Changes in food expenditures, 1969--73: findings from the retirement history study.

From 1969 to 1973, average expenditures for food at home reported by Retirement History Study respondents increased by almost the same proportion--30 percent--as did the food component of the consumer price index. Changes in these expenditures were not very responsive to changes in income, but income had greater power in explaining total food expenditures. The analysis was based largely on a regression technique that identifies the factors most important in explaining the variation in food expenditures. Size of household was the most important predictor of both the total level of household food expenditures and the per person level. Residence (urban, rural nonfarm, and farm), a proxy variable for home-produced food, was also generally significant. With size of household and income taken into account, a number of socioeconomic variables--including race, sex, age, morale, health, education, and homeownership--were found not significant.