Adaptation, specialization, and the theory of the firm foundations of the resource based view

Cambridge University Press978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm Birger Wernerfelt Frontmatter More Information Figures 2.1 Most efficient mechanisms

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

in the field of management and economics Wernerfelt revisits hisclassic articles, including an extensively revised “A Resource-BasedView of the Firm” (1984), which have been updated and synthesized

to provide precise and accessible concepts and predictions By offeringfuture directions for research and practice, this book will be of interest

to students and scholars of management and economics alike

birger wernerfeltis the J C Penney Professor of Management atthe MIT Sloan School of Management

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

Adaptation, Specialization, and the Theory of the Firm

Foundations of the Resource-Based View

Birger Wernerfelt

MIT Sloan School of Management

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Information on this title: www.cambridge.org/9781107595781 DOI: 10.1017/9781316466872

© Birger Wernerfelt 2016 This publication is in copyright Subject to statutory exception and to the provisions of relevant collective licensing agreements,

no reproduction of any part may take place without the written permission of Cambridge University Press.

First published 2016 Printed in the United Kingdom by Clays, St Ives plc

A catalogue record for this publication is available from the British Library Library of Congress Cataloging-in-Publication Data

Names: Wernerfelt, Birger, author.

Title: Adaptation, specialization, and the theory of the firm : foundations of the resource-based view / by Birger Wernerfelt, J C Penney Professor of Management, MIT Sloan School of Management.

Description: Cambridge, United Kingdom : Cambridge University Press, [2016] | Includes index.

Identifiers: LCCN 2016017702| ISBN 9781107134409 (Hardback) | ISBN 9781107595781 (Paperback)

Subjects: LCSH: Industrial organization (Economic theory) | Economic specialization | Diversification in industry.

Classification: LCC HD2326 W467 2016 | DDC 338.5–dc23 LC record available at https://lccn.loc.gov/2016017702

ISBN 978-1-107-13440-9 Hardback ISBN 978-1-107-59578-1 Paperback Cambridge University Press has no responsibility for the persistence or accuracy

of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

For Bjoern

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

3 Adaptation Costs in One Dimension: Firms, Contracts,

4 Adaptation Costs in Three Dimensions: Firms, Markets,

5 All Adaptations Are Not the Same: The Scope of Firms

vii

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

12 Adaptation Frequency and the Boundary of the Firm 201

16 On the Endogenous Amplification of Small Differences 284

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

Figures

2.1 Most efficient mechanisms and changing needs 1 page18

4.1 Most efficient mechanisms by frequency of change and

4.2 Most efficient mechanisms by frequency of change and

4.3 Most efficient mechanisms by frequency of change

8.1 Conditional posterior density of clif the seller did not

8.2 Conditional posterior densities of clif seller found, but

8.3 Conditional posterior densities of chif seller offered the

11.1 Schematic screen-shot of the bargaining environment

11.A-1 Frequency of settling times by experimental treatment 200

14.1 {k, r │The buyer will cede decision d to the seller} 265

ix

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

4.A-1 Components of efficiency and performance

12.3 Probability of pair-wise co-production by adaptation

12.A-1 Adaptation frequencies from interviews and

12.A-3 Probability of pair-wise co-production by

12.A-4 Sum of internalized adaptation frequencies by

x

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

12.A-6 Sum of squared production system residuals (SSR)

15.1 Number of rounds when one member uses code A

15.2 Number of rounds when both members use code A 278

15.3 Number of rounds when both members use code B 278

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

Preface

This book is based on fifteen papers written over a thirty-year period

Two of them, “A Resource-Based View of the Firm” (proposing what is

now called the RBV) and “On the Nature and Scope of the Firm”

(proposing what is here called the “Adaptation Cost,” or just AC, theory

of the firm1) were written a few years apart in the 1980s While both

papers aim to characterize the sets of activities that should and should not

be undertaken by firms, I long considered them incompatible One was

about excess capacity of certain productive assets and the other was

about the costs of adapting contracts Only when writing what is now

Chapter 4 did I realize that the sub-additive bargaining costs underlying

the AC theory could make productive assets de facto indivisible and

capable of existing in excess capacity, thus taking me directly to the RBV

The main purpose of the book is to clarify the AC–RBV relationship

and this is done in two steps In “Adaptation Costs in Three Dimensions:

Firms, Markets, and Contracts,” I derive the AC theory under gains

from specialization, and in “All Adaptations Are Not the Same: The

Scope of Firms and the Size of Markets,” I develop the resulting theory

of the scope of the firm and compare it to the RBV To make the linkage

as clear as possible, the four key chapters have been placed in sequence

at the start of the book The remaining eleven chapters look,

theoreti-cally and empiritheoreti-cally, at important implications and foundations of

the AC/RBV

Most chapters have their basis in a single stand-alone paper, some

published several years ago All have been rewritten for consistency

of notation and terminology and updated to reflect other literature

published between their original publication date and now I do not flag

the parts that are most extensively rewritten, but the reader should be

aware that there may be significant differences between the original

1 In its first incarnation, the theory was, at the suggestion of a referee, labelled the

“Adjustment Cost” Theory However, that term is also used in the macro literature

where it denotes a different phenomenon.

xiii

Cambridge University Press

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Since the ideas in the book have been developed over such a long time,

I have benefited from a very large number of colleagues, readers, seminaraudiences, referees, and editors The most important and consistent ofthese have been Bob Gibbons, Oliver Hart, and Steve Tadelis I havelearned a lot from them and the work is profoundly influenced by theirwillingness to repeatedly discuss on my premises I would also like tothank Tore Ellingsen, Nicolai Foss, Drew Fudenberg, Bengt Holm-strom, Patrick Legros, Jin Li, Niko Matouschek, Roger Myerson, MikePowell, Peter Norman Soerensen, Jean Tirole, Mike Whinston, andOliver Williamson for one or more discussions, questions, or commentsthat ended up being important to some of the ideas presented here Moregenerally, I would like to thank everybody who has been a member of thevibrant Organizational Economics community in Cambridge at any pointduring the last twenty-five years

Three of the chapters are based on papers coauthored with colleaguesand I would like to acknowledge the collaborations of Boris Maciejovsky,Sharon Novak, and Duncan Simester A couple of papers would not havebeen possible without the help of research assistants and I particularlywant to thank Carol Meyers for her help with what is now Chapter 12

In the later stages of the manuscript preparation, Xinyu Cao providedexcellent proofreading and checking assistance

As mentioned above, most of the Chapters are based on previouslypublished work Chapter 2 is based on “Small Forces and Large Firms:Foundations of the RBV” (Strategic Management Journal, 34, no 6,

pp 635–43, 2013); Chapter 3 is based on “On the Nature and Scope

of the Firm” (Journal of Business, 70, no 4, pp 489–514, 1997); ter 4 is based on “The Comparative Advantages of Firms, Markets, andContracts” (Economica, 82, no 236, pp 350–67, 2015); Chapter 6 isbased on “A Resource-Based View of the Firm” (Strategic ManagementJournal,5, no 2, pp 171–80, 1984); Chapter 7 is based on “Why Shouldthe Boss Own the Assets?” (Journal of Economics and Management Strat-egy, 11, no 3, pp 473–85, 2002); Chapter 8 is based on “BargainingBefore or After Communication?” (Journal of Institutional and TheoreticalEconomics, 164, no 2, pp 211–29, 2008); Chapter 9 is based on “RobustIncentive Contracts” (Journal of Institutional and Theoretical Economics,

