Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Individuals use a variety of strategies to cope with health risks. They can adopt averting behaviors to reduce the probability that an event occurs or they can mitigate the risk to reduce the severity of the event if it occurs, for example through the purchase of insurance. Averting and mitigating activities can be funded by individuals, funded publicly by a governmental organization or funded through a mixture of private and public sources. The success of policies designed to address public health risks depends in part on the willingness of individuals to adopt averting or mitigating behaviors, in particular behaviors that complement public policies and contribute to society's resilience in the face of the risk.

Models of the spread and economic impacts of serious public health threats such as a pandemic flu or a bioterrorism attack on the food supply require estimates of the willingness of individuals to adopt self-protection behaviors in order to produce accurate forecasts. Some secondary data on individual self-protection decisions in response to public health threats exists; including both observed data and self-reported changes in protection behavior as infectious diseases become more prevalent. However, future threats and individual responses to these threats will differ in important ways from past influenzas or outbreaks of infectious disease. We conducted a stated-preference conjoint survey to examine the willingness of individuals to adopt self-protection behaviors in the context of the risk of food borne illness. Stated-preference conjoint surveys represent a rigorous and systematic method to collect data on factors that influence individual choices. These surveys can provide data on self-reported behaviors in the face of new public health threats or scenarios that fall outside the range of our observed experience.

The survey posed a choice of three plans, all subsidized by a tax increase, to address the risk of food borne illness that varied in the cost, the risk of illness and the severity of illness. The tax would be used to subsidize a private averting behavior, a private mitigating behavior or for the government to hire more inspectors. We employed a split sample design to test two different formats for the conjoint question – a forced choice between two plans followed by a vote on the plan and the new tax and a vote among three choices – the two plans and an "opt-out" option. A total of 1,606 interviews were collected from the Knowledge Networks web-panel in September 2007.

Results suggest that respondents were able to understand the attributes of each proposed plan, and the significant attributes had the expected signs. However, the two different formats yielded interesting and different results in terms of preferences across the three plans and the opt-out. Along with other data on risk perceptions and household expenditures collected as part of the survey, the estimated parameters could be used to inform an economic model of public health problems like pandemic influenza for which data on the elasticity of individual self-protection behavior are scarce.