Thursday, May 3, 2012

A CRITIQUE OF KEYNES CONCLUSION

All of the modern efforts to manipulate, manage, shape, control, and direct an essentially private capitalist economy rest on this Keynesian theory, with its subjective psychological foundations. In the tradition of British philosophy and political economy from which he derives, Keynes takes subjective propensities as given data – impenetrable, inexplicable, beyond argument or appeal. They are, to use the term of which economists are fond, exogenous variables, which is to say they come from outside the system. They are given in exactly the same way that the laws of nature and the resources of the earth are given.

It is worth pausing for a moment to reflect on how far economics has moved from its classical assumptions by the time we come to Keynes. Originally, all agents are assumed to maximize gain in an environment of perfect certainty and complete knowledge. Prices, wages, profits, rents, and the rate of economic growth are, under these conditions, functions of two factors: the objective technology of production, which determines what combinations of inputs are required for specified outputs; and the relative strength of the several classes of the society, which determines how the annual net social product will be divided up among the workers, the capitalists, and the landlords.
 
As the world becomes, and is recognized to have become, more complex, the simplifying knowledge assumptions of the classical model must give way to the acknowledgement of risk, uncertainty, and finally subjective expectation. At the same time, as it becomes clear that workers do not live permanently at the level of subsistence, that landlords are not mere idle consumers, and that capitalists, for a variety of reasons, are not perfect accumulators, the elegant theorems derived from the simple behavioral assumptions of classical political economy must be given up.
 
Faced with the intrusion of subjective non-rational elements into the process of economic choice and decision, post-Keynesian economists are forced to alter fundamentally the way in which they seek to understand a capitalist economy. Instead of a priori analysis built on elementary assumptions of profit-maximization, they offer econometric models in which dummy variables and functions stand for the several elements of the decision-making process. A variable for liquidity preference; a variable for the propensity to consume; a function separating a worker’s leisure/labour trade off, which is to say the proportion of total available labour-time that the worker prefers to devote to leisure, expressed as a function of income. And so on and on. Any of you who have taken even an elementary course in macroeconomics will be aware of the extent to which the subject, as now taught, rests on this sort of model-building.

The result is that economics has become an extremely elegant, complex, mathematically sophisticated way of guessing at the shadows on the wall of the cave. In Plato’s REPUBLIC, you will recall, Socrates relates an allegory of the human condition. We are to imagine, he says, that a group of men are chained to the floor of a dark cave, so that they can only look to their front at a blank wall. Behind them, fires are lit, and unseen attendants walk before the fires, carrying small scale models of physical objects and people. The light casts shadows of these objects on the wall, where they flicker in fantastic distortion. At first, the captives are simply mystified by the succession of shadows, but after a while, some of them, those best adapted to a troglodytic existence, begin to discern repetitions and patters in the images. They formulate theories about what shadows will appear next, and the most skillful among them acquire considerable reputations for their ability to anticipate by a few moments the next images. Some, we may even imagine, extending the story a bit beyond Plato, become tenured professors of shadow-guessing, and a few whose theories of the shadow world have risen to heights of mathematical elegance even win Nobel prizes for shadow-guessing.
 
One of the captives, Socrates tells us, driven by some obscure instinct that the world holds more than shadows, works himself free of his bonds and crawls painfully to the mouth of the cave. Dazzled by the bright sunlight, he slowly acclimates himself to the brilliant light, and sees for the first time the real physical objects whose twisted and distorted shadows he has all these years observed. At last he realizes that these are the reality of which the shadows are more imperfect reflections or appearances. Rushing back into the cave to bring this momentous news to his fellows, he is temporarily blinded by the darkness, and staggers about as though mad. Naturally, the remaining captives simply laugh at his insistence that their shadows are inferior appearances, behind which lies a truer reality. Puffed up by their skill at shadow-guessing, they consider merely comic the claims by their former comrade that they are enmired in unreality.

Marx had a name for the masters of shadow-guessing. He called them Vulgar economists and contrasted them with the classical economists -- Petty, Quesney, Smith, Ricardo -- whom he considered serious students of economic reality. Paul Samuelson, the greatest of the shadow-guessers, has returned the compliment by characterizing Marx, in a famous essay, as a “minor post-Ricardian,” and, worst of all, “an auto-didact.” (To be self-taught, I suppose, is from Samuelson’s standpoint even worse than to be a minor follower of the wrong economist, for if it should turn out that one can teach oneself to understand economics, that will put an end to the hegemony of the profession.)
 
The dummy variables and functions of econometric model-building refer to nothing at all that can be directly studied. There is no way that we can get at an individual’s “propensity to consume,” for the purpose of constructing a more adequate theory of consumer behavior. Nor can anything useful be said about the inner determination of the capitalists’ expectations for future gain, so as to lay the foundations for a scientific theory of investment and growth. Instead, economists are forced to amass countless time-series of date, on which, in a manner that would have made David Hume proud, they can perform simplistic extrapolations.
 
