Jump to ContentJump to Main Navigation
Cabinets, Ministers, and Gender$

Claire Annesley, Karen Beckwith, and Susan Franceschet

Print publication date: 2019

Print ISBN-13: 9780190069018

Published to Oxford Scholarship Online: September 2019

DOI: 10.1093/oso/9780190069018.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 28 March 2020

Explaining Gendered Patterns of Cabinet Formation

Explaining Gendered Patterns of Cabinet Formation

Chapter:
(p.1) 1 Explaining Gendered Patterns of Cabinet Formation
Source:
Cabinets, Ministers, and Gender
Author(s):

Claire Annesley

Karen Beckwith

Susan Franceschet

Publisher:
Oxford University Press
DOI:10.1093/oso/9780190069018.003.0001

Abstract and Keywords

Chapter 1 introduces the three research questions guiding the book and outlines the patterns of timing, magnitude, and persistence of women’s cabinet inclusion in Australia, Canada, Chile, Germany, Spain, the United Kingdom, and the United States. It identifies the year of appointment of the first woman to cabinet; the year of the last all-male cabinet; and addresses the questions of cross-country and cross-time variation in numbers of women in cabinet. The chapter identifies formal and informal rules as forces shaping women’s opportunities for cabinet appointment, and introduces the concept of the “concrete floor,” the minimum proportion or number of women for the cabinet team to be perceived as legitimate.

Keywords:   female minister, gender, cabinet, gendered cabinet formation, women’s cabinet representation

Cabinets lie at the center of governing power. For politicians, appointment to cabinet may be the highpoint of a career, providing additional prestige and salary benefits, as well as the opportunity to set policy agendas and make important political decisions. In both presidential and parliamentary democracies, cabinet ministers enjoy considerable status and prestige, and are often more visible and well known than members of legislatures. Ministers are “visible, glamorous and important” (Blondel 1985, 3).

Ministers are also gendered. Examining cabinet appointments across time reveals that historically most ministers have been men, as have the heads of government who appointed them. In this book, we explore three questions: Why are more men than women appointed to cabinet? Why are women more likely to be appointed to cabinet in some countries than others? Why has women’s presence in cabinet varied across time? Answering these questions reveals how the process of cabinet formation is gendered, providing men and women with different (and unequal) opportunities to be appointed ministers. In this book, we reveal the gendered dimensions of cabinet appointments through an in-depth analysis of seven countries: Australia, Canada, Chile, Germany, Spain, the United Kingdom, and the United States. We constructed a data set of ministerial appointments to post-election cabinets when a prime minister or president has the opportunity of forming a cabinet de novo. We track all cabinets formed from the point at which the first female minister was appointed in each country to the last post-election cabinet formed in each country before the end of 2016. Across all seven countries in this time frame, 127 cabinets were formed. Of these, 33 included only men as ministers, or 26.82 percent of all cabinets. Although a total of 2,229 ministerial appointments were made, of these, just 294 appointments went to women, or 13.19 percent of ministerial appointments. It is clear, though unsurprising, once women were included in cabinets, men have nonetheless predominated in cabinets in these seven countries for multiple decades. (p.2) Among our country cases, only Germany has consistently included at least one woman in post-election cabinets for more than 50 years.

These seven countries are useful cases for understanding women’s presence in newly formed cabinets and, we argue, for understanding the gendered process of cabinet formation. We discuss our case justification in greater depth in Chapter 2. For now, while our data show that men are predominant in cabinet appointments within each of these countries, the data also reveal significant cross-national variation in women’s representation in cabinets. For example, in terms of timing, women joined the ranks of ministers as early as 1929 in the United Kingdom and the 1930s in the United States and Spain. Elsewhere, women were entirely absent from cabinets until the late 1950s (Canada) and early 1960s (Germany).

Variation also appears in terms of the magnitude of women’s cabinet presence. We categorize the magnitude of women’s presence as an ordinal variable, ranging from very low (0 to 9.9 percent of cabinet), low (10 to 19.9 percent), medium (20 to 29.9 percent), high (30 percent or more), and parity (50 percent +/–1). In Australia, the United Kingdom, and the United States, women’s presence in cabinet has remained in the low to medium category, while in Canada, Chile, Germany, and Spain, women’s representation has reached high magnitude. More important, in this latter group of countries, parity has been achieved at least once in each country. These varying patterns challenge us to explain why more women are appointed to cabinet in some countries than in others.

Our cases also reveal important differences in the persistence of women’s representation in cabinets. The appointment of the first female minister was not always a decisive turning point that marked women’s subsequent cabinet inclusion. Even after the historic breakthrough of a first female minister, all-male cabinets continued to appear in all countries, with the exception of Germany.1 As Table 1.1 shows, in countries such as the United Kingdom, whole decades passed between the appointment of the first woman to cabinet and subsequent appointments of female ministers. Our data also show more positive patterns of persistence. Once women’s presence has increased beyond one female minister, especially since the 1980s, it has tended to do so in a continuous fashion. Increases in magnitude tend to be sustained, with minimal backsliding. We call this phenomenon the “concrete floor.” The term (p.3) “concrete floor” describes the minimum proportion or number of women for that ministerial team to be perceived as legitimate.

For each of our cases, we identify a point at which a selector (or selectors) increases, often significantly, the proportion of women in cabinet, unleashing a dynamic whereby subsequent selectors, regardless of party, perceive costs to violating the threshold set by their predecessors. We operationalize the “concrete floor” as a three-step process: (1) a selector initiates the appointment of women (in any magnitude), which (2) is confirmed at the same or higher magnitude by the subsequent selector, and (3) is sustained at the same or higher magnitude by a third selector. Each step in the process requires an election that brings in a new selector. We identify concrete floors in each of our country cases and explain how they emerge and how they function as mechanisms for continuity in women’s representation. A final goal of our study is to explain why concrete floors vary across our countries.

