Historical institutionalism is a research tradition dedicated to the study of the

Transcription

1 Historical Institutionalism in Political Science Orfeo Fioretos, Tulia G. Falleti and Adam Sheingate 1 May 2013 I. Introduction Historical institutionalism is a research tradition dedicated to the study of the origins, evolution, and consequences of political institutions from the local to the global level. Growing out of an institutional turn in the social sciences, historical institutionalism is distinguished by a theoretical commitment to the temporal dimensions of politics: how the timing and sequence of past events generate lasting legacies that shape the scope, character, and consequences of governing authority. Understood as the rules, norms, and practices that organize and constitute social relations, institutions are at the center of this research tradition for their role in creating constraints and opportunities for political action, in distributing political power, and in shaping political preferences over time. Attention to the temporal character of institutions is crucial if political scientists wish to answer questions such as why policies take their particular form, why institutions privilege certain outcomes, or why inequalities endure. The scope and sophistication of historical institutionalism has grown over the past twenty-five years, extending throughout the discipline of Political Science. In 1 Draft introduction to the Oxford Handbook of Historical Institutionalism, edited by Orfeo Fioretos, Tulia G. Falleti, and Adam Sheingate (New York and London: Oxford University Press, forthcoming 2015). The introduction is the product of a collaborative partnership that began in The editors thank Dominic Byatt of OUP for his early and enthusiastic support of the project. 1

2 comparative politics, historical institutionalism has been particularly influential and shapes research agendas in virtually every substantive area, from research on the modern state, capitalism, law, and economic development to the study of political regimes, political parties, organized societal actors, and public policy. Historical institutionalism is central to the study of American political development, focusing on the elusive character of the American state and the legacy of struggles over race and citizenship that animate much of U.S. politics. In the area of European politics, historical institutionalism is central to the study of political parties, the power of business, the attributes of welfare states, and the process of European integration. Within international relations, historical institutionalism informs seminal contributions on state sovereignty and foreign economic policy, as well as research in international security, political economy, law, and global governance. With its growing empirical reach, historical institutionalism developed a distinct conceptual toolbox for understanding the causal mechanisms that underpin processes of institutional durability and change. Its theoretical range extends along several continuums, such as from materialist accounts of institutional politics to explorations of the role of ideas in preference formation, and from structural explanations of political outcomes to narratives that highlight the transformative capacity of human agency. Historical institutionalism has contributed to vibrant methodological debates as well, sitting at the forefront of multi-method research. In particular, historical institutionalism has fostered a productive exchange between scholars who embrace deeply historiographical methods and scholars who draw more explicitly on the public choice tradition as well as large-n statistical studies. Finally, historical institutionalism has 2

3 forged links with allied disciplines and facilitated exchange between Political Science, History, Economics, and Sociology, among others. As the world again struggles to understand the short and long-term effects of unexpected events like financial crises, social revolutions, and redistributions of global power, historical institutionalism is poised to make new contributions. This volume takes stock of the accomplishments of historical institutionalism and identifies promising new areas of research. Chapters explore how historical institutionalism has revisited conventional wisdoms, resolved long-standing puzzles, and opened new areas of inquiry. They discuss historical institutionalism s contributions to the study of politics, areas where historical institutionalism complements other work in institutional analysis, and the extent to which the tradition itself has responded to criticisms directed its way. In this introduction we set the stage for the analyses, arguments, and assessments that follow. We begin by detailing the origin and crystallization of historical institutionalism, discuss its analytical core, survey some of its empirical findings, and assess its promise for addressing both enduring and new debates in Political Science. II. The Emergence and Crystallization of Historical Institutionalism Historical institutionalism has deep roots in Political Science, going back to classics in political economy that traced the emergence and development of capitalism and democracy (Polanyi 1944, Gerschenkron 1962, Moore 1966). Early studies paid close attention to the role of timing and sequence in shaping the diverse trajectories of nation-states. In the late 1970s and 1980s, as efforts to reinvigorate the state as an object of study dovetailed with a renewed interest in institutions, scholars developed a 3

