Quasi-oracle estimation of heterogeneous treatment effects

Flexible estimation of heterogeneous treatment effects lies at the heart of many statistical applications, such as personalized medicine and optimal resource allocation. In this article we develop a general class of two-step algorithms for heterogeneous treatment effect estimation in observational studies. First, we estimate marginal effects and treatment propensities to form an objective function that isolates the causal component of the signal. Then, we optimize this data-adaptive objective function. The proposed approach has several advantages over existing methods. From a practical perspective, our method is flexible and easy to use: in both steps, any loss-minimization method can be employed, such as penalized regression, deep neural networks, or boosting; moreover, these methods can be fine-tuned by cross-validation. Meanwhile, in the case of penalized kernel regression, we show that our method has a quasi-oracle property. Even when the pilot estimates for marginal effects and treatment propensities are not particularly accurate, we achieve the same error bounds as an oracle with prior knowledge of these two nuisance components. We implement variants of our approach based on penalized regression, kernel ridge regression, and boosting in a variety of simulation set-ups, and observe promising performance relative to existing baselines.

• Publication year

  2017

• Venue

  Biometrika

• Publication date

  2017-12-13

• Fields of study

  Medicine, Mathematics, Economics, Computer Science

• Identifiers
  DOI 10.1093/biomet/asaa076arXiv 1712.04912
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Cross-Validation, Risk Estimation, and Model Selection: Comment on a Paper by Rosset and Tibshirani

  2020cited by this paper
• Bayesian Regression Tree Models for Causal Inference: Regularization, Confounding, and Heterogeneous Effects (with Discussion)

  2020cited by this paper
• Multivariate Adaptive Regression Splines

  2019cited by this paper
• Some methods for heterogeneous treatment effect estimation in high dimensions

  2017influential reference
• Double/debiased machine learning for treatment and structural parameters

  2017cited by this paper
• Efficient Policy Learning

  2017influential reference
• Kernel-Based Regularized Least Squares in R (KRLS) and Stata (krls)

  2017cited by this paper
• Randomized Experiments

  2017cited by this paper
• HIGHER ORDER ESTIMATING EQUATIONS FOR HIGH-DIMENSIONAL MODELS.

  2017cited by this paper
• Selective inference for effect modification via the lasso

  2017cited by this paper
• Beyond prediction: Using big data for policy problems

  2017influential reference
• Metalearners for estimating heterogeneous treatment effects using machine learning

  2017cited by this paper
• Automated versus Do-It-Yourself Methods for Causal Inference: Lessons Learned from a Data Analysis Competition

  2017cited by this paper
• Bayesian regression tree models for causal inference: regularization, confounding, and heterogeneous effects

  2017influential reference
• Generalized random forests

  2016cited by this paper
• Optimal Individualized Treatments in Resource-Limited Settings

  2016cited by this paper
• Estimating individual treatment effect: generalization bounds and algorithms

  2016cited by this paper
• Decomposing Treatment Effect Variation

  2016cited by this paper
• XGBoost: A Scalable Tree Boosting System

  2016cited by this paper
• TensorFlow: A system for large-scale machine learning

  2016cited by this paper
• Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction

  2016cited by this paper
• Super-Learning of an Optimal Dynamic Treatment Rule

  2016cited by this paper
• MINIMAX ESTIMATION OF A FUNCTIONAL ON A STRUCTURED

  2016cited by this paper
• Predicting the Future - Big Data, Machine Learning, and Clinical Medicine.

  2016cited by this paper
• High-dimensional regression adjustments in randomized experiments

  2016cited by this paper
• STATISTICAL INFERENCE FOR THE MEAN OUTCOME UNDER A POSSIBLY NON-UNIQUE OPTIMAL TREATMENT STRATEGY.

  2016cited by this paper
• Tree-based methods for individualized treatment regimes.

  2015cited by this paper
• Estimation and Inference of Heterogeneous Treatment Effects using Random Forests

  2015influential reference
• Who should be Treated? Empirical Welfare Maximization Methods for Treatment Choice

  2015cited by this paper
• Recursive partitioning for heterogeneous causal effects

  2015influential reference
• Estimating treatment effect heterogeneity in randomized program evaluation

  2013influential reference
• Program evaluation and causal inference with high-dimensional data

  2013influential reference
• Estimating Individualized Treatment Rules Using Outcome Weighted Learning

  2012cited by this paper
• A Simple Method for Detecting Interactions between a Treatment and a Large Number of Covariates

  2012cited by this paper
• Estimating optimal treatment regimes from a classification perspective

  2012cited by this paper
• Statistical Applications in Genetics and Molecular Biology Super Learner

  2011cited by this paper
• Doubly Robust Policy Evaluation and Learning

  2011cited by this paper
• The International Journal of Biostatistics Targeted Maximum Likelihood Learning

  2011cited by this paper
• Bayesian Nonparametric Modeling for Causal Inference

  2011influential reference
• Targeted Learning: Causal Inference for Observational and Experimental Data

  2011cited by this paper
• Regularization in kernel learning

  2010influential reference
• Regularization Paths for Generalized Linear Models via Coordinate Descent.

