It can be unfair to take someone's most famous dicta out of context and rebut them. But then, Keynes triumphed partly through clever soundbites. For example, consider two of them.
1. "In the long run, we are all dead."
This claim is false.
What could I possibly mean by that?
I do not mean, here, that we shall all be resurrected at the last day, though in another context I would say so. Rather, one must understand how "long run" and "short run" are used in economics. Long run means a time period over which all factors can be adjusted. For example, it might refer to the time period over which wages, or the capital stock, can equilibrate after a shock. In the short run, something is fixed. For example, one might say that short-run supply is upward-sloping, but long-run supply is flat, meaning that in the short run, entrepreneurs cannot adjust their capital stocks, so they will expand labor only and produce inefficiently, but in the long run entrepeneurs can get capital and labor right.
Estimates of the short run tend to be two years or less; of the long run, five years or more, with 2-5 years sometimes vaguely called the "medim run." Of course, this depends on the particularly characteristics of a given market. But a period of say, twenty years, would always be called the long run. Well, guess what? In twenty years, most of us will still be alive.
The serious point is that "in the long run, we're all dead" is used as an excuse for focusing on the short run. The rebuttal "no, in the long run, most of us are still alive" might serve as a useful reminder of where economic policy should really focus.
Possibly there was a dark double entendre here. Living in a time of revolutionary turmoil, perhaps Keynes half-meant: "In the long run, if we don't give people jobs soon, there will be a communist revolution, and I and all the other bourgeois capitalists will end up in front of a firing squad. We need policies that can ease the short-run pain long enough to avert revolution, and never mind the ultimate cost." OK, but if that's what Keynes meant, why is he so influential now, when capitalist legitimacy is secure?
2. "When the facts change, I change my mind. What do you do, sir?"
It is true only in a superficial and misleading sense that people should change their minds when the facts change. If the light changes from green to red, I change my belief from "The light is green" to "The light is red." But the two beliefs are not really contradictory. We might illustrate this by adding an implicit time-qualifier to each statement, so that my beliefs are (a) "The light is green at time t1" and (b) "The light is red at time t2." And of course, I do not abandon belief (a) when belief (b) appears.
So it is not always the case that one should change one's mind when the facts change. Of course, if you have theories that make predictions extending into the future, and if the future falsifies those predictions, you should modify or abandon those theories. It is a mark of a good theory, however, that it does not need to be modified or changed when new facts appears, for it has anticipated them, or at least, it has not anticipated the contrary.
The epistemology of this Keynesian dictum, then, is irresponsible and complacent. It suggests a person who is always inventing new theories to fit new facts, and discarding them as easily, and feeling no compunction about it. Theorists must aim higher than that.
While Keynes often appealed to "the facts" in order to rebut the classical economists, it is naive to think there are facts without theory. The raw data of experience is always interpreted through the lens of theories, and a person who thinks he relies on "just the facts" is merely a person too naive to recognize that he uses theories, and therefore unable to think critically about the theories he is using. Keynes seems sometimes to go astray because of this just-the-facts fallacy. For example, the concept of "propensity to consume," plays an important role in Keynesian theory, tacitly supposes that consumption is a function of income alone. A theorist starting from individual rationality would know better, and Milton Friedman and others later showed that consumption depends mainly on lifetime rather than current income.