“But I believe many EAs are resistant to raising salaries as a way to close talent gaps because they conflate willingness to work for low pay with fit for a job”
The claim I have heard most frequently is not this but rather that labor supply is inelastic below market rates. E.g. there are people who really want to work for you (because of your mission or the prestige or whatever), and to them being paid 60% market rate is basically the same as being paid 90% market rate. So raising your compensation from 60% to 90% market rate won’t actually attract more candidates.
(I modified your picture here to show this – you can see that the quantity supplied Qs is very close to the equilibrium quantity supplied Qs*.)
I don’t know if this is actually true (it seems simplistic, at the least), but it seems consistent with everything you’ve written above yet still doesn’t imply EA organizations should focus on raising salaries.
It makes sense to me that labor supply would be relatively inelastic below market rates. But there are two (at least) nuances we should be aware of.
1) The degree of inelasticity should depend on where pay sits relative to market. If we assume candidates view money as offering decreasing marginal returns, that suggests raising pay from 60% to 70% of market will be more beneficial than raising it from 70% to 80% of market. But to understand what strategies will be effective, we need to collect money to understand where EA orgs are starting from.
2) Some people can afford to work for 60% of market, some people can’t even if they believe in the mission. As Khorton notes, that kind of pay will systematically exclude “the people who needed to care for their aging parents or pay off massive student loans or had other significant financial constraints.” I think the EA community has gotten caught up in the observation that “there are a lot of smart people willing to work for way below market” and lost track of the question of “what’s the right way to structure EA compensation to maximize impact?”
I think the EA community has gotten caught up in the observation that “there are a lot of smart people willing to work for way below market” and lost track of the question of “what’s the right way to structure EA compensation to maximize impact?”
Not sure I fully understand this. You’re saying something like: “it might be true that increasing wages will have only a small increase in the number of candidates, but those new candidates are unusually impactful (because they are from underrepresented groups) so it’s still worth doing?”
Thanks for flagging this, please let me know if this clarifies:
There are a bunch of smart people in the EA community for whom “being paid 60% market rate is basically the same as being paid 90% market rate.” Let’s call them Cohort 1.
But there are also a lot of people for whom it makes a huge difference whether they’re paid 60% or 90% of market; for them it might make all the difference in whether they can even consider the job. Let’s call them Cohort 2. In some cases, people in Cohort 2 look like people in Cohort 1 but with more student loans, poorer parents, or other differences that have nothing to do with ability or mission alignment. In other cases, people are in Cohort 2 for reasons that make them systematically different in important ways. For example, I’d expect Cohort 2 to have more people with high earning power than Cohort 1.
To oversimplify, I think a lot of EAs believe that organizations should hire people from Cohort 1 because that means they can get smart, cheap, mission-aligned people. I agree that gets you smart, cheap, mission-aligned people, but think that hiring from Cohort 2 (or a Cohort 1.5 that’s somewhere in between) might still be a better strategy. In other words, we should be asking questions like “does paying 60% or 90% of market lead to more impact over the long-term?”
But are you saying a) there are lots of people in cohort 2 who are great candidates so we should go after them, or b) there are not very many people in cohort 2, but because they are from some underrepresented demographic we should still go after them?
Thanks for writing this Jon!
“But I believe many EAs are resistant to raising salaries as a way to close talent gaps because they conflate willingness to work for low pay with fit for a job”
The claim I have heard most frequently is not this but rather that labor supply is inelastic below market rates. E.g. there are people who really want to work for you (because of your mission or the prestige or whatever), and to them being paid 60% market rate is basically the same as being paid 90% market rate. So raising your compensation from 60% to 90% market rate won’t actually attract more candidates.
(I modified your picture here to show this – you can see that the quantity supplied Qs is very close to the equilibrium quantity supplied Qs*.)
I don’t know if this is actually true (it seems simplistic, at the least), but it seems consistent with everything you’ve written above yet still doesn’t imply EA organizations should focus on raising salaries.
Thanks Ben!
It makes sense to me that labor supply would be relatively inelastic below market rates. But there are two (at least) nuances we should be aware of.
1) The degree of inelasticity should depend on where pay sits relative to market. If we assume candidates view money as offering decreasing marginal returns, that suggests raising pay from 60% to 70% of market will be more beneficial than raising it from 70% to 80% of market. But to understand what strategies will be effective, we need to collect money to understand where EA orgs are starting from.
2) Some people can afford to work for 60% of market, some people can’t even if they believe in the mission. As Khorton notes, that kind of pay will systematically exclude “the people who needed to care for their aging parents or pay off massive student loans or had other significant financial constraints.” I think the EA community has gotten caught up in the observation that “there are a lot of smart people willing to work for way below market” and lost track of the question of “what’s the right way to structure EA compensation to maximize impact?”
Not sure I fully understand this. You’re saying something like: “it might be true that increasing wages will have only a small increase in the number of candidates, but those new candidates are unusually impactful (because they are from underrepresented groups) so it’s still worth doing?”
Thanks for flagging this, please let me know if this clarifies:
There are a bunch of smart people in the EA community for whom “being paid 60% market rate is basically the same as being paid 90% market rate.” Let’s call them Cohort 1.
But there are also a lot of people for whom it makes a huge difference whether they’re paid 60% or 90% of market; for them it might make all the difference in whether they can even consider the job. Let’s call them Cohort 2. In some cases, people in Cohort 2 look like people in Cohort 1 but with more student loans, poorer parents, or other differences that have nothing to do with ability or mission alignment. In other cases, people are in Cohort 2 for reasons that make them systematically different in important ways. For example, I’d expect Cohort 2 to have more people with high earning power than Cohort 1.
To oversimplify, I think a lot of EAs believe that organizations should hire people from Cohort 1 because that means they can get smart, cheap, mission-aligned people. I agree that gets you smart, cheap, mission-aligned people, but think that hiring from Cohort 2 (or a Cohort 1.5 that’s somewhere in between) might still be a better strategy. In other words, we should be asking questions like “does paying 60% or 90% of market lead to more impact over the long-term?”
Cool. I think that is a helpful segmentation.
But are you saying a) there are lots of people in cohort 2 who are great candidates so we should go after them, or b) there are not very many people in cohort 2, but because they are from some underrepresented demographic we should still go after them?
I’m arguing for point A.