Dissertation Title： "Health Risk Behavior, Preferences and Policy"This dissertation address health behavior in response to risk, and public preferences regarding risk. Health behavior is explored in two papers on the topic of child seating behavior in motor vehicles. Seating position is the behavior of interest, in response to the health risk to children posed by passenger air bags.
Paper #1 is an empirical study of child seating behavior in vehicles traveling in New England. From a sample of 503 drivers observed and interviewed in 1998, we found that the presence of passenger air bags in vehicles was independently associated with a decreased chance of children being seated in the front seat. This result suggests that drivers were adjusting their behavior in response to the hazard posed by air bags to children.
Paper #2 reports on an accompanying study that analyzed trends in child seating behavior from 1990 through 1998 in vehicles involved in fatal crashes. Using a national database of fatal crashes to identify predictors of child front seating, we found that while the frequency of front seating declined in all vehicles over this period, greater declines were associated with the presence of both a passenger air bag and younger children in the vehicle. These results confirmed and extended our New England findings, implying that the behavioral response to the air bag hazard was greater for younger children. Public preferences regarding risk were explored in the context of the role that risk characteristics play in preferences regarding health policies.
Papers #3 and 4 present the development and implementation of a preference elicitation instrument. We describe the application of a constructivist approach to preference elicitation, as well as measures of survey effectiveness and the use of internal verification mechanisms in data interpretation. Our results demonstrate that two particular risk attributes, voluntariness and personal controllability, are important influences on preferences. In the health scenarios we explored, the public preferred to reduce risks of equal magnitude that were involuntary or uncontrollable, compared to those that were the converse. Our findings suggest that economic analyses of health programs that involve risk should consider the effect of risk attributes on preferences.