To my mind, the main question isn't whether there will be poor people with the same marginal utility in consumption as those who exist now, but whether the people trading off the consumption of the poor today with their consumption in the future are discounting the future too heavily. If interest rates are positive, their marginal utility in consumption can fall, just more slowly than the interest rate lets your money grow.
One thing to add is that charitable opportunities can be unforseen and temporary, so saving money to spend on e.g. natural disasters is good.
In the end the solution might be something like having uncertainty across all parameters and allocating money to each time period proportional to the expected value (and updating as time goes on).
The two reasons I donate now: 1) future me is not present me, and 2) healthcare interventions might end up with fewer low-hanging fruit.
On point one, I consider myself to be somewhat playing "against" my future self. See for example how a lot of vegans -- committed, ethically-centered vegans who think "meat is murder" -- stop being vegan sometime in the future. Likewise, even though I'm very into global health *now*, there is a very high chance that 40-year-old-me has values that diverge strongly from twentysomething-present-me. If I save to donate later, there is not only the regular boring kind of risk (I invest poorly and end up with less to give, there is some major geopolitical event that changes the whole landscape that makes my savings worthless, etc), but also the much bigger risk that future-me is unaligned to present-me's desires, and therefore spends my savings on a house or a local (inefficient) food pantry or something. Therefore, since someone who is plausibly unaligned to what I want is getting control of my invested money, I'm better off donating now (or as soon as possible so my values have minimal time to drift).
This isn't an issue with personal-consumption savings; presumably future-me will still want food and heating and a roof over their head, so I don't need to worry about major drift in personal-consumption earmarked savings.
On point two, there are people dying of malaria right now who could be helped by insecticide-treated bed nets, but it seems plausible that within a few years a vaccine or other wide-scale intervention makes antimalarial treatments significantly less necessary. As the current low-hanging fruit is used up, my future altruism purchases might get less return per dollar.
OP already addresses your two points. On the first, he pointed that you could put your money in a trust that is mandated to invest then donate. On the second, it relates to the hypothesis that there will still be poor people in the future. You convincingly challenge this hypothesis.
Perhaps I misread the post, but I didn't see it as fully understanding my objections.
On the first, a Donor Advised Fund (the one way I know of locking money up for future charitable donations in the US) wouldn't stop future me from using it all in ways very unaligned with my current EA goals, like an inefficient local food pantry as I mentioned.
For the second, the fact that there will always be poor people does not tell me whether the rate at which I can purchase QALYs will change significantly. There will almost certainly be poor people in 2040, but (unless things go really poorly) I see no reason to believe that I could still save a life for the equivalent of $500 in 2025 USD, and therefore any gains from investing could be wiped out by future opportunities being less tractable.
Another point in favour of saving: Capital investment has good side effects. When you put $10000 into index funds, that indirectly creates more ability for businesses to invest in more capital. That allows them to provide more goods and services, and that itself makes the world a better place.
If you could find a way to invest your money profitably in Africa, that can very well create significantly higher utility for Africans
Would we not expect the marginal utility of philanthropic dollars to go down over time, even if the absolute numbers of the poor stay the same (assuming the philanthropic sector grows along with the economy)? A growing philanthropic sector (or at least the 'effective' part of it) would exhaust more 'low hanging fruit' like curing easy-to-cure infectious diseases and move onto lower-marginal-benefit activities like perhaps treating harder-to-cure non-communicable diseases.
the argument that future charity is equal in moral worth to current charity isn’t just flawed, it’s wrong. The reason is that saving someone’s life today pays returns tomorrow. Everyone that they go on to help, every good that they go to produce, every economy they go on to stimulate is a return on that investment. And we know that the value of investment compounds with time above all else
I don't think this is true. For example, here is a quote from Rory Stewart (head of GiveDirectly and former head of the UK's foreign aid agency).
"For, let us say, $100,000 [in direct cash transfers] for a typical village in Rwanda, you can find within about three months: the village goes from 40% electrification to 80% electrification; that livestock ownership goes up to 70%; that everybody ends up with the latrine. Everybody ends up with government health insurance, education. Enrollment goes up, nutrition goes up, small businesses are created."
It seems impossible to me that this wouldn't have persistent growth effects at the level of the village, and if scaled up, at the level of the country.
More importantly, as I mentioned in my other comment, I don't think that we shouldn't think about long-run growth rates as the primary thing that matters. I think long-run rates and levels are fixed at the trajectory of frontier growth, and instead we should be trying to stimulate catch up growth to that trajectory as quickly as possible.
I beg to differ. It's reasonable to assume that by investing in education and health in deprived areas, you'll have higher returns on these donations than by investing in already well-funded S&P500 firms.
