Governance in a Partially Globalized World PDF 2000 ASP Presidential Address
Document Details
Uploaded by Deleted User
2000
Robert O. Keohane
Tags
Related
Summary
This document is a presidential address by Robert O. Keohane to the American Political Science Association in 2000 discussing the challenges of designing governance institutions in a partially globalized world. The author outlines some desirable institutional features, the limits of rational egoism, and the role of beliefs and persuasion in governing effectively.
Full Transcript
American Political Science Review Vol. 95, No. 1 March 2001 Governance in a Partially Globalized World Presidential Address, American Political Science Association, 2000 ROBERT O. KEOHANE Duke University...
American Political Science Review Vol. 95, No. 1 March 2001 Governance in a Partially Globalized World Presidential Address, American Political Science Association, 2000 ROBERT O. KEOHANE Duke University F acing globalization, the challenge for political science resembles that of the founders of the United States: how to design institutions for a polity of unprecedented size and diversity. Globalization produces discord and requires effective governance, but effective institutions are difficult to create and maintain. Liberal-democratic institutions must also meet standards of accountability and participation, and should foster persuasion rather than rely on coercion and interest-based bargaining. Effective institutions must rely on self-interest rather than altruism, yet both liberal-democratic legitimacy and the meaning of self-interest depend on people’s values and beliefs. The analysis of beliefs, and their effect on institutional outcomes, must therefore be integrated into institutional analysis. Insights from branches of political science as diverse as game theory, rational-choice institutionalism, historical institutionalism, and democratic theory can help political scientists understand how to design institutions on a world—and human—scale. T alk of globalization is common today in the press interdependence include war. As international rela- and increasingly in political science. Broadly tions “realists” have long recognized, interdependence speaking, globalization means the shrinkage of and lack of governance make a deadly mixture. This distance on a world scale through the emergence and Hobbesian premise can be stated in a more positive thickening of networks of connections— environmental form: Globalization creates potential gains from coop- and social as well as economic (Held et al. 1999; eration. This argument is often seen as “liberal” and is Keohane and Nye 2001). Forms of limited associated with Adam Smith and David Ricardo, but it globalization have existed for centuries, as exemplified is actually complementary to Hobbe’s point. The gains by the Silk Road. Globalization took place during the of cooperation loom larger relative to the alternative of last decades of the nineteenth century, only to be unregulated conflict. Both realists and liberals agree reversed sharply during the thirty years after World that under conditions of interdependence, institutions War I. It has returned even more strongly recently, are essential if people are to have opportunities to although it remains far from complete. We live in a pursue the good life (Hobbes 1967; Keohane partially globalized world. 1984; Keohane and Nye 2001). Globalization depends on effective governance, now My second premise is that institutions can foster as in the past. Effective governance is not inevitable. If exploitation or even oppression. As Judith Shklar it occurs, it is more likely to take place through (1984, 244) expresses it, “no liberal ever forgets that interstate cooperation and transnational networks than governments are coercive.” The result is what I will call through a world state. But even if national states retain the governance dilemma: Although institutions are many of their present functions, effective governance of essential for human life, they are also dangerous. a partially—and increasingly— globalized world will Pessimistic about voluntary cooperation, Hobbes firmly require more extensive international institutions. Gov- grasped the authoritarian horn of the governance ernance arrangements to promote cooperation and dilemma. We who are unwilling to accept Hobbe’s help resolve conflict must be developed if globalization solution incur an obligation to try to explain how is not to stall or go into reverse. effective institutions that serve human interests can be Not all patterns of globalization would be beneficial. designed and maintained. We must ask the question It is easy to conjure up nightmare scenarios of a that Plato propounded more than two millenia ago: globalized world controlled by self-serving elites work- Who guards the guardians? ing to depress wages and suppress local political au- Clearly, the stakes are high: no less than peace, tonomy. So we need to engage in normative as well as prosperity, and freedom. Political science as a profes- positive analysis. To make a partially globalized world sion should accept the challenge of discovering how benign, we need not just effective governance but the well-structured institutions could enable the world to right kind of governance. have “a new birth of freedom” (Lincoln 1863). We My analysis begins with two premises. The first is need to reflect on what we, as political scientists, know that increased interdependence among human beings that could help actors in global society design and produces discord, since self-regarding actions affect the maintain institutions that would make possible the welfare of others. At worst, the effects of international good life for our descendants. In the first section of this essay I sketch what might Robert O. Keohane is James B. Duke Professor of Political Science, be called the “ideal world.” What normative standards Duke University, Perkins Library 214, Durham, NC 27708-0204 should institutions meet, and what categories should ([email protected]). we use to evaluate institutions according to those For comments on earlier versions of this address, I am grateful to Ruth Grant, Stanley Hoffmann, Peter J. Katzenstein, Nannerl O. standards? I turn next to what we know about real Keohane, and Joseph S. Nye, Jr. For insights into the issue of institutions—why they exist, how they are created and accountability, I am particularly indebted to Professor Nye. maintained, and what this implies about their actual 1 Governance in a Partially Globalized World March 2001 operation. In the concluding section I try to bring ideal objectives for which liberal democracy, or polyarchy, is and reality together to discuss institutional design. Are designed at the national level. there ways by which we can resolve the governance dilemma, using institutions to promote cooperation Consequences and create order, without succumbing to exploitation or tyranny? We can think of outcomes in terms of how global governance affects the life situations of individuals. In outlining these outcome-related objectives, I combine DESIRABLE INSTITUTIONS FOR A Amartya Sen’s concept of capabilities with Rawls’s PARTIALLY GLOBALIZED WORLD conception of justice. Sen (1999, 75) begins with the Democratic theorists emphasize that citizens should Aristotelian concept of “human functioning” and de- reflect on politics and exercise their collective will fines a person’s “capability” as “the alternative combi- (Rousseau 1978), based on what Jurgen Haber- nations of functionings that are feasible for her to mas (1996, 296) has called “a culturally established achieve.” A person’s “capability set represents the background consensus shared by the citizenry.” Gov- freedom to achieve: the alternative functioning combi- ernments derive their just powers from the consent of nations from which this person can choose” (p. 75). the governed, as the American Declaration of Inde- Governance should enhance the capability sets of the pendence proclaims, and also from their reflective people being governed, leading to enhancements in participation. their personal security, freedom to make choices, and To the potential utopianism of democratic thought I welfare as measured by such indices as the UN Human juxtapose what a former president of this association, Development Index. And it should do so in a just way, who was also my teacher, called the liberalism of fear which I think of in the terms made famous by Rawls (Shklar 1984). In the tradition of realistic liberalism, I (1971). Behind the “veil of ignorance,” not knowing believe that the people require institutional protection one’s future situation, people should regard the ar- both from self-serving elites and from their worst rangements for determining the distribution of capa- impulses, from what James Madison ( 1961) in bilities as just. As a summary set of indicators, J. Federalist 10 called the “violence of faction.” Madison Roland Pennock’s (1966) list holds up quite well: and Shklar demonstrate that liberalism need not be security, liberty, welfare, and justice. optimistic about human nature. Indeed, at the global scale the supply of rogues may be expected to expand Functions with the extent of the market. Institutional protection from the arbitrary exercise of coercion, or authoritative The world for which we need to design institutions will exploitation, will be as important at the global level as be culturally and politically so diverse that most func- at the level of the national state. tions of governance should be performed at local and The discourse theory of Habermas restates liberal national levels, on the principle familiar to students of arguments in the language of communicative ration- federalism or of the European Union’s notion of ality. Legitimacy, in this view, rests on institutionalized “subsidiarity.” Five key functions, however, should be procedures for open communication and collective handled at least to some extent by regional or global reflection. Or, as Habermas (1996, 304) quotes John institutions. Dewey ( 1954, 208), “the essential need is the The first of these functions is to limit the use of improvement of the methods and conditions of