M-A-D Prison Ministry

Will Employers HIRE Ex-Offenders?

 

At current incarceration rates in the United States, the Bureau of Justice Statistics (BJS) estimates that approximately 9 percent of all men will serve some time in state or federal prisons. Moreover, for certain subgroups of the population, the proportion likely to serve time is quite large. For example, nearly 30 percent of African American men and 16 percent of Hispanic men will serve prison sentences at some point in their lives (U.S. Bureau of Justice Statistics, 1997). The BJS also estimates that the median time served for prisoners released during the late 1990s was less than 2 years. Together, these two pieces of information (high incarceration rates and a relatively short median sentence) suggest that a large minority of noninstitutionalized men have served time in state or federal prisons. Although state and federal prison populations appear to be leveling off, the rapid increase in incarceration rates over the past two decades indicates that the numbers of ex-offenders among the noninstitutionalized population are likely to be large into the foreseeable future. The successful reintegration of this growing population depends in part on the employment potential of ex-offenders. There are several reasons to suspect that serving time may adversely affect future earnings and employment prospects. The incarcerated do not accumulate work experience and may experience an erosion of skills while serving time. Furthermore, any ties to legitimate employers are likely to be severed by an initial arrest and by a prison spell. From the viewpoint of employers, a criminal history record may signal an untrustworthy or otherwise problematic employee. Employers may avoid such workers due to a perceived increased propensity to break rules, steal, or harm customers. Several studies have analyzed the labor market consequences of involvement in the criminal justice system by testing for direct effects on future employment and earnings of being arrested (Grogger, 1995) or of serving time (Kling, 2000). In this paper, we take an alternative tack. We characterize the demand side of the labor market for ex-offenders by analyzing employer self-reported preferences with respect to applicants with criminal histories. Specifically, we seek to answer the following questions. To what extent are employers willing to hire workers with criminal backgrounds? Does this willingness to hire vary with different job and firm characteristics and over time with the tightness of the labor market? To what extent do employers act on these preferences by investigating the criminal history records of applicants? Does the propensity to check also vary with specific job and firm characteristics? Finally, how do these measures of employer demand for ex-offenders compare with their potential supply to the labor market, and what do these comparisons imply about employment prospects for ex-offenders in the current environment and near future? We begin with a discussion of the possible reasons why employers may be unwilling to hire workers with criminal backgrounds, as well as a discussion of the means by which employers may glean information about the criminal records of applicants. Next, we use a representative sample of employers from four large metropolitan areas to examine the extent to which employers would consider hiring ex-offenders and the extent to which employers check for criminal backgrounds during the hiring process. We then analyze the degree to which the willingness to hire and the checking of backgrounds vary with firm and job-specific characteristics and with the tightness of labor markets. We also compare these estimates of demand with the numbers of ex-offenders who might at least potentially enter the labor market on the supply side. Finally, we discuss the implications of the findings for the employment prospects of ex-offenders and, more generally, for job-seekers from disadvantaged groups. EMPLOYERS’ WILLINGNESS TO HIRE EX-OFFENDERS AND THEIR USE OF CRIMINAL HISTORY RECORDS Many of the demand-side factors that adversely affect the employment prospects of low-skill workers are also likely to affect the employment prospects of ex-offenders, given the overlap between these two groups. For example, “hard” and “soft” skill requirements by employers, transit and locational factors, slack urban labor markets, employer discrimination, and weak social networks are all likely to adversely affect the employment and earnings of both low-skill workers in general and the ex-offender population in particular (Freeman and Rodgers, 2000; Stoll, Holzer, and Ihlanfeldt, 2000; Moss and Tilly, 2000; Holzer, 1996). Nonetheless, ex-offender status may create additional barriers to employment above and beyond those experienced by other low-skill workers. Employers may be unwilling to hire workers with criminal backgrounds (all else held equal) due to several considerations. To start, certain occupations are legally closed to individuals with felony convictions under state and, in some cases, federal law (Hahn, 1991). Examples include jobs requiring contact with children, certain health services occupations, and employment with firms providing security services. In addition, employers may place a premium on the trustworthiness of employees, especially when the ability to monitor employee performance is imperfect. Jobs involving significant customer contact or the handling of cash or expensive merchandise will require dependable, honest employees. To the extent that past criminal activity signals something less, employers may take such information into account when making hiring decisions. Furthermore, in many states employers can be held liable for the criminal actions of their employees under the theory of negligent hiring. Legally, negligence is premised on the idea that one who breaches a duty of care to others in an organization or to the public is legally liable for any damages that result (Glynn, 1988). Under the theory of negligent hiring, employers may be liable for the risk created by exposing the public and their employees to potentially dangerous individuals. As articulated by Bushway (1996), “employers who know, or should have known, that an employee has had a history of criminal behavior may be liable for the employee’s criminal or tortious acts.” Thus, employers may be exposed to punitive damages as well as liability for loss, pain, and suffering as a result of negligent hiring. Employers have lost 72 percent of negligent hiring cases, with an average settlement of more than $1.6 million (Connerley, Arvey, and Bernardy, 2001). The high probability of losing, coupled with the magnitude of settlement awards, suggests that fear of litigation may substantially deter employers from hiring applicants with criminal history records. The ability of employers to effectively act on an aversion to hiring ex-offenders, and the nature of the action in terms of hiring and screening behavior, will depend on the accessibility of criminal history record information to entities outside the