Dynamic Labour Force Participation of Married Women and Endogenous Work Experience

Abstract

This paper presents and estimates a dynamic model of married women's labour force participation and fertility in which the effect of work experience on wages is explicitly taken into account. Because current participation alters future potential earnings, the investment return to work will be an important factor in the current work decision in any forward-looking behavioural model. The model is estimated using the National Longitudinal Surveys mature women's cohort. We use the estimates of our model to predict changes in the lifecycle patterns of employment due to changes in schooling, fertility, husband's income, and the magnitude of the experience effect on wages. We find that although work experience increases the disutility of further work, the effect is overwhelmed by the positive effect of experience on wages, leading to persistence in the employment patterns of these women. In addition we find that an increase in young children and in husband's income substantially reduces participation while increased schooling has a powerful positive impact on participation. This paper presents an estimable structural dynamic model of married women's labour force participation and fertility in which wages are stochastic and work experience or cumulative participation is endogeneous. The model is structural in the sense that the parameters which are estimated are contained in the fundamental relationships governing behaviour, namely the utility function and the constraints. The model is contained in the class of models which describe the life-cycle capital accumulation process with endogeneous labour supply such as Weiss (1972) and Heckman (1976). It is closest in spirit to that of Weiss and Gronau (1981).1 The basic feature of their model and ours is that labour market participation affects future wages, which then affects future participation. The investment return to current work will necessarily be taken into account in any forward-looking optimizing model. As Weiss and Gronau note, estimates of labour supply models have ignored the inherent behavioural dynamics associated with a positive wageexperience profile. There is no adequate empirical treatment of the human capital investment dimension of the labour force participation decision in the literature. Heckman and Willis (1977) have studied a sequential discrete choice model of the labour force participation of married women in a reduced-form framework. Their work