Summary: County poverty rates are highly correlated with murder rates in the United States (r=0.52), with counties in the poorest decile having about 6 times the murder rate of counties in the least poor decile. The article argues that the correlation between poverty and violent crime is not primarily due to the fact that poverty causes violence, but to other factors creating the correlation. There is reverse causality: criminal charges cause an immediate drop in income and employment. Crime also worsens the local economic situation, causing high costs and forcing productive residents and businesses out of the region, which mechanically concentrates poverty in high-crime areas. Selection is key: people with characteristics that impede economic success are also more likely to commit violent crime, so poverty and violence are concentrated in one place, even without the poverty→violence effect. A small minority of individuals account for a huge share of welfare use and convictions (e.g., 22% account for 81% of convictions in Caspi et al. 2017), consistent with the high concentration of antisocial outcomes. Low cognitive ability predicts violent crime, including in self-reported offenses, and this relationship persists largely within families in sibling comparisons. In Frisell et al. (2012), men in the lowest cognitive ability category had about a 20-30% risk of being convicted of violent crime, compared to <2.5% in the highest category. Mental disorders predict poorer educational and labor market outcomes, even after accounting for sibling fixed effects, and predict violent crime and recidivism, even after accounting for family-related confounding factors. These risk characteristics are heritable, which the article considers as a contributing factor to intergenerational persistence and genetic confounders in the poverty-crime link. A large Scandinavian sibling study found that after accounting for unobservable family-related confounders through sibling comparisons, parental income or neighborhood poverty had virtually zero, non-significant impact on violent crime convictions. A study of more than 500,000 people found that moving to a neighborhood with a different level of poverty changes the risk of violent crime only slightly, with a within-individual impact of about 3% of the cross-sectional association. Studies of released prisoners placed in more- or less-poor neighborhoods found no significant effect of neighborhood socioeconomic characteristics on violent crime recidivism in within-individual comparisons. Reviews of employment and income transfer programs indicate that they may affect property crime, but have little or no systematic effect on violent crime. Macroeconomic data show no consistent relationship between economic downturns/poverty changes and trends in violent crime, including a decline in homicides during the Great Depression and no systematic increase in violent crime globally after the economic shock of the COVID pandemic. Time-based studies of the wealth cycle indicate that financially motivated crime increases when money runs out, but violent crime does not, or if it does, it trends somewhat in the opposite direction. Evidence on lottery wealth (random prize size) does not indicate a significant effect of increased wealth on criminal behavior and excludes effects greater than about one-fifth of the cross-sectional relationship between wealth and crime. The conclusion of the article is that the correlation between poverty and violent crime can be largely or entirely explained by selection and reverse causality, and that the effect of poverty on violent crime is minimal or virtually zero.