Historical EA funding data: 2025 update

NOTE: Original post updated with additional data, which impacts some conclusions /​ observations—see “Updated data” section for further details.

Long time lurker, first time poster—be nice please! :)

I was searching for summary data of EA funding trends, but couldn’t find anything more recent than Tyler’s post from 2022. So I decided to update it. If this analysis is done properly anywhere, please let me know.

The spreadsheet is here (some things might look weird due to importing from Excel to sheets)

Observations

  • EA grantmaking appears on a steady downward trend since 2022 /​ FTX.

  • The squeeze on GH funding to support AI /​ other longtermist priorities appears to be really taking effect this year (though 2025 is a rough estimate and has significant uncertainty.)

  • I am really interested in particular about the apparent drop in GW grants this year. I suspect that it is wrong or at least misleading—the metrics report suggests they are raising ~$300m p.a. from non OP donors. Not sure if I have made an error (missing direct to charity donations?) or if they are just sitting on funding with the ongoing USAID disruption.

(Edit note: I no longer endorse some of these observations; see “Updated data” section at the bottom of this post)

Methodology

  • I compiled the latest grants databases from EA Funds, GiveWell, OpenPhilanthropy, and SFF. I added summary level data from ACE.

  • To remove double counting, I removed any OpenPhilanthropy grants that were duplicated in GiveWell’s grant database. Likewise for EA Funds.

  • I inflation adjusted to 2025 $ based on the US CPI data from WorldBank.

  • For 2025 data, I made a judgement call on how much data was “complete” and pro-rated accordingly—e.g. from GiveWell, it looks complete up until the end of June, so I excluded any grants made in H2 and doubled the sum.

Notes

My numbers are a bit different from Tyler’s. I’ve identified the following reasons:

  • Inflation adjustments (i.e. an upward boost from using 2025$)

  • I’ve used GiveWell’s grant database rather than their metrics reports,

  • Different avoidance of double counting (I removed from OP, Tyler removed from GW. I also went through more manually—from what I can see Tyler removed any GW grant that has OP as a donor, but this will remove too much for many grants that have multiple donors.

  • Tyler excludes “Other” grants, I’ve left them in.

Thanks for reading, hope this is interesting!

Updated data, with additional sources

A few people helpfully shared additional data points in the comments (Thanks @Jason, @MHR🔾, @Andreas Jessen🔾) I have included these in the below graph.

It provides a fuller picture for 2024 /​ 2025, but is much less comparable with earlier years. However, I do think it contradicts the previous tentative conclusion that overall EA funding is declining (also contradicted by Luke and Sjir’s helpful memo). It also shows that GiveWell are maintaining their funding (though not growing in line with overall EA funding), which I think aligns better with what is suggested by their metric reports.

Changes made:

  • Updated OpenPhil figures for 2024 and 2025 to align with this, better reflecting won’t-be-published and yet-to-be-published grants (note this means they are really not comparable with earlier years)

  • Updated GiveWell’s 2025 estimate to match their estimate here

  • Added data for Longview and Founders Pledge based on this