DOCUMENT RESUME ED 296 785 PS 017 420 AUTHOR Levinger, Beryl School Feeding Programs in Developing Countries: An TITLE Analysis of Actual and Potential Impact. AID Evaluation Special Study No. 30. Agency fcr International Development (IDCA), INSTITUTION Washington, DC. Bureau for Food and Voluntary Assistance. REPORT NO PN-AAL-060 PUB DATE Jan 86 125p.; This report was prepared in collaboration with NOTE the Horace Mann-Lincoln Institute of Teachers College, Columbia University. PUB TYPE Information Analyses (070) EDRS PRICE MF01/PC05 Plus Postage. DESCRIPTORS *Academic Achievement; *Attendance Patterns; Developed Nations; *Developing Nations; Enrollment; *Food; Foreign Countries; Nutrition; Program Design; Program Effectiveness; *Research Design; Research Methodology; Research Needs; *School Activities; School Involvement *Food Distribution Programs IDENTIFIERS ABSTRACT Reviewed are empirical studies of relationships among school feeding programs (SFPs), school attendance, enrollment, cognitive development, and academic performance. Recommendations for SFP design are specified. An agenda for a systematic operations research project on the influence of SFPs on these variabla and on nutritional status is described. A brief introductory section discusses the background and assumptions of the study. Section 2, which focuses on school attendance and enrollment, reviews retrospective analysis, comparative, and noncomparative studies, as well as studies that identify determinants of school attendance and enrollment, and other relevant studies. Section 3 reviews studies on the relationship between diet and cognitive development, and between SFP participation and cognitive development in developing and industrialized countries. Studies on short- and long-term behavioral effects are also covered. Each study is reviewed in terms of its findings and methodology for the purpose of identifying implications for the design of SFPs and for future research. Section 4 identifies approaches that maximize the educational impact of Agency for International Development-supported school feeding programs. Section 5 outlines an operations research project on the impact of SFPs. (RH) *********************************************************************** Reproductions supplied by EDRS are the best that can be made from the original document. **************************P********************t*********************** EI 4 9 6 a , . 't +WV'S, .1A r SCHOOL FEEDING PROGRAMS IN DEVELOPING COUNTRIES: AN ANALYSIS OF ACTUAL AND POTENTIAL IMPACT AID EVALUATION SPECIAL STUDY NO. 30 by Beryl Levinger, Ph.D. U.S. Agency for International Development January 1986 This report was prepared in collaboration with the Horace Mann-Lincoln Institute of Teachers College, Columbia University under a contract with the Office of Evaluation, Bureau for Food The and Voluntary Aid, the Agency for International Development. views and interpretations expressed in this report are those of the author and should not be attributed to the Agency for International Development. 3 TABLE OF CONTENTS Page v Acknowledgments vii Summary 1 Introduction 1. School Attendance and A Review of the Literature: 2. Enrollment 6 Retrospective Analysis Studies 8 2.1 15 Comparative Studies 2.2 32 Noncomparative Studies 2.3 Determinants of School Attendance and Enrollment 2.4 35 Studies The United States 38 Other Relevant Studies: 2,5 40 Conclusions 2.6 l .terature on Cognitive Development A Review of the 3. 42 and School Achievement Studies on the Relationship Between Diet and 3.1 43 Cognitive Development Studies on the Relationship Between SFP Partici- 3.2 pation and Cognitive Development in Developing 57 Countries Studies on the Relationship Between SFP Partici- 3.3 pation and Cognitive Development in 66 Industrialized Nations 66 Studies on Short-Term Behavioral Effects..,. 3.3.1 71 Studies on Long-Term Behavioral Effects 3.3.2 76 Conclusions 3.4 78 Desigh Implications 4. 78 Introduction 4.1 79 SFPs and Enhanced Internal Efficiency 4.2 83 SFPs and Enhanced External Efficiency 4.3 85 Conclusions 4.4 86 Conclusions and Next Steps 5. Bibliography -v- ACKNOWLEDGMENTS Many people helped in the preparation of this study. To thank them all would add considerably to the weight of an However, there are a few individuals whose already hefty work. contributions are so significant that it is impossible to omit mention of them without misrepresenting the origins of this report. The idea for the study originated with Judith Gilmore and Hope Sukin, both of the evaluation office in the Agency for International Development's (AID's) Bureau for Food and Voluntary They identified much of the literature that is re- Assistance. viewed in the body of the text, questioned and probed as needed so that I could sharpen my insights, and were marvelously suppor- tive throughout the project. This study was undertaken in collaboration with the Horace Mann-Lincoln Institute of Teachers College, Columbia University, where I was employed as senior research associate at the time of A special thank-you is due to Gary Bridge, Director its writing. of the Institute, for enabling me to incorporate the project into my other responsibilities at the Institute and for his guidance in analyzing some of the cognitive development and academic Ron Lane typed and achievement issues discussed in the text. edited the manuscript with superb skill, despite the many demands Diane Brede, an Insti- made by other projects at the Institute. tute research associate, rendered invaluable service by preparing the matrices that appear as appendixes to the study and by assembling