The Empirics of Consumer Risk Attitudes and Genetically Modified Foods

The debate over genetically modifiedorganisms (GMOs) and GM foods has gained greater prominence recently. However, potential health and environmental risks, lack of trust in regulatory mechanisms, social and ethical concerns associated with GM foods have led to a rejection of GM foods by some consumers worldwide. In this study, we examine consumer attitudes and purchase intentions towards GM foods, using survey data of 624 consumers in Taiwan in 2002. Survey results show that reduced use of pesticide was considered the most important benefit of GM products, and unknown health problems associated with GM foods was the most important risk concern. The results also show the majority of respondents preferred GM product labeling to be mandatory, and of positive labeling. About 45 percent of respondents conceptually accepted GM foods. Given GM food prices were lower than those of non-GM foods, however, 73 percent of respondents revealed to be potential buyers. To quantify the effect of demographic, socioeconomic, and attitudinal factors on GM food consumption, a joint probit and ordered probit model is estimated, taking account of joint decision making by consumers, i.e., if or not one is willing to purchase GM foods by paying less; and how much less to pay. The results of joint estimation, however, show the estimate of r (cross-equation correlation) is not significantly different from zero. Separate estimation results of probit model indicate that respondents who care more about GMOs and GM products issues after being informed of benefits and potential risks of GMOs and GM products, are more likely to be willing to purchase GM foods, while older consumers are less likely to be willing to make purchase of GM foods. For the ordered probit model of price premium (how much less to pay relative to non-GM foods), the parameter estimates of all socioeconomic and demographic variables are not significant. In order to better understand consumer conceptual acceptability, we formulate and estimate another ordered probit model. The estimation results show that consumers who rank unknown health risk or allergic problems as top two risk concerns or who have attained a higher educational level are less likely to conceptually accept GMOs and GM products.