Last week, the leading Democratic contenders for president came together on one debate stage for the first time. While they debated health care, education, gun violence and immigration, no one mentioned economic inequality. That’s unlike the previous debates, during which presidential hopefuls decried rampant inequalities in wealth and income.

Perhaps the candidates are concerned that inequality is not a promising platform for a presidential bid. Research by political scientists Kenneth Scheve and David Stasavage supports this fear: U.S. voters are not easily alarmed by growing disparities between the rich and the poor, the 1 percent and the rest.

However, Americans seem to feel uneasy about income gaps that grow from circumstances — such as gender, race or the family one is born into — that are hard or impossible to change. In a meritocratic society, these are not supposed to affect your ability to rise. In future debates, presidential hopefuls might wish to start a conversation around such income gaps.

Why income gaps matter

Although the United States regularly has political debates about inequality, in the past decades the country has launched few effective political initiatives to reduce it. One reason might be that debates focus on differences between the rich and the poor. Economists Ilyana Kuziemko, Michael I. Norton, Emmanuel Saez and Stefanie Stantcheva have found that giving Americans information about such economic differences has little effect on their political preferences. When confronted with these differences, Americans find ways to justify them, political scientist Kris-Stella Trump has shown.

But learning about other aspects of inequality does influence Americans’ political points of view. Stantcheva and economists Alberto Alesina and Edoardo Teso found that Americans, at least on the political left, respond to information about intergenerational mobility, which indicates how strongly economic fortunes depend on one’s family background — presumably because it contradicts U.S. ideals about economic fairness and equality of opportunity, or the American Dream.

These American ideals suggest that other circumstances beyond individual control should also not affect one’s prospects. The latest census numbers reveal factors such as one’s gender or race continue to manifest in income gaps.

How I did my research

In a recently published article, I explored what Americans know and think about income gaps that reflect circumstances beyond an individual’s control. The article is based on an online survey, conducted in 2016 and 2017, with respondents recruited over the Amazon platform MTurk. (Amazon chief executive and founder Jeff Bezos owns The Washington Post.) Over 400 people participated. While they represent a broad range of demographics, the study is not nationally representative.

Respondents were asked to guess the size of these circumstance-based income gaps in the U.S. labor market — between men and women, whites and nonwhites, and Americans with university-educated parents and those without. They could indicate their guesses on three different sliders. In the same interface, half the participants were then shown the actual size of the gaps. Afterward, both groups were asked about their support for government redistribution to reduce inequality in general.

Most participants strongly underestimated the income gaps. This was most pronounced for the gender gap, with over 98 percent guessing something lower than the actual gap. Based on 2010 data, the actual gender gap in annual income, including full- and part-time employees and before tax, was about $27,300. Men on average suspected the gender income gap to be smaller than women did, $8,300 versus $10,000.

The racial income gap in annual income was $17,800. White Americans on average guessed it to be $10,500, while nonwhites’ average guess was $13,800. Participants also underestimated the income gap between individuals with university-educated parents and those without; the figure is $18,700, while the average guess was $14,100.

When asked about government redistribution measures such as increasing taxes, participants who guessed the gaps to be larger were more likely to agree. And after being informed about their actual size, those who underestimated the gaps were most likely to increase their support for such measures, by up to 35 percentage points. Those effects varied by subgroup, with men and respondents without university-educated parents being least inclined to increase their support.

A year later, I successfully followed up by email with more than a third of the respondents — and found that the information left a lasting impression. Those who had been told about the actual size of the gaps were still more accurate in their guesses, and — if they had underestimated the gaps previously — were still more likely to support government redistributive measures.

What Democratic presidential hopefuls can learn from this

Americans dislike income gaps that reflect circumstances beyond individual control. And yet they underestimate them dramatically. That suggests that focusing on such gaps may be useful for Democratic presidential candidates. Further, discussing such gaps invites a broader conversation about structural inequality.

Gaps exist beyond the labor market. In last week’s debate, the audience applauded former congressman Beto O’Rourke for pointing out that, “in Texas, a 5-year-old child in kindergarten is five times as likely to be disciplined or suspended or expelled based on the color of their skin. In our health-care system … there’s a maternal mortality crisis three times as deadly for women of color, [and] there’s 10 times the wealth in white America than there is in black America.”

My work suggests that such statements about income gaps that grow from characteristics beyond individuals’ control affect how Americans form political preferences, more so than statements about generic inequality. Of course, not everyone responds to such information in the same way; mentioning these gaps might even provoke a cultural backlash, as political scientists Pippa Norris and Ronald Inglehart argue. There’s still a lot to be learned about how the way we talk about inequality affects us and how politicians can discuss inequality effectively.

Bastian Becker (@beckerbastian) is a postdoctoral researcher at the University of Bremen’s Research Center on Inequality and Social Policy in Germany.