debate, large-scale violence. Warfare has been endemic in discussion, and persuasion.” The ideal that Habermas, modern world politics, and modern “total warfare” all John Rawls (1971), Robert Dahl (1976, 45– 6), and but obliterates the distinction between combatants and many other political philosophers have upheld is that noncombatants, rendering the “hard shell” of the state of rational persuasion— changing others’ minds on the permeable (Herz 1959). All plans for global gover- basis of reason, not coercion, manipulation, or material nance, from the incremental to the utopian, begin with sanctions. Persuasion in practice is much more com- the determination, in the opening words of the United plex than this ideal type, but seeking to move toward Nations Charter (1945), “to save succeeding genera- this ideal seems to me to be crucial for acceptable tions from the scourge of war.” governance in a partially globalized world. The second function is a generalization of the first. With these normative considerations in mind, we can Institutions for global governance will need to limit the ask: What political institutions would be appropriate negative externalities of decentralized action. A major for a partially globalized world? Political institutions implication of interdependence is that it provides op- are persistent and connected sets of formal and infor- portunities for actors to externalize the costs of their mal rules within which attempts at influence take place. actions onto others. Examples include “beggar thy In evaluating institutions, I am interested in their neighbor” monetary policies, air pollution by upwind consequences, functions, and procedures. On all three countries, and the harboring of transnational criminals, dimensions, it would be quixotic to expect global terrorists, or former dictators. Much international con- governance to reach the standard of modern democra- flict and discord can be interpreted as resulting from cies or polyarchies, which Dahl (1989) has analyzed so such negative externalities; much international cooper- thoroughly. Instead, we should aspire to a more loosely ation takes the form of mutual adjustment of policy to coupled system at the global level that attains the major reduce these externalities or internalize some of their 2 American Political Science Review Vol. 95, No. 1 costs (Keohane 1984). Following the convention in the 1865 is an important historical example. Yet, the international relations literature, I will refer to these expansion of concern about human rights during the situations, which resemble classic prisoners’ dilemmas, past two decades has been extraordinary, both in the as collaboration games (Martin 1992; Stein 1983). scope of rights claimed—and frequently codified in UN The third function of governance institutions is to agreements—and in the breadth of transnational advo- provide focal points in coordination games (Fearon cacy movements and coalitions promoting such rights 1998; Krasner 1991; Martin 1992; Schelling 1960). In (Keck and Sikkink 1998). Concern about poverty, situations with a clear focal point, no one has an however, has not been matched by effective action to incentive to defect. Great efficiency gains can be eliminate the source of human misery (World Bank achieved by agreeing on a single standard—for mea- 2000). surement, technical specifications, or language commu- nication. Actors may find it difficult, for distributional Procedures reasons, to reach such an agreement, but after an institutionalized solution has been found, it will be Liberal democrats are concerned not only with out- self-enforcing. comes but also with procedures. I will put forward The fourth major function of governance institutions three procedural criteria for an acceptable global gov- for a partially globalized world is to deal with system ernance system. The first is accountability: Publics need disruptions. As global networks have become tighter to have ways to hold elites accountable for their and more complex, they have generated systemic ef- actions. The second is participation: Democratic prin- fects that are often unanticipated (Jervis 1997). Exam- ciples require that some level of participation in mak- ples include the Great Depression (Kindleberger ing collective decisions be open to all competent adults 1978); global climate change; the world financial crisis in the society. The third is persuasion, facilitated by the of 1997–98, with its various panics culminating in the existence of institutionalized procedures for communi- panic of August 1998 following the Russian devalua- cation, insulated to a significant extent from the use tion; and the Melissa and Lovebug viruses that hit the and threats of force and sanctions, and sufficiently open Internet in 2000. Some of these systemic effects arise to hinder manipulation. from situations that have the structure of collaboration Our standards of accountability, participation, and games in which incentives exist for defection. In the persuasion will have to be quite minimal to be realistic future, biotechnology, genetic manipulation, and pow- in a polity of perhaps ten billion people. Because I erful technologies of which we are as yet unaware may, assume the maintenance of national societies and state like market capitalism, combine great opportunity with or state-like governance arrangements, I do not pre- systemic risk. sume that global governance will benefit from consen- The fifth major function of global governance is to sus on deep substantive principles. Global governance provide a guarantee against the worst forms of abuse, will have to be limited and somewhat shallow if it is to particularly involving violence and deprivation, so that be sustainable. Overly ambitious attempts at global people can use their capabilities for productive pur- governance would necessarily rely too much on mate- poses. Tyrants who murder their own people may need rial sanctions and coercion. The degree of consensus to be restrained or removed by outsiders. Global on principles— even procedural principles, such as inequality leads to differences in capabilities that are so those of accountability, participation, and persuasion— great as to be morally indefensible and to which would be too weak to support decisions that reach concerted international action is an appropriate re- deeply into people’s lives and the meanings that they sponse. Yet, the effects of globalization on inequality construct for themselves. The point of presenting ideal are much more complicated than they are often por- criteria is to portray a direction, not a blueprint. trayed. Whereas average per-capita income has vastly Now that these normative cards are on the table, I increased during the last forty years, cross-national turn to some of the positive contributions of political inequality in such income does not seem to have science. In the next section I ask how we can use our changed dramatically during the same period, although knowledge as political scientists to design sustainable some countries have become enormously more institutions that would perform the functions I have wealthy, and others have become poorer (Firebaugh listed. In the final section I explore how these institu- 1999). Meanwhile, inequality within countries varies tions could facilitate the democratic procedural virtues enormously. Some globalizing societies have a rela- of accountability, participation, and persuasion. These tively egalitarian income distribution, whereas in oth- issues are all part of one overriding question: How can ers it is highly unequal. Inequality seems to be complex we design institutions that would facilitate human and conditional on many features of politics and soci- functioning, in the sense of Aristotle or Sen? ety other than degree of globalization, and effective action to enhance human functioning will require INSTITUTIONAL EXISTENCE AND POWER domestic as well as international efforts. Whatever the economic effects of globalization, so- How can authoritative institutions exist at all? This is a cial globalization certainly increases the attention paid question that Rousseau ( 1978, I, 1) claimed not to events in distant places, highlighting abuses that are to know how to answer and that students of interna- widely abhorrent. Such issue advocacy is not new: The tional politics have recently debated. No student of transnational antislavery movement between 1833 and international relations is likely to forget that institu- 3 Governance in a Partially Globalized World March 2001 tions are fragile and that institutional success is prob- public good (North 1981, 68; 1990, 93). Indeed, rent- lematic. seeking coalitions have incentives to resist socially To address this issue, I begin with the contributions beneficial institutional innovations that would reduce of rational-choice institutionalism, which has insis- their own advantages (Olson 1982). tently sought to raise the question of institutional Functional solutions to the problem of institutional existence and has addressed it with the tools of equi- existence are therefore incomplete. There must be librium analysis (Shepsle 1986). To design appropriate political entrepreneurs with both the capacity and the and legitimate global institutions, we need to fashion a incentives to invest in the creation of institutions and rich version of institutionalist theory, which uses the the monitoring and enforcement of rules. Unless the power of the rationality assumption without being entrepreneurs can capture selective benefits from their hobbled by a crude psychology of material self-interest. activities, they will not create institutions. And these But before discussing such a theory, it is important to institutions will not be effective unless sufficient com- indicate briefly why a simple functional answer is not pliance is induced by a combination of material and sufficient. normative incentives. To use economic language, prob- lems of supply (Bates 1988; Shepsle and Weingast 1995) as well as demand have to be solved. The Inadequacy of Functional Theories Mancur Olson’s (1965) analysis of the logic of col- One can imagine a simple functional theory of global lective action has two major