Anonymous asked:
What are QALYs and DALYs?
They’re two different ways of assessing how bad particular disabilities are. Imagine that the average person would trade a year of depressed-person life for six months of nondepressed life. If that’s true, then depression is worth half a DALY/QALY.
A lot of disabled people don’t like them because they imply that disabled lives are worth less than nondisabled lives. I like them because they’re a first attempt to empirically Do Utilitarianism.
The actual methodology behind these numbers (asking non-disabled people how they think they would feel if disabled) has been shown to lead to inaccurately low quality-of-life estimates. (This article from the British Medical Journal has a ton of links.)
Using DALYs or QALYs also leads to (seemingly?) wrong conclusions. Let’s say we can rescue one of two groups: 190 neurotypical students, or 200 students with Asperger’s*. (These groups are the same in every respect other than quantity and neurotype. Their average remaining life expectancy is 65 years.) This Global Burden of Disease paper generously hosted by jefftk gives the disability weight of Asperger’s syndrome as 0.110, but let’s be cautious and use the lower value of their 95% uncertainty interval, which is 0.073.
Now, let’s see what the DALY-maximizing action is. Saving the 190 neurotypical students has an expected value of (190 students)*(65 life-years/student)*(1 - disability weight 0)=12350 DALYs. Saving the 200 students with Asperger’s has an expected value of (200 students)*(65 life-years/student)*(1 - disability weight 0.073)=12051 DALYs. If we’d used the actual weight given for Asperger’s, we’d have gotten 11570 DALYs as the expected value. Saving the 190 neurotypical students and letting the 200 students with Asperger’s die is therefore the DALY-maximizing action.
DALY-maximizing leads to 10 more deaths than acting in accordance with other methods of morality would have. I consider this a pretty terrible conclusion.
I suppose it’s a good first start if you keep the limitations in mind and don’t build a decision-making agent that treats “maximize DALYs” as its utility function. Nobody’s actually suggesting that the one true utility function that leads to the most moral actions is “maximize DALYs”, AFAIK, so this is a little bit of a strawman argument.
*Note: I prefer person-first language with regards to autistic people, and consider myself and other people previously diagnosed with Asperger’s to be autistic, but Asperger’s and autism had different disability weights in the report, and “Asperger’s students” sounds like the students are in Hans Asperger’s class.