Chap-160, no 4, pp 545–54, 2004); Chapter 10 is based on “Delegation,

978-1-107-13440-9 — Adaptation, Specialization, and the Theory of the Firm

Birger Wernerfelt

Frontmatter

More Information

Committees, and Managers” (Journal of Economics and Management

Strategy, 16, no 1, pp 35–51, 2007); Chapter 11 is based on “Costs of

Implementation: Bargaining Costs Versus Allocative Efficiency” (with

Boris Maciejovsky in Journal of Economic Behavior and Organization, 77,

no 3, pp 318–25, 2011); Chapter 12 is based on “On the Grouping of

Services into Firms: Make-or-Buy with Interdependent Parts” (with

Sharon Novak in Journal of Economics and Management Strategy, 21,

no 1, pp 53–77, 2012); Chapter 13 is based on “Determinants of

Asset Ownership: A Study of the Carpentry Trade” (with Duncan I

Simester in Review of Economics and Statistics, 87, no 1, pp 50–58,

2005); Chapter 14 is based on “Renegotiation Facilitates Contractual

Incompleteness” (Journal of Economics and Management Strategy, 16,

no 4, pp 893–910, 2007); Chapter 15 is based on “Organizational

Languages” (Journal of Economics and Management Strategy, 13, no 3,

pp 461–72, 2004); and Chapter 16 is based on “Why Do Firms Tend to

Become Different?” (Chapter 9, pp 121–33 in Constance Helfat (ed.),

Handbook of Organizational Capabilities,Malden, MA and Oxford, UK:

Blackwell, 2003) I am grateful for the respective publishers for granting

permission to use their articles in this volume and to Jason Clinkscales

of MIT for collecting those permissions

I would like to thank the University of Michigan, Northwestern

University, the University of Copenhagen, and particularly MIT for

providing me with stimulating work environments And last, but not

least, Paula Parish of the Cambridge University Press for believing in

the project and helping me through the writing and publication process

Part I

Agenda

1 Introduction

“A Resource-Based View of the Firm” (RBV), a paper I wrote in 1984,has gone on to become very influential in the management literature Itintroduced ideas that are taught in strategy, personnel, marketing, andoften several otherfields, in virtually every MBA program in the world.The implications resonate with practicing managers, have intuitiveappeal, and are easy to apply At the same time, the paper has spurned

a vigorous academic debate about its theoretical foundations (Foss,1998; Priem and Butler, 2001)

Meanwhile, another foundational debate, over what exactly a“firm” is,has been raging in economics Although two Nobel prizes have beenawarded for answers to this question, the only agreed-upon proposition isthat we, as of 2016, do not have a commonly accepted theory of thefirm.Drawn to foundational topics, I have struggled with the question over thelast thirty years, gradually refining the “Adaptation Cost” (AC) theoryand making several other contributions to the economic literature on thefirm While these papers are written for an economics audience, the mainforces at work will seemfirst order to managers, and it turns out that thetheory portraysfirms as acting in accordance with the RBV

In contrast to the influence enjoyed by the RBV, my economic papershave remained on the fringes of the debate in thatfield There are at leasttwo reasons for this First, economists rightly have very conservativebeliefs about what constitutes first order effects Second, I failed tostress to the profession that the theory rings true to the people modeled–those who make decisions about the scope of realfirms In my defense, it

is not an easy case to make Economists value that kind of externalvalidity in a theory, but most know relatively little about how top man-agers think about the scope of thefirm, and find it hard to interpret thefacts in a way that allows them to bridge the gap to theory My hope isthat the RBV can serve as an intermediate point on such a bridge

Taken as a whole, the collection of papers contained in this volumecould thus contribute to two debates, about the foundations of the RBVand the theory of the firm At the moment, almost no management

scholars know about my work in economics and economists are often

unfamiliar with the RBV

Compared to other recent theories of the firm, the essential

assump-tion behind the AC theory is that bargaining costs, incurred when

adapting contracts, are sub-additive This leads to a concept of

employ-ment as a cost-minimizing mechanism and explains why labor inputs

may be economically indivisible The other important assumption, which

is much less controversial, is that workers are more efficient when they

specialize There are advantages of specialization in both business and

service dimensions and this means that production costs are higher when

workers have to change from one business and/or service to another

These two types of adaptation costs, contractual and productive, are

both incurred when demand shortages make complete specialization

impossible, and the AC theory depicts economic institutions as attempts

to minimize the total costs of adaptation In particular, the employment

relationship, which is used to define the firm, is found to be more

efficient when frequent adaptations are needed A firm has incentives

to expand its scope when sub-additive bargaining costs makes it most

efficient to realize gains from specialized use of inputs inside the firm.This means that the firm can enhance efficiency by expanding intobusinesses that are new but“related” in the sense that the same expertise,

in business and service dimensions, can be used with little loss of efciency in all of them The RBV looks at the special case in which

fi-expertise differs betweenfirms, but the AC theory does not require that.1Since the paper forming the basis for Chapter 2, “Small Forces and

Large Firms: Foundations of the RBV,” contains a sketch of the ment made in this book, I will use the present chapter to place the

argu-individual components in the context of previous and current literature

1.1 Strategy, Human Resource Management, and Marketing

“A Resource-Based View of the Firm” did not, as some have thought,

start as a generalization of Edith Penrose’s (1959) theory about firms

growing to utilize excess capacity of managerial time Rather, it was a

reaction, inspired by game theory, to Michael Porter’s “five forcesanalysis” of industries (1980) My thinking was simply that it is norma-tively inconsistent to recommend the same industries to everybody

Since competition will erode profits, recommendations can only be

consistent with equilibrium if they send different firms in differentdirections Seen in a broader perspective, this reaction and the timing

of it were not surprising at all: Porter’s book followed the lines of JoeBain’s (1956) Industrial Organization economics and this branch ofeconomics was, in the mid-eighties, being turned upside down by econo-mists newly armed with the tools of game theory (e.g., Tirole, 1988).2

To generate different operational recommendations for differentfirms,

I took as a starting point thatfirms have different productive assets andthat this matters for their production possibilities More precisely,

I assumed that firms differ in terms of their economically inalienableproductive assets – their resources A productive asset is economicallyinalienable from afirm f if it is inefficient to have any part of its capacityused by afirm other than f Sub-additive price determination costs maymake it inefficient to trade fractions of an asset, such that it cannot be

efficiently exploited by trade, although it could be used by the firm itself

It is not required that the firm use 100 percent of the asset’s capacity,only that it is inefficient to have a fraction of the capacity used by anotherfirm So these are the assets of which the firm may have excess capacityeven in equilibrium.3

The RBV paper was not an immediate success and I did not promote it

or even cite it in thefirst several years However, things changed whenthree later papers proposed closely related ideas with more or less differ-ent terminology: Prahalad and Hamel (1990) on Core Competencies,Barney (1991) on Resource-Based Theory, and Teece, Pisano, andShuen (1997) on Dynamic Capabilities These papers all share theassumption thatfirms are different and suggest that they should leverageexcess capacity of what they are good at The strategy community real-ized the similarities, and the success of my paper is in no small measuredue to the“sales job” done by these other authors

These similarities notwithstanding, the RBV paper differs significantlyfrom the later contributions in its treatment of thefirm’s development ofnew resources This is a tricky issue because symmetric competitionwould result in rents from this type of innovation being competed away

In the RBV paper, the initial asymmetry in resources creates asymmetricpositions in both output and resource markets, and the firm should

Strategy, Human Resource Management, and Marketing 5

leverage both of these So the result is thatfirms should do what they aregood at and get better at things they are good at getting better at In

Teece, Pisano, and Shuen (1997), it is posited that some firms have aspecial resource that allows them to be good at learning (thus the name

“dynamic capabilities”) It may be hard to empirically distinguish

between the two theories, but the fact that firms tend to grow withinthe same general area would seem to suggest that any generalized learn-

ing somehow is tied to thefirm’s initial resources

Very soon after the RBV had diffused into strategy, it was applied to

human resource management (personnel), and is now often thought of as

forming the foundation of that field For example, Allen and Wright

(2007, p 90) write that the RBV is “the guiding paradigm on which

virtually all strategic HRM [human resource management] research

is based.”