criminal justice system. Information on arrest, conviction, and time served for nonfederal offenses are compiled by the state where the offense occurred. Each state and the District of Columbia maintains a central repository where this information is housed and from which criminal history information is disseminated. All law enforcement agencies within a state are required to report arrest and disposition information to the central repository for all serious offenses (U.S. Department of Justice, 1999). The information in the repository is the source used to generate rap sheets for law enforcement officials and the source for criminal history records for non-criminal-justice purposes. In its most recent review of state privacy and security legislation, the U.S. Department of Justice concludes that criminal history record information is becoming increasingly more available to noncriminal-justice users, although the degree of openness varies from state to state (U.S. Department of Justice, 1999). Nearly all states make a distinction between arrest records and conviction records. In general, states are less likely to freely disseminate information on arrest records, especially arrests for cases that are still open or have occurred within the previous year. States tend to place fewer restrictions on non-criminal-justice access to conviction records. Currently, 23 states have some form of public access or freedom of information statutes that pertain to some aspect of criminal history information. Given the availability of criminal history records to non-criminal-justice users, employers are likely to check the criminal backgrounds of potential employees. Certainly, the extent to which they do so is likely to be in part a function of their aversion to ex-offenders, for all the reasons we discussed. In addition, the propensity to check the backgrounds of applicants is likely to be related to the size of the local ex-offender population as a proportion of the local labor force. This proportion varies from state to state due to interstate differences in sentencing and other criminal justice policies. To be sure, employers can act on an aversion to hiring ex-offender in the absence of information on an applicant’s criminal history record by screening job applicants on the basis of characteristics that are predictive (either in actuality or in perception) of previous criminal activity. For example, if employers believe that African Americans, welfare recipients, or workers with unaccounted-for breaks in their employment histories are more likely to have past criminal convictions, employers may statistically discriminate against such individuals. Although imperfect information will clearly lead to instances of “false positive” and “false negative” assessments of previous criminality, basing employment decisions on such discriminatory rules of thumb may minimize the likelihood of hiring ex-offenders. DESCRIPTION OF THE DATA To examine whether employers are willing to hire ex-offenders and to conduct criminal background checks, we use an establishment survey collected through the Multi-City Study of Urban Inequality (MCSUI). The survey includes slightly over 3,000 establishments and was conducted between June 1992 and May 1994 in the Atlanta, Boston, Detroit, and Los Angeles metropolitan areas. The sample of firms is drawn from two sources: the employers of the respondents to a household survey conducted in conjunction with the survey of establishments that provided approximately 30 percent of the observations, and a sample of establishments generated by Survey Sampling Incorporated (SSI). The SSI sample is a randomly stratified sample where the initial lists are stratified by establishment size and firms are sampled according to the proportion of metropolitan area employment accounted for by their respective size categories. Hence, the SSI sample is representative of the set of establishments faced by a job seeker in any of the four metropolitan areas. We use sample weights in all calculations to account for the nonrepresentative portion of the sample from the household survey. Establishments were screened according to whether they had hired an employee into a position not requiring a college degree within the previous year. The response rate for firms that passed the initial screen was 67 percent. This compares favorably with other establishment surveys (Kling, 1995).7 Telephone surveys were conducted with individuals in charge of hiring at the firm. The survey includes two question vital to the current analysis: one on employer preferences with respect to workers with criminal histories and another on whether employers use criminal background checks. The question on employer preferences refers to the most recently filled position and reads, “Would you accept for this position an applicant who had a criminal record?definitely will, probably will, probably not, absolutely not.” For criminal background checks, the question reads, “For the last position hired into, how often do you check the applicant’s criminal record?always, sometimes, or never.” CHARACTERIZING THE DEMAND FOR EX-OFFENDERS A. Are Employers Reluctant to Consider Ex-Offenders? Figure 1 presents the distribution of employer responses to the question inquiring about the likelihood that the employer would be willing to accept an applicant with a criminal record. Over 60 percent of employers indicate that they would “probably not” or “definitely not” be willing to hire an applicant with a criminal record, with “probably not” being the modal response.8 Only 38 percent of employers indicate that they definitely or probably would consider an applicant with a criminal history, with 12.5 percent indicating that they definitely would consider hiring an ex-offender. Hence, the simple distribution of the responses to this question reveals a relatively widespread aversion to applicants with criminal histories. To put these employer responses into perspective, Figure 2 presents the distributions of employer responses to similarly worded questions concerning the likelihood that employers would accept applications from other groups of low-skill and possibly stigmatized workers. The figure illustrates employer preferences with respect to the acceptability of welfare recipients, applicants with a GED but no high school diploma, applicants with spotty work histories, and applicants who have been unemployed for a year or more. Approximately 92 percent of employers indicate that they definitely or probably would hire former or current welfare recipients, 96 percent indicate that they definitely or probably would hire workers with a GED in lieu of a high school diploma, 59 percent indicate that they would hire workers with a spotty employment history, and 83 percent indicate that they would be likely to consider an application from an individual who has been unemployed for a year or more. In contrast, only 38 percent of employers indicate that they definitely or probably would accept an application from an ex-offender. Hence, in addition to the fact that the aversion to