1: \begin{abstract}

2: This paper presents a novel application of graph neural networks for

3: modeling and estimating network heterogeneity. Network heterogeneity

4: is characterized by variations in unit's decisions or outcomes that

5: depend not only on its own attributes but also on the conditions of

6: its surrounding neighborhood. We delineate the convergence rate of

7: the graph neural networks estimator, as well as its applicability

8: in semiparametric causal inference with heterogeneous treatment effects.

9: The finite-sample performance of our estimator is evaluated through

10: Monte Carlo simulations. In an empirical setting related to microfinance

11: program participation, we apply the new estimator to examine the average

12: treatment effects and outcomes of counterfactual policies, and to

13: propose an enhanced strategy for selecting the initial recipients

14: of program information in social networks.

15: \end{abstract}

16: