentralized firms have responded to this problem by creating and fostering a shared culture that encourages individualism and boldness rather than groupthink, but this is clearly a delicate balance that may be very difficult for firms with decentralized decision-making to get right.

130

Employees and entrepreneurship

The bottom line: decentralization of decision-making within firms has been introduced in order to encourage employees to engage in creative problem-solving which results in innovations. We are more likely to see such structures in highly competitive markets where it is not enough to be the most efficient or lowest-cost producer of goods which may quickly be rendered obsolete by innovative competitors. Successful decentralization of decision-making within firms is able to harness creativity, innovation, and discovery of new opportunities. However, much work still remains before we will have a full understanding of decentralization of decision-making within firms. I will present some unanswered questions next. Probably the most important theoretical question in this area concerns the operations of cross-functional teamwork. Almost all firms with significant degrees of decentralization of decision-making rely on crossfunctional ‘project’ teams consisting of members of undifferentiated rank. Why? I believe there are two reasons for this: 1) These teams are an effective way to build close relationships among employees which will increase the effectiveness of peer-pressure as a motivating factor; in addition, team members can closely monitor each other. 2) Working in small teams will bring about more effective structures of shared meaning, as all relevant relationships become face to face. Though I believe these are correct, I do not believe they are complete. It does not explain the relationship between teams and entrepreneurship. It is possible that cross-functional project teams may facilitate ‘team entrepreneurship’ (Sautet 2000, pp. 107, 129) where the combined knowledge and differentiated typifications of multiple employees can create innovations that the independently working teammembers would not have introduced. We must be able to explain this process in much greater detail. There is an important unanswered question in regard to a firm-relevant market structure: how exactly does the competitive environment affect the extent of decentralization of decision-making? In other words, will there be any stable relationship between the intensity of market competition and the amount of decentralization that we observe? I have suggested that there is, but further theoretical development as well as empirical evidence is necessary to answer this question to the full extent. Also, what is the relationship between the form of ownership and the extent of decentralization? Are we more likely to see decentralization among proprietorships or corporations? In a proprietorship the decision to engage in decentralization comes down to the single owner who is not responsible to anyone else and can stick to his vision even in the face of temporary setbacks; in a corporation it may be more difficult to decentralize decision-making, since the management must justify it to the board of directors and stockholders, parties who may

Conclusion

131

not be enthusiastic supporters of decentralization and who may not understand it very well. Therefore, I expect that we would observe decentralization of decision-making in proprietorships more than in corporations, but much work is required before this question can be fully settled. Further questions that remain to be answered are, what rules, formal and informal, are required to have successful decentralized decision-making within firms? Are these rules general and transferable or specific to each firm? My analysis would indicate that most of the rules are specific to each firm, but we don’t understand this yet. In addition, how important is it that employees have a perception of fairness on the part of both the rule structure and the management/owners? Again, my studies would lead me to say that the perception of fairness is absolutely key to accomplish organizational loyalty and self-identification with the firm’s goals. Also, are profitsharing and employee stock ownership plans (ESOPs) necessary or merely optional features of decentralization of decision-making within firms? All of the firms with significant decentralization of decision-making that I have examined put a lot of importance on profit-sharing and/or ESOPs, as a way both to tie the rewards to employee output and to create a sense of ownership over the firm which would bring about greater emotional involvement in its performance. But are they necessary? There are obviously many potentially fruitful areas of inquiry that remain ahead of us. Once we can answer all these questions, we will have not only a much stronger theory of decentralization of decision-making within firms, but also a stronger theory of the firm in general.

