and actual behavior of firms that always check backgrounds are not completely similar, and that
such checks may not limit the actual hiring of ex-offenders.
On the other hand, background checks may not influence whether a firm hires a least one
ex-offender, but they may limit the overall number of them hired at the firm. One possibility is
that background checks are used not necessarily to exclude all ex-offender applicants (at least for
jobs where ex-offenders are not legally barred from employment), but perhaps in many cases to
provide information to employers about specific offending backgrounds to help guide
employment decisions. An alternative explanation is that it is likely that fewer ex-offenders
apply to firms that always conduct background checks for understandable reasons. However, we
are not able to distinguish with our data which of these explanations dominates.
We also find a similar pattern with respect to the racial composition of applicants to the
firm. We include this factor in the analysis since blacks and Latinos are overrepresented among
the ex-offender population (BLS, 2001), and thus higher percentages of applications from these
groups will likely correlate with increased applications from ex-offenders. The average
percentage of applications from black males and Latinos are nearly identical across the
willingness-to-accept categories. However, the percentage of applications from black males and
Latinos for firms that have hired ex-offenders is significantly higher than for those firms that
have not.13 This suggests that though blacks and Latinos are overrepresented among ex-
offenders, firms’ willingness to hire ex-offenders seems not to be influenced by the extent to
which they received applications from these groups and vice versa. However, whether or not
offenders in which the categories included willing to accept, depends on crime and unwilling to accept was also estimated, but produced qualitatively similar results to those shown in Table 1 as well.
13 This result for Latinos is likely driven by male Latinos since ex-offenders are overwhelmingly male. However, our data does not include a variable that measures the percentage of applicants that are male or female Latinos.