I co-founded the Lead Exposure Elimination Project (https://leadelimination.org/), an international NGO that works to avert childhood lead poisoning.
I’ve also published academic papers in the global health space, as well as on X-Risk, with a focus on emerging technologies and totalitarianism.
Hi Misha! I totally agree. Tractability could make a significant difference on the expected cost-effectiveness of a particular country. In the coming weeks we’ll be having a full blog post on our site unpacking our country selection in detail, but I can give you a quick summary here.
To identify the most viable target countries we’ve assessed every country worldwide (with more than 300,000 births in the last 5 years), and ranked them on a number of factors. These factors included size of lead burden, absence of lead paint regulation, stability (which works as a loose indicator of tractability—or at least of intractability in extreme cases), and neglectedness by other actors.
Based on this, our other top tier countries included Madagascar, Angola, Sierra Leone, Liberia, Burkina Faso, Niger, Zimbabwe, Lesotho, Haiti, Guatemala, El Salvador, and Honduras.
Hi Linch, Great question. Probability estimates about the future are always difficult—we can give you some loose indications of what we expect, though these should be taken with a grain of salt.
1. A two month timeline may be hard to estimate, as things can often run more slowly when starting in a new country. However, we can give more confident estimates on a three month timeline (which offers some buffer room).
We’d attach a reasonably high probability to having enough information to make these decisions within the next 3 months (~80%). Our decision-relevant data gaps at the moment are a) about the level of lead paint use in Malawi, and b) about the tractability of meeting with Malawian politicians. However, we have ways to test both of these things, and identify whether they should be a disqualifying factor, and have strong leads on the connections required to get this information.
2. Our probability of piloting in Malawi is also quite high, as the early indicators of burden and tractability seem quite strong. We’ve managed to find a 2017 study on the level of lead in paint in Malawi which indicates that this country would be highly promising to target. The study found that 56% of paint tested in Malawi had more than 90ppm, and 37.5% had more than 600ppm - this makes us think that it is quite likely that lead is a significant contributor to the overall burden of lead poisoning in Malawi. On the tractability side, making contacts within Malawi has also been much easier than expected. This makes us think that our probability of piloting in Malawi is better than even, around 60%.
3. Good question. This intervention has the potential to have very large-scale benefits, and very high cost-effectiveness, but I don’t think I could attach a reasonable probability estimate that I’d trust this far out.
Interesting list! One important cause area that I think may have missed is preventing/avoiding stable longterm totalitarianism.
Toby Ord and Bryan Caplan have both written on this—see “the Precipice” for Ord’s discussion and “The Totalitarian Threat” in Bostrom’s “Global Catastrophic Risks” for Caplan’s.
It may be worth adding these to the list as it seems that totalitarianism is fairly widely accepted as a cause candidate. Thanks for the post as well, lots of interesting ideas and links in here!
LEEP is now live on everyorg.com too! Thanks for alerting us to this @WilliamKiely
https://www.every.org/leadexposureeliminationproject
Really interesting thoughts! Thanks for putting this together. So useful and practical to have a priority list of this kind!
I was quite surprised that policy around nuclear weapons isn’t rated even higher. Looking at your spreadsheet, I think the ‘Australia ROI’ score for nuclear war might be quite strongly underrated at present. It is currently rated on that metric as a 2 (out of 5) meaning “Compared to other countries, Australia has little control and influence on this issue. Australia also has a poor international reputation on this issue and rarely implements best practices. Australia has no comparative advantages for addressing this issue.”
This seems to be a mistake to me. You’re right that Australia has little sway on the likelihood of nuclear war, or on the behaviour of major nuclear powers, and it isn’t a nuclear power itself. However, it still has huge advantage compared to most countries in addressing existential risk—the reason being that (according to a number of experts like Brian Martin and Toby Ord) Australia seems far more likely than almost any other country to survive a nuclear war. So while there likely isn’t much work to be done in Australia to prevent a nuclear war from occurring, it seems like there is a huge amount that could be done in Australia, and maybe only in Australia, in order to prevent the actual extinction risk attached to nuclear war.
