Monetary Policy
Unexpected effects

In Discussion Paper No. 291, Research Fellow Neil Rankin explores the effects on the long-run path of an economy of the processes by which economic agents learn, or form their expectations. In particular he examines the implications of different expectational regimes for the ability of monetary policy to affect real output.
Rankin builds on the idea that imperfect competition may provide useful insights into macroeconomic behaviour. This idea has been extensively explored in recent literature, much of which, including Rankin's previous work in Discussion Paper No. 205, has used the framework of a monetary temporary equilibrium model. In such models, money does play a role but expectations are assumed to be formed in an essentially subjective fashion, as arbitrary functions of current and past values of economic variables. This may be appropriate in a static, short-run analysis, Rankin argues, but it leaves open questions as to what would happen over a longer period, during which it might be reasonable to suppose that all agents would have learned enough about the workings of the economy to form expectations which are, on average, correct, i.e. to form expectations `rationally'.
Rankin uses a model of imperfect competition, in which dynamics are introduced via an `overlapping-generations' model of the type familiar in monetary theory. In general, output and employment are below their full employment levels. The model is analysed under three different hypotheses regarding expectations formation, `adaptive', `monetarist' and `pure rational'. Rankin finds that these three formulations of expectations formation lead to different steady states, with contrasting output responses to the rate of monetary growth, despite constraining expectations to be `rational' in the steady state.
Rankin therefore demonstrates how, although agents eventually form expectations which are `rational', the process by which they arrive at these expectations does affect the course of the economy in the long run. This contrasts with most of the recent literature, in which it is assumed that as long as agents eventually learn to hold rational expectations, the ultimate course of the economy does not depend on the precise manner in which they learn to do so. While this assumption is often defended as a convenient simplification, Rankin concludes that it can be misleading: different expectations formation processes lead to different steady states, even though expectations converge to correctness under all these processes. His analysis shows that different mechanisms of expectations-formation cannot be ignored as being of only transient importance.

Imperfect Competition, Expectations and the Effectiveness of Monetary Policy
Neil Rankin

Discussion Paper No. 291, March 1989 (IM)