Источник: Eastern Economic Journal, Spring97, Vol. 23 Issue 2, p231, 6p
SAY'S LAW OF MARKETS: WHAT DID IT MEAN AND WHY SHOULD WE CARE?
It was Keynes who planted the idea that all economists before him subscribed to Say's Law of Markets--that was indeed his definition of a "classical economist"--and by Say's Law of Markets he meant the doctrine that the economic system is always, or nearly always, operating at its full-employment ceiling; since "supply creates its own demand", recessions and depressions are necessarily short-lived and self-correcting. He agreed that Say's Law as such had long been abandoned by most economists but he insisted that they had never ceased to believe in the working assumption of full employment and full-capacity production.
I think that Keynes was quite right thus to characterize Keynesian economics, but to label this as a belief in something called "Say's Law of Markets"-- a term that had long disappeared from the economic literature--and then to hang it on a catch-all definition of "classical economics" was to invite hopeless confusion about the history of economic thought. John Hicks added mayhem to mischief with the title of his famous IS-LM paper, "Mr Keynes and the Classics: A Suggested Interpretation." And who were Hicks' classics? Adam Smith, Ricardo, and John Stuart Mill, and/or Marshall and Pigou. Hicks never explained his use of the term because by now (1937) "classical economics" had become almost a term of abuse.
There was a time in the long distant past when classical economics meant English Classical Political Economy from Smith to Mill, say, 1776 to 1867. Did those classical economists believe in Say's Law? Yes, they did and what they meant by it was the proposition there was a world of difference between what Ricardo as early as 1817 called "sudden changes in the channels of trade" and what Mill in 1848 called periodically recurring "commercial crises." Commercial crises were "speculative states" that were always followed by "quiescent states". But general overproduction -- an inability to sell all goods at cost-covering prices, including the going rate of return on capital--might well be irremediable. In short, they recognized the existence of cycles of booms and slumps--and increasingly so in the 1840s after the slumps of 1825,1836 and 1847--but they distinguished these from a state of secular stagnation, what we would now call a"low-level equilibrium trap." This was no idle academic issue because the Industrial Revolution had raised fears in the minds of many contemporaries about the ability of any economic system to absorb the ever-increasing output of an industrial economy.
Unfortunately, in the anxiety to win the intellectual argument, the classical economists fell into the trap of propounding Say's Law as the logical impossibility of general overproduction. But there is no logical impossibility of general gluts: all that is needed is an upward shift in the liquidity preference function--an increase in the demand-for-money to hold--because of a collapse of confidence or for any reason whatsoever and then there will be a glut of goods in general, a surfeit of at least some goods that cannot be sold at cost-covering prices. Whether such a state will persist or whether it will involve an adjustment process that will soon restore equilibrium is another question and here all the classical economists with the exception of Sismondi on the left and Malthus and Chalmers on the right agreed that gluts are temporary and that slumps and crises always give way in time to booms and financial quiescence. This was what the classical economists really meant by Say's Law of Markets, namely, that a free enterprise capitalist economy has an inherent tendency to return to full employment, which is indeed its normal state of economic activity.
Let us now translate all that has been said in modern jargon. Oskar Lange, in an influential article on Say's Law in 1942, laid down the tram-lines within which the discussion has run ever since. Without so much as mentioning Keynes, except in two footnotes, Lange expressed the meaning of Say's Law as Walras would have interpreted it were he alive in 1942. Say's Law of Markets could either mean Say's Identity or Say's Equality. It followed from what he called Walras's Law that the total value of all goods and services demanded in an economy must equal the total value of all goods and services supplied if and only if the demand for money is equal to the supply of money. Thus to assert in the manner of many classical economists that general overproduction in a monetary economy is logically impossible is tantamount to the assertion that the money market is always in equilibrium regardless of either relative or absolute prices and this is what Lange labelled "Say's Identity." However, there is also what he labelled "Say's Equality", namely, the proposition that an excess demand for money, or what is the same thing, an excess supply of goods, is self-correcting through automatic price and interest variations, in particular what Patinkin has taught us all to call "the real-balance effect."
