Say's Law and the Keynesian Revolution: How Macroeconomics Lost Its Way.

By Steven Kates. Cheltenham, UK: Edward Elgar Publishing, 1998. Pp. ix, 252. $80.00.

The meaning and the significance of Say's Law have spawned numerous controversies, and the issues at stake have shaped theoretical models and policy advice for generations of economists. Unfortunately, as explained by both Skinner (1967) and Clower and Leijonhufvud (1973), some of the bickering over Say's Law has been based on little more than obtuseness and misinterpretations. Most importantly, the misrepresentations of the past still haunt our conceptual framework and our textbooks. For one thing, incessant repetitions notwithstanding, the supply-creates-its-own-demand version of Say's Law had nothing to do with Say's actual arguments.

The title of Kates's book suggests that the book will examine the legacy of the Say's Law controversies. In fact, the book is mostly a series of short and somewhat disparate vignettes that summarize the arguments of various writers on Say's Law. The book is divided into twelve chapters, the first of which presents Kates's view of the role of Say's Law in Keynes's General Theory. Chapters 2-4 then condense the arguments of key classical economists (Say, Malthus, Torrens, Ricardo, and the Mills). Chapters 5 and 6 sum up various figures ranging from McCullogh through various neoclassical economists (here labeled "English Classical") to some of Keynes's own contemporaries. Chapters 7 and 8 then present Kates's view of how Keynes came to his conception of Say's Law. Chapters 9-11 summarize various writers after Keynes, and finally, Chapter 12 presents a short conclusion.

While the book is made up of summaries of many writers, for some reason it does not even mention Say's most boisterous critic, Karl Marx. This is curious in view of the fact that Kates actually reproduces Marx's (1929, 1952) famous commodity-money-commodity (or CM-C) transformation argument (pp. 23, 75). In any case, while Marx was certainly influential in deriding Say and while Marx's critique of Say does anticipate Keynes's naive supply-creates-its-own-demand version of Say's Law, the term "Say's Law" seems to have been first coined by Taylor (1909) when Say's Traite was about as old as Marshall's Principles are today.

As defined by Taylor, Say's Law was the proposition that "if we can assume that producers have directed production in true accord with one another's wants, total demand must in the long run coincide with the total product or output of goods produced for the market" (Taylor 1909, p. 94). This was the original definition of Say's Law and, while it postdated classical economics, it did capture one of Say's key insights, namely, that as long as sellers only offer those goods that buyers intend to buy, then it stands to reason that all goods will eventually be sold and that all markets will clear. Albeit, there was more to Say than this simple inference.

As explained by Clower and Leijonhufvud (1973), Say's basic observation was that all successful trade depends on quid pro quo offerings. That is, people cannot "buy any articles whatever to a greater amount than that which they have produced either by themselves, or by means of their capitals and lands" (Say 1821, p. 3). In short, Say started out with the familiar budget constraint that we now take for granted, and most of his varied arguments on markets are simply aggregate-consistency arguments that are inferable from this underlying constraint. Rather than denying the existence of gluts and unemployment, he provided the fundamental insight necessary for understanding all recessions, that the effective demand for goods is limited by wealth and income, which in turn are limited by the successful sale of goods.

In this context, Kates's arguments on Say's Law and the Keynesian Revolution are somewhat peculiar. He starts his book out by denouncing Keynes's misrepresentations of Say and the classical economists, which he claims ultimately caused macroeconomics to lose its way (hence the subtitle of the book). But, surprisingly, Kates then provides his own definition of Say's Law as "the proposition that failure of effective demand does not cause recession" (p. 1). Of course, when Kates uses the term "Say's Law," it means just what he chooses it to mean - neither more nor less. Quibbling about Kates's definitions is less fruitful than making sure that one understands what he means. The real problem is that Kates attributes his own peculiar definition of Say's Law to the classical economists and that he argues that Keynes (mistaken as he otherwise may have been) was really attacking this (i.e., Kates's) version of Say's Law.