There are two central problems with this mode of theoretical operation, and together they have brought economics to its present sad condition. The first problem is that there is no stable set of psychological propensities or motives about which reliable knowledge can be accumulated. As I pointed out when discussing Mill’s introduction of the factor of habit or custom, these labels or placeholders -- habit, custom, propensity to consume, liquidity preference, leisure/labour trade-off, and the rest -- are merely summary names given to whole congeries of heterogeneous and shifting motivations. Some consumers may be guided in their decisions about savings versus consumption by a consideration of present versus future pleasures; others may be influenced by the uncertainty of unemployment; still others may be reacting to the experience of seeing savings shrink in value under high rates of inflation. And some consumers may even have been influenced by the sorts of public service advertising that first surfaced during the Eisenhower years, when Americans were exhorted to buy on credit as a way of showing their faith in the American system

Economists extrapolate from past behavior, only to find that the present deviates from the past. As they make mid-course corrections in their econometric estimates, reality continues to shift beneath them. The problem is not that modern economic reality is complex. Their formal models are more than adequate to handle a high level of complexity. The problem is that their theories are theories of appearances, surface manifestations, and hence give no genuine insight into the causes of the shifting shadows.

The second problem, more serious even than the first, is that economics is a study of human choice and decision, not of inanimate nature or animal behavior. The consumers, investors, and entrepreneurs whose preferences and propensities are modeled by the econometricians are themselves self-conscious agents increasingly aware of and influenced by the descriptions, predictions, hopes, and anxieties of economists, public figures, and social commentators.

All of you are familiar from today’s newspapers with the ways in which the interplay of economic prediction and private investment or consumption decision wreaks havoc with the efforts of the central government to manage the national economy. Businessmen postpone the expansion of productive capacity in the expectation that high interest rates will slow recovery, balloon the federal deficit, send the government into the money market for even greater borrowing, and thereby maintain the high rates. The government enacts a tax designed to draw a larger share of income into savings, but the stock market discounts the effects of the act six months before it goes into operation, condemning it to failure.
 
Under these circumstances, madcap schemes acquire respectability, and establishment politicians and economists react by labeling them “voodoo economics.” But in fact, this is merely the effort by licensed witch doctors to drive their upstart competition back into the bush.

What is the heart of the problem? I should like to suggest to you that Marx’s style of diagnosis still retains considerable merit. The underlying problem, as he would have argued, is that the forms of capitalist development have become fetters. In the early stages of capitalism, the pressures of a permanent excess supply of workers, together with the almost total lack of collective or legal protections, held wages at or near subsistence. Severe competition among large numbers of small businesses forced firms to strive for maximum growth if they were to survive at all. The effect was rapid, although uneven, expansion of the productive capacity of the economy.

With the rise in worker standard of living and the unrelenting amalgamation of small capitals into large, with the advance of collective bargaining and the development of a system of money and banking sophisticated enough to support the new capitalist order, growth and prosperity increasingly came to depend on the appropriate rate of savings and investment and the proper management of the expansion of effective demand for the consumer goods and capital goods being produced.
 
The new economics of Keynes and his followers sought to preserve an essentially private economic order, in which not only mere legal ownership, but more importantly the effective control and management of the means of production, remained lodged in countless uncoordinated private firms. What this meant -- and means still today -- is that the ultimate determination of the allocation of the collective social product rests with the irrational and impenetrable subjective preferences, propensities, and expectations of private individuals. It is scarcely surprising that even the most agile shadow-watchers are unable to either predict or to control the flow of images on the wall of our cave. What is to be done? There are three possibilities, each corresponding to a different conception of the relationship between the economic theory and the practical management of the economy. The first possibility is to press forward as we have been doing, with more elaborate and refined macroeconomic models, more complex fiscal and monetary programs, all imposed on a private economy organized on the pursuit of profit. This course we may call the persistence of the actual, and as metaphysicians have often observed, the actual occupies a privileged place in our experience and beliefs. I am absolutely convinced that this course of action is doomed to failure, but I rather imagine it will take another decade or two of the economic disaster before the American people are prepared to scuttle the conventional wisdom and adopt some better means of social decision.
 
The second possibility is to turn the clock back two centuries and try, by an act of faith and will, to re-instantiate a world in which the classical economic theories work. This I shall call the yearning for the impossible, and it is now, with the advent of the theory of rational expectations and so-called supply-side economics, reaching the height of its brief resurgence. The impossible has always exerted a powerful attraction on credulous souls, sometimes even eclipsing the actual by its beauty and simplicity. In earlier days, when society was less thoroughly interconnected, it was possible to embrace the impossible for quite some time before being brought up short against reality. Now, unfortunately for Mr. Reagan, the impossible succumbs to the actual in about as long as it takes to get from a presidential to a mid-term election. I think we can with confidence conclude that rational expectations, supply-side economics, and the revival of the theory of the free market have already peaked and are on the decline.
 
