Class Attitudes in the 2018 Midterms

By Jon Green (@_Jon_Green) and Spencer Piston (@SpencerPiston)

The class reputations of the two major parties, create, on balance, a reliable advantage for the Democrats. Over the past few decades, political scientists have asked the public what they like and dislike about the Democratic Party and the Republican Party. This research has routinely found that the Democratic Party is perceived to be more likely than Republicans to help the poor, the working class, and the middle class. The opposite is true of the Republican Party, which is negatively associated with assisting the rich.

Consistent with these findings, Class Attitudes in America, recently published by one of us (Spencer), shows that substantial proportions of the American public view poor people with sympathy and rich people with resentment. In recent elections, these attitudes have been associated with voting behavior; sympathy toward the poor and resentment toward the rich both increased Barack Obama’s vote share in 2008 and 2012. But in the 2016 election, class attitudes diminished in importance and did not explain variation in major party vote choice, controlling for other factors.

There are a variety of explanations why Hillary Clinton did not benefit from these class attitudes to the same extent that Obama did: In 2008 and 2012, Obama had gone out of his way to paint both of his opponents, McCain and Romney, as out-of-touch millionaires likely to support policies that would transfer resources to the wealthy. Clinton, however, did not pursue this strategy in 2016, instead focusing her criticisms on Donald Trump’s character. Trump muddied the waters further by claiming on multiple occasions that rich people are unfairly advantaged in America and that only a truly rich person, such as himself, could take on the entrenched interest groups who had created these conditions.

However, Trump’s actions as president quickly undercut whatever notions political observers may have had that he would put any sort of dent in the wealthiest Americans’ bank accounts. The most consequential piece of legislation he signed into law during his first two years entailed a massive upward redistribution of income that was never popular with the public. And while the Russia investigation and Donald Trump’s personal absurdities may have dominated midterms discussion on Twitter, by all accounts Democratic congressional candidates remained strikingly disciplined in keeping their focus on this and other unpopular, regressive Republican policy positions such as attempting to repeal the Affordable Care Act.

Given the class dynamics at play in the 2018 midterms, we included a two-item battery on Data for Progress’s What the Hell Happened survey that featured in the work cited above:

For each of the following groups, please say whether most people in the group have more money than they deserve, less money than they deserve, or about the right amount of money:

  • Poor people

  • Rich people

Response options ranged from 1 (A lot more money than they deserve) to 7 (A lot less money than they deserve).

Responses to these items correlate strongly with one another, which could itself be an indication that economic inequality was a salient issue in 2018. Respondents who said that one group had about the right amount of money tended to say that the other group did as well, while respondents who said that the rich have more money than they deserve also tended to say that the poor have less. Very few respondents said that the poor have more money than they deserve or that the rich have less than they deserve.

To test class attitudes’ possible role in the 2018 midterms, we modeled the relationship between class attitudes and voting for a Democratic House candidate in 2018, controlling for other factors such as political identity (partisanship and ideology), demographics (age, gender, income, college education, and status as a born-again Christian), and group attitudes (social dominance orientation, racial resentment, sexism, and nativism) among white respondents. Due to correlations between some of these independent variables, each model includes either sympathy toward the poor or resentment of the rich as the key explanatory variable, and includes one of the group attitudes controls.

The below plot shows results from each of the models that used nativism as the group attitudes control (for more on nativism in 2016, click here). This result is shown because, of the group attitudes controls, nativism was the strongest independent predictor of House vote choice – though each was, independently, statistically significant. Importantly, the inclusion of different group attitudes controls does not meaningfully affect the relationship between class attitudes and vote choice.

For each class attitudes item, we generated predicted probabilities of reporting a Democratic House vote given each response to the class attitudes item of interest, holding all other independent variables at their racial group median. Triangles represent the results for the model based on sympathy for the poor; diamonds represent results for the model based on resentment toward the rich; each have bootstrapped 95% prediction intervals. Bars in the background of the plot reflect the distribution of responses to each of the class attitudes items among whites – lighter bars for sympathy toward the poor; darker bars for resentment toward the rich.


The results clearly show that class attitudes were meaningfully associated with major party preference in the 2018 midterms. Otherwise-equal respondents with more sympathy toward the poor or resentment toward the rich were significantly more likely to favor Democratic candidates, while respondents who prefer the status quo income distribution were more likely to prefer Republicans.

These findings could prove instructive to Democratic candidates heading into the 2020 elections. The ample evidence that Republicans in Congress and the Trump Administration prefer an increasingly unequal distribution of income appears to have benefitted Democrats this year. Outside of the 2016 Election, class inequality has been, and remains, a winning issue for Democrats who are perceived to be looking out for those with less income. Looking ahead to 2020, it seems reasonable to expect that Democrats who promise to use government to help those who need it most, rather than take money out of their pockets to give it to millionaires, will have an easier time getting elected. Those who deliver on that promise will have an easier time staying elected.

Jon Green (@_Jon_Green) is a Ph.D. candidate in political science at Ohio State and co-founder of Data for Progress.

Spencer Piston (@SpencerPiston) is an assistant professor of Political Science. His scholarship examines racial and economic inequality by analyzing the influence of attitudes about social groups on public opinion and political behavior.

Ethan Winter