Dealing with limited overlap in estimation of average treatment effects
2009/01/24 by R. K. Crump, Richard K. Crump, V. Joseph Hotz +5 · 997 citations
Mathematics · #Advanced Causal Inference Techniques #Computer science #Economics #Estimation #History #Library science #Management #Miami #Statistical Methods and Inference #Statistical Methods in Clinical Trials
paper · open access · doi:10.1093/biomet/asn055
published in Biometrika 96(1), 187-199 (Oxford University Press)
openalex publication_date 2009/01/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/04
Abstract
Estimation of average treatment effects under unconfounded or ignorable treatment assignment is often hampered by lack of overlap in the covariate distributions between treatment groups. This lack of overlap can lead to imprecise estimates, and can make commonly used estimators sensitive to the choice of specification. In such cases researchers have often used ad hoc methods for trimming the sample. We develop a systematic approach to addressing lack of overlap. We characterize optimal subsamples for which the average treatment effect can be estimated most precisely. Under some conditions, the optimal selection rules depend solely on the propensity score. For a wide range of distributions, a good approximation to the optimal rule is provided by the simple rule of thumb to discard all units with estimated propensity scores outside the range [0.1,0.9].
Citations
Cited by
- Treatment Effects in the Presence of Unmeasured Confounding: Dealing With Observations in the Tails of the Propensity Score Distribution--A Simulation Study
- Propensity Score Matching in Accounting Research
- Propensity Score Weighting and Trimming Strategies for Reducing Variance and Bias of Treatment Effect Estimates: A Simulation Study
- Estimating treatment effects with machine learning
- Recent Developments in the Econometrics of Program Evaluation
- Potential Outcome and Directed Acyclic Graph Approaches to Causality: Relevance for Empirical Practice in Economics
- A Human-Augmenting Agentic Workflow for Observational Causal Inference
- Contamination Bias in Linear Regressions
- Start-up subsidies for the unemployed: Long-term evidence and effect heterogeneity
- Noise-induced randomization in regression discontinuity designs
- LLM-Extracted Covariates for Clinical Causal Inference: Rethinking Integration Strategies
- Data-Poisoning Audits for Causal Effect Estimation
- Generative Synthetic Data for Causal Inference: Pitfalls, Remedies, and Opportunities
- Posterior Predictive Treatment Assignment Methods for Causal Inference in the Context of Time-Varying Treatments
- Can social assistance reduce violent conflict and civil unrest? Evidence from a large-scale public works programme in Ethiopia
- Balanced Linear Contextual Bandits
- Automated versus Do-It-Yourself Methods for Causal Inference: Lessons Learned from a Data Analysis Competition
- Causal Inference: A Missing Data Perspective
- A Survey on Causal Inference
- A Sensitivity Approach to Causal Inference Under Limited Overlap
- Nonparametric Heterogeneous Treatment Effect Estimation in Repeated Cross Sectional Designs
- Propensity Score Weighting for Causal Subgroup Analysis
- The Confounder Trap: Treatment-Encoding Representations in Causal Inference with Text
- Matching Methods for Confounder Adjustment: An Addition to the Epidemiologist’s Toolbox
- The Balance‐Sample Size Frontier in Matching Methods for Causal Inference
- Mapping the effectiveness and risks of GLP-1 receptor agonists
- The wealth of nations and the health of populations: A quasi-experimental design of the impact of sovereign debt crises on child mortality
- Optimal Matching of an Optimally Chosen Subset in Observational Studies
- Effect of Cigarette Price and Tax Increases on Smoking in Europe: A Difference-in-Differences Study with Double Machine Learning
- Optimal Estimation of Generalized Average Treatment Effects using Kernel Optimal Matching
- Learning density ratios in causal inference using Bregman-Riesz regression
- The Oaxaca–Blinder Unexplained Component as a Treatment Effects Estimator
- Estimating Treatment Effects Under Bounded Heterogeneity
- Overlap-Adaptive Regularization for Conditional Average Treatment Effect Estimation
- Fixed Effects and the Generalized Mundlak Estimator
- International environmental agreements and imperfect enforcement: Evidence from CITES
- Estimating Treatment Effects for Depression in Longitudinal Therapy Switching Settings
- The Blessings of Multiple Causes
- Double propensity-score adjustment: A solution to design bias or bias due to incomplete matching
- Balancing vs modeling approaches to weighting in practice
- On variance estimation of the inverse probability‐of‐treatment weighting estimator: A tutorial for different types of propensity score weights
- Causal inference in randomized clinical trials
- A framework for causal estimand selection under positivity violations
- Combining Non-probability and Probability Survey Samples Through Mass Imputation
- Combining Non-Probability and Probability Survey Samples Through Mass Imputation
- Bounding Treatment Effects by Pooling Limited Information across Observations
- Difference-in-Differences with Multiple Time Periods
