This file is also available in Adobe Acrobat PDF format
PROBLEMS OF INSTITUTIONAL ANALYSIS
INSTITUTIONS ARE the foundation of social life. They consist of formal and informal rules, monitoring and enforcement mechanisms, and systems of meaning that define the context within which individuals, corporations, labor unions, nation-states, and other organizations operate and interact with each other. Institutions are settlements born from struggle and bargaining. They reflect the resources and power of those who made them and, in turn, affect the distribution of resources and power in society. Once created, institutions are powerful external forces that help determine how people make sense of their world and act in it. They channel and regulate conflict and thus ensure stability in society.
Without stable institutions, life becomes chaotic and arduous, as people learned, for example, following the demise of the communist regimes in Eastern Europe after 1989. The sudden weakening of old political constitutions, property rights, law enforcement, and other institutions generated tremendous confusion as these countries began to move toward capitalism and democracy. National politics were inundated suddenly with dozens of political parties and interest groups vying for power in unstable coalitions. Economies were besieged with all sorts of mafia, scam artists, and other shady characters that made commerce unpredictable and sometimes dangerous. Even a simple thing like taking a train from one city to another was suddenly fraught with possible attacks by bandits and uncertainty about schedules. As a result, building new political and economic institutions was one of the most urgent tasks for the postcommunist countries (Elster et al. 1998). This should not be surprising. After all, Max Weber, Karl Marx, Emile Durkheim, Karl Polanyi, and many other scholars have shown that capitalism itself is impossible without a solid institutional base.1 This is one reason why institutional analysis has become such an important perspective in the social sciences in North America and Europe.
As this example suggests, institutions change. Sometimes they change radically. Sometimes they evolve more incrementally. Sometimes they are quite stable. This book is about institutional change, how to recognize different types of institutional change when we see them, and how better to understand some of the forces that make institutional change happen or not. As we shall see, scholars from various intellectual and social science backgrounds share common conceptual, theoretical, and methodological problems when it comes to these issues. Indeed, institutional analysis has been criticized repeatedly for having an inadequate understanding of institutional change (e.g., Lieberman 2002; Scott 2001, 181; Thelen and Steinmo 1992, 15). Among other things, we sometimes have trouble recognizing institutional change for what it is, mistaking evolutionary developments for more revolutionary shifts and failing to appreciate fully how much continuity there often is during episodes of institutional change, even across big events like those that occurred in postcommunist Europe after 1989. As such, this book is intended to confront and help rectify some of the most important analytic problems facing institutional analysis today. In doing so, it also offers a new theory of institutional change that combines insights from several versions of institutional analysis.
This book also confronts, however, a more concrete problem: the phenomenon of globalization. Much has been written about globalization and much of this writing suggests that the forces of globalization have triggered what amounts to a revolutionary change in national political and economic institutions.2 Some scholars intimate that this threatens the utility of institutional analysis for understanding today's world. I disagree. Much of this literature has overestimated the extent of institutional change associated with globalization. Why? Because much of it either has failed to take institutional analysis into account, or has failed to elaborate institutional analysis along the lines suggested in this book.
Broadly speaking, then, this book does two things. First, it identifies serious problems that are central to institutional analysis and offers some conceptual, methodological, and theoretical tools for solving them--tools that help improve the analytic power of institutional analysis, especially for understanding change. Second, to demonstrate the utility of these tools, it applies them to the phenomenon of globalization to show how they can help us improve our understanding--and reduce our misunderstanding--of how phenomena like this produce institutional change, or not, in the world around us.
But what do I mean by institutional analysis? Today there are three major versions of institutional analysis: rational choice, organizational, and historical institutionalism.3 To be precise, these are often referred to as new forms of institutional analysis, or neoinstitutionalism, because each one descends from an old version of institutional analysis. Each one also cuts across sociology, political science, and economics. Their utility is evident from the extraordinarily wide range of things that they have been used to explain: the development of art museums and other cultural organizations, the rise of modern corporations, the organization of labor unions, the adoption of affirmative action policies, the diffusion of business management models, the progression of forms of corporate governance, the rise of health care systems, the creation of markets, the transformation of property rights, the growth of economies and nation-states, the making of economic, fiscal, social, technology, and other public policies, the growth of social movements, the ability of groups to preserve natural resources, the variation that occurs in national economic performance and national responses to economic crisis, the operation of national and international security regimes, the diffusion of citizenship rights, the dissemination of world culture, the effects of globalization, and much more.4
Despite the important insights of the three new institutionalist paradigms, there has been much bickering among their supporters as to which is the best approach and what each one's strengths and weaknesses are (e.g., DiMaggio and Powell 1991; Kiser and Hechter 1991; Thelen and Steinmo 1992). Indeed, some of us even disagree about what institutions are in the first place. For instance, some rational choice institutionalists define institutions as a strategic equilibrium. By strategic equilibrium they mean a situation where no persons would unilaterally choose to alter their current behavior given the available alternatives and given their expectations about how others might respond if they began to behave differently (Bates et al. 1998, 8-10; Calvert 1998, 57-60). Other rational choice institutionalists view institutions as sets of formal and informal rules and the monitoring and sanctioning mechanisms that cause actors to comply with these rules (North 1990, 3). Similarly, historical institutionalists take institutions to be the formal and informal rules and procedures that structure conduct (Thelen and Steinmo 1992, 2). Finally, although organizational institutionalists recognize that institutions include formal rules, many of them see institutions as informal, common, and taken-for-granted cultural frameworks, scripts, and cognitive schema (Jepperson 1991).5
Regardless of these and other disagreements, beginning in the late 1990s scholars began calling for a second movement in institutional analysis, that is, a more constructive dialogue that explores the ways in which these paradigms might complement and connect to each other (e.g., Campbell and Pedersen 2001b; Hall and Taylor 1996; Immergut 1998; Peters 1999, 149-51; Suchman 1997; Thelen 1999). The hope was that such a conversation might lead to a reconciliation and perhaps a new and more unified approach for studying institutions.
The premise of this book is that an important part of a second movement in institutional analysis involves recognizing that these three paradigms share a common set of problems that need to be resolved. By recognizing these problems, institutionalists should see that they have much more in common and at stake together than they thought previously. Moreover, for reasons discussed later, unless these problems are resolved, institutional analysis is not likely to advance much beyond its current stage of development. At least three important analytic problems require attention: the problems of change, mechanisms, and ideas. Globalization constitutes a fourth problem, but of a much different order.6 This book examines each one of these problems and suggests how we might begin to rectify them. To my knowledge, although many scholars have compared the strengths and weaknesses of one institutionalist paradigm to another, virtually no one has examined in detail the common problems that all three paradigms share. Nor have they tried to offer solutions for them. Let me elaborate briefly on each of these problems and how the book is organized around them.
