The Principles of Political Economy

Henry Sidgwick


Chapter 3

The Method of Economic Science

§2. Why then, it may be pertinently asked, does Mill say---as he certainly does say---that Political Economy is essentially an abstract science, that its method is essentially ``the method à priori'', and that it ``has been so understood and taught by all its most distinguished teachers''? The only answer I can give is that in this and similar passages Mill is thinking, not of the Theory of Production as he himself conceives and expounds it, but of the Theory of Distribution and Exchange: and primarily of that portion of this latter subject which he distinguishes as ``statical'' and not ``dynamical''---i.e. that which treats of the determination of the reward of services and the value of products in the existing condition of industry. This is the part of the subject to which, since Ricardo, the attention of economic theorists has been chiefly directed (though they have often not distinguished it clearly from other parts): and it is easy to show how a method largely different from that adopted in treating the question of Production naturally suggests itself here. The broad and striking fact which originally led and still leads reflective minds to discuss the question ``how a nation is made wealthy'' is the vast difference between the amounts of wealth possessed by different nations and by the same nation at different periods of its history; especially the great increase in the most recent times, in consequence of what we speak of vaguely as ``advance of civilisation'', ``progress of arts and sciences'', ``development of trade and commerce'', &c. Hence in our method of dealing with this question induction from historical facts is naturally prominent; though a certain amount of deduction inevitably comes in when we analyse the combined play of the forces of economic change whose effects history presents to us. And we may, of course, examine the phenomena of Distribution from the same point of view of Comparative Plutology; we may ask why the share of wealth annually obtained by an English miner is larger than that obtained by a German miner, or why English landowners now obtain higher rents than they did 100 years ago: and if in our answers we ``include directly or remotely, the operation of all the causes'' that have combined in causing the differences investigated, it seems evident that our method of investigation must be---just as in the case of Production---a primarily inductive and historical one. We shall have to note and explain differences and changes in national character generally, in the habitual energy, enterprise, and thrift of special classes, in law and administration and other political circumstances, in the state of knowledge, the state of general and special education, and other social facts; and in this explanation the ``method à priori'' can evidently occupy but a very subordinate place.

But such questions are not, I think, those which most obviously suggest themselves in connexion with the phenomena of distribution. Here the broad and striking fact, that at once troubles the sympathy and stirs the curiosity of reflective persons, is the great difference between the shares of different members of the same society at the same time. Thus what economists have been primarily concerned to explain is how the complicated division of the produce of industry among the different classes of persons who have co-operated to produce it is actually determined here and now; and what is likely to be the effect of any particular change that may occur in the determining conditions, while the general state of things remains substantially the same. Similarly as regards the phenomena of exchange, the most natural and obvious question is why each of the vast number of articles that make up what in the aggregate we call wealth is exchanged and estimated at its present price; and how far any particular event, other things remaining the same, would tend to raise or lower its price.

It is in answering these questions that the general theory of Political Economy, as commonly treated, uses mainly an abstract, deductive, and hypothetical method. That is, it considers the general laws governing the determination of remunerations and prices, in a state of things taken as the type to which modern civilised society generally approximates, in which freedom of exchange and freedom in choice of calling and domicile are supposed to be---broadly speaking---complete within a certain range, and in which the natures and relations of the human beings composing the industrial organization are supposed to be simpler and more uniform than is actually the case in any known community. By means of this simplification we obtain exact answers to our general economic questions through reasonings that sometimes reach a considerable degree of complexity. It is obvious that answers so obtained do not by themselves enable us accurately to interpret or predict concrete economic phenomena; but I hold that when modified by a rough conjectural allowance for the difference between our hypothetical premises and the actual facts in any case, they do materially assist us in attaining approximate correctness in our interpretations and predictions.

It is, however, evident that the applicability and utility of such hypothetical reasonings will depend largely on two conditions: first on the degree of success attained in forming our original suppositions, so that they may correspond as closely as possible to the facts, without becoming unmanageably complex; and secondly on the extent to which we recognise and attend to the divergence from facts which is---in most cases---inevitable in such abstract reasonings, and the insight and skill which we show in conjecturing roughly the effect of modifying causes whose operation we cannot precisely trace. To secure success in either of these respects we require an accurate knowledge of the general characteristics of the matter with which we are dealing; and I do not see how we are to obtain this knowledge without an inductive study of economic facts. It is not perhaps necessary that the deductive and inductive investigation of any class of economic phenomena should be always carried on simultaneously, or even by the same persons; but the latter is certainly an indispensable supplement to the former.

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