Polycentric Games and Institutions
[Volume 3] March 24, 1999
Polycentric Games and Institutions:
Readings from the Workshop in Political Theory and Policy Analysis
Edited by Michael D. McGinnis

Series Foreword
Michael D. McGinnis

Michael D. McGinnis

Part I. Developing a Framework for the Analysis of Institutions

Introduction to Part I (Michael D. McGinnis)

1. Public Choice: A Different Approach to the Study of Public Administration {To Reference}
Vincent Ostrom and Elinor Ostrom

2. The Three Worlds of Action: A Metatheoretical Synthesis of Institutional Approaches {To Reference}
Larry L. Kiser and Elinor Ostrom

3. An Agenda for the Study of Institutions {To Reference}
Elinor Ostrom

4. A Grammar of Institutions {To Reference}
Sue E.S. Crawford and Elinor Ostrom

Part II. Voting, Conflict, and Leadership

Introduction to Part II (Michael D. McGinnis)

5. Votes and Vetoes {To Reference}
Roberta Herzberg and Vincent Ostrom

6. Negative Decision Powers and Institutional Equilibrium: Experiments on Blocking Coalitions {To Reference}
Rick Wilson and Roberta Herzberg

7. Policy Uncertainty and Two-Level Games: Examples of Correlated Equilibria {To Reference}
Michael D. McGinnis and John T. Williams

8. Shepherds and their Leaders Among the Raikas of India: A Principal-Agent Perspective {To Reference}
Arun Agrawal

Part III. Rules, Regulations, and Resource Management

Introduction to Part III (Michael D. McGinnis)

9. Heterogenous Players and Specialized Models {To Reference}
Eric Rasmusen

10. Governing a Groundwater Commons: A Strategic and Laboratory Analysis of Western Water Law {To Reference}
Roy Gardner, Michael Moore, and James Walker

11. Bottlenecks and Governance Structures: Open Access and Long-Term Contracting in Natural Gas {To Reference}
Thomas P. Lyon and Steven C. Hackett

Part IV. Models of Monitoring and Sanctioning

Introduction to Part IV (Michael D. McGinnis)

12. Transforming Rural Hunters into Conservationists: An Assessment of Community-Based Wildlife Management Programs in Africa {To Reference}
Clark C. Gibson and Stuart A. Marks

13. Irrigation Institutions and the Games Irrigators Play: Rule Enforcement on Government- and Farmer-Managed Systems {To Reference}
Franz J. Weissing and Elinor Ostrom

14. Coping with Asymmetries in the Commons: Self-Governing Irrigation Systems Can Work {To Reference}
Elinor Ostrom and Roy Gardner

Part V. Laboratory Experiments and Self-Governance

Introduction to Part V (Michael D. McGinnis)

15. Neither Markets Nor States: Linking Transformation Processes in Collective Action Arenas {To Reference}
Elinor Ostrom and James Walker

16. A Behavioral Approach to the Rational Choice Theory of Collective Action {To Reference}
Elinor Ostrom

Suggested Further Readings
List of Contributors

Links to the Web Pages:

  • Series Foreword: Self-Governance, Institutional Analysis, and the Workshop
  • Polycentric Governance and Development
  • Polycentricity and Local Public Economies
  • Polycentric Games and Institutions

  • 3rd Revised Draft -- January 19, 1999

    Polycentric Games and Institutions
    Readings from the Workshop in Political Theory and Policy Analysis


    Michael D. McGinnis

    People establish and use institutions to help them achieve a vast array of individual and collective goals. As a consequence, an understanding of institutions is crucial to any social science. In recent years, game theory, formal models, and experimental research have become valued components of the methodological tool-kit used by political scientists, policy analysts, economists, and other scholars interested in the analysis of institutions. These analytical tools are particularly useful for highlighting the basic structure of decision problems facing individuals and collective groups. Game models, in particular, direct attention to the basic configuration of common, conflicting, and complementary interests facing participants in different settings. Simplicity in representation is key to the success of game theory, but this strength comes with inherent limitations (McGinnis 1991). For example, simple formal representations of institutions are too abstract to capture all of the important aspects of inherently complex institutional arrangements.

    Several scholars affiliated with the Workshop in Political Theory and Policy Analysis at Indiana University have made important contributions towards bringing a richer empirical content to formal models and experimental analyses of institutions. The readings collected in this volume illustrate several varieties of institutional analysis as exemplified in the research of Workshop scholars. Each reading builds upon the foundation of game theory to use different levels of formal or experimental rigor to address similar sets of questions concerning institutions and governance.

    Contributors address such factors as the physical nature of goods, the relationship of majority voting to other procedures by which collective choices are made, the normative foundations of utility functions, alternative legal doctrines and property rights systems, the effects of government regulations on markets, the incentives facing actors engaged in monitoring and sanctioning activities, and the origins of the normative expectations upon which individuals base their choices in field settings and in laboratory experiments. Each reading illustrates different ways in which the basic principles of game theory can be used applied to deepen our understanding of particular empirical situations or to clarify the meaning of such important analytical concepts as norms, rules, and institutions. In totality, these readings illustrate the range of the contribution of scholars affiliated with the Workshop in Political Theory and Policy Analysis.

    Polycentric Games

    I coined the term polycentric games for use in the title of this volume to capture the distinctive nature of the Workshop approach to formal models of institutions. This term extends the logic behind the widely-known and influential concepts of two-level games (Putnam 1988) and nested games (Tsebelis 1990). Putnam’s model emphasizes that potential problems with the domestic ratification of international agreements can affect the process by which those agreements are reached in the first place. Tsebelis demonstrates that actions that seem irrational in one context may be perfectly understandable once analysts incorporate that actor’s strategic interactions with other actors. In both cases, the important point is that analysts must expand the scope of their models to consider the cross-effects of concurrent games.

    Polycentric games encompass an even wider array of interactions. A basic presumption is that actors involved in any single interaction can draw on a vast array of informational cues to help them understand the behavior of other actors. In particular, they can draw on commonly understood norms or rules, and they can use institutional procedures to organize their interactions. But it’s not enough to say that many games are occurring simultaneously; it’s also essential to provide some guidance concerning how to handle the resulting complexity. As the readings included in this volume demonstrate, Workshop-affiliated scholars have developed conceptual frameworks and analytical tools that bring some coherence to the study of polycentric games.

    Levels of Analysis and Arenas of Choice

    Research on two-level or nested games directs attention to connections across domestic and international levels of analysis, defined in terms of scales of aggregation, from individuals to small groups to organizations to national governments to international systems (see Russett and Starr 1996). International relations scholars use these levels to differentiate among supposedly distinct sets of explanatory factors, each of which brings a unique perspective to the study of international relations (see Waltz 1959). In this way, personality or perceptual factors at the individual level of analysis can be kept distinct from the general characteristics of organizational behavior as well as the contrasting tendencies of democratic and autocratic governments, and other factors operating at the systemic level tend to maintain a balance of power among contending Great Powers. Even though, as suggested in the emerging literature on two-level or nested games, it may be essential that any complete explanation incorporates factors at different levels of analysis, the basis of separation among levels of aggregation seems solid.

    To study polycentric games requires a shift of focus to different arenas of choice: operational choice, collective choice, and constitutional choice. The distinction between these three arenas of choice is an essential component of the Institutional Analysis and Development (IAD) framework, which is explained more fully in Chapter 2 of this volume (and in Ostrom, Gardner, and Walker 1994).

    In the operational choice arena, concrete actions are undertaken by those individuals most directly affected or by public officials. The outcomes of these actions directly impact the world in some demonstrable manner. The rules that define and constrain the operational activities of individual citizens and officials have been established by processes occurring in arenas of collective choice, and the rules by which these rules themselves are subject to modification are determined in the arena of constitutional choice. In some circumstances, constitutional choice results in preparation of a written constitution, but more generally communities develop informal shared understandings about the ways in which that community organizes itself to make collective decisions. These shared understandings are an essential component of the decisional context for collective choice and operational activities.

    Authors of the essays reprinted in this volume originally used the term “level of analysis” to differentiate among what I am calling different “arenas of choice.” However, levels of analysis are closely associated with scales of aggregation in other research literatures, and the IAD framework points towards distinctions that need not be limited to scales of aggregation. The interactions characteristic of any single arena of choice can occur at different scales of aggregation or among the same group of individuals at different points in time.

    For example, constitutional choice typically involves a wider range of participants than routine operational choices. But even the same set of individuals may interact in one way when they are deciding how they will make future decisions and in quite another (more competitive) manner when deciding the allocation of scarce resources in a single time period. In other words, constitutional choice can be analytically distinguished from times when participants are simply debating the application of commonly-held principles or rules to the current situations. The crucial point is not the number of people involved, but rather the mode of activity occurring in different settings or time periods. These distinctions are explained more fully in later chapters.

    Important analytical similarities apply to all three arenas of choice. In each arena, individual and collective choice are constrained to some range of strategic options. Actors confront an action situation with strategic options and role expectations as defined during periods of more encompassing interactions (analogous to “higher” levels of analysis). In each arena, the choices of actors jointly produce patterns of interactions and outcomes which shape the nature of their interactions in the other arenas (especially in those corresponding to “lower” arenas). Influences move back and forth across arenas in complex but understandable patterns.

    In short, institutions matter because they link arenas of choice by defining the roles that individual or collective actors fulfill. If one looks at political situations in this manner, it is clear that all three arenas (or centers) of choice are involved in any one particular application. The term “polycentric” (originally coined by Ostrom, Tiebout, and Warren 1961) nicely conveys an image of a network of overlapping and inter-linked arenas of choice. The essence of a “game” in the technical sense lies in the nature of interactions among participants with at least some difference of interest. Since games of operational choice, collective choice, and constitutional choice are occuring concurrently, the term “polycentric games” seems particularly apt as a summary statement.

    This broad conceptual framework provides an overall context for each of the more specialized analyses included in this volume. Foundational constitutional questions are rarely in doubt during the routine operational decisions that are most often the subject of game models, and yet the outcome of processes of constitutional choice cannot be ignored, for such deliberations determine who has the capability or the responsibility to participate in operational decisions. Such situations are too complex to model in their entirety, but the essays collected here demonstrate that important insights can be garnered from partial models of specific situations, provided ample consideration is given to concurrent or supplementary processes in other arenas of choice.

    Organization of This Volume

    The readings collected in this volume are organized into five parts. The essays in Part I lay out inter-related frameworks for analysis: the IAD framework mentioned above, a fuller representation of the action situations occurring in each arena of choice, and a conceptual grammar that differentiates among strategies, norms, and rules. Each of the remaining Parts directs attention to particular empirical settings or types of institutional arrangements. Part II focuses on voting institutions and other ways in which the interests of principals and agents can be reconciled. Part III investigates particular configurations of rules and regulations, especially related to the management of natural resources. Workshop scholars have demonstrated the crucial importance of monitoring and sanctioning in maintaining effective systems of resource management and of governance, and the chapters included in Part IV use different game models to investigate the expected performance of monitors with different incentive structures. The two essays in Part V provide overviews of an extensive body of experimental research that indicates that the behavior of rational individuals is more nuanced and variated than is generally appreciated. The final essay in this part, Elinor Ostrom’s recent Presidential Address to the American Political Science Association, serves as a concluding essay. In this address she points toward future development of a “second generation” of rational choice models capable of more fully incorporating the behavioral determinants of individual action, and thus of the consequences of institutions.

    As noted in the Series Foreword, this volume has been compiled in conjunction with two other edited volumes of previously published research by scholars affiliated with the Workshop in Political Theory and Policy Analysis. This particular volume includes some of the more technical pieces of research selected from a much broader array of research methodologies used by Workshop scholars. It is primarily intended for use by scholars and students with some expertise in game theory and experimental research.

    One unfortunate consequence of this decision to concentrate technical readings in a separate volume is that it may appear as if this formal or experimental work was tangential to the development of the extensive empirical research programs also associated with the Workshop. Nothing could be further from the truth, for game models and laboratory experiments have played important roles in the evolution of Workshop research programs on metropolitan governance, common-pool resources, and development (see McGinnis 1999a,b). Formal models have helped focus attention on the fundamental nature of the action situations relevant in each of these areas of application, whereas the relatively controlled environment of the laboratory setting allows researchers to isolate the effects of particular factors on the ability of groups of individuals to cooperate for their common good. The benefits of cross-fertilization across formal models, experiments, and field research have been demonstrated in Ostrom, Gardner, and Walker (1994). Formal and experimental methods supplement the strengths of field research; no one method can tell us everything we need to know. Readers of this volume are encouraged to keep in mind the broader context of the formal and experimental research reported here.

    Because of the diversity of topics covered in this volume, each Part begins with an introductory essay that places the works included in that Part within the broader context of the multi-faceted research programs implemented by Workshop-affiliated scholars. These brief introductions focus on the essays included in this volume, along with a few closely associated works that could not be included here. The remainder of this general introduction provides a quick tour of the vast expanse of the field of modern political economy. This sketch cannot hope to be comprehensive; its only purpose is to locate the unique contributions of institutional analysis within the broader context of other analytical approaches to the study of institutions, public policy, and constitutional order.

    Institutional Analysis and Related Areas of Research

    Methodological Individualism, Markets, and Games

    Institutional analysis builds on the basic principles of methodological individualism. The individual is taken to be the foundational point of analysis, since individual choice is crucial to all social outcomes. But it is equally important to understand the institutional context within which individual choices occur, as well as the ways in which individual choices shape these institutional contexts. Institutions matter, in many different ways, but, ultimately, their impacts are filtered through the choice of individuals.

    This emphasis on the joint importance of individual choice and institutions can be usefully contrasted to the common (but by no means universal) attitude among economists that markets are a natural phenomenon. In other words, a presumption is often made that, if you provide individuals with endowments of goods or skills, then they will automatically begin to engage in mutually beneficial exchanges. In the proper circumstances, voluntary exchanges of private goods do indeed result in collectively beneficial outcomes, as if guided by an “invisible hand” (Smith 1976 [1776]). But markets are themselves institutions that operate at optimal efficiency only if the goods subject to exchange are private goods and only in the context of supporting political and legal institutions. In other words, markets are themselves potential subjects of institutional analysis, even though their existence and their efficiency are typically taken for granted.

    For the most part, however, Workshop-affiliated scholars have investigated situations in which market exchanges are inappropriate or ineffective. Individuals are still presumed to be the best judge of their own interests, but they also have a much wider array of options from which to choose. Communities of individuals may establish diverse institutions to help achieve their common goals. Even in the context of cooperative institutions, however, individuals may still have an incentive to engage in opportunistic behavior.

    This ever-present tension between the potential benefits of cooperation and the temptations of individual opportunism is the central problem of modern political economy. In recent years, economists, political scientists, other social scientists (and even a few natural scientists) have come to rely on game theory as a fundamental tool for the analysis of interactions among individuals with common, conflicting, and complementary interests (see Kreps 1990a,b). By focusing attention on strategic interactions among actors who pursue their own interests, game theory helps clarify how the independent actions of individual or other actors may result in collective outcomes that none would have preferred.

