Small Area Causal Inference

We develop a framework for small area causal inference (SACI). Unlike the traditional doubly robust (DR) estimation approach, the new framework is based on the empirical best prediction (EBP) method. EBP takes into consideration not only the bias but also the variance factor, which is often a more dominant component in the mean squared prediction error (MSPE). Under a linear mixed model for SACI, we establish consistency and asymptotic normality of the maximum likelihood estimator of the model parameters. More relevant to our main objective, we establish asymptotic optimality of the EBP in terms of minimizing the MSPE of the area-specific causal-effect (ASCE) predictor. Moreover, in view of the traditional interest in causal inference, we establish asymptotic normality of the EBP of the ASCE , which is used to construct prediction intervals for the ASCEs. We carry out extensive simulation studies to investigate finite-sample performance of the proposed EBP and its comparison with the traditional DR and other existing methods. Our simulation results show that the proposed EBP method has significant advantage over the comparing methods in terms of MSPE and coverage probability and length of the resulting prediction interval. A real-data example from agricultural-animal genetics is discussed.

Authors

Publication Details

Journal
arXiv (Cornell University)
Published
2026-11-17
DOI
https://doi.org/10.48550/arxiv.2611.22339
Primary Topic
Advanced Causal Inference Techniques
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Small Area Causal Inference

Jiangshan Zhang, Yilin Li, Jiming Jiang
arXiv (Cornell University)
Advanced Causal Inference Techniques
preprint

Small Area Causal Inference

Jiangshan Zhang, Yilin Li, Jiming Jiang
preprint en

Abstract

We develop a framework for small area causal inference (SACI). Unlike the traditional doubly robust (DR) estimation approach, the new framework is based on the empirical best prediction (EBP) method. EBP takes into consideration not only the bias but also the variance factor, which is often a more dominant component in the mean squared prediction error (MSPE). Under a linear mixed model for SACI, we establish consistency and asymptotic normality of the maximum likelihood estimator of the model parameters. More relevant to our main objective, we establish asymptotic optimality of the EBP in terms of minimizing the MSPE of the area-specific causal-effect (ASCE) predictor. Moreover, in view of the traditional interest in causal inference, we establish asymptotic normality of the EBP of the ASCE , which is used to construct prediction intervals for the ASCEs. We carry out extensive simulation studies to investigate finite-sample performance of the proposed EBP and its comparison with the traditional DR and other existing methods. Our simulation results show that the proposed EBP method has significant advantage over the comparing methods in terms of MSPE and coverage probability and length of the resulting prediction interval. A real-data example from agricultural-animal genetics is discussed.

arXiv (Cornell University)
Advanced Causal Inference Techniques
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon