Neoclassical convergence versus technological catch-up : a contribution for reaching a consensus

New macro empirical evidence is provided to assess the relative importance of object and idea gaps in explaining the world income distribution dynamics. Formal statistical hypothesis tests allow us to discriminate between two competing growthmodels: (i) the standard neoclassical growth model similar to that employed by Mankiw, Romer, and Weil (1992), (ii) an extension of the Nelson and Phelps' approach (1966) that emphasizes the importance of technology transfer in addition to factors accumulation. First, the latter model better characterizes international data at an aggregate level. It cannot be rejected as a null hypothesis and is significantly preferred to a standard neoclassical model. Second, robust to sample selection evidence suggests that the high social returns to investment in equipment (as opposed to structure) reflect technology transfer mediated through capital goods. Finally, technological catch-up mostly benefits "socially" advanced economies and largely contributes to the polarization of the world income distribution. -- economic growth ; neoclassical convergence ; technological catch-up ; income dynamics