Acknowledcments I Thank Again Dr. Forcina and the Editor for a Stimulating Article, and the Nih Grants for Statistical Methods for Partially Controlled Studies, and for the Center for Prevention and Early Intervention, for Partial Support

Comment It is a pleasure to comment on the paper by Antonio Forcina. Our comment is restricted to the example, discussed in section 3.3 of Forcina's paper, of a two-armed double-blind placebo-controlled randomized experiment in which: (i) no subject in the placebo arm takes active treatment, (ii) some subjects in the treatment arm do not comply and fail to take the active treatment, and (iii) the exclusion restriction that the randomization arm has no direct effect on a binary outcome of interest except through received treatment is assumed true, owing to double blinding. In this setting, it is known that, under a counterfactual or potential outcomes causal model, the average effect of treatment on the treated is the only causal