Recently, 80,000 Hours wrote that they think only a small proportion of people should earn to give long term. They asked their team “At this point in time, and on the margin, what portion of altruistically motivated graduates from a good university, who are open to pursuing any career path, should aim to earn to give in the long term?” and got a median answer of 15%.
I’m not confident in my response, but when thinking about the movement as a whole, I’d suggest a ratio closer to 50%, if not as high as 80%.
The ratio of donation opportunity generated per staff of EA orgs suggests a need for a higher-earning to give ratio.
This consideration is definitely more speculative, but depending on how you look at the numbers, it’s possible to justify an earning-to-give ratio above 29% and possibly as high as 99%.
The Size of the Opportunity
The biggest example of generating a lot of donating opportunity per staff member is the Against Malaria Foundation. They only havetwo full-time staff, yet according to GiveWell their 2015 room for more funding was around $5M. This suggests that a very high talent EA direct worker could generate as much as $2.5M for earning-to-give people to attempt to fill with donations.
Give Directly has 25 staff and has a room for more funding of somewhere between $1M to $25M (with the potential for much more), which is somewhere between $40K to >$1M per staff member.
How Much Do People Earn to Give?
80K found in their 2014 survey that people doing earning-to-give were donating on average $13K/year for 2013, with an expectation to rise to $56K/year within three years if people's plans are taken at face value. A follow-up in 2015 found a new average of $16.5K/year for 2014, though for a more limited sample.
According to the EA Survey (focused on 2013 data), the mean donation in our sample from EAs who met the criteria for earning-to-give (>=$60K annual income and >=10% donations) was $9.5K/year.
Now Let’s Do Math
Using the best possible numbers for earning-to-give, imagine that an additional direct work person was like the AMF staff and generated $2.5M in donation opportunity. Now imagine I took the low value of earning-to-give, $9.5K/year. Using these two numbers, it would take 263 people earning to give to match one person doing direct work, suggesting that 99.6% of people should earn-to-give.
If instead I look at GiveDirectly’s low number of $40K per staff member compared to a high value of earning-to-give ($56K/year), I calculate you need 1 person earning to give for every 1.4 people doing direct work, suggesting a ratio of 29%.
The Problem With Funding Salaries
Of course, it does seem clear that marginal direct workers won’t be as productive at creating earning-to-give opportunity as AMF staff have been, so there will be less to fund per marginal direct worker.
Furthermore, AMF may not be typical of charities marginal direct workers are moving into. Charities like CEA or MIRI scale not by creating opportunities to fund programs, but rather through hiring staff, and this ends up with a lower amount of donation opportunity to staff member (equal to their salary).
Still, even if we expect earning-to-give people to be primarily focused on funding salaries, we may need more than 15%. Consider the classic story of ETG who takes a job in finance to fund two direct workers, doubling their impact. First, if salaries + overhead per staff are ~$50K/yr (though they can be much lower), a 15% earning to give ratio would require the average earning to give donation to be over $330K, which seems unrealistically high, even for the next three years. A more realistic ratio where the average (not median) earning-to-give person donates $50K/year (still more than is currently happening) would be 1:1, or 50%.
Second, the AMF model should not be ignored when thinking about donation opportunity created per marginal direct worker, and it’s quite possible that in some scenarios we may need many earning-to-give people to support the opportunities created. It’s an open question what proportion of EA resources should be going to GiveWell’s scalable health charities, which often generate millions of dollars in donation opportunity per employee, than other EA orgs where the donation opportunity per employee is quite less. Those who think the majority of EA funding should focus on GiveWell’s top charities, or that GiveWell top charities should be the default giving opportunity for the typical donor absent good reason to donate elsewhere, should also be far more inclined to think that more people should earn-to-give.
EA Orgs Have More Room For More Funding Than People Are Giving Credit For
Recently I’ve been involved in some EA projects that I’ve wanted to fundraise for. On three occasions with three separate EAs the reaction was always “Huh, I’m really surprised that hasn’t been funded already.” ...And many of those projects still need money.
There is supposed to be an explosion in earning to give. But I don’t think it’s arrived quite yet. From my point of view many plausibly important projects still want more funding.
Just to mention a few projects I’m aware of --
-
According to Jacy’s speech at EA Global, Animal Charity Evaluators is quite funding constrained and would hire more researchers if they had the money.
-
The Machine Intelligence Research Institute believes they could have as much as $5.5M in room for more funding according to their fundraiser (though I don’t know if they think of themselves as funding constrained or not).
-
Charity Science feels talent constrained but also could really use more money for planned expansions we can’t afford yet.
-
Some people close to GBS Switzerland told me they feel funding constrained.
-
Several projects currently going through EA Ventures have not yet raised the funding that they wanted.
-
And, of course, GiveWell top charities have $500K to $23.4M in room for more funding according to the latest analysis.