First, institutionalists debate about the best way to describe institutional change. For instance, some of us argue that it tends to follow an evolutionary pattern characterized by the gradual accumulation of small, incremental changes over long periods of time. Others say that it conforms more to patterns of either punctuated equilibrium or punctuated evolution. These patterns are more revolutionary than evolutionary in the sense that they are characterized initially by prolonged periods of either equilibrium and stability or evolution that are interrupted suddenly by a crisis that throws things into turmoil until a radically new set of institutional arrangements is established, which then remain in equilibrium or evolve slowly for another long period of time (Blyth 2002, chap. 1; Hay 2001; Krasner 1984). No matter what our position is on the issue, we often fail to explain adequately how to distinguish among these and other patterns of change. That is, we fail to specify one of the most important dependent variables with which institutionalists are concerned. As a result, it becomes difficult to establish the degree to which a given episode of change is actually evolutionary, revolutionary, or something else. This has been true even with respect to some of the most important institutional changes in the twentieth century. For instance, there has been much debate about the degree to which institutional change in postcommunist Eastern Europe during the 1990s constituted a sharp, revolutionary break with the communist era, or something more evolutionary that still bore a strong resemblance to past institutions (e.g., Campbell and Pedersen 1996; Crawford and Lijphart 1995; Hanley et al. 2002; Stark 1996). Chapter 2 explores what institutionalists mean by institutional change and how we might better recognize different patterns of change when they occur. Resolution of this problem is critical if we want to adjudicate among competing claims about the nature of institutional change in specific empirical cases, including globalization, and, in turn, if we want to test theories that seek to account for different patterns of change.
Second, institutionalists rely on causal concepts but often without clearly specifying the underlying mechanisms or processes by which change occurs. As a result, these concepts resemble mysterious black boxes whose contents need to be unpacked and examined. Unless we do this, vague concepts end up carrying much of the argument when, in fact, mechanisms should be doing the work. This is an important problem in the social sciences in general (Hedström and Swedberg 1998a; McCloskey 1985) and especially for institutionalists insofar as critics charge that the neglect of mechanisms undermines the empirical and theoretical credibility of institutionalists' arguments (Hirsch and Lounsbury 1997; Hirsch 1997; Knight 2001). It also undermines the credibility of arguments about globalization. Chapter 3 discusses how we might think more clearly about the mechanisms underlying some of the most important causal concepts that institutionalists use. In particular, it focuses on the concepts of path dependence and diffusion. It also addresses, albeit in less detail, other concepts, such as how institutions enable, empower, constitute, and constrain action. Resolution of this problem is critical if we want to develop tighter and more convincing causal arguments about institutional change.
Third, many institutionalists have taken a cultural turn recognizing recently that ideas, broadly construed, as well as self-interests drive change (Campbell 2002). Whether norms, intellectual paradigms, policy frames, world views, and other types of ideas affect institutional change are issues that have spawned new research agendas and much debate among institutionalists working in comparative politics, international relations, political economy, and political, economic, and organizational sociology.7 Many of us, however, have been hobbled in these efforts because our work on ideas suffers from conceptual and methodological problems (Blyth 1997; Finnemore 1996; Yee 1996). Notably, what we mean by ideas and what the mechanisms are by which ideas affect behavior are often confusing. And it is not always clear in the first place what the difference is between ideas and interests. Chapter 4 clarifies how we might better think about all this, particularly how we might better study the effects that ideas have on political, economic, organizational, and other types of behavior. Insofar as this chapter discusses the ideational mechanisms associated with institutional change, it also addresses some of the issues raised in chapter 3. Resolution of this problem is critical if we want to determine how ideas--such as norms, paradigms, and frames--as well as self-interests affect institutional change, and the degree to which they do this.
By grappling with the problems of change, mechanisms, and ideas, this volume offers conceptual, theoretical, and methodological insights that can help institutionalists better understand important social phenomena in the world around us. To demonstrate this, I apply these insights in chapter 5 to an analysis of globalization. Why pick globalization rather than some other empirical phenomenon? Many institutionalists have argued that differences in national institutions are an important cause of variation across countries in political and economic performance. Globalization is a fourth problem for institutional analysis insofar as some people now claim that the forces of globalization, notably increased international capital mobility and trade, are leveling these institutional differences across countries. As a result, they intimate that much institutional analysis is becoming irrelevant for understanding societies during the late twentieth and early twenty-first centuries (e.g., Giddens 2000; Guéhenno 1995).8 If we accept this argument, then our desire to improve institutional analysis will likely diminish and scholars will turn to alternative analytic approaches. Utilizing the lessons from chapters 2, 3, and 4, chapter 5 shows that claims about the leveling effects of globalization are mistaken and that institutional analysis can help us understand why this is so. Resolution of this problem through the application of the analytic tools developed earlier in the book is critical if we want to demonstrate that institutional analysis is still relevant for understanding how societies are changing during the globalization era.
Two clarifications are necessary about the problem of globalization. First, this problem is of a fundamentally different order than the other three. The problems of change, mechanisms, and ideas are internal to the logic, conceptual apparatus, and causal arguments of institutional analysis. The problem of globalization is external to these arguments in the sense that a change in the empirical reality of the world, that is, increased capital mobility and trade, appears to some observers to pose a threat to the relevance of institutional analysis per se for understanding societies nowadays. Second, although the analysis of globalization in chapter 5 develops insights and arguments specific to the debates about globalization, its primary purpose is to illustrate how the concepts, methods, and arguments of previous chapters can be used fruitfully in doing empirical research on important real-world problems. It is not intended to be a full-blown or exhaustive critique of the globalization thesis.
While chapter 5 demonstrates the utility of my arguments for empirical research, chapter 6 weaves them together into a theory of institutional change that integrates insights from all three institutionalist paradigms thereby advancing the second movement--that advocating a sustained three-way dialogue--in institutional analysis. In brief, this theory suggests how institutional change is a process of constrained innovation. By this I mean, on the one hand, that institutions tend to constrain the range of options from which actors are likely to choose as they engage in institutional innovation. But, on the other hand, institutions also provide principles, practices, and opportunities that actors use creatively as they innovate within these constraints. I present this theory as a series of analytic propositions about the antecedents that initially trigger episodes of institutional change and the complex processes by which it unfolds. This is a deliberate attempt to advance the second movement in institutional analysis. Based on these propositions, I also offer some suggestions about how institutionalists can fruitfully orient their work in the future.
Although the problems of change, mechanisms, ideas, and globalization are the central focus of this book, it will become evident that each one involves a variety of secondary problems that also occasionally require attention. Briefly, the problem of change requires that we think about the appropriate time frames we use in analysis and the important institutional dimensions we choose to track over time. The problem of mechanisms requires that we confront the relationship between social structure and agency and that we acknowledge the relative advantages and disadvantages associated with different levels of analysis in our research. The problem of ideas requires that we consider carefully how we should define ideas, identify the actors who use them, specify the causal mechanisms that involve them, and develop the appropriate methodologies for studying their impacts. The problem of globalization requires that we consider all of these things carefully in studying how international economic pressures may affect national political and economic institutions. The secondary problems underlying the four primary problems around which this book is organized will not all receive equal treatment. This is not because one is necessarily any more or less important than another, but because I deal with them on an ad hoc basis and only insofar as is necessary to address the four primary problems with which the book is concerned.