implications for the gov- institutions by which the demand for governance, gen- ernance of a globalized world. First, there is no guar- erated by globalism, creates its own supply. Such an antee that governance arrangements will be created account has the defining characteristic that the real or that will sustain high levels of globalism. As Western anticipated effects of a process play a causal role history reveals, notably in the collapse of the Roman (Cohen 1978). A functional account can only be con- Empire and in World War I, extensive social and vincing if the causal mechanism for adaptation is economic relations can be undermined by a collapse of clearly specified. In biology, one such mechanism is governance. At the global as well as national level, Darwinian evolution, which in its strong form implies political scientists need to be as concerned with de- environmental determinism. The selection environ- grees of governance as with forms of governance ment determines which individual organisms, or other (Huntington 1968, 1). units, survive. Although the individual units may un- The second implication of Olson’s insight is that we dergo random mutations, they do not act in a goal- cannot understand why institutions vary so much in directed fashion, and they do not fundamentally affect their degree of effectiveness simply by studying institu- the environment that selects them. But environmental tions. To focus only on existing institutions is to select determinism and the absence of goal-seeking behavior on the dependent variable, giving us no variance and are not assumptions that seem to fit human social and no leverage on our problem. On the contrary, we need political reality (Kahler 1999). Hence, evolutionary to explore situations in which institutions have not been arguments in the social sciences have mostly stayed at created, despite a widespread belief that if such insti- the level of metaphor. They certainly do not provide us tutions were created, they would be beneficial. Or we with warrant for a functionalist account of how gover- can compare situations in which institutions exist to nance arrangements for globalization would emerge, earlier ones in which they were absent (cf. Tilly 1975, since the causal mechanism for selection seems even 1990). weaker at the global level than with respect to compe- tition among states. The other causal mechanism for functional theories Institutional Theory and Bargaining involves rational anticipation. Agents, seeing the ex- Equilibria pected consequences of various courses of action, plan Rational-choice institutionalism in political science in- their actions and design institutions in order to maxi- sists that institutions, to persist, must reflect bargaining mize the net benefits that they receive. Ronald Coase equilibria of games in which actors seek to pursue their (1960) and Oliver Williamson (1985) use functional own interests, as they define them. This perspective, theory in Cohen’s sense to explain why firms exist at all. stated elegantly by Kenneth Shepsle (1986), is not new “Transactional economies” account for choices of mar- in its essentials. Indeed, in investigating the effects of kets or hierarchies (Williamson 1975, 248). That is, the constitutions, Aristotle held that those vesting author- more efficient organizational arrangement will some- ity in the middle class will promote rationality and the how be selected. protection of property rights (Politics IV, xi, 4 –15).1 He But there is a micro-macro problem here, since sought to explain variations in constitutional forms by arrangements most efficient for society are not neces- referring to variations in social conditions (Politics IV, sarily optimal for the leaders of the organization. At iii, 1– 6). And he argued that a stable constitution is not the level of societies, as Douglass North has pointed only one that a majority seeks to maintain but also one out, the history of real economies is one of persistent for which “there is no single section in all the state inefficiency, which he explains essentially in terms of the free rider problem. Even if an institutional innovation would increase efficiency, no one may have the incen- 1 All references to Aristotle’s Politics are to Barker’s (1948) transla- tive to develop it, since institutional innovation is a tion. 4 American Political Science Review Vol. 95, No. 1 which would favor a change to a different constitution” why do institutions not simply “inherit” rather than (IV, ix, 10). solve Arrow’s paradox (Aldrich 1993)? In the terms of rational-choice institutionalism, Ar- The general answer seems to be that institutions istotle was interested both in institutional equilibrium generate rules that resolve Arrow’s paradox, for exam- and equilibrium institutions. So were Smith (1776) and ple, by giving agenda-setting power to particular agents Madison ( 1961). The eighteenth century view, (Shepsle 1986) or by requiring supermajorities to which resonates with rational-choice institutionalism, change institutional arrangements. These rules ensure was that the “passions” of people in bourgeois society that majorities cannot alter them easily when the can be interpreted in terms of their interests (Hir- median voter’s preferences change. schman 1977, 110) and can be moderated by wise First-mover advantages and agenda control provide institutions. incentives for institutional innovation and help to Yet, rational-choice institutionalism has been more stabilize institutions. They operate somewhat differ- rigorous and more relentless than its predecessors in ently, however, in coordination and collaboration situ- insisting on explaining, by reference to incentives, why ations, as described above. In situations of coordina- institutions exist. Because rational-choice theorists tion, institutions, once accepted, are in equilibrium. seek to explain in formal terms why institutions exist, Participants do not have incentives to deviate unilater- they have to confront directly two critical questions. (1) ally from widely accepted standards for Internet con- Under what conditions will political entrepreneurs nectivity or airline traffic control. Institutions to solve have incentives to create institutions? (2) What makes collaboration problems are much more fragile. After such institutions stable? an agreement on institutions to solve collaboration Since institutions are public goods, they are likely to problems is reached, participants typically have incen- be underproduced and, at the limit, not produced at all. tives to defect if they expect to avoid retaliation from Hence there must be selective incentives for politicians others (Martin 1992). to invest in institutional innovations (Aldrich 1995). In Students of international relations have used this addition, significant advantages must accrue to institu- distinction to show how much more difficult it is to tional innovators, such as conferring on them control maintain collaboration institutions: Monitoring and over future rules or creating barriers to entry to enforcement are essential. Furthermore, during the potential competitors. Otherwise, latecomers could bargaining process “hold-outs” may be able to negoti- ate better terms for themselves in collaboration than in free ride on the accomplishments of their predecessors, coordination situations, since threats to remain outside and anticipation of such free riding would discourage collaboration-oriented institutions are more credible institutional innovation. Another barrier to entry for than threats to remain outside a widely accepted latecomers may be ideology. Insofar as only a few coordination equilibrium. In international relations, ideologies, quite distinct from one another, can exist, the side-payments negotiated by China and India to first movers would gain an advantage by seizing favor- join the ozone regime, and the refusal so far of able ideological ground (cf. Hinich and Munger 1994). developing countries to be bound by emissions restric- The implication for our problem of institutional design tions in a climate change regime, illustrate this point. is that first-mover advantages are essential if institu- If we keep our normative as well as positive lenses in tional innovation is to occur. focus, we will see that this apparent advantage of The European Union (EU) provides a compelling coordination institutions has a dark side. Initiators of example of first-mover advantages in international coordination institutions can exercise great influence organizations. New members of the EU have to accept, over the choice of focal points, thereby gaining an in their entirety, the rules already established by their enduring first-mover advantage over their rivals (Kras- predecessors. As a result, the innovators of the Euro- ner 1991). Collaboration institutions do not offer such pean Community—the six founding members— gain first-mover advantages, since participants can defect at persistent and cumulative advantages from having writ- lower cost. Collaboration institutions therefore provide ten the original rules. These rules are important. Even fewer opportunities, as compared to coordination in- if implementation is often slow, during the 1990s all stitutions, for coercion of latecomers. Real institutions members of the EU had implemented more than 75% usually combine coordination and collaboration func- of EU directives, and more than half had implemented tions, and therefore also contain a mixture of destabi- more than 85% (Martin 2000, 174). First-mover advan- lizing (or liberating) elements and stabilizing (or po- tages are also evident in the processes of writing tentially oppressive) ones. national electoral rules: Those who win an earlier Institutions, whether emphasizing coordination or election create rules that subsequently favor their collaboration, necessarily institutionalize bias, in favor party, policy positions, and personal careers (Bawn of groups that have agenda control or wish to maintain 1993; Remington and Smith 1996). the status quo. It is therefore not surprising that The second key question is that of stability. If advocates of social equality, such as Thomas Jefferson, institutional rules constrain majorities, why