Abstract
In recent years, considerable interest has been shown in economics in statistically testing whether one series causes another in the sense of Granger (1969). Most of this literature is concerned with bivariate relationships in a time series 'model free' setting, and little attempt has been made to consider the general usefulness of the Granger causality concept for economic modelling. In this paper, causality testing is examined in the context of a dynamic linear simultaneous equation econometric model, possibly with vector MA errors. Both the implications of Granger unidirectional causality and the procedure for testing are considered. The general multivariate autoregressive-moving average (ARMA) system
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
