Rural†Urban Migration Rates and Development: A Quantitative Note

There have been two earlier efforts to estimate how rural outmigration rates evolve over the development process. In one the sample was too small and possibilities of a non†monotonic pattern were precluded. In the other inverted U†shaped migration rate patterns were estimated on the basis of an urbanization logistic curve. It is argued here that estimates based on such a curve would tend to be biased downward at very high per capita income levels. The reason is the relatively small rural populations of most industrialized countries. In such countries relatively high rates of rural outmigration could still be consistent with relatively slow rates of urbanization. Using more direct estimation techniques inverted U†shaped patterns were confirmed, but the turning points occurred much later in the development process and the migration rate estimates were sharply higher (up to five times at the highest per capita income levels) than those of the logistic estimates or even those estimated here on the basis of an LDC only sample. Finally, when the effects of income growth rates on migration rates were estimated directly, the high sensitivity levels found in the logistic study (where growth rates were treated simply as a shift parameter) could not be replicated.