This article conceptualizes an innovative understanding and measurement of women's political leadership, theoretically justifies its application, and analyzes contemporary variation in its patterns through comparative case studies. In recent years, scholars of comparative government have studied with great interest the election of female prime ministers and presidents (e.g., Derichs and Thompson Reference Derichs and Thompson2013; Jalalzai Reference Jalalzai2013) and cross-national variation in female members of parliaments (MPs) and cabinets (e.g., Bauer and Tremblay Reference Bauer and Tremblay2011; Paxton and Hughes Reference Paxton and Hughes2017; Suraj, Scherpereel, and Adams Reference Suraj, Scherpereel and Adams2014). Yet, when it comes to regions beyond Europe and the Americas, comparative empirical analysis of women's political leadership (WPL) across national-level governments has been largely neglected. Addressing this gap in the literature, we offer a new index that we believe has multiple advantages over the most commonly applied proxies for WPL.
The approach taken here is motivated by a desire to go beyond potentially misleading indicators, such as the presence of dynastic female chief executives or parliamentary gender quotas, which frequently make it appear as though women have more power in government than they actually do. Problematically, studies that focus upon a single institution may mistakenly conclude that women's presence in one branch of government is indicative of their having power in other branches as well. However, we posit that a balance of women's leadership, meaning the presence of women leaders in all relevant institutions or branches, is more revealing in regard to women's institutional capacity to exert power. Therefore, we have developed a women's political leadership index (WPLI) to facilitate descriptive, comparative, causal, consequential, and intertemporal analyses of WPL by capturing a holistic snapshot of women's political power in a country at a given time.
Although the numbers of female politicians have increased worldwide over the past several decades, from inching close to parity in Nordic parliaments to exceeding parity in Rwanda (Adler Reference Adler, Ngunjiri and Madsen2015; Dahlerup and Leyenaar Reference Dahlerup, Leyenaar, Dahlerup and Leyenaar2013), often via “fast-track” strategies that rely on reserved seats and candidate gender quotas (e.g., Dahlerup and Freidenvall Reference Dahlerup and Freidenvall2005; Ismail, Rasdi, and Jamal Reference Ismail, Mohd Rasdi and Abd. Jamal2011; Jalalzai and Krook Reference Jalalzai and Krook2010), women still make up a small minority of presidents, prime ministers, cabinet ministers, governors, mayors, and high court judges around the world. As of January 2018, women accounted for 23.4% of parliament members globally (IPU 2018), but in many parts of the world, women are hampered by multiple barriers to entry into politics and a host of discriminatory practices once they achieve office, ranging from threats, harassment, and intimidation to gender-biased workplace norms, discourses, and hierarchies (Dahlerup and Leyenaar Reference Dahlerup, Leyenaar, Dahlerup and Leyenaar2013; Goetz Reference Goetz, Goetz and Hassim2003; UN Women 2014).
Contributing to the study of women's political leadership, we offer an approach that differs from most of the existing literature. Instead of taking individual politicians as the unit of analysis, we treat women's political leadership as contingent upon the theoretical concepts of “formal structural power” and “male dominance,” which necessarily encompass multiple apex political decision-making institutions. Drawing on this perspective and incorporating data collected from Asia, the world's most populous continent, we developed the WPLI by combining gendered measures of formal leadership across judicial, executive, and legislative branches.
As subsequently discussed, conceptualizing and measuring women's political leadership in this way may better equip us to test theoretical hypotheses about empirical relations at the systemic level and potential threshold impacts of a “critical mass.” As one of the most long-standing and debated ideas in gender politics, the critical mass hypothesis hinges upon the distinction identified by Kanter (Reference Kanter1977) between a “token” presence (in which women may be too few to change the existing patriarchal structure of an institution) and “non-token” presence (in which a sufficient number of women are present to potentially change group behavior and make an institution more friendly toward women). For example, in an influential study of parliaments during the mid-1990s, Schwindt-Bayer and Mishler (Reference Schwindt-Bayer and Mishler2005, 422) observed that women's presence mattered greatly, as they found that “the percentage of women in the legislature is a principal determinant of women's policy responsiveness and of women's confidence in the legislative process.” In other words, numbers and proportions of women may greatly impact their substantive and symbolic representation. Yet scholars have also pointed out how analyses of critical mass are often theoretically underspecified or lack sufficient empirical data to test properly (Dahlerup Reference Dahlerup, Campbell and Childs2014). Critical issues include not just how many/what share of women constitutes a critical mass but also where and when does a critical mass matter the most? To begin answering these questions, we believe it is important to move beyond merely studying the presence of women leaders in any single institution to look at the balance (or balanced presence) of women in leadership positions across all relevant institutions.
