I saw a CEA role advertised for twice the salary of an AMF job, whereas they do not seem to differ dramatically in expected level of expertise / experience. Even if one believes they can make more impact at AMF, they would have to give up 20k pounds in salary to pass on the content specialist role. We learned recently to consider earning less, but this may still be quite the conundrum. What do you think?
Hi John,
[Context: I work in People Ops for CEA and have been involved in designing our compensation strategies.]
Thanks for posting about this. I can give more context on CEA’s side, at least.
With regards to the required level of skill/experience, I’d say we consider them different. We are aiming to hire for the content role as a specialist. We currently have four broad bands (excluding the ED/CEO whose salary is set by the board), and the specialist band is second to the top . “Junior” would be the bottom band for us. (That said, I’d still expect us to offer someone joining at the junior level 10-25% more than what’s listed as the AMF salary above, if they were in the UK.)
Another potential factor that might contribute to the size of this pay gap is that I believe some orgs. also compensate their operational staff at lower rates than other staff. I don’t know whether this is true for AMF. CEA has chosen not to do this as a) we think their work is comparably valuable to non-operational staff b) deciding who was in an “operations” role is not as straightforward for as it might be for some organizations.
Another factor here might be location. We adjust salaries for location and it seems like AMF might not, although I’m not sure. If we were to hire someone in a place with a lower cost of living than the locations listed, we would offer a lower rate of pay than what’s listed above.
In general, we aim to have salaries that allow us to attract top talent, with the goal that for most roles, people don’t have to take a massive pay cut compared to their counterfactual options to join us. As we’ve wanted to attract more mid-career people, this has become increasingly important to us. Figuring where the balance point is between compensating sufficiently highly such that salaries aren’t so low they are dissuasive and conservatively enough we’re not wasting money or causing bad ecosystem effects is difficult and I’m not sure we’re getting it right.
(Looking at AMF, it appears the bulk of their staff have “operations” in their job title or appear to be operations-oriented.)
Against malaria foundation is an NGO which is EA affiliated, so would have a big market outside of EA where there is more expectation that working for an NGO means a lower salary. They may also consider a lower standard of applicant than the direct EA job.
Also they may just be very different jobs it’s not clear.
This doesn’t answer the conundrum in any way sorry!
Relatedly, a significant fraction of viable candidates for the Content Specialist job live in high COL areas like the Bay, NYC, etc. A salary like the one AMF is offering basically rules people in those areas out. That’s a bigger problem when you’re starting with a smaller candidate pool (EA-specific vs. broader NGO). It makes sense that AMF isn’t willing to pay extra to include high-COL candidates in the pool, and that CEA would be.
Many orgs hesitate to create large pay differentials on a candidate’s location, especially when there is no business advantage to location. So one effect of posting a salary that is competitive (or competitive enough) for the Bay is that the salary offered for lower COL areas will be generous by local standards.
Donors to AMF likely have different expectations about charitable spending than donors to CEA in that donors to the former likely care more about minimising overhead, while donors to CEA care more about the job being done well even if they have to pay more for it.
To everyone who downvoted this: please explain why you think it’s wrong.
(edit: the comment above was at −7 agreement when I wrote this)
A glance at AMF’s financial data, cross-checked against a recent GiveWell Form 990, suggests ~ half of AMF’s revenue flows through GiveWell. Given GiveWell’s own compensation practices, and its well-known practices for evaluating charities, the idea that GiveWell (and allied donors who route donations through GiveWell) cares more about overhead than effectiveness seems questionable. Although the source of the other ~ half of AMF’s funding is unclear to me, the assertion about AMF donor preferences needs support in light of GiveWell’s prominence and given the alternative explanations posted elsewhere on the thread. Furthermore, AMF’s administrative costs of ~ 1% are covered by a small set of private donors, reducing the probability that AMF thinks its mainline donors care as much about them on the margin.
In addition, AMF has always been funding constrained. CEA is now tightening the purse strings a bit, but my understanding is that it has felt non-financial constraints more than financial ones in the past, which would partially explain a looser hand on spending. Moreover, it is difficult to change an existing salary structure over the short run. Donor-focused explanations seem significantly less likely here than organization-focused or labor pool-focused ones to me.
CEA is widely perceived as grossly ineffective whereas AMF is perceived as highly effective.
CEA has three major projects I’m aware of, and executes poorly on all of them.
EA Forum
The glorified subreddit you’re reading this on costs $2M per year to run, ~$30 per hour of user engagement on non-community posts. By their own admission, they could run it for 10x less than this.
EA Globals
They won’t reveal what they’ve spent on them, but there’s good reason to think it’s too much. At EAG London, they had three staff standing behind each 1m x 4m table. I asked 5 of them to rate their boredom on a scale of 1-10. The lowest answer I got was 12. They insist on throwing workshops, which the vast majority of users perceive to be a waste of time—it’s the one-to-one’s that produce the value. If they embraced that fact, they’d be able to choose much cheaper venues.
Community Building and Health
Their own community builders perceive them as incompetent. Most people are too scared to criticize them under their own names because they fear reprisals. You likely underestimate how poorly they are perceived for this reason. EAs don’t trust them, which makes them the worst people you could put in charge of community health.
Other Effects
They’re directly or indirectly responsible for most of EA’s scandals, Ni****gate to Castlegate. They were in the best position to prevent FTX’s ponzi scheming and did nothing, they may even have been complicit.
Some good points here, can I nominate you for the EA criticism panel? Also I’m curious to hear more about how CEA caused the scandal where Nick Bostrom wrote an email in 1990.