[Alas, when I wrote this, I was in the grip of an optimism that Keynes would have considered a very great effusion of animal spirits.  A quarter of a century later, things have only gotten worse, which I suppose some people would conclude means that are closer to revolution!]

We are left with the third alternative, which is to confront directly the underlying cause of the failure of modern economics, and respond by changing both our theory and our practice.

As we saw, the problem is this: an economy guided by the uncoordinated decisions of private individuals must necessarily rest upon subjective, variable, and self-referentially influenceable motivations. So long as the theories of the behavior of producers and consumers become factors influencing that behavior, no stable theory can be developed by means of which the economy can be guided.

There is, however, a solution. What are now predictions by external observers of the way in which economic agents will probably behave can be transformed into collective decisions about how economic agents choose to act. Instead of attempting to shape and guide an economy on the basis of predictions about what proportion of income consumers will save, or what level of profit will draw new capital into investment, we can as a society choose a level of savings that fits our collective goals and desires. We can decided to invest at a level and in a pattern calculated to achieve whatever regeneration or expansion or transformation of our industrial plant it is that serves our collective social ends. We need not rest our hopes for full employment on the armchair psychological maxim that consumers will save a larger fraction of each additional dollar of income, nor need we count for new capital investment on the fragile, variable, semi-informed subjective expectations of individual capitalists.
 
The economic theory suited to the rational management of a modern industrial economy is not the elaborate shadow guessing of modern econometrics. Rather, it is the physical-quantities linear analysis developed by Wassily Leontief as input-output analysis, together with the theory of linear programming, and the modern mathematical reinterpretation of Marx carried out by Piero Sraffa and a host of economists around the world. This theory has its roots in the Tableau économique of Quesney and the physiocrats, and builds on the analysis of Ricardo and Marx. Theories of this sort are already being employed as tools of rational economic planning in Eastern Europe, and would have a far more powerful and effective application to an economy as advanced as that of the United States.
 
The foundation of economic planning is a set of objective data specifying the available technologies, on the basis of which one can calculate the direct and indirect physical requirements for whatever final output is collectively chosen by the society as the goal of its economic activities. The level of output and rate of growth of an economy have natural limits, imposed by the state of technology, the size and skills of the labour force, and the stock of available tools, raw materials, and machinery. Neither ideology nor economic theory can change these constraints, although over time, rational social planning can alter them in major ways. With the modern methods of analysis to which I have referred, it is now becoming possible to undertake complex society-wide planning based on objective facts and collective choices, rather than on shadow appearances and subjective preferences and expectations. The anarchy of the marketplace can finally give way to rational social management founded on a politics of public deliberation and determination of collective economic goals.
 
There is, of course, an obstacle to the triumph of the rational, as we may call this third course of action. The obstacle is not primarily theoretical, nor is it technological. Rather, it is political. Those who are systematically benefitted by the present economic order will fight to maintain their advantage. Hence the struggle for socialism, which is of course what I am talking about, must be carried on in every political arena, as well as in the literature of economic theory.
 
Nevertheless, there is a great deal to be gained by confronting established dogma and exposing its true nature as the scholasticism of shadows. The economics of the neo-classical synthesis, as it has been called, is in retreat. The master shadow-guessers of Samuelson’s generation are giving way to epigoni who struggle helplessly to reassert their authority in the face of economic disaster. Theories of collectivel rational economic choice will not substitute for political action, but they have an indispensable role to place in the social transformation that must be undertaken. Perhaps my observations this afternoon will encourage some of you to revisit the dismal science -- or to make its acquaintance for the first time -- and to explore the new theories which are arising to challenge the old.


6 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. A necessary, if not sufficient, condition of a "triumph of the rational" is the universal breaking of the psychological fetters that, according to Plato, are ingredient in the shadow-play. The success of the Enlightenment ambitions in that direction remains uncertain.

    ReplyDelete
  3. Given the well-composed analysis present in the first two parts of this essay, part 3 was disappointing, to say the least.

    First, the two "central problems" with neo-classical economics that the essay identifies are very superficial problems. At root, the second "problem" is just a hasty indictment of social science in general, and thus is completely over-blown. The first "problem," which has somewhat more merit, is that psychology is a dubious foundation for social science. The author doubts that any interesting psychological generalizations are possible that could furnish the grounds required for economic generalizations. In this regard, the author's conclusion is underdeveloped, seemingly premature, and licenses an unwarranted degree of skepticism, since it seems to demand precision of the social sciences that is not even required of the natural sciences.

