This study analyses the relation between perceived health status and intertemporal choice. We use data from experiments with real monetary rewards conduEted among students in South Africa to estimate risk and time preferences. These experimental data, based on muitiple price lists developed by Coller & Williams (1999), Holt & Laury (2002), and Harrison et al. (2002, 2005), show that HIV+ agents and participants that perceive to have a high HIV contraction risk are less risk-averse. Although the latter group displays higher discount rates, HP positive agents seem to have substantially lower discount rates, indicating longer time horizons in spite of their lowered life expectancy. However, we show that direct estimates of discount rates can be seriously biased estimators of the pure rate of time preference when other factors than just the pure rate of time preference are not considered simultaneously. We correct for differential mortality risk, risk aversion and differences in anticipated future marginal utility increases and price in these factors when calculating pure rates of time preference from observed discount rates. Once these factors are taken into account, HIV+ agents’ time preferences conform to expectations.
discount rate risk aversion perceived HIV infection risk mortality time preferences marginal utility hyperbolic discounting