Kirsten Gillibrand, Kristen Anderson-Lopez, Linda McMahon, Kristen Van Ogtrop, Tamron Hall, Charlie Kammerer, Evelyn Webster, Michelle Peranteau and Nancy Gibbs attend the TIME and Real Simple’s Women & Success event at the Park Hyatt on Oct. 1 in New York. (Larry Busacca/Getty Images for Time Inc.)

Joshua Tucker: The following is a guest post from political scientists Chris Karpowitz (BYU) and Tali Mendelberg (Princeton).


A recent flurry of publicity for Sen. Kirsten Gillibrand’s new book, “Off the Sidelines: Raise Your Voice, Change the World,” has focused on the entrenched sexism in the U.S. Senate.  Gillibrand (D-N.Y.) – one of only 20 women in the 100-member Senate – recounts that after losing 50 pounds, a male colleague “approached her, squeezed her stomach, and said ‘Don’t lose too much weight now. I like my girls chubby.’”  While one member of Congress encouraged her to continue working out at the House gym “because you wouldn’t want to get porky,” another told her not to worry about her weight because, “You know, Kirsten, you’re even pretty when you’re fat.”  Sexist comments extended outside the halls of Congress, with Senate Majority Leader Harry M. Reid (D-Nev.) calling her “the hottest member of the Senate” at a fundraiser.

Gillibrand downplayed some of these remarks as coming from older colleagues who joined the Senate when it was more of a boys’ club and have not internalized contemporary norms of professional interaction. However, our research suggests that they are not merely the result of a few bad (old) apples, but instead, at least partly the product of the norms of conduct that often develop in majority-rule institutions with few women.

These comments might not seem like sexism at first glance. After all, what is wrong with one person complementing another on their physical appearance? But as psychologists Peter Glick and Susan Fiske found, such comments can be described as “benevolent sexism” – a subtler and softer form of sexism that saps women’s authority in part by depicting women as idealized beings who need to be protected and put on a pedestal.  Such benevolent sexism seems nice enough (who can argue against chivalrous gestures and protection) and may emerge from subjectively (and genuinely) positive feelings from men, but it has a dark side that can end up disadvantaging women.  (The classic literary illustration of this is Ibsen’s “A Doll’s House.”)  For example, the male co-worker who comments on how “cute” his female colleague looks can certainly be well-intentioned, but his behavior, however well meaning, can simultaneously undermine women’s equal authority.

Moreover, what the media focus on individuals’ sexism doesn’t tell us about is the role of the situation, and, broadly speaking, of politics.  Gendered behavior is not simply a function of individual attitudes, but also of the norms and patterns of interaction that are generated in different group contexts. In our new book, “The Silent Sex: Gender, Deliberation, and Institutions” (Princeton University Press), we find that women experience far worse treatment, and end up far more disempowered, when they are the numerical minority in institutions that use majority rule, because majority rule leads to norms of adversarial or competitive behavior. That is precisely the setting in which Gillibrand and her 19 female colleagues find themselves in the Senate. It is also the setting most common in committees, town councils and school boards across the nation.

We ran an experiment to examine what happens when men and women talk about important issues and make decisions as a group.  We assembled groups in a lab setting in two very different locations. And to verify that the results hold outside the lab, we reanalyzed Katherine Cramer’s dialogue groups and sampled school board meetings from a diverse sample of school districts.  We found that in groups where women were outnumbered and the group made decisions by majority rule, the outnumbered women spoke up less often and ultimately took up much less floor time – much less than even their minority status would have led us to predict.  For example, in our lab setting, women who comprised only 20 percent of the group accounted for only 13 percent of the comments made, and this result was replicated in a sample of school boards where women were outnumbered on the board.  Moreover, in the lab, when women spoke up less often, they were also less likely to raise issues of distinctive importance to women on national surveys, such as the needs of families and children or the plight of the poor and needy, and simultaneously became less willing to advocate for generous assistance to the poor and disadvantaged.  They even became less likely to speak up on behalf of the preferences they privately endorsed prior to the discussion (and in fact, they became more likely to speak up on behalf of preferences they privately opposed prior to the discussion).  In addition, these women emerged from the discussion less likely to be confident that their pre-deliberation preferences were correct.  (Male certainty was unaffected.)