After a brief delay, the ideas made quick inroads into marketing and

now play a substantial role in thatfield – though they are less central therethan in strategy and personnel Specifically, in very few years, the dom-

inant conceptual framework in thatfield changed from being the “4P’s”(Product, Place, Promotion, and Price) to being the“3C’s and the 4P’s.”The idea is that the C’s (Company, Customers, and Competitors),

are more or less given, and that the P’s are contingent on them I do

not know why this change occurred, and but the factors raised under

“Company” are typically those that would be suggested by the RBV

Even the terminology is similar, as in the“Resource-Advantage Theory”

of Hunt and Morgan (1996), and in often used labels like“Capabilities”

or“Competencies.”

1.2 Economics: Sub-Additive Bargaining Costs

The concept of bargaining costs have proved vexing for the economics

profession One reason for this is that it has been hard to explain, in a

fully satisfactory way, why bargaining should be costly One celebrated

attempt, by Grossman and Hart (1986) and Hart and Moore (1990) was

undone by a set of sophisticated mechanisms proposed by Maskin and

Tirole (1999) Specifically, players can eliminate the bargaining ciencies exploited by Hart et al by using the so-called Moore-Rupello

ineffi-(1988) mechanisms (Yes, it is the same Moore.) Other justificationshave been based on Myerson and Satterthwaite’s (1983) result thatbargaining with two-sided incomplete information necessarily is ineffi-cient if a player cannot be forced to trade at a loss However, if the parties

can write a binding contract ex ante, they can commit themselves to trade

no matter what

There are many ways to confront this problem, none of which is fullysatisfactory Among others, we have proofs that the Moore-Rupellomechanisms are not robust (Aghion, Fudenberg, Holden, Kunimoto,and Tercieux, 2012), postulates that contracting is costly per se (Bajariand Tadelis, 2001), and appeals to bounded rationality (Williamson,1979; Hart and Moore, 2008).

In this book, I justify the reliance on bargaining costs in severaldifferent ways First, I show, in a series of experiments, that they existand are sub-additive (Chapter 11) Second, I note that bargainingrequires communication and that this is costly per se (Chapter 3).Third, I argue that some bargainers will spend resources trying to learneach other’s private valuations (Chapter 4) More generally, it makes

a difference that I define the firm by the employment relationship

In most countries, the law protects labor from overly aggressive ties, thus allowing them to opt out of contracts if the conditionsturn out to be too onerous.4Once again, I admit that none of these isfully satisfactory

penal-These theoretical discussions aside, most people believe that ing costs exist However, another barrier to their widespread use is theintuition that they are in some sense “small” and thus cannot be firstorder causes of any important phenomena This is very reasonable if one

bargain-is trying to explain large capital investments (as bargain-is done in most of theliterature on integration), but much less so when, as in this book, thefocus is on large numbers of small labor services Making this point innumerous seminars, I have never had anyone disagree that“an executivedoes not want to negotiate a separate fee for every little service his or hersecretary performs.” The AC theory highlights how the governance

of trading relationships depends on the number of adaptations neededper unit time

While bargaining costs, in various guises, have been the subject of a lot

of debate, I have never seen any mention of the second part of thepremise, that these costs are sub-additive in the number of items covered

in the negotiation And yet, this is critical Without this assumption,occasion-by-occasion bargaining by the above-mentioned executivewould be just as efficient as a once and for all negotiation covering theentire set of relevant services– an arrangement we will call an “employ-ment relationship” and use to define the firm (Chapter 3)

Fortunately, sub-additivity is intuitively appealing: Compared to thirtyitem-by-item negotiations, most people would prefer to bargain once

4

It is harder to dispute that binding ex ante contracts can be written for trade in products.

about a price for the bundle of thirty items In fact, this is exactly what

Boris Maciejovsky and Ifind in the experiments reported in Chapter 11

of this book Based on countless conversations with economists, my

strong sense is that the profession considers the sub-additivity

assump-tion relatively unproblematic compared to the resistance to any use of

bargaining cost in thefirst place

After this historical/sociological preamble, the next Chapter contains a

preview of the book

ReferencesAghion, Philippe, Drew Fudenberg, Richard Holden, Takashi Kunimoto, and

Olivier Tercieux,“Subgame-Perfect Implementation Under Information

Pertubations,” Quarterly Journal of Economics, 127, 1843–81, 2012

Allen, Matthew R., and Patrick M Wright,“Strategic Management and HRM”

in Peter Boxall, John Purcell, and Patrick M Wright (eds.), Oxford

Handbook of Human Resource Management, Oxford, UK: Oxford University

Press, 2007

Bain, Joseph S., Barriers to New Competition, Cambridge, MA: Harvard

University Press, 1956

Bajari, Patrick, and Steven Tadelis,“Incentives Versus Transaction-Costs:

A Theory of Procurement Contracts,” RAND Journal of Economics, 32,

387–407, 2001

Barney, Jay B.,“Firm Resources and Sustained Competitive Advantage,” Journal

of Management, 17, no 1, March, 89–120, 1991

Conner, Kathleen R.,“A Historical Comparison of Resource-Based Theory andFive Schools of Thought within Industrial Organization Economics:

Do We Have a New Theory of the Firm?” Journal of Management, 17, no 1,March, 121–54, 1991

Foss, Nicolai J.,“The Resource-Based Perspective: An Assessment and

Diagnosis of Problems,” Scandinavian Journal of Management, 14, no 3,

March, 133–49, 1998

Foss, Nicolai J.,“Micro-Foundations for Resource-Based Theory,” Journal of

Management, 37, 1413–28, 2011

Gibbons, Robert, and Rebecca Henderson,“What Do Managers Do?” chapter

17 in R Gibbons and J Roberts (eds.), Handbook of Organizational

Economics, Princeton, NJ: Princeton University Press, 2013

Grossman, Sanford, and Oliver D Hart,“The Costs and Benefits of

Ownership,” Journal of Political Economy, 94, 691–719, 1986

Hart, Oliver D., and John Moore,“Property Rights and the Nature of the Firm,”Journal of Political Economy, 98, 1119–58, 1990

Hart, Oliver D., and John Moore,“Contracts as Reference Points,” Quarterly

Journal of Economics, 123, 1–48, 2008

Hunt, Shelby D., and Robert M Morgan,“The Resource-Advantage Theory of

Competition,” Journal of Marketing, 60, October, 107–14, 1996

Lippman, Steven A., and Richard P Rumelt,“Uncertain Imitability: An Analysis

of Interfirm Differences Under Competition,” Bell Journal of Economics, 13,418–38, 1982

Lockett, Andrew, Rory P O’Shea, and Michael Wright, “The Development of

the Resource-Based View: Reflections from Birger Wernerfelt,”

Organizational Studies, 29, 1125–41, 2008

Maskin, Eric, and Jean Tirole,“Unforeseen Contingencies and Incomplete

Contracts,” Review of Economic Studies, 66, 83–114, 1999

Moore, John, and Raphael Rupello,“Subgame Perfect Implementation,”

Econometrica, 56, 1191–1220, 1988

Myerson, Roger B., and Mark A Satterthwaite,“Efficient Mechanisms for

Bilateral Trading,” Journal of Economic Theory, 29, 265–81, 1983

Penrose, Edith, The Theory of the Growth of the Firm, New York, NY: Oxford

Rumelt, Richard P., Daniel Schendel, and David J Teece,“Strategic

Management and Economics”, Strategic Management Journal, 12, Special

Issue, 5–29, 1991

Sappington, David, and Birger Wernerfelt,“To Brand or Not to Brand?