    Schelling (1960) defines game theory as a “theory of interdependent decision,” but it is better seen as a set of tools that can be applied to model a wide array of situations. The content of the game theoretic tool-kit has changed over time, in the sense that alternative solution concepts have been developed for application to different types of games, based on different conceptualizations of the nature of human cognition and decision. Workshop scholars have contributed to this ongoing dynamism in ways that are illustrated in the works included in this volume.

    Market Failure and the Hierarchical State

    The standard departure point for modern political economy is to treat governments as a response to market failures (see, for example, Weimer and Vining 1989: chapter 3). Markets are an efficient way to organize the production and exchange of private goods, but it is widely recognized that competitive markets may have less desirable consequences for the production of public goods or for the management of a common-pool resource.

    Workshop scholars have contributed to extensive (and on-going) empirical research programs on the provision of public goods and services in metropolitan areas and on the management of common-pool resources in mostly rural settings throughout the developing world (see McGinnis 1999a,b). In the process of these empirical investigations, Workshop scholars have also investigated the ways in which formal models can be used to expand the analytical repertoire of institutional analysts. The present volume focuses on these conceptual contributions.

    The canonical alternative to privatization is to rely on “the government,” conceptualized as a Hobbesian sovereign capable of imposing its policy decisions on recalcitrant individuals. Those who might seek to free ride on the contributions of others can, in this way, be forced to contribute in the form of taxes or similar extractions. But hierarchical structures have their own dilemmas, specifically related to agent-principal relations. How can the ruler (as principal) insure that its agents have the proper incentives to gather accurate information or to implement the sovereign’s decisions properly? Such problems are impossible to resolve completely. Miller (1992) surveys the dilemmas of hierarchical rule, and concludes that it is important for top officials to exert leadership if they are to elicit productive effort from their subordinates. Leadership by itself is not enough, but neither is reliance on contracts or coercion. (Lichbach 1996, arrives at a similarly eclectic decision in his comprehensive survey of the literatures on collective action theory.)

    From the point of view of political science, the limited ability of citizens to monitor or sanction their own rulers presents the fundamental problem of governance. Without this capacity, rulers have a tendency to shirk their responsibilities and to extract excessive amounts of resources for their own personal use. Liberal democratic institutions attempt to make rulers accountable to the general population via the mechanism of elections. The basic idea sounds simple enough: those public officials who fall short of public expectations can be removed from office in the next election. As often happens, however, this idea does not turn out to be so simple in application.

    The Limitations of Voting as a Method of Collective Choice

    Much of the modern political economy literature deals with various aspects of voting institutions. Social choice theorists (see Arrow 1963, Riker 1982) have demonstrated that there is no assurance that consistent policy will emerge from processes of majority vote, or, for that matter, from any means of aggregating individual preferences into social outcomes in a way that satisfies basic criteria of fairness. If candidates are only interested in obtaining and retaining office, then in some contexts (specifically, when the relevant policy issues can be represented on a single dimension) competing candidates will tend to offer nearly identical positions, in hopes of winning the approval of the “median voter” (Downs 1957). Conversely, under conditions of multidimensional issue spaces practically any outcome could emerge as a result of agenda manipulation or strategic voting (McKelvey 1976, Schofield 1978, Riker 1982). Myriad institutional arrangements have been devised to limit the volatility of majority rule. For example, committees may achieve coherence by limiting the range of alternatives considered (Shepsle 1979) and the nature of committees may help structure overall legislative outcomes (Krehbiel 1991).

    From the Workshop perspective, voting is only one way in which collective decisions may be made. And in many of the empirical contexts studied by Workshop scholars, specifically the management of common-pool resources, majority rule plays at best a minor role. Some Workshop scholars have focused on the implications of voting institutions (see Part II below), but, overall, a distinctive aspect of institutional analysis is the relatively low profile given to studies of voting systems. Voting is treated in the context of other institutional rules, and not as an end in itself.

    As mentioned above and detailed more fully in Chapter 2, operational choice games are directly affected by the accepted procedures for collective choice processes, which were in turn determined by some process of constitutional choice. Buchanan and Tullock (1962) suggest one distinction between constitutional and operational choice, arguing that constitutional choice requires unanimity (at least in broad outlines) whereas choices in different issue areas might be determined according to majority vote or extraordinary majority rules. But there is no reason to restrict attention to voting as a means of making collective decisions in any of these arenas.

    However decisions are arrived at, one key issue is the process by which subsequent behavior is monitored and rule violations sanctioned. Principals have to be able to monitor the behavior of their agents before they can sanction them for poor performance; monitoring is equally essential in the context of interpersonal interactions within a given community. But agents typically serve multiple principals with divergent interests, which means that decisions by political agents will help some groups and harm others.

    Successes and Failures of Collective Enterprise

    Principal-agent problems can be ameliorated (but never completely eliminated) if the social distance between principals and agents is reduced. This can occur most easily in the context of a local community or a small group of people who share some intense interest in common. For example, local communities of farmers or fishers share a common interest in maintaining the quality of their farmland or fisheries, but each of them may also seek to maximize their own income. Although these problems might seem mundane to outside observers, these are matters of life and death to the people directly involved. Thus, analysis of community efforts to manage these common-pool resources can provide important insights into the basic dilemmas of collective action.

    The Workshop approach to institutional analysis is based on the assumption that group efforts to manage common-pool resources should be granted the same status as individual or corporate rights to private property. Just as individuals are presumed to be the best judge of their own tastes, user groups should be presumed to be capable of managing their common property. A basic tenet of public policy should be that those groups who are able to manage CPRs effectively should be allowed (and encouraged) to do so, with government intervention undertaken only when user groups fail to manage their resources effectively, or if user groups violate general standards of fairness, accountability, or other widely shared concerns. Instead of presuming that governmental officials or scientific experts know best how to manage CPRs, user groups should be given the initial benefit of the doubt.

    Workshop-affiliated scholars have demonstrated that many self-governing communities in all parts of the world have proven themselves capable of managing their common-pool resources in an effective and sustainable manner (see Ostrom 1990; Ostrom, Gardner, and Walker 1994; McGinnis 1999a). Such efforts are not always successful, as there are certainly examples that more closely resemble a “tragedy of the commons” (Hardin 1968). In cases of community failure, government intervention may be essential. Furthermore, the capacity of some user groups to manage their common-pool resources in an effective manner does not imply that governmental intervention is never appropriate. Whether that intervention should come in the form of centralized management by governmental authorities or dividing up the resource into privately managed segments (or by some other institutional arrangement) is a question that can only be determined by careful consideration of the details of particular cases. The general point is that no one institutional solution will work in all empirical settings.

    How are we to determine when different policy instruments are most appropriate? As discussed above, the “market failure” model is commonly used to determine the circumstances under which government intervention is most appropriate. The basic idea is that since market exchange is an efficient way to organize the production and allocation of private goods, governments should intervene only in those circumstances for which private markets are unable to cope. Thus, governments should be responsible for providing public goods (such as national defense), which would be under-provided by private markets. Another responsibility of government is to provide the legal framework within which economic exchange occurs, to limit the exercise of private coercion and to ensure that contracts can be enforced at a relatively low cost.

    In effect, Workshop scholars assert that an analogous model of governmental intervention as response to “group failure” should be accorded comparable status. The rights of user groups to manage common property and individual (or corporate) rights to private property should have equal status in law and policy. Just as individuals are presumed to be the best judge of their own tastes, the initial presumption should be that user groups are capable of managing common property. Government intervention should occur only to correct problems of “group failure,” defined in terms directly analogous to “market failure.”

    One immediate objection to this assertion is that there is no reason to be sure that groups will manage their resources in an equitable manner. This is certainly a valid concern, for one of the most effective ways to facilitate cooperation among a group is to exclude from membership those people who have significantly different interests. It is certainly reasonable to allow government intervention to prevent gross violations of general standards of fairness, but exactly the same restrictions are typically applied to markets in private goods. Governments routinely prohibit certain kinds of exchanges in the interest of supporting community standards of morality. For example, in virtually all cultures private markets in sex services are made illegal or restricted in some fashion.

    In sum, the analogy between “market failure” and “group failure” is quite close. In both cases the initial presumption argues against government intervention, but allows such action to prohibit certain actions deemed unfair by society at large. Considerable room for debate remains open in the determination of which practices should be prohibited, and it is the responsibility of institutional analysts to provide a solid empirical foundation for these collective deliberations.

    Markets, states, and user groups should be seen as complementary institutions. None can live up to their potential without support from the others (see Lichbach 1996). Unfortunately, markets and states have received the bulk of attention in the policy literature, and the record of research by Workshop-affiliated scholars serves as a corrective to this imbalance.

    Transaction Costs, Property Rights, and the International System

    By building on an extensive set of case studies completed by scholars from all parts of the world, Ostrom (1990) demonstrated that communities throughout the world have made use of many, many alternatives to market exchange and hierarchical organizations. The diversity of institutional arrangements cannot be dismissed as a consequence of the low level of sophistication among user groups. In his extended analyses of the basic institutions of modern capitalism, Williamson (1975, 1985, 1996) has demonstrated that different kinds of nonmarket, nonhierarchical institutions are most appropriate for different empirical contexts. For Williamson, the crucial factor is the extent to which different institutional forms minimize the costs of transacting between individuals or corporate actors. The basic problem is that potential gains from trade may not be realized if the actors do not have sufficient confidence that the other side will refrain from taking advantage of any temporary or limited dependency this exchange might imply.

    Williamson was reacting to a tendency in the economic literature to focus on markets as spontaneous and firms as hierarchical forms of organization. A similar distinction has played a prominent role in defining the terms of discourse in the field of international relations. Waltz (1979) asserts that international systems are anarchical, and that these systems are composed by hierarchical states. A balance of power then emerges at the systemic level, in much the same way that economic efficiency is an automatic consequence of market competition. The Workshop perspective on international relations can be illustrated by reference to the works of Bull (1977) and Berman (1983). Bull demonstrates that the international system can be interpreted as a form of social order, in which the members of that society, sovereign states, effectively manage their own affairs. Bull does not use the term self-governance, but he clearly lays out a vision of a self-governing society of sovereign states.

    However, Bull’s near-exclusive focus on sovereign states causes him to downplay the importance of other forms of collective actors and other forms of international interaction. Berman’s (1983) analysis of the nature of Western legal systems in the early medieval period is more congruent with the logic of institutional analysis as developed in this volume. For Berman emphasizes the multiplicity of legal orders that were under development at that time, and that have continued to exist to the present day. From this perspective, the international order has always been polycentric.

    Contemporary events in the emerging markets of the former Soviet bloc have clearly demonstrated that markets do not emerge automatically, but must instead be located within a supporting institutional context. This is not a new problem. Milgrom, North, and Weingast (1990) use a game model of the medieval “law merchant” to illustrate the ways in which traders managed to overcome their dilemmas of collective action and thus facilitate long-distance trade, without establishing a central governing authority. Rulings by law merchants gave traders sufficient confidence to engage in trades. Without a similar level of confidence, any effort to establish markets will be particularly problematic.

    Potential trades may not be consummated when one party may not be convinced that the other has clear title to the goods about to be exchanged, or, more generally, whether the property rights of all relevant parties will be respected by their respective governments. Without this assurance, foregone trades will make it impossible for the economy as a whole to perform up to its potential, and economic growth will be slowed. Influential research by Douglass North (1981, 1990; North and Thomas 1973) has demonstrated that a clear definition of property rights is essential before market processes can operate at anywhere near efficient levels. Economic growth requires investor confidence, for individuals or private corporations will make investments to improve the productive capacity of their assets only if they can reasonably expect to enjoy the benefits of these investments. Governments have to convince potential investors that their property will not be confiscated without adequate compensation by some later regime.

    One theme common to Williamson’s micro-level investigations of alternative organizational forms and North’s macro-level examination of property rights and economic growth is that efficiency is the single most important criterion to be used in evaluating institutions (see also Eggertsson 1990). Research by Workshop scholars has demonstrated that more than efficiency is needed for the successful management of public economies, especially over long periods of time. Instead, as discussed throughout this volume, modelers must take into account the common cultural norms and practical knowledge shared by the participants in any ongoing game.

    Transaction Costs in Civil Society

    Institutional analysis complements well-known results from the literature on “new institutional economics” concerning the importance of property rights and transaction costs, by emphasizing the importance of clarification of property rights over commonly held assets, including the common-pool resources that have been the subject of most of the research included in this volume.

    Self-governing groups are an important component of “civil society,” which is taken to cover all activities not obviously part of either the private or public sector. However, the term civil society is limiting, for it conveys connotations of special interest groups organizing to articulate their interests to public authorities. Such activities are important, but Workshop scholars have been more interested in demonstrating the ability of groups, particularly CPR user groups, to govern their own affairs rather than simply lobbying for policy changes from the government.

    Group rights in civil society and private property rights in markets share many important characteristics. Those groups of resource users who have successfully managed their common resources have done so at the cost of establishing and enforcing rules that often call for self-sacrifice on the part of individual members of that group. They are unlikely to continue to pay those costs if there is sufficient concern that governmental officials will intervene to establish or enforce a different sets of rules. Without this assurance, group cooperation will break down, as individuals succumb to the temptations to over-exploit this resource or engage in other forms of opportunistic behavior.

    Williamson (1996) defines governance as the efforts of public authorities to shape the transactions facing individuals in market exchange. Laws make some transactions easier to arrange, while at the same time making other, socially undesirable interactions, more difficult for private actors to implement. In exactly the same way, public laws and policies shape the transactions costs facing elements of civil society. By enforcing laws and punishing the private use of coercion, governments can play an essential role in laying the foundation for successful group management of common-pool resources. In short, an effective governance structure must be supportive of group rights.

    Protection of group rights is particularly crucial if the policy goal is sustainable development, and not just economic growth per se. Resource sustainability is not a new idea: groups of fishers, farmers, and herders throughout the world have coped with sustainability problems throughout human history. Governmental officials and policy analysts should remain open to the possibility that they may be able to learn from user groups about the conditions for successful resource management.

    Rent-Seeking and Social Welfare

    In a liberal democracy, governments should be supportive of group efforts at self-governance. Unfortunately, in much of the world, governments actively undermine group efforts to organize their own activities. In totalitarian states the assertion of state control is taken to its logical extreme, but even in the petty autocracies so common in the developing world, governmental officials assert the right to grant permits that individuals, corporations, or groups need to operate as a private business or as a user group. This right to grant permits is, of course, a lucrative source of bribes and corruption. In effect, governmental office is treated as a private resource.

    Whether political authorities or bureaucratic officials are granted exclusive rights to the writing or enforcement of laws and regulations, the potential for rent creation is unavoidable. In this usage, rents are created whenever political authorities create artificial scarcities (by restricting production or entry into a market to a certain type of individuals, requiring permits, or putting restrictions on imports). Those who stand to benefit from such restrictions will engage in costly activities to obtain these benefits, and this “rent-seeking behavior” detracts from the overall welfare of society. The terms rents and rent-seeking are used to cover a multiple of sins, and space does not permit a full discussion of the associated controversies (see Mueller 1989: 229-246; Tollison 1982, 1997). The key point is that rents prevent a market economy from attaining the efficiency of perfect competition. The recipient of the specific rents will benefit, certainly in the short term, but in the long term society as a whole will suffer.