However, there is another argument to delay (the counterpart of Sam Harsimony's argument not to delay): that as time passes, I learn more/better opportunities to use my donation, so I'd better wait (perhaps until I die) to donate to the best opportunity I've found.
Are you including all the extra descendants that may result from saving someone's life, and the contributions of all the descendants?
I guess fertility rates will eventually peak and then drop below replacement everywhere, though (at least without dramatically reduced relative costs for having children, which may result from advances in technology, but either way, labour would be mostly automated).
I think you need to consider the effects of charity on fertility/population growth.
Most EA charities don't just increase consumption, they save lives, and these people will then go on to have more children (whose lives I assume are likely worth living). The whole model you wrote assumes that population is fixed, and the philanthropist chooses to improve quality of life for some number of people now or for a larger number later. But if the philanthropist is able to increase the number of lives in the future by saving one today, this assumption is wrong. Now donating early looks better, you can improve lives today and increase the number of positive lives that exist in the future.
(Your population model could survive this critique, if it assumes that we have a surplus of children, but the number that survive is determined by resource constraints. Then the population would be relatively inelastic to total births, and saving lives today wouldn't increase population in the future. But that is a very unrealistic model: saving lives increases labor and therefore resources, and more importantly this filtering mechanism seems extremely unrealistic.)
Secondly when we are modeling macro spillover effects, my instinct is that we shouldn't think just about growth rates or levels, but instead in terms of catch up growth. In this case, catching up earlier rather than later is better because it decreases the total amount of time spent far beneath the frontier.
Lastly, you make a Pareto distribution assumption, in which the proportion of poor people always remains large. I simply don't believe that is accurate. Instead, I think that over the past few decades poor people have rapidly become richer and that will continue. If true this, means that we should aggressively spend on poverty relief, because poverty will never be a bigger problem than it is now.
That last paragraph is true in the aggregate, but if some number of extremely poor people will always exist, you as a micro donor might still argue for saving in order to engage in this "charity arbitrage" in which you exchange some poverty relief today for a larger amount tomorrow. I think my previous arguments give a good reason not to, but also I think there are reasons to believe that the best opportunities for poverty relief in a world with lots of poverty are more effective than the best opportunities in a world with less poverty. If this is true, then advocates for saving need to argue that the effectiveness of poverty relief programs will decline no faster than the interest rate.
Thanks a lot for these thoughts, that gave me important insights into a question I've long been interested in. Two comments to further the reflection:
- An argument to donate now is to foster system change, by funding a progressive force that would overthrow the far right and implement global sustainable development policies.
- I am not convinced by your argument that, as long as there are extreme-poor indefinitely, you should donate the return and roll on the capital. Indeed, if returns are higher on investments than on donations, wouldn't it be welfare-enhancing to delay donations (to donate more)? The paradox is that under these assumptions, it is always better to delay, so you end up never donating anything. I think the only way out of the paradox is to admit that these assumptions do not hold in our world: e.g., as other have pointed out in comments, there may be fewer poverty in the future; or returns on donations may be higher than on investments; (or perhaps this is uncertain and you should give at least one part in case the latter conditions hold).
Interesting topic which I had to think about a lot, as I grew older & more secure, fortunate in my natural gifts. And it became clear that Charity, whilst an altruistic quality, is at best a secondary condition of a decent & cogitative being.
Humanity has a flawed economic model. It is based on fang & claw yet we presume to be more than animal, yes? So the standard Bell Curve representation of statistical "intelligence" (excluding hereditary!,) generally represents a pyramid, predator at top, prey at bottom. Yet the BULK OF ECONOMIC WEALTH, & it's distribution comes down to two variables: State & Corpoatocracy. Re-read that. Ee are just Worker drones. Yet Charity is perverted by BOTH. Weaponised. Mal-distributed deliberately.
And here we are, banging on the drum of Individuals & their paltry contributions 🤷. A far more effective contribution is for real education of HUMANS, real, not corp, not gov prop. To overturn this entire false, dog eat dog goddamn awful world we have built, 'twould be FAR more useful to BREAK THE FUCKING SYSTEM and build an economic model for a highly educated, reduced population (balanced by choice & knowledge), existing in an unpolluted, high tech Startrek future.
Now I'd buy that fer a dollar! Forget $50 to a charity, we need to SHUN & OSTRASIZE all of the fkn Banksters, BigPharma DoD scum that want the Matrix. Build our world, money from War alone would remove every bloody hungry child in the world. Just sayn.
If you’re not already familiar with it, would be helpful to check out Phil Trammel’s work on Patient Philanthropy, which makes this/ a similar point.
