- Nikhil Venkatesh
A Socialist Guide To Effective Altruism
Effective altruism is a movement that believes we should help others as much as possible. This may sound like a bland statement of the morally obvious. But one does not need to spend long in this world to see that it is nothing of the sort. Most people spend most of their time looking out for themselves and their families. When they do try to help others, most people have a bias towards the cute, or the nearby, or the well-advertised, or the traditional, or those who look like them, even when they could help others more. Many charitable and political organisations who claim to help people fail to do so. The effective altruist slogan ‘do good better’ is a response to these problems.
These are problems of which those of us on the left are well aware. And there is significant overlap between effective altruist and socialist concerns. Effective altruists have directed funds to some of the poorest people in the world; they oppose factory farming and nuclear war; they worry about the power of big tech firms. They would like to see more public investment in education, foreign aid, and pandemic preparedness. Most effective altruists describe themselves as left-of-centre. In spite of this, effective altruism’s relationship with the left has been fraught, as shown by several critiques from the left, including a recent one in these pages.
In what follows, I try to understand where such criticism might come from, and whether a reconciliation between socialism and effective altruism is possible.
Linsey McGoey’s recent critique in Jacobin was triggered by the surfacing of an email in which Nick Bostrom, one of the movement’s thought-leaders, endorsed racist views about intelligence. This is not the only effective altruist scandal of recent times. Sam Bankman-Fried, billionaire CEO of cryptocurrency exchange FTX, was a vocal effective altruist who had pledged to give away his fortune to effective charities. He is currently facing fraud charges, whilst it has also been alleged that his use of money was not always very effective or altruistic: his investments including a Bahamas mansion, and political action committees aimed at stopping democratic socialists.
It would be easy to write off effective altruism as irreparably flawed by association with such characters. However, it would be wrong. For one thing, their awfulness has been recognised by many within the movement. For another, this kind of dismissal is not something that socialists should practice. We do not write off Marx’s theories of history and alienation on the grounds of his personal flaws, nor communism for the crimes of murderous leaders who claimed its mantle. We do not refrain from showing solidarity with Palestine because such causes often attract anti-Semites. Rather, we ask whether these ideas are useful, and whether these causes are just.
We should also interrogate the question of why our traditions can sometimes attract bad people or serve to justify bad actions. In the case of effective altruism, the sources are probably threefold. Firstly, it has attracted a number of people from the ‘rationalist’ communities of the early internet which, alongside some interesting contributions, were hotbeds of reactionary social views dressed up as ‘facts and reason’. (Bostrom’s racist email came from one such group, ‘the extropians’.) They are attracted to EA, I think, for its tendency to discuss unorthodox and sci-fi ideas, its numerical approach, and its offer of a place from which to feel superior to ‘social justice warriors’ even on what is supposedly the latter’s home turf: helping others. Secondly, as McGoey rightly notes, effective altruism’s social base is wealthy white men in elite Anglophone universities. No surprise that such a movement has racism, misogyny and general douchebaggery in its ranks. Thirdly, effective altruism can justify doing bad actions (such as cryptocurrency fraud) in pursuit of good ends. No surprise that such a movement attracts some bad actors – of course, a similar problem afflicts all political movements that have ambitions to change the world, including those of the left.
However, the majority of effective altruists are not Bostroms or Bankman-Frieds, just as the majority of socialists are not Pol Pots. The majority of both groups are people who want, very deeply, to make the world a better place.
That said, there are some significant problems with effective altruism from a socialist perspective. Effective altruists are often criticised for their tendency to quantification. The mostly excellent socialist comedy podcast ‘Mandatory Redistribution Party’ accused them of:
“trying to cram the entire panoply of human experience into discrete Goodness or Badness points. This exhaustive project, once complete, would allow any of us to make any kind of moral, judicial or practical decision just by smashing out an abacus and flicking the beads around a bit. A small bowl of potpourri? Five goodness points. Asbestos? Fifty Badness points. Simple.”
Whilst it is a caricature, the effective altruist approach makes quantification hard to avoid. If we are committed to helping others as much as possible, we need to have an idea of what helps more and what helps less. From here, it is a short step to using numbers to describe the moral value of outcomes.
Socialists too find the need to quantify. As the founder of Britain’s National Health Service, and grandfather of the Labour Party left Aneurin Bevan put it: “The language of priorities is the religion of socialism.” Investment in public housing is more important than cutting taxes on the rich. Education is more important than accumulation. Redistributing wealth would help those at the bottom more than it hurts those at the top. These are all implicit quantifications. In effective altruism, though, the numbers are too often fetishized, giving the impression of precision and objectivity whilst losing sight of what they represent: people’s lives and cares and joys and suffering.
The need for precise numbers leads to a bias towards the measurable. Effective altruists often recommend giving to charities, such as the Against Malaria Foundation or GiveDirectly, whose work is relatively simple and easy to quantify, who have the resources to run randomised controlled trials, and the Western-educated staff to sell the results. As Anthony Kalulu , founder of the Uganda Community Farm, puts it, this makes it very hard for grassroots organisations, led by the poor, pursuing more holistic solutions to poverty, to be classed as ‘effective’. The same goes for political interventions (perhaps explaining McGoey’s reports of EAs working on global poverty without having heard of the IMF). As Amia Srinivasan asks, “What’s the expected marginal value of becoming an anti-capitalist revolutionary? To answer that you’d need to put a value and probability measure on achieving an unrecognisably different world… how would we go about quantifying the consequences of radically reorganising society?”