    The specific nature of the rents differs for different types of political order. In an autocratic regime, for example, the recipients of rents are likely to be members of the ruler’s family or others with a close personal connection to the ruler. Those groups (such as the military or domestic security forces) upon which the ruler is most dependent may benefit greatly from the ruler’s policies (see, for example, Wintrobe 1998). In liberal democratic regimes, the recipients of protection tend to be interest groups that reward public officials in the form of votes or campaign contributions (see, for example, Magee et al. 1989).

    Rents and other distortions to the operation of private markets are important topics for the field of public choice. In brief, public choice can be defined as the application of economic methods of analysis to the behavior of public officials. Public choice is an important corrective to the still influential tradition of welfare economics, in which a “benevolent social planner” is presumed to be willing and able to implement policies that are in the best interest of society as a whole (see Mueller 1998: chapter 19). In the public choice tradition, elected and appointed officials are presumed to be motivated by their own selfish interests, whether or not they are interested in maximizing social welfare. Niskanen (1971), for example, argues that agency heads act to maximize their agency’s budgetary resources whenever possible. Such assumptions, while simplistic, can bring important insights to the interactions of public and private actors.

    Public choice modelers often use the preferences of the “median voter” as an indicator of the likely outcome of democratic processes. In effect, the preferences of this decisive voter stand as a surrogate for the social welfare function that was so deeply undermined by Arrow’s Theorem. In this way, modelers retain the ability to make assertions concerning the relative desirability of alternative policies or institutional structures for society as a whole.

    Those of us who’ve spent time at the Workshop tend to be deeply skeptical whenever an administrator or policy analyst claims to understand or implement policy for the greatest good of society. Vincent Ostrom (1989) identifies the pernicious effects of this widely shared attitude on the field of public administration, while laying out an alternative vision of public administration within the context of polycentricity (to be discussed more fully below).

    Institutional analysis shares many of the characteristics of public choice, but its scope is wider. Self-governing groups may or may not delegate authority to particular individuals, since they may be capable of handling the relevant tasks themselves. In other circumstances, delegation may be crucial, especially for the services of monitoring and sanctioning rule violators. Concerns about rent-seeking or agent-principal relations are essential in the latter context, but may not be relevant in the former. Institutional analysis can be seen as a form of public choice where the focus is not restricted to the behavior of public officials, and citizens are not treated merely as voters or campaign contributors.

    Efficiency and Self-Governance

    Public choice theory brings to the study of politics a relentless focus on the importance of efficiency in public policy. Clearly, the efficient allocation of resources to production or consumption is a goal worthy of careful pursuit. However, individuals or communities may decide, for whatever reason, to sacrifice efficiency for the pursuit of other goals, such as accountability, fairness, or sustainable development (however these terms are understood within the relevant communities).

    One reaction is to expand the definition of efficiency to incorporate these additional goals into the underlying utility or production functions being modeled. This practice can be very useful for models of the long-term sustainability of resource management practices (see Dasgupta and Heal 1979). Workshop-affiliated scholars have tended to take a more indirect approach, to define economic efficiency more narrowly and to supplement their analyses with explicit consideration of other collective goals. For example, Ostrom (1990) lays out eight “design principles” that were common to a large number of instances of the successful long-term management of common-pool resources. Economic efficiency is not specifically included in this list, although it is implicit in the requirement for congruence between the rules-in-place and the physical nature of the goods. But the focus of her description of this congruence principle lies instead on the importance of a meaningful correspondence between the rules-in-use and the normative beliefs of that community. That is, participants must feel that the rules are “fair” according to the standards of the relevant community, or else they will not put forth the effort required to overcome their dilemmas of collective action. This is not to say that economic efficiency is irrelevant. In some institutional arrangements, especially when competitive markets in private goods are available, the criterion of efficiency takes on a unique prominence. In general, however, multiple evaluate criteria must be given equal consideration.

    The ability of self-governing groups to resolve their own practical problems, as they themselves define the problem, is the central theme that unifies all of the many research programs undertaken by Workshop scholars. But self-governance is not a simple matter of groups with homogeneous tastes forming a “club” (Sandler and Tschirhart 1980). Nonprofit organizations, for example, are an important expression of self-governance and an essential component of civil society. Hansmann (1980), Weisbrod (1977, 1988), and Salamon (1987, 1995) treat the nonprofit sector as a third sector of the economy that fills important roles left open by both the private and public sectors. From this point of view, nonprofit organizations are granted tax-exempt status in order to facilitate their contribution to collectively desired goods and services that can be most efficiently provided in that manner.

    Even so, nonprofit organizations cannot be fully understood when considered solely in efficiency terms. Individuals may receive direct benefits from participation in such groups, including the resolution of important problems. Nonprofit and nongovernmental organizations have long been of considerable interest to Workshop scholars, as expressions of the natural human tendency to form groups and to solve problems.

    Collective Action, Community, and Networks of Organizations

    Whenever assets are held in common, or maintained by some communal activities, it is necessary to make some arrangements for the monitoring of individual behavior and for the sanctioning of those individuals who do not fulfill their obligations under the agreed-upon rules. At this point a contrast with perhaps the best known approach to the study of collective action is in order. Olson (1965) argues that selective incentives are the key inducement for individual participation in collective action. This may be true for the types of groups Olson has in mind, which tend to be groups that lobby the government for the provision of some service. (Alternatively, groups may form in order to overthrow the existing government or to express their displeasure; incentives for group formation in this context are surveyed by Lichbach 1995, 1996.) In other words, these are voluntary groups that individuals may form if they choose to do so, and they may or may not choose to contribute towards the production of the collective good in question. Workshop scholars tend to focus on groups (farmers, fishers, etc.) whose very livelihood is dependent on the success or failure of their efforts to manage a salient resource collectively.

    Such groups often see themselves as a community, which shares common understandings and values even if individuals have conflicting interests. A sense of communal belonging can make an important contribution towards group success, but it must be consistent with the rational pursuit of individual interests (Hechter 1987). Lichbach (1996) demonstrates that community can play a contributing role in resolving dilemmas of collective action, but that markets, contracts, and coercion also play essential roles. No one mode is sufficient unto itself.

    March and Olsen (1984, 1989) emphasize one way in which a sense of shared understandings contribute towards the smooth operation of collective action. They stress that each institution (or organization) has associated with it a “logic of appropriateness,” meaning that only certain kinds of actions are deemed to be acceptable or appropriate for the agents of particular organizations or for interactions among members of that organization. This logic is part of the overall cultural context within which organizations and individuals interact. In this sense, then, the set of organizations in existence at any given time can be seen as a concrete embodiment of the cultural understandings shared by members of that community.

    It is also important to realize that no one organization stands completely on its own. Instead, organizations are linked together in complex networks. So are the individual members of a community. Thus, one important path towards an understanding of social interactions is to understand these networks, their inherent dynamics, and the nature of their underlying structure (Nohria and Eccles 1992).

    Market Competition and Democratic Governance

    This concern for institutional context leads us back to the issue of market failures with which this section began. If a community finds that market processes are unable to produce the kinds and levels of public goods and services that they need, but they are also reluctant to cede all authority and sanctioning power to an unresponsive Leviathan, then they will need to make some other arrangements. What types of arrangements will be made, and how will they monitor the outcomes of these arrangements and apply sanctions to their representatives who do not deliver the goods? These are the basic questions confronting Workshop scholars.

    In Tiebout’s (1956) “voting-with-the-feet” model, public entrepreneurs offer packages of public goods to their communities. Those individuals dissatisfied with the mix of public goods and services provided by their local officials have the option of moving to the jurisdiction of another group of public authorities whose policy package is more desirable than the first. If the costs of moving are relatively low, and the range of alternative officials or jurisdictions is sufficiently wide, then the result would be a close approximation of the model of market provision of private goods. That is, groups of individuals with similar tastes for public goods and the taxes needed to finance them would gather together in relatively homogeneous communities serviced by public officials who best provide for their needs.

    There are, of course, significant problems with the application of this principle in practice. Individual tastes are diverse, and there exist far more public goods or services than jurisdictions, so the match between community preferences and public goods provision is sure to be inexact. Yet, this simple model does have some important implications. Capital, for example, is typically more mobile than labor, and so the owners of capital can exert considerable control over the policies of public officials. Lindblom (1977) argues that the ability of business leaders to credibly threaten to move their capital elsewhere places severe restraints on the types of taxing and other economic policies that public officials can implement. For those officials who enact especially distasteful policies will find their economic base undermined by capital flight, which will lead to a worsening of economic conditions, which is likely to lead to the removal of the offending officials in the next election. Since public officials are rational, they will be deterred by this prospect of losing office and not enact policies too distasteful to the private sector. This situation can set up a “race to the bottom” as separate jurisdictions compete to provide favorable conditions for business. Lindblom is very concerned about this “market as prison” effect, because it greatly limits the ability of democratic officials to provide the policies desired by a majority of the public. Alternatively, this effect can be seen in a positive light, in the sense that it precludes the implementation of particularly ineffective economic policies.

    Polycentricity and Self-Governance

    In general, individuals should not have to move to other jurisdictions (or threaten to relocate their capital assets) to garner the benefits of competition among the providers of public goods or services. Ostrom, Tiebout, and Warren (1961) introduce the alternative conceptualization of a “polycentric political system” in which public officials representing a community (or the members of that community themselves) can select from alternative mechanisms for the production of public goods or services. These producers may be private firms, public agencies, or some other entity entirely. In this system the “providers” of public goods and services can make arrangements with those “producers” of the good or service that operate at the most efficient scale of production. This enables the system to achieve a degree of efficiency while still allowing communities a much wider range of choice. Ostrom and Ostrom (1977) use the term “public service industry” to denote the network of large and small organizations that are involved in the production and provision of any particular type of public good or service. In effect, then, a polycentric order consists of networks of interacting organizations.

    Another way in which polycentric order may lead to improved efficiency is by lowering rents. When multiple political authorities must interact to implement restrictions, then it will be more difficult for any one group to protect their position against market and other pressures. Conversely, agents who have a monopoly on authority in a given issue area are likely to induce significant levels of rent-seeking behavior. Recently, Wagner (1997), writing from outside the circle of Workshop scholars, provides a succinct statement of the implications of polycentric order for the study of public finance.

    But efficiency is not the primary reason to advocate establishment and maintenance of a polycentric system of interacting authorities and citizens. Instead, polycentricity is a fundamental pre-requisite for individual liberty and the ability of groups to govern their own affairs (V. Ostrom 1987, 1991, 1997). It is this connection between polycentricity and self-governance that lies at the very core of all of the theoretical and empirical research programs implemented by Workshop scholars (see McGinnis 1999a,b).

    The centrality of this concept inspired use of the term “polycentric games” in the title of the current volume. The implications of polycentricity for modeling should be evident: it is not ideal to model any one game in isolation. Simple models remain useful, especially for highlighting crucial components of particular empirical settings. However, when models are based on majority voting, economic efficiency, or any other single criterion, then it is too easy to lose sight of other equally essential aspects of the institutional context.

    Self-governance at the local level is sustainable only in the context of a supportive political and cultural environment at the constitutional level. Market exchange can lead to efficient outcomes only when political and legal institutions guarantee secure property rights and provide general access to low-cost and effective means of conflict resolution. Even the coercion so often taken as the defining characteristic of the “state” can lose its effectiveness if the rulers lose contact with local traditions and expectations, or if productive assets flow across national borders in search of more promising opportunities abroad. Markets, states, and user groups should be seen as complementary institutions. Each has its proper role to play in polycentric governance, its own unique configuration of strengths and weaknesses.

    Institutional Analysis and Inductive Research

    Research conducted by Workshop-affiliated scholars has made important contributions towards establishing the legitimacy of group management of common-pool resources. For the most part, this research has been inductive in nature, using successful instances of self-governing groups as an “existence proof” of the reality of successful CPR management. Claims are not made that user groups always achieve “efficient” outcomes, but rather that the remarkable thing is that groups are able to manage common-pool resources over long periods of time, thereby avoiding the dire straits predicted by Hardin’s (1968) tragedy of the commons.

    Ostrom (1990) is the most widely-cited summary of these research programs. Mathematical models and experimental settings have been used to complement the findings of field research (see Ostrom, Gardner, and Walker 1994; McGinnis 1999a). One way that Workshop scholars have keep the broader institutional context in view is by making use of inductive modes of research. Even simple models tend to be inspired by observation of the ingenious solutions devised by specific communities to address particular problems, rather than being derived from first principles.

    The inductive thrust of Workshop models was noted in Mitchell’s (1988) comparative analysis of what he calls the Bloomington school, defined in contrast to the Rochester school of social choice (with its emphasis on formal models of voting) and the Virginia school of public choice (with an emphasis on rent-seeking and economic efficiency). However, it is also important to keep in mind the broader institutional context that lies behind specific models in the Workshop tradition. The three arenas of choice encompassed by the IAD framework, as well as linkages among physical, cultural, and institutional factors highlighted by that framework, are two key aspects that have shaped institutional analysis.

    References (to works not included in this volume)

    Arrow, Kenneth J. 1963. Social Choice and Individual Values, second edition. New Haven: Yale University Press.

    Berman, Harold. 1983. Law and Revolution: The Formation of the Western Legal Tradition. Cambridge: Harvard University Press.

    Buchanan, James M., and Gordon Tullock. 1962. The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press.

    Bull, Hedley. 1977. The Anarchical Society. New York: Columbia University Press.

    Dasgupta, Partha S., and G. M. Heal. 1979. Economic Theory and Exhaustible Resources, Cambridge: Cambridge University Press.

    Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper & Row.

    Eggertsson, Thráinn. 1990. Economic Behavior and Institutions, Cambridge: Cambridge University Press.

    Hansmann, Henry B. 1980. “The Role of Nonprofit Enterprise,” Yale Law Review. 89 (5): 835-901.

    Hardin, Garrett. 1968. “The Tragedy of the Commons,” Science, 162: 1243-1248.

    Hechter, Michael. 1987. Principles of Group Solidarity. Berkeley: University of California Press.

    Krehbiel, Keith. 1991. Information and Legislative Organization. Ann Arbor: University of Michigan Press.

    Kreps, David M. 1990a. A Course in Microeconomic Theory. Princeton, NJ: Princeton University Press.

    ________. 1990b. Game Theory and Economic Modelling. New York: Oxford University Press.

    Lichbach, Mark I. 1996. The Cooperator’s Dilemma. Ann Arbor: University of Michigan Press.

    ________. 1995. The Rebel’s Dilemma. Ann Arbor: University of Michigan Press.

    Lindblom, Charles E. 1977. Politics and Markets: The World's Political Economic Systems. New York: Basic Books.

