Looking for a break: Identifying transitions in growth regimes

Economic growth in developing countries is characterized by frequent shifts in growth regimes. Following Pritchett (2000), there is a large empirical literature that has tried to identify the timing of these shifts in economic growth. Two distinct approaches have been developed by this literature. The first is a 'filter-based' approach that identifies growth breaks on the basis of subjectively defined rules, while the second approach is based on statistical structural break tests. The first approach is 'ad hoc' and lacks consistency across studies while the Bai-Perron method that is the basis of the statistical approach has low power, not able to discern true breaks in growth. In this paper, we propose a unified approach that combines the filter and statistical approaches, and avoids the limitations of each approach. Applying our approach to comparable GDP per capita data for 125 countries for the period 1950-2010, we are able to identify a much larger number of plausible breaks in GDP per capita than a pure statistical approach. More importantly, our approach is able to identify more breaks from countries with volatile growth paths, and hence has a larger proportion of breaks from developing countries, compared to other studies that use the pure statistical method of Bai-Perron.