Why EA’s Talent Bottleneck is a Barrier of its own Making

Today I want to say something quite unpopular - it is my belief that the talent bottleneck is a barrier we have created ourselves. Not only are we failing to attract talent from a wide variety of sources, but we actively push away talent that approaches us.

For those who prefer to listen instead of read, there is a video reading of this blog located at the bottom of the page. It also contains extra exploration of the themes for those interested. Want to read more articles? Here is another one!

Introduction

We hear frequently at EA conferences, in the forums, and via meetings across the broad spectrum of Effective Altruism as a community and a philosophy that our primary bottleneck for growth has shifted from funding (of which EA now has no lack) to talent. In short, EA is struggling to get enough motivated, qualified individuals across various fields from AI to Animal Rights to Climate Restoration and everything in between. This is commonly seen as a problem intrinsic to academic endeavour. After all, there have always been fewer experts than there are problems.

Today I want to say something quite unpopular - it is my belief that the talent bottleneck is a barrier we have created ourselves. I believe that EA unconsciously, due to a mixture of innocent ignorance and the culture which it has created for itself, actively prevents a wide variety of talent from joining its ranks both as an organisation and a philosophy. Not only are we failing to attract talent from a wide variety of sources, but we actively push away talent that approaches us.

Reading criticism is often difficult - but there is a reward here. Not just knowledge, but gossip. I am going to expose parts of my life I normally keep hidden, even though it’s embarrassing to me, because I feel that the embarrassment I feel from exposing my life here will be nothing compared to the gains that can be made by learning from it.

Essentially, I want to discuss the equality and diversity bottleneck in EA. There are many barriers to many different types of people, but the one I am focusing on from personal experience is the socio-economic barriers to entry for being part of the EA community. As a disclaimer I do want to say that this will be very UK-focused, and I encourage writers and researchers from lower socio-economic backgrounds elsewhere to write pieces detailing their own experiences wherever they are. I do believe though that many of these core issues transfer well to other nations.

I want to begin by stating that these barriers are no fault of any individual. EA as a community is very welcoming, vibrant, and friendly. The thing about unconscious bias, though, is that no matter how friendly an individual or community culture, things can go wrong. This blog post is meant as an introduction to raise some themes for discussion, rather than an in-depth report. I encourage others to publish their own works detailing their own experiences across the EDI spectrum in order for the community as a whole to get a better idea of where we can improve.

The main point of my post is this: We are actively pushing away talent from poorer backgrounds whilst those working against us are not. I believe this amounts to a large number of missed opportunities, and we have potentially lost many change-makers to private corporations who are not making our mistakes.

The Good


As I mentioned, EA as a community are friendly and welcoming, helping provide support where they can. As a result, many organisers have procedures in place to make sure no-one misses out. They give free tickets to students and to those for whom finances are a barrier to attending. These tickets go for £300-£500 a pop so it’s no small discount. Travel costs are also reimbursed up to a point, and the conferences generally have partial to full catering which is a major boost.

The travel costs can make a huge difference, given that travelling to London even from within the UK can be in the multiple hundreds of pounds alone, not to mention accommodation. Being able to be reimbursed for these, even in full, is a good equaliser. EA even offers advance payments on request, so that people can buy upfront if they don’t have enough money to pay at the time.

The catering is something that probably wasn’t intentional as a socioeconomic equaliser - but it really is. For EAG London 2022, I had £4.37 in my bank account. That’s not ‘apart from savings’, I mean the entire liquid wealth I had was £4.37 including overdraft and savings. If it wasn’t for EAG offering breakfast, lunch, and dinner then I would have had to just skip meals.

Some may say here “If you’re that broke, why go to a conference?” to which my reply is simple. If Wilhelm Rontgen, Marie Curie, Robert Hooke, Elizabeth Blackwell, or Louis Pasteur (among literally millions of other great minds) had never shared their ideas with the world or formed collaborations, humanity would be far poorer for it. I am in no way putting my humble self at the level of these great men and women, but not attending a conference to share research with others in my field for the benefit of humanity just because I don’t want to be hungry for a few days seems a terrible risk/reward decision. So I keep going, despite the discomfort.

