Urban homicide: Still the same

We use 1984–1986 data to estimate lifetime risks of being murdered in each of 50 large American cities and then compare these projections to others made earlier from 1971–1972 and 1976–1977 data. We find strong constancy over time in the average urban resident's murder risk (essentially a 1 in 68 chance of eventually being slain). Moreover, we find a high stability in the dispersion of risk by region, race, and city size and in the relative rankings of the 50 cities by murder risk. We comment briefly about such recent phenomena as the proliferation of crack and the resumption of capital punishment.Throughout this paper, we use the words murder, homicide, killing, and slaying interchangeably, meaning in all cases murder and nonnegligent manslaughter as defined by the FBI.