Money, Output, Rational Expectations and Neutrality: Some Econometric Results for the UK

Abstract

A substantial body of theoretical literature associated with the new classical school of macroeconomics demonstrates that, if real variables, such as output and employment, deviate from their market-clearing, natural, rates only because of random exogenous shocks or unanticipated changes in policy variables, such as the money stock (i.e. if structural neutrality (SN) holds), and if economic agents form their expectations according to the rational expectations hypothesis (REH) (see Muth, 1961), then economic policy will be ineffective (see for example Lucas, 1972, 1973, 1975; Sargent and Wallace, 1975, 1976; and Sargent, 1979). This policy ineffectiveness hypothesis has recently been the subject of intensive econometric investigation, which began with the seminal empirical studies on the effectiveness of monetary policy in the United States by Barro (1977, 1978). Barro's studies, and those of Barro and Rush (1980), again for the United States, Wogin (1980), for Canada, and Attfield, Demery and Duck (1981a, b) for the UK, all support the hypothesis that money is neutral. All these studies have in common the feature that they test for the joint hypothesis of structural neutrality and rational expectations (RESN). In contrast to these studies, Leiderman (1980) has shown that, within the framework of the type of model that has been used to test for the effectiveness of monetary policy, the separate hypotheses of RE and RESN are uniquely nested within the general structure of these models, and therefore that the nested hypothesis testing methodology, discussed by Mizon (1977a, b), may usefully be adopted to gain more information about the validity of the constituent hypotheses of RE and SN. Leiderman found that, in an extended version of Barro's (1977) model of money growth and unemployment, the hypotheses of RE and SN could not be rejected separately or jointly. This finding was taken by Leiderman as confirming the new classical hypothesis that monetary policy is neutral. The purpose of this paper is to report the findings of an empirical study for the United Kingdom similar to that undertaken by Leiderman for the United States. Our findings for the UK are at variance with those of Leiderman for the USA, and also with those of Attfield et al. (1981a, b) on the validity of the joint hypothesis of RESN for the UK. We reject the view that monetary policy can have no effect on real output in the short run. The paper is organized as follows. In Section I we present the model, or maintained hypothesis, against which the restrictions, implied by RE and SN, are tested. Section II contains a report of the empirical tests of the new classical policy ineffectiveness hypothesis and its component hypotheses. Some concluding comments are given in Section III.