A Cross-Country Test of the Natural Rate Hypothesis

This paper provides some evidence concerning the applicability of the Natural Rate Hypothesis (NRH) to the determination of Canadian real GNP. Lucas (1973) and Barro (1976) develop the proposition that anticipated money supply shocks will not affect real output or employment. Moreover, they show that unanticipated money supply shocks can affect real economic variables in that they cause agents to confuse absolute and relative price changes. Barro (1977, 1978) has tested the NRH by estimating a forecasting model of the money supply and using it to decompose observed money supply growth into its anticipated and unanticipated components. Forecasted money becomes anticipated money and the residuals become unanticipated money; these components are then used as explanatory variables in regression models of selected real economic variables. The money neutrality hypothesis is tested by applying classical testing procedures to the null hypothesis that the coefficients on current and lagged anticipated money supply growth are jointly equal to zero in these models. Barro was unable, on this basis, to reject the NRE for the U.S.