MULTIVARIATE-ANALYSIS OF INTERSTATE VARIATION IN FERTILITY OF TEENAGE GIRLS

A multiple regression model was used to determine the correlates of state-to-state variation in fertility of teenage girls. The independent variables were the abortion-to-live birth ratio, contraception/conception ratio, racial composition, adult personal income, percent urban population, and adult educational attainment. The model explained 69 percent of the interstate variation in 1974 fertility. Both the racial composition and the abortion ratio were significantly (p <.01) correlated with fertility. With 1970-1974 change in fertility as the dependent variable, the model explained 80 percent of the variation, with income as the most strongly correlated variable. The contraception/conception ratio was also significant (p <.01) as was racial composition (p <.05). Factor analysis of the independent variables showed that the six variables were well accounted for by three factors representing culture/education status, abortion availability, and contraception availability. The analysis showed that utilization of both contraception and abortion was important in determining the fertility of a state's teenagers. Increased availability of publicly subsidized contraception in low income areas and increased availability of abortion in low income and rural areas might be expected to result in decreased fertility of teenage girls.