Valuing food-borne risks using time-series data: The case of E. coli O157:H7 and BSE crises in Japan

This study evaluates changes in consumers' concerns on food safety after the outbreaks of E. coli O157 and bovine spongiform encephalopathy (BSE) in Japan using household consumption time-series data. A food demand system for Japanese households is estimated using the linear approximate almost-ideal demand system (AIDS) model to evaluate the willingness to accept (WTA) compensation for risk. The Kalman filtering method is applied to produce estimates without a priori assumption regarding timing of the changes. The WTA value rises immediately after a food safety crisis occurs and declines in a short time. However, it does not return to previous levels for an extended period. A possible explanation for remaining effects of a crisis might be that they are the results of habit formation and learning effects of consumption. [EconLit citations: D12, D18, Q13]. © 2006 Wiley Periodicals, Inc. Agribusiness 22: 219-232, 2006.