The fundamental independence from war influences of the developments that brought about a new period of economic analysis can be easily established by listing them. First, there was the unprecedented wealth of statistical facts. Second, there were new results that grew out of working the old apparatus. Third, there was the development of dynamics. Fourth, there was the new relation between economic theory and statistical methods (Econometrics). It is these four—obviously interdependent—aspects of contemporaneous work that will be discussed in the chapters that are to follow. The rest of this chapter will be devoted to the discussion of a matter of ‘atmosphere.’ Our time is one of transition, not only in the sense in which any time is of necessity transitional, but also in the specific sense defined by rapidity, and by universal awareness and expectation, of actual and impending social change of a fundamental nature. Few will deny this. It will be convenient to state at once the two ways in which that fact bears upon the scientific work in our field. The first things to occur to most of us are the new patterns and the new problems. But so far as these are concerned, it is more important to realize the extent to which they are but old friends in new sociological garb than it is to realize the extent to which we are really facing new scientific problems. For to begin with, we may repeat for recent economic history what was said a moment ago for the economic history of the First World War. Social patterns, economic and other policies, economic situations are all quite different, but this does not in itself imply that new economic principles are either suggested by them or required in order to understand them. Thus foreign policies, economic and other, that strike the good old liberal as novel heresies and more enthusiastic observers as great discoveries would, as we have seen, have looked very familiar to Malynes and Misselden. The labor contract is no longer ‘free,’ but not only has it never been so except during comparatively short spells of history but also this does not mean a novel problem to the analyst—all he has to do is to take a different model from his box of tools. Political rents—payments from public funds to particular groups to which no specific economic service corresponds—are a salient feature of modern society; but they were not less important in the society of Louis XV: the fact that the recipients carried different class connotations is, for purposes of purely economic analysis, less important than it might seem. Friends and foes of the New Deal agreed in looking upon it as new. So it was in more than one sense. But not in our sense: practically every measure that is covered by that slogan had been observed and fully analyzed before. Nor is this all. There are possible events that would create historically novel situations. Fullfledged (not bolshevist) socialism, adopted by modern industrial society, is an example. But to the economist it would not present a new problem. The theory of a socialist economy lies ready at hand, fully worked out in part by thoroughly bourgeois economists at a time when there was no hope or danger of that exercise in pure theory ever being put to practical use. In this respect, economists were better than was their repute. Nevertheless, there are patterns and problems that are analytically novel. Among the examples, I might have mentioned ‘going off gold,’ devaluation, depreciation, exchange control, and other features of monetary management as instances of time-honored devices—only they were not always ‘honored’ and hence they were known under uncomplimentary names. But this would be only substantially, and not quite, correct. We do see other sides of them that were not seen before, and we have learned to reason about them in different ways. Moreover, theory tends to become—in part intentionally, but still History of economic analysis 1112 more unintentionally—specialized when, by tacit agreement, theorists look for a long time at the same social and economic pattern. Its features are then taken for granted, and many propositions are framed to fit even the least persistent of them. If central banks are practically treasury departments; if other banks have almost lost all functions but the clerical ones of cashing checks and buying government bonds; if the market rate of interest does not mean a thing and money-market and stock-exchange mechanisms are well-nigh paralyzed; if the profit motive of the industrial family is rapidly vanishing; if salaried employees administer the most important concerns; if private thrift and private investment have ceased to function, and income generation through government expenditure is looked upon as a normal element of the economic process, in which taxation absorbs the higher incomes; and so on—then the relative importance of the various pieces of the capitalist engine is so thoroughly affected (many of the pieces do not work at all, whereas others that could be justifiably neglected before assume a dominant role) that all the ‘applied’ fields will naturally acquire an entirely different complexion. And theorists will redistribute their emphasis upon their various models and work out some of them more fully while shelving others. But it is important to realize that this is all and that from the standpoint of analytic technique it means much