I’m currently a co-director at EA Netherlands (with Ricardo van Mildert). We’re working to build and strengthen the EA community here.
Before this, I worked as a consultant on urban socioeconomic development projects and programmes funded by the EU. Before that, I studied liberal arts (in the UK) and then philosophy (in the Netherlands).
Hit me up if you wanna find out about the Dutch EA community! :)
Not sure if it’s yet the worst in its history but I agree this is serious.
Based on publicly available info, CEA’s marcomms spend in 2025 was about 4% of its total, so roughly $0.8–0.9M. That’s about the same in absolute terms as its Communications budget in 2023 ($0.89M). The share has gone up mainly because the rest of the budget shrank. (The two teams may not be scoped identically, so treat this as approximate.)
My 2 cents: I think marcomms should be closer to 10% of spend, maybe 15%. On current numbers that’s roughly $2–3M.
I’d take most of it from events. Events were about half of CEA’s budget in 2023, the last year with a public breakdown, and they’re largely internal-facing. EA is pretty well coordinated as a community. Our comms are not good. I think we could spend a bit less on coordination and a lot more on being understood.
CEA says the constraint is hiring, not money. That may well be true. But EAN brought in comms capacity and got results we’re happy with, on a central budget of under €200k. So I’d like to understand more about what’s making this hard at CEA’s end, and whether there are routes around it, e.g. agencies, contractors, or funding comms capacity in national groups. Asked in the nicest possible way.
Thanks for the work you’re doing Fran!
I guess ‘significant partisan gap’ is too vague a phrasing. However, I think it’s fair to say there is much more of a gap than there was.
Post Coxon-resignation, polls on the American general public suggest (1) more Americans than ever are prioritising AI harms and (2) there’s an emerging partisan gap in concern about AI, though a gap is not new among ‘very concerned’ Americans.
On (1), compared to the same survey two weeks BC (Before Coxon), there was a significant jump in the number of Americans saying “AI harms” should be a top priority for government: AI harms as a top priority increased from 36% to 43% (+7 points). Overall, 80% of Americans now say “AI harms” should be a priority for the government.
On (2), for the first time since June, there’s a significant partisan gap in concern about AI, with 80% of Democrats saying they’re concerned and 63% of Republicans. The gap is driven largely by Democrats becoming more concerned, while Republicans’ concern remains high (63%) and relatively unchanged.
Thanks!
Could you extend this back in time to pre-FTX? November 2022 and beyond? A lot of EA engagement metrics started declining around then. If the forum is the same, my guess would be this is due to a broader trend, rather than anything specific to do with the forum.
Yes, partly. I want it visible that I think this is bad. A community with no formal membership has very little else. There’s no board that can sanction Manifund and no membership to revoke, so what people say in public, under their own names, is close to the whole enforcement mechanism for community norms.
I certainly hope I could be convinced this is fine/none of my business!
For example, if the media covers this and don’t draw connections with EA, that would update me towards ‘none of my business’ (right now I think there is a reasonably strong connection, but if informed outside observers disagree with me on this I’d become less certain). Or if someone could present me with case studies of individuals who were convicted of fraud and then went on to have successful finance-y careers without further misconduct, that would update me towards ‘this is fine’.
I’m co-director of Effective Altruism Netherlands, writing in a personal capacity. I think this decision is clearly wrong.
Hiring someone with a recent fraud conviction into a finance role at a grantmaking organisation is what ordinary safeguards exist to prevent. Justifying the bypassing of ordinary safeguards because the people involved are smart and well-intentioned is a fair summary of how FTX happened.
I think if she really was overcompensating she wouldn’t go for this role because a) the ethics of hiring her are clearly contestable (check out the number of comments on this post) and b) Manifund has made contestable decisions in the recent past.
I agree she’s now likely to be less vulnerable than she was—I assume she’s learned something from her experience. However, I think she’s still more likely to be vulnerable in this way than someone who has no track record, or has a positive track record. People aren’t static, but they’re also far from being perfectly fluid.
I disagree with you on (2). I’m staying with friends at the moment. One of them is working in an EA-adjacent organisation but has never been to an EAG or anything like that. She’d already heard yesterday about ‘some EA org hiring one of the convicted FTX people’ and brought it up with me. She’s not a student; she’s just online.
I think Nick Bostrom left a reasonable impression on a subset of people who are actually pretty relevant to the project of effective altruism. Friends whom I’d put in the general category of ‘informed readers of the New York Times’ still say things to me like, ‘What about that Bostrom guy? Seems pretty racist...’. It’s the same with FTX, the EA castle, blah blah.
And then people I’ve spoken with who work with ultra-high-net-worth individuals say they have to be very careful about associating with EA because it has a terrible reputation amongst a large proportion of these individuals, particularly in the US, largely because of FTX.
(FWIW, I think Austin’s decision is a bad one in and of itself—reputation stuff put to one side. My reading of Caroline’s court transcripts, Going Infinite, etc., suggests to me she’s a smart, hardworking, well-intentioned individual who is vulnerable to doing things she disagrees with when encouraged to do them by a sufficiently ‘charismatic’ individual. In other words, not someone who should be in a high-leverage role).
I picked one claim from this more or less at random to check—the fourfold assumption about government spending—and I don’t think it’s accurate.