    Along these latter lines, I found myself pondering the analogy between Brownian motion and the author's comments about variables representing the individuals' motivations as "summary names given to whole congeries of heterogeneous and shifting motivations." Just as the fact that any single molecule in a fog may follow an extremely complicated course does not upset generalizations about the overall descent of the fog, it seems that any single economic actor may follow a course of action dictated by multiple motivations without upsetting generalizations about the economic behavior of the society as a whole. In the end, the author's skepticism about the neo-classical framework is so general that as a consequence it licenses skepticism about all of social science and a considerable amount of natural science -- which is just a reductio ad absurdum.

    Second, the proposed alternative framework to neo-classical economics is mysterious and defended with a lot of hand-waving. How *exactly* does this framework differ in its *analysis* from neo-classical economics (and not just in is conclusions)? It's really hard to get a sense of that from the essay. Moreover, severe doubts start to set in when references are made to "rational central planning" in Eastern Europe, without any recognition of the incredible black market in drugs that flourished there, not to mention the famous Eastern exploding tv sets and desultory food stores with names like "Milk" and "Bread." This kind of selective attention to detail made me feel like I was reading the Marxist version of Robin Hansen.

    What really seems to be missing from this essay is any recognition that market transactions are one -- but only one -- form of genuine cooperation between people. Therefore, there is nothing evil or wrong with market transactions by themselves, and there is no reason to stamp them out completely and replace them with central planning. Problems only arise when market transactions are thought to be the *only* form of cooperation, such that every other form of social cooperation is modeled on and understood in terms of market transactions. This latter development is the real unfortunate legacy of the neo-classical economic framework. Keynes is such an important figure in resisting this framework because his work showed that other, non-market forms of cooperation are required to make market cooperation possible and efficient.

    ReplyDelete
  4. I wonder if Prof. Wolff has seen Leni Riefenstahl's film "Triumph of the Will"? Is "Triumph of the Rational" an allusion, or an accident?

    ReplyDelete
  5. I know about the film, of course, but in fact have never seen it. The apparent allusion was an accident.

    ReplyDelete
  6. Unfortunately, I only stumbled upon this comment thread today (18 July 2014, 1:25 PM AEST), so I am probably going to get little in the way of feedback from the relevant parties.

    At any rate, I'll address Chris Langston's comment, specifically what regards what Prof. Wolff calls the "second problem, more serious even than the first", of which Langston says "At root, the second 'problem' is just a hasty indictment of social science in general, and thus is completely over-blown".

    I, of course, can't speak for Prof. Wolff and if I am mistaken, he should correct me immediately (hopefully, he keeps track of comment), but what he refers here as the second problem was identified by Robert E. Lucas, Jr. (who won the 1995 Nobel Prize for this finding, among others) and is widely known as the Lucas critique.

    Judge by the conclusion to Lucas' 1976 article "Econometric Policy Evaluation: a critique" (it's easily found with Google):

    "This essay has been devoted to an exposition and elaboration of a single syllogism: given that the structure of an econometric model consists of optimal decision rules of economic agents, and that optimal decision rules vary systematically with changes in the structure of series relevant to the decision maker, it follows that any change in policy will systematically alter the structure of econometric models.
    "For the question of the short-term forecasting, or tracking ability of econometric models, we have seen that this conclusion is of only occasional significance. For issues involving policy evaluation, in contrast, it is fundamental; for it implies that comparisons of the effects of alternative policy rules using current macroeconometric models are invalid regardless of the performance of these models over the sample period or in ex ante short-term forecasting".

    This leads me to my second comment: I am rather perplexed by Langston's claim that the "second problem" is a "hasty indictment of social science in general", when neither Prof. Wolff spoke of social science in general, nor social scientists in general design economic policy.

    As Langston did not argue this point, and his statement is far from obvious, perhaps he should clarify. I, like him, am not fond of hand-wavings.

    Furthermore, for good or for ill, the Lucas critique was used precisely as an indictment against economic policy. If Prof. Wolff's "second problem" was over-blown, as Langston claims, it wasn't him who over-blew it. According to the Nobel Committee's press release announcing the 1995 Nobel prize:

    "The Lucas critique has had a profound influence on economic-policy recommendations. Shifts in economic policy often produce a completely different outcome if the agents adapt their expectations to the new policy stance."
    http://www.nobelprize.org/nobel_prizes/economic-sciences/laureates/1995/press.html

    Finally, I found rather surprising (hastily chosen?) Langston's comparison of "molecules in a fog" with economic agents: the point of Lucas critique and (if I understood him) Prof. Wolff's "second problem" is that we cannot simply compare molecules and people: people, unlike molecules, are free in their way to react to policies.

    Obviously, Langston may have good reasons to express himself so categorically. So, if he is still reading, I'd be interested in judging his economic argument.

    ReplyDelete