This thesis includes three independent chapters.
The first connects inter-generational asset transfers and education. Using Italian microdata, I show that (i) youths expecting inheritances are more likely to pursue higher education, and that (ii) inheritance expectations and receipts strongly predict bequest intentions. I join these two findings in a simple analytical model where dynastic altruism acts as a patience multiplier. Inheritance-expecting youths bear the short-term costs of education to reap the long-term benefits in terms of earnings – hence ability to fund a bequest. A quantitative life-cycle model embedding this feature shows that (i) heterogeneity in bequest motives explains around a third of the student gap between those between those expecting and not expecting inheritances, and (ii) the link between dynastic altruism and education is stronger when the present value of the returns to education is lower.
The second chapter, co-authored with Andrej Mijakovic, documents a positive association between within-country income inequality and current account balances. This is driven by differences in saving rates arising in response to permanent income inequality. A two-country heterogeneous agent model with non-homothetic preferences reproduces saving rates increasing in permanent income, leading to capital outflows from the less unequal country. In addition, the model accounts for the U.S. exception through the effects of financial liberalisation, and illustrates how inequality variation across and within countries, rather than its absolute levels, shapes global imbalances.
The third chapter, co-authored with several colleagues, investigates how feedback on blood test results affects participation in biomeasure sample collection in a longitudinal study. Using a randomized study within the UK Understanding Society Innovation Panel, we find that offering feedback significantly increases participants’ willingness to provide blood samples, especially in web-based surveys. This shows how simple interventions can enhance data collection, with important implications for improving the quality and quantity of biomedical data.