Abstract
A recent article by Gray, Stockard, and Stone contended that the increase in the proportion of births to unmarried women since 1974 in the United States was not caused by any major change in underlying fertility behavior, but rather by a decrease in the proportion of women who are married, which increased both the population at risk and the birth rate of unmarried women relative to that of married women. In this comment, I argue that the statistical test of this explanation used in the article is invalid because the variables in the analysis are not stationary time series. Correct statistical tests reject the explanation. In particular, I demonstrate persistent, nonstationary deviations from the relationships predicted by the theory advanced by Gray et al. For long periods, the proportion unmarried played only a small role in the changes in the ratio of nonmarital to marital birth rates, contrary to the theory.
