I suppose that philosophy is still queen of the social sciences, but I fear that economics has become the knave. The awed regard in which the discipline was once held has given way to something bordering on disrespect. This change in public esteem primarily arises from the inability of the economics profession to give a coherent explanation for, and therefore to prescribe a workable policy against, inflation. But this is by no means the only problem on its hands. Following the advice of the profession, the dollar was finally allowed to “float” in 1971, but far from removing the dollar crisis, floating seems to have worsened it. Heeding the best economic strategies, we have sought for an entire generation to spur economic development, but the World Bank has recently admitted that 40 percent of the population of the “developing” lands still lives in what the Bank calls “absolute poverty.”1 Meanwhile economists confess to their ignorance and impotence with regard to the decline in American productivity, the emergence of an alienated and unemployed underclass, spreading urban blight, threatening technological displacement, and murky economic prospects.
Many economists blame the disappointing performance of their discipline on the constant intervention of government in economic affairs, to the despair of predictors and policymakers alike. There is, of course, some merit to this contention. But I believe a deeper reason for the poor showing of the profession is the extreme fragility of the theoretical foundations on which economic prediction and policy both ultimately rest. At least part of the recurrent failure stems from the fact that economists conceive of the world in terms that fail to grasp its essential characteristics or that seriously misrepresent them. It is this shaky underpinning that I wish to examine before discussing a new departure in economic thinking that is the subject of the book at hand.
A good place to start is with the words that economics has imported from physics and made fashionable throughout social science: “micro” and “macro.” Microeconomics is concerned with aspects of the economy that are centered in the act of choice, allocation, decision-making. Macroeconomics is devoted to the performance of the economy as a whole, especially with regard to employment and output and inflation. This seems, on the surface, like a very convenient way of examining the economy from two different vantage points, micro yielding a worm’s eye view, macro a bird’s eye view. But what is strange is that there is no way of going from one view to the other. One would think that by opening up the worm’s eye lens one would eventually take in the entire flow of output or employment that originates in the “micro” acts of individuals or firms—but no such comprehensive view emerges, only a blur. Conversely, it would appear that by closing down the macro lens we could bring into sharp focus the individual actions that are the constituent elements of the flow of output or the rise in prices, but again no such picture emerges: the macro lens simply cannot distinguish the individual actors. Thus macro and micro are not the complementary slides of a stereopticon giving us a single complete picture from two incomplete ones. They are, rather, two quite different pictures that cannot be combined. What, then, is economics?
A similar confusion surrounds what is surely the single most important concept in economics, the idea of price. Prices enable us to make rational choices, or guide entrepreneurs to making a profit. Economics is sometimes even called “price theory.” But what is a price? It is certainly not just a dollar sign affixed by the workings of history to a bushel of wheat or a ton of steel, for such prices, established by tradition or sheer power or mere accident, would have no logical structure, and could not therefore serve as guides to rational action. But economists have always been fascinated with prices just because they have believed that prices were not established by tradition or chance, but reflected something real though hidden about the social world, something that imposed an orderly arrangement on economic life, just as atoms and molecules impose orderliness and regularity on the world of matter.
What is that real, but hidden, structure? The classical economists, from Adam Smith through Marx and Mill, thought it was the amount of labor-energies embodied in various commodities, so that the prices of things measured their “weight” in terms of labor input. This concept brought an immense clarification into the social world, but also immense complication. For one thing, not all labor was alike, so that one had to find a way of comparing the surgeon’s input with the street sweeper’s. For another, there were always difficult problems in assessing the effect of capital and resources on price. Not even the purest “labor value” theorist ever held that labor alone determined prices.