4 conceptually more precise understanding of the causal impact of history and institutions for political life. By the early 1990s, historical institutionalism crystalized into a distinctive tradition of institutional analysis that addressed an expanding array of topics and problems in Political Science. As cruder variants of behavioralism, pluralism, and Marxism were called into question during the 1970s and 1980s for treating formal arrangements of political authority simply as arenas within which actors and groups competed for power and as epiphenomenal of economic relations, Political Science experienced a new institutional turn. Scholars began to highlight the causal role of formal rules in constituting actors and constraining their behavior, and pointed to the role of institutions in ordering political life through regulative, normative, and cognitive mechanisms. There emerged, however, important differences in how scholars studied institutions and in the claims they made about their impact (March and Olsen 1984; Hall and Taylor 1998; Immergut 1998). For scholars working in a rationalist tradition, institutions induced stability by limiting the range of alternatives actors confront. Scholars working in a sociological tradition emphasized the normative and cognitive dimensions of institutions, such as logics of appropriateness that dictated roles and scripts followed by actors. While sharing some of these theoretical commitments, scholars developing the historical institutional variety placed particular emphasis on the temporal dimension of politics, or on how the distributional effects of institutional changes and continuities shape political behavior and outcomes over time. The new wave of historical institutional scholars built considerably upon previous work that examined the emergence of the modern state and its significance for 4

5 understanding contemporary politics. What the newer generation of scholars accomplished was to foreground the state as an historical construct of time and place. Thus, Nettl s (1968) early discussion of the state as a conceptual variable anticipated and informed later work in comparative politics that sought to bring the state back in (Evans et al, 1985), helped launch the study of American political development focused on the emergence of a new American state (Skowronek 1982), and stimulated scholarship in international relations that highlighted how historical variations in statesociety relations shaped foreign economic policies (e.g. Krasner 1976; Katzenstein 1978; Ikenberry 1988). A central theme running through much of the early work on the state was that it was a partly autonomous actor working above and through society, rather than simply being a reflection of a pluralist process (Nordlinger 1981). Specifically, the degree to which states approximated the Weberian ideal of a rational, modern bureaucracy was highlighted as a source behind diverse policy outcomes within and across states (Skowronek 1982, Skocpol and Finegold 1982). In making an ontological claim about the status of the state as an object of inquiry in Political Science, these scholars also made a theoretical claim (sometimes implicitly) about the historical processes that shaped the nature of the state, its formation over time, and its capacities. For example, Theda Skocpol s (1979) early study of social revolutions explored how the timing and nature of peasant revolts influenced patterns of regime breakdown and shaped the possibilities for post-revolutionary state-building. From this point of departure, scholars began to pay attention to the timing of events, patterns of political development, and the legacy effects of political structures created in the past. 5

6 As scholars deepened their study of the state and gradually expanded their empirical focus beyond it, they began to write more self-consciously about the temporal dimensions of politics. The 1992 publication of Structuring Politics is an important turning point in this regard. More than simply coining a term, the contributors to the volume executed an important shift in emphasis from a historically oriented focus on institutions to an explicitly historical institutionalism. 2 This required staking out a set of fundamental and foundational claims about the operation of institutions and their effects on preference formation, coalition building, policy evolution, and political dynamics (Steinmo and Thelen 1992). Put another way, from a focus on patterns in state formation across time and space, scholars set to work developing an analytical toolbox for the diachronic study of politics and the elaboration of claims about processes of institutional creation, reproduction, and change. An important development contributing to the crystallization of historical institutionalism was growing skepticism with the way rational choice institutionalism took individual preferences as given. Thelen and Steinmo noted in their introduction to Structuring Politics that one, perhaps the, core difference between rational choice institutionalism and historical institutionalism lies in the question of preference formation (1992, 9, original emphasis). They argued that individual preferences are not given and constant, but endogenous to historical processes that distribute resources and structure power through institutions. Thelen and Steinmo further argued that if 2 Steinmo and Thelen note that they borrow the term historical institutionalism from Theda Skocpol, to distinguish this variant of institutionalism from the alternative, rational choice variant (1992, 28, n. 4). Steinmo dates the actual coining of historical institutionalism to 1989: The term came out of a small workshop held in Boulder, Colorado in January 1989 (2008, 136, n. 1). 6

7 institutions could strengthen, weaken, or transform individual s preferences and goals, they could also alter the structure of political coalitions with potentially significant effects for policy outcomes. In other words, articulating what role institutions had in shaping preferences held promise for answering why particular political coalitions formed, when these would change and why, and how the behavior of coalition members would impact changes or continuities in institutions themselves over time. Skeptical of behavioral and pluralist models which assumed that government choices reflected general population-wide preferences in relatively consistent fashion across institutionally diverse contexts, historical institutionalists gave particular attention to organizations. Industry, bank, and employers associations, unions, environmental and other advocacy networks, as well as administrative bureaus of the state, were highlighted for their role in representing and mobilizing support for particular and particularistic policy preferences. In a spirited defense of a focus on organizations as structures that aggregate the endeavor of many individuals, Peter Hall also stressed their importance as suppliers of information and vendors of interpretation, their role in the implementation of public policy, as well as their potential role in ultimately alter[ing] the preferences of distinct groups (1986, 233). Organizations themselves were therefore studied as they were embedded within formal institutions, principally political and economic ones, for as Steinmo (1989, 502) noted: [n]either interests nor value have substantive meaning if abstracted from the institutional context in which humans define them. In theorizing the origin of preferences, a group of historical institutionalists highlighted the generative role of ideas in shaping interests and goals. In their contributions to Structuring Politics, Peter Hall, Desmond King, and Margaret Weir 7

