Spatial Development and Mobility Frictions in Latin America : Theory-Based Empirical Evidence

Using fine-grained spatial data and a dynamic spatial general equilibrium model, this paperassesses the magnitude of mobility frictions in Latin America as well as the effects of their reduction on spatialdevelopment in the region. The results suggest that in most Latin American countries, migration frictions calibratedbased on spatially differentiated initial utility are on average smaller and less dispersed than those obtainedassuming uniform within-country initial utility. A reduction in trade costs due to optimal investments in roadinfrastructure in most Latin American countries increases the present discounted value of real per capita income onaverage in the region by 15.1 percent. This effect is larger than the effects obtained with static quantitative trademodels because of substantial dynamic gains. By contrast, a reduction in migration entry costs in the most productiveand more populous locations in the Latin American countries has a negligible effect on the present discounted value ofthe region’s real per capita income, reflecting the relatively small dispersion in domestic migration frictionsand their relatively low levels in top locations. In both counterfactuals, the welfare increases are significantlylarger than the increases in real per capita output because the reductions in mobility frictions allow people torelocate to areas with better amenities and therefore derive higher utility. These results suggest that trade costs, notmigration barriers, represent a major constraint to theefficient spatial distribution of economic activity and growth in Latin America.