Project Syndicate - What Really Works Against Youth Crime? Oct 15, 2025 Jennifer Doleac

 Project Syndicate 

What Really Works Against Youth Crime?

Oct 15, 2025

Jennifer Doleac

Research shows that, when it comes to deterring crime, the severity of punishment – the length of a prison sentence – matters far less than whether punishment is swift and certain. This finding holds true for adults, but the difference is especially stark for teenagers, given their higher impulsivity.

HOUSTON – After US President Donald Trump declared a “crime emergency” in Washington, DC – pointing to “roving mobs of wild youth” – his administration introduced a package of legislation to overhaul policing and criminal justice in the district, not least by subjecting young people to harsher punishments. The House of Representatives proceeded to pass bills that would restrict judges from giving lighter sentences to young people and allow children as young as 14 to be tried as adults.

The optimal “age of majority” in the criminal-justice system has long been the subject of heated debate, and not only the in the United States. Around the world – including in Australia and the United Kingdom – policymakers are grappling with how harshly to punish young offenders. Are longer sentences for “youth crime” necessary to keep their constituents safe?

While crime is a young person’s game, with those in their teens and early twenties the most likely to offend, youth is typically treated as a mitigating factor in court. Judges are reluctant to lock young people up for long periods, because they know that adolescents are often impulsive and, with their brains still developing, poorly equipped to appreciate the long-term consequences of their choices. A 16-year-old who steals a car today might be a responsible worker and parent in 5-10 years. If that kid is sentenced to years in prison, however, a poor decision can become a lifetime of disadvantage.

This leaves us with difficult, even existential, questions: Are teenagers sufficiently responsible for their actions to merit adult consequences? If punishing them harshly is not justified, do the rest of us simply have to live with their criminal behavior? 

But these questions miss the point. Extensive research shows that, when it comes to deterrence, the severity of punishment – the length of a prison sentence – matters far less than whether punishment is swift and certain. This finding holds true for adults, but the difference is especially stark for teenagers, given their heightened impulsivity and shortsightedness. For a 15-year-old deciding whether to participate in a burglary, the expectation of immediate accountability, however limited, could provide an effective deterrent, but a vague, low-probability threat of eventual jail time is unlikely to mean much.

As it stands, however, a vague, low-probability threat of eventual jail time is as compelling as deterrence gets in most jurisdictions, because “clearance” rates – the share of reported crimes that police actually solve – are very low. In the US, fewer than half of violent crimes result in an arrest. For property crimes, the clearance rate is closer to 15%. In other words, in the vast majority of cases, perpetrators face no punishment whatsoever.

Instead of arguing about how long kids should remain locked up, we should focus on increasing the probability that they will be held accountable at all. Police departments should be given the resources and incentives they need to solve more cases quickly, not encouraged to make a few high-profile arrests. Likewise, courts should be able to ensure that consequences – fines, community service, or stints in custody – follow swiftly, even after a relatively minor offense. What matters for deterrence is that punishments are prompt and predictable, not that they are draconian.

But deterrence is only part of the story. Young people also need tools to help them make better decisions – whether before they offend or after they complete their punishment. Here, too, the evidence marks a clear path forward. Programs grounded in cognitive behavioral therapy (CBT) have consistently been shown to help adolescents slow down, regulate their emotions, consider the consequences of their actions, and make deliberate choices. These skills are essential not only to reduce crime among adolescents, but also to support young people’s development into responsible adults.

CBT is not a magic bullet, but it reliably reduces violent behavior – on the order of 40%. Unfortunately, instead of scaling up CBT programs and preventing impulsive choices that harm others, policymakers continue to channel their attention and resources toward harsher sentencing laws.

The debate over youth sentencing is not going away. But these conversations distract us from what works. The evidence is overwhelming: if we want to reduce crime and support young people, we should spend less time arguing about the “right” sentence length and more time ensuring that the criminal-justice system is working predictably and efficiently, and expanding access to CBT programs. Critics might argue that this approach is “soft on crime.” But what could be softer on crime than a system in which most offenses go unpunished?

Swift, predictable punishment and CBT won’t stop every young person from engaging in violence, and sometimes locking someone up is the only way to keep the community safe. But prison is costly – to families, communities, and taxpayers. We should use it as sparingly as possible.1

Featured

Supply-Chain Economics Beats Tariff Politics

Oct 21, 2025Jun Du

Must Our Globalized World Be a Suicide Pact?

Oct 2, 2025Antara Haldar

When Do We Die?

Nov 1, 2023Peter Singer

What Africans Want from COP30

Oct 23, 2025Carlos Lopes

The Trump Administration’s Latest Assault on Workers

Oct 24, 2025Sandeep Vaheesan

Jennifer Doleac

Jennifer Doleac

Writing for PS since 2025

1 Commentary

Jennifer Doleac, Executive Vice President of Criminal Justice at Arnold Ventures, is the author of the forthcoming The Science of Second Chances: A Revolution in Criminal Justice (Henry Holt and Co., 2026).