Over 50% of US homicides by blacks but only 20% of poor are black. Blacks have 9.8x higher unadjusted homicide victimization than whites. At same income levels, blacks have substantially higher homicide rates than whites. Affluent blacks have higher homicide risk than poorest whites. Black homicide rates 13.1 per 100k above $40k income vs whites 2.5 per 100k. Controlling for poverty reduces disparity by 29% but race effect stays at 18.20 per 100k. All SES measures (income, education, labor participation) yield similar persistent black-white gaps. Hispanics' homicide rates near whites' after poverty adjustment. Asians have lowest rates even after SES controls. Black victimization rates proxy offending rates – actual black offending disparity larger due to more black-on-white homicides. In low-poverty counties, blacks die at 12.1 per 100k vs whites 2.4 per 100k. Selection (personality, intelligence, antisocial traits) drives poverty-crime link more than poverty causation. Controlling SES overcontrols for selection effects, overstating poverty's causal role.