5 ms·

That isn't how the various things called "value of a life" are calculated. The Wikipedia article linked above describes some ways to do it, and if you look at

by gjm11 6y ago

That isn't how the various things called "value of a life" are calculated. The Wikipedia article linked above describes some ways to do it, and if you look at them you will see that with one exception they are not estimating, nor trying to estimate, expected future earning power or anything like it. (Though that might affect the answer.) The first approach they describe: take N people and ask each of them how much they would pay to reduce their chance of dying in the next year by 1/M. Then the average of M times this figure is the group's estimate of the value of their lives. (The description in the article simplifies the calculation by taking N=M, but there is no need to.) The second approach: look at what people are willing to forgo in order to reduce their chance of dying a bit, or willing to increase their chance of dying a bit in order to have. If you're willing to increase your chance of dying by X in order to get Y, that suggests you value your life at no more than (the value to you of Y) / X. Look at lots of different X and take some sort of average of the resulting estimates. The third approach is looking at future earnings. The page adds this caveat: "Another potential issue when using wages to value life is that the calculation does not take into account the value of time that is not spent working, such as vacation or leisure." The fourth approach is more or less the same as the first. It's obvious that aside from the second these are not the same as a person's future earnings. And we shouldn't expect them to be; people generally value other things about their future lives besides the money they may earn. (A kinda-artificial example that I think makes the point. Suppose the following things happen: 1. Economic growth stops or slows sufficiently that no one expects investments to grow appreciably an more. 2. You get rich. 3. You retire, intending to supply your needs and wants simply by spending some of your mountain of cash. In this situation your expected future income is zero. But I bet you would still be willing to pay something to reduce your chance of early death.) Also, though this is a less important point: Life insurance is not only for paying off some fraction of your theoretical lifetime income. E.g., you can buy life insurance policies even after you have retired, even if you have no income. Obviously one reason why you would buy life insurance is to make up for loss of future income; but it could be e.g. that one thing you do is to care for other family members, and that if you died they would need someone else to do that who would need paying, and if you were buying life insurance that would be something you would take into account.

vlovich123 6y ago
If you ask a billionaire how much they’re willing to spend to save their life (or significantly improve their health) and you ask someone with a $1000 dollars in their bank account, is the answer to that question measuring the worth of a life or how much that capital is worth? Very wealthy people would be more willing to trade wealth for health/happiness whereas poorer people are less willing to make that trade off since they need that capital for more basic necessities (eating, rent, having kids, etc).
- drdeca 6y ago
  I imagine one could, in principle, look at how this varies (on average) with how much money a person has and/or their income. Hm, if one did, then perhaps one could use this as a basis for expressing how much people value other things? Though, I suppose there's the question of like, how would someone trade off risk of dieing within N years vs dying within 2N years. Like, if something would increase one's chances of dieing within 2N years by some quantity, but decrease one's chances of dieing withing N years by some other quantity, at what values would people make or not make that choice? (or, in the other direction). Depending on how that works out, maybe that would mean that this isn't a great foundation for translating between "how much do they value this (in dollars), at their current income level" and "how much do they value this (in some other unit which is hopefully more natural in some sense)". Is there something else that would be a better basis?
- gjm11 6y ago
  It's measuring the value of their life relative to the value of money, for them. In many situations that isn't the statistic you actually care about. (For the avoidance of doubt: my previous comment wasn't trying to claim that the "statistical value of a life", measured in such a way, is a good measure of how much a person values their life in non-monetary terms, or anything like that. I was just objecting to the specific claim I was responding to, namely that the "statistical value of a life" is measuring only its value as a future income stream.) Very very crudely, the value of a small gain in money is something like inversely proportional to the amount you have already, which means that the value of a given amount of wealth increases something like logarithmically with the amount of money, and either of these means that financial institutions like markets care about people's interests roughly proportionally to their wealth. Everything in this paragraph is pure handwaving, but I find it a useful intuition.

by gjm11 6y ago

vlovich123 6y ago

If you ask a billionaire how much they’re willing to spend to save their life (or significantly improve their health) and you ask someone with a $1000 dollars in their bank account, is the answer to that question measuring the worth of a life or how much that capital is worth? Very wealthy people would be more willing to trade wealth for health/happiness whereas poorer people are less willing to make that trade off since they need that capital for more basic necessities (eating, rent, having kids, etc).

drdeca 6y ago
I imagine one could, in principle, look at how this varies (on average) with how much money a person has and/or their income. Hm, if one did, then perhaps one could use this as a basis for expressing how much people value other things? Though, I suppose there's the question of like, how would someone trade off risk of dieing within N years vs dying within 2N years. Like, if something would increase one's chances of dieing within 2N years by some quantity, but decrease one's chances of dieing withing N years by some other quantity, at what values would people make or not make that choice? (or, in the other direction). Depending on how that works out, maybe that would mean that this isn't a great foundation for translating between "how much do they value this (in dollars), at their current income level" and "how much do they value this (in some other unit which is hopefully more natural in some sense)". Is there something else that would be a better basis?
gjm11 6y ago
It's measuring the value of their life relative to the value of money, for them. In many situations that isn't the statistic you actually care about. (For the avoidance of doubt: my previous comment wasn't trying to claim that the "statistical value of a life", measured in such a way, is a good measure of how much a person values their life in non-monetary terms, or anything like that. I was just objecting to the specific claim I was responding to, namely that the "statistical value of a life" is measuring only its value as a future income stream.) Very very crudely, the value of a small gain in money is something like inversely proportional to the amount you have already, which means that the value of a given amount of wealth increases something like logarithmically with the amount of money, and either of these means that financial institutions like markets care about people's interests roughly proportionally to their wealth. Everything in this paragraph is pure handwaving, but I find it a useful intuition.