Number of EAs per capita by country

Yeah, given the GDP-per-capita graph shared by William the Kiwi, and Estonia being the home of one of EA's most generous donors, I suspect that availability of philanthropic funds may be the bottleneck for local EA growth.

It wouldn't surprise me if low-GDP countries have many stories like this one or this one, of motivated people struggling to obtain funding due to lack of connections with philanthropists. If true, that would be an unfortunate situation -- lower GDP per capita often goes with a lower cost of living, meaning you get more bang for your philanthropic buck, all else equal.

Maybe we need a new sort of "digital nomad EtG/grantmaker EA" who travels around, networking with EAs in the developing world to identify great giving opportunities. This could go well with a cause area like AI alignment which can be done anywhere and could benefit from fresh thinking:

Alignment researchers’ general models of the field are well-calibrated, but the fact that they don’t think the field is on track to solve alignment suggests that additional approaches should be pursued beyond what is currently being undertaken—a view which was also echoed continuously throughout alignment researchers’ free responses. We think this results lends additional credence to pursuing neglected approaches to alignment.

https://forum.effectivealtruism.org/posts/eToqPAyB4GxDBrrrf/key-takeaways-from-our-ea-and-alignment-research-surveys

Nice! Seems like from the interest generated on this post there is a fair bit of appetite for this.

Wait up, the USA didn't even make the list...?!

Surprising, although I guess this is just because there are plenty of EAGs in the US already.

I think the survey team didn't do a per capita visualisation because response rates will probably vary a lot between countries for reasons other than the number of EAs per capita. 

Yeh, as we note here:

We have reported this previously in both EAS 2018 and EAS 2019. We didn't report it this year because the per capita numbers are pretty noisy (at least among the locations with the highest EAs per capita, which tend to be low population countries).

The last time we reported this was 2020, with the caveat that "Iceland, Luxembourg and Cyprus, nevertheless have very low numbers of EA (<5) respondents. This graph doesn't leave out any countries with particularly high numbers of EAs, in absolute terms, though Poland and China are missing despite having >10."

We'll discuss the details more in the post we are putting together on this (hoping to release this month), but there is indeed quite a lot of noise when you look at EAs per capita, and in particular the highest EAs per capita countries, due to small populations and small numbers of respondents, often close to zero (e.g. small countries can jump in and out of the top rankings based on having 5 or 0 respondents in a year). In the full post we'll additionally examine results for composites of years (e.g. 2020-2022), and which countries outperform what a model would predict (though that will be heavily caveated).

Nice! Yes good idea to use EAG data - I wonder if it would interact with which countries have an EAGx of their own, e.g. are some people in Germany or Netherlands or wherever else less likely to come to EAG when they know they can go to a conference domestically later in the year? Perhaps it would only be a small effect.

I agree with that.

Also, notice that the top countries are pretty small. That may be because random factors/shocks may be more likely to push the average up or down for small countries. Cf:

Kahneman begins the chapter with an example of data interpretation using cases of kidney cancer. The lowest rates of kidney cancer are in counties that are rural and vote Republican. All sorts of theories jump to mind based on that data.  However, a few paragraphs later Kahneman notes that the data also shows that the counties with the highest rates of kidney cancer are rural and vote Republican. The problem is that rural counties have small sample sizes and therefore are prone to extremes. 

NickLaing @ 2024-06-30T07:18 (+11)

The classic fact about variance in small populations, from the start of "thinking fast and slow". Love it!

OscarD @ 2024-06-30T08:19 (+2)

True!
Of course if we had all the data we could run a fancier statistical test. I suppose my observation is limited to the fact that the English-speaking vs European ranges seem similar rather than e.g. all the Anglosphere countries being distinctly higher than all the European countries.

JoshuaBlake @ 2024-07-01T07:10 (+2)

You could do a funnel type plot where your y-axis is EAs/capita and your x-axis is 1/sqrt(population), which is sort of what you'd expect the standard deviation to look like.

OscarD @ 2024-07-01T07:29 (+2)

OK this is what we get, using the 25 EAs cutoff for the red line.

JoshuaBlake @ 2024-07-09T15:07 (+4)

Hmm.... That cutoff is really making it hard to assess what's going on here IMO. Everything is kinda clustered close to the line making me suspect the selection effect is important.

Yeah, I made a quick chart comparing anglosphere vs non-anglosphere

True, I don't have access to the raw data sadly, only to the data from the EA Survey forum post which has a minimum number of EAs cutoff.

<Sarcasm>

New cause area: increase global GDP by multiple trillion to increase EA recruitment

</Sarcasm>

Here is the original chart adjusted for GDP/c. Estonia, New Zealand, Israel, and UK have more EAs than expected, the US has less. What might predict this difference?

Thanks, good idea to make this graph, and indeed that is quite a high R^2! I wonder if a log-log version of this graph would be more informative, particularly for the y-axis?

William the Kiwi @ 2024-07-05T10:44 (+4)

Here is a log-log version, with trendline equations.

Link to doc https://docs.google.com/spreadsheets/d/1VETGm-EjAvQCsGToQCnYwqFzwVfU2TNEyjOtAriMbpc/edit?usp=sharing

OscarD @ 2024-07-05T12:01 (+2)

Nice, yes quite a clear trend. Re David Moss's comment, it will be interesting to see what the fuller data shows, including countries with <25 EA survey respondents.

Difficult to say. I think it's just a lucky case of a small country that managed to get a few active organizers early on (and managed to get members fill out the survey 😁).

AFAIK, we haven't done anything significantly differently to other groups. It's been a fairly standard Intro Course, Career Course, Public Events, Core Volunteer Team setup.

Whether it's got something to do with the Estonian cultural landscape affecting people's readiness to join EA is hard to assess, as I have little experience recruiting outside Estonia :)

But most people I've spoken to in Intro Courses have been pretty much immediately sold on EA ideas, and have been familiar with animal welfare and global health concerns.

Yeah this sounds right to me -- I suspect in many cases a high number will be a result of a few awesome and motivated organizers in that particular country.

[Edit: I've become less certain]

If I was on the CEA community-building team, I would be using data to try and identify top local EA organizers, then interview them and try to figure out what they are doing which works so well, then put together a guide based on that.

FWIW I think the Norwegian main talent pipeline is from Norwegian University of Science and Technology (NTNU), e.g. see this post, rather than University of Oslo. The community there has been going strong there for 10 years.

Thanks, I just found myself thinking 'Australia seems to be a large net exporter of EAs', being in that category myself, so then went looking for data on where EAs are from or live now. That's a reasonable hypothesis, and is perhaps consistent with the fact that there is more variance in EA-ness of smaller countries. I am guessing there isn't a particularly deep reason Estonia has more EAs than Latvia and Lithuania, for instance, and just for historically contingent reasons there happened to be some good organisers in Estonia a while ago? Overall I suppose I don't want to infer too much from this data, but indeed it is interesting to theorise, we would just need more/better information to test various explanatory hypotheses.

I don't think one can infer that without having the whole distribution across different countries. It may just be that small countries have greater variance. (Though I don't know what principle the author used for excluding certain countries.)

Good idea, I don't know of this dataset existing, but if someone has it that would be interesting!

Thanks good idea, I have not done this myself but if anyone does, let us know the results!