A study of the determinants of migration: the case of Greek migration to West Germany 1960-1982

In the period following the end of World War II, Western European countries have experienced rapid economic growth. In the second half of the fifties, labour shortages emerged, obliging developed countries to have recourse to foreign labour in order to maintain high growth rates. During the sixties, bilateral agreements between European industralised countries (West Germany, France, Sweden, Belgium ... ) and less developed Mediterranean countries (Spain, Portugal, Greece, Yugoslavia, Turkey ... ) produced large-scale migration in Western Europe. The main bulk of Greek emigration has been directed towards West Germany, reaching a peak in 1971, while the reverse flow of returning migrants exceeded emigration from 1974 up to 1981. Data concerning these two flows, from 1960 to 1982, give us the opportunity to test the determinants of both outward and return migration using models based on the Neo-classical, the Keynesian and the Human Capital theories. Under the Neo-classical assumptions about labour and product markets, migration of labour is explained by income differentials prevailing between two regions. The Keynesian model adds unemployment as a cause of migration. Because of the static framework concerning the above models, expectations about future income resulting from migration have been introduced to make the model dynamic. Under the Human Capital theory, migration will occur if the present value of the expected benefits exceeds the present value of the expected costs resulting from migration. Empirical tests of the above model's using OLS or other methods attempting to overcome econometric problems, are presented. Logarithmic forms of emigration equations present the best results. The logarithmic form implicitly assumes that emigration is of a Cobb-Douglas type function. Because of the weaknesses concerning Cobb-Douglas type functions, a translog type emigration function is determined and tests are applied in order to find the best estimation provided by the two functions. Next, we consider migration decisionmaking at the level of an individual who seeks to maximise his welfare in conditions of uncertainty. Introducing utility functions and risk coefficients, the maximisation of welfare yields a stochastic migration function. Furthermore, we examine the migration decision in a binary choice model context. The potential migrant has to decide whether to migrate or not, and an application of the binary logit probability model enables us to estimate the probability that an individual drawn at random from the population will choose to migrate. Finally, we estimate emigration and return migration functions together with employment (or unemployment) and wages functions in a simultaneous equations system in order to avoid simultaneous bias resulting from interdependence between migration and other variables used as explanatory in the previous models.