No 2019/10 Why bother about region-specific growth patterns and how to identify them?

Abstract

Understanding the causes of regional growth has been of key concern for policy makers and scholars in economic geography and regional science. Regional growth models estimate the effect of various regional factors on the growth of an average region. However, these models leave a large share of regional growth unexplained. This is to be expected, as the embeddedness of regions at various spatial scales from the local to the global invariably leads to growth that is highly region-specific. However, existing quantitative approaches to regional growth are not able to cope with this, treating outliers in growth models as stochastic noise rather than as cases of empirical interest. The paper proposes a method to carve out the region-specific growth component in regional growth models and illustrates this empirically using data for Sweden from 1990-2016. We find robust patterns of periodic region-specific growth outweighing the effect of generic structural factors. This has important implications for how we should think about and empirically address regional growth.