Metropolitan governance, residential segregation, and mortality among African Americans.

OBJECTIVES: This study tested the hypothesis that the degree to which local government is metropolitanized is associated with mortality rates for African Americans and with residential segregation, which has itself previously been shown to be positively associated with mortality among African Americans.
METHODS: One hundred fourteen US standard metropolitan statistical areas were examined. The primary dependent variable was the age-adjusted, race- and sex-specific all-cause mortality rate, averaged for 1990 and 1991. The 2 primary independent variables were residential segregation, as measured by the index of dissimilarity, and metropolitanization of government, as measured by the central city's elasticity score.
RESULTS: Mortality rates for male and female African Americans were lower in metropolitan statistical areas with more metropolitanized local governments and lower levels of residential segregation. Mortality for male and female Whites was not associated in either direction with residential segregation. White male mortality showed no association with level of metropolitanization, but lower White female mortality rates were associated with less metropolitanization.
CONCLUSIONS: This study suggests the need for further research into whether policy changes in areas not traditionally thought of as "health policy" areas can improve the health of urban minorities.