Contents - Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Most studies analysing Brazil's school-lunch programme focused on operational efficiency. In other words, their primary objective was to evaluate the process of buying and distributing food. Until now, it has been generally believed that the programme promotes an increase in food consumption at school, assuming that households benefiting from participation do not modify purchases and distribution of food among household members.
To assess the effects of food transferred to the beneficiaries, the following questions were posed: How do the participating families perceive the programme? Does the programme augment food availability for the beneficiaries? Can the programme enhance nutrient consumption among low-income families? Can it act as a powerful instrument to protect poorer households from a deterioration in nutritional status arising from the economic crisis or processes of economic adjustment?
Perspective of the beneficiaries
The lunch programme and school enrolment are strongly interrelated. About 60% of the 253 &mikes claimed that it was an important factor in deciding to send their children to school (table 6). When families not benefiting from the programme are excluded because their children are not in school or are enrolled in one without a lunch programme, the percentage increases to 76%. Nearly 90% of the households whose children attend preschool and receive at least two meals claimed that the programme was important in the decision. Thus the effects are not limited to nutrition, but extend to education as well.
TABLE 6. Importance of the school-lunch programme to the decision of whether to send a child to school by income level (percentages of 253 families)
|Monthly income (MW)||Total|
TABLE 7. Children's opinion of the school meals by income level (percentages)
|Monthly income (MW)||Total|
Another important feature is the children's opinion of the school meals. The sample showed a definite predominance (59%) of children who liked the meals (table 7). Considering only the preschoolers who received more than one meal, the results are even more satisfactory. About 82% said they like the lunch programme, 10% indicated indifference, and only 8% said they did not like it.
As shown in table 7, a strong predominance of lower-income children claim to like the meals. Various factors can contribute to this result. First, it is probable that, because of insufficient food consumption, these children value the meals more than do those whose families have greater purchasing power and who receive a more adequate diet. A second important element is the fact that approximately 36% of the children interviewed belonging to the income group below 2.5 MW did not have breakfast before coming to school.
Statistical analysis of the nutrition effects
The statistical analysis focused on the income elasticity of calorie and protein consumption for lowincome families, with the intention of addressing issues concerning the relevance of nutrition programmes during periods of economic crisis and policy adjustment, and the nutrition impact of the school-lunch programme.
The analysis undertaken used daily per capita calorie and protein consumption as the standard indicator of family nutrition. In calculating this measure, food consumed at home was adjusted by an estimate of the household members' consumption away from home because of the difficulty of obtaining reliable information about food consumed away from home.
First, the average consumption per capita per meal for food consumed at home (Cpch) was calculated according to the following equation:
Cpch = total calories (protein) consumed at home / number of meals consumed at home
Next, the Cpch was multiplied by the total number of meals eaten within and outside the home by all household members in order to obtain the total calorie and protein consumption (Ct) for each household:
Ct = Cpch x total number of meals.
Finally, the measure of per capita consumption (Cpc) for each household was calculated as follows:
Cpc= Ct/household size.
In sum, the procedure adopted to construct the measures of per capita calorie (CprC) and protein (CpcP) consumption used average consumption of calories and protein per meal eaten at home as a proxy for average consumption per meal eaten outside the home.
The income elasticity and the impact of the school-lunch programme on the calorie and protein consumption of schoolchildren was calculated using a set of equations with the following specification:
Cpc= f(Ypc. D1, D2).
where Cpc is the measure of per capita calorie or protein consumption for schoolchildren, Ypc, is income per capita, and D1 and D2 are dummy variables that test the nutrition impact of the standard meal plan (provided to children who remain at school for only one period) and the complete meal plan (provided to children who stay at school for a full two periods).
The relationship between the availability of calories and protein and (1) per capita income and (2) the type of school-lunch programme was estimated with the following functional form by ordinary least-squares regression:
Cpc-i = a+bYpc+cD1+dD2,
where i = C for calories and P for protein.
When the model is specified in logarithmic form. b estimates the income elasticity. If the model is specified in linear form, b estimates the slope, from which the elasticity may be calculated at any point or for any interval of the function. The estimates of the coefficients of the variables D1 and D2 allow the nutrition impact of the standard plan (primary school) and complete plan (preschool) to be evaluated individually.
