A jackknife bias correction for nonlinear network data models with fixed effects

Abstract

I introduce a new method for bias correction of dyadic models with agent-specific fixed effects, including the dyadic link formation model with homophily and degree heterogeneity. The proposed approach uses a jackknife procedure to deal with the incidental parameters problem. The method can be applied to both directed and undirected networks, allows for non-binary outcome variables, and can be used to bias correct estimates of average effects and counterfactual outcomes. I also show how the jackknife can be used to bias correct fixed-effect averages over functions that depend on multiple nodes, e.g. triads or tetrads in the network. As an example, I implement specifica- tion tests for dependence across dyads, such as reciprocity or transitivity. Finally, I demonstrate the usefulness of the estimator in an application to a gravity model for import/export relationships across countries.