History of Economic Analysis part 118

Chia sẻ: Mi Sa | Ngày: | Loại File: PDF | Số trang:10

Thêm vào BST

Báo xấu

47
lượt xem 3
download

Download Vui lòng tải xuống để xem tài liệu đầy đủ

History of Economic Analysis part 118. At the time of his death in 1950, Joseph Schumpeter-one of the major figures in economics during the first half of the 20th century-was working on his monumental History of Economic Analysis. A complete history of humankind's theoretical efforts to understand economic phenomena from ancient Greece to the present, this book is an important contribution to the history of ideas as well as to economics.

Chủ đề:

Bình luận(0) Đăng nhập để gửi bình luận!

Lưu

Nội dung Text: History of Economic Analysis part 118

History of economic analysis 1132 ‘business cycle research without theory,’ which flared up occasionally, was of a nature similar to the one that arose about the work of Mitchell and the National Bureau of Economic Research and will hence be touched upon together with the latter. The importance of the work of Wesley Clair Mitchell and the National Bureau of Economic Research, which he led and inspired, has been emphasized already. Just as Professor E.Wagemann said somewhere that the publications of his Institut für Konjunkturforschung were simply the second volume of his Allgemeine Geldlehre (1923), so Mitchell might have said that (most of the) publications of the National Bureau formed all together a huge second volume to the first he had published in 1913. And his volume of 1927, Business Cycles: The Problem and its Setting, was, like Wagemann’s Konjunkturlehre of 1928, an organizing survey of problems, viewpoints, and materials— for work which it was given to him to carry, if not to completion, at least as far as his (and A.F.Burns’s) monumental Measuring Business Cycles (1946). We cannot enter into a discussion of what is known as the National Bureau method of depicting cycles statistically. All we can do is to point out that this effort of establishing and of marshalling a vast amount of (primarily) statistical material essentially continues the plan that was partially executed in the book of 1913 and owes nothing to macrodynamic theory, though it may eventually set problems and provide important checks for it: the work of Mitchell and his group aims primarily at showing what it is we have to explain and, beyond this, suggests viewpoints from which to do so. I use this opportunity for a brief comment on the little controversy about methodology adverted to above. Mitchell might have done something toward preventing it, had he distinguished more clearly between theory in the sense of explanatory hypothesis and theory in the sense of analytic apparatus. Most of us would agree with him if he felt that the formulation of explanatory hypotheses should wait upon acquisition of a fuller command of the facts and that the explanatory hypotheses, so far offered, old and new, lacked proper substantiation and might be unable to stand up in the light of the facts he was going to assemble. Even so he displayed no active hostility to the many ‘theories’ of business cycles which he listed in his book of 1927 with perfect detachment. But in addition he cared little for the technical refinements of ‘theory’ in the instrumental sense of the term, just as he cared little for the modern refinements of statistical method. His early associations with Veblenite tendencies did the rest to make him appear, in the eyes of the profession, as more of an anti-theorist than he was—and still more so in the eyes of those votaries of macrodynamics for whom economic theory and the mathematical model tend to be synonymous. But actually, in intention as well as in fact, he was laying the foundations for a ‘theory,’ a business cycle theory as well as a general theory of the economic process, but for a different one. Similarly, the Harvard Committee, in professing to proceed without theory, really meant no more than that they did not intend to be guided in their factual work by preconceived explanatory hypotheses. But business cycle research is research into sequences of business situations which are also the subject matter, or part of the subject matter, of macrodynamics. Co-operation between the two was thus obviously indicated. All students of business cycles, not debarred by mathematical disabilities, should have recognized this from the first. The formal logic of lags, rates of change, cumulations, and of the oscillations they may produce is bound to be helpful in the interpretation of the observed behavior of time-
