Working paper no. 1995-25
This paper investigates the effect of parental benefits on the timing and spacing of births and on employment dynamics of Swedish women. Using microdata on wages and incomes, I estimate a reduced-form multistate duration model of the bivariate life cycle fertility and employment process. The paper provides an example of the application of empirical procedures to develop and evaluate multistate hazard models. Estimation results are mixed with estimated wage effects generally consistent with theoretical predictions while estimated effects of the benefits are not. The evaluation procedures offer valuable diagnostic information and suggest several avenues for future research.