THE TITLE OF THIS BOOK TAKES OFF FROM Simone de Beauvoir’s classic tome The Second Sex, originally published in 1949. That book was the first ambitious piece of research on the problem of women’s lower status in society. Although we have come a long way since 1949, women remain second-class citizens in reality if not in law. How this status plays out in a key arena of politics and society—public discussion—and what to do about it is the focus of our book. Women are not the “silent sex” in all domains. However, we show that they are less likely than men to talk and to influence others when discussing matters of common concern. This gender gap is far larger than women’s tremendous progress in Western society would indicate. In the most powerful settings where common affairs are decided, where women are often few and discussion is governed by competitive norms, the average woman participates at only two-thirds of the average man’s rate. Along with this quiescence, women are perceived as less influential, women set the agenda less effectively, and women more often fail to move the decision toward their preferences. That is, women tend to enter—and to exit—with less authority than men.
Authority is the expectation of influence, and society typically grants less of it to women than to men.1 So men and women tend to enter the room with different levels of expected influence. However, a person can act in ways that enhance or detract from their authority. By the same token, others can act in ways that enhance or detract from another’s authority. In both cases, these actions affect others’ and one’s own expectation that one can or should carry influence. Therefore, the actions that people exchange during discussion affect the authority gender gap.2
And those actions are in turn shaped by the rules, procedures, and the social composition of the group. Some procedures and compositions can affect the interaction in ways that elevate women’s status in the group, thereby remedying the authority deficit women face. But other rules and compositions create dynamics of conversation that further erode women’s low standing, and then, women tend to exit with even less than they had when they began.
The final twist is that the rules interact with the number of women present, so that the same rule works very differently depending on who is in the room. That is, the discussion group’s gender composition and decision rule matter a great deal, but the effect of the former depends on the latter. When groups are composed of many women and the group uses majority rule, women’s authority rises and sometimes equals that of men. But even when women are few, the right rules for collective decision making can ameliorate the gender gap in authority. Women can achieve a higher status and authority in part by their own actions—namely, by participating and articulating their preferences at rates equal to men. They also achieve this status and authority in part as a consequence of others’ actions: women’s authority is boosted when the degree of warmth and affirmation they receive rises. When women’s status has been instantiated in these ways during discussion, women are more likely to voice their distinctive concerns and preferences and to gain influence in their own and in others’ eyes. And this increased voice has an effect on collective outcomes: the group sets policies that are more generous toward the poor and vulnerable. Thus, attending to the joint consequences of who is in the room and the rules by which groups make decisions allows us to offer new insights about gender and authority.
Gender is a key theme of the book, and we use the concept of gender in several ways. First, and most obvious, gender refers to a social identity as a man or a woman. However, gender is not simply a characteristic of the individual man or woman, nor is it reducible merely to average differences between men and women. It is also a dimension of the style of interaction between individuals, which can be more masculine or more feminine. When the interaction tilts toward a style associated with men, or toward a style associated with women, it is “gendered.” Similarly, gender is also a characteristic of settings. When a setting is structured in a way that emphasizes overt competition, or in a way that stresses cooperation, it is gendered, since these modes are associated with men and women, respectively.3 These gendered aspects of interactions and of settings in turn can have an additional gendering effect by affecting the authority of women. Some of the gendered effects of interactions and settings affect women’s authority by shifting the behavior of women, and some affect that authority indirectly, by shifting the behavior of men. Women can enhance their authority by speaking more; men can lend women authority by validating their speech and supporting what they say. Our goal is to bring together these various ways in which gender matters to better understand the effects of gender on authority. We aim to do as Burns recommends: “The most successful work combining context and gender has done two things simultaneously. It has employed theories and measures placing gender in context, and it has deployed theories and measures of individual-level mechanisms, usually from psychology. In the end, this work has helped us understand not just that context enables gender to matter, but it has allowed us to begin to specify why and exactly how” (Burns 2007, 115–16).4
However, this book also addresses a number of issues beyond gender. First, we are exploring how people talk about economic redistribution, and therefore, our study sheds light on the determinants of policy on a salient issue of our time. Topics ranging from income inequality to poverty, to taxes, to the obligations of government, to the needs of children—all these are implicated in the discussions we investigate, and so are the policies that address these themes.
As a study of politics and public discussion, our book also speaks to those concerned with public opinion and political participation. We find that group discussion is a formidable force in shaping public opinion. But unlike many of the proliferating studies of deliberation, we attend to one very important fact about discussion: it occurs in small groups.5 Our definition of small groups draws on a part of Verba’s: “groups in which face-to-face communication is possible among all members” (1961, 12). This definition includes gatherings attended by as few as three and as many as dozens of people, and it is thus useful for capturing the dynamics of a large number of settings affecting many people. From the school board to the church committee, small groups are a defining characteristic of civic life. Yet while small groups are ubiquitous in our politics, studies of public opinion have neglected the impact of the characteristics of small groups on citizens.
