*This paper investigates the heterogeneity of take-up rates by social backgrounds. In our model, parents optimally transmit a work ethic to their child, which increases utility in employment to the expense of decreasing utility while unemployed. Once adult, the child either works or is unemployed. When unemployed, individuals have to decide whether or not to take their social benefits up. The model can predict that individuals from higher social backgrounds are less likely to take their benefits up because of a higher work ethic. We find that it is optimal for some workers not to take their rights in terms of unemployment benefits if unemployed. As a consequence, while some non take-up is due to inefficiency in the administrative procedure or to the lack of information, the non take-up due to endogenous preferences should not worry policy-makers. We thus investigate the part of the non take-up rate which is explained by preferences, in order to see whether or not the non take-up is a problem of inefficiency, which should be corrected, or simply a consequence of rational behavior, in which case nothing should be done about it.

**In developing countries with limited markets for credit and insurance, parents generally choose to invest in their future in the form of children as children potentially become a source of income for old age. This paper investigates, theoretically and empirically, the effect of parental investment in children’s education on their old-age support. In particular, do such incentives vary by child’s gender in a society with bride price payments? To do so, we exploit individual data on marriage, monetary transfers and in-kind services between parents and adult children from Indonesia. We explore the human capital variations created by a large school construction program by the Indonesian government during the 70s to estimate the differential effect of children’s education on support to elderly parents. We find that education increases monetary transfers from married children, regardless of their gender. We then focus on two mechanisms explaining the relation between children’s human capital and assistance to elderly. The first one is the impact of education on expected wages, that we name labor-market returns to education. The second is the impact of education on the quality of the spouse and on the bargaining power within the newly formed household, the marriage-market returns to education.