Institutional Theory: Problems and Prospects B. Guy Peters 69 Reihe Politikwissenschaft Political Science Series 69 Reihe Politikwissenschaft Political Science Series Institutional Theory: Problems and Prospects B. Guy Peters July 2000 Institut für Höhere Studien (IHS), Wien Institute for Advanced Studies, Vienna Contact: Prof. Dr. B. Guy Peters University of Pittsburgh, Department of Political Science 4L01 Forbes Quadrangle, Pittsburgh, PA 15260 (: 001-412- 648-7250 email: bgpeters@vms.cis.pitt.edu Founded in 1963 by two prominent Austrians living in exile – the sociologist Paul F. Lazarsfeld and the economist Oskar Morgenstern with the financial support from the Ford Foundation, the Austrian Federal Ministry of Education, and the City of Vienna, the Institute for Advanced Studies (IHS) is the first institution for postgraduate education and research in economics and the social sciences in Austria. The Political Science Series presents research done at the Department of Political Science and aims to share “work in progress” before formal publication. It includes papers by the Department’s teaching and research staff, visiting professors, graduate students, visiting fellows, and invited participants in seminars, workshops, and conferences. As usual, authors bear full responsibility for the content of their contributions. Das Institut für Höhere Studien (IHS) wurde im Jahr 1963 von zwei prominenten Exilösterreichern – dem Soziologen Paul F. Lazarsfeld und dem Ökonomen Oskar Morgenstern – mit Hilfe der FordStiftung, des Österreichischen Bundesministeriums für Unterricht und der Stadt Wien gegründet und ist somit die erste nachuniversitäre Lehr- und Forschungsstätte für die Sozial- und Wirtschaftswissenschaften in Österreich. Die Reihe Politikwissenschaft bietet Einblick in die Forschungsarbeit der Abteilung für Politikwissenschaft und verfolgt das Ziel, abteilungsinterne Diskussionsbeiträge einer breiteren fachinternen Öffentlichkeit zugänglich zu machen. Die inhaltliche Verantwortung für die veröffentlichten Beiträge liegt bei den Autoren und Autorinnen. Gastbeiträge werden als solche gekennzeichnet. Abstract Institutional theory in political science has made great advances in recent years, but also has a number of significant theoretical and methodological problems. The most important of these problems is the generally static nature of institutional explanations. Also, there is a nagging problem of the difficulties in measuring institutional variables in other than simplistic, nominal categories. As well as discussing these problems this paper addresses the static nature of institutional theory by examining the concept of “institutionalization”, and the creation (and tearing down) of institutional structures. The paper argues that by considering institutionalization as a continuous variable rather than a nominal variable we can begin to understand better the dynamics of institutions themselves, and therefore also develop better institutional explanations for other social and political phenomena. Zusammenfassung Die theoretische Perspektive des Institutionalismus war in den letzten Jahren sehr erfolgreich, weist aber immer noch einige wesentliche theoretische und methodologische Probleme auf. Eines dieser Probleme ist, daß Institutionalismus oft sehr unbeweglich wirkt. Ein weiteres Problem bezieht sich auf die Schwierigkeiten Institutionen zu messen. Diese Probleme werden im vorliegenden Papier diskutiert und das Konzept der “Institutionalisierung” kritisch beleuchtet, mit dem Ziel die Dynamik des Vorganges besser zu begreifen und bessere Erklärungen für soziale und politische Phänomene mit Hilfe des Institutionalismus zu erzielen. Contents Zusammenfassung 5 1. Conceptual Confusion: Perhaps Not so Great? 2 1.1 Approaches to Institutionalism...................................................................................2 1.2 How Much Difference is There?.................................................................................4 2. Institutional Change 7 2.1 Distinguishing Two Types of Institutional Change.........................................................7 2.2 Systems Theory and Institutionalization......................................................................9 2.3 Changing Values and Structures .............................................................................10 3. Measurement: What is An Institution? 11 4. Independent or Dependent Variable, or Both? 13 5. Summary: What Does it Explain? How Does it Explain? 14 6. References 15 I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 1 The past decade and a half have seen a major reassertion of institutional theories in the social sciences, and especially in political science. The March and Olsen (1984) article in the APSR was the beginning of the revolution against the methodological individualism of both behavioralism and rational choice approaches. Following from that and their subsequent publications (1989; 1994; Brunsson and Olsen, 1993; Olsen and Peters, 1996) there has been a proliferation of institutional theories and applications of those theories. Similarly, in economics (North, 1990; Alston, Eggerston and North, 1996; Khalil, 1995) and in sociology (DiMaggio and Powell, 1991; Scott, 1995; Zucker, 1987) there has been a birth (or more appropriately a resurrection) of institutional approaches to the basic questions in these disciplines. While this proliferation of interest is gratifying for those of us who never really gave up on the utility of the institution as an explanatory concept, it also presents some important analytic and theoretical problems. This paper will examine the nature of contemporary institutional theory and will focus on questions concerning its coherence and its potential utility as a central organizing approach for political science. There is no need for the domination of one approach or another in a large and diverse discipline dealing with equally diverse versions of government. It is clear, however, that the advocates of rational choice approaches are attempting to impose that orthodoxy on the discipline. The question then becomes whether institutionalism has sufficient analytic power to be a worthy counterpoint to that one attempt to create hegemony. It is not only the apparent theoretical inconsistency within institutionalism that presents problems for extending this mode of political analysis. There are also important empirical problems when we attempt to utilize institutionalism as an organizing theory for political science. In particular, measurement of institutions and variations in their characteristics pose perhaps the greatest challenge to the use of these several theories in a more systematic manner. We assume a priori that institutions make sense as an explanatory variable; the problems arise when attempt to specify what about them matters, and how they exert their influence on the dependent variables (usually policy). As we will point out below, for some versions of institutionalism the measurements are obvious and border on the trivial; the question of measurement is simply what are the formal structures and what can differences among those structures predict. For other versions of this approach the possibilities for measurement are more ambiguous and apparently more remote from any common sense meaning of institutionalism. If we are to continue to follow the canons of the conventional social sciences, and especially develop a comparative political theory, then we must be able to cope with institutions on more than an abstract, theoretical level and develop measures that enable testing of the predictions derived from institutional theory. 