Unemployment and Homicides: Evidence from Individual Level U.S. Data

Using state-level data, many studies examine how the homicide rate evolves across the business cycle as measured by the unemployment rate. We consider a different approach and utilize individual level data, allowing us to control for individual characteristics of those who died between 1989 and 2004. We then compare those who died of homicide versus those who died due to some other type of death, analyzing if the incidence of homicides rises disproportionately relative to other types of death during different stages of the business cycle. We do not obtain uniform findings. The probability of homicide relative to natural death rose with state unemployment during the late 1990's but no strong associations were found for preceding or subsequent periods.