Notes to Pages 198–204the level of reported happiness, which if generally true would support
cardinality.
- For a more formal exposition, see online Annex 15.
- We come later to the issue of what weight to give to the happi-
ness of different individuals. For the moment we can assume that we
simply add them up. - The position is more complicated when there are large mutu-
ally exclusive projects. - Produced by the National Institute for Health and Care Excel-
lence (NICE). - NICE has used a cut- off between £20,000 and £30,000— say
$35,000. This is the cut- off for a healthy life year. It is broadly in line
with the Department of Transport’s estimate of the human cost of a
fatality, which in 2008 was £1.65 million (see Deloitte LLP [2009], table
1). This value, though based on willingness to pay, presumably reflects
the discounted value of the remaining years of life of the typical traffic
accident victim. The US Department of Transportation uses a some-
what higher figure. - For the present we assume that one unit of LS equals 0.1 QALYs.
By questionnaires to the public, a team at Sheffield has established
that people are willing to sacrifice approximately ten units of life-
satisfaction for one more year of life at LS = 10. - Disposable income is about £20,000 per head, or say $20,000 ×
1.5. Thus ∂H/∂Y = 0.2/(20,000 × 1.25) = 1/125,000. - See Chapter 2.
- Kahneman, Ritov, and Schkade (2000).
- We obtain the monetary value of some nonmonetary experi-
ence X by running a happiness equation (where H is happiness and
Y income), H = a 1 logY + azX. Then the equivalent variation of income
when X changes is given by ∆Y = a^2 a^ Y 1 ∆X. - According to Rawls (1971) we should simply focus on the very
bottom. - We ought to mention some other more data- intensive ap-
proaches that focus directly on misery. One approach is to focus specifi-
cally on negative emotion as measured by replies to questions like “How
sad/worried/frustrated/angry were you yesterday?” (For data on replies
to these questions see Helliwell, Layard, and Sachs [2012], chapters 2
and 3.) Alternatively we could use time- use data, where individuals are
asked about each episode in the previous day, with questions for each
episode about the extent of various positive and negative emotions. As