Tests of Causality

The Message in the Innovations


G. William Schwert

University of Rochester, Rochester, NY 14627
and National Bureau of Economic Research


Carnegie-Rochester Conference Series on Public Policy, 10 (Spring 1979) 55-96


This paper describes new time series techniques and shows the advantages and disadvantages of these techniques compared with more traditional methods. Cross-correlation analysis of residuals from autoregressive-integrated- moving average (ARIMA) models to determine the predictive relations between two variables. I conclude that it is important to consider the power of this procedure before putting much faith in empirical results that seem to find a "lack of relationship" between economic variables.

Key words: Causality tests, ARIMA

JEL Classifications: C22


Cited 53 times in the SSCI through April 1996

© Copyright 1979, Elsevier
Return to Publications Page