Individuals, Institutions, and Conservatism in Moral Philosophy
How many effective altruists does it take to screw in a lightbulb?
None. There are better uses of their time.
How many critics of effective altruism does it take to screw in a lightbulb?
Who knows? They’re still waiting for structural reform to take care of it.*
A few years back, the movement known as “effective altruism” began to make a splash, and over the past year there has been a series of splashing back from critics. Now wading into the muddy waters to clean things up a bit is Jeff McMahan (Oxford). He has written a short piece, “Philosophical Critiques of Effective Altruism,” in which he takes a look at how the debate has unfolded, along with one of its historical precedents—Peter Unger’s Living High and Letting Die, its review by Martha Nussbaum (Chicago) in the London Review of Books, and the follow-up to that review.
One of the central issues is whether the appropriate locus of attention is on individual voluntary actions, or on what our political, economic, and social institutions are like, with critics of effective altruism complaining that its emphasis on the former is misplaced given the urgent reforms needed in regards to the latter. In one passage on this, McMahan writes:
I am neither a community nor a state. I can determine only what I will do, not what my community or state will do. I can, of course, decide to concentrate my individual efforts on changing my state’s institutions, or indeed on trying to change global economic institutions, though the probability of my making a difference to the lives of badly impoverished people may be substantially lower if I adopt this course than if I undertake more direct action,unmediated by the state.
It is obviously better, however, if people do both. Yet there has to be a certain division of moral labor, with some people taking direct action to address the plight of the most impoverished people, while others devote their efforts to bringing about institutional changes through political action. To suppose that the only acceptable option is to work to reform global economic institutions and that it is self-indulgent to make incremental contributions to the amelioration of poverty through individual action is rather like condemning a doctor who treats the victims of a war for failing to devote his efforts instead to eliminating the root causes of war.
That some philosophers work to understand what our individual duties might be against a background of malfunctioning institutions does not free ‘the philosopher’ from trying also to understand issues of global justice and institutional reform. No philosopher I know is looking for reasons to avoid working to achieve an enhanced moral understanding. Yet if others who are not philosophers become persuaded that Srinivasan and Herzog are right that the appropriate agents for addressing problems of global poverty are communities, classes, and states, they are likely to be quite content to leave the problems to these entities and not bother with them themselves.
McMahan has more fundamental things to say about moral philosophy, too, especially about the conservatism in moral philosophy typically expressed by critics of effective altruism. These critics “repeatedly appeal to sturdy common sense, to what all right-thinking people believe, which they appear to assume is immune to rejection or revision in response to philosophical argument.” McMahan’s view is different:
I am not a fan of the rhetorical move McMahan inserts at the end of the piece:
It is salutary to recall that the early efforts of those we now recognize as having been in the vanguard of moral progress – abolitionists, campaigners for women’s rights and female suffrage, vegetarians and opponents of vivisection – have always been fiercely resisted and ridiculed by those to whom it was inconceivable that the common sense view at the time might be mistaken.
It is also salutary to recall that people on the vanguard of moral evil—European colonizers who displaced and exterminated millions–were fiercely resisted and ridiculed by their enemies, on the grounds that it was inconceivable that they had the right to appropriate land and destroy ways of life. All deep change is resisted on common-sense grounds, so that resistance itself cannot in any way serve as evidence for progress. Since we are complaining about rhetorical style: it is typical of the defenders of this movement to speak as if their opponents are on the wrong side of history, and that is no less regrettable than the deployment of Williams’ “mocking and disdainful manner”.
McMahan is not suggesting (as you say) that “resistance” is “evidence of progress.” Rather, McMahan is suggesting that we should not assume common sense already contains the truth about some moral matter. (You can tell this is McMahan’s intent by looking at the topic sentence of the paragraph in question, namely: “It might seem self-serving for me, a moral philosopher, to express skepticism about the supposition that the truth about morality is already contained in the common sense moral beliefs of ordinary people.”)
