When I was in college, I read a paper written by Amanda Agan and Sojna Starr titled “Ban the Box, Criminal Records, and Racial Discrimination: A Field Experiment.” This was the article that convinced me to pursue a career in economic analysis.
This paper found that when employers were not allowed to ask potential job applicants whether or not they had a criminal record (job applications are not allowed to have the namesake box that people with criminal records need to check), they began making racially-biased assumptions. This led to fewer applicants of color receiving offers for interviews. This is an oversimplification, but I encourage anyone who is interested to read the full paper. It really is worth your time.
What I like most about this paper is that the authors do a good job of explaining that while “ban the box” policies have unintended racial bias problems, they are not strictly bad policies. They exist to address a major problem: that people with criminal records have a substantially more difficult time finding employment than people who don’t.
We know that people with criminal records have a hard time entering the labor force. What happens if these people are able to get past the initial barrier? Recently, I came across a new working paper that helps answer this question.
This paper used comprehensive Swedish register data from 1990 to 2015 to examine the labor market effects of a criminal record. Sweden, unlike the United States, has historically limited public access to criminal records, potentially lessening the likelihood that employers would toss out applicants due to their history.
The study looked at workers before and after they were charged with a crime and matched them with similar individuals suspected of a crime but not charged in order to uncover a casual relationship between criminal records and labor market outcomes. The authors found that a first-time criminal charge led to a 5% decrease in earnings and a 2% drop in the number of months worked each year.
These effects lasted for over a decade, even after the record was expunged. Interestingly, most of the decline wasn’t because people lost their jobs or went to jail. It was because they ended up working at lower-paying firms, often earning less even within those firms compared to other employees.
They also found a lot of variation in how willing companies are to hire people with criminal records. Small firms, firms that didn’t advertise background checks, and firms led by managers who had personal experience with the criminal justice system were more likely to hire these workers.
Some firms during this period changed managers, switching from people who were less likely to hire employees with criminal backgrounds to managers that were more likely to hire these people. Importantly, this shift didn’t hurt the firm’s performance, meaning the lower pay was not necessarily reflecting a lower quality of work.
If it is true that workers with criminal records are not less efficient (which would push back against some commonly held knowledge about human capital loss during incarceration) then there may be a market failure that public policy could correct.
Consider this scenario: An individual gets released from prison, and has to decide whether they will enter the labor market or return to criminal activity. In order to choose the labor market, they would need to earn the fair market wage for the work they could provide, say $25,000 per year.
Because of their criminal record, they can only earn $20,000 per year, so they choose to return to crime. If the public could subsidize the difference there then they could keep this particular individual attached to the labor force and contributing to the broader economy.
Both of these studies highlight the tough reality that people with criminal records face when trying to reenter the workforce. They also show why it’s so important to approach policies around criminal records and hiring carefully.