This essay was submitted to Open Philanthropy's Cause Exploration Prizes contest.
If you're seeing this in summer 2022, we'll be posting many submissions in a short period. If you want to stop seeing them so often, apply a filter for the appropriate tag!
Summary: AMC for climate transition (chemicals), cultured meat and public goods as a possible cause area
I view this as a very very shallow investigation into the possible impact of using advanced market commitments (AMC), possibly prizes, for chemical industry transition, climate transition in general, cultured meat, and (large value) public goods in general (I look at sleep/wake cycle productivity, education, and sleep/wake drug development). I mention two meta/other ideas of possible interest on the power of story/art/narrative, and on cohort training for cause evaluation
I almost did not write this up because I concluded this was not likely to meet EA / Open Philanthropy (OP) hurdles as I understand them (which might not be very well, as being mostly an outsider to EA). But, given the guidelines that even short notes might be valuable, even if not as complete as I would like. My conclusions are tentative.
In reading and evaluating the area it did occur to me that AMCs and AMC-like “pools of funding” could usefully be directed at other public goods and externalities beyond the healthcare arena that has been looked at. OP has indicated this a possible area to investigate further, and my conclusion is I would move this up the priority for a researcher to properly evaluate. I do this on basis that (1) the AMC mechanism is tractable, has economic backing, and can combine private and public capital and (2) it is feasible from political positions (both left/right) unlike eg carbon tax. And (3) there are many possible areas this could open up, though unclear if 1000x return bar would be met - hence further research necessary.
In doing this basic cause evaluation it occurred to me that OP could fund basic cause evaluation training which might be a win-win for young altruists and OP.
Brief summary on impact, tractability and neglectedness
Potentially tractable. In most areas, I judge the problem as not easily practically tractable although possible (cf see Stripe etc. on Carbon markets AMCs).
Impact probably 10x to 100x, though could be 1000x. Under most assumptions the impact is only moderate - around 10x to 100x returns. However, if you give large weight to animal suffering, or give weight to the systemic nature of certain goods eg chemicals are in 90%+ of human materials, or large value to certain public goods, you can calculate large values of 1000x. Very uncertain.
Neglectedness ranges across possible cause areas. The chemical industry problem is understood, and arguably moderately neglected. Cultured meat is not neglected. Climate transition is not neglected. Certain public goods eg sleep/wake cycle learning might be moderately neglected (as politically difficult).
There is much uncertainty. My intuition is that there are high return cause areas out there for researchers to find, as when going through papers and areas new ideas were regularly occurring to me, although none reaching 1000x hurdles at first glance.
Brief articulation of AMCs and the problems AMCs solve
What is the problem? Uncertainty in the value of a market, leads to lower R&D and investment into an idea/goods/service than may be ideal for positive human (or animal) welfare. Public goods can be valuable but with limited incentives for private actors to develop innovations.
Advance Market Commitments (AMCs) are market creation mechanisms that provide the incentives and guarantees needed to ensure sufficient returns on investment typically by private sector actors, but could apply for any actors, and combination of actors. They can possibly combine with grant making or prizes.
Best known for vaccine development, and low commercial value, but high health burden diseases, and developing world; and more recently negative carbon emissions technology (although this was proposed in the UK by policy makers as far back as 2009).
Other areas that have emerged in academic papers are quantum computing, various carbon technologies (eg carbon capture, negative emissions technologies), infrastructure for self-driving cars; rockets (NASA’s COTS programme).
The question is are there areas potentially suitable (also tractable, impactful, neglected) ?
Economists have tended to suggest:
- Public goods
- 5 - 10 year technology road maps
- Nascent markets which can be jump started by stable repeat customers
Are suitable for AMCs
There is a debate as to how proven the technology needs to be before AMCs can work, and how much understanding of the area funders of AMCs should have.
Climate areas: On the climate challenge, there are “hard to abate” sectors where AMCs might help. These are sectors where technology has started but is 10 years away from commercialisation although time lines are coming in here. These include:
- Chemical industry processes, plastics and other industrials
- Industrials process in general, certain building materials: paint, steel, aluminium, copper, lithium, fertilisers, cement etc.; industrial gases both niche eg neon, and wide spread eg nitrogen, hydrogen etc.
