In this essay I lay out how historical institutionalism (HI) could serve as an important tool for feminist political scientists, highlighting the potentially distinctive contribution and advantages of a feminist historical institutionalism (rather than a feminist institutionalism) for feminist political science and particularly for a feminist comparative politics. The potentially significant contribution of HI is to help us answer some “big questions,” in particular, how and why institutional change occurs. This, in turn, can help us understand how positive gender change, such as improvements in women's descriptive and substantive representation, can come about. This latter question is obviously not only of academic but also of political concern for feminists who wish to lessen gender inequality.
The issue of whether the explicit adoption of HI would help feminist political science find a place nearer to the center of the discipline is an important but separate one. Suffice it to say that its overall approach and underlying assumptions (methodologically pluralist, problem driven, and historically focused) make HI potentially more open to incorporating gender into its frameworks than other forms of institutionalism. And there is even some evidence that this sometimes happens, for example, in Theda Skocpol's (Reference Skocpol1992) work on the development of the American social policy and in some of Paul Pierson's (Reference Pierson1996) work on European integration, in which he examines gender equality in the European Community.
While not appropriate for every research question, HI approaches can therefore solve some of the problems that currently hamper feminist political analysis in answering some big questions, such as how certain institutions and regimes are gendered, how they came into being, and how change can come about, as well as in understanding the relationship between different actors and the institutional context. The rest of this essay demonstrates how analyses grounded in HI could facilitate better understandings of the interaction of structure with agency and of why and how gender positive change can occur in certain contexts and not others.
The Historical Institutionalist (HI) Framework
Rather than rehearse the nature of HI in great detail, I outline some characteristics associated with one strand of historical institutionalism that makes it a useful approach for many feminist political scientists and particularly those developing a comparative politics of gender (Pierson and Skocpol n.d.; Thelen Reference Thelen1999; Thelen and Steinmo Reference Thelen, Steinmo, Steinmo, Thelen and Longstreth1992).
Historical institutionalists address real world puzzles – explaining variations in important or surprising patterns of events. In providing explanations, HI takes history seriously. But it does not just “tell the story.” Through structured comparison and historical process tracing, it aims to identify the causal mechanisms that lie behind particular empirical processes. But it does not do this through functional explanations that argue, for example, that institutions/regimes grow up because they are appropriate or by emphasizing the tendency toward equilibrium as rational choice institutionalism does (Thelen Reference Thelen1999). Instead, historical institutionalists use various notions of path dependence, ranging from the very loose to the more rigidly determining, to look at the ways in which slow-moving causal processes are linked (Pierson and Skocpol n.d.; Pierson Reference Pierson, Mahoney and Rueschemeyer2003; Pierson Reference Pierson2004; Thelen Reference Thelen, Mahoney and Rueschemeyer2003). Critical junctures, “as a period of significant change, which typically occurs in distinct ways in different countries . . . and which is hypothesized to produce distinct legacies” (Collier and Collier Reference Collier and Collier1991, 29), play a central role in setting these path-dependent processes in motion.
Contesting claims that HI is biased toward structures and institutional continuity (Hay and Wincott Reference Hay and Wincott1998), Kathleen Thelen (Reference Thelen, Mahoney and Rueschemeyer2003; Reference Thelen2004) is interested in three different questions concerning how institutions are created, reproduced and changed: what are their origins, why do they persist, and how can we explain their demise? Focusing on understanding institutional change and evolution and taking issue with deterministic explanations that rely on constant causes, Thelen (Reference Thelen, Mahoney and Rueschemeyer2003; Reference Thelen2004) argues that institutions may persist for reasons that are different from those responsible for their creation. Unforeseen collisions and interactions, resulting in unintended consequences, often mean that outcomes are not always what might be predicted through strong path dependence. But at the same time, the importance of developmental pathways mean that institutions are not always replaced or redesigned. Often, institutional layering—new institutions added on to existing ones, particularly if coalitions lack sufficient support to innovate—or institutional conversion, for example, if new groups are being incorporated, takes place. And institutions are reconfigured and evolve as a result of endogenous factors. Thelen (Reference Thelen, Mahoney and Rueschemeyer2003, 213) claims that there is a need for more structure in HI analyses at the “front end,” namely, showing how structures limit actors' choices at “choice points” or critical junctures, and more agency at the “back end” of arguments, emphasizing the ways in which institutions operate not just as constraints but also as strategic resources for actors.
