Scanner Data and the Measurement of Inflation
探讨如何利用零售商的条形码扫描仪数据来测量通货膨胀，这些数据提供了每种商品的价格、数量和交易额，从而可以估计替代偏差并比较不同的质量调整方法。
This paper outlines the potential use of bar-code scanner data from retailers for the measurement of inflation. The source benefits from its extensive coverage in providing data on prices, quantities and values of transactions of each model of a good sold. Relative weights can thus be ascribed to price changes in both base and current months at a highly detailed level which allows us to estimate substitution bias. Methods of adjusting for quality changes can be considered. The dummy variable hedonic approach is compared with a superlative, exact hedonic approach and a matching technique akin to that used by statistical offices.