Even if health is but one factor influencing the utility of the individual, it may though be an especially important one. Yet, economic models describing the individual’s optimal demand for health and factors that affect the individual’s health-related behaviour (from which the individual’s demand for health care, etc., could be derived) are relatively few. This is, though, one of the fundamental issues in the area of health economics, and such knowledge is vital in order both to understand risk behaviour, such as smoking and illicit drug abuse, and to create a basis for normative evaluation of health care programmes based on individual preferences.
The groundbreaking work in the area of individuals’ demand for health was done by Michael Grossman and was published in the early 1970s, building on Gary Becker’s work on human capital theory (Becker, 1965). In his seminal article on the individual’s demand for health (Grossman, 1972), Grossman treated the demand for health in a dynamic setting where the individual was born with a certain level of health that depreciated over time, unless the individual invested in health through various activities. A decade after his original paper was published, Grossman pointed out a number of directions in which his original model could be further developed (Grossman, 1982). However, even though his original model has been used for empirical applications a number of times, the theoretical contributions have been limited. Some of the perhaps most notably theoretical contributions to the original model so far have been the introduction of uncertainty (see, e.g. Cropper, 1977. Dardanoni and Wagstaff, 1990; Liljas, 1998; Picone et al., 1998; Zweifel and Breyer, 1997), insurance (Liljas, 1998, 2000; Tabata and Ohkusa, 2000), and the family as a producer of health (Bolin et al., 2001; Jacobson, 2000). For a good overview of many of these developments, see Grossman (2000).
Another simplifying assumption made in the original model was to allow the individuals to borrow and save money at the same (constant) interest rate without any restrictions or transaction costs, i.e. implicitly making the constrained maximization problem an isoperimetric problem (i.e. when there are integral restrictions, an assumption also made by Grossman’s successors 1). Thus, by applying an integral constraint, Grossman thereby assumed that the budget only had to be binding over the individual’s entire lifespan. This was a very