Testing Granger Non-Causality in Panels with Cross-Sectional Dependencies

Lenon Minorics, Caner Turkmen, David Kernert, Patrick Bloebaum, Laurent Callot, Dominik Janzing
Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, PMLR 151:10534-10554, 2022.

Abstract

This paper proposes a new approach for testing Granger non-causality on panel data. Instead of aggregating panel member statistics, we aggregate their corresponding p-values and show that the resulting p-value approximately bounds the type I error by the chosen significance level even if the panel members are dependent. We compare our approach against the most widely used Granger causality algorithm on panel data and show that our approach yields lower FDR at the same power for large sample sizes and panels with cross sectional dependencies. Finally, we examine COVID-19 data about confirmed cases and deaths measured in countries/regions worldwide and show that our approach is able to discover the true causal relation between confirmed cases and deaths while state-of-the-art approaches fail.

Cite this Paper

BibTeX
@InProceedings{pmlr-v151-minorics22a, title = { Testing Granger Non-Causality in Panels with Cross-Sectional Dependencies }, author = {Minorics, Lenon and Turkmen, Caner and Kernert, David and Bloebaum, Patrick and Callot, Laurent and Janzing, Dominik}, booktitle = {Proceedings of The 25th International Conference on Artificial Intelligence and Statistics}, pages = {10534--10554}, year = {2022}, editor = {Camps-Valls, Gustau and Ruiz, Francisco J. R. and Valera, Isabel}, volume = {151}, series = {Proceedings of Machine Learning Research}, month = {28--30 Mar}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v151/minorics22a/minorics22a.pdf}, url = {https://proceedings.mlr.press/v151/minorics22a.html}, abstract = { This paper proposes a new approach for testing Granger non-causality on panel data. Instead of aggregating panel member statistics, we aggregate their corresponding p-values and show that the resulting p-value approximately bounds the type I error by the chosen significance level even if the panel members are dependent. We compare our approach against the most widely used Granger causality algorithm on panel data and show that our approach yields lower FDR at the same power for large sample sizes and panels with cross sectional dependencies. Finally, we examine COVID-19 data about confirmed cases and deaths measured in countries/regions worldwide and show that our approach is able to discover the true causal relation between confirmed cases and deaths while state-of-the-art approaches fail. } }
Endnote
%0 Conference Paper %T Testing Granger Non-Causality in Panels with Cross-Sectional Dependencies %A Lenon Minorics %A Caner Turkmen %A David Kernert %A Patrick Bloebaum %A Laurent Callot %A Dominik Janzing %B Proceedings of The 25th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2022 %E Gustau Camps-Valls %E Francisco J. R. Ruiz %E Isabel Valera %F pmlr-v151-minorics22a %I PMLR %P 10534--10554 %U https://proceedings.mlr.press/v151/minorics22a.html %V 151 %X This paper proposes a new approach for testing Granger non-causality on panel data. Instead of aggregating panel member statistics, we aggregate their corresponding p-values and show that the resulting p-value approximately bounds the type I error by the chosen significance level even if the panel members are dependent. We compare our approach against the most widely used Granger causality algorithm on panel data and show that our approach yields lower FDR at the same power for large sample sizes and panels with cross sectional dependencies. Finally, we examine COVID-19 data about confirmed cases and deaths measured in countries/regions worldwide and show that our approach is able to discover the true causal relation between confirmed cases and deaths while state-of-the-art approaches fail.
APA
Minorics, L., Turkmen, C., Kernert, D., Bloebaum, P., Callot, L. & Janzing, D.. (2022). Testing Granger Non-Causality in Panels with Cross-Sectional Dependencies . Proceedings of The 25th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 151:10534-10554 Available from https://proceedings.mlr.press/v151/minorics22a.html.

Related Material