EA’s decisions in trying to prevent money being a barrier are very helpful, as are their funds such as community grants which allow groups in poorer areas (like my own EA group, EA on the Tyne) to act well without funding as a major challenge.

However, EA suffers from a bit of a homogeneity problem in that most of the EA decision-making leadership tend to be from more stable, wealthier backgrounds who have not had the same experiences as many of those on the bottom of society’s ladder. This means that though some of the ideas are helpful, they have pitfalls no-one would have noticed in the decision-making meetings. For example, for those of us on benefits such as Universal Credit, receiving any funds into our bank accounts will be counted as 'earnings' and deducted from our next payment. That £400 we just got for train fare? Deducted from our next 'paycheck' and now we can't make rent. It's something many wouldn't think of, but those of us in poverty would.

Here are a few more of them.

The Bad

Effective Altruism as a community is not controlled by the ‘powers that be’, because the main organisers try not to have ‘control’ of the community and allow it to grow organically to increase the chance of new, unorthodox ideas. It’s an unusual stance, but in my opinion a good one. However due to sheer human nature they still shape the community. And it’s not always for the better.

Barriers in the Early Pipeline

Let us focus first on where the major EA numbers are based. You can pick any EA organisation at random, or any EA group from the list of largest or most active chapters, and see an extremely focused geographical spread. Essentially, they’re based around London, Oxford, and Cambridge. In the US-based organisations they are typically amongst the same 3-5 universities as well.

This is important for a few reasons. Firstly, EA does most of its recruiting both for organisations and for the general community from the ‘top’ schools in the UK. This seems like a good idea on the surface, but much like the oft-repeated EA story of the deworming medicine being the solution for education scores, what seems obviously helpful can not always have the desired effect.

First of all let’s pick the major UK universities EA draw from - Oxford and Cambridge. These are ‘top’ Universities, so surely drawing from students there results in the best talent? A quick look at their websites, however, shows a pattern emerging.

Cambridge University

“Taught students undertaking taught Masters courses (for example, the MPhil by Advanced Study) are normally expected not to work during term-time. Students should consult their Faculty and Department for further details regarding vacation dates when they may be able to work.” [Source]
“Research students undertaking a course of more than 12 months or the MPhil by Research may undertake paid work up to a maximum of ten hours per week with the approval of both their Supervisor and College Tutor. The work undertaken should be either academic-related,  related to a student's professional or career development, outreach work undertaken on behalf of the University or related to covid-19.” [Source]

Oxford University

“Study at Oxford is very demanding. We would strongly advise against you relying on income from employment to fund your studies as this may have an adverse effect on your ability to complete your course to your full potential.” [Source]
“An undergraduate degree at the University of Oxford is a full-time course. During term-time you will have very little time to take on a part-time job in addition to your studies and other activities. There are some opportunities to do a limited amount of paid work within colleges, but you should not rely on this in your budgeting” [Source]

Are these rules stuck to? Apparently so, according to research by Cherwell. It found that the national average for students working part-time was 57%, with 90% of those working as many as 20 hours per week. In Oxford? Only 20% of students are employed and the majority work less than five hours per week.

So how are these funded? There are a variety of scholarships available, but some quick sums on a calculator show these don’t quite add up. Oxford is one of the UK’s most expensive cities, and even on the maximum student loan amount you’re still £2000 - £5000 short each academic year. Cherwell found that to make this up, a student would need to work 850 hours a year - which amounts to full-time for 22 weeks.  Essentially, to not only excel but even participate at these universities you need savings - often from family or sponsors. I will say this, because it’s important: many students from low income families do go to these universities and manage to make it work. So if you’re reading this and this is you, don’t be mad. However, you can acknowledge you are an outlier.

To bring this back to EA, what does this mean? It means we’re not selecting from the brightest or most talented - we’re selecting from the wealthiest. How much exceptional talent are at universities in the poorer areas of the UK? Particularly the North of the country? EA has a major focus on the ‘top’ universities because there’s a general assumption it’s where the ‘best’ type of people are. Society has long had a subconscious belief that poor people are less intelligent, because capitalism convinces them that we live in a meritocracy where the hard working and smart become wealthy and the lazy become poor, when in reality that isn’t the case.

