. 1
( 18)



>>

ISBN:0521893135
© 2002 (584 pages)
!
" #$ % &
%% & & % % "



The Economics of Contracts-Theories and
Applications
Eric Brousseau
Jean-Michel Glachant
CAMBRIDGE UNIVERSITY PRESS

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGE
The Pitt Building, Trumpington Street, Cambridge, United Kingdom

CAMBRIDGE UNIVERSITY PRESS
The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa
http://www.cambridge.org

Copyright © Cambridge University Press

This book 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 2002
Typeface Plantin 10/12 pt System LATEX2 [TB]
A catalogue record for this book is available from the British Library
Library of Congress Cataloguing in Publication data
The economics of contracts: theories and applications/edited by Eric Brousseau and
Jean-Michel Glachant.
p. cm.
Includes revised and translated versions of chapters which appeared in a special issue
of Revue d'©conomie industrielle (2002, 92)
Includes bibliographical references and index.

ISBN 0 521 81490 1
0-521-89313-5

(pb.)
1. Contracts-Economic aspects. I. Brousseau, Eric. II. Glachant, Jean-Michel.
K840.E28 2002
338-dc21 2002019300

ISBN 0 521 81490 1 hardback
ISBN 0 521 89313 5 paperback
Contributors

PHILIPPE AGHION is Professor of Economics at Harvard University, His main fields of
interest are the theory of contracts and the theory of growth. Together, both theories
allow a better understanding of the links between technical change and institutional
evolutions. His most recent research focuses on the relationship between competition
and growth, and he is currently starting a project on contracts and growth.
ASHISH ARORA is Associate Professor of Economics at Heinz School, and Research
Director at the Carnegie Mellon Software Center, both at Carnegie Mellon University.
Arora's research focuses on the economics of technological change, intellectual property
rights, and technology licensing. He has published extensively on the growth and
development of biotechnology and the chemical industry. His most recent book is
Markets for Technology (MIT Press).

BENITO ARRU‘ADA is Professor of Business Organization at Universitat Pompeu
Fabra, Barcelona, Spain. In addition to retailing, his research deals with contractual
practices in the franchising, auditing, healthcare, public administration, construction,
trucking, fishing, and conveyancing industries, as well as the impact of different legal
rules, such as those on payment delays, multidisciplinary professional firms, corporate
governance, and land registration.

MATTHEW BENNETT received his PhD from the department of economics at the
University of Warwick. His dissertation centered around competition and regulation policy.
He is currently working in the University of Toulouse under a Marie Curie training grant,
on the interaction of license auctions and optimal regulatory contracts.

ERIC BROUSSEAU is Professor of Economics at the University of Paris X. He works
with two research centers: FORUM (University of Paris X), where he is the director of the
department of industrial organization, and ATOM at the University of Paris (Panth©on-
Sorbonne). His area of interest is the economics of coordination, mainly contractual and
institutional economics. His applied fields of research include the economics of
intellectual property rights and the economics of the digital economy.
GÉRARD CHARREAUX is Professor in Management Science at the Universit© de
Bourgogne. He is presently coordinator of the research program in finance, governance,
and organizational architecture for the Laboratoire d'Analyse et de Techniques
Economiques (Latec-Cnrs) and editor of the review Finance Contrôle Strategie. His main
fields of research are corporate governance and corporate finance.
RONALD COASE is Clifton R. Musser Professor Emeritus of Economics at the
University of Chicago Law School. He was editor of the Journal of Law and Economics
1964-82. In 1991, he was awarded the Alfred Nobel Memorial Prize in Economic
Sciences.

GODEFROY DANG-NGUYEN is Professor and Head of the Economics and Human
Sciences Department, ENST Bretagne. He is currently doing research on the impact of
information technology on corporations, institutions, and public policy. He is invited
Professor at the College of Europe in Bruges.

M'HAND FARES is a research fellow at INRA, ESR-Montpellier (UMR MOISA) and
member of the research center ATOM-GENI (University of Paris I). His major fields of
interest are contracting and organization, law and economics, and organization of the
agro-food industry.

OLIVIER FAVEREAU is Professor of Economics at the University of Paris X. He is also
the head of FORUM, a research unit which develops an institutionalist program of
research, in four fields: money and macroeconomics, industrial economics, employment
systems, and transition and development studies. His own work deals with conventions
and institutions on the labor market.
ANDREA FOSFURI is Assistant Professor at the Business Department of the University
Carlos III, Madrid, and research affiliate at CEPR, London. He has published in several
leading economics and management journals, and co-authored with Ashish Arora and
Alfonso Gambardella the book Markets for Technology: Economics of Innovation and
Corporate Strategy (MIT Press).
EIRIK G. FURUBOTN is Research Fellow at the Private Enterprise Research Center,
Texas A&M University, College Station, Texas, USA. He is also Honorary Professor of
Economics at the University of Saarland, Saarbrucken, Germany, and member of the
Advisory Board of the Journal of Institutional and Theoretical Economics.
JACQUES GHESTIN is Professor Emeritus of Law at the University de Paris I
(Panth©on-Sorbonne). He is the main author of several treatises dedicated to civil law
and contractual law. He is also a lawyer, and practices international contracting and
arbitration.

JEAN-MICHEL GLACHANT is Professor of Economics at the University of Paris Sud. He
was formerly the director of the research center ATOM at the University of Paris
Panth©on-Sorbonne, and currently heads the economics department of the research
center ADIS at the University of Paris XI. He is also member of the Economic Advisory
Council of the French electricity regulation commission (CRE).

MICHEL GLAIS is a Professor of Economics at the University of Rennes. He specializes
in anti-trust and competition law, business strategy, and corporate finance and
assessment. He lectures for several Universities in Europe and America: Herriot-Wyatt
(Edinburgh); Boston College; University of New Hampshire; La Sapienza (Roma); Baltic
Business School (Sweden). He is also a chartered expert at Court, and involved in
private consultancy for several major corporations.

VICTOR P. GOLDBERG is the Thomas Macioce Professor of Law and the Co-Director
of the Center for Law & Economic Studies at Columbia University. His current research
focus is an application of economic reasoning to contract law cases and doctrine.

OLIVER HART is the Andrew E. Furer Professor of Economics at Harvard University,
and a Research Associate of the National Bureau of Economic Research (NBER). He is
also Centennial Visiting Professor at the London School of Economics. He works on the
theory of the firm and financial contracting.

GUY L. F. HOLBURN is an Assistant Professor at the University of Western Ontario,
Richard Ivey School of Business. Prior to joining Ivey, he completed his MA in
Economics and PhD at the University of California, Berkeley. His research focuses on
utility and regulation issues, particularly as applied to the electricity industry. He has also
worked as a consultant for Bain and Co. and for the California Public Utilities
Commission.

PAUL L. JOSKOW is the Elizabeth and James Killian Professor of Economics and
Management at the Massachusetts Institute of Technology (MIT), Director of the MIT
Center for Energy and Environmental Policy Research, and a Research Associate of the
National Bureau of Economic Research.

CLAUDIA KESER is research staff member at the IBM T. J. Watson Research Center,
Yorktown Heights, New York and associated fellow of the Centre Interuniversitaire de
Recherche en Analyse des Organisations (CIRANO), Montreal.

BENJAMIN KLEIN is Professor of Economics at UCLA and President of Economic
Analysis at LLC, an economic consulting firm based in Los Angeles. His research
interests focus on the law and economics of contractual arrangements and anti-trust
policy, including vertical distribution arrangements, vertical integration, and competitive
marketing policies.

FRANCINE LAFONTAINE is Professor of Business Economics and Public Policy at the
University of Michigan Business School. She is also a Faculty Research Fellow at the
National Bureau of Economic Research (NBER). Her research focuses on incentives
issues and contracting practices, with special emphasis on the franchise and trucking
industries.

GARY D. LIBECAP is Anheuser Busch Professor and Professor of Economics and Law
at the University of Arizona in Tucson. He also is Director of the Karl Eller Center and
Research Associate in the National Bureau of Economic Research. His research
interests focus on the issues of property rights, economic behavior, and resource use.
W. BENTLEY MACLEOD is Professor of Economics and Law at the University of
Southern California, a Director of the Center for Law, Economics and Organization, and
is currently visiting Professor of Economics and Law at the California Institute of
Technology. His recent research concerns the theoretical and empirical implications of
bounded rationality for contract form.

ERIC MALIN is Professor of Economics at the University of La R©union. He is member
of GREMAQ (Toulouse) and CERESUR (La R©union). His main research interests
include network economics, price discrimination, and health economics.
DAVID MARTIMORT is Professor at the University of Toulouse. He is also member of
the Institut d'Economie Industrielle (IDEI) in Toulouse and CEPR, London. His work
concerns collusion in organizations and mechanism design in multiprincipals'
environment. He has authored a textbook with Jean-Jacques Laffont on incentives, The
Theory of Incentives: The Principal-Agent Model. He has been invited to teach contract
theory at Harvard, Pompeu Fabra, and Universit© de Montreal.

SCOTT E. MASTEN is Professor of Business Economics and Public Policy at the
University of Michigan Business School. His research interests include contracting
practices, contract law, and their relation to economic organization.