Kates's difficulty, it seems, is that he does not understand how the classical economists used the terms "demand" and "effective demand." Thus, he conflates the classical term "effective demand" and the more modem Keynesian notion of the aggregate demand function. To see how pre-Keynesian writers used this term, let us quote Keynes's father from the original Palgrave (this quote was suggested to me by Bob Clower): "By 'demand' in political economy is meant what may be more distinctively called effective demand, that is, not the mere desire for anything, but desire accompanied by the offer of something valuable in exchange" (Keynes 1894, p. 539). In other words, in pre-Keynesian economics, it is the existence of quid pro quo offers that constitutes the effective demand for a good. Hence, if such offers are not forthcoming, this represents, by definition, a failure of effective demand. Moreover, following Say, the classical economists believed "that demand and supply considered as aggregates are strictly interdependent, and that neither can increase or diminish without necessitating a corresponding increase or diminution of the other" (Keynes 1894, p. 540). In other words, the classical economists believed that effective demand is constrained by successful sales.

To recap this important point, Kates's definition of Say's Law is based on a misunderstanding of how economists before Keynes defined the term "effective demand." Of course, this is not a problem unique to Kates. Ahiakpor (1997, 1998) has demonstrated how Keynes's unorthodox use of terms, and the subsequent adaptation of Keynes's terms by the economics profession, has made it hard for most modem economists to grasp the language of classical arguments.

In any case, consider how Say presented the standard classical effective demand argument: Production does not always and automatically create revenues that can be used to buy other goods since "the existence of these revenues depends on the production having exchangeable value, which it can only have in consequence of the want which there is for such production in the actual state of society" (Say 1821, p. 13). If the producer "evinces inexperience in his affairs, he may gain nothing; he may very probably be a loser" (Say 1821, p. 14). Hence, gluts may arise from "the ignorance of producers or traders on the nature and extent of the demand in the places to which goods have been consigned" (Say 1821, p. 59). The underlying cause of these gluts is the lack of "Data . . . to serve as the foundation of good calculations" (Say 1821, p. 59). In short, gluts arise when buyers and sellers do not succeed in coordinating their trading plans and when this in turn affects their effective demand for other goods.

Kates does mention classical arguments to the effect that unsuccessful sales of some goods may mean that "demand for other products would be lower than was originally anticipated" (p. 4), and he also quotes Ricardo to the effect that demand depends on the means to purchase goods and that this in turn means that everything that affects production will thereby 'affect demand (p. 44). Even so, he does not see the effective demand argument contained in the words he cites. Kates then relies on second-hand authorities who tell him that the classical economists rejected all demand failure arguments. To reconcile all of this, Kates insists that classical theories of business cycles focused on "the structure rather than the level of demand" (p. 19). In short, by attributing his peculiar definitions to the classical economists, Kates alters their original arguments.

In truth, Kates has a tendency to read things out of context. Consider, for example, Kates's distinction between James Mill and Say: "Mill's conception is that one buys with one's own products, which is profoundly different from Say, who conceived of production as opening opportunities for others to sell" (p. 35). Kates insists that this represents an important difference between the two, in spite of the fact that both Say and Mill explicitly pointed out that these propositions are in fact two sides of the same coin. Say even quoted himself on this in his Letters to Malthus:

I had said 'As each of us can only purchase the productions of others with his own productions - as the value we can buy is equal to the value we can produce, the more men can produce the more they will purchase.' Thence follows the other conclusion . . . 'that if certain goods remain unsold. it is because other goods are not produced; and that it is production alone which opens markets for produce.' (Say 1821, p. 3)

Moreover, on the very page from Mill that Kates cites in order to contrast Mill and Say, Mill actually writes: "The production of commodities creates, and is the one universal cause which creates a market for the commodities produced" (Mill [1808] 1966, p. 135).

