No abstract

Damages from tropical storms are enormous and predicted to increase with climate change. This study uses high-frequency data to investigate typhoons' effects through information channels on Taiwan's wholesale vegetable market during the 1996-2014 period. We first identify effects on prices and quantities during storms' striking periods and further separate price effects into those due, respectively, to supply and demand shifts. The results show that for typhoons that make landfall, prices rise significantly during the striking period, and the decomposition results indicate that most of the price effects during periods with warnings but no landfall are due to demand shifts, which supports evidence of precautionary purchases by consumers. However, during the landfall period, the price effect is mainly driven by decreased supply. In addition, we find that typhoon effects differ between specific vegetables, and the magnitude of precautionary purchase is correlated with expected damage to those vegetables. Consumers also make more precautionary purchases when they face higher-intensity typhoons or learn typhoons will make landfall within 24 h, and such prior information related to intensity or landfall urgency can also amplify early harvest effects by farmers.

This research uses a difference-in-differences framework to investigate the effect of new risk information on housing prices in Taiwan. The results show that this information changed individuals' subjective risk perceptions, so that housing prices in the highest-risk areas dropped, but only temporarily in the first three months after the disclosure. Also, this information effect happened for those apartments lacking certain earthquake-resistant characteristics, while there was no effect for the apartments with such characteristics. In addition, we investigate the dynamics of the effect around the boundary. We demonstrate that individuals can form continuous risk beliefs based on discrete information, as the housing prices dropped more sharply for those apartments located closer to the center of the risk area. Furthermore, individuals updated their risk beliefs differently for apartments with different earthquake resistance qualities. For those apartments lacking such qualities, the immediate price drops were relative larger, and the housing prices returned to normal more slowly, relative to the apartments with such qualities, thus considered "safer". The effect did not disappear at all for those apartments lacking earthquake-resistance that were also located in the center of the risk area.

No abstract

No abstract