Entropy Balancing for Causal Effects: A Multivariate Reweighting Method to Produce Balanced Samples in Observational Studies
Open accessThe DOI matches its Crossref record; 55 of the 55 references it cites were found in a registry or an index.
Abstract
This paper proposes entropy balancing, a data preprocessing method to achieve covariate balance in observational studies with binary treatments. Entropy balancing relies on a maximum entropy reweighting scheme that calibrates unit weights so that the reweighted treatment and control group satisfy a potentially large set of prespecified balance conditions that incorporate information about known sample moments. Entropy balancing thereby exactly adjusts inequalities in representation with respect to the first, second, and possibly higher moments of the covariate distributions. These balance improvements can reduce model dependence for the subsequent estimation of treatment effects. The method assures that balance improves on all covariate moments included in the reweighting. It also obviates the need for continual balance checking and iterative searching over propensity score models that may stochastically balance the covariate moments. We demonstrate the use of entropy balancing with Monte Carlo simulations and empirical applications.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Moreover, we match treatment and control groups on a wider range of individual- and school-level characteristics using reweighting on the propensity of treatment (68) and maximum-entropy balancing (69).”
From Learning loss due to school closures during the COVID-19 pandemic (2021) · cites this paper as [69]
In the citing paper
“This approach matches treatment and control groups based on an extended set of covariates and creates balancing weights that differ as little as possible from unitary weights (Hainmueller, 2012).”
From Double-edged sword: persistent effects of Communist regime affiliations on well-being and preferences (2023) · doi:10.1007/s00148-022-00930-0 · cites this paper as Hainmueller, 2012
Citations
Cites
The central role of the propensity score in observational studies for causal effects
ROSENBAUM, P. R., & RUBIN, D. B. (1983) · Biometrika · doi:10.1093/biomet/70.1.41 · Cited by 31,241 · cited 3×
“The conventional solution to this problem is to assume ignorable treatment assignment and overlap (Rosenbaum and Rubin 1983), which implies thatY(0) ⊥ D”
InfluentialLarge Sample Properties of Generalized Method of Moments Estimators
Hansen, L. P. (1982) · Econometrica · doi:10.2307/1912775 · Cited by 13,467
Information theory and statistics
(1959) · Journal of the Franklin Institute · doi:10.1016/0016-0032(59)90368-0 · Cited by 7,094
A generalization of sampling without replacement from a finite universe
Horvitz, D. G., & Thompson, D. J. (1952) · Journal of the American Statistical Association · doi:10.1080/01621459.1952.10483446 · Cited by 4,293 · cited 2×
“…sample with the treated observations in the reweighted data.6 The idea is similar to the classic Horvitz–Thompson adjustment used in the survey literature where units are weighted by the inverse of the inclusion probabilities that result from the sampling design (Horvitz and Thompson 1952).”
Matching as Nonparametric Preprocessing for Reducing Model Dependence in Parametric Causal Inference
Ho, D. E., Imai, K., King, G., & Stuart, E. A. (2007) · Political Analysis · doi:10.1093/pan/mpl013 · Cited by 4,205 · cited 7×
“5 A variety of such preprocessing procedures have been proposed (Imbens 2004; Rubin 2006; Ho et al. 2007; Sekhon 2009).”
