This is part three of a series on criminal justice reform.

The First Step Act was one of the last laws passed by the 115th Congress and signed by President Trump before the current federal government shutdown. The act, which largely focuses on recidivism reduction through prison reform and some sentencing reforms, is also notable for the generally bipartisan support it received.

In this finalpart of a three-part series on criminal justice reform, we’ll consider some of the implications of the act in light of the motivations for reform (addressed inpart one of the series) and what existing economic research has to say about crime and justice (addressedinpart two).

What are the reforms?

The new law includes a collection of elements that are almost certainly wins for human dignity. Among these are a prohibition on restraints on pregnant prisoners, increased access to feminine hygiene products, and restrictions on the use of solitary confinement for juveniles. But three main areas of the reform—within prisons, in sentencing, and leading up to re-entry—are plicated and merit additional analysis.

First, what are the main reforms?

1. Within Prisons

The First Step Act establishes an incentive system for offenders to participate in “evidence-based recidivism reduction” programming. One of the more consequential incentives it provides is the award of time credits for eligible prisoners. Time credit, when applied, moves offenders out of prisons and into pre-release custody (e.g.,home confinement, residential reentry center, etc.). Offenders who are determined to be of medium- or high-risk of recidivism are credited 10 days for every 30 days of “successful participation” in a qualifying program. For those deemed minimum- or low-risk, the credit increases to 15 days for every 30 days of successful participation. Even prisoners who are ineligible for time credits are allowed other incentives for program participation. (Ineligible offenders include those convicted of particularly violent crimes, kidnapping, terrorism-related acts, mitted with firearms, and other especially serious felonies.) Their incentives may include additional phone and visitation privileges, transfer to a facility nearer to their “release residence,” missary spending limits, additional email access, and other incentives as determined by the Bureau of Prisons.

2. In Sentencing

This federal law lowers the mandatory minimum for repeat drug offenders, allows judges more discretion in sentencing low-criminal history offenders, and makes retroactive a law from 2010 that reduces disparities in sentencing for crack and powder cocaine offenses. For example, before the act, a second felony drug offense would require a federal judge to sentence the offender to at least 20 years but now that mandatory minimum is set at 15 years. For those convicted of a third (or higher) felony drug offense, what was previously a mandatory life sentence is reduced to a 25-year minimum sentence. This increases the discretion of judges at the time of sentencing as does an expansion of the so-called “safety valve” that allows judges more discretion in sentencing low-criminal history offenders. In addition, prisoners convicted of drug crimes who were sentenced before the Fair Sentencing Act of 2010 can now request resentencing, increasing the opportunity for those with crack cocaine convictions to have sentences reduced to parable to powder cocaine sentences, a parity that act initially established.

3. Anticipating Reentry

The act requires that prisoners be placed “as close as practicable to the prisoner’s primary residence” and within 500 driving-miles of that residence. It also requires that the Bureau of Prisons assist prisoners in obtaining both government identification and a birth certificate before release. Both of these changes have a potential impact on post-incarceration es, by both moderating the effects of physical separation munity and family and assisting former offenders in forging new labor-market attachments upon release.

What are the likely effects?

While fewer than 10% of offenders are sentenced and incarcerated at the federal level, on net, I expect the law to have a positive impact on that population of former offenders and for society.

First, the law brings the focus of criminal corrections to recidivism reduction. Considering that the vast majority (over 90%) of incarcerated individuals will eventually reenter society, it is necessary to consider the likelihood of future criminal activity by former offenders.

Second, the law accepts and reflects that incentives matter for former offenders, consistent with the notion that offenders aren’t simply irrational or criminal types. And the incentives structure established by the law is not just limited to the explicit ones surrounding recidivism reduction programming. For example, placing offenders closer to munities and families is likely important to maintaining any existing personal connections that can be beneficial to a re-entering person. Likewise, assisting prisoners in obtaining identification is important for reemployment. We might say these efforts increase the feasibility of staying on the “right side of the law.” Economists would say that where munity, and labor market connections are strong, the opportunity costs of crime and future incarceration are high and, therefore, recidivism is less likely. In other words, families and meaningful work reduce the incentive mitting crimes.

Finally, if sentencing reform is needed to balance public budgets or, as some might say, reflect justice more than vengeance, drug offenses are probably the best place to start for reducing sentence severity. If longer sentences are less likely to reduce recidivism among substance-related felonies than mon crimes, it may be that corrections budgets can be reduced, families can be preserved, and former offenders can get back to work with little to no consequences for future criminal activity by former drug offenders.

Reasons for cautious optimism

But my optimism about the law is a cautious one for two reasons.

First, while the letter of the law repeatedly emphasizes “evidence-based” programming, it is less clear what would constitute high-quality evidence in practice. I am encouraged that the legislation specifiesempirical evidence and requires that two of five members of the independent mittee involved in identifying programs musthave “published peer-reviewed scholarship about risk and needs assessments.” It is critical that mittee members fully recognize the challenges of making predictions plex human behaviors where data is likely unavailable on an array of contributing factors. mittee should consider the large and growing empirical economics literature on criminal justice policy and reform and should employ some of the major players there. Even then, it’s a hard row to hoe, as high-quality empirical work requires massive data, multiple analyses, and significant effort. Time constraints and political pressures are likely to be especially binding in light of the First Step Act’s accelerated implementation goals. (For example, the law requires the “development of risk and needs assessment system” within 210 days of the law’s enactment.)

Finally, it is unclear how large the effect of even high-quality programs will be in light of the particular incentive structure. It is encouraging that the law suggests a thoughtful array of programs, including those that encourage family relationship building, vocational training, and cognitive behavioral treatment. The law also makes explicit a prohibition of discrimination toward faith-based programs.

But even if the “right” programs are put in place, it is not clear that the incentive structures will encourage the “right” people to join those programs. Recall that the time credit is larger for minimal- and low-risk offenders than higher-recidivism-risk offenders, and that the time credit isn’t even available to those convicted of the most serious felony offenses.

Criminal sentences should reflect justice, meaning that longer initial sentences are merited on these grounds for more series crimes. But to the extent that political pressures result in incentives toward rehabilitation that differ by crime- or criminal-type, the programs for recidivism reduction may not reach the offenders for whom the returns—to the offender and, importantly, to society—would be greatest.

Image: President Donald J. Trump Announces the “First Step Act” (Public Domain)