On the Mutual Dynamics of Interregional Gross Migration Flows in Space and Time

Abstract

This paper applies spatial dynamic panel data models to analyse the labor market dimension of interregional population flows among German federal states in the period 1993–2009. Making use of recent improvements in the estimation of space-time dynamic panel data models and the computation of meaningful scalar summary measures for the obtained regression coefficients, the empirical results show that the network of German interregional gross migration flows is subject to serial and spatial autocorrelation patterns, which affect the interpretation of the considered regional labor market signals. Using a time-dynamic spatial Durbin model as preferred empirical specification, the results indicate that regional differences in real income growth, the labor participation rate and real-estate prices impact on interregional out-migration flows. The estimated coefficients signs of the obtained space-time summary measures thereby hint at the validity of the neoclassical migration model in predicting gross out-migration flows. In order to take a closer look at the dynamic evolution the direct and indirect network effects over time, cumulative multipliers for a time horizon of up to 10 years have been computed for each variable. The results show that the speed of adjustment of the migration response towards the long-run impact of labor market signals is fast and mostly occurs within the first 3 years. Regarding the interplay of the direct and indirect effects, the estimation results uniformly hint at additive linkages for all variables. Overall, the obtained results underline the importance of a decomposition of the total effects of labor market signals on interregional migration flows by means of their spatial and temporal network dynamics.