CLAUDE MÉNARD is Full Professor of Economics at the University of Paris Panth©on-
Sorbonne and director of the Center for Analytical Theory of Organizations and Markets
(ATOM). He is President (2001-02) of the International Society for New Institutional
Economics (ISNIE). His fields of interest are mainly the economics of organization and
the economics of regulation/deregulation.

THIERRY PÉNARD is Professor of Economics at the University of Rennes I and
affiliated to CREREG. His fields of specialization include the economics of networks,
game theory, and anti-trust policy. His current research focuses on the economics of
telecommunications and the Internet.

EMMANUEL RAYNAUD is a researcher at INRA-SADAPT (National Institute of
Agronomical Research) and a member of the Center for Analytical Theory of
Organizations and Markets (ATOM) (University of Paris Panth©on-Sorbonne). His field of
specialization includes the economics of contracts and organization. His current research
focuses on product quality and vertical coordination in European agro-food industries
and on franchising (design of contracts and dual distribution in franchise chains).

PATRICK REY is Professor of Economics at the University of Toulouse and Research
Director at the Institut d'Economie Industrielle (IDEI). He has also been Associate
Professor at the Ecole Polytechnique since 1991. His fields of interest cover industrial
organization, public economics, competition law and policy, regulation of natural
monopolies, corporate finance, banking and financial intermediation, contract theory, and
theory of the firm and of organizations.

STÉPHANE SAUSSIER is Professor of Economics at the University of Nancy II. He is
also Deputy Director of the Center for Analytical Theory of Organizations and Markets
(ATOM) (University of Paris Panth©on-Sorbonne). Specializing in the economics of
organizations and contracts, he has been working on several fields of application such
as technology licensing agreements, water supply, coal contracts, and franchise
contracts, focusing on contractual choices and make-or-buy decision.
ALAN SCHWARTZ is Sterling Professor of Law and Professor of Management, Yale
University. He has been Editor of the Journal of Law, Economics and Organization,
President of the American Law and Economics Association, and Chair of the Section on
Contracts of the Association of American Law Schools. He currently is Director of the
Yale Law School Center for the Study of Corporate Law and serves on the boards of two
publicly traded companies.

PABLO T. SPILLER is the Joe Shoong Professor of International Business and Public
Policy, and chairs the Business and Public Policy Group, at the Haas School of Business
of the University of California, Berkeley. Prior to joining Berkeley, he was on the faculties
of the University of Pennsylvania, the University of Illinois, and the Hoover Institution at
Stanford University. He has published more than eighty articles and five books, and has
received numerous awards from the National Science Foundation, the Olin Foundation,
the Bradley Foundation, the Ameritech Foundation, and the National Center for
Supercomputer Applications.

CATHERINE WADDAMS PRICE is Professor of Regulation and Director of the Centre
for Competition and Regulation, University of East Anglia, Norwich, United Kingdom. Her
main research interests are in economic regulation of markets, the introduction of
competition, and the distributional consequences. She formerly acted as an economic
expert for the UK energy regulator, and is now a member of the UK Competition
Commission.

BERNARD WALLISER is Professor of Economics at Ecole Nationale des Ponts et
Chauss©es and Director of Studies at the Ecole des Hautes Etudes en Sciences
Sociales (Paris). Involved in economic methodology, he is now leading a program in
"Cognitive Economics," concerned with the study of agents' beliefs, reasoning methods,
and learning processes when these agents are involved in social networks.
OLIVER E. WILLIAMSON is the Edgar F. Kaiser Professor of Business, Professor of
Economics, and Professor of Law at the University of California, Berkeley. He is the
founding co-editor of the Journal of Law, Economics and Organization and a member of
the National Academy of Science.

MARC WILLINGER is Professor of Economics at the University Louis Pasteur
(Strasbourg, France). His current research activities contribute to the development of
experimental economics, with applications to contract design, efficiency of environmental
policy instruments, decision-making under uncertainty, and the dynamics of cooperation.
Acknowledgments
This book draws partially from a special issue of the Revue d'Economie Industrielle
entitled "The Economics of Contracts in Prospect and Retrospect," (92, 2000) in which
Eric Brousseau and Jean-Michel Glachant edited earlier versions of some of the
chapters of that book.
The publishers and editors would like to thank Les Editions Techniques et Economiques
for permission to publish revised or translated versions of the chapters 1-4, 6-8, 10, 13-
22, and 24, which appeared in the special issue of Revue d'Economie Industrielle (2000,
92).
The editors are also grateful to the board of the Revue d'Economie Industrielle, and
especially to its Editor in Chief Jacques de Bandt for having facilitated the publication of
these chapters in that book.

The editors are also indebted to Marie-Line Priot (FORUM, University of Paris X) for
secretarial support and to Paul Klassen for translation.

The publisher has used its best endeavours to ensure that the URLs for external
websites referred to in this book are correct and active at the time of going to press.
However, the publisher has no responsibility for the websites and can make no
guarantee that a site will remain live or that the content is or will remain appropriate.

A contract is an agreement under which two parties make reciprocal commitments in
terms of their behavior to coordinate. As this concept has become essential to
economics in the last thirty years, three main theoretical frameworks have emerged:
"incentive theory," "incomplete-contract theory," and "transaction-costs theory." These
frameworks have enabled scholars to renew both the microeconomics of coordination
(with implications for industrial organization, labor economics, law and economics, and
organization design) and the macroeconomics of "market" (decentralized) economies
and of the institutional framework. These developments have resulted in new analyses of
firms' strategy and State intervention (regulation of public utilities, anti-trust, public
procurement, institutional design, liberalization policies, etc.). Based on contributions by
the leading scholars in the field, this book provides an overview of the past and recent
developments in these analytical currents, presents their various aspects, and proposes
expanding horizons for theoreticians and practitioners.
Eric Brousseau is Professor of Economics at the University of Paris X and member of
the Institut Universitaire de France. He is the director of the department GIFT of FORUM
(University of Paris X and Centre National de la Recherche Scientifique), and associate
researcher at ATOM (University of Paris I). He coordinates a CNRS research consortium
on Information Technologies and the Society, and organizes the European School on
New Institutional Economics. He is member of the Boards of the International Society for
New Institutional Economics and of the Schumpeter Society.
Jean-Michel Glachant is Head of the Department of Economics at the University of
Paris XI. He is a member of the International Society for New Institutional Economics,
the International Association for Energy Economics, and the Association Fran§caise de
Science Economique, as well as head of the Electricity Reforms Group at the ADIS
research centre.
Introduction
Part I:
Chapter 1: The Economics of Contracts and The Renewal of Economics
The Economics of Contracts and The
Chapter 1:

Renewal of Economics
Eric Brousseau, Jean-Michel Glachant
1 Introduction
To an economist, a contract is an agreement under which two parties make reciprocal
commitments in terms of their behavior - a bilateral coordination arrangement. Of course,
this formulation touches on the legal concept of the contract (a meeting of minds creating
effects in law), but also transcends it. Over the course of the past thirty years, the
"contract" has become a central notion in economic analysis (section 2), giving rise to
three principal fields of study: "incentives," "incomplete contracts," and "transaction
costs" (section 3). This opened the door to a revitalization of our understanding of the
operation of market economies and of the practitioner's "toolbox" (section 4).
The goal of this chapter is to provide an overview of recent developments in these
analytical currents, to present their various aspects (section 5), and to propose
expanding horizons (section 6). The potential of these approaches, which have
fundamentally impacted on many areas of economic analysis in recent decades, is far
from exhausted. This is evinced by the contributions in this book, which draw on a variety
of methodological camps and disciplines.
2 The central role of the notion of the contract in economic
analysis
Even though the notion of the contract has long been central to our understanding of the
operation of decentralized social systems, especially in the tradition of the philosophie
des lumières, only recently have economists begun to render it justice. Following in the
footsteps of Smith and Walras, they long based their analyses of the functioning of
decentralized economies on the notions of market and price system. This application of
Walrasian analysis, in which supply meets demand around a posted price, does not
satisfactorily account for the characteristics of a decentralized economy (cf. Ronald
Coase's chapter 2 in this volume). First, and paradoxically for a model of economic
analysis, it does not account for the costs of operating the market. Next, it assumes the
pre-existence of collective coordination (implicitly institutional) - the properties of the
traded merchandise are fixed in advance, all market actors effectively participate in the
atonnement process, etc. - in contradiction with the idea that the market is truly
decentralized. Finally, this model is unrealistic because, in practice, agents exchange
goods and services outside of equilibrium and in a bilateral context, i.e. without
knowledge of the levels and prices at which other agents are trading, and without
knowledge of whether these prices clear the market.
Contract economics was born in the 1970s from a twofold movement of dissatisfaction
vis-à-vis Walrasian market theory:
On a theoretical level, new analytical tools were sought to explain how
economic agents determine the properties, quantities, and prices of the
resources they trade in face-to-face encounters. If these agents are subject to
transaction costs, if they can benefit from informational advantages, or if there
are situations in which irreversible investments must be made, then it is
reasonable to expect that one will not see the same goods traded at the same
price and under the same rules as on a Walrasian market. Price theory and, by
extension, the analysis of the formation of economic aggregates (prices, traded
quantities and qualities, etc.), were fundamentally affected by the work of Akerlof
(1970), Arrow (1971), and Stiglitz (1977), among others.
On an empirical level, problems associated with the regulation of competition
drove a renewal of economic thinking. The analysis of certain types of inter-firm
contracts, such as selective distributorship agreements, long-term cooperation
agreements, etc., was revamped. Previously considered anti-competitive, the
beneficial welfare effects of these arrangements had been ignored. The devices
available to public authorities for creating incentives and controlling producers of
services of public interest were also subjected to a reexamination. Economic
theory had not considered the possibility that either party could appropriate the
rent from monopolistic operation of such services. Demsetz and Williamson,
Baron and Laffont, to name only a few, renewed the approach to these issues of
"regulation."

