“Settings where citizens get together to talk and make decisions are ubiquitous parts of American political life and always have been.”

Among the most important—and distressing— trends in American politics is how elected officials and political institutions respond, or fail to respond, to public preferences. In an age of rising economic inequality, it appears that the preferences of wealthy citizens are far more influential on elite decision-makers and, ultimately, policy outcomes than those of the middle class or poor.1→Larry M. Bartels, Unequal Democracy: The Political Economy of the New Gilded Age (New York: Princeton University Press, 2008); Larry M. Bartels, Unequal Democracy: The Political Economy of the New Gilded Age, 2nd ed. (New York: Princeton University Press, 2016).
→Martin Gilens, Affluence and Influence: Economic Inequality and Political Power in America (Princeton, NJ: Princeton University Press, 2012).
→Martin Gilens, and Benjamin I. Page, “Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens,” Perspectives on Politics 12, no. 3 (2014): 564–81.
→See also Stuart N. Soroka and Christopher Wlezien, “On the Limits to Inequality in Representation,” PS: Political Science & Politics 41, no. 2 (2008): 319–27.
But in a democracy, disparities of influence are not merely a problem for representation—the vertical relationship between ordinary citizens and their elected officials—but also for citizens themselves as they convene in discussion groups, committees, boards, clubs, associations, town hall meetings, or other similar gatherings. Settings where citizens get together to talk and make decisions are ubiquitous parts of American political life and always have been. Whether in the United States or elsewhere, the discussions and decisions of groups of citizens comprise the bustling heart of democratic life. For this reason, what happens in these meetings merits careful study. If such gatherings merely replicate society’s background inequalities, with those who already have economic or social power having the consistent advantage, then democracy suffers.2→Jane Mansbridge, Beyond Adversary Democracy, new edition (Chicago: University Of Chicago Press, 1983).
→Lynn M. Sanders, “Against Deliberation,” Political Theory 25, no. 3 (1997): 347–76.
→Iris Marion Young, “Activist Challenges to Deliberative DemocracyPolitical Theory 29, no. 5 (2001): 670–90.

The (potentially) good news is that group discussion is a dynamic process that might, under the right circumstances, do more than merely replicate existing power relationships.3Dennis F. Thompson,“Deliberative Democratic Theory and Empirical Political Science,” Annual Review of Political Science 11 (June 15, 2008): 497–520. Authority, defined as the expectation of influence, is not a static characteristic: it can be built or undermined, developed or dissolved in the give-and-take of conversation and group decision-making. Thus, what happens in group meetings matters not just for the substantive decisions that emerge from them (though who wins and who loses is no doubt important, especially if the same group repeatedly wins), but also because such meetings are places where citizenship and community (broadly conceived) are constructed. The key question—and one that deserves considerable additional attention—is what factors help disadvantaged individuals participate more actively and achieve greater influence when they talk with their neighbors, community members, or fellow citizens?

Examining small-group dynamics

Answering this question has been the core aim of much of our recent research.4→Christopher F. Karpowitz, and Tali Mendelberg, The Silent Sex: Gender, Deliberation, and Institutions (Princeton, NJ: Princeton University Press, 2014).
→Christopher F. Karpowitz, Tali Mendelberg, and Lee Shaker, “Gender Inequality in Deliberative Participation,” American Political Science Review 106, no. 3 (2012): 533–547.
→Tali Mendelberg and Christopher F. Karpowitz, “Power, Gender, and Group Discussion,” Advances in Political Psychology 37, no. S1 (2016): 23–60.
→Tali Mendelberg and Christopher F. Karpowitz, “Women’s Authority in Political Decision-Making Groups,” The Leadership Quarterly 27, no. 3 (2016): 487–503.
→Tali Mendelberg, Christopher F. Karpowitz, and Nicholas Goedert, “Does Descriptive Representation Facilitate Women’s Distinctive Voice? How Gender Composition and Decision Rules Affect Deliberation,” American Journal of Political Science 58, no. 2 (2014): 291–306.
→Tali Mendelberg, Christopher F. Karpowitz, and J. Baxter Oliphant, “Gender Inequality in Deliberation: Unpacking the Black Box of Interaction,” Perspectives on Politics 12, no. 1 (2014): 18–44.
Our specific focus has been on gender inequalities in groups, though as we indicate below, examination of other types of social inequalities is also needed. To explore authority in group settings, we assembled participants at two different sites in the United States, then randomly assigned them to five-person groups. We varied the number of women and men in the group and also varied the decision-rule (majority or unanimity) required for a group decision. After the participants privately answered some questions about themselves and their opinions, we brought them together as groups and asked them to discuss and make a decision about how, if at all, they would redistribute their earnings from a task they would complete later in the experiment. Some participants were likely to do very well on the task, we informed them, but some were likely to do quite poorly. We carefully recorded what happened during the group’s discussion. Afterwards, we asked the participants to tell us privately about their impressions of what happened and how they felt about it.