less than the layman is inclined to believe. [Unfinished; many shorthand notes and then a sentence, which began: ‘The other way in which awareness of actual and impending change influences scientific work…’] Introduction and plan 1113 CHAPTER 2 [Developments Stemming from the Marshall- Wicksell Apparatus] [1. THE MODERN THEORY OF CONSUMERS’ BEHAVIOR AND THE ‘NEW’ THEORY OF PRODUCTION] THE MODERN THEORY of consumers’ behavior developed almost wholly during the last quarter of a century so far as the doctrine is concerned that is actually used and taught by the sector of the economics profession that is primarily interested in theory in this sense. But what was and is being developed consists in methods and results, mainly associated with the work of Fisher, Pareto, Barone, Johnson, and, if we do not mind adding a paper that remained practically unknown for a decade or more after its publication, Slutsky. This means that the fundamental ideas were present before the close of the First World War, not in the form of embryonic suggestions only, but well worked out, mainly by authors of international reputation, in forms accessible, so we should think, to every professional theorist. They had only to sink in and to be clarified, amplified, applied, and occasionally straightened out in the process. But little had to be added to them that was fundamentally new. The situation was much as it was in the automobile industry: in spite of all the improvements and new gadgets, a modern motor car is still much the same kind of thing as the motor car of 1914. 1 Exactly the same holds for what may still be called the new theory of production. And the concept of 1 While I think it necessary to emphasize this fact because it is quite essential in order to understand the present situation, I do not wish it to be misunderstood. Such a misunderstanding would be involved for instance in any impression the reader might conceive to the effect that emphasis upon that fact implies derogation either of the performance of our age or of the talent that went into it: a physicist writing in 1730 might have been equal in mental stature to Newton; but it was ‘objectively’ impossible for him to produce another work like the Principia Mathematica (1687): he would have had to bend to ‘objectively’ and ‘relatively’ lesser tasks. Similarly, it is no reflection upon either Frisch or Samuelson to include their performances in the category of elaborating or continuing work. On the contrary, both performances illustrate well the types of originality that were possible in the field of the theory of consumers’ behavior at the time they wrote: both performances produced novelties but novelties like the self-starter and not like the Otto motor. elasticity of substitution illustrates well not only what has been done but also what could be done in this field under the given circumstances. 2 A historian, inspired by his observations about similar events in the past, might have expected that Walras would have missed the boat, that is, that his work, in an epoch that was able to understand him at last, would have been thrown into the limbo where dwell the works that, inadequately appreciated in their own time, were condemned on technical inadequacies of their apparatus when their real time had come. This was not so, however. The work on consumers’ behavior and on production that can be fitted into his system and that, in part, was fitted in by Pareto, instead of preventing him from taking his proper place, produced rather a modernized Walrasian system. This process extends from 1924, when Professor Bowley’s Mathematical Groundwork of Economics made Walras’ equilibrium system internationally accessible—already modernizing it in many spots—to 1939 when Professor Hicks’s Value and Capital, or the first two parts of it, completed the task. 3 To some extent this book was particularly successful in unearthing Walrasian problems of which Walras himself had not been aware. And, partly in its wake and partly independently, a rich stream of contributions was released from which I shall merely ‘read by name’ the works of Lange, Metzler, Mosak, and—only alphabetically last— Samuelson. Much or most of this work pivots around questions of determinateness and around stability conditions and thus constitutes the bulk of the work of our day in the field of fundamental theory or even Grundlagenforschung. 2 For a general survey of the nature and uses of total and partial elasticity of substitution, see again Allen, Mathematical Analysis, pp. 341–5, 372, 504, and 512. The concept, first introduced in its simplest form by Hicks (Theory of Wages, 1932) and Joan Robinson (Economics of Imperfect Competition, 1933), was immediately put to good use by both authors in the formulation of propositions that acquire delightful simplicity thereby (e.g., see also J.R.Hicks, ‘Distribution and Economic Progress: A Revised Version,’ Review of Economic Studies, October 1936). For a time the concept was therefore deservedly popular but this popularity was soon impaired by the related facts that it ceases to be so simple so soon as there are more than two goods or ‘factors’ under consideration and that it works with difficulty when applied to statistical data. I regret my inability to survey the results of the considerable literature to which the concept has given rise. See however the discussions in Review of Economic Studies, February 1934 and February 1936). Another example of this type of gadget is A.P. Lerner’s measure of monopoly power (‘The Concept of Monopoly and the Measurement of Monopoly Power,’ Review of Economic Studies, June 1934). 