Sunday, February 12, 2017

Cash and Stranger Lives Stated versus Revealed Preferences

My roommate recently posed a hypothetical dilemma: Imagine I am a mischievous powerful alien, or if realism helps, a social scientist at a university with the world's worst IRB process.

I offer you $100,000 with the following condition: one (randomly chosen person) will die upon your taking custody of the cash.  Do you take it?

To emphasize: the person is chosen at random from the entire global population.  A twenty-something Indian is more likely than a nonagenarian Norwegian, but no guarantees either way.  Given an average of 150,000 deaths everyday, you would only be increasing that by 0.0007 percent, but that 0.0007 percent is on you and your $100,000.

What do you do?

Go ahead and think before you read on.  I'm going to do a few spacer paragraphs so your eye doesn't jump ahead to my thoughts and bias you.

Spacing ...

Spacing ...

Ready?  Need a few minutes?  No rush here.

Spacing ...

More spacing ...

Okay, here we go.

I've repeated the question a few times in conversation, and so far no one has said they would take the cash.  Yet consider this: a number of highly-effective non-profits have estimated their marginal costs of saving a life to be a few thousand dollars.  (Granted, taking non-profits' word for it is not objection-free, and there is some thorniness to measuring what's known as a Quality-adjusted Life Year in the academic lit, but I don't think the numbers wildly attached from reality.)

I find this remarkable.  Why might it be someone would turn down $100,000 they don't have to save a life (in theory), but aren't willing to part with $5,000 they do have to do the same (in practice)?

1.  We're not as altruistic as we would like (others?) to think we are, and we really would take the $100,000 if the offer were credible.
2.  It's psychologically easier to opt out of committing an evil act than is to commit to an altruistic one.
3. It's psychologically easier to part with money never owned than money previously earned.
4. People don't have good information on the costs of saving a life, overestimating them by a few orders of magnitude.

These aren't mutually exclusive, but I'd like to think #4 is both important and easily correctable.  With that in mind, I'll make a plug  for the effective altruism movement, which encourages people to give more money specifically to effective charities.  The website The Life You Can Save points to several.  Unsurprisingly, these cluster in health care in South Asia and sub-Saharan Africa.

No comments:

Post a Comment