Hacker News Reader

Comment by saas_co_de | Hacker News Reader

Unjust as in not justified on a rational economic basis.

Lotteries may be "fair" as you describe them but they would be an irrational way to distribute resources on a broad scale.

For instance, if your salary each month were a random amount taken out of a general pool distributed by your employer you would only want to stay in your job if you were a really low productivity worker with no other opportunities and would be desperately seeking alternative employment otherwise.

View on HN

Lotteries can be a very rational way to distribute resources broadly. In an emergency situation where several people need a resource but there is only a limited amount available randomly distributing the resource is the fairest and the most rational way to get it to those in need in order to prevent other activities such as hoarding, scalping and favoritism.

A trivial counterexample is an organ transplant. It doesn't make sense to flip a coin when deciding whether to give a healthy liver to an octogenarian or a 3-month-old.

There are different forces at play here (not 'ambition-sensitive' ones), but it's a case where randomness fails miserably. Obviously, the 3-month-old should get it -- indeed, this is how scarce organs are actually distributed.

My grandmother has earned it. She did a great job raising a family of 7, all of whom will miss her when she dies. She does not have a criminal record and is very unlikely to get one.

The 3-month-old has done nothing of value. Most likely, he will be just typical. There is a good chance that he will be a loss for society. His parents can probably just make another kid, and that kid would be a lot less sickly.

I think the conclusion is obvious. The octogenarian should get the organ.

Your first-year-philosophy-student contrarian argument is cute, but it's pretty obvious that the grandmother can only gain 10-15 years out of the liver, whereas a newborn can gain an entire lifetime. Getting a transplant is about extending life. Choosing someone where longevity is maximized makes the most sense from a medical standpoint.

Obviously it's also chosen based on moral "merit" where cases are less clear-cut -- say a 40-year-old rapist on death row vs. a 40-year-old family man and pastor. Again though, it's pretty obvious that a coin-flip isn't any good. Most people would choose the latter over the former.

You're choosing the goal when I thought the question here is how to set the goal in the first place.

You can say to use a utilitarian standard, but that simply begs the question of whose utility, as these can be in direct conflict and in practice, people choose their own or their in group's utility.

Also, even when we set up some political system to enforce this inequality, we often have ended up making a different concentration of power in the positions that actually run that system, destroying the very equality that was supposed to be created. Money is just one form of power and different kinds of power are generally convertible.

Your mistake is to treat a year of person X's life as equal to a year of person Y's life.

Her 10 or 15 years is worth more, because we can already know that she is a good person. The young person could haunt us for decades.

Also, why should we extend life? Well, some people are worthy of it. So again, my grandmother wins.

I see no benefit the maximizing the value of random abstract people. Merely maximizing total longevity fails to distinguish desirable people from undesirable people. This is partly why death row exists; in that case we certainly aren't maximizing longevity, and we are declining the obvious choice to extend life.

If maximizing life were really the goal, we could ban birth control or even force people to become pregnant.

Past accomplishments count. My grandma has earned the organ.

There is no need for relentless condescension when all you’re trying to say is what people consider “fair” should not actually be fair at all, but rather skewed to favor whatever world view they subscribe to.

The coin flip is no good if you believe wicked people should not be held in equal regard to those who have done nothing wrong. But it is perfectly fine if you believe the only thing that matters is a truly fair outcome.

IT's a trivial counterexample because it's so easy to see the ramifications int aht case. In many cases foresight is quite limited and people wildly overestimate their predictive powers.

But what if the octogenarian really wants to live? It may be justified to give it to the 3 month old, but that is not fair to the octogenarian and incredibly ageist.

This argument is terrible and has been put to rest by Nozick in 1974. Look up the utility monster.

There's nothing intrinsically wrong with being ageist. This is a good example why.

Then again the octogenarian already lived all those years and the child obviously had no such opportunity.

How is giving to somebody who had something and used it and not giving it to somebody who hadn't had the same opportunity fair?

For a moment in time, the octogenarian and the 3 month old are both in a competition for the same resource, if they both have the same chance of getting it, and the octogenarian happens to get it, the outcome is still fair within this context. It could just as easily have gone to the 3 month old.

It is not fair to count the octogenarian's past good fortune against them in what should be a fair contest. But we do it, because when handing out organs we don't care about fairness, we care about viable lifetime remaining.

When searching for fairness a lot of people seem to believe that good luck in the past means it should be harder for you to have good luck in the future. Ironically, this is very unfair.

"Fair" does not mean "increases entropy" in this context. We can split hairs about what exactly "fair" is supposed to be, it's probably central to this whole nebulous disagreement, but a lowest common denominator would be something like "has higher value than opportunity cost". This may even increase entropy in aggregate. A fair die does not make a fair game.

Half the problem is that, in practice, 'fair' often ends up defined as "in favor of my and my in-group's interests" by choosing a principle that coincides with those interests to maximize.

It matters whose costs we're measuring here and it matters how we measure that. See what responses you get when, e.g., you change the problem to be deciding who gets the liver among two twins who are identical except that one is, say, a Republican and the other a Democrat, or some other situation where the only difference in utility is which of two opposed groups would be aided by their survival.

Then compare the responses you get to how much that person approves or disapproves of the respective beliefs.

It is not fair because the octogenarian really wants it, perhaps more than the 3-month old?

Does the octogenarian want to live more or less than the 3 month old?

> Obviously, the 3-month-old should get it.

Octogenarian Einstein/Tesla vs random 3-month-old? Is it still obvious?

Reply on news.ycombinator.com