Household scanner data contain rich information on household demographics and transactions in actual markets over a long time period. To more fully understand the characteristics of these data, we conducted an analysis to determine whether household expenditures in the Nielsen Homescan panel are similar to the Bureau of Labor Statistic's Consumer Expenditure Diary Survey. We found that many differences in reported expenditures across the two datasets can be explained by such household demographics as female head, income, and household size, for example. The largest degrees of discrepancies across datasets occur for food categories containing more random-weight foods without universal product codes.