the bibliography. Three other Teachers College colleagues were also most Ann Boehm and Mary Alice generous with their time and insights. White, members of the Psychology Department, were instrumental in helping me to identify procedures that could be used to detect changes in cognitive functioning that might result from partici- Joan Gussow, a nutrition pation in school feeding programs. educator, helped elucidate the linkages among malnutrition, poverty, and school performance. Last, but not least, are the contributions made by staff members at CARE, Catholic Relief Services, and Checchi and Without their help it would have been impossible for Company. me to gain access to many of the documents so critical tc an adequate review of school feeding programs. While help has come from many sources, the point of view expressed and all recommendations and conclusions are my own. Although this study was produced under a contract with AID's -vi- Bureau of Food and Voluntary Assistance, Office of Evaluati I received full latitude to praise or damn school feeding programs as the evidence and my inferences would permit. I In the end I found neither unfettered praise nor utter damnation to be appropriate. ; The resultant work, therefore, seeks more to explain and interpret ambiguities than to offer a definitive judgment on the program's worth. Beryl Levinger, Ph.D. Assistant Executive Director CARE SUMMARY Three objectives are commonly associated with school feeding (1) to increase school enrollment and atten- programs (SFPs): dance among school-age children; (2) to improve the nutritional status of children in school; and (3) to improve the cognitive This study examines or academic performance of these children. and assesses empirical evidence regarding the hypothesized rela- tionships among SFPs, school attendance, enrollment, cognitive It also uses research development, and acadeEIC15Tformance. Finally, the findings to derive SFP design recommendations. study proposes an agenda for a needed operations research project on how SFPs influence attendance, enrollment, cognitive develop- ment, academic performance, and nutritional status. Four types of studies have been conducted to assess the retrospective impact of SFPs on attendance and enrollment: analyses, comparative studies, noncomparative studies, and studies examining the determinants of school attendance and To date, retrospective analyses (of which three are enrollment. reviewed) have not yielded results in which decision-makers can Most fail to use enrollment ratios based on have confidence. solid demographic data, lack data on contextual variables that might influence school attendance, and do not report longitudinal Because of the inherent weaknesses in this type of changes. study and the inconclusive nature of their findings, they do not lend support for the hypothesized relationships among SFPs, attendance, and enrollment. Six studies are reviewed that examine the impact of SFPs by comparing data on attendance and enrollment between SFP and non- The evidence suggests, Most were inconclusive. SFP schools. however, that SFPs may be most effective in meeting their attend- ance-related objective in settings where attendance is not already high and where children come from rural, relatively low Several of the studies also point to socioeconomic backgrounds. the need for program regularity to achieve an impact on chil- These findings suggest the impor- dzen's school-going habits. tance of targeting practices that take into consideration both need and the probability that program regularity can be maintained. Eight studies covering eleven different countries examine the impact of SFPs on attendance and enrollment using primarily Most of these failed impressionistic data drawn from teachers. Significantly, seven of the eight to provide control groups. noted a positive programmatic impact on attendance and enroll- However, their ment, whereas only one drew mixed conclusions. methodological imprecision makes these findings suspect. Two studies examined how nutritional status influences school attendance and enrollment without specifically assessing the impact of SFPs. In one study (Guatemala), the researchers concluded that when economic and family backgroind factors were held constant, size (a proxy for nutritional status) and health of children acted as independent, positive determinants of both attendance and performance. This study suggests that where SFPs can be designed to have an impact on nutritional status, impacts on attendance and performance will also be achieved. Thus, proper targeting and_ the provision of an adequate ration become design issues related not only to changes in nutritional status, but to attendance and performance outcomes as well. The second study (Terai, Nepal) reported similar findings; nutritional status contributed positively and significantly to the probability of a child's being enrolled in school. This study also found that the influence of nutritional status variables on enrollment may differ between boys and girls. The general conclusion one can derive with respect to SFPs, enrollment, and attendance is that feeding programs seem to make a difference when there is a good fit between SFP design and the environment in which the program operates. Judging from the literature, however, this fit is sometimes not present or the evidence to support it is inconclusive. SFP impact appears to be a function of program ecology; nevertheless, crisp guidelines and incontrovertible findings are not available to aid decision- makers in formulating policy for a variety of