2 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S 1. Conceptual Confusion: Perhaps Not so Great? Potentially the most important impediment to a more central position for institutionalism is that the term means so many things to different scholars, and that some of the alternative approaches are not only different but even contradictory (see Hall and Taylor, 1996; Kato, 1995). If one adopts some versions of the institutional approach he or she may have very different empirical evidence, and make very different predictions about behavior, than if one were doing research using another version. This is true even though this research is conducted using what is nominally the same theoretical approach. A central question then is just how much of an impediment these internal differences are and what if anything can be done to generate a more unified approach for institutional theory. 1.1 Approaches to Institutionalism We should begin here with some description and analysis of the basic institutional approaches. I have already done this at book length (Peters, 1999a), but a brief description of the major approaches is required at this point in order to make the rest of the discussion more comprehensible. The above mentioned book dealt with seven versions of institutionalism, but we will only deal with six of those here, and focus attention on only four. In this paper we will not discuss sociological institutionalism, given that it is concerned more with the questions of another, if allied, discipline. 1 We will also spend relatively little time in describing and discussing international regimes (Rittenberger, 1993) and interest groups (Knoke, Pappi and Tsujinaka, 1996; Kickert, Klijn and Koopenjaan, 1997) as institutions. These two versions are, in many ways, more applications of institutional logics to particular settings, just as there also are papers in this conference devoted to applications of the basic ideas of institutionalism to a number of different settings. The first of the major approaches to institutional analysis is the normative approach advocated by March and Olsen (1984; 1989; 1996). They argue that the best way to understand political behavior (seemingly both individual and collective) is through a “logic of appropriateness” that individuals acquire through their membership in institutions. They contrast this normative logic with the “logic of consequentiality” that is central to rational choice theories. That is, March and Olsen argue that people functioning within institutions behave as they do because of normative standards rather than because of their desire to maximize individual utilities. Further, these standards of behavior are acquired through involvement with one or more institutions and the institutions are the major social repositories of values. 1 Further, the March and Olsen approach draws heavily on some aspects of sociological institutionalism, particularly the work of Selznick and his students. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 3 Although March and Olsen might find it difficult to accept, there is also a rational choice version of institutionalism; in fact there are several variations on this general theme. 2 The underlying logic of rational choice institutionalism is that institutions are arrangements of rules and incentives, and the members of the institutions behave in response to those basic components of institutional structure. Unlike individuals in normative institutionalism the preferences of the occupants of these structures do not have their preferences modified by membership in the institution. Rather, the individuals who interact with the institutions have their own well ordered sets of preferences that remains largely unchanged by any institutional involvement they may have. Historical institutionalism is the third of the approaches to be discussed in this paper. The argument of this approach is that the policy and structural choices made at the inception of the institution will have a persistent influence over its behavior for the remainder of its existence (Steinmo, Thelen and Longstreth, 1992). This idea of “path dependency” is the central explanatory principle for the historical institutionalists, although they are also very interested in the ideas that help to shape and to sustain the directions of policy (Hall, 1986; King, 1996). This approach is obviously well-suited to explaining the persistence of policies but is much less promising as a means of explaining change in policies or structures. The final version of institutionalism we will be discussing will be termed empirical institutionalism. This term is employed to describe a body of literature that asks the deceptively simple question of whether institutions make any difference in policy choices, or in political stability. The definition of institutions utilized in this literature is rather a common sense one, emphasizing the formal structures of government. In particular, this literature has focussed on the differences between presidential and parliamentary governments (Weaver and Rockman, 1993; Von Mettenheim, 1996). This is a very old debate in political science, but the institutionalist discussion has elaborated the discussion and demonstrated its possibilities for theoretical development. Like the rational choice version of institutionalism there is yet another version of empirical institutionalism. This version, associated initially with Samuel Huntington, is even more basic than the presidential/parliamentary debate. 3 Huntington argued for the importance of institutions, without specifying clearly what those institutions were (1965; 1968). Huntington's concern was with the creation of structures that intermediated between the generation of demands in society and the government itself. Huntington appeared to be concerned primarily with formal government institutions but the same logic appears in some of Robert Putnam's arguments (Putnam, Leonardi and Nanetti, 1993) for the importance of structures 2 I argue that there are three distinct versions of rational choice institutionalism: principal-agent models (Horn, 1995); game theoretical approaches (Calvert, 1995), and rules-based models (Ostrom, 1990). 