CHAPTER 1
The Problem
NEW JERSEY’S GOVERNOR CHRIS CHRISTIE was a darling of the Republican Party when he rolled into the town of Princeton for a public relations blitz in November 2011. The governor had just declined calls to run for the highest office in the land and was riding high in opinion polls. On this day, he was putting his political capital to work in a town meeting with some 150 residents. Christie lumbered into the local public library’s common room and spoke in his trademark pugnacious style. He voiced some themes of his agenda and then opened the floor to the residents of Princeton. The lowliest, the poorest, and the most powerless people in the town now had the chance to tell the governor directly whatever they wished him to know.
Christie chose a public meeting as the optimal setting to publicize his agenda and score political points. Even in today’s mass media society, meetings are a backbone of politics. Not only do they air pressing matters, they also symbolize the heart and soul of a community. In a recent national survey, when Americans were asked “the best way for officials to learn what the majority of people in our country think about important issues,” they gave a resounding approval to town meetings, choosing them as their favorite means. By a ratio of 2 to 1 they preferred town meetings to the second favorite, opinion polls.1 Americans want their political system to be a meeting democracy.
This is not merely wishful talk. Between a quarter and a third of American citizens participate in deliberative settings in any given year—as many as contribute to campaigns, far more than the number who work for a political campaign or engage in protest, and nearly as many as the percentage who contact a public official.2 Aside from all this discourse directed by or at government, it seems that everywhere Americans turn, a meeting of their fellows is waiting for them.3 Americans talk in groups with their compatriots at work, at their religious institutions, or in their voluntary associations—clubs, charities, homeowner associations, neighborhood groups, professional and trade organizations, and PTAs. Of the three-quarters of Americans who belong to an association of some kind, over a quarter serve on its board (Burns, Schlozman, and Verba 2001, 76).4 Most Americans affiliate with a church, and most of them serve on its board at some point (90).5 Americans are committed to committees. Long ago, the famous French observer Alexis de Tocqueville noted that Americans love to gather together, talk, and act for a common purpose (Tocqueville [1835] 2006, 192).6 The iconic moment in the birth of our nation, the Boston Tea Party, came to life at a public meeting in Old South Meeting House (Young 2011). Americans are still joiners, lamentations of declining bowling leagues notwithstanding. But they are more than joiners; they are meeters. The organizations they join still rest on the foundation of the public meeting. And what people do at meetings is talk.
But what no one noticed at the meeting with Governor Christie is that in this poster child of democracy, where all could speak their mind to all, one-half of the population spoke twice as much as the other half. That quiescence mattered; the louder half received over twice the governor’s time in response.7 And perhaps surprisingly, this pattern had nothing to do with race or class or political party; the quieter half—the population that spoke little and to whom little was said—was the female half.
Two speakers illustrate the dynamic at the meeting, and the pattern we will document in this book. One speaker was a twenty-something student, liberal, articulate, and passionate. The other speaker was also a twenty-something student, also liberal, articulate, and passionate. Both voiced clear positions and stated relevant facts. Both struck a civil tone as they challenged the governor on a salient issue of the day. But the female speaker took only 63% of the floor time of the male. And she received 29% of the response time devoted to the male. While all the citizens in attendance were equal, some were more equal than others—but not for the reasons we commonly suspect.
It is one of the best documented facts of American politics that groups with less power and authority in society are less likely to participate in politics (Verba, Schlozman, and Brady 1995). But we typically think of these groups as defined by lines of race, ethnicity, language, income, age, or education. In modern America, observers of politics focus on the political disadvantages of groups with less resources: those left out of the privileges of class, those still stigmatized by race, those struggling as new arrivals from far-off places. But the quiescent half at the meeting was none of these groups. To the contrary, as residents of one of the wealthiest towns in America, they were members of the country’s moneyed and social elite. They were the “1%,” or close to it. But they were the female “1%.”
The women of Princeton are members of some of the top households of the most powerful nation in the world, yet in the setting that Americans idealize as the heart of democracy, they took half the floor time of the men and got half the attention of the highest official of their state. When Ralph Waldo Emerson described the meetings in his town of Concord, Massachusetts, he wrote that every person “has his fair weight in the government” (Mansbridge 1983, 133). By our notions of fairness, Emerson was wrong about the men of Massachusetts, most of whom were denied the ability to carry any weight in government by exclusions of property, race, or age (132–33). Clearly, at a time when women were excluded from eligibility, he was not even thinking of women. The America of today is not the America of Emerson. But what goes with little notice is that even now, in a historical age when women have made great strides in American society, and when women vote in greater numbers than men, America’s female citizens greatly underparticipate relative to men in key areas of politics and other public settings of decision making.8 Women lack a “fair weight in the government.”
Princeton is not your typical American town, but its gender inequality is far from an anomaly. The only comprehensive study of “open, face-to-face, legislative democracy in general purpose governments” is Frank Bryan’s book, Real Democracy (2004, 216). Bryan, a political scientist and fan of New England, took it upon himself to crisscross every last acre of Vermont out of sheer passion for town meetings. He and his students can recount in loving detail what transpired in each of a jaw-dropping 1,389 meetings in that state. Bryan found that although women composed 46% of attenders, they contributed only 28% of the speaking turns to the average meeting. Most men speak; most women don’t.9 In only 8% of the meetings Bryan studied were women more likely to speak than men.10 Attending the meeting is not the same as speaking up. Speech is an act of political participation in its own right. And while women are dutifully showing up, they are not actively participating.
GENDER INEQUALITY AND DESCRIPTIVE REPRESENTATION IN DELIBERATIVE SETTINGS
Women make up 50.1% of the world’s adult population, but 20% of its leaders.11 At first thought, this massive gap is easily discounted as yet another unfortunate dysfunction of the many third-world countries that populate the globe. But the gap does not naturally go away as a society achieves economic and social progress.12 Twenty percent is just as accurate in the United States as in Turkmenistan.13 And 20% is the number not only in the highest level of national government, where the ranks of women might be thinnest, but at the local level as well, in towns all across the United States and many other developed nations (Crowder-Meyer 2010).14
Sounding the alarm, the United Nations declared that all member states should adopt 30% minimum targets for women in all political bodies by 1995.15 A who’s who of leading international institutions has issued similar proclamations. The European Union, the Organization of American States, and the African Union have all declared that their continent should adopt specific compositions of women in their official decision making (Krook 2009).