Hence, we offer the WPLI as a more refined and comprehensive alternative to previous attempts to measure women's political leadership cross-nationally, which, for the most part, have neither specifically aimed at nor succeeded in measuring WPL collectively. Probably the closest and most widely known attempts currently in operation are (1) the “political power” component of the United Nations Development Programme's (1995) Gender Empowerment Measure, which includes female seats in parliament, and (2) the “political empowerment” component of the World Economic Forum's (2015) Global Gender Gap Index, which is based on female seats in parliament, female cabinet ministers, and years with a female head of state out of the last 50 years. While we concur that cabinet ministers are indicative of executive branch leadership, we find these measures to be deficient overall, for several reasons. First, the number of women in parliament fails to capture leadership within parliament and in other government branches. Second, simply looking at chief executives over time does little to illuminate the nature of women's leadership as myriad factors influence a chief executive's rise to power. Third, the judiciary is an important component of political leadership ignored by most existing metrics. Hence, our approach goes beyond women's presence in any single political leadership arena to incorporate the balance of women's presence across multiple apex political decision-making positions.
The rest of the article is structured as follows. Starting with a review of some salient findings on WPL from previous research, we explain and theoretically justify the empirical composition of our leadership index. We then apply it to 21 diverse territories in Asia. To demonstrate the utility of our approach, we conduct a series of comparative country case studies to illustrate four prominent patterns of WPL: exclusion, illusion, imbalance, and balance. Lastly, we conclude by discussing broader implications of presence and balance in women's political leadership and the potential benefits of globally expanding the scope of the WPLI.
LITERATURE REVIEW
Conceptualizing and measuring women's political leadership are both unique and worthwhile endeavors. As the growing literature on women political leaders demonstrates, the landscape facilitating or preventing women from entering politics is complex and not yet fully understood. Among possible causal explanations, cultural factors (including patriarchy, egalitarianism, performance orientation, collectivism, and power distance) often statistically trump socioeconomic variables and political institutions. For example, a recent analysis of 181 countries concludes that “national culture explains as much of the variance in women's participation in politics as all other factors combined” (Bullough et al. Reference Bullough, Kroeck, Newburry, Kundu and Lowe2012, 407). A common conjecture is that traditional societies favor men as political leaders but this bias dissipates in postindustrial societies, where many women are well educated and employed in the formal sector (Norris and Inglehart Reference Norris and Inglehart2004). As more time passes after women first enter leadership positions, a society familiarized with women exercising political power may be more willing to s/elect women into office (Reynolds Reference Reynolds1999, 567).
Yet, while this sort of predictable trajectory may prove true at local levels of government or even in national parliaments, cultural associations between “authority and masculinity” might continue to prevent women from reaching the highest positions in politics (Ismail, Rasdi, and Jamal Reference Ismail, Mohd Rasdi and Abd. Jamal2011, 388), in line with beliefs that masculine qualities are synonymous with the duties of state and its leadership (McDonagh Reference McDonagh2009, 6). While many studies conclude that egalitarian societies are more favorable to women's advancement (e.g., McDonagh Reference McDonagh2009; Norris and Inglehart Reference Norris and Inglehart2004), even in postindustrial societies with relatively higher levels of gender equality, prevailing prejudices that women are “less competent” or “less worthy” leaders (Carli and Eagly Reference Carli and Eagly2001, 631) tend to exclude women from top political leadership positions (Jalalzai Reference Jalalzai2008, 229).
Just as social and cultural factors influence the ability of women to enter politics, the structures of political systems and parties impact women's ability to obtain leadership positions. Women chief executives are more likely to appear in parliamentary as opposed to presidential systems, especially in dual executive systems, where power is shared with another executive or otherwise dispersed (Jalalzai and Krook Reference Jalalzai and Krook2010, 12). In short, women are less likely to become the chief executive when that position commands a greater share of authority. In terms of political parties, ideology matters greatly. Egalitarian-oriented parties on the left tend to support more women candidates than parties on the right (Joshi and Kingma Reference Joshi and Kingma2013). Additionally, the organizational structure of political parties and “gender-related candidate rules” impact the number of women nominated, their ranking on party lists, and, ultimately, the proportion elected to parliament (Caul Reference Caul1999, 79). While gender quotas can increase the numbers of women in office, this depends on quota type and degree of enforcement (Jalalzai and Krook Reference Jalalzai and Krook2010; Krook Reference Krook2009). Even in progressive parties, leadership of the women's wing is often the highest position that a woman can reach (Derichs Reference Derichs, Derichs and Thompson2013, 303), and there is little effort by most parties to push women into higher leadership positions or assist women in obtaining the minor offices necessary as progressive steps up the political ladder (Genovese Reference Genovese, Genovese and Steckenrider2013, 340).
CONCEPTUALIZING AND MEASURING WOMEN'S POLITICAL LEADERSHIP
While the research discussed above offers a range of interesting hypotheses about political leadership, it often treats individual leaders or institutions as the unit of analysis. By contrast, our focus is not on individual “female political leaders” or a single political institution but on “women's political leadership,” which we define as the presence and balance of women leaders across the full range of formal political apex decision-making institutions in a given polity. This conceptualization derives from structural theories of institutional power, and we understand patriarchy to be nothing less than a structure of power. Einspahr (Reference Einspahr2010, 12) explains patriarchy as a structure of domination
which systematically reproduces a gendered power imbalance and which systematically privileges men as a group in relation to women as a group, constituting a background condition of domination whether or not individual men “choose” to exercise their power and whether or not individual women experience the most overt forms of such power in their lifetimes. (emphasis added)
In this respect, leadership of political institutions matters since leaders play a key role in perpetuating, reforming, and transforming gender-biased institutions, which may be difficult or slow to change on their own (Mackay Reference Mackay, Campbell and Childs2014). The gender of political leaders is significant because of the pervasiveness of male dominance, which is a relationship among collectivities or “the relative positions of individuals (as members of groups) vis-à-vis other individuals (as members of groups) and institutions” (Einspahr Reference Einspahr2010, 4).
