Longevity Insurance
Let’s say we (as a country) ban life insurance and health insurance as separate packages [1] and require them to be combined in something I’ll call “Longevity Insurance”. The idea is that as a person/consumer, you can buy a “life expectancy” of 75 years, or 90 years, or whatever. In addition, you specify a maximum dollar amount that the longevity insurance will ever pay out—say, $2 million. If you have any medical issues throughout your life, up to the life expectancy threshold, the insurance plan will pay for your expenses. If it fails to keep you consciously alive for the duration of your “life expectancy”, then upon your death, the policy guarantees that the company will pay the full remaining amount to your next of kin.
As an example, suppose you (let’s say you’re a woman) had purchased a 75-year policy, but you had a car accident. The paramedics tried to save you, and the hospital bill came to $100k, but even after that noble effort, you still died. As a result, your husband and children get $1.9M. Alternatively, if in our hypothetical situation they succeed in resuscitating you, the company would keep the $1.9M for future medical bills, and, if they fulfill their promise of life expectancy, they pocket the remainder as profit on your 75th birthday.
It seems like this arrangement would put all of the right incentives [2] in place for both companies and individuals. Most individuals would want to avoid trivial medical expenses in order to maximize payout to family in case of accidental death. Companies would want to maximize health and longevity in order to profit from the end-of-life payout. And our society would have a way to rationally consider the value of life without resorting to arguments that essentially conclude “life is of infinite value,” and in doing so, prevent sensible gerontological triage. To put it into perspective, it makes little sense that we spend $1M (as a society) trying to save a 92-year-old when that same amount could have saved 10 teenagers.
Longevity Insurance companies would be incentivized to become heavily involved in medical research that prevents disease, prolongs life, and keeps people healthy. I can imagine a whole array of things that make sense in this context. For example, it would be the right place to fund studies on genetics, it could be the right vehicle for getting ‘free’ immunizations, and it could even make public funding for “health insurance” easier to pass—simply set the bar low enough that everyone can agree on an age that society will extend a policy for. Do we all agree that everyone in our society should live to age 50? Super! The government will cover Longevity Insurance up to age 50.
[1] We could also just allow Longevity Insurance as a free-market alternative, but for the sake of argument, let’s ban its competitors.
[2] The one incentive that Longevity Insurance does not seem to address well is the possibility of next-of-kin killing their loved one just prior to the end of an insurance policy. One option would be to require a one-year moratorium in the case where someone dies within a year of their policy ending. This would give time for an investigation before awarding large sums of money.
* crosspost from my blog, http://halfcupofsugar.com/longevity-insurance
It sounds like you’re trying to get them to maximize the probability of you living to a certain age. It would be better if you could make the incentive more along the lines of maximizing your lifespan. For example, you insure 100 years for $10,000 per year. If you die when you’re 50, then your closest relative gets reimbursed for the remaining 50 years, so they get $500,000 back. Also, you don’t have to set it as constant. For example, you could say that you will pay $10,000 per year until you’re 50, and then $5,000 per year after that. That way, you get more money if you’re not retired. You could also set it so you pay $5,000 per year, but have an additional $5,000 per year before 50 if you’re capable of doing your job. And maybe specify it more so that if you’re disabled your life isn’t worth as much.
I notice that in your example, if the money will pay off tomorrow, they will pay up to the remaining amount to keep you alive one more day.
I just noticed something else that would improve this.
You only really want the money back for when you need to support your family and such. The rest is just to give them incentives. The obvious thing to do would be to have them pay you a certain amount back no matter what so you don’t have to absorb that risk, but that would ruin their incentive. To fix this, add a second company. They pay a portion of your premiums, and if you die, they get the money that you don’t need.
Suppose early on in my life I incurr $1,999,999 in expenses. Then, later on I have an unexpected accident that costs $2. Prima facie the system should still fund this. However, your system, by not focusing on the margins, does not satisfy this.
Also, you probably want a smoother tail. Your system would spend $2m to extend the life of a 74 year old by a year, but nothing to extend the life of a 75 year old by 20 years.
Also, I don’t think that, in equilibrium, health insurance companies are actually incentivised to fund research into longevity. This would only make policies cheaper if consumers hadn’t already taken into account the increases to lifespan.
Finally, I think this post would be improved if you hadn’t mentioned its making public healthcare easier—Politics is the Mind-Killer.
Interesting idea, but I see two problems :
“Most individuals would want to avoid trivial medical expenses” that’s actually a huge problem. In our societies, people tend to postpone going to see doctors (because it costs money, even if most is paid back by social security (like it is here) or private insurances, and because it takes precious time). But for many disease, the earlier they get treated, the easier is the treatment. You have a small weird thing on your skin, you don’t go see a doctor for that, and it happens to be a skin cancer and you die, or it happens to be a wart, and when you finally go see a doctor, it has multiplied and you may even have contaged someone else. The key for efficient healthcare is frequently going to the doctor to check you don’t have anything. Maybe there could be a number of checks/examens that are not taken from your $2 millions ? A visit to the dentist and general practitioner for a checkup a year and a few things like that ?
You only consider “life”, but not in which condition. So how does your mechanism handle issues like eyesight ? How does it handle artificial limbs for people who require them ? The interests of the person and of the insurance will be opposite in those issues—the insurance doesn’t care if you’re blind and limbless, as long as you’re alive, but the person may prefer to be able to see and move around, even if it means a risk of dying earlier (surgery always has a risk).