Actors, often described as interests in HI, therefore play a key role, whether within institutions (such as bureaucracies, governments, and political parties) or outside of them (such as interest groups and civil society organizations). Actors are strategic, seeking to realize certain complex, contingent, and constantly changing goals, and they do so in contexts that will favor certain strategies over others (Hay and Wincott Reference Hay and Wincott1998, 954). Change through contestation is often significant, and political struggles are inevitably mediated by the institutional context in which they take place. Bringing together the analysis of actors and institutions to understand the ways that institutions can shape not only actors' strategies and their outcomes but also their goals is, therefore, a key part of HI (Thelen and Steinmo Reference Thelen, Steinmo, Steinmo, Thelen and Longstreth1992). Historical institutionalists argue that their understanding of interests does accommodate notions of inequality of power and resources between actors. Institutions have distributional effects. They reflect, reproduce, and magnify particular patterns of power. Moreover, political arrangements and policy feedbacks actively facilitate the empowerment of certain groups, and these factors will impact on actors' goals and strategies (Thelen Reference Thelen1999, 394).
The significance of the ideational in institutionalist analyses has recently become more contested as the relationship of the ideational with the actor's behavior has loomed large (Blyth Reference Blyth1997). Although institutional analysis, particularly sociological institutionalism, does consider the role of norms and rules in creating and maintaining both formal and informal institutions, some analysts have claimed that HI does not emphasize the ideational sufficiently, particularly in times of crisis and change. Clearly, the role of ideas, norms, rules, and procedures in influencing the mutually constitutive relationship between actors and institutions must be a key part of HI. Finally, historical institutionalists emphasize the importance of the overarching context. Because of their emphasis on complex causal patterns, historical institutionalists often take a macro sweep and, even in single case studies, put them in a wider context. HI's overall approach facilitates analyses that can demonstrate that, because in different national contexts, different mechanisms of reproduction sustain different institutions, common international trends can have differing domestic consequences. Additionally, sequencing and timing can play a central role in ensuring that different factors play out differently in different contexts.
Toward a Feminist Historical Institutionalism?
What can we take from this approach to inform feminist historical institutionalist research strategies? Following Thelen (Reference Thelen1999; Reference Thelen, Mahoney and Rueschemeyer2003) I would argue for a middle notion of path dependence that, while recognizing the importance of critical junctures, does not invest too great a determining role in such factors as self-reinforcing positive feedback mechanisms or increasing returns that ensure that certain paths once adopted are hard to deviate from (Pierson Reference Pierson2000). According to Thelen (Reference Thelen, Mahoney and Rueschemeyer2003), too much emphasis on effects that reinforce particular trajectories are in danger of overdeterminism and of obscuring ongoing political contestation. So while historical institutionalist approaches can be useful in explaining how particular institutions and regimes arose, how they are gendered, and why it is often so difficult to change them, they also need to be able to explain how and why institutions can or cannot be renegotiated in different contexts, focusing both on the formal and informal variants and the ways in which institutions have gendered norms and logics (Chappell Reference Chappell2006).
We must also consider how feminist historical institutionalists should answer “big questions.” How big and bounded should their analyses be? Comparative work is inevitably at the center of much HI. Until recently, many feminist political scientists seemed reluctant to undertake systematic comparative analyses of more than two cases. Part of this reluctance may have been due to the lack of work on how to undertake gendered comparative analysis, as well as to some of the methods associated with medium and large-n studies. But one consequence is that some analytical problems, for example, regarding the treatment of institutions, structure and agency as outlined previously, remain unchallenged in the absence of a rigorous body of work engaging in systematic and contextualized comparison over a range of cases. However, in common with mainstream HI, it is likely that small-n studies will predominate in feminist HI work because tracing complex causal patterns over time does not easily lend itself to large-scale quantitative studies. Because of their emphasis on solving real-world puzzles, historical institutionalists are sometimes accused of selecting on the dependent variable, that is, choosing their cases because of the outcome, and this is obviously potentially more problematic with a single case study—making the case for feminist comparative work that uses more than one case even more compelling.
It is hard to assess how much feminist work located within a HI framework, either explicitly or implicitly, already exists. The comparative politics of gender and feminist comparative politics are young fields, but there is now a growing body of work that can be identified as such (Tripp Reference Tripp2007). However, there are far fewer works that either explicitly identify themselves as historical institutionalist or have been identified as such by others (see, for example, Charrad Reference Charrad2001; Htun Reference Htun2003; O'Connor, Orloff, and Shaver Reference O'Connor, Orloff and Shaver1999).
Without a substantial existing body of feminist HI work, there are few guidelines on how to proceed. The rest of this essay outlines how an HI approach could be used for understanding substantively important gender outcomes and institutional continuity and change in transitions to democracy (Waylen Reference Waylen2007). It begins with a real-world puzzle: Why do the outcomes of some transitions to democracy seem more positive (defined as improvements in women's descriptive and substantive representation) than others, and how can we explain these varying gender outcomes?
