The original results are hosted on a site that no longer works, so the results have been moved here: https://rethinkpriorities.org/s/EASurvey2015.pdf

Excellent job, thanks for taking the time to analyze the data so well. Kudos!

I'm glad the survey asked questions about how welcomed people feel. It's really important to help people feel welcomed in order to motivate them to engage with EA well. It's concerned to me that 36% percent felt that the EA community was not actively welcoming, and of those a quarter felt that the EA community was actively unwelcoming.

This number is very similar to the outcomes of another question that makes me highly concerned, "Do insecurities about not being "EA enough" sometimes prevent you from taking action or participating more in the EA community?" The fact that 37% answered yes make this an obvious area of concern, as the survey report correctly identified. I'd be highly interested in seeing what is the correlation between these two questions, and would be willing to bet money the correlation is not insignificant.

Of course, the likely number for these 2 questions is much higher, since many more of the people who feel this way would not take the survey. I think Julia's recent post has some great ideas about how to address this issue.

I'd also like to see the correlation between those people who feel this way and the amount and proportion of income of donations. Like others, I'm concerned that the median number is only $330. Are the people who perceive a high obligation to give and insecurities about not being "EA enough" also the ones not donating much or are they donating more than the average EA? That would impact how to address the question of how to get the median donation up.

On the positive side, I see that the average proportion donated by EAs is 7.5% of their annual income, which is 350% higher than the average donated by Americans. Nice to see that number.

Overall, I think that being more welcoming would cause more people to be more positively engaged with the EA community, identify more strongly as EAs, and be more eager to give their money and time. This should be a win-win!

For the analysis of donations, which asked about donations in 2014, I'd like to see the numbers for people who became EAs in 2013 or earlier (including the breakdowns for non-students and for donations as % of income for those with income of $10,000 or more).

37% of respondents first got involved with EA in 2015, so their 2014 donations do not tell us much about the donation behavior of EAs. Another 24% first got involved with EA in 2014, and it's unclear how much their 2014 donations tell us given that they only began to be involved in EA midyear.

I noticed a contrast between the framing of a couple different parts of the survey results:

[T]he median donation was $333. Certainly good, [...]

It’s clear that EA has a significant problem with gender diversity

I think it would be more reasonable to say that the median donation is a "significant problem", and the gender ratio is suboptimal but I'm not terribly concerned about it. I believe we as a movement should generally be less concerned about how many of us are men and more concerned about how much people are donating.

Certainly many people have good reason for only donating a little, but on average, I'm sure we can donate more than $333, and we should encourage ourselves to do better.

I have two questions.

(1) The number of donors per charity was included and even broken down by referral type, but the total amount of money moved per charity was not (I believe) disclosed. I'm wondering if this is data you intentionally left out of the analysis (for privacy reasons) or just happened not to include. I would be interested in that data since it can help convey the seriousness of donations. If privacy considerations are an issue, you might wish to disclose the number only for the charities that got at least ten donors.

(2) Somewhat related to Dan Keys' point: the feedback time between the year of donations (2014) and the time the results are released (July 2016) is larger than ideal. You mentioned that the reason for having to do it this way was that the survey was conducted prior to the end of 2015. I'm wondering if it might make more sense to conduct the survey around the end of January, and then have the results released a little before the Giving Season of the next year. Is this something you've considered, and/or are there other ways you expect the feedback loop to be shorter in subsequent years?

Here are a few ideas for improvement that may be worth considering:

This Year's Survey Results Document

1) Add page numbers for ease of reference

2) Revise the document so that the Key Findings on page 2/3, the table on page 11/12, the table on page 12/13, the table on page 18/19, the table on page 28/29, and the abbreviations on page 31/32 each display on a single page

3) Add percentages to the age histogram on page 14, the location table on page 15, the career table on page 18/19, and the graph on page 22

4) For the table on page 26, add a "Total" column and a "Total" row that show both the number and the percentage of respondents choosing each option for each of the two questions

Future Surveys

1) Ask people for the dollar amount they donated to each broad cause area to get an estimate of what percent of money donated by EAs goes to each cause area

2) Ask people who changed their career or career plan because of EA which career type (earning to give, direct work, research) they are now pursuing instead

3) It might be interesting to ask people whether they tell non-EAs about EA and if yes, the number of non-EAs they have convinced to become an EA

4) It might be interesting to have a marital status question as well as a question about whether those who are married are married to another EA

Overall it's a great survey with very useful information. Thank you for all of your effort!

