A Straightforward Approach to Derive Individual Consumption by Age from Household Expenditure Survey Data, Using Adult Equivalent Scales

摘要

A common method to derive individual food consumption by age in conducting cohort analysis is to divide household data classified by age of household head by the number of persons included in the household. This practice severely underestimates individual consumption by young parents, particularly in their 20s and 30s, since young children consume substantially less than their parents. We propose an alternative, adult equivalent scales (AES) approach to temper possible biases in estimation of individual consumption by age, and compare it to simple division by the number of household members, and to previous efforts to derive individual consumption using more sophisticated and data-intensive methods. In these comparisons, estimates of age/cohort effects in consumption of selected food products using AES are reasonably close to estimates using more sophisticated, but difficult, methods, while correcting the bias introduced by the simple division method. We illustrate the practical difference that the choice of method makes by projecting consumption levels using the various approaches.