Table 1.1 First Female Ministers and Last All-Male Cabinets, by Country

Country

  • Year

  • First Female Minister in Cabinet

  • Year

  • Last All-Male Post-Election Cabinet

Years between Appointment of First Woman to Cabinet and Last All-Male Cabinet

Germany

1961

1957

+4

Canada

1957

1968

–10

Chile*

1952

1970

–18

Australia

1949

1993

–44

United States

1933

1981

–47

Spain*

1936

1986

–50

United Kingdom

1929

1987

–58

(*) Includes only democratically elected governments

Why Are More Men Than Women Appointed to Cabinet?

The existing scholarship on gender and cabinet ministers provides substantial insight into ministerial appointments. First, this research confirms that few ministers are women. The relative absence of women from cabinets has meant that gender scholars have focused more attention on women’s representation in parliaments, where numbers grew (p.4) significantly by the 1980s in much of the world. More recent gains in women’s cabinet representation have prompted an expanding research agenda examining all aspects of women’s involvement in cabinet, from gendered patterns of appointment and removal,2 to gendered patterns of portfolio allocation (Barnes and O’Brien 2018; Escobar-Lemmon and Taylor-Robinson 2005), and gendered styles of policymaking (Atchison and Downs 2009; Atchison 2015; Escobar-Lemmon, Schwindt-Bayer, and Taylor-Robinson 2014).

Despite a rapidly expanding body of scholarship, there are two aspects of existing work that our book seeks to address. First, the findings of existing studies are mixed and therefore inconclusive when it comes to determining which factors account for larger proportions of women in cabinet. Second, along with scholars of ministerial recruitment, gender scholars have paid less attention to the process of cabinet appointment and more attention to outcomes, leaving us with gaps in knowledge about why fewer women than men become ministers.

Gender and politics scholars have employed supply and demand models, developed in research on legislative recruitment and its gendered consequences (e.g., Norris and Lovenduski 1995) to test four factors related to women’s cabinet representation. In translating supply and demand models to cabinet, scholars operationalize supply factors as (1) the proportion of women in the legislature and (2) the type of political system (presidential or parliamentary). Demand factors include (3) party ideology and (4) the sex of the selector. The evidence that supply factors affect women’s cabinet representation is mixed. While most studies find a positive correlation between the percentage of women in the legislature and the percentage of women in cabinet (Bauer and Okpotor 2013; Claveria 2014; Davis 1997; Krook and O’Brien 2012; Stockemer 2017), others find a relationship only in parliamentary but not presidential systems (Whitford, Wilkinson, and Ball 2007), and Bego, examining post-communist countries in Central and Eastern Europe, finds that larger percentages of women in parliament only predict the inclusion of at least one woman in cabinet (2014, 354). Adams, Scherpereel, and Jacobs (2016) examine women’s legislative and cabinet representation (p.5) in sub-Saharan Africa, finding a complex rather than straightforward relationship between women’s parliamentary presence and the number of female ministers. The evidence with respect to system type and cabinet appointments is likewise inconclusive. Most studies find that women’s cabinet presence is greater in non-parliamentary systems (Krook and O’Brien 2012) or countries with specialist recruitment norms, whereby ministers are selected for their policy expertise (Bauer and Okpotor 2013; Claveria 2014; Davis 1997), but a recent study of 194 countries finds that parliamentary systems have more women in cabinet than presidential systems (Stockemer 2017). Reynolds’ (1999) comparative research involving 180 countries found no effect for political system type.

Findings about the impact of demand-side factors are likewise inconclusive. Most studies find that leftist selectors appoint more women (Claveria 2014; Krook and O’Brien 2012; Siaroff 2000), but Davis finds that party ideology has no effect (1997, 58), explaining that leftist leaders appoint more women to cabinet because there are more women in parliament when left parties are in government. She concludes that “ideology plays an indirect role . . . it does not appear to play a direct role in recruitment of women to cabinet office” (1997, 65). Bego’s study of post-communist cases likewise finds no effect for party ideology (2014, 354–355). Studies of how selector sex affects women’s representation point in different directions. Davis (1997) and Reyes-Housholder (2016) find that female selectors appoint more women to their cabinets, but other studies challenge such findings (Krook and O’Brien 2012). O’Brien and her coauthors (2015) find that female party leaders in parliamentary democracies include fewer women than do their male counterparts. In sum, despite significant efforts by gender scholars to determine the correlates of women’s cabinet representation, the research findings are inconclusive.

Research into ministerial recruitment that does not focus on gender shares many similarities with studies of political elites more generally, focusing attention on who legislators and ministers are, rather than how they achieve their office. Studies of cabinet ministers emphasize their shared personal and political characteristics, including age, gender, years of parliamentary experience or former political experience, and, above all, party allegiance. Most studies use data on the social characteristics and political backgrounds of ministers to identify the common characteristics of ministers in specific countries or regions (Blondel and Thiebault 1991; Dowding and Dumont 2009, 2015; Escobar-Lemmon and Taylor-Robinson 2016; Vogel 2009). (p.6) Such studies confirm that men are vastly over-represented among ministers (Blondel 1985; Davis 1997).

Such studies tell us little about why the process of ministerial recruitment produces such skewed outcomes. Looking only at outcomes, that is, who ministers are, does not explain why some individuals are deemed qualified for ministerial office but others are not. The ministerial outcomes approach looks at ministers as an aggregate—appointments across time or across multiple countries, but this approach cannot take into account the fact that ministers are recruited as part of a discrete team. Ministers’ individual experience or expertise matters not only for its own sake, but also because of what it contributes to that team. That means some qualified individuals might be ruled out because there is someone with a similar profile already in cabinet. Looking primarily at outcomes also fails to help us explain the gendered composition of cabinets. As early as 1985, Blondel expressed disappointment that, despite pressures from second wave feminism, few women had made it to the ranks of cabinet (1985, 30), and he remained pessimistic that the numbers of female ministers would increase.