WPL is a phenomenon that challenges one particular (but structurally very important) aspect of male dominance—that is, who (in terms of sex) holds political leadership positions. As Duerst-Lahti (Reference Duerst-Lahti and O'Connor2010, 22, 25, 28) argues,
In general, the more power—especially structural power—there is associated with a particular position, the less likely a woman is to hold the position … Institutions are not particularly conducive to women, but the only way the institutions will change is through the politics of their presence … their very presence means that they inevitably change the institution itself.
Therefore, in order to understand the degree, type, origins, and impacts of WPL, we first need to measure it across an ensemble of institutions that possess and delegate power at the national level. As Bourque (Reference Bourque, Freeman, Bourque and Shelton2001, 85) points out, “political institutions structure public and private life as well as the distribution of resources and opportunities,” therefore “access to leadership in those institutions is critical for any group that wishes to shape the public agenda, or … reshape the distribution of resources and public goods.”
Following this conceptualization, in order to measure women's political leadership systematically in cross-cultural contexts, we examine the proportions of women in government bodies with the decision-making capacity to substantively influence the rules and norms that have a significant impact on people's everyday lives. As Goertz (Reference Goertz2006, 6) argues, since most important social science concepts are multidimensional and multilevel, one should first provide a coherent definition of the underlying concept, as done here, and then break down the concept into empirical components before identifying variables for which data can be collected for each of these dimensions. Thus, we operationalize the concept of “women's political leadership” with respect to the total apex government leadership positions filled by women vis-à-vis men, incorporating both women's presence among leaders and the balance of women leaders across institutions. Notably, our approach is motivated not by what kind of data are easily available but by a political and sociological understanding of power and patriarchy as structural in nature, which implies that the question of political leadership must likewise be seen in structural terms—especially following Michels's (Reference Michels1915) insight that political leadership in any society or organization necessarily takes an oligarchic form.
Presence and balance mean that we take into account not only the absolute number of women compared to men in positions of leadership but also their balanced representation across multiple leadership bodies. What matters is the share of political leadership positions held by women relative to men in all important formal political decision-making branches of government (both appointed and elected posts), including the legislative, which makes laws; executive, which implements policies; and judicial, which interprets laws and adjudicates conflicts. Thus, we examine the leadership positions that formally exercise the most power among the executive, judicial, and legislative branches.
Because we conceptualize WPL in terms of both the presence of women in leadership positions and the balance of their representation across institutions, for the executive branch of government, we examine the sex of all cabinet-level ministers as opposed to a lone chief executive.Footnote 1 Cabinet ministers represent the highest levels of formal leadership across the most powerful departments and agencies in the state's bureaucracy, thus each of these ministers holds a position of leadership with potential to have considerable influence in introducing or vetoing new legislation and regulations (Bauer and Tremblay Reference Bauer and Tremblay2011). As Atchison and Down (Reference Atchison and Down2009, 6) highlight, cabinet ministers “control substantial portions of the national budget and are responsible for not only initiating but also implementing policy” as well as advancing their own policy agendas and stifling those they oppose.
For the legislative branch, we examine the sex of leaders (chairpersons) of parliamentary committees as opposed to all MPs. This is because the committee structure of parliaments generally bestows committee leaders with greater power over legislation compared to ordinary MPs. Parliamentary committees are present in almost all national assemblies (Longley and Davidson Reference Longley and Davidson1998), and committee chairpersons have a much greater ability to introduce, alter, or veto legislation (Heath, Schwindt-Bayer, and Taylor-Robinson Reference Heath, Schwindt-Bayer and Taylor-Robinson2005, 421). Despite the importance of these positions, research on women's committee assignments is notably lacking (Baekgaard and Kjaer Reference Baekgaard and Kjaer2012, 466), and this indicator has thus far not been incorporated into any widely disseminated cross-national measure of women's political leadership.
Within the judiciary, we consider the sex of members of the national supreme court or highest appellate court in the country. Since the decisions of the highest court cannot be appealed to any other judicial authority in most countries, supreme courts have considerable power to shape public policy and the agendas of other political actors (Yanus Reference Yanus and O'Connor2010). As Hoekstra (Reference Hoekstra2010, 482) notes, when attempting to understand women's political leadership, “equal emphasis should be placed on understanding the lack of gender diversity on courts, especially high courts.”
As our women's political leadership index aims to capture both the presence of women in leadership positions and the balanced representation of women leaders across three branches of government, we have developed two primary indicators: (1) an “index score” consisting of a modified geometric mean of women's proportion of collective leadership across three components—the national judiciary (supreme court), executive (cabinet), and legislative branch (parliamentary committee leaders)—and (2) a country's minimum share on any of these three components. The former provides us with a singular measure to broadly compare balanced leadership across countries, while the latter provides additional insight into which of the four patterns of leadership discussed later best describes WPL in a given country at a given time.
