Although primarily known as an economist, Keynes produced one philosophical classic, the Treatise on Probability (1921). This develops the theory of probability and confirmation theory on the basis of an objective, logical relation of degrees of implication amongst propositions. Although subsequent work has not been kind to such a notion, Keynes’s working-out of its possibilities remains, together with the work of Carnap, the main showpiece of confirmation theory. In particular, Keynes realized that in order for induction to increase the probability of generalizations as progressive evidence eliminates potential falsifications, a ‘principle of limited independent variety’ must be assumed, giving the necessary structure for probability to increase.

In economics, Keynes held that markets were very unlike simple physical systems. There is no reason to suppose that they can attain an equilibrium to which they return after disturbances. They are in general highly unpredictable, vulnerable to irrational swings of confidence and fear, and their uncertainties defy analysis in terms of probabilities and computable risks. Keynesianism labels the view that recessions are at least as much due to failure of demand as to failure of supply, so that in order to counteract a recession or depression the government should increase its own expenditure. The opposite view, that governments need instead to contract their expenditure became widely held in the late twentieth century, largely on the back of the monetarist doctrines, including the unrealistic assumption that even in fearful times agents would prefer to invest money in productive enterprises rather than to hoard it. See also animal spirits, fiscal policy, market fundamentalism, Say’s law.