A philosophical introduction to effective altruism

Hello Will,

Thank for linking the article, which I enjoyed. I'll jump straight to the three objections I have. I'll first state these briefly and then explain them in greater detail. First, the Duty of Beneficence, as you state it, seems both ad hoc and under-demanding. Second, your first argument for Maximising Beneficence does not provide the support you claim it does. Third, insufficient explanation is offered for the relationship between, one the one hand, assessing problems by their scale, neglectedness, and tractability and, on the other, cost-effectiveness. I'm (obviously) an EA enthusiast, so I offer these in the spirit of trying to improve the arguments for EA.

1. The Duty of Beneficence

You state "Most middle or upper class people in rich countries have a duty to make helping others a significant part of their lives" but why is it the case that "Most middle or upper class people in rich countries" have the duty, rather than everyone? Suppose if I'm rich but in a poor country, or poor but in a rich country. Do I have no duty of beneficence? That seems implausible. What if I am rich and currently live in a rich country, but move to a poor country - does my duty of beneficence disappear?

Let's call your version The Middle-Class-Rich-Country Duty of Beneficence (MCRC) and distinguish that from the General Duty of Beneficence: all individuals have a duty to make helping others a significant part of their lives. As you don't defend the stronger, General Duty, I assume you're taking the position that there is no General Duty, just MCRC.

I suppose you're taking this as an argumentative strategy (rather than you actually believe it) and doing this because you assume your readers would accept MCRC but not the General duty: presumably many people think that, at the very least, those who are globally fortunate ought to help others. Your (clever) move is to point out the reader is themself likely part of this elite.

One problem with this move is that MCRC is arbitrary. I do not think you provide an justification for this particular specification, nor do I think there is one. Second, as noted, it is grossly under-demanding to those who are not both middle-class and in rich countries.

I worry the response to your argument from those who dislike Effective Altruism will be to claim 1. MCRC is implausible, 2. note you haven't provided an argument for the General duty and thus 3. assume there is no general duty. I wonder if the better strategy is just to argue there must be a general duty to benefit others if the costs to us are small (a la Singer in Famine, Affluence and Morality). You would then note the costs to the world's rich are clearly small relative to the benefit they would give to others.

2. Maximising beneficence

Your first of two arguments for maximising beneficence is an appeal to cases. You write:

Suppose that, as a volunteer doctor in a resource-starved hospital in a poor country, you can do one of two things with your last day of work before you return home. First, you could perform surgery on an elderly man with prostate cancer, thereby saving his life. Or you could treat two children from malaria, thereby saving both their lives. If you had a personal attachment to the cause of fighting prostate cancer, would that give you sufficient reason to save the the life of the elderly man rather than the two children? Clearly not. The importance of saving two lives rather than one, and of saving people who have much more to gain from their treatment, clearly outweighs whatever reason a personal attachment might bring.

Immediately after you state:

Yet this is morally analogous to the decisions that we actually face when we try to use our resources to do good. The only way in which it is morally disanalogous is with respect to what’s at stake. (emphasis added)

One important way in which the case is morally disanalogous is that it is about a doctor performing their duties. We often think people, acting their professional roles, have duties that do not apply to individuals acting their private lives. To highlight this, I imagine many people have the intuition that if this same doctor were to run a marathon for charity, that doctor would be morally permitted to fundraise for whatever cause they wanted, not just whatever cause saves lives most cost-effectively (or, more broadly, does the most good). When running the marathon, the doctor is a merely a private citizen and, qua private citizen, they can do good any way they want. One could accept the doctor must save the two rather than the one while denying individuals are generally required to maximise beneficence.

Potentially, using a case where someone, acting a private citizen, can save either one life or two at the same cost - e.g. you stop the trolley killing two people rather than one - would have some the right intuitive pull whilst avoiding the objection it was morally disanalogous.

I think there might be an objection to your second argument for maximising beneficience, but I haven't been able to formulate it yet.

3. Scale, neglectedness, and tractability

You raise the problem of comparing effectiveness across different problems:

This is a legitimate concern, and effective altruists have developed an alternative heuristic framework for prioritising among causes, including when the impact within some of those causes is difficult to measure. According to this framework, the following factors are indicative of which causes are highest-priority [scale, neglectedness, tractability]

My concern is that in this essay you don't offer an explanation to the reader as to why they should find it plausible that "the following three factors are indicative of which causes are highest-priority". Why and exactly how should I use scale, neglectedness and tractability to compare farmed animal welfare to global health and development, two causes effective altruists find high priority?

You do provide a citation to your own book, Doing Good Better, but (as you and I have discussed before) it's not very clear from that book either (a) why scale, neglectedness and tractability are indicative of which causes are highest priority or (b) how precisely those factors can be used by individuals to determine what the priorities are.

The most plausible precisification of the framework is the one offered by Owen Cotton-Barratt here, where the three factors are multiplied together to give marginal cost-effectiveness (you mention this version of the framework in your Palgrave article). I presume you didn't want to get into the nitty-gritty in this essay, but I imagine some readers will be left confused about how the framework you mention in this article does the work of comparing causes. My only suggestion here is that you could consider also citing Owen's and/or 80k's explanations of the framework.

From my quick read of your Norton Introduction, it seems like you're arguing for moral realism being a prerequisite to EA. (Words like "duty" and "command" make me think this.)

Is that right?

Hey Will,

As usual, I’m a huge fan of this work of yours. Thanks for your contribution!

Writing because the argument for the Duty of Beneficence felt a little underdeveloped here. Not that there aren’t strong arguments for it, but that it needed some additional advocacy in this piece.

I’m imagining sitting down to a dinner table and being told, “You really shouldn’t eat with that fork” and incredulously responding with, “Why not?” If the response to my question is, “All of the different authorities on manners (e.g., Emily Post, etc.) all say you shouldn’t use that fork,” I’d be rather unconvinced. When I reply, “Well I just don’t care that much about manners,” the argument-from-agreement-of-different-manners-systems is not compelling to me.

I think it’s worth developing the meta-ethical warrant for the Duty of Beneficence in this piece for both philosophical reasons but also practical ones. A college student considering either buying a new iPhone or giving $1000 to charity is unlikely to be convinced by the claim that, “All of the various fans of morality all agree that you should give the money to charity, not buy a new phone.” I know this isn’t a project in meta-ethics, but I suspect that addressing this issue in motivating the Duty of Beneficence might make the piece itself more effective.

Thanks for your thoughts!