Our lab experiment allowed us to collect several key measures of authority within the group setting, including how much each member of the group talked, how often they interrupted each other, what each person said and whether it matched the views they reported to us prior to the group discussion, what the group ultimately decided, and how the participants evaluated each other after the discussion was over, including their views about who was most influential. We were then able to compare women in their groups to the men in that same group and to the men and women in other groups.

Inequality in deliberation

The results were striking. In the groups that are most common in political settings—where women were outnumbered by men and the group made decisions by majority rule—women faced a dramatic influence deficit. In groups with one woman and four men, for example, the woman makes up 20 percent of the group, so one metric of equal participation would be whether women take up 20 percent of the conversation. But in our lab experiment, women in such groups accounted for only about 12 percent of the conversation. In other words, women’s talk was barely over half of what it would have been if they were speaking up in proportion to their presence in the group. Among women who entered the conversation least confident in their ability to participate, the numbers were even worse. To be sure that this result wasn’t a function of something peculiar about the lab setting, we also collected data from school boards around the country and examined the patterns of participation across boards with different gender compositions. In the school boards data, the patterns were nearly identical to what we found in the lab: women who were outnumbered by men spoke up far less often than we would have expected, even after accounting for their lower presence in the group.

“Women were also less likely to receive positive feedback and support from men.”

The lab study revealed additional indicators of gender-based power disparities in the same groups where women were relatively silent. Women were more likely to be interrupted by men in negative ways—by expressing disapproval with words like “but,” “no,” “I disagree,” or by changing the subject. These instances showed how the interrupter expressed skepticism or opposition to what the speaker was saying or took the floor and changed the subject without any acknowledgement of what the previous speaker had been saying. Women were also less likely to receive positive feedback and support from men. Such positive feedback—words like “yes,” “uh huh,” “I see,” or “good point”—is especially important, because it helps to build the authority of the speaker, letting her know that others are listening and engaging actively with what she says. In addition, women were less likely to speak about topics of distinct concern to women, and less likely to advocate on behalf of the disadvantaged.5Previous studies showed that women are more likely than men to raise a set of policy domains that we labeled “care issues”: specifically, issues dealing with children, family, the poor, or the needy. Given these facts, it is perhaps not surprising that the group ultimately made decisions that were less friendly toward the disadvantaged. In sum, when women were outnumbered by men in groups deciding by majority rule, women’s lower status was reflected in the microdynamics of the group discussion.

As if that were not deficit enough, these dynamics were also correlated with lower levels of authority for women, as measured by the other members of the group. In groups where women were outnumbered, not only did women participate less and differently, they were also far less likely to be named by the other members of the group as the “most influential” participant. Again, these disproportionately low stores of authority and influence were present even after accounting for the fact that there were more men than women in the group. Nor were these findings a function of gender disparities in education, age, or income.

Achieving equality in small-group deliberation

Given these rather discouraging findings, where is the evidence that women can build authority and influence in group discussion? Recall that we randomly assigned individuals to different types of groups, varying both the gender composition and the group’s decision rule. These group-level features of the discussion are the key to the remedy. Nearly all of the negative dynamics we described in majority-rule groups were cancelled out when the gender composition favored women, though it often took a supermajority of women before women reached equality of participation and influence. In these groups, women were more likely to speak up, less likely to experience negative interruptions, more likely to receive positive feedback, and more likely to talk in ways that reflected women’s distinctive concerns. Finally, women were more likely to be seen as having authority and influence in the group’s discussion. That increased influence was reflected in the group’s decision, which more closely reflected women’s preferences regarding income redistribution.

“For women outnumbered by men, this signal had the effect of sharply reducing the gender disparities of authority and influence we described above.”

The other feature of the group over which we exercised experimental control—the group’s decision rule—also mattered. In groups where women were outnumbered by men but the group was instructed to decide by consensus, the dynamics differed sharply from what we found under majority rule. The unanimity requirement signals that every voice and every vote matters, which means that it empowers those who are outnumbered in the group. (One downside of a consensus rule is that decision-making may take longer, or benefit those who strongly prefer the status quo over change.) For women outnumbered by men, this signal had the effect of sharply reducing the gender disparities of authority and influence we described above. (It also boosted male participation dramatically in groups where men were outnumbered by women.)

In sum, small-group processes, structures, and institutions matter: they influence the discussion dynamics and ultimately the group norms through which people build influence and authority. In other words, the small-group settings that represent the ground-level foundation of modern democracy help to construct empowered citizens, but only if those settings are structured with disparities of influence in mind. Our results are thus simultaneously sobering and hopeful. Though gender inequalities continue to pervade many common political settings, features of the group are themselves a matter of political decision-making and are often under the direct control of meeting organizers. We can choose to structure group settings in ways that support and empower voices that in the past have too often been lost.