The movement toward women’s descriptive representation—that is, increasing women’s physical presence in decision making (Pitkin 1967)—is not mere rhetoric. Of the fifty-nine countries holding elections in 2011, seventeen mandated quotas in their legislatures.16 As of that year, fifty-two countries had introduced legal gender quotas of some kind in elections, and in approximately forty more, at least one political party uses voluntary gender quotas to choose its candidates (Dahlerup 2012, vii; see also Krook 2008, table 2; Krook 2009). Most countries have implemented a political quota for women of some kind (Pande and Ford 2011, 8). Similar efforts are being implemented in quasi-public or private domains, such as laws requiring minimal female representation on corporate boards, enacted in Spain, Iceland, the Netherlands, Norway, and France (10).17
Unlike some reforms adopted on a world scale, this one has not entirely bypassed the United States. Although the United States has fewer and weaker requirements than many other countries, public and quasi-public efforts have been made to increase women’s presence. The state of Iowa, for example, has passed several rounds of legal mandates for “gender balance” in response to the low numbers of women on its appointed boards (Hannagan and Larimer 2011a). Federal law requires that boards be representative of “social identities” (Mansbridge 1999).18 Both the Democratic and Republican parties have used gender in choosing party representatives for committees or national convention delegates (Baer 2003, 127–31).19
As public officials made a concerted effort to bring women into their representative bodies, another equally ambitious democratic reform effort was taking off around the globe. Along the length of the Americas, in island nations off the coast of Africa, in the heart of Europe, in remote villages throughout India, and even in communist China, ordinary people are finding themselves invited to participate in official attempts to bring power to the people. The setting in each case would look strikingly familiar to Emerson and his compatriots in their New England town meetings. What these groups are doing, each in their own way, is deliberating about the pressing issues of the day. Citizens have gathered in locations as disparate as Porto Alegre, Brazil; British Columbia, Canada; Rajasthan, India; São Tomé and Príncipe; Benin; and the state of Texas. Anywhere from a handful of individuals all the way up to 6% of the adult population of the country has participated in these forums. People are not merely spewing hot air at these meetings; they are actually governing. The citizens of British Columbia used a face-to-face assembly to recommend a new electoral system to their fellow citizens. Indian village meetings are constitutionally empowered to decide many of the issues that directly affect quality of life in the most basic way. And the list goes on.20
The best evidence we have for the broad scope of deliberative settings comes from the United States. As many as 97% of American cities hold public meetings, and most large cities rely on active neighborhood councils (Macedo et al. 2005, 66; Karpowitz 2006). Over one thousand towns conduct formal New England–style town meetings in the United States (Bryan 2004, 3). Some states, such as Oregon, use citizen deliberation groups to plan state budgets; environmental protection assessments routinely include citizen meetings; and important policy domains, such as medical research priorities, are informed by citizen deliberation (Gastil 2000; Ryfe 2004; Simonsen and Robbins 2000). Well over three million US citizens serve on deliberating juries each year (Gastil et al. 2010, 4), and approximately a third will have served on a jury at some point in their lives (Devine et al. 2001, 622; Gastil et al. 2010, 4, note 3). Over fifty million people belong to one of over one-quarter million homeowners associations (Macedo et al. 2005, 102).
But while deliberative democracy is being cultivated in the grass roots, women’s representation in these settings has been wallowing in the backwater. Few have applied the question of women’s equality from the arena of legislatures to these forums for the people.21 Seldom do we ask how to design these deliberations so that half the population can find equal voice.22
And when scholars ask this question, they often cannot produce a clear answer. In studies of participation in Indian village meetings, a team of researchers has concluded, “it is clear that Gram Sabhas (village meetings) are not a forum for women in their current form” (Besley et al. 2005a, 656).23 The usual variables that enhance the attendance rates of disadvantaged groups including the landless, illiterate, and Scheduled Castes and Tribes—such as village literacy—actually depress women’s participation significantly (Besley et al. 2005a, 653, table 2). Similarly, Pamela Conover, Donald Searing, and Ivor Crewe included controls on a host of possible variables that could explain why British and American women report engaging in fewer political discussions than men, but the gender gap persisted nonetheless (2002, table 7).24 Not only do we seldom study women’s representation in deliberative democracy, when we do, we are left scratching our heads. Frank Bryan, one of the few scholars to rigorously study this question in the United States, has recently declared, “for the life of me and after thirty years of research, I remain stumped when it comes to predicting women’s involvement” in public meetings (Bryan 2004, 249).25
The best hopes for women’s equal representation have come from “critical mass” theory. Elegantly formulated by Rosabeth Moss Kanter, the theory predicts that where women compose less than 15%, men’s culture dominates. Women function there as mere tokens. They have little influence or agency and must conform to masculine ways. But where women’s percentage climbs well beyond 15% and reaches up to 40%, women can begin to make a difference. Indeed, the United Nations justified its 30% female target on the premise of this notion. Its formal language states, “The figure of 30 percent forms the so-called ‘critical mass,’ believed to be necessary for women to make a visible impact on the style and content of political decision-making.”26 And should women reach a balance with men, their experience and power improves further still. Or so the theory goes.
The problem is that women’s percentage in deliberating bodies matters little—in itself. At issue is that descriptive representation may not translate into substantive representation—the actual expression and implementation of women’s preferences. Nor will it unfailingly translate into symbolic representation—the sense that women can and should exercise power. There is puzzlingly mixed support for the prediction that increasing numbers yield increasing rates of participation or influence. In fact, in some settings, such as Vermont’s town meetings, the higher women’s percentage in the deliberating body, the lower is their share of the speakers (Bryan 2004, 222). Even reaching the promised land of gender balance seems woefully insufficient. Princeton’s women made up 53% of those present at the town hall meeting with the governor, but 25% of the speakers, a pattern that mirrors Bryan’s results from Vermont town meetings (cited earlier in this chapter; Bryan 2004, 216–17). A study of New York City’s online discussions among city residents, held to consider rebuilding the World Trade Center, found similar results. While women’s proportion of participants mirrored the city’s population (around 50%), it dropped by about 10 points among active contributors to discussion—the upper quartile of participants, who posted about 80% of the posts (Trenel 2009).27 That is, even in today’s United States, where women are better educated than men and even outpace men in voter turnout, meetings are characterized by a significant gap between women’s physical presence in deliberation and their participation in it, especially their active participation.
The weak or inconsistent effect of critical mass in citizen forums is not limited to the northeastern United States. The pattern repeats in midwestern college towns, in Israeli kibbutzim, and in Indian villages.28 For example, while some studies find that village leadership quotas for Indian women “benefit their villages while providing the public goods preferred by women,” others find that increasing women’s descriptive representation matters little (Ban and Rao 2008b; 2009, 6).29
Numbers alone do not determine the active use of voice, and this is a familiar refrain in recent legislative studies of women. Specifically, larger numbers of women in the legislature do not consistently increase women’s influence, or the substantive representation of women’s distinctive priorities and perspectives (for example, Carroll 2001; Franceschet, Krook, and Piscopo 2012; Reingold 2000; 2008, 132, 140).30 The reasons include the overriding influence of party membership (and the allocation of power according to the party’s position in the legislature), women’s lack of seniority, the imperative to represent constituency interests rather than women’s distinctive concerns, women’s heavy dependence on male elites and patrons who are the gatekeepers to elected positions or resources, and male legislators’ backlash against women’s legislative gains.31
For example, even after female members of the British Labour Party crossed well over the 15% threshold of elected members of the House of Commons, they still perceived hostility from male colleagues when they attempted to use a more feminine style, reporting pressure “to conform to the traditional norms of the House” (Childs 2004a). Or in New Zealand, though women’s numbers increased to around 30%, their feminizing effect on decision making and policy remained negligible (Grey 2002). Even as women near an even split, they do not always advance. In Rwanda, women have composed nearly half the legislators, yet their balanced number made no apparent difference to policy outcomes (Devlin and Elgie 2008). That is, despite rising numbers, the institution often remains steeped in a masculine culture, and women fail to achieve the voice and authority predicted by theories of critical mass (Lovenduski 2005).
Not only do numbers often fail to advance women’s representation, but also studies in various settings document a backlash as women’s numbers increase. For example, in US state legislatures, as the proportion of women increased, male committee chairs became more verbally dominant and less inclusive of women (Kathlene 1994; Rosenthal 1998, 90). As Kathlene concluded: “women legislators, despite their numerical and positional gains, may be seriously disadvantaged in committee hearings and unable to participate equally” (1994, 572). Validating this backlash notion with an entirely different type of action—US House members’ campaign contributions to fellow members—Kanthak and Krause found that as the proportion of women increases in the member’s party, men decrease their contributions to women while increasing them to men (Kanthak and Krause 2010). These discouraging patterns obtain in a variety of countries. In New Zealand, for example, the representation of women’s issues was set back as women increased from approximately 15% to 30% (Grey 2006).