Here's the paper I think:
https://philiptrammell.com/static/PatienceAndPhilanthropy.pdf
Thanks--here's an updated version (https://philiptrammell.com/static/Dynamic_Public_Good_Provision_under_Time_Preference_Heterogeneity.pdf)
and here's a report which discusses related issues as well (https://docs.google.com/document/d/1NcfTgZsqT9k30ngeQbappYyn-UO4vltjkm64n4or5r4/).
To my mind, the main question isn't whether there will be poor people with the same marginal utility in consumption as those who exist now, but whether the people trading off the consumption of the poor today with their consumption in the future are discounting the future too heavily. If interest rates are positive, their marginal utility in consumption can fall, just more slowly than the interest rate lets your money grow.
One thing to add is that charitable opportunities can be unforseen and temporary, so saving money to spend on e.g. natural disasters is good.
In the end the solution might be something like having uncertainty across all parameters and allocating money to each time period proportional to the expected value (and updating as time goes on).
The two reasons I donate now: 1) future me is not present me, and 2) healthcare interventions might end up with fewer low-hanging fruit.
On point one, I consider myself to be somewhat playing "against" my future self. See for example how a lot of vegans -- committed, ethically-centered vegans who think "meat is murder" -- stop being vegan sometime in the future. Likewise, even though I'm very into global health *now*, there is a very high chance that 40-year-old-me has values that diverge strongly from twentysomething-present-me. If I save to donate later, there is not only the regular boring kind of risk (I invest poorly and end up with less to give, there is some major geopolitical event that changes the whole landscape that makes my savings worthless, etc), but also the much bigger risk that future-me is unaligned to present-me's desires, and therefore spends my savings on a house or a local (inefficient) food pantry or something. Therefore, since someone who is plausibly unaligned to what I want is getting control of my invested money, I'm better off donating now (or as soon as possible so my values have minimal time to drift).
This isn't an issue with personal-consumption savings; presumably future-me will still want food and heating and a roof over their head, so I don't need to worry about major drift in personal-consumption earmarked savings.
On point two, there are people dying of malaria right now who could be helped by insecticide-treated bed nets, but it seems plausible that within a few years a vaccine or other wide-scale intervention makes antimalarial treatments significantly less necessary. As the current low-hanging fruit is used up, my future altruism purchases might get less return per dollar.
OP already addresses your two points. On the first, he pointed that you could put your money in a trust that is mandated to invest then donate. On the second, it relates to the hypothesis that there will still be poor people in the future. You convincingly challenge this hypothesis.
Perhaps I misread the post, but I didn't see it as fully understanding my objections.
On the first, a Donor Advised Fund (the one way I know of locking money up for future charitable donations in the US) wouldn't stop future me from using it all in ways very unaligned with my current EA goals, like an inefficient local food pantry as I mentioned.
For the second, the fact that there will always be poor people does not tell me whether the rate at which I can purchase QALYs will change significantly. There will almost certainly be poor people in 2040, but (unless things go really poorly) I see no reason to believe that I could still save a life for the equivalent of $500 in 2025 USD, and therefore any gains from investing could be wiped out by future opportunities being less tractable.
*$5,000 USD - can't seem to edit on mobile
Another point in favour of saving: Capital investment has good side effects. When you put $10000 into index funds, that indirectly creates more ability for businesses to invest in more capital. That allows them to provide more goods and services, and that itself makes the world a better place.
If you could find a way to invest your money profitably in Africa, that can very well create significantly higher utility for Africans
Would we not expect the marginal utility of philanthropic dollars to go down over time, even if the absolute numbers of the poor stay the same (assuming the philanthropic sector grows along with the economy)? A growing philanthropic sector (or at least the 'effective' part of it) would exhaust more 'low hanging fruit' like curing easy-to-cure infectious diseases and move onto lower-marginal-benefit activities like perhaps treating harder-to-cure non-communicable diseases.
the argument that future charity is equal in moral worth to current charity isn’t just flawed, it’s wrong. The reason is that saving someone’s life today pays returns tomorrow. Everyone that they go on to help, every good that they go to produce, every economy they go on to stimulate is a return on that investment. And we know that the value of investment compounds with time above all else
This is already covered. Nobody seriously believes that it affects growth rates.
I don't think this is true. For example, here is a quote from Rory Stewart (head of GiveDirectly and former head of the UK's foreign aid agency).
"For, let us say, $100,000 [in direct cash transfers] for a typical village in Rwanda, you can find within about three months: the village goes from 40% electrification to 80% electrification; that livestock ownership goes up to 70%; that everybody ends up with the latrine. Everybody ends up with government health insurance, education. Enrollment goes up, nutrition goes up, small businesses are created."
It seems impossible to me that this wouldn't have persistent growth effects at the level of the village, and if scaled up, at the level of the country.
More importantly, as I mentioned in my other comment, I don't think that we shouldn't think about long-run growth rates as the primary thing that matters. I think long-run rates and levels are fixed at the trajectory of frontier growth, and instead we should be trying to stimulate catch up growth to that trajectory as quickly as possible.
