Should We Test The Life Cycle-Permanent Income Hypothesis With Food Consumption Data?

Most studies of household intertemporal consumption behavior use food consumption data from the Panel Study of Income Dynamics (PSID). A natural question is whether the intertemporal behavior of food consumption is typical of overall consumption. I address this question by examining the time-series behavior of food and other consumption series at the aggregate level. Unfortunately, I find that food is not typical: the life cycle-permanent income hypothesis can be easily rejected for most aggregate consumption series, but not for food. This suggests that tests of the LCH/PIH using PSID data are likely to have low powe