Doubling costs to get +10% labour doesn’t seem like a great deal
I agree in principal, but in this case the alternative is eliminating$400k-4M of funding, which is much more expensive than doubling the salary of e.g. a research assistant.
To be clear, I am more so skeptical of this valuation than I am actually suggesting doubling salaries. But conditional on the fact that one engaged donor entering the non-profit labor force is worth >$400k, seems like the right call.
If there are now 10 staff, each paid $100k, and each generating $1m of value p.a., then the net gain is $10m - $1m = $9m. The CBR is 1:9.
If we double salaries and get one extra staff member, we’re now paying $2.2m to generate $11m of value. The excess is $8.8m. The average CBR has dropped to 5:1, and the CBR of the marginal $1.2m was actually below 1.
Agreed, just a function of how many salaries you assume will have to be doubled alongside to fill that one position
(a) Hopefully, doubling ten salaries to fill one is not a realistic model. Each incremental wage increase should expand the pool of available labor. If the EA movement is labor-constrained, I expect a more modest raise would cause supply to meet demand.
(b) Otherwise, we should consider that the organization was paying only half of market salary, which perhaps inflated their ‘effectiveness’ in the first place. Taking half of your market pay is itself an altruistic act, which is not counted towards the org’s costs. Presumably if these folks chose that pay cut, they would also choose to donate much of their excess salary (whether pay raise from this org, or taking a for-profit gig).
On b), for exactly that reason, our donors at least usually focus more on the opportunity costs of the labour input to 80k rather than our financial costs—looking mainly at ‘labour out’ (in terms of plan changes) vs. ‘labour in’. I think our financial costs are a minority of our total costs.
On a), yes, you’d need to hope for a better return than a doubling leads to +10% labour estimate I made.
If we suppose a 20% increase is sufficient for +10% labour, then the new situation would be:
Total costs: $1.32m
Impact: $11m
So, the excess value has increased from $9m to $9.7m, and the CBR of the marginal $320k is about 1:3. So, this would be worth doing, though the cost-effectiveness is about a third of before. (In our case at least, I don’t think a +20% increase to salaries would lead to +10% more hires though.)
It looks like the breakeven point is roughly an 80% increase in salaries to gain 10% of labour with this simplified model. (I.e. the CBR of the marginal $1m is around 1:1). In reality I don’t think we’d want to go that close to the breakeven point—because there may be better uses of money, due to the reputation costs of unusually high salaries, and because salaries are harder to lower than to raise (and so if uncertain, it’s better to undershoot).
In reality I don’t think we’d want to go that close to the breakeven point—because there may be better uses of money, due to the reputation costs of unusually high salaries, and because salaries are harder to lower than to raise (and so if uncertain, it’s better to undershoot).
Good points, I agree it would be better to undershoot.
Still, even with the pessimistic assumptions, the high end of that $0.4-4M range seems quite unlikely.
Does 80k actually advise people making >$1M to quit their jobs in favor of entry-level EA work? If so, that would be a major update to my thinking.
Does 80k actually advise people making >$1M to quit their jobs in favor of entry-level EA work?
It depends on what you mean by ‘entry level’ & relative fit in each path, but the short answer is yes.
If someone was earning $1m per year and didn’t think that might grow a lot further from there, I’d encourage them to seriously consider switching to direct work.
I.e. I think it would be worth doing a round of speaking to people at the key orgs, making applications and exploring options for several months (esp insofar as that can be done without jeopardising your current job). Then they could compare what comes up with their current role. I know some people going through this process right now.
If someone was already doing direct work and doing well, I definitely wouldn’t encourage them to leave if they were offered a $1m/year earning to give position.
The issue for someone already in earning to give is that probability that they can find a role like that which is a good fit for them, which is a long way from guaranteed.
Shouldn’t the displacement value be a factor though? This might be wrong, but my thinking is (a) the replacement person in the $1M job will on average give little or nothing to effective charity (b) the switcher has no prior experience or expertise in non-profit, so presumably the next-best hire there is only marginally worse?
The estimates are aiming to take account of the counterfactual i.e. when I say “that person generates value equivalent to extra donations of $1m per year to the movement”, the $1m is accounting for the fact that the movement has the option to hire someone else.