In sum, this book makes six important contributions. It focuses attention on some of the most pressing theoretical and methodological problems that threaten to block the further development of institutional analysis. It suggests solutions for how these problems might be resolved. It demonstrates the utility of these solutions for better understanding how institutional change happens by applying them to the phenomenon of globalization. It offers a theory of institutional change. And in doing so it contributes to the second movement in institutional analysis by raising common issues and questions about which proponents of the three institutionalist paradigms need to think carefully in order to advance their collective enterprise. To illustrate my arguments, I give throughout the book a variety of empirical examples of policy-making and institution building in the United States, Europe, and Asia. Many of these examples are historical or cross-national and are drawn from comparative politics and political economy, economic and political sociology, organizational sociology, and, to a lesser extent, international relations. As a result, this book also offers a guide for students and scholars interested in some of the most exciting debates and empirical work being done today in institutional analysis.
THE RISE OF INSTITUTIONAL ANALYSIS AND ITS PROBLEMS
How extensive are the problems that I have identified in the new institutional analysis? This will become clear in subsequent chapters. But to begin to answer this question it is helpful to understand how each paradigm developed and what it stands for intellectually. Institutional analysis has deep roots in sociology, political science, and economics. My intent here is not to review this history in great detail or to develop an exhaustive comparison of all the important similarities and differences among these paradigms. This has been done at length elsewhere.9 Nor do I intend to elaborate the complexities of each problem here. That will be done systematically in later chapters. Instead, my purpose in the remainder of this chapter is to provide a brief thumbnail sketch of each paradigm and its origins. Because much has been written about the differences among paradigms, I discuss them only briefly.10 Instead, my primary goal is to show that all three paradigms tend to focus on similar patterns of institutional change, rely on similar causal concepts, and recognize the need for a better understanding of how ideas--such as norms, values, and cognitive structures--as opposed to self-interests affect behavior in general and institutional change in particular. Insofar as all three paradigms share common ground in this regard it should not be surprising that they also share common problems. Because, as noted above, the problem of globalization is of a fundamentally different order than the problems of change, mechanisms, and ideas, I will not discuss it in my review of the three paradigms.
The discussion is summarized in table 1.1, which represents each paradigm as an ideal type. Keep in mind, however, that each paradigm consists of a vast literature. Although I will generalize about each one, there are always exceptions to these generalizations and considerable variation within each paradigm.11 Occasionally, there are also scholars whose work cuts across paradigms (e.g., Campbell 1998; Clemens 1997; Keck and Sikkink 1998). Hence, although sufficient for present purposes, the following discussion is intended to reveal only some of the broad contours of each paradigm. And again, it tends to emphasize similarities rather than differences among paradigms.
Rational Choice Institutionalism
In economics, institutional analysis emerged during the late nineteenth and early twentieth centuries. It was based on the insights particularly of Max Weber, but also Karl Marx and other nineteenth-century scholars. Among the most influential institutional economists were John R. Commons, Thorstein Veblen, and Wesley C. Mitchell, all of whom shared a devotion to empirically based research, an interest in the changing nature of economic institutions, habits, and norms, and a belief in the ability of informed concerted action to improve human welfare (Hodgson 1994, 58-60). Indeed, several institutionalists in the United States, notably Commons, played important roles in the New Deal during the 1930s--a time during which institution building was high on the national political agenda because it was viewed as a way to extricate the United States from the clutches of the Great Depression.
This old version of institutional analysis was juxtaposed to neoclassical economics, which stressed the need for simple theoretical models that abstracted from reality through the use of formal deductive analysis and mathematics. In contrast, the old institutionalists eschewed the elegance, rigor, and parsimony of formal deductive modeling for thick description, historical analysis, and inductive reasoning that, in their view, more accurately captured the true nature of fundamental historical processes. The old institutionalists were also less inclined than the neoclassicists to make simplifying assumptions about human rationality. Neoclassicists argued that it was reasonable to assume, at least for purposes of formal mathematical modeling, that people were motivated by self-interest and operated according to a cost-benefit rationality. Institutionalists argued that human motivation and rationality was more complex, that people operated sometimes according to altruistic principles, and that the language people thought in affected how they perceived themselves and, thus, their interests. Finally, the old institutionalists insisted that one cannot understand the economy by focusing just on individual market actors. Instead, they argued for a more holistic account that also paid attention to the roles played in the economy by corporations, labor unions, trade associations, banks, and other organizations as well as the unequal distribution of wealth, power, and information among them (Yonay 1998).
The professional competition between institutional and neoclassical economics grew particularly intense during the period between the two world wars. After the Second World War, the intellectual sway of institutional economics declined. By the late 1950s it had become marginalized and neoclassical economics had become dominant in North America and Europe in part due to the rising popularity of econometrics and mathematical modeling (Yonay 1998).
However, a new brand of institutional analysis, the so-called new institutionalism in economics, emerged beginning in the 1970s in reaction to persistent problems for which neoclassical economics seemed to have either little interest or few solutions. In particular, Oliver Williamson (1975, 1985), Douglass North (1981, 1990), and others recognized that competitive markets did not always produce the most efficient economic behavior. They argued that under certain conditions, such as small numbers bargaining or limited information about suppliers and customers, markets were in fact typically inefficient insofar as monitoring and enforcing transactions could be done at lower cost through different institutions like corporate hierarchies or long-term subcontracts. By institutions they meant systems of formal and informal rules and compliance procedures (e.g., North 1990, 3; Rutherford 1994, 182). Hence, whereas the neoclassicists had pushed institutions aside and argued that unfettered markets were generally the most efficient arena for economic activity, the new institutionalists disagreed and wanted to bring the analysis of institutions back into economics.
The new institutional economics also held that although transactions could be organized through a variety of institutions, no institution was necessarily going to be optimally efficient. Why?First, because political elites are self-interested and want to maximize tax revenues and other benefits of office, they frequently create property rights and other institutions that constrain the behavior of market actors in ways that undermine long-term economic growth. For instance, North (1981, 185), who won the Nobel Prize in economics for his work, argued that the domination of politics in Western countries by pluralist interest groups, which emerged from the struggle among workers, farmers, and businesses, generated a system of property rights that favored income and wealth redistribution at the expense of economic efficiency. Second, for a couple of reasons, once institutions are established they evolve slowly in a path-dependent manner. On the one hand, institutions are expensive to create and change. On the other hand, once established, institutions tend to generate positive feedback and support from constituents as well as institutional elites who derive increasing financial returns and other benefits from them. Hence, influential people are often reluctant to modify institutions in fundamental ways--even if institutions are discovered to be inefficient relative to the possible alternatives. Instead, people tend to make only marginal adjustments to their institutions that result in path-dependent change.
The causal concept of path dependence suggests that once people are on a particular institutional path they tend to stick to it rather than jump to another one. Thus, evolutionary change becomes the norm (Arthur 1994; North 1990). According to North (1990, 90), even when major disruptions occur to the institutional status quo, such as a severe price shock or political crisis of the sort that are central to punctuated equilibrium models, the changes that result are often less discontinuous and more evolutionary than they appear on the surface as a result of path-dependent effects. Of course, this is not to say that the new institutionalists reject completely the possibility of truly revolutionary change that constitutes a sharp break from the past. Williamson (1985), for instance, suggested that a sudden increase in transaction costs will be followed by a rapid shift from markets to vertically integrated firms as the principle means of governing economic exchange. And even North (1998, 19) accepted that on at least some occasions abrupt changes may occur that resemble the punctuated equilibrium model, mentioned earlier.
