Valuing Lost Home Production of Dual Earner Households

Over the last fifty years, home production output may have changed significantly due to dramatic increases in women's time allocation to market work. It is important to quantify this change: to the extent that increases in GDP derive from new time allocation patterns, failure to measure reductions in home-production output can create a biased impression of progress. This paper proposes a new method for measuring changes in home production activity. The main idea is to compare the accumulation of retirement wealth for single- and dual-earner couples. Assuming that male time allocation stays relatively stable, a household in which the wife works spends some of its income on goods and services that substitute for home production. Purchases of day care, cleaning services, restaurant meals, and so forth leave the household with less resources available for savings. As a result, we expect dual earner households to systematically consume a larger fraction of their measured income. We estimate the value of foregone home production from the differences in savings rates among single- and dual-earner couples using panel data from the Health and Retirement Study. We can then calculate the net contribution of the increases in female market employment to GDP growth. We formulate a life cycle model where women can allocate their time between home production and market work. Working more hours in the market leads to a loss of home production output. We model this loss as a convex function of market hours. We believe that such a specification has a priori appeal: a woman's hours at home may not all be equally valuable, and, if she seeks market work, she should sacrifice the hours with the lowest opportunity cost first. We solve the model analytically and derive the regression equation that is used for the estimation. Although some of our analytical results rely on functional form assumptions, our framework has several attractive features. In particular, the model allows many dimensions of household heterogeneity, including differences in ability to earn on the market and produce at home and differences in wage and family size profiles. Another advantage is that the model can be generalized to incorporate labor force participation and human capital accumulation decisions without affecting our estimates. Our results indicate that the value of foregone home production is modest. For every dollar that she earns in the market, a woman sacrifices 20-30 cents in lost home production. Thus the net gain of female employment is 70-80 percent of the woman's market earnings. We use our estimates to calculate the impact of dramatic changes in female employment on GDP and aggregate home production. Due to a rise in female labor force participation and the closing of the wage gap, the fraction of labor income earned by women increased from 20% in 1959 to over 35% in 1999. The corresponding fall in home production is substantial: we estimate that 5 percent of 1999 GDP consisted of goods and services that used to be produced at home in 1959