opportunity cost of leisure) and non-employment income determine labor supply
decisions, and poor health lowers these factors, human capital theory predicts that poor
health will also be negatively related to the probability of labor force participation
(Chirokos and Nestel, 1984).4
The human capital theory approach to the treatment of health in labor models was
extended by Grossman (1972) and Michael’s (1973) models of household production of
commodities such as health. These models acknowledge that health is both a
consumptive and investment commodity and, unlike other forms of human capital which
do not necessarily depreciate, the large initial stock of health must be continually
replenished through the sacrifice of time and monetary resources. Thus, while earnings
are partly determined by investments in health capital, the stock of health today depends
on past (and current) investments in health which, in turn, depend upon past (and current)
earnings.5
In Grossman’s model, health does not affect an individual’s productivity but
determines the amount of time that the individual can spend earning income and
producing non-market commodities. The variation of individuals’ inputs of medical
services and time into the production of health will vary according to the opportunity
costs of the time lost to poor health. Michael suggests a similar result; however, his