Bibliography Aghion, Philippe and Jean Tirole (1997), ‘Formal and Real Authority in Organizations’, Journal of Political Economy, 105 (1), 1–29. Alchian, A.A. and H. Demsetz (1972), ‘Production, Information Costs and Economic Organization’, American Economic Review, 62 (5), 777– 95. Alchian, Armen A. and Susan Woodward (1988), ‘The Firm Is Dead: Long Live the Firm – A Review of Oliver E. Williamson’s The Economic Institutions of Capitalism’, Journal of Economic Literature, XXVI (March), 65–79. Aoki, M. (1990), ‘Towards an Economic Model of the Japanese Firm’, Journal of Economic Literature, 28 (1), 1–27. Bain, J.S. (1956), Barriers to New Competition, Cambridge, MA: Harvard University Press. —— (1968), Industrial Organization, New York: John Wiley & Sons. Bartlett, F.C. (1932), Remembering, Cambridge: Cambridge University Press. Becker, Gary (1962), ‘Investment in Human Capital: Effects on Earnings’, Journal of Political Economy, 70 (October), 9–49. Boettke, Peter (1993), Why Perestroika Failed, London: Routledge. Casson, Mark (1982), The Entrepreneur: An Economic Theory, Totowa, NJ: Barnes & Noble. Chandler, Alfred ([1962] 1990), Strategy and Structure, Second Edition, Cambridge, MA: MIT Press. Coase, Ronald H. (1937), ‘The Nature of the Firm’, Economica, 4, 386–405. —— (1960), ‘The Problem of Social Cost’, Journal of Law and Economics, 3 (October), 1–44. —— (1984), ‘The New Institutional Economics’, Journal of Institutional and Theoretical Economics, 140 (March), 229–31. —— (1988), ‘The Nature of the Firm: Origin, Meaning and Influence’, Journal of Law, Economics and Organization, 4 (1) (Spring), 3–47. —— (1992), ‘The Institutional Structure of Production’, American Economic Review, 82 (4) (September), 713–19. Colombo, Massimo and Marco Delmastro (2004), ‘Delegation of Authority in Business Organizations: An Empirical Test’, Journal of Industrial Economics, 52 (1), 53–80. 132

Bibliography

133

Cowen, Tyler and David Parker (1997), Markets in the Firm: A MarketProcess Approach to Management, London: Institute of Economic Affairs. Cyert, Richard and James March (1963), A Behavioral Theory of the Firm, Englewood Cliffs, NJ: Prentice Hall. De Paola, Maria and Vincenzo Scoppa (2006), ‘Organizational Design, Project Selection, and Incentives’, Journal of Institutional and Theoretical Economics, 162 (3), 424–49. Dessein, Wouter (2002), ‘Authority and Communication in Organizations’, Review of Economic Studies, 69 (4), 811–38. De Vecchi, Nicolo (1995), Entrepreneurs, Institutions and Economic Change, Aldershot, UK and Brookfield, USA: Edward Elgar. Ebeling, Richard (1986), ‘Toward a Hermeneutical Economics’, in I.M. Kirzner (ed.), Subjectivism, Intelligibility, and Economic Understanding, London: Macmillan. —— (1994), ‘Expectations and Expectations Formation in Mises’ Theory of the Market Process’, in Peter Boettke and David Prychitko (eds), The Market Process, Aldershot, UK and Brookfield, USA: Edward Elgar. —— (1995a), ‘Cooperation in Anonymity’, in David Prychitko (ed.), Individuals, Institutions, Interpretations, Aldershot: Avebury Publishing. —— (1995b), ‘Toward a Hermeneutical Economics: Expectations, Prices and the Role of Interpretation in a Theory of the Market Process’, in David Prychitko (ed.), Individuals, Institutions, Interpretations, Aldershot: Avebury Publishing. —— (1995c), ‘Austrian Subjectivism and Phenomenological Foundations’, Advances in Austrian Economics, 2A, 39–53. —— (1999), ‘Human Action, Ideal Types, and the Market Process: Alfred Schutz and the Austrian Economists’, in Lester Embree (ed.), Schutzian Social Science, 115–34. Dordrecht: Kluwer Academic Publishers. Foss, Nicolai J. (1994), ‘The Theory of the Firm: The Austrians as Precursors and Critics of Contemporary Theory’, Review of Austrian Economics, 7 (1), 31–65. —— (1997), ‘Austrian Insights and the Theory of the Firm’, Advances in Austrian Economics, 4, 175–98. —— (1999), ‘The Use of Knowledge in Firms’, Journal of Institutional and Theoretical Economics, 155, 458–86. —— (2001a), ‘Misesian Ownership and Coasian Authority in Hayekian Settings: The Case of the Knowledge Economy’, Quarterly Journal of Austrian Economics, 4 (4), 3–24. —— (2001b), ‘Whither Economic Organization?’, manuscript (2005), Zeitschrift für Betriebswirtschaftsforschung, (2002), 57–67.

