Effective altruism is a political nightmare
I find it genuinely tragic that posting an explainer feels justified.
Update: This post prompted a podcast with Rabbithole Investigations—give it a listen!
Ever since that really rich nerd lost a bunch of money, it seems like various corners of the internet are trying to make sense of effective altruism. A very short summary of EA is that a bunch of elite academics and billionaires in wealthy imperial nations apparently woke up one day and realized: “Wow, a lot of children are dying or don’t have enough to eat! That seems bad.”
Philosophy and economics is the nightmare combination. Then they added crypto billionaires to the mix.
Analytic philosophy meets development economics
I first encountered EA in a philosophy classroom. I took an ethics seminar with Peter Unger at NYU, whose most famous contribution to ethics is his 1996 book Living High and Letting Die. For a touch of color, Unger was the kind of pedagogue who kicks his feet up on the seminar room table and begins sentences with iconic clauses like “It’s not politically correct to say this, but…” To Unger’s credit, the book has the most amazing trolley problem illustration in the history of trolley problem memes.
That class was different from other philosophy classes at NYU in the sense that it was mostly concerned with statistics. In many class sessions, Unger would consult WHO reports to rattle off child mortality rates in the Global South. Yet unlike e.g. my Russian or South American literature classes, Unger’s class wasn’t really concerned with macro-scale structural phenomena like “capitalism” or “colonialism.” (Might child mortality rates be related to capitalist nations’ extractive and expansionary tendencies? Effective altruists don’t care.)
Much like other ethicists, Unger was concerned with arguments that can motivate individual action—his book’s thought experiments focus on why people in wealthy nations feel more comfortable ignoring a letter from UNICEF than ignoring a child drowning in a pond, despite the fact that a UNICEF donation could “save more lives.”
This was a development of Peter Singer’s famous article “Famine, Affluence, and Morality.” Singer has since gone on to more explicitly argue for effective altruism. In The Most Good You Can Do (2015), Singer argues for “earning to give,” which is the most mind-numbingly individualist framework imaginable. In short, Singer says that because kids are dying in the Global South, people in wealthy nations should be pursuing extremely high-paid professions like working at a hedge fund. The income from gigs like that could then be donated to charity. In sum: The more money you make, the more money you can donate. Singer frames this as a pragmatic response to global poverty that individuals can pursue immediately. See e.g. his Reddit AMA:
This is the kind of consequentialist calculation that lies at the core of effective altruism. Of course, philosophers aren’t personally inclined towards the quantification of social utility. Traditionally, that is the domain of economists. As sociologist Elizabeth Popp Berman has taught us, economics has made the “cost-benefit analysis” central to American politics. And that’s tragic.
Quantifying “cost-effectiveness”
GiveWell is the institutionalized form of this consequentialism. In the interest of charitability (ha ha ha) I will say that GiveWell was created in response to a real problem—international NGOs like Red Cross genuinely do set donors’ money on fire. GiveWell is a “meta-charity” that evaluates the cost-effectiveness of charitable organizations, founded by some hedge fundies in 2006.
Historically, GiveWell has promoted organizations that combat malaria. I’m not going to argue against funding those organizations. I’ve donated to Against Malaria myself.
But the evaluation framework is fascinating from a social science perspective. GiveWell was created to serve a bunch of wealthy people who want to make tax-deductible donations while understanding the efficacy of those donations in the same way they understand their own work. Again, this org was created by hedge fundies. Matthew Bishop and Michael Green coined the term “philanthrocapitalism” to describe this social phenomenon: “the new generation of billionaires […] are using big-business style strategies and expecting results and accountability to match.”
In the above screenshot, you can see that GiveWell’s recommendations are “supported by 40,000 hours of research annually.” Who are these researchers? Economists, mostly.
The cult of RCTs
I’ll spare you the suffering of going through GiveWell’s entire evaluation methodology. But here, effective altruism dovetails with another damaging phenomenon, the cult of the randomized controlled trial (RCT) in social science methodology and causal inference research design. As GiveWell staffers wrote in 2012: “We believe that RCTs have multiple qualities that make them – all else equal – more credible than other studies.” Why, though?
Sociologists Luciana de Souza Leão and Gil Eyal have a lovely article documenting the meteoric rise of RCTs in the early 2000s. But this wasn’t the first RCT movement! Another effort was made in the 1960s-80s, but it failed. So, they ask: why did the new movement succeed? Was it because the second movement more successfully explained the intrinsically superior scientific properties of RCTs?
Of course not. The first movement failed because of political resistance. Development aid programs previously relied on local governments to implement them, and RCTs necessarily involve withholding material benefits in order to create an experimental control group. It turns out that if you’re offering malaria netting to the treatment group, local program administrators think that the control group deserves it, too!
To solve this “problem,” the second wave of randomistas (academic economists and international development aid organizations) joined forces. Global NGOs are fairly new, and they aren’t accountable to the communities in which they administer programs: they do not “have to pretend to serve everyone.” Instead of local nonprofits or government organizations, the randomistas turned to global NGOs, banks, and survey firms, who don’t have to worry about political upheaval. Of course, RCTs are expensive to implement, which is why you need the philanthrocapitalists, who themselves are bewitched by the “mechanical objectivity” of RCTs as an evaluation method.