Following Munck (Reference Munck2009, 49–50), we chose the geometric mean as the best aggregation rule to apply because we believe the relationship between our three components is interactive (since one government branch may be able to veto another) and partially compensatory (since one branch can set or shape an agenda that impacts the others). However, we apply an arithmetic mean of the three components for those countries where one of the component values is zero because the geometric mean would also return a zero value, which we feel would be too punitive given the partially compensatory nature of the different institutions.Footnote 2 Hence, we label our index score as derived from a “modified geometric mean.” As our goal is to maximize both conceptual validity and reliability, we provide further details on our data collection, coding, and aggregation in Appendix A in the supplementary material online. Following Coppedge et al. (Reference Coppedge2011), our underlying disaggregated data are publicly available to other researchers who might prefer to use other aggregation methods or construct alternative indices.Footnote 3
Our focus here is also specifically on the presence and balance of women's political leadership at the top echelons of government as opposed to women's representation. We do not count the numbers of women politicians in general (descriptive representation) or examine particular actions taken by women in politics (substantive representation). Representation (in politics) has two connotations that distinguish it from leadership: first, women may be descriptively represented in a government's executive (bureaucracy), legislative (national assembly), and judicial (judges) branches while also (relatively) absent from the leadership of the executive (cabinet), legislature (chairs of parliamentary committees), and judiciary (supreme court justices). Second, as discussed by Pitkin (Reference Pitkin1967) and Phillips (Reference Phillips1995), the concept of “representation” is multifaceted and includes dimensions such as formal, descriptive, substantive, and symbolic. Here we are looking only at who holds leadership posts in the branches of government and what share of those leaders are female.
To illustrate our index, we apply it to Asia, home to the majority of the world's women. Our work incorporates insights and data from studies on gender and leadership of Asian political executives and parliaments (e.g., Derichs and Thompson Reference Derichs and Thompson2013; Joshi and Och Reference Joshi and Och2014; True et al. Reference True, Niner, Parashar and George2012) while additionally incorporating cabinet portfolios, parliamentary committees, and high court judges, which thus far have generally been absent from comparative studies of women in Asian politics. Table 1 displays the WPLI modified geometric mean and minimum component scores (between 0.00 and 1.00) for 21 territories in 2017. It also includes the share of women as a proportion of leaders in the judiciary, executive, and legislative branches. As displayed in Appendix B in the supplementary material online, we have also calculated the WPLI for 2015. Between 2015 and 2017, one can observe increases in the WPLI geometric mean (0.11 to 0.12) and women leaders in the judiciary (0.13 to 0.16) and legislature (0.10 to 0.14) but a slight decline of women leaders in the executive (0.12 to 0.11).
Women's political leadership index and its components, 2017
Source: Authors’ data set.
Table 2 ranks these territories by 2017 modified geometric mean WPLI index scores and categorizes countries in regard to both presence and balance. Women are considered to have more than a token presence in leadership positions if their overall modified geometric mean is 0.125 or greater.Footnote 4 Similarly, we consider women's leadership in a country to be balanced if the component score is greater than 0.125 for each of the three branches of government. Taking both presence and balance into consideration, we label each territory examined here as having either “balance,” when women have a non-token leadership presence across all three government branches (Philippines, Taiwan, Macau); “imbalance,” when women have a non-token presence overall but not in all branches (Kazakhstan, Georgia, Japan, Indonesia); “exclusion” (the most common pattern in Asia), when women have at most a token presence in leadership as a whole; or “illusion,” a special subset of exclusion cases in which the public has the impression that women are more active in leadership because one or more women are or were prominent in the leadership of one branch of government, such as a current or previous chief executive (i.e., prime minister or president as in Hong Kong, Bangladesh, South Korea, India, and Pakistan).Footnote 5 We now turn to a comparative case analysis to make greater sense of these dynamics.
2017 WPLI scores and four categories of leadership presence and balance
Source: Authors’ data set.
As Table 2 reveals, the difference between female members of parliament (FMPs) and WPL varies across countries, but in most cases, women's share of parliament is higher than the WPLI index score; the most extreme cases are Afghanistan and Pakistan. We observe a positive correlation between the WPLI index score and FMPs (r = 0.57) and a modest correlation between women leaders in the judiciary and legislature (r = 0.32) but essentially no relationship between women leaders in the judiciary and executive (r = 0.01) or the executive and legislature (r = 0.02).Footnote 6 These results indicate that the three government branches have some independence from each other and WPL cannot be fully captured by indicators that examine only one or two of these branches.
COMPARATIVE CASE ANALYSIS
Our comparative case analysis explores a descriptive typology for the overarching concept of WPL based on the two dimensions of presence (token or non-token representation in women's collective share of leadership) and balance (equal or unequal presence of non-token women leaders across government branches), as illustrated in the 2 × 2 matrix in Table 3. Here we draw upon Kanter's (Reference Kanter1977, 966) notion that a presence of less than about one-sixth (about 15%) of a group's members often relegates them to “token” status within a group; however, since we are looking at top leadership positions, we consider it reasonable to lower this threshold slightly to one-eighth (12.5%). Our four categories or “ideal-types” aim to be collectively exhaustive and mutually exclusive, but, as pointed out by Collier, Laporte, and Seawright (Reference Collier, Laporte, Seawright, Box-Steffensmeier, Brady and Collier2008, 157), we recognize that sometimes real-world cases are situated at the intersection of categories.
