Ranking German regions using interregional migration - What does internal migration tells us about regional well-being?

This research project ranks German regions with help of interregional migration data instead of gross domestic product, household incomes, unemployment or quality of life estimates. Therefore we estimate regional utility differentials for German states and planning regions following the approach of Nakajima & Tabuchi (2011). The estimation of regional utilities with respect to aggregate interregional migrate are based on a discrete choice of model of migration decisions. Given some assumptions on the nature of migration costs and a structural model using double exponential extreme value function, the specification allows omitting distance-related migration costs in the estimation equitation. In contrast to other migration studies an overall standard of living / regional utility measure is derived. The purpose of this paper is not to explain observed migration with respect economic determinants or non market goods at regional level. These estimates and derived regional rankings are strongly correlated to other standard rankings of well-being in case of Germany. In addition, the development of regional utilities shows considerable variation over time for single regions. Not surprisingly, East German regions have relative low utilities and ranks compared to West Germany. But the results also indicate between substantial utility differentials between West German regions. We also do not find evidence for regional convergence for Germany as a whole, given that our utility measures incorporate economic potential and non-market amenities like natural amenities or publicly provided goods. However, East German regions have considerably improved for the last twenty years but this result may be misleading given the massive outmigration of the past. Thus, the reduced outmigration from East Germany may simply correspond to a lack of potential migrants today. A potential shortcoming of the approach is that we do not take account of selective migration explicitly. Further we ask whether observed the migration and thus utiliies are consisten with a spatial equilibrium. This is rejected here by comparing compensating income differentials and actual income differential. This result is expected with respect to German reunification and also missing evidence for a spatial equilibrium in other country studies not surprising. We think the use of migration date instead of regional income or output data as proxy is a promising approach. Especially, migration data has a high quality and need not to be imputed at lower regional level like GDP data.