A Theoretical and Empirical Question,” Journal of Business, 58, no 3, July,

279–93, 1985

Teece, David, Gary Pisano, and Amy Shuen,”Dynamic Capabilities and

Strategic Management,” Strategic Management Journal, 18, no 7, August,

509–33, 1997

Teece, David J.,“Towards an Economic Theory of the Multi-Product Firm,”

Journal of Economic Behavior and Organization, 3, no 1, March, 39–63, 1982.Tirole, Jean, The Theory of Industrial Organization, Cambridge, MA: MIT Press,1988

Wernerfelt, Birger,“A Resource-Based View of the Firm,” Strategic ManagementJournal, 5, 171–80, 1984

Wernerfelt, Birger,“Umbrella Branding as a Signal of New Product Quality: AnExample of Signaling by Posting a Bond,” RAND Journal of Economics, 19,

no 3, Autumn, 458–66, 1988

Williamson, Oliver E., Markets and Hierarchies: Analysis and Antitrust Implications,New York, NY: The Free Press, 1975

Williamson, Oliver E.,“Transaction-Cost Economics: The Governance of

Contractual Relations,” Journal of Law and Economics, 22, 233–61, 1979

Small Forces, High Frequencies, and Large Firms

2.1 Introduction

This book is aboutfirms Many readers will be teaching about firms and/

or doing research about them And yet, we have no generally accepted

answers to the most basic questions about them What is afirm? Whenshould we use them? How do they work differently than markets? Why do

some of them diversify?

Most research in business and economics is conducted by taking a

trading mechanism, typically a firm, a market, or a contract, as given

The researcher then imposes some subset of a standard set of

assump-tions, and answers questions about the resulting actions and outcomes

While this procedure has proven very fruitful, it raises some deeper

questions: What determines the choice of mechanism in the first place?And why arefirms, markets, and contracts so commonly used? On these,

we are well short of a theory which is unified in the sense that it canexplain all three mechanisms and do so without relying on assumptions

different than those normally imposed in analyses taking each of the

mechanisms as a given

In the following chapters, we propose an Adaptation Cost (AC) theory

of thefirm that speaks to these questions and at the same time provides

a micro-foundation for the Resource-Based View (RBV) of the Firm

(Chapter 6)

The argument starts with a worker who has provided a particular

service for an entrepreneur whose needs now have changed The

worker’s productive efficiency will suffer if he changes to another service

or another entrepreneur and in either case some costs may be incurred in

the process of reaching agreement on the terms of the new trade We look

for the most efficient way to balance these three sources of adaptation

costs (inefficiency) in three stages In Chapter 3, we hold production

costs constant and consider bargaining costs only This allows us to

compare bilateral mechanisms, including employment and contracts

In Chapter 4, we introduce advantages of specialization such that

changes between services or entrepreneurs result in higher productioncosts This allows us to expand the set of alternatives to include markets,

as well asfirms We finally, in Chapter 5, introduce a distance metric onthe sets of services and businesses such that some changes are morecostly than others By thus treating firms and markets in a somewhatsimilar way, we are able to derive predictions about the optimal scope offirms and the optimal size of markets Let us now discuss the three types

of adaptation costs in a bit more detail

We start with a “small” force; the difference in bargaining costsbetween negotiating a single average price for a lot of services versusnegotiating separate prices for each individual service It is not counter-intuitive to think of bargaining costs as being subject to economies ofscale in this sense: When faced with the service of trading thirty smallitems, most people would prefer to negotiate once over the bundle ratherthan thirty times on an item-by-item basis In Chapter 3, we show howsub-additivity of bargaining costs can help explain why we use theemployment mechanism to govern trading relationships in need of fre-quent adaptation

The costs of worker adaptation are not just those of bargaining overpayments It is also costly, in the form of lower productivity, to changebetween services and to transition from one business to another While it

is possible for workers to provide different services for different neurs in each period, they have to adapt less if they specialize in providing

entrepre-a single service, stentrepre-ay in entrepre-a bilentrepre-aterentrepre-al relentrepre-ationship (specientrepre-alize in working forone business), or do both So an advantage of either type of specialization

is that it reduces the amount of adaptation needed In Chapter 4, we usethis insight to extend the AC theory to explain the comparative advan-tages of markets, relative tofirms and contracts

A worker is most efficient if he can be “doubly specialized,” ally providing the same service to the same business If this is impossible,because demand facing the business is too small to use a full-time servicespecialist, it will often be second best to specialize in one of the twodimensions and deal with the occasional adaptation in the other Somesuch adaptations, whether between two services or two businesses, aremore costly than others In Chapter 5, we therefore define a measure ofdistance or “relatedness” between pairs of services or businesses So aworker who is specialized in one dimension, say a service (business), canapproximate double specialization by focusing on a small set of relatedbusinesses (services) This perspective allows us to further extend the ACtheory and predict the scope offirms and the size of markets

continu-To see how this leads to a theory of the scope of thefirm, consider aworker who cannot be doubly specialized because the business has too

little demand for the service in question Assume further that

sub-additive bargaining costs make it prohibitively expensive for the worker

to hold two employment positions at the same time In this case the

worker could use his extra time performing the specialized service for

another business operated by the same entrepreneur or he could match

up with a different entrepreneur through the market Since the former

may entail more bargaining costs, the two businesses operated by the

focal entrepreneur must be related in the sense that not too many gains

from business specialization are lost going back and forth between them–

compared to what could be expected if the other business is found in the

market In this scenario, then, the scope of the firm is expanded to

leverage excess capacity of productive assets up to the point where the

component businesses become too different

We have now arrived at the central message of the RBV Define

“resources” as productive assets that are economically inalienable in

the sense that it is more efficient to use excess capacity inside theboundary of thefirm (More precisely, an asset is economically inalien-

able from afirm f if it is inefficient to have any part of its capacity used by

afirm other than f.) Just like the AC theory, the RBV says that firms with

excess capacity of resources should look to leverage them and deploy

them in related businesses (We normally think of the RBV as being

concerned with the case in whichfirms have different resource

endow-ments, but the argument coming out of the AC theory also applies to

homogeneous firms Furthermore, the resources do not have to beontologically indivisible, we just require that sub-additive bargaining

costs makes it inefficient to divide them.)

Since the example concerns a single worker, it may seem unappealing

in light of the very largefirms that dominate modern economies ever, we can readily extend the argument to groups of employees and

How-other productive assets, as long as they are subject to sufficiently

sub-additive bargaining (price determination) costs With this preamble,

Chapter 6 then contains an updated version of the RBV paper with more

explicit emphasis on the point that resources matter to the extent that the

firm has excess capacity

The relationship with the RBV is an important strength of the AC

theory of thefirm, because we know that the RBV rings true to many of

the managers who make decisions about the scope of their firms – thevery decisions that all theories of thefirm try to reconstruct Beyond that,the predictions of the theory are consistent with many stylized facts about

firms: They are used when more frequent and diverse adaptations are

needed, when workers’ benefits of specialization in individual services are

smaller, and when it is costly to switch from one business to another

Having thus developed the AC/RBV theory of the firm, we look, inChapters 7–10, at several implications about differences between firmsand markets: Who should own the productive assets? How does thesteepness of incentives compare? How do the patterns of communicationdiffer? And who should make which decisions?