Dynamics and business cycle research 1133 series material. Macrodynamics should be not less helpful in any attempt to put the existing theoretical material into a more promising shape, for instance, in deciding questions of determinateness and in formulating conditions of damping or explosiveness and the like. Problems of the mechanisms by which impulses are propagated through the economic system may be cleared up by macrodynamic methods, which therefore may contribute substantially, among other things, to our understanding of cyclical turning points.17 The star example by which to demonstrate the usefulness of these methods is the theory of oscillators, that is, of factors that create fluctuations in the system, although they are perfectly steady—free from fluctuations—themselves.18 ‘Literary’ students of business cycles will not easily see the possibility of this. They will be prone to argue that no factor can contribute to cyclical fluctuations unless its own time series is oscillatory. And so they might be expected to show some signs of gratitude to macrodynamics for thus widening their horizon as they should in other instances for having their arguments sharpened and corrected. If they do not always do so, this is no doubt primarily owing to mathematical disabilities. But there is also another reason which it is important to state. It has been said above that macrodynamics helps us to understand mechanisms of propagation. It will perhaps assist the reader if he will look upon the economic system as a sort of resonator, which reacts to the impact of disturbing or ‘irritating’ events in a manner that is partly determined by its physical structure. Think for instance of a violin that ‘reacts’ in a determined manner when ‘irritated’ as the player applies the bow. Understanding the ‘laws’ of this reaction contributes to a complete ‘explanation’ of the phenomenon that we call a violin concert. But evidently this contribution, even if reinforced by the contribution of the neurophysiologist, does not explain the whole of it: aesthetic evaluation and the like apart, there is a range of purely scientific ground that acoustics and physiology are constitutionally unable to cover. Similarly macrodynamics, while quite essential to an explanation of cyclical phenomena, suffers from definite limitations:19 its cyclical models are what acous- 17 An instructive example of this type is the discussion between Professors R.Frisch and J.M.Clark concerning the relation between the turning points of consumption and the production of capital goods in the cycle (‘The Interrelation between Capital Production and Consumer-Taking,’ ‘Reply,’ ‘Rejoinder,’ and ‘A Further Word,’ Journal of Political Economy, 1931–2). 18 A mechanical model will illustrate this phenomenon. Let an electric clock be placed upon a somewhat rickety table. The electric current that keeps the clock going is perfectly steady. Yet it may produce an oscillatory movement of the table. 19 The simile limps, of course, like all similes. And so does the following suggestion which is not a simile. Cycles run their courses in the historical evolution of the capitalist economy. Even neglecting all the economic sociology that must therefore inevitably enter into their explanation, we cannot help recognizing that their theory or, to avoid this word, their analysis must be largely bound up with the theory or analysis of evolution rather than with dynamics, which is the theory or analysis of sequences that do not carry any historical dates. No doubt there are certain mechanisms that played as great a part in 1857 as in 1929. And these must be taken account of in any observed
History of economic analysis 1134 tic models of resonators are for the violin concert. But its votaries will not see this. They construct macrodynamic models that are to explain all there is to explain, for economists, in the cyclical phenomena. The very attempt to do so involves several definite errors of fact.20 And flimsy structures based upon arbitrary assumptions are immediately ‘applied’ and presented as guides to policy, a practice that of course completes the list of reasons for irritation in the opposite camp. One sometimes has the impression that there are only two groups of economists: those who do not understand a difference equation; and those who understand nothing else. It is therefore a hope, rather than a prognosis to be presently fulfilled, which I am expressing if I venture to say that this entirely unnecessary barrier—but one which is no novelty in our science—to fertilizing interaction will vanish by virtue of the logic of things. I have still to advert to a promising branch of dynamics that is not indeed microeconomics, because it does not reach the individual deciding agents, but is not macroeconomics either, because its models do not embrace the whole of the economy: it is akin to Marshall’s partial analysis and is (mostly) concerned with individual industries. The famous corn-hog cycle is the best-known instance: if farmers, under the influence of a favorable relation between the price of hogs (pork) and the cost of rearing them (price of corn), all decide at roughly the same time to increase their hog production and if, as will be the case in this instance, they all come out at roughly the same time with an increased supply of hogs, this may cause a sharp drop in the price of pork (and also a rise