A neighbor called on a Saturday morning to tell me Larry had died. The day before, he’d stopped to play with my dog on my morning walk, then given me a big hug goodbye. The day he died, he called me from the roof of his house to tell me how much he loved that dog — then went inside and, for the last time, used.

Larry was a drug dealer. The steady stream of people nodding off on his porch, the socks worn instead of shoes to hide the track marks between their toes, the monthly police raids — none of it left much room for doubt. None of it kept him from his job for more than a day, either.

He was one of the few people on my street I ever talked to about my work. When I mentioned Adverse Childhood Experiences, he’d always say some version of “we all survive shit as kids” — with a small, melancholy catch in his voice every time.

It’s tempting to sort people like Larry into simple categories: addict, dealer, criminal. Poverty, homelessness, and mental illness get treated as though they explain crime on their own, when each one is really a downstream symptom of something further back. Research tied to the CDC-Kaiser Adverse Childhood Experiences (ACE) study has repeatedly found that ACEs — abuse, neglect, and household dysfunction before age 18 — predict a wide range of adverse outcomes in adulthood, including contact with the criminal justice system.^1,2 Here’s what the evidence actually shows about who ends up there, and why prevention has to start a long time before anyone commits a crime.

Poverty and crime don’t move in lockstep the way people assume

Poverty and crime are often talked about as if they’re the same problem. They aren’t, and the relationship between them is far less predictable than that framing suggests. Decades of criminology research have found the poverty-crime link to be inconsistent across neighborhoods, cities, and time periods — shaped much more by inequality, social cohesion, and access to legitimate opportunity than by income alone.³

Larry lived in Section 8 housing. He didn’t drive, and he picked up odd jobs for cash — but he always had enough money for dog treats for every dog on the block. I never asked him how he squared what he did with what he believed, but other people who’ve sold drugs to support an addiction have told me they think of themselves less as criminals than as reluctant pharmacists, providing something people need to avoid getting sick.

For people with few resources and even fewer safety nets, crime isn’t chosen from a menu of equal options — it’s often what’s left after the legitimate ones have already been priced out of reach: a job that requires a resume, an interview, the right clothes, reliable transportation. Public assistance programs exist to close exactly that gap, but they’re perennially targeted for cuts over a fraud rate research consistently puts in the low single digits, even as the same research shows those programs measurably improve outcomes for the next generation.

Untreated mental illness gets blamed for violence it doesn’t explain

Mental illness has become a default explanation for violent crime — a framing that’s usually wrong and actively harmful, because it stigmatizes people who need care rather than suspicion, and it starves the systems that could actually help them. What the data does show is a revolving door: people with serious mental illness are incarcerated at far higher rates than the general population, not because a diagnosis makes someone dangerous, but because jails and prisons have become a default landing place for untreated illness that community care never reached.⁴

The fix isn’t more suspicion of people with a diagnosis. It’s keeping people out of the justice system in the first place — through accessible community-based treatment, crisis-response teams trained to de-escalate rather than arrest, and real screening and care for the people who end up incarcerated anyway.

Thinking back on Larry’s last days, I’ve wondered whether his overdose was at least partly deliberate. He was a fixture in the neighborhood, but there was a loneliness under it — the way he talked about his adult children, underplaying how estranged he actually was from them. They grew up with a father in active addiction; I understand why they kept their distance. Whatever pain that caused him, it likely traces back further than his addiction did.

Homelessness isn’t a moral failure, and criminalizing it doesn’t fix it

On a single night in January 2020, an estimated 580,000 people were experiencing homelessness in the United States.⁵ Homelessness and crime get discussed together constantly, but the relationship isn’t what most people assume. It isn’t illegal to lack a home — though a growing number of cities have quietly made it illegal to exist in public without one, citing people for sleeping, sitting, or storing belongings in places that have run out of other ideas.

For people surviving without stable housing, especially those also managing active addiction, crime can start to look like the only option left. The alternative isn’t more enforcement — it’s permanent supportive housing with no arbitrary time limit, and reaching people before they lose housing in the first place, not after.

Substance-use disorders and unaddressed ACEs feed each other

Larry’s addiction and his crimes were inseparable — that’s true for most people who deal or steal to fund a habit. It’s also true that very few people incarcerated with a substance use disorder ever receive treatment for it while they’re inside, which all but guarantees the same cycle continues after release: national drug-testing data has found that a large majority of arrestees test positive for at least one drug at the time of arrest.⁶ Among adolescents in the justice system specifically, ACE exposure is strongly and repeatedly linked to the severity of substance use and psychiatric symptoms that follow.⁷

None of that is really about willpower. Research on ACEs has consistently found a graded relationship between a person’s ACE score and the odds of later substance dependence — the more adversity in childhood, the earlier drinking or drug use tends to start, and the harder it becomes to stop.⁸ Treating addiction after the fact matters. Treating the childhood trauma underneath it, before it becomes addiction, matters more.

Crime is a complicated problem, made by complicated lives — and it gets easier to address once you stop mistaking the downstream symptoms for the cause. Investment in communities living in poverty, real care for people with mental illness, permanent housing for people without it, and treatment for people in active addiction would all measurably lower crime. So would something further upstream from all of them: stopping the childhood trauma that, for a lot of people, set every one of those dominoes falling in the first place.

Bisma Abedin contributed reporting to this piece.