Sales and the (Mis) Measurement of Price Level Fluctuations

Abstract

Using a large set of scanner data from grocery and drug stores located throughout the US, I show that temporary price reductions (“sales”) have a large impact on average price paid. Even after aggregating across goods and markets, the month-to-month change in average price paid is four times as volatile as that of the “regular” price because of changes in sale frequency, depth, and quantity response. The impact of sales is positively correlated with the local unemployment rate during recessions. CPI item indexes do not fully account for variation in sale activity and since these indexes are used to compute real consumption expenditure, it will contain measurement error that is correlated with macroeconomic shocks.