CEA staff have so much karma that they can tank any comment with a single downvote. Don’t think that the agreement vote score is representative of the community’s opinion.
A recurrent disappointment with EA Forum comments related to the karma system is the frequency with which they assume that a particular explanation of the karma score of a given comment is correct, without even considering alternative explanations. One doesn’t have to be very creative or contrarian to think of plausible reasons users may have disagreed-voted with the statement “CEA is widely perceived as grossly ineffective whereas AMF is perceived as highly effective” other than that “CEA staff have so much karma that they can tank any comment with a single downvote”.
In my limited experience any post about lower salaries and overhead is likely to get disagree votes, even if the post itself makes a lot of sense and isn’t necessarily supporting the lower salaries.
They don’t? You didn’t include job descriptions, but from the titles alone I would guess content specialist is the harder job to find good candidates for.
The following statement by the acting head of CEA Online (the branch which includes the CS) may help explain some of the difference:
So I think the CS salary may be pulled upward by the need to keep people who live in expensive US cities in the applicant pool, and (possibly) by a desire to avoid significant disparities between tech and non-tech employees at CEA.
Are we sure the domains being compared are similar in supply and demand? Ex. Low expertise/experience swe make more than similarly experienced ui designers
Different organizations have different compensation philosophies and compensation strategies, which results in different tactics. It is hard to say anything with confidence unless we know all the details of each of these two roles, including the context they function in. If we believe that the world is a perfectly fair and just place, then there are probably various compensable factors present in one job that the other lacks (maybe it is higher stress, or has higher standards of performance, or requires a wider array of knowledge). But there is also a lot of ‘noise’ in reality, so there are probably some arbitrary factors that affect the pay also.
Compensation is a complicated topic that everyone has an opinion about and yet which very few people have done any research/reading about.[1] I’ve only read a little bit about it so far. But if you would like to learn a bit more the book Pay, by Kevin F. Hallock is a very good introduction. The review article Compensation, Benefits, and Total Rewards: A Bird’s-Eye (Re)View also gives a decent overview if you’d rather read 20 pages than 200.
So in that sense it is similar to raising children, or finding love, or other topics that people have widely varying views on.
The funny thing is, (I don’t have any inside info here, this is all pure speculation) I wouldn’t be shocked if the AMF position ended up being more competitive of the two, despite the lower salary.
Normally non-profit salaries tend to be lower than for-profit salaries even when skills are equivalent because people are altruistic and therefore willing to work for less if the work is prosocial and meaningful and fits their interests. For example, the position of a professor is more competitive and has a higher skill requirement than industrial R&D jobs in the same field, but the latter is more compensated.
I believe that in the EA community this effect is much more pronounced. Some (not all) EA orgs are likely in a somewhat unique position where even if you offer 28-35k and less, you may still be getting applications from people with degrees from top schools who could be making double or triple or quadruple that on the open market, and at some point when you notice that your top applicants are mostly motivated by impact and not money you might become uncertain as to whether or not offering more money actually further improves the applicant pool.
In such an environment, salaries aren’t exactly set by market forces in the same way as a normal job. Instead, they are set by organizational culture and decision making. This is likely all the more true for remote roles, where lack of geography constraints makes expected pay among equally skilled candidates even more variable.
Some people see the situation and think “so let’s spend less money and be more cost effective, leaving more money for the beneficiaries and the impact and attracting more donations from our reduced overhead”. They aren’t in it for money, and they figure all the truly committed candidates who make the best hires aren’t either.
Other people see the situation and and think “nevertheless, let’s pay people competitive rates, even if we could get away with less” either out of a sense of fair play (e.g. let’s not underpay people doing such important work just because there are some people who are altruistic enough to accept that, that’s not nice) or the golden rule (e.g. they themselves want a competitive salary and would feel weird offering an uncompetitive one) or or because they figure they will get better candidates or more long term career sustainability and personnel retention that way.
One of these perspectives is probably the correct one for a given use-case, but I’m not sure which one and reasonable people seem to diverge.
(Of course it’s not just personal preference—some organizations, and some positions inside organizations, have more star-power than others and so have this option more. And it’s also a cause area thing—some cause areas perceive themselves as more funding-bottlenecked where every $2 in salary is one less mosquito net, while others aren’t implicitly making that comparison as much, pouring extra money into their project wouldn’t necessarily improve their project and the true bottleneck is something else.
As far as the personal conundrum goes, I guess you have to ask yourself how much you value earning more, and consider if you’d be willing to pay the difference to buy the greater impact that you’d achieve by taking the position you believe is higher impact to take.
Should the hypothetical candidate consider absolute impact in the two positions, or impact relative to the next-best candidate? My hunch is that the additional marginal impact of having the best candidate in the CEA role (as opposed to the second-best) may be higher than the additional marginal impact of having the best candidate in the AMF role (as opposed to the second-best). This is in part due to my prediction that the AMF position’s candidate pool is much larger than that of the CEA position.
I would agree that impact calculations are only improved by considering concepts like counterfactual / additional marginal and so on.
I would caveat that when you’re not doing calculations based on data, but rather more informal reasoning, it’s important not to overweight this idea and assume that less competitive positions are probably better—it might easily be that the role with higher impact when your calculation is naïve to margins and counterfactuals, remains the role with higher impact after adding those calculations in, even if it is indeed the more competitive role.
I think for most people when it comes to big cause area differences like CEA vs AMF, what they think about the big picture regarding cause areas will likely dominate their considerations. Your estimate would have to fall within a very specific range before adjustments for counterfactual additionality on the margin would be a consideration that tips the scale, wouldn’t it?