An investigation of the spatial correlations for relative purchasing power in Baden-Wurttemberg

The relative purchasing power-i.e., the purchasing power per inhabitant-is one of the key characteristics for businesses deciding on site selection. Apart from that it also plays a major role in regional planning, pricing policy and market research. In this study we investigate the spatial correlations for relative purchasing power of the townships in Baden-Wurttemberg. In particular, changes in relative purchasing power are analysed for three different time intervals, 1987-1993, 1993-1998 and 1998-2004, by means of distance-dependent characteristics like the mark-correlation function, the Simpson indices alpha(r) and beta(r) and by tests on random labelling. It is shown that there are positive correlations for small distances between different townships but that these positive correlations become weaker over the years until they are almost nonexistent (in the sense that hypotheses of random labelling are no longer rejected). A conclusion from this loss of spatial correlations with time is that the relative purchasing power might become more and more purely random. This means that the relative purchasing power in a township is less and less influenced by the relative purchasing power of townships nearby. We further analysed these changes in the Bodensee-Oberschwaben and Stuttgart regions to compare the development of the relative purchasing power in both urban and rural environments.