The main tension to the [effective altruist] movement, as I see it, is one that many movements deal with. A movement that was primarily fueled by regular people – and their passions, and interests, and different kinds of provenance – attracted a number of very wealthy funders [and came to be driven by] the funding decisions, and sometimes just the public identities, of people like SBF and … a few others.
Benjamin Soskis, The Guardian
There’s a near consensus that [effective altruism] needs funding diversification but with Open Phil[anthropy] accounting for ~90% of [effective altruist] funding, that’s just not possible due to some pretty basic math.
Marcus Abromovitch, “Funding diversification for mid-large ea organizations is nearly impossible in the short-medium term“
1. Introduction
This is Part 8 of my series on billionaire philanthropy.
Part 1 introduced the series. Part 2 looked at the uneasy relationship between billionaire philanthropy and democracy. Part 3 examined the project of patient philanthropy, which seeks to create billionaire foundations by holding money over a long period of time.
Part 4 looked at philanthropic motivations. Part 5 looked at sources of philanthropic wealth. Part 6 looked at extravagant spending.
Part 7 began a discussion of donor discretion. Today’s post rounds out my discussion of donor discretion by discussing three ways in which donor discretion combines with a lack of funding diversification to pose epistemic and practical challenges to the effective altruism movement.
In writing this post, I find myself in the odd position of saying something that many if not most effective altruists already agree with. I think that is a good sign, not only that these views are on the right track, but also that many effective altruists are committed to doing better in this regard. It is good to seek agreement when possible, and if it turns out that we agree on many of these views, so much the better.
2. Epistemics and finance
Here is a story that the effective altruism movement would like to make true of itself. Cause-neutral prioritization research combined with widespread community reflection on the results of that research lead to shared beliefs about the relative priorities of philanthropic interventions. Those beliefs are then translated by funders into tangible donations. Call this the epistemic story.
Here is another story. Funders have money. Organizations and researchers need money. In a market dominated by a small handful of funders, changes of opinion by one or a small handful of key decisionmakers lead to radical changes in the allocation of funding between causes. This, in turn, leads researchers and organizations to form or pivot to address newly fundable causes. The resulting increase in visibility, prestige, and quality of work on these causes leads community members to conclude that they are worth funding. Call this the financial story.
As is so often the case, the truth likely lies comfortably away from either extreme. Certainly I would not go so far as to suggest that the financial story explains the vast majority of cause prioritization. But I would also not be so quick to dismiss the financial story. Nor would I be the first commentator to suggest this story. For example, the historian Benjamin Soskis writes:
The main tension to the [effective altruist] movement, as I see it, is one that many movements deal with … A movement that was primarily fueled by regular people – and their passions, and interests, and different kinds of provenance – attracted a number of very wealthy funders … [and was then steered by] the funding decisions, and sometimes just the public identities, of people like SBF and Elon Musk and a few others.
There is, moreover, a neat way of telling the financial story through the recent history of effective altruism. Neat stories are a dime a dozen, so the mere existence of a plausible story should not be accounted definitive proof. But it is, I think, at least suggestive. Here is the story.
The effective altruist movement has had roughly three waves. The first, dominated by medium-sized donations steered by organizations such as GiveWell and Giving What We Can, favored short-termist interventions such as global health and anti-poverty work. The second, dominated by Open Philanthropy and the FTX Future Fund, favored a broad spectrum of longtermist causes. The third, dominated by Open Philanthropy, focuses predominantly on challenges and opportunities posed by artificial intelligence.
The first two of these waves can be easily seen in existing funding data. For example, here is Tyler Maule’s analysis of funding by cause area from 2014 to 2022:
Here, we see a sharp turn towards longtermism. There is also a corresponding change in the predominance of funders. Here again is Maule’s analysis:
I don’t have sufficient data to support the third-wave story about a rapid increase in AI-related spending, though I suspect that such data could be found. I can support the third-wave claim about funder dominance. For example, Marcus Abromovitch estimates that in 2023, Open Philanthropy gave 691 million dollars, GiveWell gave 318 million dollars, and no other funder exceeded 50 million dollars.
Some degree of coincidence between funding trends and community beliefs and actions is to be expected. After all, funders share many beliefs with community members and may be moved by the same arguments as other members of the community. At the same time, the strength and rapidity with which funding trends have tracked community beliefs and organizational mission statements suggests a nontrivial degree of truth to the financial story.