The differentiation that Kates makes between Mill and Say suggests that Kates does not understand the logic that unified Say's and Mill's arguments on markets, gluts, and the coordination of trade. This, in turn, also explains Kates's presentation of Say's diverse and yet logically congruous arguments as disparate. Kates does not understand how Say applied his aggregate-consistency insights in different contexts, and this explains why Kates sees the various arguments as separate, logically unrelated, and even contradictory. This conclusion, that Kates does not really understand Mill and Say, is further reinforced by Kates's misinterpretation of Say's arguments to the effect that gluts of some commodities are generated by shortages of other commodities. Kates is so confused about the argument at hand that he actually talks about Say's "search for phantom goods which were not produced which undermines his perceptions of the nature of recessions" (p. 31).

Kates is equally obtuse in his treatment of Clower and Leijonhufvud. Consider the fact that, after reciting some of Clower and Leijonhufvud's (1973) seminal article on the logic of Say's arguments, Kates complains, "[W]hat they do not do is explain the implications for economic theory and policy if Say's Law is valid" (p. 205). He even bemoans the fact that there "is no mention of Say's Identity or Say's Equality" (p. 200). This in spite of the fact that the logic of the article in question has shown these two propositions to be logically fatuous, based on out-of-context quotes, and, as Schumpeter had earlier put it, "quite irrelevant for Say's purposes" (Schumpeter 1954, p. 619). In short, Kates misses the point.

The fable of Say's Law we usually find in the textbooks is based on little more than a series of escalating misquotes, wherein each subsequent writer has carried the misinterpretations of the originals a bit further. The succession of perpetrators has included some of the greatest authorities of 20th century economics. The provocative title of Kates's book notwithstanding, he is too respectful of these authorities. He is consistently willing to rely on authority over careful reading of original sources. Paradoxically enough, this book is well worth reading because it puts these authorities on display. There is a lot of "he wrote" in Kates's book, and a critical reader, armed with a good understanding of Skinner (1967) and Clower and Leijonhufvud (1973) and who is willing to dig out original sources, will find the book very revealing.

References

Ahiakpor, James C. W. 1997. Full employment: A classical assumption or Keynes's rhetorical device? Southern Economic Journal 64:56-74.

Ahiakpor, James C. W. 1998. Keynes on the classics: A revolution mainly in definitions. In Keynes and the classics reconsidered, edited by James C. W. Ahiakpor. Kluwer Academic Publishers, pp. ??-??.

Clower, Robert W., and Axel Leijonhufvud. 1973. Say's principle, what it means and doesn't mean. Intermountain Economic Review. Fall. Reprint, Money and markets: Essays by Robert W. Clower, Chapter 12. Cambridge: Cambridge University Press. 1984.

Keynes, J. Neville. 1894. Demand. In Dictionary of political economy, Volume 1. edited by R. H. Inglis Palgrave. pp. 538-42.

Marx, Karl. 1929. Capital. (Translated from the 4th edition of Das Kapital, Book 1, Der Produktionsprozess des Kapitals, by Eden and Cedar Paul.) New York: International Publishers.

Marx, Karl. 1952. Theories of Surplus Value. (Translated from Theorien uber den Mehrwert by G. A. Bonner and Emile Burns.) New York: International Publishers.

Mill, James. [1808] 1966. Commerce defended. In James Mill: Selected economic writings, 2nd edition. edited by Donald Winch. Edinburgh: Oliver and Boyd, pp. ??-??.

Say, Jean-Baptiste. 1821. Letters to Mr. Malthus on several subjects of political economy and on the cause of the stagnation of commerce. Translated by John Richter. London: Sherwood, Neely, and Jones. Reprint, New York: Augustus M. Kelley, 1967.

Schumpeter, Joseph A. 1954. History of economic analysis. New York: Oxford University Press.

Skinner, A. S. 1967. Say's law: Origins and content. Economica 34:165-6.

Taylor, Fred M. 1909. Principles of economics. Ann Arbor, MI: Privately printed.

Petur O. Jonsson Fayetteville State University