To unravel this puzzle, I did a systematic and contextualized small-n comparison of eight “third wave” transitions that included transitions from state socialism (Hungary, Poland, Czechoslovakia) and authoritarianism (South Africa, Chile, Brazil, Argentina, El Salvador) to some loosely defined form of democracy. The n was large enough to have more cases than variables and some variation on the dependent variable. Indeed, their gender outcomes did vary considerably. For example, in 2006 levels of women's legislative representation ranged from around 9% to 36%. Although women's policy agencies (WPAs) were widely established, their effectiveness varied considerably, and gender policy outcomes also differed tremendously—policies against domestic violence were widely adopted, but in only one case were abortion laws liberalized and in two cases became more restrictive.
To explain these gender outcomes within an HI framework, several things have to be done. First, in order to identify the causal sequences and developmental pathways that led to these varying outcomes, attention to process over time is important. The time frame of the study spanned 20 to 30 years and included the previous nondemocratic regime and the transition itself, as well as 10 or more posttransition years, thus allowing us to see whether the different institutional legacies of authoritarian and state socialist regimes had a significant impact. Second, transitions to democracy, particularly the actual transition—defined as the period between the collapse of the nondemocratic rule and the founding election—are identified as a form of critical juncture. But in keeping with looser versions of HI, structural constraints, even in the most open transitions, still weighed heavily on actors, affecting their strategies, goals, and outcomes; and institutional legacies remained. In some cases, for example, in some pacted transitions, constraints were more determining than in others. A feminist HI analysis needs to examine the nature of the constraints that help to determine actors' goals and strategies. Third, a feminist HI analysis would explore to what extent outcomes emerge as a result of path-dependent processes set in motion at the point of transition, and how far they are also a result of ongoing contestation.
If we consider how structural constraints affect actors, we see that while women's organizing during all stages of transition seems a necessary condition for positive gender outcomes to occur, on its own it is not sufficient to achieve change in the short term. The nature of that organizing—how far movements are cohesive and interact with the state and the conventional political arena—is important. Although women's organizing is frequently the key to articulating gender issues, several other factors help to determine whether certain gender issues, once articulated, make it onto the agenda of the transition during the critical juncture and are ultimately translated into improvements in women's citizenship in the posttransition period.
Key women, particularly feminists, are often central. They are present not just within organized women's movements, which are usually crucial in articulating issues, but also within legislatures, political parties, governments, and bureaucracies in the posttransition period. And these key women often form a range of insider and outsider alliances with other important actors. The headway that key women can make, whether from women's movements or other institutions, depends on the openness of both the transitions and institutions. Pacted and relatively drawn-out transitions with negotiation processes that are relatively open, transparent, and accountable, though often more constrained, also appeared more likely to be accessible to women actors. But powerful participants in negotiations, particularly within the opposition, also have to be open to gender concerns, and feminists have to be already present or have access to those arenas. Therefore, political parties, such as the African National Congress in South Africa, can also play a key role.
Furthermore, it helps if gender issues are framed strategically in ways that resonate in that political climate. In South Africa, feminists could insert gender into a broader discourse of rights; in the Latin American cases, women's issues were positioned as part of the opposition to dictatorship and state violence and the struggle for democracy. Goals, for example in the area of reproductive rights, as well as strategies were also shaped by the varying structural constraints in different contexts.
Using a relatively loose notion of path dependence that allows for unpredictable outcomes and ongoing contestation, we see that in order to maintain gender issues on the policy agenda and obtain positive outcomes in the posttransition period, several other conditions must be fulfilled. The subsequent policy environment needs to be sufficiently open. It is helpful if some important building blocks are already in place. New institutions such as gender-friendly constitutions can form one such block. A government in power that is sympathetic to gender issues with feminists in key positions in the party hierarchy, executive, and legislature, as well as institutions such as an effective WPA, is also critical. There is some evidence to show that in different contexts, activist women will strategically target those arenas—electoral, bureaucratic, or constitutional/legal—most accessible to them, affecting not only their goals but also the outcomes they can achieve. At one end of the spectrum, key women were present in all three arenas in South Africa, and institutional change took place in all three.
There was a considerable degree of institutional continuity in the posttransition period and much gendered institutional change, such as the establishment of WPAs and quotas, that resulted from ongoing contestations in which women were fighting for incorporation. These changes took the form of institutional layering and conversion. The WPAs that were set up in the aftermath of transitions were often positioned as units within other ministries with unclear and overlapping briefs, and only a few were created to stand alone with any power and autonomy. The implementation of electoral quotas to increase women's representation was another example of a partial renegotiation of some elements of electoral institutions, leaving others in place. Outcomes therefore play out differently in different contexts as a result of varying degrees of continuity from the previous regime, the extent of institutional layering, and differing levels and types of contestation at different points.
A feminist HI approach, therefore, enables us to answer important questions by combining the analysis of actors, particularly key actors in insider and outsider alliances, with their institutional context in ways that are mindful of institutional legacies, as well as institutional change and the importance of ideas and framing. Sensitively used, an HI framework and tools such as critical junctures and path dependence can prove very helpful for feminist political science.
