Candidate Gender Quotas in France and the United Kingdom
Embedded Reform: Legal Setbacks and
Candidate Gender Quotas in France and the United Kingdom
Mona Lena Krook
ESRC Postdoctoral Fellow
Department of Politics
University of Bristol
10 Priory Road
Bristol BS8 1TU
Paper presented at the Political Studies Association Women and Politics Conference,
University of Bristol, February 19, 2005.
Embedded Reform: Legal Setbacks and
Candidate Gender Quotas in France and the United Kingdom
Candidate gender quotas have become increasingly popular in recent years as measures to increase the proportion of women in elected office. These policies – which include reserved seats, political party quotas, and legislative quotas[1] – have now been debated, adopted, or repealed in more than one hundred countries around the world. The rapid spread of these provisions has been accompanied by a literal explosion of research on gender quotas, which has sought to explain how and why quotas are adopted and, more recently, why some quota policies are more effective than others in facilitating women’s access to political office. Although increasingly more sophisticated, most of these studies treat each country as a tabula rasa when discussing quotas as strategies for increasing women’s political representation. More specifically, scholars theorize the features of the electoral system, the political party system, the quota provisions, and the coalitions of actors for and against quotas that, they argue, play a role in facilitating or undermining quota implementation. While privileging the role of the political context, this research implicitly assumes that these insights can be generalized to explain other cases, irrespective of time or place. A growing number of quota campaigns, however, have experienced serious setbacks when quota policies have been overturned on the grounds that they are unconstitutional or illegal. These cases highlight the normative aspects of quota reform, but even more crucially, call attention to the temporal dimensions of quota strategies, which often proceed in a context where earlier attempts to increase women’s political representation have succeeded or failed. Whereas countries that fall into the former category can build on existing gains to legitimize further action, thus strengthening quota provisions over time,[2] those that fall into the latter category face a much more restricted environment in which earlier setbacks set the boundaries on future action, thus weakening the prospects for further quota reform. All the same, prior failures suggest new directions for change not available to all quota campaigns by publicly exposing specific barriers to the selection of women as political candidates.
In this paper, I explore these dynamics to outline an approach for examining quota reforms in their broader temporal context. After briefly reviewing existing explanations for variations in quota impact, I examine recent work on temporality and distinguish between two emerging perspectives on institutional stability and change that offer potentially strong parallels with quota reforms. Adopting the second perspective, which theorizes how existing political arrangements shape the parameters of future policy development, I then apply a new technique based on sequences of reiterated problem solving (Haydu 1998) to analyze how legal setbacks shaped the subsequent paths of quota campaigns in two particular cases, France and the United Kingdom. Although these cases are unusual among countries where quotas have been declared unconstitutional or illegal, since they are the only ones where these judgments have not entirely paralyzed efforts to institute quotas, they offer valuable insights into the processes whereby earlier events influence both the form and content of subsequent strategies. Indeed, they not only reflect two different types of policy change – revision of the constitution and the sex discrimination act, respectively – but also embody two distinct dynamics of embedded reform that, following Thelen (2003, 2004), I label normative conversion and normative layering. This comparison thus provides crucial tactical lessons for on-going and failed quota campaigns, as well as new theoretical perspectives on the concept of ‘bounded innovation’ in historical institutionalist analysis, by tracking iterated shifts in formal norms of equality and representation.
