Working Paper 2016-398
Most non-monetary development indicators are bounded and many of them are presented in terms of either attainments or shortfalls. Whether an absolute approach or a relative approach should be undertaken to assess cross-country convergence of these indicators has been a subject of debate. Revisiting this debate, we provide three arguments explaining why a relative approach is misleading and, instead, an absolute approach is more appropriate. We assess the presence of absolute convergence across countries in several non-monetary development indicators by applying a number of absolute inequality indices. Although we find numerous instances of absolute convergence, these are rarely robust to alternative specifications of indices. We additionally contribute to the available methodological toolkit of convergence analysis by employing absolute-Lorenz curves to assess the robustness of absolute cross-country convergence, which is rarely conducted in the literature, and never to date with absolute-Lorenz curves. We also clarify the relationship between different relevant notions of egalitarian progress and elucidate how progress in these indicators relates to changes in their convergence using absolute Lorenz curves.
Authors: Suman Seth, Gaston Yalonetzky.