The sentence: “they also argue that the resources governments contribute would have been used less efficiently elsewhere, often assuming a fourfold difference in effectiveness (GiveWell, 2018b).”
I don’t think the 2018 post cited contains that assumption. The only 4:1 ratio I can see comes from a hypothetical: a charity buys $5,000 of “magic pills”, the government spends $5,000 distributing them, 1,000 lives are saved, and without the charity the government’s money would have saved 250.
The parameter GiveWell actually states in that post is that the counterfactual value of government funds is ~75% as cost-effective as GiveDirectly.
That’s also quite an old blog post, so I checked the current figure used in GiveWell’s SMC analysis. There, each dollar of in-kind government resources is estimated at around 1⁄20 of the value of SMC spending, not 1⁄4.
They also explain where the 1⁄20 comes from. Interestingly, it doesn’t imply a particularly low opinion of government effectiveness.
GiveWell guesses the displaced resources split 80/10/10 across health, education and social security, then estimates the value of each separately.
Health comes out at 0.0056 units of value per dollar. The composition is from Uganda’s National Health Expenditure Accounts 2013-14 — roughly 31% HIV/AIDS, 22% malaria, then maternal conditions, respiratory infections, perinatal conditions, injuries, nutritional deficiencies, diarrhoeal disease. Each of the eight largest categories gets a guessed cost per life saved ($30,000 for HIV/AIDS, $10,000 for malaria, $20,000 for most of the rest) and is classified as saving either adult or under-five lives. Categories without a guess, about a quarter of spending, are dropped and the rest rescaled. That produces a weighted $12,254 per child life and $30,000 per adult life, split 45⁄55, which GiveWell converts using its moral weights of 116 for a child death and 73 for an adult death.
Education (0.0028) and social security (0.0026) are done similarly, benchmarked against GiveDirectly rather than against lives saved. Weighted 80/10/10, that gives an overall estimate of 0.005 units of value per dollar.
Cash transfers sit at 0.0034 in the same model. So government spending is being valued at roughly 1.5x unconditional cash transfers — above GiveWell’s unit of account, though well below the 10x bar it requires to fund something.
The gap is large because SMC is modelled as exceptional, not because domestic institutions are rated poorly.
I’ve only checked this one claim, so I can’t say anything about the rest. I know Dirk-Jan a bit—he spoke at EAGxAmsterdam in December—and Martijn is on EAN’s board (for full disclosure). I’ll look through this in more detail and get in touch with Dirk-Jan. I haven’t sent this comment to Martijn to check; as he mentions in his comment, most of the writing was done by Dirk-Jan.
Edit: To be clear, none of this necessarily undermines the broader argument—you could read it as strengthening it. But it does look like a factual error, and I’d argue “GiveWell assumes a fourfold difference” is a mischaracterisation. It isn’t fourfold, and it isn’t an assumption: it’s an estimate, built up from stated components, and in the SMC case it comes out around 1⁄20.
The version I’d defend is closer to: GiveWell estimates a large difference, around twentyfold in one case.
It would be cool to hear from Dirk-Jan how he’d improve on that estimate!
Good question! I’ve updated the readme with the info
I used Claude to build a booking web app for the EA Netherlands coworking office and open-sourced it (MIT), in case you run a similar space: https://github.com/ea-netherlands/ean-office
It’s removed lots of small, fiddly jobs on our side, and our users seem to love it!
It does desk booking (full or half days), QR check-in at the door, a trial day for first-time visitors that an admin approves, automatic reminders and a no-show escalation ladder, member-proposed evening events and whole-office co-working days — and it produces the occupancy and attendance figures our funder report asks for. It replaced an Airtable form, an email approval chain, and a shared Google Sheet. Hosting is €0 on free tiers.
My take: have you spoken with key decision-makers about incentivising them to write more on the forum instead of discussing things internally in Google Docs or whatever?
IIRC, 80k stopped using their DIPY(?) metric several years ago.
If you look through their recent reports, they mostly share metrics related to engagement with their programmes/services. And then occasionally there is an additional survey, e.g., the job board survey in their H2 2025 report.
From what I understand, CG likes grantees to report individual narrative case studies, and then these are coded. At least, that’s how it works for the CEA’s CBG programme. So maybe that’s what 80k does as well? And then this is supplemented with engagement metrics.
If that’s the case, it’s pretty similar to CEA’s system.
Be grumpy! I think the forum lacks spice.
Tier 5 is about career changes. That seems like a sensible thing to be tracking, no?
But it’s a lagging indicator so to me it also makes sense to look at engagement at other points in the funnel.
Yes, I think any increase in marcomm efforts should be spent on external-facing stuff; not spin, just constantly explaining what we’re about so there isn’t a comms vacuum and people first hear about EA from EA-aligned orgs, rather than people who dislike EA-aligned orgs.
80k and others have been excellent at this for years. GWWC seem to be getting much better at it. EA Australia as well.
For comparison, 80k has spent around $2–3M a year on marketing since 2022 (roughly 11–16% of its budget), and GWWC around $300k (roughly 13%), with plans to scale to millions. Both figures are from their recent job ads.
On hiring, from what I’ve heard from speaking to some people at CEA, the difficulty has been finding people who are good at comms and who fully understand EA. Honestly, this feels like letting perfection get in the way of good enough. But I can of course sympathise, hiring is hard, comms is hard, and a mess-up here has big implications.