Charles Dillon
Bio
Experienced quant trader, based in London. Formerly a volunteer at Rethink Priorities, where I did some forecasting research. Interested in longtermism and meta causes.
Posts 7
Comments113
This conceptually seems similar to the meat eater problem argument against global health interventions.
I think a lot of this coordination is implicit rather than explicit, and I don't think it's very well publicised (and there's room for marginal donations to change whether the org gets funded to their high Vs medium target for example, and signalling value that individuals think this is good, so I do not mean to say that this is the only consequence of a donation).
I think there is a misconception here - when it is said that these charities will be fully funded anyway, what that can mean is that they will try to fundraise for a certain budget (perhaps with high/medium/low targets) and larger donors will often choose to fill the remaining gap in their fundraising late in the fundraising process.
This means you are often not really giving the charity extra on top of their budget, but in practice funging with the largest donors. The largest donors will then often give slightly less to them and give to their next best option instead.
As an individual, you are in this case redirecting funding from an organisation which agree with your priorities to whatever their next best option is.
For example, I personally made some donations to animal welfare charities this year which very likely funged to some extent with the EA Funds animal welfare fund. What that means is that the counterfactual effectiveness of my donation might be equivalent to whatever the last thing they chose to fund was (which I think is probably quite good in expectation).
Suggestion: pre-commit to a ranking method for forecasters. Chuck out questions which go to <5%/>95% within a week. Take the pairs (question, time) with 10n+ updates within the last m days for some n,m, and no overlap (for questions with overlap pick the time which maximises number of predictions). Take the n best forecasters per your ranking method in the sample and compare them to the full sample and the "without them" sample.
I think the layout of this post is quite reader unfriendly.
I strongly suggest you start with a full summary rather than just an intro, and don't bury your conclusions midway between the post and some very long appendices which are unlikely to be very useful to 90% of readers.
As it is, anyone wishing to respond in depth would basically have to do the work of summarizing the post themselves, which increases the friction on feedback.