Does economic opportunity reduce crime?

The short answer is yes, with conditions. Across decades of studies — correlational, quasi-experimental, and randomized — a legitimate, decently paid connection to the labor market lowers the kind of crime that is about money and idle time: property crime, and youth violence. The effect is real but bounded. It is strongest at the economic margin, largest for acquisitive crime, and it depends heavily on job quality, not just on employment. A subsidized slot that ends in eight weeks and pays less than the illegal alternative does much less than a stable job in a sector that pays a real wage.

01 — The theoryCrime as a choice under constraints

The modern economics of crime starts with Gary Becker's 1968 paper "Crime and Punishment: An Economic Approach." Becker's move was to treat an offense as a decision: a person weighs the expected gain from a crime against the expected punishment (discounted by the odds of being caught) and against the opportunity cost — the legitimate income and future they would be risking. Raise any of those costs and, at the margin, some people choose differently.

That framing produced a long-running political split. One camp read it as a case for tougher, more certain punishment — raise the price of crime. The other read it as a case for jobs and wages — raise the price of not committing crime, until the legal path is simply the better deal. Decades of empirical work have been, in effect, a test of the second claim: does giving people a better legitimate option actually move behavior?

02 — The correlationUnemployment, wages, and property crime

The aggregate relationship is old and reasonably robust: when unemployment rises, property crime rises. The methodological problem is that crime and unemployment move together for many reasons, so the harder studies use instruments — outside forces that shift local employment but have no other channel to crime.

Raphael and Winter-Ebmer (2001) used state defense-contract spending as an instrument and estimated that a one-percentage-point increase in unemployment raises property crime by roughly 2 to 5 percent, with weaker and less consistent effects on violent crime. Gould, Weinberg and Mustard (2002) went further and separated wages from joblessness: looking at less-educated men from 1979 to 1997, they found that wage trends and unemployment together explained more than half of the movement in property-crime rates — and that the wage of low-skill work mattered at least as much as whether work was available at all.

Put differently, "a job" is not the whole story even in the aggregate data. A labor market that offers low-skill workers a real, rising wage does more to suppress property crime than one that merely offers them a paycheck.

03 — Losing workThe natural experiment of the layoff

The cleanest individual-level evidence comes from mass layoffs, where a person loses a job for reasons that have nothing to do with them personally. Register data from Scandinavia lets researchers follow the same individuals before and after.

A Norwegian study of displaced workers found that criminal-charge rates rose by about 20 percent in the year of displacement, with the increase concentrated in property crime — consistent with people trying to replace lost earnings — and fading over subsequent years. A Danish study (Bennett and Ouazad) put the increase in the probability of committing a crime at roughly 30 percent after displacement, and found something policy-relevant: crime was lower while workers were on active labor-market support and spiked when benefit eligibility ran out. The support wasn't just income; the structure and obligation mattered.

04 — Gaining work or incomeThe interventions that were actually tested

Summer jobs for youth. In a randomized trial published in Science, Sara Heller studied Chicago's One Summer Plus program: about 1,600 students from high-violence neighborhoods, 25 hours a week of minimum-wage work plus an adult mentor, for eight weeks. Violent-crime arrests among participants fell by roughly 43 percent over the following 16 months. Crucially, most of the drop happened after the jobs ended, so it was not simply a matter of keeping teenagers busy. A separate evaluation of Boston's summer youth program found violent-crime arraignments down about 35 percent and property-crime arraignments down about 57 percent in the year-plus after the program.

Transitional jobs for people leaving prison. Here the evidence is genuinely mixed, and that matters. A random-assignment evaluation of New York's Center for Employment Opportunities (CEO) — immediate paid transitional work, then help finding a permanent job — found lower rates of reconviction and reincarceration in the first two years, with the largest effects for the highest-risk participants. But the Transitional Jobs Reentry Demonstration, which tested a similar model across four Midwestern cities, raised employment in the subsidized jobs and then found no effect on any measure of recidivism at two years. The difference appears to lie in implementation, the strength of the bridge to unsubsidized work, and who is served.

