Contents - Previous - Next

This is the old United Nations University website. Visit the new site at

Health and nutrition considerations in education planning. 2. The cost and effectiveness of school-based interventions


Dean T. Jamison and Joanne Leslie



This paper discusses intervention packages and their costs, and then briefly overviews, from an economic perspective, the strength of the claim of health and nutrition interventions for school-age children on scarce resources. The authors conclude that, given what is known about the probable effect of health and nutrition interventions for learning and attendance, and given the relatively modest cost of a carefully designed, carefully targeted programme, the implication for educational planners is clear: more investment in child health and nutrition will pay off well for education. Cost-benefit analyses suggest that appropriate health and nutrition interventions in the schools are likely to prove to be very high-yield investments.



In the preceding paper [1] we reviewed the main health and nutrition problems likely to be important for school-age children, and summarized available information about linkages between these problems and each of three significant educational concerns: children who are unprepared to begin school at the usual age, the failure of many students to learn adequately while in school, and the unequal participation in schooling by girls. It was clear from that discussion that more research is needed concerning the extent of the adverse impact of health and nutrition conditions on school outcomes. Nonetheless, despite the research gaps, there is already sufficient evidence that poor health, malnutrition, and premature fertility are constraining educational achievement to justify active discussion and development by education planners of ways to reduce the burden of these conditions among school-age children-both those who are in school and those who are not.

The authors are affiliated with the University of California at Los Angeles, Los Angeles, California, USA.

For many of the conditions affecting school-age children, there exist effective, safe, and relatively inexpensive medical interventions, a number of them only recently available. The task of developing systems to deliver these interventions (and establishing the role that schools should play within those systems) is an increasingly pressing one, in part because of the opportunities that the medical technologies offer, and in part because of increased understanding of the educational significance of successful intervention.

In this paper we deal with approaches to intervention. We begin by developing a list of intervention vehicles available to education authorities, then review information available on the cost and effectiveness of these mechanisms, look at the claim of such interventions on scarce resources, and conclude with a summary of the implications of the preceding paper and this one for education planners.


Alternative intervention vehicles

It is important to be clear at the outset that intervention packages should typically not be condition-specific but rather include activities that might easily address more than one condition. For example, it would make no sense to have a team of nutritionists visiting schools annually to administer oral iodine supplements, a team of health workers visiting to dispense albendozole to deal with intestinal helminths, and still a third team dispensing praziquantel for schistosomiasis. Hence we introduce the notion of intervention vehicle (in the example given, the vehicle would be what we call a mobile school health team) that can carry a range of logistically related interventions as appropriate to local epidemiological circumstances. The following subsections describe intervention vehicles relevant to the different classes of educational problems that were discussed in the preceding paper.


Preparing children to enrol

Table 1 describes several intervention vehicles available to education planners and school officials at the local and district as well as national levels to help prepare and encourage children to enter school. The main educational goals of interventions targeted to pre-school-age children would be to prevent the serious disabilities that make it difficult or impossible for children to enrol in regular schools, and to improve the readiness of those who have suffered delayed physical and cognitive development.

Community-based efforts to improve the health and nutrition of pre-school-age children are already a priority activity of the health sectors in most developing countries. (See UNICEF's annual State of the World's Children series for both the main interventions being advocated and success to date in implementing them [2].) The role of education planners would be to lend the weight of their support to such efforts and, in some cases, to advocate the allocation of resources to interventions that are particularly important in terms of later school outcomes [3]. Such interventions would certainly include, for example, addressing problems of micronutrient-deficiency disorders, which is not now a component of the standard child-survival package.

Along with advocacy of community-based efforts to improve the health and nutrition of pre-school children, education planners will want to look closely at both developing and developed countries" experience with early childhood development programmes, which have been found to have modest but positive effects on both enrolment and achievement in school [4; 5]. Although such programmes appear to be relatively expensive, their potential for achieving important health, nutrition, and education goals, as well as their intrinsic appeal in being targeted to young children, means that they may be able to attract support from a range of both public and private agencies.