This twofold origin explains the remarkable development of contract theory and its key
contribution to a fundamental redesign of all areas of economic analysis, from the study
of microeconomic interactions to that of macroeconomic aggregates (such as the labor
market), passing on the way the various domains of applied economics, finance,
international trade, industrial organization, etc.

This success is essentially attributable to the analytical power of the notion of contract.
On the one hand, the idea of contract focuses attention on elementary social structures,
those that regulate coordination at a bilateral level. On the other hand, despite its
simplicity as a concept, the contract allows us to examine a number of key issues. We
can point to at least four:
First, the analysis of contracts allows us to reexamine the exact nature of
difficulties associated with economic coordination, while deepening our
understanding of the functioning and the basis of coordination mechanisms.
Second, this approach illuminates the details of various provisions for
coordination: routines, incentives, the authority principle, means of coercion,
conflict resolution, etc.
Third, analysis of the origins of contracts sheds light on how agents
conceptualize the rules and decision-making structures that frame their behavior.
Finally, studying the evolution of contractual mechanisms helps us understand
changes in the structures that frame economic activity.
The contractual approach thus allows us to analyze coordination mechanisms within a
simplified but rigorous framework. It not only illuminates the properties of contracts, but
also those of other harmonization instruments, such as markets, organizations, and
institutions (cf. Oliver Williamson's chapter 3 in this volume). These collective
arrangements reveal mechanisms comparable to those typical of contracts (participation
incentives, allocation of decision rights, provisions to give credibility to commitments,
etc.).

It should be noted that the analysis of contracts must also be clear on the limits of this
approach to economic activity. Specifically, this is true for organizations and institutions
that are not reducible to the notion of the contract. On the one hand, organizations and
institutions have a fundamentally collective character: an individual will join them without
negotiating each rule governing the relations between members. Moreover, the evolution
of this relational framework cannot be controlled by any individual acting alone. On the
other hand, the properties of organizations' and institutions' collective arrangements do
not derive uniquely from the content of the bilateral relationships linking each of their
elements, but also from the communal articulation of these arrangements - in other
words, the topology of the interaction networks.
The contractual approach is also relevant because of the exchanges it makes possible
with other disciplines. These include law, of course, but also management, sociology,
anthropology, political and administrative sciences, and philosophy. The notion of the
contract is simultaneously broader in scope and more general than the notion of the
market. This has allowed the economic analysis of the contract to export some of its
results, notably the difficulty of creating perfect incentive mechanisms, the incentive-
insurance dilemma, or the impossibility, under many conditions, of drafting complete
contracts (cf. Alt and Shepsle 1990). But the contractual approach has also provided a
gateway for imports that have proven indispensable to advances in economic analysis
(cf. section 6). Other intellectual and methodological traditions have allowed us to extend
the economics of contractual coordination. Legal analysis, for example, specifies the role
of various mechanisms that ultimately guarantee the performance of contracts and
brings to light their "embedding" into the general rules that give them meaning and
complete them. Management sciences emphasize that economic agents concretely act
on the complementary relationship between contracts and imperfect incentive provisions
to resolve coordination problems (e.g. Koenig 1999).
3 Three principal currents
3.1 Origins

While we can speak of "contract economics" in general, it is worthwhile to distinguish
between several branches of contract theory, into which various analytical traditions
have converged that were themselves renewed in the process. While these currents all
sprang from dissatisfaction with the standard analytical model of the market, different
methodologies gave rise to them.

One of the new models derives from the lineage of the standard model. Arrow's work on
the functioning of insurance markets (Arrow 1971), and that of Akerlof (1970) on the
market for used automobiles, led to the theory of incomplete information. Challenging the
assumption that all actors on a market have access to symmetrical, or identical,
information, the authors drew attention to the consequences of one individual having an
informational advantage. They emphasized the importance of implementing disclosure
mechanisms to limit the ability of the "informed" to take advantage of the "under-
informed." This line of research dates from the 1960s.

As early as the 1930s, however, other foundations of modern contract analysis were laid.
Coase was the first to enunciate the idea that the existence of coordination costs on the
market justifies resorting to various coordination mechanisms in a decentralized
economy, especially hierarchical coordination within firms (cf. Coase 1937, 1988). Some
forty years later this analysis was taken up and expanded by Williamson.

But Coase was not the only influence on Williamson. The latter's early work in the 1960s
represented the Carnegie behaviorist school, along with Cyert and March (Cyert and
March 1963). Here we find the lineage of theories of the firm whose formulation began in
the 1930s, but whose full development occurred primarily in the 1950s. Managerial and
behaviorist approaches to the firm (from Berle and Means 1932 to Simon 1947, passing
over Hall and Hitch 1939), as well as the controversies surrounding their development (cf.
Machlup 1967), permitted considerable advances in the understanding of non-price
coordination. Starting in the 1970s, many of these advances were revisited by
economists interested in the properties of contractual, organizational, and institutional
means of coordination.

Another "school" had a profound influence on contemporary contract theory: property
rights (Alchian 1961, Demsetz 1967, Furubotn and Pejovich 1974). In a certain sense,
Coase also laid the foundations for this approach with his analysis of the problem of
externalities (Coase 1960), which brought to light the implications of property-rights
definitions for the issue of efficiency. This contribution then merged with further
developments from the Chicago school. Comparative analysis of alternate property-
rights systems revealed that the allocation of residual rights (the right to determine the
use of resources and to appropriate the ensuing income) may, or may not, motivate an
efficient use of resources. This approach yielded essential elements of theories of the
firm and of contracts (Alchian and Demsetz 1972, Klein, Crawford and Alchian 1978).
Under certain types of relational arrangements, only a reallocation of property rights can
overcome economic agents' propensity to be opportunistic. This school also focused
economists' attention on the specific consequences of the manipulation of incentive
systems.

Finally, it would be impossible to ignore the contributions of other disciplines. Economic
analysis of the law has concentrated on certain aspects of contractual relationships. It is
also noteworthy that one of the primary concepts in the economic analysis of contracts,
the notion of the "hybrid form" proposed by Williamson (1985), drew directly on Macneil's
(1974) socio-legal analysis. On another level, economic views of non-market
coordination were profoundly influenced by developments in management sciences, by
sociology and psycho-sociology, by administrative sciences, and by the history of
organizations, as is evinced by the frequency of references to Barnard, Simon, and
Chandler (Barnard 1938, Simon 1947, Chandler 1962). As to the economics of
institutions, which develops an analysis more concerned with the role of the institutional
environment on the design and the performance of contracts, it traces its roots to history,
to political science, and to ethnology (cf. Eggertsson 1990, North 1990).

Arising from these precursors, three schools dominate the field of contract economics
today: incentive theory (IT), incomplete-contract theory (ICT), and transaction-costs
theory (TCT). These are distinguished by differences in their underlying assumptions,
leading them to emphasize different problems. The standard models of these three
theories are described in the appendix to this chapter by M'hand Fares.

3.2 Incentive theory

Incentive theory (IT) draws on several of the traditional hypotheses of Walrasian
economic theory. Notably, it assumes that economic agents are endowed with
substantial, or Savage, rationality (Savage 1954), that they possess complete
information concerning the structure of the issues they confront along with unlimited
computational abilities, and that they have a complete and ordered preference set.

The information available to these agents is "complete" in the sense that, even though
they cannot precisely anticipate a future that remains stochastic, they do know the
structure of all the problems that may occur. What they cannot know, where applicable,
is what issues will in fact arise, nor in what sequence. Thus, they envision the future on
the basis of probabilities (objective or subjective). This links to the notion of risk, as
described by Knight (1921) (even though Knight did not account for subjective
probabilities). Given this theoretical framework, agents imagine the most efficient
solutions as functions of the different possible states of nature and compute their
expected values. These calculations are possible since agents are endowed with
unlimited abilities in this area. In other words: calculating costs them nothing in terms of
time or resources. Finally, since agents' preference functions are complete and stable
over time, they effectively choose optimal solutions.
The assumption that diverges from the Walrasian universe is that the two contracting
parties do not have access to the same information on certain variables. This is an
evolution toward a more realistic conception. In a decentralized economy, there is no
reason why one party should know, ex ante, the private information of the other (such as
her preferences, the quality of her resources, her willingness to pay, or her reservation
price). Depending on whether the variable on which there is asymmetric information is
exogenous “ i.e. not subject to manipulation during the exchange by the party
possessing it “ or endogenous “ i.e. vulnerable to such manipulation “ we speak of
models of adverse selection or moral hazard, respectively. Adverse selection, for
example, is exemplified by a potential employer's uncertainty concerning a job seeker's
level of competence, while moral hazard refers to uncertainty about the level of effort the
latter will supply.

