The Sustainability of Optimal Policies: Analysis and Empirical Assessment Using the London Business School Model

Abstract

The paper addresses the Kydland and Prescott (1977) critique of the use of optimal control methods for macroeconomic policy design. The critique is that the optimal policy in models with rational expectations are time inconsistent i.e., it becomes suboptimal through the mere passage of time even in the absence of unanticipated shocks.It is argued that time inconsistency undermines the credibility of the optimal policy in the eyes of the private sector who will come to expect the policy-maker to reoptimise. Optimal policy must then be constrained to have the property of time consistency which, for many models, is a serious restriction.In Barro and Gordon (1983) a different approach to the time inconsistency problem is offered. They examine whether reputational consideration can restore credibility for policy-makers and hence avoid the inferior outcome of the time consistency constraint. they assume that policy-makers suffer a loss of reputation if they renege on earlier commitments. With this ‘punishment’ mechanism in place, Barro and Gordon show that policies superior to the time consistent policy exist which can be credible and sustainable.The Barro-Gordon analysis is model s ecific and in particular applies to static models only. In Currie and Levine (1987) and Levine 6988) a generalisation of their analysis to structurally dynamic, stochastic linear rules is provided.The paper draws on this work together with the linearisation procedure of Christodoulalus et al (1988) to investigate, empirically, the sustainability of optimal policies on the London Business School Model.