Therefore, two additional constraints must be satisfied in addition to the conditions required
by PZ’s model. First, the utility value of a prevented illness must depend only on
the size of the health improvement, and second, the ability of an individual to enjoy wealth
cannot be affected by illness. The latter of these conditions is highly counter-intuitive and is
not supported by empirical evidence (see Evans and Viscusi, 1993 and Sloan et al., 1998),
making any link between CBA and CEA of the sort Johannesson and Meltzer propose
untenable.