In contrast to the complex interactive effects favored by much recenttheoretical literature on thefirm, many of the arguments offered here rely

on comparatively simpler and more direct forces This simplicity oftenhas the effect of turning received wisdom on its head Instead of“I amthe boss because I own the assets” (Grossman and Hart, 1986), it is

“I own the assets because I am the boss.” Employment contracts are powered, not because employees are supposed to, by themselves, allocatesome time to aspects of the job that are not contractable (Holmstrom andMilgrom, 1991), but because the boss might intervene and order them to

low-do something not covered by the incentives There is less cation between than withinfirms, not because firms can commit to treatbad news gently, but to protect the firm’s bargaining power Decisionrights are not yielded to enhance investment incentives (Aghion andTirole, 1997), but because it is too costly to agree on everything

communi-Part IV of the book contains three chapters that test the most uniqueaspects of the theory Thefirst reports on a series of experiments (Chap-ter 11) showing that subjects act as if they incur sub-additive bargainingcosts In the second (Chapter 12), we use interview data to conduct avery general make-or-buy study of several supply chains in the automo-bile industry Aiming to test the AC theory directly, we show thatcomponents needing more frequent mutual (coordinated) adaptationare much more likely to be produced by the samefirm The third (Chap-ter 13) contains a test of the theory of asset ownership– a key differencebetween the AC and Property Rights theories (PRT) Again using inter-view data, we show that employees own productive assets when theiractions are critical determinants of the rate at which these assetsdepreciate

In Part V of the book, we discuss three critical assumptions Chapter 14

is about incomplete contracts It is clear that we cannot have a theory ofthefirm without some sort of contractual incompleteness, and while mostpeoplefind it natural to assume that not all contingencies can be antici-pated, it is still considered an ugly assumption by many economists Wemake use of bargaining costs to turn the argument on its head If it is costly

to negotiate agreements about what should happen in each contingency,parties may prefer to leave it unspecified So rather than, “since contractsare incomplete, they can be renegotiated” (Hart and Moore, 1990), it is

“contracts can be left incomplete, because they can be renegotiated.”

Chapters 15 and 16 are aboutfirm heterogeneity – the defining erty of the resources on which the RBV is focused The (business)

prop-adaptation costs that play such a crucial role in the AC theory can be

justified with identical firms, but they are a lot more natural if we rely on

heterogeneity The heterogeneity assumption has received a lot of

atten-tion in the RBV literature, often under the heading “Why are Firms

Different?” (Rumelt, Schendel, and Teece, 1994) The mystery is whyless successfulfirms fail to perfectly imitate their high performing peers

In Chapter 15 we propose a multiple equilibrium model of

“organiza-tional languages.” The equilibria are different, yet have identical

per-formance ex ante, and we would naturally expect different firms to usedifferent“languages.” Even if one language ex post turns out to be betterthan the others, imitation will be hampered by the difficulty of migratingall players to the new equilibrium

In both the economics and strategy literatures, the question of firm

differences is often phrased as one of explaining“Persistent Performance

Differences” (Gibbons and Henderson, 2013) Several proposals have

been made, but many suffer from aggregation problems Specifically, ifindividuals are different, we would expect any differences to have washed

out by the law of large numbers In Chapter 16 we therefore propose a

theory with a very different flavor: Once a small difference has been

noticed, thefirm will tend to deepen it, much like water running down

a mountain side

2.2 Sub-Additive Bargaining Costs

Using language from physics, economists often refer to bargaining costs

as“frictions” in the workings of the economy (Williamson, 1981) The

analogy to air resistance, etc., suggests two implications: That the small

force can be safely ignored for many purposes (apples dropping), and

that it is critical in others (flight) Economists have largely embraced the

first implication, but not the second Bargaining costs are almost

univer-sally abstracted from In particular, perfect and costless “Coasian gaining,” though originally introduced as part of a reductio ad absurdum(Coase, 1960), is now a standard assumption (Hart, 2008)

bar-The demonstrated success of frictionless models is certainly part of the

reason for this, but the amorphous nature of bargaining costs is another

While it often does not matter exactly why bargaining is costly, the costs

themselves come in many guises The most immediate examples are that

bargaining is unpleasant, takes time, and could be subject to strategic

inefficiencies (Myerson and Satterthwaite, 1983) But in many casesmore substantial costs may be incurred in the form of investments in

bargaining power (Tullock, 1980), time spent preparing to bargain(Busse, Silva-Risso, and Zettelmeyer, 2006), and post-bargaining aggrie-vement (Hart and Moore, 2008) Sources of bargaining costs differfrom situation to situation and no single source is universally important.Furthermore, apart from the quite unsatisfactory“fixed cost of bargaining,”there is no unified way to model all of these types of costs So it is perhapsnot surprising that bargaining costs have a bad name in economics.

Conceding that bargaining costs generally are“small,” they are mostlikely to matter where a lot of bargains have to be struck in a shortamount of time For example, the foremen studied by Guest (1956)attend to more than one activity per minute Consistent with this, theanalysis tells us that bargaining costs matter most in situations in whichpairs of agents have frequent needs for mutual adaptation (If manyagents are involved, it can reasonably be argued that market disciplineeliminates the scope for, and thus any cost of, bargaining.)

Intuitively, it would seem to be “expensive” if managers have tobargain with their subordinates at each turn, negotiating payments forevery order given So it is logical to ask whether bargaining costs are sub-additive, such that the parties can economize on them by pooling severalbargains into one In Chapter 11 we experimentally address the dualissues of existence and sub-additivity of bargaining costs Subjects aregiven the role of buyer or seller, paired up, and asked to bargain severaltimes In each bargain, the sellers’ costs and buyers’ values are randomdraws from commonly known distributions The draws are commonknowledge and at the end of the experiment subjects are paid, for allconsummated trades, the difference between negotiated prices and owncosts/values The novel aspect of the experiment is that subjects, at anytime, can make two kinds of offers: one applying to the current tradeonly, and one applying to all current and future trades So the lattereffectively pools the remaining bargains into one Pooling is generally notcostless because the rules specify that all bargains covered by a poolingagreement have to be consummated As the supports of costs and valuesare partially overlapping, costs may exceed values in some bargains.These can be avoided if the subjects bargain on a trade by trade basis,but not if they pool The article reports on six variations of the experi-ment The results, and their consistent pattern across treatments, showthat bargaining costs exist, matter, and are sub-additive in the experi-mental setting

The experiment does not admit several frequently mentioned sources

of bargaining costs Subjects cannot expend any resources preparing tobargain, they have no opportunity to burn surplus with inefficient post-bargaining behavior, and there should be no distortions from incomplete

information (all the treatments involve full information bargaining).

This does not imply that the above types of bargaining costs do not

exist; we believe that they do, but it suggests that simply “not liking

to spend one’s time bargaining” is sufficiently important to show up

in the data

2.3 An Adaptation Cost Theory of the Firm

Before describing the proposed AC theory, we would like to make some

more general observations about its relationship to Transaction-Cost

Economics (TCE) and PRT

First, since firms are “common,” in the sense that there are many of

them, their existence should ideally be explained by some equally common

factors In particular, even though “hold-up” is an out-of-equilibrium

phenomenon in both PRT and TCE, one wonders how widespread it is,

and thus whether the threat of it can explain the multitude offirms in the

economy Bargaining frictions, in contrast, are ubiquitous

Second, since we can imagine a wide variety of institutions, it would

seem unlikely that two or more problems would have the same solution,

as when TCE portrays thefirm as a response to any mixture of ex anteand ex post inefficiencies In contrast, PRT focuses entirely on ex anteinefficiencies and we concentrate on ex post inefficiencies

Third, when thinking of thefirm, it is possible to define its essence in

terms of joint asset ownership or the employment relationship TCE

suggests that the two are driven by “substantially the same forces”(Williamson, 1975, p 99) and PRT gives primacy to asset ownership

with employment following by definition (Grossman and Hart, 1986)

We will use employment relationship as the basis for the theory of the

firm and show in a later part (and Chapters 7 and 13) that asset

owner-ship is likely to be allocated to the boss

Fourth, legal and governmental definitions of employment do not

provide precise standards by which to judge a theory These definitionsare vague and imprecise – probably reflecting the lack of theoretical

clarity in the area A complementary, and possibly better, standard is

given by“everyday language” use of the terms “firm” and “employee.” In

the end, we probably cannot develop a theory that exactly rationalizes

one of these standards However, some theories will clearly come closer

than others

The primary source of inspiration for our theory of the firm is the

literature on optimal trading mechanisms combined with the powerful

intuition that it would be massively inefficient for bosses to negotiate anew fee each time they issue a different instruction To make this hold up

to theoretical scrutiny, we need the positive and sub-additive bargainingcosts discussed in Part II.