Taking stock of the research on women’s numbers, Beckwith and Cowell-Meyers argue that “critical mass theory is both problematic and under-theorized,” its mechanism “unspecified,” and the power of small numbers of women “neglected” (2007, 553). Dahlerup finds that “the number of women in parliaments is probably not the most crucial factor” in women’s ability to exercise political power (2006, 520).32 Htun and Weldon find that feminist movements and organizations in civil society affect social policy much more than “intra-legislative political phenomena such as … women in government” (2012, abstract). And as Franceschet, Krook, and Piscopo sum it up, raising women’s numbers “may have positive, mixed, and sometimes even perverse effects on women’s political representation” (2012, 13). That is, increasing numbers do not always help, and sometimes backfire.
Not only do women fail to reach consistent gains by exceeding the 15%, or even the 40% threshold, they may accomplish more when they are few. In some state legislatures, female “tokens” are more likely to sponsor and enact women’s issue bills than men in that legislature, and may be as successful as the more numerous women in more gender-balanced legislatures (Bratton 2005). A critical mass may do little that a few exceptional “token” women accomplish as well or even better, according to some cross-national studies (Childs and Krook 2006; Crowley 2004; Kittilson 2008, 324). Such findings lead some to suggest that scholars “give up” on the notion that critical mass matters and instead study “critical actors” (Childs and Krook 2006). In sum, small numbers may not be fatal, just as large numbers may not be sufficient.
When do numbers change the masculine environment, give women a more equal effective voice, and lead to meaningful increases in women’s substantive representation? Why do small numbers of women sometimes succeed?
The common denominator from many of these studies is the notion that the features of political institutions and contexts can determine whether, and how, women’s numbers matter. For example, larger numbers help when women use institutional mechanisms to work collaboratively (Duerst-Lahti 2002a, 380; Thomas 2005, 253–54; Thomas and Welch 2001). Conversely, when the institution’s incentives or structures tie women’s fortunes to male party elites and make a women’s coalition unlikely or weak, women have trouble turning their larger numbers into change (Franceschet, Krook, and Piscopo 2012). Numbers also help when they are accompanied by explicit efforts to empower women, such as through the advance of feminist movements and organizations, or when the discourse environment is receptive to women’s voices (Beckwith and Cowell-Meyers 2007, 557; Studlar and McAllister 2002; Grey 2006). That is, the effects of numbers depend on institutional features. Not every study explicitly draws this conclusion, but these studies’ evidence is consistent with it.
In this book we show why numbers do not matter on their own in formal discussions and why the combined effects of numbers and institutions help us to solve the puzzle of women’s representation in deliberating bodies. Large numbers need particular institutional norms and rules in order to work. Moreover, we explain when and why small numbers can work quite well. The key is that the factors of numbers and institutional norms or rules jointly shape the status of deliberators within the group. Deliberation is a site of power and authority. It is one way that people with power exercise that power, and one place where people without authority can gain it. The numbers and the rules shape women’s position in deliberation, and the very nature of deliberation itself, such that women’s authority falls, or rises, accordingly. By unpacking the black box of deliberation, we can come to understand women’s participation, influence, and authority.
We began with a puzzle. Women are more likely to be absent than present in the settings where public decisions are discussed and made. Moreover, even when they sit at the table in increasing numbers, they often participate and influence at low rates. Despite considerable study, the key to the puzzle has remained elusive. The problem has gained urgency as the effort to democratize has cultivated more opportunities for citizen deliberation, supplementing discussion in official bodies. The main remedy pursued so far is to raise the number of women to a critical mass. But that remedy seems to fail on its own. We show how, why, and what to do about it.
WHY DOES WOMEN’S LOW VOICE AND AUTHORITY MATTER?
Women’s underrepresentation matters for a number of reasons. These reasons are grounded in one big starting point: women and men tend to differ in important respects. To be sure, the large variance within gender tends to exceed the typically small differences between the two sexes (Sapiro 2003). Yet despite the considerable diversity of women’s views, a host of scholarly findings point to persistent average gender differences in some priorities and perspectives. Regardless of the sources of these differences, they are bolstered by the fact that society continues to be structured in part by gendered social roles, occupations, and divisions of labor. People tend to internalize the expectations that correspond to their assigned gender. Women and men are socialized from birth into respective feminine and masculine ways of acting, thinking, and feeling. Women’s formal and implicit roles, their values, and even their core motives and personality are thus shaped in the mold of what society defines as feminine.33
Feminine habits of mind tend to be defined as cooperative and caring for others, so it is no surprise that women on average exhibit a stronger tendency in that direction. Women tend to have more empathy than men (Baron-Cohen 2003).34 They are more likely than men to cooperate when there are no strategic incentives to do so (Eckel and Grossman 1998) and generally display somewhat more communal orientations (Huddy, Cassese, and Lizotte 2008, 33). Women are much more reluctant to endorse force and coercion than are men (Conover and Sapiro 1993). Women tend to be more religious and more oriented to community standards and morals (Burns, Schlozman, and Verba 2001; Clark and Clark 1993; Eagly et al. 2004; Huddy, Cassese, and Lizotte 2008, 44) and more responsive to social sanctions (Barr 2004).
Consequently, while women and men share much in common, and while women disagree with one another on many issues, women nevertheless tend to have some political priorities that differ from men’s. Women, more than men, tend to favor, and to prioritize, the needs of those who are vulnerable, disadvantaged, poor, exploited, or stigmatized (Andersen 1996; Gilens 1988; Huddy, Cassese, and Lizotte 2008; Shapiro and Mahajan 1986; Sears and Huddy 1990).35 These tendencies hold true for elites, such as officials in legislatures, as well as for ordinary citizens (Childs 2004b; Miguel 2012; Poggione 2004; Swers 1998; Thomas 1991, 1994).
Such preferences represent an important set of political viewpoints that are valuable in their own right, and if women do not participate fully, their distinctive concerns and perspectives may not find full representation. When women’s voices are absent, those distinctive considerations are likely to be lost, to the detriment of the group.
By extension, greater participation from women could affect the decisions and outcomes that emerge from institutions. Policy can look very different when made by women—the needs of vulnerable populations tend to be placed front and center, and the poor become less destitute. When women are empowered under the right conditions, they are more likely than men to speak for, sponsor and advocate, or vote for policies that assist women, children, families, stigmatized or disadvantaged minorities, and the vulnerable in society, for more generous policies on a number of social service measures, or for policies that serve the common good (Beaman et al. 2009; Bolzendahl 2011; Childs 2004b; Duflo and Topalova 2004; MacDonald and O’Brien 2011; Miguel 2012; Poggione 2004; Swers 1998; 2002; Thomas 1991, 1994; Walsh 2002). To be sure, exceptions exist. Margaret Thatcher, for example, was well known for her masculine style and policy goals. She and a few others like her remind us that the effect of individual gender is far from deterministic or inconstant. However, on average, women do tend to exhibit a central tendency that differs from men’s priorities.36 Gender matters because if women are not adequately represented, their distinctive preferences cannot enrich the mix of views and shape collective decisions.
Full participation by women can affect not only what decisions are made but also how they are made. Put differently, if institutions were to adequately include women, the very nature of the institution could change (Carroll 2001; Phillips 1995). Some feminist theorists argue that “women make representative institutions more democratic” (Dovi 2008, 157) or deliberative (Mansbridge 1994; Norris 1996). As Norris put it, women “introduce a kinder, gentler politics,” one that is “characterised by co-operation rather than conflict, collaboration rather than hierarchy” (1996, 91). Women are said to be “more willing to listen to the other side” and “less adversarial” (Bochel and Briggs 2000, 66). They even “make less noise in the [legislative] chamber” (G. Young 2001, 8; quoted in Childs 2004a, 3); as one study summed it up, “heckles are primarily a men’s affair” (Grünenfelder and Bächtiger 2007, 16).37 Raising the authority of women can thus alter a decision-making institution to make it more democratic (in the sense of more inclusive) and also more deliberative—more focused on listening, cooperating, and collaborating (Grünenfelder and Bächtiger 2007).