The next chapter in the story was the 1952 paper by Becker and Baumol, "The Classical Monetary Theory: The Outcome of the Discussion." This drew the distinction between Say's Identity and Say's Equality even more explicitly than Lange and demonstrated that John Stuart Mill was the only classical economist who stated this distinction so clearly as to leave no doubt that he recognized the difference between an identity and an equilibrium condition. To quote from Mill's Unsettled Questions in Political Economy, "In order to render the argument for the impossibility of an excess of all commodities applicable to the case in which a circulating medium is employed, money itself must be a commodity. It must undoubtedly be admitted that there cannot be an excess of all other commodities and an excess of money at the same time" . But "the utility of money" consists precisely of the possibility of being able to postpone the act of purchase and so, Mill continued, "it may very well occur, that there may be, at some given time, a very general inclination to sell with as little delay as possible, accompanied with a general inclination to defer all purchases as long as possible. This is always actually the case in those periods which are described as periods of general excess. And no one, after sufficient explanation, will contest the possibility of general excess, in this sense of the word" [Blaug, 1997]. If Keynes (or Lange) had quoted this statement instead of similar but more ambiguous statements in Mill's Principles, the long confusing debate about what the classical economists meant by Say's Law could have been entirely avoided. It is true that Mill does not state the real-balance effect in so many words: for him a fall in absolute prices lowers the public's demand for cash, not because of its effect in raising the real value of cash balances, but because of the expectation that the fall in prices will not continue indefinitely. Nevertheless, an automatic equilibrating mechanism is contemplated and to that extent, Mill is subscribing to Say's Equality, not Say's Identity. To be sure, nothing as lucid as this was ever said by any other classical economist: Ricardo defended Say's Law as valid irrespective of price-interest variations in the same chapter of his Principles in which he explained how variations in the rate of interest govern the demand for investment funds, and Say criticized Ricardo's dogmatic defense of the law of markets, emphasizing the function of the interest mechanism for equilibrating saving and investment as if this was an idea that had never occurred to Ricardo. Similarly, James Mill and McCulloch discussed the issue of the impossibility of general gluts at various places in their writings, sometimes expressing Say's Identity but frequently hinting at Say's Equality without, however, showing the younger Mill's awareness of the logical structure of the problem.
In brief, I think it is fair to say that when a classical economist asserted the so-called "impossibility of general overproduction", he had in mind not periodic crises but secular stagnation. Could the capitalist system absorb the constant increases in productive capacity without breakdown from limits inherent in the system? Say's Equality supplied an affirmative answer to this question: with flexible prices, the system is forever tending to full employment, full-capacity equilibrium. The classical economists never established this proposition with any rigor but they did appeal at least obliquely to what is a perfectly valid comparative static argument, namely, the real-balance effect.
We have at least two powerful voices in the secondary literature that would reject this conclusion. Firstly, Patinkin in an appendix to his great book, Money, Interest and Prices , objected to the very term Say's Equality as misleading because no classical economist, not even John Stuart Mill, ever clearly specified the dynamic equilibrating mechanism that would validate Say's Equality. This is perfectly true but the same purist insistence on analytical rigor would lead one to deny that there is such a thing as a classical theory of long-run prices, since no classical economist ever really demonstrated the existence, determinacy and stability of equilibrium in a single commodity market. Secondly, Thomas Sowell's full-length study of Say's Law. An Historical Analysis  denied that Say's Law in the classical period had anything to do with secular stagnation, insisting instead that it was a debate about short-run problems. But the frequent recognition of business cycles in the literature of the 1820s, 1830s and increasingly so in the 1840s, not to mention Mill's focus throughout his discussion of the law of markets on the Sismondi-Malthus predictions of permanent overproduction, throws cold water on Sowell's interpretation. Finally, when Baumol came to re-examine "Say's (at least) Eight Laws" in 1977, he concluded that what Say's Law had ultimately meant for Say himself was Say's Equality, that is, that "supply and demand are always equated by a rapid and powerful equilibrating mechanism." I rest my case!
And that brings me, at long last, to Steven Kates' paper on "The True Meaning of Say's Law." Kates' earlier paper, "Crucial Influences on Keynes' Understanding of Say's Law" , finally pin-pointed the precise source of Keynes' reference to the then obscure doctrine of Say's Law in The General Theory including the phrase "supply creates its own demand" to denote Say's Law, words which were Keynes' own invention. However, in the paper under examination Kates goes well beyond questions of doctrinal exegesis to argue that the classical economists were quite right to deny that recessions or depressions are ever caused by excessive production in the sense of insufficient aggregate demand relative to supply; in short, recessions always mean misdirected production but never general overproduction. The doctrine of the possibility of general gluts, Kates argues, is not just fallacious as a proposition in comparative statics but false even as a proposition in aggregate dynamics; or, as he himself puts it, "Say's Law was the classical proposition which ruled demand failure out as a theory of recession" and Say's Law is absolutely correct. Kates, like William Hutt in A Rehabilitation of Say's Law , would have us turn Keynes on his head to argue that what is wrong is not Say's Law in any of its versions but Keynes' refutation of Say's Law and Keynes' belief that an insufficiency of effective demand can ever be the cause of unemployment. It follows from the Keynesian doctrine that a modern economy or a recession can spend its way to full employment, irrespective of whether that extra spending is public or private. On the other hand, if Keynesian doctrine is erroneous, it will require a change in the pattern of demand and, the structure of production, and hence a change in relative rather than absolute prices, to produce a recovery. I take it that is the view that Kates endorses.