In addition (or as an alternative) to early childhood development programmes, ministries of education may wish to consider adding a preparatory year before the regular first year of primary school. The objective of such a year would be as much, if not more, to improve the health and nutritional status of children before they attempt regular school activities as to enhance their social and cognitive readiness. In situations where adding a year would not be feasible (for either school-related or household-related reasons), it may be worth considering targeting a fairly intensive health and nutrition intervention to children during their first regular year of primary school.


Facilitating learning and attendance

TABLE 1. Pre-school intervention vehicles

Vehicle, and target age group Nature of interventions
Advocacy of interventions to promote preschool child health and nutrition -ages 0-5 years Priority interventions to be advocated might include:

-growth monitoring with targeted, on-site supplementary feeding;

-prevention of iron-deficiency anaemia through antihelminthic agents, iron fortification of food, iron supplements, and/or ascorbic acid supplements;

-periodic oral doses of vitamin A to prevent blindness, and of iodine to prevent iodine-deficiency disease;

-immunizations, particularly against polio and measles.

Early childhood development programmes -ages 3-5 years Programmes may be undertaken by ministries of education, ministries of health, or non-governmental organizations. They standardly emphasize one or more of the following:

-provision of nutrition and health care,

-early social and cognitive stimulation,

-child care. Most experience to date in developing countries has been with small-scale experimental programmes.

Pre-school preparatory year -ages 4-6 years Experience with this approach has been primarily in industrialized countries, although favourable experiences have been reported from Brazil. It would be somewhat similar to the (usually half-day) pre-kindergarten year that has become fairly common in the United States and a number of other higher-income countries, but would place much greater emphasis on nutritional supplementation, treatment of helminthic infections, and screening for vision and hearing impairment, along with provision of early stimulation and teaching social and communication skills.

TABLE 2. School-based intervention vehicles

Vehicle, and

main conditions

to be addressed

Nature of interventions
School location and



-poisoning and






Special schools or transporta

tion services would allow

greater access to education

by disabled school-age chil


Appropriate maintenance and

location of schools away

from busy streets would re

duce poisoning and injury.

Provision of water supplies and

latrines would facilitate

hygiene and improve con

venience of schools.

Single-sex schools, schools

closer to home, or schools

with better security would

reduce concerns about un

wanted pregnancy and sex

ual violence, and help girls to

remain in school.

Infant-care facilities would en

courage girls' participation

by allowing for care of youn

ger siblings or the student's

own child.

School health


-iron deficiency

-malaria (non-





of disabled

children [6]

-premature fer-

tility (through

provision of


-minor injuries

The school health worker could

be a nurse, school health

master, or teacher desig

nated as responsible for

health issues. The worker's

main role would be to pro

vice contact with and refer

rals to local health and

family-planning services and/

or the mobile school health

team (see below).

Depending on training and

need, the health worker, in

addition to dealing with the

conditions listed at left,

could provide health, nutri

tion, and family-planning

education to students (as

well as to other school staff).

Mobile school

health team


and iodine





The mobile school health

team-under the auspices of

the ministries of education

and/or health-would visit

schools on a regular basis,

probably once or twice a

year, to provide any booster

immunizations needed


TABLE 2-Continued

Vehicle, and

main conditions

to be addressed

Nature of interventions
Mobile school

health team



-dental caries


micronutrient supplements

that can be given in infre

quent periodic doses, such as

iodine and vitamin A, treat

ment for a range of helmin

thic infections, and preven

tive fluoride treatments for

dental caries.

Teams would be available to

provide health, nutrition,

and family-planning educa

tion sessions and to consult

with the school health work

er, students, or parents

about individual student

health needs.

The teams should serve all

school-age children in the

community-not just those

enrolled in school.

School feeding





School meals that are quantita

tively and qualitatively

adequate to significantly re

duce protein-energy mal

nutrition or micronutrient

deficiencies are expensive,

though even nutritionally

less adequate school feeding

programmes have been

found to have a significant

positive effect on attend


Efforts targeted primanly at re

lieving short-term hunger are

less expensive and should

focus on providing breakfast

or a small snack shortly after

students arrive at school.