hiring applicants with criminal histories is widespread, this aversion is considerably stronger than the aversion of employers to hiring applicants from other commonly stigmatized groups of workers. Do employer attitudes concerning the employability of ex-offenders differ across the metropolitan areas represented in our sample? Figure 3 presents results comparable to those in Figure 1 for each of the four metropolitan areas (Atlanta, Boston, Detroit, and Los Angeles). Employer responses are comparable across areas. In all metropolitan areas, approximately 60 percent of employers indicate that they definitely or probably would not consider applicants with criminal histories. This uniformity is particularly striking considering the interarea differences in size, demographic composition, and economic conditions that exist. Given that the data pertain to employers who have recently hired low-skill workers (employers who are perhaps the most likely to employ ex-offenders), the simple distributions in Figures 1 and 3 imply that the large majority of employers are unwilling to hire exoffenders and that this unwillingness appears uniform across cities. An alternative indirect means of gauging employer aversion to applicants with criminal histories is to assess whether employers investigate the criminal backgrounds of their potential employees. Figure 4 presents the distribution of employer responses to the question concerning the frequency with which employers check the criminal backgrounds of job applicants. Approximately 32 percent of employers in our sample say that they always check, 17 percent indicate that they check sometimes, and 51 percent indicate that they never check criminal backgrounds. Though the period covered in our sample precedes the widespread availability of internet services providing low-cost criminal background checks, the use of criminal checks by a sizable minority of employers is evident in the data. To the extent that the question concerning employer willingness to hire ex-offenders and the question concerning the use of background checks are both measuring employer aversion to hiring applicants with criminal histories, the responses to the two questions should be systematically related. Figure 5 explores this possibility. The figure presents the distribution of the responses to the criminal background checks question by the responses to the question concerning the likelihood that the employer would consider hiring an ex-offender. There is a clear association between the responses to these two questions. Only 19 percent of employers who say they definitely would consider hiring an ex-offender check criminal backgrounds, but 56 percent of employers who say that they would definitely not hire an applicant with a criminal history conduct background checks. Conversely, while 32 percent of the least willing to hire employers never check criminal backgrounds, the comparable figure for those employers most willing to hire ex-offenders is 61 percent. Hence, employer responses to the two questions are largely consistent with one another and suggest considerable employer reluctance to hire ex-offenders. Concerning patterns across metropolitan areas, there are some significant differences in the propensity to check criminal backgrounds across areas, in contrast to the uniformity in the responses to the “willingness to hire” question depicted in Figure 3. Figure 6 presents this comparison for the four metropolitan areas in our sample. The figure shows that the propensities to check criminal backgrounds among firms in Atlanta, Detroit, and Los Angeles are generally comparable, but that employers in Boston were the least likely to use criminal background checks as a screening device. Potential explanations for this difference may lie in interarea differences in the size of the population of ex-offenders or in differences in the ease with which employers can access criminal justice information for non-criminaljustice purposes. To investigate the first potential explanation, Table 1 presents estimates of the incarcerated and ex-offender populations for the four states in which the metropolitan areas are located during the period covered by our sample.10 As can be seen, the incarceration and probation rates, as well as the proportion of the population that represents recently released offenders (all expressed per 100,000 state residents), are considerably lower in Massachusetts than in the other three states. Hence, the lower propensity of Boston firms to investigate the criminal background of employees may be due in part to the relatively small ex-offender population of the host state. Concerning differences in the stringency of state laws governing non-criminal-justice uses of criminal history records, examination of the most recent Compendium of State Privacy and Security Legislation: 1999 Overview (U.S. Department of Justice, 1999) does not reveal glaring interstate differences in who can legally obtain access to criminal history records. However, in a review of earlier issues of the compendium, Bushway (1996) concludes that criminal history records were considerably less available in Massachusetts to non-criminal-justice entities around the beginning of the 1990s. To summarize, Figures 1 through 6 and Table 1 indicate clear patterns. Through their stated hiring preferences and as revealed by their actions (i.e., running background checks), employers reveal considerable reluctance to hiring workers with criminal histories. This aversion appears to be stronger  than the aversion to hiring workers from other commonly stigmatized groups. Moreover, the correlation between the stated preferences and checking behavior are consistent with one another. We now turn to a description of the correlates of these measures of employer demand for ex-offenders. B. Which Employers Are Most Likely to Avoid Applicants with Criminal History Records and Which Employers Check? Employer attitudes toward applicants with criminal histories are likely to depend on the nature of the employer’s business and on the nature of the position that is being filled. One would expect a priori that employers who cannot perfectly monitor their employees, or employers who hire workers into positions that deal frequently with the public, would be more averse to hiring ex-offenders. One might also suspect that employers filling jobs that require the handling of expensive merchandise, large amounts of cash, or costly equipment may be differentially averse to applicants with a criminal past. To explore these possible differences, Table 2 presents averages of establishment characteristics, recruiting and screening methods, required job tasks and qualifications, and desired employee characteristics for the sample of employers stratified by the four possible responses to the question concerning willingness to consider ex-offenders. Establishment characteristics include the number of employees at the establishment, industry, the percentage of workers represented by a union, dummy variables indicating that the hiring agent is black and that the firm