Jonathan Fuhr, Philipp Berens, Dominik Papies
arXiv 21 Mar 2024 · Statistics — Machine Learning · 6 citations (OpenAlex)
arXiv:2403.14385 · PDF · DOI · OpenAlex · Extracted main text
The estimation of causal effects with observational data continues to be a very active research area. In recent years, researchers have developed new frameworks which use machine learning to relax classical assumptions necessary for the estimation of causal effects. In this paper, we review one of the most prominent methods - "double/debiased machine learning" (DML) - and empirically evaluate it by comparing its performance on simulated data relative to more traditional statistical methods, before applying it to real-world data. Our findings indicate that the application of a suitably flexible machine learning algorithm within DML improves the adjustment for various nonlinear confounding relationships. This advantage enables a departure from traditional functional form assumptions typically necessary in causal effect estimation. However, we demonstrate that the method continues to critically depend on standard assumptions about causal structure and identification. When estimating the effects of air pollution on housing prices in our application, we find that DML estimates are consistently larger than estimates of less flexible methods. From our overall results, we provide actionable recommendations for specific choices researchers must make when applying DML in practice.
appendix boundary found by appendix_command · 89% of the source is main text. Read the extracted text to check this.
The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.
| Reference | Intensity | Mentions | Sections | Main text | |
|---|---|---|---|---|---|
| 1 | Hernán, M. A. and Robins, J. M (2020) Causal Inference: What If | 1.000 | 7 | 3 | 100% |
| 2 | Imbens, G. W. and Rubin, D. B (2015) Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction | 1.000 | 7 | 3 | 100% |
| 3 | Cinelli, C., Forney, A., and Pearl, J (2022) A Crash Course in Good and Bad Controls | 1.000 | 6 | 4 | 100% |
| 4 | Chernozhukov, V., Chetverikov, D., Demirer, M., Duflo, E., Hansen, C… (2018) Double/debiased machine learning for treatment and structural parameters | 0.979 | 33 | 7 | 94% |
| 5 | Belloni, A., Chernozhukov, V., and Hansen, C (2014) High-Dimensional Methods and Inference on Structural and Treatment Effects | 0.941 | 6 | 4 | 83% |
| 6 | McConnell, K. J. and Lindner, S (2019) Estimating treatment effects with machine learning | 0.928 | 5 | 3 | 80% |
| 7 | Wooldridge, J (2012) Introductory Econometrics: A Modern Approach | 0.928 | 4 | 3 | 100% |
| 8 | Hastie, T., Tibshirani, R., and Friedman, J (2009) The Elements of Statistical Learning: Data Mining, Inference, and Prediction, Second Edition | 0.874 | 6 | 2 | 100% |
| 9 | Mullainathan, S. and Spiess, J (2017) Machine Learning: An Applied Econometric Approach | 0.874 | 5 | 2 | 100% |
| 10 | Athey, S (2019) The Impact of Machine Learning on Economics | 0.843 | 3 | 3 | 100% |
Showing the top 10 of 86 scored citations.
arXiv econ.EM papers that cite this one, ranked by how heavily they lean on it.
| Citing paper | Intensity | Mentions | Sections | |
|---|---|---|---|---|
| 1 | Anytime-Valid Inference for Double/Debiased Machine Learning of Causal Parameters | 0.405 | 1 | 1 |