Explaining why women are disadvantaged requires a focus on the process of cabinet appointments. While the ministerial recruitment literature has little to say about process, the government formation literature is deeply concerned with the procedural and temporal dimensions of forming cabinets, although few scholars working in this area focus on the gendered outcomes of government formation. Researchers have examined the process through which certain combinations of parties come together to form a government in relation to coalition building. Game theoretical studies of government formation such as Riker (1980) and Laver and Shepsle (1996) see coalition formation as a process strongly guided by preferences and strategic decisions made by rational actors for office-seeking or policy-seeking motives. For them, the focus is on how politicians as rational actors bargain and negotiate the process of government formation to secure their party in the new government. This model for the process of government formation is most explicitly concerned with the partisan identity of those who form government (Laver and Shepsle 1996, 57), wherein politicians are primarily viewed as agents of the party to which they belong (Laver and Shepsle 1996, 42). They are less interested in the selection of individual ministers once the partisan composition of government has been decided. Although these studies help us to understand general strategic calculations involved in cabinet formation, they do not explain the gendered nature of the process or the gendered outcomes.

(p.7) While government formation studies tend to emphasize the strategic nature of the process, wherein party leaders and elites are rational actors seeking most the favorable outcome for their party, some scholars emphasize the rule-bound nature of this process. Strøm, Budge, and Laver challenge as “hardly realistic” (1994, 307) the premise of cabinet formation in coalition theory that treats political parties as if they are “unconstrained players in an institution-free world” (1994, 303). Instead, they argue that coalition formation is often “severely constrained by institutional arrangements and prior commitments” (1994, 303). For them, “government formation and maintenance are highly structured processes, and a variety of institutional features impinge on the choice set available to government formateurs” (1994, 305). The primary institutional constraints on coalition formation relate to the size or composition of the executive, rules of government investiture, and recognition rules (1994, 310–312).3

We find this recognition of the institutional context of the process of government formation useful. Although constitutions are relatively silent on the formation of cabinet, the absence of written rules does not mean that the process of cabinet formation operates devoid of rules. While Strøm, Budge, and Laver (1994) focus on formal rules, albeit distinguishing between hard and soft constraints (309), other authors acknowledge the importance of informal rules. According to De Winter, “the process of government formation is governed by a restricted number of formal provisions and a wide variety of informal rules” (1995, 122). Similarly, Mershon’s study of coalition formation acknowledges that, in the process of coalition bargaining, informal rules emerge alongside—and often at variance with—formal constitutional constraints. She argues, “politicians create informal rules to alter formal institutions that do not function to their benefit” (1994, 40). Politicians have incentives to invent informal rules.

Blondel (1985) also recognizes the importance of formal and informal rules in the selection of ministers. Rules are established “in order to give political advancement a more orderly outlook”; there must be “some recognition of the routes that should be followed to become a minister” (1985, 13). Blondel alludes to informal rules when he identifies various “profiles” of routes to ministerial office. These profiles “acquire the value of myths, in that they are felt to reinforce certain ‘ideals’ or ‘principles’ that political systems (p.8) and its leaders attempt to extol” (1985, 14–15). Significantly for understanding how gender functions in cabinet construction, Blondel recognizes how these profiles “also constitute many filters, which let some individuals come to the top while others are prevented from being selected” (1985, 15).

Drawing on the insights from existing scholarship, we investigate and explain ministerial recruitment as a process, and, more important, a process in which gender, understood as ideas about men’s and women’s entitlement to political power, operates and shapes outcomes at each stage. The process of cabinet appointment begins with the results of an election and ends with the formal announcement of a cabinet by the president or prime minister.4 We consider the entirety of the process to be strongly bound by rules, although, as we explain later, many of these rules are not codified in constitutions or statutes. Throughout the book, we identify all the sets of rules that determine the process of cabinet formation in seven democratic countries. We include rules that are codified and officially enforceable, as well as rules that are unwritten yet routinely observed. Some of these rules are at the system level, and others are at the party level. We identify the components of these sets of rules that advantage men and inhibit women’s access to ministerial office—in other words, rules that actively gender the construction of cabinets.

We further view ministerial recruitment as a process involving two sets of political actors: the selectors, including all actors with the authority to decide who will be invited to join the ministerial team; and the ministrables, defined as those political actors who are both eligible and qualified to serve in cabinet as ministers. While these political actors are strongly rule-bound, they are nonetheless able to use their agency strategically to work with or around the rules to shape outcomes in ways that reinforce or challenge gender hierarchies. Explicating the rules of ministerial recruitment will help us answer our two other research questions.

Why Are Women More Likely to be Appointed to Cabinet in Some Countries Than in Others?

Existing studies, and our own data, tell us that more women are appointed to cabinets in some countries than others. But we do not yet fully know (p.9) why. Some studies offer rich detail on a series of individual countries, including detail about the presence of women, but they are not strictly comparative (Dowding and Dumont 2009, 2015). In the gender and politics literature, many studies are comparative, exploring variation in women’s appointment to cabinet across countries. Some studies compare within a single region (Adams, Scherpereel, and Jacobs 2016; Bego 2014; Davis 1997; Bauer and Okpotor 2013; Escobar-Lemmon and Taylor-Robinson 2005, 2009). Most recently, Escobar-Lemmon and Taylor-Robinson’s Women in Presidential Cabinets (2016) provides a comparative focus of the backgrounds, connections, and credentials of cabinet ministers in five presidential democracies in the Americas: Argentina, Chile, Colombia, Costa Rica, and the United States. Other studies offer a more global comparative perspective (Bauer and Tremblay 2011; Claveria 2014; Krook and O’Brien 2012; Reynolds 1999; Scherpereel, Adams, and Jacobs 2018; Siaroff 2000; and Whitford, Wilkinson, and Ball 2007). These studies have considerably expanded our knowledge about women’s presence in ministerial office. But overall there is no single approach or consensus in findings, as scholars adopt different research designs, test different hypotheses, and operationalize differently the variables they believe to be relevant to explaining cross-national differences in women’s cabinet presence.