This combination of more inclusive and more deliberative interaction can create a feedback loop for women’s representation, further increasing the authority of the women who are present. Norms of interaction that emphasize warmth and cooperation can encourage women to participate actively and to exercise influence. Some studies have suggested that predominantly male settings tend to produce a masculine style of interaction that negatively affects women, even among political elites (Kathlene 1994; Mattei 1998). The masculine forms of interaction in these settings can send a strong negative signal to women about their expected role in the discussion and thus may lower women’s participation in it. If interaction becomes more feminine—that is, more deliberative and democratic—then women’s authority can rise. One way to transform interaction is to increase women’s descriptive representation in combination with the appropriate decision-making rule. Thus women’s low numbers matter in part because they allow a less deliberative and democratic interaction style, which in turn depresses women’s authority. Elevating the numbers can change that interaction style, increasing the probability that women will not be mere bystanders, but will instead become active, authoritative voices in the group.
In addition, women’s low representation matters for democratic legitimacy and public engagement. Increasing women’s participation in political institutions affects how observers—especially women—think and feel about those institutions and their place in them. For example, women’s descriptive representation can elevate the legitimacy of the political institution that purports to represent them and can strengthen some citizens’ democratic attitudes or civic orientations (Mansbridge 1999). Men and women view government as more legitimate and trustworthy where women are more numerous in it (Schwindt-Bayer and Mishler 2005). Women who are represented by women tend to view their members of Congress in a more positive light (Lawless 2004). Where women are visible on the ballot, or when the number of women in office increases, women tend to have higher levels of political efficacy, political knowledge, and engagement in a variety of political jurisdictions and countries.38 When Americans were asked in 2000 whether they think the country would be governed better or worse if more women held public office, most said “governed better,” and only a small minority said “governed worse” (Simmons 2002, cited in Dolan and Sanbonmatsu 2009).39 Thus increasing the number of women in office may increase the level of political engagement by women and the overall legitimacy of government in the eyes of both women and men.
A rise in the number of women in government can also raise the status of women in the polity more generally. For example, as women occupy more leadership roles in Indian villages, voters’ negative stereotypes of female as inadequate to lead decrease, and their willingness to vote for women rises (Beaman, Pande, and Cirone 2012). A study of twenty-five countries found that women’s descriptive representation increases women’s belief in “women’s ability to govern” even after controlling for other competing explanations (Alexander 2012, 460). Women’s authority in the eyes of others may rise as they prove by their cumulative performance that they can in fact govern competently.
More broadly, women’s participation in public settings carries important symbolic meaning. Participation in the life of one’s community and its public affairs signifies not only specific power over particular decisions but also a more general authority in public life, and authority in turn shapes the most fundamental sense of human worth. That sense goes by various names—honor, respect, standing, esteem, and so on. We see these as all referencing the notion of worth embodied most famously in the Declaration of Independence, which, modified, continues to resound today: all people are created equal. Women will not be viewed as equal in that basic sense of worth until they carry equal authority in the public discourse that governs their community. To participate fully is to take one’s place in the ranks of those who govern, and in a democracy, where every member of the community governs, one must govern to be a member of the community. Mansbridge justifies the use of descriptive representation in part precisely on these grounds; when a social group has been excluded from governance, and its ability to govern questioned, then measures should be taken to include its members in deliberation so as to establish its “ability to rule” (1999, 628).
The notion that full participation in deliberative institutions constitutes one’s standing in the community applies not only to the formal forums that directly govern, such as juries, formal community boards and councils, or committees selected by the state and invested with governing power; it also applies to the informal associations that make up civil society. Theorist Mark Warren writes that civil society is not divorced from the operation of government and politics (2000, 32–33). When citizens gather for private purposes their actions resonate outward. The life of neighborhood associations, parents’ organizations, and even entertainment clubs shapes the orientation that the members adopt toward the role of government and their own efficacy to affect their political system (Eliasoph 1998). Civic associations can powerfully determine citizens’ ability to affect their government, for example, by the way that they distribute and organize their own power internally (Skocpol, Ganz, and Munson 2000). The power relations of informal civic life—including the norms and patterns of discursive interaction in these settings—shape authority in the more formal institutions of politics and government. Thus women’s standing in civic associations also bears on their standing in the polity.
Fundamentally, gender inequality is worth studying because gender continues to be a dimension of political underrepresentation.40 As Burns, Schlozman, and Verba (2001) note, politics is still a “man’s game.” To be sure, women now enjoy equal formal legal rights in the United States and many countries. They have increased their income, occupational status, workforce participation, and most of all, education. Despite huge progress, individuals are still heavily sorted into gendered roles based on their biological features, and this sorting has profound implications for how political power is understood and used. As the psychologist Susan Fiske argues, with these roles come a bifurcation of social esteem into two opposing dimensions of regard; women’s roles confer upon them love, or appreciation, or liking, or attraction—but not authority. Women may find “feminine” ways to get what they want, by leveraging their distinctive gender roles, yet still lack authority.
The problem for women is that while they can achieve some of their aims by being liked for their sociability, loved for their nurturing, appreciated for their cooperation and teamwork, or desired for their sexuality, they may still lack in basic respect and authority in public life and group decision making.41 Women are expected to talk, and carry influence, on topics directly dealing with women’s domains of expertise. But not so in other domains. It is no surprise, then, that gender gaps reveal themselves in many forms of political participation and representation. Even when women gain ground in society, their representation does not necessarily follow, because society continues to revolve around highly gendered notions of where and how women should exercise overt influence and who deserves authority.
Given that participation in public discussions has far-reaching implications, women’s status in these discussions is of the utmost importance. Often, women are not perceived, or perceive themselves, as influential or valued to the same extent as men during public discussions. This low status interferes with women’s representation in a myriad of significant ways.
DELIBERATIVE THEORY AND THE QUESTION OF SOCIAL EQUALITY
Women’s low authority is not only a problem for the ideal of democratic inclusion. It is also an Achilles heel for advocates of deliberation.
Political philosophy has experienced an explosion of interest in deliberative democracy. Many normative theorists have advocated more citizen deliberation.42 As Gutmann and Thompson write, “no subject has been more discussed in political theory in the last two decades than deliberative democracy” (2004, vii; see also Dryzek 2007, cited in Thompson 2008).
The most influential theorist and advocate of the ideal of deliberative democracy is the eminent philosopher Jürgen Habermas (1996). Habermas’s core concept is reason. For him the ideal speech situation is one that establishes reason through “intersubjectivity,” the exchange and development of logic and evidence among deliberators. Justifying one’s position through reasons that anyone can accept is the basis of deliberation. As Chambers puts it, in Habermasian fashion, deliberation “ought to spark active reasoning and thoughtfulness” (2009, 335). Deliberation can encourage people to question, to think more deeply about, and to articulate their fundamentals—why they believe what they do and wish what they want. In deliberation people are prompted to examine their presuppositions and their assumptions, their values and beliefs, and revisit them, reconstruct them, and come to a better understanding of what to do with them. The discourse of “practical reason” carries the Enlightenment project forward and forms the foundation of democratic practice. Habermas’s theory is thus grounded in the notion that deliberation occurs through justification and rational discourse (1996; Chambers 1996).
Inspired by Habermas, a number of normative theorists have listed required preconditions for deliberation. One of the key dimensions of desirable deliberation as articulated by many theorists is social equality in participation and influence. A minimum requirement is equal access to the floor, but that is only the beginning. Knight and Johnson offer the standard of equal opportunity to influence the discussion, not merely the formal opportunity to deliberate (1997). For most theorists, equality does not require that each and every person should speak the same amount and carry the same influence. But following Habermas, many theorists posit that each person should speak and influence the same amount unless their arguments differ in quality. And this implies that factors other than quality should not play a role, whether they are based on concrete resources or more intangible status.
