I think point one is simultaneously really important and also a really risky thing for EA to pursue. I think a lot of the current discussion around EAâs reputation has been really defensive/âreactive. Part of that is because weâre trying to put out controversy fires, but I also think that EAs are much more comfortable not interacting at all with illegible, traditional systems and undervalue what playing the game well could get us.
As a dramatic example, the UN spent $47 billion in 2021 on a hilariously imbalanced set of 17 development goals. These goals are revised every 15 years. We can try to grab a seat at the table in 2030 or 2045 and align the new goals with EA principlesâthat could mean billions dollars per year diverted to EA causes assuming current funding levels. (This is a very very conservative assumption btwâfunding appears to be increasing steadily YoY)
But on the flipside, would an organization that can get a seat at the UN table still be recognizably EA? Or will we have destroyed the heart of it to get there?
EA has some essential weirdnesses that will mean that itâll always be a black sheep in the nonprofit industrial complex, and I really donât want to see us lose those weird things for the sake of increasing reputation/âfunding. So itâll be a delicate balancing act.
I personally think that the ideal (but very difficult) way forward is to try to be (and also seem[1]) so staunchly ethical that it warps public opinion away from the rest of the nonprofit ecosystem and towards us, instead of trying to become part of the trad nonprofit in-group. I think the anti-slavery activist and quaker, Benjamin Lay, is an inspirational figure for this path.
To be explicit, this means stopping or setting much higher bars for doing things that are effective and actually are the rational things to do in a social vacuum but burn social capital. Buying castles, having really nice office spaces, paying above market rates for employees at EA charities, overly enthusiastic and conspicuously well funded university groups, etc.
My sense is that a lot of EAs think that trading off reputation to do these effective things is worth it, but my claim is that they think so because they donât realize how good the upside of having a good reputation is. I certainly didnât think of positive reputation as having any value before I started working at Samaritans; my model was that reputation is something that you strive to keep non-negative and then youâre basically good.
I also do want to state explicitly that even with the correct model of the upside of having a stellar reputation, it could still be more optimal in the long run to continue to do things that are slightly offputting. âReputation is pricelessâ is not literally true; it could be more effort than itâs worth to pursue for EA considering that we have a pretty deep stock of like, google softengs who feel very alienated from the trad nonprofit world etc
Man, I really wish I had more to say about point 4! Itâs so deeply counter-intuitive that people would literally risk eviction or starvation than go to places that treat them kind of crappily, but thatâs honestly the extent of the observation.
Could it be that these places actually told them that they couldnât help them, or tried to help but werenât able to solve their issues, rather than they decided to leave because they werenât being spoken to nicely enough or something? I too would predict that they would prioritise solving their urgent issues over feeling comfortable. It could also be that showing a lot of care indicates that you are correctly understanding their issues and needs and are better equipped to give helpful advice, so their avoidance of places that treat them badly does not necessarily signify distorted priorities.
To go into the weeds a little, a lot of this goes back to funding restrictions. As an example, if youâre helping people find employment, you generally operate using funding for that from the state, and the funding generally comes with stringent quotas and reporting requirements.
The expectation is, and what organizations generally do, is deal with the reporting requirements by requiring people to fill out a 3+ page intake form before they will book an appointment to talk to a person[1]. Samaritans chooses to talk to people right away, subtly probing for answers during the meeting (since itâs all stuff they need to know to help with the job search anyways) and have their staff fill out the forms themselves afterwards. This truly creates a lot more work on their end; you can maybe intuit that by choosing to internalize filling out the intake forms, Samaritans are spending 30% more time per client for the same amount of funding.
This sort of âinternalizationâ is standard across Samaritans ventures. It is a giant pain, but they do it because it removes barriers for so many people to access much needed services.
In a truly unfortunate number of cases too, the forms are only available in English and the clients are new immigrants/ârefugees with limited English skills.
This also then loops right back to point 1; one reason they can afford to spend more time per client is that people are willing to be paid under market rate to work at Samaritans, because of its well deserved great reputation.
One more tactic for reputation management: they have a salary policy where everyone in the entire org has salaries that are within 15% of each other, and the founders, who work regular 60 hour weeks, earn around 55k a year (CAD). The messaging around this is that they both want more of their funding to go towards helping others, and that they want to pay around the Canadian median salary, and it helps with increasing solidarity between Samaritans workers and the people they help. Donors love this.
EAs have this take that you should pay people well, and itâs a good take. It unfortunately also burns reputation.
Very interesting! I would want to see more evidence before concluding that EA salary practices affirmatively burn reputation, though. GiveWell and Open Phil look a bit higher than U.S. federal government to me, even after accounting for the fact that the feds donât adjust salaries enough in very high COL areas. But I sense that most other orgs pay at a somewhat lower scale . . . so on the whole EA pay may be roughly similar to US Government employee pay? Itâs not obvious to me how that rate of pay would affirmatively burn reputation.
Many, many people think of government workers as lazy parasites who are overpaid and hard to fire for no reason, so Iâm not sure thatâs a super useful comparison lol. I worked in the federal public service for a bit and experienced some of this firsthand :â)
Fair enough, although I think thats a function of their perception of employee quality and about of effort than the rates per se. I used it as a comparison point because the USG underpays at the high end of the talent market (so itâs under market rate for good talent) and because the pay scale is more uniform than the private sector and thus easier to compare with.
Do you have any suggestions or takes about ways to move towards the sketches for EA integration for points 1 and 4?