As noted earlier, cross-national variation in women’s cabinet presence is often attributed to differences in system type or modes of qualification. Researchers hypothesize, for example, that women are more likely to be recruited to cabinets in presidential systems because selectors are not limited to appointing from among members of parliaments (where there are always fewer women than men) when appointing ministers. According to Whitford et al. (2007, 564), “in a presidential system, the executive is less constrained in the appointments they make to cabinet positions and is free to place women in cabinet-level positions.” Scholars who opt instead to distinguish generalist from specialist recruitment norms are looking at something rather different, namely, whether ministers are recruited primarily for their political skills and experience, or for their policy-specific expertise.

Some cabinet scholars, including Rebecca Davis, acknowledge the artificiality of the generalist versus specialist distinction, noting, “in every cabinet system at any given time, there is a mix of ministerial backgrounds” (1997, 38). Davis also notes that “[d]ifferentiation between generalist and specialist systems does not mean that specialization does not occur in generalist systems or that generalists to not serve in specialist systems” (1997, 38–39). Her (p.10) study of women’s presence in cabinet in 15 Western European countries uses the generalist–specialist variable, but she ranks rather than dichotomizes her cases. Other researchers have used system type (Siaroff 2000) or rules about incompatibility of holding a cabinet portfolio and sitting in parliament (Bauer and Okpotor 2015) as a proxy for distinguishing between generalist and specialist recruitment practices. Whether scholars use political system type of parliamentary versus presidential, or recruitment criterion of generalist versus specialist, both parliamentary and generalist recruiting methods limit the supply of women.

Our data set includes seven countries: two presidential democracies (Chile and the United States); two parliamentary democracies (Germany and Spain); and three Westminster parliamentary systems (Australia, Canada, and the United Kingdom). There is no clear cross-national pattern of women’s cabinet inclusion in regard to political system type or recruitment criteria. Chile, with a presidential system, has had a parity cabinet, as have Germany and Spain, parliamentary systems, and Canada, with a Westminster system. In the US presidential system, rates of women’s recruitment to cabinet have remained comparatively low, as they have in the Westminster systems of the United Kingdom and Australia. Clearly, there is much that we do not know. Explicating the rules of ministerial recruitment specific to each of these counties, and the parties that form governments, as well as how the rules are gendered, will give us a more nuanced explanation for cross-national patterns of women’s presence in cabinets and for the gendered process of cabinet formation across countries.

Why Has Women’s Presence in Cabinet Varied across Time?

Because many gender and politics studies are cross-sectional rather than longitudinal, increases in women’s presence in cabinet have yet to be fully explained. Many longitudinal studies (Claveria 2014; Escobar-Lemmon and Taylor-Robinson 2005) begin their analysis in the 1980s or 1990s, when women’s cabinet presence was becoming more common, but they do not seek to explain the long periods of time when women were absent from cabinet and when women’s presence ranged between zero and one minister.

(p.11) By identifying the rules of ministerial recruitment and how they are used and shaped by empowered and strategic actors, we hope not only to show what the rules are, but also to identify whether, when, and how they have changed over time and what impact rule changes have on women’s cabinet presence. To understand how rules change, we draw insights from scholars who have been explicitly longitudinal in their approach. For example, Juan Rodríguez Teruel (2011a) has found that, over time, more ministers in Spain are being recruited from outside parliament and from regional political structures. Similarly, Kaiser and Fischer (2009) note the growing importance of regional political structures as a route into Germany’s federal cabinet. Other studies demonstrate the growing power of presidents and prime ministers over time (e.g., Poguntke and Webb 2005). Although these studies do not deal with gender, they explain changes in rules of ministerial selection and qualification, which, in turn, may help to explain shifting patterns of women’s recruitment.

Other studies address over-time variation in women’s presence in cabinet (Claveria 2014; Davis 1997; Escobar-Lemmon and Taylor-Robinson 2005), finding that patterns of women’s ministerial appointment are more volatile than those for parliamentary elections (e.g., Scherpereel, Adams, and Jacob 2018). Election outcomes rarely lead to the complete removal of all members of parliament in a single party; election loss is usually only partial for any parliamentary party.5 Although an election loss may still leave a party with a degree of continuity of members in parliament, at the cabinet level an election loss means a complete change of ministerial personnel. The volatility of cabinets compared to parliaments means that conclusions about women’s cabinet presence focusing on a single point in time are problematic, especially if system-level factors are being tested (Bauer and Okpotor 2013; Krook and O’Brien 2012; Reynolds 1999; Siaroff 2000; Whitford et al. 2007).

Scherpereel et al. (2018, 3) address volatility in women’s cabinet presence, identifying differences as “backsliding” and “climbing.” They define backsliding as “the difference in representation between t1 and t if the change is negative; climbing is defined as the difference in representation between t1 and t if the change is positive.” The authors find differences between advances in women’s legislative representation and women’s cabinet presence, with “women’s legislative representation . . . characterized by a (p.12) ratchet effect” and women’s cabinet representation characterized by a “see-saw effect—an asymmetric process in which one year’s representational gains erode in subsequent years” (2018, 1; emphasis in original).

In contrast to Scherpereel et al. (2018) we find little evidence of significant backsliding across time in women’s inclusion in cabinets across our cases. Rather, we find that over time, gains in women’s ministerial recruitment tend to be sustained across initial cabinets formed post-election. Once a prime minister or president appoints a woman (or some women) to cabinet, subsequent selectors in the same country tend to appoint at the same magnitude; some even increase the number of women included in cabinet. As noted earlier, we call this lack of volatility in numbers of women appointed to cabinet as the “concrete floor.”

Data, Approach, and Main Contributions

The existing scholarship on ministerial recruitment equips us with the initial tools for explaining the patterns we see in our data: women are under-represented in cabinet; under-representation varies cross-nationally; and it changes over time. The following section sets out in more detail the patterns our data reveal, the approach we adopt to explain them, and the contribution we make to the scholarship on cabinets and executives.

