The Principles of Political Economy

Henry Sidgwick


Chapter 3

The Method of Economic Science

§1. The result arrived at in the last chapter may be summed up thus. The Science---as distinct from the Art---of Political Economy, of which the general principles will be expounded in the first two books of this treatise, deals mainly with the laws or general facts of the Production, Distribution, and Exchange of wealth; and also with the general facts of the Consumption of wealth, so far as these are connected with the former. This definition of the subject coincides broadly with that adopted by most English writers; but there exist considerable differences of opinion as to the method by which the subject should be investigated: differences which---as was before observed---have been brought into special prominence in recent controversies. These controversies have turned mainly on two fundamental questions, which it will be convenient to consider together, since they are closely connected. It is disputed, first, whether the science of Political Economy can be advantageously treated separately from the general Science of Society: and secondly, whether its method is properly deductive and à priori, or inductive and historical.

It does not appear to me, that any instructive result can be attained by discussing either of these points, unless we carefully distinguish between different inquiries which, as we have seen, have been included under the name of Political Economy; and examine each separately in relation to the questions above raised. If we attend to this distinction I think it will appear that, though the divergences of view above noticed are likely always to exist to a certain extent, the controversy arising from them may at any rate be reduced to a much smaller space than it at present tends to occupy.

Let us begin, then, by considering the two questions of method above mentioned in relation to the Theory of Production. If we ask whether the investigation of the causes, by which the labour of any society is rendered more or less productive of wealth, can be properly separated from other parts of the general science of society, it is, I think, clearly wrong to answer such a question in an absolute way, either negatively or affirmatively. On the one hand, every economist ought to admit with Mill ``the universal consensus of social phenomena, whereby nothing which takes place in any part of the operations of society is without its share of influence on every other part''; on the other hand, this is no reason why the study of the industrial or wealth-producing organization of society should not be pursued as a ``separate though independent branch of sociological speculation'', just as in the natural body we study separately the physiology and pathology of each of the principal organs and tissues, though every one is acted on by the state of all the others.

And it is to be observed that the relations of industry to other factors of social life vary indefinitely in closeness and importance; so that the question how far it is needful to investigate them is one which has to be answered very differently in relation to different economic enquiries. Thus, in considering generally the causes of the improvement in the productive powers of labour, the importance of a healthy condition of social morality must not be overlooked; but it is not therefore the economist's duty to study in detail the doctrine or discipline of the different Christian churches: if, however, we are studying historically the causes that have affected the interest of capital, the views of Christian theologians with regard to usury will require careful attention. So, again, the conditions and development of the Fine Arts will not generally demand more than a very brief and summary treatment from the economist: if, however, we are investigating the share taken by a particular community in the international organization of industry, the special artistic faculties and sensibilities of its members may become a consideration of much importance. Similarly the influence exercised on industry by government has often been an economic factor of the first magnitude: still it is obvious that, in modern European communities, at the existing stage of social development, changes in the industrial organization of the civilised part of mankind are largely independent of changes in their political organization. For instance, in the present century, we have seen France pass from Absolute Monarchy to Limited Monarchy, from Limited Monarchy to Republic, from Republic to Empire, and from Empire to Republic again; and yet none of these changes---except the third during a transient crisis---have appreciably affected its industrial system; whereas this latter has been materially modified during the same period by causes unconnected with politics, such as the invention of railways and of electric telegraphs. At the same time I should quite admit that most English economists a generation ago hardly foresaw the extent to which political conditions would continue to affect industry up to the present date: and, similarly, the relations between the development of industry and other factors of social life, such as the progress and diffusion of knowledge, and the changes in national character or in the habits and sentiments of special classes, have hardly met with due consideration. Still the modifications which appear to me necessary on this score are of a subordinate kind; they do not amount to a fundamental difference of treatment.

If now we ask whether the method of such an investigation as we have been considering should be `inductive' and `historical' or `deductive' and `à priori', it again seems to me clear that there is not really room for much controversy. At any rate, I know of no economist who has attempted to ascertain the ``causes of the improvement in the productive powers of labour'' by a method purely---or even mainly---à priori and unhistorical. A certain amount of deductive reasoning, no doubt, has commonly been introduced into this investigation; but this seems inevitable. In particular, we require for the comprehension of economic facts some interpretation of the motives of human agents; and this has necessarily to be supplied, to a large extent, from our general knowledge of human nature---modified, of course, by any special knowledge that we may be able to gain as to the peculiar mental characteristics of the class of persons whom we are considering. But in the general analysis of the conditions favourable to effective production, which Mill and other writers who have followed him have given in the first part of their exposition, the deductive element has always been quite subordinate; and so far as the method adopted is different from what would ordinarily be called `inductive', it is not because it is in any sense an à priori method; but because it chiefly consists in getting a clearer and more systematic view, through reflective analysis, of general facts which common experience has already made familiar.

Thus, when Mill in his first six chapters states the requisites of production, labour, capital and natural agents; when he defines the notion of labour, considers its relation to the natural agents on which it operates, and classifies the different kinds of labour and the different species of utility produced by it; when he makes clear the notion of capital, as wealth diverted from the purpose of directly satisfying its owner's needs, and employed, whether in the form of instruments or labourers' necessaries, in producing other wealth: when he points out how capital is continually consumed and reproduced, but with various degrees of rapidity, according as it is fixed or circulating;---it is obvious that all these results, however interesting, are obtained by merely analysing and systematizing our common empirical knowledge of the facts of industry. So, again, when he goes on to consider the conditions on which the degree of productiveness of different productive agents depends; his method is again merely that of comparing and generalising from observed facts. Thus he studies quite à posteriori the differences in the natural advantages of different countries; the differences among human beings in habits of energetic work, in capacity of exertion for distant objects, in keenness of desire for wealth, and in other intellectual and moral qualities; and the differences in the security afforded ``by government, and against government'' at different times and places. So further, in the discussion of the advantages of division of labour, and in the comparison of production on a small scale with production on a large scale, his argument though partly deductive still relies greatly on specific experience. Then again, when he states the law of the increase of labour, the causes that actually counteract the capacity of increasing population inherent in human beings, and the extent of their operation, are investigated inductively (Ch. x.); and so are the actual variations in the ``effective desire of accumulation'', which causes the increase of capital (Ch. xi.). In both these cases we could, no doubt, without conscious induction, lay down certain incontrovertible abstract propositions; but in the former case we should hardly get beyond the truths of elementary arithmetic, and in the latter case we should hardly get beyond such trivial maxims as that ``wealth is increased by industry and thrift'', &c.

I have gone into these details, not because I wish to lay stress on Mill's authority, but because none of the ``orthodox'' critics of his widely-read book has ever attacked his general method of treating the Theory of Production. What therefore we have to remark is not merely that Mill's treatment of this part of his subject is mainly inductive and analytical; but that it never seems to have occurred to any ``à priori'' economist that it ought to have been different.

[Back to:] [Forward to:] [Up to:]