Gutmann and Thompson state, for example, that deliberation rightly constructed will “diminish the discriminatory effects of class, race, and gender inequalities” (2004, 50). Not only do theorists wish to see the equal opportunity to deliberate, they also posit an ideal where status or power inequalities in society do not affect the deliberation itself. In the article that in many ways sparked the deliberative turn in political theory, Joshua Cohen writes that in ideal deliberation the “existing distribution of power and resources does not shape [deliberators’] chances to contribute to deliberation, nor does that distribution play an authoritative role in their deliberation” (1996, 74; emphasis added). Dennis Thompson further specifies that “the discussion is better deliberation to the extent that the participation is equally distributed,” and “most [theorists] agree that the more the deliberation is influenced by unequal economic resources and social status, the more deficient it is” (2008, 501 and 506; emphasis added). Thus several influential theorists of deliberation offer a standard of social equality of actual participation and influence, and not merely equal opportunity to talk.
However, a chief concern about implementing deliberation is that it will fall far short of equality and perhaps even exacerbate social inequalities. Some critics have argued that deliberation as actually practiced tends to reflect or even magnify existing disadvantages of status or power (Sanders 1997; Young 2000). In response, the later writings of advocates acknowledge that “unequal resources are likely to produce unequal participation in the deliberative forum” (Thompson 2008, 509). Still, as Thompson puts it, “we do not … know whether it is true, as some theorists plausibly argue, that under many conditions deliberation is less affected by prevailing inequalities than power-based modes of decision making” (2008, 509). So while advocates recognize that social inequalities are a barrier in actual deliberation, they do not explore the most serious ways that inequalities might undermine participation and influence, and some argue that actual deliberation is still more egalitarian than actual alternative modes of politics.43 While the advocates make a strong case for the need to have gender-egalitarian deliberation, they have not yet provided an analysis of how or why patterns of inequality in society interfere with egalitarian deliberation, or what the appropriate solutions might be.
The critics have provided such an analysis, drawing attention to the specific mechanisms of voice and authority (Fraser 1992; Sanders 1997; Williams 2000; Young 1996; 2001). These critics worry that groups with less status or authority in society may speak less, be heard less, and, relatedly, carry less authority in deliberation than members of groups with more authority in society (Mansbridge 1983). Moreover, they argue that the very nature of language, through its rhetorical elements, either enhances inequality during deliberation or can remedy it. They claim that marginalized groups are more likely to express emotion, to offer personal testimony, and to structure their contributions as narratives rather than as logical “if-then” statements; consequently, their contributions may not seem to be based on the right kind of evidence, to be objective and universally relevant, and otherwise reasonable.
So it is not simply a problem of women’s contributions being valued less because women speak them. Rather, the types of considerations women tend to articulate, and how they articulate them, are valued less because they reflect ways of thinking and self-expression that have been socially constructed as less authoritative (Sanders 1997; Williams 2000; Young 1996, 2000). As the philosopher Fricker (2007) argues, not only might women be devalued for who they are, but they also may be devalued for how they think, feel, and express themselves (see also Allen 2012, 22). In other words, because their social identity as women carries less authority, and because women’s psychological modes and social styles tend to carry less authority, women as a whole will not be able to participate or be heard in the way that advocates of deliberation would wish. So while Habermas values the transformative potential of deliberation, deliberation will fail to achieve that transformation, because not all reasons, forms of expression, and modes of interaction are socially constructed to be equally authoritative. Thus the normative requirement of equal participation and influence can only be achieved when the devalued communication styles of marginalized groups are validated. From our perspective, this critique is extremely important because it argues that deliberation must enable groups such as women, who enter the deliberation with less authority, to build their authority during deliberation.
To their credit, many theorists have incorporated this critique of deliberation. They have included the types of rhetorical styles said to characterize the cultures of marginalized groups as part of their definition of what counts as quality discourse (Chambers 2003). For example, Dryzek expands deliberation to allow for rhetorical forms such as gossip and storytelling (2005, 224; so do Mansbridge et al. 2010). As Chambers summarizes the current state of the theoretical landscape: “Gone, with only a few exceptions … is the narrow, highly rationalistic view of reason-giving that stresses a model of impartiality rising above all difference” (2003, 321).
However, even with these more capacious understandings of deliberation, the deliberative vision remains incomplete and inadequate. What is still missing is a full reckoning with the problem of how authority gaps can be overcome. Both the advocates and the critics have yet to explain how to build equal authority within deliberative settings. A group that comes to the deliberation with lower stores of authority will not elevate its authority in the eyes of high-authority deliberators by using its distinctive styles of communication, because those styles signal low authority. For example, why would a low-authority member who tells a story be given a hearing by high-authority members inclined to dismiss this form of communication as nonauthoritative? Women are viewed as less authoritative when they express emotion, so why would their use of emotion remedy their lack of authority? Testimony and emotional expression may create a sense of authentic self-expression for disadvantaged groups, but if such expressions are not also a means to build real authority among discussion partners, then the goal of deliberative equality is not likely to be realized.
For that reason, encouraging devalued forms of deliberation may fail to close the authority gap, and may even reinforce it. That is, deliberation may not merely fail to live up to its promise; unless it guarantees equal authority within itself, it reinforces social inequality and risks becoming an actively negative force. As Honneth puts it, “recognition must precede cognition” (2005, 119). When it does not, the attempt to generate an exchange of reason, and perhaps even an exchange of emotions and personal experience, may become antiegalitarian. This is the point we seek to highlight in this book. By accepting that deliberation rests on authority, that unequal authority infiltrates deliberation itself, and that deliberation can thus reproduce and enhance unequal authority (at least under some conditions), advocates can then move to exploring ways to remedy this problem.44
In addition to authority, we also highlight the importance of another dimension: social and emotional connection. Not only does social inequality come from, and in turn affects, authority, it also shapes the nature of social interaction during discussion. Women are socialized to be more interdependent than men. While there is much variance within genders, there are also cultures of gender, as we mentioned and will elaborate in chapter 3. Consequently, women may not feel able or motivated to articulate their perspectives when the form of interaction lacks an interdependent character. This dimension of discussion may affect men’s contribution on the whole less negatively. In other words, women may be less likely to feel comfortable and competent than men do in competitive discussions that lack active signals of emotional warmth and connection between deliberators. Deliberative equality thus requires attention to the style of discussion, and specifically to social warmth.
The position Anderson and Honneth articulate takes these arguments into account and provides a useful starting point. They put it as follows: “the conditions for autonomously leading one’s own life turn out to be dependent on the establishment of relationships of mutual recognition. Prominent among these relationships are (1) legally institutionalized relations of universal respect for the autonomy and dignity of persons (central to self-respect); (2) close relations of love and friendship (central to self-trust); and (3) networks of solidarity and shared values within which the particular worth of members of a community can be acknowledged (central to self-esteem)” (2005, 131).
Deliberation can offer the interweaving of Anderson and Honneth’s three strands. Specifically, by interacting under the right circumstances, people can: (1) be subject to institutional conditions that encourage universal respect for each person’s autonomy and dignity, (2) create relations of warm regard, and (3) establish solidarity and hence affirm the worth of each member of the group. In our view, the central concept of deliberation, and what it can offer the cause of equality, is “empathetic engagement” (to use Honneth’s phrase [2005, 113]). But it is not enough for listeners to adopt a private mentality of empathy; they should convey to their conversation partners that they are experiencing empathy for the speaker. The speaker needs to sense that the listeners are empathizing with her. That is, empathy is not only necessary as a quality of the listener but also as a quality of the communication—the listener must signal it effectively enough for the speaker to perceive. Through this process, deliberation must first establish a socially connected and socially valued person before it can lead that person to engage in the pursuit of rationality. When deliberative settings carry out this social function, they not only enhance their mission of pursuing reason, they also enhance the equality of women with men. Empathy is not only the precursor to reason, it is also the precursor to equality.