- Navigated Weighting to Improve Inverse Probability Weighting for Missing Data Problems and Causal Inference
- Why Propensity Scores Should Not Be Used for Matching
- Labour supply effects of early retirement provision
- Estimating quantile treatments without strict overlap
- Using Balancing Weights to Target the Treatment Effect on the Treated when Overlap is Poor
- Posterior Predictive Treatment Assignment for Estimating Causal Effects with Limited Overlap
- Impacts of a School‐Wide, Peer‐Led Approach to Sexuality Education: A Matched Comparison Group Design
- Counterfactual Representation Learning with Balancing Weights
- Causal Inference without Balance Checking: Coarsened Exact Matching
- Identifying Causal-Effect Inference Failure with Uncertainty-Aware Models
- Variable Selection for Causal Inference via Outcome-Adaptive Random Forest
- Causal inference in the context of an error prone exposure: air pollution and mortality
- Community Fact-Checks Do Not Break Follower Loyalty
- Modern causal inference approaches to improve power for subgroup analysis in randomized controlled trials
- Certificates without Electrons? Theory and Evidence on Impacts from AI-Driven Power Demand
- Martingale Doppelgänger-Eval: An Identification Framework for Auditing Candlestick Understanding in Vision-Language Models
- The impacts of same and opposite gender alumni speakers on interest in economics
- The Privatization Origins of Political Corporations: Evidence from the Pinochet Regime
- MOSIC: Model-Agnostic Optimal Subgroup Identification with Multi-Constraint for Improved Reliability
- The effect of unemployment benefits on re-employment rates: Evidence from the Finnish unemployment insurance reform
- Incremental Intervention Effects in Studies with Dropout and Many Timepoints
- Matching Methods for Causal Inference: A Review and a Look Forward
- Early parenting program as intervention strategy for emotional distress in first-time mothers: a propensity score analysis. [europepmc]
- Investing in communities: evaluating the added value of community mobilization on HIV prevention outcomes among FSWs in India. [europepmc]
- Likelihood of Unemployed Smokers vs Nonsmokers Attaining Reemployment in a One-Year Observational Study. [europepmc]
- Matching Weights to Simultaneously Compare Three Treatment Groups: Comparison to Three-way Matching. [europepmc]
- Building the evidence base for REDD+: Study design and methods for evaluating the impacts of conservation interventions on local well-being. [europepmc]
- Decomposition Analysis to Identify Intervention Targets for Reducing Disparities. [europepmc]
- Perils of scaling up: Effects of expanding a nutrition programme in Madagascar. [europepmc]
- Design and Evaluation of an External Control Arm Using Prior Clinical Trials and Real-World Data. [europepmc]
- Identification of the Differential Effect of City-Level on the Gini Coefficient of Health Service Delivery in Online Health Community. [europepmc]
- Instrumental variable estimation of truncated local average treatment effects. [europepmc]
- Applied comparison of large-scale propensity score matching and cardinality matching for causal inference in observational research. [europepmc]
- Core concepts in pharmacoepidemiology: Violations of the positivity assumption in the causal analysis of observational data: Consequences and statistical approaches. [europepmc]
- Prevalence, Characteristics, and Outcomes of Emergency Department Discharge Among Patients With Sepsis. [europepmc]
- Propensity score methods for observational studies with clustered data: A review. [europepmc]
- Two-stage matching-adjusted indirect comparison. [europepmc]
- Comparison of the blood pressure management between sodium-glucose cotransporter 2 inhibitors and glucagon-like peptide 1 receptor agonists. [europepmc]
- Impact of Malakit intervention on perceptions, knowledge, attitudes, and practices related to malaria among workers in clandestine gold mines in French Guiana: results of multicentric cross-sectional surveys over time. [europepmc]
- Denosumab and incidence of type 2 diabetes among adults with osteoporosis: population based cohort study. [europepmc]
- Recovery and symptom trajectories up to two years after SARS-CoV-2 infection: population based, longitudinal cohort study. [europepmc]
- A note on semiparametric efficient generalization of causal effects from randomized trials to target populations. [europepmc]
- Platelet Transfusion and Death or Neurodevelopmental Impairment in Children Born Extremely Preterm. [europepmc]
- Use of sensitivity analyses to assess uncontrolled confounding from unmeasured variables in observational, active comparator pharmacoepidemiologic studies: a systematic review. [europepmc]
- Using overlap weights to address extreme propensity scores in estimating restricted mean counterfactual survival times. [europepmc]
- Stigma toward people living with HIV among healthcare providers in midwifery hospitals in Shenzhen, China from 2020 to 2023. [europepmc]
- Tranexamic acid is associated with post-injury mortality in a resource-limited trauma system: Findings from the epidemiology and outcomes of prolonged trauma care cohort study. [europepmc]
Related