″ I’d like to hear if there’s been any relevant work done on this topic (either within EA organizations or within general academia). Increasing returns is a fairly common topic within economics, so I figure there is plenty of relevant research out there on this. ”
These are my key reasons (with links to academic EA and other discussions) for seeing diminishing returns as the relevant situation on average for EA as a whole, and in particular the most effective causes:
If problems can be solved, and vary in difficulty over multiple orders of magnitude (in required inputs), you will tend to see diminishing returns as you plot the number of problems solved with increasing resources; see this series of posts by Owen Cotton-Barratt and others
Empirically, we do see systematic diminishing returns to R&D inputs across fields of scientific and technological innovation, and for global total factor productivity; but historically the greatest successes of philanthropy, reductions in poverty, and increased prosperity have stemmed from innovation, and many EA priorities involve research and development
In politics and public policy the literatures on lobbying and campaign finance suggest diminishing returns
In growing new movements, there is an element of compounding returns, as new participants carry forward work (including further growth), and so influencing; this topic has been the subject of a fair amount of EA attention
When there are varied possible expenditures with widely varying cost-effectiveness and some limits on room for more funding (eventually, there may be increasing returns before that), then working one’s way from the most effective options to the least produces a diminishing returns curve at a scale large enough to encompass multiple interventions; Toby Ord discusses the landscape of global health interventions having this property
Elaborating on the idea of limits to funding and scaling: an extremely cost-effective intervention with linear or increasing returns that scaled to very large expenditures would often imply impossibly large effects; there can be cheap opportunities to save a human life today for $100 under special circumstances, but there can’t be trillions of dollars worth of such opportunities, since you would be saving more than the world population; likewise the probability of premature extinction cannot fall below 0, etc
So far EA is still small and unusual relative to the world, and much of its activity is harvesting low-hanging fruit from areas with diminishing returns (a consequence of those fruit) that couldn’t be scaled to extremes (this is least true for linear aid interventions added to the already large global aid and local programs, and in particular GiveDirectly, but holds for what I would consider more promising, in cost-effectiveness, EA global health interventions such as gene drive R&D for malaria eradication); as EA activity expands more currently underfunded areas will see returns diminish to the point of falling behind interventions with more linear or increasing returns but worse current cost-effectiveness
Experience with successes using neglectedness (which in prioritization practice does involve looking at the reasons for neglect) thus far, at least on dimensions for which feedback has yet arrived
″ Ideally, we would like to not simply select causes that are neglected, but to select causes that are neglectedfor reasons other than their impact. ”
Thanks for this comment! The links were helpful. I have a few comments on your points:
″ Empirically, we do see systematic diminishing returns to R&D inputs across fields of scientific and technological innovation ”
After reading the introduction of that article you linked, I’m not sure that it has found evidence of diminishing returns to research, or at least that it has found the kind of diminishing returns that we would care about. They find that the number of researchers required to double GDP (or any other measure of output) has increased over time, but that doesn’t mean that the number of researchers required to increase GDP by a fixed amount has increased. In fact, if you take their Moore’s law example, we find that the number of transistors added to a computer chip per researcher per year is 58000 larger than it was in the early 70s (it takes 18 times more researchers to double the number of transistors, but that number of transistors is about a million times larger than it was in the 70s). When it comes to research on how to do the most good, I think we would care about research output in levels, rather than in percentage terms (so, I only care how many lives a health intervention would save at time t, rather than how many lives it will save as a percentage of the total amount of lives at time t).
″ In politics and public policy the literatures on lobbying and campaign finance suggest diminishing returns ”
I’m struggling to see how those articles you linked are finding diminishing returns. Is there something I’m missing? The lobbying article says that the effectiveness of lobbying is larger when an issue does not receive much public attention, but that doesn’t mean that, for the same issue, the effectiveness of lobbying spending will drop with each dollar spent. Similarly, the campaign finance article mentions studies that find no causal connection between ad-spending and winning an election for general elections and others which show a causal connection for primary and local elections. I don’t see how this means that my second dollar donated to a campaign will have less expected value than my first dollar.
