Insurance contracts are regulatory processes driven by prediction markets. If you have liability plus required insurance then different insurance companies compete with regulations that you have to follow to be eligible for their insurance contracts and the insurance companies trade on various futures to manage their risks.
Yeah. So to paraphrase, you could have a system where you can release something but only if you can find someone who will sell you insurance, then if your drug gives people cancer on a 20 year time delay the insurer has to pay for the chemo, and more for the damages that can’t be repaired.
How do these systems address the fact that companies and their human executives often just wont viscerally care about consequences that’re 20 years away? Or that sometimes the downsides will be so adverse that one org wont really be able to weigh them in proportion to the upsides via monetary incentives? (you can’t punish an org with anything worse than bankruptcy, but sometimes the downsides are so bad that you need to.) With venture granters I tried to think of ways of increasing accountability beyond what bankruptcy allows. One example is prison sentences for insurers who can’t pay liabilities, regardless of whether their mistakes were malicious. Not sure why we don’t have that already. Another is network structures between insurers: they can make agreements to share liabilities to increase their ‘load rating’, but now they’re going to have formalized incentives to audit each other, which I’ve heard via development economist Mushtaq Khan ( https://80000hours.org/podcast/episodes/mushtaq-khan-institutional-economics/ ) is a great way to keep an industry straight and to get the industry to participate fruitfully in generating effective regulation.
We don’t have many problems with insurance companies going broke and not being able to pay out.
The bigger problem is that “Ask a jury of your peers whether or not the drug was responsible for the cancer” is a crappy way to determine causality.
Clarify “Compete with regulations”?
Different insurance companies have different policies about what the buyer of the insurance company has to do to be eligible for them. Those policies are regulations.
We don’t have many problems with insurance companies going broke and not being able to pay out.
How do you know this? (I’d be surprised if this was true, unless it was a result of a) insurance companies’ contracts maintaining a bottomless barrel of excuses to not provide the service they sold, which is very common b) a result of forgiveness and bailouts which should not have happened or c) it basically does happen, but the brand gets sold to new owners and most of the staff and facilities are kept and no one notices. In which case I wouldn’t be surprised at all, but I’m still right, these are all very problematic phenomena.)
Different insurance companies have different policies about what the buyer of the insurance company has to do to be eligible for them. Those policies are regulations.
That makes sense, so the government sets some low-level requirements that’re risky to take on, simple, and abstract, something like, idk, “you’ll be destroyed if The Weighers find that your approvals produce more harm than good” or something, and the insurance company builds specific instructions on top of that like “nothing bad will happen to you as long as you do these specific things”?
Insurance is exactly a mechanism that transforms high-variance penalties in the future into consistent penalties in the present: the more risky you are, the higher your premiums.
Then insurance as you’ve defined it is not a specific mechanism, it’s a category of mechanisms, most of which don’t do the thing they’re supposed to do. I want to do mechanism design. I want to make a proposal specific enough to be carried out irl.
Insurance contracts are regulatory processes driven by prediction markets. If you have liability plus required insurance then different insurance companies compete with regulations that you have to follow to be eligible for their insurance contracts and the insurance companies trade on various futures to manage their risks.
Yeah. So to paraphrase, you could have a system where you can release something but only if you can find someone who will sell you insurance, then if your drug gives people cancer on a 20 year time delay the insurer has to pay for the chemo, and more for the damages that can’t be repaired.
How do these systems address the fact that companies and their human executives often just wont viscerally care about consequences that’re 20 years away? Or that sometimes the downsides will be so adverse that one org wont really be able to weigh them in proportion to the upsides via monetary incentives? (you can’t punish an org with anything worse than bankruptcy, but sometimes the downsides are so bad that you need to.) With venture granters I tried to think of ways of increasing accountability beyond what bankruptcy allows. One example is prison sentences for insurers who can’t pay liabilities, regardless of whether their mistakes were malicious. Not sure why we don’t have that already. Another is network structures between insurers: they can make agreements to share liabilities to increase their ‘load rating’, but now they’re going to have formalized incentives to audit each other, which I’ve heard via development economist Mushtaq Khan ( https://80000hours.org/podcast/episodes/mushtaq-khan-institutional-economics/ ) is a great way to keep an industry straight and to get the industry to participate fruitfully in generating effective regulation.
Clarify “Compete with regulations”?
We don’t have many problems with insurance companies going broke and not being able to pay out.
The bigger problem is that “Ask a jury of your peers whether or not the drug was responsible for the cancer” is a crappy way to determine causality.
Different insurance companies have different policies about what the buyer of the insurance company has to do to be eligible for them. Those policies are regulations.
How do you know this? (I’d be surprised if this was true, unless it was a result of a) insurance companies’ contracts maintaining a bottomless barrel of excuses to not provide the service they sold, which is very common b) a result of forgiveness and bailouts which should not have happened or c) it basically does happen, but the brand gets sold to new owners and most of the staff and facilities are kept and no one notices. In which case I wouldn’t be surprised at all, but I’m still right, these are all very problematic phenomena.)
That makes sense, so the government sets some low-level requirements that’re risky to take on, simple, and abstract, something like, idk, “you’ll be destroyed if The Weighers find that your approvals produce more harm than good” or something, and the insurance company builds specific instructions on top of that like “nothing bad will happen to you as long as you do these specific things”?
It’s the sort of problem that’s newsworthy. If we would have many problems like that, people would be angry about it.
Or at least the amount of bad that happens is manageable for the insurance company.
If you look at malpractice insurance, some insurance providers require Board certification of doctors or give those that have it better rates.
Insurance is exactly a mechanism that transforms high-variance penalties in the future into consistent penalties in the present: the more risky you are, the higher your premiums.
Then insurance as you’ve defined it is not a specific mechanism, it’s a category of mechanisms, most of which don’t do the thing they’re supposed to do. I want to do mechanism design. I want to make a proposal specific enough to be carried out irl.