New Population-Level Measures of Economic Dependency and Gender Inequality

Economic dependency is typically understood as a household-level relationship in which shifts toward greater dependency for one member are commensurate with shifts toward less dependency for the other member(s), and equality is defined as equal contributions. We propose a new measure that defines economic dependency separately for men and women and defines gender equality as equal levels of dependency for men and women in the population as a whole (rather than equal contributions in individual households). Thus the new measure defines economic dependency and gender equality as population-level characteristics much like income inequality. Like other population measures, the proposed measure is accessible on one level while containing detailed decomposition properties that show how family formation practices (such as assortative mating, family structure, and marriage patterns) are associated with economic dependency, often in ways that are the inverse of the association with income inequality. A range of alternative and counterfactual estimates are provided but all show very different and non-commensurate levels and trends in economic dependency for men and women between 1970 and 2000, with virtually no change for men during this period despite the large increase in wives’ earnings. INTRODUCTION Although women’s financial contribution to family income has grown dramatically over the past several decades, men’s contributions continue to predominate. According to figures presented below, the average share of white couples’ earnings provided by wives doubled from 1970 to 2000, from 14 to 31 percent. This leaves 70 percent accounted for by men’s contribution. The distribution of married couples tells a similar story. In 2007, men’s earnings exceeded women’s earnings in 85 percent of all married couple families (U.S. Census Bureau 2008). Using a more detailed classification, Raley, Mattingly, and Bianchi (2006: 11) found that “husbands were still the sole (25 percent) or major provider (39 percent)” in the large majority of couples (in 2001), where major provider is defined as contributing 60 percent or more of family income. Wives, by contrast, were the primary or sole earners in 12 percent of couples. When examined over time, even fewer wives maintain their status as primary earners for a period of five consecutive years (Winslow-Bowe 2006). Men’s large contribution to family income raises the questions of exactly how much men’s reliance on spousal earnings has increased, exactly how much women’s has decreased, and whether the two have shifted to the same degree (though in opposite directions). It is difficult to answer this question with existing measures of economic dependency. The most widely used measure – women’s own income as a share of couples’ total income – is defined for the couple as a whole rather than for each partner separately. For example, an increase of 10 1 Similarly, the influential variant of this measure developed by Sorensen and McLanahan (1987: 663) is scaled so that one member in the couple is defined as dependent, the one who contributes less than half of combined income, and the other as independent. The measure ranges from -1.0