PART V CONCLUSION A SKETCH OF MODERN DEVELOPMENTS CHAPTER 1 [Introduction and Plan] 1. PLAN OF THE PART ONCE MORE we change our rules of procedure. The surveys presented in the three preceding Parts were indeed far from complete. But though incomplete, they aimed at conveying fairly comprehensive pictures. So far as scientific economics in the usual sense is concerned, no significant man or work or movement was left out—not intentionally, at least—and I have done what I could do within this volume to touch upon the more important framework and frontier questions. In this Part, we shall not go on with this plan. In a sense, our inquiry ends, at the foothills of the Marshall-Wicksellian mountain range, with the last glance at the classical situation around 1900. If we go on at all, it is with a different and much more restricted purpose. It seemed desirable, first, to show how the work of that period fared in our own time; 1 second, to point out some roads that are leading away from and beyond it; and, third, to attempt diagnosis and prognosis of contemporaneous efforts. This will at best give us a bird’s-eye view of just a few great contours with all the details and all the frontier districts left out. More than that, this view will have to be highly selective. I cannot even list all that I am going to leave out. But I will illustrate it by mentioning two men: Gottl and Spann. The widely different messages of these men, as is indeed obvious from the considerable body of literature produced by their followers, have shaped many a mind. In this sense, they are possibly 1 [To a considerable extent, J.A.S. had already done this in Part IV. It will be recalled that, when outlining the plan of the book in Part I, he wrote: ‘Part IV will present an account of the fortunes of analytic or scientific economics from the end of the “classic” period to the First World War though the history of some topics will (for the sake of convenience) be carried to the present time…. Part V is merely a sketch of modern developments, relieved of some of its cargo by the anticipations in Part IV that have just been mentioned, and aims at nothing more ambitious than helping the reader to understand how modern work links up with the work of the past.’] more important than any two high-powered technicians of economic theory. But they are not important for us. We are concerned with the technicians. He who writes a history of, say, agricultural technology does not thereby prove that he thinks it more important than the history of religion. Only so far as those authors—or any other of the same type— actually attempted analytic work in the sense adopted for the purposes of this book, does our failure to deal with them carry implications to which they or their adherents could object. Is this quite understood? [J.A.S. never finished this introduction to the Conclusion nor did he cover some of the topics he planned to include. In place of the introduction he would have written, there is presented in the next section a summary of five lectures, which were outlined at the same time that he was planning Part V and the last two or three chapters of Part IV and which presumably summarize what J.A.S. considered to be the main lines of advance in the recent period. What he actually intended to cover in the Conclusion (in addition to chs. 2, 3, 4, and 5 below) can only be guessed at from two pages of abbreviated notes (mostly in that maddening shorthand), which are reproduced in the Appendix. Among the ‘things still entirely lacking in (Part) V’ he listed: (1) Morgenstern and von Neumann, Theory of Games and Economic Behavior (1944) (2) Leontief’s Linear Programming (3) Income Analysis—Social Accounting (4) …Chenery (Engineering Production Functions)…Frisch (5) (Several lines of shorthand notes) From the second paragraph of section 3 (this chapter), it is obvious that J.A.S. also intended to comment on the ‘unprecedented wealth of statistical facts’ and on Econometrics, ‘the new relation between economic theory and statistical methods.’] 2. THE PROGRESS OF THEORETICAL ECONOMICS DURING THE LAST TWENTYFIVE YEARS 1 (a) Introductory Lecture on the Scope of the Course. The First World War caused a complete change in the economic policies of all nations which has persisted ever since. This was due, first, to the fact that all nations have had to face new problems arising out of political and economic situations in which they had never found themselves before. But, second, this change in policies was due also to the fact that the war had thoroughly upset the previous distribution of political weights. Thus, we observe not only new problems and new situations but also new attitudes toward them. 1 [J.A.S. delivered a course of five lectures in January 1948 at the School of Economics, University of Mexico, on this subject, which coincides roughly with what was intended for Part V (partly anticipated in Part IV). What follows is a brief summary of those lectures, written in advance for translation into Spanish and presented here in place of the Introduction and Plan, which were never Introduction and plan 1105 completed. The summary is printed in full despite certain repetitions; references in square brackets show where the subjects are treated in the History. These lectures were, of course, planned for a mixed audience and