I think my if-the-stars-align minimum is probably around £45k these days. But then it starts going up once there are suboptimal circumstances like the ones you mention. In practice I might expect it to land at 125% to 250% of that figure depending how the non-salary aspects of the job look. 

I'm curious about the motivation of the question; FWIW my figure here is a complicated function of my expenses, anticipated flexibility on those expenses, past savings, future plans, etc. in a way that I wouldn't treat it as much of a guide to what anyone else would or should say. 

It does indeed depend a lot. I think the critical thing to remember is that the figure should be the minimum of what it costs to get a certain type of talent and how valuable that talent is. Clean Water is worth thousands of dollars per year to me, but if you turned up on my doorstep with a one-year supply of water for $1k I'd tell you to stop wasting my time because I can get it far more cheaply than that. 

When assessing the cost of acquiring talent, the hard thing to track is how many people aren't in the pool of applicants at all due to funding constraints. That sounds like it's Abraham's position and I think it's more common than often given credit for; there's something very low-status in EA about saying 'I could be doing this more impactful thing, but I won't because it won't pay me enough'. 

Funding isn't the only constraint on salaries of course; appearances matter too. Once your org is paying enough that you can't really pay more without getting a lot of sideways glances you don't want to get, that's when I would mostly stop calling you funding-constrained* and then I imagine this number can get really really high; cost of talent becomes ~infinite and we're back to looking at 'value'. Open Phil's hiring is perhaps in approximately that position. 

If you are still in a position where you could raise salaries if it weren't for funding constraints, I tend to think this number struggles to make it out of low six figures. Possible exceptions are positions that want a very specific combination of skills and experiences, like senior leadership at central EA orgs. 

*Assuming you are mostly turning money into people into impact, rather than e.g. money into nets into impact. 

I got very lucky that I was born in a city that is objectively one of the best places in the world to do what I do, so reasons to move location are limited.

More generally I don't feel like I'm doing anything particularly out of the ordinary here compared to a world where I am not donating; I like money, more of it is better than less of it, but there are sometimes costs to getting more money that outweigh the money. Though I would say that as you go up the earnings curve it gets easier and easier to mitigate the personal costs, e.g. by spending money to save time. 

Perhaps the biggest risk is that if I set my marginal tax + donation rate too high I am insufficiently incentivised to earn more, to the detriment of both me and the world. Still working on that one.

This really depends how broadly I define things; does reading the EA Forum count? In terms of time that feels like it's being pretty directly spent on deciding, my sense is ~50 hours per year. That's roughly evenly split between checking whether the considerations that inform my cause prioritisaition have changed - e.g. has a big new funder moved into a space - and evaluating individual opportunities. 

I touched on the evaluation question in a couple of other answers. 

It's either my 2014 donations to 80k or my 2015 donations to Charity Science, which eventually evolved into AIM. Both orgs were pretty small at the time, from memory we were >15% of their budgets in those years.

My views have not changed directionally, but I do feel happier with them than I did at the time for a couple of reasons:

• I thought and continue to think that the best argument is some version of 'clever arguments aside, from a layperson perspective what you're doing looks awfully similar to what caused the GFC, and the GFC was a huge disaster which society has not learned the lessons from'.
  • If you talk to people inside finance, they will usually reject the second claim and say a huge amount has changed since the GFC.
  • In particular, regulatory pressure shifted many 'interesting' risks from too-big-to-fail banks to hedge funds and firms like Jane Street (JS), where I used to work. JS arguably has much better incentives to keep its house in order than the big banks did, and it shouldn't have any call on public funds if it fails to do so.
  • But of course there was a reasonable question of whether JS and its ilk would actually succeed in doing this. And if they failed, would society pick up the tab somehow. As more time passes, the GFC looks more like the outlier event here.
• On the positive side of the ledger, most of my work at JS was improving the pricing of equity ETFs. When I started there I felt like almost nobody I spoke to outside JS knew what an ETF was and when I explained it they couldn't really see the point. Now I feel like virtually all UK personal financial advice I see will mention ETFs as a solid option; a cheap and simple way to invest in a diversified fashion. I'm fine with having been a very small part of what made that happen. 

