Abstract
A proposal is outlined for reconciling the use of many indicators or instruments in setting monetary policy, with the need for clarity and openness. The proposal draws on the literature of unobserved component models (Stock and Watson, 1991), constructing the unobserved component using the Kalman Filter, from a set of unobservable series. The technique used allows for cointegration between variables and for the appropriate specification of stochastic trends (Garratt, Hall and Henry, 1992). In the empirical application a single measure of the tightness of monetary policy is derived from a range of observable indicators based on monthly UK data for the period 1978–92.