Most of all, though, it is also salutary to recall that the early efforts of those we now recognize as having been in the vanguard of moral *institutional* progress—abolitionists, campaigners for women’s rights and female suffrage, desegregationists, antimonarchists, anticolonialists, labor activists, revolutionaries on the right side—were definitely not maximizing expected utility. Donating directly to the starving surely had a higher expected payoff than a single person’s activism in pursuit of any of these major institutional reforms, just like it does today. Effective altruists would not have been at the forefront of any of these movements. That’s the whole point.
That’s at best not obvious. If you multiply a very low probability by a very high utility you can get a quite significant chunk of expected utility.
I’m not sure McMahan can make that argument, though; as he points out, the arguments for effective altruism are not exclusively utility-maximizing arguments; Singer and Unger argue that we have a duty to save lives where we can. In Singer’s thought experiment, we would surely condemn someone who refused to save a drowning child in order not to miss a ___ blank rights meeting, even if they argued that their presence at the ____ rights meeting had a one-in-a-million chance of producing an amount of utility equal to saving the lives of two million children. (And yes, that seems like a fairly silly way to think about the moral calculus, but how else are we to balance competing concerns if we are to rely on maximizing expected utility in order to make our moral decisions?)
And I’m not at all sure that “low probability of high utility” is the best way to think about what activists accomplish. Activists accomplish things by acting collectively and encouraging collective action, and it’s not clear that the best way of thinking about it is to say that each activist’s individual action has a certain very small chance of tipping things over into massive change. But McMahan (in the quote above) seems to disclaim thinking about collective action as opposed to individual action.
I don’t really disagree with any of this. My point was pretty narrow: that WP’s claim that those on the vanguard of institutional progress “were definitely not maximising expected utility” isn’t as obvious as it looks.
Matt’s second point is a big part of what I had in mind: when these movements succeed they succeed as a collective, and it’s hard to see how *a single person’s activism* contributed that much. Looking at the probability that you’re the “tipping point” doesn’t seem to get it right.
I don’t really think it’s obvious—I don’t think expected utility calculations ever are. it’s completely unclear to me how to make sense of contributions to these kinds of collective actions in individual expected utility calculations. But I don’t see any way the kind of action needed at the beginnings of a movement is going to maximize expected utility unless you’re capturing effects that are so far downwind that it’s no longer a workable way for creatures like us to decide what to do.
Most of all, though, if people *in our position* maximize expected utility by supporting immediate crisis aid rather than devoting themselves to institutional reform, I think the same’s true of the people who participated in all these movements. Unless you want to put the whole house on the efficiency of modern aid organizations radically increasing the utility of crisis aid, in which case… good thing we got women’s suffrage before oxfam came along? That doesn’t seem any better.
EA is not just about crisis aid, a point I stress in my reply to Leiter, here:
http://www.philosophyetc.net/2016/04/effective-altruism-radical-politics-and.html
I would be interested to hear more about specific proposals for “a workable way for creatures like us to decide what to do”. (Many EAs are interested in such heuristics. But merely to reject expected utility calculations is, of course, not yet to have a better heuristic in hand.)
“Should we prioritize lobbying for ‘massive forced redistribution’? Maybe! It sounds to me like the kind of proposal that EAs would be interested in considering. Of course, we’d want to hear more from reputable economists about the likely effects and possible risks (gotta be careful not to destroy the global economy, after all), and also more from skilled analysts on tractability: what our chances are of achieving such a goal. My initial hunch is that this is unlikely to be the best bet out there, for broadly ‘structural’ reasons I’ll explain below, but I’m certainly open to hearing more. (In particular, it’d be helpful to have some specific recommendations of putatively effective organizations pursuing this goal, along with their explanations of how they would use more funding, what kind of evidence they can offer that they’re a well-run organization that would use the funding wisely, etc.)”
I have to admit that I find this kind of funny. There’s, um, a big disconnect here.
The thought is this: When it comes to global inequity, we’d end up with the best results by far via major institutional reform. We can do that; we’ve done it before. But we have no “evidence based” way to go about it it, especially not as individual actors, and it would be hard to find one: It involves a huge number of people acting in disorganized ways, and there’s no straightforward way to attribute payoffs. It involves spending a lot of time on projects that don’t have a clear causal connection to the end goal. It requires major attitudinal change, which we poorly understand. And it’s very rare, so we don’t have much data.