Wages and tax credits. Studies of state Earned Income Tax Credit expansions find meaningful reductions in violent crime — on the order of 10 percent — with little effect on property crime. Agan and Makowsky, studying more than 200 minimum-wage increases and 21 state EITC programs against the release records of millions of former prisoners, found that EITC availability reduced recidivism, but only for women. And Crystal Yang, following more than four million people released across 43 states, found that being released into a county with higher low-skilled wages significantly lowered the risk of returning to prison — a one-percent wage increase cutting the recidivism hazard by roughly half a percent, with larger effects for Black and first-time releasees, and the effect concentrated in well-paying sectors such as construction.

05 — Not any jobJob quality is doing much of the work

The reentry literature keeps landing on the same qualification: the level of employment matters less than the quality of it. Yang's data show that only jobs in industries that pay well moved recidivism. Other work finds that high-quality post-prison employment is associated with lower reincarceration while low-quality employment is not. And there is a cautionary result on the other side: when minimum-wage increases price some young workers out of jobs, those workers show higher property crime — the income channel running in reverse.

So the operative variable is not "has a job." It is closer to "has a stable connection to the legal economy that pays more than the illegal alternative and has somewhere to go." That is a higher bar, and it explains why some jobs programs succeed and others don't.

06 — MechanismsMoney, time, and structure

Three channels are usually invoked, and the evidence suggests all three are live, in different mixes for different crimes:

Opportunity cost of time. Work raises what an hour is worth in the legal economy, which raises what is forfeited by an hour of crime or an arrest.

Income. Wages and transfers reduce the material pressure behind acquisitive crime — the layoff studies and the property-crime specificity point here.

Structure and social control. A job imposes routine, obligation, supervision, and an identity. The Danish benefit-exhaustion spike, and the fact that the summer-jobs effect grew after the paychecks stopped, both suggest this channel is real. A Boston Fed analysis of the summer-jobs mechanism found the crime reduction tracked gains in emotion regulation and conflict resolution more than "job readiness" or simple incapacitation. And a purely behavioral program with no jobs component — Chicago's Becoming A Man — cut violent-crime arrests by roughly 45 percent, with benefit-cost ratios estimated between 5:1 and 30:1. Structure and skills carry weight independent of the wage.

What does not respond much: white-collar and many categories of violent crime that aren't about money or adolescent conflict. Economic opportunity is a lever on a specific and important slice of crime, not a general theory of it.

07 — The economicsWhat it costs, next to the alternative

The comparison that matters for policy is against the default tool, incarceration, which is expensive and — at current U.S. levels — delivers a low marginal return in deterrence. Reviews by the Washington State Institute for Public Policy and the 2016 Council of Economic Advisers report both conclude that additional spending on employment, education, and policing is more cost-effective at the margin than additional spending on prison.

Specific numbers: WSIPP estimates in-prison vocational training at about $1,960 per participant against roughly $12,000 in downstream benefit — a return near 7:1 — and residential drug treatment at about 2.7:1. The summer-jobs and behavioral-skills programs above show benefit-cost ratios from the crime reduction alone that range from several-to-one into the double digits. None of these is a windfall, and the reentry results are uneven, but the direction is consistent: at the margin, buying opportunity is a better deal than buying more incarceration.

08 — The through-lineWhat the evidence supports, and what it doesn't

Supported: economic opportunity reduces acquisitive crime and youth violence; job loss increases them for the person who loses the job; wage level and job quality matter distinctly from employment; and pairing work with structure or skills outperforms a bare paycheck. This is one of the more consistent findings in the field.

Not supported: that any jobs program works, that the effect is large for all crime types, or that opportunity alone is a complete substitute for the rest of the criminal-justice system. The programs that fail usually fail on quality and follow-through, not on the premise.

Why neverapply publishes this. Our mission is a labor market where more people are found for good work instead of applying into the void. The research above is part of why we treat that as more than a convenience: at population scale, whether people can get and keep a decently paid job is a public-safety input, not only a personal one. neverapply doesn't reduce crime — but the thing it's trying to make easier, getting connected to legitimate work, is a lever the evidence takes seriously.