were of necessity rather general and elementary.] Economics and Political Economy. Economists moved with the times and a significant change occurred in their views about practical questions. The sum total of these views together with the schema of social values that underlies these views we shall call Political Economy. Accordingly, we say that a new Political Economy arose after 1918. But, however interesting it would be to describe this new Political Economy and to inquire into its sociological roots, this is not our task in this course. The new views on economic policy will be considered only so far as they are relevant to the development of scientific economics. By Scientific or Analytic Economics, in contrast to Political Economy, we shall mean the stock of facts and methods that economists collect with the purpose of explaining the phenomena of economic life. The difference between this analytic economics and political economy can be illustrated by analogy with the difference between the subjects that are taught at a faculty of medicine. Such a faculty has professors of surgery, internal medicine, and so on who teach the practical art of treating patients. But there are also professors of chemistry, physiology, biology who teach the scientific foundations of that art but not that art itself. It is with the analogues of the latter that we are concerned. Economics and Economic Theory. We shall restrict our subject still further. Perhaps the most important progress that has occurred in scientific economics is the vast increase in our command over facts. All types of information about facts have increased beyond the boldest dreams of past generations but our epoch has been particularly characterized by an increase of statistical information which was so great as to open up quite new possibilities for scientific research. In step with this increase of statistical material, there has been an equally important development of statistical methods. But we shall disregard all this and concentrate our attention on the developments in that restricted field which is called Economic Theory. So many misunderstandings still prevail about the nature, use, and limitations of Economic Theory that it is necessary to explain our conception of it. There was a time when Economic Theory meant precisely what we have called Political Economy above: there was a ‘liberal’ or ‘socialist’ or ‘mercantilist’ theory, and all those theories more or less meant political doctrines or at least practical recommendations. This is not the modern view. The modern economist considers theory simply as an instrument of research. This instrumental character of economic theory will be illustrated by examples which will also explain the relation that nevertheless exists between economic theory and economic policy. Precisely because economic theory is only an instrument of research, it cannot produce concrete results without the facts that are supplied by statistics or non-statistical description. This had been realized already by the Spanish economists of the sixteenth and seventeenth centuries. But the alliance between statistics and theoretical economics was not complete until the emergence of modern Econometrics. The Main Lines of Advance within Economic Theory. The most obvious way in which sciences advance is by new departures, that is, by the discovery of new facts, or new aspects of old facts, or new relations between facts. Examples will be given from the history of physics and of economics. But there is another way. When we use the concepts History of economic analysis 1106 and theorems that we have inherited from our predecessors, these concepts and theorems—which we call the analytic apparatus of a science—change in our hands. We add here and correct there and so this apparatus slowly develops into a different one. It will be our first task to describe how, approximately between 1890 and 1914, a system of economic theory consolidated itself and how this system formed the basis of later work, which started in the early 1920’s and transformed it without intending to do so [Part V, ch. 2 of this History]. Then we shall see how a new analytic apparatus developed which is known as economic dynamics [Part V, ch. 4 of this History], Another new departure, mainly associated with the name of Lord Keynes will be considered next [Part V, ch. 5]. And, finally, we shall sum up what has been accomplished and what might be expected for the near future. (b) The Marshall-Wicksell System and Its Development. Scientific economics found its systematic form in the eighteenth century (Beccaria, A.Smith, Turgot) and, after various ‘revolutions,’ in the Principles of Political Economy of J.S.Mill. This system was in turn revolutionized by the introduction of the marginal utility principle (Jevons, Menger, Walras). But another process of consolidation took place between 1890 and 1914, and a theoretical system of apparatus emerged which is embodied in the standard works of A.Marshall and K.Wicksell. A few minutes will be devoted to describing the salient characteristics of this system and the extent to which it was accepted by the professional theorists of all countries [Part IV of the History]. We shall then proceed to discuss the main lines of advance that started from this system. The Theory of the Individual Firm and