Long-term multivariate prediction of migration patterns

Abstract

This paper using 1977-1981 data gathered from the Central Bureau of Statistics examines the effect that 4 social indicators have on in-migration and out-migration rates of 56 Israeli cities with a population of 5000 or more and over a 5 year period. In response to previous work done by Berman and Eaglstein in 1983 this study reveals possible differential changes in the predictive structure of these indicators over time. The 4 social indicators: crime rate percent unemployment population size and distance of a city from a major population center were entered into 2 stepwise regressions along with a dependent variable of the in-migration rate and another of the out-migration rate. The results reveal that none of the 4 social indicators reliably explain long term in-migration patterns. However crime rates unemployment and distance significantly and differentially increase prediction ability of long term out-migration patterns. Of the 4 indicators tested crime proves to be the most consistent predictor of out-migration. Contrary to previous findings unemployment rates and distance of a city from a major population center reveal no predictive power with respect to in-migration. With all 4 social indicators some have short term influences and some have longer term effects. The study results provide a useful tool for gaining an understanding of long term effects on migration patterns and can be of value for long term social planning.