Typology of women's political leadership
Drawing from the countries for which we have constructed the WPLI, we offer four illustrative case studies, each of which captures a different pattern of women's political leadership: exclusion, illusion, imbalance, and balance. Here we employ descriptive-exploratory (as opposed to explanatory) case studies to uncover empirical nuances, identify relevant context, and probe potential sources of causation as a means of generating hypotheses to be tested in future research (Yin Reference Yin2009, 18–20). Using a “structured, focused comparison” (George and Bennett Reference George and Bennett2005) and “replication design” (Yin Reference Yin2009, 53), each case study provides a contextual description with broader information and greater detail on women political leaders than what the WPLI alone provides. It then probes possible causal factors, such as culture, institutions, and socioeconomic modernization (see Schwindt-Bayer and Squire Reference Schwindt-Bayer and Squire2014) emerging from scholarly analyses by leading experts on WPL in these countries. By using the method of paired comparisons (see Varshney Reference Varshney2002), we are also able to control to some extent for dominant religious and cultural influences, namely, Islam in Southwest Asia and Confucianism in Northeast Asia.
Exclusion
In some countries, women are generally excluded from political leadership. In such instances, a lack of women in leadership roles also means there can be no balance of women leaders across branches of government. A representative case is the Islamic Republic of Iran, with a 2017 WPLI index score of 0.00. From 1979 to 2017, all of Iran's presidents, supreme leaders, chief justices, and parliamentary committee chairs were men. In 2017, no women served on Iran's supervisory 12-member Council of Guardians or its Assembly of Experts, which is composed of 86 high-ranking mujtahed clerics who are entitled to interpret Islamic law. In the executive branch, there were no women among 18 cabinet ministers, although women held three vice chair positions in minor ministries.
Prevalent in the professions, women make up one-third of Iranian government employees, two-thirds of university students, and are present on elected municipal and provincial councils (Vakil Reference Vakil2011), but they make up only 6% of Iranian parliament (Majlis) members. Iranian women are legally entitled to vote and drive, and roughly three-quarters of Iranians favor women having full equal rights with men (Vakil Reference Vakil2011, 18).
How, then, do we explain the consistent absence of women from national-level political leadership positions in Iran over the past four decades? According to Vakil (Reference Vakil2011, 8), “the state poses the biggest obstacle” and has refused to introduce any gender quotas. As Osanloo (Reference Osanloo2009, 31) notes, the regime's political structure is based on gender segregation, “reconditioning” women's status in order to “distinguish post-revolutionary Iran” from the previous government's “perceived capitulation” to the West. Although women actively participated in the 1979 Revolution, women have been subjected to widespread and significant restrictions based on the application of Sharia. Women have made considerable gains in education while government reforms have addressed inheritance, insurance, custody, and divorce laws, but aside from a brief rise in women holding higher-level political posts during Mohammad Khatami's presidency (1997–2005), the state continues to support polygamy, pro-natality policies, and the intimidation of women's rights activists (Vakil Reference Vakil2011, 9). As a result of patriarchal interpretations of Islamic law, it has been difficult for women to hold political leadership posts, since even acts such as refusing to wear a chador in parliament have disqualified female MPs from their position (Vakil Reference Vakil2011, 71–72).
Illusion
A second pattern of women's political leadership occurs when women have a prominent leadership position in only one branch of government but have merely a token presence or are entirely absent from leadership positions in other branches, thus lacking balance. In such cases, women leaders in the single branch often have highly visible roles, such as the chief executive, thereby creating the “illusion” that women's political leadership is strong when in fact political leadership is still overwhelmingly in the hands of men.
This pattern is exemplified by the case of Pakistan, where a woman, Benazir Bhutto, twice served as prime minister (1988–1990, 1993–1996). Women currently make up 21% of Pakistan's lower house MPs, and one-third of all local government seats are reserved for women. However, despite Bhutto's tenure as prime minister and Pakistan's long history of reserved legislative seats for women dating back to the 1950s (Krook Reference Krook2009, 60), there is widespread exclusion of women from top leadership positions across all branches of government. In 2017, there was not a single woman among the country's 20 cabinet members, 31 leaders of parliamentary committees, or the 15-member Supreme Court of Pakistan, resulting in the same WPLI score (0.00) as Iran.
Unlike Iran, however, Pakistan has gender quotas for its local and national assemblies, it has competition between political parties, and its state is not a theocracy. But experts have argued that the overwhelming paucity of women leaders in all branches of Pakistan's government is related to a male-dominated political culture, in which 60% of all Pakistani voters are men (UN Women 2014) and resistance to women in politics is framed under the rubric of culture, with tactics ranging from “special announcements over mosque loudspeakers that voting by women was un-Islamic” in some areas to “passing local laws permitting the houses of female voters—not even just candidates—to be demolished” in the North-West Frontier Province (Krook Reference Krook2009, 73, 78).