As antonin_broi mentioned in another comment, political causes seem to have increasing returns built in to them. You need a majority to get a law passed or to get someone elected, so under complete certainty there would be zero (marginal) value to convincing people to vote your way until you reach the median voter. After that there will once again be zero marginal value to buying additional votes.
″ In growing new movements, there is an element of compounding returns, as new participants carry forward work (including further growth), and so influencing; this topic has been the subject of a fair amount of EA attention ”
I agree that this is important for growing new movements, and I have seen EA articles discuss a sort of “multiplier effect” (if you convince one person to join a group they will then convince other people). But none of the articles I have seen, including the one that you linked, have mentioned the possibility of increasing returns to scale. Increasing returns would arise if the cost of convincing an additional person to join decreases with the number of people that are already involved. This could arise because of changing social norms or due to increased name recognition.
″ historically the greatest successes of philanthropy, reductions in poverty, and increased prosperity have stemmed from innovation, and many EA priorities involve research and development ”
This brings up one potentially important point: in addition to scaling effects that you mentioned, another common source of increasing returns is high research and development requirements. High R&D requirements mean that the first units of output are very expensive (because in addition to the costs of production you also have to learn how to produce them) compared with following units. To apply this to an EA topic, if Givewell didn’t exist, then to do a unit of good in global health we would either have to fund less cost-effective charities (because we wouldn’t know which one was best) or pay money to create Givewell before donating to its highest recommended charities. In the second scenario, the cost of producing a unit of good within global health is very high for the first unit and significantly lower for the second. The fact that innovation seems to be one of the more effective forms of philanthropy increases the possibility that we are in a world where increasing returns to scale are relevant to doing good. However, I’m not completely sure on my reasoning here. I may be missing something.
″ Experience with successes using neglectedness (which in prioritization practice does involve looking at the reasons for neglect) thus far, at least on dimensions for which feedback has yet arrived ”
I think this would be a very important piece of evidence. Can you give me some detail about the successes so far?
″ I’d like to hear if there’s been any relevant work done on this topic (either within EA organizations or within general academia). Increasing returns is a fairly common topic within economics, so I figure there is plenty of relevant research out there on this. ”
These are my key reasons (with links to academic EA and other discussions) for seeing diminishing returns as the relevant situation on average for EA as a whole, and in particular the most effective causes:
If problems can be solved, and vary in difficulty over multiple orders of magnitude (in required inputs), you will tend to see diminishing returns as you plot the number of problems solved with increasing resources; see this series of posts by Owen Cotton-Barratt and others
Empirically, we do see systematic diminishing returns to R&D inputs across fields of scientific and technological innovation, and for global total factor productivity; but historically the greatest successes of philanthropy, reductions in poverty, and increased prosperity have stemmed from innovation, and many EA priorities involve research and development
In politics and public policy the literatures on lobbying and campaign finance suggest diminishing returns
In growing new movements, there is an element of compounding returns, as new participants carry forward work (including further growth), and so influencing; this topic has been the subject of a fair amount of EA attention
When there are varied possible expenditures with widely varying cost-effectiveness and some limits on room for more funding (eventually, there may be increasing returns before that), then working one’s way from the most effective options to the least produces a diminishing returns curve at a scale large enough to encompass multiple interventions; Toby Ord discusses the landscape of global health interventions having this property
Elaborating on the idea of limits to funding and scaling: an extremely cost-effective intervention with linear or increasing returns that scaled to very large expenditures would often imply impossibly large effects; there can be cheap opportunities to save a human life today for $100 under special circumstances, but there can’t be trillions of dollars worth of such opportunities, since you would be saving more than the world population; likewise the probability of premature extinction cannot fall below 0, etc
So far EA is still small and unusual relative to the world, and much of its activity is harvesting low-hanging fruit from areas with diminishing returns (a consequence of those fruit) that couldn’t be scaled to extremes (this is least true for linear aid interventions added to the already large global aid and local programs, and in particular GiveDirectly, but holds for what I would consider more promising, in cost-effectiveness, EA global health interventions such as gene drive R&D for malaria eradication); as EA activity expands more currently underfunded areas will see returns diminish to the point of falling behind interventions with more linear or increasing returns but worse current cost-effectiveness
Experience with successes using neglectedness (which in prioritization practice does involve looking at the reasons for neglect) thus far, at least on dimensions for which feedback has yet arrived
″ Ideally, we would like to not simply select causes that are neglected, but to select causes that are neglected for reasons other than their impact. ”
Agreed.
