Causality: Some New Thoughts on an Old Topic

No 351, Econometric Society 2004 Australasian Meetings from Econometric Society

Abstract: Traditional "Granger-Causality" (henceforth just G-causality) concerned the conditional mean. It required that the causal variable Yt preceded the causal variable Xt+1 in time and also that Yt contained special information about Xt+1 which would be shown in the conditional mean E[Xt+1|Yt]. There is an immediate forecasting implication. Later, in terms of conditional distributions, Yt did not cause Xt=1 in distributions, if Yt was conditionally independent of Xt+1. Some new implications of this definition will be presented and the links between the distributions in mean and distribution explored. Since the appearance of these definitions a number of alternative forms have appeared, due to Hoover, Pearl, White, and others and they will be discussed and compared. There will be no conclusion

Keywords: Causality (search for similar items in EconPapers)
JEL-codes: C53 (search for similar items in EconPapers)
Date: 2004-08-11

There are no downloads for this item, see the EconPapers FAQ for hints about obtaining it.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Access Statistics for this paper

More papers in Econometric Society 2004 Australasian Meetings from Econometric Society
Contact information at EDIRC.
Series data maintained by Christopher F. Baum ().