Neighborhood effects in economic growth

One of the most striking features of the world economy is that wealthy countries are clustered together. This paper theoretically and empirically explains a mechanism for this clustering by extending the Acemoglu and Ventura model so that it takes real geography into account. Countries close to fast growing economies experience faster growth in aggregate demand for their exports, stimulating faster domestic growth. As a result, a poor country that is surrounded by other poor countries finds it more difficult to grow because its terms of trade shift against it. When this model is estimated on data for 1965 to 1985, we find statistically and economically significant effects. If the typical European country were located in Africa, these terms of trade effects would have lowered its growth rate by almost 1 percentage point per year. The results strongly suggest that it is very difficult to raise income in poor countries without dealing with regional problems.