While reserved seat quotas create the impression of inclusion, Pakistan's political system is permeated by “highly gendered and patriarchal structures,” particularly the political parties, military, Jirga councils, clergy, police, judiciary, and single-member district electoral system (Müller Reference Müller, Ayaz and Fleschenberg2009, 166). Violence against women politicians is pervasive, and few women enter formal politics. Those who do are almost exclusively members of elite political families and dynasties and beneficiaries of reserved seats (UN Women 2014). Roughly four out of five women in parliament since independence have held reserved seats. In the 2002 elections, when reserved seats were reintroduced and increased in the provincial legislatures (to 15% of seats), National Assembly (to 60 of 342 seats), and Senate (to 17 of 100 seats), few women contested them because reserved seats are allocated by political party leaders.
In a context in which mostly all-male Central Executive Councils of political parties determine candidate nominations, women are regularly sidelined to the relatively powerless “women's wing” of their respective political parties (Müller Reference Müller, Ayaz and Fleschenberg2009, 174). As a result, the women nominated to reserved seats in the National Assembly or the provincial assemblies “tend to be proxy representatives and are invariably relatives of politicians,” who often lack political and professional experience, and frequently party leadership limits their meaningful participation in assemblies once elected (Ali Reference Ali, Ayaz and Fleschenberg2009, 189). As Fleschenberg (Reference Fleschenberg, Derichs and Thompson2013, 73) notes, “Handicapped by the concept of purdah, which permeates all spheres of women's lives in Pakistan, women are discouraged from active participation in public places, such as decision-making … Female political representation and participation follows the principle of co-opting women through reserved seats and is characterized by an elite-oriented appendage phenomenon.”
Imbalance
A third pattern of women's political leadership is “imbalance,” where women have a non-token presence among government leaders on the aggregate, but a lack of balance exists because their non-token presence is limited to one or two government branches. Japan, with a 2017 WPLI score of 0.14, provides a good illustration of this pattern. In 2015, women had a non-token presence on Japan's Supreme Court (20%) and cabinet (22%) but chaired no committees in the lower house of its parliament (the Diet). As of 2017, two parliamentary standing committees (Education, Culture, Sports, Science and Technology; Economy, Trade and Industry) are now chaired by women, but they are a clear exception, as 27 out of 29 total standing and special committees in the Japanese House of Representatives have male chairpersons. The numbers of female leaders are similar but their proportions are higher in the Supreme Court (2 out of 15 judges) and cabinet (3 out of 18) ministers with women holding the portfolios of Defense; Internal Affairs and Communications; and the Tokyo Olympic and Paralympic Games.
What might explain Japan's pattern of imbalance? Prime Minister Shinzo Abe has publicly claimed to support increasing female representation in the ruling Liberal Democratic Party (LDP), which has governed the country for most of the period since 1955. Yet experts argue that “quota adoption has not been a mainstream priority,” and the LDP “continues to marginalize women in the party” (Gaunder Reference Gaunder2015, 176). As in Pakistan, there is a weak pipeline for women in Japan to gain leadership experience. Japanese women make up a small share (less than 10%) of both the national parliament and local government assemblies, with 40% of local councils and prefectural assemblies outside major cities featuring no female legislators (Shin Reference Shin, Vianello and Hawkesworth2016, 348). Very few chief executives are women. In 2009, only 3 out of 47 governors were women, and a 2003 national tally of female mayors recorded only 6 in 677 cities and 3 in 2,562 towns and villages (Gelb Reference Gelb and O'Connor2010, 390). Marginalization of women also occurs in the bureaucracy (a common feeder position for top political leadership positions in Japan), where “only about 1 percent of female officials hold posts in the three highest grades” (Iwanaga Reference Iwanaga and Iwanaga2008, 102).
As in Pakistan, the high costs of election campaigns in Japan limit women's chances to become MPs, but unlike in Pakistan, women have made up the majority of Japanese voters since the 1960s and there have been increased opportunities for women in the proportional representation tier of parliamentary elections since Japan's shift to a parallel (mixed-member majoritarian) electoral system in 1994. However, scholars argue that Japan's “1.5 party” system traditionally has not been particularly supportive of including and promoting women among its leadership (Iwanaga Reference Iwanaga and Iwanaga2008, 115). Although the LDP now claims to support the proliferation of women among its ranks and did experience some gains under Prime Minister Junichiro Koizumi (2001–2006), “the party organization does not contain any institutional mechanisms to nominate women candidates,” and within Japan, “the weakness of the Left has reduced the pressure to legislate quotas to increase the number of women candidates” (Gaunder Reference Gaunder2015, 182, 177). For example, the Japanese Communist Party nominated many women candidates in 2012 but won few seats, and the center-left Democratic Party of Japan brought 26 new women into parliament in 2009, but none was reelected in 2012 (Gaunder Reference Gaunder2015, 181). Thus, while some Japanese prime ministers may put forth the image of a gender-balanced government, closer analysis reveals a lack of genuine initiatives to promote women's leadership in a diversity of political bodies.