In particular a) having plans set up for survival of some portion of the Australian population so that nuclear war doesn’t result in extinction (At present, it’s my understanding that the Australian government has no contingencies in place for population survival after nuclear war at this point) and b) ensuring that Australia doesn’t arm themselves with nuclear weapons/doesn’t take actions that lead it to become a major nuclear target in future. As being a nuclear target would jeopardise one of the places most likely to survive. These two approaches seem like they could be powerful ways to avoid the extinction risk associated with nuclear weapons, and Australia would be better positioned to implement this than almost any other country (possible exceptions/other viable countries might be New Zealand, or possibly Argentina).
Any change of score here—even a change by one point on this one metric—would make nuclear risk the #1 highest priority option to address in Australia in your spreadsheet. As it currently scores as the equal most important option in you list.
It seems valid using your metrics too, as rating it a 3 would mean that “Compared to other countries, Australia has some control and influence on this issue.” and ” “Australia has few comparative advantages for addressing this issue.” (instead of no comparative advantages), which, even under conservative assumptions seems highly likely to be true.
Given Australia’s uniquely good positioning with this issue, is there a reason it wasn’t rated higher on this metric? Perhaps there’s something I’ve missed.
Thank you for these interesting answers. Do you think the creation of new fields is also subject to diminishing returns? e.g. are new fields harder to find as well? Or do you think that only technologies are subject to diminishing returns?
On this note, do you think progress is likely to be open to us indefinitely, or would you expect that eventually we will reach a level of technological maturity where all meaningful low-hanging fruit (be they individual technologies or S curves) have been picked and there is little further technological progress? If so, why? If not, why not?
Hi Matt, thanks for your comment! We haven’t looked into this, but if we do we’ll let you know. As yet, we have not focused on water as a source of lead exposure because it is less of a contributor in lower-income countries where lead burdens are highest.
Thanks so much Ula!
Thanks very much Pete!
This was a great post! Really enjoyed reading it!
That said, the ordering in the shortlist in this post is a touch confusing—it doesn’t seem to map to the total scores in your spreadsheet? e.g. AI scores as equal 4th-6th most important in your spreadsheet, but is listed 1st in the short list. Meanwhile nuclear war and pandemic risk score as equal 1st-3rd, but are listed 9th and 10th in the shortlist. I wonder if it would make more sense to list the options that scored the highest first (as presumably, they should be prioritised the most)?
Ahhhh I see! Thanks for clarifying!
And thanks again—this was a great read.
Thanks Larks! Glad you enjoyed it!
This was a fantastic post Jenny. Very insightful. Thank you!
Ah! I can definitely see how that might have been confusing, thanks for letting us know. I’ll make sure that this is reworded to be as clear as possible. Good catch!
And that sounds fantastic. It’s likely we’ll seek to hire a dedicated operations staff member as we scale, perhaps in our second or third year. In case you’d like to keep an eye out for when such positions crop up, all future job postings for LEEP will be announced on our website, in our newsletter, and on the 80,000 Hours Jobs board.
Hi Brian! Thanks for commenting. Here is a link which lists countries which do not currently have lead paint regulation. We didn’t come across any ranking of most promising countries to target, but we have collated this information ourselves. We’ll be publicly releasing our country selection spreadsheet in the coming weeks and this will likely provide you with the information that you need.
Until then, I can pass on some relevant information about our findings. From our research, we found that the Philippines does currently have legislation (as you mentioned), and is not currently neglected by other actors (EcoWaste Coalition and IPEN are active there).
We found that there are still significant levels of lead poisoning in the Philippines (see annex), but we are not confident that this necessarily indicates that the current legislation is not well enforced. It could be an indicator that there is significant exposure to lead from other sources, or it could just be that the effects have been delayed because legislation takes time to have impact. Paint bans need to be in effect for a few years before health impacts can be noticed.
On the other hand, it could well be an enforcement issue. Controls were introduced in 2015 to ensure that paint has lead levels below 90ppm, but in 2017 it was found that the Philippines still had 23% of their paint with lead levels exceeding 90ppm, 16% of their paint with lead levels exceeding 600ppm, and 12% of their paint with lead levels exceeding 10,000ppm. That might imply that the controls aren’t being followed perfectly. However, it’s also been three years since that study was done, so things may have changed.
Hope this was useful!