Almost-Matching-Exactly for Treatment Effect Estimation under Network Interference
This addresses treatment effect estimation under network interference, a problem in causal inference for social and network sciences, but it appears incremental as an extension of matching methods to network settings.
The authors tackled the problem of estimating direct treatment effects in randomized experiments where units are connected in a network and can influence each other, which biases traditional estimators. They proposed a matching method that matches units almost exactly on counts of unique subgraphs in their neighborhood graphs, showing empirically that it performs better than existing methods.
We propose a matching method that recovers direct treatment effects from randomized experiments where units are connected in an observed network, and units that share edges can potentially influence each others' outcomes. Traditional treatment effect estimators for randomized experiments are biased and error prone in this setting. Our method matches units almost exactly on counts of unique subgraphs within their neighborhood graphs. The matches that we construct are interpretable and high-quality. Our method can be extended easily to accommodate additional unit-level covariate information. We show empirically that our method performs better than other existing methodologies for this problem, while producing meaningful, interpretable results.