Recent Migration Patterns in the Developed World: A Clarification of Some Differences between our and NASA's Findings

Abstract

This paper is a reply to two recently published critiques of our finding of a discontinuity in the recent internal migration patterns of Europe, Japan, and North America. Using data from the HAS A Human Settlements Systems Task, Hall–Hay and Gordon both fail to detect any significant narrowing in the differential between the growth rates of metropolitan areas and the growth rates of rural areas in Europe and Japan over the period 1950–1970 (they concede that this difference has disappeared, and has even been reversed in the United States). Our rejoinder here consists simply of a clarification of our own independent research on regional population change in these same countries. Unlike the IIASA project, this research has been confined, in the case of Europe and Japan, to a study of the trends in net internal migration to their politically and economically dominant core regions, for which data are available for the post-1970 period as well. Most of the disagreement over the presence or absence of a discontinuity in the regional population trends in the countries of western Europe and Japan can be explained by this simple difference in the principal orientations of the two studies, the first towards all metropolitan areas in these countries for the period 1950–1970, and emphasizing the total population growth of these areas, the other towards their densest, richest, and generally most important regions for the longer period 1950–1980, and emphasizing net internal migration to these regions rather than their overall population growth. For there is little doubt, as we demonstrate here, that there has been an abrupt and precipitous reduction in net internal migration towards the core regions of many countries in the developed world in the 1970s, though a comparable reduction may not have taken place to all metropolitan areas in the aggregate. Gordon's and Hall–Hay's claim to have rebutted our thesis is thus seen to be based on a misconception of the subject of our study.