is located in the central city, a variable measuring the average distance to blacks in the metropolitan area, and a variable measuring the average distance to whites in the metropolitan area. Our measures of recruiting and screening methods are a set of dummy variables equal to 1 if the employee regularly uses the described methods. Similarly, the daily job tasks are the means of dummy variables describing daily tasks performed on the job, such as interacting with customers, reading, writing, and using a computer. “Job qualifications” refers to applicant qualifications that the employer expresses are either absolutely necessary or strongly preferred, and “desired characteristics of new employees” refers to employee traits that employers deem important. Also included is “types of applicants who would probably not be hired.” Several clear patterns emerge in Table 2. First, the establishment size distribution among the least-willing employers is skewed toward smaller firms, while large firms are disproportionately represented among employers most willing to hire workers with criminal histories. Among employers willing to hire ex-offenders, manufacturing firms are disproportionately represented, whereas establishments in the finance, insurance, and real estate sector and the services sector are underrepresented. The opposite pattern holds among the least-willing employers. There is little relationship between the remainder of the establishment characteristics and employer aversion to hiring ex-offenders, although employers located in the central city are slightly more likely to be among the firms willing to hire. Concerning recruiting and screening methods, firms that are averse to hiring ex-offenders are less likely to consider applications from walk-ins, to post help-wanted signs, to consider referrals from state and community agencies, or to use affirmative action in recruiting for the position. Positive relationships are also seen between employer aversion and several of the screening methods, such as using aptitude and personality tests, checking criminal backgrounds, and verifying educational attainment and references. One of the strongest associations evident in Table 2 is the positive relationship between employer unwillingness to hire ex-offenders and whether the recently filled job involves frequent customer contact. Among employers most willing to hire ex-offenders, 52 percent of the positions required customer contact. Among employers least willing to hire ex-offenders, 71 percent of the positions required customer contact. Weaker positive correlations are also found between unwillingness to hire and several of the other job tasks, including phone conversations, reading, and writing. Employers least willing to hire ex-offenders are more likely to require high school degrees, recent and specific work experience, references, and some vocational education. Such employers are also more likely to indicate that physical attractiveness, neatness, politeness, motivation, verbal skills, and the ability to speak English well are very important employee characteristics. Hence, Table 2 indicates that smaller, nonmanufacturing firms whose employees interact with customers are the most averse to hiring ex-offenders. The patterns also indicate that averse employers are less likely to use informal recruiting techniques (walk-ins, for example). These patterns are sensible and support our earlier discussion of the potential reasons why employers may be reluctant to hire ex-offenders. Table 3 presents average characteristics for the sample of establishments stratified by the response to the question concerning employer use of criminal background checks. The variables presented are exactly the same as those presented in Table 2. There are notable differences in the size and industrial distributions between employers that check and employers that do not. Despite the greater reluctance of smaller employers to consider applicants with criminal histories, smaller employers are most represented among establishments that never use criminal background checks. Nearly 40 percent of employers that never check have fewer than 20 employees, compared with 24 percent for employers that always check. Concerning industry, manufacturing firms are the least likely to use criminal background checks while establishments in finance, insurance, and real estate and in services are the most likely. There is a negative relationship between a firm’s proximity to black neighborhoods and use of criminal background checks. In addition, more-unionized establishments screen criminal history records more than do lessunionized establishments. Several interesting patterns are seen in the difference in the use of various recruitment methods. Employers that check criminal backgrounds are more likely to use informal recruiting methods (accepting walk-ins and posting help-wanted signs), to accept referrals from state and community agencies, and to use affirmative action in recruiting. In Table 2 we saw that firms unwilling to hire workers with criminal histories were less likely to use these tools. The combination of these two findings suggests that in the absence of explicit inquiries into an applicant’s criminal history, employers use these alternative recruiting and screening methods to avoid applicants who are ex-offenders. Concerning screening methods, use of background checks is strongly associated with the use of other forms of tests (such as drug and aptitude tests) and with the likelihood that the employer verifies the stated educational attainment of the applicant and the applicant’s references. Employers filling positions that require customer contact are more likely to check criminal backgrounds, consistent with the patterns in Table 2. In addition, employers who check are more likely to require high school diplomas, though the association is considerably weaker than that with the variable measuring employer unwillingness to hire ex-offenders. The positive association with firm size, unionization rates, and use of other screening tests suggests that employers with more formal human resources systems are more likely to run background checks. The positive association between checking and positions with customer contact indicates that the determinants of employer aversion, and possibly state law, are also important determinants of whether one checks. However, several patterns in Table 3 indicate that the differences in the averages between employers who check and employers who don’t are opposite of what one might predict from the patterns in Table 2. Hence, a further dissection of the data may better illuminate the relationship between employer aversion, the use of criminal background checks, and recruiting and screening. Table 4 provides this more detailed cross-tabulation of the data. We first classify all employers who say that they definitely or probably would hire applicants with criminal records as willing to hire exoffenders, and employers who respond definitely not or probably not as unwilling to hire ex-offenders. Next, we dichotomize