In practice, most orgs are practicing threshold hiring, where if someone is clearly above the bar, they’ll create a new role for them (which is what we should expect if there’s a funding overhang).
I agree in principal, but in this case the alternative is eliminating$400k-4M of funding, which is much more expensive than doubling the salary of e.g. a research assistant.
To be clear, I am more so skeptical of this valuation than I am actually suggesting doubling salaries. But conditional on the fact that one engaged donor entering the non-profit labor force is worth >$400k, seems like the right call.
Not sure I follow the maths.
If there are now 10 staff, each paid $100k, and each generating $1m of value p.a., then the net gain is $10m - $1m = $9m. The CBR is 1:9.
If we double salaries and get one extra staff member, we’re now paying $2.2m to generate $11m of value. The excess is $8.8m. The average CBR has dropped to 5:1, and the CBR of the marginal $1.2m was actually below 1.
Agreed, just a function of how many salaries you assume will have to be doubled alongside to fill that one position
(a) Hopefully, doubling ten salaries to fill one is not a realistic model. Each incremental wage increase should expand the pool of available labor. If the EA movement is labor-constrained, I expect a more modest raise would cause supply to meet demand.
(b) Otherwise, we should consider that the organization was paying only half of market salary, which perhaps inflated their ‘effectiveness’ in the first place. Taking half of your market pay is itself an altruistic act, which is not counted towards the org’s costs. Presumably if these folks chose that pay cut, they would also choose to donate much of their excess salary (whether pay raise from this org, or taking a for-profit gig).
On b), for exactly that reason, our donors at least usually focus more on the opportunity costs of the labour input to 80k rather than our financial costs—looking mainly at ‘labour out’ (in terms of plan changes) vs. ‘labour in’. I think our financial costs are a minority of our total costs.
On a), yes, you’d need to hope for a better return than a doubling leads to +10% labour estimate I made.
If we suppose a 20% increase is sufficient for +10% labour, then the new situation would be:
Total costs: $1.32m
Impact: $11m
So, the excess value has increased from $9m to $9.7m, and the CBR of the marginal $320k is about 1:3. So, this would be worth doing, though the cost-effectiveness is about a third of before. (In our case at least, I don’t think a +20% increase to salaries would lead to +10% more hires though.)
It looks like the breakeven point is roughly an 80% increase in salaries to gain 10% of labour with this simplified model. (I.e. the CBR of the marginal $1m is around 1:1). In reality I don’t think we’d want to go that close to the breakeven point—because there may be better uses of money, due to the reputation costs of unusually high salaries, and because salaries are harder to lower than to raise (and so if uncertain, it’s better to undershoot).
Good points, I agree it would be better to undershoot.
Still, even with the pessimistic assumptions, the high end of that $0.4-4M range seems quite unlikely.
Does 80k actually advise people making >$1M to quit their jobs in favor of entry-level EA work? If so, that would be a major update to my thinking.
It depends on what you mean by ‘entry level’ & relative fit in each path, but the short answer is yes.
If someone was earning $1m per year and didn’t think that might grow a lot further from there, I’d encourage them to seriously consider switching to direct work.
I.e. I think it would be worth doing a round of speaking to people at the key orgs, making applications and exploring options for several months (esp insofar as that can be done without jeopardising your current job). Then they could compare what comes up with their current role. I know some people going through this process right now.
If someone was already doing direct work and doing well, I definitely wouldn’t encourage them to leave if they were offered a $1m/year earning to give position.
The issue for someone already in earning to give is that probability that they can find a role like that which is a good fit for them, which is a long way from guaranteed.
That all seems reasonable.
Shouldn’t the displacement value be a factor though? This might be wrong, but my thinking is (a) the replacement person in the $1M job will on average give little or nothing to effective charity (b) the switcher has no prior experience or expertise in non-profit, so presumably the next-best hire there is only marginally worse?
The estimates are aiming to take account of the counterfactual i.e. when I say “that person generates value equivalent to extra donations of $1m per year to the movement”, the $1m is accounting for the fact that the movement has the option to hire someone else.
In practice, most orgs are practicing threshold hiring, where if someone is clearly above the bar, they’ll create a new role for them (which is what we should expect if there’s a funding overhang).