This study utilizes reference lotteries on life and death to establish
a death-equivalent metric for valuing long-term health effects. We us
e a computer-based survey approach to elicit choices among residential
locations that pose different risks of chronic disease and dying in a
n automobile accident. From paired choices between different locations
, we infer their rates of trade-off between reducing the risks of chro
nic diseases and the automobile death risk. The values of reducing the
risks from two diseases, a nerve disease (peripheral neuropathy) and
lymphoma (cancer of the lymph system), are measured in terms of both t
rade-off rates with the risk of an automobile death and with dollars.
While the use of reference lotteries for monetary outcomes to establis
h a utility metric is well established for monetary outcomes, our resu
lts suggest that reference lotteries on life and death can also be app
lied with decision-makers facing realistic choices to construct a util
ity metric for valuing health status. The results were corroborated by
a strong positive correlation between the risk-risk trade-off values
and relative aversion scores for the different health outcomes, as wel
l as by the relative values of avoiding the three diseases in our stud
y.