the criminal background checks variable by defining employers who say they check sometimes or always as checks, and employers who say they never check as doesn’t check. The table presents means for the variables in Tables 2 and 3 for the four categories defined by these two dichotomized variables. Stratifying the sample in this manner reveals several patterns that are masked in Tables 2 and 3. First, the firm size distributions indicate that small firms constitute a very large portion of firms that are unwilling to hire and that do not check criminal backgrounds. Fully 45 percent of establishments in this category have fewer than 20 employees. Concerning the distributions by industry, manufacturing establishments are quite likely to be among firms that are willing to hire ex-offenders and that never review criminal history records, while establishments in the services and the finance, insurance, and real estate sectors are most likely to be unwilling and to check. Retail and wholesale trade establishments are disproportionately represented among firms that will not hire ex-offenders and that never check (a pattern consistent with the size distributions). Skipping ahead to daily job tasks, customer contact is positively associated with checking among both firms that are willing and firms that are unwilling to hire ex-offenders. In addition, employers who check criminal backgrounds are more likely to demand certain job qualifications of applicants, relative to employers who do not check. Perhaps the most interesting patterns in Table 4 are found in the differences in the means of the dummy variables indicating types of applicants that the employer avoids. Among firms that are willing to hire ex-offenders, checking is basically unrelated to these variables. Among firms that are unwilling to hire ex-offenders, firms that check are less averse to hiring these types of workers than firms that do not check. For both firms that check and firms that don’t check, unwillingness to hire ex-offenders is associated with a greater unwillingness to hire the types of applicants described by the dummy variables. However, this differential aversion is greatest among employers who do not run criminal background checks. Specifically, among employers who do not check criminal history records, the difference between those who are unwilling to hire ex-offenders and those who are willing is 12 percentage points for the “not willing to hire welfare recipients” dummy, 6 percentage points for the “unwilling to hire workers with a GED” dummy, 24 percentage points for the “spotty work history” dummy, and 14 percentage points for the “unemployed for a year or more” dummy. The comparable differences among firms that do check are 7, 2, 13, and 9 percentage points, respectively. These differences among employers who check are uniformly lower (and by considerable magnitudes) than those for employers who do not. These patterns indicate that employers who do not review criminal history records and who are unwilling to hire ex-offenders are more likely to exclude from consideration applicants with characteristics that may be indicative of a criminal history. The results in Table 4 suggest that educational attainment, prior participation in public assistance programs, and gaps in an applicant’s employment history are perceived by employers as such a signal. An alternative signal that may be taken into account by employers is race. Specifically, African Americans are considerably more likely to have past criminal convictions than are members of other racial and ethnic groups. Moreover, employers may overestimate the average incidence of prior conviction among blacks due to prejudice or a general lack of experience with black employees. To explore this possibility, Table 5 presents average values of a dummy variable indicating that the last worker hired into a noncollege position is black. We present averages for the whole sample, for the sample stratified by whether the firm checks criminal backgrounds, for the sample stratified by whether the employer is willing to hire ex-offenders, and for the four categories defined by the cross of these two variables. The final row of the table presents the differences in means between firms that are unwilling to hire ex-offenders and firms that are willing, while the final column presents differences in means between establishments that check criminal backgrounds and establishments that do not. For the sample overall, there is no discernible overall difference in the likelihood of hiring a black worker between employers who are willing to hire ex-offenders and employers who are unwilling. There is a large significant difference, however, between employers who check criminal backgrounds and employers who do not. Relative to employers who do not check, employers who check are 8.5 percentage points more likely to have hired an African American applicant into the most recently filled position. This difference is statistically significant at the 1 percent level of confidence. Among employers willing to hire ex-offenders, this difference is 4.8 percentage points and is marginally significant. Among employers who are unwilling to hire ex-offenders, this difference is 10.7 percentage points and is highly significant. Moreover, the difference between these two differences (shown in the last row and last column) is statistically significant at the 8 percent level. This latter finding indicates that the relatively larger positive effect of background checks for firms that are unwilling to hire ex-offenders is larger and statistically distinguishable from the effect for firms that are willing. Hence, the patterns observed for the groups of stigmatized workers in Table 4 are reproduced for the hiring outcomes of African Americans in Table 5. The statistical discrimination implied by the results in Tables 4 and 5 raises interesting policy questions with respect to state laws governing non-criminal-justice access to criminal history records. To the extent that employers substitute such screening for actual information on criminal histories, there will be some workers who will be unfairly discriminated against. In addition, workers from groups with high rates of previous criminal activity (for example, young men, African Americans, workers with gaps in their employment history) are likely to be considerably impacted by such discrimination. Hence, while more liberal access policies may adversely impact the employment prospects of applicants with criminal histories, the information infusion may positively impact the employment prospects of workers from groups with high incarceration rates. This is a provocative trade-off that needs to be explored in further detail. C. Does Employer Demand for Ex-Offenders Vary over Time or with the Business Cycle? Regardless of whether employers check backgrounds, the data presented above indicate relatively limited demand by employers for ex-offenders. However, given that these data are based on surveys administered during a period of relatively slack labor markets earlier in the