It is important to note that scholars in this tradition occasionally discuss another causal concept that is closely related to path dependence. That is the concept of diffusion (e.g., Coleman et al. 1966). For instance, Brian Arthur (1994, chap. 5) explained that under some circumstances the diffusion of an economic practice through a market may be driven by "information contagion," that is, informal word-of-mouth information sharing about what the best practice is. He argued that buyers who seek such information tend to learn about practices that previous buyers have already adopted and with which they are already familiar. Hence, buyers are likely to learn more about a frequently used practice than one with few previous users. In this way, certain practices are more likely to diffuse through the market than others as a result of an information feedback process that is much like those that are associated with path-dependent change. However, researchers who are influenced by the tradition of the new institutional economics and who are sympathetic to diffusion arguments point out that little attention is paid in this literature to whether the practice in question is transformed in the process of diffusing from one context to another and, if so, how this happens (Kiser and Abel 2002). As we shall see, this is also a problem in other versions of institutional analysis.
The new institutional analysis in economics did not constitute a complete rejection of the neoclassical tradition. In particular, North, Williamson, and their followers embraced methodological individualism--the idea that micro-level individual actions give rise to institutions. They also accepted in modified form the neoclassical assumption that individual action is motivated by an instrumental rationality, that is, a decision-making logic based on an interest in maximizing benefits relative to costs. The modification was that rational decision making was limited by the availability of relevant information and the cognitive ability of actors to process it (Langlois 1986; Rutherford 1994, chap. 3). Institutions, they argued, are built to advance actors' self-interests. By self-interests they meant an actor's concern with improving his or her well-being (Coleman 1990, 28-29; Scharpf 1997, 40).12 Once institutions are built, actors continue to pursue their interests as best they can within the constraints imposed on them by these institutions. In contrast to the old institutionalism, the new institutionalism was more abstract, formal, individualist, and oriented toward rational choice and economizing models (Rutherford 1994, 4). As a result, this view represented a rational choice approach to the study of economic institutions.
In sociology and political science the development of rational choice institutionalism was heavily influenced by these ideas. In sociology it was also a reaction against the legacy of Emile Durkheim (1933, 1938) and the structural functionalists, notably Talcott Parsons (1951), who argued that social structure and institutions generate individual action, not the other way around. Rational choice institutionalists in sociology rejected this view in favor of methodological individualism (Coleman 1990; Hechter 1987; Nee 1998). So did their counterparts in political science. Furthermore, in political science, rational choice institutionalists reacted against an earlier institutionalist tradition of the late nineteenth and early twentieth centuries, illustrated, for instance, by Woodrow Wilson's scholarly work. Like the old institutionalists in economics, the old institutionalists in political science were intent on describing in rich detail and understanding the development of formal institutions like constitutions, cabinets, and parliaments in non-abstract and often atheoretical terms. They did so with a strong normative emphasis that was often rooted in moral philosophy, and their work paid little attention to the development of testable propositions (Peters 1999, chaps. 1, 3; Scott 2001, 6-8). Given its reaction against the older views, it is not surprising that much rational choice institutionalism in political science, and to a lesser extent in sociology, is now characterized by formalized and abstract game-theoretic modeling (e.g., Calvert 1998).
Rational choice institutionalists in sociology and political science also paid much attention to how individuals build and modify institutions to achieve their interests (e.g., Kiser and Kane 2001; Knight 1992, 1998). Some accepted that patterns of punctuated equilibrium occur occasionally and cause fairly revolutionary and abrupt institutional changes (Moe 1987; Shepsle 1986). However, many others maintained that path-dependent evolutionary change is the norm (e.g., Ingram and Clay 2000, 527). For example, in a discussion reminiscent of North's objections to punctuated equilibrium models, political scientist William Riker (1998, 121) argued that even brand new and apparently revolutionary institutions are not created de novo, but contain vestiges from the past. Riker showed that the U.S. Constitution evolved from the Articles of Confederation, the national government of the Continental Congress, the thirteen state governments, and British colonial administration. But regardless of whether they focus on evolutionary or revolutionary change, rational choice institutionalists find that both norms and more formal institutions emerge and are enforced as a result of self-interested behavior (e.g., Brinton and Nee 1998, part I; North 1981, chap. 5; Ostrom 1990, chap. 6).
Perhaps the most important contribution that sociologists and political scientists have made to the rational choice paradigm is to develop the so-called choice-within-constraints approach--a view that takes seriously that institutions, including informal norms as well as formal rules and regulations, limit the range of choices individuals are likely to make as they pursue their interests (Alt and Shepsle 1990a; Nee 1998; Scharpf 1997). For example, Margaret Levi (1988) argued that political rulers are predatory insofar as they are interested in maximizing the revenues they collect from citizens. Their ability to maximize revenues is constrained, however, by the type of political institutions within which they operate. During the Middle Ages, English monarchs eventually had to negotiate with parliament over tax policy. This lent tax policy a degree of credibility and legitimacy that was absent in absolutist states, such as France, and so English rulers were able to tax more effectively than the French. Variation in institutional constraints made the difference.
It is worth mentioning briefly that rational choice theorists who are especially interested in game theory do not always follow the choice-within-constraints approach. Instead of viewing institutions as rules that constrain behavior, they argue, as noted earlier, that institutions are strategic equilibria--situations where no one sees an advantage in changing his or her behavior. As game theory has become increasingly popular among political scientists, even those who initially adopted the choice-within-constraints approach, such as Margaret Levi, have shifted toward this new definition of institutions (e.g., Bates et al. 1998, 8-10).
Finally, there has been much talk among rational choice institutionalists about the need to better incorporate an analysis of "ideas" into their work (e.g., Garrett and Weingast 1993; Goldstein and Keohane 1993b). Many have acknowledged that cognitive structures, belief systems, and other sorts of ideas influence how actors perceive their interests and options in the first place, including which institutions they might prefer to build and sustain under different circumstances. Notably, the concept of bounded rationality, which rational choice theorists invoke frequently, often alludes to these sorts of cognitive constraints. Bounded rationality refers to the limited capacities of actors to collect and process important information and use it to make well-informed decisions (Jones 1999). Yet, according to some rational choice institutionalists, although interest in the issue of cognition is increasing, their colleagues have failed to incorporate adequately an analysis of these sorts of ideas into their work (Knight and North 1997).13
In sum, although there are several varieties of rational choice institutionalism in the social sciences (e.g., Alt and Shepsle 1990b; Knight and Sened 1998b), they all tend to share basic features that relate to the problems in institutional analysis raised earlier. First, many rational choice institutionalists examine how institutional change typically follows an incremental and evolutionary pattern rather than a punctuated equilibrium pattern. Others attend to more dramatic and abrupt changes. In either case, however, little attention has been paid to how we can tell whether change is evolutionary or not. Indeed, a recent review of the literature on rational choice institutionalism makes no mention of this issue (Ingram and Clay 2000). Second, rational choice institutionalists believe that change tends to be evolutionary because it occurs through the path-dependent mechanisms of feedback, increasing returns, and choice within constraints (Knight 1992, chap. 4; North 1998). Some also extend the discussion of path-dependence to the process of diffusion. Nevertheless, reflecting on the state of the art, North (1998, 21) complained that rational choice institutionalists "simply are ignorant" about the mechanisms involved in path-dependent institutional change. As noted earlier, the same seems to be true for diffusion insofar as they neglect how institutional practices may be transformed as they are adopted into local contexts. Third, according to North, a better understanding of these mechanisms requires rational choice institutionalists to pay much closer attention to the normative and cognitive processes by which actors perceive the range of choices available to them. In other words, they need to take ideas as well as self-interest much more seriously than they have to date. In particular, although some rational choice institutionalists have recognized increasingly that cognitive structures, belief systems, and the like constrain change, most theorists, in North's view, have paid rather limited attention to how ideas like these come into play and understand very little about how they operate (North 1990, 42-44; Knight and North 1997).
