Rationality in the mind's eye: An alternative test of rational expectations using subjective forecast and evaluation data

Abstract

The purpose of this paper is to examine if business managers' forecasts are rational in a constructionist sense. Adherents to constructivism believe that human senses do not convey “real reality”, but rather a set of experiences that are conditioned by beliefs and differences in innate cognitive attributes to yield a constructed and inherently subjective perception. We examine whether managers' forecasts, when measured against their own perceptions of reality, are rational. The Conference Board provides a data base that asks managers for a subjective prediction of the future, and a subjective evaluation of the past. Both the forecast and the evaluation are subjective, conceived in the mind of the forecaster. If rational expectations is a theory that accurately represents the decision process, then it should be reflected in data where the forecaster computes not only the prediction, but also the value of the parameter that was predicted. Thus, this data allows forecasts to be tested for rationality using managers' own perceptions as the basis for judgement. If rational expectations is an accurate representation of how economic actors form forecasts, then a mind's eye evaluation should show evidence of this rationality. Yet, though the managers generate both the forecast and the evaluation, they failed four of the five traditional tests for rationality.