Recent Developments in Policy Optimisation for an Open Monetary Economy

In this paper I shall outline some of the unresolved issues in the application of optimisation techniques in the choice of monetary and fiscal policies for an open economy. Many of these issues are independent of the methods used to choose policies, but most have implications for the use of optimal control techniques. Nevertheless, there is a sense in which the great efficiency of policy selection by optimisation techniques forces many of the issues into the open. Why is this? It is because any powerful optimisation technique pushes hard at all the constraints of the system, and tests all conceivable combinations of instruments. If there is a falsely loose constraint in the system, then the optimal controls will take full advantage of it. If one group of decisionmakers in the model is falsely depicted as having static expectations, then the optimally-chosen policies will unmercifully exploit their gullibility. Pindyck (1973, pp. 151–3) has emphasised that this exploitive power of optimisation techniques makes them very useful in exposing ‘… characteristics of the model that would otherwise remain submerged’. Until this process is over, and there is no clear way of telling when it is over, the optimal control techniques must be regarded primarily as part of the model-building process rather than as an established procedure for policy selection.