Incentive theory (IT) starts from a canonical situation in which an under-informed party “
called the "principal"“ puts into place an incentive scheme to induce the informed party “
the "agent"“ to either disclose information (adverse-selection model) or to adopt
behavior compatible with the interests of the principal (moral-hazard model). The
incentive scheme consists of remuneration being conditional on signals that result from
the agent's behavior (such as the choice of an option from a list of propositions
considered a "menu" of contracts or as the visible result of the effort supplied when the
effort itself is not observable).

The existence of such an incentive scheme relies on two key assumptions:
While the principal is under-informed, not knowing the true value of the
hidden variable, she does know both the probability distribution of this
variable and the agent's preference structure. The principal can thus put
herself "in the place" of the agent to anticipate the latter's reactions to the
set of conceivable remuneration schemes, and then select the one she
prefers from those acceptable to the agent.
There is an institutional framework, hidden but competent and
benevolent, which ensures that the principal respects her commitments.
Thus, any proposition made by the principal is credible to the agent.
Moreover, the proposed remuneration scheme is based upon "verifiable"
information, i.e. observable by a third party.

The solution to adverse selection problems relies on the design of a "menu of contracts"
that will induce self-revelation by the agent of her private information. The principal
designs a set of optional contracts “ i.e. a set of payment formulae linked to various
counterparts by the agent. While he does not know the agent's private information, he
knows the set of possible values it may take. Since he also knows her preferences, she
is able to design a contract that maximizes the agent's utility for each possible value of
that private information. When the agent faces the resulting set of possible options, she
spontaneously chooses the contract that maximizes her utility, allowing the principal to
infer private information. Of course, the principal's interest is to obtain this revelation in
exchange for the lowest possible payment.

The canonical moral-hazard problem occurs when one relevant dimension of the agent's
input is not observable by the principal “ one dimension is costly to the agent, and that
affects the principal's welfare. For instance, an employer cares about an employee's
productivity. However, he cannot deduce the efforts she actually supplied from the
observed productivity, because the productivity of a single agent depends on many other
variables that are not under her control and not observable to the principal (coworkers'
efforts, the productivity of capital, randomness in the production process, etc.). To incite
the agent, the apparent optimal remuneration mechanism would be to linearly index her
wage on her observed productivity. However, if the agent is risk averse, she will not
accept such a payment scheme, as it could provide her with negative or very low
remuneration, even when the poor outcome would not be attributable to her own level of
effort. Because of risk aversion, the agent would prefer to be paid a fixed wage. However,
in that case she would not be motivated to provide her best effort. To solve this
"incentive versus insurance" dilemma, the optimal payment scheme combines a fixed
base pay and a variable bonus indexed on the observed result; yielding a nonlinear
payment scheme.
Into this analytical framework, which was formulated during the first half of the 1980s,
many refinements were subsequently incorporated that considerably extended its reach
(cf., for example, Salani© 1997). First, the theories of adverse selection and moral
hazard were combined. Subsequent extensions included teaming one principal with
several agents, letting informational asymmetry apply to several variables, repeating
interactions over time, etc. Chapter 10 in this volume by Eric Malin and David Martimort
provides a good overview of the analytical strength of this theoretical framework.

3.3 Incomplete contract theory

Incomplete contract theory (ICT) is the most recent. Its initial purpose was to model
some of Williamson's propositions about vertical integration (Grossman and Hart 1986),
but subsequent developments led it in different directions. ICT thus came to examine the
impacts of the institutional framework on contract design, though its roots lay in the study
of the effects of property-rights allocations on the distribution of the residual surplus
between agents and on their incentives to invest.
In terms of its assumptions, ICT is also close to "standard" neoclassical theory. In
particular, agents are deemed to possess Savage rationality. However, it is distinguished
from both Walrasian theory and incentive theory by a key hypothesis. ICT postulates that
complete contracting of agents' future actions is impossible when no third party can
"verify," ex post, the real value of some of the variables central to the interaction between
the agents. Here the institutional framework is no longer implicit. On the contrary, the
issue here is that the "judge," symbolizing the authority that ultimately ensures the
performance of the contract, is incapable of observing or evaluating some relevant
variables “ such as the level of effort or of some investments. It follows that contracting
on unverifiable variables is useless, and other means must be found to ensure efficient
coordination.
To focus on the issues arising from non-verifiability (failure of the institutional framework),
ICT assumes that there is no asymmetry in the parties' information. Both observe all the
available information during each period of trade, while the "judge" cannot verify some of
it, which is therefore non-contractible. Uncertainty arises because each agent has to act
on the non-contractible variable in the absence of complete information on the outcome
of his behavior since he cannot anticipate with certainty what the other will do. Formally,
this is represented by contracting over two periods. During the first period the agents
realize non-verifiable investments. The second period is devoted to trade, the
characteristics of which, in terms of price and quantity, are the only verifiable variables.
This generates a dilemma: since it is possible to contract only on verifiable variables,
agents can commit only on the characteristics of their trade in the second period. Now,
the level of investment realized by the parties in the first period depends upon this
contracted level of trade. However, once the actual level of the investments is known by
the end of the first period, along with the state of nature in which the trade will take place,
the ex ante contracted level of trade is no longer optimal. Ex post, it would thus be
optimal to renegotiate the amount of the trade. But, if the agents anticipate this
renegotiation, they will no longer have an incentive to efficiently invest ex ante (since the
contracted amount of trade is no longer credible).
The solution to this coordination dilemma consists of signing a commitment constraining
the scope of the ex post negotiations in order to provide an incentive to each party to
invest optimally ex ante. This arrangement assigns a unilateral decision right to one of
the parties to determine the effective level of trade ex post, while a default option
protects the interests of the second party by establishing a minimal level of trade. Two
families of models have been created deriving from this framework. The first is
represented by the work of Hart and Moore (1988). An efficient level of investment is not
obtained from the beneficiary of the default option, since this option is insufficiently
sophisticated to motivate him to invest at the optimal level under all conditions. The ex
ante inefficiency follows from the fact that the default option is contingent on the state of
nature that materializes. The second family is an extension to the work of Aghion,
Dewatripont and Rey (1994), who postulate that the default option may provide an
incentive for the beneficiary to invest optimally. They assume that the judge will be
capable of verifying, and of rendering enforceable, default options of great complexity
and that he will oppose any renegotiation of these provisions.

ICT thus establishes a direct link between the ability of judicial institutions to observe or
evaluate the nature of implementable contracts and their efficiency. When some
variables are unobservable, contracts are incomplete. Thus, the capabilities of judicial
institutions determine the level of sophistication of the default clause, which motivates
efficient behavior on behalf of the party that does not benefit from renegotiation rights (i.e.
the right to decide and to the residual surplus).
Though ICT has been the subject of a vast literature it remains less well developed than
IT. This is partly attributable to the dispute between its proponents (especially Oliver Hart)
and those of IT (especially Jean Tirole) and TCT. Tirole (1999) points out a logical
inconsistency between the assumption of agents' perfect rationality and their inability to
implement a revelation mechanism, ex ante, that will force them to reveal to the judge
the true level of their investments, ex post (thus de facto eliminating non-verifiability).
Hart, and other advocates of ICT, reject this criticism. For such a revelation mechanism
to work, it should not be renegotiable ex post. They maintain further that if it were, this
would be tantamount to imputing verification abilities to the judge that he generally lacks.
As to transactions-costs economists, they acknowledge the usefulness of the analytical
framework suggested by IT, but emphasize that it does not draw all the conclusions
implied by the rationality constraints imputed to the judge. If the judge's rationality is
irremediably bounded, as ICT de facto assumes in postulating that he is unable to verify
certain variables, why assume that the contracting parties' rationality escapes similar
limitations? It would be more consistent to resort to a hypothesis of bounded rationality
for all the actors “ the parties and the judge “ as is the case in the TCT (Brousseau and
Fares 2000).
Chapter 11 by Oliver Hart in this volume nicely points out how ICT considerably enriches
the economic analysis of the firm and provides stimulating insights into law and
economics since it is able to account for the impact of the institutional framework upon
the economics contractual practices. Chapter 12 in this volume by Philippe Aghion and
Patrick Rey focuses on the allocation of control rights under various circumstances
among parties facing wealth constraints. It points out how participation constraints
interact with efficiency considerations in designing optimal incomplete contracts.

3.4 The new institutional transaction costs theory
TCT is based on the assumption of non-Savage rationality. This rationality is "bounded"
in the sense of Simon (1947, 1976). This means that agents have limited abilities to
calculate, but also that they operate in a universe in which they do not know, a priori, the
structure of the set of problems that may arise. These agents are confronted with
"radical" uncertainty (in the sense of Knight 1921 or Shackle 1955), rendering them
unable to compose complete contracts.