3 Even in those two parts, Hicks did much more than modernize Walras. So far as mere modernization goes, he also modernized Marshall, and I do not mean to suggest that ‘modernization’ describes those two parts of Value and Capital adequately. On the other hand, Hicks’s treatment is much too brief to accomplish the modernization of Walras and Marshall completely; it should be said rather that he produced essential material for it. Developments stemming 1115 [2. THEORY OF THE INDIVIDUAL FIRM AND MONOPOLISTIC COMPETITION] Equally important, however, and much more important as regards direct applicability to practical questions and hence for the economic profession as a whole, is another development that objectively stems from Marshall—the Theory of the Individual Firm and in connection with it, the Theory of Monopolistic or Imperfect Competition. 4 Everyone knows that this new arm of the economist’s analytic engine was added, in different forms, by English and American authors who worked independently of one another—a striking proof of the intellectual, still more than practical, need for this type of theory and a not less striking illustration of how the logic of the scientific situation may drive different minds along similar lines of advance. 5 In the United States, The Theory of Monopolistic Competition sprang, without any warning, fully armed from Professor E.H.Chamberlin’s head in 1933 6 and met with a 4 Still more than in other cases, I am anxious in this one to divest my emphasis upon a historical filiation from any semblance of derogation. This emphasis seems imperative because of two different sets of facts. First, Marshall, on the strength of his frequent use of the concept of the (small) individual industry, to which in particular most of his diagrams refer, has been sometimes accused of having neglected the economics of the individual firm. But as we have seen, and as analysis of his argument (and of such concepts as a firm’s special market or internal economies) could prove, he gave, on the contrary, quite unusual attention to the problems of the individual firm and offered suggestions that indeed called for development but force us, precisely because of this, to look upon later work, especially by Marshallians, as an offshoot from his. Second, Marshall’s concepts and treatment of the individual industry and of increasing returns invited criticism: their very shortcomings were fertile; they spoke with so certain a voice that the critic’s constructive task was cut out for him. 5 This and the fact that we have here to do with a broad movement in which many participate, though only few make the decisive hits that history records, stand out still more clearly if we take account also of the related literature on oligopolistic patterns. We then discern a similar movement in the Northern countries (see especially F. Zeuthen, ‘Mellem monopol og konkurrence,’ Nationalokonomisk Tidsskrift, 1929, and Problems of Monopoly and Economic Warfare, 1930) and in Germany (see von Stackelberg, Marktform und Gleichgewicht, 1934, who noticed and discussed most of the German as well as the non-German contributions). 6 Chapter 8 (on distribution), the contents of which were first presented in a paper read before the American Economic Association (at the meeting in Philadelphia, 1933) and which was then published in extenso in Explorations in Economics (in honor of F.W.Taussig, 1936), was added to the 2nd ed. of the book (1937). Chapter 7—the second chapter on selling costs—was omitted from the Ph.D. thesis handed in at Harvard on April 1, 1927 in order to meet the time limit, although it was fully worked out by then. The thesis does not differ in any essential from the 1st ed. of the book and, since it was in the stage of final revision for several months before, does not owe anything to Sraffa’s article (‘The Laws of Returns under Competitive Conditions’), which appeared in the Economic Journal, December 1926. The author proposed the subject for a Ph.D. thesis as early as 1921, when he was a student at the University of Michigan (author’s communication). In spite of subconscious influences that may have come from early Marshallian training, we therefore have here a striking instance of History of economic analysis 1116 corresponding success, which was as much due to the force and brilliance of his exposition as it was to the maturity of the scientific situation. The work claimed to reconstruct the whole of value theory by blending or fusing the hitherto separate theories of monopoly and competition. Nor was this all. It also claimed to teach a new economic Weltanschauung from the standpoint of which practically all economic problems appear in a new light. In any case the most important original contributions of the work—mainly contained in Chapters 4–7 on product differentiation and selling costs—met with very little fundamental dissent, if indeed with any. But a whole literature that amplified and applied these contributions followed in its wake. In England, Mrs. Joan Robinson’s Economics of Imperfect Competition, also in 1933, impinged upon a less unprepared profession and, for this and other reasons, was less spectacularly successful. As we know, Piero Sraffa, in 1926, had thrown out the idea that appeal to the theory of monopoly was the remedy for the difficulties about equilibrium that had arisen in connection with increasing returns. In doing so he had already suggested that actual conditions in industry will in general lie in the intermediate zone between monopoly and competition and that, since it was the competitive theory which held the field, it was then necessary ‘to turn towards monopoly.’ Finding monopoly thus released ‘from its uncomfortable pen’ (Robinson, op. cit. p. 4) in which it had existed, in seclusion from the main corpus of economic analysis, Mrs. Robinson proposed to reconstruct the theory of value by allowing monopoly to ‘swallow up the competitive analysis’—every firm being a monopolist, that is, a single seller of its own product, and competition coming in by bits until we reach the limiting case where a large number of such single sellers of perfectly substitutable products sell in a perfect market, and the demand for the product of each of them becomes perfectly elastic, the case usually described as perfect competition (op. cit. p. 5). 7 It should be observed that this concept subjective and objective originality—and of originality of the purely theoretical type that owed nothing to ‘the collection of direct empirical evidence,’ though a ‘guiding principle’ certainly was to create a theory that would fit facts better than what Chamberlin conceived to be the theory of competition current at that time (author’s communication). It seemed worth while to abandon, in this case, the principle of sketchiness that governs our exposition, especially in this Part, not only because of the importance of the book that—next to Keynes’s General Theory and with Hicks’s Value and Capital and Hayek’s contribution—must certainly be considered as one of the most successful books in theoretical economics that the period since 1918 has produced, but also because its author is not, like most authors mentioned in this book, beyond reach of personal interview. And personal contact, though only one of several methods for studying the ways of the human mind and especially the manner in which original work emerges and takes effect, is an important one and particularly useful in providing a check on the others. Three elements of scientific achievement are particularly obvious in this case: the maturity of the scientific situation; the ability to grasp an important idea with force and enthusiasm; and the ability to stay with it and to shut oneself off from the disturbing effects of other scientific ideas or aspects. 7 In her Foreword and Introduction, Mrs. Robinson not only acknowledged obligation to Marshall and Pigou but gave ample credit to Sraffa, to whose papers—both the Developments stemming 1117 of monopoly is not the traditional one. In fact the traditional concept can be satisfactorily defined only by the criterion that it admits the application of the Cournot-Marshall theory of monopoly. But this theory in turn presupposes the existence of a demand curve that is independently given and immune to influences from other firms upon the behavior of the one under consideration. Hence the traditional theory of monopoly is constitutionally unable to ‘swallow up’ any cases where these influences cannot be neglected, and hence the traditional concept of monopoly becomes inapplicable. English and the Italian contribution mentioned in Part IV, ch. 7, sec. 8d—it is therefore necessary to recur in all questions that touch upon her fundamental analytic intentions. This is rendered more difficult by the fact that Sraffa (see next sentence in our text) did not use the word Monopoly in the Robinsonian but in the usual sense. But she also acknowledged indebtedness to, or a sort of partnership in the spirit with, a number of other fellow economists of whom we must in particular notice three. There was Harrod, whose share in the analysis of impurely or imperfectly competitive patterns must be valued more highly than his papers (including his ‘Doctrines of Imperfect Competition,’ Quarterly Journal of Economics, May 1934; ‘Imperfect Competition and the Trade Cycle,’ Review of Economic Statistics, May 1936; and ‘Price and Cost in Entrepreneurs’ Policy,’ Oxford Economic Papers, May 1939) would in themselves indicate, especially considering the dates of their publication. And there were Shove and Kahn, whose names may, at some future time, owe the greater part of their recognition to Mrs. Robinson’s generous tributes. These tributes were fully deserved (as was Keynes’s tribute to Kahn, see below ch. 5). Both are scholars of a type that Cambridge produces much more readily than do other centers of scientific economics or rather of science in general. They throw their ideas into a common pool. By critical and positive suggestion they help other people’s ideas into definite existence. And they exert anonymous influence— influence as leaders—far beyond anything that can be definitely credited to them from their publications. I take this opportunity to mention a point on which Mrs. Robinson lays great emphasis in her Foreword and indeed throughout her book, the ‘marginal revenue curve.’ She gives credit, for both the thing and the word, to several of her contemporaries, particularly to Mr. Harrod and Professors Yntema and Viner. It is quite natural that use of this convenient tool suggested itself at that time to many (including Chamberlin), especially to those who had previously struggled with the clumsier Marshallian total curves. We must not, however, forget that the tool was first used by Cournot, and no author of the 1920’s or 1930’s can have any objective claim to it. History of economic analysis 1118 CHAPTER 3 [Economics in the ‘Totalitarian’ Countries] * NO EXPLANATION SHOULD be needed of what to some readers may seem to be an unjustifiable neglect of ‘totalitarian’ economic literatures. However I do wish to state that such neglect has nothing to do with political prejudice. I have no intention of neglecting any analytic work that has been done or is being done in ‘totalitarian’ countries, and the mere fact that such work is presented in the wrappings of a ‘totalitarian’ philosophy or even intended to serve and to implement it is no more reason for me to neglect it than my strong personal aversion to utilitarianism is a reason for neglecting the analytic work of Bentham. The various totalitarian philosophies themselves, however, are excluded—just as has been the utilitarian philosophy qua philosophy—not because they are ‘totalitarian’ but because they are ‘philosophies,’ that is, speculations that live outside the sphere of empirical science. In this respect we are merely carrying out a principle that has been followed all along and has been fully discussed in Part I, where the distinction between analytic economics and political economy was introduced mainly to give effect to it. Since this view is at variance with deeply rooted beliefs, the reader is invited to refresh his memory about what was said there. The principle above does not, however, fully explain why the economic literature introduced in totalitarian countries will not figure greatly in the sketch that is to follow. There are two other reasons for this: first, some of the most important contributions, such as von Stackelberg’s Marktform und Gleichgewicht (1934), or part of Del Vecchio’s work on money, have already been mentioned in Part IV, where we carried the histories of a number of topics down to the present; second, material of the kind that belongs in a history of economic analysis has not been plentiful under totalitarian regimes. For the rest, the cases of the three main totalitarian countries, Germany, Italy, and Russia, 1 are too different to be covered by a single generalization. * [This is the only chapter written for Part V, the subject matter of which is not mentioned in the Mexican lectures (ch. 1, sec. 2 above).] 1 Japan and Spain never were ‘totalitarian’ in any meaningful sense of the term. But as regards Japan, it should be observed that the interruption of contacts during the war and my ignorance of the language have created0 a lacuna which, in the time at my disposal, I have been unable to fill. All that prewar contacts enable me to say is that the importance of this lacuna is certainly not negligible and may be considerable. [In the two years since his death, former Japanese students have published translations or arranged for translations of all the books and long essays of J.A.S. This includes 1. GERMANY In Germany, methods of teaching and research had been rapidly improved in the period of the Weimar Republic (1918–32). Historical work and work on current problems (of the kind cultivated by the Verein für Sozialpolitik) went on as before; as noticed in Part IV, chapter 4, these types of work gradually lost their anti-theoretical methodological bent, and both interest and competence in ‘theory’ increased, the spreading use of Cassel’s treatise 2 being equally significant as an effect, a cause, and a symptom; in addition there were the autochthonous messages of such teachers as von Gottl, Liefmann, Oppenheimer, and Spann, to which even their most severe critic cannot deny the merit of having stimulated many minds; and there were, more accessible to Anglo-American understanding (and to my own), the performances of Diehl, Eucken, and others and, above all, those of Spiethoff and Sombart. The Viennese group, under the leadership of Professor L.von Mises, though it retained a vital individuality until it was, for the time being at all events, dispersed in the 1930’s, entered into closer relations than before with the rest of German economists and was thus in a position to assert its own distinctive doctrines. Two tendencies toward Americanization cannot be left out of this sketch. One was the inexorable progress of specialization. Though the comprehensive courses on general economics, economic (and social) policy, and public finance remained in their dominant positions, specialized groups began to acquire more and more definite existence; in a more significant sense than before, it the early work in German and this History. Das Wesen und der Hauptinhalt… (1908) and Theorie der Wirtschaftlichen Entwicklung (1912) were translated in 1936 and 1937.] 