News that Barbara Mikulski, the longest-serving woman in Congress, will not seek re-election in 2016 prompts reflection on both her four decades of pioneering service and the cause of women’s representation more generally. Despite some progress, women remain woefully under-represented in both Congress and most state legislatures, relative to their proportion of the population. In the 114th Congress, for example, only about 19 percent of legislators are women (20 in the Senate and 84 in the House), despite the enduring fact that women comprise half of the nation’s population. And this sort of disparity is common in other settings, including town councils, judicial bodies, and many businesses and work groups, as well.
How might things change if our legislative committees and other decision-making bodies included more women or rules designed to empower women? Our focus is on the substantive issues women raise and discuss. Specifically, what are the conditions under which women’s distinctive concerns find full expression in decision-making groups, and when do women’s concerns take a backseat to other interests and perspectives? To answer those questions, we focus on two important features of groups: the number of women present and the rules groups use to make decisions.
Previous studies have shown that while variation exists within each gender, women privilege issues of care and compassion for vulnerable groups in ways that differ from men. Melody Crowder-Meyer shows that when given a chance to respond to an open-ended question about the nation’s “most important problem,” for example, women are dramatically more likely than men to mention poverty, homelessness, and children. Men, on the other hand, tend to focus on jobs, energy prices, and taxes more than women do. These differing priorities can be seen in other ways as well. For example, Michele Swers finds that women in Congress are more likely than men to prioritize issues related to women, children, and families, including in committee deliberations. To be sure, men and women share some important priorities (such as the economy, or health care), but because they occupy somewhat different roles in society, they also hold some distinct priorities.
However, in our AJPS article, “Does Descriptive Representation Facilitate Women’s Distinctive Voice? How Gender Composition and Decision Rules Affect Deliberation”, we argue that women’s distinctive priorities find less expression where women’s status in the group is low. Two factors that affect that status are women’s relative number, and the group’s procedures. To explore how those features affect the content of group discussion, we conducted an experiment in two very different places – one religious and conservative, the other secular and liberal. We recruited students and community members and assigned them to five-member groups. Importantly, we randomly assigned the group’s gender composition and its rules for making binding decisions, which allows us to examine the causal effects of these two factors.
Following the basic outlines of a study by Frohlich and Oppenheimer, participants were asked to meet with the other members of the group, have a “full and open discussion,” and make a decision about how to redistribute the money they would earn later in the experiment. They weren’t told what they would be doing to earn money, so they could not predict if they would be poor or well off. Thus, members had to decide whether or not to institute a “safety net” for the poorest performing members, whoever they might be, and how generous that safety net would be. Groups randomly assigned to the unanimous condition needed all members to vote for the same redistributive rule, while those in the majority condition required just three votes for a binding decision. In the end, the money they actually took home from the experiment was based on both their individual efforts and the group’s decisions about redistribution.
We conducted a systematic content analysis of words spoken by each participant, looking specifically for whether participants raised issues of “care” by talking about children, families, the poor, and the needy (issues that previous work showed are more likely to be the distinctive concerns of women), as opposed to “financial” issues, such as salary or taxes (which tend to be male priorities). To rule out the possibility that participants raised care issues in unsympathetic ways, we classified each mention as sympathetic, neutral, or negative and found that mentions were negative only about 5% of the time.
Most importantly, we found that the content of the discussion varied dramatically with the group’s composition and rules. In groups where women’s status was lowest – a lone woman was surrounded by four men and the group was instructed to decide by majority rule – women used words related to “care” issues at a very low rate. Lone women in groups with four men invoked care an average of only 4 times for every 1,000 words and never mentioned family. In all groups where women were outnumbered by men, women raised care issues at a rate of only 6 words per 1,000. Even then, they almost always waited until men had raised the issue first. Thus, in the context that is most common in many political and business settings – majority-rule groups with a minority of women – women were simply far less likely to raise topics of distinctive concern to women.
In majority-rule groups where women outnumbered men, however, women invoked care issues much more often – at a rate of about 15 words per 1,000. (That number represents the average per individual, so when multiplied by the greater number of women in the group, it represents a substantial change in the focus of the group conversation.) Another way of examining these data is to explore the ratio of the frequency of care issues to financial issues for each group member. When men formed the majority in a group deciding by majority rule, the average ratio was a paltry 0.44, but it rose to 2.7 (a more than 6-fold increase) when women outnumbered men. When groups decide by majority rule, our data shows, the conversation turns to care issues only when women are in the majority.
But our findings also reveal another way to achieve greater voice for women. In the groups assigned to decide by unanimity, women raised care issues even when they were in a minority – at a rate of over 10 words per 1,000 no matter how many women were in the group (though the rate is highest in all-female groups). In other words, in mixed-gender groups, women spoke about issues of distinctive concern to them at the same rate whether the group included few or many women. This means that in groups with few women, deliberators raised care issues much more often under unanimity than under majority rule, but groups assigned to unanimity also saw no dramatic increase in care issues as the number of women increased.
Why does unanimity have such a profound effect? When every vote is needed to come to a decision, every voice is valuable, and women in groups with many men have a greater ability to set the group’s agenda and the content of its discussion. By that same token, unanimity empowers any minority, male or female, so it also facilitated increased discussion of male priorities in groups where women predominated. That is why, under unanimity rule, having more women in the group didn’t substantially increase the focus on issues of distinctive concern to women.
All of these results are robust to the presence of controls for the experimental location, the ideology of the participants, and the group members’ pre-discussion preferences about specific principles of redistribution. The combination of gender composition and decision rule makes a profound difference in the emphasis of group conversations, even when we account for other aspects of the group or the individuals who comprise it. Together, rules and numbers create norms of group interaction that empower women – or fail to do so – and this empowerment can be seen in women’s ability to turn the conversation to issues they tend to prioritize most.
About the Authors: Tali Mendelberg is a Professor in the Department of Politics at Princeton University, Christopher F. Karpowitz is an Associate Professor of Political Science at Brigham Young University and Nicholas Goedert is a Visiting Professor of Government and Law at Lafayette College. Their article “Does Descriptive Representation Facilitate Women’s Distinctive Voice? How Gender Composition and Decision Rules Affect Deliberation” appeared in the April 2014 issue of the American Journal of Political Science.