    Magee, Stephen P., William A. Brock, and Leslie Young. 1989. Black Hole Tariffs and Endogenous Policy Theory: Political Economy in General Equilibrium. Cambridge: Cambridge University Press.

    March, James G., and Johan P. Olsen. 1984. "The New Institutionalism: Organizational Factors in Political Life," American Political Science Review, 78: 734-749.

    ________. 1989. Rediscovering Institutions: The Organizational Basis of Politics. New York: Free Press.

    McGinnis, Michael D. 1991. “Richardson, Rationality, and Restrictive Models of Arms Races," Journal of Conflict Resolution, 35: 443-473.

    ________, ed. 1999a. Polycentric Governance and Development, Ann Arbor: University of Michigan Press.

    ________, ed. 1999b. Polycentricity and Local Public Economies, Ann Arbor: University of Michigan Press.

    McGinnis, Michael D., and John T. Williams. 1998. Compound Dilemmas: Democracy, Collective Action, and Superpower Rivalry. Indiana University, Bloomington: draft book manuscript.

    McKelvey, Richard D. 1976. “Intransitivities in Multi-dimensional Voting Models and Some Implications for Agenda Control,” Journal of Economic Theory, 12: 472-482.

    Milgrom, Paul, Douglass North, and Barry Weingast. 1990. “The Role of Institutions in the Revival of Trade: The Law Merchant, Private Judges, and the Champagne Fairs,” Economics and Politics, 2:1-23.

    Miller, Gary J. 1992. Managerial Dilemmas: The Political Economy of Hierarchy, New York: Cambridge University Press.

    Mitchell, William C. 1988. “Virginia, Rochester, and Bloomington: Twenty-Five Years of Public Choice and Political Science,” Public Choice, 56: 101-119.

    Mueller, Dennis C. 1989. Public Choice II. Cambridge: Cambridge University Press.

    Niskanen. W. A., Jr. 1971. Bureaucracy and Representative Government. Chicago: AVC.

    Nohria, Nitin, and Robert G. Eccles, ed. 1992. Networks and Organizations: Structure, Form, and Action. Boston: Harvard Business School Press.

    North, Douglass C. 1981. Structure and Change in Economic History. New York: Norton.

    ________. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press.

    North, Douglass C., and Robert Paul Thomas. 1973. The Rise of the Western World: A New Economic History. Cambridge: Cambridge University Press.

    Olson, Mancur. 1965. The Logic of Collective Action. Harvard University Press.

    Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

    Ostrom, Elinor, Roy Gardner, and James Walker, with Arun Agrawal, William Blomquist, Edella Schlager, and Shui Yan Tang. 1994. Rules, Games, and Common-Pool Resources, Ann Arbor: University of Michigan Press.

    Ostrom, Elinor, James Walker, and Roy Gardner. 1992. "Covenants with and without a Sword: Self-Governance is Possible." American Political Science Review, 86: 404-17.

    Ostrom, Vincent. 1987. The Political Theory of a Compound Republic: Designing the American Experiment. 2d rev. ed. San Francisco: ICS Press (first edition 1971)

    ________. 1989. The Intellectual Crisis in American Public Administration. 2nd ed. University of Alabama Press (first edition 1973)

    ________. 1991. The Meaning of American Federalism: Constituting a Self-Governing Society. San Francisco: ICS Press.

    ________. 1997. The Meaning of Democracy and the Vulnerability of Democracies: A Response to Tocqueville’s Challenge. Ann Arbor: University of Michigan Press.

    Ostrom, Vincent, and Elinor Ostrom. 1977. “Public Goods and Public Choices.” In Alternatives for Delivering Public Services. Toward Improved Performance, ed. E. S. Savas, 7-49. Boulder, CO: Westview Press. Reprinted in McGinnis 1999b.

    Ostrom, Vincent, Charles M. Tiebout, and Robert Warren. 1961. “The Organization of Government in Metropolitan Areas: A Theoretical Inquiry.” American Political Science Review 55 (Dec.): 831-42. Reprinted in McGinnis 1999b.

    Putnam, Robert D. 1988. "Diplomacy and Domestic Politics: The Logic of Two-Level Games," International Organization, 42, 427-460.

    Riker, William H. 1982. Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice. San Francisco: W. H. Freeman.

    Russett, Bruce, and Harvey Starr. 1996. World Politics: The Menu for Choice, 5th edition. San Francisco: W. H. Freeman.

    Salamon, Lester M. 1987. “Partners in Public Service: The Scope and Theory of Government-Nonprofit Relations,” in Walter W. Powell, ed., The Nonprofit Sector: A Research Handbook, New Haven: Yale University Press, pp. 99-117.

    ________. 1995. Partners in Public Service: Government-Nonprofit Relations in the Modern Welfare State. Baltimore: Johns Hopkins University Press.

    Sandler, Todd, and John T. Tshirhart. 1980. “The Economic Theory of Clubs: An Evaluation Survey,” Journal of Economic Literature. 18: 1488-1521.

    Schofield, Norman. 1978. “Instability of Simple Dynamic Games.” Review of Economic Studies, 45: 575-594.

    Schelling, Thomas C. 1960. The Strategy of Conflict. New Haven: Harvard University Press.

    Shepsle, Kenneth. 1979. “Institutional Arrangements and Equilibrium in Multi-dimensional Voting Models,” American Journal of Political Science, 23: 27-59.

    Simon, Herbert A. 1957. Models of Man. New York: Wiley.

    ________. 1964. Administrative Behavior: A Study of the Decision-Making Processes in Administrative Organizations. New York: Macmillan.

    ________. 1997. Models of Bounded Rationality: Empiricially Grounded Economic Reason. Cambridge, MA: MIT Press.

    Smith, Adam. 1976 [1776] An Inquiry into the Nature and Causes of the Wealth of Nations, ed. Edwin Cannan, Chicago: University of Chicago Press.

    Tiebout, Charles M. 1956. “A Pure Theory of Public Expenditures,” Journal of Political Economy, 64 (October), 416-424.

    Tocqueville, Alexis de. 1969 [1835]. Democracy in America, edited by J.P. Mayer, translated by George Lawrence, Anchor Books.

    Tollison, Robert D. 1980. “Rent Seeking: A Survey.” Kyklos 35 (4): 575-602.

    ________. 1997. “Rent Seeking.” In Dennis C. Mueller, ed., Perspectives on Public Choice: A Handbook, Cambridge: Cambridge University Press, pp. 506-525.

    Tsebelis, George. 1990. Nested Games. Berkeley: University of California Press.

    Tullock, Gordon. 1965. Politics of Bureaucracy. Washington, DC: Public Affairs Press.

    von Wright, Georg H. 1951. “Deontic Logic,” Mind, 60: 58-74.

    Wagner, Richard E. 1997. “Reconsidering the Fiscal Role of Government,” American Economic Review, 87 (2): 160-.

    Waltz, Kenneth N. 1959. Man, the State, and War, New York: Columbia University Press.

    ________. 1979. Theory of International Politics. Reading, MA: Addison-Wesley.

    Weimer, David L., and Aidan R. Vining. 1989. Policy Analysis: Concepts and Practice, 2nd edition, Englewood Cliffs, NJ: Prentice-Hall.

    Weisbrod, Burton A. 1977. The Voluntary Nonprofit Sector: An Economic Analysis. Lexington, Mass,: D.C. Heath.

    ________. 1988. The Nonprofit Economy. Cambridge: Harvard University Press.

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

    ________. 1985. The Economic Institutions of Capitalism. New York: Free Press.

    ________. 1996. The Mechanisms of Governance. New York: Oxford University Press.

    Wintrobe, Ronald. 1998. The Political Economy of Dictatorship. Cambridge: Cambridge University Press.

    Introduction to Part I

    In “Public Choice: A Different Approach to the Study of Public Administration” (Chapter 1), Vincent Ostrom and Elinor Ostrom survey the analytical foundations of the Workshop approach to institutional analysis. In this essay from Public Administration Review, first published in 1971, they lay out a general method of analysis that has some important differences from the mainstream approach to public choice. Mueller (1989: 1), defines public choice as “the economic study of nonmarket decision making, or simply the application of economics to political science.” Although this ever-growing body of public choice research has always been an important source of inspiration for scholars associated with the Workshop, there have been many other sources that push in somewhat different directions.

    For example, Workshop scholars are skeptical of claims that a single model of rational behavior is valid for all individuals in all institutional contexts. Rational choice theory, grounded in methodological individualism, provides the general approach to research, but different models may be relevant for different situations. For example, the same individual might follow Simon’s (1957, 1997) satisficing procedure in one decision context while engaging in more extensive information search and evaluation in other situations. Selective pressures may be strong enough to eliminate habit-driven behavior in some contexts but not in another. There is no reason to presume that any one individual acts exactly the same in all circumstances; still, there is merit in trying to locate the relevant range of decisional procedures within the context of a common explanation.

    In Chapter 1 the Ostroms review alternative approaches to the study of public administration. They begin by critiquing the standard view of public administration as an exercise in planning, a view most clearly articulated by Woodrow Wilson and other major figures in the American reform tradition (see E. Ostrom 1972; V. Ostrom 1989). They then discuss Herbert Simon’s (1957, 1964) influential critique of the assumption that individuals or organizations can reasonably be expected to have the capacity to plan at the necessary level of detail and comprehensiveness. They also discuss efforts by Tullock (1965) and other public choice theorists to interpret the behavior of bureaucratic officials as driven by an unremitting pursuit of their own selfish interests in power, prestige, or resources. However, they argue that the same official, if placed in a different institutional context, might be less inclined (or less able) to take advantage of that position.

    The authors argue that the “nature of the good” is an important analytical category that cannot be overlooked in political discourse or empirical evaluation. Different institutional arrangements will be most effective for the resolution of problems associated with the production, provision, or allocation of private goods, public goods, common-pool resources, and toll goods. (For an extended discussion of these distinctions and their implications for institutional analysis, see Ostrom and Ostrom 1977a and other essays reprinted in McGinnis 1999b.)

    The authors conclude with a brief overview of the merits of polycentricity as an alternative approach to public administration. Rather than seeing governance as a question of the implementation of a single, uniform law, allowance should be made for local communities to form associations at whatever scale of aggregation that they feel is most appropriate for the solution of their common problems. Public officials are then seen as providers of public goods and services to these communities. As providers, they do not need to produce the goods or services themselves, but can instead choose to make arrangements with private producers, other public agencies, or even encourage the communities to produce the relevant goods or services themselves. The potential benefits of such a system of polycentric order are described in more detail elsewhere (see especially Ostrom, Tiebout, and Warren 1961).

    In Chapter 1 the Ostroms briefly discuss their interpretation of the classic dilemmas of collective action as laid out by Olson (1965) and Hardin (1968). They emphasize the importance of including constitutional order within the purview of the analysis of particular policy problems, and insist that voting is only one of many ways in which communities can arrive at collective decisions. All in all, they sketch institutional analysis as a distinctive approach to the study of public policy and political economy (see Mitchell 1988).

    The Institutional Analysis and Development Framework

    One of the distinctive emphases in institutional analysis is the recognition that rational choice theory is not the only way in which collective choices can be conceptualized. In addition, there may not be a single process of “rational choice” that applies equally to all institutional contexts. Because of these complexities, an institutional analyst seeking to model or analyze a particular empirical setting must first determine what explanatory factors are most likely to be useful in those circumstances. Through a long series of ongoing discussions among faculty, students, and practitioners, an overall “framework” of analysis has been developed. This framework lays out a general conceptual schema that encompasses all of the factors most relevant to alternative theories of choice and to particular models of specific empirical situations.

    This sequence of framework-theory-model helps clarify the scope of disputes between advocates of different modes of analysis. In some cases analysts may share the same theory but disagree over what particular assumptions should be built into a model of a specific situation. Similarly, analysts who are less comfortable with rational choice theory as it is typically conceptualized may still share a common framework of analysis with rational choice modelers.

    To craft this framework of analysis, emphasis has been placed on an understanding of the overall action situation confronting individuals and groups. In “The Three Worlds of Action: A Metatheoretical Synthesis of Institutional Approaches” (Chapter 2), Larry Kiser and Elinor Ostrom summarize the Institutional Analysis and Development (IAD) Framework, an organizing schema that emerged out of extensive discussions at the Workshop. Two later formulations (Oakerson 1992 and Ostrom, Gardner, and Walker 1994: chapter 2) include further refinements, but this chapter remains the canonical presentation of the IAD framework precisely because it portrays it as a work in process. A wide array of factors must be included in any full-scale analysis of the implications of particular institutions, and keeping a handle on this diversity is an ever-shifting task.

    Distinctions among action situations in arenas of operational, collective, and constitutional choice recur throughout all Workshop research programs, as do discussions of interactions among physical or material conditions, attributes of the community, and the rules-in-use that shape the behavior of individuals or organizations. The IAD framework provides a shared language for a wide array of institutional analyses, thus facilitating comparisons among more specific theories and models of particular phenomena.

    However, one aspect of this language can be potentially confusing. As discussed in the general introduction to this volume, the term “levels of analysis” should be reserved for processes operating at different scales of aggregation, as in international relations theory (Waltz 1959; Russett and Starr 1996). The term “arena of choice” seems a better reflection of the basic idea behind the “three worlds” in the title of this essay.

    Outcomes in one arena of choice define the nature of the games being played concurrently in other arenas. For example, constitutional decisions define the processes by which organizations are expected to interact. Similarly, collective choices specify the operational rights and responsibilities of specific actors. All three arenas interact in any one situation.

    A Subtle Difference of Emphasis

    In her Presidential Address to the Public Choice Society, “An Agenda for the Study of Institutions” (Chapter 3), Elinor Ostrom provides a fuller explanation of the “action situation” that lies at the core of the IAD framework. In doing so, she lays out a research program meant to generalize standard conceptualizations of game models. Rather than assuming that all actors are identical in interests or capabilities, Ostrom lays out a set of categories that jointly define the roles different types of players fill as well as the rules by which their interactions are structured. Although much of game theory has been focused on models of actors with symmetric interests or capabilities, Elinor Ostrom argues that it is crucially important to incorporate at least some of the ways in which actors differ. In particular, institutions define new roles, thus imparting to certain individuals a set of interests, capabilities, and responsibilities that should be incorporated in any effort to model the effects of institutional arrangements. She argues that all institutions share some foundational set of components; any one institution’s consequences are determined by the configuration of these structural elements. In this way she hopes to move the field beyond a preoccupation with supposedly “institution-free” settings to fuller and more useful formal representations of institutions.

    Ostrom presents her concept of an “action situation” as a generalization of standard game models. However, a typical reaction of game theorists is that they also have to define each of the components that Ostrom lists in her framework, or else their game models would not be fully specified. Technically speaking this is correct, but there remains an important difference in emphasis. For Ostrom, and for other Workshop scholars, it is essential to keep in mind the extent to which actors’ preferences as well as the choice options available to them are determined by the institutional arrangements that define their position or that shape their perceptions and options. It is essential to remember that concurrent games in other arenas of choice interact in subtle ways with any ongoing process of interaction. The payoffs and menu of choices available to participants in operational games have been defined by collective choice processes. Games over collective deliberations are in turn shaped by the positions and interests defined or manifested in the constitutional choice arena. In a strict sense this may be no different from a complete specification of a game model, but in practice this concern with simultaneous consideration of multiple choice arenas inspires Workshop-affiliated scholars toward a more inductive mode of analysis.

