46th Congress of the European Regional Science Association: "Enlargement, Southern Europe and the Mediterranean", August 30th - September 3rd, 2006, Volos, Greece
In this study, using econometric methods, we examine whether internal migration in the last 30 years in Turkey has had any effect on the speed of convergence across Turkish provinces. According to standard neoclassical theory, migration across regions is conducive to faster convergence in income per capita: migration occurs from regions with low per capita income towards regions with higher per capita income, thus per capita income in in-migration regions would fall while that in out-migration regions would tend to rise, holding all else constant. In this study, we first test for absolute convergence across 67 Turkish provinces for 1975-2000 using non-linear least squares method. We find that there occurs no absolute convergence, meaning that provinces with initial-low-income per capita had no tendency to grow at a faster rate than provinces with initial-higher-income per capita. This result may be due to the fact that there are significant structural differences among provinces. To test this likelihood, regional dummies and sectoral shares in gross provincial product variables (agriculture, industry and services) are added to the convergence regressions. As expected, when we control for regional and sectoral differences across provinces, convergence across provinces occurs. Lastly, in order to assess the contribution of migration to convergence, we include net migration rates as explanatory variables to convergence regressions. We use the Instrumental Variables method in order to control for endogeneity between growth in per capita income and migration. According to our preliminary results, contrary to the predictions of the standard neoclassical theory, for 1975-2000, internal migration is not conducive to faster per capita income convergence across provinces in Turkey. One probable reason is that the marginal returns to capital in most net out-migration provinces and regions are relatively lower than those in the net in-migration provinces and regions in Turkey. Accordingly, the incentives to invest in capital in net-out migration regions may well be less than those in the net in-migration regions. Faced with lower investment in gross capital formation, and thus lower economic growth, net out-migration provinces and regions may not benefit from out-migration in terms of convergence in per capita income.