TL;DR: As you begin community building in the EA space, it can be unclear how many enaged members you can expect to produce, in this post I estimate that only 0.0026% of the population of people who have "been outreached to" become active EAs, for most university groups this implies about 1-2 active EAs across your whole population. This is an underestimate  (reasons detailed below), as such I encourage you to see this as as an absolute lower bound (and attainable target) for production of level 3-5 engaged Effective Altruists.

Introduction

Below is a quick take on the number of members taking significant action that you can expect to have in your university group. This is based on proportions of engagement with EA following outreach and makes lots of assumptions. Where I am aware of them I have tried to highlight these assumptions and indicate the extent to which they imply the figure is an over or underestimate.

Simply, I first estimate the number of people who have heard of EA (engaged with the ideas), I then estimate the number of people taking significant action, Express this as a proportion and multiply it by the population of my university (Durham), this provides an estimated upper bound for the number of people taking significant action in my university. 

# of people that have heard of EA

Using the inclusive stringent definition of having heard of EA  figures from this report by Rethink Priorities. Which is still believed to be an overestimate by RP, 2.6% of the US adult population has heard of EA. (based on answers to definition based questions, rather than self-reporting qualitative judgements of levels of engagement. In particular by selecting for answers one would be very unlikely to be able to give they had not actually encountered effective altruism))

This is still taken to be an overestimate because people can guess at a passable definition of EA by simply knowing the meanings of the words Effective and Altruism - evidenced by many surveyors self-reporting no prior knowledge of EA but still meeting stringent criteria in later answers.

Extending this 2.6% globally is also a significant overestimate for many reasons, such as: There is correlation between Higher levels of education and engagement with EA; Proximity bias (San Fran hub etc.). Instead estimate US, Canada and Europe number with this figure and apply a 30% discount rate to ROW. (this is an inutitive guess based on proportion of highly educated populations and the concentration of EA groups)

Global Population = ~8.2 billion

US population = ~340.1 million

Canada population = ~ 41.9 million

EU population = ~744.3 million

Rest of World population = ~7.07 billion

Therefore the number of people who have heard of EA in a stringent manner can be roughly 29,283,000 ((340.1+41.9+744.3)million *0.026) for the US, Canada and Europe and 128,674,000 (7.07 billion*2.6%*0.7) for the Rest of the world. This gives a total outreached population of 157,957,000.

How many Engaged EAs are there?

Method 1: Rethink Priorities and 80k Study extrapolated 

In 2019 there were an estimated 6500 active EAs at the level 3-5 level (moderately engaged and above), In the 80k post about community building and allocation of resources, Rob estimates a 14% growth rate for 2020. I will assume this growth rate remains consistent (I have no reason to update or believe my estimate will be better, and I am unable to evaluate any changes in CEA strategy since 2020). This means that as of today we have $6500\ast {1.14}^{4.67}$ = 11986 (4 and two thirds years since 2020, this is important because I'm using current population)

Method 2: Forum Members

To estimate the number of people that respond to the EA survey I have sent polls into my GCP group chat and in the UK G.O.R. asking whether they responded.

The 2024 survey had 1916 respondents. (There is no 2025 survey, I will assume a 14% growth rate). Approx 60% of the small people I surveyed who are active EAs said that they filled in the survey. if 1916 represents 60% of active EAs, then there are 1916/0.6 = 4786 active EAs in 2024. 

This is an underestimate for two significant reasons, the first being that EA community builders love surveys and the majority of people I asked were community builders and so are more likely than "the average EA" to respond to a survey. The second that responses to this poll were public so there is some social embarrassment that comes with answering no, so people are much more likely to answer yes. These together imply to me that I should estimate the percentage of respondents 40% higher than above. 

Applying this internal validity discount as well as the 14% growth rate we estimate 4786*1.4*1.14=7638

This survey reached very few  people so we can have only low credence in assuming this represents an accurate response rate to the survey. 

Weighted Average Number of active EAs

We can have only low credence in method 2, the sample was very small and a large validity discount is a good proxy for huge uncertainty, I will therefore trust Rob (& the 80k team) and weight the methods 70% to 30% leaving a weighted average estimate for the number of active EAs as 70%*11986+30%*7638 = 10,682.

Conclusions

Proportion of people that, when outreached to, become an Active EA is therefore roughly 10,682/157,957,000 = 0.0068%

For my university student population of 20,000, this means that if we successfully managed to outreach to every single member, we could expect 20,000*0.000068 = 1.35 active EAs.

There are several noteworthy reasons this is a significant underestimate and I think it is useful to think about them:

1. This assumes a completely equal distribution in quality of outreach globally - this is of course wrong. People are better or worse at outreach and in general group organisers receive training and guidance on effective outreach, this will non-trivially increase the proportion of engagement with the EA principles and community
2. This treats all people as having similar capacity to commit time, money or other resources to Effective Altruism, university groups are necessarily engaging (in general) wit students that have more of these resources, so will have higher uptake (at least during their time at university).

Some less of note reasons I think this figure might represent an underestimate is that 2.6% is too high even for the US, Rethink mention false positives. And my discount rate is likely too low based on the spread of EAs globally but I didn't want to model the proportion of outreached EAs based on the # of active EAs in those locations as this seems circular!

p.s. Hopefully this can serve as motivation to university organisers - you're doing okay! :)