I am not saying the EA community members from these universities aren’t smart. They are. The universities have significant grading requirements and so any graduate from there deserves their academic chops. The issue is that having an organisation and community overwhelmingly made up of people from comfortable backgrounds means that we have huge blind spots in our intellectual diversity - especially considering how often poverty features in our work. We sit and talk about poverty for days without 99% of the community having any experience of it. That’s not only morally wrong, it opens us up to huge research blunders due to assumptions that are rarely correct.

So what do we do about it? We urgently need greater EA community investment in the poorer areas of the UK (and I assume other nations too, though their residents will know best). This means basing organisations across a wider area, hosting formal events across a wider area, and advertising EA/offering talks across a wider range of universities. However, there’s no point attracting talent from these areas if we have barriers stopping them from participating and sharing their ideas.This brings us to the late pipeline.

The Ugly

Barriers in the Late Pipeline

So we’ve attracted EA community members from lower socio-economic strata. Job done? Not quite. There are still a few significant barriers in EA that stop people contributing, and I can show this via (very embarrassing) personal anecdotes.

EA frequently hosts its events in Oxford and London, which makes sense because there’s a lot of EA infrastructure there and they’re easy to get to for foreign participants flying in from abroad. It does, however, raise a significant issue for many in our community. Lists of the UK’s most expensive cities to visit change quickly, but London and Oxford are always in the top 5, and mostly in the top 3. Currently, travel site TripAdvisor lists Oxford as the most expensive and London as second, but as mentioned this changes.

For someone from poverty, this is a colossal barrier to participation. I have visited three or four major EA events in the past, and for two of those I slept ‘rough’ for at least one night. I feel bad saying ‘rough’ because I wasn’t in a doorway, but I was pretty close. For EAG London 2022 I slept on a bus for one night as I could only afford 1 night’s worth of hotel with the travel reimbursement offered to me by EA. I got no sleep (which made it around 32 hours spent awake), but I was warm and safe overnight. The next afternoon I cut short my schedule and missed some sessions to get to my hotel early as I was exhausted from no sleep and lots of walking, and fell asleep at 4pm, then slept through my morning alarm and missed my first few 1-on-1s the next day. I sent apologies and lied saying I’d had transport trouble. Saying ‘sorry, I was exhausted from having to use public transport as a hotel so I didn’t get murdered in a doorway’ seemed a bit…reputation damage-y. So I lied and masked. As I commonly do. As others in my situation do. It’s another reason I’m including these anecdotes. If you’re a researcher who’s had to lie to hide embarrassment of the things you've had to do to contribute to EA’s research environment - you’re not alone. I see you. I hear you. The downside of masking is it’s lonely.

Post-Publish Edit: I brought this story up in the forum and EA travel grantmakers said anyone in this situation in future should contact them. It appears they do care deeply, and I don't want to make it seem they don't, but this may also raise an issue that people worry about pushing their luck or don't know they can ask for more :)

The issue is that there’s no real way to tell EA that you’re not just asking for reimbursement but massively relying on it. Obviously everyone is entitled to some financial help to bridge the gap between their thoughts and the community, but some rely on it more. A ‘hardship fund’ to apply to where EA can buy hotel rooms on your behalf, or can increase the reimbursement would go a long way to helping solve this.

That brings me to another point - at one point I had to email EA and ask to speed up my reimbursement because I was destitute. It had to be forwarded between 3 staff to sort out. That’s humiliating. They were really kind and helpful, but it still makes you feel embarrassed. There needs to be a single individual in EA who deals with issues like this.

Post-Article Edit: Feedback from some readers indicated on an organisational level this may be unavoidable. This is addressed in a follow-up blog of ideas.

Another example of the issues with funding is that hotels closer to the venue fill up very quickly and shoot up in price, which means cheaper hotels are further away. For those on a tight budget, the high costs of transport in cities like London or Oxford mean long walks. At EAG London 2022, I had to walk an hour and a half from my coach stop to the venue (after sleeping on the Megabus, which isn’t really EA’s fault), and around 45 minutes from my hotel to the venue. Carrying my belongings. I hope my 1-on-1s like sweat patches. Oh, also EA in general - because I have them in the photos. Nice.

