This essay proposes an integrated discursive institutionalism as a framework for feminist political analysis. Both historical institutionalism and discourse analysis have merits and limitations, and both perspectives complement each other and offer solutions to their respective deficiencies. Traditionally there has been a strong demarcation between the two perspectives. A common way to divide both approaches is between investigating “causal regularities” and “understanding meaning.” I argue that a feminist institutionalism needs to deconstruct the dichotomy of causal explanation versus meaning and description and to reformulate the concept of causality. There is no adequate explanation without “meaning,” and the stretching of institutionalism toward “ideas” exemplifies this inadequacy.
Rather than emphasizing their differences, I stress that institutionalism and discourse theory share important epistemological insights that facilitate their convergence into an integrated approach. Both theoretical perspectives emerged in response to the economic determinism in materialist theory. Both approaches use concepts of temporality, relationality, and contextuality to analyze political developments. Their respective research designs start from real-world puzzles and are problem-driven, rather than aiming at a general theory (Thelen Reference Thelen1999, Torfing Reference Torfing, Horwarth and Torfing2005). Both approaches have in recent years moved closer toward one another. Not only has institutionalism incorporated “ideas” into its frameworks, but discursive analysis in political science also operates with a clear notion that political discourses are embedded in institutional contexts, which in turn structure the flow of possible articulations.
An integrated feminist discursive institutionalism has four major advantages. Firstly, it enables a relational conceptualization of gender and thus conceives gender as a relevant analytical category, even if women's agency is deemed an irrelevant causal factor. Secondly, it implies a reflexive notion of the political, indispensable for feminist scholarship, whereby political processes are not reduced to a fixed set of state institutions and rational utility maximizing actors; instead, the intersections between public and private are taken seriously, and policy problems and actors' preferences are not treated as given but as something to be explained. Thirdly, it allows for a reformulation of causality as contextual in keeping with a configurative strategy of research, which fits into a feminist perspective of situated knowledge. Fourthly, it better equips feminist scholarship to analyze the current transformations toward governance, a mode of ruling to which the discursive political processes and knowledge production are central.
I first sketch the merits and limitations of gender-sensitive historical institutionalism before discussing the prospects and challenges of a gendered discursive institutionalism.
The Merits and Limitations of Gender-Sensitive Historical Institutionalism
The major contribution of institutionalism is not simply adding a new set of variables of state capacity and structure, as suggested by protagonists of a feminist variable approach (see, for example, Mazur Reference Mazur2002). Instead, the central new insight of institutionalism for comparative politics is its reflexive perspective on the political, which allows it to move beyond determinist and rationalist conceptions of causality. Historical institutionalism displaces the structure/action divide based on universal causal laws (and its historically specific forms of realization) in favor of the focus on social and political orders as relational networks and temporal processes (Pierson Reference Pierson2004). Here, institutions are conceived of as conditions for action that can make a certain course of action more or less “appropriate” or “promising.” Historical institutionalism opens a space of contingency and investigates the interplay between institutions and agency.
In addition, the concept of causality is reformulated, linking it to a further key epistemological insight regarding theorizing temporality. The significance of specific phenomena is not due only to what they are but also to when they took place and their situatedness in relation to other factors and processes over time (Pierson Reference Pierson2004). Attention to timing and sequence highlights that political decisions are made with regard to the previous policies and their effects. This feedback mechanism can cause a counterreaction, a revision or strengthening of policies as such, constituting a path dependency.
This dynamic concept of causation fits well into a feminist framework. It radically abandons the notion of master categories, which make it impossible to grasp the formative “causal” role of gender relations. The contextualized and temporalized concept of causality corresponds to the feminist epistemology of situated knowledge. In addition, the emphasis on relationality enables researchers, at least in principle, to investigate a major concern to recent feminist theorizing—the intersection of gender, class, race, and ethnicity.
