This paper studies how initial economic conditions shape refugees’ long-term integration. I leverage the inability of refugees to time their arrivals to local labor market conditions—due to quasi-random geographic placement and long processing times. Using administrative landing records linked to annual tax data from 1995-2021, I follow them over their first decade in Canada. Refugees initially placed in regions with higher unemployment are persistently less likely to work and earn lower incomes. In contrast, looking at the second generation—who landed as children—I find no lasting effects on education, employment, or social assimilation outcomes such as citizenship and intermarriage. To reconcile these patterns, I show that adverse initial conditions induce parental adjustments: higher unemployment rates at arrival raise rates of secondary migration to stronger labor markets and increase post-arrival educational investments. These parental responses are linked to improved outcomes among the second generation.

We study whether neighborhood peers influence immigrants’ access to high-paying employers. Using longitudinal employer–employee tax data linked to immigration records, we measure employer quality with firm-specific earnings premiums and exploit variation across neighborhoods within larger locales. We find that immigrant-peers raise employer quality, especially for job switchers, while native-peers have little effect. Peer effects are nearly twice as large when peers share a country of birth or mother tongue. Results are robust across specifications, underscoring identity-based neighborhood networks as a channel of immigrant labor market integration.

In this paper we provide new evidence on the health spillover effects of health shocks within couples. Using administrative data from the Netherlands and a matching event-study framework, we estimate the causal effect of experiencing a health shock within a couple on the health of the initially unaffected partner. Our findings reveal a significant deterioration in the partner’s health outlook, characterized by substantial increases in hospital visits, overnight stays, and mortality. The health decline is broad in scope, encompassing higher risk of infections, accidents, and digestive and cardiovascular conditions. This deterioration is accompanied by substantial increases in stress, anxiety and depression for both men and women, as well as sleep disorders for women. These effects are not driven by a heavy caregiving load, financial distress or worsening of health behaviors. On the contrary, the adverse outcomes persist despite suggestive positive changes, including increased exercise for both men and women, and reduced alcohol consumption among women. 

Online Appendix

This paper examines the opportunity cost of pursuing a PhD by tracing the earnings trajectories of graduate students from undergraduate study through doctoral training and into the labour market. Using linked Canadian administrative and census data, we compare PhD graduates to those who complete a master’s degree, to professional degree holders, and to individuals who enter but do not complete a PhD. We find that PhD graduates earn significantly less than their peers early in their careers due to delayed labour market entry. Over time, their earnings recover and eventually surpass those of master’s graduates—but primarily among those who obtain academic positions and remain employed later in life. This “double premium” reflects both higher earnings conditional on full-time academic employment and longer labour force attachment. By contrast, the most substantial penalties accrue to non-completers who withdraw late from PhD programs. Finally, we document worsening outcomes for recent PhD graduates, driven largely by declining rates of academic employment. These findings highlight the central role of career timing, labour force attachment, and access to academic positions in shaping the economic returns to doctoral education.