The effects of group contexts can also be seen in the behavior of men toward women.  The same women who spoke little, rarely mentioned their distinctive concerns, and advocated less clearly for their preferences also experienced more hostile interruptions when they did talk. When women were outnumbered by men in groups deciding by majority rule, women received a high proportion of negative interruptions from men. Conversely, when women’s numbers grew, men’s behavior toward women changed – they became much more likely to interrupt with positive expressions of support – a cue that audience is actively engaged in what the speaker has to say.  These basic trends can be seen in Figure 1, which shows the average negative proportion of interruptions received by women from men.  Women’s likelihood of receiving hostile interruptions drops dramatically as the number of women in the group increases – for groups instructed to use majority rule.  (Notably, women’s experience under unanimous rule follows a very different pattern.)  Interruptions toward men, by contrast, did not change with the gender composition of the group; only women’s did.


Figure 1: Proportion of Negative Interruptions Received by Women from Men. Note: Dependent variable is the proportion of all negative and positive interruptions received by women from men that are negative (group-level average).  Dots are averages for each experimental condition.  Lines are predicted values generated from a regression model controlling for the number of comments made by group members. Data Source: The Deliberative Justice Project. Figure: Chris Karpowitz and Tali Mendelberg/The Monkey Cage  
Figure 1: Proportion of Negative Interruptions Received by Women from Men. Note: Dependent variable is the proportion of all negative and positive interruptions received by women from men that are negative (group-level average).  Dots are averages for each experimental condition.  Lines are predicted values generated from a regression model controlling for the number of comments made by group members. Data: The Deliberative Justice Project. Figure: Chris Karpowitz and Tali Mendelberg/The Monkey Cage

Why do seemingly trivial speech acts such as floor time and interruptions matter? Because they have a profound effect on women’s status and authority. In the lab, individuals who spoke more were more likely to later believe that their opinions were influential and that their voices were heard. Women’s sense of their personal influence also increased when they received fewer negative interruptions.  (Men’s self-efficacy, by comparison, was not affected by others’ interruptions.)  Moreover, speaking up and receiving positive validation from someone else builds influence in the eyes of fellow group members.  Men and women who held the floor for a greater percentage of the group’s conversation were dramatically more likely to later be identified by their fellow group members as the “most influential” group participant. Similarly, those who received more positive interruptions from their fellow group members were also more likely to be seen as influential.  The silent but authoritative woman is the exception, not the rule.

Ultimately, women’s influence was closely tied to the group context.  Women who found themselves as a distinct minority in groups deciding by majority rule – the setting in which they talked less, received less positive feedback, and felt lower levels of efficacy – were in fact less effective in altering the group’s eventual decision.  When women did not comprise a distinct minority in an adversarial setting or when the group’s decision rules empowered numerical minorities, the group’s discursive dynamic was far different – and so too was women’s ability to move the group’s outcome.

In sum, women’s authority in groups depends not just on the attitudes of men (whether sexist or not) but on the conditions of discussion – numbers and rules – and the norms that develop as men and women deliberate together.

What can we do in the face of findings like these?  While the results show just how great the deficit of authority is in common political settings – majority rule with few women – they also give some reason for hope.  After all, group contexts are not set in stone.  They are very much under the polity’s collective control.  In our research, when we altered those conditions by increasing the number of women in majority-rule groups or by using a different decision rule when women were few, the patterns of talk time, the content of speech, interrupting behavior, and ultimately, women’s authority within the group – changed, too.  In other words, when the public conversation focuses too narrowly on the attitudes of men (whether sexist or not), and ignores the group context in which collective decisions are made, we miss an important opportunity to understand how group norms and practices contribute to women’s status and authority.