In planning the territorial allocation of economic activity one of the most important factors to be taken into consideration is the availability of manpower in the various region. However, during the early phases of central planning, this factor was somewhat neglected by planners. Two causes of this neglect may be mentioned. First, it was assumed that manpower supply would exactly follow the demand for manpower, partly by migration, and partly by an optimal development of the birth rate. Second, it was also assumed that demographic development was governed by deterministic laws which made it possible to forecast birth rates, the extent of migration, etc., without great difficulty. The demographic and economic history of Hungary, however, did not conform to these simplifying assumptions. Demographic processes, such as natural increase and migration, proved to be influenced by a great number of factors, not solely economic ones, but also factors of a sociological and psychological nature. This will be illustrated by research findings undertaken by the Demographic Research Institute of the Central Statistical Office. 2. BACKGROUND TO THE STUDY The birth rate in Hungary fell in the years 1962-1965 to the lowest level in the World (around 13 per thousand population). It should be mentioned, however, that since 1966 some improvement can be observed. Studies were performed by the Demographic Research Institute on the causes for the decline and low level of the birth rate, and particularly the decline of fertility. One of these studies used the method of correlation and regression analysis based on fertility data in the various regions of the country. This method has been employed on a world scale by Weintraub [8] and Adelman [1], and by Vielrose [7] who studied the causes of fertility differences in Poland. Such a territorial cross-section method seems to be better suited to the analysis of time-series data than correlation and regression analysis in view of the methodological problems associated with the latter, e.g. the problem of lags, the lack of data for sufficiently long periods, etc. The data for 191 districts and towns of Hungary for 1960 (the year of the last census) were used as a basis for the computations. The variables taken into consideration were: * The author is associated with the Demographic Research Institute of the Central Statistical