In response to the second point, perhaps one could insure, not a number of years of life, but instead some number of QUALYs, perhaps to be had within a certain time-frame.
So then, if you go blind (which I think has a disability weighting of about 0.5), the insurance company would have to cough up the proportion of the money corresponding to the QUALYs you’d lose.
So suppose I’m in a contract for 80 QUALYs up to the age of 80, and the value of the policy is $2 million. Then if I go blind at 35, I’ve lost (40/2) = 20 QUALYs, which is 1⁄4 of my policy, and so I’d get $500,000.
Perhaps if there was a prospect of, say, your blindness being reversed, the company might pay out each year you were blind, or something.
Your second point is one I hadn’t considered. I suppose for some conditions, there would be a correlation between lowered quality of life and early death, in which case a ‘longevity insurance’ company would calculate quality of life as a correlated factor of longevity. For example, blindness probably has a higher rate of accidental death, so there would be some incentive for the insurance company to help, but the incentive is not as strong as the individual’s desire for the improved quality of life that sight would bring.
As an aside, it seems to me that evolution has factored these correlations in as well. We have two eyes, presumably, because an insurance policy against blindness improves the chance of procreation. Unfortunately, as individuals, we don’t necessarily want to maximize posterity, but some unique mix of longevity and posterity.
Robin Hanson’s proposal is pretty similar.
You should really call this something else, since longevity insurance is already a thing.
(Don’t feel bad; I actually invented them independently before finally learning that it was already a mature industry. Oops.)
To ask the main question that the first link brings to mind: What prevents a person from paying both a life insurance company and a longevity insurance company (possible the same company) relatively-small amounts of money each in exchange for either a relatively-large payout from the life insurance if the person dies early and a relatively-large payout from the longevity insurance if the person dies late?
To extend, what prevents a hypothetically large number of people to on average create this effect (even if each is disallowed from having both instead of just one or the other) and so creating a guaranteed total loss overall on the part of an insurance company?
Well, nothing, I would imagine; but keep in mind you are locking away a lot of money for a very long period of time, and the payouts are constantly adjusted with age—so unless the companies are outright screwing up and allowing an arbitrage opportunity, you reduce your expected returns either through not getting as much ‘relatively’ as you seem to expect or by opportunity cost (the companies returning your money, but having profited off the ‘float’ - which is how insurance companies have long been able to pay out ‘more’ than they should, because they made their profit off investing your money and not taking a percentage of your premiums).
I assume that the insurance company won’t sell a policy that is unfavorable to them in expectation. The way insurance companies make money is to set their rates so that they win on average. If you buy both life insurance and longevity insurance, you’ll find that the payments you put in exceed the value of the payout, at least in expectation.
Put another way: you’re dutch-booking yourself, not them.
Or have I missed a nuance here?
An appealing idea. I would also consider such an insurance.
I also see a problem, similar to kilobug’s second point: How would such an insurance handle a person refusing some recommended treatment? Especially if the treatment is effective (say in terms of life expectancy), but also has nasty side effects (for a while/for the rest of your life/etc). The parallell obvious to me is life insurance and suicide, but life insurance does not pay out for death due to suicide. The policy becoming void if treatment is refused (without caveats) would, however, be a dealbreaker for many people (me included). As far as I can tell, Robin Hanson’s proposal also does not address this.
I can imagine wanting to say something about what risk/quality of life/life expectancy tradeoffs I would want to make (and thus what treatment I would refuse), but it’s not clear to me how something like this could be specified.
In your example there is very strong inclination to kill yourself the day before 75th birthday, and very strong inclination to ‘spend’ all the money trying to save you after that suicide attempt even if it is 100% hopeless (better to put money through the hospital than to give them away to next of kin). Note: you can’t reliably recognize suicide. The worst forms of suicide (closing eyes and steering onto opposite lane on the road for example) are indistinguishable from accidents, and so you will treat accidents as suicides or suicides as accidents.
A sensible policy really needs to avoid discontinuities and special cases.
Financially, this sounds very similar to whole life insurance. Conceptually, you pay premiums over time, and the policy builds in value (as opposed to term life insurance, which does not build equity). At any time, you can borrow against the value of the policy (i.e. to pay medical expenses). Eventually, the policy matures and can simply be cashed in.
If that financial instrument would fulfill the described purpose, then great. But when I went shopping for that product, it was incredibly expensive. I pay ~$1500 for term life insurance, and I think that life insurance that built value was about an order of magnitude more expensive.
Interesting idea. I would consider this option in a free market. I wonder though, about what the natural price would be. It may be that there are so many very high cost ways to try to keep somebody alive that this plan would be very expensive. That is, it may be that so many people end up getting million dollar cancer treatments that the company has to pass this onto consumers in the form of a high monthly premium.
Sounds interesting, but I’m not clear on how it would work with treatment that has a much larger effect on quality of life than life expectancy.
I don’t think we actually do that, once you take into account how much we care about the individuals in question. That is, while I’m willing to believe that we’ll spend $1M on Grandpa rather than 10 teenage strangers, I’m not at all convinced we would make the same choice when it’s Grandpa versus 10 teenage family members in criticial condition. (And there’s no way we would if it were 10 teenage family members versus a 92-year-old stranger.)
What this complaint is really about, it seems to me, is that people spend their resources on people close to them rather than on others who “need” them more.
And then if you die, the insurance company could have you frozen to avoid paying out however much is left in your coverage. Of course, they would first have to lobby for cryopatients to be legally recognized as non-dead.
But then again they would have little incentive to reanimate you later.