market prospects of ex-offenders in greater detail by describing the specific employment barriers
that confront this growing population of mostly young men.
2. Data and Definitions of Main Variables
Our data were collected using 20-minute telephone surveys administered to 619
establishments in Los Angeles. The survey was administered between May 2001 and November
2001. We chose to survey employers in Los Angeles for a variety of reasons. It is a large and
populous metropolitan area in a state with a large incarcerated population in which nearly a third
of recently released prisoners return to Los Angeles County, the geographic boundaries of our
study area. At the time of the survey its regional economy registered some of the lowest
unemployment rates in thirty years and appeared relatively strong while the national economy
had dipped into a recession.
However, while the survey was in the field, the Los Angeles economy began to weaken,
particularly in the manufacturing sector; and, of course, the events of September 11 took place.
These events are likely to have affected employer responses to questions about their willingness
to hire ex-offenders, perhaps in the negative direction, which we explore later in this paper.
By and large, the sample of establishments drawn and other survey methods were
borrowed from the earlier survey of employers that we have analyzed in previous papers.4
Employers were drawn from lists complied by Survey Sampling Inc. (SSI), primarily from
telephone directories. To the extent possible, the phone interviews were conducted with the
person in the establishment who is responsible for entry-level hiring. Establishments were
4 Harry J. Holzer developed and administered this survey, called the Multi-City Employer Survey (MCES). MCES includes observations on 3,220 employers in four cities (approximately 800 per city): Atlanta, Boston, Detroit, and Los Angeles. The Los Angeles portion of this survey used the identical geographic sampling unit as that used in the survey we report on here. See Holzer (1996) for an extensive discussion of the survey methods and data.