Political contexts of quota reform
Initial research on gender quotas focused primarily on explaining quota adoption, arguing alternatively that women mobilize for gender quotas to increase women’s representation, political elites recognize strategic advantages for pursuing gender quotas, gender quotas are consistent with existing or emerging notions of equality and representation, and gender quotas are supported by international norms and spread through transnational sharing (Krook 2004a). A new wave of studies, however, has turned increasingly to the question of quota implementation, seeking to understand the factors that magnify or constrain the impact of these policies on women’s political representation. Undertaking fine-grained study of individual cases, this research as a whole concludes that the political context – that is, features of the electoral system, the political party system, the quota provisions, and the coalitions of actors for and against quotas – explains why some quotas are more effective than others in bringing more women into political office. Scholars observe, for example, that quotas tend to have a much greater impact in proportional representation electoral systems with closed lists and high district magnitudes than in systems characterized by majority elections, open lists, or low district magnitudes (Caul 1999; Gray 2003; Jones 1998; Nordlund 2003). Some go one step further and identify idiosyncratic features of particular electoral systems that negatively affect quota implementation, including the possibility for parties to run more than one list in each district (Costa Benavides 2003), the existence of distinct electoral systems for different types of elections (Jones 1998), and the chance for parties to nominate more candidates than the number of seats available (Htun 2002; Sacchet 2003). Others seek to nuance these explanations, however, by pointing to crucial differences across political party systems and political parties themselves. They note that quotas are more likely to be implemented in party systems where several parties co-exist and larger parties respond to policy innovations initiated by smaller parties (Kolinsky 1993; Stykársdóttir 1986; cf. Matland and Studlar 1996), although they also find that the tendency to implement quotas is closely connected with left-wing ideology and the ability of the party leadership to enforce party or national regulations (Davidson-Schmich 2002; Htun 2002).
A third group of scholars examines the details of quota provisions themselves and links their impact to the wording of the quota, whether the language used in the policy strengthens the quota requirement or reduces ambiguity or vagueness regarding the process of implementation (Araújo 2003; Baldez 2004; Chama 2001; Giraud and Jenson 2001); the requirements of the quota, whether the policy specifies where female candidates should be placed and to which elections the policy applies (Green 2003; Htun and Jones 2002; Meier 2004); the sanctions of the quota, whether the policy establishes organs for reviewing and enforcing quota requirements and procedures for punishing or rectifying non-compliance (García Quesada 2003; Guldvik 2003; Jones 1996); and the perceived legitimacy of the quota, whether the policy is viewed as legal or constitutional from the point of view of national and international law (Mossuz-Lavau 1998; Russell 2000; Yáñez 1998). A last set of explanations, finally, focuses on the actors who support and oppose quotas and their respective roles in guaranteeing or undermining quota implementation, namely political elites who comply with or seek to subvert gender quotas (Dahlerup 2001; Peschard 2002), women’s organizations who pressure elites and train female candidates (Durrieu 1999; Lokar 2003; Schmidt 2003), national and international courts who provide an arena to challenge non-compliance (Carrio 2002; Schmidt 2003; Villanueva Flores 2003), and ordinary citizens who engage in public scrutiny of parties’ selection practices (Baldez 2004; Kolinsky 1991).[3] While these accounts pay careful attention to the minutiae of individual cases, the way their findings are presented suggests that these insights can be generalized to explain other cases, regardless of place or time. Although some recent studies discover distinct patterns across developed and developing countries, and consequently criticize attempts to extend frameworks elaborated in the study of the former to analyze the latter (Bruhn 2003; Yoon 2004), virtually no research examines the possible role of temporal effects, which may serve a similar function in generating different types of causal effects, depending on a case’s location in time or its experiences within a longer trajectory of temporal processes (but see Krook 2004b; Krook 2005). Approaching each country as a ‘blank slate’ may thus produce misleading findings, as the scope for quota adoption and implementation may not be the same across national or even political party contexts.
Temporal contexts of quota reform
Cases where temporal processes loom especially large are those where previous attempts to increase women’s representation have already taken place. In these countries, earlier efforts shift the environment in which all subsequent strategies proceed, shaping not only their form and content but also, to a great degree, their prospects for success. As such, similar adjustments in the political context will not necessarily achieve the same effect across all cases, simply because each case is embedded in its own particular trajectory of reform. Campaigns in some countries can draw on existing gains to legitimize further action and strengthen quota provisions over time, while campaigns in others must navigate current setbacks that weaken the prospects for further reform, but that also point to new strategies for change. Focusing on the latter set of cases, which offer a potentially greater range of insights into the complicating role of temporal processes, I review perspectives on ‘time’ in social scientific analysis and then outline two approaches to political stability and change that incorporate temporality as a central element in the study of policy reform. The first theorizes that earlier reforms constrain later reforms through a punctuated equilibrium model of change, while the second suggests that latter reforms are an extension of earlier reforms through a bounded innovation model of change. I argue that the second perspective presents a more useful lens for analyzing quota campaigns, as it offers a more flexible judgment on the ways in which existing political arrangements shape the parameters of future policy development.
