In this paper, I address the impact of home-ownership and social renting rates on interregional migration in the Netherlands. I focus especially on their relation with natives’ migration out of the larger and more popular Dutch urban regions. By applying a multilevel social relations model I am able to control simultanously for (i) both region-specific effects of origin and destination, (ii) dyad (regional pair) specific effects, and (iii) the impact of the housing market structure in both the region of origin and the region of destination. I find positive and high elasticities of social renting (0.8) and homeownership (1.8) rates on out-migration, while homeownership rates have a smaller and negative impact (−0.5) on in-migration—pointing to the significant role the housing market has on the crowding out of natives in the larger urban regions. On top of that, I find that regional specific in- and out-migration flows are highly correlated (0.9) just as regional specific dyadic flows (0.8), showing each region’s and dyad’s ideosyncratic migration pattern. Finally, I show that the probabilistic model proposed is able to accurately predict migration flows both within and out-of-sample.
Keywords: Gravity model — housing market — interregional migration — multilevel social relations model — urban regions