    Specifically, Ostrom calls for a shift away from the analytical practice of considering changes in the rules of the game in isolation. She argues that the implications of sets of rules must be understood, since they act together in a configural manner. Interactions among rules were emphasized in classic work by John R. Commons (1959), and Ostrom applies this idea to contemporary game theory. In this essay she develops several examples from the theoretical and experimental literature on public choice to illustrate the configural nature of rule changes.

    This general point that rules need to be explicitly specified in game models is developed further in Roy Gardner and Elinor Ostrom (1991), in which several game models are used to represent the implications of a series of alternative rules for the assignment of fishing spots (see also Ostrom, Gardner, and Walker 1994). They begin by laying out Hobbes’ conceptualization of a “state of nature” as a “default condition” which might apply to some situations, but in nearly all social settings there are differences among actor interests and capabilities which can be attributed to the existence of institutional rules. This essay serves as a great follow-up piece to the more abstract presentation of the abstract principles laid out in Elinor Ostrom’s “Agenda for the Study of Institutions,” but because the specific models discussed there were incorporated into the analysis presented in Ostrom, Gardner, and Walker (1994), it did not seem appropriate to reprint that material here.

    Ostrom (1989) delves deeper into the nature of the actors who interact to form action situations. She assumes that individuals can learn, but only imperfectly. Individuals are subject to a multitude of influences, but individuals located within the supportive environment of polycentric order are more likely to successfully adapt to their situations. In effect, this essay can be seen as a more micro-level investigation of the core components of the IAD framework. Elinor Ostrom re-visits these issues in a more comprehensive manner in her Presidential Address to the American Political Science Review, included as the last selection in the present volume.

    These essays on interactions between institutions and individuals also reflect the influence of a community of scholars at the Center for Interdisciplinary Research at Bielefeld University in Germany, where the Ostroms served as visiting scholars during 1981. Elinor Ostrom later spent a semester there studying game theory with Reinhard Selten, a Nobel Prize laureate with a long-standing interest in innovative ways of interpreting the behavioral basis of game theory. His influence can be seen especially in the game models section of E. Ostrom, Gardner, and Walker (1994) and in essays Elinor Ostrom co-authored with Franz Weissing (including chapter 13 of this volume). A collection of essays edited by Kauffman, Majone, and Ostrom (1986) includes chapters by several policy analysts (including Paul Sabatier) and prominent game theorists Reinhard Selten and Martin Shubik. It includes two chapters by Elinor Ostrom laying out the IAD framework as it was understood at that time and three chapters by Vincent Ostrom (including the essay co-authored with Roberta Herzberg reprinted in this volume) on voting systems and the nature of constitutional order. This Bielefeld volume is a fascinating illustration of diverse but related efforts to comprehend interactions among systems, networks, hierarchies, institutions, and norms.

    Incorporating Normative Considerations

    Another distinction from the international relations literature is useful at this point. In the constructivist approach to international relations, a distinction is made between obligatory and constitutive norms (see Katzenstein 1996). Obligatory norms are moral prescriptions that specify what individuals or organizations should do in specific empirical contexts, whereas constitutive norms define the very nature of these organizations. In this sense both the organizations and the nature of their interactions are “socially constructed.” In terms of the IAD framework, these constitutive norms are equivalent to the interactions between arenas of choice discussed above, whereas obligatory norms are manifested in the ways in which actors interact in any single choice arena. However, Workshop scholars would cringe at the term “obligatory,” since no norm is self-enforcing but must instead be supported by some form of monitoring and sanctioning (see Ostrom 1990).

    V. Ostrom (1976) summarizes one of the most important inspirations behind this effort to bring normative expectations within the purview of institutional analysis. John R. Commons (1959) was concerned about directing attention to what he calls “going concerns” rather than remaining fixated on the details of organization charts or constitutional provisions. In a “going concern” individuals are engaged in a series of transactions, which are shaped by the overall structure of legal rules and moral expectations. Commons’s basic framework has recently inspired an extensive body of research on the ways in which the costs associated with any form of transaction between two or more actors can be reduced in order to facilitate the likelihood that mutually beneficial transactions would be undertaken. This area of “transition cost economics” is most closely associated with Williamson (1975, 1985, 1996); related areas of “new institutional economics” are surveyed in Eggertsson (1990).

    Vincent Ostrom asks us to step back to re-examine the “old institutional economics” of John R. Commons. Ostrom draws on distinctions Commons makes concerning relationships between the rights and responsibilities of individuals serving in different legal capacities. For example, any provision that a public official “must” take a certain action necessarily implies a responsibility for some other official, or for the citizenry as a whole, to undertake an effort to monitor the behavior of the first official, to make sure the original provision is being satisfied (see also Ostrom and Ostrom 1972). Without this additional structure, the first legal provision stands in isolation and cannot be expected to be realized in actual conditions. Mutually reinforced sets of rules, rights and responsibilities constitute what Commons calls “working rules,” and it is to this type of rules that Workshop scholars have directed their attention. Knowing the law is not enough -- it is also essential to understand how those rules are implemented in specific action situations.

    In “A Grammar of Institutions” (Chapter 4), Sue Crawford and Elinor Ostrom push this line of argument further. They use a sequence of definitions to try to make sense out of the multiple meanings of such conceptually slippery terms as institutions, rules, and norms. This conceptualization builds on many sources, including the deontic logic of von Wright (1951) and the legal framework of John R. Commons (1959). One important component concerns the consequences of the extent to which individuals internalize the norms of their community. This internalization makes certain socially undesirable actions costly to the relevant individuals, but these internal costs are rarely sufficient to prevent completely the occurrence of prescribed behavior. Thus, finding some way to monitor and sanction rule-breaking behavior remains essential to the maintenance of social cooperation. In this way this conceptualization provides a logical foundation for the observed importance of monitoring and sanctioning in successful CPR regimes (E. Ostrom 1990; McGinnis 1999a).

    The internal costs included in their representation of actor incentives act like the “obligatory norms” discussed above, in the sense that they help define what actors consider to be acceptable and unacceptable behavior. Meanwhile, broader institutional arrangement (from other arenas of choice) define the options available to the actors themselves, by specifying which acts are prohibited, permitted, or required. In game models analysts select which actions are available to the various actors included in the model, but institutional analysts build on the realization that the set of options available to actors in any action situation is determined in games in other arenas of choice. Similarly, game theorists can posit whatever incentives they want for their modeled individuals to pursue, but in applications it is important to understand the source of those payoffs. Hence the need to consider the broader context of “polycentric games.”

    References (to works not included in this volume)

    Commons, John R. 1959. The Legal Foundations of Capitalism. Madison: University of Wisconsin Press.

    Eggertsson, Thráinn. 1990. Economic Behavior and Institutions, Cambridge: Cambridge University Press.

    Gardner, Roy and Elinor Ostrom. 1991. "Rules and Games." Public Choice, Vol. 70, No. 2 (May), 121-49.

    Hardin, Garrett. 1968. “The Tragedy of the Commons,” Science, 162: 1243-1248.

    Katzenstein, Peter J., ed. 1996. The Culture of National Security, New York: Columbia University Press

    Kaufmann, Franz-Xaver, Giandomenico Majone, and Vincent Ostrom, eds. 1986. Guidance, Control, and Evaluation in the Public Sector. Berlin and New York: Walter de Gruyter.

    McGinnis, Michael D., ed. 1999a. Polycentric Governance and Development, Ann Arbor: University of Michigan Press.

    ________, ed. 1999b. Polycentricity and Local Public Economies, Ann Arbor: University of Michigan Press.

    Mitchell, William C. 1988. “Virginia, Rochester, and Bloomington: Twenty-Five Years of Public Choice and Political Science,” Public Choice, 56: 101-119.

    Mueller, Dennis C. 1989. Public Choice II. Cambridge: Cambridge University Press.

    Oakerson, Ronald J. 1992. "Analyzing the Commons: A Framework." In Making the Commons Work: Theory, Practice, and Policy, ed. Daniel W. Bromley et al., 41-59. San Francisco, CA: ICS Press.

    Olson, Mancur. 1965. The Logic of Collective Action. Harvard University Press.

    Ostrom, Elinor. 1972. “Metropolitan Reform: Propositions Derived from Two Traditions.” Social Science Quarterly 53 (Dec.): 474-93. Reprinted in McGinnis 1999b.

    Ostrom, Elinor. 1989. “Microconstitutional Change in Multiconstitutional Political Systems.” Rationality and Society, 1(1), (July), 11-50.

    ________. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

    Ostrom, Elinor, Roy Gardner, and James Walker, with Arun Agrawal, William Blomquist, Edella Schlager, and Shui Yan Tang. 1994. Rules, Games, and Common-Pool Resources, Ann Arbor: University of Michigan Press.

    Ostrom, Vincent. 1976. “John R. Commons’s Foundation for Policy Analysis.” Journal of Economic Issues 10(4) (Dec.): 839-57.

    ________. 1989. The Intellectual Crisis in American Public Administration. 2nd ed. University of Alabama Press (first edition 1973)

    Ostrom, Vincent, and Elinor Ostrom. 1972. "Legal and Political Conditions of Water Resource Development." Land Economics 48(1) (Feb.): 1-14. Reprinted in McGinnis 1999a.

    ________. 1977. “Public Goods and Public Choices.” In Alternatives for Delivering Public Services. Toward Improved Performance, ed. E. S. Savas, 7-49. Boulder, CO: Westview Press. Reprinted in McGinnis 1999b.

    Ostrom, Vincent, Charles M. Tiebout, and Robert Warren. 1961. “The Organization of Government in Metropolitan Areas: A Theoretical Inquiry.” American Political Science Review 55 (Dec.): 831-42. Reprinted in McGinnis 1999b.

    Russett, Bruce, and Harvey Starr. 1996. World Politics: The Menu for Choice, 5th edition. San Francisco: W. H. Freeman.

    Simon, Herbert A. 1957. Models of Man. New York: Wiley.

    ________. 1964. Administrative Behavior: A Study of the Decision-Making Processes in Administrative Organizations. New York: Macmillan.

    ________. 1997. Models of Bounded Rationality: Empiricially Grounded Economic Reason. Cambridge, MA: MIT Press.

    Tullock, Gordon. 1965. Politics of Bureaucracy. Washington, DC: Public Affairs Press.

    von Wright, Georg H. 1951. “Deontic Logic,” Mind, 60: 58-74.

    Waltz, Kenneth N. 1959. Man, the State, and War, New York: Columbia University Press.

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

    ________. 1985. The Economic Institutions of Capitalism. New York: Free Press.

    ________. 1996. The Mechanisms of Governance. New York: Oxford University Press.

    Introduction to Part II

    Essays in Part II show how the general frameworks presented in Part I can be used to address specific questions of voting, conflict, and leadership. Despite their many differences in substance and method, all of these essays deal with important aspects of the problems faced by any community seeking to agree upon some collective decision.

    Voting and Constitutional Order

    When North Americans of the revolutionary era engaged in an explicit exercise in constitutional choice, they designed a multi-faceted system of checks and balances. Vincent Ostrom (1987) details the basic conceptual framework upon which the Founders built this constitution as an act of creative artisanship (Ostrom 1980). In “Votes and Vetoes” (Chapter 5), Roberta Herzberg and Vincent Ostrom demonstrate the natural connection between voting as a positive act and the many veto points included in the constitutional order of checks and balances. They argue that the ability of diverse minority interests to block the implementation of policies which they strongly oppose has played an important role in sustaining the American experiment.

    Voting is typically taken to be the defining characteristic of modern democracy. Several voting institutions are included in the U.S. constitution, but the essential foundation for the subsequent success of the American experiment in democracy lies in the polycentric nature of the American constitutional order (Ostrom 1991, 1997). Voting institutions, per se, receive considerably less attention in this book than would be expected by most political scientists. Voting is one important way in which individual preferences can be said to be aggregated into public policy, but communities need not rely on formal mechanisms of voting. Normative theorists have long emphasized that an attitude of respect towards the perceptions and interests of other citizens is a fundamental requisite of a sound democracy, and discussions in informal settings are often sufficient for small groups to arrive at a consensus. In short, democratic governance requires more than voting procedures and representative institutions.

    One of the most important lessons of the Workshop for political science in general has been to highlight the importance of forms of political interaction that do not involve voting. The point is not that voting is inconsequential, but rather that it may have received proportionally too much attention by political scientists (see McGinnis 1996).

    Reliance on majority rule has a tendency to direct political agents to conceptualize politics in terms of a contest in which one either wins or loses, rather than approaching politics from a problem-solving orientation. In so doing, political scientists run the risk of losing sight of the ways in which real communities of people manage to work out their own problems. It is this later form of politics that lies at the core of the Workshop approach to policy analysis.

    An over-emphasis on majority voting can contribute to the reinforcement of a Hobbesian conceptualization of sovereignty that is fundamentally at odds with sovereignty as conceptualized by the founders of the American republic. As Vincent Ostrom (1987, 1991, 1997) emphasizes, the capacity of self-governing communities of individuals to resolve their own problems is the key ingredient in the perpetuation of a democratic system of governance. Still, voting remains one way in which groups can make collective decisions, and as such voting institutions are fully deserving of careful analysis. And, indeed, voting has played a central role in the research programs of several scholars associated with the Workshop. Full consideration of this topic would take us too far from the central theme of this volume, but to ignore it entirely would result in an incomplete portrayal of Workshop research programs.

    Social Choice and Laboratory Experiments

    The central result of social choice theory identifies the problems of logical coherence that plague any effort to use majority voting to resolve questions of public policy. Arrow’s (1963) Theorem tells us that voting cycles can occur in any system of aggregating preferences that satisfies the full set of fairness conditions laid out in that work. Another central result of social choice theory concerns the instability of majority voting processes (see Riker 1982). This instability can result from the general ability of participants to find some policy proposal that will defeat the status quo.

    Since very little in the real world of legislative behavior corresponds to this impression of endless chaos, many researchers have sought to uncover the ways in which chaotic outcomes are avoided. Perhaps the most well-known effort is the “structure-induced equilibrium” of Shepsle (1979) in which specific legislative rules are shown to limit the ability of agenda setters to raise certain issues. Those scholars who study voting institutions from the Workshop perspective have focused on more subtle forms of institutional arrangements that might have many of the same effects as the details of legislative procedures.

    In "Negative Decision Powers and Institutional Equilibrium: Experiments on Blocking Coalitions" (Chapter 6), Rick Wilson and Roberta Herzberg represent negative voting power in terms of a spatial voting model. They report on laboratory experiments that demonstrate that giving one individual blocking power tends to help that individual get a more desirable outcome from the voting game. They did not detect any evidence that blocking power was able to uniquely define any particular point in the policy space as the “natural” outcome of the game, except to reinforce a tendency to remain at the status quo point. For if the individual with blocking power prefers the status quo to any offered alternative, then the status quo is going to be preserved. In this way they were able to demonstrate, in a laboratory setting, the import of the factors emphasized in the preceding selection.

