This is a Draft Amnesty Week draft. It may not be polished, up to my usual standards, fully thought through, or fully fact-checked.
This is a quick writeup of an idea I was thinking about a couple of years ago, the goal here is to give you the gist of the idea as a starting point for researching it more thoroughly.
In my opinion the relevant research to be done is into the minimum cost per person-year of stored food, properly accounting for storage costs, financing, spoilage, and creating a nutritionally survivable diet.
There are other angles you could also look at like how to deal with distribution, how much food is already (de facto or deliberately) stockpiled, or how to pitch this idea to governments.
The idea
Governments (or someone else) could potentially store a lot of food for the long term quite cheaply. Back of the envelope maths says something like $10 per person-year of stored food, or around 0.05% ($3B) of US GDP to ensure a year of food for every person in the US.
This would be beneficial across a wide range of catastrophe scenarios, and could be a complement to or a replacement for the idea of resilient foods (like turning wood into digestible carbs). In the event of a disaster, it might take a long time (months/​years) for resilient foods to be scaled up to a sufficient level, meaning a large stockpile would still be valuable.
Additionally many sorts of food-reducing disasters may not last more than a few months or years anyway (especially when adjusted for the fact that it could be more like a reduction in crop yields rather than a total lack of food), in which case this could be a sufficient solution and not just a complimentary one.
Nuclear winter particularly is generally predicted to result in at most years (not decades) of reduced (but not eliminated) crop yields, and so storing a few years worth of food could be a sufficient and in fact more robust solution than the large amount of research/​coordination required to scale up resilient foods. That said I’m pretty sceptical of nuclear winter[1] anyway, but there are other sorts of catastrophes (great power conflict, widespread crop blight) where it certainly wouldn’t hurt the situation to have several years of food for everyone on the planet stored somewhere.
Given that the $10 per person-year is likely optimistic, I would guess that this is a quite good but not fantastically great intervention in marginal cost effectiveness terms. E.g. LEEP is ~$14 per DALY, and this would be ~$10 per ~40 DALYs if a year of food is enough to tide someone through a disaster in which they otherwise would have starved. But this is obviously only realised if such a disaster occurs.
I think it could be something like PEPFAR though where it’s very simple and scalable so you could convince governments to adopt it and bear most of the cost. Also you might imagine that it reduces the risk of civilisational collapse or extinction more than the marginal $/​DALY number implies.
The back of the envelope maths
I’m assuming you would store some food that is cheap + calorie dense + has a long storage life in big, government-operated, centralised locations to keep the storage costs down.
Taking palm oil as an example of a high calorie/​$ food, and assuming that you can extend the shelf life by freezing it/​putting it in an inert environment/​not worrying about the rancid taste because you would otherwise starve to death, we get:
Shelf life: 1 year, I’m going to assume you could extend that to 20 years by preserving it. I don’t really see how it could go off in a low oxygen environment but maybe there’s something I don’t understand about how food goes off
Assume a person needs 1800 kcal/​day to stay alive (likely pessimistic), so 660,000 kcal/​year. It would cost $73 to buy that much palm oil, $3.7/​year if you stored it for 20 years. This doesn’t account for storage costs and financing, hence the $10 per person-year estimate above.
If this was operated at country-scale this would also drive up food prices, I don’t expect this to make too much difference though because if you are storing it for 20 years then you would only need 5% more calories produced, assuming you destroy it at the end of the 20 years. Also it might make sense to do something other than storing food until it’s completely unusable.
Research project idea: food stockpiling as a GCR intervention
This is a quick writeup of an idea I was thinking about a couple of years ago, the goal here is to give you the gist of the idea as a starting point for researching it more thoroughly.
In my opinion the relevant research to be done is into the minimum cost per person-year of stored food, properly accounting for storage costs, financing, spoilage, and creating a nutritionally survivable diet.
There are other angles you could also look at like how to deal with distribution, how much food is already (de facto or deliberately) stockpiled, or how to pitch this idea to governments.
The idea
Governments (or someone else) could potentially store a lot of food for the long term quite cheaply. Back of the envelope maths says something like $10 per person-year of stored food, or around 0.05% ($3B) of US GDP to ensure a year of food for every person in the US.
This would be beneficial across a wide range of catastrophe scenarios, and could be a complement to or a replacement for the idea of resilient foods (like turning wood into digestible carbs). In the event of a disaster, it might take a long time (months/​years) for resilient foods to be scaled up to a sufficient level, meaning a large stockpile would still be valuable.
Additionally many sorts of food-reducing disasters may not last more than a few months or years anyway (especially when adjusted for the fact that it could be more like a reduction in crop yields rather than a total lack of food), in which case this could be a sufficient solution and not just a complimentary one.
Nuclear winter particularly is generally predicted to result in at most years (not decades) of reduced (but not eliminated) crop yields, and so storing a few years worth of food could be a sufficient and in fact more robust solution than the large amount of research/​coordination required to scale up resilient foods. That said I’m pretty sceptical of nuclear winter[1] anyway, but there are other sorts of catastrophes (great power conflict, widespread crop blight) where it certainly wouldn’t hurt the situation to have several years of food for everyone on the planet stored somewhere.
Given that the $10 per person-year is likely optimistic, I would guess that this is a quite good but not fantastically great intervention in marginal cost effectiveness terms. E.g. LEEP is ~$14 per DALY, and this would be ~$10 per ~40 DALYs if a year of food is enough to tide someone through a disaster in which they otherwise would have starved. But this is obviously only realised if such a disaster occurs.
I think it could be something like PEPFAR though where it’s very simple and scalable so you could convince governments to adopt it and bear most of the cost. Also you might imagine that it reduces the risk of civilisational collapse or extinction more than the marginal $/​DALY number implies.
The back of the envelope maths
I’m assuming you would store some food that is cheap + calorie dense + has a long storage life in big, government-operated, centralised locations to keep the storage costs down.
Taking palm oil as an example of a high calorie/​$ food, and assuming that you can extend the shelf life by freezing it/​putting it in an inert environment/​not worrying about the rancid taste because you would otherwise starve to death, we get:
9000 kcal/​kg
$1/​kg
Shelf life: 1 year, I’m going to assume you could extend that to 20 years by preserving it. I don’t really see how it could go off in a low oxygen environment but maybe there’s something I don’t understand about how food goes off
Assume a person needs 1800 kcal/​day to stay alive (likely pessimistic), so 660,000 kcal/​year. It would cost $73 to buy that much palm oil, $3.7/​year if you stored it for 20 years. This doesn’t account for storage costs and financing, hence the $10 per person-year estimate above.
If this was operated at country-scale this would also drive up food prices, I don’t expect this to make too much difference though because if you are storing it for 20 years then you would only need 5% more calories produced, assuming you destroy it at the end of the 20 years. Also it might make sense to do something other than storing food until it’s completely unusable.
For basically the reasons outlined in these posts