Probability, Econometrics and Truth: The Methodology of Econometrics

Rating:
Summary: The Best Book Written On Econometrics,Philosophy,andJMKeynes
Review: This book is already a classic.Keuzenkamp(K)blends the philosophy and history of probability and statistics with that of econometrics in such a skillful fashion that the technically trained reader(upper division courses in mathematical statistics,probability,econometrics and philosophy)will want to start reading it over again as soon as he finishs the book the first time.K correctly realizes that the goal of econometrics can only be that of estimating and not testing.K correctly shows that econometrics should be founded in the positivist(logical empiricist)tradition.Finally,an academic economist gives J M Keynes's arguments,made by Keynes in his debate with Tinbergen over the ability of econometricians to Test(with a capital T as in Truth) different theories of the business cycle,their just due.There is only one small(very small)quibble.On pages 272-273,K states:"Furthermore, an investigator may try different prior distributions in order to obtain upper and lower limits in probabilistic inference(this approach is proposed by Good and Leamer)".In fact,J M Keynes was the first advocate of this approach in Part V of A Treatise on Probability.The interested reader can find Keynes's interval approach in chapters 5,10,15,16,17,20,22,29 and 30.K has destroyed the myth,created by many philosophically illiterate economists and econometricians over the last 60-65 years,that Keynes's knowledge about mathematics and statistics was "rusty" and that he was just trying to shoot the sound ideas of other researchers "full of holes" because he didn't understand what they were doing.Keynes understood only to well.