3 In sociology one can find some of the same sort of analysis in the work of S. N. Eisenstadt (1963; 1967). 4 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S in civil society for developing stable and effective democracy. Although heavily empirical also, this version of institutional analysis does not have the rather simplistic (Weaver and Rockman, 1993) concern with only a single type of institutional variation, but rather spreads the empirical net much wider. 4 In some ways the empirical version of institutionalism is similar to rational choice versions in that institutions are conceptualized as exogenous to the values of the individuals functioning within them. This statement means that it is assumed that individual values will not be altered by involvement with the institution. Behavior will change in response to the assortment of opportunities and constraints presented by the structure, but the values that condition that behavior are assumed to be unaffected by the institution. This assumption is found in both versions of empirical institutionalism discussed above, although more explicit in the simpler arguments of the presidential v. parliamentary government debate. 1.2 How Much Difference is There? The discussion to this point has concentrated on the characteristics of the individual institutional approaches, and therefore has necessarily stressed the differences existing among the approaches. The most basic question, therefore, is whether there is a sufficient intellectual core to justify an argument that there is indeed a single institutional approach within the several variations. The question then becomes one of judging whether that core can be serve as a unifying approach to the discipline of political science. The answer to this basic question must be both yes and no. There are some features that do unify the approach but at the same time there are also important differences. The unifying features are, I will argue, adequate to provide the beginning of a more unified approach, but there is still the need to develop that approach and to think more explicitly about the theoretical core (for one, highly normative, version see Immergut, 1998). This is not an argument to spend all our time theorizing about institutions; quite the contrary. The need is to interact the theoretical concerns with empirical and sectoral analyses. as is being done in this conference, and to use the interaction to refine the theory as well as the empirical analysis. The most important argument binding these various approaches is that structures – however defined – do matter. At their simplest these structures are the formal structures of government, defined as presidential or parliamentary. The historical institutionalist version is a little more complex, but still is a rather common sense notion of a formal structure that persists across time. For March and Olsen structure appears to be both a formalized 4 A trawl through the papers at the APSA this year yielded literally dozens of papers that purported to provide “ institutional explanations” but most merely meant that there were some minor differences in structure, rather than any theoretical understanding of institutions. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 5 organizational apparatus and also the pattern of the values that those structures possess and inculcate into new members (March and Olsen, 1989); the historical institutionalists also are very much concerned with the influence of ideas on institutions and may speak of structures in ideational terms. For the rational choice advocates, on the other hand, structures are composed of incentives and/or constraints on behaviors. Being rational animals individuals are assumed to respond to those factors in the structure of the institution. Another factor that tends to unify institutional analysis is that structures persist while individuals come and go. Thus, even though individuals appear to be the primary animator in the rational choice version of institutionalism, the institutions in question do appear to enjoy some existence outside of those individuals (but see below). This persistence is obviously central to the historical institutionalists, to the point of creating a perhaps excessively static conception of institutionalism. Similarly, the March and Olsen version of institutionalism assumes that institutions persist and that they attempt to replicate themselves by socializing new members into the values that define the institution. The important point is that we must what motivates political behavior over time, as well as understanding the more immediate pressures for action and change. Following from the above two points, institutionalism also argues that structures (institutions) create greater regularity of human behavior than would otherwise exist and therefore enhance the explanatory and predictive capacity of the social sciences. Even if the rational choice approach depends primarily on individual utility functions and rational calculations based on those utilities the presence of on-going institutions enables the external researcher to predict behaviors and to test hypotheses that might not be possible without the presence of the structures. This is, of course, also the more practical point that Huntington was making about political development – institutions create the predictable, regular behavior necessary for a peaceful and effective political system. All the above about the unifying aspects of institutional theory having been said, there are some obvious theoretical problems in this literature. These problems arise because of the multiple understandings of what an institution is, and about the factors that shape behavior within institutions. One central issue that arises here concerns the source of preferences, and the ways in which individuals and institutions interact in making decisions and forming judgements about “good” policies. On the one hand March and Olsen argue vigorously that preferences are endogenous, based on the experiences of the individual within the institution. At the other end of an implicit spectrum the rational choice approach argues that preferences are exogenous, and do not change appreciably because of involvement with an institution. The other two approaches are less specific on this matter, although the historical institutionalists appear closer to the March and Olsen view and the empirical