Now or later? Anomalies in Intertemporal Choice and the Demand for Health Care

An intertemporal state-dependent expected utility model - adding up the anticipated fl‡ows of utility from wealth and from temporary emotions induced by the resolution of uncertainty - is used to explain ‘procrastinative’ versus ‘over-anxious ’demand for health care, both in comparison to the classical expected utility of wealth maximizer’s behaviour. The emotional expected utility (EEU-)maximizer, balances the marginal ‘costs ’of delaying a diagnosis (as seen from the pure expected utility of wealth (EU-) maximizer’s point of view), against
the marginal bene…ts of lowering the disutility of anticipated post-outcome disappointment and pre-outcome anxiety. If an EU-maximizer (the ‘cool ’subject) and an EEU-subject with the same - sufficiently high - rate of time preference have the choice to receive a critical diagnosis and treatment ‘now’ or ‘later’, the EEU-subject will decide in favour of ‘later’, while the EU-maximizer prefers ‘now’. For low discount rates, however, the EEU-type expresses an even stronger preference for diagnosis ‘now’ than the ‘cool’ type. In a state of sickness, the same subject, who was willing to procrastinate in the …rst place, can be lured into ‘unfair’t herapy bets promising an improbable and delayed success. If the EEU-subject knows that the health status cannot be changed after receiving the diagnosis, not knowing the status will be strictly preferred, while the EU-subject will be just indi¤erent. The model also implies that propagating preventive care should rely on ‘enlightening’ and not on ‘frightening’arguments, as the latter might raise the attractiveness of procrastination.