SHOULD CRIMINAL HISTORY RECORDS BE UNIVERSALLY AVAILABLE?

Abstract

At current incarceration rates, roughly 9% of all males born today will serve time in prison at some point in their lives. By race and ethnicity, the comparable figures are 4% for white males, 16% for Hispanic males, and 28% for black males (U.S. Bureau of Justice Statistics, 1997). Ample research now documents that having spent time in prison worsens one’s employment prospects through several potential channels. First, employers are strongly averse to hiring ex-offenders, a finding that has been born out in both audit studies (Pager, 2003) as well as employer interviews (Holzer et al., 2004). Second, spending time in prison may prohibit the accumulation of valuable work experience and even erode one’s labor market skills. In conjunction with the large increases in U.S. incarceration rates in recent decades, these facts suggest that criminal history records represent an increasingly important barrier to employment for young, less-skilled men. Moreover, given racial disparities in incarceration rates, the increasing salience of this characteristic is likely to exacerbate racial inequality in employment and earnings. In their provocative analysis, Kurlychek et al. (2006) raise the important question of whether unfettered employer access can be justified by the legitimate concerns of employers. More specifically, they assess whether the rate at which young offenders desist from offending with time since the last offense merits limiting employer access to arrest and conviction information for sufficiently distant past offenses. This question is important and timely given the large increase in the pool of ex-offenders (by some estimates nearly 5% of the adult male population). Moreover, as policies governing access have liberalized and information systems have advanced to the point where now several private firms offer cheap and quick criminal background checks, employers are checking the criminal history records of applicants at rates that are historically unprecedented (Holzer et al., 2004). Kurlychek et al. (2006) demonstrate that for a cohort of young men in Philadelphia the likelihood of a repeat offense declines precipitously with the time that has elapsed since the last offense. This pattern is consistent with both a causal effect of staying clean as well as a remaining population of former offenders that becomes increasingly selected with time since the