LEARNING RATIONAL-EXPECTATIONS IN A POLICY GAME

Rational expectations is a maintained assumption in the analysis of economic policy. Here we examine how two types of learning rational expectations (rational and econometric) affect the time profile of optimal policy. In both cases the government adopts policies which delay convergence to rational expectations. There is also a reduction in the inflationary bias, in one case permanently, in the other temporarily. The results also show that reputation effects will be explained by models of learning. © 1991.