    In another work, Herzberg (1992) uses a spatial voting model to explore the logical consistency of an argument made by John Calhoun concerning the effects of concurrent majorities. That essay stands as a good example of the ways in which formal game representations can help clarify the strengths and weaknesses of arguments typically made in a more informal manner. Jillson and Wilson (1994) apply modern spatial voting models to the decisions made by the first American congress, during the Articles of Confederation era, thus demonstrating that these technical advances can also help improve our understanding of our past historical eras.

    In a series of related works, Herzberg and Wilson used laboratory experiments to explore several other aspects of voting institutions. Herzberg and Wilson (1991) demonstrate that the potential for agenda manipulation can be limited if legislators incur costs by adding issues to the agenda. By treating the gathering of certain kinds of information as costly, a degree of stability can be imposed on collective outcomes. It is often argued that the capacity of individuals to misrepresent their preferences via strategic or sophisticated voting can serve to counteract the manipulations of agenda setters (see Riker 1982). The experimental results of Herzberg and Wilson (1988) show how difficult it is for experimental subjects to achieve their desired results via sophisticated voting. Of course, this result does not necessarily imply the nonexistence of sophisticated voting in other contexts, when participants have at their disposal a wider range of ways to gather information or to coordinate their behavior on matters of great concern to them at the time. Still, these results do point out the difficulty of sophisticated behavior in a relatively sparse institutional context. In a related vein, Haney, Herzberg, and Wilson (1992) explore some of the difficulties associated with the efforts of leaders to garner useful advice from a small group of advisors. Again, however, the specific nature of the laboratory setting may not have captured the rich institutional context of foreign policy making in the real world. Nonetheless, laboratory experiments have proven to be a very useful means of investigating the implications of alternative institutional arrangements.

    Conflict and Governance

    The dilemmas of collective action in the international arena take center stage in the next selection. In “Policy Uncertainty and Two-Level Games: Examples of Correlated Equilibria,” (Chapter 7), John T. Williams and I examine situations in which a diverse array of information is readily available for policy makers to consider. We argue for a broader interpretation of relationships between national and international politics, by justifying a formulation of “two-level games” that differs in substantial ways from Putnam’s (1988) more influential formulation. Whereas Putnam specifies a particular sequence of events in which the results of international negotiations are presented to domestic legislative institutions for approval or ratification, we argue that policy advocates in all states involved in international interactions are going to have access to extensive amounts of information that should help them predict the likely behavior of other governments. As a consequence, we argue that Aumann’s (1987) correlated equilibrium may be a more appropriate tool for modeling the outcomes of these two-level interactions. This essay differs from much of the game theoretic tradition by not specifying the steps by which individuals obtain information, but it does present an interesting alternative to more standard representations based on games of incomplete information (or signaling games).

    Aumann’s notion is disarmingly simple. Even though game theorists may be uncertain which of the many possible equilibria will be selected by the players, the particpants have access to a wide array of informational cues that they can use to help them coordinate on a particular solution. This is not to say that coordination problems are trivial; far from it. As we show in this chapter, a failure of coordination may occur with some positive probability even for the case of correlated equilibrium.

    We do not specify the sources of informational cues in our general analysis of the relevance of correlated equilibrium to two-level games, precisely because the nature of the relevant cues will differ in different contexts. However, we could have placed more emphasis on the importance of institutions as providing cues to the players as they seek to coordinate their behavior, either explicitly or implicitly. For example, CIA estimates of Soviet military spending played a very important (and controversial) role in helping sustain the behavior patterns underlying rivalry between the two superpowers during the Cold War (see McGinnis and Williams 1998).

    The concept of correlated equilibrium is equally relevant at both of the levels of analysis included in a domestic-international game. Competing policy advocates within any one country have a common interest in arriving at a collectively sound policy, but conflicting interests over the content of that policy. A similar mixture of common and conflicting interests recurs at the level of international interactions, for the leaders of rival powers may share a common interest in avoiding war or excessively high arms expenditures. Elsewhere (McGinnis and Williams 1998), we demonstrate that under conditions of international rivalry we can expect the efficient use of information in policy debates to resolve domestic coordination problems while simultaneously exacerbating problems of international cooperation. From the perspective of polycentric games, it is important to understand multi-level dilemmas of collective action and other cross-level linkages.

    Under our interpretation of Aumann’s correlated equilibrium solution concept, participants are conditioning their behavior on the common signals generated from the vast amount of information that becomes available in debates over contentious public issues. No one individual has access to all relevant information, and reasonable people are likely to interpret this same information as support of contradictory policy initiatives. Still, at the aggregate level, there may be circumstances under which information use may be so efficient as to approximate the behavior expected of a unitary rational actor. Game models in the international relations literature often make this assumption, and our research was one effort to delineate the circumstances under which that assumption is most warranted.

    On the other hand, if very little information is generated, then we are inclined to be much more skeptical of the appropriateness of treating collective entities as fully rational actors. There are also reasons to wonder whether individual humans are fully capable of the prodigious feats of rational calculation routinely attributed to them by many game theorists. Under conditions of competitive markets it may prove reasonable to assume that only the most efficient actors survive, but in many other decisional contexts there remains plenty of room for individuals to indulge other predilections.

    The potential instability of processes of majority voting (Arrow 1963; Riker 1982) are a particularly pernicious source of coordination dilemmas in democratic polities. For this reason, we focused on voting examples in our discussion of the implications of correlated equilibria in two-level games.

    Voting Over Resource Management Issues

    A natural next step at this point would be to draw out explicit linkages between voting institutions and the management of common-pool resources. Vincent and Elinor Ostrom (1977) apply Buchanan and Tullock’s (1962) classic analysis of constitutional choice to the management of common-pool resources. Their basic question is this: If a community is confronted with the problem of managing a CPR, how might they go about setting up a means of voting on these collective problems? Buchanan and Tullock point out the costs associated with finding out the preferences of all individuals (as is necessary under a rule of unanimity) and the dangers of having a collective decision imposed against one’s will (as happens to minorities under a system of majority rule). It turned out that voting per se played a more minor role in the CPR management than was originally expected. Instead, a wide array of cases studied by field researchers demonstrated that communities used other, less formal mechanisms to arrive at collective decisions.

    After investigating many cases from the field and from laboratory experiments, it became apparent that majority voting may have even more negative connotations than was originally suggested in this early work. In Walker, Gardner, Herr and Ostrom (1997), the authors report on an experimental analysis of the ways in which majority voting schemes can be used to resolve the distributional conflicts associated with the management of common-pool resources. The results of these experiments are troubling, for those subjects who were in the minority felt they were dealt with unfairly. Basically, members of a winning coalition divided all of the positive gains amongst themselves, leaving nothing for the other players. Subject comments make it clear that the ways in which majority voting was used in this experimental setting violated deeply felt notions of fairness and justice on the part of many of the participants. These experiments demonstrate that simple, majority rule voting is not the best way to deal with some problems.

    Leadership and Principal-Agent Relations

    Disputes concerning which rules are appropriate for application to specific decisions are an important component of interactions in all arenas of collective choice. In the specific context of groups dealing with the management of common-pool resources, discussions often center on which of the relevant allocation schemes is most appropriate or fair in any given situation. In more fully articulated constitutional orders, alternative allocation schemes are enshrined into law (see Chapter 10 below). Concern for interactions between law and alternative institutions for resource management harkens back to research completed by Workshop scholars long before the Workshop was even established (e.g., Ostrom 1953, 1976, Ostrom and Ostrom 1972).

    In the developing world, legal arrangements are less well established, but many of the same concerns are relevant. In “Shepherds and their Leaders Among the Raikas of India: A Principal-Agent Perspective” (Chapter 8), Arun Agrawal shows how a self-governed group of pastoralists structure the process of choosing and monitoring an individual to act in the group’s behalf. This agent is authorized to select times and locations for temporary settlement and migration patterns and other matters for which some means of coordination is required. Agrawal uses a normal-form game matrix to examine the circumstances under which the group as a whole will be able to monitor the activities of this agent. He concludes that situations that facilitate direct monitoring by the group as a whole are most conducive to long-term success of the group, both in economic terms and in terms of fostering the sense of community that such groups need.

    It may be useful to conclude the introduction to this part by acknowledging the diversity of the essays included here. The substantive topics of these essays range from majority voting (primarily in the United States) to leadership of pastoral communities (in India) to the multi-level nature of international relations (throughout the world). The methods include formal models illustrated with abstract examples, laboratory experiments, field research, and philosophical investigations. As a set, these essays represent a microcosm of the Workshop approach to institutional analysis. Despite this diversity in methods and substantive topics, certain common concerns permeate all of these works. How do groups make collective decisions? What information is relevant to making these decisions? How are the interests of potentially aggrieved groups taken into account when evaluating policy changes? And, especially, how can communities restrain the opportunistic behavior of their leaders and agents? None of these questions is answered fully in any of these readings, but each provides a piece of the overall picture.

    References (to works not included in this volume)

    Arrow, Kenneth J. 1963. Social Choice and Individual Values, second edition. New Haven: Yale University Press.

    Aumann, Robert J. 1987. “Correlated Equilibrium as an Expression of Bayesian Rationality” Econometrica 55:1-18.

    Buchanan, James M., and Gordon Tullock. 1962. The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press.

    Haney, Patrick J., Roberta Herzberg, and Rick K. Wilson. 1992. “Advice and Consent: Unitary Actors, Advisory Models, and Experimental Tests.” Journal of Conflict Resolution 36(4):603-633.

    Herzberg, Roberta. 1992. "An Analytic Choice Approach to Concurrent Majorities: The Relevance of John C. Calhoun's Theory for Institutional Design." Journal of Politics, 54 (1): 54-81.

    ________. and Rick K. Wilson. 1988. “Results of Sophisticated Voting in an Experimental Setting.” Journal of Politics 50(2):471-486.

    ________. 1991. “Costly Agendas and Spatial Voting Games: Theory and Experiments on Agenda Access Costs.” In Laboratory Research in Political Economy, ed. Thomas R. Palfrey, 169-99. Ann Arbor: University of Michigan Press.

    Jillson, Calvin, and Rick K. Wilson. 1994. Congressional Dynamics: Structure, Coordination, and Choice in the First American Congress, 1774-1789. Stanford University Press.

    McGinnis, Michael D. 1992. “Bridging or Broadening the Gap? A Comment on Wagner's ‘Rationality and Misperception in Deterrence Theory’.” Journal of Theoretical Politics 4(4):443-57.

    ________. 1996. “Elinor Ostrom: A Career in Institutional Analysis,” PS: Political Science & Politics, 29 (4), 737-741.

    McGinnis, Michael D., ed. 1999a. Polycentric Governance and Development, Ann Arbor: University of Michigan Press.

    ________, ed. 1999b. Polycentricity and Local Public Economies, Ann Arbor: University of Michigan Press.

    McGinnis, Michael D., and John T. Williams. 1998. Compound Dilemmas: Democracy, Collective Action, and Superpower Rivalry. Indiana University, Bloomington: draft book manuscript.

    Ostrom, Vincent. 1953. “State Administration of Natural Resources in the West.” American Political Science Review, Vol. 47, No. 2 (June), 478-493.

    ________. 1976. “John R. Commons’s Foundation for Policy Analysis.” Journal of Economic Issues 10(4) (Dec.): 839-57.

    ________. 1980. “Artisanship and Artifact.” Public Administration Review 40(4) (July-Aug.): 309-17. Reprinted in McGinnis 1999a.

    ________. 1987. The Political Theory of a Compound Republic: Designing the American Experiment. 2d rev. ed. San Francisco: ICS Press (first edition 1971)

    ________. 1991. The Meaning of American Federalism: Constituting a Self-Governing Society. San Francisco: ICS Press.

    ________. 1997. The Meaning of Democracy and the Vulnerability of Democracies: A Response to Tocqueville’s Challenge. Ann Arbor: University of Michigan Press.

    Ostrom, Vincent, and Elinor Ostrom. 1972. "Legal and Political Conditions of Water Resource Development." Land Economics 48(1) (Feb.): 1-14. Reprinted in McGinnis 1999a.

    ________. 1977. “A Theory for Institutional Analysis of Common Pool Problems.” In Garett Hardin and John Baden, eds. Managing the Commons, San Francisco: W. H. Freeman and Company, pp. 157-172.

    Putnam, Robert D. 1988. "Diplomacy and Domestic Politics: The Logic of Two-Level Games," International Organization, 42, 427-460.

    Riker, William H. 1982. Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice. San Francisco: W. H. Freeman.

    Shepsle, Kenneth. 1979. “Institutional Arrangements and Equilibrium in Multi-dimensional Voting Models,” American Journal of Political Science, 23: 27-59.

    Walker, James, Roy Gardner, Andy Herr, and Elinor Ostrom (1998) “Collective Choice in the Commons: Experimental Results on Proposed Allocation Rules and Votes.” Re-submitted to The Economic Journal, Dec. 1998. Paper presented at: (1) the 1997 Annual Meeting of the Western Political Science Association, Tucson, AZ, March 13-15, 1997; (2) the Department of Economics, Krannert School of Management, Purdue University, Lafayette, IN, March 27, 1996; and (3) Conference on "Game Theory in the Behavioral Sciences," Tucson, AZ, October 11-12, 1995. Working Paper, Workshop in Political Theory and Policy Analysis, Indiana University.

    Introduction to Part III

    This Part combines a general defense of game theory with two models of the ways in which rules and regulations shape community efforts to manage their natural resources.

    Games and Models

    In “Heterogenous Players and Specialized Models” (Chapter 9), Eric Rasmusen defends game theory as a method that introduces analytical discipline into the efforts of social scientists to understand specific situations. (This essay was originally included in a special issue of Rationality and Society on “The Use of Game Theory in the Social Sciences.”) Game theory is often dismissed as being unable to handle situations in which the participants have heterogenous interests or capabilities. Rasmusen argues that only those overly simple game models that claim comprehensive coverage fall victim to this complaint. He argues that the real strength of game theory lies in its ability to trace out the implications of interactions among actors with divergent interests and capabilities, provided that each model is designed specifically to apply to particular empirical situations. With such detailed models analysts can investigate the likely consequences of changes in the parameters of that situation.

    Although this chapter’s running example of teams choosing between offensive and defensive strategies in a football game may seem trivial, the general point is an important one. Rasumusen wrote this essay before he became closely associated with the Workshop, but his comments reveal an attitude very much in line with the characteristic Workshop emphasis on models and empirical analyses of narrowly defined empirical situations. For it is only in the context of rich empirical settings that the implications of alternative institutional arrangements can be fully understood. The other two selections in this part of the volume present detailed models (formal or informal) of particular empirical contexts related to resource management.

