Using matched employer-employee data from Italy, Germany, and Austria, we document a U-shaped relationship between within-job wage changes and separations: workers experiencing both wage cuts and wage gains are more likely to leave their employer. The pattern disappears once wage changes are standardized within tenure groups, pointing to a compositional job-ladder mechanism. To interpret this fact, we develop and estimate a random-search model with worker and firm heterogeneity, learning, persistent match productivity, and on-the-job renegotiation. History-dependent learning drives heterogeneity in wage-change volatility, while renegotiation capital stabilizes wages as workers move up the ladder. Learning frictions reduce aggregate output by 1.3%.

Governments increasingly tie clean-technology subsidies to social and industrial aims, hoping to accelerate decarbonization while nurturing domestic production. France’s 2024 reform embeds a life-cycle environmental score in its electric-vehicle (EV) purchase bonus (based on estimated emissions during the production process). Linking exhaustive French registration records to assembly plants and battery origins, we exploit the sharp January-2024 eligibility shock in a difference-in-differences design. Treated EV models experience 60% decline in sales relative to EVs remaining eligible. We combine these estimates with survey-based stated preferences in order to estimate a nested-logit demand system. Model-based counterfactuals suggest that the reform reduced total EV penetration by about 1 pp, implying slower fleet decarbonization. The reform cut both public spending and consumer surplus and raised tank-to-wheel emissions enough to offset manufacturing-stage CO2 gains, resulting in more emissions. A budget-neutral variant that recycles the fiscal savings into larger bonuses for eligible models would restore the lost electrification, and would cost about €4,500 of consumer surplus per reshored battery while generating positive net environmental gains.