Politics & Policy

A Better Way to Fight Recidivism

An innovative program gives ex-offenders a fairer shot to rebuild their lives.

Set a man up for a modicum of success, and he might just make a better life for himself.

As many as two-thirds of the 650,000 inmates released from prisons and jails in the United States each year will be re-arrested within three years. America Works is trying to change that.

A new report from the Manhattan Institute looks at AW’s successful approach, which centers around providing more “enhanced job-readiness training and job-search assistance” to nonviolent offenders than the typical re-entry program does. If the welfare-reform successes of the ’90s taught us anything, it’s that the best welfare program is a job. By rapidly attaching such offenders to work, AW aims to minimize the chance that they’ll re-offend later.

As the report puts it:

America Works is condensed into an intense one- or two-week period. It uses a tough-love approach, stressing interpersonal communication and such “soft” skills as time and anger management. It places special attention on teaching practical skills that many former inmates never acquired, such as résumé preparation, search strategies, and interview techniques. And it uses a network of employers, who are open to hiring ex-offenders and with whom it has long-term relationships, to place clients. Its goal is not only to help former inmates find jobs but also to keep jobs, and it provides follow-up services for six months. In 2005, the program provided job-readiness classes to 1,000 ex-offenders, placing 700 in jobs.

Aaron Yelowitz, one of the authors of the report, is associate professor at the department of economics at the University of Kentucky, and he talks below about the research and its findings.


Kathryn Jean Lopez: First of all, two-thirds of those released from jail commit a crime within three years?

Aaron Yelowitz: Recidivism rates are enormously high. The footnote in the paper, using information from Bureau of Justice Statistics, suggests somewhat lower — but still very high — recidivism rates. It’s a real issue, one that probably has something to do with lack of labor-market opportunity.


Lopez: Is it criminal for law enforcement policymakers, religious communities, and neighborhoods to not address this?

Yelowitz: The findings in our study certainly suggest it is beneficial — not just from the ex-offender’s perspective, but the locality’s perspective — to reintegrate the person into society. We find strong evidence that nonviolent ex-offenders are less likely to go back to criminal activity with enhanced job training.


Lopez: What set you out on this study?

Yelowitz: As an economist, I am deeply interested in answering public-policy issues convincingly using careful data analysis. Given my expertise in data analysis, when Manhattan Institute approached me about assuming responsibility for a project that was already started on prisoner re-entry, one that used a randomized controlled trial, I found the idea intellectually appealing. (The first two paragraphs of section IV of the study get into this in more detail.)


Lopez: Why does a Kentuckian care about prisoners at Riker’s Island in New York?

Yelowitz: There are several core lessons from the America Works randomized controlled trial that extend well beyond New York. First, the type of training matters. Comprehensive assistance is more expensive, but, at least in the New York context, [it] led to significant reductions in recidivism, and the cost of those crimes would have been more than the training’s costs. Second, and perhaps not surprisingly, training matters differently depending on the ex-offender’s history. Nonviolent ex-offenders had large reductions in recidivism, but this was not the case for violent ones.


Lopez: What’s the most important finding and consideration?

Yelowitz: As mentioned above, the most noteworthy finding is that enhanced training matters, but only to certain types of ex-offenders, in particular, nonviolent ones. This may not be surprising, since there’s only so much one can do to improve a person’s job prospects in two weeks, and that is a difficult mountain to climb for someone with an especially problematic history.


Lopez: The report finds that “programs offering enhanced job assistance are far from the norm.” Why is this?

Yelowitz: I would speculate it is because the programs tend to be more expensive. Policymakers are focusing on the absolute costs, not the net costs (costs minus benefits). There is a significant benefit — reduced recidivism.


Lopez: What’s unique about America Works?

Yelowitz: The America Works approach is very “hands on” and teaches various “soft skills” in a manner that other programs do not.


Lopez: Who is behind America Works? What motivates them?

Yelowitz: America Works describes themselves this way: “America Works, a New York–based private employment company with operations in seven states and the District of Columbia, has been committed to the mission that employment leads to self-sufficiency and self-assuredness.” Please note that private companies like America Works do get paid for their successful job placements, and that serves as an important economic motivation.

As an irrelevant sidenote, the latest season of House of Cards on Netflix has an “America Works” proposal that is in no way related to this program.


Lopez: [According to your report] “Only 31.1 percent of nonviolent ex-offenders who received enhanced training were arrested during the 18 to 36 months in which they were tracked, compared with 50 percent of similar participants who received standard training.” What’s not working for that 31 percent?

Yelowitz: The unfortunate reality is that — even with enhanced assistance — there will still be a significant portion of ex-offenders who will find themselves back in the criminal-justice system. Clearly there are longstanding issues that job-search assistance can’t deal with that hamper the ability of some ex-offenders to re-integrate into society. Nonetheless, the enhanced job-search assistance certainly moved the needle for nonviolent ex-offenders.


Lopez: “While the benefits to society from averted crimes are very hard to calculate in dollar terms, the study estimated average savings of about $231,000 for each nonviolent ex-offender who received extra help, based on the lower crime record posted by the group as a whole, following training,” the report states. It goes on: “This figure represents a 46-fold return on the cost of the training, not counting impossible-to-quantify benefits to individuals involved, their families, and communities.” Do you have any idea what it means for families?

Yelowitz: The savings mentioned is due to crimes that were never committed because of the enhanced job services. Thus, for the typical family, neighborhoods are somewhat safer than before.


Lopez: You mention child care at one point in the report — how many of those who go through the America Works programs are active parents?

Yelowitz: Table 1 shows that around half of the participants (55 percent) had children. But some of those surely had adult children.


Lopez: Who are the kinds of employers who hire ex-prisoners?

Yelowitz: One limitation of our study is that we have difficulty tracking the labor-market transition for ex-offenders. Instead, we can only track their re-entry into the criminal-justice system. Thus, it is impossible to say from the data we currently have.


Lopez: Why is it good for all of us that they do?

Yelowitz: Although we don’t have data tracking the actual labor-market outcomes, the obvious transmission mechanism to reduced recidivism would be through better legitimate labor-market opportunities.


Lopez: Does this mean anything for violent offenders?

Yelowitz: Our non-findings for violent ex-offenders does matter. It shows that one should strategically target resources. The money devoted to enhanced training for violent ex-offenders didn’t offer the same return as the same money targeted to nonviolent ones. Our study doesn’t address what might work for violent ex-offenders, but it does say that the type of training offered in this randomized controlled trial wasn’t effective.


Lopez: Did anything surprise you? Anything you want to know more about?

Yelowitz: As mentioned, our data analysis came in after the project was already off the ground. Before doing the data analysis, my prior belief was that the program might offer a modest benefit, but perhaps not one that we would be able to detect statistically, given the sample size. The size of the effect for nonviolent offenders is certainly impressive, and it surprised me. I would certainly like to dig much deeper into the labor-market mechanisms and reliance on public assistance.


Lopez: Who must read your report and what section would you send them directly to?

Yelowitz: In an era of tight budgets, tough decisions have to be made on how to allocate money. I would think anyone who has discretion over funds for re-integrating ex-offenders into society would find the study important, and the section showing the effects for nonviolent offenders would be essential.


The Latest