Nearly two years after the 2016 election, the parties continue to fight about race and gender. In Brett Kavanaugh’s Supreme Court confirmation process, Democrats felt that allegations of sexual assault had been brushed aside while Republicans felt that accusations were being unfairly used to further a political agenda. In states’ voter roll purges, Democrats saw suppression aimed at black voters while Republicans saw the racialization of a purely political dispute. How is it that the two parties see issues of race and gender so differently?
A huge body of research has shown that voters were more divided by race and gender views in the 2016 election than they were in previous elections. But it turns out that rather than voters supporting the party that best represents their views about race and gender, the effect may more often work the other way — the parties may be shaping voters’ personal beliefs. Candidates and elected officials can drive a person to change their views, or loyalty to a party may dictate both a person’s beliefs and their candidate preferences.1
Several new papers use panel studies, which map changes in opinion by re-interviewing the same voters repeatedly over time, to see which came first: Did voters select their party or candidate based on their prior race and gender views, or did voters update those views after choosing their party or candidate?2 One working paper used data from seven different panel surveys to examine the attitudes that a nationally representative sample of white Americans held about race and partisanship. It found that during the last three presidential cycles, there was more evidence that white voters changed their views of racial discrimination to match their political party than that those voters changed their partisanship to match their views. The 2016 campaign in particular produced large swings in racial attitudes that moved to match voters’ partisanship, especially among voters who paid close attention to politics. Donald Trump’s incendiary remarks — and Hillary Clinton’s efforts to highlight them as disqualifying — probably moved more Democrats to perceive discrimination as an important problem that needs to be addressed (with some Republicans adopting the opposite view).
The cause-and-effect relationship is less clear when survey participants were asked directly about their general feelings toward various racial groups. The same seven-survey study examined how warmly participants felt toward white people and racial minorities. White Americans who felt more warmly toward white people than black people in 2016 became more Republican (though partisanship also led some to change their views of the two racial groups). Another longer-term panel study by political scientist and FiveThirtyEight contributor Daniel Hopkins followed the same white voters over nine years and found that prior anti-black sentiment predicted voting for Trump.
But voters’ views of specific racial issues, rather than their broader feelings about minority groups, were more likely to follow their candidate preferences. A study by Peter Enns at Cornell University found that Trump and Clinton voters changed their views on controversies like the Black Lives Matter movement to match their candidate’s views, rather than choosing their candidate based on their views about this issue.
We have less evidence on how gender attitudes changed, but voters may also have shifted their attitudes toward women during 2016 to match their candidate’s stance. Survey experiments and a panel study show that Trump’s comments about women made his supporters more willing to both tolerate and express sexist views — changes that were still evident in 2018. Meanwhile, Democrats’ belief that gender discrimination is a big problem has increased since Trump’s election.
Although media attention has largely focused on Trump voters, a working paper showed that it was actually Clinton voters who underwent the more dramatic partisan shift in 2016 (echoing other findings). The largest changes in views of race and gender occurred among white liberals; their perceptions of racial and gender discrimination increased, their feelings toward minorities improved, and their support of policies aimed at increasing diversity, like affirmative action and allowing more immigration, rose. Voters who consistently voted Democratic moved to the left on these questions, especially young voters. That means studies that show an increased association between Trump support and conservative views on race and gender might in part actually reflect Democrats becoming more liberal on these questions.
Thus, the darkest portrayals of Trump riding a growing wave of racism and sexism to victory may be somewhat overstated. Research on the 2016 election does not support the idea that Trump benefited from an overall increase in racist and sexist attitudes among white voters; rather, the evidence shows that liberal-leaning voters moved away from his views faster than conservatives moved toward them.
Americans are now even more divided on questions about the extent of discrimination in the country and the benefits or drawbacks of counteracting it. Since Trump took office, his administration has sparked regular controversies about immigration, sexual assault and racial intolerance. What’s more, experiments show that as president, Trump is able to change Republicans’ views on a variety of issues, suggesting that divisions may continue to grow. The panel studies I’ve reviewed, along with other studies, found evidence that partisanship and race and gender divisions can reinforce one another over time.
As the 2018 midterm election approaches, Republicans and Democrats can now be counted on to adopt their party leaders’ views on issues like the Kavanaugh nomination and voter roll purges. The Trump administration, liberals’ resistance to it, and the 2018 campaign cycle give us no reason to believe that the divisions in our society will narrow anytime soon.