A Comparison Contrast of Keynes's 1933-1935 IS-LM Model Incorporating Degree of Confidence (The State of the News) with Hicks' 1937 IS-LM Model without Degree of Confidence: D. Champernowne Got It Right in 1936

Abstract

Between 1933 and 1935, Keynes used his own original version of the IS-LM model in his lectures concerning his new approach to economic theory, the General Theory. This model explicitly incorporated the degree of confidence that decision makers have in their calculated expectations, which is a probabilistic concept, in both the IS curve and LM curve in the variable termed the “State of the News”. Keynes did not explicitly emphasize this model in the General Theory because he could not explicitly make use of his interval valued approach to probabilities, which occurs if there is any state less then complete confidence. Instead, he substituted a generalized version, his D-Z model, of AC Pigou’s special theory of an aggregate demand function and an aggregate supply function, for his original IS-LM model. Keynes carried out this analysis in chapter 3, the appendix to chapter 19, chapter 20, and chapter 21 of the General Theory. This allowed him to specify his Aggregate Supply Curve, which is a locus of all expected, possible D-Z intersections that satisfy the necessary and sufficient first and second order conditions for maximizing economic profits in a purely competitive world in the short run with one variable input, labor, and one fixed input, capital. Keynes was able to do this because he replaced Pigou’s realized prices and profits with a set of expected prices and expected profits, expectations, that incorporated the idea of definite, numerical, as well as indefinite non-numerical (intervals) probability. Champernowne explicitly incorporated the State of the News into both of his IS-LM functions as independent variables and explicitly showed that his variable, which would shift both curves, was missing from the classical and neoclassical models where expectations are strictly based only on a concept of precise probability. Hicks read Champernowne’s paper before he submitted his own paper for publication in Econometrica in 1937. He deliberately left this variable out because he knew that it did not fit into the classical model he had created for comparison with Keynes’s model and did not understand what Keynes was talking about because his concept of probability interpreted the probability estimate as being the confidence expressed by a decision maker. Nevertheless, Hicks then attempted to rectify this situation by specifying a completely, infinitely elastic left tail for his LM curve that did reflect, but in an ad hoc manner, the lack of confidence of decision makers in the state of the news.