Contractual incompleteness in TCT can be considered "strong," since it has another
source: institutional failure (Williamson 1985, 1996). As is the case in ICT, institutions
that are ultimately responsible for ensuring the performance of contracts cannot enforce
those clauses that pertain to unverifiable variables. Moreover, judges are also prisoners
of their bounded rationality. They may take a long time before pronouncing judgment,
refuse to rule, make mistakes, etc. Thus, the performance of contracts is not guaranteed
by external mechanisms.
Consequently, the bounded rationality of agents and judges combine to explain the
acceptance of contracts that remain incomplete. To ensure coordination despite the
incompleteness of their contracts, agents must, on the one hand, make provision for
procedures to dictate the actions of each, ex post, and, on the other hand, implement
means to ensure the ex post performance of their commitments. In this case the contract
allocates decision rights to: (a) one, or (b) both of the parties (negotiation procedures), or
(c) to a third party (distinct from the judge). It also puts into place a series of supervisory
and coercion mechanisms to ensure that the parties respect their mutual commitments.
The contract thus creates a "private order," by virtue of which the parties will be able to
ensure each other's cooperation ex post.
TCT facilitates analysis of how economic agents combine commitment constraints “
designed to guarantee the realization of specific investments “ with flexibility constraints
“ needed because of the impossibility of perfectly foreseeing the coordination modes
that would be optimal ex post. Olivier Favereau and Bernard Walliser in chapter 14 in
this volume draw on an analysis formulated in terms of option values to propose an
innovative rereading of the "commitment“flexibility" dilemma originally presented by
Simon (1951). TCT, however, assumes a broader approach, in that it simultaneously
deals with the efficiency of adjustments ex post and constraints on the performance of
contracts:
TCT insists on safeguards to protect each party from the potential for
opportunistic behavior on behalf of the other and to provide incentives to
commit to the transaction. In this regard, it emphasizes the manipulation
of the costs of breaking the agreement “ using security deposits
("hostages") or irreversible investments “ and the length of the
commitment.
The longer this duration, the more difficult it becomes to predict efficient
future adjustments. It thus becomes necessary to redefine the parties'
obligations over the course of the performance of the contract. We here
observe a paradoxical aspect to contractual incompleteness with respect
to the credibility of the commitment: since the parties know that revisions
are possible in the future, they are less inclined to violate their
commitments when the contract does not provide them with an efficient
(or satisfactory) outcome.
Finally, TCT insists on private conflict resolution mechanisms. Since
commitments are open-ended and specific, conflict resolution cannot be
efficiently ensured by outside authorities. Under these conditions, the
contracting parties must agree beforehand on bilateral procedures for
resolving disagreements.
However, owing to the bounded rationality of the agents who design and implement them,
all these bilateral coordination devices remain imperfect. They are also costly to devise
and manage, so the contracting parties will, as much as possible, fall back on collective
provisions emanating from the institutional framework. This latter plays two essential
roles:
First, it provides a basic set of coordination rules, freeing agents from the
need to invent, or reinvent, all of them within their contractual
relationships. For example, external technical standards eliminate the
need to compose a voluminous specification manual, while "common
knowledge" specific to a profession dispenses with the requirement to
formally describe the criteria defining certain characteristics, or behavior,
as "standard" or "fair."
Second, the institutional framework lends credibility to sanctions
guaranteeing the performance of contractual obligations. Reputation, the
selfregulating systems of some professions, and public authorities' power
to regulate and coerce, all provide further support for the contracting
parties.

This has important consequences for the analysis of contracts. On the one hand, the
nature of implementable contractual arrangements is highly dependent on the real
characteristics of the institutional framework, particularly on the makeup of its failings. On
the other hand, the institutional framework cannot be reduced to its public components,
such as the legal environment and the judiciary. Formal collective institutions (such as
professional codes of conduct or "self-regulations" enforced by corporations or
professional associations) join with their "informal" analogs (including behavioral rules
imposed by relational networks such as professions, social and ethnic groups, etc.) to
flesh out the full complement of relevant properties of the institutional framework (North
1990).

3.5 The three base models and their ramifications
The three base models (IT, ICT, TCT) can be represented schematically and juxtaposed
with the Walrasian model (WT is Walrasian Theory) (table 1.1).


Table 1.1: Schematic representation of the different approaches
Contracting External
Principal
parties' institutio
Theory Rationality issue
information ns

WT Savage Complete Perfect Centralized
and (precludin and
symmetric g simultaneo
deviations us
from the establishm
announce ent of all
d plans) equilibrium
prices and
traded
quantities
IT Savage Complete Perfect Disclosure
and (guarantee and
asymmetric ing the incentives
performan ensured by
ce of payment
commitme schemes
nts)
ICT Savage Complete Imperfect Allocation
Table 1.1: Schematic representation of the different approaches
Contracting External
Principal
parties' institutio
Theory Rationality issue
information ns

and (unable to of decision
symmetric verify rights and
some residual
variables) surplus to
motivate
non-
contractibl
e
investment
s
Creation of
TCT Simon Incomplete Very
procedures
and imperfect
for
asymmetric (unable to
decision
verify
making ex
some
post and of
variables
mechanis
and
ms to
subject to
render the
bounded
commitme
rationality)
nts
enforceabl
e

The three alternatives to the Walrasian approach shown in table 1.1 have given rise to
various offshoots or hybrids. In applied economics, in particular, the nature of the issues
dealt with have often made it necessary to move away from the canonical forms of the
three theories. While these theories are somewhat competitive, they should also be
viewed as complementary to the extent that they do not emphasize the same dimensions
of contracts. To simplify, IT focuses on remuneration schemes, ICT relates to
renegotiation provisions that are framed by default clauses, and TCT deals with how
rights to decide, control, and coerce are allocated between the parties. Sometimes a
combination of several approaches is called for to explain a real phenomenon, as was
demonstrated by the work of Holmström and Milgrom (1994) on the internal governance
of firms.
Positive agency theory (Jensen and Meckling 1976, Fama 1980) constitutes one of the
archetypes of these hybridizations. As G©rard Charreaux points out in chapter 15 in this
volume, this theory aims to analyze relationships within organizations on the basis of
assumptions that are quite realistic. Thus, it shares with TCT the notion that efficient
(rather than optimal) coordination results from the combination of several imperfect
contractual and institutional mechanisms. However, positive agency theory emphasizes
the coordination of the allocation of decision rights and the mechanisms governing
remuneration and the assignment of residual incomes (in the tradition of the analysis of
Alchian and Demsetz 1972) and thus also draws on incentive theory.
4 Many fields of application
The application of contract theory to various branches of economic analysis has
generated a multiplicity of results: on the microeconomic level for the analysis of different
types of contractual practices (sub-section 4.1); in macroeconomic reexaminations of the
properties of a truly decentralized economy (sub-section 4.2); and, finally, for the
regulation of interdependence in relationships between individuals within a given
institutional environment (sub-section 4.3).

4.1 A rereading of microeconomic interactions
Recognition of the contract as an object of economic analysis was expanded by the
study of different categories of contractual relations. These studies allowed the theory to
be extended so as to better characterize the coordination regimes effective in certain
industries and to clarify the choices of some economic decision-makers. In management,
for example, studies on efficient methods of coordination with suppliers, partners, or
distributors are legion (cf., for example, in the Strategic Management Journal). In
economics, this research has accompanied the redesign of public policy, especially
related to competition and the regulation of services of general interest (also known as
"public services" or "utilities").

Issues relating to industrial organization have motivated the greatest number of such
studies. In a break with traditional approaches, which focused on anti-competitive
consequences of bilateral relationships, systematic investigation of inter-firm contracting
practices has sought to illuminate their contributions to economic efficiency.
One of the most-studied practices has undoubtedly been contracting between firms and
their suppliers. Subsequent to the landmark case of the relationship between General
Motors and Fisher Body “ one of its suppliers in the 1920s (Klein, Crawford and Alchian
1978; cf. also Benjamin Klein's chapter 4 in this volume and the Journal of Law and
Economics (43 (1), April 2000) that dedicates several papers to this case) “
contemporary industries, especially automobile manufacturing, have seen their
contractual practices repeatedly scrutinized (e.g. Aoki 1988). These analyses have
differentiated between various categories of sub-contracting and partnership
relationships and have examined their impact on firm and industry competitiveness.
During the 1990s comparative analysis of the vertical-integration decision and
partnership contracts provided the frame of reference for tracing the evolution of
corporate practices: be they outsourcing policies resulting from a refocusing on the core
business, or the development of industrial partnerships to increase flexibility in
production and follow the acceleration of the pace of innovation (e.g. Deakin and Michie
1997).