134

Bibliography

Foss, Nicolai J. (2001c), ‘Internal Disaggregation in Oticon: An Organizational Economics Interpretation of the Rise and Decline of the Spaghetti Organization’, unpublished manuscript. —— (2003), ‘Selective Intervention and Internal Hybrids: Interpreting and Learning from the Rise and Decline of the Oticon Spaghetti Organization’, Organization Science, 14 (3), 331–49. Foss, Nicolai and Peter Klein (2005), ‘Entrepreneurship and the Economic Theory of the Firm: Any Gains from Trade?’, in Sharon Alvarez, Rajshree Agarwal and Olav Sorenson (eds), Handbook of Entrepreneurship Research: Disciplinary Perspectives, New York: Springer. Foss, Nicolai and Volker Mahnke (eds) (2000), Competence, Governance, and Entrepreneurship, Oxford: Oxford University Press. Foss, Kirsten, Nicolai Foss and Peter Klein (2006), ‘Original and Derived Judgment: An Entrepreneurial Theory of Economic Organization’, manuscript, (2007), Organization Studies, 28 (12), 1893–912. Frey, B. (1997), ‘On the Relationship between Intrinsic and Extrinsic Work Motivation’, International Journal of Industrial Organization, 15, 427–39. Frey, B. and I. Bohnet (1995), ‘Institutions Affect Fairness: Experimental Investigations’, Journal of Institutional and Theoretical Economics, 151, 286–303. Gable, Wayne and Jerry Ellig (1993), Introduction to Market-Based Management, Fairfax, VA: Center for Study of Market Processes. Garrouste, Pierre (2002), ‘Knowledge: A Challenge for the Austrian Theory of the Firm’, in Nicolai Foss and Peter Klein (eds), Entrepreneurship and the Firm, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Grossman, Sanford J. and Oliver D. Hart (1986), ‘The Costs and Benefits of Ownership: A Theory of Vertical and Lateral Integration’, Journal of Political Economy, 94 (4), August, 691–719. Harper, David A. (1996), Entrepreneurship and the Market Process, London: Routledge. Hart, Oliver and John Moore (1999), ‘On the Design of Hierarchies: Coordination versus Specialization’, Journal of Political Economy, 113 (4), 675–702. Hayek, Friedrich A. ([1945] 1980), ‘The Use of Knowledge in a Society’, in Individualism and Economic Order, Chicago: University of Chicago Press, Midway Reprint. —— (1944), The Road to Serfdom, Chicago: University of Chicago Press. —— ([1948] 1980), Individualism and Economic Order, Chicago: University of Chicago Press, Midway Reprint. —— (1973), Law, Legislation and Liberty, Vol. 1: Rules and Order, Chicago: University of Chicago Press.

Bibliography

135

Holmstrom, Bengt and John Roberts (1998), ‘The Boundaries of the Firm Revisited’, Journal of Economic Perspectives, 12, 73–94. Jensen, Michael and William Meckling (1992), ‘Specific and General Knowledge, and Organization Structure’, in L. Werin and H. Wijkander (eds), Contract in Economics, Cambridge: Basil Blackwell. Kahneman, D., J. Knetsch, and R. Thaler (1986), ‘Fairness as a Constraint on Profit Seeking’, American Economic Review, 76, 728–41. Kirzner, Israel (1985), Discovery and the Capitalist Process, Chicago: University of Chicago Press. —— (1992), The Meaning of the Market Process, London: Routledge. Knight, Frank (1921), Risk, Uncertainty, and Profit, New York: August M. Kelley. Koch, Charles (2007), The Science of Success, Hoboken, NJ: John Wiley & Sons. Kohak, Erazim (1978), Idea and Experience, Chicago: University of Chicago Press. Konow, J. (1996), ‘A Positive Theory of Economic Fairness’, Journal of Economic Behavior and Organization, 31 (1), 13–55. Kreps, D.M. (1990), ‘Corporate Culture and Economic Theory’, in J. Alt and K. Shepsle (eds), Perspectives on Positive Political Economy, Cambridge: Cambridge University Press. Lachmann, Ludwig M. (1971), The Legacy of Max Weber, Berkeley, CA: Glendessary Press. Langlois, Richard N. (1986a), ‘The New Institutional Economics: An Introductory Essay’, in Richard Langlois (ed.), Economics as a Process, Cambridge: Cambridge University Press. —— (1986b), ‘Coherence and Flexibility: Social Institutions in a World of Radical Uncertainty’, in I.M. Kirzner (ed.), Subjectivism, Intelligibility, and Economic Understanding, London: Macmillan. —— (1992), ‘Orders and Organizations: Toward an Austrian Theory of Social Institutions’, in B.J. Caldwell and S. Boehm (eds), Austrian Economics: Tensions and New Directions, Boston: Kluwer Academic Publishers. —— (1994a), ‘Do Firms Plan?’, unpublished manuscript, (1995), Constitutional Political Economy, 6, 247–61. —— (1994b), ‘The Boundaries of the Firm’, in P.J. Boettke (ed.), The Elgar Companion to Austrian Economics, Aldershot, UK and Brookfield, USA: Edward Elgar. Lavoie, Donald C. (1985a), Rivalry and Central Planning: The Socialist Calculation Debate Reconsidered, Cambridge: Cambridge University Press.

