When the Asquith Coalition Government took office in May, 1915, the new First Lord of the Admiralty was Mr. Arthur James Balfour. His appointment to that particular position at that time was due to a number of influences, among which political exigency was by no means the least important. In the Cabinet reconstruction, Asquith had insisted that Grey at the Foreign Office and Kitchener at the War Office were not to be changed, and these reservations materially lessened the higher Cabinet posts which were available for prominent Conservatives. As a result of this delicate weighing of leaders and offices, Balfour became the most obvious choice for the Admiralty, a position for which he was well qualified on at least two other grounds. He not only held a seat in the Commons, but his experience as an unofficial consultant and member of the War Council during the previous seven months had given him a more comprehensive knowledge of the problems of the fighting services than any of his Conservative colleagues. The suitability of his appointment, moreover, was endorsed by his predecessor, Mr. Churchill, who suggested that inasmuch as Balfour had recently been in unusually close touch with events at the Admiralty, he could take over its duties with very little break in the continuity of the administration.

Less than justice has been done to Sir William Temple as a political scientist. It is true that his reputation now stands higher than it did in the nineteenth century when Courtenay, his biographer, observed, “Neither in science nor in politics, were his opinions founded upon an accurate combination of facts and principles,” and added, “Of Political Economy Temple knew nothing” and when Macaulay dismissed the essay on government as “exceedingly childish.” Three modern studies have done much to revise this unflattering estimate, but some aspects of Temple's political thought have remained unconsidered or confused. Professor Herriott in his analysis of Temple's ideas on government has pointed out the novelty of those ideas but has given little attention to the significance of Temple's approach and method. Dr. Marburg's excellent book, though it does not deal primarily with Temple's political ideas, has made a valuable contribution towards their understanding. But her views on Temple's attitude to science and his use of its method are open to question. Professor Woodbridge, in his recent very full study of Temple, praises “the astonishing originality” of the essay On the Original and Nature of Government, and remarks as its great merit that its argument is based “upon historical probability, observation and experience,” but he does not emphasize the inductive aspect of Temple's work nor discuss the relation of his political thought to the scientific movement. Apart from Professor G. N. Clark's explicit but passing references to Temple's significance as one of the first of the inductive political thinkers, the extent to which Temple participated in the prevailing scientific climate of opinion and was influenced by it has been generally overlooked or greatly underestimated.

This article is an attempt to apply to certain problems of shifting equilibria some techniques which are familiar to readers of Professor Schumpeter. The problems to be examined are: the effect of a technical advance on the equilibrium position of a single firm which is in a position to adopt the new techniques; the effect of a technical advance on the optimum and typical size of firms in the industry; the social dispersal of benefits from a technical advance; the relation of a technical advance to the rate of investment and the rate of interest; and the implications of technical advance for social policy and the level of welfare.

I wish first to consider the position of a firm in imperfect competition experiencing a technical advance over a period of time. Now, whatever the nature of this firm, there will be a point, at any moment, beyond which increases in output can be achieved only at an increased unit cost. The greater the proportion of fixed to variable costs the further to the right (graphically speaking) this point will be. In industries like railroading the point may be to the right of the relevant sector of the curves because the proportionate decline in fixed costs per unit of output is absolutely so great as to offset rising variable unit costs for any practicable output. However diminishing returns inevitably exist in the rigorous sense that the intensification of the use of fixed plant, at any moment of time, will eventually result in diminished increments of product per unit of variable factor employed, and this implies increasing marginal variable cost. Total unit costs will increase when, with each increase in output, the absolute increases in variable costs exceed the absolute decrees in fixed costs per unit; and we should remember that, though the decline in fixed costs is uniform as a proportionate decline (the fixed unit cost curve is always a rectangular hyperbola of the type xy=k), the absolute decline, as output is increased, becomes less and less. Thus, and this is a commonplace of economic theory, any change in the proportion of fixed to variable costs of the firm, will have a definite effect on the optimum cost point and on whether or not the firm is operating, within any practicable range of output, under increasing or decreasing costs.