During the last fifteen years carefully prepared population forecasts have appeared for nearly every important country possessed of reasonably adequate vital statistics and census records. With the general recognition of essential differences between the behaviour of bugs in a bottle and homo sapiens “in leaky national containers”, the usual procedure in the more recent studies has been to isolate the principal component factors in population growth, predict the future trend in each, and thus estimate the probable population ten, twenty, fifty, or more years hence. Already the majority of even the more conservative forecasts are proving too high and almost invariably because of over-estimation of the birth-rate.
Errors of this sort are not unexpected in view of the inherent difficulties of the problem. Their number is legion; it will suffice to note three. First, there is the difficulty of foreseeing changes in the social and particularly in the economic environment. Even in so-called normal times, each year brings unanticipated and sometimes substantial modifications in the complex of social forces making for greater or smaller rates of population increase. The duration and severity of such abnormal phenomena as the present depression are, of course, incapable of measurement in advance. Second, the nature and extent of human reactions to environmental change vary greatly both from country to country and from time to time. How a particular people will behave in a new situation is, therefore, usually a matter of considerable uncertainty. This uncertainty is in part at least attributable to a third difficulty. While many excellent studies, especially during the last decade, have contributed materially to a more adequate understanding of the forces controlling population growth, the existing state of knowledge still falls appreciably short of permitting precise quantitative measurement of many important influences on the birth-rate.