Logical Differencing in Dyadic Network Formation Models with Nontransferable Utilities

TL;DR

This paper introduces a novel logical differencing method for semiparametric dyadic network formation models with nontransferable utilities, effectively controlling for unobserved heterogeneity without requiring additive separability.

Contribution

It develops a new logical differencing approach to handle unobserved heterogeneity in NTU network models, expanding the toolkit for social network analysis.

Findings

Consistent estimator for the model is proposed.

Simulation studies demonstrate the estimator's performance.

Application to Nyakatoke networks illustrates practical utility.

Abstract

This paper considers a semiparametric model of dyadic network formation under nontransferable utilities (NTU). NTU arises frequently in real-world social interactions that require bilateral consent, but by its nature induces additive non-separability. We show how unobserved individual heterogeneity in our model can be canceled out without additive separability, using a novel method we call logical differencing. The key idea is to construct events involving the intersection of two mutually exclusive restrictions on the unobserved heterogeneity, based on multivariate monotonicity. We provide a consistent estimator and analyze its performance via simulation, and apply our method to the Nyakatoke risk-sharing networks.