Modern economics has wholly eschewed the idea that a common substratum of labor might give coherence to the array of prices. Instead it has elaborated the view first put forward in the last half of the nineteenth century by Hermann Gossen, W. Stanley Jevons, Carl Menger, and Léon Walras—a quadrumvirate dubbed the Gang of Four by one of my students. The Four, working independently, asserted that the hidden rationale of prices was not their labor inputs but their psychological outputs—the “utilities,” or satisfactions, that consumers derived from commodities. Thus Bentham’s principle of universal pleasure and pain became the central unifying concept of economics, and prices were seen as a representation in dollars of the Felicific Calculus that Bentham believed to be the hidden regulator of all human behavior.
The utility theory of price got rid of all the difficulties of a labor theory of value but it brought grave problems of its own. For the utilities of different goods, unlike their embodied labor, were not located within the goods themselves. The utilities of diamonds or water did not, so to speak, radiate out from them because of their properties as things. Rather, utility was clearly understood to be a psychological relationship between the commodity and its user, so that water might possess a great deal of utility in the desert and none at all in a puddle. But this relational aspect of utility now posed a very serious issue: how was economics to add up the pleasures of commodities, all of which resided in the relationship of goods to people and none of which resided in the goods themselves? In other words, how was the utility of my drink of water to be added to yours? And if economics could not add up these utilities, what objective meaning could it attach to a collection of objects with price tags attached, such as the collection we call GNP? Gross national product is certainly a sum of something. What is that something?
From the work of the Four emerged yet another unifying, but ultimately obfuscating, idea. This was the vision of a network of prices, each reflecting the utilities and disutilities that are the psychological realities behind demand and supply—a network that could be pictured as perfectly balancing the claims of each individual with every other. In such a condition of “general equilibrium,” the flux of the market would finally come to rest. Every person would have attained a position that yielded as much net pleasure as that person’s initial “endowments”—his or her initial wealth or bargaining power—made possible. Moreover, it would be shown that the tendency of the market place, left to its own devices, was to move in this direction. However imperfect the world might be, the market mechanism was the means for attaining as much material well-being as that imperfect world offered.
Not surprisingly, this vision required special training before it could be discerned. For example, the awkward question of initial endowments—the fact that some actors on the marketplace wore tags as “labor” and others as “landlords” or “capitalists”—was ignored: that was a matter for historians, not economists. Difficult questions about envy and power, or vexing problems about the damage done by treading on others’ toes during the market scramble, were ignored. Perhaps most amazing, the sheer difficulties of achieving a general equilibrium, although recognized, were tossed aside. The French economist Léon Walras, to whom we owe the feat of formalizing the requirements of such an equilibrium, was forced to resort to the supposition that an imaginary auctioneer cried out various prices to which everyone responded with bids and offers, with the understanding that if the resulting supplies and demands did not interlock perfectly, all deals would be called off and the imaginary auction begin again!
But the main—and I think pernicious—problem raised by general equilibrium theory is not the extreme unrealism of its assumptions. All models are abstract and “unrealistic.” The critical problem is that general equilibrium rules out the very process that economics should be trying to explain and explore—the process of disequilibrium created by an economic system that is constantly revolutionizing its technical means of production and seeking to accumulate more wealth. It is all very well for abstraction to rid us of the distraction of reality, but it is not well when abstraction cuts out the matter that should be the focus of analysis.
Perhaps this is enough to indicate how flimsy are the foundations on which the structure of conventional economics is built today. But I will mention for good measure one additional problem: that economics does not have a clear definition for a word that it constantly uses—the word capital. Economists are always talking, for example, about the productivity of capital. But what do they mean by “capital”? One thing they do not mean is the physical embodiment of past labor and resources, such as a shovel or a computer. These are capital goods which are certainly invaluable for production. Let us consider a shovel for a moment. A shovel may multiply the productive capacity of a worker by ten. But is the shovel itself—the inert thing—“productive”? Can any thing be productive, no matter how much it enhances the working capacity of a person? Without the man will the shovel dig? Without engineers, mechanics, and operators will a computer “work”? Thus when we talk about shovels we are not talking about the productivity of “capital,” but about how capital goods, which are built out of labor and resources, enhance the productivity of human labor.