136

Bibliography

Lavoie, Donald C. (1985b), National Economic Planning: What Is Left?, Cambridge, MA: Ballinger Publishing Company, Harper & Row. Malone, Thomas (2004), The Future of Work, Boston: Harvard Business School Press. Marshak, Jacob (1968), ‘Economics of Inquiring, Communicating, Deciding’, American Economic Review, 58 (May), 1–18. Marshall, Alfred (1925), Principles of Economics, 7th edn, London: Macmillan. —— (1948), Principles of Economics, 8th edn, New York: Macmillan. Menger, Carl ([1883] 1985), Investigations into the Method of the Social Sciences with Special Reference to Economics, trans. Francis J. Nock, ed. Lawrence White, New York: New York University Press. Minkler, Lanse (1993a) ‘Knowledge and Internal Organization’, Journal of Economic Behavior and Organization, 21, 17–30. —— (1993b), ‘The Problem with Dispersed Knowledge: Firms in Theory and Practice’, Kyklos, 46, 569–87. —— (2002), ‘Shirking and Motivations in Firms: Survey Evidence on Worker Attitudes’, University of Connecticut Department of Economics Working Paper Series, Working Paper 2002–40, September 2002. Mises, Ludwig von ([1944] 1983), Bureaucracy, Cedar Falls, IA: Center for Futures Education. —— (1949), Human Action, Chicago: Henry Regnery Company. Mookherjee, Dilip (2006), ‘Decentralization, Hierarchies, and Incentives: A Mechanism Design Perspective’, Journal of Economic Literature, 44 (2), 367–90. Nelson, Richard R. and Sidney G. Winter (1982), An Evolutionary Theory of Economic Change, Cambridge, MA: Belknap Press of Harvard University Press. Nevins, A. and F. Hill (1954), Ford, Vol. 1: The Times, The Man, The Company, New York: Scribner’s. O’Driscoll, Gerald P., Jr. and Mario J. Rizzo (1985), The Economics of Time and Ignorance, Oxford, UK: Blackwell. Penrose, Edith T. (1959), The Growth of the Firm, Oxford: Basil Blackwell. Pinchot, Gifford, III (1985), Intrapreneuring, New York: Harper & Row. Polanyi, Michael (1962), Personal Knowledge: Towards a Post-critical Philosophy, New York: Harper & Row. Putterman, Louis and Randall S. Kroszner (eds) (1996), The Economic Nature of the Firm: A Reader, Second Edition, Cambridge: Cambridge University Press. Rajan, Raghuram and Julie Wulf (2006), ‘The Flattening Firm: Evidence on the Changing Nature of Firm Hierarchies from Panel Data’, Review of Economics and Statistics, 88 (4), 759–73.

Bibliography

137

Rajan, Raghuram and Luigi Zingales (2001), ‘The Firm as a Dedicated Hierarchy: A Theory of the Origins and Growth of Firms’, Quarterly Journal of Economics, 116 (3), 805–51. Richardson, G.B. (1972), ‘The Organisation of Industry’, Economic Journal, 82, 883–96. Ricketts, Martin (2002), The Economics of Business Enterprise, Third Edition, Cheltenham, UK and Northampton, MA, USA: Edward Elgar. Rivkin, Jan and Nikolaj Siggelkow (2003), ‘Balancing Search and Stability: Interdependencies among Elements of Organizational Design’, Management Science, 49 (3), 290–311. Sautet, Frederic (2000), An Entrepreneurial Theory of the Firm, London: Routledge. Schumpeter, Joseph ([1934] 1993), The Theory of Economic Development, New Brunswick, NJ: Transaction Publishers. —— (1942), Capitalism, Socialism and Democracy, New York: Hanper Torchbooks. —— ([1947] 1989), ‘The Creative Response in Economic History’, in R.V. Clemence (ed.), Essays on Entrepreneurs, Innovations, Business Cycles, and the Evolution of Capitalism, New Brunswick, NJ: Transaction Publishers. —— ([1949] 1989), ‘Economic Theory and Entrepreneurial History’, in R.V. Clemence (ed.), Essays on Entrepreneurs, Innovations, Business Cycles, and the Evolution of Capitalism, New Brunswick, NJ: Transaction Publishers. Semler, Ricardo (1989), ‘Managing without Managers,’ Harvard Business Review, September–October, 76–84. Shackle, G.L.S. (1988), Business, Time and Thought, New York: New York University Press. Shipper, Frank and Charles C. Manz (1998), ‘W.L. Gore & Associates, Inc. in 1998’, in Arthur A. Thompson and A.J. Strickland III (eds), Strategic Management: Concepts and Cases, Tenth Edition, New York: Irwin/McGraw Hill Publishing. Simon, H.A. (1957), Models of Man: Social and Rational, London: John Wiley. —— (1961), Administrative Behavior, Second Edition, New York: Macmillan. —— (1991), ‘Organizations and Markets’, Journal of Economic Perspectives, 5 (2), 25–44. Sliwka, Dirk (2001), ‘On the Costs and Benefits of Delegation in Organizations’, Journal of Institutional and Theoretical Economics, 157 (4), 568–90.