I beg to differ. It's reasonable to assume that by investing in education and health in deprived areas, you'll have higher returns on these donations than by investing in already well-funded S&P500 firms.
However, there is another argument to delay (the counterpart of Sam Harsimony's argument not to delay): that as time passes, I learn more/better opportunities to use my donation, so I'd better wait (perhaps until I die) to donate to the best opportunity I've found.
Are you including all the extra descendants that may result from saving someone's life, and the contributions of all the descendants?
I guess fertility rates will eventually peak and then drop below replacement everywhere, though (at least without dramatically reduced relative costs for having children, which may result from advances in technology, but either way, labour would be mostly automated).
I think you need to consider the effects of charity on fertility/population growth.
Most EA charities don't just increase consumption, they save lives, and these people will then go on to have more children (whose lives I assume are likely worth living). The whole model you wrote assumes that population is fixed, and the philanthropist chooses to improve quality of life for some number of people now or for a larger number later. But if the philanthropist is able to increase the number of lives in the future by saving one today, this assumption is wrong. Now donating early looks better, you can improve lives today and increase the number of positive lives that exist in the future.
(Your population model could survive this critique, if it assumes that we have a surplus of children, but the number that survive is determined by resource constraints. Then the population would be relatively inelastic to total births, and saving lives today wouldn't increase population in the future. But that is a very unrealistic model: saving lives increases labor and therefore resources, and more importantly this filtering mechanism seems extremely unrealistic.)
Secondly when we are modeling macro spillover effects, my instinct is that we shouldn't think just about growth rates or levels, but instead in terms of catch up growth. In this case, catching up earlier rather than later is better because it decreases the total amount of time spent far beneath the frontier.
Lastly, you make a Pareto distribution assumption, in which the proportion of poor people always remains large. I simply don't believe that is accurate. Instead, I think that over the past few decades poor people have rapidly become richer and that will continue. If true this, means that we should aggressively spend on poverty relief, because poverty will never be a bigger problem than it is now.
That last paragraph is true in the aggregate, but if some number of extremely poor people will always exist, you as a micro donor might still argue for saving in order to engage in this "charity arbitrage" in which you exchange some poverty relief today for a larger amount tomorrow. I think my previous arguments give a good reason not to, but also I think there are reasons to believe that the best opportunities for poverty relief in a world with lots of poverty are more effective than the best opportunities in a world with less poverty. If this is true, then advocates for saving need to argue that the effectiveness of poverty relief programs will decline no faster than the interest rate.
Important post
Thanks a lot for these thoughts, that gave me important insights into a question I've long been interested in. Two comments to further the reflection:
- An argument to donate now is to foster system change, by funding a progressive force that would overthrow the far right and implement global sustainable development policies.
- I am not convinced by your argument that, as long as there are extreme-poor indefinitely, you should donate the return and roll on the capital. Indeed, if returns are higher on investments than on donations, wouldn't it be welfare-enhancing to delay donations (to donate more)? The paradox is that under these assumptions, it is always better to delay, so you end up never donating anything. I think the only way out of the paradox is to admit that these assumptions do not hold in our world: e.g., as other have pointed out in comments, there may be fewer poverty in the future; or returns on donations may be higher than on investments; (or perhaps this is uncertain and you should give at least one part in case the latter conditions hold).
Interesting topic which I had to think about a lot, as I grew older & more secure, fortunate in my natural gifts. And it became clear that Charity, whilst an altruistic quality, is at best a secondary condition of a decent & cogitative being.
Humanity has a flawed economic model. It is based on fang & claw yet we presume to be more than animal, yes? So the standard Bell Curve representation of statistical "intelligence" (excluding hereditary!,) generally represents a pyramid, predator at top, prey at bottom. Yet the BULK OF ECONOMIC WEALTH, & it's distribution comes down to two variables: State & Corpoatocracy. Re-read that. Ee are just Worker drones. Yet Charity is perverted by BOTH. Weaponised. Mal-distributed deliberately.
And here we are, banging on the drum of Individuals & their paltry contributions 🤷. A far more effective contribution is for real education of HUMANS, real, not corp, not gov prop. To overturn this entire false, dog eat dog goddamn awful world we have built, 'twould be FAR more useful to BREAK THE FUCKING SYSTEM and build an economic model for a highly educated, reduced population (balanced by choice & knowledge), existing in an unpolluted, high tech Startrek future.
Now I'd buy that fer a dollar! Forget $50 to a charity, we need to SHUN & OSTRASIZE all of the fkn Banksters, BigPharma DoD scum that want the Matrix. Build our world, money from War alone would remove every bloody hungry child in the world. Just sayn.