The purpose of this paper is to study how progress in home production technologies and in medical technologies influences gender differences in labor market outcomes and the household division of labor, in an economy with endogenous gender roles. We consider a model in which incentive problems on the labor market play a crucial role in the emergence of unequal labor market outcomes across genders. In turn, the differential labor market outcomes influence gender roles within the household, thus generating a feedback mechanism. Households in our economy efficiently choose how to optimally allocate time of each spouse to the production of a home public good, produced with home hours and market goods. We view children as a crucial component of this good. Biological gender differences, namely women's ability to bear children, will be reflected in the spouses' relative productivity in the production of the home public good. We model this by allowing women's marginal productivity to be greater than men's in the home production function. We assume that female and male workers are equally productive in market work. Women's comparative advantage in home production can be interpreted as tied to the decreased market productivity of women during and after pregnancy. Alternatively, it captures women's higher contribution to the nourishment of children via breast feeding. A consequence of women's comparative advantage is that they will optimally allocate more time to home work than men. The extent of this difference, however, also depends on the degree of substitutability in the spouses' home hours and on other features of the home production technology, such as the contribution of market goods and the level of technological progress in home production. Each individual in our model is also employed by firms to produce market goods, which requires time, market hours, and effort. As argued by Becker (1977) and Holmstrom and Milgrom (1991), if an agent increases the amount of time or attention devoted to one activity, the marginal cost of other activities increases. Hence, we assume that higher home hours increase the marginal cost of market hours and effort. We also assume that effort is private information, while market hours and output are observed. Given the higher marginal cost of market work for workers with higher home hours, the optimal labor contracts will assign lower market hours, effort and earnings to these workers. Hence, women's higher home hours will determine a negative differential in earnings, market hours and effort, relative to men. The incentive problem in the labor market, due to the unobservability of effort, provides an amplification mechanism for the relative productivity differences across genders. As a result, the equilibrium gender differences in earnings will be greater than the relative productivity difference. The feedback mechanism between labor market outcomes and household decisions on home production determines the equilibrium allocation of home versus market hours across genders and the gender earnings differential. We interpret this process as an endogenous attribution of gender roles. Since earnings are decreasing in home hours, the households' opportunity cost of home hours is decreasing in home hours. Hence, the labor market outcome also influences the households' trade-off between market goods versus home hours in home production. We analyze the effects of progress in home production and in medical technologies in the context of our model. There was a rapid diffusion of consumer durables, such as home appliances, in the post-war period, as documented by Greenwood, Seshadri, Yorukoglu (2005). This can be interpreted as a reflection of technological progress in home production. Advancement in medical technologies with an impact on home production occurred prior to this. Specifically, in the mid 1930's, childbirth became increasingly medicalized, leading to a sharp decline in maternal mortality rates between 1936 and 1956, which reduced the physical cost of pregnancy. Over the same period, there was also a rapid decline in the price of breast milk substitutes. This reduced mothers' required contribution to the nourishment of infants. We argue that both these developments determined a decline in women's comparative advantage in home production. Improvements in these medical technologies continued after 1955, however, at a more gradual pace. On this basis, we conclude that the introduction of new home appliances likely played a dominant role after 1955. Greenwood, Seshadri, Yorukoglu (2005) argue that the introduction of home appliances is a critical determinant of rising female labor force participation in the course of the twentieth century. However, the widespread diffusion of household appliances only started in the 1950's. Before World War II, household appliances were primarily adopted by wealthy households, and their labor savings effects mostly resulted in a reduction in the number of servants (Cowan, 1983). Female labor force participation increased from 24.8% in 1930 to 29.8% in 1947. This increase is arguably mostly related to the changes in medical technologies rather than to progress in consumer durables. Female labor force participation continued to increase in the post-war period, rising from 35.7% in 1955 to 46.3% in 1975. The effect of improved home production technologies should dominate that of medical technologies over this period. We use our model to show how these different sources of technological progress impact gender differentials in labor market outcomes and home hours. We find that improvements in medical technologies determine a significant decline in female home hours and gender differentials on the labor market. The endogeneity of earnings differentials and their negative dependence on home hours, however, dampens the equilibrium effects of improvements in home production technologies.
