How banning boxes encourages discrimination

Each year, more than 637,000 people are released from jail. These newly released ex-convicts often struggle to reintegrate into society. High recidivism demonstrates that many fail: about two-thirds of ex-convicts are re-arrested within 3 years of release. One of the major barriers to reintegration is the inability of ex-convicts to find a job. To improve hiring outcomes, 24 states and the federal government have passed “ban-the-box” policies. These policies generally prevent employers from asking about a candidate’s criminal history at the earliest stages of the application process. A recent study finds that, even as these policies improve outcomes for ex-convicts, they may worsen outcomes for minorities—particularly low-skilled Black and Hispanic men.

In “Does ‘ban the box’ help or hurt low-skilled workers? Statistical discrimination and employment outcomes when criminal histories are hidden” (National Bureau of Economic Research, working paper 22469, July 2016), Jennifer Doleac and Benjamin Hansen examine whether ban-the-box policies are a net positive or net negative for racial minorities. The authors suggest that, even in jurisdictions that ban the box, employers are biased against hiring ex-convicts. So when denied explicit information on conviction status, employers may use other characteristics—including age, skill level, race, and gender—to “guess” which candidates are ex-convicts and deny those individuals employment interviews. For employers who engage in this guessing game, young, low-skilled, Black and Hispanic males are the least attractive candidates. If enough employers follow this practice, ban-the-box policies could increase racial disparities.

To determine whether ban-the-box policies had the effect of increasing discrimination, Doleac and Hansen examined data for 855,772 men ages 25 to 34, about 59 percent of whom were low skilled (i.e., no college degree) and 46 percent of whom lived in ban-the-box jurisdictions. They calculated the employment probability for these men on the basis of their race/ethnicity and geographic location, then compared that probability with the actual percentage of men from each group who were employed, according to U.S. Bureau of Labor Statistics Current Population Survey data. According to the researchers’ findings, ban-the-box policies do disproportionately affect Black and Hispanic workers.

Doleac and Hansen found that ban-the-box policies reduce the probability of employment for young Black men without a college degree by 5.1 percent. The effect on young Hispanic men was less but still statistically significant at 2.9 percent. These figures confirmed the authors’ initial hypotheses as well as their broader assumption that “well-intentioned policies that remove information about negative characteristics can do more harm than good.” In support of this latter statement, the authors point to other studies showing that allowing employers to drug test and check credit histories makes them more likely to hire Black workers. These studies suggest that in the absence of such information, employers use race as a proxy and assume drug use and poor credit for Black candidates.

Despite the somewhat disheartening confirmation of racial bias in the hiring process, the study results have important policy implications, particularly for advocates hoping ban-the-box policies will help racial minorities, who are disproportionately arrested and convicted. If nothing else, the findings give policymakers more information when judging the utility of such laws.