Medical Ethics
Michael Green
Manuel Vargas
phone, office information

11 May. Numbers don't count.
13 May. Aggregation and Kamm.
13 May. Glover
15 May. How to measure quality in a QALY
15 May. QALYs and fairness
22 May. QALYs in macro and micro contexts
27 May. Rakowski and Post on preferring the younger.
29 May. How great a loss is death?
3 June. Consent as a way out?

Medical Ethics: 18 May. The veil of ignorance argument


This argument is supposed to show that allocating health care according to the number of QALYs that will be produced is fair, despite the apparent cases of "double jeopardy" that Harris notes.

Watch the professor add!

Here's where I went wrong. I took the probability of being D or E (2/3) and I multiplied it by their combined QALY scores (10 + 10 = 20).

You all were absolutely right: as far as a chooser behind the veil of ignorance is concerned, all that's relevant is his probability of getting the QALYs of one person. These are added together in determining the expected utility of a policy to an individual chooser behind the veil of ignorance; the value of the overall outcome and its probability are not relevant to the individual chooser. Consequently, I shouldn't have added D and E together as I did.

Thanks for catching my mistake.

Lesson: write out all the steps. I cut corners in preparing my notes and made a simple error.

Singer, et. al., Harris, and Taurek

Singer, et. al.: ration health care so as to maximize QALYs, throughout the population as a whole.

Harris: save as many lives as you can.

Taurek: give everyone an equal chance of living.

In the last example, Singer, et. al. would favor C over D and E. Harris would have favored D and E over C. Taurek would have insisted that C, D, and E all have equal chances of being saved.

Does Harris have a consistent position? Doesn't he attack Singer, et. al. with an argument like Taurek's? But since that's so, how can he favor saving as many lives as possible as opposed to giving everyone and equal chance?

Another question about Taurek

How would giving people an equal chance work in this case?

We can save either C or we can save D and E: there's a drug and C needs it all whereas either D or E could get by with half of it.

Suppose we give each a 1/3 chance of being saved -- we draw from a tank of three ping pong balls.

If C wins, D and E lose.

But what if D wins? We'd still have half the drug left. It's no use to C (he needed 100%). But shouldn't we give it to E? If we do give it to E, though, D and E would have a 2/3 chance of winning -- if D wins, E does too (and vice versa).

Solution: we'd have to increase the number of ping pong balls with C's name on them to make the decision truly random. D and E have, in effect, two balls (since E benefits if D's ball is drawn and vice versa). So, C should have two balls and D and E should have one each.

Previous | Next

This page was originally posted on 5/27/98; 12:08:41 PM and was last built on 6/3/98; 2:44:13 PM with BBEdit and Frontier 5 on a Macintosh running System 8.1.