decade (1992–1994), the figures presented above may substantially understate this component of labor demand. It is widely known that tight labor markets tend to disproportionately increase demand for less-skilled and disadvantaged workers (Hoynes, 2000), and that African Americans benefited particularly from the recent boom (Freeman and Rodgers, 2000). It is also quite possible that any improvements in demand that initially result from cyclical factors could generate secular changes in employer behavior. For example, employer attitudes and hiring behavior may adjust with experience in a manner that reduces the stigma attached to a prior conviction.13 Thus, the timing of the survey used to generate the results presented above begs the question, Has such demand improved over time, particularly given the dramatic tightening of U.S. labor markets that occurred over the remainder of the 1990s?14 To shed some light on this question, we turn to an additional source of data, a more recent set of surveys that were administered to roughly 3,000 employers by phone in several metropolitan areas in 1998–99. The metropolitan areas in which the latter surveys were administered are Chicago, Cleveland, Los Angeles, and Milwaukee.15 A certain amount of overlap exists in the questions asked on each survey. Hence, we can make some limited inferences about changes over time in employer demand for ex-offenders and responsiveness of such demand to business-cycle conditions. Specifically, the latter survey contains the same questions as the earlier survey concerning willingness to hire ex-offenders and other stigmatized workers into the last noncollege job filled.16 The more recent survey also includes some additional questions on current willingness to hire ex-offenders. Specifically, employers were asked whether they would be willing to hire any ex-offenders at the current time, and if so how many. Tabulations using these data are presented in Table 6. The upper portion of the table presents the percentages of employers in both surveys who answered that they would “definitely” or “probably” have hired persons with criminal backgrounds into their last-filled noncollege job. The data are presented for the overall sample of establishments and also for subsamples by metropolitan area (Milwaukee versus other) and by selected industries. The lower portion of the table then presents data from the more recent survey only on current willingness to hire ex-offenders. The table presents the percentages of employers who indicated they would hire ex-offenders currently as well as the percentages of all jobs in the sample establishments that could potentially be filled by ex-offenders. Tabulations are presented for the total sample of establishments and the same subsamples used in the upper portion of the table. Comparison of results from the two surveys indicates a small increase between 1992–94 and 1998–99 in the proportion of all employers willing to hire ex-offenders, from 38 to 41 percent. However, all of the increase is accounted for by the relatively greater willingness of employers in Milwaukee to hire ex-offenders (49 percent). By industry, we find some substantial increases in the willingness of retail trade establishments to hire ex-offenders, and much smaller changes within the manufacturing and services sectors, though the relative rankings of industries in terms of willingness to hire ex-offenders do not change. How should we interpret these numbers? The relatively greater willingness of employers in Milwaukee to hire ex-offenders no doubt at least partly reflects the very tight labor market conditions of Wisconsin in the late 1990s, though other factors might also be in play here.17 The tightening of labor markets in other areas should have led to at least some improvements in demand for ex-offenders as well. However, the characteristics of firms and jobs in our samples of employers changed across the two surveys in a manner that is likely to partially offset any increases in demand for ex-offenders caused by the business cycle.18 Moreover, since labor markets nationwide continued to be very tight at least through the end of 2000, it is also possible that our data (which reflect the period only through early 1999) do not fully capture the extent of rising demand for ex-offenders during this period.19 Some additional light is shed on this issue by the data on current willingness to hire ex-offenders from the 1998–99 survey. Overall, just under 20 percent of employers expressed a current willingness to hire ex-offenders, and 1.4 percent of all jobs in their establishments would be available. Once again, those in Milwaukee are more willing to hire ex-offenders than those elsewhere, though the distinction is much clearer among proportions of employers than among proportions of all jobs open. By industry, however, we now find a relatively greater current availability of jobs to ex-offenders in retail trade than in manufacturing or the services. This pattern is consistent with the much higher rates of gross hiring and job vacancies in the former sector.20 Indeed, the data on the current willingness of employers to hire ex-offenders suggest a somewhat greater sensitivity to demand conditions than we saw in the earlier measure of employer demand based on the last-filled job. This seems largely to reflect differences in the sampling used to generate each measure. Since each establishment in the sample reports one recently filled noncollege job, that measure is based on a sample of employers weighted only by their number of employees.21 In contrast, current job availability measures differences across these establishments in hiring and job vacancy rates as well. The latter measure may thus be better suited than the former for evaluating the effects of demand conditions on employer hiring of ex-offenders and other groups. The latter measure also more accurately captures the shift in hiring toward retail trade and services that may impede the demand for ex-offenders (and other less-skilled workers) over time. Taken together, the data available do suggest that employer willingness to hire ex-offenders is at least partly responsive to business cycle conditions and that it did increase somewhat over the course of the 1990s. However, the magnitudes of these increases were modest at best, and employer demand for exoffenders remained quite limited even in the boom conditions at the end of the decade. D. Prisoners, Ex-Offenders, and Potential Supply to the Labor Market The relatively limited employer demand for ex-offenders that exists even in very tight labor markets raises another set of questions. Is such demand sufficient, at least in the short run, to absorb the numbers of offenders being released from prison? Moreover, who among the population of individuals either currently or previously incarcerated might contribute (at least potentially) to the supply of exoffender seeking legitimate employment in the labor market? Some data on the numbers of currently incarcerated prisoners, ex-prisoners, and ex-felons appear in Table 7. The data are