$ε$-Identifiability of Causal Quantities

Ang Li, Scott Mueller, Xin Shu, Judea Pearl
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:4978-4986, 2026.

Abstract

Identifying the effects of causes and causes of effects is vital in virtually every scientific field. Often, however, the needed probabilities may not be fully identifiable from the available data sources. This paper shows how approximate identifiability is still possible for several probabilities of causation. We term this $\epsilon\text{-identifiability}$ and demonstrate its usefulness in cases where the behavior of certain subpopulations can be restricted within sufficiently narrow bounds. In particular, we show how unidentifiable causal effects and counterfactual probabilities can be $\epsilon\text{-identified}$ when such allowances are made. Often, these allowances are easily measured and reasonably assumed. Finally, $\epsilon\text{-identifiability}$ is applied to the unit selection problem.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-li26j, title = { $ε$-Identifiability of Causal Quantities }, author = {Li, Ang and Mueller, Scott and Shu, Xin and Pearl, Judea}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {4978--4986}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/li26j/li26j.pdf}, url = {https://proceedings.mlr.press/v300/li26j.html}, abstract = { Identifying the effects of causes and causes of effects is vital in virtually every scientific field. Often, however, the needed probabilities may not be fully identifiable from the available data sources. This paper shows how approximate identifiability is still possible for several probabilities of causation. We term this $\epsilon\text{-identifiability}$ and demonstrate its usefulness in cases where the behavior of certain subpopulations can be restricted within sufficiently narrow bounds. In particular, we show how unidentifiable causal effects and counterfactual probabilities can be $\epsilon\text{-identified}$ when such allowances are made. Often, these allowances are easily measured and reasonably assumed. Finally, $\epsilon\text{-identifiability}$ is applied to the unit selection problem. } }
Endnote
%0 Conference Paper %T $ε$-Identifiability of Causal Quantities %A Ang Li %A Scott Mueller %A Xin Shu %A Judea Pearl %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-li26j %I PMLR %P 4978--4986 %U https://proceedings.mlr.press/v300/li26j.html %V 300 %X Identifying the effects of causes and causes of effects is vital in virtually every scientific field. Often, however, the needed probabilities may not be fully identifiable from the available data sources. This paper shows how approximate identifiability is still possible for several probabilities of causation. We term this $\epsilon\text{-identifiability}$ and demonstrate its usefulness in cases where the behavior of certain subpopulations can be restricted within sufficiently narrow bounds. In particular, we show how unidentifiable causal effects and counterfactual probabilities can be $\epsilon\text{-identified}$ when such allowances are made. Often, these allowances are easily measured and reasonably assumed. Finally, $\epsilon\text{-identifiability}$ is applied to the unit selection problem.
APA
Li, A., Mueller, S., Shu, X. & Pearl, J.. (2026). $ε$-Identifiability of Causal Quantities . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:4978-4986 Available from https://proceedings.mlr.press/v300/li26j.html.

Related Material