In developing countries, where formal credit and insurance markets are incomplete or non-existent, households use informal mechanisms to increase their consumption, savings and insurance possibilities. This dissertation studies two of these informal devices: participation in Rotating Savings and Credit Associations (roscas), and internal migration. I mainly follow an empirical approach, using the Indonesia Family Life Survey dataset (IFLS). In particular, I analyze the effect of mothers’ participation in roscas on children secondary or tertiary school enrollment: taking into account the endogeneity of maternal participation in roscas, I find that maternal participation significantly increases the probability of children’s school enrollment. The results show a negative correlation between the unobservables determining rosca participation and the unobservables determining schooling. This can be interpreted as evidence that in households that are more vulnerable to shocks, mothers invest less in their children’s education, while increasing their participation in roscas, as an insurance and savings device. The subsequent chapter analyzes the determinants of internal migration in Indonesia, distinguishing between ‘individual migration’ and ‘family migration’, where the first relates to movements which individuals undertake alone, while the latter indicates movements they undertake with their families. The estimates, performed on men and women samples separately, highlight that individual and family migration have very distinct determinants. For instance, an increase in household wealth (proxied by the value of land owned) does not influence the probability of individual migration, while significantly reducing the probability of family moves. The results also show that, among men, individuals whose households have intermediate levels of income are less likely to move alone, compared to those in the extreme of the households' income distribution. This is consistent with the predictions of the New Economics of Labour Migration. The final chapter (joint with M. Mendola, Università degli Studi di Milano Bicocca) studies the impact of internal migration on the consumption growth of origin households. We are able to disentangle the impact of current and return migration, while taking into account the endogeneity of internal migration. We find that having current internal migrants does not exert a statistically significant impact on origin households’ consumption growth, while having return internal migrants significantly reduces per capita consumption growth at origin. The IFLS data suggest that the negative effect of return migration on consumption growth is likely to be explained by a lower income potential of returning migrants with respect to current migrants and non-migrants.