Temporality in social science
Historians have long incorporated notions of temporality and sequencing in their analyses, recognizing that all causal processes unfold within particular temporal contexts in which earlier events shape and influence paths to later outcomes. The ‘historic turn’ in sociology imported these ideas to the general study of social phenomena, inspiring scholars to theorize more explicitly the temporal dimensions of causal relationships located at specific moments in time and separated by varying degrees of temporal proximity. Reacting to tendencies within sociology and political science to extrapolate general theoretical statements from the study of cases occurring in particular places at specific moments in time, it explored instead how temporal contexts affect causal propositions, causal processes, and causal mechanisms in predictable and contingent ways (Griffin 1992; Somers 1996). As such, this work recognizes three broad approaches to the relationship between time and causality: a teleological approach, in which history embodies an inherent logic of social development that explains individual events by reference to some future outcome; an experimental approach, in which history is frozen into artificially exchangeable units and compared across cases by means of ‘natural experiments’; and an eventful approach, in which history is path dependent but unexpected events may suddenly shift the relationships between causal factors in radical and unanticipated ways (Sewell 1996). The first two approaches encompass most traditional work in the social sciences, while the third approach – despite some classic contributions – has grown increasingly popular in recent years as scholars have tackled puzzling historical patterns by rethinking conventional perspectives on the temporal scope and distance of causal effects.
Drawing on the insights of history, these researchers explore the importance of periodization in causal analysis to determine whether causal relations differ across temporal contexts and thus undermine the generalizing claims of a particular theory (Griffin 1992), or whether breaks between two periods can serve to bolster claims about cause and effect by delimiting the start or end of a theorized causal relationship (Lieberman 2001). Others investigate the relationships between distinct historical periods, asking whether earlier causal paths shape later causal relations through the dynamic of world time, in the sense that certain historical events change the causal prerequisites for developments that take place at a later period in time (Gerschenkron 1962; Moore 1966; McMichael 1990). Still others criticize the notion of immediate effects and note that causes and outcomes have variable time horizons that – in the case of slow-moving processes – may involve cumulative shifts, threshold effects, or the unfolding of extended causal chains before the impact of a given factor can be observed (Abbott 2001; Aminzade 1992; Pierson 2003). These insights largely undermine the assumptions underlying the experimental approach, which imply that when and in which order events happen is irrelevant to causal analysis (Rueschemeyer and Stephens 1997). All the same, they suggest various types of causal sequences: linear sequences, where one action leads inexorably to another towards a specific outcome; convergent or conjunctural sequences, where several actions come together in largely unanticipated ways to bring about a particular outcome; and path dependent sequences, where one crucial – and often contingent – action strongly shapes the path of later actions in the direction of a specific outcome (George and Bennett 2005; Griffin 1992; Mahoney 2000; Ragin 2000). The third type of sequence has received a great deal of theoretical attention in recent years, with scholars applying the concepts of ‘critical junctures’ and ‘path dependence’ to understand a range of political outcomes, particularly those where initial conditions do not appear to explain existing patterns of social interaction (Collier and Collier 1991; Mahoney 2001; Marcussen et al 1999). New work by earlier proponents of these concepts, however, criticizes this ‘punctuated equilibrium’ model of change and seeks to replace it with a more fluid model of ‘bounded innovation’ centered around dynamics of continuity and development (Thelen 2004; Pierson 2004).