One way organizational institutionalism differs from rational choice institutionalism is in its much greater emphasis on normative and cognitive ideas. Organizational institutionalism's roots stretch back to Weber, Durkheim, and several mid-twentieth-century organization theorists (DiMaggio and Powell 1991; Perrow 1986, chap. 5; Scott 2001, chaps. 1, 2). Prominent among them was Philip Selznick (1949), whose classic study of the Tennessee Valley Authority (TVA) is frequently identified as a cornerstone of the old institutionalism in organizational analysis. In constructing a detailed case study of the development of the TVA, Selznick argued that organizations are not always rational. That is, they do not always act independently in the interest of accomplishing their official goals. Instead, organizations often change in response to the pressures and values imposed on them by powerful constituents in their environment. As organizations are infused with these values, their real objectives diverge from their officially mandated goals and they come to serve the values and interests of these constituents. In turn, as organizations become more important for actors in their environment who depend on them, they become institutionalized along with the values they embody.
According to Selznick, the TVA was established ostensibly to facilitate regional development through democratic planning. However, he showed that to maintain control, consent, and ultimately its authority, the organization coopted local citizens and included representatives of powerful agricultural interests among its leadership. As a result, contrary to the organization's initial charge, its programs ended up serving more affluent and powerful agricultural constituents at the expense of its official and more populist goals. Following Selznick, a number of early organizational institutionalists documented that organizations respond to environmental pressures by altering their initial goals in order to survive (e.g., Zald and Denton 1963).
What is important here is that much of the old institutional analysis in organization studies maintained that norms and values were an important part of institutional life. Organizations responded to the norms and values in their environment in ways that resulted in behavior that did not fit traditional theories of organizational decision making, which rested on rationalist and utilitarian assumptions. These were the same assumptions that underpinned neoclassical economics and the new institutionalism in economics (March 1996, 281-83). Thus, according to organizational institutionalists, organizations sought to act appropriately vis-à-vis their cultural environments rather than instrumentally vis-à-vis their official goals.14
The difference between a logic of appropriateness and a logic of instrumentality is central to new versions of organizational institutionalism (March and Olsen 1989, chap. 2). Building on the idea that organizations act in response to the norms and values in their environment, John Meyer and Brian Rowan (1977, 340) argued that
organizations are driven to incorporate the practices and procedures defined by prevailing rationalized concepts of organizational work and institutionalized in society. Organizations that do so increase their legitimacy and their survival prospects, independent of the immediate efficacy of the acquired practices and procedures.
In other words, organizations adopt whatever practices they believe their institutional environment deems appropriate or legitimate regardless of whether these practices increase organizational efficiency or otherwise reduce costs relative to benefits, as rational choice institutionalists contend.
Meyer and other new organizational institutionalists have generated a vast number of empirical studies arguing the basic point that the behavior and formal structure of business firms, schools, nation-states, and other organizations are defined by their institutional environments and that these organizations seek legitimacy from their environments to ensure their survival (e.g., Boli and Thomas 1999a; Scott and Meyer 1994; Thomas et al. 1987). In contrast to the old institutionalists like Selznick, who focused on individual organizations, the new institutionalists shift their attention to entire fields or populations of organizations within an institutional environment. For them, institutions include informal and common cultural frameworks, symbolism, and taken-for-granted cognitive schema as well as formal rule systems. Importantly, institutions also contain routines and processes that sustain these frameworks, symbols, schema, and rules and, therefore, reproduce themselves over time (Jepperson 1991, 145). Moreover, because organizations of a common type, such as art museums or business firms, share a common institutional environment, they all tend to adopt similar structures and practices over time. Hence, fields or populations of organizations gradually become isomorphic or homogenous. This is not because they all instrumentally seek the most efficient way of operating or try to maximize benefits relative to costs, as rational choice institutionalists would expect, but because they seek to conform to the culturally appropriate scripts, schema, and organizational models in their environment. For instance, organizational institutionalists have argued that the field of state governments in the United States adopted very similar enabling legislation for the establishment of health maintenance organizations because they perceived that it was the legitimate thing to do within their political environment (Strang and Bradburn 2001).
The distinction between normative and cognitive factors is important for organizational institutionalists. Whereas much of the old organizational institutionalism emphasized that actors respond to the normative and value-laden pressures in their environments, and do so deliberately and self-consciously, new versions of organizational institutionalism additionally stressed cognition and how actors' perceptions of what is an appropriate practice depend on the taken-for-granted scripts, schema, habits, and routines that they possess and through which they interpret the world (Scott 2001, chap. 2). Indeed, actors may not have a clear sense of what their self-interests or goals are at all, particularly during periods of great uncertainty and information scarcity, and so they may act in unreflective ways according to these taken-for-granted cognitive structures. Hence, institutionalized scripts and routines are said not only to constrain action but also to enable or constitute it by providing actors with models to guide their behavior (Clemens and Cook 1999). In this regard, the new organizational institutionalism also differs sharply from much rational choice institutionalism and draws heavily from phenomenology, ethnomethodology, and cognitive psychology (DiMaggio and Powell 1991; DiMaggio 1997).15
Reliance on normative and cognitive factors has not been without its problems. For example, despite its importance within this paradigm, some organizational institutionalists have admitted that the distinction between normative and cognitive ideas is often blurred; that it is not clear when norms become so taken for granted that they become part of a cognitive structure (Clemens and Cook, 1999, 445-46). Critics, such as Paul Hirsch (1997), have also suggested that the mechanisms by which normative and cognitive structures affect behavior are often poorly specified.16 In any case, insofar as taken-for-granted institutions are said to constrain behavior, on the one hand, and to enable or constitute it, on the other, in ways that lead to the unreflective reproduction of behavior over time with only minor variations, some organizational institutionalists have embraced the idea that institutional change generally occurs in path-dependent evolutionary ways (Powell 1991, 189-94).
Another branch of organizational theory, although one not always associated with institutionalism, also argues that institutional change tends to follow an evolutionary pattern. Population ecologists maintain that principles and practices become institutionalized as they diffuse through populations of organizations and thus become taken for granted cognitively as the legitimate mode of operation (e.g., Carroll and Hannan 1989). Adopting concepts from evolutionary biology, the population ecology view holds that diffusion is driven largely by competitive selection. Organizations adopting the principles and practices that best fit their environmental niche will be the ones most likely to survive. Although most organizational institutionalists distance themselves from this approach, ecologists argue that the two views are compatible once we realize that populations of organizations may evolve as they seek to adopt principles and practices that confer legitimacy on them as social actors (Hannan and Freeman 1989, 34-35). The problem, they admit, is that while they know a good deal about selection processes, they know much less about the processes that lead to the creation of new organizational principles and practices in the first place (Hannan and Freeman 1989, 20-21). In other words, this approach neglects an analysis of creative innovation, entrepreneurial activity, and agency. The same criticism has been leveled against organizational institutionalists who talk about how cognitive schema and other institutions constrain change in path-dependent ways (Hirsch 1997).
