Thanks! I was going to write an EA Forum post at some point also trying to clarify the relationship between the debate over “patient vs urgent longtermism” and the debate over giving now vs later, and I agree that it’s not as straightforward as people sometimes think.
On the one hand, as you point out, one could be a “patient longtermist” but still think that there are capacity-building sorts of spending opportunities worth funding now.
But I’d also argue that, if urgent longtermism is defined roughly as the view that there will be critical junctures in the next few decades, as you put it, then an urgent longtermist could still think it’s worth investing now, so that more money will be spent near those junctures in a few decades. Investing to give in, say, thirty years is still pretty unusual behavior, at least for small donors, but totally compatible with “urgent longtermism” / “hinge of history”-type views as they’re usually defined.
I was going to write an EA Forum post at some point also trying to clarify the relationship between the debate over “patient vs urgent longtermism” and the debate over giving now vs later, and I agree that it’s not as straightforward as people sometimes think.
It seems to me that there are roughly three relevant confusions/sources of confusion in discussions around patient philanthropy, patient longtermism, and investing to give. I’ll try to briefly describe them, and I’d be interested to hear if you or others think this is accurate.
1. “Patient philanthropy” has been confused with “philanthropists should invest to give later”
People—including me, sometimes, and 80k in the podcast description here—have used the term “patient philanthropy” to refer to the position that we should “invest” resources for future altruistic “spending” (rather than “spending” now). Or people have said things like “the arguments for patient philanthropy” when meaning the arguments for that position.
But in your write-up, you actually used the term “patient philanthropy” as something like a starting assumption, saying:
We will call someone “patient” if he has low (including zero) pure time preference with respect to the welfare he creates by providing a good.
You did argue that a patient philanthropist should “invest [resources] for use on future philanthropic projects”, but that wasn’t what patient philanthropy meant, and one could contest those arguments. (I’m not saying I do contest them; I think I remember mostly being in agreement.)
2. “Patient longtermism” wasn’t clearly defined, and one thing it might mean is quite distinct even from what people think patient philanthropy means
Ben Todd then introduced term “patient longtermism”, seeming to mean either:
“what people think Trammell meant by patient philanthropy, but now for longtermism”: i.e., longtermists should invest resources for use on future philanthropic projects
Something quite distinct: i.e., that now is not an especially “hingy”/”high-leverage” time.
I think Owen’s post is primarily arguing that this view is distinct from the idea that longtermists should invest resources for use on future philanthropic projects. I definitely agree. I think this view can potentially be a factor in whether longtermists should “invest” vs “spending” now, but there are various other factors as well.
Todd’s post on the topic seems to initially mean the second, then switch to the first, then switch to both:
One of the parts of effective altruism I’ve found most intellectually interesting recently is ‘patient longtermism’.
This is a school of thinking that takes longtermism seriously, but combines that with the idea that we’re not facing an unusually urgent threat to the future, or another urgent opportunity to have a long-term impact. (We may still be facing threats to the future, but the idea is that they’re not more pressing today than the threats we’ll face down the line.)
Broadly, patient longtermists argue that instead of focusing on reducing specific existential risks or working on AI alignment and so on today, we should expect that the crucial moment for longtermists to act lies in the future, and our main task today should be to prepare for that time.
It’s not a new idea –- Benjamin Franklin was arguably a patient longtermist, and Robin Hanson was writing about it by 2011 — but there has been some interesting recent research.
3. “Investing” has sometimes been interpreted as only “financial investments”
You explicitly stated in your write-up and your 80k interview that “investment” in this context doesn’t have to mean just financial investments. But it’s sometimes been interpreted as if it does just mean that.
Personally, I like to use one of the following framings:
Should we financially invest in order to give later, or give now?
I meant to define patient longtermism in terms of when you think the hinges are.
This will usually correspond to where you think the balance of object-level spending vs. investing/meta should be, but can come apart (e.g. uncertainty arguments could favour investing even if you think hinginess is going down).
I don’t think it should be defined in terms of not having a pure rate of time preference, since the urgent longtermists don’t have a pure rate of time preference either.