do these and democrats such as Rousseau, are often suspicious majorities not simply change the institutional rules to of institutions. Barriers to competition confer monop- remove the constraint? If they do, what happens to the olistic privileges and therefore create normative prob- “structure-induced equilibrium” that solves Arrow’s lems. Yet, institutions are essential for the good life. paradox of social choice (Riker 1980)? In other words, Normatively, those of us who believe in Shklar’s 5 Governance in a Partially Globalized World March 2001 (1984) “liberalism of fear” both support institutions tially unlimited number of equilibria appear in all and are cautious about them. We support them be- interesting games. When the equilibrium rabbit is to be cause we know that without well-functioning political pulled from the hat, we are as likely to get a thousand institutions, life is indeed “nasty, brutish, and short.” rabbits as one. Equilibria multiply like rabbits in Aus- But we are suspicious, since we understand how self- tralia and are about as useful. As Elinor Ostrom (1998, serving elites can use institutions to engage in theft and 4) commented in her address to this association three oppression. In a partially globalized world, we will need years ago, the assumptions of rationality, amoral self- institutions of broader scope. But as in national democ- interest, and lack of influence from social norms lead to racies, eternal vigilance will be the price of liberty. explanatory chaos: “Everything is predicted: optimal Rational-choice theory has led to fruitful inquiries outcomes, the Pareto-inferior Nash equilibria, and into the issue of why institutions exist, because it everything in between.” relentlessly questions any apparent equilibrium. The In addressing the problems of institutional design, it skeptical question—why should institutions exist at is a good thing that people are not purely egotistical. It all?— has ironically led to a deeper understanding of would be difficult to understand the creation of major institutions than has the assumption that we could take political institutions—from the U.S. Constitution to their existence for granted and focus on how they work. UN-sponsored human rights agreements—if we took egotism and the free-rider problem too seriously. We are indeed wise to assume that institutions, to be in The Limits of Rational Egoism long-term equilibrium, must be broadly consistent with Commenting on Toqueville, Albert Hirschman (1977, the self-interest of powerful actors. But we cannot 125) has pointed out a normative problem with the understand the origins of institutions if we banish emphasis on self-interest that I have thus far empha- principled action from our analytic world. sized: “Social arrangements that substitute the inter- Egoists have a hard time overcoming problems of ests for the passions as the guiding principle of human mistrust, because they know that everyone has an action for the many can have the side effect of killing incentive to disguise his or her preferences. Only costly the civic spirit.” There is also an analytical problem: signals will be credible; but the cost of signalling We know from a variety of work that this egoistic reduces the prospective value of cooperation and limits picture is seriously incomplete. the agreements that can be reached. Egoists also have Rationalist theory often carries with it the heavy difficulty solving bargaining problems, since they do not baggage of egoism. People are viewed as self-interested recognize norms of fairness that can provide focal individuals whose incentives are strictly shaped by their points for agreement. Cool practitioners of self-inter- environment, including the rules of the institutions in est, known to be such, may be less able to cooperate which they are located. The most sophisticated version productively than individuals who are governed by of this argument does not make the essentialist claim emotions that send reliable signals, such as love or that “human nature” is fundamentally egoistic but reliability (Frank 1988). In Sen’s (1977, 336) phrase, gives priority instead to an instrumentalist logic. Hans people in purely rational-choice models are “rational J. Morgenthau (1967, chap. 1), for example, argues that fools,” incapable of distinguishing among egoistic pref- since power is a necessary means to other goals in erences, sympathy, and commitments. international politics, we can analyze leaders’ behavior As Sen makes clear, rejecting the premise of egoism in terms of power even if they do not seek power for its does not imply rejecting the assumption of ration- own sake (see also Wolfers 1962, chap. 7). For ration- ality—more or less bounded in Herbert Simon’s (1985) alist students of American and comparative politics, sense. Nor does it imply altruism: People can empa- political leaders may have a multiplicity of goals, but thize with others without being self-sacrificing. What it since continuation in office is a necessary means to does is demand that norms and values be brought back achieve any of them, it can be regarded as a universal into the picture. Committed individuals, seeking policy objective of politicians, whether purely instrumental or goals as well as office for its own sake, and constrained consummatory (Geddes 1994; Mayhew 1974). by norms of fairness or even by more transcendental Thoughtful theorists of rational choice recognize values, can nevertheless calculate as rationally as the that the assumption of egoism oversimplifies social egoists of economic theory. reality. Norms of reciprocity and fairness often affect In thinking about a partially globalized world, one social behavior (Levi 1997; Ostrom 1990). The theo- might be tempted to dismiss half the governance retical predictions derived from the assumption of dilemma by pointing out that because international egoism encounter serious predictive failure in experi- institutions are very weak, they are unlikely to be mental settings (Ostrom 1998). And survey research oppressive. For example, contemporary opponents of shows that citizens evaluate the legitimacy of the legal globalization and associated international institutions system on the basis not only of their own success in have sought to portray the World Trade Organization dealing with it but also of their perceptions of its (WTO) as some sort of bureaucratic monster, although procedural fairness (Tyler 1990). my own university’s budget allocates more money in Sometimes the assumption of egoism is defended on two weeks than WTO spends in a year.2 the ground that only with such simple models can solutions be found to strategic games. But the folk 2 The budget of the WTO for 2000 is 127,697,010 Swiss francs, or theorem of game theory demonstrates that an essen- approximately $73.8 million, at September 2000 exchange rates. 6 American Political Science Review Vol. 95, No. 1 True as it is, this appeal to institutional weakness is nants of action. They depend heavily on trust, reputa- not fully convincing. The problem is not that interna- tion, and reciprocity, which depend in turn on networks tional organizations are huge and oppressive but that of civic engagement, or social capital. Building such they are seen as serving the vested interests of the networks is an incremental process. Engagement in a powerful and privileged. And they do. Indeed, they are just set of social relations helps create personal integ- institutions of the privileged, by the privileged, and all rity, which is the basis for consistent principled action too often for the privileged. There are severe restraints (Grant 1997). Networks of civic engagement are not on the powers of the international civil servants who easily divided into “international” and “domestic” but, lead these organizations, but few such checks limit the rather, cross those lines (Keck and Sikkink 1998). ability of the strongest states, such as the United States, Rational strategic action depends on the expectations to dictate policies and veto personnel. Yet, in the and incentives that these networks create. absence of such institutions, dictation by strong states Until recently, students of international politics paid would be even more direct, less encumbered by rules. too little attention to beliefs. The realists insisted on Like Churchill’s aphorism about democracy, an insti- the dominance of interests and power, which they tutionalized world is probably the worst form of gov- traced to material factors. Marxist and neoclassical ernance— except for the alternatives. political economists also relied on material forces for Ironically, it is the privileged who often appeal to their explanations. Students of institutions, such as I, altruism—their own, of course—as the guarantee sought to gain credibility by showing that our theories against the abuse of power. Political scientists have are as realistically based in interests and power as those spent too much time debunking altruism as a general of our realist adversaries, that we are not tainted by the motivating factor in politics to be detained long by such idealist brush. Ironically enough, however, the theory claims. Anyone my age has lived through the disastrous of strategic interaction on which we all rely has insis- failures of social systems, notably in Russia and China, tently argued that beliefs are crucial to understanding based on the premise that human nature can be any game-theoretic situation (Morrow 1994; Wendt remolded. The reality is that the worst people thrive 1999). under the cover of such grand visions. In any event, the The fact that strategic action depends on expecta- heterogeneity of the world’s population makes it im- tions means that understanding historical and cultural possible to imagine any single ideology providing the context is critical to any analysis of how institutions basis for a coherent, value-based system of global operate. Peter Katzenstein (1993, 1996) has used the governance. The answer to global governance prob- differing responses of Germany and Japan to military lems does not lie in revelation. defeat and economic revival to make this point in a Faced with the governance dilemma, those of us cogent and forceful