And what about the return on capital? Will the maker of shovels earn a return (called profit) from the productivity of his shovels? We have just seen that the shovels in themselves have no productivity. Or is the shovel-capitalist entitled to a return because his shovels so greatly boost the productivity of those who use them? In that case, will not competitive entrepreneurs crowd the market for shovels until they sell for nothing but their cost of production, plus whatever tiny margin might be required to induce entrepreneurs to continue to produce them?
So “capital” is not shovels; nor is the profit on capital the return on shovels, which is only a reflection of their abundance or scarcity, not of their productivity. Is capital then some kind of fund of money, out of which shovels (or computers) emerge? But is such a money fund “productive”? What does it produce? Indeed, what is this fund other than a claim on labor and resources? Capital thus remains a mysterious, near metaphysical concept for economists. Marx said that capital was the form that class power assumed in a world of monetized and commodified economic relations, but most economists do not read Marx.
There are additional objections to the basis on which conventional economics is built, but I hope that I have indicated some general reasons for the malaise that is spreading within its ranks.2 And the malaise is spreading. Scolding addresses by eminent economists are now a regular feature of the meetings of the American Economic Association.3 Meanwhile, two groups have already spun off from the main body of orthodox thought. One of them, the Association for Evolutionary Economics, taking its guidance from Thorstein Veblen and John R. Commons, has called for a return to a more historically minded approach that would put greater emphasis on social institutions. The other group, the Union of Radical Political Economists, has turned in the direction of Marxian economics. Neither body, however, has taken upon itself the task of systematically exposing the weaknesses of the theoretical underpinnings of conventional thought. This objective has become the aim of a third movement, not formally disassociated from the reigning orthodoxy, but clearly distinguishing its members as belonging to a new “Post-Keynesian” school of thought.
Post-Keynesian economics is an amalgam of the ideas of three economists. From Keynes it has taken its main vocabulary and its emphasis on instability and uncertainty as the inescapable starting points for analysis. From Michal Kalecki (pronounced Kaletsky), a Polish émigré to London whose work in the 1930s went almost unnoticed at the time, it has accepted a view of the economic system as reflecting two conjoined—but often uncomfortably conjoined—processes. One of these is the process by which current business investment determines the rate of expansion, and the rate of expansion in turn determines the level of profits. The second is the process by which firms seek to finance future investment by applying a suitable mark-up to their prices. From these two simple sequences Kalecki was able to construct a model that enabled him simultaneously to consider the macro question of growth of the entire economy and the micro question of how firms set their prices.
Third, post-Keynesian economics rests on the work of Piero Sraffa, a remarkable Italian scholar, long a Cambridge don. Sraffa’s output has been sparse, refined, and fertile. In 1925 he published an article that was perhaps the first to lay bare the flimsy theoretical basis behind the conventional depiction of the price system. This was then succeeded by a labor of decades in producing the definitive edition of Ricardo’s work. There followed in 1960 one of the shortest, clearest, and yet most difficult books in economics—Production of Commodities by Means of Commodities. This was an enormously abstract depiction of an economic system in unfamiliar terms—as a system not bound together by supply and demand but by technological linkages: so-and-so many bushels of corn and tons of iron needed to produce commodity X or commodity Y.
Sraffa’s rarefied model was important because it contained a radically different conception of the economic universe from that explicit in the ideas of the Four. Sraffa’s economic world continued the earlier visions of the classical economists. It was built on the actual process of production, not the timeless process of exchange. It was concerned with the nature and disposition of a surplus—a flow of production greater than that needed merely to restore the system to its starting point—rather than with the allocation of a “given” output in which there was no surplus. It dispensed with the need for the ectoplasm of capital and contented itself with various kinds of capital goods: its conception of profit depended on technology and the share of wages in the total product, not the marginal productivity of the ectoplasm.