8 explored the intersection of ideas and institutions, and identified conditions under which new and old, local and international ideas were reflected in the policy and institutional choices of countries. Pointing to the role of ideas in shaping institutions and interests, this literature grew quickly as studies addressed long-standing puzzles in Political Science, such as why common founding myths of citizenship generated diverse identities (Smith 1997), why economic openness persisted despite demands for closure (Goldstein 1994), and why states extended significant governing authority to international organizations (Ikenberry 1992). 3 A Conceptual Core As historical institutionalists re-conceptualized the process of preference formation with reference to institutional contexts and stressed the role of organizations and ideas in shaping interests and coalitions, they developed a theoretical and methodological toolkit to study temporal effects in politics. Whereas this toolkit for the study of diachronic processes is large, we focus here on critical junctures, path dependence, intercurrence, and modes of gradual institutional change as four analytically related concepts that provide a window into how historical institutionalists understand the causal mechanisms behind patters of institutional creation, reproduction, and, more recently, change. 3 Some founding voices of the tradition expressed strong support for the focus on organizations, but early skepticism of the literature on ideas. Skocpol noted, for example: I do not think that institutions are simply or primarily systems of meaning or normative frameworks. Group identities for me are grounded on organizational linkages, access to resources, and some sense of success over time in political undertakings (1995, 105). 8

9 In Shaping the Political Arena, Collier and Collier made a significant contribution to the historical institutionalist tradition by highlighting the causal effects of critical junctures. Defining a critical juncture as a period of significant change, which typically occurs in distinct ways in different countries (or in other units of analysis) and which is hypothesized to produce distinct legacies (1991, 29), Collier and Collier argued that variation in the unfolding of critical junctures across contexts holds the key to explaining divergent political legacies and outcomes. They stressed the importance of specifying the duration of the critical juncture as well as the effecting historical legacies (1991, 31-34) and highlighted that the timing of the critical juncture, in relation to other developments, was consequential to subsequent politics. Unlike other types of historical causes, they maintained that critical junctures generate legacies that can reproduce themselves without the enduring presence or recurrence of their causes. In the language that would quickly take root, critical junctures marked the beginning of path-dependent processes. Scholars have debated the extent to which critical junctures themselves can be explained by reference to institutions or other antecedent causes (Pierson and Skocpol 2001; Slater and Simmons 2010), and the degree of agency that stems from these critical moments. Considering the agency effects of critical junctures, for example, Capoccia and Kelemen (2007, 348) argue that critical junctures should be understood as periods of time that are significantly shorter than the path-dependent processes stemming from them. If critical juncture periods are too long, they note, the substantial influence of agency that is expected in these periods will be constrained (again) by reemerging institutional constraints. Theirs is thus a plea to refine the critical juncture framework by paying greater attention to the permissive conditions behind the opening of a specific juncture, 9

10 which is mainly about how previous mechanism of reproduction are upended and agency is given room to create new or modify existing institutions. To this plea, Soifer (2012) adds one for greater attention to the distinction between the permissive and productive conditions of critical junctures, or between the conditions that initially loosen the constraints on agency and those that ultimately close the room for it. Critical junctures are important in the historical institutionalist tradition because they may be initial markers of path-dependent processes. After the openness of the critical juncture moment, which enables relatively free agency, a process or sequence of events ensues in which institutions exert their causal force. However, some historical institutionalist scholars have stressed contingency (rather than critical junctures) in the origins of path-dependent processes. They argue that the early events that trigger pathdependent processes must be accidental ones that were neither anticipated nor intended. Understood as a stochastic reality in which parts cannot be explained by available theories, the contingency criterion is highlighted for it enables explanations of how apparently random, accidental, and small events have major consequences over time (Mahoney 2000; Mahoney and Schensul 2006, 461). 4 Whether it is agency in the context of disrupting critical junctures such as revolutions, wars, and economic crises, or it is contingency that marks the beginning of path dependent processes, what is unique in such processes is not that time matters, as in the general statement early events affect later ones, but that the causal impact of early events is significantly stronger than that of subsequent events. It is as if the early events 4 Paradigmatic examples of stochastic events include the adoption of the QWERTY keyboard (David 1985) and the first step in Polya urn processes (Arthur 1994). 10