way. Historical explorations of interested in governance on a world scale could retreat institutional phenomena and negotiations may draw to the pure self-interest model. With that set of as- effectively on rational-choice theory, but they must go sumptions, we would probably limit world governance. well beyond its premises to describe multidimensional We would sacrifice gains that could result from better human behavior (Bates et al. 1998). Indeed, political cooperation in order to guard against rule by undem- scientists have quite a bit to learn from international ocratic, self-serving institutions responsive, in opaque law, which studies rational strategic action in the ways, to powerful elites. If we were successful, the context of rules and rule making, deeply structured by result would be to limit global governance, even at the interests and power but also reflecting the influence of expense of greater poverty and more violent conflict. ideas on interests and on how power is exercised We might think ourselves wise, but the results would be (Grewe 2000). sad. Due to excessive fear, we would have sacrificed the A major task before our discipline is how to connect liberal vision of progress. rational strategic action with beliefs and values. In her presidential address three years ago, Ostrom (1998) linked rational-choice theory with the laboratory ex- Institutions, Expectations, and Beliefs periments of cognitive science to show that institu- It may seem that we are at an impasse. Sober reliance tional incentives, fundamental norms of trust, and the on limited institutions based on pure self-interest could practice of reciprocity (Axelrod 1984; Ostrom 1990) all lead to a low-level “equilibrium trap.” But we may be provide crucial foundations of cooperation. “At the tempted to settle for such an equilibrium rather than core of a behavioral explanation,” Ostrom (1998, 12) accept oppressive global institutions. said, “are the links between the trust that individuals There may be a way out of this impasse. That path is have in others, the investment others make in trustwor- to pay more attention than we have to expectations of thy reputations, and the probability that participants how others will behave and, therefore, to underlying will use reciprocity norms.” That is, principled values, values and beliefs. Expectations are critical determi- “congealed” in institutions, provide the basis for mean- ingful rational actions and direct such actions in ways that we can describe and explain (Riker 1980). ($.578 per Swiss franc, September 26, 2000). The total operating Robert Putnam’s Making Democracy Work exempli- expenses of Duke University for the last year available (1998 –99) were $1,989,929,000, or approximately $38.2 million per week. For the fies a productive analysis of the connections among WTO budget, see http://www.wto.org. For the Duke budget, see the values, social norms, and rational behavior. Putnam Annual Report of Duke University, 1998 –99. argues that “networks of civic engagement” produce 7 Governance in a Partially Globalized World March 2001 better government. Why does he think so? Not because rights? Probably not, unless they had compelling inter- engaged people necessarily work altruistically for the nal reasons, as tyrants, to do so. Ordinary egoists, common good but because these networks increase governing nontyrannically, would have interests in costs of defection, facilitate communication, and create mimicking the principled leaders whom they succeed. favorable expectations of others’ likely actions (Put- Furthermore, the egoists would face serious collective nam 1993, 173– 4). action problems in overturning norms of human rights: Understanding beliefs is not opposed to understand- Their counterpart egoists would have an interest in ing interests. On the contrary, interests are incompre- defending those rights in order to enhance their repu- hensible without an awareness of the beliefs that lie tation as principled agents. As a result, egoistic self- behind them. Indeed, even the financial self-interest so interest would counsel them to uphold the norms dear to political economists implies acceptance of established and even to bear some costs in order to norms that would be incomprehensible in many soci- send credible signals that they believe in the norms eties, whether those imagined by Jean-Jacques Rous- (even though, by assumption, they do not). The effect seau or studied by twentieth-century anthropologists. of former principles would persist for a while, although The values and beliefs that are dominant within a it would eventually fade. society provide the foundations for rational strategy. What this thought-experiment illustrates is a simple Even beliefs about beliefs can be as solid as any but fundamental point. Beliefs in norms and princi- material interests. As Barry O’Neill shows brilliantly in ples— even beliefs only held in the past— can pro- a book awarded the Woodrow Wilson Prize, prestige foundly affect rational action in the present. Joseph refers to “beliefs about beliefs”—whether people think Schumpeter ( 1950, 137) made the famous argu- that others hold a high opinion of someone (O’Neill ment sixty years ago that capitalism requires precapi- 1999, 193). Prestige is a “social fact,” like a dollar bill talist values: “The stock exchange is a poor substitute (Searle 1995): Although it is genuinely real, its impor- for the Holy Grail.” The facile response to his argu- tance does not lie in its material manifestation but in ment at millenium’s end could be: “Yes, but he didn’t the beliefs people hold. Both money and prestige take into account NASDAQ.” More seriously, how- matter a great deal in politics, but only insofar as ever, the varieties of capitalism in the world today, people hold beliefs about others’ beliefs. from Japanese corporatism to American legalism to To see how beliefs relate to issues of institutional Russian organized theft, make it clear that what is design, think about two possible worlds of the future. economically rational in each context varies enor- In one of them, the “normative anarchy” portrayed by mously. Schumpeter was wrong about the staying the “political realism” of the late twentieth century power of capitalism but right about the dependence of (Waltz 1979) prevails. That is, there is no consensus institutions, capitalism included, on beliefs. about principled beliefs on the basis of which gover- nance across national boundaries can take place and INSTITUTIONAL DESIGN: BRINGING transnational networks of people with similar beliefs IDEALS AND REALITY TOGETHER are virtually nonexistant. The only norm on which there is general agreement is the “antinorm” of sovereignty: I began by sketching an ideal vision—a liberal and the principle that the rulers of each state are supreme democratic vision— of how institutions should work. internally and independent from external authority On the liberal side, it includes what one might call the (Bull 1977). Since I expect self-interested agents to liberalism of progress, represented by such eighteenth- continue to dominate among politicians, I would ex- century thinkers as Smith and Madison. But it also pect, in this world, familiar patterns of modern West- includes Shklar’s liberalism of fear, which emphasizes ern international politics to persist. Rationally egoistic the potential depravities of human nature and the politicians would have few incentives to fight for prin- pathologies of human institutions and is deeply cogni- ciples of human rights, since to do so they would have zant of imperialism, totalitarianism, and the Holocaust to overcome both collective action problems and ridi- (Arendt 1958). The liberalism of fear is horrified cule from “realistic” statesmen and academics. by the atrocities of Rwanda and Bosnia, but these Now consider another world, in which certain prin- atrocities do not shake its liberalism, which was forged ciples have become generally accepted—as opposition in the searing recognition that human action can be to slavery became generally accepted in the nineteenth horrible. century and as certain human rights seem to be becom- The liberalism of progress and the liberalism of fear ing accepted now (Keck and Sikkink 1998). In this are two sides of the same coin. They both seek to world, transnational advocacy networks would be ac- understand how otherwise unattractive human pas- tive. Behavior in this world would, of course, be sions can nevertheless promote the general good. different from that in the first world. Even those who Madison is the American father of such a realistic do not subscribe to these principles would have to liberalism, but it has deep roots both in English utili- calculate the costs of acting counter to them. tarianism, going back to Hobbes, and in French Now let us go a step farther and imagine that the thought (Hirschman 1977; Keohane 1980). Neither principled innovators of the new principles, and the Madison nor Smith indulged in the more utopian value-based transnational networks, disappear, to be dreams of the liberalism of progress. Even though replaced by purely rational egoists. Would the egoists potential gains from trade, combined with advancing seek immediately to overturn these norms of human technology, make it possible for all economies to 8 American Political Science Review Vol. 95, No. 1 prosper simultaneously, the Hobbesian desire for of domestic accountability of governments to their “power after power” gets in the way. So does greed. publics. Such practices can reinforce accountability People often seek to gain distributional advantages not insofar as transparency ensures that people within the by being more productive but by gaining control of several states can make judgments about their own public policies in order to capture rents. Nevertheless, governments’ performance. mercantilist theory has been proved bankrupt, and the Nonelectoral dimensions of accountability also ex- institutions of liberal democracy have limited, although ist.3 Some international regimes seek to regulate the they have not eliminated, the success of rent-seeking. activities of firms and of governments, although they Smith and Madison would not be