approach appears closer to the rational choice perspective. The historical institutionalists, for example, 6 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S focus on the role of values in defining institutions and implicitly (and sometimes explicitly) focus on the way in which those values Another important question is the extent to which the different approaches consider institutions to be malleable. We will spend quite a bit of space below discussing institutional change, but the question here is not about how does change occur but rather about the ease with which change may be brought about. On this issue also the approaches can be very far apart. On the one hand the rational choice version tends to see change as quite easy; all one needs to do is to change the incentives (e.g. the pay-off matrix of the game) and behavior will almost immediately change. This assumption help to explain why there has been so much greater interest in institutional design among rational choice theorists (see Weimer, 1995; Goodin, 1995; Keman, 1996). If individuals can be expected to respond to incentives and constraints in predictable ways then design is rather simple At the other end of this dimension of malleability the historical institutionalists assume that change is difficult at best, and likewise that change is very difficult to plan and design. In the historical view change comes about primarily through “punctuations” in the equilibrium that characterizes institutions in this view (see Krasner, 1984; 1988). Institutional design, given the uncontrollability of change, is a virtual impossibility. The other two approaches fall between these two extremes in the ease of change. The normative version of institutionalism tends to see change as rather difficult, involving as it does changing values that have taken some time to build into the institutions (see Brunsson and Olsen, 1993), while the empirical approach actually has very little to say about the malleability of institutions. Thus, the degree of integration (and potentially therefore the degree of utility) of institutionalism may depend upon one sees the glass as half full or half empty. There are some important unifying features, and these help to press home the most central argument that structure does matter. On the other hand, however, it is not at all clear how structure does matter, and how those structures affect the behavior of individuals functioning within them. One way to proceed in developing institutional theory, therefore, is to accept the inherent ambiguity of this body of theory as it currently exists, and indeed to revel in that diversity rather than deploring it. Just as Allison (1971) did for approaches to decisionmaking in organizations so too can we utilize these various versions of institutionalisms as a set of lenses to illuminate different aspects of political structures and political behavior (see for example Peters, 1999b). They all say something important about what institutions are, and they each also have important blind spots. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 7 2. Institutional Change We have already undertaken some discussion of institutional change, but the problem is a sufficiently important one in this body of theory to devote some greater attention to it. One standard critique of institutional theory has been that it is almost inherently static while the world of politics, which it seeks to explain, is almost inherently dynamic. While that critique easily can be exaggerated, there is a certain amount of validity in it. Institutional theories do tend to be “variance” theories, and are better at explaining differences among types of institutions than in explaining the development of one or another individual institution. 5 The critique of an inherently static nature is perhaps especially applicable to the empirical approach, as the epitome of Mohr's “variance” theory of organizations. The question being asked here is simply if existing types of structure make any differences in the types of decisions taken by those institutions. On the other hand the normative version of institutionalism cultivated by March and Olsen tends to focus more on the individual institution and the way in which it develops its internal, common “logic of appropriateness”, This version is not most fundamentally concerned with change but it does have a very clear concern with institutional development. 2.1 Distinguishing Two Types of Institutional Change The above comment about development within an institution points to the need to consider two types of institutional change. One is internal development of the institution, or a process of institutionalization (and also possibly deinstitutionalization). The other type of change is change in the values and/or structures that are assumed to characterize the institutions. Thus, especially in the rational choice version, an institution can remain well-institutionalized but through changing the incentives being offered it can alter its nature and its impact on the individuals with which it interacts. The institutionalization argument, therefore, is that institutions must become institutions; being an institution is a variable not a constant, and not all are as fully institutionalized as are others. For example, a new civil service system created in a post-socialist country may have had some of the formal structural attributes of similar institutions elsewhere, but would not yet have developed the value structure or the autonomy that would enable us to typify it as fully institutionalized civil service system. Selznick (1957) argued that institutionalization involves “infusing a structure with value”, so in the case of emerging civil services it would be argued that the structures must be animated by the appropriate values, not just have formal 5 Mohr (1982; see also Zucker, 1977) distinguishes organizational theories as being variance theories or process theories. The latter type of theory 8 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S structures than could be recognized as being like those in long-standing democracies (Verheijen, 1999). The assessment of the state of an institution must, of course be more than a judgment call and we therefore need to develop some criteria to assess the extent of institutionalization. One set of criteria of institutionalization has been advanced by Samuel Huntington, who argues that there are four dimensions through which we can judge the level of institutionalization of any structure that we observe: autonomy, adaptability, complexity, and coherence. These four have been applied to several types of institutional arrangements (Polsby, 1968; Ragsdale and Theis, 1998) and they do provide one avenue for understanding the transformation