In the midterm elections in 2018, the power distribution in the House of Representatives was dramatically shifted from Republican to Democratic control. Many expected this given the relatively low approval of President Trump. Historically, this meant that the presidential party would lose many home races. The elections before the election largely confirmed the likely takeover by the Democrats.
But what we did not know yet: Which groups supported the Democrats in this election? How do you compare these patterns with previous elections?
Below are five diagrams that explain what happened. These charts are largely based on data from the Cooperative Congressional Election Study (CCES), a large-scale academic survey conducted since 2008 in each election year. For the 2018 CCES analysis, we used selected interviews with weighted respondents to be nationally representative of the adult population. We then used a probable election model that was trained in previous election cycles to produce estimates for 2018 voters.
1. How did different age groups vote?
Let's first look at the voting patterns between different age groups in house races over the past ten years. While all age groups have been more democratic this year than in 2016, under-50's have shifted more. In particular, 18 to 29-year-old voters in 2018 elected the Democratic candidates by a margin of 2: 1 over the Republicans. And although we will not know it for some time, there are some indications that they have made a bigger impression on the electorate than on a typical midterm election – in other words, young people have voted for particularly high votes.
2. How did suburbs choose?
Suburban districts were among the most contested battlefields in this campaign. The CCES data shows that Democrats perform well in these districts. The following table shows the house vote among suburban people divided into US regions. Suburban voters supported the Democratic House candidates with a healthy lead over Republican candidates in all regions except the South, where the party failure was even balanced.
3. As expected, women and men voted very differently
Another pattern everyone observed was the gender gap. This, as the next chart shows, is the largest we've seen in at least a decade. While nearly 60 percent of women who voted for one of the two major parties voted for Democratic candidates, only 47 percent of men voted. This is a gender gap of 13 points.
4. Break up women and men by race and education
Of course, women and men are incredibly broad groups that consist of every US population. Which subgroups of women and men were the most distant from each other? The next chart shows the two-party vote share among white voters (since voters of color have gone mostly for democratic candidates, regardless of gender, sex and education), gender and education.
As you can see, white women without a university degree modestly moved towards the Democrats, more than white men without university degrees. But a far greater proportion of white college-educated women came to the Democrats, even beyond their previous support for Democratic candidates in previous elections. How much did they swing? In 2018, white women with college education increased their support for Democratic candidates by 8 percentage points compared to 2016. In previous cycles, this group accounted for about 15 percent of voters, so that large democratic margins among this population certainly contributed to success in 2018.
5. Why did the white women who were trained in the school migrate to the Democrats?
What explains the huge shift of white graduates? The last chart comes from an analysis I did for Data for Progress. In this article, I compare the role of the electorate to women in this election with the role it played in 2018.
Among other things, I've looked at what researchers call "hostile sexism," a set of antagonistic attitudes towards women based on the belief that women want to control men. While hostile sexism was a strong indicator of Trump's support, he had no influence on the election of people at their home races in 2016. That changed in 2018.
The graph below shows how a higher level of sexism is related to the election of the Republican candidate in 2018 and 2016 and influences other factors such as ideology, partisanship, racial behavior and demographics. In 2016, the approval or disagreement of a sexistist voter (found on the X-axis) was not relevant to the Republican candidate's election. However, in 2018, people who were more likely to disagree with sexist claims (ie, had less hostile sexism) were much less in favor of Republicans.
In essence, lesser-voted voters punished Republican House candidates in ways they did not do in 2016. In addition, the Republicans have won no sexist voters more to compensate for this loss.
Overall, these five charts suggest that Republicans may want to worry about being tied to a president whose rhetoric is often disagreeable and offensive. This means eliminating younger voters at historic rates and at the same time driving out women (especially those with a university degree). As the Republican Party brand becomes increasingly synonymous with Trump, these patterns may persist until 2020 and beyond.
Brian F. Schaffner (@b_schaffner) is Newhouse Professor of Citizenship at Tables College and the Faculty of Political Science at Tufts University.