CoChat
Sign Up
Cited by 3,563Advanced Causal Inference TechniquesQualitative Comparative Analysis ResearchStatistical Methods and Bayesian Inference

Causal Inference without Balance Checking: Coarsened Exact Matching

Open access
Iacus, Stefano M., King, Gary, Porro, Giuseppe·2012·Political Analysis
DOI 10.1093/pan/mpr013
Open this paper in CoChat — free accountOpen PDF

Abstract

We discuss a method for improving causal inferences called “Coarsened Exact Matching” (CEM), and the new “Monotonic Imbalance Bounding” (MIB) class of matching methods from which CEM is derived. We summarize what is known about CEM and MIB, derive and illustrate several new desirable statistical properties of CEM, and then propose a variety of useful extensions. We show that CEM possesses a wide range of statistical properties not available in most other matching methods but is at the same time exceptionally easy to comprehend and use. We focus on the connection between theoretical properties and practical applications. We also make available easy-to-use open source software for R, Stata , and SPSS that implement all our suggestions.

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/pan/mpr013

Cites

55 of 55 references verified0 citation errors · 0 not checkable

Statistics

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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?

  • Because pruned data are used, the treatment effect cannot be claimed for whole population; estimand is local average treatment effect for treated.

    “Since we are using pruned data, we cannot claim to measure the treatment effect on the whole population.Rather, the estimand in our statistical models is the local average treatment effect for the treated (Iacus et al., 2012).”

    From Relative Feedback Increases Disparities in Effort and Performance in Crowdsourcing Contests (2022) · cites this paper as Iacus et al., 2012

    Section 6.3.3 defines the estimand after matching as the average treatment effect for the matched people, which the paper calls a “local SATT.” But Section 2.3 says pruning only control units leaves the target quantity unchanged; the shift to a local estimand depends on pruning treated units, not merely on using pruned data. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • Using CEM improves estimation by reducing covariate imbalance between treated and control groups.

    “To remedy this, we use Coarsened Exact Matching (CEM) to improve the estimation by reducing the imbalance in covariates between treated and control groups (Blackwell et al. 2009, Iacus et al. 2012).”

    From Skill-Biased Technical Change, Again? Online Gig Platforms and Local Employment (2025) · cites this paper as Iacus et al. 2012

  • The key property of CEM is that it is in a class of methods called Monotonic Imbalance Bounding, which bounds the maximum imbalance in some feature of the empirical distributions through an ex-ante choice by the user.

    “The key property of CEM is that it is in a class of methods called Monotonic Imbalance Bounding, which bounds the maximum imbalance in some feature of the empirical distributions through an exante choice by the user.”

    From Drivers of green bond issuance and new evidence on the 'greenium' (2021) · cites this paper as Iacus et al. 2012

  • CEM is considered a better alternative to PSM because CEM reduces the model dependence of the matching that is inherent in PSM.

    “In our study, we employ coarsened exact matching (CEM), which is considered to be a better alternative to PSM, because CEM reduces the model dependence of the matching that is inherent in PSM (Iacus et al. 2012).”

    From Drivers of green bond issuance and new evidence on the 'greenium' (2021) · cites this paper as Iacus et al. 2012

  • CEM reduces imbalances of covariates between treated and control groups.

    “CEM reduces imbalances of covariates between treated and control groups.”

    From Drivers of green bond issuance and new evidence on the 'greenium' (2021) · cites this paper as Iacus et al. 2012

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Cites

Flagged first, then by citations

Research works printed but not linked: 4

  1. [19]19. ---. 2011b. Replication data for: Causal inference without balance checking: Coarsened Exact Matching. Murray Research Archive [distributor] V1 [version]. http://hdl.handle.net/1902.1/15601.
  2. [24]24. Imai, Kosuke, Gary King, and Elizabeth Stuart. 2008. Misunderstandings among experimentalists and observationalists about causal inference. Journal of the Royal Statistical Society, Series A 171, 2:481-502. http://gking.harvard.edu/files/abs/matchse-abs.shtml (accessed 2008).
  3. [34]34. Lalonde, Robert. 1986. Evaluating the econometric evaluations of training programs. American Economic Review 76:604-20.
  4. [43]43. ---. 2006. Matched sampling for causal effects. Cambridge, UK: Cambridge University Press.

Cited by

Sorted by citations

Related

Authors
Iacus, S. M., King, G., & Porro, G.
Affiliations
  • University of Milan (IT)
  • Harvard University (US)
  • University of Trieste (IT)
Corresponding author
King, Gary
ORCID
Venue
Political Analysis
Publisher
Cambridge University Press (CUP)
Published
2012
Volume
20
Issue
1
Pages
1-24
ISSN
  • 1047-1987 (ISSN-L)
  • 1476-4989
DOI
10.1093/pan/mpr013
OpenAlex
W2029941278
Semantic Scholar
69bc307169754c186f224dcaabf38f7959afe9a1
Cited by
3,563
References
55
Citations by year
2026: 283 · 2025: 364 · 2024: 360 · 2023: 360 · 2022: 350 · 2021: 395 · 2020: 330 · 2019: 249 · 2018: 219 · 2017: 141 · 2016: 155 · 2015: 102 · 2014: 109 · 2013: 69 · 2012: 42
FWCI
67.11
Type
journal article
Language
en
Open access
Licence
www.cambridge.org/core/terms (unspecified, 1830 d)
Locations
  • Political Analysis (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • Archivio Istituzionale della Ricerca (Universita Degli Studi Di Milano) (submittedVersion, other-oa)
  • IrInSubria (University of Insubria) (submittedVersion)
Topics
  • Physical Sciences › Mathematics › Statistics and Probability › Advanced Causal Inference Techniques
  • Social Sciences › Sociology and Political Science › Qualitative Comparative Analysis Research
  • Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
Keywords
coarsened exact matching, causal inference, matching methods, 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