For their part, young economists were drawn to the development subfield (and RCTs) because they were working in the shadow of the “credibility revolution,” which basically meant searching for an instrumental variables identification strategy. I work on some IV stuff myself, and let me tell you, it is extremely tenuous and limiting. I’m sure it’s much worse for economists, though: “RCTs are a much narrower' ‘golden ticket,’ namely a hall pass that shields a young scholar from [… a] situation where the consensus underlying disciplinary objectivity has collapsed.”
This was a quick gloss of their article. You should read it in full if this piques your interest.
Today, if you are doing any sort of quantitative social science research, you’ll quickly discover that RCTs are (wrongly) considered the “gold standard.” This raises a separate argument, which I highly recommend to anyone interested—Nancy Cartwright and Angus Deaton have written a lot of great stuff on this. For people who don’t care about that, what I’ll say is that RCTs are definitionally narrow. They measure one thing being changed for a specific group at a specific point in time. By contrast, e.g. the Green New Deal or prison abolition would change a lot of stuff, all at once. The applied economists cannot cleanly measure the effects of big-picture change. (Hmm, I wonder why social scientists advocate incremental policy changes?)
The funny part, for me, after learning all of this is that the analytic philosophers were sort of a late entry to the party, providing a moral imperative to join forces with EA through shame. “It’s so messed up that you ignore the letter from UNICEF, don’t you care about the starving children? Why aren’t you donating more to Against Malaria?” It’s like the academics tried to graft morality onto the Faustian bargain between development economists and philanthrocapitalists. But you, dear reader, should not be seduced by their professional advancement strategy.
We see this deception pretty directly in the Bankman-Fried text messages that have been making the rounds. “So the ethics stuff - mostly a front?” “Yeah. I mean that’s not all of it, but it’s a lot.”
Who cares?
Anyway, why are people talking about EA all of a sudden? Obviously the crypto nerds are one reason why. Bankman-Fried’s trading firm was composed of a bunch of effective altruist acolytes, and a bunch of their proceeds allegedly fed back into EA-approved charities. I’m not really interested in speculating about the personal-psychological incentives here, but a fair way to sum up these recent developments is that EA was always appealing to people with way too much money. The crypto people were just new to the scene, and EA was a ready-to-hand cover tactic.
Another reason why EA is in the news, it seems, is that Bankman-Fried was bankrolling media organizations. So these outlets are writing about effective altruism because the EA hivemind is paying them. “Of course they’re newsworthy, they pay my rent!”
I only felt the need to speak up about this because I actually did feel convinced by some of Unger’s arguments back at NYU in 2015. Notably, though, 2015 was also when the first Bernie Sanders POTUS campaign occurred. I’m grateful for that coincidence. Without a political framework to address the harms of capitalism, I may have bought Singer and Unger’s arguments—I could have tried to be a software developer and “earn to give.” But I didn’t. I organized a benefit concert for Bernie and canvassed. Of course, Bernie still lost, and lost hard. But something I learned in the process is that taking action to fight capitalism is something you do with other people. As sociologist and abolitionist Brendan McQuade puts it: “How we will share labor to meet our shared needs?”
I wrote this essay to provide a counter-argument for people with vaguely left or liberal political inclinations. Of course, this doesn’t mean that paying your rent through a nonprofit (like I did before), a university (like I do now), or a political campaign means you’re Doing the Most Good™. This is a separate argument, but I think radical work tends to be unpaid. People need to eat, and that’s what funded work is for.
The point I’d like to leave you with here is that the effective altruists have the first part right: Global poverty is fucked up, and must be urgently remedied. But tax-deductible donations to NGOs are, at best, a band-aid on an arterial wound. At worst, the “earn to give” ethical imperative actually encourages individuals who are alarmed by capitalist crises to reproduce the world-imperial capitalist system. EA tells people who aren’t already rich that they need to get rich through the same structure that causes child mortality in Sub-Saharan Africa. And if you’re already a billionaire, EA gives you an intellectual, quantified veneer of defensibility: “Look, I’m not evil! Look how much I donate to these cost-effective charitable organizations!” (It seems like other philanthropic orgs see this kind of thinking as so threateningly effective that they’re mounting counter-arguments.)
What about criminal justice reform?
Another reason why I care about this stuff (and why I think you should care) is that criminal justice reform nonprofits are primarily funded by philanthrocapitalists. For three years after I left NYU, most of my salary at two different organizations was paid by Arnold Ventures, which is actually a perfect microcosm of all of this stuff rolled up into one organization: a hedge fundie who thinks they have the answers to reform capitalism, an institutional structure facilitating “social investment” and evaluation, and the cult of RCTs. “Arnold Ventures” itself is a rebranding from “The Laura and John Arnold Foundation,” matching an institutional shift from a charitable foundation to an LLC. They take the social investment ethos so seriously that they changed their institutional structure accordingly.
