# Rescuing double robustness: safe estimation under complete misspecification

arXiv: https://arxiv.org/abs/2509.22446 · HTML: https://mlsanity.com/paper/2509.22446/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 21 of its 86 references that are in the map.
- First posted on arXiv 26 Sep 2025.

## Builds on (most cited first, 21 of 21)

- [Prediction-Powered Inference](https://mlsanity.com/paper/2301.09633/) (2023)
- [Semiparametric doubly robust targeted double machine learning: a review](https://mlsanity.com/paper/2203.06469/) (2022)
- [Stability](https://mlsanity.com/paper/1310.0150/) (2013)
- [PPI++: Efficient Prediction-Powered Inference](https://mlsanity.com/paper/2311.01453/) (2023)
- [Veridical Data Science](https://mlsanity.com/paper/1901.08152/) (2019)
- [Cross-Prediction-Powered Inference](https://mlsanity.com/paper/2309.16598/) (2023)
- [Semi-supervised Inference: General Theory and Estimation of Means](https://mlsanity.com/paper/1606.07268/) (2016)
- [Efficient and Adaptive Linear Regression in Semi-Supervised Settings](https://mlsanity.com/paper/1701.04889/) (2017)
- [Sharp instruments for classifying compliers and generalizing causal effects](https://mlsanity.com/paper/1801.03635/) (2018)
- [A Unified Framework for Semiparametrically Efficient Semi-Supervised Learning](https://mlsanity.com/paper/2502.17741/) (2025)
- [Optimal and Safe Estimation for High-Dimensional Semi-Supervised Learning](https://mlsanity.com/paper/2011.14185/) (2020)
- [The s-value: evaluating stability with respect to distributional shifts](https://mlsanity.com/paper/2105.03067/) (2021)
- [Off-policy estimation of linear functionals: Non-asymptotic theory for semi-parametric efficiency](https://mlsanity.com/paper/2209.13075/) (2022)
- [PCS-UQ: Uncertainty Quantification via the Predictability-Computability-Stability Framework](https://mlsanity.com/paper/2505.08784/) (2025)
- [Challenges of the inconsistency regime: Novel debiasing methods for missing data models](https://mlsanity.com/paper/2309.01362/) (2023)
- [Semi-Supervised U-statistics](https://mlsanity.com/paper/2402.18921/) (2024)
- [PCS Workflow for Veridical Data Science in the Age of AI](https://mlsanity.com/paper/2508.00835/) (2025)
- [Doubly Robust Estimators with Weak Overlap](https://mlsanity.com/paper/2304.08974/) (2023)
- [Augmented Doubly Robust Post-Imputation Inference for Proteomic Data](https://mlsanity.com/paper/2403.15802/) (2024)
- [C-Learner: Constrained Learning for Causal Inference](https://mlsanity.com/paper/2405.09493/) (2024)
- [Benchmarking Estimators for Natural Experiments: A Novel Dataset and a Doubly Robust Algorithm](https://mlsanity.com/paper/2409.04500/) (2024)

## Abstract (arXiv)

Double robustness is a major selling point of semiparametric and missing data methodology. Its virtues lie in protection against partial nuisance misspecification and asymptotic semiparametric efficiency under correct nuisance specification. However, in many applications, complete nuisance misspecification should be regarded as the norm (or at the very least the expected default), and thus doubly robust estimators may behave fragilely. In fact, it has been amply verified empirically that these estimators can perform poorly when all nuisance functions are misspecified. Here, we first characterize this phenomenon of double fragility, and then propose a solution based on adaptive correction clipping (DR+ACC). We argue that our DR+ACC proposal is safe, in that it inherits the favorable properties of doubly robust estimators under correct nuisance specification, but its error is guaranteed to be bounded by a convex combination of the individual nuisance model errors, which prevents the instability caused by the compounding product of errors of doubly robust estimators. We also show that our proposal comes with no reduction in semiparametric efficiency compared to doubly robust estimators, and thus valid inference based on asymptotic normality can be conducted when nuisances are well-specified. We showcase the efficacy of our DR+ACC estimator both through extensive simulations and by applying it to the analysis of Alzheimer's disease proteomics data.