The determinants and consequences of long-term contracts have been researched in
other industries, notably those belonging to the energy sector. They have provided a
better understanding of the economics of negotiation mechanisms and of private conflict
resolution, as well as of the comparative efficiency of contractual adjustment
mechanisms in various contexts. Moreover, the analysis of long-term contracts “ often
associated with the initial phase of the deployment of transportation networks and the
exploitation of new mineral deposits “ has yielded a better understanding of the
feasibility of liberalizing network industries once the initial investment has been
recuperated or the interconnections have multiplied (Joskow and Schmalensee 1983).
Three important results have been obtained in this area. First, contrary to intuition, many
long-term contracts are relatively flexible (Goldberg and Erickson 1987, Crocker and
Masten 1991). Second, these contracts are central to the provision of those utilities that
are indispensable to modern economies “ water, gas, electricity, etc. Third, to some
extent these contracts have proven compatible with other modes of coordination (such
as spot markets), allowing flexibility, security, and freedom of choice to coexist.
Distribution agreements linking manufacturers, wholesalers or the creators of
commercial concepts with distributors have also stimulated a large body of work,
especially on franchising. The franchisor, having created a business model distinguished
by a brand, delegates the actual implementation of this model to others (the franchisees).
Horizontal externalities are generated between the distributors (since the behavior of
each impacts on the shared brand image) as well as vertical externalities between the
franchisor and the franchisees (either of whose actions affect the level of sales). The
franchise system is designed to internalize these externalities as much as possible. This
results both from the specific form of each contract, as well as from the general
architecture of the contractual network, as is underlined in chapter 18 in this volume by
Francine Lafontaine and Emmanuel Raynaud.
Distribution agreements also encompass looser relationships between manufacturers or
wholesalers and distributors “ comprising the wide array of "vertical restrictions." They
are so designated to the extent that these vertical contracts do not limit themselves to an
agreement on the unit price of the goods traded, but also impose de facto behavioral
constraints on the buyer, i.e. the distributor. Price constraints (regressive pricing,
systems of rebates and volume discounts, binding retail prices, etc.) or "non-price"
restrictions (service requirements) implemented in vertical contracts allow various pricing
issues to be resolved (the double-marginalization problem): provision of services related
to sales (consulting, after-sales service), management of competition between points of
sale and between networks. Klein and Saft (1985) and OECD (1994) provide interesting
summaries underlining the complex impact of these practices on social welfare and on
the division of surpluses between distributors and their partners. Benito Arru±ada in
chapter 19 in this volume provides an opportune reminder that the distributor himself
may impose constraints upstream, which may be designed to increase economic
efficiency and not necessarily reveal a desire for more market power.
Another very interesting family of contracts deals with trade in technology and, more
generally, intangibles. In an economy increasingly based on knowledge and information,
arrangements for immaterial transactions become essential. The specific interest of the
case of technology licensing agreements is that it applies to resources that are complex
and imperfectly protected by the body of laws governing intellectual and industrial
property rights. The implementation of efficient contractual mechanisms requires
recourse to specialized collective devices that simplify and secure such transactions (cf.
Bessy and Brousseau 1998). The analysis of the dynamics of trade in technology allows
us to understand how these market infrastructures are progressively assembled. Chapter
21 in this volume by Ashish Arora and Andrea Fosfuri provides an account of such a
dynamic in the chemical industry. The experience acquired by the contracting parties,
the appearance of intermediaries, and the standardization of practices explain the fall in
transaction costs and the multiplication of agreements that foster the dissemination of
information over time.
Agreements governing interconnections between network operators also merit attention
because of their implications for the organization of markets and for competition. As
Godefroy Dang-Nguyen and Thierry P©nard emphasize in chapter 20 in this volume,
these agreements raise issues pertaining to the financial management of externalities
(interconnection tariffs) arising, and from the allocation of property rights to operators.
These questions are now being asked in all networked industries, but they have a wider
relevance since they apply to interdependence between producers of complex product-
services. Production organized as the assembly of elementary components is gaining
ground in many industrial sectors (e.g. computers, automobile) and services (tourism,
banking and insurance).
Finally, a great deal of attention has been paid to the delegation, or concession “
interpreted as contractual (Goldberg 1976) “ by public authorities to private operators of
the production of certain goods or services in a non-competitive environment
(armaments, infrastructure, public goods). Baron and Myerson (1982), Baron and
Besanko (1984), and especially Laffont and Tirole (1993) bolstered the study of
regulation by emphasizing the informational asymmetries between public trusteeship and
regulated firms, galvanizing a search for new regulatory practices. Confronted with the
difficulty of implementing efficient regulations (cf. chapter 23 in this volume by Matthew
Bennett and Catherine Waddams Price), there has been a movement toward opening
the provision of these services to competition. In some cases, however, establishing
competition between operators has proven a difficult task, owing to either the degree of
specialization of the required investment (degree of "specificity," Williamson 1976) or to
the necessity of maintaining a direct, centralized coordination between the supply of, and
the demand for, these services (Glachant 1998, 2002). Public authorities must then
contract efficiently with service providers in a monopoly position. In chapter 24 in this
volume on urban water supply systems, Claude M©nard and St©phane Saussier analyze
the profusion and complexity of choices that arise.

All in all, given that contracts are tools of coordination whose flexibility and adaptability
allow them to be tailored to the exact conditions of their use, contract analysts have been
able to raise doubts about the applicability of traditional theoretical approaches and the
policies they support. The relevant level of analysis is more sub-microeconomic than
traditional microeconomics, because it examines in detail the management of
transactions. The unit of analysis is no longer the market or the industry, but the
transaction. This change in perspective has enriched industrial economics and, more
recently, inspired a renewal in law and economics:
In industrial economics, we are freed from a conception of behavior
exclusively dictated by the structure of the market or of the industry.
Conceptualizations of the nature of the limits of the firm have been
overthrown, and traditional assumptions about the primacy of
technological determinants vigorously contested. A new type of
organizational arrangement has been identified: the "hybrid form."
Relationships between firms are no longer exclusively market based, but
may also draw on a private order, which is relatively stable and
organized in networks (e.g. M©nard 1996).
Studies in the area of law and economics were energized as traditional
beliefs about the efficiency of seeking redress in court, and by extension
in the legislature, in legal rulings and in judges, were called into question
in light of the concepts of bounded rationality and transaction costs.
Several alternative systems of law are now recognized for the
implementation of and enforcement of contracts. The efficiency of
recourse to the law and the judge is now challenged by that of recourse
to "private orders" and private conflict-resolution mechanisms.
This renewal of theoretical analysis has extended even into the domains of economic
decision-making and of public policy design. For example, Victor Goldberg in chapter 8
in this volume emphasizes how legal principles must draw on economic reasoning to
evaluate the legitimacy of some contract clauses that may appear unorthodox at first
glance. But not only contract law is impacted “ similar changes have swept competition
policy. Chapter 22 in this volume by Michel Glais provides an opportune reminder that
the definition of pertinent regulatory exemptions remains open in European Community
(EC) law. We could enumerate other areas of law and public policy, such as insurance,
health, and environmental protection, etc., to which the economic analysis of contracts
can be applied not to mention many dimensions of management.

4.2 The analysis of the functioning of a decentralized market economy

The contractual approach to coordination has had repercussions far beyond the analysis
of bilateral interactions. It is at the root of a renewed analysis of the functioning of a
decentralized economy. Efforts have been made to comprehend the consequences of
substituting the concept of a Walrasian market model with one in which agents meet and
contract in a truly decentralized manner. The economics of labor and employment
constitute the preferred field of application of these new approaches, which are
particularly suited to explaining the rather paradoxical operation of the labor "market"
(e.g. Shapiro and Stiglitz 1984). The theory of implicit contracts prepared the way,
followed by several other approaches “ notably the efficiency wage and labor market
segmentation “ explaining the disequilibria in labor markets on the basis of incentive
contracts.

The theory of implicit contracts (Azariadis 1975) signaled the abandonment of the idea
that economic agents could design a complete system of contingent markets to cover all
eventualities in future states of nature. The wage relationship is understood as a risk-
sharing contract between employees and employers. This implicit contract establishes
wage and employment levels that do not correspond to those of competition market
equilibrium. Despite its flaws, this theory deserves credit for opening a breach in the
preceding orthodoxy.
The theory of efficiency wages represented a second wave beginning in the early 1980s
(Akerlof 1984, Yellen 1984), which ultimately provided new foundations for labor
economics and modern macroeconomics. In the presence of informational asymmetries
between employers and workers, firms cannot rely exclusively on competition or on
internal controls to attract the best professionals and guarantee the required levels of
effort and quality. Incentive contracts fulfill this role by paying an informational rent to the
employee to resolve issues of adverse selection and moral risk. It follows that the price
of labor is higher than its Walrasian value (equal to the marginal productivity of labor)
and that, consequently, labor demand is below supply. This generates an endogenous
disequilibrium in the market on the basis of microeconomic behavior that is perfectly
rational. These results were reinforced by theories of labor market segmentation.

Not only the labor market experiences spontaneous disequilibria, but also markets for
goods and services. This is reinforced when they are characterized by imperfect
competition owing to a concentration of industries, to differentiation strategies, or to price
discrimination. The New Keynesian Economics (Mankiw 1990, Romer 1991) traces from
inter-individual interactions to the formation of global equilibria and macroeconomic
aggregates in order to analyze the properties of market economies and to generate
consequences for economic policy. In general terms, since markets do not
spontaneously move to equilibrium, they appear to have Keynesian properties that,
under certain circumstances, may justify public intervention in order to alleviate the
shortfall in global demand. The great contribution of contract economics is to underline
that price formation at a bilateral level may prevent spontaneous market adjustment. This
failure to adjust is not attributable to external constraints (of a regulatory nature), but
rather to the decentralization of decisions. This is not to suggest, of course, that
regulations and public intervention are exempt from any distortionary effects.

