Women Candidates and Election Aversion

By: Kristin Kanthak and Jonathan Woon

One of the bigger stories to come out of the 2012 election was the number of women – 20 – whom voters elected to the Senate. This number received its notoriety not because it was so large; indeed, women still comprise only 20 percent of the body. Rather, it is a reflection of how far women’s representation has come. At the same time, though, there is little to indicate that the proportion of women in office will reach parity on its own, largely because the number of women willing to run for elective office seems to have stagnated. Our work, for which we have received a National Science Foundation grant, attempts to explore why women don’t run, and perhaps to pinpoint some of the factors that may increase women’s participation in electoral politics.

We take the question of women’s candidate emergence to the social scientific laboratory, where we can control important factors such as the costs of running for election and can measure otherwise slippery quantities such as “candidate quality.” Our design is simple. We ask our experimental subjects to perform a task, and we pay them based on how well they did. We then tell them they will be doing the task again, and will again be paid based on their performance. But this time, they can also volunteer to make their performance count for all members of their randomly-chosen group. In one treatment, we ask them if they are willing to be selected at random to represent the group. In another, we ask them if they are willing to run for an election that will select the group’s representative.

What we find is dramatic and surprising: Women are far less likely to indicate a willingness to run in an election – even our very simple laboratory election – than are men, an effect we call “election aversion.” Perhaps even more startling, women tend to ignore completely their own ability to perform the representative’s job when deciding whether or not to run in the election. In other words, women don’t run, but smart women really don’t run. Furthermore, we see no such gender effect in the random selection treatment. Women are willing to volunteer to represent a group – and they can effectively take their own abilities into account when making that decision – but when you ask them to run in an election, they balk.

These results are rather chilling for those who care about the election of women representatives. There is a whole list of suspects that might help account for why women run less than men. For example, professional women are more likely to be married to other professionals than are professional men, which makes it harder for their careers to fit with those of their spouses. And women tend to take on more responsibility for child care, which makes it harder for their careers to fit with their family life. Yet our research indicates that even if one were effectively to balance all of those other external forces, we would still likely see gender differences in willingness to run. 

One potential next step in our research is to see if we can determine if external forces might be able to help mitigate gender-based election aversion. What if we told subjects about gender-based election aversion? What if we got them to think about women leaders who had made the choice to run? It is our hope that our continued research may help to answer some of these questions, while also helping to inform our more general understanding of how elections – and representation – work.