in the price of corn), which might induce a majority of them to contract their production, which would recreate favorable conditions that would in turn lead to another expansion in hog production. The resulting cycle may of course be damped, explosive, or stationary, and a very simple general model can be set up to describe this mechanism that is indeed observable, not only on the hog market but in a large class of cases.21 cycle by more or less generally applicable macrodynamic schemata, just as must, on a lower level of technique, the ordinary theory of supply and demand. But they are only tools and do not in themselves suffice, even if supplied with all conceivable time series, to reconstruct the phenomenon as a whole and, of course, still less its long-run outcomes. 20 Three of these may serve as illustrations. They will at the same time show why the respective objections do not tell against the models themselves but only against the claim alluded to. (1) Macrodynamic models, presented with that claim, involve the proposition that the ‘causes’ of business cycles must be found in the interaction between the social aggregates themselves, whereas it can be proved that business cycles arise from sectional disturbances. (2) With the same proviso, macrodynamic models carry the implication that the structural changes that transform economics historically have nothing to do with business cycles, whereas it can be proved that cycles are the form that structural changes take. (3) Constructors of macrodynamic models, almost always, aim at explaining all the phases of cycles (and the turning points) by a single ‘final’ equation. This is indeed not impossible. But it spells error to assume that it must be possible and to bend analysis to this requirement. 21 The reader is referred to M.Ezekiel, ‘The Cobweb Theorem,’ Quarterly Journal of Economics, February 1938 (reprinted in the volume, Readings in Business Cycle
Dynamics and business cycle research 1135 Another famous instance, displaying the phenomenon for durable goods, is Professor Tinbergen’s shipbuilding cycle.22 On the one hand, it stands to reason that no great confidence can be placed in the results that such schemata yield—more apparently than really—and that extreme care is imperative in applying them, if indeed they be applicable at all to any practical cases at all. Thus, readers of Professor Tinbergen’s paper will note with concern the formidable list of assumptions contrary to fact that they are asked to accept. But even if they accept them all, they will find it difficult to reconcile themselves to the complete neglect of all the influences upon shipbuilding that other industries and general business conditions are bound to exert; and they may see in the basic graph (op. cit. p. 154) more traces of the business cycle than of the mechanism that the schema isolates. On the other hand, however, schemata of this kind are first steps toward a more perfect dynamic theory and must therefore be listed as pioneer ventures of first-order importance: the same reader who is impressed with their shortcomings—much as he would be in reading a description of Columbus’ flagship—should be also impressed with the fact that an element of the mechanism they describe is undoubtedly present in almost every practical case and, furthermore, with the host of well-defined tasks that they suggest for further work on the same line. Such work cannot at present claim to be more than exploratory. But it explores the ground on which a new structure will stand some day. Theory), where he will find all that is necessary including almost all the relevant literature. 22 J.Tinbergen, ‘Ein Schiffbauzyklus?,’ Weltwirtschaftliches Archiv, July 1931. The model is very interesting. Let currently available tonnage of freight-carrying vessels be represented on a time axis. Call it f(t) and assume, as a first approximation, that it varies only in consequence of new tonnage produced, which we can therefore denote by f′(t). Postulate that freight charges will be high (low) when tonnage is low (high) relatively to its trend which will stimulate (discourage) orders for new tonnage, the execution of which will discourage (stimulate) further orders and so on. Increase of tonnage at any point of time will thus depend upon the relative scarcity or abundance of tonnage some time (say years) before: where a is a constant that represents the intensity of reaction. This is a mixed difference and differential equation, the first of its type to enter economic theory. Its solution will describe the development in time of the tonnage (theoretically ever after), if the development in an initial interval be given. According to a standard method much used by physicists, we get the solution by means of the (tentative) substitution, f(t)=eαt+β. Mathematically trained readers will notice that this solution will be periodic if we make a an imaginary number (Euler’s relation: eiat=cos αt+i sin αt). [See J.A.S., Business Cycles p. 533.]
