Flexible Imputation of Incomplete Network Data

2026-04-03

AI Summary

This study addresses the challenges of information loss and estimation bias arising from network data sampling by proposing a nonparametric imputation method that dispenses with parametric assumptions or low-rank constraints. The approach integrates covariate projection with local two-way fixed-effects regression, flexibly combining observed covariates and unobserved heterogeneity to achieve robust reconstruction of missing network links. Theoretical analysis establishes element-wise convergence rates for the imputed matrix and demonstrates the consistency of subsequent generalized method of moments (GMM) estimators. Empirical evaluations on both simulated and real-world microfinance network data confirm the method’s superior imputation accuracy and enhanced performance in downstream parameter estimation.

Abstract

Sampled network data are common in empirical research because collecting full network information is costly, but using sampled networks can lead to biased estimates. We propose a nonparametric imputation method for sampled networks and show that empirical analysis based on imputed networks yields consistent parameter estimates. Our approach imputes missing network links by combining a projection onto covariates with a local two-way fixed-effects regression, which avoids parametric assumptions, does not rely on low-rank restrictions, and flexibly accommodates both observed covariates and unobserved heterogeneity. We establish entrywise convergence rates for the imputed matrix and prove the consistency of GMM estimators based on the imputed network. We further derive the convergence rate of the corresponding estimator in the linear-in-means peer-effects model. Simulations show strong performance of our method both in terms of imputation accuracy and in downstream empirical analysis. We illustrate our method with an application to the microfinance network data of Banerjee et al. (2013).

Problem

• incomplete network data
• sampled networks
• biased estimates
• network imputation
• missing links

Innovation

• nonparametric imputation
• network sampling
• two-way fixed effects
• unobserved heterogeneity
• GMM consistency