  2010cited by this paper
• Optimal Rates for Regularized Least Squares Regression

  2009cited by this paper
• Subgroup Analysis via Recursive Partitioning

  2009cited by this paper
• Asymptotics for Statistical Treatment Rules

  2009cited by this paper
• Support vector machines

  2008cited by this paper
• FAST RATES FOR ESTIMATION ERROR AND ORACLE INEQUALITIES FOR MODEL SELECTION

  2008influential reference
• Optimal Rates for the Regularized Least-Squares Algorithm

  2007cited by this paper
• CONSISTENCY OF CROSS VALIDATION FOR COMPARING REGRESSION PROCEDURES

  2007cited by this paper
• Comparing Experimental and Matching Methods Using a Large-Scale Voter Mobilization Experiment

  2006influential reference
• Semiparametric Theory and Missing Data

  2006cited by this paper
• The Cross-Validated Adaptive Epsilon-Net Estimator

  2006cited by this paper
• Empirical minimization

  2006cited by this paper
• The Generic Chaining

  2005cited by this paper
• A Generalization Error for Q-Learning

  2005cited by this paper
• Optimal Structural Nested Models for Optimal Sequential Decisions

  2004influential reference
• Stacked regressions

  2004cited by this paper
• Unified Cross-Validation Methodology For Selection Among Estimators and a General Cross-Validated Adaptive Epsilon-Net Estimator: Finite Sample Oracle Inequalities and Examples

  2003cited by this paper
• Statistical treatment rules for heterogeneous populations

  2003cited by this paper
• ESTIMATING THE APPROXIMATION ERROR IN LEARNING THEORY

  2003cited by this paper
• On the mathematical foundations of learning

  2001cited by this paper
• About the constants in Talagrand's concentration inequalities for empirical processes

  2000cited by this paper
• Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models

  1999cited by this paper
• Efficient and adaptive estimation for semiparametric models

  1998cited by this paper
• New concentration inequalities in product spaces

  1996cited by this paper
• Regression Shrinkage and Selection via the Lasso

  1996cited by this paper
• Semiparametric Efficiency in Multivariate Regression Models with Missing Data

  1995cited by this paper
• The asymptotic variance of semiparametric estimators

  1994cited by this paper
• Efficient and Adaptive Estimation for Semiparametric Models.

  1994cited by this paper
• Sharper Bounds for Gaussian and Empirical Processes

  1994influential reference
• Stacked generalization

  1992cited by this paper
• ROOT-N-CONSISTENT SEMIPARAMETRIC REGRESSION

  1988influential reference
• On Asymptotically Efficient Estimation in Semiparametric Models

  1986cited by this paper
• The central role of the propensity score in observational studies for causal effects

  1983cited by this paper
• Estimating causal effects of treatments in randomized and nonrandomized studies.

  1974cited by this paper

• SpatialCausal : a spatially-aware causal inference deep learning model for out-of-hospital cardiac arrest survival prediction

  2026cites this paper
• Integrating Unstructured Text into Causal Inference: Empirical Evidence from Real Data

  2026cites this paper
• How Business Agglomeration Affects Individual Points-of-Interest: A Causal Effect Estimation Perspective

  2026cites this paper
• Multi-treatment uplift evaluation on non-random assignment biased data

  2026cites this paper
• Preference-based Conditional Treatment Effects and Policy Learning

  2026cites this paper
• Learning from Biased and Costly Data Sources: Minimax-optimal Data Collection under a Budget

  2026cites this paper
• Estimation of the interpretable heterogeneous treatment effect with causal subgroup discovery in survival outcomes

  2026cites this paper
• Using Meta-Learners and Propensity Score Matching to Optimize Customer Retention

  2026cites this paper
• Causal Effect Estimation with Latent Textual Treatments

  2026cites this paper
• Direct Doubly Robust Estimation of Conditional Quantile Contrasts

  2026cites this paper
• B&S2Vec: Mapping market structure in two-sided platform based on consumers' purchase trajectories

  2026cites this paper
• Policy learning under constraint: Maximizing a primary outcome while controlling an adverse event

  2026cites this paper
• Optimal and Structure-Adaptive CATE Estimation with Kernel Ridge Regression

  2026cites this paper
• Generalized Bayes for Causal Inference

  2026influential citation
• Effects of sustainable soil fertility management practices on maize income in rural Zambia: a double machine learning approach

  2026cites this paper
• Bayesian causal forests with covariate balancing propensity score: a novel approach for heterogeneous treatment effects estimation

  2026cites this paper
• Measures for Assessing Causal Effect Heterogeneity Unexplained by Covariates

  2026cites this paper
• Beyond the Numbers: Causal Effects of Financial Report Sentiment on Bank Profitability

  2026cites this paper
• Observationally Informed Adaptive Causal Experimental Design

  2026cites this paper
• Strategic Governance for Sustainability: The Role of Internal Controls in Shaping Financial Performance Through ESG in Emerging Markets

  2026cites this paper
• Targeted Synthetic Control Method

  2026cites this paper
• Enhanced inference for distributions and quantiles of individual treatment effects in various experiments

  2026cites this paper
• Estimating optimal interpretable individualized treatment regimes from a classification perspective using adaptive LASSO

  2026influential citation
• Evaluating the long-term urban effects of high-speed rail in Japan: An integrated approach using synthetic difference-in-differences and double/debiased machine learning

  2026cites this paper
• From Emotion to Action: AURORA’s Sentiment-Based Model for Tourism Decisions

  2026influential citation
• Robust X-Learner: Breaking the Curse of Imbalance and Heavy Tails via Robust Cross-Imputation

  2026cites this paper
• Causal Identification in Multi-Task Demand Learning with Confounding

  2026cites this paper
• Joint Graph Learning for Robust Causal Inference over Knowledge Graphs

  2026cites this paper
• Detecting and Explaining Unlawful Insider Trading: A Shapley Value and Causal Forest Approach to Identifying Key Drivers and Causal Relationships

  2026cites this paper
• Bayesian Supervised Causal Clustering

  2026cites this paper
• Good Allocations from Bad Estimates

  2026cites this paper
• A transformer-enhanced causal analysis framework for inter-stop energy consumption of electric bus

  2026cites this paper
• When Rail Matters Most: Policy Impacts in China's Freight "Modal Shift from Road to Railway"

  2026cites this paper
• Causal Inference with Complex Treatments: A Survey

  2026cites this paper
• Rank-Learner: Orthogonal Ranking of Treatment Effects

  2026cites this paper
• Heterogeneous scars in later life: The economic impact of early labour market opportunities

  2026cites this paper
• Precision Mental Health: Predicting Heterogeneous Treatment Effects for Depression through Data Integration.

  2025cites this paper
• Optimal averaging estimator of heterogeneous treatment effects for single-index models

  2025cites this paper
• Causal Clustering for Conditional Average Treatment Effects Estimation and Subgroup Discovery

  2025influential citation
• CTRL Your Shift: Clustered Transfer Residual Learning for Many Small Datasets

  2025cites this paper
• Digital lead generation platforms: Rightsizing the seller base

  2025cites this paper
• Counterfactual Probabilistic Diffusion with Expert Models

  2025cites this paper
• Decision Theoretic Subgroup Detection With Bayesian Machine Learning

  2025cites this paper
• Federated causal inference from observational data

  2025cites this paper
• Bridging Structural Causal Inference and Machine Learning: The S-DIDML Estimator for Heterogeneous Treatment Effects

  2025cites this paper
• Learning Treatment Policies From Multimodal Electronic Health Records

  2025cites this paper
• Optimal Targeting in Dynamic Systems

  2025cites this paper
• Shrinkage-Based Regressions with Many Related Treatments

  2025cites this paper
• Estimating the Local Average Treatment Effect Without the Exclusion Restriction

  2025cites this paper
• LILI clustering algorithm: Limit Inferior Leaf Interval Integrated into Causal Forest for Causal Interference

  2025cites this paper
• Cross-World Assumption and Refining Prediction Intervals for Individual Treatment Effects

  2025cites this paper
• The Critical Challenge of using Large-Scale Digital Experiment Platforms for Scientific Discovery

  2025cites this paper
• An Algorithm for Identifying Interpretable Subgroups With Elevated Treatment Effects

  2025cites this paper
• Direct Effect Modeling Methods to Estimate Heterogeneity of Treatment Effects.