The Patterns in the Data: Timing, Magnitude, and Persistence

We collected data on every post-election cabinet in seven countries, from the date of the appointment of the first female cabinet minister in each county to the end of 2016. By post-election cabinet, we mean the first cabinet appointed following a general election. The decision to focus only on post-election cabinets stems from our starting assumption that cabinets are not simply a collection of unrelated individuals; rather, a cabinet is a team made up of a collectivity of persons who bring different skills and talents to government. As a result, in constructing our data set, we focus on those ministers who are appointed to a cabinet together, as a team, at a one specific point in time, namely, following an election when a leader is forming the initial cabinet. When forming their initial cabinets, selectors (p.13) are making and balancing numerous appointments, as many as 30 in some cases. Cabinet appointments made after the initial cabinet formation often involve considerations beyond the control of the prime minister or president and may involve factors that are unrelated to the politics of cabinet appointment, such as the death or resignation of a minister. Moreover, cabinet reshuffles normally involve only a handful of aspirant ministers; hence, competition for a small number of posts may be higher than is the case in formation of post-election cabinets.

Finally, for parliamentary systems, our data set includes only those cabinets formed by prime ministers following a general election. We are not counting those cases where new prime ministers emerge following the resignation of a prime minister or a party leadership contest. We exclude cabinets formed, for example, by Prime Minister Kim Campbell in Canada in 1993 or Prime Minister Theresa May in the United Kingdom in 2016. Both leaders assumed the prime ministership after winning party leadership contests. Neither leader was positioned to appoint a completely new cabinet, nor were they authorized to do so by the general electorate. When new prime ministers emerge via party leadership contests and subsequently win a general election, we include the cabinet formed after that election, such as the 2010 election of Australian prime minister Julia Gillard.

We count ministers rather than portfolios. For this study, we are more interested in the gendered inclusion of women in cabinet and less concerned with the distribution of cabinet responsibilities;6 where a cabinet minister holds more than one portfolio, we count that minister, male or female, only once. We acknowledge that in parliamentary systems, prime ministers are also members of the cabinet; however, we do not include prime ministers in our count of cabinet members. Prime ministers achieve their position through a different procedure than ministers and are selected as party leaders by their party selectorate, authorized thereafter by success in a general election. Prime ministers are the selectors of ministers for their cabinet, but they do not select themselves.7 We focus on the appointment process and on the selector separately, as an actor who works with the rules to select ministers rather than as a member of the cabinet.

(p.14) For the purposes of this book, we count and compare the number of women appointed to each post-election cabinet and calculate the share of women in each cabinet as a percentage. Our data set contains the size of cabinet and number of women appointed following all democratic elections upon the appointment of the first female cabinet member. This means that the number of cabinets formed in each country varies according to (1) the date of the first female minister appointed, (2) the onset and the duration of democracy, and (3) how many elections have been held. We coded each post-election cabinet by the party allegiance of the main selector, as left, right, or center and by the selector’s sex, either male or female.

An important goal of this book is to identify changing patterns in the recruitment of women to cabinet across time. Although most longitudinal studies begin their analysis in the 1980s or 1990s, when women’s cabinet presence was becoming more common (Claveria 2014; Escobar-Lemmon and Taylor-Robinson 2005, 2016), we take a longer historical view of gendered appointments over time, including long periods of time when women were absent from cabinet and when women’s presence ranged between zero and one minister. Our temporal frame of analysis begins with the appointment of the first female minister in each country (see Table 1.1). Women’s first inclusion in cabinets ranges from as early as 1929 in the United Kingdom and 1933 in the United States to as recently as 1957 in Canada and 1961 in Germany. These dates are significant turning points in the composition of each country’s cabinet, representing the first time the practice of appointing only men to cabinet is broken. It is important to note, however, that even after a woman was initially appointed to cabinet, all-male cabinets continued to appear regularly in many countries, and it took decades for women’s inclusion to move beyond a single token female minister (see Table 1.1). To detect and explain patterns over time, we classify women’s cabinet representation based on three dimensions: timing, magnitude, and persistence.

By timing, we are interested in how early or how late women’s inclusion into cabinets beyond a token presence occurs. In other words, we want to know when selectors stop appointing cabinets with zero or one female minister. By magnitude, we mean the percentages and numbers of women and men in cabinet. We categorize the magnitude of women’s cabinet presence as an ordinal variable, ranging from (1) very low (0 to 9.9 percent of cabinet), (2) low (10 to 19.9 percent), (3) medium (20 to (p.15) 29.9 percent), to (4) high (30 percent or more) and (5) parity 50 percent +/–1 minister. We construct these categories from the research on women’s parliamentary presence, where “the story of the story of critical mass” (Dahlerup 2006) identified 30 percent as an important turning point for women’s political representation, although not necessarily one that would produce substantive changes in political outcomes. At the lower end, we categorize women’s cabinet representation as “very low” if women constitute fewer than 10 percent of cabinet ministers (see Dahlerup and Leyenaar 2013). This category reflects the reality of many government teams worldwide, where only one woman sits in the cabinet. Given the comparatively small size of cabinets (e.g., 15–25 ministers), once two or more women are appointed to cabinet, most country cases fall into the “low” or “medium” magnitude categories. Finally, by persistence, we are interested in how much backsliding occurs, that is, once reaching a new magnitude of women’s inclusion, say, moving from the category of “low” to the “medium” category, does a country regress to a lower magnitude?

We categorize cabinets both by the number of women and the percentage of women within each, across time. Because cabinets are small (the average size of a cabinet in our data set is 18) and the size of cabinet can vary within countries, the number of ministers is as important to note as the percentage. A small positive change in the number of women can lead to a large change in their percentage representation within a cabinet, and a reduction in the size of the cabinet can lead to an increase in the percentage of women, even if the number of women remains constant. Finally, given the small size of cabinets (e.g., relative to parliaments), cabinets of uneven numbers cannot produce pure parity cabinets, that is, those in which numbers of women and men in cabinet are equal. As a result, we define parity cabinets as those in which women (or men) constitute a range of 50 percent +/–1 minister. For example, in Germany, Chancellor Gerhard Schröder’s 2002 cabinet included six female ministers in a 13-member cabinet (46 percent). Because cabinets often contain an odd number of ministers, a perfect 50 percent representation of women (or men) is not possible. The addition of one more woman in cabinet would have increased the presence of female ministers to 53 percent. By identifying gender parity cabinets as a range of 50 percent +/–1, we are able reasonably to classify gender parity cabinets of odd as well as even numbers of ministers. As a result, we include Schröder’s 2002 cabinet as an example of a gender parity cabinet.

