STATE-LEVEL CLUSTERING OF SAFETY MEASURES AND ITS RELATIONSHIP TO INJURY MORTALITY

This article proposes a social model of investigating injury mortality . The authors hypothesize that (1) state-level laws and regulations on safety cluster together in one or more groupings; (2) groupings of sa fety measures play a significant role in injury mortality; and (3) inj ury mortality is very highly associated with social structural variabl es. There is a clustering of safety policies, with five factors explai ning 67 percent of variance, although no ''master factor'' was discove red. The strongest factor, explaining 21 percent of variance, includes three gun laws and low speed limits before the 1973 federal law. One factor is the most global in that it taps three distinct areas, includ ing helmet laws, minor blood alcohol levels, and smoke detectors, thou gh it only explains 7.5 percent of variance. The only factor that rema ins in a regression for injury mortality is one that includes strong s eat belt laws and strong enforcement of those laws, though in the dire ction opposite to that hypothesized. This factor, along with percentag e rural and environmental spending per capita, is significant for both motor vehicle and non-motor vehicle mortality. For motor vehicle mort ality alone, deaths are higher in states with higher percentages of Hi spanics and fewer people receiving food stamps and AFDC. Many factors that usually predict individual injury mortality do not hold at the st ate level, suggesting the usefulness of looking at social factors for new insights into injury mortality and prevention.