2.2. On Regional Convergence: Two Disciplines at Stake

Two academic disciplines, economics and geography (the neoclassical school of economics and the human geography to be more precise), predict two opposite outcomes on the same topic: regional convergence for the former and regional divergence for the latter. On the one hand, the Solow model suggests that two economies with similar technology, savings rates, and population growth rates should converge to the same capital-labor ratio and level of per-capita income in the long run. This model finds empirical support through numerous publications like Barro and Sala-i-Martin’s (1995). Barro and Sala-i-Martin applied the neoclassical model to the U.S. states. “The overall evidence weighs heavily in favor of convergence: poor states tend to grow faster in terms of per capita income and product, and within sectors as well as for state aggregates.” In that sense, the growth literature validates the income inequality debate fueled by Kuznets (1955).

On the other hand, the new economic geography, similarly concerned with the understanding of the origins and effects of regional growth differences, levels criticism at the way neo-classical theory handles the analysis of income inequality at the sub national scale: “neoclassical theory is based on diminishing returns in which any spatial structure (such as the creation of rich and poor regions) is self-canceling (through convergence)”. The economists of the new economic geography attempted to bridge the gap at the theoretical level, substituting the assumption of diminishing returns to that of increasing returns to scale (Krugman 1995, 1996, 1998). Despite those recent improvements, the opposition against regional convergence persists, as the researchers from both disciplines marshal empirical evidence supporting the increase in regional inequality in the United States from the late 1970s on, and conclude on economic polarization and regional divergence (Danziger and Gottschalk, 1993; Amos, 1988; Braun 1991, Coughlin and Mandelbaum, 1988; Rowley, Redman and Angle, 1991, Fan and Casetti 1994).

The next section looks more closely at the arguments debated in both disciplines.