Health, nutrition

and family-




Students, teachers, and parents

should be explicitly targeted

by school-based information

al and motivational cam

paigns based on local needs.

Appropriate health, nutrition,

and family-planning in

formation should be inte

grated into the regular

school curriculum.

a. Educational efforts might particularly address prevention of conditions that are difficult or impossible to reverse (e.g. HIV infection, pregnancy) and addictions (e.g. to tobacco, alcohol).

Table 2 describes a number of approaches that could be taken to delivering school-based health and nutrition services, or to modifying school structure or policy in order to reduce health and nutrition problems that affect school-age children. Since many school-based intervention vehicles have the potential both to facilitate overall learning and attendance and to improve the participation of girls, the interventions listed in table 2 form the basis of the discussion in both this and the following subsection. Whereas school-age children are the main targets for the interventions discussed, teachers and other school staff are also potential beneficiaries, and their needs should be borne in mind when designing interventions.

Substantial resources are allocated by developing countries to the continuing construction and renovation of school facilities. Therefore, it is well worth while to consider ways in which the location, facilities, and types of schools being built can reduce the negative effects of certain health and nutrition problems on education. Locating schools close to where children live could reduce not only the time cost but also the substantial energy cost associated in some cases with getting to and from school. Children in urban parts of the developing world are exposed to rapidly growing risks of both lead poisoning [7; 8] and injury associated with vehicular traffic. Modest but significant reductions in such risks could be achieved by building schools away from busy streets and by designing smaller, more local schools in both rural and urban areas. Providing clean water and latrines at schools would improve the health of students and teachers.

Many schools in developing countries already have a school health worker, but even more do not. Depending on the availability of trained health personnel and the size of schools, it may be more appropriate to have a teacher with some health and nutrition training rather than a full-time school health worker (or a full time worker may be shared among several schools). In any case, it is potentially important for someone in each school to be assigned to screen for easily recognized health and nutrition problems, to make referrals to local health facilities, and to provide minimum first aid and perhaps some basic medications [9].

An approach with which experience is limited in most parts of the developing world but that often warrants serious consideration may be called the mobile school health team. Such a team may be staffed by regular ministry of health personnel or by personnel from a special health and nutrition unit within the ministry of education. Services that require more clinical expertise than can be provided by a school health worker, or that are effectively carried out periodically (such as vitamin-A and iodine supplementation, treatment of intestinal helminths and of schistosomiasis, preventive dental care, and vision and hearing screening) can be provided by the health team. It is important that all school-age children in the school catchment area should be encouraged to participate in the periodic visits of the team, both to address health or nutrition problems that may be causing absenteeism or non-enrolment and to provide a point of contact between school officials and nonattenders.

Among the intervention vehicles listed in table 2, school feeding has been the most frequently evaluated [7; 10-13]. In general, such evaluations have found that school feeding programmes appear to have a significantly positive effect on school attendance in spite of limited evidence of a positive effect on nutritional status. In addition, several studies from the United States [Meyers et al., cited in 7; 14] and one from Jamaica [15] show feeding children breakfast or a snack early in the school day to have a significant positive effect on school performance. Current evidence suggests that programmes designed to alleviate short-term hunger (e.g. breakfast or snack programmes) are less expensive than more elaborate lunch or multimeal programmes designed to address chronic protein-energy malnutrition and equally or more beneficial in terms of school outcomes [4].

There has also been considerable experience with school-based health, nutrition, and family-planning education, although much of this remains unevaluated in terms of effects on either health or education [16]. Students, teachers, and parents should all be seen as potential targets and potential providers of health, nutrition, and family-planning information. The wide distribution and prestige of teachers and schools in developing countries means that they will often be the most effective vehicles for reaching not only students but also their families with information about health conditions and services [2]. Since almost all school-age children will be of child-bearing age themselves within five to ten years, investments in increasing their knowledge of health, nutrition, and family planning can benefit not only themselves and their siblings but also their children within a relatively short time.


