II. The Problem of Application

CHAPTER II THE PROBLEM OF ApPLICATION "'The treating as constant what is variable is the SOUTce of most of the fallacies in Political Economy."-F. Y. Edgeworth. It shQuld be clear from the beginning that the fundanlental problem of economic policy consists in the applieation of economic theory. The modern theory of kno\vledge leaves no doubt that it is impos sible to grasp reality without the construction of theoretical formulre, and that every neglect of theory is attended by grave consequences. This fundamental truth is expressed with incomparable clarity by Pro fessor Whitehead when he says: " The parado.x is now fully established that the utmost abstractions are the true weapons with which to control our thought of concrete facts."· And it was Goethe who coined the immortal phrase: "Das Hochste ware, zu begreifen, dass alles Faktische schon Theorie ist." t This is a truth which should constantly be impressed upon those who are responsible for economic policy. For if official economic policy is using and applying theory in this and every sense, then it comes down into the arena where the theoretical economist can challenge it, and it should allow itself to be put right, which it is much too unwilling to do.

The application of theory means acting in a *" Science and the Modern 'Vorld," New York, 1925, p. 48. t" The greatest thing would be to realize that everything factual implies a statement of theory." 6 THE PROBLEM OF APPLICATION 7 systematic way. Now, in the modern money economy everything in the nature of a social-economic occur rence consists in human actions and behaviour and the phenomena by which the occurrence is recognized prices, quantities of commodities, production, &c.-are merely reflections of this behaviour. Measures of economic policy are therefore in the last analysis directed towards influencing this human behaviour. Generally, however, they have to reach that end not by directly affecting human behaviour, say, by advice or command, but by setting up or removing barriers or by conditioning complex economic quantities, as, for example, by fixing prices in such a way as to cQmpel the desired behaviour. The path followed by economic policy is therefore always an indirect one even where " direct" measures are applied. * Systematic economic policy is based on the one side on definite expectations as to the psychological effect which its measures will have on the human beings affected by them, and Qn the other on expectations as to. the behaviour of objective data of the external world.

Here we come to the first gap which economic theory has had to bridge for itself by the process of isolation. The human mind is in reality subject to innumerable other influences, over and above those set out here, and they must all be taken into account. This gap, which is only quoted as an example, must not however be used as an arguluen t against economic theory, for it has to do with things belonging to an entirely different level of abstraction. * This" indirect" type of economic policy should not, however, be confused with that explicitly indirect policy such as is represented by the case where a lowering of real wages is sought not by decreasing money wages, but by raising tariffs or depreciating the currency.

8 THE LIMITS OF ECONOMICS Pure economic theo.ry consists of a number of pro positions of a very high degree of abstraction. The same is true also of other sciences which have some practical application, but in these other sciences the gap between theory and practice can usually be bridged very quickly. The reason for this is that these (natural) sciences possess a number of constants on which they can rely as props to give them support in the practical sphere. Such constants-ranging from boilinp; points, through the constant velocities of light and sound and the dispersion of lines in the spectrum to Planck's coefficient h - lend a certain solidity to mathematical equations in the natural sciences. In the social sciences, and therefore in economics, they are lacking. There is a word which takes their place and eschews all difficulties-the word relat'l:ve. The Inost that can be done in economic theory is to show relations between quantities (which themselves only represent relationships), and even this can only succeed under the assumption of a strict ceteris pari bus which gives the illusion 0.£ the stability of constants but which is in reality o.nly equivalent to the constancy of the degree of abstraction. Constants have to be replaced by historical data if we want to give even the simplest illustration of a proposition, and immediately we come to treat historical data an " isolation" in the strict scientific sense is impossible.

