Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Recent literature on cointegration and unit roots has focused attention on the distribution of test statistics frequently used to test efficiency in rational expectations models. This paper concentrates on the permanent income hypothesis of real consumption. The authors illustrate, by using the proper asymptotic theory and small-sample approximations, the cases in which tests of such a hypothesis are biased towards rejection and cases where they have the correct sizes. Their results serve to interpret numerous Monte Carlo studies in the literature on this issue. Special emphasis is placed on the distinction between "weak" and "semistrong" rationality tests. Copyright 1988 by Royal Economic Society.