Feminist scholars using historical institutionalist frameworks show that national variations in the strategies and goals of political actors (including those of women's movements), as well as policy outcomes, can be largely explained by investigating the interplay between the institutional matrix and processes of group formation (for example, Skocpol Reference Skocpol1992; Sainsbury Reference Sainsbury2001). However, both early and more recent feminist work using this approach illustrates the limitations of institutionalism, which sees the formation of collective actors and their strategies and goals as analytically relevant, but does not have the theoretical tools to fully grasp such processes (Kulawik Reference Kulawik1999; Reference Kulawik2009). Furthermore, the reduction of the category of “gender” to women is the norm in this literature (O'Connor, Orloff, and Shaver Reference O'Connor, Orloff and Shaver1999; Skocpol Reference Skocpol1992). Most work to date has been insufficiently attentive to the contradictions, exclusions, and trade-offs in the construction of policy discourses. While authors such as Georgina Waylen (Reference Georgina2007) and Susan Franceschet (Reference Franceschet2008) stretch their analysis toward issue framing and its resonance with a broader discourse and policy style, they fail to treat these discursive elements systematically. Such work illustrates what Vivien Schmidt (Reference Schmidt2006), one of the protagonists of a “discursive institutionalism” in mainstream comparative research, calls the limits of arguments related to “institutional fit.”
Unbounding the Political: Feminist Political Discourse Analysis
What does discourse analysis offer that other approaches do not? First, discursive approaches advance the reflexive concept of the political and abandon mechanistic and rationalistic notions of the political subscribed to by large parts of political science and which has been recast by historical institutionalists in important ways. Some scholars working within the institutionalist tradition have, as already mentioned, for quite some time sought to incorporate “ideas” into their analysis (Blyth Reference Mark1997). The main division between the ideational and the discursive perspectives is that the latter deems discourse as constitutive of societal relations, whereas for the former, ideas present only one possible variable of a variety within competing explanatory factors (Fischer Reference Fischer2003, 27–45; Kulawik Reference Kulawik1999, 46). The ideational approach maintains a dichotomy of ideas versus “interests” (Schmidt Reference Schmidt2006). According to constructionist discourse theory, those very schemata on which interests are based—such as economic utility and means–ends rationality—have to be seen as resulting from discourse. Taking into account the masculinist gender bias of the account of interests, the deconstruction of such schemata is an important task for feminist analysis.
There is, however, no correct way to apply discourse analysis to the political field (Bacchi Reference Bacchi2005, 198). In the 1980s, two feminist scholars, the political scientist Jane Jenson and the philosopher Nancy Fraser, made important attempts to apply discourse theory in feminist political analysis. Both began by reformulating the concept of the political by unbounding it: Politics is not simply understood as a conflictual process between actors in pursuit of particular interests within an institutionalized decision-making process; it is more a struggle of interpretation over who, what, and how to “politicize” (Fraser Reference Fraser1989, 166). Thus, politics is not just a strategic dispute over who gets what but a struggle for representation of needs, problems, and identities. The “universe of political discourse” (Jenson Reference Jenson1986; Reference Jenson1989) is constituted by a variety of actors and discourses, as well as by mechanisms of inclusion and exclusion.
Within this framework, political power is not only composed of strategic strengths, such as organizational membership or the number of seats in the parliament; it is also the ability to put one's own interpretation of social relations and problems on the political agenda and thus to push for one's own solutions and proposals. The discursive reformulation of the political implies a shift in analytical perspective on the object of study itself. Policies were usually investigated as a reaction to “objective” problems, yet, through the lens of discourse, problems are no longer taken as given but perceived as a result of interpretation. Accordingly, countries differ not only in terms of their policy institutions but also in in the way that problems and their causes are interpreted, which, in turn, influence the solutions that are deemed appropriate.