Monopolistic Competition. Neither Marshall nor Wicksell had neglected the task of analyzing the behavior of individual firms. But their theorems, except in the case of monopoly, referred mostly to a whole group of firms (industry) or even to the whole organism of the social economy. They hardly realized the necessity of investigating more closely the behavior of the individual units that combine to produce the phenomena which we associate with an industry or the social economy. In analyzing this behavior theorists soon discovered that the case of perfect or pure competition was a rare exception rather than the rule, and that the economic organism, especially in cases of decreasing average cost, does not function as it would under perfect or pure competition. From this a new body of theorems arose, the theory of Imperfect (Robinson) or Monopolistic (Chamberlin) Competition, the main features of which will be briefly characterized [Part V, ch. 2 of the History]. Indifference Varieties. In spite of the protests of Pareto and others, the theorists of the Marshall-Wicksell generation used uncritically the concept of marginal utility. During the 1920’s and 1930’s this concept was rapidly discarded in favor of the ‘indifference-curve’ approach. The reasons for this and the advantages of the indifference-curve approach will be discussed briefly (See Hicks, Value and Capital, 2nd ed., 1946) [Part IV, ch. 7, sec. 8 and App. and Part V, ch. 2]. The consequences of the passing of the old marginal-utility theory for Welfare Economics can only be touched perfunctorily [Part IV, ch. 7, Appendix: Note on the Theory of Utility]. Other Improvements of the Marshall-Wicksell Apparatus. With increasing scientific rigor and especially with the increasing use of mathematics in economic theory, the theorists have in the last twenty-five years been able to develop many of the doctrines taught by Introduction and plan 1107 Marshall and Wicksell and to correct others. An example of these developments is the theory of substitution which created the concept of elasticity of substitution. This conception is useful in settling in a few lines many problems that filled pages and even volumes in the past (for instance the problem of the influence of the introduction of machines upon the interests of labor). Corrections have been mainly applied to the old theory of production by means of a closer analysis of the properties of Production Functions [Part IV, ch. 7, sec. 8]. (c) Economic Dynamics. We call a relation static if it connects economic quantities that refer to the same point of time. Thus, if the quantity of a commodity that is demanded at a point of time (t) is considered as dependent upon the price of this commodity at the same point of time (t), this is a static relation. We call a relation dynamic if it connects economic quantities that refer to different points of time. Thus, if the quantity of a commodity that is offered at a point of time (t) is considered as dependent upon the price that prevailed at the point of time (t−1), this is a dynamic proposition. These definitions of the terms ‘static’ and ‘dynamic’ must be carefully distinguished from others that have been used and are still used sometimes. The Marshall-Wicksell system was essentially static. The Importance of a Dynamic Theory. The necessity of developing a dynamic theory rests upon three facts: (1) It is obvious that most quantities demanded and offered, both of finished commodities and of factors of production, as well as prices and incomes are in reality related to other economic quantities that belong not to the same moment but to the past or to the expected future. It is particularly obvious that monopolists want to maximize gains not for the moment but over a stretch of time. (2) It is not so obvious but it is nevertheless true that this makes a great deal of difference to results. If we drop the hypothesis that each element of the economy depends only on the other elements as they are at the same point of time, quite different results and quite new phenomena emerge, for instance the phenomenon of endogenous fluctuations. (3) Finally, the task of developing a dynamic theory is very difficult and cannot be accomplished simply by adding dynamic qualifications to static theory. It requires new techniques and raises fundamental problems of its own. An example of the new techniques required is the theory of difference equations. An example of the new fundamental problems is economic equilibrium, which, if considered from a dynamic standpoint, appears in a new light. An Illustration: the Cobweb Problem. When farmers observe current prices of, e.g., pork and fodder, they will decide to produce more or less hogs according as hog production is or is not profitable at this current relation between the prices of pork and fodder. But this decision cannot take effect before a certain period has elapsed. The resulting supply of pork will then impinge on the market and change the pre-existing relation between the prices of pork and fodder. This will induce new decision by the farmers and so on. This ‘cobweb problem’ or ‘hog cycle’ will be discussed, under simplifying assumptions, by means of a simple diagram. A similar problem is the so-called shipbuilding cycle studied by Tinbergen (Weltwirtschaftliches Archiv, 1931). [All the problems