EconBase
← All papers

Doubly-Valid/Doubly-Sharp Sensitivity Analysis for Causal Inference with Unmeasured Confounding

Jacob Dorn, Kevin Guo, Nathan Kallus

arXiv 21 Dec 2021 · Statistics — Methodology · publishedJournal of the American Statistical Association (2024) · 4 citations (OpenAlex)

arXiv:2112.11449 · PDF · DOI · OpenAlex · Extracted main text

Abstract

We consider the problem of constructing bounds on the average treatment effect (ATE) when unmeasured confounders exist but have bounded influence. Specifically, we assume that omitted confounders could not change the odds of treatment for any unit by more than a fixed factor. We derive the sharp partial identification bounds implied by this assumption by leveraging distributionally robust optimization, and we propose estimators of these bounds with several novel robustness properties. The first is double sharpness: our estimators consistently estimate the sharp ATE bounds when one of two nuisance parameters is misspecified and achieve semiparametric efficiency when all nuisance parameters are suitably consistent. The second is double validity: even when most nuisance parameters are misspecified, our estimators still provide valid but possibly conservative bounds for the ATE and our Wald confidence intervals remain valid even when our estimators are not asymptotically normal. As a result, our estimators provide a highly credible method for sensitivity analysis of causal inferences.

Citation extraction

73
references
134
in-text mentions
73
distinct cited
8
self-citations
9,250
main-text words

appendix boundary found by appendix_command · 39% of the source is main text. Read the extracted text to check this.

Most heavily cited references

The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.

ReferenceIntensityMentionsSectionsMain text
1Zhao, Q., D. S. Small, and B. B. Bhattacharya (2019) Sensitivity analysis for inverse probability weighting estimators via the percentile bootstrap1.00093100%
2Tan, Z (2006) A distributional approach for causal inference using propensity scores1.00064100%
3Rockafellar, R. T. and S. Uryasev (2000) Optimization of conditional value-at-risk0.9285380%
4Athey, S., J. Tibshirani, and S. Wager (2019, 04) (2019) Generalized random forests0.92843100%
5Dorn, J. and K. Guo (2022) Sharp sensitivity analysis for inverse propensity weighting via quantile balancing self0.90912675%
6Yadlowsky, S., H. Namkoong, S. Basu, J. C. Duchi, and L. Tian (2022) Bounds on the conditional and average treatment effect with unobserved confounding factors0.8434375%
7Chernozhukov, V., D. Chetverikov, M. Demirer, E. Duflo, C. Hansen, W… (2018) Double/debiased machine learning for treatment and structural parameters0.81142100%
8Jin, Y., Z. Ren, and E. J. Candès (2021) Sensitivity analysis of individual treatment effects: A robust conformal inference approach0.81142100%
9Belloni, A. and V. Chernozhukov (2011) $_1$-penalized quantile regression in high-dimensional sparse models0.73732100%
10Soriano, D., E. Ben-Michael, P. J. Bickel, A. Feller, and S. D. Pime… (2021) Interpretable sensitivity analysis for balancing weights0.73732100%

Showing the top 10 of 73 scored citations.

Cited by, within the corpus

arXiv econ.EM papers that cite this one, ranked by how heavily they lean on it.

Citing paperIntensityMentionsSections
1A General Approach to Relaxing Unconfoundedness0.92844
2Sensitivity Analysis for Linear Estimators0.84344
3Policy Learning under Unobserved Confounding: A Robust and Efficient Approach0.84343
4A Sensitivity Analysis of the Surrogate Index Approach for Estimating Long-Term Treatment Effects0.82293
5Treatment Effect Risk: Bounds and Inference0.64422
6Partial Identification of Causal Effects for Endogenous Continuous Treatments0.64422
7What's the Harm? Sharp Bounds on the Fraction Negatively Affected by Treatment0.51121
8Conformalized Lee Inference: Distribution-Free Individual Treatment Effect Intervals under Monotone Sample Selection0.51121
9Inference on Strongly Identified Functionals of Weakly Identified Functions0.40511
10Robust Design and Evaluation of Predictive Algorithms under Unobserved Confounding0.40511