Forecasting government policy: An example of the importance of time inconsistency

Abstract

This paper examines the importance of the concept of time inconsistency for the forecasting of government policy. It is argued that practising macroeconomic forecasters should assume that policy-makers are attempting to maximise an objective function and that this type of behaviour can be captured using optimal control techniques. The analogy is drawn with the conventional assumption of rational expectations in forecasting private sector behaviour. Given this approach, the crucial role of the assumptions regarding the time consistency of policy makers' behaviour is emphasised. Empirical examples using the National Institute macroeconomic model of the U.K. economy are presented to illustrate the argument.