The Impact of the Great East Japan Earthquake on Commodity Prices: New Evidence from High-Frequency Scanner Data

Using high-frequency point-of-sales data, we investigate the impact of the Great East Japan Earthquake on prices. In contrast to the official Consumer Price Index that shows little impact, our price index based on scanner data reveals a modest yet notable increase in within-store commodity prices following the disaster in eastern Japan. We find that the disaster had a large and lasting effect in reducing the frequency and magnitude of bargain sales at retail stores especially in Tokyo. However, under severe commodity shortages, the standard index based on within-store same-commodity price changes may not capture the actual changes in the cost of living. To overcome this problem, we propose two alternative price indices based on (1) commodity price without distinguishing stores and (2) unit price without distinguishing commodities (i.e., brands) within the same category. Focusing on the food categories for which the disaster induced large excess demand, we find that rise in the alternative price indices was considerably greater than that in the standard price index. We further show that observed price elasticity between stores (as well as between brands) declined significantly after the disaster, indicating that, in the face of commodity shortages, the same commodity sold at different stores (or even different brands within the same category) became closer substitutes for consumers.