This thesis consists of a summary and four self-contained papers. Paper [I] examines whether fathers influence the time their children spend in subsidized child care. Two non-nested models of family child care demand are estimated. The parameter estimates indicate that several characteristics of the father are associated with the time his child spends in child care. J-tests and bootstrapped J-tests also show that a model where the father’s characteristics are excluded can be rejected in favour of a model where his characteristics are included. Paper [II] considers the effects of the Swedish child care fee reform on public expenditures and taxation in the municipalities. A difference-in-difference approach is employed where outcomes are compared with respect to the municipalities’ pre-reform fee systems. The results show that pre-reform characteristics determine taxes and child care expenditures in the post-reform period. It is also found that changes in child care quality were not connected to the pre-reform systems characteristics. Paper [III] provides evidence of the effect of college quality on earnings in Sweden. The results suggest that the link between college quality and earnings is weak. A small positive effect is found for individuals that are likely to work full time. Controlling for region of work affects the estimated effects, indicating a correlation between choice of college quality and choice of labour market region. In Paper [IV], earnings differences between transfer and non-transfer students are analysed. The results show that earnings, during the first years after leaving the university, are significantly lower for students who change universities compared to students who do not change. The earnings differences decrease significantly over time and over the earnings distribution.