However, organizational institutionalists are less sanguine about evolutionary change than many rational choice institutionalists. To be sure, many argue that evolutionary change is the norm and that even apparently revolutionary institutional changes may, upon closer inspection, stem from the gradual accretion of many incremental path-dependent adjustments (e.g., Van de Ven and Garud 1993)--an argument very much like Riker's rational choice analysis of the U.S. Constitution. But others argue that when change occurs, it is more likely to follow a punctuated equilibrium or punctuated evolution pattern. That is, change is episodic and marked by a brief period of crisis or critical intervention followed by a longer period of stability or path-dependent evolutionary change (Powell 1991, 197).
For example, Neil Fligstein (1990) showed that the development of corporate America followed a pattern of punctuated equilibrium. Periodic bouts of market and price instability, severe competition, and overproduction triggered a series of crises for U.S. corporations throughout the late nineteenth and twentieth centuries. In response to each crisis, corporate managers shifted from one conception of corporate control to another. For Fligstein (10), conceptions of control were "totalizing worldviews" through which managers interpret their situations. Direct control of markets through holding companies and horizontal mergers gave way in the early 1900s to a manufacturing conception of control that was based on vertical integration and price leading. This was replaced by a sales and marketing approach during the 1930s that emphasized product diversification. Then, during the 1960s, a financial conception of control emerged that privileged conglomerate mergers. In each case, a crisis upset the institutional equilibrium and precipitated a search for a new corporate order that, once institutionalized, ushered in a new period of institutional equilibrium.
Regardless of whether they subscribe to an evolutionary, punctuated equilibrium, or other view of change, the new organizational institutionalists generally agree that fields of organizations adopt common institutionalized practices as a result of three essential processes by which practices diffuse through the field. First are mimetic processes in which organizations facing uncertainty lack a clear idea of what they should do and so copy the practices of other, apparently successful, organizations in their field. Second are normative processes in which organizational leaders have been professionalized to share similar views on what constitutes appropriate organizational practice. Third are coercive processes in which organizations adapt to pressure from other organizations around them to conform to institutionalized standards (DiMaggio and Powell 1983; Mizruchi and Fein 1999). Diffusion is the operative concept here. So, for example, Pamela Tolbert and Lynne Zucker (1983) argued that civil service reform began to diffuse during the late nineteenth century across municipal governments in the United States because it was seen as a way to reduce political conflict and improve government operations. However, as the reform model was adopted by an initial set of governments, it became institutionalized. Later, when other governments adopted it, they did so more to obtain social legitimacy and because they simply took for granted that it was the appropriate way to organize their operations. As noted earlier, the problem with the diffusion approach, according to critics, is that it appears to be a mindless and excessively mechanical transfer of principles and practices from one organization to another. In other words, the underlying mechanisms of transfer are underspecified, and again there is little room for an account of actors and agency (Hirsch 1997; Rao et al. 2003).17
Organizational institutionalism was developed and elaborated primarily by sociologists. To be sure, a few economists have relied in part on notions of culturally specific, taken-for-granted routines, practices, and norms to explain variation in economic structure and performance across organizations and countries and over time (e.g., Best 1990; Hodgson 1988, chap. 5; Lazonick 1991; Nelson and Winter 1982; Whitley and Kristensen 1997). However, outside of sociology it is probably in the area of international relations that this approach has been most influential (e.g., Finnemore 1996; Katzenstein 1996b). For instance, Alexander Wendt (1992; see also Jepperson et al. 1996) argued, in contrast to more rational-choice or realist views, that states interact with one another through their representatives and through these interactions they develop a set of intersubjective understandings about themselves and each other. That is, they develop identities and worldviews at the cognitive level through which they define situations, their interests, and subsequent courses of action. According to Wendt, through this identity formation process states decide whether to pursue self-help security strategies, such as defending themselves without assistance from other states, or more cooperative strategies, like NATO, in which they work with other states for mutual defense. Furthermore, work in international relations has suggested that international norms also influence state security strategies and other decisions, such as whether states should procure and deploy nuclear or chemical weapons systems (e.g., Eyre and Suchman 1996; Price and Tannenwald 1996).
To review, first, while some organizational institutionalists accept that institutional change follows a punctuated equilibrium or perhaps punctuated evolutionary pattern, others argue that institutional change often tends to be evolutionary. Second, both path dependence and diffusion are important concepts that organizational institutionalists use to account for change. Proponents of both concepts have been criticized generally, however, for overemphasizing the importance of institutional structure, failing to demonstrate the mechanisms by which these structures affect actors, and, therefore, neglecting that actors and agency are important ingredients in episodes of institutional change (Finnemore 1996, 339; DiMaggio and Powell 1991; Hirsch 1997; Hirsch and Lounsbury 1997; Stinchcombe 1997). Third, the new organizational institutionalists emphasize the importance of cognitive schema and normative principles as determinants of behavior and change. In contrast to rational choice institutionalists, they often put much more emphasis on ideas like these than they put on self-interest as motivational forces. Still, as we have seen, this has not been without its problems--notably, a failure to differentiate clearly between normative and cognitive ideas.
Historical institutionalism is derived from classical political economy, particularly the historical materialism of Marx and the comparative institutional history of Weber. It has beendeveloped primarily by political scientists who have studied how political and economic decision making is affected by the institutional arrangement of states, including the organization of government agencies, parliaments, constitutions, and electoral rules; by the institutional arrangement of economies, including the organization of labor unions and business associations; and by the institutional connections between states and economies, including economic policy, business regulations, and property rights (e.g., Katzenstein 1978). In this view, institutions are sets of formal and informal rules and procedures, such as those codified in the law or deployed by bureaucratic organizations like states and business firms (Thelen and Steinmo 1992, 2).
Much historical institutionalism has been developed in a macro-level comparative context. For example, many scholars have shown how variation in national political institutions affected the level and progressivity of tax codes across countries (Steinmo 1993), the development of welfare states (Weir and Skocpol 1985; Skocpol 1992), differences in national health care systems (Immergut 1992), and the responses of different governments to international economic crises (Gourevitch 1986). Of course, some researchers have made important insights by focusing on a single state, such as Stephen Skowronek's (1982) analysis of the development of the U.S. civil service, regulatory, and military institutions during the early twentieth century, and how these developments affected the capacity of the federal government to pursue its goals. In all of these instances, and in contrast to organizational institutionalism, historical institutionalists have focused more on broader societal and state structures than on organizations per se. The important point in much of this work is not that institutions directly determine outcomes like these but, more modestly, that institutions constrain them. In a classic statement of the position, Kathleen Thelen and Sven Steinmo (1992, 3) declared that, "what is implicit but crucial in this and most other conceptions of historical institutionalism is that institutions constrain and refract politics but they are never the sole 'cause' of outcomes." This view resembles the choice-within-constraints approach of rational choice institutionalism, discussed earlier.