4.3 The analysis of institutions and of the institutional environment

Another field stimulated by the economic approach to contracts has been the analysis of
institutions. Contractual relationships develop in the presence of ground rules that
facilitate their appearance and stability and determine the modalities and the conditions
of their efficiency. These institutions, which define the "rules of the game" and its frame,
constitute what the New Institutional Economics calls the "institutional environment."
Agents enter into contracts on the assumption of the upstream existence of laws that
establish their ability to contract. Consequently, a favorite extension of contract analysis
is the study of the nature and diversity of property-rights regimes. The study of these
regimes' attributes extends well beyond simple legal or administrative rules. It covers all
provisions contributing to the definition of the characteristics of rights of use (measure) or
responsible for limiting access to resources to authorized economic agents (enforcement)
(cf. Barzel 1989). As pointed out and illustrated in chapter 9 in this volume by Gary
Libecap, contract analysis and property-rights analysis can be matched according to two
different approaches. On the one hand, the delineation and distribution of property rights
provide an explanation for why contracting sometimes does, and sometimes does not,
lead to an efficient outcome under various circumstances. On the other hand, contract
analysis sheds light on the circumstances under which a decentralized process can
enable economic agents to establish an efficient allocation and delineation of property
rights. Such analyses are essential for a better understanding of how to manage
economic reforms (e.g. agrarian reforms) and design property-rights regimes for new
economic resources (e.g. information in the digital world).

The study of contractual relationships also relies on the analysis of institutions designed
to assist in their enforcement, be they formal (administration, legal system, but also
professional associations), or informal (culture, traditions and customs). Here economic
analysis joins with other disciplines, especially law, sociology, administrative and political
sciences.
One of the great empirical questions revolves around the viability and efficiency of
transposing contractual arrangements into institutional environments of a fundamentally
different nature. These transpositions may result from expansion of industrial or financial
operators beyond the boundaries of their home countries, or from a transformation of the
institutional environment (i.e. the implementation of the single-market regulatory
framework in the European Union (EU), or the institutional reconstruction of the countries
of the former Communist Bloc). One of the fields that has been most subject to empirical
examination is that of regulated activities (telecommunications, water, electricity, etc.)
(e.g. World Bank 1995, Levy and Spiller 1996, Glachant and Finon 2000). Based on the
analysis of reforms to the electricity sector in various countries, two chapters in this
volume nicely review the issues at stake in the design of so-called "deregulation"
processes (that should more precisely be qualified as "liberalization" processes). Paul
Joskow in chapter 26 emphasizes the idea that the efficient outcome of such processes
relies mostly on the design of an institutional framework able to limit contractual hazards.
Indeed, self-regulation by competitive pressure cannot be sufficient in these industries
characterized by huge fixed costs (and therefore concentration) and interoperability
constraints (resulting in interdependencies and coordination needs among operators).
Guy Holburn and Pablo Spiller in chapter 25 address the problems raised by the need to
design such an efficient institutional frame. Since the instances in charge of regulating
industries are part of a broader institutional set that comprises formal and informal
institutions, the design of devices aimed at monitoring and supervising an industry (or the
competitive process) has to be consistent with the institutional framework within which it
is embedded. Optimal "deregulation" can therefore vary widely across countries, and at
the same time may require broad political or social reforms.

This backdrop to contracts is important because institutions determine the rules of the
game for each relationship. It is also important, however, because contractual
coordination is incomplete by construction. Neither the formation of agents' capacity to
contract, nor their provisions for negotiating, formalizing, or implementing contracts could
exist without the support of other coordination modes. Contractual relationships rest on
informal and incalculable arrangements, such as convention (Orl©an 1994), as well as on
rules or norms controlled by formal institutions. On the whole, contracts do not constitute
a closed universe, and an essential element of the interplay in contractual relationships
comes from their institutional environment (e.g. M©nard 2000).
This broadening of perspective lends some legitimacy to a rehabilitation of public
intervention in the management of relationships between economic agents. It is not a
matter of substituting for them, as was sometimes the case in the past, but rather of
developing efficient infrastructures to promote these interactions. In these matters
conceptions of the role of the public authority have also evolved, since contractual
approaches have contributed to underline the capacity of actors to adapt and organize
themselves. The government should not treat all the structures emanating from agents'
actions as arrangements to be subverted or nationalized, but rather as provisions with
efficiency aspects that should be promoted and deleterious aspects to be curbed (e.g.
collusion). Chapter 7 in this volume by Alan Schwartz outlines the vast research program
opened up by that perspective.
5 Different theories, different methods
Extensions to these various approaches to the field of contract economics have followed
diverging paths. Essentially more hypothetical and deductive, IT and ICT primarily strive
to develop a formal view of the relationships between contracting parties using the most
generic models possible. TCT was developed more from empirical work. However, there
have been several formalizations of TCT, and some tests of IT. Developments in
modeling (sub-section 5.1), on the one hand, and in empirical work (sub-section 5.2), on
the other, thus raise issues addressing all economic approaches to contracts.

5.1 Differences in methodological perspective

Given their foothold in perfect rationality, IT and ICT have not presented any significant
obstacles to the construction of formal models representing the interactions between
agents and the manner in which they conceptualize payments or renegotiation schemes
to resolve issues of asymmetric information or incomplete contracting.

Progress in modeling IT has primarily consisted of refining tools that are increasingly
generic (moving from discrete to continuous cases, moving from models separating
adverse selection and moral hazard to models associating them, moving from models in
which asymmetries pertain to a single variable to multitask models with asymmetries on
several variables, moving from two-party models to models of a principal, an agent, and
an intermediary-supervisor, etc.). In general terms, the evolution of these models has
revealed that the more complex the problem to be solved, the more complicated the
optimal incentive scheme, leading to second-best solutions very distant from the first-
best (i.e. the amount of the informational rent abandoned to the agent increases). As
Arrow pointed out (Arrow 1985), this result is surprising since, in practice, incentive
schemes that are actually used are relatively "rustic" compared to those in the theory.
Moreover, from a normative perspective, these complex schemes are not easily
implementable in the real world. Thus, assumptions have been explored that generate
theoretical contracts closer to observed incentive schemes and that generate simpler
recommendations. This is the goal, for example, of the article by Holmström and Milgrom
(1991) on the fixed wage.

ICT followed a different path. In an effort to replicate the predictions of Coase and
Williamson concerning the vertical-integration decision on the basis of Savage rationality,
it was initially constructed on a collection of purely ad hoc hypotheses. It later evolved
around the search for more generic assumptions that could generate the results of
contractual incompleteness and optimality. This process gave rise to a theory very
different from Williamson's.

TCT was built on a different methodology, being more inductive. It proceeded by
categorization, identifying different classes of solutions to coordination problems. Thus,
three generic categories came into being: "markets," "hierarchies," and "hybrid forms,"
but also a multitude of sub-categories of contract classes (see pp. 16“20 above). The
value of this method is well known, and it underlies the "empirical success story" that is
TCT, according to Williamson. The theoretical propositions of TCT are constructed on
the basis of empirical observations, facilitating the subsequent elaboration of
propositions that are testable on observable variables. However, it also harbors
concealed flaws. On one hand, there is a proliferation of categorizations and typologies
unique to each author, sometimes creating a certain conceptual ambiguity. On the other
hand, TCT must assume that observed contracts are subject to selection processes that
obey the theory's conjecture “ the minimization of transaction costs. This underlies the
claim that the contract types observed most frequently under given circumstances are
those that are relatively most efficient. Now, to be rigorous, it would be necessary to
substantiate the contention that the selection process is capable of eliminating forms of
coordination that generate excessive transaction costs.

Two principal reasons can be given for the methodological features of TCT. First, it does
not rest on a definition of bounded rationality that would allow the decisions of the
contracting parties to be axiomatized. Rationality in TCT is defined only as an absence of
Savage rationality. In this matter the theory remains inductive. Also, TCT does not derive
from a detailed analysis of selection processes that could compensate for the absence of
a specific decision-making model while accounting for the behavior of a representative
agent subject to a selection process, as is the case with evolutionary economics.

5.2 Empirical verification: case studies, econometrics, and experiments

While Hart and Holmström (1987) expressed regret at the absence of empirical
verification of the economics of contracts, such studies have in the meanwhile
proliferated to the point of making an exact count impossible. A survey by Shelanski and
Klein (1995) counted over 150 papers dealing exclusively with the field of transaction
costs (cf. also Coeurderoy and Qu©lin 1997). The two principal characteristics of these
empirical verifications are the coexistence of econometric tests and case studies, and
the large proportion dedicated to the issue of transaction costs.
Questions that have been tested econometrically can be grouped into three families (cf.
chapter 16 in this volume by Scott Masten and St©phane Saussier, as well as Crocker
and Masten 1996). First is the issue of contracts other than those defining a "pure"
commercial transaction. A variant on this approach isolates the duration of the contract
as the relevant variable: Why contract for a non-null duration? For several successive
transactions rather than for only one? Second, the "make or buy" issue is examined:
Why have a good or service supplied internally rather than from an external source?
Finally, econometric tests are also applied to the determinants of the variety of clauses in
contracts: price formulas, guarantees, attribution of decision or supervision powers,
conception of arbitrage mechanisms, etc. Overall, TCT has presented the largest
number of testable propositions for these three types of empirical verification. For the
aforementioned methodological reasons, it is sometimes the only theory with anything to
say on the subject. Moreover, so far its propositions have successfully withstood many
attempts at econometric refutation. IT, however, has yielded explanations of the
incentive effects of different forms of land rental (i.e. farming versus sharecropping;
Stiglitz 1974) or remuneration provisions in franchise contracts (while at the same time
finding its propositions pertaining to the risk-aversion hypothesis discredited). Salani©
(1999) presents the econometric literature, of which there is still a dearth, on IT. These
differences between the treatment of TCT and IT are attributable to the restrictive
assumptions of the latter, which make it difficult to formulate testable assumptions on
empirical data. As to ICT, so far it has been the object of only a handful of tests, limited
exclusively to the issue of vertical integration. There, again, very strict assumptions
render econometric testing delicate.

