Introduction

While there is general agreement on the short-term effect of monetary policy, economists tend to agree that money has no effect on production and unemployment in the longer term characterised by wage and price flexibility and rational expectations.

This chapter surveys the discussion in the literature as to whether money can influence production in a model setting characterised by four features: (i) economic relationships result from the agents' optimisation, (ii) agents are able perfectly to forecast future levels of economic variables or, in the case of stochastic variables, the stochastic distributions of economic variables, (iii) agents have preferences which include only real variables, and (iv) there are no costs associated with changes in nominal variables. We speak of ‘monetary neutrality’ if money has no impact on production in such a model setting.

This point of departure leaves out a number of channels where a monetary impact on production would seem largely to have an effect over a shorter time horizon. We exclude, for example, adjustment costs associated with changes in prices. Menu costs associated with price changes play a major role in ‘New Keynesian Economics’. We also exclude informational deficiencies which may cause wage setters or other agents to over- or underestimate the real wage. This means that we do not consider the slow adjustment of expectations which form part of the early natural unemployment hypotheses, cf. Friedman (1968) and Phelps (1968).