- Long distance shipping, long distance flying
These areas tend to have these characteristics with the relevant “green” technology behind decarbonising those process are at an early (pre-commercial viability) but often proven to work (the process can be reproduced reliably but not at scale or cheaply) stage.
Chemicals
Chemicals is one area. Here one would need to have industry (1) electrify energy processes and (2) replace fossil feedstock with green hydrogen or green methanol or some other low carbon process.
Individual actors in the chemicals sector would lose money if they tried to transition first - by investing in new capital equipment and processes - without having a guaranteed market and also without coordination that this would be industry wide and with end product guaranteed markets.
What philanthropy could do: In theory, a coordination meta-organization could help establish this market with an AMC and potentially in conjunction with the chemical industries’ biggest buyers (consumer products companies) there could be some AMC to guarantee this transition. Ultimately consumers would pay for the costs, but these are likely only a small part of the overall costs of the product (think yogurt pots, plastics: the Energy Transition Commission seems to have an estimate maybe as low as 1% eg 1c on price of plastic drinks bottle) but I am unsure how much to trust this). Even so, ensuring demand for emissions-free chemicals would be an important part of the solution. This is what a foundation/philanthropy could help build, this meta-coordination-organisation.
Ball park 500x return on $10m. Say it would cost $10m to create a meta-organization, coordinate the work, and convince the end users of plastics industry (or maybe governments other companies) to commit to AMCs (rather than find money itself) and this leads to abatement 30 years earlier than with no work by this org.
Currently around 920 mt CO2 pa from chemicals industry. Assume 500 Mt is abated at a price of eg $100 t CO2, that would be about $50bn in value (and part of the climate transition pathway enacted), around 500x return on the $10m. With great uncertainty this is towards the level of a 1000x return.
This process likely needs to happen and likely would happen in time as the cost of technology drops (assuming rate of innovation holds), but spot checks with the industry put this closer to a 40 to 50 year time line as current processes are profitable and the demand for low-carbon produced chemicals is unclear (there is demand, but not at great premiums, as in if the cost would be with 10%, buyers would transition cf. Bill Gates book) [My view, BY]
Does not pass OP bar. Ultimately, this is important, but I do not think a meta-organisation is easy to create or easily viable to create and run - so I rate the idea as moderate. A major hurdle is a “leader” equivalent to a start-up founder who would lead this organisation. On the other hand, if there was a leader then building a new organisation and one which would pass learnings on to other EA/OP aligned organisations could be of further value.
In theory, similar ideas could be applied to other hard to abate areas eg aluminium, steel, cement where a guaranteed market or co-ordinating meta-organisation would be helpful.
There are certain initiatives towards some of that (eg “green” or low carbon steel initiatives) but these could be accelerated. There may also be more neglected areas eg industrial neon which are worth exploring by a researcher. Here an EA/OP aligned organisation could pass on learnings to other similar types of challenges.
Cultured Meat
Another area would be cultured meat. In particular for pigs, and maybe chicken, fish if you weight non-human suffering; and in particular, cows, ruminants, if you weight climate emissions challenges. Cultured chicken and beef have nascent likely markets, but in other animal proteins the markets are even further behind. This could be an area to explore. The data on the climate impacts and the animal suffering impacts are not well understood. For climate impacts much depends on the energy source and land use assumptions. The animal suffering implications are potentially greater. I highlight this as there are animal welfare gains as well as climate gains.
Estimate of suffering. Eg estimates are of 1.5 billion pigs, 60 billion+ chickens, 600m+ turkeys, 300m+ cattle killed a year. A 10% saving of animal, enabled - for instance - 10 years earlier due to an AMC would still be 500m+ animal lives spared a year, or about 5bn over period of 10 years.
If you valued an animal life at eg $1 this is a value of $5bn, so it would a 1000x return would be on a investment of $5m. This seems to me too low to achieve too much. But it might be possible for a meta-coordination-organisation to pool together animal welfare organisations and climate organisations to potentially find $50 to $500m to pool. This level of value, I estimate would be enough to catalyse especially in some of the smaller protein markets eg lamb.