11 of the process would set its tone with greater strength than later ones. Pierson describes path dependence as dynamic processes involving positive feedback. For this reason, he claims attention to the sequence in which developments occur is crucial in determining why one outcome among multiple possible outcomes prevails (2004, 20). In other words, not only what event happens in a process or sequence of events, but when it happens in relation to the other events in the sequence, is highly consequential. Thus, for example Falleti (2010) finds that the order in which different types of decentralization (administrative, fiscal, or political) unfolded over time after the dismissal of the developmental state reshaped intergovernmental balances of power. This outcome was due to early reforms that produced positive feedback effects which strengthened the bargaining power of those territorial actors who set in motion the decentralization process. Attention to causal mechanisms such as positive feedback effects that link the constitutive events of a sequence has helped historical institutionalists specify why and how constituent events are connected. Although there is debate about the precise definition of causal mechanism, historical institutionalists agree that such mechanisms ought neither be reduced to intervening variables as is the case with other theoretical and methodological traditions (King et al 1994, 85-87), nor to the measurement of the presence or degree of a certain attribute. 5 Instead, they emphasize the mechanism by which causes and effects are linked. Thus, in Falleti and Lynch s definition causal mechanisms are relatively abstract concepts or patterns of action that can travel from one specific instance [ ] of causation to another and that explain how [and why] a 5 It is worth noting that Mahoney (2001: ) identified twenty-four definitions of causal mechanisms in the social sciences literature. 11

12 hypothesized cause creates a particular outcome in a given context (2009, 1145) Understood this way, causal mechanisms identify relationships (or the micro-level processes) connecting conditions and outcomes such as in the processes of learning, competition (Pierson 2004, 39-41, 124-9), or boundary control (Gibson 2005, 2012). Attention to causal mechanisms has also helped scholars distinguish between different types of path-dependent processes. Making a distinction between selfreinforcing and reactive mechanisms, Mahoney (2000) introduces the notion of pathdependent processes of change or reactive sequences. For Mahoney, reactive sequences are characterized by mechanisms of reaction and counter-reaction that originate in a contingent breakpoint. To the fairly linear narrative of self-reinforcing mechanisms within a singular sequence of events, Mahoney adds the complexity of intersecting sequences that result in contingent breakpoints, which in turn produce causally connected reactive and counter-reactive events that lead to the outcome of interest (2000, ). The extensive attention to path-dependent processes often leads to characterizations of historical institutionalism as a tradition devoted particularly strongly to the study of institutional stability. Yet, at least since Structuring Politics underscored the importance of examining the politics of institutional dynamism, change has been embedded in the historical institutionalist approach. Early on, for example, Karen Orren and Stephen Skowronek (1994) encouraged scholars to move beyond the iconography of order. Using the term intercurrence to represent the complexity of multiple overlapping processes of institutional creation, reproduction, and change, they pointed to the layered structure of institutional action and stressed that institutions created at different times generate mosaics of authority. More precisely, Orren and Skowronek argued that the 12

13 non-simultaneity of institutional creation produces an accumulation of competing controls within institutions of government, such that the normal condition of the polity will be that of multiple, incongruous authorities operating simultaneously (Orren and Skowronek 2004, 108). From this perspective, institutions that are sometimes ill-fitted to one another or governed according to contradictory imperatives give rise to contestation and conflict between ordering mechanisms, enabling actors to exploit tensions and contradictions to promote new forms of power and authority. For this reason, politics should be understood and studied as the governance, not of single but overlapping sources of authority. The shift from an emphasis on critical junctures followed by periods of relative institutional stability to an emphasis on the layered nature of institutional orders and competition among authorities characterizes a rapidly growing literature on gradual institutional change. Pointing to the layered quality of institutions and the varying levels of discretion they confer to individuals to interpret and enforce rules, this literature brings attention to the causal mechanisms that produce variations in patterns of incremental change. Building on their own (Thelen 2004; Streeck and Thelen 2005) and related historical institutionalist scholarship (Hacker 2005, Schickler 2001), Thelen and Mahoney (2010b) give particular prominence to four modes of gradual institutional change: displacement or the removal of existing rules and the introduction of new ones; layering or the introduction of new rules on top of or alongside existing ones; drift or the changed impact of existing rules due to shifts in the environment; and conversion or the changed enactment of existing rules due to their strategic redeployment (2010, 15-22). 13