CHAPTER 5 [Keynes and Modern Macroeconomics]1 IN A HISTORY of economic analysis, it is from the standpoint of modern macro- economics that we must look upon the greatest literary success of our epoch, J.M.Keynes’s General Theory of Employment, Interest and Money (1936), and it is from this standpoint only that we can attempt to do justice to it. From any other view, this inevitably spells injustice. Like most of the great economists whose messages reached the general public, especially like A.Smith, Lord Keynes was much else besides being a worker in the field of economic analysis. He was a forceful and dauntless leader of public opinion, a wise adviser to his country—the England that was born in the First World War and that afterwards kept, with deepening lines, the social physiognomy then acquired— and a successful representative of her interest, a man who would have conquered a place in history even if he had never done a stroke of specifically scientific work: he would still have been the man who wrote The Economic Consequences of the Peace (1919), bursting into international fame when men of equal insight but less courage and men of equal courage but less insight kept silent.2 His General Theory, in a sense, was a similar feat of leadership. It taught England, in the form of an apparently general analysis, his own personal view of her social and economic situation and also his own personal view of ‘what should be done about it.’ In addition, impinging as it did upon the moral atmosphere created by the depression and upon a rising tide of radicalism, the message of the book, issued from the vantage ground of Cambridge and propagated by many able and faithful disciples, met with equal success elsewhere and particularly in the United States. Considering that Lord Keynes’s attitude 1 [This was the last thing written by J.A.S. for his History. It was left behind to be typed when he departed from Cambridge for the Christmas vacation, December 1949. It was not typed until after his death. Hence there was no opportunity for corrections or modifications.] 2 In Parts II, III, and IV, I have occasionally attempted to sketch personalities as personalities. This cannot be done in this brief survey. Therefore I shall merely add that the tribute above fails to convey a picture of the man or even the wealth of his interests. Even his purely scientific work will not enter our picture in all its aspects. I have described the words above as a tribute. But behind this tribute there is a much ampler one that remains unwritten here. [See ‘John Maynard Keynes (1883– 1946),’ written by J.A.S. for the American Economic Review, September 1946, reprinted in Ten Great Economists (1951).]
Keynes and modern macroeconomics 1137 was rather conservative in many respects, especially in matters touching freedom of enterprise, this might seem surprising. But it must not be forgotten that he rendered a decisive service to equalitarianism in an all-important point. Economists with an equalitarian bent had long before learned to discount all other aspects or functions of inequality of income except one: like J.S.Mill they had retained scruples concerning the effects of equalitarian policies upon saving. Keynes freed them from these scruples. His analysis seemed to restore intellectual respectability to anti-saving views; and he spelled out the implications of this in Chapter 24 of the General Theory. Thus, though his scientific message appealed to many of the best minds of the economic profession, it also appealed to the writers and talkers on the fringes of professional economics who gleaned nothing from the General Theory except the New Economics of Spending and for whom he brought back the happy times of Mrs. Marcet (see Part III, ch. 4) when every schoolgirl, by learning the use of a few simple concepts, acquired competence to judge of all the ins and outs of the infinitely complex organism of capitalist society. Keynes was Ricardo’s peer in the highest sense of the phrase. But he was Ricardo’s peer also in that his work is a striking example of what we have called above the Ricardian Vice, namely, the habit of piling a heavy load of practical conclusions upon a tenuous groundwork, which was unequal to it yet seemed in its simplicity not only attractive but also convincing. All this goes a long way though not the whole way toward answering the questions that always interest us, namely, the questions what it is in a man’s message that makes people listen to him, and why and how. However, our only task is to insert into our survey Keynes’s contribution to our analytic apparatus. But the importance of his work seems to impose the duty, before doing this, of presenting a few comments on its wider aspects. [1. COMMENTS ON THE WIDER ASPECTS OF KEYNES’S WORK] First, Keynes’s work presents an excellent example for our thesis that, in principle, vision of facts and meanings precedes analytic work, which, setting in to implement the vision, then goes on hand in hand with it in an unending relation of give and take. Nothing can be more obvious than that in the beginning of the relevant part