that structures must make in order to survive, and to be able to influence their members and their environment. Taken rather briefly and simply these four concepts can be understood in the following ways: 1. Autonomy represents a concern with the capacity of institutions to make and implement their own decisions. Arguably, to the extent that they are not dependent upon another organization or institution they can be said to be institutionalized. This concept might be operationalized in terms of budgets and autonomous sources of revenue; 2. Adaptability taps the extent to which an institution is capable of adapting to changes in the environment, or more importantly capable of molding that environment. As with open systems approaches to social life the institution should be able to continue to import needed resources despite changes in the relevant environment; 3. Complexity demonstrates the capacity of the institution to construct internal structures to fulfill its goals and to cope with the environment. Again, this conception is analogous to thinking in systems theory and structural-functionalism that discussed the importance of structural differentiation (see Almond and Powell, 1967); 4. Coherence represents the capacity of the institution to manage its own workload and to develop procedures to process tasks in a timely a reasonable manner. This also represents a capacity of the institution to make decisions about its core tasks and beliefs and to filter out diversions from those. These four attributes may help in the difficult task of measuring the degree of institutionalization, and hence of measuring institutions themselves. That having been said, several of the variables mentioned above may be more manifestations of institutionalization having occurred, rather than the indicators of the concept itself. That is, organizations become autonomous because they are well institutionalized, so that autonomy indicates the existence of the concept “institutionalization” but it is not a measure of that concept – fever is not the disease, only an indication that some disease or another exists. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 9 Institutionalization itself may be better represented by the complexity and coherence dimensions that indicate the internal management capacity of the structure in question. Another way to conceptualize institutionalization may be the standardization of procedures and the routinization of practices within the organization. Brunsson (1999), for example, argues that this process of standardization of procedures is central to the creation and maintenance of institutions. So long as there substantial internal variation there can not be said to be an institution operative in the field. Although the four concepts derived from Huntington's disucssion of institutinal development, as well as from the other students of institutionalization are useful they may not capture all the complexity of institutionalization. Two additional factors (see Goetz and Peters, 1999) appear important in defining institutionalization in contemporary public sector structures. These two factors are: 5. Congruence. This variable reflects a concern with the extent to which relationships within political institutions match the social relations they are supposed to regularize and maintain. If political institutions are not congruent then they can not be expected to survive and be effective. Also, this reflects a concern about differences between the values of poltiical elites and the values of the masses. 6. Exclusivity. The final variable related to the intensity of functional competition among institutions. When there is little or no competition then the institutions can expect longterm survival. When there are multiple institutions attempting to perform the same tasks then the competition will have to be resolved in some way, often by the termination of one or more institutions. 2.2 Systems Theory and Institutionalization Another way to address institutionalization is to revive the systems theory that has been used in organizational analysis (Katz and Kahn, 1978), as well as in a number of other areas of social science (Easton, 1970). In this theory, organizations are defined in terms of their capacity to create and maintain a boundary from the environment and from other organizations. All organizations must import resources from the environment but the more institutionalized are hypothesized to have effective gatekeepers to control those transactions. We will return to the measurement question in a later section of the paper, but the conceptual issue is more important here – one aspect of institutional change is in the creation of institutions, and not all institutions are as fully formed as are others. Likewise, institutions may deinstitutionalize, and become less autonomous, or less coherent. These changes in the attributes of institutions may predict something of their own behavior, as well as their effects on policy. 10 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S 2.3 Changing Values and Structures The second aspect of institutional change is changing the content of what institutions do, or what they believe. Here there are marked differences among the approaches, with the rational choice versions standing out in their assumptions of the ease of change. For the other approaches, however, change is a real problem if institutional analysis is to be an organizing concept for political science. The stereotypical conception of institutionalism is that it is at best static, and even if it can provide useful explanations of different policy outcomes or different types of decisions, it can not provide an explanation for change. The above discussion is to some extent a restatement of the familiar structure/agency problem in constructing social theory (Callinicos, 1989; Hay, 1995). Should we focus on structure and thereby run the risk of hyperstability in our explanations, or do we focus on agents of change and thereby encounter the risk of unpredictability and instability in our understanding of the social world? Both of these horns of the dilemma can be overstated, but broadly this is the type of choice that social theory inevitably must confront. We have elsewhere (Peters and Pierre, 1998) attempted to address some of the theoretical issues that the potentially static nature of institutions presents. The argument there was that the dichotomy between structure and agency was usually phrased much too starkly, especially by the critics of institutional theory. While structure may be seen as constraints on unfettered creativity of action by individuals also shape structures. The effects of the Clinton administration on the office of President of the United States is an obvious example. The office will never be the same, but the same could be said off the effects of a number of other presidents as well. These political structures have evolved along with the individuals who inhabit and animate them, and both institutions and individuals are shaped in the interaction. 