But overall all these definitions are pretty up in the air. It would be great if someone would like to take a more rigorous look.
But overall all these definitions are pretty up in the air. It would be great if someone would like to take a more rigorous look.
Agreed. When writing my post on these matters, I found it surprisingly hard to even just describe what I was talking about and what claims I was making/summarising. It’d be handy to have settled, consistent, clearly defined terms.
I don’t think it should be defined in terms of not having a pure rate of time preference, since the urgent longtermists don’t have a pure rate of time preference either.
Yeah, I agree with this too. I was just pointing out that, when introducing the term that “patient longtermism” seems to deliberately mirror, Trammell defined it as being about having no pure time preference. So if a term deliberately mirrors that one but then uses a different meaning, I think that can create confusion.
That said, I think how people have interpreted Trammell’s term—as about having the bottom-line belief that one should invest for future altruistic projects rather than spending now—is arguably more intuitive than how he defined it.[1] I say this because it seems to me more intuitive that “patient” should be a matter of waiting and doing something later (so a low overall discount rate, not just a low pure time discount rate). But someone with zero pure time preference could, under certain conditions, still want to spend on direct work now rather than waiting. (Not sure if I’ve explained that well.)
And it does seem good to have a term for the bottom-line belief that one should invest for future altruistic projects rather than spending now. So maybe we should just accept that “patient altruism” has evolved to mean that. (Though there is the issue that it implies the opposite is “impatient”, which sounds rude/dismissive.)
But then believing hinginess will be higher in future is a separate (though related) matter again. And I don’t think it’s as intuitive to call that “patience”, for the same reason I don’t think it’s super intuitive to call zero pure time preference “patience”; someone who believes hinginess will be higher in future could still want to spend on direct work now. I might want to say that believing hinginess will be higher in future will tend to push towards more “patience”, but that it isn’t itself “patience”.
It could perhaps be good to have a separate shorthand term for believing hinginess will be higher in future. (And btw, I do think it’s valuable that you/80k have highlighted the matter of how hinginess will change over time as a key uncertainty, and highlighted that not all longtermists believe the hingiest period is now/soon.)
[1] Maybe “patience” is an established term with a meaning similar to Trammell’s in philosophy/econ? I’m not sure.
This will usually correspond to where you think the balance of object-level spending vs. investing/meta should be
Do you mean you think beliefs about how hinginess* is changing over time should be, or in practice is, the primary factor driving how much someone thinks we should allocate to object-level spending relative to meta/investing?
If so, that doesn’t seem obviously true to me. As I mentioned above and explore further in this post, I think there are a bunch of other important factors.(These factors have been prominently considered in prior work on this topic, so it’s not a new thing I’m throwing into the picture; e.g., Michael Dickens’ post on what the philanthropic discount rate should be included some hinginess-related-things as factors, but also had other major factors.)
In brief (for explanation of terms, see the post):
How effectively can we “punt to the future”?
What would be the long-term growth rate of financial investments?
What would be the long-term rate of expropriation of financial investments? How does this vary as investments grow larger?
What would be the long-term “growth rate” from other punting activities?
Would the people we’d be punting to act in ways we’d endorse?
Which “direct” actions might have compounding positive impacts?
Do marginal returns to “direct work” done within a given time period diminish? If so, how steeply?
It seems to me like it could be totally reasonable to think that hinginess is going to decrease somewhat over time in expectation, but that it’s still worth pushing more towards meta/investing because:
One thinks we can get such a good “interest rate” on financial or movement-building-ish investments, and/or
One thinks because marginal returns to more direct work within a time period diminish steeply, and we’ve already hit that point for the upcoming time periods (the low-hanging fruit will be taken), but longtermism might fizzle out later so some people should do investing/meta so we can still take the low-hanging fruit then if that happens
And conversely, it seems to me like it could be totally reasonable to think that hinginess is going to increase somewhat over time in expectation, but that it’s still worth pushing more towards object-level spending because:
One thinks the “interest rate” would be too bad, perhaps because one thinks value drift will be too high, or our movement-building investments would be too tied to specific areas and we’re too unsure if those are the right areas to focus on
One thinks object-level spending is both good for the relatively low level of near-term hinginess (e.g., reducing x-risk in the next 10 years) and has substantial meta/investment-like benefits
I don’t think this would be best described like “They’re doing something that looks like object-level work, but they’re doing it for meta purposes” (though I definitely think that could sometimes happen), because the person is doing it for both purposes.
