We propose the dynamic network effect (DNE) model for the study of high-dimensional multivariate time series data. Cross-sectional dependencies between units are captured via one or multiple observed networks and a low-dimensional vector of latent stochastic network effects. The...