²⁰This point goes to the heart of the debate about the meaning of the conflict between the much larger increases in consumption inequality measured in the CE’s diary survey (which covers a two week period) versus the interview survey (which covers a three month period). One interpretation of the discrepancy is that with the spread of “big box” retailers, households may be making fewer trips to the store but buying more when they do go. It is possible that over some appropriately extended period (like a year) they buy precisely the same amount as before the advent of the big box stores, but the shift to more diary-survey periods with zero expenditures and fewer with larger expenditures would look like an increase in inequality. See Attanasio, Battistin, and Ichimura (2004) for a discussion.