Are men more likely to attend EA London events? Attendance data, 2016-2018.

[Kirsten]

Summary

Using EA London attendance data, David Nash and I found that women* are just as likely to attend one EA event as men. However, women are less likely to return to future EA events. Women and men are about equally likely to attend most learning-focused events, like talks and reading groups. Women are much less likely to attend socials and strategy meetings. I discuss our methodology and implications for EA communities.

Introduction

Over the past three years, diversity in EA has been a recurring theme on the EA Forum. Alex Gordon-Brown suggested four types of diversity: diversity of talent, experience, opinion, and appearance. He argues that although bonds built on visual attributes like race and gender are normally weaker than bonds based on shared experiences or opinions, they are still important because they’re built so quickly. If there are only a few women in a room full of men, we should expect them to be uncomfortable, even if no one has been unpleasant to them.

Suggestions on improving diversity in Effective Altruism communities have been made here and here. Suggestions on how to have these conversations constructively, without tokenizing or minimizing the people already in our community, have been made here.

EA London’s events tend to attract more men than women; our typical monthly pub social would be 70% male and 30% female. Is this because women are less likely to attend EA London events for the first time, less likely to return, or both?

David Nash and I found that women are just as likely to attend one EA event as men, but are less likely to return to future EA events. Women and men are about equally likely to attend most learning-focused events, like talks and reading groups, but women are much less likely to attend socials and strategy meetings.

In this article, I describe our research methods and limitations; present the data EA London has collected; discuss implications for EA communities; and suggest next steps for EA gender research.

Methods and Limitations

David compiled attendance data from 5/​6/​2016 to 8/​3/​2018. He used a software program to code names as likely male or female; he then corrected some of the entries manually.

It’s possible that we’ve misgendered some attendees, especially if they only attended one or two events. Because we’re using aggregate data about hundreds of people I’m not particularly worried—I think the general trend is still useful.

Obviously, it’s impossible to know if the trends we find in London apply anywhere else. I’m sharing them here as a starting point for research in other communities and discussion about gender in EA spaces.

Data

Below, we have attendance data by gender for all attendees between June 5^th 2016 and March 8^th 2018. The column to the left shows how many events a given individual has attended; for example, there are 94 men and 37 women who have attended four or more events in the last two years. As a percentage, 28.24% of people who attended four or more EA London events since June 5^th 2016 are women.

# of events	Female Percentage	Male #	Female #
Just 1 event	50.40%	473	481
1+ event	46.88%	681	601
2+ event	36.59%	208	120
3+	33.88%	121	62
4+	28.24%	94	37
5+	28.26%	66	26
6+	25%	54	18
7+	21.43%	44	12
8+	21.28%	37	10
9+	23.08%	30	9
10+	22.86%	27	8
11+	21.88%	25	7
12+	23.08%	20	6
13+	25%	18	6
14+	23.81%	16	5
20+	26.67%	11	4

This table represents the same data in a different way, showing how women are much less likely than men to have attended more than seven EA London events in the last two years.

Over the past two years, how many attendees have been to exactly…?
	% Female	Male	Female
1 event	50.40%	473	481
2-3 events	42.13%	114	83
4-6 events	33.33%	50	25
7-19 events	19.51%	33	8
20+ events	26.67%	11	4

This table compares attendance at EA London social events (particularly our monthly pub socials) to all other events we host (including talks, reading groups, and career-specific networking events). This data shows that women are much less likely than men to attend socials, but about as likely to attend non-social events.

	Year	Attendance	Female	Percentage
Total	2016	732	372	51%
	2017	1743	689	40%
	2018	424	177	42%
Social	2016	240	97	40%
	2017	812	270	33%
	2018	208	71	34%
Non Social	2016	492	275	56%
	2017	931	419	45%
	2018	216	106	49%

Attendance data by event type (not pictured) shows that EA London strategy meetings and monthly pub socials are much more likely to be attended by men than women.

Implications for EA Communities

In London, women are as likely as men to attend an EA event for the first time. Marketing events differently to attract more women seems unnecessary. We don’t have a problem attracting women; we are just less likely to retain them.

