50th Congress of the European Regional Science Association: "Sustainable Regional Growth and Development in the Creative Knowledge Economy", 19-23 August 2010, Jönköping, Sweden
We identify the causal effect of a reduction in military personnel on a number of socioeconomic indicators within the periphery of the military base. The base realignments and closures (BRACs) within the German armed forces is an exogenous source of variation that allows for the estimation of the causal effect of a particular demand shock on household income, output, unemployment, and crime rates within a specified buffer zone around each base. The analysis covers 129 communities for the period 2003-2007. Consistent with evidence found elsewhere, we find that these base adjustments have a marginal impact on the local community in which the bases are located. Within the context of the current political discussion over BRACs in Germany, this study provides policy guidance by examining the costs and benefits to the surrounding community. It also gives an indication of how quickly a local economy can adjust to an exogenous reduction in demand. The data on crime were obtained from the Criminal Statistics of the German Police. Other socioeconomic variables are drawn from the federal and state statistical offices of Germany. The data are recorded at the level of the Kreis (NUTS 3). The data on military bases, closures, and core personnel were taken from the Deployment Concept of the Federal Armed Forces of Germany 2004. This includes a variable that spatially situates the base at the Gemeinde level (LAU 2/NUTS 5). To incorporate the information from the home and surrounding Kreise, we use a Geographic Information System (GIS) application to draw a circular buffer around the centroid of the selected Gemeinde. The area of the buffer's overlap with each of the Kreise contained therein is calculated and then divided by the total area of the buffer. This quotient is used to construct a weighted sum of the information in each of the surrounding Kreise for the variables used in the analysis. One advantage of this approach is that, by adjusting the size of the buffer, it allows us to readily test the robustness of the results according to the scale of the analysis.