Such neglect is all the more puzzling because several large literatures in economics, sociology, and psychology point to the importance of a group level of analysis. One of the most robust and strong findings of the literature on social dilemmas and games is that small group discussion converts a loose collective of individuals acting in their narrow self-interest into a cooperative unit acting for the common good of its members (Mendelberg 2002; Ostrom 1998). That literature finds that the group unit of analysis is crucial; for example, the group’s voting rules and agenda-setting procedures significantly affect the behavior of its members (Meirowitz and Landa 2009). Using quite different paradigms, social psychologists have documented over decades the crucial impact of group factors on individual members, including the group’s initial distribution of views and the persistent norms it develops (Mendelberg 2002). A group’s dynamics thus operate independently of the individuals in it and can shape those individuals in important ways. The group is, to paraphrase Kurt Lewin (1951), more than the sum of its parts.
Most importantly for scholars of political behavior, as we unpack the processes of deliberation and small-group interaction, we come to understand how people communicate and how this process of communication shapes not only attitudes and policy preferences but also the fundamental bedrock of political power—the authoritative exercise of voice. We provide the largest-scale analysis to date of how the specifics of interaction build or erode power, and these insights point to the importance of social factors in constructing the currency of politics. The key insights we offer scholars of political behavior are: first, the specific nature of interaction matters by shaping the authority of group members and, thereby, the weight of their opinion in the discussion; second, social identity must be taken into account, because who is interacting with whom also matters; and third, institutions and their procedures deserve a central place in attempts to understand how group dynamics construct authority and how authority affects group decisions.
This book’s focus on the social and institutional construction of authority in groups thus contributes not only to our understanding of gender and of political behavior but also to analyses of institutions. It investigates how institutions reinforce, or eliminate, social inequalities. Institutional procedures that appear neutral can in fact be systematically biased in favor of half the population. However, some rules can inoculate against the disadvantages created by society and level the playing field of political power. People walk through the door with stable differences shaped by social mechanisms, and institutions must take them as they are, but if the institution understands the nature of the social inequality it is dealing with, it can shape the way people interact, for good or for bad. The key is to recognize what scholarship on institutions seldom does: that institutions carry different effects depending on the composition of their members, and that the effects vary depending on the social identity of the member. We show this by undertaking an unusually systematic analysis of how members interact under the constraints of different rules and different social compositions.
Our focus on the group and on authority also allows us to speak to advocates —and critics—of democratic deliberation. Deliberation provides a potential remedy for the troubling ills of modern democracies. The average citizen’s woefully low levels of political knowledge, reasoning, and interest in politics call out for reform. Inviting people to participate in town-meeting-style forums and revitalizing the vibrant grassroots associations of the past are promising ideas that rightly give observers and practitioners hope. But deliberation rests on the foundation of the small group. Small groups, as scholars have long known, have their own serious pitfalls. We show that one of the pitfalls most advocates of deliberation neglect is gender inequality. Women are highly disadvantaged in many deliberative settings, and this disadvantage affects everything from how long they speak, to the respect they are shown, to the content of what they say, to the influence they carry, to their sense of their own capacity, and to their power over group decisions. All the more troubling, these problems can emerge even when the terms of discussion do not, at first glance, appear to disadvantage women. The problem is not that women are disliked or formally discriminated against; rather, the problem is that while women are liked, they are not given equal authority.
We demonstrate how to use the potential of deliberation without incurring its liabilities. Deliberation can come closer to the aspirations of its advocates, but only if its practitioners attend closely to the nature of the group and the rules by which group members interact. The key lies in who is deliberating and the rules and norms of deliberation. Deliberation requires careful institutional design, and we offer guidelines for that design.
To move beyond the existing literatures on deliberation, we rely on speech-act theory rather than on the philosophical assumptions of deliberation’s advocates. This theory is a general approach to the study of language that emphasizes its social meaning and functions (Holtgraves 2002, 5–6). While scholars of deliberation tend to focus on rationality, and thus on the truth-value, logic, and evidence that talk communicates, we take a different tack. We are not studying the types of phenomena that scholars of political behavior tend to study: knowledge gains, increasing tolerance of cultural difference, the group’s polarization to an extreme decision, or the effects of diverse preferences. Because we treat language as an act that conveys social meaning, such as status, or human connection, we analyze its social impact rather than the quality of the information or the level of reasoning contained in language.6 Because speech is a social act, it shapes rapport and connection as well as efficacy and influence. We examine speech as a form of symbolic political or civic participation that may reflect and contribute to the sense of political efficacy and authority—in short, as a political act that creates civic standing.
Our book brings not only renewed attention but also innovative methods to the question of how to create such standing for women. While some leaders and moderators of such groups, and scholars of small group settings, have long assumed that gender matters, and taken steps to include women, our findings can make several new contributions.7
First, this book uses controlled experiments that randomly vary the conditions of discussion, using many groups. Those results are validated with a study of naturally occurring groups. These methodological moves represent an advance, showing rigorously the depth of the problem and the great success of the solutions. While some practitioners and scholars understand the perils of the gender inequality we document, many do not fully do so. The most recent study on this subject, by Hickerson and Gastil, concludes that gender inequality is not a big problem, and that its findings “shift the burden of evidence somewhat by challenging deliberation’s critics to find compelling examples of difference effects in well-structured and consequential deliberative spaces” (2008, 300). Our evidence establishes that gender inequality is deep and pervasive despite the steps commonly assumed to guard against it.