I’m not going to try and provide a political history of Arnold’s criminal justice portfolio here, but I will note that they were heralded by GiveWell affiliate Open Philanthropy as a leader in the criminal justice reform philanthropy space. Some other major funders in the space are the Koch brothers (yes, really) and the Ford Foundation. The earliest resource I know of on this topic is INCITE!’s The Revolution Will Not Be Funded. Kay Whitlock and Nancy Heitzeg also have a new book that explicitly provides the political history you might want, Carceral Con. I also strongly recommend this historical treatment of Ford’s involvement in the #CloseRikers campaign.
I’m not pointing this out to say, “The philanthropists are evil!” I’m personally implicated in this stuff, as I’ve pointed out, as is my professional network. I merely want to point out that the kinds of arguments and social-scientific practices that are used to persuade billionaires to donate money are not the same kinds of arguments and practices that should inform political organizing against criminalization. For a really specific example, the Houston Institute has a poignant argument about how Arnold-funded RCTs can be socially damaging in the criminal justice reform context. The short version is: Why do we deprive the “control group” of the benefits of a research intervention when we know they will benefit from it? Well, because the philanthrocapitalists insist on an RCT, and the intervention won’t be funded without that initial deprivation.
The takeaway: We don’t need to limit ourselves to narrow, incremental interventions that can be measured by Ivy League social scientists (myself included). We don’t need to wait for the program evaluations to be conducted. We can start building the future we want, together, right now. The social scientists, program officers, and philosophers are welcome to describe, evaluate, and opine from the sidelines.
If effective altruism is encouraging people to donate a kidney, then sure, whatever, good for them. I’m not totally sure why wealthy people (myself included) needed to read an analytic philosophy argument or an economics paper to justify altruistic behavior. Ultimately, though, the kind of action that is required to truly address the harms of capitalism is collective and political. As I’ve written before, the question is not “What’s the most good I can do?” The question is: “What’s the most that we can do for each other, together?”
***
Update 11.21.22: A Twitter mutual pushed back on my RCT section here:
I wasn't persuaded by your methodological critique, since I felt like you didn't engage with the model's obvious virtues: It's potential to disambiguate causality by isolating a variable. How can alternative designs deliver this same good?
I responded with the following argument, which they found persuasive. I thought it might be interesting for future readers:
There are plenty of credible causal inference research designs, e.g. instrumental variables, regression discontinuity, difference-in-differences. All of these aspire to the genuine treatment manipulation in RCTs through "natural experiments." They come with extra assumptions and are more complicated from a modeling perspective, but can be applied in many more contexts than RCTs. However. All of this causal inference stuff, RCTs included, tends to focus on "internal validity," while what policy analysts care about is "external validity" - how will this intervention affect a different population at a different point in time? RCTs tell you literally nothing about this. Really the argument is "RCTs are no better than other very limited causal inference tools, but emphasizing them leads to research ethics quandaries surrounding equipoise" which leads to the deprivation stuff I was talking about in the piece. Particularly in a political advocacy framework that requires evaluation before $$$/political capital investment.
***
You might be thinking—okay, Jon, that’s all fine, but what kind of political stuff do you even want us to be doing? I’d argue: Americans should be laser-focused on reparations, and in that framework things like the Green New Deal and Medicare for All are a great start. I’m a PIC abolitionist, so that’s also a good thing to read about. Micah Herskind has the most extensive abolitionist reading list.
You don’t have to share my politics to feel convinced by my argument, though. All I’m saying is that political praxis demands collective action, not (just) tax-deductible donations. That might mean doing get-out-the-vote work for a politician, organizing with an abortion fund, or doing jail support.
Man, thanks for writing this. Your thoughts are always so organized and easy to follow, I really appreciate it!
The only thing I really knew about EA until recently was the GiveWell angle. A lot of charities skim too much off the top, a lot of charities target feel-good causes that are over-funded leaving more urgent-yet-boring causes underfunded, give your money wisely. I grew up poor and only recently started receiving a really good salary once I began my current career, so I wanted my newfound ability to donate to charities to go as far as possible, so I would've considered myself an effective altruist until recently. I didn't realize a lot of the worse aspects of EA that you point out in this.
Do you still consider the evaluation of cost-effectiveness of charities a useful tool? Are there organizations that don't rely on RCT that someone like me could use instead? I'm currently decently politically active back home (mostly in the lgbtq+ community, though I'm really interested in getting more into prison abolition once I return stateside), and am specifically working in the field of renewable energy engineering to try to do my own little part to "make the world a better place", so I do appreciate that there are more dimensions to living a moral life than donating money and feeling smug about it. But I still want whatever money I do give to go as far as possible, and appreciate the EA ethos on that front. Or am I still missing the point?
Thanks again!
If you think participating in US politics is the best way to do the “most that we can do for each other, together”, I have some oceanfront property in Nebraska to sell you. Of course we should vote and canvas for progressive candidates, and billionaires like SBF certainly don’t make their fortunes ethically, but to suggest that EA is somehow wrong to advocate for people to donate to effective causes is ridiculous.