Income Inequality and Homicide in North America: A Contextual Analysis of Canada, the United States and Mexico

ABSTRACT This study examines whether income inequality predicts homicide rates in Canada, the United States and Mexico. Drawing on relative deprivation and violent competition theories, we assess whether the inequality–violence relationship varies across national contexts and whether economic growth reduces homicide. Using panel data from 1990 to 2020, we estimate a spatial generalized additive model that allows non‐linear and country‐specific effects of inequality, GDP per capita and economic growth on homicide rates. The model includes spatial smooths, country and year fixed effects and country‐specific functional forms. The results show that income inequality is positively associated with homicide rates only in Canada, offering partial support for theoretical expectations. In the United States and Mexico, inequality does not significantly predict homicide levels. In Mexico, however, current and historical inequality display contrasting patterns consistent with Simpson's paradox. Economic growth does not reduce homicide in any country. The findings indicate that the inequality–violence relationship is neither universal nor linear. National context conditions whether structural pressures become lethal violence, implying that inequality‐reduction policies must address institutional and competitive constraints.