Criminal justice policy is often shaped by political rhetoric, moral instinct and competing ideologies. Economist Jennifer Doleac says that approach misses the most important question: what actually works.
In a recent interview about her book The Science of Second Chances, Doleac said policymakers should rely less on assumptions and more on measurable outcomes when designing reforms. She believes that many well-intentioned ideas fail, some create unintended harm and durable progress requires testing policies against real-world evidence.
“We need to be trying lots of things, but we need to be humble about it and recognize that most of the good ideas we have aren’t going to work,” Doleac said. “Human behavior surprises us and the problems we’re trying to solve are really complex.”
She said that reality should not discourage reform, but insteadshould change how reform is pursued.
“My general pitch to policymakers is aim to fail fast rather than pretending you’re not going to fail at all,” she said.
Doleac, now with Arnold Ventures, entered the field as an economist interested in how empirical tools could improve public policy. She said her early focus on law enforcement DNA databases came from noticing that states had widely different rules on whose DNA could be collected and stored.
Those differences created what researchers call a natural experiment — a real-world setting where policy variation can reveal cause and effect. Rather than relying only on theory or simple correlations, economists often study how institutional differences shape outcomes across otherwise similar groups.
“You find that coin flip that we would use if we were able to run our ideal experiment, but we find that coin flip in the wild,” she said.
One of Doleac’s central arguments, following from recent scholarship, is that certainty of consequences matters more than severity of punishment. She said criminal justice systems have often overinvested in sentence length while underinvesting in solving crimes quickly and consistently.
“If we want to deter crime, the best way to do that is to increase the probability of getting caught or the swiftness and certainty of consequences,” she said. “It doesn’t really matter so much what those consequences are.”
According to Doleac, people most at risk of offending are often not making long-term calculations about future punishments. A small chance of a severe sentence may carry less deterrent effect than a high chance of immediate accountability.
Perhaps among the more surprising findings in the book is her research on DNA databases.
Using state policy expansions, researchers compared similarly situated people convicted just before and after new collection laws took effect. The key difference was whether they entered the database and therefore faced a greater likelihood of being identified if they reoffended.
The result, she said, was substantial.
“What happens is about a 40% reduction in recidivism for people who are in the database,” Doleac said. “So it’s a huge impact.”
She acknowledged concerns about privacy but said many people misunderstand what is actually stored. Rather than health or family-trait data, she said the system uses identifying markers that function more like a unique code for matching purposes.
Doleac contrasted that kind of targeted prevention with incarceration, which she described as costly and often overused.
“If our goal is reducing crime, prison should be the last resort because it’s so expensive,” she said.
She noted that communities may still choose incarceration for reasons such as punishment or moral accountability, but she distinguished those values-based judgments from empirical questions about public safety returns.
The interview also addressed one of the most prominent reentry reforms of the last decade: Ban the Box laws, which remove criminal history questions from initial job applications.
The policy was designed to help people with records get interviews before being screened out. But Doleac said the evidence shows the results were often the opposite of what supporters intended.
She pointed to the concept of statistical discrimination, in which employers denied direct information use proxies instead. In hiring, that can mean making assumptions based on race, age or educational background.
“If an employer is really worried about a recent conviction, they might discriminate then broaden the discrimination to young Black men with lower education levels,” she said.
According to Doleac, studies found Ban the Box increased racial disparities in hiring and did not improve employment outcomes for people with criminal records.
“It broadened the discrimination instead of reducing it,” she said.
She said newer Clean Slate record-sealing policies have also shown limited employment effects so far.
“The evidence we have so far shows it has zero impact on employment,” she said.
Instead of focusing only on hiding information, Doleac said reformers should identify what employers fear and solve those specific problems through better training, incentives or risk-sharing systems that make hiring more attractive.
Another area she highlighted was the connection between environmental conditions and crime. She cited research linking real-time air pollution exposure to worse cognitive functioning, lower academic performance and increases in violent behavior.
In one study involving Chicago, she said shifts in wind direction carrying highway exhaust were followed by shifts in where violent crime increased.
“The violent crime is following the wind direction,” she said.
She also pointed to research suggesting air filters can improve student test scores, indicating that relatively inexpensive environmental interventions may have broader long-term social benefits.
For Doleac, those examples show that progress does not always require sweeping revolution. Smaller gains at multiple stages of the system can add up.
“We don’t have to wait for big structural change,” she said.
The title of her book, she explained, refers not just to release from prison but to intervention opportunities throughout the system — during first contact, supervision, incarceration and reentry. A real second chance means changing trajectories, not simply repeating failed responses.
She also emphasized that many people naturally age out of crime, which raises questions about extremely long sentences.
“You’re locking someone up for something they did when they were 20, and then by the time they’re 50, they’re a totally different person,” she said.
Near the end of the interview, Doleac returned to what she sees as the greatest obstacle to smarter policymaking: fear of admitting when an idea did not work.
“The first thing we’re trying is at the very least not the best solution,” she said. “Our goal is to solve the problem, our goal is not to implement the policy.”
Even amid polarized debates over crime and punishment, she said evidence can still create common ground.
“There’s a lot of other stuff we could be doing that both sides could get behind and that we could implement in the real world,” Doleac said. “So I am super optimistic about this space.”
Follow the Vanguard on Social Media – X, Instagram and Facebook. Subscribe the Vanguard News letters. To make a tax-deductible donation, please visit davisvanguard.org/donate or give directly through ActBlue. Your support will ensure that the vital work of the Vanguard continues.
