In general, people are paid more when their work is highly valuable and scarce. Is this still true within the EA non-profit space? I’d imagine this data is harder to interpret because successful wages often require paying people enough to get them to work for a non-profit rather than a for-profit company. Also, different cause areas receive different amounts of funding, which might make this tricky. Still, if I’m doing career research to find high-impact jobs, would this be a good thing for me to look at?
To decompose your question into several sub-questions:
Should you defer to price signals for cause prioritization?
My rough sense is that price signals are about as good as the 80th percentile EA’s cause prio, ranked by how much time they’ve spent thinking about cause prioritization.
(This is mostly because most EAs do not think about cause prio very much. I think you could outperform by spending ~1 week thinking about it, for example.)
Should you defer to price signals for choosing between organizations within a given cause?
This mostly seems decent to me. For example, CG struggled to find organizations better than Givewell’s top charities for near-termist, human-centric work.
Notable exceptions here for work which people don’t want to fund for non-effectiveness reasons, like politics or adversarial campaigning.
Should you defer to price signals for choosing between roles within an organization?
Yes, I mostly trust organizations to price appropriately, although also I think you can just ask the hiring manager.
I think this is an interesting question! I think you’re right to point out some of the factors that influence it including cause area, role type (and payment norms for them). I also think organizational cultural norms affect it quite heavily.
My guess is that if you had a large enough dataset and controlled for enough factors, salary would predict ‘role leverage’ quite well. But I don’t expect it to be very useful when choosing between roles to apply for, because the correlation will be weak, your dataset is too small etc. Basically, there are too many predictors and too much noise for it to be very informative. I think you’re better off just reading the descriptions or using other heuristics like cause area, job title etc if you’re trying to filter quickly.