14 The literature on gradual institutional change represents a new interpretation of the origin and causes of institutional innovation and development. From this perspective, critical junctures that generate new institutions that can reproduce themselves for long periods are not the only, not even the main, determinants of institutional development. Though the emphasis falls on incremental patterns of change, this literature underscores that the cumulative effects of prolonged periods of gradual change can have transformative effects on institutions. Moreover, the new literature notes that piecemeal change can lead to the adaptation of old rules to new environments (as in processes of drift and conversion), as well as to the adoption of new rules that sideline or operate on top of old ones (as in processes of displacement or layering). Identifying the conditions under which these processes are manifest and what their cumulative consequences are over time has become central in studies of modern polities at the local, national, as well as international levels (Mahoney and Thelen 2010a; Moshella and Tsingou 2013). This is a promising program of research that may provide a new approach to longstanding questions of historical institutionalism, such as those about the origins of institutional order versus flux and the causal weight of structure and agency in institutional development. Empirical Contributions In the past two decades, scholars have deployed and developed historical institutionalism s analytical toolbox and used it across a wide empirical terrain. In comparative politics, variations and developments in the character of the modern state has been an intense area of research (e.g., Kohli 2004). Critical junctures and institutional 14

15 legacies have been central to the study of the welfare state (e.g., Esping-Andersen 1990, Lynch 2006). Historical institutionalism continues to inform a dynamic research agenda on state and non-state actors social welfare provision (e.g., Huber and Stephens 2001, Cammett and McLean 2011). The scholarship on labor has continued to identify the legacies of institutional arrangements on unions and industrial relations (Swenson 2002; Caraway 2008). The study of political regime emergence and change, of democratization processes as well as of the origins and endurance of authoritarian regimes, has also been fertile ground for historical institutional accounts (e.g., Bellin 2002; Gibson 2012; Grzymala-Busse 2002; Levitsky and Way 2010). Within the field of American politics, the concepts of path dependence and feedback effects figure prominently in explanations for the distinctive character of U.S. social policy, such as the heavy reliance on private, arms-length instruments for the provision of health care and other government benefits (Hacker 2002, Howard 1999, Morgan and Campbell 2011). The study of American politics has also been central to the development and elaboration of theories of gradual institutional change such as layering, conversion, and drift (Schickler 2001, Hacker 2005, Staszak 2010). Finally, attention to the intercurrent character of institutional arrangements is also a common feature of work in American politics, perhaps because of the fragmentation of the American political system (Orren and Skowronek 2004). Partial, overlapping patterns of authority illuminate the contradictory tendencies and impulses in American politics, such as the coexistence of surprisingly robust anti-discrimination policies alongside a sprawling criminal justice system that disproportionately incarcerates racial minorities (Lieberman 2008, Frymer 2008, Gottschalk 2006, King and Smith 2005). 15

16 Historical institutionalism features widely in the study of European politics. It is particularly central to the large literature on the welfare state and models of capitalism, which entail much attention to positive feedback effects across a broad spatial, policy and historical scope (e.g. Hall and Soskice 2001; Culpepper 2005; Deeg 2005; Lynch 2006). Studies of democracy, likewise, have engaged the temporal sensibilities of historical institutionalism and document the role of ideas and ideology in shaping the evolution of diverse forms of political governance (e.g. Berman 2006; Capoccia and Ziblatt 2010). From a small number of early accounts of how policy and institutional legacies shaped European integration (e.g. Pierson 1996; Leibfried and Pierson 1995), historical institutionalism informs a growing and diverse set of studies examining novel supranational and intergovernmental institutions. Exploring the consequences of critical junctures, positive feedback effects, and the sources of a wide variety of incremental change, this literature documents institutional continuities and changes in supranational representation, intergovernmental executive powers, courts, regulatory systems and more (e.g. Meuiner and McNamara 2007; Büthe 2012; Kelemen 2004; Thatcher and Coen 2008; Posner 2009). Critical junctures and path dependence also figure in the international relations literature. Studies account for the relative stability, yet progressive nature, of the modern international system and structures of global governance with reference to concepts central to historical institutionalism (Ikenberry 2001; Fioretos 2011). Historical conjunctures and positive feedback effects have been central in shaping the evolution and varied nature of territorial and sovereign states (Spruyt 1994; Krasner 1995/96). Similarly, the effects of sequencing in shaping states security and economic doctrines 16

17 has been explored (Solingen 2007; Farrell and Newman 2010), along with the role of a mixture of largely immutable and rapidly evolving national political and economic institutions in shaping the governance of international trade and monetary institutions (Barton, Goldstein, Josling, and Steinberg 2010; Helleiner 2010). The structured and creative agency of national and international activists has also been studied, including in shaping the evolution of human rights norms, the expansion of international courts, and in bringing about novel forms of transnational regulation (Finnemore 2003; Alter 2010; Büthe and Mattli 2011; Newman 2008). III. The Future of Historical Institutionalism The consolidation of historical institutionalism into a conceptually sophisticated and empirically wide-ranging tradition does not mean there is uniformity in the perspectives of its practitioners, that internal debates have ended, or that the tradition s empirical footprint is equally strong across areas of research. Indeed, there is ongoing debate among scholars about the degree of dynamism within institutions, the role of actors in institutional accounts, and the relative weight of material and cognitive factors in the formation of preferences and the explanation of outcomes. Such debate is a mark of a lively research agenda and demonstrates that the development of historicalinstitutionalism continues. As scholars continue to puzzle through these core theoretical questions, opportunities exist to adopt new analytical techniques and methods, integrate historical institutionalism more closely with related social science disciplines, and expand the empirical scope of research within Political Science. 17

