3 ms·

I'm probably missing something, but I have some serious concerns about the effectiveness of this approach to charity. First of all, without some coordinated ef

I'm probably missing something, but I have some serious concerns about the effectiveness of this approach to charity. First of all, without some coordinated effort to spread out donations, won't some of the more effective charities still suffer? If everyone gives to the top 4 charities, for example, won't the 5th most effective one suffer? I don't believe this movement yet has the traction behind it to have this kind of negative effect, but won't it get worse as the movement grows? Second of all, who decides what is the most effective cause (or who interprets the studies that indicate which cause is most effective)? If I could spend 1 dollar and know it would save 5 lives or prevent 50 people from living with horrible mental illness, I would have trouble deciding which one to give to, even if given empirical evidence as to which is cause is the most beneficial to society. Any field dealing with this many human variables is bound to produce at least some skewed results. To me, it seems like the trolley problem with a limitless number of tracks with different situations. Third of all, some of the causes that are not effective might still benefit from some amount of involvement. The example they give - volunteering at a soup kitchen - is one such example. Say giving to the homeless is effective, but volunteering at a soup kitchen is not. Someone still needs to volunteer at the soup kitchen, even if it's not many people.

Kalium 8y ago
I can see you don't miss much. You've made some excellent points! > First of all, without some coordinated effort to spread out donations, won't some of the more effective charities still suffer? This is almost certainly true. However, it's also true that this is the case in any scenario where there are less than infinite resources. The starvation issue is very real today. With that said, I think adherents to EA hope that their movement will incentivize effectiveness in charities. Some might postulate that there are worse possible outcomes than charities competing for effectiveness. > Second of all, who decides what is the most effective cause (or who interprets the studies that indicate which cause is most effective)? If I could spend 1 dollar and know it would save 5 lives or prevent 50 people from living with horrible mental illness, I would have trouble deciding which one to give to, even if given empirical evidence as to which is cause is the most beneficial to society. This is another excellent point! How people pick causes is a real concern. It's also one people face today. I think it's largely still a matter of personal choice and priorities. The EA approach tries to put a framework in place for explicitly reasoning about what people currently tend to do implicitly. It absolutely is a trolley problem with infinite tracks and infinite victims. Yet, we already live that trolley problem. > Third of all, some of the causes that are not effective might still benefit from some amount of involvement. The example they give - volunteering at a soup kitchen - is one such example. Say giving to the homeless is effective, but volunteering at a soup kitchen is not. Someone still needs to volunteer at the soup kitchen, even if it's not many people. Not to nitpick your example, but a sufficiently well-funded soup kitchen could likely afford to hire some of the homeless to staff it. Then it could be free of the need for volunteers and lift people off of having to live on the streets! It wouldn't be many people, but it also seems like it might be preferable to staffing the kitchen entirely with unpaid volunteers. More broadly, you're once again absolutely right. This is a real and pressing concern in a scenario where Effective Altruism becomes literally the only way any person thinks about charity. However, some might postulate that we are not currently in that scenario.
Paul-ish 8y ago
I suspect, but don't know for sure, that most charities experience diminishing marginal utility for each dollar received after some point. Therefore if people give too much to a single charity, it's rank will fall. If there is not diminishing utility, you would have a utility monster[1], which would cause the problem you pointed out. [1] https://en.m.wikipedia.org/wiki/Utility_monster https://en.m.wikipedia.org/wiki/Utility_monster
- gatheredprior 8y ago
  > I suspect, but don't know for sure, that most charities experience diminishing marginal utility for each dollar received after some point. This is absolutely true. If you're a charity and your primary activity is distributing anti-malarial bed nets, once you have enough bed nets to meet demand, your going to have to move on to other activities that won't necessarily be exactly as effective as the bed nets. In a contrived thought experiment with "perfect" information about the effectiveness of activities, EA would suggest a "Greedy" (algorithm) style of allocation. For each incoming dollar, you send it to whichever activity has the greatest utility per dollar give, and repeat that for as many dollars as you have.
readams 8y ago
GiveWell actually uses this in their rankings; the charity needs to have the ability to scale up with more money.
lkbm 8y ago
Your first paragraph is something I think about from time to time, but as you point out, we're not to the point where it's a problem. Right now we're in a competitive market where we select for emotional impact and marketing prowess. EA, for me, is about moving to a market where we select for effectiveness. We should, once we have the resources as a movement, include funding for experimenting with new organizations. FWIW, it's worth noting that Givewell.org (the EA organization I pay the most attention to) includes "room for more funding" as a major decision point in its recommendations. Funding is a constraint for most organizations, but they recognize that dropping a hundred million dollars on a small organization isn't always the best thing, even if that organization does the best work. > The example they give - volunteering at a soup kitchen - is one such example. Say giving to the homeless is effective, but volunteering at a soup kitchen is not. Someone still needs to volunteer at the soup kitchen, even if it's not many people. 80000hours.org talks about this some, mostly in the consideration of working in direct service vs. earning to give. Part of this is what you're passionate about -- if high paying professions would make you miserable, working in direct service might be for you. Part of it is opportunity. Not everyone is cut out to be a lawyer/doctor/engineer. I do disagree with "someone still needs to volunteer". Someone needs to volunteer or get paid to work at the soup kitchen. If the choice is between having a doctor spend their time volunteering, or having that doctor donate an extra $100k/year, and $40k of that paying someone to replace the missing volunteer time, the soup kitchen is still up $60k, and will probably get better results from having someone experienced doing the work. In the way we traditionally evaluated charities, this would be a worse outcome, because the overhead just went way up. One of EA's main things is to stop worrying about overhead. When I buy a car, I want the best car per dollar, not the car from whoever pays the CEO the least. When I buy good, I want the most good I can get for my dollar. I disagree with other EAs in a ton of areas, but what makes us EAs is that one principle.

