Paul Davidson Comments To Posner Article...

I am pleased that Posner has read what I (Davidson) wrote. Yes P{osner is correct that I have used Keynes as the basis for my analysis of uncertainty and Post Keynesian theory. When Keynes wrote, unfortunately, the theory of stochastic probability theory [ ergodic theory] had not been developed in the English language. Thus Keynes’s uncertainty is not terchnically defined and can easily be thought to be the same of Knight’s concept of uncertainty – but they are quite different. Knight is using an epistemological definition of uncertainty, Keynes has an ontological uncertainty concept. This makes a big difference in understanding how financial institutions and money contracts, especially loan contracts i.e., leverage making contracts, affect financial crisis. What I have done was to use technical language to show Keynes’s general theory rejects several axioms that underlie classical theory. The Classical theory's ergodic axiom underlies Friedman’s monetarism, Lucas’s rational expectations, , andthe laissez-faire economic philosophy Fama’s efficient market theory. An axiom is defined as a universal truth that needs not be proven.The ergodic axiom presumes that the same probability distribution that governed past economic outcomes governs all future outcomes. Thus, given the ergodic axiom, the future is statistically predictable– and rational decision makers know (in the actuarial sense at least) what the future is when they make a decision today. Thus, a rational economic person would never sign a loan contract unless he/she “knew” they could service this debt within their known future income and budget constraints. In a classical theory, there can never be a default by optimizing rational people, there can be no foreclosures, and no insolvencies. Hence the theories based on the ergodic axiom cannot develop a useful policy to solve these financial systemic problems when they occur in the world of experience. Keynes rejected the ergodic axiom as a basis for his general theory. Thus, in his general theory, there need not exist any current objective probability distribution that decision makers “know” will govern future outcomes. Without going into details, Knight’s unique events that cause uncertainty is the equivalent of Taleb’s black swan – an occurrence that occurs in an ergodic system but that will have a very low, but still fixed probability , of occurrence. So with a big enough sample one can predict the existence of a black swan financial disaster in a Knight system. In a nonergodic system there is no probability (as Keynes stated in his 1937 article) on which to estimate future outcomes. Thus the necessity to seal economic transaction with monetary contracts that FIX payments into the future!

Read more: http://www.newyorker.com/online/blogs/johncassidy/2010/01/interview-with-richard-posner.html#ixzz0dBN5g3Hs