Thanks for this comment! The links were helpful. I have a few comments on your points:
″ Empirically, we do see systematic diminishing returns to R&D inputs across fields of scientific and technological innovation ”
After reading the introduction of that article you linked, I’m not sure that it has found evidence of diminishing returns to research, or at least that it has found the kind of diminishing returns that we would care about. They find that the number of researchers required to double GDP (or any other measure of output) has increased over time, but that doesn’t mean that the number of researchers required to increase GDP by a fixed amount has increased. In fact, if you take their Moore’s law example, we find that the number of transistors added to a computer chip per researcher per year is 58000 larger than it was in the early 70s (it takes 18 times more researchers to double the number of transistors, but that number of transistors is about a million times larger than it was in the 70s). When it comes to research on how to do the most good, I think we would care about research output in levels, rather than in percentage terms (so, I only care how many lives a health intervention would save at time t, rather than how many lives it will save as a percentage of the total amount of lives at time t).
″ In politics and public policy the literatures on lobbying and campaign finance suggest diminishing returns ”
I’m struggling to see how those articles you linked are finding diminishing returns. Is there something I’m missing? The lobbying article says that the effectiveness of lobbying is larger when an issue does not receive much public attention, but that doesn’t mean that, for the same issue, the effectiveness of lobbying spending will drop with each dollar spent. Similarly, the campaign finance article mentions studies that find no causal connection between ad-spending and winning an election for general elections and others which show a causal connection for primary and local elections. I don’t see how this means that my second dollar donated to a campaign will have less expected value than my first dollar.
As antonin_broi mentioned in another comment, political causes seem to have increasing returns built in to them. You need a majority to get a law passed or to get someone elected, so under complete certainty there would be zero (marginal) value to convincing people to vote your way until you reach the median voter. After that there will once again be zero marginal value to buying additional votes.
″ In growing new movements, there is an element of compounding returns, as new participants carry forward work (including further growth), and so influencing; this topic has been the subject of a fair amount of EA attention ”
I agree that this is important for growing new movements, and I have seen EA articles discuss a sort of “multiplier effect” (if you convince one person to join a group they will then convince other people). But none of the articles I have seen, including the one that you linked, have mentioned the possibility of increasing returns to scale. Increasing returns would arise if the cost of convincing an additional person to join decreases with the number of people that are already involved. This could arise because of changing social norms or due to increased name recognition.
″ historically the greatest successes of philanthropy, reductions in poverty, and increased prosperity have stemmed from innovation, and many EA priorities involve research and development ”
This brings up one potentially important point: in addition to scaling effects that you mentioned, another common source of increasing returns is high research and development requirements. High R&D requirements mean that the first units of output are very expensive (because in addition to the costs of production you also have to learn how to produce them) compared with following units. To apply this to an EA topic, if Givewell didn’t exist, then to do a unit of good in global health we would either have to fund less cost-effective charities (because we wouldn’t know which one was best) or pay money to create Givewell before donating to its highest recommended charities. In the second scenario, the cost of producing a unit of good within global health is very high for the first unit and significantly lower for the second. The fact that innovation seems to be one of the more effective forms of philanthropy increases the possibility that we are in a world where increasing returns to scale are relevant to doing good. However, I’m not completely sure on my reasoning here. I may be missing something.
″ Experience with successes using neglectedness (which in prioritization practice does involve looking at the reasons for neglect) thus far, at least on dimensions for which feedback has yet arrived ”
I think this would be a very important piece of evidence. Can you give me some detail about the successes so far?