(This would perhaps be somewhat analogous to people who agree that there will or could be far more events/experiences that are morally relevant in the long-term future than in the present/near-term future, but still think we should focus on improving the present/near-term future, because predictably improving the long-term future is too hard.)
It also seems worth noting that a lot of people’s beliefs about hinginess might be very uncertain or not very action-relevant; they might expect hinginess to slightly increase, expect it to slightly decrease, expect it to go up and down in ways that are hard to predict or notice, or have no real expectations about it. That’d make it easier for other factors to play the dominant role.
I definitely do think how hinginess will change over time is one of the key factors. And it’s plausible to me that it’s the primary one. But it’s not obvious to me that it’s the primary one, and I imagine how important each factor is will vary between people.
*Personally, I prefer “leverage over the future”, or in second place “pivotality”.
But I’d also argue that, if urgent longtermism is defined roughly as the view that there will be critical junctures in the next few decades, as you put it, then an urgent longtermist could still think it’s worth investing now, so that more money will be spent near those junctures in a few decades.
Yes, I totally agree with this. Indeed a large part of what I was trying to say was that I’m more sympathetic to this strategy right now for “urgent longtermists” than “patient longtermists” (although it happens that I mostly still think it’s beaten by non-financial investment opportunities which will pay off soon enough).
[LMK if you found something I wrote confusing; I could consider editing to improve clarity.]
Indeed a large part of what I was trying to say was that I’m more sympathetic to this strategy right now for “urgent longtermists” than “patient longtermists”
For what it’s worth, as someone who read this post fairly quickly, I thought this point only came across as a minor aside in the grand scheme of your post (I think you only mention it in the last paragraph before your “some caveats”?).
I think you make the point here as well but I’m not sure if it came across as clearly as it could:
My current guess is to recommend spending rather than saving money at current margins to both patient and urgent longtermists; neither recommendation feels robust, but I’m actually a little more pro-spending for patient longtermists.
By the way I have strongly upvoted this post! I was aware of your overall point already but it was definitely useful to read your explanation and some of the more specific points. I agree that people have been too quick to jump to giving later if they identify as patient longtermists.
I guess “large part of what I was trying to say” was an overstatement; it’s an illustrative facet of the large thing (which is that there isn’t the obvious coupling).
Anyhow thanks for the pointer, I made some small language tweaks to the second part quoted (which hopefully help).
Sorry, no, that’s clear! I should have noted that you say that too.
The point I wanted to make is that your reason for saving as an urgent longtermist isn’t necessarily something like “we’re already making use of all these urgent opportunities now, so might as well build up a buffer in case the money is gone later”. You could just think that now isn’t a particularly promising time to spend, period, but that there will be promising opportunities later this century, and still be classified as an urgent longtermist.
That is, an urgent longtermist could have stereotypically “patient longtermist” beliefs about the quality of direct-impact spending opportunities available in December 2020.
In the abstract I agree that you could think that. But I’d make some of the same claims for the urgent longtermist as the patient longtermist: that some of the best investment opportunities are probably non-financial, and we should be trying to make use of those before going on to financial investments. (There’s a question about whether at current margins we’re already using them up.)
I think there are some principled reasons to be unsurprised if the best available non-financial investment opportunities are better than the best available financial investment opportunities. Financial investment is a competitive market; there are lots of people who have money and want more money, and so for a given risk tolerance (and without lots of work) you can’t expect to massively outperform what others are making.
