Adaptive expectations

Abstract: This chapter discusses money-wage dynamics and labor-market equilibrium. A generalized excess-demand theory of the rate of change of the average money-wage rate has been developed for frictional labor markets that allocate heterogeneous jobs and workers without having perfect information and market clearance by auction. There are two explanatory variables: the vacancy rate and the unemployment rate. The unemployment rate and the rate of change of employment are shown to be joint proxies for the vacancy rate. Hence, generalized excess demand can be regarded as a derived function of the unemployment rate and the rate of change of employment. This relationship is the augmented Phillips curve. Some of its properties are deduced. Equilibrium entails equality between the actual and expected rates of wage change. The steady-state equilibrium locus is implied to be a vertical line at a unique steady-state equilibrium unemployment rate. This is consistent with the usual theory of anticipated inflation. But if there are downward money-wage rigidities, then, up to a point, every one percentage point increase of the expected rate of wage change produces less than a one percentage point increase of the actual rate of wage change. The steady-state equilibrium locus will then have the characteristic negative slope of the Phillips curve in the range of large unemployment rates. But at sufficiently small unemployment rates, equilibrium is impossible and under the adaptive expectations theory, an explosive hyperinflation will result.

Abstract: Presenting an analysis of the rise and fall of U.S. inflation after 1960, this book examines two broad explanations for the behaviour of inflation and unemployment in this period: the natural-rate hypothesis joined to the Lucas critique and a more traditional econometric policy evaluation modified to include adaptive expectations and learning. The text begins with an explanation of how American policymakers increased inflation in the early 1960s by following erroneous assumptions about the exploitability of the Phillips curve - the inverse-relationship between inflation and unemployment. In subsequent chapters, it connects a sequence of ideas, such as self-confirming equilibria, least-squares and other adaptive or recursive learning algorithms. The author synthesizes results from macroeconomics, game theory, control theory, and other fields to extend both adaptive expectations and rational expectations theory, and he explains postwar inflation in terms of drifting coefficients. He interprets his results in favour of adaptive expectations as the relevant mechanism affecting inflation policy. This book is intended for academics, graduate students, and professional economists.

Abstract: I. Introduction, 227. — II. Akerman's argument, 229. — III. Reformulation in terms of adaptive expectations, 230. — IV. An alternative but equivalent formulation, 233. — V. Some empirical results, 236. — VI. More complicated models, 238. — VII. Conclusions, 240.

Abstract: Organizations have been modeled as goal directed systems which use simple decision rules to adapt behavior in response to performance feedback. This paper examines the formation of organizational goals, or aspiration levels, over time in groups of individuals representing top management teams of simulated organizations. The analysis compares the empirical validity of an adaptive attainment discrepancy model with models derived from rational and adaptive expectations theories. The results suggest that the attainment discrepancy model, which is based on a simple decision rule of adjustment to performance feedback, provides the most robust description of aspiration formation. They are also informative with regard to the application of expectation models to aspiration formation: There is a great deal of similarity between these results and those of prior studies on expectation formation. In addition, the study finds that there tends to be an optimistic bias in aspiration formation, that adaptation is not consistently incremental, and that adaptive learning may, over time, lead to behavioral outcomes that are consistent with rationality.