The Keynes-Kalecki-Sraffa school soon found its foremost expositor in Joan Robinson, a star pupil of Keynes and widely acknowledged (save by the Nobel committee) as one of the most important and innovative economists of our day. Following her lead, the post-Keynesians soon came into head-on conflict with the prevailing neoclassical school under the redoubtable leadership of Paul Samuelson. The battle was fought with the most arcane terms—“reswitching,” “surrogate-production functions,” “meccano sets,” “putty”—but its purpose was deadly serious. It was to attack (on the part of Cambridge, England) and to defend (on the part of Cambridge, USA) the internal consistency and external relevance of the prevailing ideas of capital and of the price system itself. When the war between the Cambridges was finally resolved in favor of Cambridge on the Cam, the hitherto impregnable structure of neoclassical thought had received a serious, perhaps even fatal, blow.4
Over the last decade, post-Keynesian economics has gradually become more assertive, moving from a critique of received theory to an exposition of its own theoretical views and of the policies that flow from them. A Guide to Post Keynesian Economics, edited by Alfred Eichner, is a kind of status report on the present position of the movement. A collection of essays by a dozen contributors, it covers many fields of economics: price theory, labor markets money, resources, and the like. The prose is clear and nontechnical (for economists) and with some thoughtful reading the message is accessible to other social scientists or to informed laymen. For anyone who wishes to know what is going on in the world of economic thought, it is, I believe, indispensable. Since I cannot consider the separate essays of the book here (although I would like to call attention to a brilliant chapter on production theory by Richard X. Chase), I shall try to restate, in somewhat abrupt fashion, some main elements of the new post-Keynesian view.
- Unbalanced growth, not equilibrium, is the norm for any theoretical consideration of capitalism. Disequilibrium is the condition that must be considered to prevail at any moment, and also the condition that will prevail in the future. There is no such thing as a tendency to equilibrium.
- Prices are not mainly determined by supply and demand. In some markets, such as commodities, the conventional diagram of criss-crossed demand and supply curves may usefully depict the forces at work in setting prices. But in the all-important markets of the oligopolistic core—the markets in which the Fortune Five Hundred sell their wares—prices are virtually unaffected by shifts in the public’s demand, at least so far as the short run is concerned. Prices are determined by a mark-up over costs, much as Kalecki suggested, and guided by the hope of generating the profits needed to finance future investment. These investment plans may of course go astray, but they are nonetheless the key consideration by which firms establish their selling prices.
The distribution of income is integrally connected with the expansion of income. This is summed up in the rather cryptic proposition that “workers spend what they get and capitalists get what they [the capitalists] spend.” I cannot demonstrate the logic of this proposition here, but perhaps I can convey an intuitive sense of what it is about. There is a flow of spending that pays wages and another flow that becomes profits. The wage flow is all spent (“workers spend what they get”), and thereby returns to firms to meet the next round of payrolls. Meanwhile the flow into profits is (for purposes of argument) all saved and thereby becomes the source of investment. This investment spending returns to firms as a flow of revenues over and above what they have laid out as payrolls (“capitalists get what they spend”). This proposition has a number of unexpected corollaries, such as that taxes on profits in the short run, provided that they are spent by government, will be recouped by firms and will not actually reduce profits. I offer this probably incomprehensible idea merely to suggest that novel policy consequences follow from the post-Keynesian assumptions.5
There is no tendency in the economic system toward an orderly social outcome of economic processes, so that the natural inclination of post-Keynesian economics is toward intervention, not laissez-faire. The reason is not that post-Keynesians reject the idea that markets “work.” It is that they believe that markets often work perversely, aggravating rather than resolving imbalances between supplies and demands.
There is no labor market, in the sense that there is a market for wheat or for stocks and bonds. As a consequence, the volume of unemployment cannot be primarily blamed on labor pricing itself out of the market, although of course high wages will cause employers to seek labor-saving methods of production. The main determinant of the level of unemployment will be the volume of output, and the main factor determining the incidence of unemployment will be the segmentation of the work force into sheltered enclaves and exposed occupations. Moreover, these differentiations of the labor “market” are not considered mere imperfections, as a neoclassical economist would describe them. They are held to be inextricably connected with the structure of the business system, itself segmented into an oligopolistic core and a competitive periphery.
