Uma investigação empírica da relação entre escravidão e violência letal no Brasil contemporâneo

Abstract

Since the 1980s, the escalation of crime has motivated investigations into the causes of violence in Brazil. While most quantitative studies focus on contemporary factors, sociological literature points to slavery as a historical determinant of the social structure that would engender violence in the country. This study fills the gap in empirical research on the historical causes of violence in Brazil by verifying the existence of positive, significant, and robust correlations between the percentage of enslaved people in Brazilian municipalities in 1872 and municipal homicide rates in 2000. Since the variable representing slavery assumes different values among municipalities in the treated group, the generalized propensity score weighting method is employed, which generates robust correlational estimates of the relationship between continuous treatment and the outcome variable. The sample of municipalities in the Northeast region showed positive but non-significant correlations. Samples from the Southeast region and the combined Southeast and Northeast regions exhibited consistently positive correlations, whose significance levels varied across model specifications. The estimates are positive, significant, and robust for the sample encompassing the entire national territory, controlling for state fixed effects. The results obtained were robust to the inclusion of income inequality controls and its interaction with average household per capita income. Due to different dynamics between inequality, income, and violence in Brazilian municipalities, the interaction term was highly significant. The positive, significant, and robust correlation between slavery intensity in 1872 (measured during its period of decline) and lethal violence in 2000 for the sample comprising all geographic regions provides evidence of spillover effects of slavery's influence on contemporary violence. The results suggest that the pernicious effect of slavery on violence "spilled over" significantly from slave-holding municipalities—the only ones occupied in 1872—to the entire national territory.