The basic theory is developed in Chapters 3 and 4 Both chapters arefocused on workers supplying services to businesses with changingneeds Businesses are operated by entrepreneurs who in each perioddiscover the needs and then try to get them met Wefirst consider threebilateral mechanisms in Chapter 3: Employment, Sequential Contract-ing, and Price Lists.1 In that chapter, our analysis of adaptation costsfocuses on the price determination costs only; production costs areheld constant

In the Employment mechanism, the parties negotiate, on a for-all basis, a wage and a large set of services to be supplied on demand.Examples of services covered could be all the many things a secretary or asuperintendent may be asked to do Since a lot is covered by the agree-ment, the initial bargaining costs may be large, but in equilibrium, nofurther costs are incurred, while gains from trade are realized in everyperiod If one of the parties initiates a later re-negotiation of the wage,bargaining costs are incurred again and there is a corresponding loss

once-and-of efficiency

Under Sequential Contracting, a new price is negotiated wheneverthe business’s needs change, and bargaining costs are incurred on eachoccasion However, these bargains are simpler than those required foremployment contracts; only a single, known service is involved and theoverall stakes are lower The per-occasion bargaining costs are thuslower

With Price Lists, a set of prices are agreed upon ex ante and thenreferred to as needed As in Sequential Contracting, the per-servicebargaining costs are again fairly low However, the mechanism is not

efficient if the parties have to negotiate a very, very long price list So thediversity of needs (how long a price list would have to be) plays animportant role in the relative attractiveness of alternative mechanisms

When needs for service adaptation arise with sufficient frequency,the folk theorem allows us to assume that all trades are efficient inEmployment and under Sequential Contracting, while the Price Listcan be assumed to implement all efficient trades covered by the list.2This

1 These were named “Hierarchy,” “Negotiation-as-Needed,” and “Price Lists,” respectively,

in Wernerfelt (1997).

2 The original version of this chapter contained an explicit dynamic model with two-sided incomplete information along with a folk theorem proving the limiting ef ficiency of ex post trade Most of this formality was taken out of the reduced-form version eventually published.

means that the only bargaining costs are those associated with the process

itself – there are no trading inefficiencies Given this, the performances

of the three mechanisms only differ because of the different costs of

adapting contracts to changes in needs In the Employment

mechan-ism, these are the one-time costs of negotiating a wage agreement; with

Sequential Contracting, these are the per-change costs of agreeing on

new prices; and in the Price List, these are the one-time costs of

negotiating the price list plus the loss of gains from trades not covered

by the list Since no mechanism can govern change with lower variable

costs of adaptation than Employment (just a verbal instruction), there

exists a region of the parameter space (when needs change frequently)

in which it weakly dominates all other mechanisms in a very large space

In Figure 2.1, we have plotted the most efficient mechanisms in terms

of the relative importance of frequent and diverse changes in needs

Looking at the figure, Price Lists are good when they can be keptrelatively short, Sequential Contracting is good when needs change

infrequently, and Employment is good when needs change frequently

and many diverse adaptations are needed

In Chapter 4, we add adaptation costs caused by less efficient

production Specifically, we assume that there are gains from ization (absence of adaptation) in two dimensions: businesses and

special-services The former is captured by assuming that a worker has to

incur an adaptation cost (for example, to cover costs of

transporta-tion, coordinatransporta-tion, learning, and billing) each time he wants to serve a

different business We model the latter by making the standard

assumption that different sellers are good at different services In

this context, Chapter 4 compares three mechanisms: Employment,

Diversity of services

needed

Frequency of changes in needs

Sequential contracting Employment

Price list

Figure 2.1 Most efficient mechanisms and changing needs 1

Sequential Contracting, and Markets To facilitate comparison, weagain assume that trade is ex post efficient in all three mechanismsand concentrate on minimizing total adaptation costs So the Employ-ment and Sequential Contracting mechanisms are exactly the same as

in Chapter 3 discussed earlier The performance of Employment isgains from trade every period minus the one-time costs of negotiatingthe employment contract, and the performance of Sequential Con-tracting is gains from trade minus bargaining costs every period Thenewly introduced mechanism, the Market, is different because sellerscan specialize in the services at which they are most efficient Forexample, instead of being superintendents they can be plumbers,carpenters, or electricians So while gains from trade are higher, thedownside is that workers have to incur business adaptation costs (losegains from specialization) each time they change to serve a newbusiness.3 We show that there exist three regions in the parameterspace in which each of these three mechanisms weakly dominates allothers in a very large class The relative performance depends on thefrequency with which needs change, the gains from specialization in

an individual service, the business adaptation costs, and bargainingcosts These are illustrated in Figures 2.2 and 2.3

Gains from business specialization

Frequency of changes in needs

Sequential contracting

Employment Market

Figure 2.2 Most efficient mechanisms and changing needs 2

3 It is worth noting that this is not an argument about hold-up What matters is the cost of the speci fic investment, not the extent to which it is made.

In Figure 2.2, Markets are good when workers’ between-business

adaptation costs (gains from business specialization) are low, Sequential

Contracting is again good when needs change infrequently, and

Employ-ment is good when the cost advantages of service specialists are small and

needs change fast

Figure 2.3 highlights another parameter: the gains from specialization

in meeting a particular need (“service specialization”) It shows that theMarket is good when service specialists are much more efficient than

employees doing many different jobs

The central empirical prediction of the theory, that more frequent

changes in needs make Employment more attractive, is tested in

Chap-ter 12 We use data from the automobile industry While a car is made up

of thousands of parts, the industry divides them into thirty-six“systems”(body in white, interior, etc.) These systems are interdependent to

differing degrees and our hypothesis is that pairs of systems needing

more frequent mutual adaptation are more likely to be designed and

produced by a single firm Interviews with a number of executives andexperts in the global automobile market helped us produce a 36×36

matrix of these interdependencies It turns out to be extremely hard to

find the optimal structure of production (the number of possible

solu-tions is on the order of the number of seconds since the big bang)

However, the data suggests that the automotive industry does a very

Frequency of changes in needs

Gains from service specialization

Market

Employment Sequential contracting

Figure 2.3 Most efficient mechanisms and changing needs 3

good job: compared to random assignment, pairs of systems needingmore frequent mutual adaptation are much more likely to be housed inthe samefirm.

2.4 The AC Theory of the Scope of the Firm and the RBV

In Chapter 5, we explore the implications of the AC theory for the scope

of thefirm and find that it revolves around excess capacity of ally inalienable assets and thus incentives to expand into related busi-nesses The argument starts by refining the model from Chapter 4.Instead of postulatingfixed production costs of adaptation (gains fromspecialization), we introduce measures of distance (“relatedness”)between pairs of businesses or pairs of services In the simplest model,this is applied at the level of an individual worker Gains from perfectspecialization work as in Chapter 4: A worker is more efficient atplumbing if he does more of it, more efficient at providing repair servicesfor the building at #50 Main St if he does it more, and even more

economic-efficient if he specializes in both plumbing and #50 Main St However,depending on the service, not all businesses need it often enough tooccupy a full-time specialist The idea is, then, that afirm can combineseveral businesses (or services) that are related in the sense that workingfor (performing) one makes you more efficient at working for (perform-ing) the other If you cannot do plumbing for #50 Main St on a full-timebasis, it may still be good to split your time between plumbing at #50 andplumbing at #60, or carpentry and plumbing at #50 (At least compared

to doing plumbing and accounting in two very different buildings.)

