Differences-in-Differences with multiple Treatments under Control

Numerous quasi-experimental identification strategies making use of the differencein- differences setup suffer from multiple treatments which can be separated into sequential and simultaneous treatments. While for causal inferences under sequential treatments a staggered difference-in-differences approach might be applied, for causal inferences under simultaneous treatments the standard differences-indifferences approach is normally not applicable. Accordingly, we present an adjusted differences-in-differences identification strategy that can neutralize the effects of additional treatments implemented simultaneously through the definition and the specific composition of the control group and an amended common trend assumption. Even though the adjusted difference-in-differences strategy identifies the average treatment effect on the treated, we also show that the adjusted strategy is capable of identifying the average treatment effect under stronger common trend assumptions and the absence of interaction effects between the treatments.