If we’re committed to using our resources as effectively and efficiently as possible based on the best available evidence, I agree with you that we won’t be pursuing it. I take that to be a serious blind spot, and reason to think that we would not get the best results by effective altruist decision making.
Do you think that effective altruism would have directed people toward the radical activism that made major institutional moral progress possible in the past? Every time? If not, how is that not a problem?
As for how to decide what to do, I think the kind of community-based considerations that activists normally work with are pretty good—the ones that emphasize not getting your hands dirty and often celebrate symbolic victories. But I’m not the one advocating for a major change in moral decision making. (some hypocrisy here with a lot of effective altruists: “systemic change is risky! hey everyone—drastically change your moral decision making!”)
“If we’re committed to using our resources as effectively and efficiently as possible based on the best available evidence… we would not get the best results”
Assuming your latter judgment is supported by (e.g. historical) evidence, then it could feed into an evidence-based argument for typical activist behaviour. Make a rough estimate of how much good activists have done, divide it by the time and other resources that they’ve collectively committed, check whether there’s any reason to think that you’re in an unusually good or bad position to be effective as an activist (if so, weight as appropriate), and – voila! – you have the makings of a rough-and-ready estimate of the expected value of typical activist behaviour, which may or may not suggest it to be more worthwhile than, say, saving a life via donating to AMF.
It’s obviously only a very rough and tentative conclusion that could be drawn from this, but it seems more action-guiding than nothing. And if you lack the evidential basis for judging that activism is, on the whole and in the long run, likely to be more worthwhile than alternative forms of do-gooding (supported by more robust evidence, e.g. AMF, deworming, etc.), well that would just be to say that you have no basis for criticizing the standard EA recommendations, right? I mean, either your view is supported by evidence, in which case evidence-based reasoning can take it on board, or it isn’t, and we shouldn’t. But either way, complaining about the very idea of evidence-based practical reasoning just seems kind of nonsensical.
(In fairness, I take it the real worry is that many EAs have too restrictive a notion of allowable ‘evidence’ — requiring double-blind trials and such. But I don’t think that’s universally the case — again, just look at the very speculative sort of reasoning that Bostrom et al. engage in. It’s more that in the particular case of selecting between traditional charities, we’ve general evidence that most have historically not been that successful, but a few really do have great robust supporting evidence for them, so we’ve excellent grounds for preferring them over other traditional charities that don’t have any chance of being orders of magnitude more effective than AMF and the like. But more speculative evidence could be considered in cases with far greater potential up-sides. Or so it seems to me.)
I appreciate the response!
I think there’s some version of the two “core commitments” I would accept, but it wouldn’t look very much like EA anymore. I don’t think people should be reforming their moral thought to focus on maximizing effectiveness and efficiency, and so I don’t think there’s much value in proselytizing about it.
My impression of the EA/leftist debate from outside the EA sphere: A group of well off white men start the EA movement. They make specific recommendations about how to do best that are very embedded in the framework of global capitalism, and they give little thought to leftist critiques or the possibility of structural overhaul—they don’t consider it and decide it’s not’s best, they’re just not engaging with it. But they have a framework where they can always say “if that would be effective, that’s EA too!” Then critics say “you’re totally ignoring the need for structural reform.” EA activists say “if that would be effective, that’s EA too!” and continue to not engage with leftist and other political literature that’s not specifically addressed to them. Sometimes, as a bonus, they suggest their critics, often women and people of color, are being irrational.
This could be completely wrong; maybe some EA people have really extensively dealt with this stuff, and I’d really like to see that. But from the outside, I keep seeing these responses that won’t engage any deeper than what was included in a specific criticism, and that’s frustrating. There’s this weird element of “that’s the kind of thing we’d totally want to consider! … as long as it’s spoon fed to us, and you didn’t do enough.”
Also, re: “They generally seem more interested in signalling how very Left they are by complaining about how “bourgeois” or “consumerist” EA seems to them”… Effective altruism *is* super bougie. Sure, the theoretical content can bear on what people of all income brackets should do, but EA is also a movement, and as a movement, it is completely unwelcoming to low income people. Take a look at Giving What We Can or 80000 Hours and imagine you’re living close to paycheck to paycheck with little opportunity to change that. How much of their content is geared to people like you? None. Literally—none. 80000 Hours will helpfully suggest you get a dream job and the closest Giving What We Can can imagine is “students / the unemployed / the retired.” Not a peep of advice for anyone who might be facing significant economic instability themselves.