of Keynes’s work stood his vision of England’s aging capitalism and his intuitive diagnosis of it (which he followed up without the slightest consideration of other possible diagnoses): the arteriosclerotic economy whose opportunities for rejuvenating venture decline while the old habits of saving formed in times of plentiful opportunity persist. This vision was clearly formulated in the first pages of the Economic Consequences of the Peace (1919) and adumbrated with increasing clearness in successive works, especially in the Tract on Monetary Reform (1923) and the Treatise on Money (1930), Keynes’s most ambitious purely scholarly venture. This Treatise, though no failure in the ordinary sense of the term, met respectful but damaging criticism and, above all, failed to express Keynes’s vision adequately. Thereupon, with admirable resoluteness, he determined to throw away the impeding pieces of apparatus, and bent to the task of framing an analytic system that would express his fundamental idea and nothing else. The result, given to the world in 1936, seems to have satisfied him completely, so much so that he felt himself to have led economics out of 150 years of error into the land of definitive truth—a claim that cannot
History of economic analysis 1138 be put to test here but was as readily accepted by some as it discredited his work in the eyes of others. Next, we must record Keynes’s acknowledgments of indebtedness, which in all cases can be independently established, to Mrs. Joan Robinson, Mr. R.G. Hawtrey, Mr. R.F.Harrod, but especially to Mr. R.F.Kahn, whose share in the historic achievement cannot have fallen very far short of co-authorship. I take this opportunity to rescue from threatening oblivion, in addition to his share in Keynes’s General Theory and in the theory of imperfect competition, another contribution of Kahn’s. Marshall, though offering plenty of material about the theory of short-run processes, always emphasized primarily the properties of the long-run normal without, perhaps, making it sufficiently clear that what he really meant was the pure logic of the economic process rather than any state of things that will actually emerge at any future time. It was necessary to realize that what in fact emerges and can be observed is the result of a succession of short-run events and short-run responses to them and will in general bear little resemblance to the perfect equilibrium that would emerge if time were given for everything to work itself out without any further disturbances occurring meanwhile. This point of view, obviously very important for the improvement of economic analysis, has been taken consistently by Mr. Kahn, more consistently and consciously than by anyone else, I believe, although I am unable to put my finger on any particular publication of his that would substantiate this assertion. (On the possible relation of this scientific contribution to the short-run philosophy of our age, see above Part IV, ch. 7.) Third, Keynes must be credited or debited, as the case may be, with the fatherhood of modern stagnationism. In itself stagnationism is practically as old as economic thought. In any prolonged period of economic malaise economists, falling in like other people with the humors of their time, proffer theories that pretend to show that depression has come to stay. We have had instances before. But so far as our own epoch and scientific literature are concerned, this attitude can be traced as we have seen to Keynes’s Economic Consequences of the Peace. In the United States, very naturally, it did not ‘catch on’ until the crisis of 1929–32, but in the aftermath of this crisis it caught on with a vengeance. A group that might almost be called a school and that found resonance in almost all the strata of public opinion—the opinion of the harassed business community included—rose to scientific importance under the brilliant leadership of Professor Alvin H.Hansen, who amplified and expanded the doctrine of the mature or stagnating economy in part on different grounds than Keynes. We cannot attempt a critical analysis of it and shall confine ourselves to noting that it stood up better than might have been expected in the face of apparently contradicting evidence for three reasons: (1) because the new economic opportunities that are unfolding themselves may be, in part with justice, attributed to the consequences of the Second World War and thus be interpreted as an intermezzo that is irrelevant to questions of fundamental trend; (2) because every span of prosperity, however prolonged, displays setbacks that it will always be possible to interpret as manifestations of that trend; (3) because some workers who are not ‘stagnationists’ in either Keynes’s or Hansen’s sense nevertheless arrive at a similar result for reasons of their own.3 It sometimes looks as if we ought to speak not of stagnationists and anti-stagnationists but rather of two different lines of a single stagnationist argument—at least if we neglect all those anti-stagnationist writers who confine themselves to criticizing individual stagnationist arguments.