InfluentialCausal Inference without Balance Checking: Coarsened Exact Matching
Iacus, S. M., King, G., & Porro, G. (2012) · Political Analysis · doi:10.1093/pan/mpr013 · Cited by 3,563
A Generalization of Sampling Without Replacement From a Finite Universe
Horvitz, D. G., & Thompson, D. J. (1952) · Journal of the American Statistical Association · doi:10.2307/2280784 · Cited by 3,011
Nonparametric Estimation of Average Treatment Effects Under Exogeneity: A Review
Imbens, G. W. (2004) · Review of Economics and Statistics · doi:10.1162/003465304323023651 · Cited by 2,994
Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score
Hirano, K., Imbens, G. W., & Ridder, G. (2003) · Econometrica · doi:10.1111/1468-0262.00442 · Cited by 2,355 · cited 4×
“In this regard, entropy balancing provides a generalization of the propensity score weighting approach (Hirano, Imbens, and Ridder 2003) where the researcher first estimates the propensity score weights with a logistic regression and then computes balance checks to see if the estimated weights…”
InfluentialVariable Selection for Propensity Score Models
Brookhart, M. A., Schneeweiss, S., Rothman, K. J., Glynn, R. J., Avorn, J., & Stürmer, T. (2006) · American Journal of Epidemiology · doi:10.1093/aje/kwj149 · Cited by 2,336
Research works printed but not linked: 2
- Oh H. L. , and Scheuren F. J. 1978. Multivariate ratio raking estimation in the 1973 exact match study. Proceedings of the Section on Survey Research Methods XXV: 716–22.
- Zaslavsky, 1988, Representing local reweighting area adjustments by of households, Survey Methodology, 14
Cited by
Items 1,001–1,007 of 1,007
Environmental change and migration aspirations: Evidence from Bangladesh
Rudolph, L., Koubi, V., & Freihardt, J. (2025) · Global Environmental Change · doi:10.1016/j.gloenvcha.2025.102966 · Cited by 18
“We use entropy balancing (Hainmueller, 2012) to achieve balance in the distribution of the first, second and third moments of these covariates for both the erosion control and treatment groups and the flood erosion and treatment group.”
Andreicovici, I., Jeny, A., & Lui, D. (2025) · The Accounting Review · doi:10.2308/tar-2020-0437 · Cited by 8
(2021) · RePEc: Research Papers in Economics · Cited by 6
Clayton, K., Ferwerda, J., & Horiuchi, Y. (2022) · doi:10.31235/osf.io/wn32u · Cited by 6
Artificial intelligence (AI) and corporate governance: Evidence from board size
Chatjuthamard, P., Jiraporn, P., & Lee, S. M. (2026) · Journal of Behavioral and Experimental Finance · doi:10.1016/j.jbef.2026.101144 · Cited by 4
Wladis, C., Sencindiver, B., Hachey, A. C., Myszkowski, N., Offenholley, K., Samuels, J., & Aly, G. (2026) · AERA Open · doi:10.1177/23328584261419497 · Cited by 1
Robots and Firms’ Labor Search: The Role of Temporary Work Agencies
Beneito, P., Garcia de la Vega, M., Vicente-Chirivella, Ó., & Wilemme, G. (2025) · doi:10.2139/ssrn.5928552
Metadata
- Authors
- Hainmueller, J.
- Affiliations
- Massachusetts Institute of Technology (US)
- Corresponding author
- Hainmueller, Jens
- ORCID
- Hainmueller, Jens
- Venue
- Political Analysis
- Publisher
- Cambridge University Press (CUP)
- Published
- 2012
- Volume
- 20
- Issue
- 1
- Pages
- 25-46
- ISSN
- 1047-1987 (ISSN-L)
- 1476-4989
- OpenAlex
- W3122812581
- Cited by
- 6,026
- References
- 55
- Citations by year
- 2026: 1134 · 2025: 1230 · 2024: 954 · 2023: 686 · 2022: 541 · 2021: 430 · 2020: 322 · 2019: 232 · 2018: 164 · 2017: 124 · 2016: 77 · 2015: 54 · 2014: 42 · 2013: 20 · 2012: 11
- FWCI
- 21.58
- Type
- journal article
- Language
- en
- Open access
- green
- doi.org
- Locations
- Political Analysis (publishedVersion)
- SSRN Electronic Journal (acceptedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Social Sciences › Decision Sciences › Management Science and Operations Research › Optimal Experimental Design Methods
- Keywords
- entropy balancing, covariate balance, reweighting, observational studies, binary treatment, causal effects, propensity score, model dependence, sample moments, treatment effect estimation
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.