The Law of One Price in Scandinavian Duty-Free Stores
利用斯堪的纳维亚三家免税店的数据，检验一价定律是否成立。尽管产品相同且无贸易壁垒，但名义价格不连续调整而汇率每日波动，导致一价定律被拒绝，但大偏差会触发价格调整，限制套利机会。
Many empirical studies have rejected the law of one price. That prices of a good differ across locations has been explained by differences in product attributes and costs of local inputs, transport costs, trade barriers, and that buyers have imperfect information about prices in different locations; see Penelopi K. Goldberg and Michael M. Knetter (1997) for a survey. We examine the law of one price in situations where none of the mentioned reasons for its failure can be invoked. It has also been suggested that deviations from the law of one price are a consequence of rigid nominal prices and that different countries typically have different currencies. We explore whether this can contribute to our understanding of deviations from the law of one price. Our data are taken from three Scandinavian duty-free outlets, where each product (at the same location) has price tags in at least two currencies. Hence, a consumer has the option to choose between several prices for the same identical good. In such a setting there is a strong prior that the law of one price (LOP hereafter) holds well. However, the potential for arbitrage will arise because nominal prices are not continuously adjusted while exchange rates fluctuate daily. Based on standard tests, we reject LOP at all duty-free outlets. Given that LOP does not hold here, it is less surprising that many previous studies have found that prices of similar products at different locations differ significantly. Nevertheless, the main conclusion of the paper is that in this natural experiment, LOP remains a useful guide to the behavior of relative prices. As deviations become large, nominal prices are adjusted to reduce the deviations from LOP, thereby limiting arbitrage opportunities. The patterns at the duty-free outlets suggest that there is a band of inaction so that small deviations from LOP may persist (for almost a decade in one case), but that large deviations quickly lead firms to adjust relative prices. The findings are consistent with costly arbitrage and fixed costs of adjusting nominal prices.