Convergence and spatial patterns in labor productivity

"This study examines convergence in aggregate labor productivity levels, using data on the 67 provinces of Turkey over the period 1975-1990. Markov chains, a nonparametric approach, are applied to characterize the long run tendency of labor productivity. Evidence shows polarization: the observation that the majority of provinces tend to move towards low productivity level, while few move toward high productivity level. These two groups form convergence clubs around the upper and lower tails of the distribution. Furthermore, nonparametric regression results, in conformity with the results obtained from the Markov chain model, reveal a persistent spatial pattern in labor productivity: a pattern of high productivity, which has lasted for the period 1975-1990, around three highly industrialized provinces"-- Economic Research Forum for the Arab Countries, Iran and Turkey web site.