138

Bibliography

Stein, Jeremy (2002), ‘Information Production and Capital Allocation: Decentralized versus Hierarchical Firms’, Journal of Finance, 57 (5), 1891–921. Vaughn, Karen I. (1987), ‘The Invisible Hand’, in John Eatwell, Murray Milgate, and Peter Newman (eds), The New Palgrave: A Dictionary of Economics, London: Macmillan. —— (1994), Austrian Economics in America: The Migration of a Tradition, Cambridge: Cambridge University Press. Wieser, Friedrich von ([1926] 1983), The Law of Power, Lincoln, NE: Bureau of Business Research, University of Nebraska. Williamson, Oliver E. (1975), Markets and Hierarchies: Analysis and AntiTrust Implications, New York: Free Press. —— (1985), The Economic Institutions of Capitalism: Firms, Markets, Relational Contracting, New York: Free Press. —— (1986), Economic Organization, New York: New York University Press. —— (1991), ‘The Logic of Economic Organization’, in O.E. Williamson and S.G. Winter (eds), The Nature of the Firm: Origins, Evolution and Development, Oxford: Oxford University Press. —— (1996), The Mechanisms of Governance, New York: Oxford University Press. —— (2000), ‘Strategy Research: Competence and Governance Perspectives’, in Nicolai Foss and Volker Mahnke (eds), Competence, Governance, and Entrepreneurship, Oxford: Oxford University Press. Witt, Ulrich (1998), ‘Imagination and Leadership: The Neglected Dimension of an Evolutionary Theory of the Firm’, Journal of Economic Behavior and Organization, 35, 161–77. Zabojnik, Jan (2002), ‘Centralized and Decentralized Decision Making in Organizations’, Journal of Labor Economics, 20 (1), 1–22.

Index Note: Page numbers in bold indicate main entry 3M 100, 101, 102, 106, 107, 112

Coasian firm 21, 58 Colombo, Massimo 90 command-and-control 2, 43, 92, 103 communication distortion 32, 34 see also Williamson, Oliver competition 9, 13, 20, 27, 39, 42, 43, 44, 64, 68, 77, 86, 114, 125, 130 contract 20 22, 24, 26, 27, 28, 29, 30, 31, 35, 36, 40, 44, 50, 57, 88, 91, 95, 96, 111, 113, 121 contractual hazards 31 corporate culture 89, 91, 106, 108, 113 corporate personal consumption 33 corporate slack 11, 12, 101 cost minimization, cost-cutting 9, 13, 44 Cowen, Tyler 42, 69, 73, 87, 88 customer/consumer 5, 8, 9, 14, 41, 43, 58, 66, 76, 77, 78, 80, 88, 92, 100 Cyert, Richard 40

agency problem/agency theory see principal–agent problem Aghion, Philippe 65, 66, 67, 68 Alchian, Armen 20, 24, 28, 29, 57, 93, 111 Aldikacti, Hulki 107 Aoki, M. 89 Asea-Brown-Boveri 53, 63 asset-specificity 27, 28, 33, 35 asymmetric information 54, 55, 57, 65, 66, 111 asymmetric knowledge see knowledge, asymmetric Austrian school 4, 22, 42, 43, 44, 45, 46, 50, 51, 56, 62, 63, 72, 88, 90, 112, 114, 117, 126 Bain, J.S. 20 Bartlett, F.C. 40 Becker, Gary 41 bilateral dependency 27 see also Williamson, Oliver black box of the firm 21, 31, 43 Boettke, Peter 42 Bohnet, I. 111 bounded rationality see rationality, bounded capital 67, 72, 73, 76, 122, 123 capital, human 41, 89, 114, 115, 124 Casson, Mark 1 central planning 4, 16, 22, 23, 24, 25, 40, 42, 43, 63, 76, 87 Chandler, Alfred 46, 65 Coase, Ronald 4, 5, 16, 20, 21–4, 25, 26, 30, 32, 34, 39, 40, 42, 43, 44, 45, 46, 47, 48, 54, 55, 57, 69, 71, 72, 73, 77, 78, 80, 84, 88, 90, 91, 92, 112