Helping girls remain in school

The relationship between girls' health and nutrition status and their participation and achievement in school is bidirectional. Enhancing girls' health and nutrition status (including reducing their own or their parents' concerns about unwanted pregnancy and sexual violence) can increase their participation and achievement in school. At the same time, more years of schooling and completion of a higher level of schooling offer many important health and nutrition benefits to girls, among which are those of delayed childbearing [17] Although our focus here is primarily on the potential educational benefits to be derived from reducing health and nutrition problems among school-age girls, the health and nutrition benefits, both direct and indirect, are likely to be substantial and must be kept in mind.

*The limited impact of school feeding programmes on protein-energy malnutrition is probably due to the relatively small number of school feeding days and the small ration sizes in most programmes. However. no evaluations of these programmes of which we are aware have considered separately those schoolage children who are forming tissue rapidly (either because they are recovering from moderate to severe malnutrition or because they have entered the hormonally driven pubertal growth spurt), in whom the nutritional impact of school feeding might be more apparent.

There are a number of steps that education planners and school officials can take to address health and nutrition problems that may negatively affect girls' participation or achievement in school. Concerns about sexual safety, for example, can partially be addressed through a combination of constructing schools closer to where girls live, establishing more single-sex schools (particularly at the secondary level), increasing the number of female teachers, and improving security on and around the school grounds [18].

In addition, certain school-based services could be helpful. Sex education, rape-prevention education, family-planning information, dispensing contraceptives, and abortion referrals all need to be considered based on as much open dialogue as possible among teachers, parents, and school officials. The benefits that such interventions offer in terms of reduced absenteeism and dropout due to premature fertility, sexually transmitted diseases, and fears about sexual safety warrant substantial efforts at implementation, in part because most school-age girls have no alternatives [19].

TABLE 3. Costs of selected interventions (illustrative estimates)

Vehicle Cost per targeted child per year


Pre-school pre- paratory year Might range from 40% to 150% of cost of a year of regular primary school. Costs at low end of range assume part-time attendance (half the standard school day, or 2-3 days per week). Costs at high end assume full-time attendance, a daily meal, and perhaps somewhat reduced class size. Costs would obviously be reduced if the "year" were much less than that, perhaps only several months.
Improved school facilities Depends entirely on what is included in package. Access to piped, stored, or well water might cost in the range of $0.25-$2.00 per child per year. Access to functioning latrine might cost $0.10-$0.50 per child per year. School gardens cost perhaps only $250 to start up [22] but are expensive in terms of student and teacher time and energy. Special schools (sex-segregated; schools for the handicapped) may be more costly than regular schools if they require boarding facilities, smaller classes, or special transportation arrangements. Distance education in small local schools can perhaps reduce costs while improving security for females and educational quality [23].
School health worker A substantial commitment of time - e.g. 25% of time of one teacher per 500 students-would add 1%- 2% to annual cost per child. Supplies might cost 50.50-SI.OO per child per year for chloro- quine, a similar amount for iron, and somewhat less for first aid. (Not all schools or children within a school would need iron supplements or chloroquine. ) Main cost would be released time for a teacher to supervise facilities and hygiene, liaise with local health service, con duct health and nutrition education, dispense iron supple ments and chloroquine, and provide first aid. If half of a teacher's time were made available per 500 students and the initial student/teacher ratio were about 40, costs would increase by about 3%-4%; a more modest commit ment of 25% of a teacher's time per 500 students would generate a 1%-2% increase. Alternatively, if no additional staff were hired, an increase in class size from 40 to 42 stu dents would result from the more substantial intervention. Some training costs would also be incurred.
Mobile school health team On the order of $0.50-$2.00 per child per year, depending on number of conditions treated per child. Assuming the team visited 200 schools per year at a cost of $ 10.000 for vehicle and fuel and $10,000 for salaries, the cost would be $100 per school visit or $0.20-$0.40 per child per year for schools in the range of 250-500 students. Costs of medicine, on the basis of annual administration, Would be perhaps $0.25 per affected child per year per con dition for control of iodine deficiency. intestinal helminths, and schistosomiasis.
School feeding Small breakfast snack (300-400 kcal/day) might cost $20-$40 per child per year. A more substantial meal would cost more, but probably not proportionately. Costs of school feeding will generally be high, but will depend Substantially on the mechanisms for food preparation and delivery. An important consideration is the extent to which benefits leak'' to the rest of the child's family, thereby providing an income redistributional effect that should be separately accounted from school feeding cost per se. Cost analyses have rarely been included in available evalua tions of school feeding programmer, but estimates of $12 per beneficiary per year have been reported from Brazil and $0.33 per beneficiary per day. or perhaps $50 per beneficiary per year. from Chile [13]. The Chile programme was excep tional in scope-breakfast snack plus hot lunch for a total of 700 kcal of energy and 20 g of protein per child per day. Berg [24] finds feeding programmes to cost $0.50-$0.80 per 1.000 kcal delivered.
Health, nutri- tion, and family- planning education Something useful could be done in the range of $0. 10-$0.50 per child per year.  

