The social development of Brazilian States has generally not been followed by a reduction in homicide rates. A positive exception is the State of São Paulo, which experienced a clear reduction in its homicide rate from 1999 to 2005. There was a marked increase in the incarceration rate that preceded and then coincided with an equally important decline in the homicide rate. A bivariate Granger causality test showed that past incarceration rates helped predict homicide in the State of São Paulo from 1999 to 2005. Although it was not possible to conclude that this relationship is causal, a clear temporal association was observed, consistent with the interpretation that the increase in the incarceration of criminals had the effect of disabling or deterring crime in São Paulo. Other factors not assessed in this study may have caused the reduction in homicides in the State of São Paulo from 1999 to 2005.
Homicide; Prisons; Prisoners; Violence
