Previous literature shows that internal migration rates are strongly procyclical. This would seem to imply that geographic relocation does not help mitigate negative local economic shocks during recessions. This paper shows that this is not the case. I document that net in-migration rates decreased in areas more affected by the Great Recession. Using various IV strategies that rely on the importance of the construction sector and the indebtedness of households before the crisis, I conclude that internal migration might help to alleviate up to one third of the effects of the crisis on wages in the most affected locations. This is due to a disproportionate decrease in in-migration into those locations rather than an increase in out-migration. More generally, I show that differences in population growth rates across locations are mainly explained by differences in in-migration rates rather than in out-migration rates. I introduce a model to guide the empirical analysis and to quantify the spill-over effects caused by internal migration.