-
And other projects I’m aware of are on the horizon and may make more public pitches soon.
...Of course, it’s possible you may not believe in supporting these projects. If so, depending on what you want to support, it’s possible the remaining room for more funding may be quite small. But the belief I’ve been hearing from some that all EA projects are getting fully funded all the time and that EA Ventures will simply take care of the rest is not currently true.
Earning to Give May Have Better Career Capital
Career capital, as 80,000 Hours defines it, are the skills, credentials, connections, and other benefits you get from a career that help you have more impact in the future. Since what matters is your total impact and not your current impact, it’s quite plausible that young people (including myself) should be focusing much more on career capital than having an impact now.
Earning to give careers usually have good career capital opportunities. Usually the most important skills (80K suggests computer programming, web development, statistics, machine learning, independent work, self-motivation, sales, communication, and management) are typically found in abundance in good for-profit careers that also happen to have good earning to give potential.
Yes, you can certainly find this capital in non-ETG careers, and I want to be careful to not fall into a charity vs. for-profit ETG dichotomy when many additional options exist (e.g., research, academia, and politics). But I generally believe that it is easiest to find career capital within earning-to-give careers and it becomes much easier to transition out of earning to give and into other careers than vice versa (with maybe the exception of academia).
Earning to Give Fits More People
Furthermore, psychologically, earning-to-give seems to me to be a better fit for the average EA than direct work. Many EAs are already working in a company and can simply move to donate more of their salary or focus on increasing their salary, rather than quit their job and start a new one. Furthermore, EA direct work is frequently concentrated in certain cities that require relocation, which can be a big choice.
Another advantage to earning to give is it’s often easier to accomplish for people who have less altruistic motivation or direct work talent. Of course, this certainly misses 80K’s point because they were talking about people on the margin and people who were high-talent and high-motivation who were open to many career paths. That is a different question. But in the actual movement, it’s important to note that many people are not in the top 0.1% of drive and talent. Yet they still have a valuable thing to contribute -- a share of their job through earning to give. It’s often much easier to get, keep, and do an earning-to-give job than it is to do EA direct work.
Lastly, earning to give frequently gives much more career capital than other jobs early on in one’s career, which would allow for people to launch more successful non-ETG careers in the future. It’s generally a lot easier to move from earning to give to not earning than vice versa, given constraints on personal savings and on various industries.
Conclusion
Obviously the choice between earning to give or something else is a personal one that is sensitive to your personal skills and fit. But I think if you have the ability to enter a particularly high-earning career, there are good reasons to do so, and more people should be considering it than the current wisdom seems to suggest.
There are certainly good reasons to think 80K’s argument is correct -- there is already a lot of money in EA through Good Ventures and other high net-worth individuals and there is a potential that existing earning-to-give EAs may start donating much more soon. I’m excited how much the EA movement might grow through more direct work. However, I think we are still a ways off until earning-to-give people start reliably funding the remaining 85% of the movement.
But until things change, I’d suggest that 15% seems to me to be quite small, and I’d suggest an earning-to-give ratio at 50% or even as high as 80% for the movement as a whole.
I agree it's not intuitive when you put it that way, but I think it makes sense:
Imagine a strange world where the highest impact thing to do is to turn this magical crank. Also forget about meta opportunities -- there are only two possible roles: turning the crank yourself (direct work) or funding the salaries of people who turn the crank (earning to give).
Forgetting a moment about psychological constraints and thinking only about pure impact, you ought to turn the crank yourself, because the more people turning the crank the more impact there is, and earning money itself doesn't turn any cranks.
However, the people turning the crank need to not starve to death, and they're requesting a frugal $25K/year to fund their lifestyles.
Currently there's $0 in this crank movement (two different puns intended), so we need an ETG person to fund some salaries.
Now imagine that the highest salary you can get is $50K/yr -- you take $25K/year for yourself and can fund one person full-time turning the crank. So for every one person turning the crank, you need one person ETG, or else nothing would happen. The next person to join the crank movement will do ETG, the person after that to join will turn the crank directly, and so on.
Now imagine that someone gets a superjob and can earn $50M/yr -- they also take $25K/year for themselves, but they have $49,975,000 left over for donations; enough to fund 1999 people to turn the crank. Now the next 1999 people to join should turn the crank directly, because we don't need any more money. None of the next 1999 people should earn to give.
When earning to give was worse, we needed more people doing it. When it got better, we needed much less people doing it. Thus, the better earning to give is, the less you should do it, generally speaking.
Does that make sense?
Thanks Peter, I agree that your insight about "when it's worse more people should do it" is correct and your analogy is helpful.
But what really confuses me is the specific example I quoted where you are looking at what an AMF staff person does versus what an E2G person does. It seems like in both cases you are comparing "donation opportunity", yet you are choosing the one which is worse.