The solution to the assignment problem still leaves open the question

of governance: Should the worker be independent or an employee? If themost efficient mechanism involves being a service specialized employee,the worker’s time is economically inalienable and the optimal scope ofthe firm will include a set of businesses that are related in the abovesense If the most efficient mechanism involves being a service special-ized independent, we can use analogous arguments tofind the optimalsize of the market

For our present purposes, the former case is the most interesting.One way of looking at it is the following: Suppose that the worker–entrepreneur pair negotiates an employment contract for one service inone business, but that the business needs are too small to occupy theworker full time In this case very few extra costs would be incurred ifthe worker is asked to use his excess time in a business that is related inthe sense discussed above: Both extra bargaining costs and productivitylosses from adaptation would be small In contrast, the bargaining and/or

The AC Theory of the Scope of the Firm and the RBV 21

adaptation costs incurred in a“sale” of any fraction, negotiating a secondemployment contract with a similar firm or adapting to a possibly verydifferent firm in the market, would be quite large Such an asset (the

worker’s time) is therefore economically inalienable in the sense that it isinefficient for more than one firm to use it So it is as if the entrepreneurhas excess capacity of the asset, and as a result, he will often find it

efficient to expand his firm into related businesses

Depending on how big the advantages of specialization are, afirm mayfind it profitable to hire a specialist even if it cannot use him in that

capacity 100 percent of the time (Assuming that the rents are shared

between the worker and thefirm.) Such a firm will be “good at” the type

of services provided by the specialist If there is another business that

meets the criteria for relatedness, it could well be profitable for the firm

to add that business to its scope So afirm that is good at plumbing will

be in businesses that use plumbing and are otherwise similar, and be

likely to expand into more businesses like them to the extent that it has

excess plumbing capacity.4Note that the argument does not depend on

the specialist being efficient in an absolute sense, nor on match quality It

does require that firms are different to the extent that one can justifydifferences in between-firm adaptation costs

This result, that it may be most efficient for the firm to expand itsscope to more fully utilize the excess capacity of some workers’ time, is anexample of the central point made by the RBV In brutally abbreviated

form, the RBV argument goes like this: Start by defining resources asproductive assets that are“economically inalienable” for a focal firm, by

which we mean that it is inefficient for any other firm to use any part ofthem The inalienability means that the focal firm can have excess

capacity of the resource If so, it can reap efficiencies by entering new

businesses that are “related” to its current scope in the sense that the

adaptation costs between them and existing businesses are small and they

use the asset (resource) in question.5

While most applications of the RBV invoke resources that are

consid-erably more complex than the time of individual employees, trade in all

these assets is subject to sub-additive adaptation costs (or more generally

4 This emphasis on specialization is also found in Foss, 1996.

5

The general idea that excess capacity of assets can drive expansion of the firm goes back at least to Penrose (1959) For example, TCE (Williamson, 1975; Teece, 1982) explains multiproduct firms by indivisibilities and various “market failures.” The argument here is

different First, even an ontologically divisible asset would be economically inalienable if suf ficiently sub-additive bargaining costs made it uneconomical to sell or rent excess

capacity Second, sub-additivity is the critical assumption: If bargaining costs were

positive but super-additive, it would be more attractive to trade small increments.

price determination costs) Consider, for example, groups of employees–like teams of scientists or marketers who can work particularly efficiently

by specializing in afirm’s businesses The AC logic applies to such teamsbecause it would be very expensive to negotiate a price for occasional use

of the entire team.6

2.5 Judging the AC Theory Against Stylized Facts

2.5.1 Asset Ownership

Perhaps the central stylized fact to be faced by any theory of thefirm isthat the boss, the person being obeyed by the employee, generally ownsthe productive assets used by the latter As its name suggests, the PRT(Grossman and Hart, 1986, Hart and Moore, 1990) has taken thisstylized fact particularly seriously and in fact starts by analyzing whoshould own the assets and ends by defining that person as the boss (Soyou are my boss if and only if I work with your assets.) While thisrepresents a major advance over previous theories that more or lessignored the issue, there are cases in which employees own some assets–i.e., many craftsmen own their own tools There are also cases in whichemployees appear to work with no assets at all This then suggests that it

is necessary to use two different theories to explain asset ownership andemployment; although these theories still should be able to explain thatthe boss does own the assets in the vast majority of cases

Since the AC theory of thefirm does not depend on assets at all, wehave many degrees of freedom in pairing it with a theory of asset owner-ship, as long as the latter uses the boss vs employee roles as a significantinput One such theory is described in Chapter 7 The idea is simple andnatural: Since the boss decides how an asset should be used, she shouldbear the consequences of these decisions For example, if the boss asksthat a machine be used on a service that puts a lot of wear and tear on it, it

is normally desirable that she pays the resulting externality in the form ofdepreciation costs An important exception is one in which actions of theasset user (the employee) are a more important determinant of depreci-ation (“Is a carpenter using a chisel as hammer?”) The theory depictsoptimal ownership as striking a balance between the boss’s incentiveswhen deciding what the asset is used for, and the employee’s incentiveswhen deciding how to use the asset It thus allows for cases in which theemployee owns assets as well as for cases in which no assets are used

6 As long as the adaptation costs can grow suf ficiently large, the scope of the firm is bounded even if the resource has in finite capacity.

We test and refine this theory in Chapter 13 Using questionnaire data

onfifty carpenters and forty-one tools, we find that employee ownership

is more likely when the tool is easily lost or stolen (events that are under

the control of the operator) and less likely when the boss’s serviceassignment affects depreciation Consistent with the idea that individual

incentives for responsible use gets diluted when several employees use

the same asset, we also find that shared-use assets are more likely to beowned by the boss In contrast, wefind no effect of asset-specific human

capital investments (although we cannot rule out that these play a role in

other industries)

2.5.2 Communication within and between Firms

It is widely believed that there is more communication within than

between firms (Bolton and Dewatripont, 1994; Simester and Knez,2002) To look at this, we consider, in Chapter 8, a bilateral trading

game in which the seller may discover an improved, but possibly more

costly, widget design After the discovery phase, he offers one, and only

one, design to the buyer So if the seller has discovered the new design,

he has the choice between offering it or the“old” design We will think of

him offering the new design as communicating about its discovery In

contrast, tofind the new design but offer the old will be thought of as notcommunicating In theex ante bargaining mechanism, meant to representEmployment, the parties agree on a price before the seller learns about

the existence, and if so the cost, of an improved design In the ex postbargaining mechanism, meant to represent Sequential Contracting, the

parties bargain after the seller has learned about the new design and had a

chance to inform the buyer about its existence In either case we assume

that bargaining consists of the buyer making a take-it-or-leave-it (TIOLI)

offer While the seller’s costs for the two designs are his private

infor-mation, it is possible for the buyer to make some inferences about these

costs by inverting the seller’s communication strategy The key point

is now that a proposal to trade the improved design allows the buyer

to make some inferences about the relative costs of the old and the

improved designs A proposal to trade the old design may also reflect

relative costs, but could just be driven by the seller not having found the

improved design So while“communication” – proposing the improveddesign – may allow the seller to share in a larger total surplus, it also

reveals more information and allows the buyer to make a, for her, better

TIOLI offer Communication thus diminishes a player’s bargainingpower, and there is less incentive to communicate under sequential

contracting This concern does not play a role in the employment

relationship, where the wage has been settled on ex ante, making ex postbargaining power irrelevant So consistent with stylized facts, we haveless communication between than within thefirm.