We can now spell out the characteristics of settings that will produce gender equality. Settings that promote interactions high on social affiliation and “warm regard” are beneficial to women. These interactions should create for women the sense of “solidarity” and “friendship” Anderson and Honneth identified as important. In addition, these settings should create authority for women, the dimension on which women are particularly disadvantaged. Settings that do so would meet the requirement of “universal respect for the autonomy and dignity” of each member. In sum, we now have the guideposts that will chart our path here—women need settings that foster rapport and respect, because these affirm their sense of belonging and their competence, and these are prerequisites for women’s authoritative participation and representation—in short, for their influence.
In sum, philosophers justify deliberation on various laudable grounds, but our argument privileges one of these justifications in particular. We view deliberation as a practice that cultivates citizenly virtues, that raises the knowledge, reasoning, and general cognitive competence of the typical disengaged, low-capacity voter. Nor is its most important justification its ability to cast democratic government in the glow of legitimacy. More important, egalitarian deliberation can establish and maintain the equal authority of each person as a citizen. The deliberative setting is not only a school for democracy that informs the masses, pacifies intolerance, and instills in the free rider a sense of duty to the community. Deliberative meetings are also a social institution that delineates the individual’s standing as a respected member of that community.
And it is through that standing that the other beneficial effects of deliberation can come about for disadvantaged groups. Yes, deliberation can raise the citizen’s motivation to engage in the life of the community, so crucial to the person’s level of knowledge and interest in public affairs. But that motivation is rooted deeper, in the person’s sense that she is the kind of person who is entitled to do so, expected to do so, capable of doing so, valued for doing so, and, therefore, does so. So deliberation can engage the unengaged, and thereby school them in the skills and attitudes of democracy, when it recasts who is entitled to carry authority in deliberation. By giving women authority, deliberation can signal to women that they are worthy of the full rights of citizenship and raise their competence in the process. When that occurs, deliberation can shape in fundamental ways the life, the dynamics, and ultimately the decisions of the polity.
THE ARGUMENT OF THIS BOOK
The key questions we aim to answer should now be clear: under what conditions will women achieve an effective voice in deliberating bodies?45 Does equalizing the number of women help to equalize women’s participation and representation? What can institutions do to facilitate women’s authority? What needs to happen in order for the distinctive priorities of women—priorities such as the inclination to help others—to be heard and to influence fellow deliberators?
Our answer is: higher numbers of women do aid women’s representation, but it all depends. It depends on the decision rule and the norms of discussion that such rules generate. The reason that having more women helps women is the same reason that decision rules matter: discussion is an act of communication, and therefore it is governed by implicit but powerful norms. These norms are rooted in gendered cultures, in conventions of interaction among women, among men, and between men and women, and they are rooted in unspoken rules of social inclusion or competition. Unanimous rule signals the need to include everyone. Majority rule produces a more competitive dynamic that aids each gender when it is the majority in the group. While we know a great deal about how these rules affect the inclusion of a generic individual’s preferences and ideas, we know very little about how they affect gendered interaction and the inclusion of women’s preferences and ideas. Our study explains why these decision rules affect men and women as men and women.
We set the stage for our analysis in chapter 2 by detailing why women continue today to be the “silent sex” in political participation broadly, and in authoritative discursive acts specifically, despite substantial gains in education, concrete resources, and formal rights. We derive expectations for women’s participation and influence in group discussion in chapter 3.
As we explain in chapter 4, we use an experimental design because we wish to avoid the unavoidable pitfall of observational studies—the effect of greater numbers of women, or of institutional arrangements, is shaped by other factors, making it impossible to know if what appear to be the effects of numbers or institutional rules are in fact due to something that causes these or that is correlated with them. We vary two features of deliberation: the group’s gender composition and the group’s decision rule. We create a norm of inclusive interaction by assigning groups to deliberate with unanimous rule, and we contrast these groups with others assigned to deliberate with majority rule. We replicate our results in two very different sites, one a religious and socially conservative city in the Mountain West (Provo, Utah), and the other a secular, wealthy, predominantly liberal town in the eastern seaboard (Princeton, New Jersey). The basic patterns of gendered interaction hold regardless of location.
In chapter 5 we show that numbers and rules jointly elevate or depress women’s participation in discussion, and that the more women speak, the more influential they become. Conditions that prompt women to speak thus equalize their influence. The reason that speech is a civic act that reflects and creates one’s standing in the community become clear in this chapter—speech affects the sense of personal efficacy and others’ views of the person’s authority. In chapter 6, we explore the reasons for this gender gap in speech participation, testing various statistical moderators to see what it is about a person’s gender that causes the inequalities. We show in chapter 7 that settings that equalize women’s voice also tilt the agenda toward talk of women’s distinctive concerns: children, family, schools, the poor and the needy. Deliberation matters because it sets the terms of the decision, and women’s lack of standing robs them of their ability to shape that agenda. In chapter 8, we show how the minute dynamics of conversational engagement deny equal status for women in the group. But we also demonstrate how these dynamics can foster a warm climate of social affiliation that helps women. In chapter 9, we find that the same conditions that foster equal participation, equal perceived influence, and an agenda devoted to women’s distinctive concerns also produce more compassionate policies for the poor, vulnerable, and disadvantaged. Women’s influence on collective generosity is measurable in precise dollar amounts, and settings that empower women generate significant increases in the standard of living of the poorest members of the society.
Our main findings are that majority rule is bad for women’s substantive, symbolic, and authoritative representation as long as women are the gender minority. Conversely, majority rule is a boon for women when they make up a large majority of the group. Unanimous rule aids women when women are the gender minority, as they benefit from the norms of inclusion that unanimous rule produces. However, unanimous rule provides no advantage to women when they are the gender majority, as it also empowers the male minority. Putting these findings together, we conclude, in a departure from prior studies, that the effects of one factor—gender composition—depend upon the other factor—rules and procedures. That is, the effect of gender and the level of gender inequality depend on the structure of the group setting, and that structure is shaped by who deliberates and under what rules.
What these findings tell us is that rules that are unbiased on their face actually produce deep social inequalities, but rules can also remedy the inequality people bring with them into deliberation. Who deliberates matters, because it creates norms of social interaction and ways of talking, but deliberators are not trapped in the constraints society imposes on them.
Despite drawing heavily on our controlled experiment, we can take some comfort that our results hold up in the real world. To beef up the confidence in our findings—to increase our external validity—we replicate the key results in a natural setting. There are over 14,000 school boards in the United States. Among these we find an enormous variety of gender compositions. Many boards have an even mix of men and women, but many are lopsided, including all-female and all-male groups, what we call gender “enclaves.” A sample of meetings from these boards allows us to test our lab findings in locations small and large, in rural, suburban, and urban settings, in places where gender roles have been slow to change and where women have advanced far. We present these findings in chapter 10.
While this book is about gender inequality, its framework can be used to study other forms of inequality. Some dynamics are probably special to the case of gender, as we explain in chapters 2 and 3. But other dynamics may not be. Unfortunately, we are unable to give a full treatment to the question of race, class, and other inequalities. In fact, school boards contain very small numbers of nonwhites, as do many other deliberative settings in the United States (Jacobs, Cook, and Delli Carpini 2009).46 Although at the end of chapter 10 we do provide some small hints about the gender dynamics in racially diverse groups, our experiment is confined to Anglo whites in the United States. We hope that future studies carry on the important work of investigating the interaction dynamics of race, class, and other forms of inequality and marginalization.