6 The argument about structure and agency can also run the other way. Agency is not as unbridled as the advocates of individual agency as an explanatory device might have us believe. Few, if any, public sector actors can act without confronting powerful constraints from the institutions within which they operate. Individuals may have substantial charisma and other political resources but they can only be successful if the can channel those resources within known and effective institutions. There are the relatively few exceptions who have been able to revolutionize the political system but even then, in good Weberian fashion, that charisma does become institutionalized. 6 Grafstein argues that one of the central paradoxes of institutionalism is that institutions are the products of individual action but yet we assume that individuals somehow are powerless against institutions. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 11 The fundamental point being made here is that institutions and individuals are involved in an ongoing process of interaction that produces change and even replacement of existing institutions (see Peters, 1999, 48–50). It is appropriate to focus on the stability of institutions and their capacity to explain policies through time, but it is probably inappropriate to dwell on that stability and assume that institutions impose a mortmain on the possibility of change. What is now required, however, is some means of measuring and describing institutions, and the nature of their values and incentives, in order to be able to assess not only the level of institutionalization but also the likely effects whey will have on policy. 3. Measurement: What is An Institution? We argue above that measurement becomes crucial for addressing some of the problems of institutional theory. If we can sort out the conceptual confusion about what an institution actually is, we are still left with a very real problem about how to measure an institution. One obvious answer to this question is that each of the approaches implies a different form of measurement, so that there is no possible answer. The empirical institutionalists, for example, have been largely content with their simple dichotomy, although the political world is almost certainly more complex than that (Duverger, 1980; Lijphart, 1984); in some ways this is little changed from the “old institutionalists” who concentrated on formal-legal definitions of institutions. The historical institutionalists are responsible primarily for identifying what is the existing institutional arrangement, whether in policy or structural terms, so that this can be argued to persist or not. Even this measurement task, however, may not be so simple as it appears, given that we may be left with the familiar incrementalist problem of deciding how big a change is needed for there to be a change (Dempster and Wildavsky, 1980). Both the rational choice and the normative versions of institutional theory present more difficult measurement problems. The rational choice version requires identifying just what are constraints, imposed on, or the incentives available to, the participants in an institution. The notions of the payoffs available in, for example, game theory are generally hypothetical. Calvert (1995), for example, assigns some numbers to the payoff matrices in games of institutional design that are firmly anchored in thin air. There have been some interesting efforts to identify and calculate the incentives that motivate members of institutions in a rational choice framework (see Brehm and Gates, 1999), but these are relatively scarce. Alternatively, if the rules-based version of institutions is being used then there is the need to measure constraint, and with that the need to ask when is a rule really a rule? That is, tax laws in many Southern European countries may be rules but if they are not obeyed, and it is widely known and accepted that they are not obeyed, then are they really rules? Thus, the 12 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S means of measuring compliance are needed to determine if the institutions (conceptualized as rules) do in fact exist in anything other than a paper form. 7 This problem, in turn, makes the existence of an institution almost tautological, and hence any measurement of content or degree becomes almost meaningless. The normative conceptualization of institutions requires measuring the values that exist within the institution, and the extent to which they are common throughout the institution. A useful exemplar for this variety of measurement would be the existing literature on organizational cultures (Martin and Siehl, 1983; Martin and Frost, 1996). This literature has provided the intellectual means for cataloging organizational cultures (Hofstede, 1990) and also has examined the extent to which a common culture exists within an organization or institution. 8 Further, some literature has examined the ways in which organizational cultures are produced and reinforced in both the public and private sectors. The research required to measure institutions within the normative framework could be conducted with questionnaires, through more in-depth interviews, or by using organizational artifacts (training manuals, internal communications, etc.). It is, however, more than possible to assess the existence of a institutional culture and to typify it. Typifying that culture may involve examining the management content and/or at the policy aspects, especially for public sector institutions responsible for making and implementing particular types of policy. Both of these dimensions help to understand how the institution functions and how it relates to its external environment. The measurement question is in many ways a continuation of the discussion about levels of institutionalization presented above. That is, if we believe think we know what an institution is we can begin to develop a group of measures to capture the most important aspects of that social reality, but if we do not have a clear view of the concept then measurement is stifled. Assuming simply that an institution exists, or that it does not, does not appear to offer sufficient guidance for constructing social measurement schemes. On the other hand, beginning to develop and actually use measures of institutions, albeit preliminary and potentially flawed, may be necessary to amplify the meanings of the theories more completely. Although Giovanni Sartori (1991) and other scholars have long argued that measurement without adequate conceptualization is at best unlikely to yield useful results, there is something to be said for bashing ahead and at least attempting alternative ways of measuring an institution and its attributes. So long as the study of institutions remains at the extremely abstract level as it now stands then the multiple approaches all remain equally 7 There is a large literature on compliance in organizations, as well as about the manner in which citizens respond to public laws such a taxes. 