Opposite effect

Ban the box: Why laws to stop hiring discrimination actually make it worse

Larry French/Getty

President Barack Obama introduced it on a federal level. The Department of Education put it into effect for higher education across America. More than 150 cities and counties did it, according to the National Employment Law Project.

Yet, a growing body of research suggests that “ban the box”—a popular progressive policy meant to help people with criminal records get jobs—should be nixed. Studies say it’s creating even wider racial hiring gaps, which advocacy groups like the National Association for the Advancement of Colored People (NAACP) have said “ban the box” would help close.

For the uninitiated, “banning the box” means that an employer won’t ask job applicants about their criminal records until later in the hiring process—the logic being that smart, hirable people are being turned away before they’re able to show off their qualifications. Discouraged applicants are more likely to reoffend, the thinking goes, since it’s harder for them to find a legitimate paycheck. And since black and Latinx people are caught in the criminal justice system at disproportionate rates, “ban the box” aims to cut unemployment among these groups.

Except, research suggests, it’s only making things worse.

“Ban the box” laws vary. Some apply only to governments, others to governments and major employers. Meanwhile, in Chicago, New Jersey, and New York City, they apply to most private employees within each jurisdiction. When the New Jersey and New York City laws were set to go into effect last year, researchers Sonja Starr, a professor at University of Michigan Law School, and Amanda Agan, an economics professor at Rutgers University, saw an opportunity to create a field experiment: How does “ban the box” actually affect people?

Starr and Agan created and submitted 15,000 fake online job applications to employers in New York City and New Jersey, both before and after the laws went into effect. They randomly assigned criminal histories to the applications, as well as information that could be interpreted as a sign of a record (e.g. GED or one-year gap in employment). Starr and Agan then assigned stereotypically “black-sounding names” (like Tyree, Terrell, and Daquan) and stereotypically “white-sounding names” (like Scott, Dylan, and Cody) to the applications.

Studies say they’re creating even wider racial hiring gaps.

Before “ban the box” was introduced, white applicants were getting 7% more callbacks from employers than similar black applicants. But after the policy went into effect, that number jumped to 45%, according to the field experiment’s findings.

“Ban the box came to bring about very great improvements for white applicants with criminal records because once the box is banned, employers seem to make the positive assumption that they don’t have a criminal record,” Starr told me over the phone.

“But for black applicants without criminal records, their callback rates went substantially down. And that seems to be because if they don’t get to tell the employer that they don’t have a criminal record, employers seem to make negative assumptions about them.”

Without information about an applicant’s criminal record, researchers hypothesize, employers are assuming that since black men are statistically more likely to have criminal records, they must have criminal records. What’s more, white men—even the ones who had large employment gaps and GEDs, the red flags meant to hint at a history of criminality—were still getting more callbacks for interviews. “Neither characteristic significantly affects callback rates [for white men],” according to the findings.

So, instead of discrimination based on criminal histories, “ban the box” encourages statistical discrimination based on the little information that employers have—and unfortunately, the latter paints with a much broader, racially tinged brush.

“A lot of jurisdictions have passed ‘ban the box’ as part of a racial equality measure,” Starr said. “But actually, you’re seeing a tradeoff with criminal records-based discrimination and race-based discrimination that I’m certain is not the consequence that policymakers were hoping for.”

In a separate working paper, Benjamin Hansen, an economics professor at University of Oregon, and Jennifer Doleac, a public policy and economics professor at University of Virginia, made similar conclusions. “These findings support the hypothesis that when an applicant’s criminal history is unavailable, employers statistically discriminate against demographic groups that include more ex-offenders,” Hansen and Doleac wrote in the latest version of their unpublished paper from this month.

For young, low-skilled black men and Latinx men, “ban the box” decreases their chances of getting hired by 5.1% and 2.9%, respectively, they found.

“For black applicants without criminal records, their callback rates went substantially down.”

- Sonja Starr, University of Michigan Law School professor

“The fact that both of these papers are lining up together is what makes for a really compelling narrative with respect to what’s happening with the policy,” Hansen told me, adding that Starr and Agan’s field experiment is “top-notch.”

For Hansen, it all fits into part of a larger narrative that suggests giving employers more information—and not less, as “ban the box” proposes—is actually good for job applicants of color. Contrary to what researchers thought would happen, he noted, when employers review credit scores during the hiring process, it actually helps black workers get hired, according to a Massachusetts Institute of Technology study from earlier this year. Why? Although black people have, on average, lower credit scores than white people, they’re freed of the stigma of having a perceived low credit score when this information is available at the individual level. In other words, when black job applicants can offer employers information that undermines a stereotype, they benefit. Researchers have found similar results for mandatory drug tests and in full background checks during the hiring process.

But that still leaves us with the problem of how to help people with criminal records who have a hard time finding legitimate work. “Employers, for whatever reason, have determined that a prior criminal history is an important signal to them,” Hansen said, adding that lawmakers are responsible for addressing this issue.

Several states, such as Tennessee, Georgia, and Texas, have in recent years passed laws that give liability protection to employers who hire applicants with past convictions. Some have even issued official “certificates of employability” to these people, helping assuage any concerns the employer might have. Hansen said that, anecdotally, he’s heard such laws are having a positive impact on hires.

Since Starr and Agan published their field experiment about the adverse effects of “ban the box” in June, Starr said she’s heard criticism from pro-“ban the box” groups; they argue that these findings are no reason to abandon the laws because racial discrimination is the real problem.

“I don’t necessarily disagree with that, but finding ways to address racial discrimination at that stage in the hiring process is really difficult,” she told me. “We’ve been trying to figure out how to enforce it for half a century.”

“Ban the box,” research shows, only exacerbates this problem. Some may see it as progress for job applicants with criminal records, but so far, science is calling it the equivalent of taking one step forward and two steps back.