CEA, EA Funds, and the missing US politics option
A draft Facebook post sent in spring 2017 to fellow EA-critical bloggers Ben Hoffman and Sarah Constantin for review before publishing. The Centre for Effective Altruism (CEA) had recently launched EA Funds, a donation-routing platform, and had decided in passing not to include a US politics fund despite outside donors offering large sums to seed one. CEA’s explanation was tucked into a longer blog post; this post argues that’s a transparency problem regardless of whether the underlying decision is right.
I’m concerned about a recent statement by the Center for Effective Altruism (CEA): “During YC, we received a lot of pressure to add a US politics fund but did not go ahead, despite promises of large amounts of funding. We did not think we could find a suitable fund manager, and we worried offering such a fund might dilute the overall effectiveness of donations moved through the platform.” [1] Since many smart, altruistic people are focused on US politics, I was very surprised that CEA couldn’t find a good fund manager. I myself can think of several good candidates. I emailed a CEA employee to ask if they could explain more, or if they had any thoughts I should include in a Facebook post. They replied: “We considered a few people, but still thought that the best recommendations they might come up with would not compare favourably to other cause areas. I’d be interested to see a discussion, but would be quite surprised if a discussion lead to many beneficiary organisations being suggested that I thought were as effective as other current uses of EA Funds donations.”
I think other causes being more effective than US politics is a fairly plausible position, but the way this decision was made seems problematic. It’s a pretty important decision — EA Funds, even at its early stage, already directs millions a year in donations. [2] We don’t know what the “large amounts” mentioned were, but it seems safe to assume that it was in the millions, maybe more. [3] A theme discussed elsewhere in the post was transparency, with CEA saying: “We’ve had feedback from the community that we are not transparent enough (…) A commitment to increased transparency is costly, mainly because of the additional time and attention (…) However, given the extent to which our work impacts the entire community, we have underinvested in this in the past. We have decided to dedicate significantly more resources to external communications this quarter.”
However, making important decisions about what causes are effective — decisions that could direct tens, potentially hundreds of millions in spending over the next decade — without explaining the reasoning isn’t a transparent process. The decision itself was only mentioned obliquely, as part of one sentence inside a long blog post. [4] Donors should be fully informed about where their money is going, and why. If a fund picks A as a default over B, a donor doesn’t know that this choice was made, and the donor chooses A, when they would have chosen B had B been the default, the fund is effectively “choosing for them” without their full knowledge. [5] CEA also defines itself, and by extension its websites, events, and so on, as “cause neutral”. [6] Cause neutrality is critical to EA as it currently exists. If Bob Jones thinks X is the most important cause, and John Smith thinks Y is the most important cause, then if Bob excludes Y from his groups and John excludes X from his groups, EA soon stops existing; every cause just goes off in its own direction. If there’s already significant support for a cause, and it’s excluded from a seemingly cause-neutral forum (in this case EA Funds), that raises the question of what else is being excluded, or whether a cause is included in EA only if its proponents are savvy at internal politicking.
Sources: What CEA is currently focusing on and why Annualized extrapolation of figures in Update on Effective Altruism Funds CEA recently participated in YC, and YC leaders Paul Graham and Sam Altman have made seven-figure political donations: TechCrunch on YC’s Sam Altman IMO, blog posts are problematic if they’re the only place to find major, long-term policy decisions, because they quickly go out-of-date. Someone who finds CEA in, say, 2023 isn’t going to read through all the archives, but shouldn’t be left unaware of major decisions or policies.
One could reasonably argue that there are thousands of niche causes, and since CEA can’t cover them all, some choosing is necessary. However, since major donors specifically lobbied CEA for a US politics option, that doesn’t apply here.
Understanding cause neutrality