International Price Comparison using Scanner Data

Cross-country price indices are crucial to compare living standards between countries. An accurate measurement of these price indices has proven to be an extremely difficult task because the consumption patterns of different countries do not overlap. We construct a unique data on prices and quantities at the barcode-level across two countries with different income level, United States and Mexico. At the aggregate level, we identify heterogeneity in quality and variety as important sources of bias in international price comparisons. In addition, we compute income- and country-specific price indices using a non-homothetic preference structure. We find that the price indices crucially depend on both the income distribution of each country and on the local producers' choice of quality. Our results indicate that previous studies importantly understate real income inequality between countries.