Identification of causal effects in the presence of nonignorable missing outcome values

We consider a new approach to identify the causal effects of a binary treatment when the outcome is missing on a subset of units and dependence of nonresponse on the outcome cannot be ruled out even after conditioning on observed covariates. We provide sufficient conditions under which the availability of a binary instrument for nonresponse allows us to partially identify causal effects in some latent subgroups of units, defined by their nonresponse behavior in all possible combinations of treatment and instrument, named Principal Strata. Examples are provided as possible scenarios where our assumptions may be plausible; they are used to discuss the key role of the instrument for nonresponse in identifying average causal effects in presence of non-ignorable missing outcomes and provide new guidelines on study designs for causal inference. [Working Paper 2011/05, Department of Statistics, University of Florence, Italy]