The generalization of Walter Bagehot that successful administration “depends on a due mixture of special and non-special minds” has lost none of its importance with the passage of time, for the problem of combining competent administration with democratic control is more urgent today than ever before. Bagehot contended that the Cabinet Minister, the “non-special mind,” served a double purpose: he not only enabled a democracy to control the civil service, but he also made a genuine contribution to its administrative efficiency. The specialized civil servant, if left to his own devices, tended to become narrow in outlook, careless of the public convenience, and restricted by departmental routine; and the important secondary function of the Minister was to correct these failings by supplying a fresh mind and a different point of view.

The government of Great Britain still furnishes, as in Bagehot's time, the best practical application of this principle, and the high standards maintained by many of the departments bear testimony as to its essential truth. The need for this interplay of “special and non-special minds” is perhaps best seen in the history of the War Office, a department which by its very uniqueness in several respects presents in exaggerated form both the difficulty of the problem and the virtue of the remedy. The War Office is thus considered not merely as a sample department, but one which places, as it were, a magnifying glass over the general problem of departmental administration elsewhere.

The economist ordinarily draws a series of supply curves designed to show the relationship between cost per unit and units produced on the assumption of static conditions and usually without regard to the frequency distribution of units produced in relation to cost per unit. Accordingly, he focuses his attention on definitions of the marginal unit, of the marginal cost per unit, and of the marginal producer. This is largely because he wants to equalize price and marginal cost to obtain a rational theory of value under his own premises. It is true that the economist will not ignore variations about the margin and will even concede that, under dynamic conditions, you may arrange producers in series from low cost to high cost and that for the time being differential entrepreneurial costs will become manifest. It is assumed, however, that these variations will tend to disappear the closer dynamic conditions settle down to static or equilibrium conditions. But the variations do not disappear and we would be extremely surprised if they did; indeed, the probability of uniformity of unit costs of production for any given number of producers is mathematically remote. The important thing, however, is the degree of variation that actually occurs under certain conditions or for a certain number of producers or for a certain industry. If we approach the problem in this way we may bring reality to economic analysis, improve the working tools of marginalism or correct its inadequate theoretical structure.

The purpose of this article is to make available in somewhat more detailed form than has heretofore appeared, the results of a population projection recently made by Mr. M. C. MacLean and the present writer and to discuss some of the more important implications of the indicated changes in the Canadian population structure.

The Basis of Projection. In making the projection specific, birthrates were computed by relating the average number of births over the six years 1931-6 to the estimated number of survivors of the 1931 adult female population at the mid-point of that period. The projection was carried forward by five-year intervals for the two sexes separately using birth-rates so computed and mortality rates derived from the official all-Canada life tables for 1931. These rates were held constant throughout and no account was taken of the possible future effects of immigration and emigration or of war.

The reason tor this procedure as well as tor the selection oi the above-mentioned birth- and death-rates may be stated briefly. During the post-war decade (1921-31) the crude birth-rate in Canada as reflected by the ratio of children 0-4 to women 15-44 experienced a quite unusual and drastic decline. It fell by as much as 14.4 per cent in the ten-year period as compared with an average drop of approximately 5 per cent per decade in the preceding fifty years. About half of the decline in the last intercensal decade may be attributed to postponed marriages. A further drastic decline was registered during the depression years which undoubtedly resulted in many more postponed marriages. It seemed reasonable, therefore, to regard the existing low rates as somewhat abnormal and cyclical in character, and to anticipate at least a moderate though perhaps temporary revival as soon as economic conditions permitted the resumption of marriages on anything approaching a normal level. At the same time experience in Canada, as well as in other countries, suggests that the underlying trend in Canadian birthrates is likely to be downward for some time to come. Instead of attempting corrections for trend and cyclical movements which at best would involve a good deal of guess-work, it was decided to use throughout the low rates prevailing during the first part of the present decade (1931-6) on the assumption that their use would offset in some measure the neglect of the long-term downward trend.