Working Wives and Dual-Earner Families.

Whaples, Robert

A century ago, only one married woman in twenty participated in the labor market. In 1940, fewer than one married woman in seven was working outside the home. Today more than half are. Because of this, the "dual-earner family is now more prevalent than the one-earner, with husband and wife employed in almost 60 percent of all married-couple families. Only 17 percent of married couples are one-earner, with the rest having other or no earners". In Working Wives and Dual-Earner Families, Rose Rubin and Bobye Piney seek to comprehensively analyze the differences between dual-earner and one-earner families. They review a vast body of economic theory and empirical evidence, adding new empirical findings of their own. Chapters are given to labor force participation, tax policies, earnings levels, expenditure patterns, wealth levels, and the distribution of income.

The most solid contribution of the book is in laying out the economic theories surrounding dual-earner families in a simple, straight-forward manner and in reviewing previous research on the subject. These sections will be valuable to students and other newcomers to the field. Among the important findings reported are that for "most retired wives who have been in the labor force, the 50 percent of the husband's (Social Security) benefits is higher than their own earned benefits ... In 1982 only 40 percent of ever-married women received benefits derived from their own work and contributions" to Social Security; that most of the differences between the spending patterns of one-earner and dual-earner families relate to income level rather than to wife's work status ("most researchers have found significant differences only in expenditures for child care and transportation"); that the 1990 median income of one-earner families, $30,265, was only two-thirds that of dual-earner families, $46,777, and the gap is widening; and that for a woman earning $20,000 per year, whose husband earns $30,000, net earnings (gross earnings minus taxes and work-related expenses such as transportation and child care) are as little as 35 percent of gross earnings when there are two children in day care.

The book's greatest weakness is its lack of ambition to cover more ground. The authors are not very serious about their claim to examine historical trends. For example, the work of Claudia Goldin (e.g., Understanding the Gender Gap: An Economic History of American Women, New York: Oxford University Press, 1990) is not cited in the section on "Growth in Employment of Women and Wives." Despite its title, "families" are almost missing from the study. At least two additional chapters would be welcomed - one on child care arrangements and another, using recent time diary studies, on the allocation of time within the household. The role of the family could also be examined in a broader discussion of women's choices between part-time and full-time work. Finally, the section titled "Policy Implications" encompasses only one page. A bit underweight.

In addition, there are problems with aspects of the authors' empirical work. The authors have severely limited themselves by basing their empirical findings almost exclusively on the Consumer Expenditure (CE) Surveys of the BLS. In some places they have used this data uncritically, such as when they comment that the share of health care expenditures decreased between 1972-73 and 1984. The CE surveys omit much of households' health care expenditures because they are "paid for" by the employer. Similarly, the discussion of asset levels is misleading because the CE figures which they report measure only checking accounts, savings accounts, stocks and bonds, and U.S. bonds. The authors realize that something is wrong with their table which contends that average assets (in constant dollars) declined by over two-thirds between 1972-73 and 1986 for both full-time working wife (FWW) and nonworking wife (NWW) families. Use of alternative data sets would clear up the confusion, show increasing wealth levels, and allay suspicion about comparisons between FWW, NWW, and other families that use this flawed measure. A third instance of uncritical data use is in measuring real wage trends. The CPI is known to systematically overstate inflation. Thus, using it understates recent real wage increases. Unfortunately, this shortcoming is shared by many economists. When will we learn?

Finally, in the empirical sections many of the model specifications are inadequately discussed and defended. For example, echoing other studies, Rubin and Riney find that "if nonworking wives begin to work full-time, the degree of equality of family income distribution is expected to increase". However, they reach this conclusion by comparing the Gini coefficients of a number of population subgroups, rather than the more conventional technique of examining changes in the whole-group Gini. Why have they chosen this method? More explanation is needed. More importantly, the discussion of income inequality ignores their earlier model and simulation which made the strong case that the net earnings of working wives are considerably below their gross earnings. By incorporating the concept of net earnings into their Gini coefficient calculations, the authors could have made a much more substantial contribution to the understanding of how working wives effect economic inequality.

Robert Whaples Wake Forest University