outlined for this lecture on Economic Dynamics are treated in Part V, ch. 4 of the History.] History of economic analysis 1108 (d) Income Analysis. We have a strong scientific interest in reducing the number of the economic variables with which we have to deal. If we tried to write down the equations that determine the static equilibrium of millions of firms and households, we should never accomplish the task. In particular, we could never marshal the statistics that would be the necessary complement of such a system. This suggests the idea of reducing the number of variables to a few great social aggregates. This idea is very old. From the first economists have tried to reason on national income, national sum total of wages, and the like. But it was only during the last quarter of a century that this idea has been systematically followed up. It is clear that we should be in a much better position to apply theory to statistics and statistics to theory if we could, for some purposes or for all, confine ourselves to such variables as National Income, National Consumption and Investment, Quantity of Money, Employment, and Interest Rates. Analysis which attempts to do so is called Macroanalysis (R.Frisch). Because the National Income is the central variable in which we are particularly interested, it is also called Income Analysis. The Keynesian Theory. The most successful of all the theoretical systems that have been inspired by this wish to simplify the structure of economic theory is the static system that is associated with the name of the late Lord Keynes. Many others have also been constructed, for example, by Amoroso, Frisch, Kalecki, Pigou, Tinbergen, Vinci. Lord Keynes used only four variables explicitly: quantity of money (deposits), consumption, investment, and interest rates. Income enters also, but is simply identical with consumption plus investment. The price level is eliminated by the use of ‘wage-units’ or labor hours in which all quantities are expressed. Employment is wedded to income by the assumption that it is strictly proportional to income expressed in wage-units. The variables are linked together by three relations: the liquidity-preference function, the consumption function (which implies the famous ‘multiplier’), and the investment function, all of which will be briefly explained. Discussion of the Keynesian Theory. Keynes presented his theory as a macrostatic system. But it is possible to turn it into a macrodynamic system without great difficulty. It is much more serious that Keynes assumed not only that methods of production remain unchanged but also that the amount of industrial equipment does not vary. This restricts his analysis to very short periods of time (3–10 months). Moreover, since technological change is the essence of the capitalist process and the source of most of its problems, this assumption excludes the salient features of capitalist reality. The novelty in Keynes’s theory of saving consists simply in this. Before Keynes, economists used to take it for granted that, normally, savers invest whatever they save. Keynes assumed that people save without having any definite intention to invest and that, when they have saved, they may decide not to invest at all but to keep their savings in the form of money (General Theory of Employment, Interest, and Money, 1936, pp. 165–6). On this alone rests the peculiar features of his theory of interest. But saving without investment occurs only in deep depressions, that is to say, in about one year out of every ten on the long-time average. The concept of marginal efficiency of capital is not the same as the old marginal productivity of capital, but essentially it expresses the same facts. Keynes’s theory of wages is interesting because it seems to supply an explanation of permanent unemployment as distinguished from cyclical unemployment. But it does so Introduction and plan 1109 only by means of the assumption that monetary wage rates are rigid. And nobody has ever denied that unemployment may persist indefinitely in this case. The Success of the Keynesian Theory. We have seen that fundamentally Keynes accepts the Marshallian apparatus of economic theory and that he only adjusted it in a number of points. But these points were very important for the explanation of the depression of the 1930’s and therefore rightly attracted attention. Moreover, his simple system that considers only a few aggregates was easy to master and to manipulate. From these factors of scientific success we must, however, distinguish a much more potent factor of political success. Keynes seemed to present an argument that saving, the great virtue that the majority of bourgeois economists from A.Smith on had always extolled, was really a vice that was the cause not of capital formation but of unemployment and capital destruction. This attracted many people who had for other reasons renounced allegiance to the values of capitalist society, and thus made Keynesian doctrine—not quite logically—the banner of economic radicalism. [All the salient points of this Lecture on Income Analysis will be found developed in Part V, ch. 5, ‘Keynes and Modern Macroeconomics.’] (e) Summary of the Course. It is impossible to foresee what future generations will think of the work in economic theory from 1920 to 1945. We can survey the points