8 One theoretical question we are not tackling here is the difference between institutions and organizations (See Zucker, 1987; Khalil, 1995; Peters, 2000). I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 13 valid. Some more detailed testing of propositions and assumptions may be required before we can fully clear away the conceptual swamp in which this area of the discipline often appears to be mired. The measurement question is especially important as we think about institutionalism as a basis for developing comparative political theory (assuming for the moment that there is any other kind). We have noted above that a good deal of institutional theory could be characterized as “variance theories”, meaning that the structural differences among institutions are assumed to predict differences in policy choices, or the stability of the governmental structures. If we are to move beyond simple dichotomies and to be able to make more precise predictions about performance then more precise measures will the necessary. For example, if we are interested in the impact of executive politics on policy and performance then we will need to go beyond the difference between presidential and parliamentary systems (see Goetz and Peters, 1999). 4. Independent or Dependent Variable, or Both? The general assumption in the literature has been that institutions are independent variables. In the empirical version of the approach there has been a good deal of analysis attempting to demonstrate that differences in institutions do make a difference in policy (Weaver and Rockman, 1995). The nature of political theory is indeed to provide explanations for a broad range of political phenomena, and much of the discussion in this paper has been about the explanatory capacity of institutional theory. We should, however, also think about institutions as dependent variables, and question what can explain the choice, or autonomous development, of an institution. This need to think about institutions in both ways is not dissimilar to some aspects of behavioral theory; attitudes, for example, are both independent and dependent variables. The purpose of thinking about institutions as dependent variables is linked with the question of institutional development and institutionalization mentioned above. In particular, are these structures inherently the product of an evolutionary process, or are they amenable to design and manipulation. Further, if they are the product of a more evolutionary process, then what explains that process? Associated with possibilities of thinking about institutions in political life is the need to conceptualize the relationship between the political environment and institutions. All versions of institutionalism we have discussed have a more or less central conception of institutions acting rather autonomously in making policy; the measures of institutionalization discussed above assume that autonomy is an essential element of institutionalization. On the other hand, however, in democratic systems institutions must also be closely connected with the 14 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S environment and respond to the legitimate pressures coming from outside. Thus, although internally institutionalization may be a positive virtue then . 5. Summary: What Does it Explain? How Does it Explain? We have now had a discussion of a number of issues in institutional theory, but we must return to the central question of whether institutional theory can function as an organizing approach for the discipline. Posing this question does not mean that it would necessarily be the only or even the dominant approach, but is there sufficient power and generality to think about institutionalism in these terms? In other words, can institutional analysis explain a range of phenomena in political life, and do so in a parsimonious manner? The answer to the above question is a definitive yes and no. On the one hand institutions can be discerned existing within a wide range of political settings, and performing a wide range of public tasks. Also, institutions appear volte face to be associated with differences in behavior of individuals and differences in decision-making outcomes. Institutions also help in reducing variance in political behavior and therefore help to improve the possibilities of prediction. Likewise, institutions are more readily identifiable (at least under most definitions of the term) so that they comprise a useful point at which to begin the analysis. The critics would argue, however, that even if institutional theory does constitute a good place at which to begin the analysis, they are not such a good place to end it. There are a number of problems in the theory itself, and in its ability to provide coherent explanations of political phenomena, that limit its utility as a central framework for the discipline. One of the more important problems is the difficulty of measuring institutions; we know they exist but how do they vary? Further, such explanations as may be available from institutional theory may be excessively static, and be incapable of coping with the dynamism and complexity of the contemporary political world. I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 15 6. References Alston, L. J., T. Eggerston and D. C. North (1996) Empirical Studies of Organizational Change (Cambridge: Cambridge University Press). Allison, G. T. (1971) Essence of Decision (Boston: Little, Brown). Almond, G. A. and G. B. Powell (1967) Comparative Politics: A Developmental Approach (Boston: Little. Brown). Brehm, J. and S. Gates (1999) Working, Shirking and Sabotage: Bureaucratic Response to a Democratic Public (Ann Arbor: University of Michigan Press). Brunsson, N. (1999) Standardization as Institutionalization, in M. Egeberg and P. Laegreid, eds., Organizing Institutions: Essays in Honour of Johan P. Olsen (Oslo; Scandinavian University Press). Brunsson, N. and J. P. Olsen (1993) The Reforming Organization (London: Routledge). Callinicos, A. (1987) Making History: Agency, Structure and Change in Social Theory (Cambridge: Polity). Calvert, R. L. (1995) The Rational Choice Theory of Institutions: Implications for Design, in D. Weimer, ed. Institutional Design (Dordrecht: Kluwer). Dempster, M.A.H and A. Wildavsky (1980) On Change: Or, There is No Magic Size for an Increment, Political Studies 27, 371–89. Dimaggio, P. and W. Powell (1991) The Iron Cage Revisited: Institutional Isomorphism and Collective Rationality in Organizational Fields, American Sociological Review 48, 147– 60. Duverger, M. (1980) A New Political System Model: Semi-Presidential Government, European Journal for Political Research 8, 165–87. Easton, D. (1970) A Systems Analysis of Political Analysis (New York: Wiley). Goetz, K. H. and B. G. Peters (1999) Institutional Theory and Political Executives: Creating Executive Organizations East and West, Paper presented at Conference on Institutional Theory, Ross Priory, Dunb., Scotland, October 18–19. 