In the eagerness to vindicate the classical economists and to shield them once and for all from Keynes' accusations, Kates has totally lost sight of the difference between the type of economic system that provided the historical context for the considerations of the classical economists and that which provided the backdrop to The General Theory. I once distinguished between Keynesian and Marxian unemployment in an attempt to describe that differences Keynesian unemployment denotes a situation in which the flow of private plus public investment is insufficient to mop up the savings that would be forthcoming at full employment levels of income; because of relative overabundance of physical capital, rates of return are too low to call forth the volume of investment required to produce full employment. Marxian unemployment, on the other hand, is the result of capital scarcity relative to the labor supply; inappropriate resource endowments and the limited technical possibilities of substituting labor for capital makes it impossible to absorb all the labor that is available even when the capital stock is used to capacity. Marxian unemployment is the result of both excessive population growth and income levels too low to produce an adequate flow of savings, combined with a primitive, rigid technology. Too little thrift, not insufficient aggregate demand, impedes the expansion of output. Marxian unemployment is a structural, not a cyclical, problem and for that reason an easy monetary policy and an expansionary fiscal policy, effective in curing Keynesian unemployment, is much more likely to produce inflation without leading to full employment.
The symptom of unemployment is the same in either case, but the successful cure is different since the nature of the illness is quite different. It follows that the actual or disguised unemployment of the now overpopulated underdeveloped countries of Asia and Africa must be cured by methods different from those designed to attack underemployment in a mature capital economy.
To ignore this distinction is to imagine that when Ricardo and John Stuart Mill analyzed the causes and cures of general overproduction in slumps they had in mind the same sort of economy that Keynes had in mind when he attacked Say's Identity travelling in the guise of Say's Law. But Keynes was culpable as a historian of economic thought because of a pronounced tendency to engage in historical generalizations as, for example, --"that the weakness of the inducement to invest has been at all times the key to the economic problem." It was this that caused him to ascribe his own peculiar definition of Say's Law to every economist before him and which has ever since given Say's Law an importance out of all proportion to its actual role in classical and certainly in neoclassical economics. Kates seems to me to have fallen victim to the same tendency by arguing that if Ricardo and Mill were right in the first half of the 19th century, they would also have been right in the second half of the 20th century.
I contend instead that the classical economists grasped the fact that they were confronted with Marxian, not Keynesian unemployment, and that their misgivings about Sismondi and Malthus calling for more "unproductive consumption" by landlords were well-taken. But we in the western world are faced with Keynesian unemployment, which is indeed due to insufficient effective demand and which is indeed curable by standard demand management, albeit assisted by supply-side policies.
Baumol, W. J. Say's (at Least) Eight Laws, or What Say and James Mill May Really Have Meant. Economica, May 1977, 145-61.
Becker, G. and Baumol, W. J. The Classical Economic Theory: The Outcome of the Discussion. Economica, 1952, 355-76.
Blaug, M. Economic Theory Retrospect, 5th ed. Cambridge University Press, 1997.
------., ed. Pioneers in Economics, No 15, Jean Baptiste Say, 1776-1832, Edward Elgar, 1991.
Hutt, W. A Rehabilitation of Say's Law. 1947.
Kates, S. Crucial Influences on Keynes's Understanding of Say's Law. History of Economics Review, Winter 1995.
Lange, O. Say's Law: A Restatement and Criticism, in Papers in Economics and Sociology, by Lange, Oskar. Oxford: Pergamon Press, (1942) 1970.
Mill, J. Unsettled Questions in Political Economy, 1844.
Patinkin, Money, Interest, and Prices, 1956.
Sowell, T. Say's Law: An Historical Analysis. Princeton: Princeton University Press, 1972.