for 1999, to facilitate comparisons with the demand-side data for 1998–99 in the previous table. The current inmate populations of federal/state prisons as well as the jail population are presented here. For the latter, the table presents figures for both total jail inmates and the subset of jail inmates who have been convicted of felonies.22 The numbers of prisoners currently on parole as well as the numbers of convicted felons on probation are also given. The table also presents the flow of prisoners released during 1999 and estimates of the total current stock of ex-prisoners and convicted exfelons. In addition to the raw numbers in each category, we also present percentages of the civilian labor force in 1999 that each number would constitute. These can then be compared with the percentages of all jobs available to ex-offenders on the demand side of the labor market that appear in the previous table. The data in Table 7 indicate that there are currently close to 2 million individuals incarcerated in prisons and jails in the United States, of whom about 1.4 million are convicted felons. Including those on parole and felony probation would add roughly 2.7 million more to those counts. The total stock of exprisoners and ex-felons is estimated to be close to 3 million and 9 million, respectively. If the numbers of current parolees and felons on probation are added to the latter figures, the result would be nearly 6 million ex-prisoners and nearly 12 million ex-felons at least potentially available in the labor market. In addition, the data indicate an annual flow of roughly 600,000 from prison each year. Combining annual flows of new parolees and those on felony probation with those of released prisoners generates as many as 2 million individuals added to the potential labor force each year. As percentages of the civilian labor force, these numbers are quite striking. For instance, those currently incarcerated represent 1 percent or more of the labor force. The totals currently on parole or felony probation together constitute nearly 2 percent of the labor force, while the total stock of exprisoners and ex-felons represents over 2 percent and 6 percent, respectively. When those on parole or felony probation are combined with the overall stock of ex-prisoners and ex-felons, the totals constitute about 4 or 8 percent of the civilian labor force, respectively. Considering annual flows out of prison rather than stocks, we find that the flow of new prisoners constitutes just 0.4 percent of the labor force each year; but when combined with the flow of new parolees and felons on probation, they generate estimates of well over 1 percent of the labor force being released each year. How can we compare these estimates with those of jobs that employers would fill with exoffenders? Combining ex-prisoners or ex-felons with those currently on parole or probation, and remembering that many ex-prisoners and ex-felons are already likely working, we estimate that the total stock of nonemployed ex-felons at any point in time might be 4–6 million, or 3–4 percent of the labor force; the stock of ex-prisoners might be 2–3 million, or 1.5–2 percent of the labor force. At least the latter figures (i.e., those for ex-prisoners) do not suggest a huge imbalance between the potential labor supply from ex-offenders and the aggregate job availability they might face (at 1.4 percent of all jobs), though the former figures (i.e., those for ex-felons) are somewhat less reassuring. Alternative approaches might consider annual labor market flows on the two sides of the market. A net new flow of ex-prisoners to the labor market of over 1 percent per year for each of the next several years almost certainly exceeds the average net flow of new jobs available to them, which by our estimates would be about 0.6 percent per year. On the other hand, relative to the gross flow of new hiring (which reflects hiring generated by turnover as well as by net new job growth), a great deal more employment might potentially be available—along with a good deal more competition from other low-skill workers in the workforce. Of course, any such exercise in which quantities on the supply and demand sides of the labor market are compared is suggestive at best and based on many strong assumptions.28 Even in the very short run, the magnitudes of any potential “mismatches” in the labor market will be sensitive to factors such as the numbers of ex-felons or ex-prisoners who would actually be seeking work, the extent to which employer aversion to hiring them varies with whether they were incarcerated, the number and nature of their offenses, their experiences since conviction or incarceration, etc. Furthermore, new supplies of workers over time will help generate new jobs, as wage rates adjust in response to the growing supplies of workers. Indeed, estimates of the ability of labor markets to absorb an additional 3 million welfare recipients before the fact of their entry were frequently too pessimistic (Burtless, 2000) and ignored the dynamic nature of that market (as well as the extraordinary labor market tightness that existed during that time). On the other hand, the data in our surveys indicate a much greater reluctance of employers to hire ex-offenders than welfare recipients. Furthermore, it seems as though the labor market that the exoffenders will enter during the next few years will be much less tight than it was during most of the previous decade, while the flow of new ex-offenders will remain quite large. Also, that flow is highly concentrated in poor minority neighborhoods and among minority (especially African American) men, so even an aggregate balance of potential supply and demand in the short run may overstate the true availability of jobs facing these ex-offenders. And, of course, the very limited skills and work readiness of this population mean that even potentially available jobs may be out of reach for a large part of this population (Travis, Solomon, and Waul, 2001). In sum, there may or may not be an aggregate imbalance between the numbers of ex-offenders returning to neighborhoods and jobs potentially available to them. But, given their personal characteristics and their concentration in poor neighborhoods and minority groups, it is likely that job availability facing this group will be quite limited. 