This paper uses the quantitative spatial model with heterogeneous locations linked by costly goods trade, migration and commuting developed in Monte et al. (2018) to address the workings of local labor markets in Germany. One key contribution concerns the analysis of the role of the expenditure share of housing in the economy. We provide arguments that, in accordance with Rognlie (2015), for an economy-wide quantitative exercise, this share should be chosen lower than stipulated in much of the extant research. Our analyses show that the local general equilibrium employment and resident elasticities with respect to local productivity shocks are significantly higher with a lower housing share. Moreover, simple ex-ante observable commuting measures have very little predictive power for these general equilibrium elasticities when the housing share is small. The size of the housing share turns out to play no crucial role for two further results, however. First, employment and resident elasticities are very heterogeneous across German local labor markets, irrespective of the housing share. Second, the housing share has only little influence on the welfare effects and location patterns of counterfactual commuting cost reductions.