Jane Junn is Professor of Political Science at the University of Southern California
In the wake of Donald Trump’s election as President of the United States, one piece of data from voter exit polls has been particularly surprising for Clinton supporters: 53% of white women voted for Trump compared with 43% for Hillary Clinton. This statistic has been met with disappointment and criticism: “Fellow white women, I’m done with you,” (Sarah Ruiz-Grossman, Huffington Post), “Self-loathing. Hypocrisy. And, of course, a racist view of the world that privileges white supremacy over every other issue.” (L.V. Anderson, Slate).
That white women supported Trump despite allegations of his sexual abuse of white women, particularly when the alternative on the ballot was also a white woman, results in a sense of puzzlement among commentators.
These stories and others about white women voters and their support for Trump over Hillary Clinton seem to assume that it is novel for this group to vote for Republican Party candidates. It is not at all unusual, and astute observers in the popular press were writing about this phenomenon after the last election. In 2012, John Cassidy published “What’s Up With White Women?” in The New Yorker, but few took notice because Obama won and the “gender gap” remained robust.
Defined as the difference between the proportion of women (55%) and men (45%) supporting the Democratic candidate, the gender gap was 10 percentage points in 2012, demonstrating yet again that American women supported the Democratic Party candidate. All the same, and as Cassidy noted in 2012, exit poll data revealed that 56% of white women voted for Romney compared with only 42% for Obama, a +14 percentage point margin for the Republican Party candidate among white females.
While the limitations of exit poll results in 2016 for minority populations have been documented, (Gary Segura and Matt Barreto,Huffington Post ), exit poll estimates in previous elections for white voters have hewed relatively closely to those reported in high-quality post-election studies such as the American National Election Studies.
Using these data, political scientists have observed this pattern of behavior over a much longer period, and anyone who looks at voting data can see that white women also supported the Republican Party nominee in 2008 by a margin of +7 percentage points, when 53% of white women voted for McCain and 46% supported Obama. My colleagues and I have demonstrated these patterns in a series of papers about the dynamics of gender and race in voting in the United States, and it is important to recognize that the category of white women does not represent a political monolith.
Perhaps you are thinking these 3 elections are anomalies; 2016 being unprecedented in strangeness of all kind and 2008 and 2012 odd given the presence of the first African American candidate from a major political party running for President.
2016, 2012, and 2008 are indeed different from those they preceded for precisely these reasons, but what has remained consistent is the support of white female voters for Republican Party candidates. In how many presidential elections between 1952 and 2012 have white women supported Democrats more than Republicans? The answer is two. We can now extend the time series to 2016, and the number of times white women voted more for Democratic candidates over Republicans remains two.
The elephant in the room is white and female, and she has been standing there since 1952. This result has been hiding in plain sight, obscured by a normative bias that women are more Democratic than men. They are, and it is also true that white women are more supportive today of Democratic Party candidates than white men. But this does not mean that white women are more Democratic overall. They are not.
While the white female vote is often closely split between the two major parties, white women have only voted more Democratic than Republican twice in the 17 U.S. Presidential elections since 1952 (in 1964 and 1996). Instead, it is the introduction and steady growth of minority voters in the U.S. electorate over the last six decades that drives higher overall proportions of female support for Democratic Party candidates.
The elephant in the room is white and female, and she has been standing there since 1952. This result has been hiding in plain sight, obscured by a normative bias that women are more Democratic than men.
The gender gap in voting first appeared in the mid-1980s, and prior to then, women and men showed similar preferences for Democratic and Republican Party candidates. The divergence is not only due to potential conversion among women voters based in gender consciousness, or to the movement of white men away from the Democratic Party, but instead the beginning of a tipping point in the composition of the female electorate. Two decades after the Civil Rights Act, the Voting Rights Act and the Immigration and Nationality Act, women of color entered the electorate, voted strongly Democratic, and helped to create the gender gap, carrying white women with them.
Women of color now make up nearly a third of female voters, and support Democratic candidates by wide margins, with African American women the stalwart of the Democratic Party. It is thus compositional change – a new set of voters entering the electorate and becoming a larger proportion of American voters – driving the patterns we observe in gender, race, and electoral politics today.
Analyzing the dynamics of elections by taking gender and race into account is a step in the right direction. But to better understand the outcome of the 2016 election, analysts would be smart to look backward in order to better see forward.