There’s obviously a reason EA is targeting well off people, but we have to be real about the fact that EA is a movement of incredibly privileged people who think inequality in the western world is not very pressing.
Well, just speaking personally, I don’t find anti-capitalist economic views substantively plausible. But that’s just a judgment about the merits of a view, not anything about EA that bars such a view from consideration in the first place. So I find it misguided when people complain that EA is somehow incompatible with taking Leftist critiques seriously. I think the more honest complaint to make would just be that most EAs aren’t personally inclined towards radical Leftism. (But of course, framed like that it doesn’t seem such a powerful complaint.)
Yes, I think most EAs would agree with you (though the replies I see also suggest they have not read very much about it), and I wish they would stop pretending like they agree with their leftist critics on the value of structural reform but have decided that other routes are more efficient.
I hope I’ve explained why taking leftist critiques seriously makes me unenchanted with EA: it is reason to doubt EA as a decision procedure, and it puts you seriously at odds with EA as a movement. And that is plenty of reason to critique EA, even if you could agree with a minimal version of the thesis, and EA supporters should engage with these criticisms instead of retreating to their minimal commitments and feeling misunderstood.
For what it’s worth, I was really into Singer and Unger as an undergraduate and it’s probably a big part of how I got into philosophy. I remember googling classified ads for people seeking kidney donors after reading Singer’s book and thinking “oh wow I gotta do this.” (I still kind of think that, but I’m not an acceptable donor.) It’s not like I’ve never given EA the time of day.
Hang on, I thought we’d established (in yesterday’s 5:42 comment) that there was no reason to “doubt EA [i.e. evidence-based altruistic practical reasoning] as a decision procedure”?
And I’m not sure what “pretending” you have in mind — it’s not as though overthrowing capitalism is the only kind of systemic reform around, after all. My blog post flagged a few alternatives that EAs tend to consider genuinely promising.
Is their lack of enthusiasm for overthrowing capitalism sufficient to put others “seriously at odds” with you? Unless you’re actively opposed to incremental improvements (deworming children and the like), then I confess I don’t really see why that would have to be so. As noted in my post, there’s plenty of internal disagreement about what the absolute best means to improving the world is; that doesn’t put the various wings of EA “seriously at odds” with each other. So I do remain puzzled by anyone’s opposition.
Well, I guess you’ve argued that, but I haven’t accepted it (see my reply to yesterday’s 5:42 comment). I want people to follow the decision procedure that will most effectively make the world a better place (second order EA?). I think that there’s good reason to think that is not EA-style evidence based practical reasoning—and that EA-style practical reasoning, at least as currently done, is going in exactly the wrong direction (more impersonal and less community oriented than people’s pretheoretic moral reasoning). I took your 5:42 comment to be largely arguing for a “second order” EA justification (from the context I thought “typical activist behaviour” included the decision procedure), but now I’m thinking I misread you.
What puts me seriously at odds with EA as a movement is (I’d hope obviously?) not that I don’t like deworming children, or that I don’t like incremental improvements. It’s that it’s *centrally* a “do this, not that” movement, and the latter half involves a fair bit of disparaging of things (incremental improvements!) that I think are incredibly important. Like when you say “First do your bit for the really important stuff. After that, you can relax, engage in local politics, sign petitions, occupy Wall Street, whatever floats your boat. But don’t pretend that the latter is an adequate substitute for the important stuff.” (http://www.philosophyetc.net/2015/05/moral-priorities.html)
“The important stuff” might be donating to crisis aid, and it might be donating to stop the robot apocalypse, but it’s definitely not trying to improve the US as a global model and actor by working with OWS to change attitudes towards income inequality. Ugh. 😐
FWIW, even as a person who shares all the EA rationalist impulses, I find it very difficult to engage with people in the movement. I can’t think of anywhere else where the fact that someone disagrees would be taken as near-incontrovertible evidence that they must be either confused or irrational.
