Individuals exhibit systematic tendencies to overstate the risks of unlikel
y lethal events. If the risks of passive smoking are overstated in this man
ner, and if passive smoking is not harmful to adult health, then passive sm
oking by adults should have a discernible effect on subjective evaluations
of health status, but no corresponding effect on health. This idea is exami
ned empirically below using data from the National Health Interview Surveys
.
The empirical results can be summarized as follows. Passive smoking is asso
ciated with assessments of significantly poorer health. Poorer health asses
sments are associated with significantly greater medical resource use. Howe
ver, direct estimates of the effects of passive smoking on health care use
indicate no statistical association whatsoever. These results are consisten
t with a model whereby individuals systematically overestimate the effects
of passive smoking on their health and where the short-term effects of pass
ive smoking on adult health care costs are negligible.