There are also markets (broadly understood) competing for buy-in to worldviews. At first glance these might look less attractive to enter into, since they seem to be (roughly) zero-sum. But unlike the financial case, capital is not fungible across worldviews, so we shouldn’t assume that market forces mean that the returns from the best opportunities can’t get too good (or they’d be taken by others). And I’m not concerned about the zero-sum point, because I don’t think that the longtermist worldview is just an arbitrary set of beliefs; I think that it has ~truth on its side, and providing people with arguments plus encouraging them to reflect will on average be quite good for its market share (and to the extent that it isn’t, maybe that’s a sign that it’s getting something wrong). This is a pretty major advantage and makes it plausible that there are some really excellent opportunities available. Then I think growth over the last few years is evidence that at least some of the activities people engage in have really good returns; the crucial question is how much there are really good ones being left on the table.
Thanks! I was going to write an EA Forum post at some point also trying to clarify the relationship between the debate over “patient vs urgent longtermism” and the debate over giving now vs later, and I agree that it’s not as straightforward as people sometimes think.
On the one hand, as you point out, one could be a “patient longtermist” but still think that there are capacity-building sorts of spending opportunities worth funding now.
But I’d also argue that, if urgent longtermism is defined roughly as the view that there will be critical junctures in the next few decades, as you put it, then an urgent longtermist could still think it’s worth investing now, so that more money will be spent near those junctures in a few decades. Investing to give in, say, thirty years is still pretty unusual behavior, at least for small donors, but totally compatible with “urgent longtermism” / “hinge of history”-type views as they’re usually defined.
It seems to me that there are roughly three relevant confusions/sources of confusion in discussions around patient philanthropy, patient longtermism, and investing to give. I’ll try to briefly describe them, and I’d be interested to hear if you or others think this is accurate.
1. “Patient philanthropy” has been confused with “philanthropists should invest to give later”
People—including me, sometimes, and 80k in the podcast description here—have used the term “patient philanthropy” to refer to the position that we should “invest” resources for future altruistic “spending” (rather than “spending” now). Or people have said things like “the arguments for patient philanthropy” when meaning the arguments for that position.
But in your write-up, you actually used the term “patient philanthropy” as something like a starting assumption, saying:
You did argue that a patient philanthropist should “invest [resources] for use on future philanthropic projects”, but that wasn’t what patient philanthropy meant, and one could contest those arguments. (I’m not saying I do contest them; I think I remember mostly being in agreement.)
2. “Patient longtermism” wasn’t clearly defined, and one thing it might mean is quite distinct even from what people think patient philanthropy means
(See also this comment exchange.)
Ben Todd then introduced term “patient longtermism”, seeming to mean either:
“what people think Trammell meant by patient philanthropy, but now for longtermism”: i.e., longtermists should invest resources for use on future philanthropic projects
Something quite distinct: i.e., that now is not an especially “hingy”/”high-leverage” time.
I think Owen’s post is primarily arguing that this view is distinct from the idea that longtermists should invest resources for use on future philanthropic projects. I definitely agree. I think this view can potentially be a factor in whether longtermists should “invest” vs “spending” now, but there are various other factors as well.
Todd’s post on the topic seems to initially mean the second, then switch to the first, then switch to both:
3. “Investing” has sometimes been interpreted as only “financial investments”
(See this comment and this post for more discussion.)
You explicitly stated in your write-up and your 80k interview that “investment” in this context doesn’t have to mean just financial investments. But it’s sometimes been interpreted as if it does just mean that.
Personally, I like to use one of the following framings:
Should we financially invest in order to give later, or give now?
And if giving now, should we give to investment-like giving opportunities or “regular” giving opportunities?
Should we “punt to the future” or support “direct work now/soon”?
(See this post for discussion of those terms.)
And if punting to the future, should we do this via financially investing in order to give later, or in other ways (e.g., movement-building)?
Either framing gets at the same ideas.
But I think sometimes people try to condense that all into one question, which I think creates confusions.
I meant to define patient longtermism in terms of when you think the hinges are.
This will usually correspond to where you think the balance of object-level spending vs. investing/meta should be, but can come apart (e.g. uncertainty arguments could favour investing even if you think hinginess is going down).