Estimates in US dollars.

School-based nutrition interventions may also help girls to remain in school. The finding that school feeding programmes have a larger positive effect on enrolment rates of girls than of boys is particularly interesting in this regard [10]. A large-scale evaluation of school feeding in India, for example, found that both the presence of such a programme in a district and its intensity (defined as the ratio of beneficiaries to the number of students enrolled) had larger positive effects on the enrolment of girls than on the enrolment of all students [20].

Iodine deficiency and iron deficiency have been shown to affect cognitive function and school achievement, and are also known to increase in females with the onset of puberty. One study from Bolivia found that the relationship between reduction in goitre and improvement in intelligence quotient scores was even more significant among school-age girls than school-age boys [21]. It seems possible, therefore, that not only school feeding but also some micronutrient supplementation could offer particular benefits in terms of girls' participation and achievement in school.


The economics of intervention


The intervention vehicles that have been described typically are capable of addressing more than one of the health and nutrition conditions important for school-age children. The costs of operating them not only will be country specific but obviously will depend on exactly what conditions are addressed. Thus, including specific cost estimates here would be inappropriate. Nonetheless, enough is known about costs, or can be inferred from available experience, to provide a set of illustrative estimates. These estimates are intended only to give a broad sense of the placement of different intervention vehicles relative to one another and for comparison with education interventions of an altogether different character, such as reducing class size.

Table 3 provides these illustrative estimates for the costs of one of the interventions from table 1 (the preschool preparatory year) and for all five of the school-based interventions described in table 2. The table is self-explanatory, but two amplifying points are worth making. First, the costs are expressed in terms of cost per beneficiary per year; clearly, to the extent that the intervention is targeted, its implementation would raise the average cost per student in the system by an amount smaller than the cost per beneficiary. This point, while obvious, is worth stressing because most other schooling interventions (e.g. textbooks) must, by their very nature, be provided to all students. Second, although in principle the costs indicated in table 3 need not be borne by the central government (or by external aid), there will in fact usually be a strong case for non-local assumption of the burden. To ensure high cost-effectiveness, health and nutrition interventions should be targeted to schools where the prevalence of malnutrition and morbidity is highest. These will usually (not always) be poor schools serving poor families in poor communities; local ability to pay may therefore be lacking precisely where the cost-effectiveness of intervention is high. Hence the particular desirability for national (or international) assumption of the financial burden.


Quantitative determinants of learning outcomes

Just as the costs of intervention will be specific to the country or region, so efforts to link interventions to their effects in terms of short-term outcomes or long-term economic and other benefits must, ultimately, be country-specific. Nonetheless, if such efforts are to be successful, we expect they will share an important common element: they will focus far more than is usual in the education literature on quantitative determinants of learning outcomes (e.g., enrolment, attendance, and continuation rates) and relatively less on their effect on learning rate.

While it is well known to education planners, most health professionals would probably be surprised to learn how limited the available information is concerning the quantitative dimensions of the education system in most developing countries [4]. The concept of installed capacity to provide educational services (analogous to hospital bed availability) is used occasionally; however, the concept of capacity utilization as a factor influencing quantitative efficiency has not been of sufficient interest to promote collecting even the most routine data on teacher and student attendance or on physical-plant employment.