(p.16) Our data reveal two gendered patterns in terms of timing, persistence, and magnitude. In one group of countries—Canada, Germany, Spain, and Chile—women’s cabinet inclusion beyond a single female minister is early, persistent, and of relatively high magnitude. Among this group of countries, the era of cabinets with zero or one female minister ended decisively, without regressing and regardless of the party of the selector. The last German post-election cabinet where women’s presence was lower than 10 percent was in 1983; in Canada it was in 1980, and in Spain it was in 1986. Chile, which did not have democratically elected governments in the 1980s, left behind the era of very few women in cabinet in 1990, which was the last time when women constituted less than 10 percent of ministers. These countries also demonstrate a consistently upward trajectory since surpassing the “very low” threshold, reaching the “high magnitude” category of at least 30 percent by 1998 in Germany, 2000 in Chile, 2004 in Spain, and 2008 in Canada. Notably, each of these countries has had at least two cabinets with more than 30 percent women and each has had at least one gender parity cabinet.

The second group of countries—Australia, the United Kingdom, and the United States—displays a different gendered pattern. Although the United Kingdom (1929) and the United States (1933) experienced the earliest inclusion of a woman in cabinet, the appointment of more than one woman in any subsequent cabinet occurred significantly later. The inclusion of women in cabinet has been more uneven, in terms of persistence, and, with the exception of the United Kingdom’s cabinet formed in 2015, has not reached the “high” category in terms of magnitude. Australia, the United Kingdom, and the United States were also the last countries to leave the “very low” category, with women constituting less than 10 percent of cabinet ministers as recently as the 1990s. Australia continued to have cabinets where women constituted fewer than 10 percent of ministers into the 1990s and as recently as 2013; women were present at the lowest levels in post-election cabinets in the United Kingdom until 1997, and in the United States until 1993.

It is also worth noting that in both Australia and the United States, a pattern of including zero or just one female minister was not broken decisively at one point in time. The US case is one of slow startup, with several all-male post-election cabinets (from 1961 to 1977, and in 1981); only in 1977 and from 1993 to the present have post-election US cabinets included more than one woman. None of these three countries has had a gender parity cabinet. (p.17) These two patterns of women’s inclusion in cabinet by timing, magnitude, and persistence are visually summarized in Figure 1.1.

Explaining Gendered Patterns of Cabinet Formation

Figure 1.1. Women s cabinet appointments (magnitude percentage), across time by country.

Although our cases differ in terms of timing, magnitude, and persistence of women’s inclusion in cabinets, the data also show at least one striking similarity: women’s presence is increasing across all countries. Australia is an outlier in that it regressed to the “very low” magnitude category, dropping below 10 percent of women in cabinet in 1998 and 2013. Across all our other countries, every post-election cabinet since the early 1990s has included at least two women, and by the late 1990s, all countries (except Australia) have no post-election cabinets with fewer than three women. In fact, we find a clear pattern of moving with relative consistency from low to medium to high magnitude of women’s presence in cabinet. There is no evidence of backsliding across more than one category (for example, from high back down to low). Likewise, although parity cabinets rarely follow each other in a consecutive fashion, especially when government changes hands, parity cabinets are likely to be followed by cabinets in which women are represented at the “high” or very close to high magnitude. Although the United States has yet to reach the “high” magnitude category, women’s presence in post-election cabinets has ranged between three and four ministers for a period of more than 20 years (1993–2016). We are thus confronted with an important puzzle: Why do we see so little volatility in women’s appointments to cabinet within each country given the structural potential for tremendous volatility?

Three Dimensions of Cabinet Appointments: Process, Actors, and Rules

To explain gendered patterns of ministerial recruitment cross-nationally and over time, we developed a model of cabinet appointment composed of three dimensions: (1) ministerial appointment is a dynamic process (2) involving two sets of actors, namely, selectors and ministrables, both of which are (3) governed by rules (many of which are unwritten). Explaining who becomes a minister requires that we shift the analytical focus away from the outcome and to the process instead. This process of ministerial recruitment starts with the announcement of election results, by which point the selector/s are known, and ends with the formal announcement of a cabinet. We propose that this process is strongly gendered, and throughout the book (p.18) we illustrate the manifold ways that ideas and expectations about how men and women should behave structure the process and shape who is selected as minister. The cabinet formation process involves two sets of actors: those who have the authority to select ministers, whom we call selectors, and those who are eligible and are seeking to qualify for ministerial office, whom we call ministrables. We argue that the final construction of a cabinet can be explained by the actions of two sets of actors, by the interaction between them, and by the interaction of rules that prescribe, prohibit, and permit certain courses of action.8 Moreover, we acknowledge that these actors themselves are gendered (Gains and Lowndes 2014). As men and women, they experience the gendered assumptions and expectations of other actors, and they themselves hold views about the appropriate roles of men and women in private and public life.

We also acknowledge that these actors are not merely behaving strategically to maximize benefits to themselves or their parties. Rather, we argue that the process of cabinet construction is very much rule-bound, despite the fact that most rules are unwritten. Formal (codified) and informal (unwritten) rules in political systems and political parties construct the institutional context within which ministers are selected. Rules empower and constrain the capacity of selectors to choose autonomously, and rules determine who is eligible and, crucially, qualified for ministerial office.

Finally, we show how these rules affect men and women differently. Following Gains and Lowndes (2014), we consider whether there are rules about gender, rules with gendered effects, and how gendered actors work with the rules of ministerial recruitment to produce specific outcomes. Although the process of cabinet appointment has few rules that are specifically about gender, we show that many of the rules that determine the process of ministerial recruitment have profoundly gendered effects. Specifically, we show how rules allocate power in ways that have historically prioritized men’s appointment to ministerial office over women’s. Our goal in the book is to identify the components of these rules that have functioned to advantage men and to inhibit women’s access to ministerial office.

