abstract: 'This study argues that employment programs for individuals exiting prison can benefit society even if they do not directly reduce recidivism, by helping to identify quickly and efficiently those desisters who are ready to work. We make the following basic claims: 1. Individuals exiting prison have poor work experience, low levels of education, and generally qualify for only low-skill, entry-level jobs. Moreover, the majority will recidivate within 3 years. Employment training programs are designed to ameliorate these deficits, but to date, they have demonstrated only limited potential to improve employment prospects and recidivism risk. 2. Despite a poor track record for employment-based reentry programming, a substantial minority of individuals exiting prison has desisted from crime and has the capacity to maintain stable employment. 3. Growing evidence suggests that this desistance process occurs quickly-almost instantaneously-and is driven by decisions on the part of the individual to change. 4. This type of instantaneous, agent-based change is difficult to predict using static risk prediction tools. As a result, desistance is fundamentally unobservable to employers and others who might wish to identify good employees from the group of people who have criminal history records. In lieu of additional information, one''s true desistance state will only be revealed through time. This situation is a classic case of a market with asymmetric information. 5. Although growing numbers of employers refuse to hire individuals with criminal history records, some are in fact willing to hire from this pool of workers. More might be willing to do so if they could reliably identify desisters. The current legal environment is increasingly hostile to across-the-board bans on hiring individuals with criminal history records without documentation of business necessity. 6. Program participation, completion, and endorsement from a training organization can provide a reliable signal to employers that a given individual has desisted and is prepared to be a productive employee, as long as the cost to program completion is high for those who have not desisted, and low for those who have desisted. Effective signals must be voluntary. Requiring program completion, or graduating all participants, renders the signal useless. 7. Existing evidence demonstrates that program participants (or program completers) do in fact recidivate less often and have better employment outcomes than program nonparticipants (or program dropouts), even in cases where the program does not seem to ``work{''''} in a causal sense. This evidence can be taken to suggest that program completion provides valuable information-a signal-to the labor market. 8. Limited anecdotal evidence suggests that some employers-among those willing to hire individuals with a criminal history record-may already be using completion of employment training programs to identify ``good employees{''''} among the pool of low-skill labor. 9. The development of effective signals could create a net gain to society if, in the absence of signals, employers will largely avoid hiring individuals with criminal history records. Evidence suggests that individuals with prison records are exiting the labor market at higher rates than in the past. 10. The signaling approach is different than risk prediction because it relies on actions taken by individuals to reveal information about them that is, by definition, unobservable. Information about program completion can be valuable even if the program has not caused individuals to change. 11. Other actions besides completion of employment training programs also could function as useful signals in domains other than employment. Policy Implications Reframing the problem of reentry as a case of asymmetric information could potentially have dramatic implications for policy makers struggling to deal with the growing number of individuals with criminal history records, who are increasingly disconnected from the labor market. This disconnection occurs, at least in part, because this group is more readily identifiable through the use of criminal background checks. Although restricting the use of background checks may be infeasible in the current legal climate, policy makers are actively working to create standards for hiring individuals with criminal history records. For example, the Equal Employment Opportunity Commission is currently revising its guidance for hiring individuals with criminal history records. It is hard to overstate the level of interest, by both advocates and employers, in these ongoing discussions. Research insight could be incorporated into government statutes that currently bar individuals with criminal history records from certain types of employment. Indirectly, such guidelines also would help individuals with criminal history records trying to identify themselves to employers as ``good bets.{''''} Key elements of a research plan needed to develop this idea further include: 1. Formalizing the argument with a theoretical model that can be explicitly parameterized. Key elements of the argument depend crucially on factors such as the size of the desisting population, the outcome in the absence of effective signals, and the magnitude of the correlation between the cost of the signal and desistance. Proper specification of the requirements for effective signals in this context could then inform empirical tests of the model. . 2. Empirical testing for evidence that employers are already using factors such as program completion as signals. This testing can include surveys of employers who hire individuals with criminal history records to develop some idea of how they discriminate between individuals with criminal history records. Other potential methods include attempts to compare labor market outcomes of individuals with otherwise similar skill levels, one who has identifiably completed a program and one who has not. Empirical research testing the strength of the link between the concept of crime desistance and work productivity also would be valuable. 3. Calculating the relative costs of programs that provide signals with more traditional risk prediction tools that take advantage of currently available information. Creating these programs to generate signals only can be justified if the additional information generates savings over and above what can be gained by more passive methods. 4. Better understanding the trade-offs between maintaining voluntary programs to generate signals and creating mandatory programs, like Project HOPE, that might enhance rehabilitation. Although signaling and rehabilitation are not competing concepts, the requirement that signals be voluntarily acquired could potentially conflict with mandatory rehabilitation programs. In the short term, it might not be necessary to wait for the completion of this research before policy makers can make progress in this area. We are aware of one set of programs, often called Certificates of Relief, Rehabilitation, or Good Conduct, by which policy makers explicitly identify individuals with criminal history records who have met certain requirements, including program completion. In the strongest cases, these certificates carry with them explicit removal of statutory restrictions on individuals with criminal history records. In our view, these government-run programs are an attempt to create an explicit signal for employers that these individuals have desisted from crime. However, we are not aware of attempts to validate the standards used to qualify individuals for these certificates, nor are we aware of attempts to verify whether these signals work to create better opportunities for the involved individuals. We urge those involved in these programs to redouble their efforts to validate these promising programs.' affiliation: 'Bushway, SD (Corresponding Author), SUNY Albany, Sch Criminal Justice, 135 Western Ave, Albany, NY 12222 USA. Bushway, Shawn D., SUNY Albany, Sch Criminal Justice, Albany, NY 12222 USA. Bushway, Shawn D., SUNY Albany, Rockefeller Coll Publ Affairs \& Policy, Albany, NY 12222 USA. Apel, Robert, Rutgers State Univ, Sch Criminal Justice, Piscataway, NJ 08855 USA.' author: Bushway, Shawn D. and Apel, Robert author-email: sbushway@albany.edu author_list: - family: Bushway given: Shawn D. - family: Apel given: Robert da: '2023-09-28' doi: 10.1111/j.1745-9133.2012.00785.x eissn: 1745-9133 files: [] issn: 1538-6473 journal: CRIMINOLOGY \& PUBLIC POLICY keywords: Signaling; Prisoner reentry; Desistance; Employment programs keywords-plus: RISK; TRAJECTORIES; METAANALYSIS; RECIDIVISM; FUTURE; WORK; AGE language: English month: FEB number: '1' number-of-cited-references: '72' pages: 17-50 papis_id: 7f2afc34a4fc36ab4e46f4a77006c562 ref: Bushway2012signalingperspective researcherid-numbers: Apel, Robert/ABC-4270-2020 times-cited: '165' title: 'A Signaling Perspective on Employment-Based Reentry Programming: Training Completion as a Desistance Signal' type: Article unique-id: WOS:000313553000003 usage-count-last-180-days: '0' usage-count-since-2013: '117' volume: '11' web-of-science-categories: Criminology \& Penology year: '2012'