De Paola, Maria 90 De Vecchi, Nicolo 119 decision rights, decision-making rights, rights of control 15, 66, 72, 76, 77, 78, 79, 80, 81, 82, 87, 90, 115, 121 decision-maker 2, 11, 37, 59, 61, 67, 77 delegation 3, 4, 15, 17, 65–8, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 89, 90, 91, 94, 95, 99, 114, 121–3 Delmastro, Marco 90 Deming, William 62 Demsetz, Harold 24, 57, 93, 111 Dessein, Wouter 90 diminishing returns to management 23 discovery 12, 18, 22, 41, 44, 45, 50, 56, 86, 91, 113, 124, 125, 126, 130 double Hayekian knowledge problem see knowledge problem, double Hayekian

139

140

Index

Ebeling, Richard 17, 93, 103, 104, 105, 106, 110, 112, 117, 118, 119, 120, 122, 123, 125, 126, 129 efficiency 30, 35, 39, 71, 73, 76, 112, 123 eidetic imagination 125, 126 Einstein, Albert 114 Ellig, Jerry 42, 87 employee stock ownership plan 91, 131 entrepreneur-coordinator/ entrepreneur-promoter 4, 21, 22, 23, 52, 57, 58, 74, 88 entrepreneurship as creativity in action 18, 114, 116 epistemic conditions 42, 56 epistemology 8, 42, 51, 52, 60 equilibrium/disequilibrium16, 36, 41, 71, 73, 75, 90, 91 evolution, of firms, of rules 10, 11, 46, 47, 49, 52, 76, 86, 104, 110, 117 evolutionary-entrepreneurial stimulus 86 see also Schumpeter, Joseph expectations 32, 103, 105, 112, 117, 124 experience 31, 33, 38, 39, 56, 70, 80, 87, 104, 107, 117, 118, 119, 120, 125, 127, 129 experiment/trial-and-error 3, 4, 5, 7, 8, 9, 11, 12, 14, 16, 31, 36, 39, 40, 61, 70, 76, 80, 84, 85, 98, 102, 124, 128 fairness/reciprocity 32, 98, 99, 129, 131 flattening see structure, flat flexibility 9, 13, 30, 34, 50, 60, 61, 63, 64, 65, 68, 69, 70, 71, 105, 108, 109, 113, 117, 123, 127, 128 flexibility–efficiency tradeoff 71, 85 Ford, Henry 91 Foss, Kirsten 121, 122, 123 Foss, Nicolai 13, 14, 15, 16, 17, 39, 40, 43, 44, 45, 50, 53, 54, 55, 56, 61, 62, 63, 69, 71, 72–85, 88, 89, 90, 91, 109, 112, 113, 121, 122, 123 franchising 20, 36, 41, 88 free-riding/free-rider 10, 11, 94, 96, 102 Frey, B. 98, 111 Fry, Art/Post-it 100, 101, 112 see also 3M functional theory of entrepreneurship 115 see also Schumpeter, Joseph

Fundamental Transformation 27, 29 see also Williamson, Oliver Gable, Wayne 42, 87 game theory 81 Garrouste, Pierre 72 General Motors 62, 107 Google see Schmidt, Eric/Google Gore, Bill/W.L. Gore & Associates 5, 6, 7, 8, 13, 14, 15, 45, 65, 75, 83, 108 governance structures 27, 30, 31, 40, 49, 88, 91, 9 see also Williamson, Oliver Grossman, Sanford J. 35 groupthink 129 Harper, David A. 49, 88, 91, 125 Hart, Oliver D. 35, 65 Hayek, F.A. 37, 39, 41, 42, 44, 45, 48, 49, 51, 52, 55, 68, 70, 74, 122 see also knowledge problem Hill, F. 91 Holmstrom, Bengt 65 Husserl, Edmund 125 hybrids 20, 30, 36, 37, 39, 40, 41, 49, 73, 81, 88, 91 IBM 64 ideal types 104, 112, 118, 119, 129 imagination 8, 86, 119–20, 122, 125, 126 information, hard 67 information, soft 67, 68 integration, vertical 32, 33, 34, 35, 40, 88 interdependence/teamwork 62, 94, 99, 102, 103, 106, 113, 127, 130 interpretation/interpretive process 70, 75, 104, 112, 117, 118, 119, 120, 122, 123, 125, 129 intersubjective meaning 17, 70, 104, 105, 106, 108, 110, 113, 117, 119, 125, 129 intrapreneurship 11, 12, 55, 106, 107 invisible hand 9, 25 Jensen, Michael 65, 66, 67, 68, 90 joint ventures 20, 36, 88

