Imperfect Information, Simplistic Modeling and the Robustness of Policy Rules

Snower, D. & Wierzbicki, A.P. (1982). Imperfect Information, Simplistic Modeling and the Robustness of Policy Rules. IIASA Working Paper. IIASA, Laxenburg, Austria: WP-82-024

Preview

Text WP-82-024.pdf Download (2MB) | Preview

Abstract

The paper presents a methodology for dealing with the problems of imperfect information or simplistic modeling in macroeconomic policy problems. The methodology permits to choose a robust policy from a given set of candidate policies--that is, a policy that makes the social welfare least sensitive to various potential modeling errors. This can be achieved even if the potential modeling errors are related to model structure or delays in model equations--without requiring that the models with more complicated structure or delays are fully solved and optimized.

The particular example chosen to illustrate the methodology is a macroeconomic model of intertemporal optimization of monetary control of inflation and unemployment. The conclusions for this particular model are two-fold. Firstly, neglected delays or other modeling errors cannot, in general, substantiate rigorously the constant monetary growth rule that is usually advanced because of such modeling inaccuracies. In fact, by choosing an appropriate feedback policy formulation it is possible to obtain reasonable results of an active policy even if the underlying model used for policy derivation is very simple and the economic reality to which the policy is applied is much more complicated. Secondly, rigorous case can be made against 'impetuous' policy making with regard to inflation and unemployment, that is, against policies that by attaching a small weight to unemployment attempt to approach rapidly long-run targets for inflation. Such a policy strategy may induce instability, either through delay effects, or by making the macroeconomic system very sensitive to other modeling errors.

Actions (login required)

View Item