MANY AMERICAN EMPLOYERS include questions about past criminal convictions in their application materials, or require background checks as part of the employment process. But as rates of incarceration in the U.S. have risen, many job seekers have found their prospects hampered by these screening practices. In response, several local governments and some large private sector employers have decided to, eliminate such questions – to “ban the box.” Daniel Shoag, associate professor of public policy at Harvard Kennedy School, and Stan Veuger of the American Enterprise Institute for Public Policy explore this phenomenon in a working paper titled “No Woman, No Crime,” published this spring.
Background checks, the authors note, affect “a significant chunk of the population: as many as 65 million people are estimated to have been arrested and/or convicted of criminal offenses.” This is especially true for Hispanic and African-American men, who are convicted of crimes at much higher rates than women of any race.
Because “banning the box” laws are relatively new, research on their effects is limited so far. The authors used LEHD Origin-Destination Employment Statistics (LODES) data on employment outcomes to examine the bans’ effects on the labor market prospects of various groups.
“We tried to use new data to find out whether or not these laws make a difference, and if they do, how they redistribute employment opportunities,” Shoag explained. “We also sought to study how these laws affect the hiring process and how employers shift to other signals.”
The authors combined LODES data on employment by census tract with data from the National Neighborhood Crime Study. “These data let us look at how ‘ban the box’ laws changed employment outcomes in higher crime tracts relative to similar tracts without these laws, and even how these policies changed commuting patterns to destinations with and without these bans,” Shoag said. “The detail in the data makes it possible to carefully identify a causal effect.”
Data from Burning Glass Technologies, which gathers information on millions of online job postings, allowed the authors to explore other techniques used to screen job applicants, such as skills and experience. They found that “ban the box” policies can raise the employment of residents of the top quartile of high-crime neighborhoods by as much as 4%. This increase is in large part driven by residents getting hired into relatively low-wage jobs (over $15,000 in annual wage income) and finding work in the public sector. African Americans men, who are more likely to have been convicted of crimes, benefit from these policies. But African American women, who are less likely than men to have criminal convictions, often see their employment opportunities reduced.
“We found that these laws ‘work,’ in the sense that they increase employment for those living in high-crime neighborhoods,” Shoag noted. “The impact is big. But the increase is mostly in lower wage work, and the biggest industry response is in the public sector.”
The authors also examined the data for evidence of “upskilling,” in which employers raise the education and experience requirements for job applicants after eliminating background checks or related questions. “We find that employers substitute to other markers of skill when screening applicants,” Shoag said. “When they can't screen for criminal records upfront, employers request more experience and education in their job postings.” This phenomenon may help people with criminal records, but it may harm employment prospects for women, especially black women.
“These results highlight both the importance and success of ban-the-box initiatives and some of their unintended consequences,” the authors note. As policymakers continue to grapple with issues of employment, the data on “ban the box” and other initiatives may merit further study.