I don’t think it should be defined in terms of not having a pure rate of time preference, since the urgent longtermists don’t have a pure rate of time preference either.
But overall all these definitions are pretty up in the air. It would be great if someone would like to take a more rigorous look.
Agreed. When writing my post on these matters, I found it surprisingly hard to even just describe what I was talking about and what claims I was making/summarising. It’d be handy to have settled, consistent, clearly defined terms.
Yeah, I agree with this too. I was just pointing out that, when introducing the term that “patient longtermism” seems to deliberately mirror, Trammell defined it as being about having no pure time preference. So if a term deliberately mirrors that one but then uses a different meaning, I think that can create confusion.
That said, I think how people have interpreted Trammell’s term—as about having the bottom-line belief that one should invest for future altruistic projects rather than spending now—is arguably more intuitive than how he defined it.[1] I say this because it seems to me more intuitive that “patient” should be a matter of waiting and doing something later (so a low overall discount rate, not just a low pure time discount rate). But someone with zero pure time preference could, under certain conditions, still want to spend on direct work now rather than waiting. (Not sure if I’ve explained that well.)
And it does seem good to have a term for the bottom-line belief that one should invest for future altruistic projects rather than spending now. So maybe we should just accept that “patient altruism” has evolved to mean that. (Though there is the issue that it implies the opposite is “impatient”, which sounds rude/dismissive.)
But then believing hinginess will be higher in future is a separate (though related) matter again. And I don’t think it’s as intuitive to call that “patience”, for the same reason I don’t think it’s super intuitive to call zero pure time preference “patience”; someone who believes hinginess will be higher in future could still want to spend on direct work now. I might want to say that believing hinginess will be higher in future will tend to push towards more “patience”, but that it isn’t itself “patience”.
It could perhaps be good to have a separate shorthand term for believing hinginess will be higher in future. (And btw, I do think it’s valuable that you/80k have highlighted the matter of how hinginess will change over time as a key uncertainty, and highlighted that not all longtermists believe the hingiest period is now/soon.)
[1] Maybe “patience” is an established term with a meaning similar to Trammell’s in philosophy/econ? I’m not sure.
Do you mean you think beliefs about how hinginess* is changing over time should be, or in practice is, the primary factor driving how much someone thinks we should allocate to object-level spending relative to meta/investing?
If so, that doesn’t seem obviously true to me. As I mentioned above and explore further in this post, I think there are a bunch of other important factors.(These factors have been prominently considered in prior work on this topic, so it’s not a new thing I’m throwing into the picture; e.g., Michael Dickens’ post on what the philanthropic discount rate should be included some hinginess-related-things as factors, but also had other major factors.)
In brief (for explanation of terms, see the post):
How effectively can we “punt to the future”?
What would be the long-term growth rate of financial investments?
What would be the long-term rate of expropriation of financial investments? How does this vary as investments grow larger?
What would be the long-term “growth rate” from other punting activities?
Would the people we’d be punting to act in ways we’d endorse?
Which “direct” actions might have compounding positive impacts?
Do marginal returns to “direct work” done within a given time period diminish? If so, how steeply?
It seems to me like it could be totally reasonable to think that hinginess is going to decrease somewhat over time in expectation, but that it’s still worth pushing more towards meta/investing because:
One thinks we can get such a good “interest rate” on financial or movement-building-ish investments, and/or
One thinks because marginal returns to more direct work within a time period diminish steeply, and we’ve already hit that point for the upcoming time periods (the low-hanging fruit will be taken), but longtermism might fizzle out later so some people should do investing/meta so we can still take the low-hanging fruit then if that happens
And conversely, it seems to me like it could be totally reasonable to think that hinginess is going to increase somewhat over time in expectation, but that it’s still worth pushing more towards object-level spending because:
One thinks the “interest rate” would be too bad, perhaps because one thinks value drift will be too high, or our movement-building investments would be too tied to specific areas and we’re too unsure if those are the right areas to focus on
One thinks object-level spending is both good for the relatively low level of near-term hinginess (e.g., reducing x-risk in the next 10 years) and has substantial meta/investment-like benefits
I don’t think this would be best described like “They’re doing something that looks like object-level work, but they’re doing it for meta purposes” (though I definitely think that could sometimes happen), because the person is doing it for both purposes.
