Testing the rational expectations and structural neutrality hypotheses: Some further results for the U.K.

In this paper we test directly for the restrictions implied by rational expectations and structural neutrality. The tests are direct in the sense that they employ theLucas output equation father than the approximations to it, which replace the lagged output term by lagged monetary shocks, commonly considered in the literature. The direct approach is considered preferable because it avoids these ad hoc approximations and saves significantly on degrees of freedom. The latter permits us to pursue a fully nested testing procedure, which was not possible in earlier work employing postwar U.K. annual data. The main result is that the rational expectations restrictions are not accepted.