Moreover, when effective altruists move away from the easily measurable, they tend to direct their focus not to ambitious grassroots projects or radical political interventions, but to a particular set of highly uncertain but potentially highly impactful areas. If we manufacture artificial general intelligence, that could be very very good (fully-automated abundance, no work, no needs, no suffering) or very very bad (AI weapons with greater destructive power than hydrogen bombs, robots enslaving humanity). It might be very unlikely that we get such technology, or even that, if we get there, either of these scenarios occur. The values of those scenarios, and their likelihoods, are almost impossible to measure. But the potential payoffs are so high and so low, that many effective altruists believe their contribution to helping others ought to come through working in the AI sector, to try to push the dial one way rather than the other.
AI development is surely important for a great number of reasons, and socialists should recognise that – like the nuclear technologies of the twentieth-century – its immense potential makes it too risky to be left in the hands of profit-seeking capitalists and oppressive states. However, we might wonder why effective altruism finds grassroots charities and radical political change too difficult to measure, but happily funds highly speculative bets on artificial intelligence. Once again, one suspects the answer has to do with its social base: white, male, likely working in tech already, and more likely to be fans of science-fiction than of revolutionary politics.
A second reason for socialist wariness of effective altruism is its relationship with money. Some of effective altruism’s first headlines centred on its advice that people should ‘earn to give’: taking a high-paying corporate job and donating most of your salary would do more good, they claimed, than working in a charity, or in public service, or engaging in political activism. This was at least part of the motivation for Sam Bankman-Fried’s career. For many socialists, the idea that investment bankers might do more good than community organisers is repulsive enough for us to dismiss EA. Effective altruists these days are less invested in earning to give – but this is partly for reasons socialists might be even more put off by: many effective altruist causes are no longer ‘funding constrained’ because they are fully funded by billionaires.
Earning-to-give, and accepting funding from the very rich, should not be totally anathema to socialists. If Engels had not used his money from the dirty, exploitative cloth industry to support Marx’s work, our own tradition would have lost its most important ideas. However, socialists recognise that such sources of funding are highly problematic. It gives power within the movement to the already powerful, undermining democracy. It alienates those who – with very good reason – dislike the presence of very wealthy people in a world of poverty. There is a risk of capture, whereby the interests of your funders replace your moral purpose, either because you adapt your work to please them, or because they direct funding towards those whose views they share: here is another explanation for EA’s lack of radical politics. Effective altruists should be more attentive to these risks. (That I myself received funding from effective altruists during my graduate studies proves that at least sometimes they let a Marxist slip through the net.)
The earning-to-give controversy points to what is, in my view, the biggest problem with the effective altruist philosophy. Effective altruists tend to think in individualistic terms: given what others will do, how do I act such as to make my impact as good as possible? Perhaps for some of us the answer is to make a lot of money in finance and give it away. Capitalism isn’t going to go away, after all, just because you turn down a job at Goldman Sachs, but a case of malaria could be prevented simply by you buying some bednets. The socialist asks, instead: what can we – my class, my party, my movement, my generation – do, together, to make the world a better place? When we think in these collective terms, many more options become available. If we all acted together, we could overcome capitalism, we could stop climate change, we could tackle the root causes of poverty, inequality and ill health rather than simply treating the symptoms one by one.
Replacing earning-to-give as the headline-grabbing effective altruist view is something called ‘longtermism’, expounded in Will MacAskill’s recent book ‘What We Owe The Future’. The basic idea of longtermism is simple. You want to help people. There’s going to be a lot more people in the future than there are today: ten billion will live this century, but trillions or more could live between now and the end of human history. Therefore, to help people the most, you should focus on doing things that improve the long-run future.
What sort of things? Some longtermist suggestions will be familiar to those on the left. We should invest in green technology, since climate change threatens terrifying long-run damage. We should prevent nuclear war and contain pandemics, as these ‘existential risks’ threaten the very existence of a future for humanity. More idiosyncratically, many effective altruists think longtermism makes the stakes around artificial intelligence even greater. At the weirdest fringe of longtermism are the ‘transhumanists’ who believe in a future in which we use technology to transcend our biological limitations.
Many leftists are suspicious of longtermism (see a rare exception here, and the more common view here.) The most common criticism one hears is that we should focus on relieving the suffering of people now, before we look to helping future generations. Stated like that (and setting aside transhumanist futures), I struggle to see the difference between this view and that common amongst right-wing critics of foreign aid, that we should help people at home before helping people abroad. If we take seriously the idea that future people have equal value to us – if we refuse to discriminate against them simply because of when they are born – their interests might sometimes outweigh those of present-day people, just as we recognise that sometimes some people should be taxed to provide for others. Moreover, as socialists, we advocate our programme, very often, on the grounds of the better lives it promises for future generations: the short-term costs of a revolution might be justified by the great benefits to come, for instance. The weakest point in longtermism, to my mind, is rather the huge uncertainty over how our actions affect the far future.
What attitude, then, should socialists take to effective altruism? I think it should be one of openness to mutual exchange and allegiance – for instance, on issues such as climate, and nuclear weapons – whilst retaining a critical view of EA’s individualism, its links with capital, its narrow and elite social base, and the reactionaries in its ranks. After effective altruism’s recent scandals, more people within the movement have become alive to these problems themselves, and socialists should hope that they win out in internal debates.
Lastly, socialists should be confident of being able to counter effective altruism’s regressive tendencies without abandoning what we share: a commitment to helping others as much as possible. Many effective altruists like capitalism. This is not one of their core commitments: ‘help others as much as possible’ need not imply supporting capitalism. It stems from their belief that capitalism is the best system for helping people live better lives – a belief that is easy to come to with the social and financial bases that effective altruism enjoys. As socialists, we reject this belief, and we should aim to convince effective altruists of our view. We do not think that a system in which wealth is concentrated in the hands of a minority, and which depends on exploitation and selfishness maximises human well-being. We believe that a better world is possible. All those who want to help others should be encouraged to join us in building it.