This point is particularly visible in the rapid rise of the FTX Future Fund, which officially proclaimed that short-termist interventions were off the table. With a large portion of philanthropic funding now tied from the outset to longtermist projects, any number of longtermist interventions began to pop out of the woodwork. No doubt some of these were meritorious projects that would have been carried out in any case, but some seem more likely to have emerged due to the existence of funding. Did the world desperately need lavish fellowships in the Bahamas, business coaching for effective altruist organizations, a longtermist TV show, and sizable property purchases alongside costly renovations?
3. Signal leadership
Another challenge posed by the dominance of a few large funders is that these funders take on a signaling role in the rest of the community. I have often heard it said that many funders are reluctant to fund organizations that have not first been vetted and funded by Open Philanthropy, and I would not be the first to make this suggestion.
This type of signal leadership poses three concrete problems. First, it expands the influence of already powerful donors by giving them a chance to influence the actions of remaining donors. Controlling the lion’s share of funding is already a powerful form of influence, but when this is coupled with the ability to substantially influence remaining funders, it begins to make the funding landscape behave much more like an oligopoly, or in the current landscape even a monopoly. Under many conditions, this is not an efficient form of economic or social organization.
Second, it creates an epistemic asymmetry in which a few large organizations are trusted to do the majority of thorough vetting. Independent researchers and smaller organizations typically lack the time or capacity to evaluate grants in detail. In a market dominated by a small handful of large funders, it is often rational for them to pass off the work of evaluation to these funders. For example, Neel Nanda writes:
I earn enough to be a mid-sized donor, but I would be somewhat hesitant about funding an org that I know OpenPhil has passed up on/decided to stop funding, unless I understood the reasons why and felt comfortable disagreeing with them. This is both because of fear of unilateralist curse/downside risks, and because I broadly expect them to have spent more time than me and thought harder about the problem. I think there’s a bunch of ways this is bad reasoning, grantmaker time is scarce and they may pass up on a bunch of good grants due to lack of time/information/noise, but it would definitely give me pause.
But this does not merely expand the influence of funders over what is done. It also helps those funders to control what the community thinks by putting them in an epistemically privileged position that will be likely to be absorbed, even if it does reflect idiosyncratic features of the funder.
Finally, it creates winner-takes-all dynamics in which those locked out of initial funding rounds find no funding, whereas those funded first are then showered with a variety of additional opportunities. In some cases, this may be healthy: for example, some effective altruists believe that the best interventions are orders of magnitude better than their competitors, in which case it is well and good to shower them with riches. But if the effectiveness curve is not so heavy-tailed, at least within cause areas and perhaps also between them, then these dynamics will be unhealthy and inefficient.
4. Instability
Many fields have high start-up costs. That includes my own field, academic research, in which it takes years of training to establish expertise in a new sub-field.
Fields with high start-up costs need stability. For example, academic researchers want to know that the topics they devote years to studying are likely to still be important and fundable for some time to come. If they do not think that interest or funding will continue for more than a few years, they are likely to never enter the field, and if they do enter the field they will pay a heavy price when they find themselves newly minted as experts in a topic that no one cares to learn about or to fund.
One of the challenges of movements backed by a few large funders is that they move quickly. A few key decisionmakers can pivot the direction of the movement in a span of months or perhaps one to two years. This has, within effective altruism, already happened twice — first, in the shift from short-termism to longtermism, and second, in the shift from general longtermism to a specific focus on artificial intelligence.
The cost of these shifts is that many researchers will be left out in the cold. For example, funding interest in global priorities research has largely dried up, and institutes which spent years building a pipeline of PhD students and postdocs are increasingly left to fend for themselves.
There are, to be fair, ways to provide stability. Traditionally, this is done by endowing organizations with enough funds to operate for a long period of time, perhaps in perpetuity. That is not something that effective altruists have often been willing to do. And, to be fair, perhaps effective altruists are right to be wary of long-term investment when they have moderate to high confidence that they will change their minds about the value of the investment down the road.
However, the cost of a tightly controlled philanthropic market coupled with a resistance to permanent endowments is a highly unstable situation that disincentivizes exactly the sort of deep and costly investment needed to produce world-class results. Perhaps the cost is worth bearing, and perhaps it is not, but in either case it is a cost to be noted.
5. Conclusion
This post concludes my discussion of donor discretion by expanding upon an observation that many effective altruists already agree with: the funding landscape is too heavily dominated by a small handful of donors. This gives those donors the power to move actions and opinions at least partly through financial means rather than through strength of argument, provides institutions with signaling power to affect decisions elsewhere throughout the effective altruist ecosystem, and threatens the funding stability needed to rationalize high start-up costs necessary for world-class performance in many fields.
These are some reasons that more funding diversity would be beneficial for the effective altruism movement.
Leave a Reply