Supporters of Donald Trump's presidential bid are regularly described in the media as financially and economically insecure. Reporters use adjectives such as "downscale" and "struggling" to describe them without a second thought — even though there is no obvious statistical connection between Trump's appeal  and voters' incomes.

A couple of writers have recently pointed out this apparent contradiction, suggesting that racial rather than economic anxiety explains the success of the presumptive Republican nominee. "To understand the patterns of support and opposition to Trump, you have to talk about race," wrote Matthew Yglesias of Vox. "Trump, in particular, built his big primary wins on the backs of people who are economically comfortable."

The data suggest there is some truth to both accounts. In fact, it might be impossible to completely disentangle racial and economic concerns among the presumptive Republican nominee's supporters.

Researchers have found evidence that racially conservative white Americans might interpret their own economic experiences more negatively if they see members of other groups advancing toward equality. At the same time, actual economic conditions can influence racial attitudes. A poll conducted earlier this year by The Washington Post and ABC News suggests that both factors have been crucial to Trump's success.

The Post-ABC poll

In the results of the poll, there was no correlation between incomes and support for Trump among Republican and GOP-inclined independent voters, 34 percent of whom supported the front-runner. That was no surprise, given the data from exit polls. Based on those findings, Five Thirty Eight's Nate Silver has estimated that the average household income among Trump supporters is $72,000 a year, well above the national median.

Yet when respondents were asked whether they felt economically "comfortable," or whether they felt that they were "struggling" or "moving up," there was a clear pattern in the responses.

Those who said that they were comfortable or moving up were much less likely to support Trump. Just 29 percent did, while 40 percent of those who said they were struggling were in Trump's camp.

Respondents were also asked whether they felt that it was a "bigger problem" that African Americans and Latinos were "losing out because of preference for whites" or that whites were losing out because of preferences for these groups. Among those who said that it was a bigger problem that whites were losing out, 43 percent supported Trump. 

Among those GOP-leaning voters who said that it was a bigger problem that the other groups were losing out because of preferences for whites or who said that it was a problem for both groups or for neither group, just 25 percent supported Trump.

This question was designed to measure the extent to which Trump's supporters resent the advances of other racial and ethnic groups. Research suggests that some white Americans tend to see the economy and society as a game in which only one group of people can win. If a different group is gaining in wealth or status, they seem to assume that their own group must be losing.

For example, one national survey found that, on average, white Americans think that as discrimination against African Americans has decreased, discrimination against their own group has increased — to the extent that they think there is now more anti-white discrimination than anti-black discrimination.

Recent experiments have found that when white subjects were presented with data appearing to show that white incomes were not increasing while incomes for other groups were, the subjects became more likely to support the tea party.

What other research has to say

In fact, incomes for all races have declined somewhat in recent years, with no one group gaining or losing relative to the others. Over the long term, however, Trump's supporters probably have lost some of their advantage in income, and the research suggests a few ways the trends in their incomes could be related to his popularity.

Trump's supporters are not just overwhelmingly white. They are also largely male. In the Post-ABC poll, nearly two out of every three people supporting  Trump were men. White male workers have always made more money relative to women and black workers. Yet typical white male incomes for workers who work full time year-round have been more or less unchanged since their peak in 1973, while incomes for women and black workers have continued to increase.

These numbers only include workers' incomes, and they don't account for the fact that more men have given up looking for work or are receiving disability insurance now than they were a few decades ago and that many more have gone to prison. It could be that these men who have become discouraged and disconnected from the economy could be disproportionately common in the cultural and social circles of Trump's supporters, giving them a justifiable sense that things really have gotten worse.

There is some evidence to suggest that economic distress can make white Americans more hostile toward people of color. In one experiment, subjects were shown a series of male faces digitally manipulated to appear ambiguously racial. The subjects were more likely to say the faces were black if they had first read words associated with scarcity of resources. In this sense, economic conditions could generate resentment toward minorities among Trump's supporters.

On the other hand, unconscious biases with regard to race and gender could generate a feeling that the economy is declining for them. White men have lost  some of their advantage in terms of income relative to other groups, with the result that some might feel that they are falling behind financially regardless of their actual circumstances.

The data on white men can help explain why Trump's supporters are so nostalgic for bygone times. A recent poll by the Pew Research Center found that three out of every four people supporting Trump said life in America was better "for people like them" 50 years ago. Among all voters, fewer than half agreed.

Also, different people might respond to economic trends in different ways. In an analysis, The Post determined that respondents' answers to the questions about race and whether they felt financially comfortable were independent in a statistical sense — meaning that many of Trump's supporters felt that they were struggling financially but bore no resentment toward minorities.

At the same time, many people who felt that white Americans were losing out because of preferences for others felt that they were comfortable financially, or that they were moving up.

If it is hard for people to form unbiased judgments about their own economic situations, their attitudes about race are even messier. Those realities make it difficult for psychologists and pollsters to confidently explain how racial prejudice and economic insecurity are connected to a phenomenon like Donald Trump. But it seems clear that both factors are important to the real estate mogul's appeal.