The author critically assesses the theoretical and empirical bases of the popular view that marriage is a declining institution in the United States and that this decline is an inevitable concomitant of the growth of women's economic independence in industrial societies. The independence hypothesis has diverted attention from the demographic impact of the substantial deterioration in men's economic position during the past 20 years. Focusing particularly on Becker's specialization and trading model of the gain to marriage, the author investigates several facets of the theory: the historical fit between the major rises in women's employment and trends in marital and fertility behavior; the extent to which the marital behavior observed in recent years matches the marital behavior predicted by the theory; and the degree to which micro-level empirical research supports the theory. The article concludes that both macro- and micro-level evidence for the theory is weak. It suggests that parental sex-role specialization puts nuclear families at risk because there is rarely more than one specialist of each type in a family. Whereas in past times children's work provided a means of maintaining economic homeostasis over the family cycle and improving the family's living levels, this becomes increasingly impractical with industrialization. A more adaptive family strategy is one where the wife works.