Even with this important caveat, our study can provide a set of tools that enables scholars to look rigorously at how key features of the group—whether its composition, its decision rule, or other structures—shape the dynamics, content, and outcomes of group deliberation. It also fills the gap in studies of deliberation by conducting a large-scale systematic analysis of the details of political discussion itself, providing the first combined study of the words people use, the preferences they express, and the pattern of interruptions they receive. These then become data that we link to other data on participants’ views on the matter at hand, their decisions, the level of influence they carry over the final outcome, and the perceptions of that influence.47 It is only by carefully unpacking the ways in which men and women interact with each other in the process of making collective decisions that we begin to understand how gendered dynamics affect the individual and the group. Our study can offer other scholars of deliberation a set of ways to investigate authority in groups, gender notwithstanding.
Our theme is how and why women become the “silent sex” and how increased voice and authority for women are possible. We proceed to set the stage for women’s participation in politics in the next chapter.
1 By a ratio of 4 to 1 they would rather see officials talking to people in town meetings than “talking to people at malls or on the street” or to people who “call, write, or e-mail” the official. The telephone poll was conducted by the Kaiser Family Foundation and Princeton Survey Research Associates from January to March 2001, with 1,206 randomly sampled adults, and reported in the Public Perspective July/August 2001.
2 Various estimates converge to show that the percentage of Americans who attend a public meeting at least once a year is between 25% and 38%. The numbers of attendance compared to other activities are from the 2000 National Election Study; 27% of respondents report attending a meeting in the past twelve months. The 1990 Citizen Participation study finds that 77% of women and 80% of men are members of an organization (political or not), and of these, 49% and 50% respectively report attending a meeting of the organization at least once in the past year, meaning that about 38% of Americans attend a meeting of their organization at least once a year (based on Burns, Schlozman, and Verba 2001, 76, table 3.2). American attendance in public meetings has declined (Karpowitz 2006), but about 75% of Americans report attending a public meeting at some point in their lives (Karpowitz 2006, chapter 2). Conover, Searing, and Crewe estimate a similar number (29%) for Americans and a slightly lower number for British citizens (Conover, Searing, and Crewe 2002, table 3). Jacobs, Cook, and Delli Carpini (2009) found that 25% of Americans engaged in face-to-face deliberation (“attending a formal or informal meeting in the past year to discuss a local, national, or international issue,” 37–40); 68% informally discussed issues (“informal face-to-face or phone conversations or exchanges with people you know about public issues that are local, national, or international concerns” at least “a few times a month,” 36); and 81% engaged in at least one form of discursive participation within the last year (38; see generally Jacobs, Cook, and Delli Carpini 2009, 42). When they are invited to participate in a face-to-face deliberation, 25% of Americans are willing to do so (Neblo et al. 2010). By contrast, the heralded age of online discussion has not yet materialized; the best and most recent estimate is that a mere 4% of Americans participate in online deliberation (Jacobs, Cook, and Delli Carpini 2009, 40). Finally, the median meeting size is forty (Jacobs, Cook, and Delli Carpini 2009, 60). In all, then, deliberation still occurs in groups small enough for face-to-face interaction, and with sufficient frequency to make its study worthwhile.
3 This phrase is an allusion to Tocqueville, cited below.
4 Serve on the board, or been an officer, in the past five years.
5 Percent affiliated with a religious institution: 74% and 58% of women and men respectively; served on board or as officer within past five years: 49% and 61% respectively.
6 On his observational tour of the upstart new democracy across the Atlantic, Tocqueville chronicled a political culture bustling with meetings, discussion, and other indicators of vibrant civic life: “The political activity that pervades the United States must be seen to be understood. No sooner do you set foot upon American ground than you are stunned by a kind of tumult … here the people of one quarter of a town are meeting to decide upon the building of a church; there the election of a representative is going on; a little farther, the delegates of a district are hastening to the town in order to consult upon some local improvements; in another place, the laborers of a village quit their plows to deliberate upon a project of a road or a public school … To take a hand in the regulation of society and to discuss it is the biggest concern and, so to speak, the only pleasure an American knows” (Tocqueville [1835] 2006, 249–50). Contrary to this portrait, the average contemporary American taking in the weekend game or carousing on a night out does seem to know pleasures aside from discussion. But Tocqueville was on to something. We take from him the insight that democracy rests on active meetings attended by all, and that part of the key to their success is their cultivation of social bonds among participants.
7 The governor took eight minutes to reply to the quieter half, and twenty minutes to the talkative half. A later town hall meeting with the governor in a different part of the state exhibited a similar pattern, e.g., male audience members took more turns, made more follow-up comments, and received more speaking turns back. These findings are from our direct observations.
8 We review studies of women’s underparticipation in chapters 2 and 3.
9 To be exact, 52% of men speak; 34% of women do.
10 Calculated from numbers on pages 216–17. “Women’s speech exceed men’s speech in 105 meetings” (Bryan 2004, 216), combined with the overall number of meetings reported under figure 9.1 (217). Women compose 36% of the speakers (Bryan 2004, 214). The sample was drawn randomly from Vermont towns, but some towns were repeatedly sampled. The large majority of the sample has between two hundred and five thousand residents. The data set has 1,435 meetings. As a point of comparison, Mansbridge found a similar pattern of gender inequality in her study of face-to-face deliberation (1983).
11 The figure of the adult population is from United States Census Bureau—International Data Base. http://www.census.gov/population/international/data/idb/region.php?N=%20Region%20Results%20&T=1&A=both&RT=0&Y=2012&R=1&C=. The 20% figure is from Crowder-Meyer 2010. The most up-to-date figure for leaders is 19.5%. http://www.un.org/apps/news/story.asp?NewsID=41445&Cr=un+women&Cr1. (See also Dahlerup 2012, vii). http://www.cnn.com/2012/06/28/world/europe/women-politics-global-power/index.html?iid=article_sidebar.
12 Inglehart and Norris’s comparative study finds that, at least during the mid-1990s, the gender gap in political representation was closing more slowly than the gender gap in other areas of society, including education, legal rights, and economic opportunities (2000).
13 http://www.ipu.org/wmn-e/classif.htm. The United States and Turkmenistan are tied at 78 in this ranking of percentage of women in the lower or single House.
14 In some cases the percentage is closer to 25%. Significant exceptions to the 20% figure are the Nordic countries, where women compose 42% of the national legislatures, and the Arab states, with 11%. http://www.cnn.com/2012/06/28/world/europe/women-politics-global-power/index.html?iid=article_sidebar.
15 The 30% figure was endorsed by the United Nations Economic and Social Council as a goal to be achieved by 1995 (Report of the Fourth World Conference on Women, Beijing, September 4–15, 1995, http://www.un.org/womenwatch/daw/beijing/platform/decision.htm, accessed June 14, 2013). The United Nations Fourth World Conference on Women, which met in Beijing, included a reference to this 30% goal in its report and issued the Beijing Declaration and Platform for Action, signed unanimously by all 189 member states. This conference approved a resolution stating “[Governments should] commit themselves to establishing the goal of gender balance in governmental bodies and committees, as well as in public administrative entities, and in the judiciary, including, inter alia, setting specific targets and implementing measures to substantially increase the number of women with a view to achieving equal representation of women and men, if necessary through positive action, in all governmental and public administration positions” (paragraph 190, part a).
16 http://www.un.org/apps/news/story.asp?NewsID=41445&Cr=un+women&Cr1.
17 Other countries have quotas for state-owned companies, including Israel, South Africa, Denmark, Finland, Ireland, and Switzerland (Pande and Ford 2011, 11).