McMahan’s piece seems to me to contain an unfortunately high ratio of sneering and tone policing to actual discussion of the critics’ arguments. Indeed, he acknowledges something like this:
“One may think that I have harped excessively on the fact that the philosophical critics of effective altruism tend to express their objections in such a mocking and disdainful manner. But this seems significant to me, as it is suggestive of bad faith. The issues that the effective altruists are addressing are of the utmost seriousness. They should not be occasions for the scoring of debating points or for displays of cleverness, rhetorical prowess, or moral exhibitionism (as when critics, while presenting their objections, pause to reveal parenthetically how much they have donated to charity despite their theoretical misgivings).”
As though effective altruists have a monopoly on moral seriousness in this debate! And their critics have a monopoly on rhetorical prowess and moral exhibitionism!
Matt – What arguments have the critics made? Srinivasan’s review is the most substantial critical piece that I’m aware of, but her central criticism — that “If everything comes down to the marginal individual, then our ethical ambitions can be safely circumscribed; the philosopher is freed from the burden of trying to understand the mess we’re in, or of proposing an alternative vision of how things could be” — is patently false. As McMahan points out, political action is one type of action we might (as individuals) decide to undertake. (Lobbying for important policy and institutional reforms — e.g. open borders — is in fact a major area of interest within EA circles.)
Richard — I can’t claim to be familiar with all the critics’ arguments. Part of the reason I’m so disappointed in McMahan’s piece is that I’d really like to see a vigorous and substantive debate about effective altruism that’s accessible to non-specialists like myself, and McMahan’s piece didn’t seem like a contribution to such a debate.
But Srinivasan alone has more arguments than that. For instance, she argues that effective altruism isn’t an effective guide to action, because of the difficulty or impossibility of the calculations that are involved (which I think you acknowledge above, in your comment about “a workable way for creatures like us to decide what to do”). She argues that the reasoning behind EA seems like it should push us far beyond the things that MacAskill actually urges us to do. She invokes the arguments that we have special obligations to those around us. Even the argument you describe as patently false is something that I think is deeper than you allow for; her point is in part that we should not neglect moral theorizing about institutions rather than treating the individual as the sole locus of moral responsibility, whereas McMahan does seem to be calling for treating the individual as the sole locus of responsibility. (I also don’t find that Srinivasan is making the claim McMahan complains about in the above passage, that everyone should only work to reform global institutions; she seems to me to be arguing that it’s permissible, not mandatory. Furthermore, I think that Srinivasan has made a point that’s along the same lines as Clough’s criticism, which McMahan acknowledges as worthy, that philanthropy that takes oppressive power structures as given can perpetuate those structures. So it’s unfortunate that McMahan won’t acknowledge that point as worthy of engagement when Srinivasan makes it.)
But I don’t want to litigate the quality of the arguments. Just to point out that the arguments do exist, and though McMahan does spend some time replying to them, he spends far more time in rhetoric and complaints about tone. For instance, his actual response to the Williamsian point is one short paragraph making the argument that by those lights we can just straight up murder people if it’s important enough to our projects. Surely it would be better philosophical practice to take a little more space to elaborate on that, forestall possible objections, etc., and a little less space to accuse the critics of simply rehashing Williams? I can think of some replies the critics might make, and I find McMahan’s piece less persuasive for failing to address them.
And about the tone: not only is McMahan throwing stones from a glass house; compare his first (or last) sentence with Srinivasan’s last and tell me which is ruder. But also, it’s unfortunate that when a group of critics who are largely not white men (except for Gray) argue that a position pays too little attention to questions of oppression, gender- and race-based subordination, etc., they are met with criticism of their tone. This is something that happens far too often–I feel like I have to say that I don’t think that this indicates that McMahan is any sort of bad person, just that we should strive to avoid falling into tone policing in circumstances like that, if it’s at all possible.
+ 1 Matt Weiner
Suppose an impartial arbiter were to read everything that’s been written on effective altruism, for and against, and to tally up the instances of deliberate “displays of cleverness” or (attempted) rhetorical prowess or “moral exhibitionism”. I’m very doubtful that one side of the debate would come out of this exercise looking better — more attuned to the utmost seriousness of the issues — than the other.