In Thessaloniki, Greece, a man salvages an olive oil container from a trash bin to collect the remaining oil. Welfare agencies and charity groups have warned of a spike in poverty and homelessness in Greece due to the effects of the financial crisis. (Nikolas Giakoumidis/AP)

Economic inequality is on the rise and in the headlines. Thomas Piketty’s scholarly treatise on the subject tops bestseller lists. From Hillary Clinton to David Cameron, political candidates — in the United States and elsewhere — feel compelled to promise remedies.

Since the time of ancient Greece, political theorists and observers have feared that inequality leads to instability. The greater the income gap, the more the poor have to gain by taking from the rich. In democracies, the thinking goes, inequality should predispose voters to demand government redistribution. In dictatorships, the rich, fearing Robin Hood policies, should resist democratization. And the poor, locked out of power and wealth, should be more tempted by revolution.

Though these arguments have been around since Aristotle, it’s hard to find evidence for them in the real world. One recent political science journal article reports a “current consensus… that inequality does not matter for the politics of redistribution, at least not in any direct and particularly significant way.” Another notes that half a century of research has produced “little empirical support” for the notion that inequality prompts political violence. And one sophisticated study finds “no evidence that domestic inequality is related to regime outcomes” such as democratic revolutions.

Why? Scholars have suggested a variety of things that might derail political unrest. Belief that the economic system is fair, or the hope of being rich someday, or even organized religion might reconcile people to the gap between rich and poor. Or it could be that, with their assets hidden in Swiss bank accounts, the rich these days have just become too hard to expropriate.

But there’s a simpler possibility: Maybe inequality fails to trigger the expected political consequences because most people just don’t know how large the gap is between the wealthy and the rest of us. That’s the explanation we suggest in a recent paper. If people don’t know how much they stand to gain and at what cost, why would they take political action?

We looked at eight cross-national surveys to see what people believe about inequality. Time and again, large numbers of respondents had no clue what the income distribution looked like in their country, how it had been changing recently, and where in that distribution they personally fit.

For instance, in 2009, the International Social Survey Program (ISSP) asked people in 40 countries which of a set of five diagrams best described their own society. The diagrams ranged from a candlestick shape (broad base beneath a thin, tapering neck) to an upside-down pyramid. We compared respondents’ answers to various measures of their countries’ actual inequality levels. If people chose randomly among the five diagrams, they would be “right” about 20 percent of the time. Worldwide, 29 percent of respondents chose the right diagram — which is little better than random. Mostly they guessed wrong. And in the United States, almost a quarter would not even guess.

Respondents also had little idea how rich was rich. When asked what a typical chairman of a large national company was paid, respondents in six developed democracies gave figures averaging between 18 percent (Sweden) and 82 percent (the United States) of the actual level, as best we could determine. In South Africa, respondents thought top businessmen earned about $77,000 a year — when, in fact, their average pay was $1.7 million, according to a study of 56 major South African corporations.

Asked where they personally fit in the national income distribution, people tend to place themselves too close to the middle. In the 2010 “Life in Transition” survey of 29 post-communist and five developed European countries, 40 percent of respondents placed themselves in the central two deciles (which, by definition, should contain just 20 percent). Fewer than 1 percent put themselves in the top 10th, and fewer than 5 percent in the bottom one.

Oddly enough, many who were almost certainly among the richest seemed to think they were quite poor. Among respondents whose families owned a second home — putting them in the top 9 percent in the average country — 60 percent thought they belonged in the bottom half of the income distribution. Just as oddly, respondents who were almost certainly among the poorest thought they ranked higher up. In the World Values Survey (taken in 2010-14), 3 percent of respondents in Germany and 6 percent in Spain said they had “often” or “sometimes” had to go without enough food in the preceding year. Yet among that bottom 5 or 6 percent who were hungry, few thought they were in the poorest 10 percent nationwide, and more than one quarter placed themselves among the richest 60 percent.

Did respondents know how inequality had been changing? In 2013, the Pew Global Attitudes Project asked whether the “gap between the rich and the poor” had increased, decreased, or stayed the same in the preceding five years. The factual answer is that inequality had increased in six, decreased in eight, and stayed roughly the same in eight. But respondents proved generally clueless about their country’s direction of change. On average, about 35 percent got the answer right — only slightly more than would have done so if picking an answer randomly. In only six of the 22 countries — China, France, Greece, Indonesia, Spain, and the United States — did a majority of respondents get the direction correct.

Given this confusion about the facts, it’s hardly surprising that whether people favored government redistribution had nothing to do with their countries’ actual levels of inequality.

Rather, what mattered was how much inequality people believed existed, as we show below. Perceived inequality closely tracks the demand for redistribution. And it also correlates strongly with respondents’ reports of serious social tensions between rich and poor — more strongly than such reported tensions correlate with actual inequality.

In short, what matters is not the objective reality of inequality, but people’s beliefs about inequality. And those beliefs are often wrong. Although we don’t know what causes such beliefs, here are a few hypotheses. People may extrapolate from their immediate reference group, assuming that inequality nationwide mirrors what they see in their local or professional communities. The media and travel must also shape individuals’ impressions. Ideology may predispose people to “see” inequality inaccurately — socialists may assume capital is ever more concentrated, while free market liberals may suppose wealth is trickling down. And other psychological biases could distort perceptions.

Alongside more egalitarian Europeans, Americans still seem relatively relaxed about income inequality. But that might be changing. The percentage of U.S. respondents who say that the “growing gap between the rich and poor” is the greatest threat to the world — more dangerous even than nuclear weapons, infectious diseases, and religious hatred — jumped from 14 percent in 2002 to 27 percent last year. And if they knew how high inequality actually was…

Vladimir Gimpelson is a professor of economics and director of the Center for Labor Market Studies at the Higher School of Economics in Moscow. Daniel Treisman is a professor of political science at the University of California, Los Angeles.