EA London events aimed at newcomers tend to be gender balanced, while socials and strategy meetings are heavily male. This makes sense: we’d expect newcomer events to be gender-balanced because 47% of attendees at EA London events are female and 50% of attendees who only attend one event are female. Given that ¾ of regular attendees are male, we’d expect strategy meetings to be male-dominated. It’s more surprising that the monthly pub social was so heavily male, given that it was advertised to newcomers. However, a group of men who are quite involved in EA London regularly attend the monthly pub socials, which may explain why they skew male.

It’s not obvious why women are less likely to return to EA London events. I’m currently doing some qualitative research to investigate what motivates men and women to attend EA London events. I plan to publish it within the next month.

Next Steps

EA London attracts a near-equal number of men and women to its events, but men are more likely to return to future events. We don’t yet know why this happens, but it skews socials and strategy meetings male.

It could be useful for other communities to track their attendance and see if they find similar results. A series of interviews with women from different EA communities who have stopped attending events would also be helpful.

*We made rather simplistic assumptions for this research, based on people’s names and gender presentation. I recognize that not everyone feels comfortable with this binary. Nevertheless, I hope that this research will be a step toward making EA London more welcoming to a wider range of people.

[Julia_Wise🔸]

After years of speculation on where in the pipeline the difference is appearing, I’m so happy to see numbers that let us have a more concrete idea of what’s going on, at least in this group!

[evemccormick]

Hey, thanks for this :)

EA Cambridge (UK) has been tracking gender ratios at events for several years now, and we have fairly complete data for the last year. As far as I know, this hasn’t yet been written up or shared in any formal way, but that is something I would like to do in the near future.

On the topic of pub socials, I don’t find the gender bias surprising. The pub socials which we started running this year, aimed at non-students and postgrads, were almost always heavily male-dominated. Other types of pub social that we’ve run, targetted at newcomers, people somewhat engaged, and people deeply engaged in the community respectively, have similarly been male-dominated. At least two women, despite being deeply involved in the EA Cambridge community, have mentioned to me that they do not feel comfortable at pub socials. So while I think there are probably several reasons why fewer women attend socials in general (e.g. to do with women often feeling less comfortable in situations where they’re expected to talk on a topic, especially when they’re new), pub environments might compound the problem.

Very much looking forward to reading the results of your research about what motivates men and women to attend events :)

[DavidNash]

Looking at other social events not at pubs the ratio has been similar, ranging from picnics, hikes, Disney movies, restaurants, education/​animal focused topics.

[Catherine Low🔸]

You show Disney movies at EA meetups!! That is really lovely!

[Daniel_Eth]

Are women more likely to return for a second event if the gender ratio of the first event they attended was more balanced? This could tell you whether the difference is simply a result of the community being mostly male right now, or if it’s due to some other reason(s).

[DavidNash]

I looked into this and of the women that came to just 1 event, there was an average of 49% female attendance. Of women who came to 2 or more events, the first event they came to had, an average of 45% female attendance.

Also of the women who came to 5 or more events, half of them came to an event with less than 40% female attendance and half of them came to an event with more than 40%.

[KyleYork]

Some questions that come to mind:

1-- Are women who wish to speak during events given equal (de facto) opportunity to speak? (Of course they are given equal formal opportunity, but as I saw in my postgrad program, group sessions that favour more typically ‘masculine’ ways of engaging in dialogue—favouring assertiveness and competitiveness—would tend to leave the women in my program alienated). 2-- Could the gender difference in jobs and education and its relation to the sorts of activities EA emphasises play a part? 3-- Do women feel uncomfortable at events for any other reason? If they did, would it be in such a way that they could or would report it on a survey? 4-- Would the amount of women in leadership positions alter this? 5-- Have similar surveys been conducted with people of different races and ethnic backgrounds?

Cheers

[Kirsten]

Hi Kyle! A lot of these questions are about how women feel at events—I hope to be able to answer them after I finish working through my focus group data. I can tell you that EA London has two paid staff, one male and one female.

