Winner picking in urban revitalization policies : empirical evidence from Berlin

Urban revitalization programs are widespread used but empirically understudied place-based policy instruments which attract increasing attention in recent years. This study adds to the existing literature by evaluating the selection process which led to the designation of a set of five urban revitalization areas in Berlin, Germany in the beginning of 2010. Moreover, the study addresses the question whether the choice of the location of potential urban revitalization areas is influenced by a winner picking strategy, i.e. whether policy makers select areas that would prosper even in absence of the policy to make the policy seem more successful. We attempt to model this hypothesis by evaluating the influence of long run past trends of selected key characteristics (the unemployment rate and the share of residents with immigration background) on the probability of being selected as a target area of the revitalization policy, conditional on the current levels of these characteristics and a comprehensive set of housing and location controls. The empirical evidence is in line with our expectations and is quite stable across different specifications and several estimation approaches: we find evidence that policy makers indeed seem to base their choice to some extent on their expectation of the future performance of the areas. Specifically, they seem to choose the target areas from a pool of areas, which have high levels of unemployment and a high share of residents with immigration background. From this pool, however, they preferably choose areas which have undergone a positive development in the past years. While high current levels of the unemployment rate and the share of residents with an immigration background increase the probability of being selected as a target area, an increase in the six year change of the unemployment rate and the share of residents with immigration background, i.e. a further negative development, decreases the probability of being selected into a target area. This effect is interpreted as a winner picking strategy, as the local authorities do not simply choose the areas which have the greatest need for revitalization, but instead prefer areas which show first signs of a revitalization or gentrification process.