  2025cites this paper
• Optimal Transport Learning: Balancing Value Optimization and Fairness in Individualized Treatment Rules

  2025influential citation
• Hybrid causal inference framework for long-chain drought propagation networks: Combining convergent cross-mapping and generalized random forest

  2025cites this paper
• GAS: Large-Scale Heterogeneous Personalization in Social Network Applications at Meta

  2025cites this paper
• SACO: Sequence-Aware Constrained Optimization Framework for Coupon Distribution in E-commerce

  2025cites this paper
• Heterogeneous Causal Effect Analysis and Application Based on Meta-Learners

  2025influential citation
• Matching-Based Nonparametric Estimation of Group Average Treatment Effects

  2025cites this paper
• Spatial Heterogeneity in Temperature Elasticity of Agricultural Economic Production in Xinjiang Province, China

  2025cites this paper
• Estimating Heterogeneous Treatment Effects of Building Energy Efficiency Retrofits Using Machine Learning

  2025cites this paper
• Consistent Labeling Across Group Assignments: Variance Reduction in Conditional Average Treatment Effect Estimation

  2025cites this paper
• LLM-Driven Treatment Effect Estimation Under Inference Time Text Confounding

  2025cites this paper
• Late Fusion Multi-task Learning for Semiparametric Inference with Nuisance Parameters

  2025cites this paper
• Clinical efficacy evaluation method using dominance-based rough sets and temporal hybrid information systems

  2025cites this paper
• Double Machine Learning for Estimating Time-Varying Delayed and Instantaneous Effects Using Digital Phenotypes

  2025cites this paper
• Foundation Models for Causal Inference via Prior-Data Fitted Networks

  2025cites this paper
• A Diffusion-Based Method for Learning the Multi-Outcome Distribution of Medical Treatments

  2025cites this paper
• Causalteshap: discerning predictive from prognostic features for treatment effect analysis

  2025cites this paper
• Data-driven learning optimal K values for K-nearest neighbour matching in causal inference

  2025cites this paper
• TabPFN: One Model to Rule Them All?

  2025cites this paper
• Preventing Spurious Interactions: A New Inductive Bias for Accurate Treatment Effect Estimation

  2025cites this paper
• Generalizable estimation of conditional average treatment effects using Causal Forest in randomized controlled trials

  2025cites this paper
• Machine Learning for Econometrics

  2025cites this paper
• Source control or end-of-pipe treatment: How green finance policy impacts enterprise carbon intensity

  2025cites this paper
• Treatment Effect Estimation for Optimal Decision-Making

  2025cites this paper
• Adaptive sample splitting for randomization tests

  2025cites this paper
• Selective randomization inference for subgroup effects with continuous biomarkers

  2025cites this paper
• Statistical Learning for Heterogeneous Treatment Effects: Pretraining, Prognosis, and Prediction

  2025influential citation
• Dynamic Marketing Uplift Modeling: A Symmetry-Preserving Framework Integrating Causal Forests with Deep Reinforcement Learning for Personalized Intervention Strategies

  2025cites this paper
• A conceptual synthesis of causal assumptions for causal discovery and inference

  2025cites this paper
• Federated Causal Inference in Healthcare: Methods, Challenges, and Applications

  2025cites this paper
• A heterogeneity analysis of health-related quality of life in early adults born very preterm or very low birthweight across the sociodemographic spectrum.

  2025cites this paper
• Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data

  2025cites this paper
• Methods for Combining Observational and Experimental Causal Estimates: A Review

  2025cites this paper
• M-learner:A Flexible And Powerful Framework To Study Heterogeneous Treatment Effect In Mediation Model

  2025cites this paper
• A Distributionally-Robust Framework for Nuisance in Causal Effect Estimation

  2025cites this paper
• Heterogeneity in the association between internet use and dementia among older adults: A machine-learning analysis.

  2025cites this paper
• Precision ecology for targeted conservation action.

  2025cites this paper
• Estimating Misreporting in the Presence of Genuine Modification: A Causal Perspective

  2025cites this paper
• Doubly Robust Alignment for Large Language Models

  2025cites this paper
• On Efficient Estimation of Distributional Treatment Effects under Covariate-Adaptive Randomization

  2025cites this paper
• Do-PFN: In-Context Learning for Causal Effect Estimation

  2025cites this paper
• Semiparametric inference based on adaptively collected data

  2025cites this paper
• Understanding the causal effects and heterogeneity between metabolic syndrome and lung function: a nationwide prospective cohort study in China

  2025cites this paper
• Heterogeneous Causal Learning for Optimizing Aggregated Functions in User Growth

  2025influential citation
• Robust estimation of heterogeneous treatment effects in randomized trials leveraging external data

  2025influential citation
• Deep Learning of Continuous and Structured Policies for Aggregated Heterogeneous Treatment Effects

  2025cites this paper
• Dynamic Regularized CBDT: Variance-Calibrated Causal Boosting for Interpretable Heterogeneous Treatment Effects

  2025cites this paper