An increase in the supply of money, to the post-Keynesians, does not lead automatically to higher prices. This is because they do not put a conventional interpretation on the famous formula MV = PT (the quantity of money M times its velocity of use V equals the price level P times the volume of output or transactions T). Conventional economics assumes that there is an arrow of causation running MV → PT, which suggests that increasing M leads to higher P. But the post-Keynesians reverse the arrow, reading the formula MV ← PT. By this they do not wish to deny that an increase in money supply can affect the price level, especially at full employment. Rather, they call attention to the fact that increases in the money supply are usually not determined by forces “outside” the system, but by forces inside it. Specifically they assume that the supply of money will rise to whatever level is required to sustain higher outputs, because the first purpose of the banking system is to support the economy, not to control it.
The only policy that will control inflation is an “incomes policy”—a coordinated effort to hold down the main element of cost, namely wages and salaries. But wages and salaries cannot be restrained unless non-wage incomes are also controlled. Thus an effective incomes policy requires the limitation of increases in dividends, interest, rents, and personal profits, as well as in the disbursements to working people.
There are, of course, other elements of both theory and policy in post-Keynesian thought. Moreover, as the contributors to this volume make plain, the structure of this new direction in economic thought is far from complete. The relation of post-Keynesian theory to Marxian thought, for example, is largely unexamined. The questions raised by the classical economists about the usefulness of a labor theory of value have not yet been considered. The international implications of post-Keynesianism are hardly explored. And not least important, the views of post-Keynesians on policy often have a naïve ring. Incomes policy, to be successful, may require a depth and breadth of control that far exceeds the vague contractarian image that rises from its frequent but imprecise invocation.
But this is not yet the time for criticism. A new beginning for economics has appeared—a beginning that holds the possibility of rescuing the discipline from the intellectual sterility and ideological captivity that have made it the knave (or the fool) of the social sciences. It would be wrong to think that post-Keynesian economics will now equip us with the means to foresee government interventions, or to forestall the perversities of economic behavior that can upset the best laid plans of economic men. Nonetheless, we cannot begin to cope with such a world until we have set economic thinking on a far stronger base than that on which it is now erected. Perhaps we are witnessing the emergence of such a new chapter of economic thought, one that is very much needed if the discipline is to help and not hinder us as we move into a new phase of historical experience.
February 21, 1980
Directions in Economic Development, edited by Kenneth P. Jameson and Charles K. Wilbur (University of Notre Dame Press, 1979), p. 195. This small book of essays is an excellent survey of what we do and don’t know about economic development, especially the chapters by Celso Furtado and Jack Gurley. ↩
Needless to say, these critical remarks have been limited both by space and by the wish to avoid technical argument. A reader wanting to go into things a little more deeply might look into Joan Robinson’s Collected Economic Papers, Vol. V (Oxford, Basil Blackwell, 1979). Thereafter the going gets very technical indeed. ↩
Joan Robinson has recently summed up the gist of their complaints: “Mathematical operations are performed upon entities that cannot be defined; calculations are made in terms of units that cannot be measured; accounting identities are mistaken for functional relationships; correlations are mistaken for causal laws; differences are identified with changes; and one-way movements in time are treated like movements to and fro in space. The complexity of models is elaborated merely for display, far and away beyond the possibility of application to reality.” (From her introduction to Classical and Neoclassical Theories of Equilibrium by Vivian Walsh and Harvey Gram [Oxford, 1979], an elegant and important book.) ↩
The definitive, but very demanding, account is Geoffrey Harcourt’s Some Cambridge Controversies in the Theory of Capital (Cambridge University Press, 1972). ↩
See the chapter on Tax Incidence by A. Asimakopoulos in Eichner’s book, p. 65. The argument is put there more precisely than I have paraphrased it. ↩