Can neighbor regions shape club convergence? Spatial Markov chain analysis for Turkey

Abstract

This study explores the impact of neighbor regions on the club convergence for Turkey. Markov chain analyses are augmented by using the spatial lag conditioning. The central hypothesis is that, having a rich (poor) spatial proximity increases the chances to move towards higher (lower) income classes. Our preliminary evidence covers the 1975-2017 period and points out that Turkish regions are not converging on average rather converging to varying income levels. This signals the formation of convergence clubs. Our augmented analyses highlight that the club convergence process is influenced from the income level of neighbor regions. Those regions whose neighbors belong to high income classes have higher chances to move to higher income classes, whereas the peripheral regions linked with the poor ones are getting more isolated. Our results highlight that regional policy framework and local economic activity has an influence beyond the administrative boundaries of regions. This calls for spatially flexible and smart local policies to combat with regional disparities.