Economic Restructuring and the Retreat from Marriage

Our objective is to link recent U.S. marriage trends to changes in the employment and earnings of the marriage-eligible population, welfare benefit levels, and macroeconomic performance, which are played out differently across states. Specifically, we link pooled cross-sectional data from the 1986–1997 annual demographic supplements of the March Current Population Survey to state-year-specific indicators of economic performance from the Regional Economic Information System. We use these data to estimate reduced-form models of the economic, institutional, and demographic determinants of individual marriage outcomes. We also estimate endogenous switching regression models of marriage which take into account women's expected incomes inside and outside of marriage. Our results provide several conclusions. First, the “retreat from marriage” continued unabated during the economic recovery period of the 1990s. Second, our reduced-form results reveal that the recent economic recovery has, on balance, dampened the downward trend in marriage; the percentage married would be roughly three points lower in 1997 if state employment and earnings had remained at their 1986 levels. Third, the effects of state economic restructuring were highly differentiated, with negative effects on marriage, especially among the younger, less educated, and racial minority women. Fourth, results from our endogenous switching regression models provide only modest evidence for the economic model of marriage and call into question the appropriateness of strictly economic explanations of declining marriage.