    Evaluating Alternative Legal Doctrines

    In “Governing a Groundwater Commons: A Strategic and Laboratory Analysis of Western Water Law” (Chapter 10), Roy Gardner, Michael Moore, and James Walker use game models and laboratory experiments to investigate the differing implications of alternative legal doctrines. Their specific empirical referent is governance of groundwater resources in the western United States. They define four legal doctrines that are in place for different groundwater systems in these states. Each legal doctrine defines the rights that participants have to extract water from these systems. The prior appropriation doctrine, for example, asserts that those users who have already been using the water have established the right to continue to extract water in the future. For the state of California, this system was described by Vincent Ostrom (1967) as an after-effect of the practices adopted by miners during the era of the gold rush in the mid-eighteenth century.

    The authors interpret each of these legal doctrines in terms of the concepts used to model different common-pool resources. For example, the prior appropriation doctrine amounts to a restriction on entry. Thus, the crucial aspect of this legal doctrine is its value on the entry rule dimension of the rule configuration, as presented by Elinor Ostrom in chapter 3. The authors argue that this entry restriction should act to mitigate the negative consequences of strategic and stock externalities. (See also Schlager and Ostrom 1993 and Schlager et al. 1994.)

    The authors next formulate a model of the extraction of water by homogenous users. The “optimal” level of water that would be allocated by a “benevolent central planner” is used to define the benchmark of an “efficient” outcome. This outcome is unlikely to be achieved in real-world settings, because each participant is directly concerned other with their own payoffs, and does not take into account the external effects of their own actions on other players. The subgame perfect (Nash) equilibrium predicted by this noncooperative game version results in a less-efficient outcome, in the sense that more water is extracted from the system than can be sustained over the long term.

    Finally, the authors use different experimental settings to represent the implications of alternative legal doctrines. For example, the effect of entry restrictions (as implicit in the prior appropriation doctrine) is implemented in the laboratory by comparing the levels of the common-pool resource appropriated by groups of 5 or 10 subjects. The other parameters of the game are adjusted to isolate a pure size effect. The procedures followed in these experiments are discussed more fully in Ostrom, Gardner, and Walker (1994) and in Part V of this volume. The results of these experiments clearly demonstrate that institutions matter, in the sense that the systematically different outcomes are observed from the different experimental treatments representing different legal doctrines. In particular, subjects in the experimental treatment corresponding to limited entry obtained a higher level of efficiency than the unlimited entry treatment, even though in both cases the results were even worse than those predicted by the subgame perfect (Nash) equilibrium. Stock quotas had a more demonstrable positive effect, but were still well short of the overall optimum. These results are quite similar to those produced by other experiments on common-pool resources; see Ostrom, Gardner, and Walker (1994) and Part V of this volume.

    Governance and Transaction Costs

    The next chapter moves us to the private sector. In “Bottlenecks and Governance Structures: Open Access and Long-Term Contracting in Natural Gas” (Chapter 11), Thomas Lyon and Steven Hackett build upon influential research by Oliver Williamson (1975, 1985, 1996) on the many ways in which firms strive to the transaction costs of conducting business in sectors where long-term contracting is an appealing alternative to spot markets. One major solution is to engage in vertical integration. That is, by incorporating more aspects of the sequence by which raw materials are transformed into products made available to customers, a firm can lower its danger of being taken advantage of by suppliers or distributors at each step of this process. Williamson details the transition from long-term contracting to vertical integration, which incorporates both parties to these contracts within the same organization.

    The authors examine a sector in which vertical integration has obvious advantages to the firm, namely, the production and delivery of natural gas. In particular, long-term contracts are very common in the operation of gas pipelines. However, the energy industry has also been an area of active intervention by the United States government, seeking to limit the extent of vertical integration. Lyon and Hackett begin by stating that the basic problem confronting government regulators is how to keep this market as competitive as possible while still assuring the enforcement of the long-term contracts that are necessary for firms to invest in this inherently risky sector.

    Firms (or governments) that have frequent interactions with each other may agree to designate each other as most-favored customers (or nations). In this way each is assured that they will not be hurt by any subsequent contracts (or treaties) signed with other corporations (or governments). As Besanko and Lyon (1993) demonstrate, such arrangements can sustain an oligopoly’s domination of an industry. However, in an empirical analysis of the natural gas industry, Crocker and Lyon (1994) conclude that the reduction of transaction costs is the major reason behind most-favored nation clauses in natural gas contracts.

    Lyon and Hackett focus their attention on the conditions under which owners of natural gas pipelines implement a policy of open-access, which in this context means that pipelines should be treated as common carriers rather than being dedicated to the use of a single firm. (In common-pool resource settings, open-access refers to the absence of restrictions on the number of users, as discussed above.) They argue that government regulators have given insufficient consideration to the transaction cost benefits of discriminatory policies (that is, the oppositive of open-access rules). They investigate a series of hypotheses concerning the relative magnitude of spot markets and long-term contracting in this sector of the economy that should occur under different regulatory conditions. They conclude that although an increased reliance on open-access policies does reduce the threat of opportunism by those controlling the pipelines, these policies also increase the costs of transacting in this sector.

    This analysis by two Workshop-affiliated scholars draws on the standard presumption that governments should act to reduce the costs of economic transaction, and thereby contribute to the public good of economic growth. In practice, however, the rules and regulations enforced by governmental authorities have exactly the opposite effect. For example, Eggertsson (1992, 1996) investigates an “equilibrium trap” that Icelanders experienced for an incredibly long time period. Despite the ready availability of fish in Icelandic waters, a sophisticated fishing industry never arose in Iceland until well after the establishment of comparable industries elsewhere. In the intervening years, fishers from many parts of Europe came thousands of miles to Icelandic waters while local fishers stayed closer to shore.

    The title of Eggertsson’s 1996 essay alludes to the phrase “great experiments and monumental disasters” that Vincent Ostrom (1991, 1997) uses to describe the experience of the great revolutions of the twentieth century. In countries such as the Soviet Union and China revolutionary movements intended to create a more egalitarian system instead ended up increasing the misery of ordinary people. Ostrom uses this phrase to sound a cautionary note in response to repeated calls for wide-ranging and immediate reform. He argues that policy analysts, while remaining open to exploring the potential benefits of institutional change, should remain restrained in their ambitions. By trying to do too much all at once, would-be revolutionaries may make things worse.

    Eggertsson uses his trademark gentle humor to remind us that the absence of reform can have equally detrimental consequences. The inability of Icelandic fishers to exploit readily available resources so close to home significantly stunted their economic development. He shows how difficult it was for Icelandic fishers to step outside their assigned role within the closed political system of governance in Iceland. External factors also contributed, especially the interest of the Danish crown in providing certain towns with monopoly rights to trade with Iceland. In effect, then, Iceland’s equilibrium trap is an example of a “two-level game” (see Chapter 7) in which the interests of internal and external actors reinforced each other to produce a stable policy outcome. Eventually, however, Iceland emerged as a major fishing nation.

    The topic of the longer-term consequences of large-scale fishing in the North Atlantic lies outside the scope of this volume, but the chapters in this part clearly demonstrate the importance of understanding the institutional context of resource management. Similar concerns must be addressed whether the resource is managed primarily via private ownership (as in the natural gas industry) or by public agencies (as in the case of watershed management in the Western United States). Depending on its nature, the existing structure of governance can facilitate more efficient management of resources, speed the destruction of that resource, or delay development indefinitely (as in Icelandic fisheries). As Rasmusen argues, game models of specific situations can help us understand the consequences of alternative institutional arrangements. Such an understanding is a crucial part of the process of institutional design and policy evaluation.

    References (to works not included in this volume)

    Besanko, David, and Thomas P. Lyon. 1993. “Equilibrium Incentives for Most-Favored Customer Clauses in an Oligopolistic Industry,” International Journal of Industrial Organization, 11: 347-367.

    Crocker, Keith J., and Thomas P. Lyon. 1994. “What do ‘Facilitating Practices’ Facilitate?: An Empirical Investigation of Most-Favored Nation Clauses in Natural Gas Contracts,” Journal of Law and Economics, 36: 297-322.

    Eggertsson, Thrainn. 1992. “Analyzing Institutional Successes and Failures: A Millenium of Common Mountain Pastures in Iceland,” International Review of Law and Economics, Vol. 12: 423-437.

    ________. 1996. “No Experiments, Monumental Disasters: Why it Took a Thousand Years to Develop a Specialized Fishing Industry in Iceland,” Journal of Economic Behavior and Organization, 30: 1-23.

    McGinnis, Michael D., ed. 1999a. Polycentric Governance and Development, Ann Arbor: University of Michigan Press.

    ________, ed. 1999b. Polycentricity and Local Public Economies, Ann Arbor: University of Michigan Press.

    Ostrom, Elinor, Roy Gardner, and James Walker, with Arun Agrawal, William Blomquist, Edella Schlager, and Shui-Yan Tang. 1994. Rules, Games, and Common-Pool Resources. Ann Arbor: University of Michigan Press.

    Ostrom, Vincent. 1967. "Water and Politics California Style." Arts and Architecture, 84 (July/August), 14-16, 32. Reprinted McGinnis 1999a.

    _____. 1991. The Meaning of American Federalism: Constituting a Self-Governing Society. San Francisco: ICS Press.

    _____. 1997. The Meaning of Democracy and the Vulnerability of Democracies: A Response to Tocqueville’s Challenge. Ann Arbor: University of Michigan Press.

    Schlager, Edella, William Blomquist, and Shui Yan Tang. 1994. “Mobile Flows, Storage, and Self-Organized Institutions for Governing Common-Pool Resources.” Land Economics 70(3) (Aug.): 294-317. Reprinted in McGinnis 1999a.

    Schlager, Edella, and Elinor Ostrom. 1993. “Property-Rights Regimes and Coastal Fisheries: An Empirical Analysis.” In The Political Economy of Customs and Culture: Informal Solutions to the Commons Problem, ed. Terry L. Anderson and Randy T. Simmons, 13-41. Lanham, MD: Rowman & Littlefield. Reprinted in McGinnis 1999a.

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

    ________. 1985. The Economic Institutions of Capitalism. New York: Free Press.

    ________. 1996. The Mechanisms of Governance. New York: Oxford University Press.

    Introduction to Part IV

    One of the major points made by Elinor Ostrom in Chapter 3 is that institutional analysts need to incorporate the particular roles filled by different actors in empirical situations; Eric Rasmusen’s argument in Chapter 9 leads to a similar conclusion. Workshop scholars have demonstrated the central importance of monitoring and sanctioning to the successful operation of any form of governance, ranging all the way from micro-level irrigation systems to issues of global environmental change (see McGinnis 1999a). The chapters included in Part IV model the process of monitoring and enforcement in different ways and in different contexts.

    Models of Rule Enforcement

    In “Transforming Rural Hunters into Conservationists: An Assessment of Community-Based Wildlife Management Programs in Africa” (Chapter 12), Clark Gibson and Stuart Marks use extensive form game models to represent the fundamental dilemmas of wildlife management in Africa. This essay is especially useful as an illustration of the interactions among the diverse factors discussed throughout this volume of readings. Their model highlights, for example, the central question of whether actors have the requisite incentives and capabilities to carry out the monitoring and sanctioning activities needed for a successful wildlife preservation program. The advantages of having locally managed preservation schemes come through loud and clear. Yet local self-governance is by no means automatic, for local farmers have multiple incentives, many of which are in conflict with the goal of wildlife management.

    The next selection develops an abstract representation of an irrigation system to explore the implications of different forms of monitoring behavior. In “Irrigation Institutions and the Games Irrigators Play: Rule Enforcement on Government- and Farmer-Managed Systems” (Chapter 13), Franz Weissing and Elinor Ostrom build on a previous model (Weissing and Ostrom 1991). In that model, no separate roles for monitors or sanctioners are included. The implications of that analysis serve as a benchmark for comparison with models including two different types of monitors. In particular, the authors identify conditions under which farmers should be expected to have the appropriate incentives to carry out monitoring activities. As is typical in any reasonably complex model, different equilibrium conditions apply for different configurations of parameters. Without going through all the equilibria conditions here, one implication deserving emphasis is that it is indeed possible for rational actors to monitor and sanction each other, at least under some conditions. In short, this model serves as a formal demonstration that self-governance is indeed possible.

    At the same time, these models always predict a nonzero level of stealing in equilibrium. That is, some participants will, at least some of the time, violate the rules governing levels of water to be drawn from an irrigation system. In the follow-up essay included in this volume, these same authors investigate alternative means to lower the probability of rule violations.

    Weissing and Ostrom introduce distinctions between two different types of monitors: “integrated” guards whose payoffs are directed related to the physical success of the local farmers and “disassociated guards” whose payoffs are determined by some measure of violations detected. Integrated guards represent monitors who are themselves part of the community managing the irrigation system, whereas disassociated guards represent the incentives facing monitors appointed by the central government or some other authority not so directly concerned with the long-term viability of this irrigation system.

    By comparing the equilibria conditions for models of irrigation systems in which the necessary monitoring and sanctioning tasks are carried out by integrated or disassociated guards, they demonstrate that under a reasonable range of conditions the farmer managed systems are more efficient and sustainable than are systems managed by governmental agents. This result comports nicely with the observation of Lam, Lee, and Ostrom (1997) of exactly this difference between effective farmer-managed and ineffective government-managed irrigation systems in Nepal. Thus, Ostrom and Weissing have provided a formal game theoretic representation of results observed in field studies of actual irrigation systems.

    Both essays were based on a relatively simple addition to a standard model of an irrigation system, namely, the incorporation of actors who received benefits from detecting cheating. This seemingly simple change, however, necessitated a complicated series of analytical steps for the derivation of equilibria conditions. To cope with the multiplicity of possible equilibria solutions to these games, the authors chose to rely on the equilibrium selection criterion introduced by Harsanyi and Selten (1988). Not all game theorists are convinced that this particular process of selecting a unique equilibrium is valid in all empirical contexts. It is difficult to understand why actual game players should be expected to follow this excruciatingly logical thought process required to arrive at this specific criterion. At least as far back as Schelling’s (1960) classic depiction of the power of “salient” solutions or focal points, it has been widely recognized that the existence of multiple equilibria presents a difficult challenge to game theorists, specifically in understanding the ways in which the players themselves cope with the complexity of multiple solutions (see also Kreps 1990). A few alternative formulations that incorporate limitations on human cognition are discussed in Ostrom, Gardner, and Walker (1994) and in Part V of this volume.

    Rule Enforcement in Asymmetric Games

    The first two essays in this Part demonstrate that it is possible to build tractable game models in which the parameters of the model bear a clear connection to important aspects of the related real-world phenomenon. The Workshop research program on common-pool resources has been truly unique in its ability to draw close and complementary interconnections among formal game models, laboratory experiments, and field research.

