Thoughtful Economic Man: Essays on Rationality, Moral Rules and Benevolence [Hardcover]

This book of collected essays(eight essays and an introduction)edited by J.Gay Tulip Meeks(M)combines some very good essays on benevolence and morals by Sen and Hahn with some very poor essays on rationality,decision making and J M Keynes's views on probability,induction,decision making and uncertainty as presented by Keynes in the A Treatise on Probability(1921;TP) and the General Theory(1936;GT)by Broome,Matthews,and Meeks.Broome appears to be ignorant of the fact that the entire Ellsberg Paradox literature, which demonstrates the inapplicability of Savage's sure thing principle to decision making under ambiguity, was completely anticipated by Keynes in his two Cambridge fellowship theses of 1907 and 1908 and in the TP of 1921.Keynes presented the first decision rule in the history of Decision Theory that made use of decision weights.He called this rule " a conventional coefficient of risk and weight,c".Keynes explicitly uses the c coefficient to solve the general form of the Ellsberg Paradox problem,as well as various other types of anomalies,such as preference reversals, on p.315,ft.2.Matthews and Meeks are likewise deficient in failing to specify the relationship that exists between the c coefficient and the investment decision making under uncertainty discussion that Keynes presented informally on pp.239-241 of the GT(the need to discount the NPV rule applied to the mec schedule in chapter 11 of the GT for both risk and uncertainty[liquidity preference]besides the discount for time preference).This discussion is directly related to Keynes's analysis of the investment decision making process in the long run under conditions of uncertainty in chapter 12 of the GT ,which is the main topic of both the essays by Matthews and Meeks.Another severe problem arises for Meeks explicitly and Matthews implicitly in their discussions of Keynes's approach to the quantification and measurement of probabilities.Meeks explicitly follows the unsupported and unsupportable claims made by Sir Harold Jeffreys and Frank P Ramsey that Keynes's meaning of the terms "nonnumerical" and "nonmeasurable"in chapter 3 of the TP was that numbers could not be used ,in general,to measure a probability.Based on these false claims,it was then asserted that Keynes had created some kind of a strange nonnumerical probability(ies) that did not obey the probability calculus.The best a decision maker could do in some limited instances was to possibly ordinally rank some of the relevant probabilities some of the time.A careful reading of chapters 3,5,10,15,17,20, and 22 of the TP reveals that most Keynesian probabilities are in fact interval estimates that require TWO NUMBERS,not one number,in order to quantify the probability relationship.The many errors in Meeks' paper(see pp.144-149 and pp.152-155) all revolve around her belief that Keynes's approach was aimed at "...providing some(almost always nonnumerical)'probable knowledge'of what is to come..."(Meeks,p.153).All of Keynes's analysis in Part III of the TP on analogy and induction is aimed at providing numerical limits for the inductive conclusions.Meeks's claim that the best a decision maker can do is to compare probabilities ordinally some of the time would vitiate all of Part III of the TP.