Change over time is an important dimension to this analysis. The last two decades have seen dramatic gains in women’s appointment to cabinets across all countries, albeit at different magnitudes. Significantly, increases in (p.19) women’s recruitment to cabinet have been sustained across all countries in our data set. For each country, we show when and how rules change, focusing on the crucial role played by agency, ambiguity and ideas (Mahoney and Thelen 2010). As we argue in Chapters 7, 10, and 11, ideational change across time that increases the numbers of women in cabinet, and that persists, can establish powerful informal practices. Agency, ambiguity, and ideas have changed informal rules that have shaped women’s access to cabinet positions, for example, by expanding some of the qualifying criteria for becoming ministrable, to women’s benefit, creating a concrete floor for women’s cabinet inclusion.

Throughout the book, we show that patterns in cabinet appointment can be explained by (1) the rules about selection, which mostly empower presidents and prime ministers to choose their cabinet teams relatively unimpeded by other actors, (2) as well as rules about which individuals are eligible and qualified for a ministerial post. We find that rules about selection, eligibility, and qualifications are more consequential than political system type, party ideology, or sex of the selector. One of the strongest rules of cabinet appointment for most, although not all, of our cases is that selectors can choose ministers relatively autonomously, with little need to consult or negotiate with other political actors. Although we identify numerous rules about eligibility and qualifications that structure the pool of ministrables, we also find that the vast majority of these rules are non-codified. We argue that the intersection of rules about who selects and who can be selected explains how the process of cabinet formation is gendered and produces gendered outcomes, and we model this process in Chapter 2.

Because many of the rules about selection and qualifications are ambiguous and non-specific, they produce significant space for political agency. The degree of discretion enjoyed by most selectors helps to explain why deeply entrenched patterns can change rapidly and suddenly, namely, when specific selectors use their selection powers to increase the number of women in cabinet dramatically. Once a selector takes such a bold step, a new dynamic is unleashed, and the effects of a selector’s political agency at an earlier point in time may become locked in. For each of our cases (except Australia), we can identify such a turning point, brought about by a particular selector, or set of selectors, who initiate a new concrete floor in terms of women’s cabinet presence.

In cases where selectors use their political agency to take bold steps to increase the share of female ministers, they do so either out of personal (p.20) conviction, because of pressure from feminists within their party, and/or in response to changing ideas and representational rules about what a “balanced” cabinet looks like. Such ideas are often informed by changing notions of women’s roles in society and politics. When an individual selector puts new ideas into practice, by setting a new threshold in terms of women’s cabinet presence, the institutional context for subsequent selectors likewise changes, and new inclusionary base points are locked in. Once new standards of inclusion are set, subsequent presidents and prime ministers are held to account, we find, by other political actors—in particular, intraparty feminists—and by the media. It is now common, when a new cabinet is announced, for the media to note how many portfolios were given to women.

Ultimately, we have an optimistic story to tell. Across most of our cases, combinations of formal and informal rules intersect to create significant opportunities for political actors—namely, presidents and prime ministers acting as selectors, as well as feminists within parties—to bring about change. In Chapter 11, we show that instances of significant backsliding do not occur once political actors who enjoy discretion to select ministers put inclusionary norms with respect to gender into practice and set new concrete floors.

Our gendered institutionalist approach exposes the rules of ministerial recruitment in the seven advanced Western democracies countries in our study, but it also has broader applicability. This framework and method, we anticipate, can be used to identify the rules of selection and qualification for ministerial office in other regions and political systems, including countries with semi-presidential systems, and those with strong traditions of coalition government. The approach can also be used to expose the rules that determine the patterns of inclusion of other under-represented groups based on social class, race, ethnicity, sexual orientation, or gender identity. More broadly, our research provides a blueprint to guide scholars in how to analyze institutional settings that are largely devoid of formal rules, namely by specifying (1) the number of informal rules at play in an institutional setting; (2) whether rules prescribe, prohibit, or permit; (3) whether rules are specific or flexible; and (4) the degree of institutionalization of the rules.

Although we believe our model is generally applicable to all cases with governing cabinets, it is ultimately an empirical question whether the causal mechanisms producing concrete floors for women’s inclusion exist (p.21) in countries without democratic political systems. Although some authoritarian leaders have promoted women’s rights as a way to garner political support (Donno and Kreft 2018), it is possible that the incentives for leaders to follow informal rules about what a representative or democratically legitimate cabinet team looks like do not exist in highly authoritarian contexts.9 We hope that those studying women’s representation in non-democracies see value in applying our model to identify the rules of ministerial recruitment.

Mapping the Book’s Chapters

We begin in Chapter 2 by setting out our theoretical, conceptual, and methodological approaches for explaining gendered patterns of cabinet appointments. We employ a feminist institutionalist approach (see Kenny 2014; Mackay, Kenny, and Chappell 2010; Waylen 2014), emphasizing the significance of studying informal as well as formal rules, paying particular attention to how such rules create and maintain gendered hierarchies that advantage men in the cabinet appointment process; we also show how rules can change, emphasizing the importance of agency, ambiguity, and ideas; and we offer a model of the relationship among sets of rules to produce cabinets that include women. Chapter 2 also addresses the methodological implications of our feminist institutionalist approach and details our “qualitatively driven” approach (Mason 2006, 10) for analyzing qualitative data, including interviews, media data, and memoirs. Finally, we provide the justification for the selection of our cases, and present descriptions of the seven country cases.

The empirical chapters of the book expose the rules of ministerial recruitment and are structured as follows. First, in Chapters 3 through 7 we identify the rules of ministerial recruitment, explicating the variety and patterns of rules-in-use (Ostrom 2005, 20) in our seven country case studies. In Chapters 8 through 10, we demonstrate how these rules are gendered, explaining why rules of cabinet appointment produce different opportunities (p.22) for men and women to be selected as ministers and how this varies across countries and over time.

