US businesses are becoming increasingly segregated along educational lines, with college and non-college educated employees isolated from one another at work, according to new research.
The divide has been so severe in recent decades that it’s comparable to the racial segregation of residential neighborhoods in many US cities. And this workplace isolation is particularly high for young male workers without college degrees, limiting them from interacting with colleagues of different educational backgrounds and gaining valuable contacts and skills that could boost their career prospects.
“At the end of the day, people spend most of their waking hours at work, more than they spend in social settings and way more than they ever spent in high school or college. The workplace is where they’re supposed to pick up new skills,” says William Kerr, coauthor of the study and the D’Arbeloff Professor of Business Administration at Harvard Business School. “But they can’t learn those skills if they’re so isolated from others. It’s a lost opportunity.”
The workplace is where they’re supposed to pick up new skills. But they can’t learn those skills if they’re so isolated from others.
The bottom line: Workers without college degrees are at risk of being trapped in lower-paying, isolated jobs that limit their career development, Kerr finds in the article “Workplace Segregation Between College and Non-college Workers.” Kerr partnered on the paper, published in the journal AEA Papers and Proceedings in May, with Francis M. Dillon, formerly a predoctoral fellow at Harvard Business School who's now a doctoral student at the Wharton School of the University of Pennsylvania, and Edward L. Glaeser, the Fred and Eleanor Glimp Professor of Economics at Harvard University.
In a sluggish economy, the study offers lessons for companies looking to do more with their current talent. It also sheds light on another dimension of division in a polarized society.
Diving deep into the data
While the fissure along educational lines isn’t new, Kerr and his coauthors attempted to dive deeper into the subject than ever before.
“We have previously never calculated, at least at a scale like this, the workplace segregation by education levels. This is something that a lot of people have been probing, but not to this extent,” he explains.
For their study, the authors compiled data from the Longitudinal Employer-Household Dynamics (LEHD) program at the US Census and from unemployment insurance records, stretching from 1998 to 2020. The data covered 95 percent of all private-sector workers in the 22 states that provided full access to records—or slightly less than half of the total US population.
The data was so specific that the authors could pinpoint the makeup of workforces at individual business sites or groupings of nearby facilities owned by a business.
The isolation index
After analyzing the available data, the authors found that:
Workers became more educated. The percentage of US workers without a college degree in the data fell from about 68 percent in 2000 to 62 percent in 2020.
Employees without college degrees continued to be isolated. In 2000, the average non-college employee worked in a business in which 79 percent of the workers lacked degrees, a high “isolation” level indicating extreme workplace segregation. In 2020, that number dropped to 77 percent, which was surprising considering the increase in the overall number of workers with college degrees.
The isolation index for workers without college degrees rose. Non-college workers were 10 percent more isolated in workplaces than they should have been in 2000 and 15 percent more isolated than in 2020, given rising education levels.
Industrial vs. service economy
While the authors didn’t set out to determine the specific causes of workplace segregation, they identified several potential factors, including discrimination and the corporate outsourcing of jobs that used to be considered in-house positions, such as janitorial and food services work.
A shift over decades from a strong industrial economy to more of a service-sector economy has also played a role. The recent growth of five sectors and their isolation levels—restaurants, warehouse/storage, general merchandise stores, grocery stores, and services to buildings and dwellings—accounted for 43 percent of the increase in educational workplace isolation from 2000 to 2020.
Manufacturing businesses tend to be more dynamic and complex than service-sector businesses, requiring a mix of college-educated executives, managers, and engineers working in relative proximity to non-college workers who are making actual goods, Kerr says. Multi-faceted manufacturing workplaces have been increasingly replaced by service-sector jobs with traditionally high workplace isolation rates.
Demographics of the shift
Male workers, and especially young male workers, appear to have been hardest hit by the overall shift from an industrial economy to more of a service-sector economy, Kerr says.
Preliminary data indicates workplace segregation along educational lines applies to whites, Blacks, and other ethnic groups. “It’s not exactly equal along racial lines, but it seems to be close,” Kerr says of the make-up of those employed within predominantly non-college workplaces.
In all, workplace segregation along educational lines is harming already hard-pressed non-college workers, Kerr explains.
“If you are less educated and start with fewer skills, but if you are around on a daily basis with people that have greater skills and higher education levels, you might be able to absorb and learn some of those competencies,” Kerr says. “Such interactions and skills development ultimately make a non-college worker more attractive to other employers, enabling career progression in the workplace. But that’s not happening if non-college workers are isolated in jobs without exposure to others.”
Image: HBSWK with asset from AdobeStock.
Have feedback for us?
Workplace Segregation Between College and Non-college Workers
Dillon, Francis, Edward L. Glaeser, and William Kerr. "Workplace Segregation Between College and Non-college Workers." AEA Papers and Proceedings 115 (May 2025): 139–145.