Index judgment 44, 45, 70, 80, 88, 121–3, 128, 129 derived 121, 122, 123, 129 entrepreneurial 44, 45, 86 original 121, 122, 129 Kahneman, D. 111 Kirzner, Israel 18, 41, 42, 91, 124, 125, 126 Klein, Peter 20, 72, 121, 122, 123 Knetsch, J. 111 Knight, Frank 22, 121 knowledge asymmetric 57, 61, 71, 73, 78, 79, 87, 89, 128 see also double Hayekian knowledge problem dispersed 39, 42, 43, 44, 50, 52, 53, 54, 55, 56, 59, 60, 61, 62, 63, 69, 71, 85, 86, 89, 90, 114, 122, 123, 128 distributed 55, 78, 79, 113 division of 23, 118, 119, 128 empirical 56, 70 firms 53, 54, 55, 73 heterogeneity of 68 hidden 80 idiosyncratic 37, 38, 39, 41, 128 problem 1, 16, 18, 37, 42, 42–87, 95, 128 double Hayekian 58, 59, 65, 71, 73, 87, 88, 89, 128 see also asymmetric information/asymmetric knowledge subjective 41, 65, 68 tacit 38,41, 42, 43, 55, 56, 63, 68, 70, 88 task-specific 37 Koch, Charles G./Koch Industries 90, 124, 126 Kohak, Erazim 125 Kolind, Lars 13, 14, 61, 82, 109, 113 see also Oticon A/S Kreps, D.M. 113 Kroszner, Randall S. 20 Lachmann, Ludwig 70, 105, 113 Langlois, Richard 17, 43–51, 64, 69,

141

71, 72, 85, 88, 90, 100, 111, 113, 123 lattice organization 6, 13, 65, 108 see also spaghetti organization Lavoie, Donald C. 42 Lehr, Lew 101 see also 3M loyalty 96, 97, 98, 99, 102, 129, 131 Malmgren, Harald 113 Malone, Thomas 2, 114 management lower 2, 3, 67, 77, 78, 91, 115, 123, 124 middle 2, 3, 59, 84, 109 upper 67, 118 Manz, Charles C. 1, 19, 65, 108 March, James 40 Market-Based Management see Koch, Charles G./Koch Industries Marshak, Jacob 41 Marshall, Alfred 41, 111 McKinsey Consulting 5 Meckling, William 65, 66, 67, 68, 90 Menger, Carl 44, 45, 46 M-form 18, 46, 86, 89, 124 Minkler, Alanson 17, 43, 44, 45, 54, 56, 57, 58, 59, 60, 62, 63, 64, 72, 73, 78, 85, 87, 89, 97, 98, 99, 100, 111, 127 Mises, Ludwig von 4, 5, 63, 74, 76, 77, 78, 90, 91, 121, 124, 126 Modern Economics of Organization 44, 54, 55, 62, 88, 89 see also Coase, Ronald and Williamson, Oliver monitoring/supervisory enforcement/ oversight 3, 10, 17, 28, 29, 35, 57, 73, 78, 87, 89, 93, 94, 95, 96, 97, 98, 102, 106, 11, 127, 130 Mookherjee, Dilip 90 Moore, John 65 Morton, Steve/Tektronix 10 Nelson, Richard R. 1, 40, 46 Nevins, A. 91 New Institutional Economics/ Transaction Cost Economics 16, 20, 25, 30, 31, 39, 40, 55, 69, 71, 87, 93, 97, 98, 103, 106, 111, 128

142

Index

norms 75, 100, 111, 112 see also rules, informal O’Driscoll, Gerald P. 113 opportunism 2, 17, 26, 27, 28, 29, 30, 33, 35, 37, 38, 39, 65, 69, 81, 84, 87, 92, 93, 97, 110, 111, 113, 121, 123, 128, 129 Ore-Ida 11, 12 organic order/institutions/organization 45, 46, 47, 48, 49, 50, 51, 52, 55, 64, 70 Oticon A/S 13, 14, 15, 60, 61, 75, 76, 82, 83, 91, 109, 113 see also Kolind, Lars Palmquist, Phil 106 Parker, David 42, 69, 73, 74, 87, 88 path-dependence/routines 39, 47 Penrose, Edith 40 Polanyi, Michael 41, 126 Pontiac Fiero 107 Post-it see Fry, Art/Post-it pragmatic organizations 45, 47, 51, 55 price 5, 14, 17, 21, 22, 23, 24, 27, 29, 37, 41, 42, 55, 60, 73, 75, 91, 103, 117, 118, 125 pricing system 23, 25, 76, 90 pride 96, 110 principal–agent problem 17, 32, 33, 54, 55, 57, 58, 65, 66, 67, 73, 77, 78, 80, 87, 89, 90, 93, 94 profit center 3, 19 profit sharing 59, 70, 110, 111, 123, 131 property rights 73, 103 Putterman, Louis 20 Rajan, Raghuram 2, 18, 68, 90, 114 rationality 26, 98 bounded 26, 27, 28, 29, 30, 31, 37, 69, 86 hyperrationality 26 subgroup 32 see also Williamson, Oliver systems 32 see also Williamson, Oliver reciprocal trading 36, 88 reciprocity see fairness/reciprocity reputation/reputational effects 30, 35, 81, 82, 84, 111

