Understanding International Price Differences Using Barcode Data
Underlying price changes of items within indexes vary greatly across locations, even for narrowly defined product categories (for example, eggs). Because more distant locations consume more disparate sets of goods, the increased dispersion in prices seen as the distance between cities grows is largely attributable to compositional effects.
In Understanding International Price Differences Using Barcode Data (NBER Working Paper No. 14017), Christian Broda and David Weinstein use a vast sample of U.S. and Canadian barcode records to challenge the existing evidence on three key results about international price deviations that have been documented in previous work. These results are that borders invariably mean flagrant violations of the law of one price (LOP); that distance matters greatly for understanding these deviations; and that rates of convergence back to purchasing power parity (PPP) are inconsistent with the evidence on nominal price stickiness. The authors claim that most previous studies used data on price indexes and price surveys that limited the ability of researchers to compare identical goods across locations. Using prices for thousands of identical goods that are sold in the United States and in Canada, they find that none of the three stylized facts survive closer scrutiny. Instead, their work supports simple pricing models "where the degree of market segmentation across the border is similar to that within borders."
Finding micro data to use in studying the international pricing of identical goods has long been a challenge, but Broda and Weinstein take advantage of the fact that the United States and Canada use the same barcode system. The researchers work with scanner data from the ACNielsen Homescan database. The uniqueness of these data is that they include price and quantity information for every UPC purchased in a demographically representative sample of households in both Canada and the United States. These goods include most supermarket items, a subsample of durable goods, and electronic products. There are over 500,000 UPCs sold in each country every year. Since the data is available for different cities in the United States and regions in Canada, it allows for the comparison of identical goods and for their prices to vary across locations, both within and across borders. For the first time, this permits precise examination of the border effect. In particular, given that the quantities purchased are also included - a feature absent in the collection of national statistics data -- the role of compositional effects can be studied.
Broda and Weinstein find that the law of one price and purchasing power parity, in their absolute forms, hold about as well across borders as within countries. The researchers confirm that the LOP is violated flagrantly in international data, but also across cities within the same country. Thus, for example, the observation that an identical can of soda sells at different prices in different countries is not very informative about border barriers, because prices vary substantially even within borders.
The authors also find that international violations are not much larger than domestic ones. Broda and Weinstein believe that the aggregate indexes used in the earlier studies collapse the large within-country idiosyncratic variation of relative goods prices, while preserving the variation attributable to exchange rate movements. This has important implications for the interpretation of previous results.
Their analysis reveals only a small border effect with respect to deviations in the law of one price. Broda and Weinstein find that the impact of distance on the price deviations among identical goods is only about one tenth that obtained using price-index data. They show that the set of common goods across cities varies systematically across space and borders. Therefore, unless all individual prices within an index move together, price indexes will appear to deviate across space and borders merely because the underlying weights and goods are different.
Moreover, the analysis using barcode data shows that underlying price changes of items within indexes vary greatly across locations, even for narrowly defined product categories (for example, eggs). Because more distant locations consume more disparate sets of goods, the increased dispersion in prices seen as the distance between cities grows is largely attributable to compositional effects in the set of goods used to compute city-specific price indexes.
Finally, the variation of price adjustment is present both within and across borders. These results imply that the relatively small price differences generated by the typical exchange rate movement will tend to be very persistent, but that larger ones will be short lived. This may have important implications for understanding why prices sometimes seem to respond to exchange rate changes but at other times do not.
-- Matt Nesvisky