In an earlier post, I claimed that Regan’s Utilitarianism and Co-operation is important because it shows that in order for a moral theory to be plausible it must require of agents something beyond just the performance of certain actions—in other words, it must not be exclusively act-orientated. But, as Matt suggested in his comment, the idea that moral theories should require of agents something beyond just the performance of certain actions isn’t exactly groundbreaking. After all, lots of moral theorists have claimed (and well before 1980) that agents are required not only to perform certain actions but also to possess certain virtues, have certain motives, internalize certain rules, apply certain decision procedures, etc.
Now let’s consider The Buttons example from my earlier post. It seems a mistake to think that Uncoop has both a fundamental obligation to push and a fundamental obligation to desire that he and Coop cooperate. For Uncoop gets no brownie points for pushing without also desiring that he and Coop cooperate. What we need to say, then, is not that Uncoop has fundamental obligations both to push and to desire that he and Coop cooperate, but that Uncoop has a fundamental obligation to see to it that he and Coop both push and, consequently, derivative obligations both to push and to desire that they cooperate. Actually, on Regan’s view (viz., co-operative utilitarianism), what Uncoop (and all other agents) have a fundamental obligation to do is to apply a certain decision procedure (where applying this decision procedure involves more than just performing certain actions). It’s a complicated procedure, but the general idea is that each agent should (1) hold “himself ready to do his part in the best pattern of behavior for the group of cooperators,” (2) “identify the other agents who are willing and able to co-operate in the production of the best possible consequences,” and (3) “do his part in the best plan of behaviour for the group consisting of himself and the others so identified, in view of the behaviour of non-members of that group” (Regan 1980, pp. x and 135). So, for Regan, the obligation that Uncoop has to push is a derivative obligation. He has this obligation only because he has an obligation to correctly apply a decision procedure that when correctly applied directs him to push.
So what I think that Regan’s arguments show is that no moral theory that treats obligations to perform acts as fundamental as opposed to derivative will be able to avoid counterintuitive implications in examples like The Buttons. To my mind, this is truly groundbreaking, because it seems to me that most contemporary moral theories treat obligations to perform actions as fundamental obligations and not as derivative of fundamental obligations to, say, apply a certain decision procedure or (on my own view) to secure a certain possible world (the one where Uncoop and Coop both push).