This must certainly be admitted.By the nicer and more complex application of these measures, we should approach a more accurate account of welfare, so far as it is ultimately expressible in terms of quantity.If we discovered that a proposed course of national policy would not only increase the average income by 10% but would increase the lower incomes of the population in a higher ratio, we should seem to have got a scientific warrant for the policy.But even this degree of scientific authority would be purchased to some extent by an artificial simplification of the actual problem of social-economy.To the statesman no problem of actual finance is capable of being set in such distinctively quantitative terms.Not merely cannot an earthly Chancellor of the Exchequer know how much can be added to the incomes of the several classes by the expenditure of so many millions upon transport, or upon any other single service, but, if he could, he would not be much nearer to the standard he requires.There are many different ways of raising the revenue in question and an infinite number of combinations of these ways.The same holds of expenditure.To take the simplest case;the ten millions that he raises may be applied to transport, or to education, or to defence, all the sum or any proportion, to each.Each expenditure claims to be beneficial, an outlay for public welfare.But the benefit in the several outlays is not equally presentable in terms of money income, and, so far as definitely economic gains accrue, they are not equally immediate or equally assured.It is evident that no amount of possession of statistical knowledge can possibly reduce the problem entirely, or even mainly, to one of quantitative calculation.It is equally true that when the problem is solved, its solution will appear in quantitative shape, i.e., so much money for transport, so much for education, so much for defence.It will seem to have been worked out by reducing the three forms of desired benefits to common terms, and then dividing the ten millions among them so as to secure an equivalence of gains at the margins.Economists will point out triumphantly the alleged fact that the last £100 spent on education produces a national return of welfare exactly equal to that obtained by the last £100 spent on gunboats, though his assertion remains inherently insusceptible of proof.In truth, the Chancellor's mind does not work in this way.So far as his statecraft is disinterested, or even allowing for every form of bias, his mind forms an ideal of social progress, of a happier or better state of things, and allots the outlay of his ten millions in an endeavour to assist in realising this ideal.Now the ideal itself is not chiefly a product of quantitative calculus, but of his more or less informed imagination, and his more or less wholesome sympathies.His views as to the means of realising this ideal can never be purely scientific, though science may here be of considerable assistance.
If, treating expenditure more widely as an act of public policy, we consider it as an operation of the general will of the community, a true act of political economy, the problem remains essentially the same.When looked at through scientific spectacles, it is a purely quantitative and mechanically ordered act, because the scientific method by its very modus operandi ignores the qualitative factors.So the nation is supposed to balance this gain against another, and to lay out its revenue so as to get the largest aggregate of some common homogeneous stuff called 'welfare', in such a way that the last £100 spent on education is equivalent in its yield of this 'welfare' to the last £100 spent on the latest super-dreadnaught, or the last lot of old-age pensions.In truth, the common will no more functions in this fashion than the personal will of the Chancellor.
In each case Statecraft is an Art, and the financial policy is an artistic or creative work in which quantities are used but do not direct or dominate.
By this line of argument it may appear as if we had repudiated the entire utility of a scientific calculus.This, however, is not the case.
For though all the determinant acts of policy or welfare, performed by an individual or a society, involve organic unity of design, and the qualitative considerations appertaining thereto, important and indeed necessary assistance is rendered by the quantitative analysis of past acts expressed in the form of scientific generalisations.A clearer understanding of the nature and extent of this cooperation between science and art in the conduct of life enforces this truth.
§11.Science takes its stand upon a twofold application of the assumption of the uniformity of Nature, first, that all differences of composition can be treated as differences of quantity or degree, secondly, that history repeats itself.Now, just so far as these assumptions fit the facts, Science is valid for interpretation and for guidance.This explains why astronomy, physics and chemistry are more 'exact' sciences than biology or psychology, and why they are able to give more reliable and authoritative rules for the arts of navigation, engineering and drug-making, than the latter can for medicine, for breeding or for education.Edward Carpenter has remarked that astronomy is the most exact of the applied sciences, because we know least about it, i.e., because we treat its subject-matter almost entirely from the single quantitative standpoint of space relations.