ecological settings. This report also examines SFP impact on cognitive develop- ment and academic performance in school by reviewing three types of studies: (1) those that are concerned with the relationship between diet and cognitive development in general; (2) those that examine the relationship between SFP participation and cognitive development in developing countries; and (3) those that analyze the relationship between SFP participation and cognitive develop- ment in industrialized nations. Cognitive function may be defined as the ability to learn categories, to process and structure information, and to learn and react to social and environmental cues. Mild to moderate malnutrition, although probably not a cause of primary learning deficits, does appear to alter processes associated with cogni- tive function. Passivity, apathy, shortened attention span, reduced short-term memory, failure to acclimate to repetitive stimuli, and a lug in the development of sensory-integrative capacity are all associated with mild to moderate malnutrition. These dysfunctions prevent children from taking maximum advantage of their learning environments. Thus, children with protein- cogni- caloric malnutrition tend to function at reduced levels of One study, for exam- tive development and academic achievement. significant ple, showed that current diet was the single most predictor of classroom achievement. Mild-to-moderate malnutrition acts synergistically with The risks for a malnourished social and environmental factors. child, living in a culture of poverty, are multiple, inter- However, both human and animal studies active, and cumulative. show that a developmentally facilitative environment can allev- The iate the potentially harmful consequences of malnutrition. remediation when consequences are reversible and susceptible to conducive the child's environment is manipulated to make it more Although improvement in the child's diet to cognitive growth. alone can lead to cognitive changes, greater intellectual growth growth and social can be achieved when the child's psychological These findings suggest that SFPs environment also are enriched. designed as can reach their full potential only when they are developmental part of a broader intervention strategy to address lags or deficiencies in students. Four studies dealing specifically with the impact of SFPs achievement in developing on cognitive development and academic The failure to report countries are reviewed in this report. The other two baseline data renders two of them inconclusive. influence on suggest that factors exogenous to SFPs exert as much Despite this, none of school performance as do feeding programs. feature that the SFPs reviewed incorporated into its design any A need factors. might mitigate the impact of these "intervening" remedy exists to recast the SFP as a more integrated effort to malnutrition, deficits caused by the interaction among acute hunger, and a developmentally nonfacilitative home environment. notwithstanding, the The necessity of an integrated approach and importance of an SFP's impact on the alleviation of hunger be underestim- the improvement of nutritional status should not Cotten, in his analysis of an SFP in Haiti, found, for ated. be example, that 7 percent of the variance in IQ scores could He also found that where the quality explained by malnutrition. important to of education opportunity was low, it was especially place.1 alleviate hunger for student learning to take Studies on SFPs in this country have tended to look at Six studies on either short or long-term behavioral effects. Most emphasize short-term effects are reviewed in this report. 1Joel Cotten, "Evaluation Research on the PL 480 Title II USAID/Haiti, School Feeding Program in Haiti" (Port-au-Prince: February, 1982). -x- morning feedings and the effects of hunger on classroom behav- iors; as a group they yield conflicting results. It should be noted, however, that subjects were not necessarily malnourished. The studies do suggest, however, that the provision of breakfast may benefit students emotionally and enhance their capacity to work on school-type tasks. Long-term behavioral effects of SFPs were evaluated in five studies. They fail to demonstrate conclusively significaht rela- tionships between feeding and school performance. All are marred by serious methodological shortcomings. Section 4 of this report examines how SFPs can contribute to the improvement of a school system's internal and external efficiency. Particular attention is given to how feeding can become a springboard for cognitively oriented interventions that will allow the SFP to reach its full potential as an intervention strategy designed to have impact on both attendance and academic performance. The study's concluding section identifies three issues that need to be addressed syec.ematically through an operations re- search project: (1) What kinds of changes do SFPs promote and for whom? (2) To what extent are those changes interdependent? (3) Given a particular set of ecological conditions, what is the ideal SFP design to promote improvements in enrollment, atten- dance, and academic achievement? The methodology proposed con- sists of seven different treatment types (snack only; breakfast only; lunch only; and each of the above meals combined with a cognitive intervention) applied in each of four markedly differ- ent ecological settings. One country is recommended as the research site and a 3- to 5-year study duration is proposed. Such a project is needed if the real and potential impact of SFPs is ever to be assessed. J0