5. CONCLUSION In this paper, we analyze employer demand for ex-offenders. We use data from a recent survey of employers to analyze not only employer preferences for offenders but also the extent to which they check criminal backgrounds in light of the very imperfect information about the job applicants whom they consider. We investigate the firm and job characteristics that correlate with these measures of employer demand. Using data from surveys administered at different points in time, we also consider the extent to which such demand changed during the 1990s in response to tighter labor market conditions. Finally, we consider the quantities of demand for ex-offenders relative to their supply, based on a variety of estimates of total stocks and annual flows of offenders back to the civilian population. Our data lead to a number of important findings. Employer willingness to hire ex-offenders is very limited, even relative to other groups of disadvantaged workers (such as welfare recipients). Employer willingness to hire is highly correlated with establishment and job characteristics, and is much lower in financial or service jobs and in jobs involving a variety of tasks, particularly direct customer contact, than in other sectors. Employer tendencies to check criminal backgrounds also vary greatly with characteristics of the establishment, such as its size, which presumably reflects both the resources and expertise available for human resources functions. The fact that many smaller firms refuse to hire ex-offenders but also do not check criminal backgrounds suggests that they may engage in statistical discrimination against a broader range of applicants, such as less-educated young black men. Paradoxically, efforts to make background checks easier for employers to perform and less costly might therefore improve job prospects for these latter groups, even while they weaken prospects for those who actually have criminal backgrounds. Our comparisons of employer data at different points in the 1990s suggest some sensitivity to demand conditions and some very modest improvements in demand for ex-offenders over the decade, though these were to some extent offset by a continuing shift in employment away from manufacturing and toward the retail trade and service sectors, and toward jobs with direct customer contact. Comparisons between our estimated quantities of labor demand for versus supply of ex-offenders also suggest some imbalance between the two, particularly as large numbers of offenders continue to be released over the coming decade and are heavily concentrated in poor minority communities. A good deal more research is needed to understand more about employer demand for exoffenders. For instance, to what extent do employers distinguish between those convicted and those incarcerated? How important are the nature and quantities of the offenses, as well as when they occurred and offender records since then? Are employers engaging in more criminal background checks over time, and how does this affect demand for offenders as well as for other disadvantaged groups more broadly? In the meantime, we can say with some certainty that employer demand limits the job prospects facing exoffenders in the labor market, in addition to the many other disadvantages and difficulties they face. References Burtless, Gary. 2000. “Can the Labor Market Absorb Three Million Welfare Recipients?” In The LowWage Labor Market: Challenges and Opportunities for Economic Self-Sufficiency, edited by K. Kaye and D. Smith Nightingale. Washington, DC: Urban Institute Press. Bushway, Shawn D. 1996. “Labor Market Effects of Permitting Employer Access to Criminal History Records.” Working paper, Department of Criminology and Criminal Justice, University of Maryland. Connerley, Mary L., Richard D. Arvey, and Charles J. Bernardy. 2001. “Criminal Background Checks for Prospective and Current Employees: Current Practices among Municipal Agencies.” Public Personnel Management 20(2): 173–183. Craig, Scott R. 1987. “Negligent Hiring: Guilt by Association.” Personnel Administrator, October. Pp. 32–34. Extejt, M. M., and W. N. Bockanic. 1991. “Issues Surrounding the Theories of Negligent Hiring and Failure to Fire.” Business and Professional Ethnics Journal 8(4): 21–33. Freeman, Richard. 1992. “Crime and the Employment of Disadvantaged Youth.” In Urban Labor Markets and Job Opportunity, edited by G. Peterson and W. Vroman. Washington, DC: Urban Institute. Pp. 171–192. Freeman, Richard, and William M. Rodgers III. 2000. “Area Economic Conditions and the Labor Market Outcomes of Young Men in the 1990’s Expansion.” In Prosperity for All? The Economic Boom and African Americans, edited by R. Cherry and W. M. Rodgers III. New York: Russell Sage Foundation. Pp: 50–87. Glynn, T. P. 1988. “The Limited Viability of Negligent Supervision, Retention, Hiring, and Infliction of Emotional Distress Claims in Employment Discrimination Cases in Minnesota.” 24 William Mitchell Law Review 581. Grogger, Jeffrey. 1995. “The Effects of Arrest on the Employment and Earnings of Young Men.” Quarterly Journal of Economics 110(1): 51–71. Hahn, J. M. 1991. “Pre-Employment Information Services: Employers Beware.” Employee Relations Law Journal 17(1): 45–69. Holzer, Harry J. 1996. What Employers Want: Job Prospects for Less Educated Workers. New York: Russell Sage Foundation. Holzer, Harry J., and Sheldon Danziger. 2000. “Are Jobs Available for Disadvantaged Workers in Urban Areas?” In Urban Inequality, edited by A. O’Connor, L. Bobo, and C. Tilly. New York: Russell Sage Foundation. Holzer, Harry J., and Keith R. Ihlanfeldt. 1996. “Spatial Factors and the Employment of Blacks at the Firm Level.” New England Economic Review: Federal Reserve Bank of Boston, May/June. Pp. 65–68. Holzer, Harry J., and Michael A. Stoll. 2001. Employers and Welfare Recipients: The Effects of Welfare Reform in the Workplace. San Francisco: Public Policy Institute of California. Hoynes, Hilary. 2000. “The Employment, Earnings and Income of Less-Skilled Workers over the Business Cycle.” In Finding Jobs: Work and Welfare Reform, edited by D. Card and R. Blank. New York: Russell Sage Foundation. Jensen, Walter, Jr., and William C. Giegold. 1976. “Finding Jobs for Ex-Offenders: A Study of Employers’ Attitudes.” American Business Law Journal 14(2): 195–225. Kling, Jeffrey. 1995. “High Performance Workplaces.” Monthly Labor Review 118(5): 29–36. Kling, Jeffrey. 2000. “The Effect of Prison Sentence Length on the Subsequent Employment and Earnings of Criminal Defendant.” Woodrow Wilson School Economics Discussion Paper 208. Moss, Phillip, and Chris Tilly. 2000. Stories Employers Tell: Race, Skill and Hiring in America. New York: Russell Sage Foundation. Raphael, Steven, Michael A. Stoll, and Harry J. Holzer. 2000. “Are Suburban Firms More Likely to Discrimination against African-Americans?” Journal of Urban Economics 48: 485–508. Stoll, Michael A., Harry J. Holzer, and Keith R. Ihlanfeldt. 2000. “Within Cities and Suburbs: Racial Residential Concentration and the Spatial Distribution of Employment Opportunities across SubMetropolitan Areas.” Journal of Policy Analysis and Management 19(2): 207–232. Travis, Jeremy, Amy Solomon, and Michelle Waul. 2001. From Prison to Home: The Dimensions and Consequences of Prisoner Reentry. Washington, DC: Urban Institute Press. Uggen, Christopher, and Jeff Manza. 2001. “The Political Consequences of Felon Disenfranchisement Laws in the United States.” Paper presented at the annual meeting of the American Sociological Association, Anaheim, CA. U.S. Bureau of Justice Statistics. 1997. “Lifetime Likelihood of Going to State or Federal Prison.” Bureau of Justice Statistics Special Report, NCJ-160092. U.S. Department of Justice. 1999. Compendium of State Privacy and Security Legislation: 1999 Overview. Washington, DC.