fully satisfied, but are weaker than their domestic counterparts, and they they would be gratified by the partially successful do not meet democratic standards as well as the “best institutionalization of their ideas. practices” domestically. Global governance, combined Together the liberalism of progress and the liberal- with modern communications technology (including ism of fear emphasize the need for institutions. Smith’s technologies for linguistic translations), can begin to liberalism calls for institutions to promote exchange; generate a public space in which some people commu- Shklar’s for institutions to control human vices and nicate with one another about public policy without those individuals among us whose vices are most regard to distance. Criticism, heard and responded to dangerous to others. For these institutions to be mor- in a public space, can help generate accountability. ally acceptable, they must rest both on humane beliefs Professional standards comprise another form of non- and substantial mutual trust. The Mafia is not better electoral accountability. than anarchy; the people who live under either find Finally, markets provide a third dimension of non- themselves impaled on one horn or the other of the electoral accountability. Since people do not bring governance dilemma. equal wealth to the marketplace, markets are not Democratic theory is even more demanding. From a democratic. But they do hold firms and other institu- democratic standpoint it is not enough to have nonop- tions with hard budget constraints accountable to their pressive institutions that enforce rules. Accountability, consumers and investors in ways that are often more participation, and persuasion are also essential. Inter- rapid and effective than electoral democracy. Advo- national institutions will probably never meet the stan- cates of principle-based change have learned to use dards of electoral accountability and participation that markets on issues as diverse as promotion of infant we expect of domestic democracies (Dahl 1999), so at formula in poor countries, environmental protection, best they will be low on a democratic scale. It is unfair and labor standards. to demand too much of them. But in the liberal- These mechanisms of accountability exist, in frag- democratic tradition that I embrace, voluntary cooper- mented ways, at the global level, but they are disartic- ation based on honest communication and rational ulated. They do not come together in a clear pathway persuasion provides the strongest guarantee of a legit- by which laws are enacted and implemented. Chains of imate process. In this section, I return to the issues of delegation are long, and some of their links are hidden accountability, participation, and persuasion that I behind a veil of secrecy. Incentives for politicians to introduced earlier. hold leaders of other governments accountable are Accountability is not necessarily electoral, so it is lacking. Publics, professional groups, and advocacy essential to explore other forms of it if we are to networks can only punish leaders inconsistently. Gov- increase accountability in global governance. Participa- ernments, nongovernmental organizations, and firms tion will probably continue to be largely local, so global that do not rely on brand names may be immune from governance implies viable forms of local self-gover- market-based sanctions. In devising acceptable institu- nance. Finally, for global governance to be legitimate, tions for global governance, accountability needs to be global institutions must facilitate persuasion rather built into the mechanisms of rule making and rule than coercion or reliance on sanctions as a means of implementation. influence. Here there seems to be considerable scope for improvement, so I will emphasize persuasion in the following discussion. Participation Individual participation is essential to democratic gov- ernance. In the past, meaningful participation has only Accountability been feasible on a face-to-face basis, as in the New The partially globalized world that I imagine would not England town meeting, and it has been argued that, “in be governed by a representative electoral democracy. its deepest and richest sense, community must always States will remain important; and one state/one vote is remain a matter of face-to-face intercourse” (Dewey not a democratic principle. National identities are 1954, 211). Yet, the costs of communication unlikely to dissolve into the sense of a larger commu- between any two points of the world no longer depend nity that is necessary for democracy to thrive. on distance, and within 50 years we can expect the Accountability, however, can be indirectly linked to forms of such communication to change in extraordi- elections without a global representative democracy. nary ways. Although it is difficult to imagine good Control by democratic states over international insti- tutions can be exerted through chains of delegation. A 3 For a very sophisticated discussion of different forms of account- complementary measure is to strengthen mechanisms ability, see Scharpf 1999, especially chapter 1. 9 Governance in a Partially Globalized World March 2001 substitutes for the multiple dimensions—verbal, visual, their own strategies as a result of new information they and tactile— by which communication occurs when receive about the strategies of others. That is, they people are close to one another, the potential of become aware that others will not behave as they had communications technology should not be underesti- previously expected. In bargaining, a quid pro quo is mated. involved; in signalling, threats and promises may be More serious barriers to global democratic partici- unilateral. pation can be found in numbers and cultural diversity. Persuasion, as I will use it, involves changing peo- Meaningful collective participation in global gover- ple’s choices of alternatives independently of their cal- nance in a world of perhaps ten billion people will culations about the strategies of other players. People surely have to occur through smaller units, but these who are persuaded, in my sense of the word, change may not need to be geographically based. In the their minds for reasons other than a recalculation of partially globalized world that I am imagining, partici- advantageous choices in light of new information about pation will occur in the first instance among people others’ behavior. They may do so because they change who can understand one another, although they may be their preferences about the underlying attributes. They dispersed around the world in “disaporic public may consider new attributes during processes of spheres,” which Arjun Appuradai (1996, 22) calls “the choice. Or they may alter their conceptions of how crucibles of a postnational political order.” attributes are linked to alternatives. Whatever the geographical quality of the units that Unlike bargaining on the basis of specific reciprocity, emerge, democratic legitimacy for such a governance persuasion must appeal to norms, principles, and val- system will depend on the democratic character of ues that are shared by participants in a conversation. these smaller units of governance. It will also depend Persuasion requires giving reasons for actions, reasons on the maintenance of sufficient autonomy and author- that go beyond assertions about power, interests, and ity for those units, if participation at this level is to resolve (Elster 1998; Risse 2000). Karl Deutsch (1953, remain meaningful. 52) argued long ago that to be susceptible to persua- sion, people “must already be inwardly divided in their Persuasion and Institutions thought,” that there must be “some contradictions, actual or implied, among their habits or values.” These Since the global institutions that I imagine do not have contradictions, sharpened by discussion, may lead to superior coercive force to that of states, the influence reflection and even attitude change. processes that they authorize will have to be legitimate. Persuasion is a major subject of study in social Legitimacy is, of course, a classic subject of political psychology (McGuire 1985; Petty and Wegener 1998). philosophy and political science (Rousseau Thousands of experiments later, the essential message 1978, book 1, chap. 1; see also Hobbes 1967, from this field is that, even in the laboratory, it is chaps. 17–18; Locke 1967, chap. 9; Weber 1978, 214). In the liberal tradition that I embrace, difficult to find strong and consistent relationships that voluntary cooperation based on honest communication explain attitude change. As William McGuire (1985, and rational persuasion provides the strongest guaran- 304) puts it, “human motivation is sufficiently complex tee of a legitimate process (Habermas 1996; Rawls so that multiple and even contradictory needs may 1971). To understand the potential for legitimate gov- underlie any act.” ernance in a partially globalized world, we need to What we do know about persuasion in politics understand how institutions can facilitate rational per- indicates that it consistently involves various degrees of suasion. How do we design institutions of governance agenda control and manipulation. Rational or open so as to increase the scope for reflection and persua- persuasion, which occurs when people change their sion, as opposed to force, material incentives, and choices of alternatives voluntarily under conditions of fraud? frank communication, is an interesting ideal type but “Persuasion” means many things to many political does not describe many major political processes. Yet, scientists. I will define it with reference to two other the ideal is important, since it is so central to the processes, bargaining and signalling. In a bargaining liberal-democratic vision of politics. Indeed, thinking situation, actors know their interests and interact re- about persuasion helps restate the central normative ciprocally to seek to realize them. In a signalling question of this address: How can institutions of gov- situation, a set of actors communicates to an audience, ernance be designed so as to increase the scope for seeking to make credible promises or threats (Hinich