Beating Bias: Why Women Make Better Candidates and Members of Congress

Political scientists at the University of Chicago Harris School of Public Policy have built a model that sheds new light on why women remain acutely underrepresented in U.S. politics.

The findings challenge scholars’ conclusion in recent years that the problem can be chalked up almost entirely to women not running for office.

“These are really important issues where I think for a long time, political scientists were getting it wrong,’’ said Christopher R. Berry, William J. and Alicia Townsend Friedman Professor, academic director of the Center for Municipal Finance, and faculty director of the Master of Science Program in Computational Analysis and Public Policy (MSCAPP).

“The funny thing is that while political scientists may have been getting this wrong, any woman you talk to who is actually in politics is never surprised’’ to learn that voter bias is at play when women run for office, Berry said.

Working with Profs. Scott Ashworth and Ethan Bueno de Mesquita, Berry set out to create a model to further understanding of women being underrepresented in office. The model examined the role of both voter bias and “election aversion.’’  

Women are under-represented among officeholders and under-represented among candidates. But they win at about the same rate as men do, once you take away factors such as incumbency. And once women do run for office and win, they perform better than men. These facts are well established by years of political science literature.

The researchers found that only a model incorporating both voter bias and election aversion as factors supported and could explain all four of these proven facts about the US political landscape.

In recent years, academic research has focused on the importance of “election aversion,’’ a term covering a set of factors that affect women’s likelihood to become candidates, such as a difference in political ambition between men and women; the tendency for women to under-estimate their qualifications while men tend to over-estimate their own; and a more critical assessment of the likelihood of actually winning the seat.

The new research shows, however, that voter bias matters and is intertwined with election aversion in at least one key respect. Although counterintuitive, the existence of voter bias does not suggest women will win less often, but it does affect their decision-making.  Women candidates are sophisticated and strategic. When they decide whether to run, they are carefully assessing their chances of winning.

“They’re not going to run in a bunch of elections they expect to lose,’’ Berry said. “And even professional political scientists had a hard time understanding that.’’

As a result of that self-selection, the very best female candidates are running, and that allows them to win just as often as men.

The new findings build on Berry’s 2009 work with Professor Sarah Anzia of Stanford University. Their study titled “The Jackie (and Jill) Robinson Effect: Why Do Congresswomen Outperform Congressmen?’’ found that only the most talented, hardest-working female candidates overcome voter bias and their own self-assessment to win elections.

There are a record number of women on the ballot this election cycle including, notably, Sen. Kamala Harris running for vice president, though even the greatest success rate for women on November 3 cannot appreciably change the woeful measure of low representation. (In the current Congress, women hold just 23.7% of seats and in statewide offices the number is just a bit higher. Yet women are more than half of the US population.)

It is an election that could mean significant progress for women, though, for most voters, gender will not drive decisions, Berry said. COVID-19, racial equity, and the economy appear to be more at the forefront of peoples’ minds as they go into the voting booth this year.

Still, a significant question is what the Trump presidency has done to either increase or reduce voter bias against women. Berry sees two forces pushing in opposite directions.

“On the one hand, I think the Trump presidency has made all kinds of discrimination more acceptable to express, and so I can see it being much more likely that there will be an anti-women bias expressed openly in lots of elections and women doing less well as a result,’’ Berry said.

The flip side is that the Trump presidency and the #MeToo movement have motivated many to push back against this kind of discrimination.

“Both of these things are at work and the net effect for female candidates is hard to predict,’’ Berry said. “But my gut instinct is that the forces against this discrimination are going to be stronger, and on that we’re going to probably do better in this election with respect to the representation of women.’’

Berry and his colleagues expect the Harris model to be a useful to the field of study, unlocking further understanding of the issue and leading to changes in approach.

“The generation of research prior to ours was really focused on the perception gap -- the whole problem is that women aren’t running and we need to convince them to run,” Berry explained. “The answer was candidate recruitment.’’

But if gender discrimination and stereotyping are large factors too, as the new model demonstrates, then politic leaders might want to approach it in slightly different way going forward. 

The world will soon learn how the record number of women who made the difficult decision to run for office in 2020 fared, voter bias or not.