which posterity will have to judge but we cannot pronounce upon their value. One thing must be kept in mind, however. The economic theory of our own time and of all future times can never again be so fascinating to the wider public as it had been in the times when it was understandable to every educated person and when it seemed to establish directly ‘eternal laws’ and practical rules. Everyone can understand A. Smith. Only specialists can understand the matrix calculus and functional equations. Everyone is interested in free trade or protection. Only specialists are interested in questions of determinateness and stability. The Progress in Technique. The one thing that can be confidently asserted about the work of the period we have been surveying is that the theory of 1945 is greatly superior to the theory of 1900 as regards technique. Results are more reliable, proofs are more rigorous. This in itself also means more results and more specialized results that fit better the endless variety of the configurations of economic reality. At the same time it must be admitted that fundamentally new ideas have been almost wholly absent. We make much more of the ideas which we have inherited from the preceding period and often present them in a new light, but we have added little to them. As a conspicuous example, the theory of Business Cycles will be briefly discussed in order to show that all the essential ideas were developed before 1914 [Part V, ch. 4]. Economic Theory in the Service of Economic Policy. Modern theory no longer undertakes to show that free trade is the right policy for all times and places. But it shows much better than could have been shown by Smith or Mill what will be the effects of a particular measure of protection on the interests of all classes of society. Modern theory no longer undertakes to prove that perfect competition is an ideal. But it can show what the effects of given deviations from competition will be. Modern theory no longer recommends saving under all circumstances. But it gives to economic policy a complete description of the process of saving and of the effects that different kinds of saving will exert upon the economic situation of a country. Many other examples could be cited in History of economic analysis 1110 order to show that modern theorists are developing an apparatus that is indeed no longer simple but will render in the end the same service to economic policy which theoretical physics renders to engineering. Planning and Socialism. What has just been said may be applied to any kind of economic planning. Economic theory is slowly developing the mental instruments that are necessary in order to ‘rationalize’ planning and to tell planners what they must do and avoid in order to attain certain given ends. If a socialist society is defined as the perfectly planned society, then we may further say that modern theory is building the foundations of a truly ‘scientific’ socialism [Part IV, ch. 7, sec. 5]. To say that pure theory is of no interest for practice is as unreasonable as to say that pure mechanics is of no interest for building the machines we want. The ends themselves, that is to say, the kind of society or culture we want, we must choose ourselves. No science can do more than indicate the means of attaining whatever it is we want. [3. BACKGROUND AND PATTERNS] 1 Very roughly, the beginnings of what I believe to be a new period in the history of economic analysis date from the First World War. But this was a coincidence. Causally, that world war had little to do with the new tendencies, which in fact were discernible before 1914. The public, of course, was under the impression, as it always is in any epoch of striking events, that the economic phenomena it observed were entirely novel, unheard-of, and of a na-ture to upset completely the schemata of analytic economics. And it was a new experience for some people to be poor instead of rich and for others to be rich instead of poor, for some to see their interests championed by politicians instead of ignored, for others to see their interests attacked by politicians instead of defended as they used to be. But no economic fact or process observable during that war and its aftermath had anything new to teach to the scientific economist. The inflationary processes in particular fitted beautifully into the oldest of old schemata. Nor is this anything to wonder at. Economics is a very unsatisfactory science. But it would have to be much more unsatisfactory than it is if such an event as a war, however extensive and destructive, sufficed to upset its teaching. 1 [Apparently J.A.S. intended to do here very briefly what he did for the preceding period in Part IV, ch. 2.] Introduction and plan 1111 . Dynamics are treated in Part V, ch. 4 of the History. ] History of economic analysis 1108 (d) Income Analysis. We have a strong scientific interest in reducing the number of the economic variables. introduction of machines upon the interests of labor). Corrections have been mainly applied to the old theory of production by means of a closer analysis of the properties of Production Functions [Part. between facts. Examples will be given from the history of physics and of economics. But there is another way. When we use the concepts History of economic analysis 1106 and theorems that we have