For conditions 1 (”many chances”) and 2 (”recoverable loss”), though, we need to pause and reconsider.

The many chances condition is phrased in terms of law and public policy, and not morality. When it comes down to policies that have a robust amount of data — for instance, say, incidences of fatalities when the driver is not wearing their seat belt — we have tons of data, and we can take its advice in terms of a policy. But this is not to say that conditions cannot deviate (in what Wolff calls “one-off” decisions), where the situation’s factors are all unique, and all need to be evaluated carefully. All this means that 1 shouldn’t be phrased in terms of absolute necessity by the author’s standards. It is not a necessary, condition, but rather, an informative thing to have.

(And this is not to say that we must enter blind into one-off decisions. It just means that we have to scrutinize the context relatively more.)

The “recoverable loss” assumption plays a huge role in decision-making across the board. It also seems clear that Wolff has intuitions that seem viable: we need to allow behavior that carries small yet unrecoverable risks. (To quote the worries of one venerable Mr. Simpson: “What if I got into the shower and slipped on a bar of soap? OhmyGod, I’d be killed!” Well, of course we should still take showers, despite the risk of unrecoverable loss.)

Nevertheless, it seems quixotic to me, precisely for the reason that Hanno gave. Sometimes, for utilitarianism, unrecoverable losses have to be allowed to happen in order to do the right thing, because on balance the alternative is worse. And it makes no sense to say that “unrecoverable loss” has been assumed in the calculation if that’s precisely the consequence that has been predicted and endorsed. And if this isn’t allowed, then one of the core intuitions behind utilitarianism falls by the wayside, which may be enough to close the books on it for good. Of course, I dispute that.

—

I’m having trouble with one of Hanno’s comments, that “consequentialism is not summing real people up”. If I interpret it in its strongest form, in the sense that the contextual factors involved in a particular act (including the features of the actor) are not taken to be the thing to examine in a consequentialism analysis, then I would disagree. If consequentialism (and utilitarianism) couldn’t (or doesn’t) deal with real actors, then it would be wholly alienated from making context-sensitive judgments. But clearly we can make such judgments, as we regularly do in ethics classes.

Rather, what I think what Hanno meant with the above quoted comments is that consequentialism doesn’t necessarily have a particular actor or theory of action in mind. For, he writes directly afterward that “Consequentialism does not purport to be or to contain a theory of the person. However, the choice of our theory of the person may impose certain limitations on what goes into the consequentialist calculus.” That’s surely true.

Also, I think that Hanno’s point about the ideal actor is on target so long as we’re speaking about the consequentialist’s approach to the habits of persons. But if we’re talking at that level of abstraction, it seems to me that we’re bound to start talking about rights as a matter of course; not because they’re a feature of the abstract actor, but because there are certain kinds of act that are more acquainted with the production of happiness than others, and those kinds of acts sit parallel with whatever kinds of rights that we might imagine. In other words, I see rights as an outcome of a utilitarian analysis at the level of the habit, and not as a precondition for the analysis itself. (I hope that makes sense.)