2 Gustav Cassel has repeatedly been mentioned before. It seems appropriate, however, to recall in this place the stages in the career of the most influential international leader of our science in the 1920’s—for such he was, whatever his critics (including myself) may say. We recall first the three pieces of work by which he established himself (his sketch of a theory of prices in the Zeitschrift für die gesamte Staatswissenschaft, 1899; the paper, not yet mentioned, on the causes of the variations in the general price level, ‘Orsakerna, till förändringar i den allmänna prisnivån,’ Ekonomisk Tidskrift, 1905; and The Nature and Necessity of Interest, 1903). Partly owing to the advantage he held as a ‘neutral,’ he rose into international fame during and after the First World War—chiefly as an expert on money and international relations and as an assiduous participant in international conferences on these subjects (I mention only, as a sample, Money and Foreign Exchange after 1914, 1922: today’s monetary experts could do worse than study this book). Finally, he made a great success, at least outside the sphere of orthodox socialism, by his treatise, Theoretische Sozialökonomie, which, guided by chance or else very shrewd insight, he published in German (1918). I am myself the author of a review of the 4th ed. (1927) that was as far from favorable as was Wicksell’s review appended to the English ed. of vol. I of the latter’s Lectures. I do not think that either Wicksell or I said anything that ought to be retracted. But we overlooked something, viz., that the book was exactly what German economists needed. Economics in the ‘totalitarian’ countries 1120 was possible, after 1918, to speak of agricultural, or labor, or industrial economists. Again, research institutes like the National Bureau of Economic Research or the Bureau of Agricultural Economics grew up both inside and outside the government departments concerned with economic problems. It will suffice to mention the Institut für Weltwirtschaft at the University of Kiel, founded by one of the most efficient organizers of research who ever lived, Professor Bernhard Harms, and the Institut für Konjunkturforschung, the foundation of a similarly efficient organizer, Professor Ernst Wagemann in Berlin 3 —both of which added new economic journals to the existing ones. There is only one point in this picture of considerable advance and still more considerable promise that, from the standpoint taken in this book, must be registered as of sinister import. As the Weimar Republic settled down, the governments of the individual states— there were neither federal nor private universities, only state universities—yielded increasingly to the demand of political parties, mainly the Social-Democratic and the Centrist parties, that the appointments to professional office in economics should take account of the politics of candidates. The argument, discussed quite openly, ran as follows: economics, unlike physics and like philosophy, is a Weltanschauungswissenschaft, that is, a ‘science’ into the research and teaching of which necessarily enter the ultimate beliefs and allegiances of the investigator or teacher. These ultimate beliefs and allegiances were embodied in the socialist and the centrist (Catholic) parties and in an agglomeration of all other parties defined by the negative characteristic that they were neither socialist nor Catholic; hence, professorships should be divided up, as equally as possible, between members of these three political groups, though nobody advocated, to be sure, that this should be done irrespective of qualifications. There is no need to discuss this matter again. Less openly, the tendencies verbalized by the theory of the Weltanschauungswissenschaft make themselves felt under any circumstances and in all countries. Also they never prevail fully anywhere—except in modern Russia. In the Weimar Republic the resistance of the faculties and of upright members of the bureaucracy kept them in relatively narrow bounds. Under these circumstances the advent of National Socialism did not mean quite as great a break and did not cause all the damage that a foreign observer might expect. The National Socialist regime was intolerant not only of criti- 3 The latter started frankly from the Harvard barometer curves, though its work, mainly statistical, soon expanded—much as did the work of the Harvard Economic Society—far beyond what these curves imply. This institute was perhaps the most important single influence in spreading knowledge of modern statistical methods (as then understood). Its methodological work is therefore of historical importance. It is mainly contained in supplements to the institute’s journal, the Vierteljahrshefte zur Konjunkturforschung; see e.g. supplements no. 4 on The Analysis of Economic Curves (H.Hennig); no. 6 and 11 on Seasonal Variations (O.Donner); no. 9 on Trend (P. Lorenz); and no. 12 on Russian Contributions (A.L.Wainstein, S.A.Perwuschin, M.W.Ignatieff). There was close co-operation with the Federal Bureau of Statistics (Statistisches Reichsamt), which published a series of monographs in addition to its current publications. History of economic analysis 1121 . the demand of political parties, mainly the Social-Democratic and the Centrist parties, that the appointments to professional office in economics should take account of the politics of candidates type of gadget is A.P. Lerner’s measure of monopoly power (‘The Concept of Monopoly and the Measurement of Monopoly Power,’ Review of Economic Studies, June 1934). 3 Even in those two parts,. imperative because of two different sets of facts. First, Marshall, on the strength of his frequent use of the concept of the (small) individual industry, to which in particular most of his diagrams