reflection, and therefore persuasion, as opposed to and Munger 1994). Both processes essentially involve force, material incentives, and fraud? flows of information. If successful, these flows enable If governance were exercised only by those with actors to overcome informational asymmetries (Aker- direct stakes in issues, such a question might have no lof 1970) as well as private information (Fearon 1995) answer. Actors could be expected to use their resources and therefore reach mutually beneficial solutions. Nei- and their guile to achieve their desired objectives. And ther bargaining nor signalling as such involves any the institutions would themselves “inherit” the inequal- changes in preferences over attributes, that is, over the ities prevalent in the societies that produced them, as values involved in choices. Rousseau and many successors have pointed out (Al- If targets of influence change their choices as a result drich 1993; Riker 1980). Indeed, choices of electoral of bargaining and signalling, they do so by recalculating institutions can often be traced to the policy, party, and 10 American Political Science Review Vol. 95, No. 1 personal preferences of the politicians who created experiments (Frohlich and Oppenheimer 1992). Inso- them (Bawn 1993; Remington and Smith 1996). far as the consequences and functions of institutions One feature of both democratic governance and are not seriously degraded, institutions that encourage contemporary international institutions, however, is reflection and persuasion are normatively desirable that decision making is not limited to the parties to a and should be fostered. dispute. On the contrary, actors without a direct stake in the issues under consideration may play important CONCLUSION roles, as members of the mass public in democracies and legislators often do on issues arising for decision I have asked how we can overcome the governance through voting. In general, the legalization of rules— dilemma on a global scale. That is, how can we gain domestically, and more recently in international poli- benefits from institutions without becoming their vic- tics—requires the formation of durable rules that apply tims? How can we help design institutions for a par- to classes of cases and puts interpretation and rule- tially globalized world that perform valuable functions application into the hands of third parties, whose while respecting democratic values? And how can we authority depends on maintaining a reputation for foster beliefs that maintain benign institutions? My impartiality (Goldstein et al. 2000). Legalization also answers are drawn, mostly implicitly, from various increases the role of precedents. Precedents matter, schools of work in political science. not because loopholes are impossible to find or be- From rational-choice institutionalism, we learn both cause they cannot be overruled, but because the status the value of institutions and the need for incentives for quo will prevail in the absence of a decision to overturn institutional innovation. These incentives imply privi- it. leges for the elite, which have troubling implications Some third parties will have calculable interests that for popular control. closely parallel those of the principal disputants or From a variety of perspectives, including game the- advocates. Others may have strongly held beliefs that ory, the study of political culture, and work on the role determine their positions. Some may accept side-pay- that ideas play in politics, we learn how important ments or succumb to coercive pressure. But still others beliefs are in reaching equilibrium solutions, and how may lack both intense beliefs and direct stakes in the institutionalized beliefs structure situations of political outcome. Legal requirements or internalized norma- choice. tive standards may inhibit them from accepting induce- From traditional political theory, we are reminded of ments for their votes. Even more important, uncer- the importance of normative beliefs for the practice of tainty about the effect of future rules may make it politics—and for institutions. It is not sufficient to difficult for them to calculate their own interests. Rule create institutions that are effective; they must be makers face a peculiar form of “winner’s curse”: They accompanied by beliefs that respect and foster human risk constructing durable rules that suit them in the freedom. immediate instance but will operate against their inter- From historical institutionalism and political sociol- ests in the unknown future. ogy, we understand how values and norms operate in Insofar as uncertainty is high, actors face a situation society. Without such understanding, we can neither similar to one covered by a “veil of ignorance” (Rawls comprehend the varying expectations on which people 1971). In game-theoretic terms, the actors may still rationally act nor design institutions based on norma- have preferences over outcomes, but these preferences tive views. We abdicate our responsibility if we simply over outcomes do not directly imply preferences over assume material self-interest, as economists are wont strategies, since actors do not know their future situa- to do. tions. In experiments, introducing a veil of ignorance in From democratic theory, we discover the crucial prisoners’ dilemma games without communication in- roles of accountability, participation, and especially duces a dramatic increase in the willingness of subjects persuasion in creating legitimate political institutions. to cooperate (Frohlich and Oppenheimer 1996). It is These lessons are in tension with one another. reasonable to hypothesize that under conditions of Institutional stability is often at odds with innovation uncertainty in the real world, the chain of “inheritabil- and may conflict with accountability. Protection against ity” will be broken, and actors’ preferences about oppression can conflict with energetic governance; a future outcomes will not dictate their choices of alter- practical reliance on self-interest can conflict with the natives in the present. desire to expand the role of persuasion and reflection. Under conditions of authority for impartial third Governance, however, is about reconciling tensions; it parties, or high uncertainty about future interests, is Max Weber’s ( 1946) “boring of hard boards.” opportunities for persuasion are likely to appear, even As students of political philosophy, our objective if everyone is a rational egoist. Egoists have a long- should be to help our students, colleagues, and the term interest in rules that will correspond to an accept- broader public understand both the necessity for gov- able general principle, since they may be subject to ernance in a partially globalized world and the princi- these rules in the future. Various principles could be ples that would make such governance legitimate. As chosen— expected utility maximization, the maximum positive political scientists, we need to continue to principle (Rawls 1971, 152), minimax regret (Riker analyze the conditions under which different forms and 1996), or utility maximization subject to a floor mini- levels of governance are feasible. As practitioners of a mum, which is the prevailing choice in laboratory policy science, we need to offer advice about how 11 Governance in a Partially Globalized World March 2001 institutions for global governance should be consti- Elster, Jon. 1998. “Deliberation and Constitution Making.” In De- tuted. This advice must be realistic, not romantic. We liberative Democracy, ed. Jon Elster. Cambridge: Cambridge Uni- versity Press. Pp. 97–122. must begin with real people, not some mythological Fearon, James D. 1995. “Rationalist Explanations for War.” Inter- beings of higher moral capability. But we need also to national Organization 49 (Summer): 379 – 414. recognize, and seek to expand, the scope for reflection Fearon, James D. 1998. “Bargaining, Enforcement and International and the normative principles that reflective individuals Cooperation.” International Organization 52 (Spring): 269 –305. may espouse. We should seek to design institutions so Firebaugh, Glen. 1999. “Empirics of World Income Inequality.” American Journal of Sociology 104 (May): 1597– 631. that persuasion, not merely interests and bargaining, Frank, Robert H. 1988. Passions within Reason: The Strategic Role of plays an important role. the Emotions. New York: W. W. Norton. The stakes in the mission I propose are high, for the Frohlich, Norman, and Joe A. Oppenheimer. 1992. Choosing Justice: world and for political science. If global institutions are An Experimental Approach to Ethical Theory. Berkeley: University designed well, they will promote human welfare. But if of California Press. Frohlich, Norman, and Joe A. Oppenheimer. 1996. “Experiencing we bungle the job, the results could be disastrous. Impartiality to Invoke Fairness in the n-PD: Some Experimental Either oppression or ineptitude would likely lead to Results.” Public Choice 86 (1–2): 117–35. conflict and a renewed fragmentation of global politics. Geddes, Barbara. 1994. Politician’s Dilemma: Building State Capacity Effective and humane global governance arrangements in Latin America. Berkeley: University of California Press. Goldstein, Judith, Miles Kahler, Robert O. Keohane, and Anne- are not inevitable. They will depend on human effort Marie Slaughter. 2000. Legalization and World Politics. Special and on deep thinking about politics. issue of International Organization, vol. 54, no. 3 (Summer). As we face globalization, our challenge resembles Grant, Ruth W. 1997. Hypocrisy and Integrity: Machiavelli, Rousseau, that of the founders of this country: how to design and the Ethics of Politics. Chicago: University of Chicago Press. working institutions for a polity of unprecedented size Grewe, Wilhelm G. 2000. The Epochs of International Law, trans. Michael Byers. The Hague: De Gruyter. and diversity. Only if we rise to that challenge will we Habermas, Jurgen. 1996. Between Facts and Norms: Contributions to be doing our part to ensure Lincoln’s “rebirth of a Discourse Theory of Law and Democracy. Cambridge: MIT Press. freedom” on a world—and human—scale. Held, David, Anthony McGrew, David Goldblatt, and Jonathan Perraton. 