Next steps

Of course, ours is but one study, and much more remains to be done to understand the construction of influence and authority in group discussion settings6We see several productive lines of future research, though we only have space to mention a few of them here:
1) Other markers of political and social disadvantage ought to be further studied, such as class, race, and ethnicity.
2) Different structural factors and decision rules should be considered, aside from the two discussed in this essay. Every institutional decision has both benefits and drawbacks, and the effects of different interventions should be examined.
3) Democratic discussion and decision-making often involves small groups that meet repeatedly over extended periods of time, which create different circumstances and relationships that need further study.
as well as the relationship between small-group discussions and the larger system of political decision-making. Disparities of influence occur in formally empowered settings, too, as our study of gender and school board deliberation makes clear. But more remains to be learned about how the process of building influence and authority unfolds in all parts of the political system. This may mean paying close attention to how elite messages about the importance of equal authority and influence can boost disempowered voices.7Christopher F. Karpowitz, J. Quin Monson, and Jessica Robinson Preece, “How to Elect More Women: Gender and Candidate Success in a Field Experiment,” American Journal of Political Science, March 24, 2017. It also means thinking carefully about how building influence in one context transfers over to other contexts. For example, Karpowitz and Raphael argue that disadvantaged perspectives might especially benefit from the opportunity to talk in enclave groups as a first stage before deliberation in more diverse groups.8New York: Cambridge University Press, 2014More Info → Our lab study provides some evidence that groups comprised entirely of women had unique beneficial effects, though again, the effects were contingent on other features of the group, such as the decision rule.9Christopher F. Karpowitz and Tali Mendelberg, “Do Enclaves Remediate Social Inequality?” The Journal of Politics (forthcoming).

As we write, the political discussion in the United States is focused on the challenges of our nation’s history of racial and gender inequality. If progress toward the democratic promise of equality for all is to be realized, then understanding how equal and lasting authority is constructed in the discussion groups and public meetings that are the essence of democratic life ought to be a central concern.

References:

1
→Larry M. Bartels, Unequal Democracy: The Political Economy of the New Gilded Age (New York: Princeton University Press, 2008); Larry M. Bartels, Unequal Democracy: The Political Economy of the New Gilded Age, 2nd ed. (New York: Princeton University Press, 2016).
→Martin Gilens, Affluence and Influence: Economic Inequality and Political Power in America (Princeton, NJ: Princeton University Press, 2012).
→Martin Gilens, and Benjamin I. Page, “Testing Theories of American Politics: Elites, Interest Groups, and Average Citizens,” Perspectives on Politics 12, no. 3 (2014): 564–81.
→See also Stuart N. Soroka and Christopher Wlezien, “On the Limits to Inequality in Representation,” PS: Political Science & Politics 41, no. 2 (2008): 319–27.
2
→Jane Mansbridge, Beyond Adversary Democracy, new edition (Chicago: University Of Chicago Press, 1983).
→Lynn M. Sanders, “Against Deliberation,” Political Theory 25, no. 3 (1997): 347–76.
→Iris Marion Young, “Activist Challenges to Deliberative DemocracyPolitical Theory 29, no. 5 (2001): 670–90.
3
Dennis F. Thompson,“Deliberative Democratic Theory and Empirical Political Science,” Annual Review of Political Science 11 (June 15, 2008): 497–520.
4
→Christopher F. Karpowitz, and Tali Mendelberg, The Silent Sex: Gender, Deliberation, and Institutions (Princeton, NJ: Princeton University Press, 2014).
→Christopher F. Karpowitz, Tali Mendelberg, and Lee Shaker, “Gender Inequality in Deliberative Participation,” American Political Science Review 106, no. 3 (2012): 533–547.
→Tali Mendelberg and Christopher F. Karpowitz, “Power, Gender, and Group Discussion,” Advances in Political Psychology 37, no. S1 (2016): 23–60.
→Tali Mendelberg and Christopher F. Karpowitz, “Women’s Authority in Political Decision-Making Groups,” The Leadership Quarterly 27, no. 3 (2016): 487–503.
→Tali Mendelberg, Christopher F. Karpowitz, and Nicholas Goedert, “Does Descriptive Representation Facilitate Women’s Distinctive Voice? How Gender Composition and Decision Rules Affect Deliberation,” American Journal of Political Science 58, no. 2 (2014): 291–306.
→Tali Mendelberg, Christopher F. Karpowitz, and J. Baxter Oliphant, “Gender Inequality in Deliberation: Unpacking the Black Box of Interaction,” Perspectives on Politics 12, no. 1 (2014): 18–44.
5
Previous studies showed that women are more likely than men to raise a set of policy domains that we labeled “care issues”: specifically, issues dealing with children, family, the poor, or the needy.
6
We see several productive lines of future research, though we only have space to mention a few of them here:
1) Other markers of political and social disadvantage ought to be further studied, such as class, race, and ethnicity.
2) Different structural factors and decision rules should be considered, aside from the two discussed in this essay. Every institutional decision has both benefits and drawbacks, and the effects of different interventions should be examined.
3) Democratic discussion and decision-making often involves small groups that meet repeatedly over extended periods of time, which create different circumstances and relationships that need further study.
7
Christopher F. Karpowitz, J. Quin Monson, and Jessica Robinson Preece, “How to Elect More Women: Gender and Candidate Success in a Field Experiment,” American Journal of Political Science, March 24, 2017.
8
New York: Cambridge University Press, 2014More Info →
9
Christopher F. Karpowitz and Tali Mendelberg, “Do Enclaves Remediate Social Inequality?” The Journal of Politics (forthcoming).