Academic journal article The Journal of Consumer Affairs

Cohort Effects of Household Expenditures on Food Away from Home

Academic journal article The Journal of Consumer Affairs

Cohort Effects of Household Expenditures on Food Away from Home

Article excerpt

Data from 23 years of the U.S. Consumer Expenditure Survey (1982-2004) are analyzed to investigate cohort effects on food away-from-home (FAFH) expenditures using the age, period, and cohort (APC) model. Nine 10-year interval cohorts are included, from the Interbellum Generation born between 1896 and 1905 to the MTV Generation born between 1976 and 1985. Analyses reveal that later-born cohorts spend more on FAFH, both in dollar amount and in food budget share, compared with earlier-born cohorts. Significant cohort differences in FAFH remain after additional sociodemographic and economic variables are controlled.


A variety of factors have been identified as playing a role in increased rates of overweight and obesity in the United States, including behavioral, environmental, and genetic factors. The interplay of these factors has contributed to increased energy imbalance among a large percentage of the U.S. population, in that their calorie intakes are in excess of their calorie expenditures (Centers for Disease Control and Prevention 2009).

In the context of calorie intake, past research has found a positive association between food away-from-home (FAFH) consumption and increases in bodyweight (Binkley, Eales, and Jekanowski 2000; Pereira et al. 2005). Parallel to the increase in overweight and obesity rates, FAFH in the United States has increased significantly in the past several decades (Lin, Frazao, and Guthrie 1999a,b; Stewart et al. 2004; Stewart, Blisard, and Jolliffe 2006). Among the factors that are correlated with FAFH consumption and expenditures, age has been found to have a negative association with expenditures on FAFH in cross-sectional studies (Nayga and Capps 1994; Dong et al. 2000; Binkley 2006; Fan et al. 2007). This relationship is subject to two explanations: a lifecycle explanation and a cohort explanation, which are likely to coexist. The lifecycle explanation, also referred to as an age effect, suggests that people's FAFH behavior changes as a result of their lifecycle stages. They eat out more during young adulthood but reduce their FAFH as they get older. The cohort explanation, however, suggests that the age differences in FAFH expenditures can be a result of distinct food preferences among cohorts, in that later-born cohorts have more of a preference for eating out than earlier-born cohorts. As a result, younger cohorts have higher consumption and expenditures on FAFH at any stage of the lifecycle, compared with older cohorts. If such cohort effects in FAFH are consistent and persistent, further increases in FAFH by successive generations of the U.S. population would be projected, unless effective countermeasures are developed and implemented.

Although cohort effects have been studied in various contexts, including the prevalence of overweight and obesity in the United States where cohort effects have been found (Utz 2005), its application in FAFH has been limited. One exception is Blisard's study in 2001. Using data from 1982 to 1995, Consumer Expenditure Survey-Diary Survey (CEX-DS), Blisard (2001) decomposed food expenditures into age, period, and cohort (APC) effects against eight cohort groups, defined over five-year intervals with the first cohort born between 1921 and 1925 and the last cohort born between 1952 and 1956. It was found that later-born cohorts spent less on food at home (FH) compared with earlier-born cohorts on a per capita basis. However, no cohort effects on per capita FAFH expenditure were found in his study.

Since Blisard's study, more years of CEX-DS data have become available. The main purpose of our study is to investigate recent data covering a longer period to ascertain whether Blisard's previous findings are still supported. By having more years of the CEX-DS survey, not only do we have more data on the same cohorts Blisard studied, but we are also able to study additional later-born cohorts. Our study also differentiates from Blisard's study in several other important ways. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.