Ariane, can you clarify the following: In the bail example you mentioned, would you say it is unethical if you WOULD have the money to pay bail for everybody in need and you CHOOSE not to – but if you have limited resources (as I assume it is the case) I see no wrong in assigning randomly to the program?
Ariane: I am not saying whether this choice is right or wrong in absolute terms, I’m just saying that it violates the equipoise principle, which is imposed on medical RCTs. Random treatment assignment (versus nothing) may seem unfair, and it hurts those who receive nothing (i.e., the control group). The subjects who are involved in an RCT and receive nothing are typically worse off than not only the treated subjects, but also the people who are not involved in the experiment. The idea underlying equipoise is that the subjects in the control group deserve fair compensation.