residual claimant 35, 93, 111 responsibility 3, 4, 5, 10, 19, 44, 62, 95, 108, 127 Richardson, G.B. 40 Ricketts, Martin 83, 89, 125 Rivkin, Jan 90 Rizzo, Mario J. 113 Roberts, John 65 Robertson, D.H. 22 routines 40, 47, 55, 58, 89, 119, 120, 124 rules 10, 11, 12, 45, 46, 47, 48, 49, 50, 51, 52, 53, 69, 70, 86, 87, 110, 111, 112, 131 formal 10, 105, 120, 131 informal 10, 64, 110, 120, 131 Saturn 62 Sautet, Frederic 18, 23, 28, 41, 43, 44, 45, 52, 53, 56, 57, 58, 60, 65, 72, 73, 85, 86, 87, 88, 89, 91, 124, 130 Schelling, Thomas 113 Schmidt, Eric/Google 109 Schumpeter, Joseph 1, 17, 18, 20, 22, 86, 89, 90, 115, 116, 119, 120, 122, 123, 124, 125, 126 Schutz, Alfred 17, 87, 95, 103, 104, 105, 112, 113, 129 Scoppa, Vincenzo 90 selective intervention 34, 35, 36, 40, 67, 79, 80, 81, 82, 83, 84, 113 self-interest 26, 33, 34, 40, 66, 83, 99, 102 self-organization 2, 5 Semco S/A 59, 60, 82, 83 Semler, Ricardo see Semco S/A Shackle, G.L.S. 125 Shipper, Frank 1, 19, 65, 108 shirking 8, 9, 10, 89, 94, 98, 99, 102 Siggelkow, Nikolaj 90 Simon, Herbert 17, 26, 28, 93, 94, 95, 96, 97, 98, 99, 100, 106, 111, 113 Sliwka, Dirk 90 small-numbers condition 27, 28 socialism 53, 55, 62, 74 socialist calculation debate 32, 55 spaghetti organization 13, 14, 61, 91 see also Oticon A/S; Kolind, Lars; lattice organization

Index

143

Stein, Jeremy 65, 67, 68 structure, flat 6, 7, 15, 18, 59–68, 83, 84, 108, 114, 115, 127 structure, multidivisional 18, 72, 86, 89 see also M-form Structure–Conduct–Performance 20 structures of shared meaning 105, 107, 109, 110, 118, 129, 130

Transaction Cost Economics see New Institutional Economics Trump, Donald 112

Taylor, Frederic Winslow 28, 45 teams 6, 7, 11, 13, 14, 75, 89, 99, 106, 107, 109, 114, 125 team production 62, 70 teamwork 62, 102, 106, 127, 130 Tektronix see Morton, Steve/ Tektronix Texas Instruments 12 Thaler, R. 111 Tirole, Jean 65, 66, 67, 68 tit-for-tat strategy 81, 82, 83, 99 transaction cost 17, 18, 22, 23, 29, 32, 33, 39, 42, 43, 44, 54, 58, 69, 71, 92, 94, 111, 123

Weber, Max 112 Welch, Jack 1 Wieser, Friedrich von 92, 111 Williamson, Oliver 16, 17, 20, 21, 24–41, 42, 43, 44, 45, 47, 48, 54, 55, 69, 80, 81, 84, 88, 91 Winter, Sidney G. 1, 40, 46 Witt, Ulrich 72 W.L. Gore & Associates see Gore, Bill/W.L. Gore & Associates Woodward, Susan 28, 29 Wulf, Julie 2, 18, 68, 90, 114

uncertainty 22, 29, 30, 37, 49, 56, 57, 58, 63, 89, 100, 121, 122, 124, 129 Vaughn, Karen 9 venture capital 11, 64, 88, 101

Zabojnik, Jan 90