Title and Abstract
Regional convergence at the county level: The role of commuters
Commuters spend a substantial portion of their income in a different place from where they earn it, thereby constituting an important channel for cross-regional economic dependencies. In this paper, we analyze their role for economic convergence. The commuter flows are inherently asymmetric which implies a stronger shock propagation from large economic centers to rural regions than in the opposite direction. This is in contrast to the symmetric network structure implied by the conventional geographic weights in spatial econometric models that are based on contiguity or geographical distance measures. Motivated on the grounds of the neoclassical growth model, extended for spatial spillover effects, we use German county-level data from 2002 to 2014 to estimate a panel data model that is dynamic both across time and space. We find that the speed of convergence is substantially overestimated when ignoring spatial spillover effects, irrespective of the choice of the spatial weights matrix. The estimates of the spillover effects themselves are smaller using commuter weights than geographical measures due to the restricted feedback to large economic centers.