Kalium 8y ago

I can see you don't miss much. You've made some excellent points! > First of all, without some coordinated effort to spread out donations, won't some of the more effective charities still suffer? This is almost certainly true. However, it's also true that this is the case in any scenario where there are less than infinite resources. The starvation issue is very real today. With that said, I think adherents to EA hope that their movement will incentivize effectiveness in charities. Some might postulate that there are worse possible outcomes than charities competing for effectiveness. > Second of all, who decides what is the most effective cause (or who interprets the studies that indicate which cause is most effective)? If I could spend 1 dollar and know it would save 5 lives or prevent 50 people from living with horrible mental illness, I would have trouble deciding which one to give to, even if given empirical evidence as to which is cause is the most beneficial to society. This is another excellent point! How people pick causes is a real concern. It's also one people face today. I think it's largely still a matter of personal choice and priorities. The EA approach tries to put a framework in place for explicitly reasoning about what people currently tend to do implicitly. It absolutely is a trolley problem with infinite tracks and infinite victims. Yet, we already live that trolley problem. > Third of all, some of the causes that are not effective might still benefit from some amount of involvement. The example they give - volunteering at a soup kitchen - is one such example. Say giving to the homeless is effective, but volunteering at a soup kitchen is not. Someone still needs to volunteer at the soup kitchen, even if it's not many people. Not to nitpick your example, but a sufficiently well-funded soup kitchen could likely afford to hire some of the homeless to staff it. Then it could be free of the need for volunteers and lift people off of having to live on the streets! It wouldn't be many people, but it also seems like it might be preferable to staffing the kitchen entirely with unpaid volunteers. More broadly, you're once again absolutely right. This is a real and pressing concern in a scenario where Effective Altruism becomes literally the only way any person thinks about charity. However, some might postulate that we are not currently in that scenario.

Paul-ish 8y ago

I suspect, but don't know for sure, that most charities experience diminishing marginal utility for each dollar received after some point. Therefore if people give too much to a single charity, it's rank will fall. If there is not diminishing utility, you would have a utility monster[1], which would cause the problem you pointed out. [1] https://en.m.wikipedia.org/wiki/Utility_monster https://en.m.wikipedia.org/wiki/Utility_monster

gatheredprior 8y ago
> I suspect, but don't know for sure, that most charities experience diminishing marginal utility for each dollar received after some point. This is absolutely true. If you're a charity and your primary activity is distributing anti-malarial bed nets, once you have enough bed nets to meet demand, your going to have to move on to other activities that won't necessarily be exactly as effective as the bed nets. In a contrived thought experiment with "perfect" information about the effectiveness of activities, EA would suggest a "Greedy" (algorithm) style of allocation. For each incoming dollar, you send it to whichever activity has the greatest utility per dollar give, and repeat that for as many dollars as you have.

readams 8y ago

GiveWell actually uses this in their rankings; the charity needs to have the ability to scale up with more money.

lkbm 8y ago

Your first paragraph is something I think about from time to time, but as you point out, we're not to the point where it's a problem. Right now we're in a competitive market where we select for emotional impact and marketing prowess. EA, for me, is about moving to a market where we select for effectiveness. We should, once we have the resources as a movement, include funding for experimenting with new organizations. FWIW, it's worth noting that Givewell.org (the EA organization I pay the most attention to) includes "room for more funding" as a major decision point in its recommendations. Funding is a constraint for most organizations, but they recognize that dropping a hundred million dollars on a small organization isn't always the best thing, even if that organization does the best work. > The example they give - volunteering at a soup kitchen - is one such example. Say giving to the homeless is effective, but volunteering at a soup kitchen is not. Someone still needs to volunteer at the soup kitchen, even if it's not many people. 80000hours.org talks about this some, mostly in the consideration of working in direct service vs. earning to give. Part of this is what you're passionate about -- if high paying professions would make you miserable, working in direct service might be for you. Part of it is opportunity. Not everyone is cut out to be a lawyer/doctor/engineer. I do disagree with "someone still needs to volunteer". Someone needs to volunteer or get paid to work at the soup kitchen. If the choice is between having a doctor spend their time volunteering, or having that doctor donate an extra $100k/year, and $40k of that paying someone to replace the missing volunteer time, the soup kitchen is still up $60k, and will probably get better results from having someone experienced doing the work. In the way we traditionally evaluated charities, this would be a worse outcome, because the overhead just went way up. One of EA's main things is to stop worrying about overhead. When I buy a car, I want the best car per dollar, not the car from whoever pays the CEO the least. When I buy good, I want the most good I can get for my dollar. I disagree with other EAs in a ton of areas, but what makes us EAs is that one principle.