Does Poverty Cause Crime?

Does poverty cause crime? It remains one of the most contested questions in criminology. Large studies reach opposite conclusions depending on their design, and the disagreement runs through the methodology rather than through the politics.

The evidence divides along three lines:

• The raw correlation between poverty and crime is large and undisputed
• Sibling-comparison studies find the association disappears once family factors are controlled
• Natural experiments involving direct cash payments find measurable reductions in offending

The Link Between Poverty and Crime Is Not Disputed

Large population studies show the same pattern. A Norwegian total population study found children of parents in the lowest income decile were twice as likely to be convicted of a violent or drug crime as peers in the fifth decile.

Swedish national data produced a steeper gradient. Children of parents in the lowest income quintile had a 6.78-fold higher hazard of a violent crime conviction than those in the highest quintile, with a 95% confidence interval of 6.23 to 7.38.

Researchers on both sides of the debate accept figures of this kind. The disagreement concerns what produces them.

The Sibling Studies That Question Causation

Sariaslan and colleagues published a Swedish quasi-experimental study in the British Journal of Psychiatry in 2014. It covered 526,167 individuals born between 1989 and 1993, including 216,424 siblings and 262,267 cousins, followed between ages 15 and 21.

Comparing siblings raised in the same family at different income levels removed the 6.78-fold association entirely. The adjusted hazard ratio fell to 0.95, with a confidence interval spanning 0.44 to 2.03.

A Finnish replication published in the International Journal of Epidemiology in 2021 covered 650,680 individuals born between 1986 and 1996 and their siblings. Each $15,000 increase in family income at age 15 was associated with a 23% lower risk of violent crime arrest in the standard analysis. That association also disappeared in the sibling-comparison models.

The authors’ interpretation is that unobserved family factors, both genetic and environmental, produce both the low income and the offending, rather than income producing the offending.

The Cash-Payment Experiment Pointing the Other Way

The Great Smoky Mountains Study has followed 1,420 children in rural western North Carolina since 1993. In 1996, about three years in, the Eastern Band of Cherokee Indians opened a casino and began distributing profits to every enrolled tribal member on an equal per-capita basis, averaging roughly $4,000 per person per year.

The design carries unusual strength. Eligibility depended only on pre-existing tribal membership, and non-Indian households in the same counties received nothing, which created a built-in comparison group.

Akee and colleagues followed the same children into young adulthood and found less criminality for minor offenses alongside improved educational outcomes. Costello and colleagues, publishing in JAMA in 2003, found that behavioral symptoms (conduct and oppositional disorders) fell among children whose families moved out of poverty. Anxiety and depression symptoms did not change.

This is the strongest evidence that giving families money directly, as financial assistance programs do, can change outcomes for children.

Why the Two Bodies of Evidence Disagree

The methods answer different questions, and the sibling design has its own critics.

• Sibling comparisons ask whether income differences within one family change outcomes. A 2022 response in the International Journal of Epidemiology questioned whether that design can rule out a causal effect, since siblings share most of the environment being measured. The study’s authors published a reply defending it.
• Natural experiments ask what happens when income rises for a whole community at once. Community-level effects, such as reduced neighborhood strain and improved local services, do not appear in a within-family comparison.
• Income timing differs. Sibling studies measure ordinary changes in family income during childhood. The casino study measured a sustained, unconditional payment that continued for years.

Neither method is decisive on its own. Treat any source presenting one side as settled with caution.

Property Crime vs Violent Crime

The two categories behave differently, and combining them hides the pattern.

Property crime tracks economic conditions more closely than violent crime does. Theft and burglary respond to unemployment and income shocks in ways that assault and homicide generally do not.

The studies finding no causal link, including the Swedish and Finnish sibling analyses, specifically examined violent criminality. Their findings do not transfer automatically to theft and other crimes committed for gain.

Inequality vs Absolute Poverty

Several researchers argue that relative position predicts crime better than absolute deprivation. Countries with similar poverty rates and different inequality levels can show different violent crime rates.

The competing view holds that inequality measures capture the same underlying social factors as poverty measures, and that separating them statistically is unreliable.

This distinction matters for policy. Raising the incomes of the poorest addresses absolute poverty. Reducing the gap between top and bottom addresses inequality. The two require different interventions.

Where the Evidence Converges

Three findings appear on both sides of the debate:

• Concentrated disadvantage predicts crime more strongly than individual household poverty. Neighborhood-level deprivation carries its own predictive weight.
• Childhood adversity, family instability and parental substance misuse predict offending across income levels. These factors appear in both the causal and non-causal accounts.
• Income alone is rarely enough. The Oxford researchers behind the Finnish study concluded that simply raising family income, with nothing else, is unlikely to work.

What this means for helping families

The practical lesson is that money works best alongside stability and opportunity. That is why effective support combines income with educational assistance that keeps children in school, orphan support for children facing early adversity, and self-reliance projects that help families earn a steady living.

For the wider picture, read our guide on how to fix poverty, or see 10 real ways to help needy people.

Frequently Asked Questions

Do wealthy people commit crime at similar rates?

Higher-income individuals tend to commit different types of offense, including fraud and tax evasion, which appear less often in conviction statistics. Comparisons across income groups are also affected by differences in policing and detection.

Does unemployment increase crime rates?

Unemployment correlates more strongly with property crime than with violent crime. Research attributing a causal role to unemployment faces the same methodological disputes as research on poverty.

Did the US crime decline follow a fall in poverty?

Only partly. Poverty fell in the late 1990s as crime dropped, but crime kept falling for years afterward, including during the 2008 recession when poverty rose. That is one reason simple economic explanations face scrutiny.

Does welfare spending reduce crime?

Evidence is mixed. Some studies find reductions following benefit expansions, while others find no effect once local labor market conditions are accounted for.

What predicts offending most reliably?

Prior offending, childhood adversity, substance misuse and concentrated neighborhood disadvantage predict offending more consistently than household income across the available research.