Inasmuch as the complete plan includes two meals, it is would be logical to assume that the programme administered in preschools would have a larger impact than the standard plan. The initial estimate for the general equation fails to reject the hypothesis that there is no significant difference between the coefficients of variables D1 and D2. This result suggests that the impact on per capita calorie and protein consumption is not significantly different between the two programmes. All else being equal, the number of meals received at school does not significantly affect the per capita availability of calories and protein for schoolchildren.
Given these results, the general equation was respecified into a more restricted form in order to test the joint nutritional impact of standard and complete meal plans. The new specification of the model was formed as follows:
where i = C for calories and P for protein.
The equation was estimated in both logarithmic and linear forms. This last estimation gave better results in terms of the coefficients on the independent variables and r². The results were, for calories:
Cprc = 1,605 + 0.07Y + 357D, r² =
(1.76) (3.5) (2.62)
and for protein:
CpcP = 44.8 + 0.89Y + 8.5D, r² =
(2.14) (2.17) (2.75)
(The numbers in parentheses are t statistics.)
These estimated equations have low coefficients of determination, indicating that the model explains only a small portion of the variation in per capita availability of protein and calories among schoolchildren of the sample households. Similar results have been encountered in the evaluation of effects of nutrition programmes based on large samples [ 12]. A number of factors can help explain these low coefficients of determination.
First, the reduced-form equation of demand used in these studies excludes a series of variables that certainly affect consumption of food, such as physical activity and availability of resources at the family and community levels. Second, the use of available calories and protein as a proxy for consumption introduces a bias in the estimated values associated with the omission of food stocks and losses from diseases and spoilage. Third, these studies are generally based on very short data-collection periods, which can yield sample food purchases that are less than the true purchases over the reference period. For this study, these difficulties were added to the destabilizing macroeconomic forces that most certainly temporally affected the consumption and savings decisions of the households.
Despite the low r² obtained, the estimates of the coefficients were more significant for both income and the school-lunch variable. In the case of income, coefficients that measure the relationship between nutrient availability and per capita income of schoolchildren are positive and significantly different from zero at the 0.1% level for calories and 1% level for protein. The positive correlation is consistent with the low levels of household income. The values estimated for the coefficients, although significant, are small, resulting in an income elasticity of 0.08 for calorie availability and 0.05 for protein availability (assuming that this method of calculating an estimated income elasticity for schoolchildren is a good approximation for the household unit). These results indicate that changes in income are associated with much smaller percentage changes in the availability of calories and protein. This reinforces the hypothesis that the relationship between income and consumption of calories and protein is very weak.
If this is the case, contrary to the emerging strategies of economic adjustment advocated by the IMF and the World Bank, the elimination of poverty may not necessarily be a solution to the problem of malnutrition. Other approaches, such as programmes specifically oriented to increase nutrient consumption, could be equally important.
This argument is reinforced by our statistical analysis. The estimate coefficients that measure the relationship between the receipt of school meals and the availability of per capita calories and protein are positive and significant at the 0.1% level. The estimated values indicate that access to a school-lunch programme is associated with an increased availability of 357 calories and 8.5 g protein per capita. By extension, note that these values signify approximately 15% of the average per capita calorie and protein availability. We can conclude, therefore, that the programme is effective in augmenting the availability of calories and protein among recipients and so can be considered a powerful instrument for protecting poorer target groups from a deterioration in nutrition status caused by economic crisis or by the process of economic adjustment.
1. World Bank. World development report. Washington, DC: World Bank, 1981.
2. Behrman J. Nutrition and incomes: tightly wedded or loosely meshed? PEW/Cornell Lecture Series on Food and Nutrition Policy. Ithaca, NY, USA: Cornell University, 1988.
3. Pitt M. Food preferences and nutrition in rural Bangladesh. Rev Econ Stat 1983;65(1):105-14.
4. Shan D. The effect of price and income changes on food-energy intake in Sri Lanka. Econ Dev Cult Change 1988;36(2):315-40.
