Invoking Yaari's dual theory, we develop a model of individual vaccination decisions that incorporates quasi-hyperbolic discounting, risk aversion, and information. We test the resulting hypotheses for the flu season 2010/2011 using a representative German data set. We find a significant impact of time preferences on immunization decisions. The impact of the discount factor is significantly negative for exponential discounters. While present-biased individuals' demand for vaccination is not statistically different from the one of exponential discounters, future-biased individuals have a significantly higher probability to vaccinate. Stratification by gender reveals that these effects are entirely driven by men. That is, time preferences have no explanatory power for the vaccination decisions of women. This also applies to risk aversion, where more risk aversion implies a significantly higher probability to vaccinate for men but not women. All information measures turn out significant. Well-informed individuals have a much higher propensity to vaccinate than poorly informed individuals. If policy makers aim at improving immunization rates, then our results suggest that public policy should concentrate on providing easily accessible and concise information on the flu and the flu shot. Our results on time preferences and risk preferences imply a rather inactive role for public policy. Copyright © 2015 John Wiley & Sons, Ltd.
Keywords: flu shot; information; prevention; public health; quasi-hyperbolic discounting; risk aversion.
Copyright © 2015 John Wiley & Sons, Ltd.