18 Differences and Debates Historical institutionalism is characterized by ongoing debates about questions of durability and change. Whereas some scholars focus on the patterned regularities of institutional order, others emphasize processes of innovation, evolution, and transformation within complex institutions. At the center of this debate is to what extent regular and ongoing configurations of social interaction define political life and to what extent institutions are characterized by varying degrees of flux. On the one hand, order is produced and reproduced over time through a deepening of institutional mastery (e.g. learning) and a thickening of actors and organizations (Pierson 2000). On the other hand, the imperfect reproduction of rules and behavior can yield a continuous process of incremental change as the cumulative effects of subtle alterations gradually transform the character of an institution (Mahoney and Thelen 2010a). Of course, both processes may be at work; some components of an institution are more durable than others. Consequently, debates about order and change sometimes pose a false choice: institutions are never completely static, nor is everything up for grabs. The challenge is to differentiate between a relatively stable core of an institution from the periphery of rules and practices that may prove more malleable or subject to interpretation (Streeck and Thelen 2005, Streeck 2009). Questions about institutional order and flux partly map on to differences between accounts that focus on the architecture of institutions versus the agency of actors within them. Whereas some scholars emphasize how institutions structure action through regulative, normative, and cognitive constraints, others underscore the way actors exploit the ambiguity of institutions through creative action, innovation, and entrepreneurship 18

19 (Sheingate 2003). A broad continuum of work in historical institutionalism ranges from the heavily structural to the more radically agentic. For instance, Richard Bensel s (2000) work on the political economy of American industrialization in the late nineteenth century examines the correspondence or fit between the interests of capital and the institutions of government; individuals do not figure prominently in his account. On the other end of the spectrum, Gary Herrigel s (2010) work on the political economy of late capitalism analyzes institutions as a product of creative problem solving. Institutions, as Berk and Galvan (2009, 552) write, are not constraints on action, they [are] made through action. Some worry that too much emphasis on unleashed actors renders historicalinstitutionalism an auto-dissolving paradigm that says little about institutions at all (Hall 2010). However, careful attention to the interplay between structure and action opens room to identify how institutions shape actors preferences and how individuals act through the institutions they inhabit. As Ira Katznelson notes, by adopting a relational epistemology that sees particular clusters of preferences, interests, and identities not just as causes; but as causes as well as products, historical institutionalism often cross[es] the divide between structure and agency without eliminate[ing] the heuristic distinction between the two (Katznelson 1997, 104). Avoiding the voluntarism characteristic of rationalist approaches, historical institutionalism understands individuallevel phenomena as shaped by contextual factors rather than abstracted from it. Preferences are not fixed or exogenously determined; instead, historical-institutionalism emphasizes how the institutional context in which actors operate can shape and transform individual preferences. In sum, historical institutionalism emphasizes the micro-politics 19

20 of situated action: how individual-level phenomena are embedded within a diverse and evolving institutional environment. Focusing on the relationship between individuals and institutions points to a third area of productive debate concerning the role of human cognition versus the effects of material resources in the process of institutional emergence, as a source of durability, and as a catalyst for change. On the cognitive side, scholars emphasize the role of ideas as mental maps for interpreting the world and as rhetorical weapons deployed in the struggle over policy (Blyth 2002). From this perspective, ideas breathe life into institutions; they are the basic building blocks for communication, coordination, and persuasion. On the material side, scholars emphasize how institutions mediate the relationship between economic resources and political power. Hacker and Pierson s (2010) account of winner-take-all politics in the United States illustrates how a particular configuration of institutions and partisan politics has skewed public policies toward the interests of the privileged few. Much historical institutionalism scholarship successfully blends cognitive and material arguments. For instance, Hacker and Pierson describe how conservative think tanks turned opinions once considered radical into mainstream views about the virtues of unfettered markets and how economic and political elites deployed ideas strategically to further the material interests of the super-rich. More broadly, the logic of feedback effects emphasizes how institutions have cognitive and material effects as actors adjust expectations and organize interests to defend the status quo. To provide one example from U.S. social policy, Social Security not only reinforces the idea that seniors deserve public pensions; the very existence of the program gives seniors the material resources 20