Keynes and modern macroeconomics 1139 Finally, fourth, let us note the significant fact—significant in that it shows the extent to which Keynes’s General Theory was a response to widely accepted ideas—that in the 1930’s other works appeared that, each in its own way, attempted to express views that were similar to Keynes’s in important points. An enthusiastic Keynesian has for instance spoken of ‘Swedish stepping stones to Keynes’ and, if we neglect the value judgment which this phrase implies, we may indeed agree that the leading Swedish economists, in particular Lindahl, Myrdal, and Ohlin, developing certain pointers of Wicksell’s, built with similar materials according to a similar plan. However, I shall merely mention two works that will illustrate what I mean. Erik Lundberg’s Studies in the Theory of Economic Expansion (1937) appeared a year after Keynes’s General Theory, took full account of the latter, and contains explicit acknowledgment of its ‘stimulating influence.’ But no work of this range and depth can, within a single year, be formed by an out- side influence unless its author has arrived at somewhat similar conclusions by himself. In addition, Wicksell’s influence is much more obvious than Keynes’s, and Lundberg’s work, both in methods and results, differs sufficiently from Keynes’s to put his fundamental independence from the latter beyond doubt. Indeed, except for effectiveness of presentation, we might well speak of superiority, especially (but not only) because Lundberg tackled from the first the problem of sequence which had to be done for Keynes by followers. For us the book is particularly interesting because it displays the micro- and macrodynamic roots of current Keynesianism much better than did Keynes himself. And for the post-Keynesians of our day it should be particularly interesting because of the enlightening experience that it may afford of seeing ‘Keynesian’ propositions in a different light and in different connections. Carl Föhl’s book, Geldschöpfung und Wirtschaftskreislauf (1937), owes nothing to Keynes’s General Theory because, as the author stated in the preface, his manuscript had 3 Thus, it is possible to feel unconvinced by Keynes’s and Hansen’s arguments and nevertheless to predict that capitalist evolution tends to peter out—i.e. to settle down into a condition that might be just as well described as ‘stagnation’—because the modern state may crush or paralyze its motive forces. Modern taxation is only an example of the numerous factors which work that way, all of which can be established by an analysis of the present state of England. And inhibitions of this kind—which moreover can also be shown to be the inevitable outcome of capitalist history—will do as well as the factors emphasized by Keynes and Hansen: evidently it comes to the same thing, in a profit economy, whether the objective opportunities for gainful enterprise decrease or the profits after having been made are taxed away. Let us note in passing that there is, in some points, a strong affinity between the Keynes-Hansen and the Ricardo-J.S.Mill argument concerning the advent of a stationary state. This is particularly clear in the case of Keynes, who in his earlier writings repeatedly spoke of a ‘decreasing response of nature to human effort’—on the eve of a period of unsaleable foodstuffs and raw materials—and of pressure of population. This element is not only absent from Hansen’s argument but has been by him actually turned into its opposite. But the notion that opportunities for investment will in the future tend to decline as compared with people’s propensity to save, though handled in a manner different from Ricardo’s, is present with both authors [Hansen and Keynes]. The main difference is that they predicted difficulties in the process of the economy’s settling down to a stationary state that did not occur to Ricardo. [J.A.S. in his Capitalism, Socialism, and Democracy (1942) put forward the point of view ‘that capitalist evolution tends to peter out because the modern state may crush or paralyze its motive forces’]
History of economic analysis 1140 been completed in December 1935 so that he was not able to do more than to add references to the General Theory here and there. All the more striking are a number of parallelisms between his and Keynes’s propositions though, for English and American readers, the full extent of actual agreements will not be obvious at first sight. This is owing to two facts: Dr. Föhl used a different conceptual apparatus and arrived at his conclusions by methods that tend to obscure those agreements; and writing in a different environment he gave much space to problems that are no longer interesting to the American profession. Precisely because of this, the study of this book would be extremely instructive to American economists: precisely because of its apparently un- Keynesian approach, it reveals (objective) doctrinal relations and sheds what amounts to new light on several Keynesian problems, especially the problem of equilibrium underemployment. This book has had some influence in Germany and, so I have been informed by a Danish fellow economist, a considerable influence in Denmark. [2. THE ANALYTIC APPARATUS OF THE GENERAL THEORY] The analytic apparatus of the General Theory is, first, essentially static. We shall explain presently the apparent paradox that its place in the history of analysis is nevertheless bound up with the impulse it gave to macrodynamics. Nor do I mean to deny that large parts of the book—some would say, its most valuable parts—are devoted to dynamic considerations. But these were added to a skeleton4 that was severely static, so much so as to neglect, on principle, all sequences and periods.5 Second, this static theory is not the statics of long-run normals but the theory of short-run equilibria. Third, the most important point in this connection is that, of all the aspects of the investment proc- 4 This skeleton has been exactly formulated many times. We content ourselves with mentioning O.Lange ‘The Rate of Interest and the Optimum Propensity to Consume,’ Economica, February 1938, and L.R.Klein, The Keynesian Revolution (1947). 5 The outstanding example of this is the Kahn-Keynes multiplier. [J.A.S. planned to discuss this in the later portion of this chapter which was not finished.]