(This would perhaps be somewhat analogous to people who agree that there will or could be far more events/experiences that are morally relevant in the long-term future than in the present/near-term future, but still think we should focus on improving the present/near-term future, because predictably improving the long-term future is too hard.)
It also seems worth noting that a lot of people’s beliefs about hinginess might be very uncertain or not very action-relevant; they might expect hinginess to slightly increase, expect it to slightly decrease, expect it to go up and down in ways that are hard to predict or notice, or have no real expectations about it. That’d make it easier for other factors to play the dominant role.
I definitely do think how hinginess will change over time is one of the key factors. And it’s plausible to me that it’s the primary one. But it’s not obvious to me that it’s the primary one, and I imagine how important each factor is will vary between people.
*Personally, I prefer “leverage over the future”, or in second place “pivotality”.
Yes, I totally agree with this. Indeed a large part of what I was trying to say was that I’m more sympathetic to this strategy right now for “urgent longtermists” than “patient longtermists” (although it happens that I mostly still think it’s beaten by non-financial investment opportunities which will pay off soon enough).
[LMK if you found something I wrote confusing; I could consider editing to improve clarity.]
For what it’s worth, as someone who read this post fairly quickly, I thought this point only came across as a minor aside in the grand scheme of your post (I think you only mention it in the last paragraph before your “some caveats”?).
I think you make the point here as well but I’m not sure if it came across as clearly as it could:
By the way I have strongly upvoted this post! I was aware of your overall point already but it was definitely useful to read your explanation and some of the more specific points. I agree that people have been too quick to jump to giving later if they identify as patient longtermists.
I guess “large part of what I was trying to say” was an overstatement; it’s an illustrative facet of the large thing (which is that there isn’t the obvious coupling).
Anyhow thanks for the pointer, I made some small language tweaks to the second part quoted (which hopefully help).
Sorry, no, that’s clear! I should have noted that you say that too.
The point I wanted to make is that your reason for saving as an urgent longtermist isn’t necessarily something like “we’re already making use of all these urgent opportunities now, so might as well build up a buffer in case the money is gone later”. You could just think that now isn’t a particularly promising time to spend, period, but that there will be promising opportunities later this century, and still be classified as an urgent longtermist.
That is, an urgent longtermist could have stereotypically “patient longtermist” beliefs about the quality of direct-impact spending opportunities available in December 2020.
In the abstract I agree that you could think that. But I’d make some of the same claims for the urgent longtermist as the patient longtermist: that some of the best investment opportunities are probably non-financial, and we should be trying to make use of those before going on to financial investments. (There’s a question about whether at current margins we’re already using them up.)
I think there are some principled reasons to be unsurprised if the best available non-financial investment opportunities are better than the best available financial investment opportunities. Financial investment is a competitive market; there are lots of people who have money and want more money, and so for a given risk tolerance (and without lots of work) you can’t expect to massively outperform what others are making.
There are also markets (broadly understood) competing for buy-in to worldviews. At first glance these might look less attractive to enter into, since they seem to be (roughly) zero-sum. But unlike the financial case, capital is not fungible across worldviews, so we shouldn’t assume that market forces mean that the returns from the best opportunities can’t get too good (or they’d be taken by others). And I’m not concerned about the zero-sum point, because I don’t think that the longtermist worldview is just an arbitrary set of beliefs; I think that it has ~truth on its side, and providing people with arguments plus encouraging them to reflect will on average be quite good for its market share (and to the extent that it isn’t, maybe that’s a sign that it’s getting something wrong). This is a pretty major advantage and makes it plausible that there are some really excellent opportunities available. Then I think growth over the last few years is evidence that at least some of the activities people engage in have really good returns; the crucial question is how much there are really good ones being left on the table.