Second, the group’s formal procedures are often neglected in these literatures yet turn out to matter a great deal. We show when and why unanimous decision rule, and the consensus process it can generate, improves gender equality in deliberation—and when it does not do so. Third, in previous studies, group gender composition is often overlooked, or measured inseparably from individual gender, or hampered by inconclusive tests or inconsistent findings. We explain these inconsistencies and show when they disappear. Fourth, we have put all these elements together and added to the picture a fine-grained analysis of actual interaction and the contents of speech, at the level of the individual member. To our knowledge, no previous study has done so. Finally and most important, some of the practices deemed to alleviate the problem of inequality, such as a consensus process, do not work in all cases, because they do not take into account the combination of gender composition and the rules of discussion.8 We offer an explanation for the successes of procedures that have a long tradition in the practice of small group discussion—and for the problems of those same procedures, problems that many dialogue practitioners have failed to notice.
Potential applications for our findings are found in all corners of civic life. In that sense, this book intends to shed light not only on deliberation preceding a formal decision but also on dialogue groups, work meetings, investigative committees, boards of civic associations, classrooms, and many settings where people convene and converse over matters of common concern. The gender dynamics of groups play out anywhere people gather and interact with each other in more formal settings.
Our own experiences in academia have provided numerous opportunities to observe such interactions, and it is clear to us that the number of women present and the group’s norm of interaction matter. In Karpowitz’s department, women’s representation is a salient issue. In Mendelberg’s department, one notable experience reinforces this conclusion.
Every year, Mendelberg’s academic department at Princeton assembles committees of its faculty to make initial decisions about hiring and promotion. Women compose approximately 25% of the faculty in the department. On a typical hiring committee, women are a minority, and the interaction tends to be competitive. But once in a while, the gender composition veers off this average. On one committee, Mendelberg served entirely with other women. The group consisted of five people, the same size as the groups assembled for our experiment. All were tenured. None was a shrinking violet. Yet in this group, unlike any of the other committees, the interaction took on what can only be described as a norm of niceness. The members affirmed each other’s statements. They smiled and nodded. Each person spoke, and at length. The group did not lack for deep disagreements; the members each came from a different field within the department, and their research paradigms were as disparate as can be found in the social sciences. But the general tenor of the interaction was geared toward validating the speaker’s basic worth as a group member. By the end, the group had reached agreement, not through conformity, but out of mutual respect. What happened at this meeting? How exactly did the group generate warmth, and how does that rapport lead to inclusion and genuine consensus? Why was this meeting different from all the others? And how can groups composed of both men and women institutionalize procedures that replicate this dynamic? This book tells the story of this meeting writ large.
1 We depart somewhat from the standard Weberian definition of authority as the legitimate exercise of power (Weber 1947), but our definition may be closer to the original one Weber himself offered, which is “the probability that a command with a specific content will be obeyed by a given group of persons, despite resistance, regardless of the basis on which that probability rests” (Uphoff 1989). We define authority as the expectation of influence, retaining Weber’s emphasis on the probability of obedience, and following Weber, we focus on ways in which social roles and expectations confer authority.
2 We use the term gender rather than sex to refer to the comparison of men and women, following the common practice in political science.
3 See Winter (2008) for the idea that structures of discourse can be gendered without mentioning gender.
4 We treat gender as a category of analysis (Beckwith 2005; Ritter 2008; Scott 1986). The concept of gender invites attention to men as well as women, to men’s and women’s interaction within their own sex as well as with the opposite sex, to masculine and feminine styles and settings, and to ways that actions vary along a continuum from more feminine to more masculine. While it is convenient to refer in shorthand to a discrete category of femininity or masculinity, they are more useful as end points on a continuum. For example, we inquire about the conditions of deliberation that move men’s and women’s priorities and preferences from the feminine to the masculine, or the reverse. Making these conceptual moves carries the additional benefit of allowing us to avoid essentializing or oversimplifying gender as nothing more than fixed and unchangeable stereotypes. While we recognize enduring differences between men and women, we also want to understand how priorities and patterns of interaction between and among men and women change as the features of the group change.
5 For some explorations of deliberation and small group dynamics, see Gastil (1993; 2010).
6 Developments in the philosophy of language have helped to recast the notion of language as a conduit of logical meaning into the notion of language as a form of social action. Speech act theory, propounded by the philosophers John Searle and J. L. Austin, claims that language is a set of rules that people use to enact their intention. Relatedly, Ludwig Wittgenstein argued that the meaning of language derives from the associations people form between it and their social contexts. See the discussion by Maynard and Peräkylä (2003).
7 This paragraph is elaborated and supported in the conclusion chapter.
8 Specifically, consensus building and inclusive processes may actually backfire for women when they are the majority, a problem not recognized in any scholarship but one we document. For more on such processes, see Gastil (1993; 2008).