16 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S Goodin, R. E. (1995) The Theory of Institutional Design (Cambridge: Cambridge University Press). Grafstein, R. (1992) Institutional Realism (New Haven, CT: Yale University Press). Hall, P. A. and R. C. R. Taylor (1996) Political Science and the Three New Institutionalisms, Political Studies 44, 952–73. Hay, C. (1995) Structure and Agency, in D. Marsh and G. Stoker, eds., Theory and Method in Political Science (New York: St. Martin's). Horn, M. (1995) The Political Economy of Public Administration (Cambridge: Cambridge University Press). Huntington, S. P. (1965) Political Development and Political Decay, World Politics 17. Huntington, S. P. (1968) Political Order in Changing Societies (New Haven: Yale University Press). Immergut, E. (1998) The Theoretical Core of the New Institutionalism, Politics and Society 26, 5–34. Khalil, E. L. (1995) Organizations versus Institutions, Journal of Institutional and Theoretical Economics 151, 445–66. Kato, J. (1996) Review Article: Institutions and Rationality in Politics – Three Varieties of Neo-Institutionalists, British Journal of Political Science 26, 553–82. Katz, D. and R. L. Kahn (1978) The Social Psychology of Organizations, 2nd. ed. (New York: John Wiley). Keman, H. (1996) Political Institutions and Public Governance, in R. Czada, A. Hertier, and H. Keman, eds., Institutions and Political Choice: The Limits of Rationality (Amsterdam: VU Press). Kickert, W. J. M., E.-H. Klijn and J. F. M. Koopenjaan (1997) Managing Complex Networks: Strategies for the Public Sector (London: Sage). Knoke, D., F. U. Pappi and Y. Tsujinaka (1996) Comparing Policy Networks: Labor Politics in the U.S., Germany and Japan (Cambridge: Cambridge University Press). I H S — B. Guy Peters / Institutional Theory: Problems and Prospects — 17 Krasner, S. (1984) Approaches to the State: Alternative Conceptions and Historical Dynamics, Comparative Politics 16, 223–46. Krasner, S. (1988) Sovereignty: An Institutional Perspective, Comparative Political Studies 21, 66–94. March, J. G. and J. P. Olsen (1984) The New Institutionalism: Organizational Factors in Political Life, American Political Science Review 78, 738–49. March, J. G. and J. P. Olsen (1989) Rediscovering Institutions (New York: Free Press). March, J. G. and J. P. Olsen (1996) Democratic Governance (New York: Free Press). Martin, J. and P. Frost (1996) Organizational Culture War Games: A Struggle for Intellectual Dominance, in S. R. Clegg, C. Hardy and W. R. Nord, eds., Handbook of Organization Studies (London: Sage). Martin, J. and C. Siehl (1983) Organizational Culture and Counter-Culture: An Uneasy Synthesis, Organisational Dynamics 12: 52–64. Mohr, L. B. (1982) Explaining Organizational Behavior (San Francisco: Jossey-Bass). North, D. C. (1990) Institutions, Institutional Change, and Economic Performance (Cambridge: Cambridge University Press). Olsen, J. P. and B. G. Peters (1996) Introduction: Learning from Experience?, in Olsen and Peters, eds., Lessons from Experience: Experiential Learning from Administrative Reform in Eight Democracies (Oslo: Scandinavian University Press). Ostrom. E. (1990) Governing the Commons: The Evolution of Institutions of Collective Action (Cambridge: Cambridge University Press). Peters, B. G. (1999a) Institutional Theory: The “New Institutionalism” in Political Science (London: Cassells). Peters, B. G. (1999b) Institutional Theory and The Gore Commission: Alternative Views, in P. Laegreid and M. Egeberg, eds., Organizing Institutions: Essays in Honor of Johan P. Olsen (Oslo: Scandinavian University Press). Peters, B. G. (2000) Institutional Theory and Public Organizations, in G. Morand and P. Knopfel. eds., Essays in Memory of Raimund Germann 18 — B. Guy Peters / Institutional Theory: Problems and Prospects — I H S Peters and Pierre, (1998) Institutions and Time; Problems of Conceptualization and Explanation, Journal of Public Administration Research and Theory 8, 565–84. Polsby, N. (1968) The Institutionalization of the U.S. House of Representative, American Political Science Review 62, 144–68. Putnam, R. D., R. Leonardi and R. V. Nanetti (1993) Making Democracy Work: Civic Tradition in Modern Italy (Princeton, NJ: Princeton University Press). Ragsdale, L. and J. J. Theis, III (1997) The Institutionalization of the American Presidency 1924–92, American Journal of Political Science 41, 1280–1318. Rittberger, V. (1993) Regime Theory and International Relations (Oxford: Clarendon Press). Sartori, G. (1991) Comparing and Miscomparing, Journal of Theoretical Politics 3, 243–57. Selznick, P. (1957) Leadership in Administration (New York: Harper and Row). Scott, W. R. (1995) Institutions and Organizations (Thousand Oaks, CA: Sage). Steinmo, S., K. Thelen and F. Longstreth (1992) Structuring Politics: Historical Institutionalism in Comparative Analysis (Cambridge: Cambridge University Press). Verheijen, T. (1999) Civil Service Systems in Central and Eastern Europe (Cheltenham: Edward Elgar). Von Mettenheim, K. (1996) Presidential Institutions and Democratic Politics (Baltimore, MD: Johns Hopkins University Press). Weaver. R. K. and B. A. Rockman (1993) Do Institutions Matter?: Government Capabilities in the United States and Abroad (Washington, DC: The Brookings Institution). Weimer, D. L. (1995) Institutional Design (Dordrecht: Kluwer). Zucker, L. (1977) The Role of Institutionalization in Cultural Persistence, American Sociological Review 42, 726–43. Zucker. L. (1987) Institutional Theories of Organizations, Annual Review of Sociology 13, 443–64. Author: B. Guy Peters Title: Institutional Theory: Problems and Prospects Reihe Politikwissenschaft / Political Science Series 69 Editor: Peter Biegelbauer Associate Editor: Gertrud Hafner ISSN: 1605-8003 © 2000 by the Department of Political Science, Institute for Advanced Studies (IHS), Stumpergasse 56, A-1060 Vienna • ( +43 1 59991-0 • Fax +43 1 5970635 • http://www.ihs.ac.at ISSN: 1605-8003