5. Mateus A. Morocco: compensatory programs for reducing food subsidies. Washington, DC: World Bank, 1985.
6. Harbert L, Scandizzo P. Food distribution and nutrition intervention: the case of Chile. Washington, DC: World Bank, 1982.
7. Pitt M, Rosenzweig M. Health and nutrient consumption across and within farm households. Rev Econ Stat 1985;67(2):82-95.
8. Bouis H, Haddad L. Comparing calorie-income elasticities using calories derived from reported food purchases and a twenty-four hour recall of food intakes: an application using Philippine data. Washington, DC: International Food Policy Research Institute, 1988.
9. Coltro A. Doutrina agropecuária e alimentar. Brasilia: Ed. Thesaurus, 1988.
10. Homen de Melo FO. Problema alimentar no Brasil. Rio de Janeiro: Editora Paz e Terra, 1983.
11. Campino AC, Cacciamali MC, Cyrillo DC. Evolução do padrão alimentar no Município de São Paulo, 19721982. São Paulo: University of São Paulo, 1984.
12. Kumar S. Impact of subsidized rice on food consumption and nutrition in Kerala. Research Report no. 5. Washington, DC: International Food Policy Research Institute, 1979.
Alderman H. The effects of income and food price changes on the acquisition of food by low-income households. Washington, DC: International Food Policy Research Institute, 1986.
Anderson MA, Austin JE, Wray JD, Zeitlin MF. Supplementary feeding. In: Austin JE, Zeitlin MF, eds. Nutrition-intervention in developing countries: an overview. Cambridge, Mass, USA: Oelgeschlager, Gunn and Hain, 1981:25-48.
Beaton GH, Ghassemi H. Supplementary feeding programs for young children in developing countries. New York: United Nations Children's Fund, 1979.
Campino AC. A review of nutrition programs in Brazil. São Paulo: University of São Paulo-Fundação Instituto de Pesquisas Econômicas, 1987.
Chavez A, Martinez C, Yaschine R. Nutrition, behavioral development and mother-child interaction in young rural children. Fed Proc 1975;34:1574-79.
Fonseca J. Merenda escolar: uma contribuição pare o seu estudo. Doctoral thesis. Faculty of Education, University of São Paulo, São Paulo, 1987.
Fundação Legião Brasileira de Assistência (FLBA). O sistema inferno de informação do programa de complementação alimentar. Rio de Janeiro: FLBA, 1985.
George PS. Public distribution of foodgrains in Kerala: income distribution implications and effectiveness. Research Report no. 7. Washington, DC: International Food Policy Research Institute, 1979.
Kennedy E. Evaluation of the effect of WIC supplemental feeding on birth weight. J Am Diet Assoc 1982;80:220-27.
Musgrove P. Fighting malnutrition: an evaluation of Brazilian food and nutrition programs. World Bank Discussion Paper. Washington, DC: World Bank, 1989.
, Galindo O. Do the poor pay more? retail food prices in Northeast Brazil. Econ Dev Cult Change 1988;37(1):105-14.
Oliveira CG, Medeiros RP. O projeto de abastecimento de alimentos básicos em areas de baixa renda: uma avaliação. Recife: FUNDAJ, and INAN, 1985.
Peliano AM. Fome e desnutrição: as controvérsias da política de alimentação e nutrição. Fundação da Universidade de Brasilia, Curso de Especialização em Política Social. Brasilia: University of Brasilia, 1984.
Pinstrup-Andersen P. Assuring food security and adequate nutrition for the poor. In: Bell D, Reich M, eds. Economic crisis: approaches to policy in the third world. Dover, Mass, USA: Auburn House, 1987.
, Food subsidies in developing countries. Food Policy Statement. Washington, DC: International Food Policy Research Institute, 1988.
Rogers BL. Pakistan ration system: distribution of costs and benefits. In: Pinstrup-Andersen P, ed. Food subsidies in developing countries: costs, benefits, and policy options. Baltimore, Md, USA: Johns Hopkins University Press, 1988:242-52.
Contents - Previous - Next