Keynes and modern macroeconomics 1141 ess, it is only the expenditure effect of new investment which enters the model (not the book): as Keynes himself rightly emphasized, physical capital (equipment) is assumed to remain constant throughout, both in kind and quantity. This limits the theory to an analysis of the factors that determine the higher or lower degree of utilization of an existing industrial apparatus. Those who look for the essence of capitalism in the phenomena that attend the incessant recreation of this apparatus and the incessant revolution that goes on within it must therefore be excused if they hold that Keynes’s theory abstracts from the essence of the capitalist process.6 Fourth, though aggregative, Keynesian analysis—no doubt for the sake of simplicity—presupposes ‘free,’ if not actually ‘pure,’ competition in all commodity and factor markets. Fifth, everybody is supposed to react to a particular kind of ‘real’ values, namely, to prices expressed in wage-units or prices divided by an average money wage per unit of labor, which is determined by bargains between employers and employees—a well-nigh desperate measure of simplification that makes results incomparable as between two different points of time unless wage rates are the same in both. But there is an important exception to this postulate that people calculate in terms of real values in this sense: workmen do so only in so far as they save and invest but not in their bargains about their labor; when they negotiate wage contracts they consider exclusively money wage rates.7 Within the framework set by these five points, Keynesian analysis—the analysis of current national income—works five endogenous variables, that is, variables which the system is to determine: national income itself, employment, consumption, investment, and the rate of interest; and one exogenous variable that is given to the system by the action of the ‘authorities,’ quantity of money.8 Employment may be allowed to drop out on the strength of the 6 This does not preclude us from finding several points of contact between Keynesian and Marxian analysis. Fundamentally, however, they are opposites. 7 This raises three questions: (1) the question of the realism of this postulate per se; (2) the question of the warrant for making this exception from the rule that is adopted for all other transactions; (3) the question of the effects upon Keynes’s wage theory of this postulate. In the available space we cannot answer any of them—beyond pointing out that it was by virtue of this postulate that Keynes rejected the usual theory of the supply function for labor, which is based on the opposite postulate. Slightly exaggerating the importance of this point, Keynes calls it ‘our point of departure from the classical system’ (op. cit. p. 17). But it is true that, as his argument is laid out, it is this postulate which enables him to defend what is for his system a fundamental proposition, namely, that the wage contract does not in principle determine real wages. Followers have, gradually and tacitly, receded from this untenable position, which is less essential for his argument than Keynes himself thought. Observe, however, the tenable element in it: any increase or decrease of wage rates, if it increases or decreases wage incomes in a sufficiently important sector of the economy (or even locality), will have some influence upon prices and this influence may, in part or wholly, offset the effect of the change in money wage rates—a nexus which is of course worth emphasizing. 8 One of the many inadequacies of our exposition, which cannot aim at more than calling up a few essentials in the minds of readers who are supposed to be familiar with the General Theory, is that we must follow the practice of many Keynesians in