We haven’t used our attendance data to look for patterns in race/​ethnicity because it would be more difficult. We can usually make a good guess at someone’s gender from their name and appearance, but can’t always make a good guess at someone’s ethnic background from their name and appearance. We’d have to ask, and asking first-time visitors about their race seemed like a bad idea.

[DavidNash]

There is also some data from Facebook groups that might add context.

On the EA London group with 1797 members the split is 43.8% women and 54.7% men.

For the London animal group with 495 members the split is 50.6% women and 46.9% men.

[LucyW]

Thanks for taking this question seriously and doing the quant leg work—much appreciated! Just wondering if you /​ other authors working on this topic have now managed to conduct any qualitative ‘exit’ interviews (or at least feedback interviews) whose key reflections have been shared publicly?

[Kirsten]

I have done some interviews, actually—the Forum post has just gotten held up by my personal life.

[andrewpei]

Any follow up here on qualitative reasons why women are less likely to return to EA events?

[Kirsten]

I have the follow-up sitting on my computer at home, but for personal reasons I am unable to actually go home and upload it, so it might be another couple of months before I can share it.

[Dunja]

Thanks a lot for this post, that’s really interesting and highly relevant. I’d be curious to see also the proportion of women in online forums such as this one. And of course, I’m super interested in possible reasons behind the tendencies you describe.

[Peter Wildeford]

I’d be curious to see also the proportion of women in online forums such as this one.

We have that in the EA Survey data.

[Kirsten]

I’m not sure where to find that—can you link to it or post an extract here?

[Peter Wildeford]

As a baseline, the overall population of the survey was 73% male.

Of those who indicated they were members of the EA Forum, they were 79% male.

Of those who indicated they were member of a local group, they were 72% male.

Of those who indicated they were member of the EA Facebook Group, they were 72% male.

Of those who indicated they were member of GivingWhatWeCan, they were 76% male.

Of those who indicated they were a user of LessWrong, they were 85% male.

(For simplicity only, these responses drop people who don’t indicate they are EA, don’t answer either question (gender or membership), and don’t pick either “Male” or “Female” as their gender. All answers are rounded to the nearest percent. Also, keep in mind the sampling bias of needing to have seen and filled out the survey in order to be counted—this may mean that we don’t correctly reflect an unbiased census of the entire population as a whole.)

[Dunja]

Oh damn :-/​ I was just gonna ask for the info (been traveling and could reply only now). That’s really interesting, is this info published somewhere online? If not, it would maybe be worthwhile to make a post on this here and discuss both the reasons for the predominantly male community, as well as ideas for how to make it more gender-balanced.

I’d be very interested in possible relations between the lack of gender balance and the topic of representation discussed in another recent thread. For instance, it’d be interesting to see whether non-male EAs find the forum insufficiently focused on causes which they find more important, or largely focused on issues that they do not find as important.

[Peter Wildeford]

We haven’t posted a gender breakdown by group yet. I can see if there may be ways to follow this up as part of our forthcoming 2018 EA Survey work.

[Dunja]

That would be great!

[Dale]

Interesting data!

Sounds like churn is much lower with men. Unless it is much more expensive (in terms of advertising etc.) to get new men, which looking at the ‘1+’ row doesn’t seem to be the case, or women are more productive, this suggests you have a higher Customer-Lifetime-Value /​ Customer-Acquisition-Cost ratio with men. If this is the case then, to the extent you can, you should re-allocate your efforts at the relevant margin towards attracting more men.

https://​​www.klipfolio.com/​​resources/​​kpi-examples/​​saas-metrics/​​customer-lifetime-value-to-customer-acquisition-ratio

[Kirsten]

Several previous forum posts have suggested that it’s useful to try to have a balance of men and women because:

-In the longer term, only focusing on white, middle-class men who want to do good using evidence will mean we lose out on a large amount of talent and resources. source

-The focus on men might not be reversible: unbalanced events now could put off “highly productive” women, who might continue to view EA unfavourably years later. source

-Diversity will likely make conversations and viewpoints less predictable and more useful. source

[DavidNash]

It may be a bit short sighted to try and grow in a way that leads to 0.1% of people being interested in effective altruism rather than 5%.

It seems to be a mistake that both the atheist and skeptic communities have made in the past.