Nonlinearities in economic growth and club convergence

This paper deals with heterogeneity and nonlinearities in the growth process by developing a two-stage strategy to identify and estimate a club convergence model with threshold externalities. Because of identification and collinearity problems, we develop an entropy-based estimation procedure which simultaneously takes account of ill-posed and ill-conditioned inference problems. First, clubs are identified by introducing a mapping structure in a conditional convergence model. Finally, we estimate a multiple club convergence model, where clubs correspond to subsets of total observations. Our procedure is applied to assess the existence of club convergence for a large sample of countries (1965-2008).