Book Description

This thesis is composed of two related and one independent chapters. The first two chapters use rich data on subjective expectations about migration decisions of highly educated young adults from a lagging-behind region of Spain (Andalusia) that I collected myself during their migration decision-making process. In Chapter 1, I study how expected pecuniary and nonpecuniary factors influence these individuals' migration decisions. To do so, I estimate a life-cycle model of migration choice that takes expected migration duration into account. Crucially, the collected data allow me to separate preferences from beliefs and to distinguish between pecuniary and nonpecuniary factors. Although there is sorting on expected labor market outcomes, my results show that the set of nonpecuniary factors, such as being close to family and quality of social life, play a larger role in choosing whether to migrate. Additionally, counterfactual exercises reveal that a human capital acquisition strategy has a limited effect on temporary migration plans, which are primarily driven by nonpecuniary factors. Chapter 2 studies their self-selection intentions into internal and international migration. I find that individuals who plan to migrate internationally come from the highest end of the grade distribution and are from more privileged family backgrounds relative to the other two groups. Despite being positively selected, students who plan to migrate internationally have the most pessimistic views about their career prospects in their home region. With their migration plans, they expect higher labor market returns to migration than internal migrants. International migrants are more likely than internal migrants to plan a long term migration as opposed to a temporary migration. If individuals follow their plans, my results suggest a future brain drain from the region as well as from the country. Chapter 3, joint with Josep Amer-Mestre and Marta C. Lopes, studies the impact of COVID-19 school closures on differences in online learning usage by regional academic performance in Italy. Using real-time data from Google Trends, we find that students in regions with a previously lower academic performance increased their searches for e-learning tools more than those with higher academic performance. Given evidence from survey and administrative data that lower performing regions were using no less online learning before the pandemic, our results suggest that the COVID-19 shock widened the e-learning usage gap between academically lower and higher-performing regions in Italy.