Simple Tests for Social Interaction Models with Network Structures

We consider an extended spatial autoregressive model that can incorporate possible endogenous interactions, exogenous interactions, unobserved group fixed effects and correlation of unobservables. In the generalized method of moments (GMM) and the maximum likelihood (ML) frameworks, we introduce simple gradient based tests that can be used to test the presence of endogenous effects, the correlation of unobservables and the contextual effects. We show the asymptotic distributions of tests, and formulate robust tests that have central chi-square distributions under both the null and local misspecification. The proposed tests are easy to compute and only require the estimates from a transformed linear regression model. We carry out an extensive Monte Carlo study to investigate the size and power properties of the proposed tests. Our results show that the proposed tests have good finite sample properties and are useful for testing the presence of endogenous effects, correlation of unobservables and contextual effects in a social interaction model.