The wage gap between white and Black men in the United States has persisted for generations, but there are reasons to think it would be smaller today than it was 40 years ago. Legalized forms of racism that allowed employers to bar Black Americans from jobs, or to pay them less for equal work, have been banned for a half century. Indicators of bias—such as white Americans’ attitudes about whether they would vote for a Black president or whether they would support interracial marriage—have improved over time, suggesting that discriminatory preferences in some domains have diminished. And although Black men still lag white men on accumulated years of schooling, the education gap has narrowed.
Yet gains in the relative wages of Black men have stalled in recent decades. By 1980, average earnings among employed Black males, measured after accounting for differences in their years of schooling, were about 80 percent of those of working white men, up from 60 percent in the 1940s. Then the progress stalled. Black men, both employed and as a whole, have lost ground since.
Chicago Booth’s Erik Hurst, Yona Rubinstein of the London School of Economics, and MIT PhD student Kazuatsu Shimizu have been working to understand the reason for this. Their research suggests that Black men face high barriers of entry to many of the best-paying jobs, and the returns to these jobs have increased since 1980. A solution, they suggest, may be to equalize opportunities in early education.
The researchers’ argument is based on a concept that they call task-based discrimination, a nod to work by the late Gary S. Becker. In a book published in 1957, which framed discrimination in a market context, Becker argued that some prejudiced employers prefer hiring white workers rather than Black ones. He presented a model that has been known since as taste-based discrimination—as in, some employers have a “taste for discrimination” and as a result pay their Black workers less. A competing economic theory, statistical discrimination, argues that people with imperfect information end up making decisions using observable characteristics, often relying on stereotypes to make up for what they don’t know.
Neither model completely explains today’s situation, however, and Hurst, Rubinstein, and Shimizu put forward a framework that is informed by the tasks required by the modern economy. Its core idea is that every occupation involves different tasks, which in turn require certain skills to perform them. The intensity of discrimination varies depending on the mix of tasks involved.
For a half century, the Department of Labor has conducted surveys that map jobs to the tasks required, and more recently, economists including MIT’s David H. Autor and University of Zurich’s David Dorn categorized occupations by the extent to which they require three types of tasks: abstract, routine, or manual. Every occupation involves at least one of these, and sometimes all three, in varying amounts. For example, CEOs, computer programmers, engineers, and lawyers all do work that involves a higher proportion of abstract tasks.
Hurst, Rubinstein, and Shimizu added a fourth category, contact, to refer to tasks that involve a lot of personal interaction. Medical professionals, restaurant workers, sales clerks, and teachers all do jobs that involve many contact tasks. “If we’re going to look for a place where discrimination might be more important, we imagine that this might be more salient,” says Hurst.
Using data from the US Census Bureau, the researchers document that in 1960, Black men were less likely to sort into (read: end up in) jobs requiring either abstract or contact tasks. However, after 1960, the racial gap with respect to sorting into contact-heavy jobs narrowed substantially, while the gap in abstract-heavy jobs remained constant. As the economy changed, molded by forces including international trade and automation, Black men were left out of those jobs—and are now no more likely than they were 60 years ago to do work involving many abstract tasks.
That’s bad for earnings equity because wages for these tasks have risen faster, putting Black men at a disadvantage. The rising return to abstract tasks increased the racial wage gap by 7 percentage points between 1980 and 2018, the researchers calculate. If the wages paid for different tasks had instead remained equal, the racial wage gap would have shrunk instead of stagnating.
The force of those rising earnings in jobs such as information technology canceled out gains made in other areas. “Those two effects have offset each other from 1980 onward,” says Hurst. The racial gap in some skills has narrowed, and anti-Black discrimination has abated some, but the underrepresentation in jobs with rising value “has been a force pushing against their wages.”
What’s behind this persistent underrepresentation in abstract jobs? In short, the research argues, unequal opportunities early in life. The researchers looked to some tests and surveys that, while imperfect, provide information about how Black and white men measured up in terms of skills when they were young adults. Doing this, they find that barriers developed early on that made it hard for many Black men to reach high-paying jobs.
One long-running government survey allowed the researchers to compare the skill levels of Black and white teenagers in 1979 and 1997—and then to look at what kinds of jobs the teens went on to do when they were in their 30s. There was a clear path for young people with social skills into contact-heavy jobs, regardless of their race. Black and white teens showed around the same levels of social skills, and those levels strongly predicted whether the teens would go on to do contact jobs. The end result is that the teens who started off with essentially the same social skills went on to do essentially the same jobs.
Any difference in sorting into contact jobs, the researchers conclude, stemmed largely from hiring discrimination. And when the researchers analyzed data from government surveys that have regularly asked Americans questions such as whether they’d vote for a Black president, exploiting variation across states, they find that this racial gap in sorting was highly correlated with survey-based measures of discriminatory attitudes. While the racial gap in contact tasks narrowed between 1960 and 2018, discriminatory attitudes improved. “There is no doubt that taste-based discrimination is still a feature of the US economy today. It is just that the level of discrimination today is smaller than it was 60 years ago,” says Hurst.
But the story was different for abstract-heavy jobs. According to the measures the researchers looked at, Black teens lacked certain skills that projected into the likelihood of becoming CEOs, engineers, judges, or software developers. This skills deficiency is almost certainly the result of current and past discrimination, Hurst says, and it means that the economy has penalized Black men in two ways—first through discrimination in hiring, and second through barriers to the acquisition of increasingly valuable skills.
Given that the return to abstract tasks is rising, it is more important than ever to reduce early life barriers that are preventing Black men from getting jobs that require doing abstract tasks, the researchers write. Applicants need to have certain skills to qualify for those jobs, and education is key. “It is becoming even more important today to equalize opportunities in early childhood to close the racial Abstract skill gap given that the return to Abstract skills has been rising over time,” Hurst, Rubinstein, and Shimizu emphasize. With discrimination levels having fallen, this skills gap needs to be surmounted for progress on the wage gap to resume.
(Courtesy Chicago Booth Review/By Dee Gill and Emily Lambert)