There are many start ups working in this area, and there are regulatory barriers as well. This is where a meta organization could possibly help.
Idea of cultured meat restaurant chain. Another idea would be to establish a small chain of restaurants (or a blue print of) where these restaurants would commit to buying cultured meat to serve (regulation allowing). Place these in cultural hubs eg Paris, London, New York, SF. The purpose would not be profit, but to show that tasty, brilliant food can be made and to guarantee a commitment that these restaurants will buy cultured meat when available. This combines an AMC with other reputational intangibles.
Say, a chain of 10 restaurants costs $10m but is break even (profits cover costs and investments). The AMC value is small but potentially valuable to start up cultured meat company. For eg lamb, the direct value would be small (10 restaurants, 5 lamb per week), only 250 lambs a year in direct value but possibly large reputationally. The value here is not so much directly in animals saved but in the cultural signals needed.
Toy example of cultural value. This cultural value is uncertain and hard to value. As a toy example let us assume a vegan leader like Paul McCartney is attracted and promotes the idea. He has 4m Instagram followers. Assume 2m followers pay attention and mention this to 10 other friends. Out of these 20m people sensitised to the idea assume 1% try it out and convert to a more vegan / cultured meat diet. That would be 200K people influenced into this way of thinking. Find 10 Paul McCartney type figures to be involved in the cultured meat restaurant narrative and that is 2m people influenced.
On considering this, I reflected that the “cultual change narrative” (cf. women’s rights, minority rights, even slavery emanicipation) would be a critical point in enabling cultured meat to become mainstream and that OP could catalyse this. If this was of interest, I think there is a small line of research into what narratives are powerful (for instance, this Jon Ronson podcast/documentary suggests an art film was critical to catalysing US evangenicals on the abortion issue: https://www.bbc.co.uk/programmes/m0011cpq essentially it’s the power of story).
On this I note FTXF has a variety of media/story related project cause areas they are considering including YouTubes and Substacks. But what makes an effective substack to tlak about the Future or Animal Welfare or anything aimed at cultural change? While there are activist handbooks eg Rule for Radicals I do not find anything that more directly speaks to the change EA aligned orgs want.
Summary on cultured meat. Overall, I judge the problem is not very neglected, and of moderate impact on climate grounds, possibly more on animal welfare grounds. I would suggest further work only if OP weights animal suffering reasonably highly. That said, I think the cultural element is more neglected, if much less tractable, and a short piece of research might be useful there.
Valuable Public Goods
The last area I would look at is “valuable public goods”. AMCs as originally envisaged are not necessarily an exact fit. But this class of area I think is possibly impactful. There are findings/discovers that could dramatical improve welfare/productivity but where private incentives are lacking or complex. These areas may fall under “progress studies” areas but could include:
- Human productivity relating to sleep/wake cycles
- Biopharmaceutical use at sleep/wake cycle intervals
- Educational mastery systems for high talent youth
The overall idea is some valuable public good discovery that has limited incentives for research currently but could really improve human welfare/future.
For instance, sleep patterns. There is evidence that many humans are operating suboptimally with respect to sleep and sleep/wake cycles.
First in sleep deficit (eg https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5627640/)
And then with respect to knowing how to be “productive” as a “early bird” or “night owl” or seemingly dependent on your sleep/wake biology/genetics. Eg Life Time: The New Science of the Body Clock by Russell Foster
One idea behind this is because different people have different sleep/wake patterns eg night owl, early bird and that people work at odds with their pattern. There is some research behind this but seemingly no definitive work (that I find on normal searching).
Further more, it seems the time we take medications can have a huge impact on how effective drugs are eg. https://www.nature.com/articles/d41586-018-04600-8
The challenge and problem here is there is limited private incentive to provide definitive research in this area.
Toy impact example. This study suggest a cost of insufficient sleep in US alone is at $299 bn to $433bn. It’s easy to suggest a $1tn global cost here. Say, you can improve this by 5% through a combination of studies/narrative on using melatonin (considered a very safe drug) and studies on night owl/early bird productivity, 5% of $1tn would be $50bn gain (and so warrant a $50m investment at 1000x).
