Steven Horwitz
In support of George Selgin's comment defending the idea that Say's Law allows for general gluts if money is in excess demand, let me quote from my piece on Austrian economics and Say's Law that can be found here (see the section starting on p. 92). My apologies for the length.
The place of Say's Law in Austrian macro begins with the observation that virtually all exchanges in a market economy are exchanges of goods for money. That is, money is half of every exchange. As a result. changes in money can affect the entire economy in systemic ways. Excesses and deficiencies in the money supply will distort the money prices that guide economic activity by influencing, from the 'money side,' the exchanges that create those prices. The link to Say's Law is that these acts of monetary exchange are what bring together acts of production with acts of demand. In order for production to be the source of demand, sellers of products must be able to acquire money in return and then spend that money on the goods and services they wish to purchase. The Say's Law transformation of production into demand is mediated by money.
The key is that the supply of money must be 'right' in order for demand to properly reflect production. If money is not right. then gluts and shortages will occur. As was often argued in response to Say's Law, how can the interpretation that aggregate supply will always be sufficient to match aggregate demand be valid when we have so many examples of real world excesses and deficiencies in the supply of goods? If Say's Law stated that way were true, we would never see gluts or shortages. The problem with this interpretation is, as Say himself recognised, that it ignores that mediating role of money. For example. if money is in short supply, some producers will be unable to demand, as they will be unable to sell their goods for money. given that it is relatively unavailable. The result will be a glut, as goods and labour sit unsold. Conversely, an excess supply of money will, in the short run, heighten overall purchasing power even though there has been no sustainable increase in production. That is, demand will be greater than ultimately justified by the level of production. Over time, the best that an excess supply of money can do is to redistribute purchasing power in ways that prevent it from matching up with those who have created it via production. The eventual result of all of this will be shortages, and the rising prices we associate with excess supplies of money.
This is nothing more than Walras's Law - excess supplies of money must imply excess demands for goods, and excess demands for money must imply excess supplies of goods. Walras's Law is thus an implication of Say's Law, combined with the recognition that all exchanges are monetary exchanges. If production is the source of demand, and the translation of production into demand takes place via money, then too much money will mean 'too much' demand for goods, and too little money will mean deficient demand for goods.
...
This section's discussion can serve as a rebuttal to those who argue that Say's Law is violated by historical episodes of recession or general gluts (see Uchitelle 2001, for example). Say's Law does not say that general gluts or shortages can never occur. Rather it explains a principle by which markets operate. Whether the effects of that principle will be beneficial or not depends on the institutions that frame the markets in question. Just as Smith's invisible hand would still operate, but produce undesirable consequences. where property rights are not protected from public or private predation, so too will Say's Law not produce desirable outcomes when certain institutional prerequisites are not in place. This section's discussion suggests two crucial institutional prerequesites: the maximum flexibility of prices possible given that a certain level of 'stickiness' is inevitable, and the maintenance of monetary equilibrium. If those two are met, then Say's Law implies that general gluts and shortages are not possible. If they are absent, then the process Say's Law identifies will still operate, however it will not produce the benign results it would under the right institutions. Where prices are excessively sticky and/or where monetary disequilibrium is present, general gluts and shortages are possible, and their existence is a confirmation, not a refutation, of the principle expounded in Say's Law.
Like so many other important insights in economics, the organizing principle that is Say's Law provides us a framework for understanding how different scenarios will play out depending on the institutional arrangement that frames human action.