The difficulty of formulating testable propositions is only one of the problems
encountered when testing theories of contracts. Gathering data is also a significant
obstacle. First, obtaining information on in-force, or recently ended, contracts is
hampered by issues of confidentiality. Next, constructing the databases presents
methodological difficulties specific to the coding and normalization of the descriptions of
the contents of contract documents. Finally, econometric tests are stymied by the poor
quality of available data, be it on the contracts themselves or their explanatory variables.
Such are the reasons why case studies continue to play a role, universally recognized as
irreplaceable, in empirical verification. Given this context, legal scholars and managers,
being anchored in the practice of case studies both in their academic training and in the
day-to-day functioning of their professions, have occupied a prominent position with their
work.
It should be noted that econometrics is not the only discipline capable of subjecting
theoretical propositions to rigorous protocols of empirical verification. The controlled
nature of investigations conducted by the practitioners of experimental economics lends
itself to testing conjectures arising from very strict hypotheses like those of IT. Thus,
Claudia Keser and Marc Willinger in chapter 17 in this volume demonstrate that most
contracts presented in experimental tests do not respect the incentive constraint as
conjectured by IT, either in single-period or repeated principal“agent interaction
simulations. These results do not contradict the optimization assumption, but rather
reveal the presence of other motives in the contract relationship, such as equity and
reciprocity (suggesting the principles of contract law evoked in Jacques Ghestin's
chapter 6 in this volume).
6 Perspectives
The future of the economic analysis of contracts is contingent on progress in four areas:
the measurement and collection of data (sub-section 6.1); modeling bounded rationality
(sub-section 6.2); analysis of the institutional framework (sub-section 6.3); and, finally,
collaboration with professionals and scholars in other disciplines (sub-section 6.4).

6.1 Measurement and data collection

Significant improvements are expected in the availability of empirical data. One key
limitation that has hampered the evolution of the economic analysis of contracts to date
is that of collecting data appropriate to the issues it raises. Official statistical agencies
are focused on measuring phenomena whose scope are macroeconomic or pertain to
the microeconomics of markets or industries. The sub-microeconomic level, that of the
contract and the transaction, is not recognized and will not readily be recognized
because of confidentiality issues (trade secrets). A further issue of "measurement" is that
dimensions useful for the analysis of contracts are not part of the available accounting or
statistical standards. Until now, gathering the appropriate data has largely relied on
individual investigations and the voluntary participation of a few firms. The cost of these
collections and their near cottage-industry character explains the small size of the
available databases as well as their heterogeneity. In the future a more efficient
compilation could come from: first, recovering individual series already identified in
official statistical data-sheets of a microeconomic nature; second, gaining access to
databases used for voluntary inter-firm benchmarking or anti-trust purposes; third,
developing and using trade-specific databases maintained by private or public
foundations or professional associations. These types of advances can already be seen
when a scientific evaluation of professional practices is required in response to
challenges under evolving regulations.

6.2 Modeling bounded rationality
The formalization of different elements of the economic analysis of contracts and,
consequently, the generation of testable propositions, is still deficient. A major
shortcoming in this field is the modeling of bounded rationality. In the absence of models
adapted to the specification of the rationality of the contracting parties, formalized
analytical constructs rely on assumptions of hyper-rationality to deal with behavior
originating from semi-strong rationality. In this process, however, the observed behavior
and the stylized facts that should be explained are largely eliminated. An important
aspect of the future of the economic analysis of contracts thus depends on the possible
development of models of bounded rationality. Two possible avenues present
themselves. One begins with the standard model of rationality and proceeds to explore
various aspects of the degeneration of rationality. The work by Bentley MacLeod, in
chapter 13 in this volume, provides a good example of this type of approach. The other
approach explores the way in which actors' rationality is formed and how deductive
reasoning ties into collective and social patterns of behavior to model their choices,
values, and routines. Here, the contributions of psychology, sociology, and anthropology
are mobilized along with the more traditional methods of economists. Simon's work
constitutes an essential reference.

Reverting to current models of rationality will provide for a better understanding of how
contracts are conceived and evolve over time under the influence of learning and
selection processes. Special focus should be placed on the coordination difficulties that
are solved by contracts, as this will facilitate a rigorous analysis of the design and
consistency of the various contractual mechanisms. These are, indeed, "systems" that
we have not yet been able to consider with sufficient rigor (for a first attempt at this, see
Brickley 1999).

6.3 Selection processes
As pointed out by North (1990), and earlier by Alchian (1950), the processes according
to which viable contractual or institutional forms are selected is of importance as well.
While the design of contracts and institutions depends upon agents' behavior, the
competitive process validates or invalidates agents' choices. IT, ICT and TCT implicitly
(for the two former) or explicitly (for the latter) assume that selection is perfect and
eliminates less efficient (or more costly in terms of transaction costs) coordination
devices. As demonstrated by advances in evolutionary theory, both in economics and in
biology, evolution and selection processes lead neither to a unique and final equilibrium,
nor to an optimum. Economics in general, and contractual economics in particular, lacks
a satisfactory approach to selection processes, though such a theory of selection would
be essential to the definition of some efficiency criteria that would be more realistic than
the standards "maximizing revenues, minimizing costs." Indeed, "efficient" could also
mean "flexible," "favorable to innovation," "remediable," etc. In a sense, the contribution
by Eirik Furubotn, in chapter 5 in this volume, is a good example of the broadening of
perspective needed to build a more satisfactory analytical framework for the study of the
properties of a truly decentralized economy and for identifying strategies that are both
sustainable and preferable in terms of individual or collective welfare. There are,
however, other research directions to be explored. The analysis of competition among
alternative contractual and organizational forms, innovation in contract design, learning
by governing, and learning about governance mechanisms (etc.) thus open quite a wide
research agenda. This is pointed out by Ronald Coase in chapter 2 in this volume.

6.4 Institutional framework and enforcement
Significant progress is also expected from a better understanding of the effects of the
institutional framework on contract choices. A program of work along those lines has
already been initiated (cf. sub-section 4.3, but also Aoki 2001). More generally, a
multiplication of comparative studies conducted on the variety of contracts governing the
same professions within the same industry in different institutional environments can be
expected. These will doubtlessly allow a better identification of those characteristics of
the environment relevant to the conception of contractual arrangements, as well as an
analysis of factorsinfluencing the relative performance of these arrangements. In
exchange, such analyses will open the door to the design of institutional frameworks that
are more efficient while respectful of current practices.

6.5 Cross-disciplinary fertilizations
Finally, theoretical developments remain highly dependent on a better understanding
and grasp of empirical reality. The economic analysis of contracts should benefit from
closer and more promising collaborations with professionals and scholars in other
disciplines. Many professionals, in business as well as consulting, but also working in
national and international institutions, seek such exchanges ("will perform analysis in
exchange for access to data"). An entire sector, that of the legal professions “
representing an operational rather than an academic discipline, Law “ is expressing a
growing demand for economic analysis of legal cases and offering the basis for a joint
labor in "Law and Economics." Research in management, political science,
administrative science, sociology, and history should also stimulate the economic
analysis of contracts by suggesting both propositions and hypotheses or as a source
of building blocks, empirical evidence, and issues to be addressed.
Note
Chapter 1 was originally published as "Economie des contrats et renouvellements de
l'analyse ©conomique," in Revue d'Economie Industrielle (92, 2000).
Canonical Models of Theories of
Appendix:

Contract
M'hand Fares
1 Incentives theory
The objective of incentives theory (IT) is to analyze situations in which a contract is
contemplated under conditions of asymmetric information, that is, where one party (the
agent) knows certain relevant information of which the other party (the principal) is
ignorant. Usually two kind of situations are considered. In a moral-hazard situation, the
principal cannot observe the agent's actions or decisions. The solution is then to define
adequate contract terms in order to internalize incentives. In an adverse-selection
situation, before signing the contract, the agent is aware of private information on his
characteristics (his type). The solution is to let the agent choose between several
alternative contracts in order to reveal his private information.

1.1 Moral hazard
Let e represent the effort of the agent (he) and y = e + the production result observed
N(0, 2).
by the principal (she), with a parameter of agent productivity and
Following Holmström and Milgrom (1987), we assume that the principal offers a linear
incentive scheme, t(y) = B + y, to a risk adverse agent. The agent's risk aversion is
rw
captured by a CARA (Constant Absolute Risk Averse) utility function, u(w) = e ,
2
where r represents absolute risk aversion and w wealth, with w = B + y. As y N(0, ),
2
then w N(w, w). So, it is possible to evaluate the distribution of wealth using the
[1]
function



The agent utility will be given by , where g(e) represents the
cost of effort. The agent program is then


The first-order condition is


The principal is supposed to be risk neutral. Her expected profit is given by E [y t(y)] =
E [ e+ B e ] = (1 ) e B. She determines the optimal parameters
and B that maximize her expected profit.
Under symmetric information, the principal observes the agent's effort. The linear

. 1
( 18)



>>