In Chapter 3, we focus on who selects ministers, identifying all the rules that determine how empowered or constrained selectors are to choose their ministers. We show that in most cases, presidents and prime ministers are strongly empowered to choose their ministers, though in some cases selectors are constrained by having to share powers with co-selectors from other parties and within their own party organization. In Chapters 4 though 7, we turn our attention to the rules about how individuals who are in the eligibility pool for cabinet can demonstrate that they are also qualified to be selected as ministers. Challenging the idea that there are objective and universal qualifying criteria or “merit” that all individual ministers must display, Chapter 4 identifies three different types of qualifications that cabinets, as collectivities, must include: qualifications involving experiential criteria (political experience and policy expertise); affiliational criteria (membership in a selector’s personal network of friendship, trust, and loyalty); and representational criteria (membership in a relevant political, territorial, or social group).

In Chapter 5, we focus more deeply on experiential criteria, demonstrating that in all cases ministrables qualify by demonstrating their political experience or policy expertise—and sometimes both. We argue, however, that such experiential criteria are applied post hoc and strategically as “merit” to justify a selector’s choice of minister. We argue that qualification for appointment on the basis of policy expertise and/or political experience is insufficient to explain selection of ministers; if all ministrables meet some kind of experiential criteria, additional qualification criteria will be necessary to determine who is selected and who is not.

In Chapter 6, we identify the importance of affiliational criteria (membership in the selector’s personal networks of friendship, trust, and loyalty) as a historically strong and enduring mechanism for determining who in the eligibility pool qualifies for cabinet. In Chapter 7, we show that ministers can qualify for cabinet by meeting representational criteria—membership in a relevant political, territorial, or socio-demographic group—that are deemed important to legitimize the cabinet.

In Chapters 8, 9, and 10, we turn our analysis explicitly to gender in order to demonstrate how the rules we identify in Chapters 3 through 7 distribute opportunities to male and female ministrables unevenly. We show that rules for the process of ministerial recruitment have traditionally been (p.23) gendered to men’s advantage, producing patterns of cabinets in which men are represented in significantly higher numbers than women. In the context of gendered rules, selector agency, feminist activism in the parties, and changing ideas about women’s inclusion in politics have produced change across countries and over time.

In Chapter 8, we show how rules empowering and constraining selectors have gendered consequences for selectors and ministrables. The fact that, often, a single actor is empowered to select ministers means that the president or prime minister has the agency to choose to appoint women to cabinet. Indeed, we find that many commit to acting on that agency, evidenced by a pre-election pledge. Chapter 8 also examines whether the sex of the selector and increasing the number of selectors make a difference to the gender composition of ministerial teams.

Chapter 9 analyzes the gendered consequences of rules about qualification and investigates whether there are differences in the qualifications that male and female ministrables bring to a cabinet team. Although the type of qualifications of male and female ministers do not differ substantially, our data show that selectors employ experiential criteria strategically to justify their preferred appointments. Chapter 9 also addresses the gendered impact of selectors’ ability to appoint ministers on affiliational grounds. Appointment on affiliational criteria, we argue, best explains the reproduction of male predominance in cabinets over time.

Chapter 10 addresses representational criteria. Across all seven country cases, the extension of gendered representational criteria to include women has become a strong predictor of women’s inclusion in cabinet. The timing and strength of institutionalization of gender as a representational rule varies cross-nationally. Nevertheless, its presence is strong enough to shape a selector’s choices of ministers, often reducing his capacity to appoint on affiliational grounds.

In Chapter 11, we bring the analysis of rules and gender presented in Chapters 3 through 10 together to answer our three research questions: Why are more men than women appointed to cabinet? Why are women more likely to be appointed to cabinet in some countries than others? Why has women’s presence in cabinet varied across time? To do so, we explain how rules about selection and qualification have interacted to produce the specific outcomes in each country case and over time. We explicate in detail the establishment of the “concrete floor” in each country, identifying who initiated an increase in women’s presence in cabinet and (p.24) explaining the mechanisms through which the concrete floor has been sustained and maintained to become the new norm. We conclude with a discussion of the contribution and implications of our research for the practice of constructing cabinets, as well as for future executive and feminist scholarship in this field.

Notes:

(1) Among our cases, only Germany has consistently included at least one woman in initial cabinets for more than 50 years (since 1961).

(2) The preponderance of research on gender and cabinets focuses on women’s appointment to and, to a lesser extent, removal from ministerships. See Adams, Scherpereel, and Jacobs (2016); Bauer and Okpotor (2015); Bauer and Tremblay (2011); Bego (2014); Claveria (2015); Davis (1997); Escobar-Lemmon and Taylor-Robinson (2005, 2009, 2016); Krook and O’Brien (2012); Reyes-Housholder (2016); Reynolds (1999); Scherpereel, Adams, and Jacobs (2018); Siaroff (2000); Stockemer (2017); Whitford, Wilkinson, and Ball (2007).

(3) For discussion of country-specific rules of government formation, see Dowding and Dumont (2009); Rasch, Martin, and Cheibub (2015).

(4) In this book, we do not include cabinet reshuffles or cabinets formed following a change in party leadership in the absence of an election.

(5) A notable exception is the 1993 federal election in Canada. Prior to the election, the governing party, the Progressive Conservatives, held 156 seats. The party lost all but two seats in the election.

(6) For recent research on the gendered distribution of cabinet responsibilities, see Barnes and O’Brien (2018); Escobar-Lemmon and Taylor-Robinson (2016).

(7) Thus, for example, we would categorize Thatcher’s cabinets as having no women since, as selector, Thatcher is not counted among those selected.

(8) Our definition of rules as prescribing, prohibiting, or permitting certain courses of action comes from Ostrom (1986) and is further developed in Chapter 2.

(9) That said, it is notable that in 2013, Iran’s president felt compelled to give in to widespread public criticism of his failure to appoint any women to his cabinet. Following sustained criticism on social media, President Rouhani appointed a woman to his cabinet and promised to promote more women to other government posts (Moghadam and Haghighatjoo 2016).