Health Prioritarianism - Harvard University: Program in Ethics & Health

Report
Health Prioritarianism
Peter Vallentyne
University of Missouri
Background
• Universal Coverage Question: What determines whether one form of
universal coverage is morally better than another?
• More General Question: What determines whether a health
policy/system, for a given country at a given time, is morally better
than another?
• General Assumption: If one system is Pareto superior to the other (i.e.,
gives some people more wellbeing, and none less), then it is morally
better.
• Simplifying Assumptions:
– A useful proxy for individual’s lifetime wellbeing is individual
health-adjusted years of life (e.g., number of Dalys)
– Fixed population (so that total = average; and to avoid non-identity
problem)
– Certainty in outcomes (to avoid the need to appeal to probabilities)
• Health-year =df one year of “perfect” health (e.g., one Daly)
Health Utilitarianism
• Health Utilitarianism (no priority to less healthy):
– (1) Greater total health-years is better.
– (2) For same total health-years, the policies are equally good.
• Problem (no sensitivity to equity): No sensitivity to how individual
health-years are distributed among people.
• Equity (relevant distribution-sensitivity):
– Might appeal to desert
– Might appeal to equality
– Might appeal to priority for worse off.
• I shall focus on priority for worse off.
• Arguably, the appeal to wellbeing, or health-years, should be to brute
luck wellbeing or health-years (i.e., not attributable to her agency).
• For simplicity, I ignore this.
Health Prioritarianism
• Health Prioritarianism: It is morally more important to increase the
health-years of a given person by n units than to increase the healthyears of a person with more health-years by n units.
• Questions:
– (1) Is Health Prioritarianism correct?
– (2) What are some of the main forms that it can take?
• Two reasons to endorse Health Prioritarianism:
– Decreasing Marginal Impact of Health on Wellbeing
– Wellbeing Prioritarianism
• Let’s explore each.
Decreasing Marginal Impact of Health on Wellbeing
• Decreasing Marginal Impact of Health on Wellbeing: For a given
individual, all else being equal, increasing a person’s health-years by a
given number of units has a smaller impact on her wellbeing, the
higher her level of health-years is.
• Example: Suppose that increasing someone’s health-years from 1 unit
to 2 increases her wellbeing by 1 unit.
– Then, all else being equal, increasing her health-years from 2 units
to 3 increases her wellbeing by less than 1 unit.
• Health Prioritarianism does not follow from Decreasing Marginal
Impact on Wellbeing (in conjunction with Wellbeing Utilitarianism).
• The former is an interpersonal condition, whereas the latter is purely
intrapersonal.
Decreasing Marginal Impact of Health on Wellbeing
• Example:
Health
Your Wellbeing
My Wellbeing
1
10
1
2
20
2
3
25
2.5
• We each have decreasing marginal impact of health on wellbeing, but
increasing your health from 2 to 3 (increase of 5 units of wellbeing)
may be more morally more important than increasing my health from
1 to 2 (increase of 1 unit of wellbeing).
• Health Prioritarianism does follow, if we assume (1) Wellbeing
utilitarianism (maximize total), and (2) everyone’s cardinal wellbeing
is cardinally affected by health in the same way (all else being equal).
• (2) is false, but it may be a good working assumption for aggregate
measures of wellbeing for large populations.
• Let’s assume so.
Wellbeing Prioritarianism
• A second reason to endorse Health Prioritarianism:
• Wellbeing Prioritarianism: It is morally more important to increase
the wellbeing of a given person by a given number of units than to
increase the wellbeing of a person with greater wellbeing by the same
number of units.
• This does not entail Health Prioritarianism: A person with lower
health can have higher wellbeing (since health is not the only factor
for wellbeing).
• Still, for large populations, at the aggregate level, wellbeing and
health will be closely correlated.
• So, if Wellbeing Prioritarianism is correct, then there is a second
reason to endorse Health Prioritarianism in practice when dealing with
aggregates for large populations.
• Let us now consider some forms that Health Prioritarianism can take.
Weakly Prioritarian Health Utilitarianism
• Weakly Prioritarian Health Utilitarianism:
– (1) Greater total health-years is better.
– (2) For same total health-years, the policy with the greater lowest
individual number of health-years is better (and, for ties, compare
the second lowest number of health-years, etc.).
• This invokes priority only as a tie-breaker.
– This is arguably too little priority.
Additive Prioritarianism
• Assume a set of finitely decreasing priority-weights for health-year
increments. For example:
• Health-year Incr Weight Total
Priority-Weighted HY
0 to 1
1
1
1 to 2
.9
1.9
2 to 3
.8
2.7
• Additive Prioritarianism:
– (1) Greater total priority-weighted health-years is better.
– (2) For same total priority-weighted health-years, the policies are
equally good.
• This treats priority as more than a tie-breaker for the same total
health-years.
Additive Prioritarianism
• One problematic feature is that this entails that it can judge it better to
give a trivial increase in health to sufficiently many very healthy
people rather than to give a major increase in health to one very
unhealthy person.
• The severity of this problem will depend on how quickly the priority
weights decrease.
– They might decrease so slowly that in practice they are equivalent
to constant marginal weights (as with utilitarianism).
– Or they might decrease so quickly that in practice they are
equivalent to leximin (absolutely priority of the worse off; see
below).
Threshold prioritarianism
• Set a threshold for adequate health-years. A person’s truncated health-years
is the lesser of her actual level of health-years and the threshold.
– For example, if the threshold is 10, and A has 8 health-years and B has
12, then their respective truncated health-years are 8 and 10.
• Threshold prioritarianism:
– (1) Greater total truncated prioritarian-weighted health-years (which
ignores health above the threshold) is better.
– (2) For the same total truncated prioritarian-weighted health-years,
greater total prioritarian-weighted health-years (with no truncation) is
better.
• Below the threshold, this gives finite priority to those who are worse off, and
likewise above the threshold.
– Moreover, it gives absolute priority to the health of those below the
threshold over those above the threshold.
• Thus, it avoids the above problem “numbers problem”.
• This involves, however, a questionable discontinuity at the threshold.
Leximin
• Leximin:
– (1) Greater health-years for a person with the least health-years is
better.
– (2) If there is a tie, greater health-years for a person with the
second least health-years is better. Etc.
• This gives absolute priority to a less healthy person.
• It faces the problem that it deems it better to give the least healthy
person a trivial increase in health rather than to give a massive
number of people who are only slightly more healthy a massive
increase in health.
Conclusion
• If one adopts Health Prioritarianism, there is the question of how
strong the priority should be for the less healthy.
• There is an on-going investigation of these issues by moral
philosophers and normative economists.

similar documents