Monotone Ecological Inference

2025-04-20

AI Summary

This paper addresses the fundamental challenge in ecological inference—unobserved individual-level behavior—by proposing a partial identification framework grounded in monotonicity constraints to estimate causal differences between groups. The method leverages only the *signs* (not magnitudes) of conditional associations within neighborhoods and across geographic regions, employing interval identification under partial-order constraints, sign-restricted optimization, and formal sensitivity analysis to derive sharp upper and lower bounds for target parameters such as vaccination rates. Its key contribution is the first systematic integration of monotonicity assumptions into ecological inference, circumventing restrictive linear or independence assumptions inherent in classical approaches. Applied to U.S. county-level COVID-19 vaccination data, the framework substantially tightens credible bounds on the vaccination-rate difference between Republican- and Democratic-leaning voters, yielding marked improvements in estimation precision and robustness over conventional methods.

Technology Category

• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Nonmonotonic Reasoning
• Machine Learning: Causal Learning

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

We study monotone ecological inference, a partial identification approach to ecological inference. The approach exploits information about one or both of the following conditional associations: (1) outcome differences between groups within the same neighborhood, and (2) outcomes differences within the same group across neighborhoods with different group compositions. We show how assumptions about the sign of these conditional associations, whether individually or in relation to one another, can yield informative sharp bounds in ecological inference settings. We illustrate our proposed approach using county-level data to study differences in Covid-19 vaccination rates among Republicans and Democrats in the United States.

Problem

• Studying monotone ecological inference for partial identification
• Exploiting conditional associations for outcome differences analysis
• Applying approach to Covid-19 vaccination rate disparities

Innovation

• Exploits conditional outcome differences between groups
• Uses neighborhood and group composition variations
• Applies sign assumptions for sharp bounds