21 they need to defend against retrenchment (Campbell 2003). It is not a question of ideas or interests in the study of politics; rather, the contemporary condition is marked by the interaction between neoliberal ideas about states and markets and the material endowments of actors and groups (Hall and Lamont 2013). New Opportunities Combining structural, agentic, material, and cognitive approaches in new and different ways presents opportunities to expand the methodological scope and empirical reach of historical institutionalism in the years ahead. For instance, evolutionary approaches and works incorporating insights from behavioral and cognitive psychology suggest how individual cognition is a source of innovation and institutional reproduction. In the case of evolutionary approaches, individuals mental schemas and cognitive shortcuts can produce innovations that, if successful, are replicated over time (Lewis and Steinmo 2012, Lustick 2011). Meanwhile, work drawing from psychology examines how the difficulties actors face in making inter-temporal tradeoffs (e.g. short-term pain for long term gain) can lead to choices that reinforce the status quo (Jacobs 2011). In rethinking the role of ideas by bringing attention to the role of individual cognition, this work can generate new insights into the origins of institutions, the process of reproduction, and the possibilities for change. Historical institutionalism is also well-placed to rethink materialism and politics. Recent work on the evolution of American political parties explores how investments in material resources such as computer technology and information databases created institutional endowments that expanded the role and function of party organizations in the 21

22 political system (Galvin 2012). The phenomenon of institutional drift suggests further how material forces can transform institutions without any underlying change in their rules or operation (Hacker 2005, Lynch 2006). For instance, the gradual transformation of the U.S. labor market, especially the rise of the low-wage Wal-Mart economy, weakened the political role of organized labor and, today, limits the political possibilities of the Democratic Party (Warren 2010). Indeed, President Barack Obama s scope of action depends on the institutional consequences of the underlying economic conditions he confronts. More broadly, the gradual erosion of social protections in the United States increases inequality and, simultaneously, reduces the political efficacy of the poor. In this way, disparities in material resources diminish democracy by amplifying the political voice of some citizens and groups at the expense of others (Hacker, Mettler, and Soss 2007). Yet, as research on the effects of democratic institutions suggests, the empowering of previously marginalized actors can set larger processes of institutional change in motion (Smith 2009). In a very different vein, historical institutionalism could pay closer attention to the political consequences of viruses, weather patterns, and other self-organizing complex systems that exhibit a kind of material agency as they act on and through individuals and institutions (Coole and Frost 2010). More than simply a source of exogenous shock, a pandemic flu, global warming, or financial panic influences the operation of institutions and how individuals understand the world. To provide one example of this rich promise, Evan Lieberman (2009) uses government responses to HIV/AIDS to illuminate the interaction between the virus, ideas about ethnic identity or status, and the boundary 22

23 institutions that reinforce ethnic identities. A virus can change who we are, how we relate to others, and how we organize our institutions. Fully grasping the implications of agency, cognition, and materiality carries methodological considerations as well. For good reason, most historical institutional scholarship relies on archival research methods and empirically rich narratives to capture the temporal unfolding of institutional processes. Increasingly, historical institutionalism has incorporated mixed methods that combine large- and small-n approaches in ways that improve the precision of case selection and causal inference (Lieberman 2005). In addition, a growing body of work successfully employs behavioral methods of survey research and qualitative ethnographic studies in order to trace policy feedback effects and other institutional legacies on political participation, citizenship, and social relations (Campbell 2003, Mettler and Soss 2004, Morris-MacLean 2010). The recent turn toward evolutionary approaches noted previously points to other methodological extensions. According to Ian Lustick (2011), there is a clear but underappreciated affinity between historical institutionalism and agent-based approaches that use computer simulations of actors operating under various conditions and constraints in order to model how repeated interactions give rise to routinized behaviors. Using the tools of agent-based modeling, scholars can probe how actors decisions produce, reproduce, and transform institutional arrangements over time (Lewis and Steinmo 2012). Similarly, network analysis offers a way to incorporate insights from evolutionary theory by examining how relationships among actors are generative of social structures. For instance, Farrell and Shalizi (2012) offer an account of Athenian democracy as an evolutionary learning process in which the character of social networks, namely the 23

