Fair Policy Targeting
From MaRDI portal
Publication:6154028
DOI10.1080/01621459.2022.2142591arXiv2005.12395OpenAlexW4311747487MaRDI QIDQ6154028FDOQ6154028
Authors: Jelena Bradic 
Publication date: 19 March 2024
Published in: Journal of the American Statistical Association (Search for Journal in Brave)
Abstract: One of the major concerns of targeting interventions on individuals in social welfare programs is discrimination: individualized treatments may induce disparities across sensitive attributes such as age, gender, or race. This paper addresses the question of the design of fair and efficient treatment allocation rules. We adopt the non-maleficence perspective of first do no harm: we select the fairest allocation within the Pareto frontier. We cast the optimization into a mixed-integer linear program formulation, which can be solved using off-the-shelf algorithms. We derive regret bounds on the unfairness of the estimated policy function and small sample guarantees on the Pareto frontier under general notions of fairness. Finally, we illustrate our method using an application from education economics.
Full work available at URL: https://arxiv.org/abs/2005.12395
Cites Work
- Introduction to empirical processes and semiparametric inference
- Maximum score estimation of the stochastic utility model of choice
- Exact computation of max weighted score estimators
- Predicting binary outcomes
- Fairness through awareness
- Semiparametric efficiency bounds
- Robust inference on average treatment effects with possibly more covariates than observations
- Who should be treated? Empirical welfare maximization methods for treatment choice
- Statistical treatment rules for heterogeneous populations.
- Optimal Dynamic Treatment Regimes
- Semiparametric Efficiency in Multivariate Regression Models with Missing Data
- Microeconomic theory
- Title not available (Why is that?)
- An algorithm for linear programming which requires \(O(((m+n)n^ 2+(m+n)^{1.5}n)L)\) arithmetic operations
- Asymptotics for statistical treatment rules
- Estimation of best predictors of binary response
- WELFARE ECONOMICS AND EXISTENCE OF AN EQUILIBRIUM FOR A COMPETITIVE ECONOMY
- Inferring welfare maximizing treatment assignment under budget constraints
- Minimax regret treatment choice with covariates or with limited validity of experiments
- Model selection for treatment choice: penalized welfare maximization
- Statistical treatment choice based on asymmetric minimax regret criteria
- Policy learning with observational data
This page was built for publication: Fair Policy Targeting
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6154028)
