While health receives increasing attention from the economic profession, one can argue that the empirical analysis of household budgets has been motivated at the outset by an important aspect of the study of health status, connected with nutrient intakes: this is the topic of the seminal study of Engel (1857). One way to model the relation between the economic and nutritional status of households is to posit a causal relationship between these two. The aim of studies using this representation is to evaluate the Engel curve for calories. Unfortunately, the conclusions of empirical studies of this type diverge starkly, mainly for the following reasons, discussed by Bouis and Haddad (1992): (i) the data come from different populations, (ii) different estimation methods are used, and (iii) the variables included in the relation between calories and income differ. In particular, choosing calorie availability as dependent variable, as opposed to calorie intake yields biased results. However, it is difficult to obtain calorie intake data, because their construction is very demanding. Chesher (1997) proposes a model for the study of mean calorie intake as a function of sex and age, using only household purchases, and not the individual consumption. Since household budget data sets are much more common than individual consumption data sets, this make Chesher's approach widely applicable. In this study, we use that approach to study the nutrient intakes of Czechoslovak individuals as a function of age and sex, the evolution of this quantity across time (1989-1992) and the influence of current income and educational level of the wife on these curves. The nutrient we study is the daily energy intake (Kcal/d). We obtain the following substantive results. The age profiles of the energy intake of Czechoslovakians are similar for males and females. The overall energy intake increases with age until 55, is approximately constant between 55 and 70, and decreases afterwards. The contribution of carbohydrates to the energy intake is below the 50% recommended for a balanced diet, whereas the contribution of proteins is above the advisable 12% level. However, these imbalances have markedly decreased over the four-year period studied. The contribution of lipids to the total energy intake matches the advised level of 30%. In households based on a couple, the education level of the woman plays almost no role in the quantity of calories ingested, but a higher education level of the wife leads to a slight moderation of energy intake. The impact of income on the nutrient intakes is not consistently significant over time for proteins, lipids and carbohydrates. However, it is positive for beverages. Overall, the age profile of energy intake seems to follow the general variations of income over the life cycle (increasing during the working life, with income, decreasing for pensioners). Beside this substantive contribution, the paper also provides a methodological contribution, as we adapt Chesher's approach to the particularities of the data at hand in a way that should prove useful to other researchers.
Miquel, Ruth and François Laisney (2000), Consumption and Nutrition: Age-Intake Profiles for Czechoslovakia 1989-1992, ZEW Discussion Paper No. 00-63, Mannheim.