| Abstract: | This research examines trends in U.S. homicide rates at the city level during the so‐called homicide epidemic in the latter decades of the 20th century. Using spline regression techniques to locate structural breaks in city‐level time series, we model the true trends of homicide rates to identify those cities that exhibited a meaningful boom and bust cycle. We then use Tobit regressions for all cities at risk of experiencing a cycle to estimate unbiased effects of theoretically important predictors on the timing of the phase changes. Our findings reveal that larger cities were more likely to experience an epidemic‐like pattern, and that densely populated cities characterized by high levels of deprivation tended to exhibit the rise and fall in homicide rates earlier than other cities. |
