defined precisely in Section 4, should make optimal co-payments go down.
We have shown that this need not be the case. On the contrary: With
homogeneous preferences, Proposition 3 suggests that it is quite likely that
optimal co-payments increase with more concern for equity.
When preferences for health care vary across people, it is not obvious
in what direction optimal co-payments change when society’s concern for
equity increases. Proposition 2 suggests that in a country where income
differences are small, optimal co-payments should be lower the stronger is
society’s concern for equity, at least for treatments for which preferences are
rather heterogeneous.
We have used the term “preferences” and “severity of illness” interchange-
able throughout the paper to describe the variable £. This has been deliber-
ate: The way the model is set up, it seems natural to think of £ is a measure
of the severity of an illness (if untreated). On the other hand, one can simply
interpret £ as a parameter in (a special case of) a utility function where ag-
gregate consumption (y) and health care (—£) enter, and where health care
is only valued positively if one has some illness. In such a setting the value of
the parameter £ will simply represent the willingness to pay for a treatment
should one require it to avoid a specific illness. Also at the more practical
side, one can interpret the variable £ differently. Consider for instance the
case of prescription medicines. Some prescription medicines are good exam-
ples of a large heterogeneity in the population regarding the benefits of the
medicine. In many cases a new and more costly medicine will have the same
primary medical effect as a medicine already in use. However, the new medi-
cine may have weaker unpleasant side effects. Such side effects very often
vary strongly among different patients, being non-existent or weak for some,
and very severe for others. In this case it is natural to regard the variable £
as a measure of severity of an illness. A different example is a couple who can
only have children through assisted fertilization. The term £ is in this case a
variable reflecting how much worse of the couple feels without children than
with, i.e. a variable measuring the strength of the preferences for having
children.
The degree of heterogeneity of preferences in a society is likely to vary
15