Money and economic activity: An empirical investigation

New Classical monetary business cycle theories emphasize that unanticipated changes in nominal aggregates affect real output, but anticipated changes do not. Chapter one of this dissertation uses post-war time series data to test the New Classical hypothesis. The results support the predicted neutrality of anticipated nominal shocks.;New Classical models based on imperfect information (Lucas, 1972: Haubrich and King, 1991) predict that the output effect of a given nominal shock depends on the expected variance of nominal shocks. Chapter two extends the tests of chapter one to account for the output effects of a changing expected variance of nominal shocks. The empirical model specifies the expected variance as an ARCH (Autoregressive Conditional Heteroskedasticity) process. The results indicate that unanticipated nominal changes do not significantly affect output across all specifications. The estimated output effect of anticipated nominal shocks emerges as either negative or insignificantly different from zero. Hence the most naive Keynesian/Monetarist model fails to reject the New Classical misperceptions theory.;Chapter three presents a monetary theory of business fluctuations based on the work of Hayek (1933, 1975). The model predicts that only long-term monetary shocks affect output. The chapter evaluates the evidence on monetary theories using linear time series models. The Hayek model performs well relative to conventional monetary specifications, but the overall evidence does not support monetary theories.