# RCT Rejection Sampling for Causal Estimation Evaluation

arXiv: https://arxiv.org/abs/2307.15176 · HTML: https://mlsanity.com/paper/2307.15176/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 26 of its 67 references that are in the map.
- First posted on arXiv 27 Jul 2023.

## Builds on (most cited first, 26 of 26)

- [Do ImageNet Classifiers Generalize to ImageNet?](https://mlsanity.com/paper/1902.10811/) (2019)
- [Meta-learners for Estimating Heterogeneous Treatment Effects using Machine Learning](https://mlsanity.com/paper/1706.03461/) (2017)
- [Learning Representations for Counterfactual Inference](https://mlsanity.com/paper/1605.03661/) (2016)
- [Adapting Neural Networks for the Estimation of Treatment Effects](https://mlsanity.com/paper/1906.02120/) (2019)
- [CatBoost: gradient boosting with categorical features support](https://mlsanity.com/paper/1810.11363/) (2018)
- [Causal Inference in Natural Language Processing: Estimation, Prediction, Interpretation and Beyond](https://mlsanity.com/paper/2109.00725/) (2021)
- [Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game](https://mlsanity.com/paper/2102.13647/) (2021)
- [Automated versus do-it-yourself methods for causal inference: Lessons learned from a data analysis competition](https://mlsanity.com/paper/1707.02641/) (2017)
- [Removing Hidden Confounding by Experimental Grounding](https://mlsanity.com/paper/1810.11646/) (2018)
- [Deep Neural Networks for Estimation and Inference](https://mlsanity.com/paper/1809.09953/) (2018)
- [Text and Causal Inference: A Review of Using Text to Remove Confounding from Causal Estimates](https://mlsanity.com/paper/2005.00649/) (2020)
- [Approximate Residual Balancing: De-Biased Inference of Average Treatment Effects in High Dimensions](https://mlsanity.com/paper/1604.07125/) (2016)
- [RealCause: Realistic Causal Inference Benchmarking](https://mlsanity.com/paper/2011.15007/) (2020)
- [Benchmarking Framework for Performance-Evaluation of Causal Inference Analysis](https://mlsanity.com/paper/1802.05046/) (2018)
- [Challenges of Using Text Classifiers for Causal Inference](https://mlsanity.com/paper/1810.00956/) (2018)
- [The Case for Evaluating Causal Models Using Interventional Measures and Empirical Data](https://mlsanity.com/paper/1910.05387/) (2019)
- [Semiparametric Inference For Causal Effects In Graphical Models With Hidden Variables](https://mlsanity.com/paper/2003.12659/) (2020)
- [Validating Causal Inference Methods](https://mlsanity.com/paper/2202.04208/) (2022)
- [Adjusting for Confounders with Text: Challenges and an Empirical Evaluation Framework for Causal Inference](https://mlsanity.com/paper/2009.09961/) (2020)
- [How and Why to Use Experimental Data to Evaluate Methods for Observational Causal Inference](https://mlsanity.com/paper/2010.03051/) (2020)
- [Generating Synthetic Text Data to Evaluate Causal Inference Methods](https://mlsanity.com/paper/2102.05638/) (2021)
- [Deconfounding Scores: Feature Representations for Causal Effect Estimation with Weak Overlap](https://mlsanity.com/paper/2104.05762/) (2021)
- [On Testability of the Front-Door Model via Verma Constraints](https://mlsanity.com/paper/2203.00161/) (2022)
- [Bias and high-dimensional adjustment in observational studies of peer effects](https://mlsanity.com/paper/1706.04692/) (2017)
- [Statistical Testing under Distributional Shifts](https://mlsanity.com/paper/2105.10821/) (2021)
- [Randomized trials and their observational emulations: a framework for benchmarking and joint analysis](https://mlsanity.com/paper/2203.14857/) (2022)

## Built on by (most cited first, 7 of 7)

- [End-To-End Causal Effect Estimation from Unstructured Natural Language Data](https://mlsanity.com/paper/2407.07018/) (2024)
- [Hidden yet quantifiable: A lower bound for confounding strength using randomized trials](https://mlsanity.com/paper/2312.03871/) (2023)
- [Proximal Causal Inference With Text Data](https://mlsanity.com/paper/2401.06687/) (2024)
- [Position: Causal Machine Learning Requires Rigorous Synthetic Experiments for Broader Adoption](https://mlsanity.com/paper/2508.08883/) (2025)
- [Using Text-Based Causal Inference to Disentangle Factors Influencing Online Review Ratings](https://mlsanity.com/paper/2606.04286/) (2026)
- [Network Causal Effect Estimation In Graphical Models Of Contagion And Latent Confounding](https://mlsanity.com/paper/2411.01371/) (2024)
- [Exploring Sparse Autoencoders in Text-Based Causal Confounding Adjustment](https://mlsanity.com/paper/2609.01322/) (2026)

## Abstract (arXiv)

Confounding is a significant obstacle to unbiased estimation of causal effects from observational data. For settings with high-dimensional covariates -- such as text data, genomics, or the behavioral social sciences -- researchers have proposed methods to adjust for confounding by adapting machine learning methods to the goal of causal estimation. However, empirical evaluation of these adjustment methods has been challenging and limited. In this work, we build on a promising empirical evaluation strategy that simplifies evaluation design and uses real data: subsampling randomized controlled trials (RCTs) to create confounded observational datasets while using the average causal effects from the RCTs as ground-truth. We contribute a new sampling algorithm, which we call RCT rejection sampling, and provide theoretical guarantees that causal identification holds in the observational data to allow for valid comparisons to the ground-truth RCT. Using synthetic data, we show our algorithm indeed results in low bias when oracle estimators are evaluated on the confounded samples, which is not always the case for a previously proposed algorithm. In addition to this identification result, we highlight several finite data considerations for evaluation designers who plan to use RCT rejection sampling on their own datasets. As a proof of concept, we implement an example evaluation pipeline and walk through these finite data considerations with a novel, real-world RCT -- which we release publicly -- consisting of approximately 70k observations and text data as high-dimensional covariates. Together, these contributions build towards a broader agenda of improved empirical evaluation for causal estimation.