EUVIMEDEUVIMED entdecken
Crossref

Double machine learning-based programme evaluation under unconfoundedness

Michael C Knaus

The Econometrics Journal · 14.09.2022 · 25 (3) · 602-627

DOI: 10.1093/ectj/utac015 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Econometrics Journal, online erschienen am 01.06.2022. Verfasst von Michael C Knaus. Die Arbeit wurde bisher 91-mal zitiert (Crossref) und nennt 89 Literaturangaben. Gefördert von: Swiss National Science Foundation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This paper reviews, applies, and extends recently proposed methods based on double machine learning (DML) with a focus on programme evaluation under unconfoundedness. DML-based methods leverage flexible prediction models to adjust for confounding variables in the estimation of (a) standard average effects, (b) different forms of heterogeneous effects, and (c) optimal treatment assignment rules. An evaluation of multiple programmes of the Swiss Active Labour Market Policy illustrates how DML-based methods enable a comprehensive programme evaluation. Motivated by extreme individualised treatment effect estimates of the DR-learner, we propose the normalised DR-learner (NDR-learner) to address this issue. The NDR-learner acknowledges that individualised effect estimates can be stabilised by an individualised normalisation of inverse probability weights.

Förderung

  • Swiss National Science Foundation

Zitieren

APA

Knaus, M. C. (2022). Double machine learning-based programme evaluation under unconfoundedness. The Econometrics Journal, 25(3), 602-627. https://doi.org/10.1093/ectj/utac015

Vancouver

Knaus MC. Double machine learning-based programme evaluation under unconfoundedness. The Econometrics Journal. 2022;25(3):602-627. doi:10.1093/ectj/utac015

BibTeX
@article{Knaus2022,
  author = {Knaus, Michael C},
  title = {Double machine learning-based programme evaluation under unconfoundedness},
  journal = {The Econometrics Journal},
  year = {2022},
  volume = {25},
  number = {3},
  pages = {602--627},
  doi = {10.1093/ectj/utac015}
}
Kostenlos starten →