Jenson's conceptual and comparative work on early welfare state formation in France, Britain, and the United States was groundbreaking scholarship that anticipated many of the insights of Carol Bacchi's (Reference Bacchi1999) “what's-the-problem” approach over a decade later. Compared to Bacchi's, Jenson's framework has the advantage of being comparative. Nonetheless, there are two major limitations in her approach. First, she fails to pay attention to the intersectionality of gender with other social categories, also implying that gender stands for “women.” Second, despite its ability to link national discursive patterns with social policies, Jenson's analysis does not sufficiently explain how these specific interpretations came about (Kulawik Reference Kulawik1999, 68f). Her analysis focuses on content, rather than on the discursive practices that produce them. Jenson refers, for instance, to the relevance of experts and scientific knowledge in the policy discourses, but does not consider how different knowledge forms struggle for recognition or how different kinds of expertise acquired the defining power in the respective national political field. She applies discourse analysis in what might be called a causal way, according to which variations of problem interpretation account for different policy design in different countries. Her approach includes elements of critical-frame analysis that have been developed and applied in recent comparative feminist policy studies (Verloo and Lombardo Reference Verloo, Lombardo and Verloo2007).
In contrast, the strength of Bacchi's framework is that it elaborates the discursive mechanisms that constitute certain policy problems. Her lack of a systematic comparative perspective, however, means that she cannot explain why discursive practices and hegemonic interpretations are so different in various countries. This is a major conceptual problem inherent in the notion of discourse as constitutive. When everything is interrelated and constituted through discourse, it becomes impossible to distinguish between different explanatory factors and their impact for national variation (Kulawik Reference Kulawik1999, 49–52). And it is not accidental that the recent boom in feminist discourse analysis identifed by Bacchi (Reference Bacchi2005) consists mainly of single-country studies. I agree with Laurel Weldon's (Reference Weldon2006) insightful reflections on intersectionality, that comparative political analysis only makes sense if one can distinguish between categories and their independent formative effects.
Furthermore, political discourse analysis lacks conceptual clarity concerning institutions; they are frequently mentioned, but not given proper analytical attention. Trying to grapple with the agency problem, Bacchi (Reference Bacchi2005, 206) refers to discourses as “institutionally” produced schemas, without explaining what this means. Johanna Kantola (Reference Kantola2006, 34) explicitly addresses institutions but loses track when she claims that discourses and institutions are “mutually constitutive and reinforce each other.” If discourses and institutions always and only reinforce each other, then why distinguish them at all? As we must distinguish between institutions and discourses, it is then an empirical question whether or not they reinforce each other and under what conditions.
Conclusions: Linking Institutions and Discourses
In order to overcome these conceptual limitations of institutionalism and discursive analysis, I propose a gendered discursive institutionalism. The advantage of such an integrated approach lies in its reflexive concept of the political, in which political processes are not reduced to a fixed set of state institutions and rational utility-maximizing actors, but instead investigated from the perspective that the process inherently functions as boundary work between what is considered to be political and nonpolitical. Such an integrated approach allows for a reformulation of analytical concepts, focusing on the interrelations of institutional arrangements, actor constellations, and political discourse. The important question is, then, how to conceptualize the linking mechanisms and the interplay between them. An integrated approach requires a rethinking of institutions, agency, and political discourses.
Institutions
Although new institutionalism has moved beyond a formalistic understanding of political institutions, institutions are generally still understood in formal terms as procedures taking place outside of agents or “rule-following structures” (Schmidt Reference Schmidt2008). This critique also applies to feminist institutionalism that conceptualizes institutions as operating according to a normative “logic of appropriateness” and embodying different “gendered cultures” (Chappell Reference Chappell2006). I propose moving from a normative toward a discursive notion of institutions. This implies that institutions are indeed constituted by discursive struggles and can be understood as sedimented discourses. What distinguishes them from discourses is that there are relatively fixed functional units that serve certain purposes, such as making binding decisions or distributing social benefits. Political institutions are a duality of procedural and symbolic orders. These are two distinct yet related dimensions with which they interact with their wider political and social environment. Gender is part of both dimensions. The codes and schemas embedded in institutions, both normative and cognitive, may be reinterpreted, but in their daily routinized operations they are naturalized, and therefore not up for open contestation.
