This paper tests the exclusion of lagged growth rates of money and output from regression equations, with serially correlated disturbances, for the expected real interest rate. The authors empirical approach is an extension of the empirical strategies of Eugene F. Fama (1975) and Frederic S. Mishkin (1981)--which invoke the orthogonality of the inflation forecast error to predetermined regressors under the maintained hypothesis of rational expectations. They discuss the implications of their tests for simple real-business-cycle models. Copyright 1991 by MIT Press.

This paper analyzes a multiperiod monetary targeting procedure as a possible resolution to the credibility problem in policy when the monetary authority has some private information. By limiting the degree of flexibility permitted in policy, this procedure mitigates the credibility problem. As the length of the targeting horizon decreases, the severity of the credibility problem falls, but at the expense of weakening the monetary authority's ability to pursue its stabilization goals. Based on model simulations, the analysis studies the determinants of the optimal targeting horizon that balances the benefits of flexibility and discipline in policy.

Suppose that the government were to announce the economy will be booming in six months, and this announcement is based on false data. What effect would such an announcement have on future aggregate activity? This paper employs revisions of the series of leading economic indicators to test the hypothesis that such an announcement would have a positive effect on future activity. We find that the evidence is generally consistent with the hypothesis and that for the time period 1976–1988 the expectational shocks measured by these revisions explain over 20 percent of the fluctuation in the quarterly growth rate of industrial production.