Stop Mandating Rehabilitation Programs. Instead, Build a System that Lets the Best Ideas Win


Criminal justice policymakers often rely on so-called “evidence-based practices,” which aim to use empirical research to dictate criminal justice reforms, strategies, and programs on a national scale. This practice is admirable but misguided. The problems of a parole and probation program in a moderately affluent part of San Diego are not the same as those in rural Arkansas or an urban center in Detroit. No matter how much data we aggregate or how much research we conduct, there is no magic cure. The United States is an extremely diverse place. Like most public policy issues, the best parole and probation reforms will be designed by people close to the problem, who understand the unique challenges their programs face. The job of policymakers is to create a system of incentives which rewards the best ideas no matter their origin and insists on accountability to ultimate metrics. Such a system would allow local leaders to tailor best practices to their unique situations while holding those leaders accountable for achieving specific outcomes.

It makes no sense to spend years and millions of dollars searching for a single perfect solution at every step of a rehabilitation procedure. Instead, we should define the ultimate goals we hope to achieve—reduced crime, lower rates of recidivism, gainful employment for formerly incarcerated people—and allow local leaders to innovate and adapt solutions to the problems in front of them. If we hold government agencies accountable for hitting those goals—and reward them when they do—they will experiment and create better solutions than those mandated by meticulous national studies.

The Scholastic Approach

Experts and academics—including some of our friends at AEI, the Hudson Institute, and elsewhere—often spend decades researching and evaluating government program performance from endless angles in the pursuit of perfect policy recommendations. Researchers get mired in expensive, lengthy randomized control trials to determine whether particular rehabilitative programs are effective. These studies aim to find the most rigorously defensible programs, which can then be rolled out across the country, ensuring a uniform criminal justice system that delivers consistent outcomes.

The reality is that top-down solutions are rarely optimal—regardless of how well supported they may be by the academic literature. Take California’s Parole Supervision and Reintegration Model (CPSRM), for example. Policymakers who started the program in the 2010s drew from leading academic research and cutting-edge practices to design what was, at the time, considered the gold standard in evidence-based rehabilitation. The Parole Reform Task Force was confident that these best-of-the-literature-review policies would dramatically improve rehabilitation and supervision.

Instead, the parole agents who adopted CPSRM performed almost exactly the same as the agents who made no changes at all.

During the same time period, California ran a competing experiment in the form of SB 678. Instead of requiring county probation departments to adopt specific practices or programs, SB 678 created an incentive system that rewarded departments that reduced return to prison rates by whatever means they believed appropriate to their situation. Under SB 678, probation departments experimented with whichever strategies, practices, and programs worked best for their caseloads. Rewards were based solely on each individual department’s ability to achieve the desired ultimate outcome: reducing returns to prison. The results speak for themselves: since 2009, probation officers operating under SB 678 have helped well over a hundred thousand people stay out of prison, and saved the state of California more than a billion dollars.

SB 678 freed California probation departments to create their own innovative solutions, tailored to what was actually happening on the ground. When a new strategy worked, departments reaped a share of the success and invested in expanding use of that strategy. When a practice failed, departments were incentivized to stop using it immediately. This approach is a crucial lesson from the most competitive and highest performing parts of our free society. Business leaders and entrepreneurs know that running an economy top-down has disastrous consequences. Statist approaches to other areas of public policy fare no better. Instead, we attain the best results when we set clear incentives, let people freely experiment, and hold them accountable for achieving measurable results.

The Entrepreneurial Approach

Rather than mandating that a rehabilitative program that works in location A be applied to locations B-Z, policymakers ought to design systems that reward any local program that achieves the desired results. Incentive systems should align practitioners—wardens, officers, specialists, and coordinators—towards the correct ultimate metrics.

This is how my team at the Cicero Institute approaches criminal justice reform. Instead of trying to build the perfect set of rehabilitation programs, we recognize that practitioners in the field inevitably understand how to improve criminal justice outcomes better than distant academics or policymakers. We work with government leaders to construct incentive frameworks that encourage these practitioners to compete and achieve maximum results.

Policy reforms should be founded on strong evidence, and better data collection is required to even be able to track ultimate outcomes in many parts of the criminal justice system. That’s why we were early backers of Recidiviz, an organization that has since raised millions of dollars to create data platforms that will better track criminal justice outcomes. But while pushing for the correct kinds of measurement in other areas, we can make plenty of sorely needed improvements with the evidence already available.

We shouldn’t wait for bureaucrats to one day mandate perfect solutions. Instead, let’s learn from our free society and empower entrepreneurial practitioners to compete and solve the problems in their communities with the best approaches, regardless of their provenance.