We study how tax incentives affect the return migration of high-skilled expatriates to their home country, exploiting a generous income tax break for returnees in Italy. Using administrative data and a Triple-Difference design, we estimate a migration elasticity to the average net-of-tax rate just below one. Responses are sizable across the upper half of the earnings distribution, indicating that tax-induced migration is not limited to top earners. A cost-benefit analysis reveals that, while costly in the short term, the scheme pays for itself in present value if a sufficiently large fraction of returnees remains after the scheme elapses.

Competitive, project-based allocation mechanisms are a central feature of European Union public spending, governing the distribution of cohesion funds to regions, firms, and public institutions. We study how this allocation model performs when applied to schools by evaluating the impact of competitive school grants on student achievement. Between 2014 and 2020, roughly €3 billion of EU Structural Funds were allocated to Italian schools through calls that ranked applications and funded projects above a cutoff. Combining administrative data on standardized test scores with detailed information on funded projects, we exploit ranking thresholds using regression discontinuity and staggered difference-in-differences designs. Schools that marginally win grants for educational technology increase test scores, with larger effects for disadvantaged students. By contrast, grants for extra-curricular activities show no average school-level effects. At the student level, only participation in mathematics- or language-focused courses improves achievement, while non-focused activities yield no gains. Overall, cohesion resources raise performance when allocation mechanisms prioritize academically targeted investments.