18 The issue of gender representation on corporate boards has received attention recently; for example, Catalyst conducts regular counts of women’s numbers on Fortune 500 boards (in 2012, women composed 17% overall, and fewer than 20% of boards had at least 25% female directors. http://www.catalyst.org/knowledge/2012-catalyst-census-fortune-500-women-board-directors. The Securities and Exchange Commission required in 2010 that corporate boards disclose their formal efforts to achieve diversity. http://online.wsj.com/article_email/SB10001424052702303990604577368344256435440-lMyQjAxMTAyMDMwMDEzNDAyWj.html. Women’s presence on corporate boards does not automatically rise with women’s growing education or income. According to Pande and Ford, “increased education and labor participation rates among women are only weakly correlated with the number of women in leadership positions in the corporate sector. Despite the increase to over 50% women currently working in high-paying management and professional positions in the U.S., the percent of female CEOs in the Fortune 500 companies only increased from 0.2% in 1995 to 3% in 2009. While women constitute 15.2% of board directors in the U.S. and 12.2% in the UK, the percentage of women directors in the top companies (Fortune 500 and the FTSE 100) increased by less than 0.5% average per year over the last 10 to 15 years. The trends are not much different in developing countries” (2011, 5).
19 According to Krook, Lovenduski, and Squires: “Soon after women gained the right to vote in 1920, the Democratic party mandated that the Democratic National Committee (DNC) be composed of one man and one woman from each state and territory. The Republican party adopted a similar measure that same year, which they abandoned in 1952 but replaced in 1960 with a rule calling for 50–50 representation in all convention committees. … the DNC later ratified guidelines requiring state parties to select women as national convention delegates in proportion to their presence in the state population. When these reforms came under attack in 1972, the party rewrote delegate selection rules to ban ‘quotas’ in favour of ‘affirmative action.’ The Republicans, in contrast, chose not to regulate the state parties, although some states mandated 50–50 representation on their state central committees” (2009, 792–93).
20 See Ban and Rao 2009; Besley et al. 2005a; and Chattopadhyay and Duflo 2004. In 1993 the Indian constitution was amended to require the Indian states to grant full power over the allocation of expenditures to the most local level—village councils (Gram Panchayats), which decide on the provision of local services such as drinking water, irrigation, public buildings, roads, social services such as pensions for widows and the aged, and in some cases, informal education. It also required that one-third of village council seats, and of the council heads, be set aside for a woman. These are also set aside for Scheduled Castes and Scheduled Tribes, who are disadvantaged populations in India, in proportion to their district population. In every major state since then, at least 25% of council heads are female. The village council must hold village meetings open to any voter in that village (Gram Sabhas or Gram Sansads) to report its activities, bring its budget to a vote, and decide on recipients of social programs, at a minimum of two to four yearly village-wide meetings. Finally, the council head must “set up regular office hours where villagers can lodge complaints or requests” (Chattopadhyay and Duflo 2004, 1412).
21 While the British Columbia Citizens’ Assembly explicitly secured equal descriptive representation of men and women (James 2008, 110), we are not aware of reports of women’s substantive or symbolic representation.
22 It would be easy to assume that a random sample of the population will provide adequate descriptive representation, and it may do so (see Fishkin 1995), but as we show, that will not necessarily secure equal substantive or symbolic representation for women.
23 Ban and Rao (2009) studied speaking in 121 Indian village meetings constitutionally empowered to make important local distributional decisions. The transcripts were matched with data from household surveys conducted in the village before the meeting, allowing them to see whose preferences were expressed and implemented.
24 Conover, Searing, and Crewe’s (2002) measure of participation in political discussion includes discussion in public meetings or in informal situations in public places such as workplaces and churches.
25 Policy makers and nonprofit organizations have become interested in deliberation as a means of democratization and addressing corruption. In India, village meetings are constitutionally empowered to make important local decisions that affect private and public goods. The idea is that meetings can increase the transparency and accountability of government overall. In a society permeated by political corruption, these are urgent goals. But the risk of grassroots, local participation in decision making is that those with disproportionate influence will “capture” these processes and undermine the reform (Ban and Rao 2009). As Cornwall put it, women “are those most likely to lose out, finding themselves and their interests marginalized or overlooked in apparently ‘participatory’ processes” (2003, 1325; Guijt and Kaul Shah 1998; Mayoux 1995; and Mosse 1995).
26 http://www.un.org/womenwatch/daw/egm/eql-men/, accessed June 14, 2013.
27 This held particularly in bare-bones facilitated settings but not in actively facilitated settings; however, the study has methodological issues that make conclusions about facilitation difficult.
28 Bryan 2004, 214, note 2, and 216, note 6; Ban and Rao 2009.
29 Positive findings are reported by Chattopadhyay and Duflo (2004).
30 Reingold concludes that “there is no clear, positive relationship between sex ratios in legislative rosters and 1) the frequency or magnitude of sex differences in legislative behavior or 2) overall levels of policy activity or outcomes promoting women’s political interests” (2008, 140).
31 Beckwith 2007; Bratton 2005; Dodson 2006; Franceschet, Krook, and Piscopo 2012, 237–41; Walsh 2012.
32 Dahlerup typifies the conclusions of many studies in arguing that more important than critical mass are the standard factors predicting any type of policy success, such as the “political context,” the strength of relevant social movements and interest groups, and the “prevailing discourses” and frames (2006, 520). See also Htun and Weldon (2012) and Weldon (2002), who find that in policies that combat violence against women, feminist movements matter more than women in legislatures or other factors.
33 We elaborate and support these claims in chapters 2 and 3.
34 But see Huddy, Cassese, and Lizotte (2008, 39).
35 The gender gap persists even after controlling for partisanship and demographic variables (Crowder-Meyer 2010; Huddy, Cassese, and Lizotte 2008, 38). This is consistent with the framework of Carol Gilligan (1982). Women also tend to prioritize the needs of women, but inconsistently, though female legislators do appear to do so fairly consistently (Dodson 2006). We will discuss these issues more fully in later chapters, especially chapter 7.
36 We develop this discussion more fully in the conclusion chapter.
37 Studies based on interviews with leaders in various countries often verify that this view holds at least in leaders’ perceptions of female colleagues (Norris 1996; Childs 2004b).
38 Atkeson and Carrillo 2007; Barnes and Burchard 2012; Burns, Schlozman, and Verba 2001; Reingold and Harrell 2010; Smith, Reingold, and Owens 2012; Wolbrecht and Campbell 2007; but see Karp and Banducci 2008.
39 The American public even endorses gender balance in government. The plurality articulates the equal balance ideal (Dolan and Sanbonmatsu 2009, 419).
40 The arguments in this paragraph are more fully developed and documented in chapters 2 and 3.
41 See the extensive literature on ambivalent sexism and the notion that women, and other social groups, are independently rated as high on warmth but low on competence (Eagly and Carli 2007).
42 See, for example, Barber 1984; Bohman 1997; Chambers 1996, 2003; Fishkin 1995; Gutmann and Thompson 1996, 2004; Habermas 1989, 1996; Macedo 1999.
43 On this latter point, see, for example, Gutmann and Thompson 2004 and Fishkin 1995; but see Hibbing and Theiss-Morse 2002.
44 Some critics of deliberation argue that in place of deliberation, what subordinate groups need is a competition of ideas. However, this model is at least as disadvantageous to women as the criticized model of Habermasian deliberation. If the exchange resembles “a vibrant clash of democratic political positions,” as Mouffe advocates, that will disadvantage people who do better under conditions of interdependence, including many women (Mouffe 2000).
45 Chambers posed a similar question (2003, 322).
46 See chapter 10 for further discussion.
47 Partial analyses include Stromer-Galley and Muhlberger 2009; Stromer-Galley 2007. Partial qualitative studies include Walsh 2007 and Rosenberg 2007b.