With my more recent work it seems much too soon to say anything definitive about social impact, so I always try to acknowledge some chance that I'll feel bad when I look back on this.

It has varied. Giving both of us half the budget is in some ways most natural but we quickly noticed it was gameable to the extent we can predict each other's actions, similar to what is described here. At the moment we're much closer to 'discuss a lot and fund based on consensus'. 

Even with attempts to prevent it, I think annual risk of value drift for me is greater than the annual expected real return on equities, which tends to defeat the usual argument for giving later. 

Another exercise I've done occasionally is to look at my donations from say 5-10 years ago and muse on whether I would rather have invested the money and given now. So far that hasn't been close to true, and that's in spite of an impressive bull market in stocks over the last decade. Money was just so much more of an issue back then. I thought this from Will MacAskill was a good reflection on just how money-constrained things were 'in the old days', for those who didn't see it:

At the time, there was very little funding available in EA. Lunch was the same, every day: budget baguettes and plain hummus. The initial salaries offered by CEA were £15,000/yr pre-tax. When it started off, CEA was only able to pay its staff at all because I loaned them £7,000 — my entire life savings at the time. One of our first major donations was from Julia Wise, for $10,000, which was a significant fraction of the annual salary she received from being a mental health social worker at a prison. 

Of course the directly relevant question is: When I look back on this relatively abundant period in 2034, will I feel the same way? I honestly don't know, realistically the biggest factors will be what Good Ventures decides to do with their money and how well EA attracts other money. 

I sometimes think about whether we have or should have language for a mental health equivalent of Second-Impact syndrome. At the time I burned out I would say I was dealing with four ~independent situations or circumstances that most people would recognise as challenging, but my attitude to each one was 'this is fine, I can handle this'. Taken one at a time that was probably true, all at once was demonstrably false. 

Somehow I needed to notice that I was already dealing with one or two challenging situations and strongly pivot to a defensive posture to avoid taking on even more, similar to how you shouldn't risk a second concussion shortly after your first one. I think this is a more helpful takeaway than 'well make fewer missteps next time', which was my first reaction. 

I have taken that posture once or twice since then, but the nature of doing prevention correctly is that you never really know if what you're doing is actually making a difference. Still, it seems the logical thing to do.

This was a surprising question to me, because that's not how I think about my donations. I think there are a few things going on there:

• I only listed the four largest recipients that account for around 2/3rds of the total, so smaller orgs were naturally not listed.
  • As it happens another cluster I very nearly mentioned was AIM. I've donated roughly 150k (10% of donations) to AIM / AIM's predecessors / AIM-incubated charities.
• At the time I gave to 80k, in 2014-2018, they were much less of an 'established institution' and much more of a fast-expanding startup.
• Similarly, the bulk of the money I gave to GWWC was right as they were spinning up again in 2021.

Still, even with that said I think you're pointing at something real; my approach has certainly been more institution-heavy than say EAIF. So to show some cards, I'm roughly of the opinion that the 'give a promising individual a one-off grant' setup that's common in EA is pretty dysfunctional and I'm reluctant to pour much money into it. It requires a lot of evaluation time on the funder side and causes a lot of pain and heartache on the grantee side. There's a reason most people prefer to be employees not contractors, and I think even contractors have way better average job security than the individuals and tiny orgs inside this model. 

Then I'm also not sold the outputs are better as they would need to be to justify those non-financial costs. In the private sector my understanding is that large orgs are generally more productive than smaller ones, at least up to sizes way larger than all EA orgs, due to fixed overheads and better ability to allow employees to specialise. 

None of this is a hard no, and clearly I have done this sometimes. AIM mitigates some of these issues about as well as I think they can be mitigated. Sometimes I'm just much closer to a grant opportunity than other funders and it makes sense for me to take it. But mostly I do end up preferring to give to institutions.