With Brexit and the election of Donald Trump, there’s a widespread belief that populism is on the rise in the developed world. Writers and thinkers darkly warn of a crisis if elites don’t accede to the demands — explicit or assumed — of the working class.

It’s very hard to define whom to consider part of the elite. That makes it difficult to establish a target for popular anger, and it means that no one knows who, exactly, is expected to respond to the masses’ demands. But there’s another, related problem I didn’t talk about: Who is making the demands? Who are these working-class people who have been wronged by the system and aren’t going to take it anymore?

This is a crucial question, because it determines what policy responses might address general discontent. But, like the elite, the working class is devilishly difficult to define.

When Karl Marx created a social-class taxonomy for his 19th-century industrial European society, it was a simpler world. A large number of people were either farmers or poor urban factory workers. When Marx classified people as peasantry, proletariat, bourgeoisie, petit bourgeoisie and so on, he didn’t have to distinguish between income and occupation — the latter basically determined the former.

But though we still use many of Marx’s terms, the modern economy is a lot more complex than the one Marx dissected. Income, occupation, education and race create multiple fracture lines that call into question the existence of a unified proletariat or working class.

The easiest way of defining the working class is by income. Income inequality has increased substantially across the developed world, and in the U.S. more than most. One of the tools used to measure income inequality — the Gini coefficient, which is higher when the distribution of income is less equal — has risen quite a bit in the U.S. since the 1970s.

advertisement | advertise on newsday

If we define working class as anyone in the lower regions of this increasingly unequal income distribution, it seems clear why they’d be angry. As recent research by economists Fatih Guvenen, Greg Kaplan, Jae Song, and Justin Weidner shows, lifetime income for most American men has been declining for decades; only by sending women into the formal labor force en masse have most American families managed to improve their material situation. Other research shows that economic mobility and opportunity are declining as well — most Americans are making less than their parents did, and those in the lower ends of the distribution tend to be stuck there.

Rising inequality, stagnating income and reduced mobility seem like a toxic brew. And in the U. K., low income did indeed predict a vote for Britain’s exit from the European Union. But interestingly, the people suffering most from these trends in the U.S. don’t seem to be joining the supposedly populist revolt represented by Trump. Lower-income voters broke for Hillary Clinton in 2016, not Trump. More recent surveys also show that, all else equal, economic anxiety tended to push voters — including white voters — into the Clinton camp.

That suggests that there are other ways to think about class in the U.S. The most obvious alternative definition is education. Many polls and surveys find that the college/non-college distinction played a major role in determining who voted for Trump.

Looking at a variety of other indicators, it makes sense to think of a college degree as the essential marker of class in today’s U.S. A college degree is an increasingly strong predictor of who gets married, who stays married, who is politically and socially engaged, and even who goes to church. Research also shows that especially among white people, college education is an increasingly important predictor of mortality. This is why the term “white working class” is often used as a short-hand not for poor whites, but for those who didn’t attend college.

Another class demarcation could be occupation. Much of the populist anger in the U.S. seems to center on the decline of manufacturing employment. Meanwhile, economists have found that routine occupations are disappearing. If the working class is defined by the kind of work done, then the devastation of manufacturing jobs certainly seems like a reason for this group to be angry. What is the working class without work?

So there are multiple ways of defining the working class, and each one leads to different policy implications. Should elites assuage the anger of the working class with income redistribution, social insurance and other policies to help the poor? Should they try to make college degrees less important for finding work? Should they try to bring back manufacturing employment and/or routine manual tasks? And are these answers different in the U.K. than in the U.S.?

It’s easy to say that elites — whoever they are — should just push for all of the above, but that dodges the question of what the priorities need to be. If the U.S. and the U.K. are really in the midst of dangerous populist revolts, there can be no delay. There are also tradeoffs involved — for example, boosting manufacturing may require diverting spending from social programs that help the poor.

So as long as confusion remains about who comprises the working class, expect the response to populist anger in the U.S. and U.K. to be chaotic and disorganized.

advertisement | advertise on newsday

Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University.