"Say's Law sank without trace."

Money: Whence It Came, Where It Went, J. K. Galbraith (1975), pp. 218-220:

The belief that the economy would find its equilibrium at full employment depended partly on what had long been called Say's Law — for J. B. Say, the French counterpart and interpreter of Adam Smith — and partly on the corrective movement of wages, prices and interest rates when there was unemployment. Say's Law, not a thing of startling complexity, held that, from the proceeds of every sale of goods, there was paid out to someone somewhere in wages, salaries, interest, rent or profit (or there was taken from the man who absorbed a loss) the wherewithal to buy that item. As with one item, so with all. This being so, there could not be a shortage of purchasing power in the economy. Movements in prices, wages and interest rates then validated J. B. Say and also ensured that the fundamental tendency of the economy would be to operation at full employment. People and firms saved from their income, and this saving had, obviously, to be spent. This happened when it was invested in housing, plant, capital equipment. If people saved more than was invested, the surplus of savings would bring down interest rates. Investment would thus be stimulated and saving (at least in theory) discouraged. So the excess of savings would be eliminated and Say sustained. Prices of goods would also fall in consequence of any short-fall in purchasing power that resulted from an excess of savings. This would encourage buying and, by reducing the income from which savings were made, also reduce savings. Again Say was sustained. Until Keynes, Say's Law had ruled in economics for more than a century. And the rule was no casual thing; to a remarkable degree acceptance of Say was the test by which reputable economists were distinguished from the crackpots. Until late in the '30s no candidate for a Ph.D. at a major American university who spoke seriously of a shortage of purchasing power as a cause of depression could be passed. He was a man who saw only the surface of things, was unworthy of the company of scholars. Say's Law stands as the most distinguished example of the stability of economic ideas, including when they are wrong. 

Supplementing Say, as noted, were the forces that kept the economy at full employment. These too were relatively straightforward. Were there unemployment, the competition for jobs would bring a fall in wage rates. Prices would be less immediately affected by the unemployment. The relationship of prices to costs would thus be made more attractive — real wages would fall — and workers whose employment was previously unprofitable to employers would now be hired. The fall in wages would not affect purchasing power; because of Say, that was always sufficient. Employment would continue to expand until the approach to full employment raised wage costs and arrested the hiring. Thus did the economy find its equilibrium at or very near full employment. From this also came the one decisive recommendation of the orthodox economists for ending unemployment. Do nothing to interfere with the reduction of wages in a depression. Resist all siren voices, including that of Herbert Hoover, who, it will be recalled, urged against wage cuts. On no matter was compassion so softheaded, for to keep up wages merely perpetuated the sorrow of unemployment and the sorrows of the unemployed. This was the doctrine, perhaps more accurately the theology, that Keynes brought to an end. There are numerous points of entry on his argument; perhaps the easiest is by way of the rate of interest. Interest, he held, was not the price people were paid to save. Rather it was what they got for keeping their assets in plant, machinery or similarly unliquid forms of investment — in his language, what was paid to overcome their liquidity preference. Accordingly, a fall in interest rates might not discourage savings, encourage investment, ensure that all savings would be used. It might cause investors to retreat into cash or its equivalent. So interest rates no longer came to the support of Say's Law to ensure that savings would be spent. And if Say's Law was no longer a reliable axiom of life, the notion of a shortage of purchasing power could no longer be excluded from calculation. It might, among other things, be the consequence of a reduction in wages. 

What people sought to save, in Keynes's view, had still to be brought to equal what they wanted to invest. But the adjustment mechanism, he argued, was not the rate of interest but the total output of the economy. If efforts to save exceeded the desire to invest, the resulting shortage of purchasing power or demand caused output to fall. And it kept falling until employment and income had been so reduced that savings were also reduced or made negative. In this fashion savings were brought into line with investment — which also, meanwhile, would have fallen but by not so much. The economic equilibrium so established, it will be evident, was now one in which there was not full employment but unemployment. Thus unemployment for Keynes was a natural condition of the economy. There was much else. And not all of Keynes's argument survived. The liquidity-preference theory of interest, for example, though it served Keynes's argument, did not gain permanent acceptance as a description of reality. But on two things Keynes was immediately influential. Say's Law sank without trace. There could, it was henceforth agreed, be oversaving. And there could, as its counterpart, be a shortage of effective demand for what was being produced. And the notion that the economy could find its equilibrium with unemployment — a thought admirably reinforced by the everyday evidence of the '30s — was also almost immediately influential.