Critical junctures and path dependence: a model of punctuated equilibrium
These two perspectives have been developed primarily within the literature on ‘institutions,’ defined as the formal features of political systems, as well as the procedures, routines, conventions, norms, and cognitive scripts that structure social, economic, and political life. Scholars approach these formal and informal rules in at least three different ways. Rational choice institutionalists view institutions as conventions created by actors seeking to solve collective action dilemmas and argue that institutions survive when they provide more benefits to the relevant actors than alternative institutional forms. Sociological institutionalists, in contrast, focus on the symbol systems, cognitive scripts, and moral templates that provide the ‘frames of meaning’ guiding human action. They emphasize the interactive and mutually constitutive character of the relationship between institutions and individual actions and understand change as an attempt to enhance the social legitimacy of a particular institution. Historical institutionalists, finally, see institutions as formal and informal procedures, routines, norms, and conventions embedded in the organization of politics, society, and the economy which create asymmetries of power that are reinforced through various types of coordination and feedback effects (Hall and Taylor 1996; Immergut 1998; see also Pierson and Skocpol 2002; Powell and DiMaggio 1991; Weingast 2002). Of these three schools, historical institutionalism draws most explicitly on temporal ordering as an analytical tool for explaining current patterns, as well as past and future trajectories of development.
Distinguishing between moments of stability and moments of change, research within historical institutionalism relies heavily on two key concepts: critical junctures and path dependence. Critical junctures[4] are major turning points in political life that cannot be predicted on the basis of prior events or initial conditions, in which the – often contingent – decisions of actors establish certain directions of change. These junctures are critical because they set institutional arrangements along particular trajectories that make previously viable alternatives increasingly remote in ways that are difficult, if not impossible, to reverse (Abbott 1997b; Collier and Collier 1991; Goertz 2003; Krasner 1984; Mahoney 2001; North 1990). Path dependence describes the periods between these junctures, when institutions lock the expectations and behavior of individuals into relatively predictable, self-reinforcing patterns. It stems from mechanisms like increasing returns, sunk costs, durable commitments, personal and social conservatism, and learning curves, that reinforce movement along a given trajectory, even when other choices appear better or more efficient (Aminzade 1992; Jepperson 1991; List 2004; Mahoney 2000; Meyer and Rowan 1991; Pierson 1996; Pierson 2000; Thelen 1999). This approach thus resembles a ‘punctuated equilibrium’ model of stability and change in which long periods characterized by path dependence alternate with brief and dramatic turns of event. As a model of exogenous change, however, it provides a greater range of insights for understanding institutional stability than the sources of institutional change (Krasner 1984; Thelen 1999).
Institutional continuity and development: a model of bounded innovation
Criticizing this strict separation between institutional innovation and institutional reproduction, new research in historical institutionalism observes that institutional continuity pervades many critical junctures at the same time that institutional change and evolution characterize many periods of path dependence. Indeed, rather than complete institutional replacement, these scholars find, institutions tend to survive through two dynamics of institutional transformation: institutional conversion, where existing institutions are directed to new purposes, and institutional layering, where some elements of existing institutions are renegotiated but other elements remain. According to this perspective, institutional stability and change are not mutually exclusive, but are instead often closely implicated with one another: conversion involves the adoption of new goals or the incorporation of new groups into existing institutions, while layering grafts new elements onto otherwise stable institutional frameworks (Thelen 2003; Thelen 2004). This work extends and elaborates earlier notions of ‘bounded innovation,’ which argues that existing institutional arrangements set certain restrictions on the types of innovation that are possible such that attempts at reform typically reproduce or redefine existing rules, practices, and ideas, even when individuals hold different and conflicting views on matters of common concern (Weir 1992). At the same time, it reflects strong parallels with research that identifies the institutional order itself as a potential source of change, which suggests that individual institutions are more vulnerable in cases where a number of different institutional orders coexist and interactions and encounters among these orders create ‘friction’ between mismatched institutional patterns, forcing political actors to find new ways of advancing their aims through new institutional forms or through adaptation to take advantage of new institutional opportunities (Lieberman 2002; Orren and Skowronek 1994; Thelen 1999).