Balance
A fourth pattern of leadership occurs when women have a non-token presence in all three important government leadership bodies. This pattern, which we label “balance,” is illustrated by the case of Taiwan (Republic of China). Taiwan has one of the highest WPLI scores (0.24) in Asia, with women political leaders present in a non-token capacity on its Supreme Court (4 out of 15 judges), in the cabinet (4 out of 28 ministers), and in parliamentary committee leadership (3 out of 8 committee chairs), partly because of a requirement that chairmanship of the latter is always convened by two MPs who are frequently rotated. Like Pakistan, Taiwan has experienced a woman chief executive (President Tsai Ing-wen, elected in 2016), a sizable share of women in both local government (33%) and parliament (37%), and reserved seats for women since the 1950s. However, whereas reserved seats have come and gone in Pakistan, they have been sustained and augmented over time in Taiwan's local and national elections (Huang Reference Huang2015). Taiwan's two major political parties also have additional candidate selection quotas for non-reserved seats stipulating a minimum percentage of female candidates (Huang Reference Huang2016, 329). Moreover, an increasing number of women political leaders in Taiwan do not come from elite political families or political dynasties (Gelb Reference Gelb and O'Connor2010, 388).
What might explain the balanced pattern of women's political leadership observed in Taiwan? A number of experts point to institutional changes spearheaded by activists affiliated with the women's movement successfully making inroads into the major political parties. As Chiang (Reference Chiang and Iwanaga2008, 81) notes, “for decades, the number of female lawmakers in Taiwan has gradually increased in the national elections … women have gradually challenged male supremacy in a male-dominated sphere in order to pursue the same opportunities as men and have thus reached their work goals as political leaders at the top levels of political institutions,” which, in turn, has “dramatically changed the political environment and landscape.” Longer-term cultural differences between the Taiwanese population and other Asian nations may also be an underlying cause. Like Japan, Taiwan has experienced considerable growth of women in education and the labor force, but Gelb (Reference Gelb and O'Connor2010, 389) finds that Taiwanese political culture has featured “greater receptivity to women's demands for greater representation than in many, more traditional Asian nations.” For instance, during the 1970s and 1980s, the opposition movement during the authoritarian era and forerunner of the current Democratic Progressive Party (DPP) featured women leaders, the most prominent of whom (Hsiu-lien Annette Lu) later became Taiwan's first female vice president. In the democratic era, the DPP also took the lead in pushing for women's advancement by adopting a 25% internal party candidate quota for women in 1996. In the context of a primarily two-party system, this put pressure on the Nationalist Party (KMT) to take similar measures after it lost the 2000 presidential election, resulting in a “contagion” effect with both parties ratcheting up their commitments to appoint women candidates and promote women within their parties’ operating structures (Huang Reference Huang2015).
Political parties also appear to have played a role in transforming the political culture. By the mid-2000s, women made up about one-third of the Central Executive Committee members for both the DPP and KMT (Gelb Reference Gelb and O'Connor2010, 388). These gains facilitated the shift to a parallel mixed electoral system and a 2005 constitutional amendment guaranteeing 15% of total seats (50% of proportional representation seats) in the parliament (Legislative Yuan) to women (Huang Reference Huang2015, 213). Other institutional changes have been set in motion since a 2004 cabinet-level gender commission appointed by the DPP “passed a resolution that all the policy consultative committees and participatory commissions at all ministries should adopt one-third gender quotas whereby each sex should hold at least one-third of the seats in those committees and commissions” (Huang Reference Huang2015, 216). Compliance with this rule has been high and is now monitored by the Department of Gender Equality, which was set up in 2012. Given that women form a critical mass (more than 30%) among leaders of the major political parties and government commissions, it may not be surprising that the DPP's leader, Tsai Ing-wen, recently became the nation's president.
DISCUSSION
The national case studies discussed here with respect to the presence and balance of women in formal political leadership capture representative instances of a “true positive” (balance) in Taiwan, where WPL is at a non-token level across all government branches; a “false negative” and “mixed case” (imbalance) in Japan, where WPL is uneven across government branches but higher than indicated by numbers of women in parliament; a “false positive” (illusion) in Pakistan, where WPL is actually minimal but may seem high because of parliamentary gender quotas and having had a female prime minister; and a “true negative” (exclusion) in Iran, where WPL is transparently absent or minimal across all government branches. In our view, such differences in the presence and balance of WPL are important features in their own right (especially if one branch can veto another or set an agenda that impacts others), but this is obscured if we look only at a single institution.
These case studies also reveal how the gender of chief executives and women's share of parliament members are misleading indicators of women's political leadership, whereas the WPLI gives us a better indication of male dominance in political systems.Footnote 7 Using the World Economic Forum's (2015) Gender Gap Index “political empowerment” scores, Pakistan (0.127) and Iran (0.037) look very different, but in fact the two countries are quite similar in that women as a whole are basically excluded from most political leadership positions in both countries. For our second pair of cases, superficially Japan (0.103) and Taiwan (not scored by the World Economic Forum) might seem similar, but, as illustrated here, Taiwan is more inclusive of women in political leadership positions than Japan. Uncovering these discrepancies suggests that the WPLI makes a positive contribution and advancement over previous efforts to measure WPL.
