BACKGROUND: In November 1986, a Detroit, Michigan city ordinance requiring mandatory jail sentences for illegally carrying a firearm in public was passed to preserve "the public peace, health, safety, and welfare of the people." METHODS: We conducted a set of interrupted time-series analyses to evaluate the impact of the law on the incidence of homicides, hypothesizing that the ordinance, by its nature, would affect only firearm homicides and homicides committed outside (e.g., on the street). RESULTS: The incidence of homicide in general increased after the law was passed, but the increases in non-firearm homicides and homicides committed inside (e.g., in a home) were either statistically significant or approached statistical significance (p = .006 and p = .070, respectively), whereas changes in the incidence of firearm homicides and homicides committed outside were not statistically significant (p = .238 and p = .418, respectively). We also determined that the ordinance was essentially unenforced, apparently because of a critical shortage of jail space. CONCLUSIONS: Our findings are consistent with a model in which the ordinance had a dampening effect on firearm homicides occurring in public in Detroit. The apparent preventive effect evident in the time series analyses may have been due to publicity about the ordinance, whereas the small nature of the effect may have been due to the lack of enforcement.
BACKGROUND: In November 1986, a Detroit, Michigan city ordinance requiring mandatory jail sentences for illegally carrying a firearm in public was passed to preserve "the public peace, health, safety, and welfare of the people." METHODS: We conducted a set of interrupted time-series analyses to evaluate the impact of the law on the incidence of homicides, hypothesizing that the ordinance, by its nature, would affect only firearm homicides and homicides committed outside (e.g., on the street). RESULTS: The incidence of homicide in general increased after the law was passed, but the increases in non-firearm homicides and homicides committed inside (e.g., in a home) were either statistically significant or approached statistical significance (p = .006 and p = .070, respectively), whereas changes in the incidence of firearm homicides and homicides committed outside were not statistically significant (p = .238 and p = .418, respectively). We also determined that the ordinance was essentially unenforced, apparently because of a critical shortage of jail space. CONCLUSIONS: Our findings are consistent with a model in which the ordinance had a dampening effect on firearm homicides occurring in public in Detroit. The apparent preventive effect evident in the time series analyses may have been due to publicity about the ordinance, whereas the small nature of the effect may have been due to the lack of enforcement.