On the Immense and Severe Confusions and Muddles About the Nature of Subjective and Objective Probability in the Theory of Rational Expectations: A Subjective Probability (Distribution) Can Never Become an Objective Probability (Distribution) or a True or a Correct Probability

Abstract

Starting with J. Muth’s unsupported and unsupportable claims, originally made in 1961, that “rational expectations” were subjective probability distributions that were distributed around a known, true, objective probability distribution, various economists have provided the same type of unsupported and unsupportable claim to assert that subjective probability distributions become equal to objective probability distributions. This is impossible since the only restriction allowed in the subjective theory of probability is that the subjective probabilities must be additive, so that they are coherent, which means that they conform to, and are consistent with, the mathematical laws of the probability calculus. Any other restrictions added, as done by Muth initially, and later by rational expectationist adherents like Lucas and Sargent, to this requirement of additivity, are rejected.

The many restrictions incorporated by rational expectationists lead to the limiting frequency theory of probability, which can only hold in the very long run as one approaches infinity. Rational expectationists, however, claim that the decision makers can know the limiting, convergent behavior in the short run of the long run series of observations. They commit the logical fallacy of conditional a priorism, or long runism, that was examined repeatedly by the philosopher of science, Nicholas Rescher, in the 1970’s. None of the claims made by rational expectationists can ever hold in the short run. For instance, the Rational expectationist claim that the Phillips Curve is vertical in the short run is a perfect example of the logical fallacy of conditional a priorism.

Keywords: Rational Expectations, Savage Ramsey De Finetti, Objective Probability, Subjective Probability

JEL Classification: B10, B12, B14, B16, B20, B22