Our exploratory case studies have thus unearthed several important findings. We have demonstrated how alternative approaches overestimate WPL in countries such as Pakistan and underestimate it in countries such as Taiwan. They may also breed unrealistic expectations that quotas in a single branch of government alone will propel major changes across the political system. Additionally, they may underestimate the power of organized religion and a conservative clerical establishment in deterring WPL through different means—that is, directly through control of the state (Iran) or indirectly through cultural and social influence (Pakistan). Other approaches might also overestimate the power of socioeconomic modernization and secularization, which have opened greater opportunities for women in Taiwan than Japan. Though further study is needed to test our preliminary findings here, in general, our case analyses point toward paying greater attention to change and stasis in political culture and gender ideology, which may explain the behavior of political parties and direction of institutional reforms.Footnote 8
As the expansion of the WPLI to the global level to include all countries would be a valuable next step, we have already anticipated and conceived the following responses to possible critiques of our approach. First, one may argue that our conceptualization of women's political leadership excludes representation or participation of women who are not from the political elite. Our response to this challenge is that women's leadership is not the same as women's empowerment (see Sundström et al. Reference Sundström, Paxton, Wang and Lindberg2017). The former concept indicates that those who are leaders happen to be women, whereas the latter concept refers more broadly to (all) women gaining power in (all) spheres of society. While the two are related, they are not the same, and the former is no guarantee of the latter.
Second, some may say that focusing on formal government positions ignores other avenues by which power is exerted in society. Our response is that while dominant social norms or nongovernmental and informal actors may exert power, formal leadership positions in government still matter as they are recognized structural positions of power. As Kellerman (Reference Kellerman, Ngunjiri and Madsen2015, viii) notes, “we tend to associate the word leadership with formal positions of authority in domains such as politics,” and as Einspahr (Reference Einspahr2010, 14) recognizes, “the state is instrumental in maintaining the conditions of women's domination as well … States assure male domination by excluding women (and other oppressed groups) from political processes altogether.” While we readily acknowledge the contribution of women's movements and other activist networks to social change, we believe it is important for analytical purposes to distinguish civil society from formal political positions in government, as the two are not identical.Footnote 9
Third, some may contend that women who are political leaders do not necessarily support the women's movement or feminist causes. While this may be true, we make no assumption that women holding formal leadership positions will advance the interests of women more than men. As noted earlier, we conceptualize “women's political leadership” as distinct from “feminist political leadership” and “women's political effectiveness.”Footnote 10 The extent to which women in leadership positions are politically effective and champion women's interests are falsifiable empirical hypotheses deserving of careful study, especially since we have theoretical reasons to believe they are linked. Though as Genovese (Reference Genovese, Genovese and Steckenrider2013, 343) notes, even after rising to the highest office, no female chief executive “has challenged in any fundamental way, the patriarchal power structure of her society,” it might be the case that women political leaders will be more supportive of women's interests when they are present in greater numbers, which is something we can test with the help of the WPLI.
Fourth, one may assert that using a uniform quantitative measure of women's leadership may be problematic since the meaning of women's leadership varies across societies. As Chant (Reference Chant2006, 210, 211) points out, the numerical bias of quantitative measures can “occlude important dimensions of meaning and quality” and limit components “not only to those that are observable and quantifiable, but that are actually quantified.” Our response is that the first point is applicable to all quantitative indicators and means only that we should always approach them with caution rather than summarily dismissing all numerical measures. Regarding the second point, we are driven primarily by theoretical considerations and have gone out of our way to observe, quantify, and compare leadership components (particularly legislative committee chairs) that have not been previously studied in the region we examine. Furthermore, our WPLI explicitly incorporates a more diverse and robust set of quantifiable measures of women in leadership compared with widely deployed indices, which helps account for varying meanings of women's leadership across society.
CONCLUSION
In this article, we have introduced the WPLI as a means to measure and compare countries in terms of women's political leadership at the national level. In doing so, we have conceptualized and measured WPL to include both the presence of women leaders and the balance of women leaders across formal government posts, and we have developed a corresponding index to improve over previous attempts to compare WPL across countries. Here we are explicitly trying to move away from a singular measure of one branch, such as the number of women in parliament, because such measures fail to capture a holistic picture of women's political leadership. Measures such as the number of women in parliament can be easily swayed by quotas or other imposed rules that artificially inflate the number of women in such positions but do not necessarily translate into women in top leadership positions or WPL being more highly valued.
Upon expanding the global reach of the WPLI, we hope that it can be used in the future to support the enterprise of better understanding the causes and consequences of women's political leadership. We believe it might also help us make progress in developing, testing, and refining theories about putative sex ratio proportions or critical mass thresholds beyond which descriptive representation may differentially impact substantive representation (Beckwith Reference Beckwith2007). It may also aid us in understanding how WPL relates not only to institutions and specific individuals, but also to culture and power structures more broadly (Hazarika Reference Hazarika2008). The WPLI presented here is an important first step to aid further studies on the social, cultural, and structural factors that limit or otherwise impact women's political leadership overall.
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S1743923X18000120.