24 degree of similarity or difference among actors and the nature and number of links between them, influenced the diffusion of certain norms and behaviors that eventually gave rise to democratic institutions. Applications of network analysis are not limited to explicitly evolutionary accounts. For instance, Carpenter s (2001) work on the nineteenth century American state shows how agencies in the federal bureaucracy expanded their authority and nurtured a degree of autonomy from elected officials by constructing overlapping support networks composed of multiple civil society groups. This element of Carpenter s argument draws on earlier work by Padgett and Ansell (1993) who showed how Cosimo de Medici s political power depended on a robust network of overlapping financial and kinship ties. There is, as sociologist Roger Gould (2003) noted, a strong elective affinity between network analysis and historically-oriented social science as it combines careful archival work with a more formal analysis of social structures. As historical institutionalism displays greater methodological pluralism, further opportunities arise to connect the field with allied disciplines. Of course, historical institutionalism has always been interdisciplinary. The study of American political development operates at the intersection of Political Science and History to understand the manifold character of the American state and the evolution of governing authority from the nineteenth century to the present (Balogh 2009; Novak 2008). Similarly, economists (and economic historians) have contributed a great deal to the study of institutions, especially the relationship between institutional design and economic growth (North 1990). 24

25 However, it is the field of Sociology where the study of institutions is perhaps closest to historical institutionalism, yet the full range of opportunities for collaborative engagement has yet been tapped. In part, this is because sociologists sometimes employ a different vocabulary to describe institutions or their temporal effects (for helpful reviews see Campbell 2009 and Clemens 2007). For instance, Lawrence and Suddaby s (2006) concept of institutional work examines the varieties of situated action that contribute to the creation, maintenance, and disruption of institutions. Similarly, Fligstein and McAdam (2011) examine how skilled social actors construct and transform strategic action fields, meso-level social orders such as organizations, clans, supply chains, or governments in which actors with various skill endowments challenge and defend their position. Finally, Padgett and Powell (2012) draw on concepts from biochemistry like autocatalysis to examine the process of institutional emergence, focusing on how dynamic relations between diverse networks of actors produce repeated social interactions that become self-sustaining institutions and organizations. Scholars of historical institutionalism will develop a richer grasp of social structure and the dynamics of organizational forms by paying closer attention to the way sociologists approach questions of institutional origin, reproduction, and decay. Finally, historical institutionalism can extend its empirical reach in the future by straddling the traditional divides in our discipline. In the field of political economy, for example, research spanning international relations and comparative politics examines the changes and continuities in the regulation of modern market economies (e.g. Culpepper 2011; Hall 2011), including how old ideas and historical legacies have weighed heavily on government policies and global regulatory responses to the Great Recession (Blyth 25

26 2013; Farrell and Newman 2010; Büthe and Mattli 2010). Other areas where historical institutionalism can contribute to deeper, cross-field engagement is in the political and economic consequences of informal institutions (Helmke and Levitsky 2006; Tsai 2006) and the form and variety of authoritarian transitions (Slater 2010, Levitsky and Way 2010), including those in nominally democratic regimes such as the United States (Mickey 2013). In a world of unraveling dictatorships, systemic crisis in the financial system, widening inequalities, and the redistribution of global power, historical institutionalism is well-placed to expand its empirical scope even further. IV. Conclusion Within a generation, historical institutionalism has become a large and diverse analytical and empirical research tradition in Political Science that has tackled core puzzles in the discipline, reinvigorated the study of institutions and history, and developed new areas of research. It is firmly established in areas of research within Comparative, American, European, and International Politics. It is both empirically rich and analytically sophisticated, eschewing convenient trade-offs between these two sides of the social science coin. The chapters that follow very amply demonstrate the growth and vibrancy of historical institutionalism since it crystallized a generation ago. Indeed, and somewhat ironically, the strongest testament to its growth is the reality that not even a comprehensive volume such as this one can fully cover all relevant analytical and empirical developments in historical institutionalism. Nevertheless, it provides broad and deep coverage of large areas of research within historical institutionalism. 26

27 The volume is divided into five parts, each collecting a set of contributions on distinct aspects of historical institutionalism. The next section details conceptual and methodological foundations of historical institutionalism. Contributors revisit what is meant by structured politics, explore insights into patterns of institutional change, the role of critical junctures, the exercise of political power, and the relationship between ideas and interests. They further discuss methodological developments in historical institutionalism, including how scholars in this tradition wrestle with causality and make productive use of experimental designs. The first section also explores normative issues in historical institutionalism. The next four sections are devoted to major areas of research within Political Science. Section II explores historical institutionalism s contributions to research in comparative politics, and includes chapters on a wide range of political constructs with empirical illustrations from all corners of the world. It was within comparative politics that historical institutionalism was first structured as a distinctive theoretical approach to study the effects of institutions on politics. The volume highlights topics that since then have been at the core of the comparative historical institutional agenda, such as the study of the modern state, democratization, political parties, and organized labor. This section also includes chapters on topics that have more recently entered the comparative research agenda, such as the study of competitive authoritarianism, non-state provision of social welfare, informal institutions, and property rights in developing economies. In section III, contributors explore historical institutionalism in American Politics, which has been a major incubator for this approach in Political Science. Often associated with the study of American political development, historical institutionalism has both 27