Finding network effect of randomized treatment under weak assumptions for any outcome and any effect heterogeneity

TL;DR

This paper develops a nonparametric framework to estimate network effects of treatments, capturing diverse outcomes and heterogeneity, and critically assesses linear models for potential biases under weak assumptions.

Contribution

It introduces a network-based causal reduced form that is nonparametric and broadly applicable, enabling almost model-free estimation of network effects.

Findings

Reveals biases in linear models due to restrictive assumptions

Provides a flexible nonparametric approach for diverse outcomes

Enables almost model-free inference for network effects

Abstract

In estimating the effects of a treatment/policy with a network, an unit is subject to two types of treatment: one is the direct treatment on the unit itself, and the other is the indirect treatment (i.e., network/spillover influence) through the treated units among the friends/neighbors of the unit. In the literature, linear models are widely used where either the number of the treated neighbors or the proportion of them among the neighbors represents the intensity of the indirect treatment. In this paper, we obtain a nonparametric network-based "causal reduced form (CRF)" that allows any outcome variable (binary, count, continuous, ...) and any effect heterogeneity. Then we assess those popular linear models through the lens of the CRF. This reveals what kind of restrictive assumptions are embedded in those models, and how the restrictions can result in biases. With the CRF, we conduct almost model-free estimation and inference for network effects.