Noah Millman sets me straight on Keynesianism:

Japan has been mired in a deflationary depression for 10 years, and has been priming the pump furiously, to no obvious effect.
The Keynesian National Recovery Act and other New Deal legislation did little or nothing to end the Great Depression, and may have prolonged it. What the WPA and the like did accomplish was charitable (it put people to work and thereby kept them from hunger) and psychological (it told people someone was doing something for them, and thereby kept political stability). But the New Deal did not bring economic recovery. What finally brought about a solid and sustained economic recovery in the 1950s was the revival of global trade after WWII.
If you buy into the basic assumptions of monetarist and/or supply-side economics, you can’t also buy into Keynesian economics. They are mutually exclusive. Monetarists believe that inflation and deflation are purely monetary phenomena. Supply-siders think what drives growth in the long term is the marginal return on invested capital. Keynesians think that what drives both inflation and growth is “aggregate demand” which they think the government can manipulate through deficit spending and/or running a surplus. If growth is low and inflation low, a Keynesian would say that government should spend more to increase aggregate demand. A supply-sider would say that this will actually decrease growth because the increased spending will require raises in taxes or in borrowing (which implies hikes in future taxes, which would be discounted back to the present by the market), which in turn would reduce the marginal return on invested capital. And further, a supply-sider would say, this deficit spending would increase inflation because, by reducing returns on investment, it would encourage consumption over investment. With an uptick in inflation and a drop in real growth, you could get a rise in nominal growth, which would look like a recovery to debtors and other folks at the bottom of the economic barrel, but for the economy as a whole you’d have a net loss.
Of course, you can also not buy into ANY economic theory, and simply conclude that economics is a branch of psychology, and therefore whatever works works. This was basically the Clinton/Rubin economic stance: don’t worry about the theory, just do what the markets seem to want and all will be well. That works fine until the market decides that you really have figured out how to ensure permanent low-inflation growth, at which point you’ll have a speculative blow-off and a deflationary crash. Which is what just happened to us. This is the Austrian “malinvestment” thesis of where depressions come from–the market periodically “overinvests” after which there’s a deflationary recession to work off the excess capacity thus created – and the standard critique from economists is that it isn’t an economic theory (why should the market “overinvest” sometimes but not other times?) but a psychological one. Which is true. But that doesn’t disprove the theory.