    In “Coping with Asymmetries in the Commons: Self-Governing Irrigation Systems Can Work” (Chapter 14), Elinor Ostrom and Roy Gardner combine a simple model with the results of field research to demonstrate that farmer-managed irrigation systems can be sustained even when the community includes two groups of farmers with conflicting interests. In this case, farmers located at the head or tail end of an irrigation system have different interests. Specifically, “headenders” appear to have an inherent advantage, because they can extract their needed water before the downstream “tailenders” get a chance. But in the situations modeled here, the headenders are themselves unable to shoulder the burden of maintenance of the entire irrigation system. Thus, headenders must find a way to encourage tailenders to contribute to the system’s maintenance, which can only happen if the headenders can restrain themselves from extracting too much water. This is a very powerful form of asymmetry, one that is often found in real-world irrigation systems.

    That particular model is less elaborate in a technical sense than the other models included in this part, but it implies a very important conclusion. Despite a fundamental asymmetry in interests, there are conditions under which we can expect rational individuals of both actor types to cooperate in the maintenance of this irrigation system. Too often formal models or laboratory experiments assume or impose a condition of symmetry on participants, and this chapter is a good illustration of the potential benefits of an alternative mode of analysis in which more of the most important aspects of actual situations can be incorporated into formal models or experimental studies. In addition, this essay demonstrates that formal models and field research can be cross-integrated in productive ways. Similar results have been observed in experimental studies, which is the topic of the concluding part of this volume. (See also Ostrom, Gardner, and Walker 1994).

    References (to works not included in this volume)

    Harsanyi, John C., and Reinhard Selten. 1988. A General Theory of Equilibrium Selection in Games. Cambridge, MA: MIT Press.

    Kreps, David M. 1990. Game Theory and Economic Modelling. New York: Oxford Univeristy Press.

    Lam, Wai Fung, Myungsuk Lee, and Elinor Ostrom. 1997. "The Institutional Analysis and Development Framework: Application to Irrigation Policy in Nepal." In Derick W. Brinkerhoff, ed. Policy Studies and Developing Nations: An Institutional and Implementation Focus, vol. 5. Greenwich, CT: JAI Press, 53-85.

    McGinnis, Michael D. ed. 1999a. Polycentric Governance and Development, Ann Arbor: University of Michigan Press.

    Schelling, Thomas C. 1960. The Strategy of Conflict. New Haven: Harvard University Press.

    Weissing, Franz J., and Elinor Ostrom. 1991. “Irrigation Institutions and the Games Irrigators Play: Rule Enforcement without Guards.” In Game Equilibrium Models II: Methods, Morals, and Markets, ed. Reinhard Selten, 188-262. Berlin: Springer-Verlag.

    Introduction to Part V

    The two readings in Part V provide general overviews of a large number of laboratory experiments that have investigated the ability of groups of experimental subjects to achieve a collectively desirable outcome. The relatively controlled environment of a laboratory experiment allows researchers to isolate the effects of particular factors on the ability of groups of individuals to cooperate for their common good. In effect, these laboratory settings can be seen as experiments in self-governance: under what conditions can a group of individuals brought together on a relatively random basis learn to cooperate to achieve a common goal? From this perspective, an experimental laboratory is a reflection in microcosm of the dilemmas faced by groups of humans throughout history and throughout all realms of social life.

    Cooperation in Laboratory Experiments

    In “Neither Markets Nor States: Linking Transformation Processes in Collective Action Arenas” (Chapter 15), Elinor Ostrom and James Walker survey a long series of research reports prepared by Workshop scholars and their collaborators, especially related to the behavior of experimental subjects in public goods settings. They compare experimental treatments that vary the size of the group, the magnitude of their endowments, the similarity or heterogeneity of interests among participants, and especially the amount of information available to the experimental subjects as they make their decisions.

    In a typical baseline laboratory set-up, subjects have no opportunity to communicate or make any kinds of deals with other participants. The authors’ review of a large body of evidence highlights the ability of groups of experimental subjects to achieve better outcomes simply by being allowed to talk with each other, even if only briefly at the beginning of the play of the game. This result would be considered surprising by many game theorists because rational actors should not consider costless communication, so-called “cheap talk,” to be effective when each individual can obtain a higher payoff (at least in the short run) by ignoring these signals. Yet, these experiments provide rigorous and replicable evidence that even cheap talk does seem to help in social dilemma situations. Laboratory subjects can do even better if they are given the opportunity to make arrangements to sanction the behavior of others. Since this experimental treatment comes the closest to reflecting the basic structure found in field settings, it should not be surprising that these results sound so similar to the field research reported elsewhere.

    These experimental results address the central concern of politics: the nature of governance. Ostrom, Walker, and Gardner (1992) use a series of laboratory experiments to investigate the ability of experimental subjects to devise their own means of governance. (The results of this and related essays are summarized in Chapter 16.) One might think that it should come as no surprise to American political scientists that self-governance is possible; after all, that is the whole basis for the American system of constitutional order. Yet, what was surprising is how easily a modicum of self-governance could be introduced into the admittedly artificial situations of a laboratory experiment. Even in these stark situations, where long-term considerations were absent, experimental subjects demonstrated an ability, even an eagerness, to form their own system of governance, to monitor each other’s behavior (to the limited extent made possible by the experimental design), and to enforce their agreement by punishing transgressors. What was remarkable was that all this was so easy to achieve, simply by allowing the subjects a limited opportunity to communicate and to sanction each other. If self-governance is possible in such a starkly limited environment, then it is certainly relevant to more consequential interactions in the real world. These experimental results remind us of the crucial importance of self-governance.

    Rethinking the Nature of Rational Choice

    Although often dismissed as merely a technical tool, game theory can force scholars to confront foundational issues of the nature of human cognition and common understandings. One important outgrowth of this use of game theory to understand cognition is the alternative version of game rationality presented in the conclusion of Ostrom, Gardner, and Walker (1994). In an effort to explain the results of their laboratory experiments, they introduce a “measure-for-measure” behavioral strategy in which individuals react in a measured way to the rule violations of other participants. (This strategy was also influenced by field research demonstrating the importance of graduated sanctions as a contributor to successful management of common-pool resources; see Ostrom 1990.)

    Several game theorists have been struggled to develop more realistic models of the cognitive processes of real strategic players, as an alternative to the “hyper-rationality” characterizing the standard approach to game theory (Binmore 1990; Gardner 1995; McGinnis 1992; Rubinstein 1998; Samuelson 1997; Scharpf 1997). This remains very much a work-in-progress, but it is likely that future game models are likely to become much more sophisticated and realistic in their depictions of individual cognition and institutional arrangements.

    In her Presidential Address for the American Political Science Association, “A Behavioral Approach to the Rational Choice Theory of Collective Action” (Chapter 16), Elinor Ostrom uses the results of experimental research to suggest a new conceptualization of the nature of individual actors. In what she describes as a “second generation model” of rational choice, she assumes that most individuals are predisposed to use reciprocity norms in many circumstances and to pay careful attention to the behavior of other individuals in all circumstances. Thus, their interactions unfold in different directions depending on the nature of these initial expectations and the intermediate play of the game. Nonetheless, certain regularities of behavior emerge from years of experimental research, regularities that do not always comport with the expectations of rational choice theorists.

    Much remains to be done to formalize this “second generation” behavioral model of rational choice and to fully test its implications in diverse institutional settings, but even at this preliminary stage it is clear that the type of individual actors posited in this speech sound very much like the farmers, fishers, and other groups of individuals who have been found to have the capacity to organize themselves to achieve common goals in real-life situations. If we can find even a glimmer of a capacity for self-governance among anonymous individuals randomly thrown together in the confines of a computer laboratory, then we should hardly be surprised when communities of individuals who have come to know each other well are able to cooperate on tangible and important problems. Unfortunately, many social scientists continue to be puzzled by the human capacity for collective self-governance in the absence of governmental direction or coordination. Familiarity with the research reported in this book should suffice to dispel that sense of puzzlement.

    References (to works not included in this volume)

    Binmore, Kenneth G. 1990. Essays on the Foundations of Game Theory. Cambridge, MA: Blackwell.

    Gardner, Roy. 1995. Games for Business and Economics. New York: Wiley.

    McGinnis, Michael D. 1992. “Bridging or Broadening the Gap? A Comment on Wagner's ‘Rationality and Misperception in Deterrence Theory’.” Journal of Theoretical Politics 4(4):443-57.

    Ostrom, Elinor. 1990. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

    Ostrom, Elinor, Roy Gardner, and James Walker, with Arun Agrawal, William Blomquist, Edella Schlager, and Shui Yan Tang. 1994. Rules, Games, and Common-Pool Resources, Ann Arbor: University of Michigan Press.

    Ostrom, Elinor, James Walker, and Roy Gardner. 1992. "Covenants with and without a Sword: Self-Governance is Possible." American Political Science Review, 86: 404-17.

    Rubinstein, Ariel. 1998. Modeling Bounded Rationality. Cambridge, MA: MIT Press.

    Samuelson, Larry. 1997. Evolutionary Games and Equilibrium Selection. Cambridge, MA: MIT Press.

    Scharpf, Fritz Wilhelm. 1997. Games Real Actors Play: Actor-Centered Institutionalism in Policy Research. Boulder, CO: Westview Press.

    Suggested Further Readings

    The best place to begin is Rules, Games, and Common-Pool Resources (University of Michigan Press, 1994), written by Elinor Ostrom, Roy Gardner, and James Walker with the assistance of four Workshop colleagues. This is a truly unique book that fully integrates field research, formal models, and laboratory experiments, all focused on the management of common-pool resources. In short, this book it exemplifies the Workshop’s multi-faceted approach to institutional analysis. Field research is the primary topic of Elinor Ostrom’s award-winning book Governing the Commons: The Evolution of Institutions for Collective Action (Cambridge University Press, 1990), but this book also includes a brief introduction to the use of Prisoner’s Dilemma games and other formal representations of social dilemmas.

    As readers of this volume are well aware, the Workshop approach to institutional analysis draws inspiration from a wide variety of sources. For influential overviews of the closely related fields of research known as public choice, social choice, new institutional economics, and constitutional economics, readers should consult (respectively) Dennis C. Mueller, Public Choice II (Cambridge University Press, 1989), and William H. Riker, Liberalism Against Populism: A Confrontation Between the Theory of Democracy and the Theory of Social Choice (W.H. Freeman, 1982), Thráinn Eggertsson, Economic Behavior and Institutions (Cambridge University Press, 1990), and James M. Buchanan and Gordon Tullock, The Calculus of Consent: Logical Foundations of Constitutional Democracy (University of Michigan Press, 1962). Perspectives on Public Choice: A Handbook (Cambridge University Press, 1997), edited by Dennis C. Mueller, includes excellent overviews contributed by most of the major players in the public choice tradition.

    Three Workshop-affiliated scholars have written textbooks on game theory or environmental economics: Roy Gardner, Games for Business and Economics (Wiley, 1995), Eric Rasmusen, Games and Information: An Introduction to Game Theory, 2nd edition (Blackwell, 1994), and Steven Hackett, Environmental and Natural Resources Economics: Theory, Policy, and the Sustainable Society (M.E. Sharpe, 1998). John Williams and Kenneth Bickers are writing a general textbook on policy analysis, Public Policy Analysis: A Political Economy Approach (Houghton-Mufflin, forthcoming), that incorporates the Workshop perspective. Textbooks focusing on particular policy areas include Local Government in the United States (Institute for Contemporary Studies [ICS] Press, 1988), by Vincent Ostrom, Robert Bish, and Elinor Ostrom, and Institutional Incentives and Sustainable Development: Infrastructure Policies in Perspective (Westview Press, 1993), by Elinor Ostrom, Larry Schroeder, and Susan Wynne.

    Workshop-affiliated scholars have also contributed to several edited volumes in which game models have been applied to a diverse array of empirical settings, including Games in Hierarchies and Networks: Analytical and Empirical Approaches to the Study of Governance Institutions, edited by Fritz W. Scharpf (Westview Press, 1993), Social Dilemmas and Cooperation, edited by Ulrich Schulz, Wulf Albers, and Ulrich Meuller (Springer-Verlag, 1994), Understanding Strategic Interaction: Essays in Honor of Reinhard Selten, edited by Wulf Albers, Werner Guth, Peter Hammerstein, Benny Moldovanu, and Eric van Damme (Springer-Verlag, 1997), and a four-volume set edited by Reinhard Selten, Game Equilibrium Models I-IV (Springer-Verlag, 1991).

    Empirical applications and laboratory experiments have played important roles in the development of the research programs surveyed in this volume. Examples of such applications are collected in Laboratory Research in Political Economy, edited by Thomas R. Palfrey (University of Michigan Press, 1991) and Empirical Studies in Institutional Change, edited by Lee J. Alston, Thráinn Eggertsson, and Douglass C. North(Cambridge University Press, 1996). Calvin Jillson and Rick K. Wilson, Congressional Dynamics: Structure, Coordination, and Choice in the First American Congress, 1774-1789 (Stanford University Press, 1994) shows how the findings of experimental research on voting systems can inform our understanding of important historical processes.

    Earlier versions of essays included in this volume can be found in several edited volumes that also include contributions by scholars from other research institutes. The contributors to Franz-Xaver Kaufmann, Giandomenico Majone, and Vincent Ostrom, eds. Guidance, Control, and Evaluation in the Public Sector, (Walter de Gruyter, 1986) suggest several related approaches towards resolution of the conceptual difficulties involved in modeling large-scale systems. A shorter version of this report on the results of the Bielefeld interdisciplinary project was later published as The Public Sector-Challenge for Coordination and Learning (Walter de Gruyter, 1991), edited by Franz-Xaver Kaufmann. The broader implications of Workshop research for development issues are explored in Rethinking Institutional Analysis and Development, 2d ed. (ICS Press, 1993), edited by Vincent Ostrom, David Feeny, and Hartmut Picht. This book includes a mixture of micro-level and macro-level applications.

    The current volume was prepared in conjunction with two other volumes of previously published reports on research by Workshop-affiliated scholars. Polycentric Development and Governance (University of Michigan Press, 1999) focuses on the management of common-pool resources by self-governing communities throughout the world. This book includes early statements of the Workshop perspective, dating back before game theory had made much of an impact on the study of institutions. This volume complements the current one very well, because most of the models and experiments developed here are based on a generic representation of the problems associated with managing common-pool resources.

    It remains to be seen whether simple game models can be equally successful in helping us to understand the complexities of public service provision in metropolitan areas, which is the primary focus of Polycentricity and Local Public Economies (University of Michigan Press, 1999). This work includes classic essays on the nature of polycentric order and a series of research reports comparing the performance of large and small police agencies in selected metropolitan areas of the United States.

    Charlotte Hess maintains an extensive bibliography on game theory at Bibliographies on related topics are also available at Four volumes of Common-Pool Resources and Collective Action: A Bibliography have been published by the Workshop in Political Theory and Policy Analysis. Volume 1 (1989) and 2 (1992) were edited by Fenton Martin, and volumes 3 (1996) and 4 (1998) by Charlotte Hess. By the time this book appears in print we hope to have all four volumes available in a single file on CD-ROM. Finally, readers are encouraged to check the Workshop’s home page ( for recent updates and for links to other reference and teaching materials.

    Back to the Workshop Homepage
    Copyright 1999, Workshop in Political Theory and Policy Analysis
    Last updated:  March 24, 1999