Estimating causal effects using difference-in-differences under network dependency and interference

Abstract: Differences-in-differences (DiD) is a causal inference method for observational longitudinal data that assumes parallel expected outcome trajectories between treatment groups under the (possible) counterfactual of receiving a specific treatment. In this paper DiD is extended to allow for (i) network dependency where outcomes, treatments, and covariates may exhibit between-unit latent correlation, and (ii) interference, where treatments can affect outcomes in neighboring units. In this setting, the causal estimand of interest is the average exposure effect among units with a specific exposure level, where the exposure is a function of treatments from potentially many units. Under a conditional parallel trends assumption and suitable network dependency conditions, a doubly robust estimator allowing for data-adaptive nuisance function estimation is proposed and shown to be consistent and asymptotically normal with variance reaching the semiparametric efficiency bound. The proposed methods are evaluated in simulations and applied to study the effects of adopting emission control technologies in coal power plants on county-level mortality due to cardiovascular disease.