2.5.3 Low-Powered Incentives

While the boss’s ownership of assets is the strongest and most preciselymeasured stylized fact about the theory of the firm, another importantstylized fact is that employees face weaker incentives than independentcontractors (Williamson, 1985) This, and the extent to which it isconsistent with our theory, are taken up in Chapter 9

To look at the steepness of incentives, we consider a principal–agentmodel in which the seller can exert different levels of effort resulting inmore or less output According to the theory, employees may be asked

to perform a wide variety of services, each with a different relationshipbetween effort and value Since this introduces more noise in the effort–outcome relationship, it means, in the presence of risk aversion, that theoptimal contract for an unspecified service is less steep than that for aspecified service More precisely, we compare two contracting regimes

in a principal–agent model In the ex ante regime, meant to representemployment, the service is not known and parties have symmetric butnoisy information about the effort–outcome relationship Standard argu-ments then show that the optimal contract is less steep the larger thenoise In theex interim regime, meant to represent sequential contracting,the service is known, and the parties have less noisy but asymmetricinformation about the effort–outcome relationship This allows steeperincentives, but raises the possibility of inefficient bargaining break-down So the choice between employment and sequential contracting

is portrayed as a tradeoff between low-powered incentives and ing costs

bargain-2.5.4 Delegation of Decision Rights

Since the role of bargaining costs in motivating submission to authority iscentral to our theory of the employment relationship, one could conjec-ture that it also plays a role in the delegation of decision rights within thefirm We have investigated two further perspectives on this

In Chapter 14, we consider cases in which the delegating party canrenegotiate a decision before it is implemented In the presence ofbargaining costs, contingent claims contracts are more expensive themore contingencies are covered One way to avoid bargaining over deci-sions in a contingency is to explicitly or tacitly give one party the right to

decide This is particularly attractive if the other party is relatively

indif-ferent between the available options (Simon, 1951), if preferences are

similar, and if the delegating player has comparatively poor information

about the decision However, even if both parties care a lot about the

decision, delegation is possible as long as the delegating party retains the

right to renegotiate/protest the decision In the limit, if renegotiation is as

cheap and as effective as bargaining, the parties can delegate everything

The threat of renegotiation may then discipline the decision maker just

enough that actual renegotiation becomes unnecessary, meaning that the

parties do not even have to negotiate about the contingencies that do

materialize One should not think of this limit as interesting in itself –

After all, it describes a case in which there was no reason to write a

contract in the first place The point is rather that players will want todelegate many contingencies even far away from the limit So in this

model, contracts can be incomplete because the threat of renegotiation

makes it safe to leave them as such The direction of causality in this

argument is, of course, the opposite of that in the Property Rights

literature in which incompleteness makes renegotiation feasible

In Chapter 10, we also use the machinery of protests and renegotiation

to look at the case in which bargaining costs take the form of

decision-making costs As any manager will tell you, decision-making good decisions is

hard work You need to collect information, think through scenarios, and

dream up alternative ways of doing things One advantage of delegation

or ceding of decision rights is thus that the delegating player saves these

costs This makes it more attractive to give up decision rights and

effectively agree to take orders.7The model considers a case in which a

group of individuals have to make a single joint decision, but differ in

their preferences as well as in the amount of information they have If

participation is costly, it is ex ante efficient to delegate the decision to asmall committee whose members all have good information and jointly

have average preferences If, as in Chapter 14, renegotiation/protests

are possible, delegation becomes even more attractive and the optimal

committee size smaller

their strategies and behaviors.8While these facts are readily accepted byboth management scholars and economists, we have had a harder timebringing them into our theories The issue is that it has been very difficult

to understand why imitation does not eliminate all important ences.9 This book offers one explanation for enduring differences intwo steps The first shows how small differences might emerge even iffirms are ex ante identical, and the second uses the theory to show howsmall differences endogenously can turn into large differences

differ-In Chapter 15, we look at a team which wants to allocate a set ofresources to those members for whom they are worth the most Individ-ual valuations are private information, but the members may communi-cate in “rounds,” meaning that each member can send one binarymessage to a center in each round A code can be represented as asequence of nested bisections of the set of possible valuations Forexample, if the values 1, 2, 3, and 4 are possible, one code could have

“High” meaning “3 or 4” and “High, Low” meaning “3,” while anothercode could have“High” meaning “2, 3, or 4,” and “High, Low” meaning

“2 or 3.” Members choose their codes individually, but all measureperformance by the expected number of rounds needed to perfectlyallocate the resources A code is“linear” if when x < y < z and it sendsthe same message for x and z, then it also sends that message for y A set

of codes are in equilibrium if no unilateral deviation gives the team higherpayoff The main results are that all equilibria are symmetric in the sensethat all members use the same code, and that any linear code can besustained in equilibrium as long as all members use it Since linear codesdiffer in their efficiency, this means that we have multiple, more or less

efficient equilibria To the extent that equilibrium selection is random,this then suggests that ex ante identical teams can end up in differentequilibria with very different efficiencies

The persistence of heterogeneity is discussed in Chapter 16 Themodel explains howfirms may have incentives to amplify, through invest-ments, very small initial differences Much like water running down arock surface will enlarge a small crease, a small random asymmetrybetween the firm and its competitors will grow through subsequentinvestment in competitive advantage (see also Selove, 2014)

Firm heterogeneity shares two features with bargaining costs: Both arelikely to have many sources and we still lack a fully satisfactory theory of

8 While the AC theory of the scope of the firm does not depend on such differences, it does depend on workers incurring some costs when changing from one firm to another and these costs ultimately need some sort of justi fication.

9

See, however, Lippman and Rumelt, 1982.

their existence However, unlike sub-additive bargaining costs,firm erogeneity is not necessary for the AC/RBV theory of thefirm, although itmakes it more interesting.

het-ReferencesAgohion, Philippe, and Jean Tirole,“Formal and Real Authority in

Organizations,” Journal of Political Economy, 105, no 1, February, 1–29,

1997

Bolton, Patrick, and Mathias Dewatripont,“The Firm as a Communication

Network,” The Quarterly Journal of Economics, 109, no 4, November,

809–39, 1994

Busse, Meghan, Jorge Silva-Risso, and Florian Zettelmeyer,“$100 Cash Back:

The Pass-Through of Auto Manufacturer Promotions,” American Economic

York, NY: Routledge, 1996

Gibbons, Robert, and Rebecca Henderson,“What Do Managers Do?,”

Chapter 17 in R Gibbons and J Roberts (eds.),Handbook of Organizational

Economics, Princeton, NJ: Princeton University Press, 2013

Grossman, Sanford, and Oliver D Hart,“The Costs and Benefits of

Ownership,” Journal of Political Economy, 94, August, 691–719, 1986

Guest, R.H.,“On Time and the Foreman,” Personnel, 32, 478–86, 1956

Hart, Oliver D.,“Reference Points and the Theory of the Firm,” Economica, 75,

no 299, August, 404–11, 2008

Hart, Oliver D., and John Moore,“Property Rights and the Nature of the Firm,”Journal of Political Economy, 98, no 6, 1119–58, 1990

Hart, Oliver D., and John Moore,“Contracts as Reference Points,” Quarterly

Journal of Economics, 123, no 1, February, 1–48, 2008

Holmstrom, Bengt, and Paul Milgrom,“Multiservice Principal-Agent Analyses:

Incentive Contracts, Asset Ownership, and Job Design,” Journal of Law,

Economics and Organization, 7, special issue, 24–52, 1991

Lippman, Steven A., and Richard P Rumelt,“Uncertain Imitability: An Analysis

of Interfirm Differences Under Competition,” Bell Journal of Economics, 13,

418–38, 1982

Maciejovsky, Boris, and Birger Wernerfelt,“Costs of Implementation:

Bargaining Costs versus Allocative Efficiency,” Journal of Economic Behavior

and Organization, 77, no 3, March, 318–25, 2011

Myerson, Roger B., and Mark A Satterthwaite,“Efficient Mechanisms for

Bilateral Trading,” Journal of Economic Theory, 29, no 2, 265–81, 1983

Novak, Sharon, and Birger Wernerfelt,“On the Grouping of Services into Firms:Make-or-Buy with Interdependent Parts,” Journal of Economics and

Management Strategy, 21, no 1, Spring, 53–77, 2012