ABSTRACT
This chapter examines the problems posed for macroeconomic policy design by the existence of rival models of the economy. It examines a procedure for doing this. To illustrate the proposed procedures, numerical experiments are reported which use the UK models of HM Treasury and the National Institute of Economic and Social Research, as rival models. The problem of ignoring model uncertainty in policy design is very important. Even for models which are not generally regarded as very different, like the HMT and NI models, there are substantial losses in terms of objective function cost reduction, if the real world turns out to be like the other model. In using the simultaneous optimization algorithm to derive a Pareto optimal policy, the policymaker is free to assign weights to the individual objective functions.
