Inference under Staggered Adoption: Case Study of the Affordable Care Act

2024-12-12

AI Summary

This paper addresses the heterogeneous treatment timing arising from the staggered state-level implementation of the Affordable Care Act’s Medicaid expansion. To tackle this challenge, we propose a novel causal inference method for panel data. Methodologically, we develop the first computationally efficient matrix estimation inference framework with rigorous finite-sample coverage guarantees, supporting estimation of both individual- and aggregate-level treatment effects under arbitrary bilinear functional forms. Our approach integrates matrix completion, doubly robust estimation, and bootstrap resampling to construct robust confidence intervals. Theoretically, our framework provides a statistically grounded, general-purpose inference tool for staggered policy evaluation. Empirically, we find that Medicaid expansion significantly reduced uninsurance rates and infant mortality, but had no statistically significant effect on total healthcare expenditures.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Matrix & Tensor Methods
• Intelligent Robots: State Estimation

Application Category

• Security and Privacy: Large-scale security measurements
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Panel data consists of a collection of $N$ units that are observed over $T$ units of time. A policy or treatment is subject to staggered adoption if different units take on treatment at different times and remains treated (or never at all). Assessing the effectiveness of such a policy requires estimating the treatment effect, corresponding to the difference between outcomes for treated versus untreated units. We develop inference procedures that build upon a computationally efficient matrix estimator for treatment effects in panel data. Our routines return confidence intervals (CIs) both for individual treatment effects, as well as for more general bilinear functionals of treatment effects, with prescribed coverage guarantees. We apply these inferential methods to analyze the effectiveness of Medicaid expansion portion of the Affordable Care Act. Based on our analysis, Medicaid expansion has led to substantial reductions in uninsurance rates, has reduced infant mortality rates, and has had no significant effects on healthcare expenditures.

Problem

• Estimating treatment effects in staggered adoption panel data
• Developing inference procedures for treatment effect confidence intervals
• Analyzing Medicaid expansion impact under Affordable Care Act

Innovation

• Matrix estimator for treatment effects
• Confidence intervals with coverage guarantees
• Analysis of Medicaid expansion effects