The circumstance that economic knowledge can only take the form of establishing relationships contributes substantially to the difficulty of winning an intelli gent following for it.. Since, however, the starting point of all theoretical analysis is experience, and since the whole object must be to revert to. experience as soon as possible, there is no reason to fear that a THE PROBLEM OF APPLICATION 9 priori constructions, which lead to empty tautologies and make no contribution to knowledge of the real world, will find any place in this discussion. They are only to be considered in so far as they have given rise to mistaken criteria of economic policy. To construct economics on the basis of experience does nothing to alter its theoretical character for it is bound to use the processes of abstraction and deduc tion in translating the ordinary facts of experience into general propositions. These generalizations are, of course, equally with the laws derived from them, constantly subject to correction as in all similarly con structed sciences. All empirical laws are subject to exceptions. The major propositions can best be dealt with by means of the axiomatic method. This means putting a minimum a.f factual statements at the top, and then rigorously deducing from them all the causal connexions which they show. Economics is thus as much inductive as it is deductive. It is, accordingly, not inferior to other sciences and certainly not as regards its " worth."

It is quite wrong to associate this appeal to experi ence and facts in any way with the long-displaced do.ctrines, and confused conceptions, of the so-called historical school. This school maintained that it was impossible to make generalizations from facts and thought it sufficient simply to describe the facts. The place that was in earlier decades occupied by history is to-day taken by statistics among the disciples of the historical school who still linger on disguised. There is no need to waste time on this old controversy, for it has long since been settled with the realization that theory is not only possible but is indispensable for an understanding of economic questions. None the less, 10 THE LIMITS OF ECONOMICS the thoroughly empirical character" of economic theory cannot be stressed too strongly. A priori theory would be very easy if it were possible to dispense with the necessity of dealing with reality and with the flux of economic events and if it were sufficient to lock oneself in a room and invent the world of facts, adopting the attitude that if theory and reality did not then agree, so much the worse for reality. "Theory" of that kind can neither be confirmed nor refuted: nothing easier could be wished for. But, unfortunately, it has nothing to. do with the real world. The limits to the use of the a priori method in economics must be strictly drawn and as strictly observed. It is, moreover, worth noting that in practice the difference is one of method only, for the few surviving apriorists are obliged in practice to make so many co.ncessions that in the actual theorems themselves they abandon their original posi.

tion, so that in the end both they and the empiricists are speaking the same language . What is really the most unfortunate result of their methodological posi tion is their tendency to identify economic theory wIth a particular system of economic policy. The first question that arises in applying the con clusions of economic theory relates to the completeness of economic science. To suppo.se that this science is already complete and only awaits practical utilization would be foolish. No science can ever possibly reach this condition. It will never be possible to do more than pluck out incomplete propositions from the ideal system. Nevertheless, the question of com pleteness is just as relevant to these partial spheres. It means that it should be determined whether the theorems are" complete" for the range which they cover-and for this range only-in the sense that they THE PROBLEM OF APPLICATION 11 are capable of comprehending and explaining all possible empirical eyents in this field. Expressing it in technical language, we must establish whether the system of economic analysis exhibits "lacunce."

Such lacunce are to be found in every level of abstrac .. tion, including the highest, in relation to the one next below. The nearer, however, one approaches to economic data and seeks to fit them into. the theoretical system the more extensive become these questions which are not covered by the generaliza tions of economic theory. The proper treatment of such lacunce presupposes an appropriate arrangement of the data beforehand. I t is in this that the real practical art lies. The latter can probably be acquired only in the course of long years of study and of laborious experimentation as in the chemical or physics laboratory. Consequently, economic policy is not at all the right thing for minds that are too " theoretical," as these are for the most part much too hasty with their judgment about the facts of reality. The handling of economic policy requires not only complete familiarity with the whole system of economic theory, but also, on account of the lacunce where theory leaves the economic politician unaided, a continual readaptation of theory. This readaptation must moreover, since time is usually pressing, be assumed ad hoc. This is true, as will become evident later, at any rate for cases of refined measures of economic policy or unusual circumstances.

