screened according to whether they had hired an employee into a position not requiring a college
degree within the previous year. However, this screen eliminated no firms from our sample.
The overall survey response rate was about 65 percent, in the range of other similar recent firm
surveys (Holzer and Stoll, 2001).
The surveyed firms were drawn from a sample that was stratified ex-ante by
establishment size. Sampling across strata was performed in proportion to the amount of
regional employment accounted for by the establishment size category. Within strata, firms were
sampled at random. Thus, the sample is representative of the distribution of the workforce in the
Los Angeles region across establishment size categories without any need for additional size-
The survey contains extensive information on the establishment’s characteristics (e.g.,
size industry, presence of collective bargaining, minority ownership status, and the racial
composition of its applicants). In addition, the survey contains information on the characteristics
of the mot recent job filled in the firm that did not require a college degree (i.e., non-college job).
This includes the screening and hiring behavior of employers in filling that job and the task and
skill requirements for the job, among other factors.
The main variables we focus on in this analysis are indicators of employers’ prospective
willingness to hire ex-offenders and their actual hiring of ex-offenders. With respect to the
former indicator, for the last job filled that did not require a college degree, we ask the employer
“Would you accept for this position an applicant who had a criminal record? definitely will,
probably will, depends on the crime, probably not, absolutely not?” In addition, we also ask
employers generally “Suppose you were contacted by an employment agency that was trying to
place young males with criminal records. Do you currently have any open positions that you