Say, there is a 20% chance that chrono-timing of cancer drugs improved drug effectiveness by 10%. That’s a 20% chance of, say 10% more patients making it to 5 year survival, or other quality of life measures. It seems like there is chance of a reasonably large (in the order o 100,000s DALYs) gain of life years.
What could a philanthropist do? I judge that first research is needed on what type of public goods is under-researched and would have the most impact. I’ve come across sleep and educational mastery as it seems to me impact and tractability are possible, but I’ve not had deep discussions over it. Then it seems to me that while a grant or price could accelerate some research here possibly better, or at least less tried, would be a programme of practical application.
For instance, if sleep sufficiency were chosen as ana rea then an organisation could organise some programme as to whether melatonin used by self-assessed sleep deprived people would improve productivity. Or, similar for productivity of ”night-owls”, some programme perhaps in a large organisation assessing people’s chronotypes and whether changing schedules to fit these chronotypes led to any welfare or productivity improvements. The organisation could then undertake to publish negative or positive studies on this and put more prototocal in to action.
These are not AMCS as such, but some form of commitment to create a practical programme of work, implement it at decent scale, and then publish promote the findings with a view that if large positive welfare was found this would have great systemic (public goods) value.
The last area here, is often thought of in EA circles and that is on different educational institutions, universities etc. My idea here is based on my reading that we don’t really seem to know how best to educate the top 1% or the top 0.1% of young people.
The first challenge is finding these people. It seems like there are many in eg rural Africa or Asia who are not being found. But once found do we really understand how best to train such people? One idea here would be to found a school perhaps for 12 to 18 year olds to maybe 21 where they are taught a very different programme that follows their interests, in much smaller group teaching (such as “mastery” techniques); or really any of the best ideas in education.
The idea is that the top 1% of people even identified early seem to go on to have a disproportionate impact on future human welfare. I have not thought through this idea very carefully and leave it as a stub idea.
I note Rincon has summarised the data here on tutoring in general. (“Tutoring in general, most likely, does not reach the 2-sigma level that Bloom suggested. Likewise, it's unlikely that mastery learning provides a 1-sigma improvement. But high quality tutors, and high quality software are likely able to reach a 2-sigma improvement and beyond.”)
Cohort training as an idea. Very lastly. My day job is in investing. I noted quite a few useful similarities between investing (and indeed some areas of policy, expected utility areas) and cause evaluation. Impact is like “market size of TAM”, neglectedness is related to “is there something the market has missed, an edge, or an externality), tractability it related to easer of execution. What can someone do is related to assessing management teams execution ability and strategy.
It occurred to me that if OP were to set up a short training programme on cause evaluation eg a 4 week period, 3x 2 hour sessions a week, a half day “demo” presentation; for a cohort of eg 15 people and made it free, but in exchange those 15 people would agree to write eg two shallow 8 page investigations, every year for eg. 5 years then this would be a way for OP to have a continual “scout” programme for cause ideas. For the trainees it would give them “expected utility/cause evaluation/pseudo-investing” type skills which I would as useful for many types of sector and future jobs. Of course, this idea can be expanded on etc. but the core idea would be to scale up and train a cohort of people who would be useful to OP but who would also learn useful skills.
Conclusions
I view this as a very very shallow investigation into the possible impact of using advanced market commitments (AMC), possibly prizes, for chemical industry transition, climate transition in general, cultured meat, and (large value) public goods in general (I look at sleep/wake cycle productivity, education, and sleep/wake drug development). I mention two meta/other ideas of possible interest on the power of story/art/narrative, and on cohort training for cause evaluation. I conclude that nothing meets the OP bar as I understand it, but some areas come a little close and I hope I have given some ideas for future exploration. In particular on an idea of future cause exploration cohort courses.
Ben Yeoh available via benyeohben@gmail, on Substack, Twitter etc. Links to bio etc. here: https://www.thendobetter.com/start-here
