I study how displaced workers make joint decisions about where to work and what occupation to pursue. Using French administrative employer-employee panel data from 2002 to 2018, I focus on workers who lost their jobs through mass layoffs to obtain exogenous variation in mobility decisions. I first document that mobility increases earnings compared to staying put. Geographic moves alone raise wages by 9 percent, occupational changes yield 10 percent gains, and combined moves produce the largest effects at 16 percent. Yet 60 percent of displaced workers remain in their original location and occupation, which suggests that mobility costs must be substantial. Through the lens of a dynamic discrete choice model where decisions are nested, with workers first choosing a location and then selecting an occupation within that location, I estimate both geographical and occupational mobility costs. Geographical mobility costs are estimated to be substantial and roughly three times larger than occupational mobility costs. Counterfactual simulations show that subsidizing both types of mobility produces larger effects on worker reallocation and better returns than subsidizing only one type. This suggests that policymakers would benefit from supporting both geographical and occupational transitions rather than focusing on one dimension alone.

We study how pre-industrial climate risk during 1500-1800 influenced historical bilateral inward migration and present-day international migration stocks in Europe. We exploit datasets with high resolution (0.25 and 0.5 degree grids). We find that one standard deviation increase in historical precipitation decreases the share of today's migrants in a given location by 0.045 percentage points and also negatively influences historical migration flows. The results only hold in historically rural locations, suggesting that these long-run relationships are driven by agriculture. We also find evidence that historical climate risk had a persistent effect on current migration patterns through differences in historical prosperity.

While residential mobility has declined in France, worker mobility has shifted to the commuting channel. From 1996 to 2016, average commuting distances increased by 5 kilometers as jobs concentrated in large labor markets while residents did not follow. I study which workers have been most affected by this spatial mismatch.