Product churn, tastes, and price indices.

Abstract

Price indices commonly used to measure inflation do not account for the entry and exit of products nor for systemic changes in consumer preferences over time. Recent developments in the theory of price indices show that these issues have significant impacts on the growth of prices that consumers experience. However, executing these developments requires access to household-level data. Analyzing the ready meal purchases of 30,000 UK households during 2013–2022, we evaluate the extent of the bias between standard price indices and those that accommodate product churn as well as changes in consumer tastes. Household data also permit the effects to be calculated across income groups. Our results show that price indices commonly used to measure inflation significantly overstate the extent of price increases due to: (i) “representative” prices used to measure the Consumer Prices Index do not adequately capture the prices of goods in household baskets, and (ii) the effect of the net entry of products and taste changes, the latter being particularly strong. Furthermore, differences in the shopping baskets across income groups suggest that inflation varies across the income distribution. We find that when product churn and taste changes are accounted for, ready meals prices rose over the period by around 3% in low-income households compared with a rise of 12% and 14% in middle- and higher-income households, respectively. These issues are important not only for measuring cost-of-living and how it varies across income groups but also in addressing policy issues more generally.