The main difficulty of theorizing-in this way for a particular O,ccasionis that such a newly formed special theoretical explanation at best only illuminates the factors connected with the single historical instance. It gives no guarantee either that it will be applicable 12 THE LIMITS OF ECONOMICS to other instances or that it is capable of immediately being fitted into any definite place in the general body of theory. If these difficulties were not so great, economic science would be comparatively simple and could soon be disposed of by means of an accurate economic history kept well up to date. It follows from what has heen said above that economics is subject to a continuous process of change, which one must hope and assume to represent progress. This is due not only to the reasons already described but also to the circumstance that economic theory is neeessarily influenced by the dynamics of our general process of acquiring knowledge. Now it is a strange paradox that the majority of economists seek to elnphasize the finality of economic doctrines-their own doctrines, it should be noted. This. might be attri buted to the weakness of human nature, but it must have deeper causes as well, which we shall not search for here. N,evertheless, one is somewhat disagreeably impressed by the lack of modesty as regards their own science and as regards the capacity of the human mind which this attitude displays. It is, however, strange that the general public, and among' them those who are most sharply contemptuous of theory, namely, the majority of business men, demand that theory sho.uld be of permanent validity. It never occurs to an engineer to demand that the results of physics shall be immutable-quite the opposite is the case. But the public demands more or less explicitly that economics should preferably have already fully explained beforehand everything that can possibly happen. Should so·me event contradict or even appear to contradict an economic propo.sition (usually in its popular formulation), or more often an economic THE PROBLEM OF APPLICATION 13 sophistry or a misinterpretation of an economic theorem, the whole science is immediately" done for"

and is held to have given a new " proof" of its use lessness. The situation of economics in this respect is probably quite unique, for there is no other science which receives a similar treatment from the majority of its students as well as from the public. It is there fore especially necessary to acknowledge quite frankly that the development of economic analysis often leads the scientist to present at a later date a different solution of a problem from that previously available. This does not mean that he is on that account either fickle or igno!ant. Quite often the necessity of changing one's views is the result of new and more complete knowledge of the data or of a totally different development of the framework of conditions adopted as " constants. " We are here leaving out of account the serious possibility that the original investigation of the data may have been hindered by external limitations or inadequate training, for the real argument is much more serious than this. I t has emerged out of the modern study of trade-cycle theory. In this field of study it soon became apparent to critical minds that the facts of economic life cannot be comprehensively described in terms of statistics. And this i~ so even if it be assumed that the economic process can be satisfac torily separated from the general social process. There are always forces at work which are masked, mainly by the time factor, and which, immediately they begin to operate, are already shaping the present and the future without yet having given any tangible mani festation. It is not only in the sphere of human psychology that this is true, although that is what 14 THE LIMITS OF ECONOMICS naturally comes to mind first. Probably even more important is the circumstance that the time element often seriously falsifies the data that are available on the basis of present knowledge, but none the less is continually operating in the direction of revealing at a later date the real forces now at work. Thus at every mom ent of time there ensues an unmasking of the immediate past. Since, however, human action takes place not in the past but in the present and is directed towards obtaining results in the future which is never fully known, a vast source of error is opened up. At any moment of time economic events, like all events, seem to be proceeding in some particular direction.

Only the future, often by bringing things to pass which are contrary to expectation, shows what numerous other possibilities at that time unrealized were present. The inadequate knowledge of data is therefore part of the very nature of economic policy. In the study of the trade cycle, for example, we can rarely say definitely at any moment in what particular phase of the cyclical wave the economic system really is. Such insufficiency of data is in great part responsible for the fact that econo.mic policy is so often lacking in rationality. This will obviously be all the more so the finer the methods which economic policy sets out to apply and, however paradoxical this may sound, the shorter the periods for which decisions have to relate. It is far less frequent for economic theory to be disproved in the long' run. The lacunm of theory, the necessity which they impose of a continual readaptation of theory for con crete purposes of economic policy, and the funda mental difficulties of analysing the data (quite apart THE PROBLEM OF APPLICATION 15 from the technical difficulties of assembling them), are thus the principal reasons why the ways and methods of economic policy are somewhat vague. It would, however, be an inexcusable error to believe that this element of vagueness would be less if economic policy were to be conducted u,ithout the help of theory and with less attention to the collection and utiliza tion of current statistical and other material. Surely a flickering torch is preferable to complete darkness.