1999. Global Transformations: Politics, Economics and Culture. Stanford, CA: Stanford University Press. Herz, John H. 1959. International Politics in the Atomic Age. New REFERENCES York: Columbia University Press. Akerlof, George A. 1970. “The Market for Lemons.” Quarterly Hinich, Melvin J., and Michael Munger. 1994. Ideology and the Journal of Economics 84 (August): 488 –500. Theory of Political Choice. Ann Arbor: University of Michigan Aldrich, John. 1993. “On William Riker’s ‘Inheritability’ Problem: Press. Preferences, Institutions, and Context.” Paper presented at the Hirschman, Albert O. 1977. The Passions and the Interests: Political Southern Political Science Association, Savannah, Georgia. Arguments for Capitalism Before its Triumph. Princeton, NJ: Prince- Aldrich, John. 1995. Why Parties? The Origin and Transformation of ton University Press. Party Politics in America. Chicago: University of Chicago Press. Hobbes, Thomas. 1967. Leviathan: or the Matter, Forme and Appuradai, Arjun. 1996. Modernity at Large. Minneapolis: University Power of a Commonwealth, Ecclesisticall and Civil, ed. Michael of Minnesota Press. Oakeshott. Oxford: Basil Blackwell. Arendt, Hannah. 1951/1958. The Origins of Totalitarianism. Cleve- Huntington, Samuel P. 1968. Political Order in Changing Societies. land: World Publishing. New Haven, CT: Yale University Press. Axelrod, Robert M. 1984. The Evolution of Cooperation. New York: Jervis, Robert. 1997. System Effects: Complexity in Political and Social Basic Books. Life. Princeton, NJ: Princeton University Press. Barker, Ernest. 1948. The Politics of Aristotle, trans. Ernest Barker. Kahler, Miles. 1999. “Evolution, Choice and International Change.” Oxford: Clarendon. In Strategic Choice and International Relations, ed. David A. Lake Bates, Robert H. 1988. “Contra Contractarianism: Some Reflections and Robert Powell. Princeton, NJ: Princeton University Press. Pp. on the New Institutionalism.” Politics and Society 16 (June- 165–96. September): 387– 401. Katzenstein, Peter J. 1993. “Coping with Terrorism: Norms and Bates, Robert H., Avner Grief, Margaret Levi, Jean-Laurent Internal Security in Germany and Japan.” In Ideas and Foreign Rosenthal, and Barry R. Weingast. 1998. Analytic Narratives. Policy, ed. Judith Goldstein and Robert O. Keohane. Ithaca, NY: Princeton, NJ: Princeton University Press. Cornell University Press, Pp. 265–95. Bawn, Kathleen. 1993. “The Logic of Institutional Preferences: Katzenstein, Peter J., ed. 1996. The Culture of National Security: German Electoral Law as a Social Choice Outcome.” American Norms and Identity in World Politics. New York: Columbia Uni- Journal of Political Science 37 (November): 965– 89. versity Press. Bull, Hedley. 1977. The Anarchical Society: A Study of Order in World Keck, Margaret, and Kathryn Sikkink. 1998. Activists beyond Borders: Politics. New York: Columbia University Press. Advocacy Networks in International Politics. Ithaca, NY: Cornell Coase, Ronald H. 1960. “The Problem of Social Cost.” Journal of University Press. Law and Economics 3 (October): 1– 44. Keohane, Nannerl O. 1980. Philosophy and the State in France: The Cohen, G. A. 1978. Karl Marx’s Theory of History: A Defense. Renaissance to the Enlightenment. Princeton, NJ: Princeton Uni- Princeton, NJ: Princeton University Press. versity Press. Dahl, Robert A. 1976. Modern Political Analysis. 3rd ed. Englewood Keohane, Robert O. 1984. After Hegemony: Cooperation and Discord Cliffs, NJ: Prentice-Hall. in the World Political Economy. Princeton, NJ: Princeton Univer- Dahl, Robert A. 1989. Democracy and Its Critics. New Haven, CT: sity Press. Yale University Press. Keohane, Robert O., and Joseph S. Nye, Jr. 2001. Power and Dahl, Robert A. 1999. “Can International Organizations Be Demo- Interdependence. 3rd ed. New York: Addison-Wesley. cratic?” In Democracy’s Edges, ed. Ian Shapiro and Casiano Kindleberger, Charles P. 1973. The World in Depression, 1929 –1939. Hacker-Cordon. Cambridge: Cambridge University Press. Pp. Berkeley: University of California Press. 19 –36. Krasner, Stephen D. 1991. “Global Communications and National Deutsch, Karl W. 1953. Nationalism and Social Communication. Power: Life on the Pareto Frontier.” World Politics 43 (April): Cambridge: MIT Press. 336 – 66. Dewey, John. 1954. The Public and Its Problems. Chicago: Levi, Margaret. 1997. Consent, Dissent and Patriotism. Cambridge: Swallow Press. Cambridge University Press. 12 American Political Science Review Vol. 95, No. 1 Lincoln, Abraham. 1863. The Gettysburg Address (November 19, Risse, Thomas. 2000. “ ‘Let’s Argue!’ Communicative Action in 1863). World Politics.” International Organization 54 (Winter): 1– 40. Locke, John. 1967. Second Treatise of Government, ed. Peter Rousseau, Jean-Jacques. 1978. On the Social Contract, ed. Laslett. Cambridge: Cambridge University Press. Roger D. Masters. New York: St. Martin’s. Madison, James. 1961. Federalist No. 10. In The Federalist Scharpf, Fritz. 1999. Governing in Europe. Oxford: Oxford University Papers, by Alexander Hamilton, John Jay, and James Madison, ed. Press. Jacob E. Cooke. Middletown, CT: Wesleyan University Press. Pp. Schelling, Thomas C. 1960. The Strategy of Conflict. Cambridge, MA: 56 – 65. Harvard University Press. Martin, Lisa M. 1992. “Interests, Power and Multilateralism.” Inter- Schumpeter, Joseph A. 1950. Capitalism, Socialism and De- national Organization 46 (Autumn): 765–92. mocracy. 3d ed. New York: Harper and Row. Martin, Lisa M. 2000. Democratic Commitments: Legislatures and Searle, John. 1995. The Construction of Social Reality. New York: International Cooperation. Princeton, NJ: Princeton University Free Press. Press. Sen, Amartya K. 1977. “Rational Fools: A Critique of the Behavioral Mayhew, David. 1974. Congress: The Electoral Connection. New Foundations of Economic Theory.” Philosophy and Public Affairs 6 Haven, CT: Yale University Press. (Summer): 317– 44. McGuire, William J. 1985. “Attitudes and Attitude Change.” In Sen, Amartya K. 1999. Development as Freedom. New York: Knopf. Handbook of Social Psychology, 3rd ed., ed. Gardner Lindzey and Shepsle, Kenneth A. 1986. “Institutional Equilibrium and Equilib- Elliott Aronson. New York: Random House. Pp. 233–346. rium Institutions.” In Political Science: The Science of Politics, ed. Morgenthau, Hans J. 1967. Politics among Nations: The Struggle for Herbert F. Weisberg. New York: Agathon. Pp. 51– 81. Power and Peace. New York: Knopf. Shepsle, Kenneth A., and Barry R. Weingast. 1995. Positive Theories Morrow, James D. 1994. Game Theory for Political Scientists. Prince- of Congressional Institutions. Ann Arbor: University of Michigan ton, NJ: Princeton University Press. Press. North, Douglass. 1981. Structure and Change in Economic History. Shklar, Judith N. 1984. Ordinary Vices. Cambridge, MA: Harvard New York: Norton. University Press. North, Douglass. 1990. Institutions, Institutional Change and Eco- Simon, Herbert A. 1985. “Human Nature in Politics: The Dialogue nomic Performance. New York: Cambridge University Press. of Psychology with Political Science.” American Political Science Olson, Mancur, Jr. 1965. The Logic of Collective Action. Cambridge, Review 79 (June): 293–304. MA: Harvard University Press. Smith, Adam. 1976. The Wealth of Nations. Chicago: Univer- Olson, Mancur, Jr. 1982. The Rise and Decline of Nations: Economic sity of Chicago Press. Growth, Stagflation and Social Rigidities. New Haven, CT: Yale Stein, Arthur A. 1983. “Coordination and Collaboration: Regimes in University Press. an Anarchic World.” In International Regimes, ed Stephen D. O’Neill, Barry. 1999. Honor, Symbols and War. Ann Arbor: Univer- Krasner. Ithaca, NY: Cornell University Press. Pp. 115– 40. sity of Michigan Press. Tilly, Charles. 1975. “Reflections on the History of European Ostrom, Elinor. 1990. Governing the Commons: The Evolution of State-Making.” In The Formation of National States in Western Institutions for Collective Action. New York: Cambridge University Europe. Princeton, NJ: Princeton University Press. Pp. 3– 83. Press. Tilly, Charles. 1990. Coercion, Capital and European States AD Ostrom, Elinor. 1998. “A Behavioral Approach to the Rational 990 –1990. Oxford: Basil Blackwell. Choice Theory of Collective Action.” American Political Science Tyler, Tom R. 1990. Why People Obey the Law. New Haven, CT: Yale Review 92 (March): 1–22. University Press. Pennock, J. Roland. 1966. “Political Development, Political Systems Waltz, Kenneth N. 1979. Theory of International Politics. Reading, and Political Goods.” World Politics 18 (April): 415–34. MA: Addison-Wesley. Petty, Richard E., and Duane T. Wegener. 1998. “Attitude Change: Weber, Max. 1946. “Politics as a Vocation.” In From Max Multiple Roles for Persuasion Variables.” In Handbook of Social Weber: Essays in Sociology, ed. H. H. Gerth and C. Wright Mills. Psychology, 4th ed. vol. 1, ed. Daniel T. Gilbert, Susan T. Risk, and Oxford, Oxford University Press. Pp. 77–128. Gardner Lindzey. Boston: McGraw-Hill. Pp. 323–90. Weber, Max. 1978. Economy and Society, ed. Guenther Roth Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions and Claus Wittich. Berkeley: University of California Press. in Modern Italy. Princeton, NJ: Princeton University Press. Wendt, Alexander E. 1999. Social Theory of International Politics. Rawls, John. 1971. A Theory of Justice. Cambridge, MA: Harvard Cambridge: Cambridge University Press. University Press. Williamson, Oliver. 1975. Markets and Hierarchies. New York: Free Remington, Thomas F., and Steven S. Smith. 1996. “Political Goals, Press. Institutional Context, and the Choice of an Electoral System: The Williamson, Oliver. 1985. The Economic Institutions of Capitalism. Russian Parliamentary Election Law.” American Journal of Politi- New York: Free Press. cal Science 40 (November): 1253–79. Wolfers, Arnold. 1962. Discord and Collaboration. Baltimore, MD: Riker, William H. 1980. “Implications from the Disequilibrium of Johns Hopkins University Press. Majority Rule for the Study of Institutions.” American Political World Bank. 2000. Entering the 21st Century: World Development Science Review 74 (June): 432– 46. Report 1999 –2000. Oxford: Oxford University Press. 13