My poverty isn’t anyone else fault or problem. Especially not EA’s. I was born into poverty, yeah, but I won a 100% fee scholarship to a private high school due to my grades elsewhere. I then took out a loan and went to university, so I now have both an AI degree and a Law degree. That’s big money potential. I could go work in private industry with either qualification and end my poverty issues at literally any time. Which brings us to the crux of this blog piece.

Inevitable Consequences

Many people take that option. When you’re poor, altruism is a luxury you can’t afford. Take this blog for example. What took you a few minutes to read took me weeks to write because in addition to my PhD, I work for below minimum wage (how corporations get away with doing this is another story for another time!). Any time I find hours to write or do my own research, those hours missing from doing (barely) paid work are reflected at the end of the month and have real lifestyle consequences next month. I know I’m not the only one - many in our EA group have similar issues. EA has not, however, at this time any funding for individual research pieces. Only longer projects and these are focused on individuals founding organisations or wide-ranging projects, not scoping projects. EA organisations, in terms of career opportunities, have very few jobs at very high salaries, and nothing for early careers. You would expect to have a number of jobs at low salaries designed for early careers individuals - but alas, no. This has real consequences for attrition. In our EA on the Tyne group, we had a very talented AI technical researcher specialising in forecasting unable to financially afford his PhD anymore so he quit and now works in forecasting for an oil company. Another, whose interest lay in using drones to prevent disease in agriculture, had a similar issue and now works for a major UK weapons manufacturer. I can't help but feel that more of a focus on linking people from less fortunate backgrounds with other options could help reduce attrition. They're still part of the EA community, but obviously do very unaligned work!

A lack of EA recruitment in many less affluent universities, a lack of early-careers job prospects (even minimum wage entry-level jobs), and fellowships that fail to take into account variable backgrounds (trying going to Oxford for 6-10 weeks and using the stipend to pay your rent, bills, and support for family who remain behind), a lack of EA organisation engagement with beginning researchers, and a lack of EA infrastructure specifically for preventing socioeconomic attrition means that we are losing talent to private industry who aren’t making that mistake. To have half your local EA group join oil companies and weapons manufacturers just to put food on their table, and then hear an EA talk decrying the ‘talent bottleneck’ is very frustrating. I’m sure some readers are judging those people right now - but ethics is easy on a full belly in a warm room.

When I travel to a conference, and sleep on a bus, and walk for 90 minutes to get to the conference, to sit in a gilded room full of oil paintings and listen to people talk loudly and often about how EA has a huge financial overhang but a talent bottleneck, I can’t help but think that maybe the reason is because we’re only interested in the talent from certain, small, demographics of our society, and we’re unwilling to use any of that financial overhang to solve the problems. I detail potential useful fixes in another blog post here. No matter that fixes we put in place, if we’re to solve our talent bottleneck we need to address where we source our talent from by casting a wider net, and we need to stop attrition for members of our community whose altruistic work is strangled by the realities of their birth lottery. There are many people within the Effective Altruism community who can barely afford to keep their contribution to humanity's future going, and instead are poached by private, non-aligned organisations who can promise a better life.

I am willing to suffer what it takes because my research is in an area I feel could save billions from future suffering, and in fact thousands are already succumbing to the early stages of this suffering risk. Whether I have to sleep on a bus, walk for hours to a conference with a rucksack on, go hungry, or whatever I'll make it work. Hell - I'd walk from Newcastle to London or Oxford if it was the only way. I do this because the alternative for my grandchildren and their grandchildren, none of whom exist yet, is harm on a whole other level than mere sore feet or lack of sleep. This isn't because I'm more altruistic than anyone else, it's just because I believe so strongly in what I'm doing. The issue is that for every person like me, there are dozens more who for various factors just couldn't make it work. It's an unacceptable source of attrition that needs addressed if we are to increase intellectual diversity and gain a better understanding of how the world's problems affect the most vulnerable communities.

Again, I know the major criticism for this piece will be along the lines of “It’s not EA’s job to solve your poverty problems”.  And that’s okay. Because BAE Systems have a different opinion, and they’re always hiring.

Below is a video reading of this article, along with extra thoughts and exploration of the themes. Already read the article and just want the extra verbal discussion? Skip to around the 21 minute mark!