Historical institutionalism differs from the old institutionalism in political science of the late nineteenth and early twentieth century. As noted earlier, the old institutionalists were concerned primarily with the formal aspects of government, especially the law, how government worked, which institutions worked best, and how government might be improved. Notably, Woodrow Wilson, one of the earliest presidents of the American Political Science Association, wrote several essays about what the U.S. government could learn from European governments that might improve democracy and governmental performance. Given its normative concerns, the old institutionalists were often engaged in progressive and good government movements. Most of their work was very descriptive and lacked much theoretical development or hypothesis testing (Peters 1999, 3-11). In contrast, the recent historical institutionalists are less concerned with these normative issues. They are more interested in studying the informal as well as the formal aspects of government, analyzing the determinants of policy and its outcomes, particularly in historical and comparative perspectives, and developing and testing theories of political institutional development and policy performance. For instance, historical institutionalists have been engaged in heated theoretical debates about the origins of welfare states (e.g., Skocpol 1992; Weir et al. 1988).
However, historical institutionalism emerged mainly in reaction to several streams of thought that had become ensconced in political science and political sociology after the Second World War. First, it took exception with the behavioralism of the 1950s and 1960s, which explained politics in terms of the behaviors of political leaders, interest groups, social classes, or other actors without considering the institutional context within which they operated and how it might affect the interests and behaviors of these actors (Skocpol 1985). Second, it was an effort to move away from functionalist, Marxist, and other forms of grand theorizing that did not adequately explain cross-national variation in how interest groups and social classes were organized and why a particular group or class might demand different policies in different places and at different times (Thelen and Steinmo 1992). Third, it was a reaction against rational choice theory. Historical institutionalists were concerned that material self-interest was not the only motivation for action and that what needed to be explained was not just how institutions shaped actors' strategies, but also how it shaped their preferences and goals in the first place (Thelen and Steinmo 1992). Fourth, much of the ground-breaking work in this paradigm was done in an effort to understand why some advanced capitalist countries were more likely to suffer stagflation than others after the collapse of the Bretton Woods agreement and the oil embargos of the 1970s. Building on earlier work in comparative political economy (e.g., Shonfield 1965), many scholars argued that differences in economic policy and national economic performance during the late 1970s and 1980s depended on whether liberal, statist, or corporatist institutions were in place (e.g., Gourevitch 1986; Katzenstein 1978; Lindberg and Maier 1985). Finally, and more recently, historical institutionalists emphasized the importance of understanding institutional change as well as stability, and thus sought to improve their own work by theorizing how institutions developed in the first place (Thelen and Steinmo 1992).
Central to this perspective is the notion that the institutions that guide decision making reflect historical experience (Hall 1986; March and Olsen 1989, chap. 2). In other words, once institutions have been established through complex struggles and bargaining among organized groups, they have a continuing effect on subsequent decision-making and institution-building episodes. For example, Ira Katznelson (1985) argued that the early provision of voting rights for white men in the United States and a decentralized federalist system of government established by the Constitution mitigated the formation of working-class consciousness and, therefore, a national working-class political party. Workers could obtain social services and other benefits through community-based political machines that traded patronage for votes and that typically organized workers along racial and ethnic rather than class lines. Coupled with the fact that workers did not have to organize politically to obtain the vote, there was little interest for organizing a national labor party. In contrast, the British working class had to struggle for the vote and had to cope with a more centralized national government to obtain government services and protections. As a result, British workers formed a national labor party. Consequently, the presence or absence of a labor party had profound long-term effects on national policy-making in both countries.
As this example suggests, historical institutionalists embrace the idea that policy-making and institutional change tend to be path-dependent processes. This is because institutions constrain the choices available to decision makers; because decision makers incrementally adjust their policies and institutions in response to feedback they receive from their constituents; and because decision makers learn gradually which policies and institutions best suit their purposes (Heclo 1974; Pierson 2000a, 2000b). Insofar as historical institutionalists have emphasized the importance of policy learning, particularly as policy models travel from one country to another, occasionally with a coercive push from powerful states, some of them have recently begun to discuss policy diffusion (e.g., Djelic 1998; Duina 1999). Especially notable here is Peter Hall's (1989a) volume on the diffusion of Keynesianism across the advanced capitalist democracies. In this regard historical institutionalists have begun to develop similarities with a sizeable literature on cross-national policy transfer (e.g., Dolowitz and Marsh 1996).
However, historical institutionalists have also been very interested in abrupt and revolutionary shifts in policy and institutions that represent sharp breaks from the past. (Indeed, Theda Skocpol , one of the major proponents of this paradigm, began her career with an important comparative study of social revolutions.) Historical institutionalists frequently describe patterns of stability and change as punctuated equilibrium (e.g., Baumgartner and Jones 1993; Hall 1993; Krasner 1984). According to some critics, this has created problems for them insofar as their explanations of rapid, revolutionary change do not square well with an analytic framework that talks much about path-dependence. Put in slightly different terms, if institutions are so important in constraining policy-making outcomes at one moment, how can they suddenly become so unimportant at another (Peters 1999, 65; Thelen 1999)? For this and other reasons, some historical institutionalists have complained that the concept of path dependence is often used rather loosely without a clear specification of the mechanisms involved (Pierson and Skocpol 2000; Pierson 2000b; Thelen 2000a).
Recognizing the importance of path dependence, some historical institutionalists have followed Douglass North in maintaining that what may appear initially as a sudden episode of policy-making or institutional punctuated equilibrium and dissonance often turns out to be one of much more evolutionary change. Certainly an interest in the evolutionary pattern of change is consistent with a long-standing tradition in political science that policy-making and institutional change are best conceptualized as being incremental in nature (e.g., Lindblom 1959; Witte 1985). Notably, Paul Pierson (1994) argued that what appeared initially to many scholars as an abrupt and radical overhaul of U.S. welfare policy in the early 1980s turned out later to be considerably less dramatic and more evolutionary. Still other scholars have argued that there is some truth to both views and that the best way to describe institutional change is as a pattern of punctuated evolution (Hay 2001).
Researchers working in the historical institutionalist tradition have long accepted that self-interest motivates behavior and institutional change. Increasingly, however, they have also emphasized that ideas and principled beliefs, such as convictions about what constitutes good public policy and good government, also influence decision making (e.g., Derthick and Quirk 1985). In this regard, they are clearly related to organizational institutionalists in recognizing that a logic of appropriateness may be just as important as a logic of instrumentality (Hall 1993; March and Olsen 1989, chap. 2). In a few cases they have made efforts to deploy an analysis of ideas to help account for those moments when institutional equilibrium is punctuated by a crisis and eventually gives way to fundamental policy and institutional change. Mark Blyth (2002), for instance, showed how struggles over foundational economic theories during moments of economic crisis resulted in intellectual paradigm shifts among academic and political elites that contributed to the overhaul of economic policy during the late twentieth century in the United States and Sweden. Following Peter Hall (1993), Blyth argued that cognitive paradigms constrain the choices available to actors until they no longer provide answers to the problems these actors face. Then a search is launched for a new paradigm that, when found, ushers in a new set of policies and institutions (see also Hay 2001). In other words, although the instrumental pursuit of interests guides reform, so does the sensitivity of decision makers to what they believe are the appropriate blueprints for reform.
