arXiv 4 Jun 2019 · Statistics — Machine Learning · 8 citations (OpenAlex)
arXiv:1906.01552 · PDF · DOI · OpenAlex · Extracted main text
Personalized interventions in social services, education, and healthcare leverage individual-level causal effect predictions in order to give the best treatment to each individual or to prioritize program interventions for the individuals most likely to benefit. While the sensitivity of these domains compels us to evaluate the fairness of such policies, we show that actually auditing their disparate impacts per standard observational metrics, such as true positive rates, is impossible since ground truths are unknown. Whether our data is experimental or observational, an individual's actual outcome under an intervention different than that received can never be known, only predicted based on features. We prove how we can nonetheless point-identify these quantities under the additional assumption of monotone treatment response, which may be reasonable in many applications. We further provide a sensitivity analysis for this assumption by means of sharp partial-identification bounds under violations of monotonicity of varying strengths. We show how to use our results to audit personalized interventions using partially-identified ROC and xROC curves and demonstrate this in a case study of a French job training dataset.
appendix boundary found by appendix_command · 67% 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 | N. Kallus and A. Zhou (2019) The fairness of risk scores beyond classification: Bipartite ranking and the xauc metric | 1.000 | 5 | 3 | 100% |
| 2 | M. Hardt, E. Price, N. Srebro, et al (2016) Equality of opportunity in supervised learning | 0.843 | 3 | 3 | 100% |
| 3 | S. Wager and S. Athey (2017) Estimation and inference of heterogeneous treatment effects using random forests | 0.843 | 3 | 3 | 100% |
| 4 | T. Kitagawa and A. Tetenov (2015) Empirical welfare maximization | 0.737 | 3 | 3 | 67% |
| 5 | C. F. Manski (2003) Partial identification of probability distributions | 0.737 | 3 | 2 | 100% |
| 6 | S. Barocas, M. Hardt, and A. Narayanan (2018) Fairness and Machine Learning | 0.644 | 2 | 2 | 100% |
| 7 | J. L. Hill (2011) Bayesian nonparametric modeling for causal inference | 0.644 | 2 | 2 | 100% |
| 8 | S. Wager and S. Athey (2017) Efficient policy learning | 0.644 | 2 | 2 | 100% |
| 9 | M. Adler Well-Being and Fair Distribution | 0.511 | 2 | 2 | 50% |
| 10 | M. Berger, D. A. Black, and J. A. Smith (2000) Econometric evaluation of labour market policies, chapter EvaluatingProfiling as a Means of Allocating Government Service, pages… | 0.511 | 2 | 2 | 50% |
Showing the top 10 of 54 scored citations.
arXiv econ.EM papers that cite this one, ranked by how heavily they lean on it.
| Citing paper | Intensity | Mentions | Sections | |
|---|---|---|---|---|
| 1 | What's the Harm? Sharp Bounds on the Fraction Negatively Affected by Treatment | 0.511 | 2 | 1 |
| 2 | Debiased Machine Learning of Set-Identified Linear Models | 0.405 | 1 | 1 |
| 3 | Generalized Lee Bounds | 0.405 | 1 | 1 |
| 4 | Treatment Effect Risk: Bounds and Inference | 0.405 | 1 | 1 |
| 5 | Debiased Machine Learning of Aggregated Intersection Bounds and Other Causal Parameters | 0.405 | 1 | 1 |