Risk preferences, risk perceptions, and risky food

This paper presents the results of a study that tests the hypothesis that the effect of risk preference on choice is a function of the specific risk-preference measure utilized. In addition, this study tests the hypothesis that the effect of risk preference on choice depends upon its interaction with risk perceptions. I elicit three distinct measures of risk preference: a standard real-money Holt and Laury measure, a hypothetical health-variant of the Holt and Laury measure, and a non-context-specific self-assessment measure. I also elicit information regarding risk perceptions (specifically, food safety). These data are combined with choice data focused on consumer preferences for raw oysters. Results indicate that, after controlling for key oyster attributes, perceived risk perception is highly significant. Additionally, the effect of risk preference is significant, and the effect depends on whether respondents held informative or non-informative risk perceptions. In a treatment that includes only named oyster varieties, I find that although respondents generally prefer named Atlantic coast oysters to named Gulf and Pacific coast oysters, those who hold non-informative risk perceptions are even less likely to choose Gulf and Pacific coast oysters as the magnitude of risk aversion increases. In another treatment that includes a generic “commodity” Gulf coast oyster, I find that although named Gulf coast oysters are generally preferred to the commodity Gulf coast oyster, respondents with non-informative risk perceptions are significantly less likely to choose named Gulf coast oysters as the magnitude of risk aversion increases.