Is NAFTA Polarizing Mexico? Or El Sur Tambien Existe? Spatial Dimensions of Mexico's Post-Liberalization Growth

Abstract

Standard parametric tests of convergence cannot capture whether the increased dispersion among state incomes is due to a steepening gradient between north and south, a few hot states randomly distributed, or as an intermediate position, the emergence of convergence clubs. This paper tests for spatial dependency in income levels and growth rates before and after the trade liberalization of 1985. Looking at levels of income per capita, we clearly identify a "South", but there is no "North" or "Center". Beyond the frontline states on the US border, we immediately enter an area as poor as the South and incomes in the central zone itself are almost randomly distributed geographically. Growth shows little evidence of spatial dependency in any period: There is only weak evidence of a South and none of a North. A strong co-movement of Chiapas and Oaxaca emerged in the 1995-00 period, but it had little historical precedent and whether it will continue cannot be foreseen.

Keywords: Regional Covergence, Growth, Spatial Econometrics, Markov Chains, Mexico

JEL Classification: O1, R11, R12, F1