I read the ‘After one year of applying for EA jobs: It is really, really hard to get hired by an EA organisation’ thread. 


From memory, EA Cambridge could fill a ~300 person auditorium for a statistics lecture. Hence my Fermi estimate for the UK EA graduate pool is as follows.  Assume 1/3 of students graduate every year from the top 20 universities each with 300 interested students. Then there is a supply on the order of 2000 competent, committed EA grads each year (fudging in a number of other things). Not all of whom will apply for every EA non-technical role, but many will. 


At the moment there are order of 10 interesting looking non-technical roles in the UK on the High Impact Job Board.

So, the math here will work something like musical chairs with (maybe 10-20% of) 2000 people and 10 chairs. Those are not good odds.


Next, I checked up some of the organisations people mentioned when discussing operations roles. E.g. Open Phil and CEA. Most organisations have ‘Team’ pages where you can see who is working in similar roles. Now of the three Operations profiles I checked all had had previous experience. One had been CEO of another charity!  Another EA org has an Executive Assistant posting open. The three administrative team members listed on their site include: someone who ran a $100m business, a barrister and someone from Formula 1. However, the job posting does not mention any experience requirements.

That’s a high bar. And it’s not explicit.

This is to say… there is no point in recent graduates beating themselves up about not winning that game. I’m saying this because many organisations are too nice, in that they don’t want to explicitly put anyone off by saying, “stellar experience required.”


Another point is that roles at well known EA focused orgs may not be an effective choice on the margin for an individual. Given the math, it seems highly unlikely that a non-technical role would go unfilled by a quality candidate. So the ‘neglectedness’ criterion may well be unmet.


Policy prescription:

  • Get technical. Compete for roles with smaller pools of qualified candidates. 
  • Get experience. Become one of the people who meet the unwritten bar, elsewhere
  • Stay elsewhere. In the context of the whole economy, there are not many EAs, and there is much left undone.

I was going to write something like ‘get creative: start your own thing’. But to do that seems to require solving the same type of problem.

Comments8
Sorted by Click to highlight new comments since: Today at 9:32 PM

I agree with the thrust of the argument here but I think your estimate for the size of the EA-aligned graduate pool is far too large.

I helped run the group at Warwick (top 10 uni) for a couple of years. For each year I was on the committee, I would be surprised and very happy if more than 5 graduates identified as 'committed EAs'. I would also say that Warwick has one of the more active groups outside of Oxbridge. My fermi for the size of the EA grad pool each year would therefore be something more like:

Oxford and Cambridge: 200 (this strikes me as high but I'll defer to you)

~10 other unis with active groups: 50

~10 other unis with small groups: 20 (most uncertain about this number)

10-20% of those graduates actually applying for non-technical EA roles seems about right so I think the number is more like 25-50. The resultant ratio is still undesirable so I've no doubt that there are many grads out there having difficulty getting hired which is saddening and made all the more visceral by the recent post.

I ran the EA Berkeley group and later the UWashington group, and even this estimate seems high to me (but it would be within my 90% confidence bound, whereas 2000 is definitely not in it).

Thanks. Interesting perspective. I'd love to know how many applications these orgs get on average for such roles. I guess my bet implies low hundreds, and I'd have to recalibrate if it were tens or thousands.

Hundreds of EA applicants? Most EA org roles don't have that... I've been in/around MIRI, Ought, FHI and many other EA orgs. It's common to have about a hundred applicants for a role (research or ops) and the number of EA applicants is usually in the tens.

How do you know out of that hundred whether the applicant is EA motivated or not? It's not the kind of thing I would put in a cover letter (when it could be substituted for a more concrete statement about why the org's work is important). Nor would it be obvious on a CV unless they had led a chapter or something. Although clearly p*100 <= 100...

Cover letters to core EA orgs from EAs generally indicate interest in EA. It's sometimes also indicated by involvement in EA groups, through a CV, by referral sources, and by interviews. You can pretty reliably tell.

After the incredible response to the "really, really hard to get hired" thread, there's definitely plenty of room for ongoing analysis and advice like this. Love your 3 policy prescriptions.

Although you say that perhaps organisations are 'too nice' in not putting people off, and yes candidates can do their own investigation, perhaps by not being more explicit EA orgs are unintentionally doing the opposite of being nice, by contributing to a lot of wasted time and emotional trauma. If an organisation is genuinely never going to higher someone without experience for a particular role, then perhaps the nicest thing would be to write

"We will not consider applicants for this job with less than 2 years experience in x or y orz"

I can imagine a scenario where one line like this could save hundreds of hours of wasted application time, and also save a lot of heartache.

I don't think many EA organizations have implicit experience barriers of this kind. The roles that have been mostly hotly contested, with the most applicants, really are "generalist" positions (e.g. research associates at GiveWell or Open Phil, where more or less everyone [I think] gets to take the first basic work test, and credentials don't seem like they matter if your work tests are strong).

Meanwhile, many positions at CEA (where I currently work) are filled by people who'd never done that kind of work in a professional setting before, or who only had experience with some (but not all) aspects of the job. My job is to write and edit, and my only prior "paid" writing/editing experience was in random freelance work here and there plus a few bits of journalism I wrote as a college student in ~2013, adding up to maybe one year of "full-time" experience.

That said, there may be non-experience "barriers" that should come into play earlier, relating to a candidate's skill in a particular area... but that seems like "raise standards for your initial round of work tests", not "try to stop people from applying in the first place".