Determinants Of Migration: Inter-european Migration, 1950-1989

It is commonly assumed that governmental regulations determine the flow of international migration streams. Time-series techniques are used to analyze the migration flows to Belgium from France, Germany, Italy and the Netherlands in the post-war period 1950-1989 in order to estimate the impact that the migration policy change in 1968 had on these flows. A survey of the predominantly cross-sectional migration literature revealed that substantial contradictions are reduced when studies are separated according to their power to detect the significant push and pull factors. The preliminary findings for the time-series analyzed indicated that these series required differencing once before achieving stationarity and that they were not cointegrated. Subsequently, the data were analyzed using both OLS regression and Seemingly Unrelated Regression. Contrary to expectations, these migration streams appear unaffected by the 1968 immigration policy change; rather the results indicated that the strongest and most consistent of determinant of the migration streams examined was the unemployment level in Belgium. Unemployment at the origin had a smaller and less significant impact. The measures used to capture the economic and social quality of life, Gross Domestic Product and Infant Mortality Rate, proved to have less of an effect on these migration flows than that of unemployment. The overall finding that particular migration streams to Belgium were unaffected by the change in European migration regulations, calls into question the popular conception of migration as fluid and primarily controlled by immigration regulations at the area of destination. Rather, it appears that, when the social and economic climate of origin and destination were controlled for, the impact of regulations was surprisingly small for this particular case study. The overall model and the adopted time-series method appeared to perform well, at least according to the diagnostic statistics employed. The model did not perform well according to substansive statistics, however. The method, and particularly the data, should next be extended to further expand our knowledge about the causes of migration. The statistical method used in this thesis is recommended for study of other international migration streams in order to increase the generalizability of the findings