I think point one is simultaneously really important and also a really risky thing for EA to pursue. I think a lot of the current discussion around EAâs reputation has been really defensive/âreactive. Part of that is because weâre trying to put out controversy fires, but I also think that EAs are much more comfortable not interacting at all with illegible, traditional systems and undervalue what playing the game well could get us.
As a dramatic example, the UN spent $47 billion in 2021 on a hilariously imbalanced set of 17 development goals. These goals are revised every 15 years. We can try to grab a seat at the table in 2030 or 2045 and align the new goals with EA principlesâthat could mean billions dollars per year diverted to EA causes assuming current funding levels. (This is a very very conservative assumption btwâfunding appears to be increasing steadily YoY)
But on the flipside, would an organization that can get a seat at the UN table still be recognizably EA? Or will we have destroyed the heart of it to get there?
EA has some essential weirdnesses that will mean that itâll always be a black sheep in the nonprofit industrial complex, and I really donât want to see us lose those weird things for the sake of increasing reputation/âfunding. So itâll be a delicate balancing act.
I personally think that the ideal (but very difficult) way forward is to try to be (and also seem[1]) so staunchly ethical that it warps public opinion away from the rest of the nonprofit ecosystem and towards us, instead of trying to become part of the trad nonprofit in-group. I think the anti-slavery activist and quaker, Benjamin Lay, is an inspirational figure for this path.
To be explicit, this means stopping or setting much higher bars for doing things that are effective and actually are the rational things to do in a social vacuum but burn social capital. Buying castles, having really nice office spaces, paying above market rates for employees at EA charities, overly enthusiastic and conspicuously well funded university groups, etc.
My sense is that a lot of EAs think that trading off reputation to do these effective things is worth it, but my claim is that they think so because they donât realize how good the upside of having a good reputation is. I certainly didnât think of positive reputation as having any value before I started working at Samaritans; my model was that reputation is something that you strive to keep non-negative and then youâre basically good.
I also do want to state explicitly that even with the correct model of the upside of having a stellar reputation, it could still be more optimal in the long run to continue to do things that are slightly offputting. âReputation is pricelessâ is not literally true; it could be more effort than itâs worth to pursue for EA considering that we have a pretty deep stock of like, google softengs who feel very alienated from the trad nonprofit world etc
Man, I really wish I had more to say about point 4! Itâs so deeply counter-intuitive that people would literally risk eviction or starvation than go to places that treat them kind of crappily, but thatâs honestly the extent of the observation.
Could it be that these places actually told them that they couldnât help them, or tried to help but werenât able to solve their issues, rather than they decided to leave because they werenât being spoken to nicely enough or something? I too would predict that they would prioritise solving their urgent issues over feeling comfortable. It could also be that showing a lot of care indicates that you are correctly understanding their issues and needs and are better equipped to give helpful advice, so their avoidance of places that treat them badly does not necessarily signify distorted priorities.
Great post btw!
Thank you!
To go into the weeds a little, a lot of this goes back to funding restrictions. As an example, if youâre helping people find employment, you generally operate using funding for that from the state, and the funding generally comes with stringent quotas and reporting requirements.
The expectation is, and what organizations generally do, is deal with the reporting requirements by requiring people to fill out a 3+ page intake form before they will book an appointment to talk to a person[1]. Samaritans chooses to talk to people right away, subtly probing for answers during the meeting (since itâs all stuff they need to know to help with the job search anyways) and have their staff fill out the forms themselves afterwards. This truly creates a lot more work on their end; you can maybe intuit that by choosing to internalize filling out the intake forms, Samaritans are spending 30% more time per client for the same amount of funding.
This sort of âinternalizationâ is standard across Samaritans ventures. It is a giant pain, but they do it because it removes barriers for so many people to access much needed services.
In a truly unfortunate number of cases too, the forms are only available in English and the clients are new immigrants/ârefugees with limited English skills.
This also then loops right back to point 1; one reason they can afford to spend more time per client is that people are willing to be paid under market rate to work at Samaritans, because of its well deserved great reputation.
One more tactic for reputation management: they have a salary policy where everyone in the entire org has salaries that are within 15% of each other, and the founders, who work regular 60 hour weeks, earn around 55k a year (CAD). The messaging around this is that they both want more of their funding to go towards helping others, and that they want to pay around the Canadian median salary, and it helps with increasing solidarity between Samaritans workers and the people they help. Donors love this.
EAs have this take that you should pay people well, and itâs a good take. It unfortunately also burns reputation.
Very interesting! I would want to see more evidence before concluding that EA salary practices affirmatively burn reputation, though. GiveWell and Open Phil look a bit higher than U.S. federal government to me, even after accounting for the fact that the feds donât adjust salaries enough in very high COL areas. But I sense that most other orgs pay at a somewhat lower scale . . . so on the whole EA pay may be roughly similar to US Government employee pay? Itâs not obvious to me how that rate of pay would affirmatively burn reputation.
Many, many people think of government workers as lazy parasites who are overpaid and hard to fire for no reason, so Iâm not sure thatâs a super useful comparison lol. I worked in the federal public service for a bit and experienced some of this firsthand :â)
Fair enough, although I think thats a function of their perception of employee quality and about of effort than the rates per se. I used it as a comparison point because the USG underpays at the high end of the talent market (so itâs under market rate for good talent) and because the pay scale is more uniform than the private sector and thus easier to compare with.