The low median donation by EAs suggests that more needs to be done to raise this level. We now have a large list of EAs (the EA facebook page) with ~11k EAs. I think it would be worthwhile to conduct some message testing experiments on the list if possible. We might send messages/articles promoting a greater level of donations to a random subset of the list then later survey the facebook group to ask how much they have donated during the time period since the messagea. Then we could compare the average treatment effects of the various messages used.

At the Effective Animal Advocacy Symposium, Garrett Broad pointed out in his talk that the 2015 Survey of Effective Altruists did not ask about race, which is worrying given how overwhelmingly white the movement is. To my knowledge this makes at least two public critiques of the movement on this specific topic.

He points out that the best way to deal with race issues is not to ignore the issue, but to bring it front and center. Could we please be sure to include a question about race on the 2016 version of this survey?

EDIT: Here's an image. I'll upload a video of his talk once ACE puts the videos of the conference online.

EDIT: The video is here. It's titled "Advocacy for Education" and Garret Broad's section of the talk begins at 33:20.

Currencies were converted into American dollars and standardized using contracted virtual assistants, and then processed and analyzed using the open source Surveytools2 R package, created by Peter Hurford.

As an update, this year I did not need to contract any virtual assistants to convert currencies, because our data was standardized enough that it could be done computationally. To convert all the currencies, I used the open source CurrencyR R Package which I also created.

Hi, I notice the content section of the pdf do not quite match up to the actual subheadings in the document. In particular the contents list a section on 'What benefits do people get out of being in a local group?' which is not in the doc. Just wanted to check that nothing has been missed out of the write-up.

Otherwise - AWESOME job guys - keep up the good work - and feel free to ask if you need funding to make this happen for future years.

Great work. I won't repeat the praises others have already given though, and go straight to my comments :-)

One question i'd like to see more explicitly addressed in such surveys is, How counterfactual has the EA movement actually been?

I've only skimmed through the report, but it didn't seem like you do much to address this question. I'm much less concerned about the $333 median donation than i am about the $6.8 million total donations. That's about 20% of what GiveWell attributed directly to their research in 2014 [http://blog.givewell.org/2015/04/13/givewells-money-moved-and-web-traffic-in-2014/] (still only about half of it, if you discount Good Ventures grants).

Your table showing that GiveWell is how the largest number of EA respondents got more involved with effective altruism suggests that there might be a lot more people out there hearing about GiveWell and whatnot from other means who are essentially effective altruists, but don't know about it, or simply don't identify with the movement for some reason. Do the responses from people who have heard of effective altruism but don't identify themselves as effective altruists shed any light on why this is so?

I'm wondering whether one could run the same survey in parallel to, say, GiveWell followers and the donor bases of some EA-vetted charities, but with the questions, 'would you consider yourself an EA' replaced by, 'what factors did you take into account in choosing the charity you donate to' --- with the term 'effective altruism' tabooed, if you will.

Of course this only touches on the 'individual donations to charity' dimension of effective altruism, and i'm not sure what fraction of the total impact of the movement that should represent --- or even how to measure it.

I hope this is helpful :-)

Great job! My suggestions would be to include questions covering: a) career path and career changes, b) ways how individuals first got involved and are currently involved (e.g. via an EA hub event, if they attend any, or get involved in the online EA community) and c) what events/resources they use and would like more of/would get them more involved. Hopefully addressing the identity/discomfort issue!

Great work! For next year, maybe include a question about time spent? This could even be disaggregated into the different ways of being effective, including volunteering for an effective organization, influencing people, contributing to the movement in general (fora, etc), enabling an effective altruist, etc. I guess if you want to solicit different definitions of EA, you could ask the people why they think they are EA.