To summarize, first, historical institutionalists debate whether institutions vacillate in patterns of evolution and incrementalism, punctuated equilibrium, or punctuated evolution. Yet, as the next chapter shows, it is not always clear for them how we should determine how much change has occurred in a given historical episode or which pattern best characterizes it. Second, although historical institutionalists embrace the notion of path-dependent change, they also recognize that much work remains to be done to specify better the mechanisms involved. As we shall see in chapter 3, the mechanisms underlying diffusion also require attention. Third, historical institutionalists can be located somewhere between rational choice and organizational institutionalists in the sense that they seek to offer a more balanced treatment in their work of the interplay between ideas and a logic of appropriateness, on the one hand, and interests and a logic of instrumentality, on the other hand (e.g., Goldstein and Keohane 1993a). Still, as critics have charged (e.g., Blyth 1997; Yee 1996) and as chapter 4 shows, they often run into difficulty conceptualizing and studying how ideas affect decision making and institutional change.
CONCLUSION AND A LOOK AHEAD
Overall, it is clear that within each of the three institutionalist paradigms scholars tend to debate whether evolutionary, punctuated equilibrium, or other patterns of institutional change are the most prevalent. To my knowledge they have not established clear guidelines for how this debate should be reconciled. They also tend to adopt similar causal concepts, notably path dependence and diffusion, to explain institutional change but have been criticized for poorly specifying the underlying mechanisms involved. Finally, in varying degrees, they have all acknowledged that ideas other than that of self-interest need to be included in accounts of decision making and institutional change, although concerns exist as to how this ought to be done. In other words, the problems of change, mechanisms, and ideas are shared by all three paradigms. Once these problems are reconciled, we will see that institutional analysis can be used to shed new light on the nature of change associated with important empirical phenomena like globalization.
Let me preview more specifically the arguments that lie ahead. In chapter 2, I explain that to determine more accurately how much change occurs in any given case and whether it is more evolutionary or revolutionary, we need to specify carefully the important dimensions along which we seek to observe change, establish the time frame during which we conduct our observations, and then see how many dimensions change during that period of time. In chapter 3, I argue that to understand better how institutional change occurs in path-dependent ways and through diffusion, we need to recognize that it rarely starts from scratch. Typically, institutional change involves the recombination of old institutional elements and sometimes the introduction of new ones as well. I refer to these processes as bricolage and translation, respectively. In chapter 4, I identify four different types of ideas that affect institutional change. I also discuss the actors that create, transform, and mobilize them, and some of the mechanisms through which each type of idea exerts its effects.
In chapter 5, I pull all of these arguments together and demonstrate their utility for understanding important empirical phenomena, in this case globalization. I show that if we identify important institutional dimensions associated with globalization and track them over an appropriate period of time, we find substantially less institutional change at the level of national political economies than much of the globalization literature suggests. Furthermore, the change we see is much more evolutionary than revolutionary. This is because national institutions affect how decision makers respond to the pressures for change that are often said to be associated with globalization. Hence, change has been much less pronounced than many people have assumed because it has happened in ways that are heavily path-dependent; it has happened in ways that involve much translation and bricolage; and it has happened in ways that illustrate how ideas as well as formal institutions facilitate and, more importantly in this case, constrain change. Finally, in chapter 6, I present a theory of institutional change based on arguments developed in the previous chapters. It explains how institutional entrepreneurs and others innovate in response to the problems they confront; how they define and frame these problems as well as possible institutional solutions; how this creative process is constrained by the social, organizational, and institutional contexts within which it occurs; and how all of this affects the probabilities that an episode of institutional change will be relatively more evolutionary or revolutionary.
Throughout this book I draw material from different institutionalist paradigms and synthesize it in various ways to make my arguments. This sort of synthetic move is necessary if we agree that it is time for a second movement in institutional analysis where we begin to think seriously about building bridges among these paradigms.
That said, my aim is not to offer some sort of all-encompassing Grand Synthesis where everything in the three institutionalist paradigms is combined into a new theory that completely dissolves the distinctions among paradigms and results in a fully unified view that all institutionalists will accept. Why? There are some real differences among proponents of the three different paradigms. For instance, some institutionalists have fundamental disagreements over the micro-foundations of social action. Some rational choice institutionalists assume that people are motivated strictly by individual, material self-interest and that institution building results from the pursuit of these interests (Ingram and Clay 2000, 529). In contrast, many organizational institutionalists assume that people are motivated by collective identities and cultural frames and that these drive institution building in ways that do not fit neatly with explanations based simply on the pursuit of rational self-interest (Fligstein 2001b, 106). In some cases, these differences may be so extreme as to be irreconcilable (e.g., Hay and Wincott 1998).
The same may be true for disagreements within paradigms. Some positions are more extreme and hard-core than others and, therefore, less susceptible to synthesis. For instance, some rational choice institutionalists make the very strong assumption, often as an admitted convenience for building mathematical models and abstract game theories, that actors behave in a fully rational manner and are in fact able to maximize benefits relative to costs. Others make the weaker, yet more realistic, assumption that actors behave only in an intendedly rational manner. In other words, actors try to behave as rationally as possible given the fact that they make decisions under conditions of bounded rationality due to cognitive and other limitations on their ability to calculate costs and benefits accurately (e.g., Williamson 1985, 44-47). The latter is more compatible with historical and organizational institutionalists who are interested in how cognitive schema, norms, and the like affect institutional change. And, of course, there are other differences that may affect the possibilities for synthesis. Some rational choice institutionalists are more interested than others in how history shapes behavior in path-dependent ways; some are more interested in getting their hands dirty with data than in simply engaging in abstract theorizing; and, as noted earlier, some define institutions as rules that constrain action whereas others insist that institutions are strategic equilibria.
My point is simply that some things may be more amenable to synthesis than others. As a result, writing this sort of a book is much like institutional change itself. It involves a process of bricolage. That is, it involves selecting various ideas from different places and combining them in ways that yield something new. Of course, this is not an unbiased process. I draw more heavily from historical and organizational institutionalism, the traditions that I have studied the longest, than I do from rational choice institutionalism. And within the rational choice tradition I tend to find the work of institutionalists like North and others who follow in his footsteps to be more useful than the work of the more abstract game theorists. This is because the former have more realistic assumptions about the micro-foundations of human behavior, are more sensitive to the path-dependent nature of institutional change, take norms and cognitive structures more seriously, have a greater affinity to empirical research rather than to abstract theory construction, and define institutions as rules rather than strategic equilibrium. As a result, the former have more in common with historical and organizational institutionalism in the first place and, therefore, offer more immediate synthetic possibilities. Other people might have a different selection bias and, as a result, might come up with different synthetic insights. I encourage them to do so. But this does not negate my basic argument that all three institutionalist paradigms share common problems and that an important step toward resolving them involves synthesizing portions of each.
Return to Book Description
File created: 8/7/2007
Questions and comments to: email@example.com
Princeton University Press