Agency
Although neither institutionalism nor discourse theory solves the problem of agency satisfactorily, a discursive account is helpful in reformulating the relation between agency and institutions in a nondeterministic manner. Institutions are not an external structure imposed upon political actors but constructs internal to actors, which they change and create. “Internal” means that institutions work in the ways that the actors “do institutions.” In turn, “doing institutions” occurs in two ways: first, as a routine in which actors rely on “background discursive abilities” (Schmidt Reference Schmidt2008) that implicate “rule following” from what can be termed as “habitus”; and second, through employing “foreground discursive abilities” in which actors problematize and deliberate problems and rules. This distinction parallels “practical consciousness” and “discursive consciousness” (Giddens Reference Giddens1984). The discursive perspective helps us reconceptualize the relationship between political opportunity structures and agency. The “objective” existence of opportunity structures is not decisive for the actual strategies pursued by actors, but how these opportunities are perceived (Kulawik Reference Kulawik1992; Naumann Reference Naumann2005). Discursive processes of collective identity formation, in which the negotiation of shared aims or the interpretation of previous experiences with political institutions or social policies takes place, are crucial for understanding how political actors mobilize and pursue their claims. The agency problem can only be solved within a conception that proceeds from the situatedness of action as a constitutive and not only contingent condition of agency (Joas Reference Joas1996, 235).
Political discourses
These represent a special kind of discourse in the universe of all societal discourses. Political discourses are concerned with the politicization of needs, problems, and identities in order to establish them in the public sphere and as legitimate in relation to state action. therefore, they have a special operational logic, which distinguishes them from cultural discourses in the public sphere or private communication. But, of course, political discourses interact with other discursive arenas in the process of politicization, at the same time they operate as “boundary work,” setting the limits of the political space by processes of marking and boundary drawing (Gottweiss Reference Gottweiss, Hajer and Wagenaar2003).
The analytic tools of feminist discursive institutionalism have to be elaborated more rigorously concerning what is meant by “discursive practices” and with regard to systematic empirical research. I propose first to distinguish between a constitutive and a causal notion of discourse and to investigate how they can be fruitfully used in comparative analysis. The constitutive dimension of discourse limits the scope of statements that can be meaningfully articulated in a given society. But there is still a variety of claims that can be made in a given national context. The outcome of such interpretative struggles depends on the discursive resources and competences of the political actors, as well the institutional arrangements. The distinction between a constitutive and causal conception allows us to turn the duality of discourse—as processes of intersubjective interpretation and systems of signification—into an empirical question. Thus, it becomes possible to evaluate under which conditions people might “use” discourse and under which conditions they are “used” by it.
Second, I suggest that feminist analysis must pay systematic attention to the articulation of different knowledge forms—such as scientific, moral, legal, economic—in national discourses and how normative and cognitive claims intersect, as well as how countries might differ in the ways they evaluate such claims. Modern states are unthinkable without the contribution of science and scientific expertise and their input into political processes. Countries differ enormously with regard to how they institutionalize expertise and assess different knowledge systems. To date, we know almost nothing about the genderedness of public knowledge regimes (Campbell and Pedersen Reference Campbell and Pedersen2007). In recent feminist scholarship, however, there is increasing realization that gender knowledge and expertise have become part of modern governance and that epistemic authority is gaining importance (for example, Bedford Reference Bedford, Rai and Waylen2008). Sheila Jasanoff's (Reference Jasanoff2005) concept of “civic epistemologies” as nationally specific ways of public knowing, in the sense of institutionalized practices to deploy and authorize knowledge claims, can offer important insights here.
My final point on the interaction between discourses and institutions is that institutions are not only sedimentations of discursive struggles; they are also locations for communication. The configuration of institutional arrangements has important implications for discursive practices because they structure access to discursive arenas and styles of communication. Empirical research shows that discourses and institutions interact, but it is too anecdotal and limited. A more encompassing body of scholarship, which allows for a continuous dialogue between the development and refinement of analytical concepts and empirical research, is needed in order to move forward a comparative politics of gender and a feminist discursive institutionalism.