One particular notion of quantitative efficiency that has been fairly widely used in the education literature in the past is the ratio of the number of years a school cycle (e.g. primary) should take to the number of student years actually required, because of dropout and repetition, to produce a graduate. This notion is peculiar in failing to define cycles in terms of learning, in neglecting time of exposure, and in treating a cycle partially completed as having no value. An attempt has been made to redress some of these problems by explicitly controlling for learning in measuring quantitative efficiency [25]. Issues of quantitative efficiency and capacity utilization are keys to understanding the economics of health and nutrition interventions; lack of attention to these variables increases the range of uncertainty that must attach to current attempts to assess the cost-effectiveness or rate of return of health and nutrition interventions for children.

FIG. 1. Health interventions and learning outcomes


Cost-effective improvement in learning outcomes

Figure 1 provides a simple structure for thinking about the determinants of learning outcomes in an age cohort, which is what we take to be the proximate purpose of investing in a school system. One can think of learning outcomes as determined from the product (or a more complicated function) of quantity of exposure (participation) and rate of learning. Most interventions directly affect either one or the other of these determining factors: rate of learning, for example, will be influenced by such inputs as learning materials and improved teacher qualifications; participation will be influenced by school location and pricing policy, as well as by selection policies in general and repetition policies in particular. (Intervention to improve the rate of learning can have an indirect effect, as indicated by the relevant arrow in figure 1, on learning outcomes by reducing dropout rates or possibly by increasing attendance.)

The most salient characteristic of health, nutrition, and family-planning interventions is that they operate on both main determinants of learning outcomes- learning rate and participation. Different interventions will, of course, have different relative impact on the two factors: fertility limitation, for example, would be expected to have little impact on rate of learning but a substantial impact on female participation, whereas correction of iodine deficiency might more equally affect both.

Although a quantitative assessment of the cost per unit learning gain of various health and nutrition interventions is not now possible, it is clear that strong learning-rate and attendance effects can be achieved at costs of several dollars per targeted student per year. This places these interventions clearly into the group that have been called "promising avenues" [4]. Further research will allow estimates of cost per unit learning gain of the sort that have been reported for various other interventions to improve learning rate [26].


Benefits from improving school capacity utilization

The empirical evidence summarized in the preceding paper [1] strongly suggests that health and nutrition interventions can significantly affect student participation in schooling, both through enhanced attendance and through reduced dropout rates. A quantitative estimate of the resulting increase in amount of schooling received would allow application of widely used methods in the economics of education to estimate the benefits of such intervention and to weigh them against costs. The total cost of an intervention's increasing the amount of schooling received by the typical student will be the sum of the cost of the intervention, the cost to the student of spending additional well time in school, and the cost to the school system of providing the additional capacity required, if any, to deal with the student's greater participation. (The intervention could increase the child's time in school both by having what was hitherto sick time at home, which we regard as having zero opportunity cost, become well time at school, and by shifting what had been well time at home, which has a real cost, to time at school.)

The benefits of increased participation can, as a first approximation, be estimated in the same way that increases in numbers of years of schooling are estimated-that is, in terms of the increment they entail in earnings (or productivity) for the individual later in life [27]. While such a cost-benefit assessment cannot be carried out with any precision given the available data, approximate values of the key variables can give a sense of the magnitude of the results that might be expected.

It is helpful in analysing costs to consider more explicitly than is typical for education administrators the relation between system capacity and system utilization. Figure 2 illustrates the relation between capacity and utilization in terms of several simple concepts. The number of student places needed to serve a catchment area is simply the population size in the relevant age cohort, here ages 5-14, times the percentage of that cohort that would ideally be in school (for this age group it would be 100%). (Provision for over-age students would shift the line for places needed to the right in the figure.) Each student place requires (given local policies on class size) a certain amount of physical plant, teacher availability, and so forth; the level of those resources actually available determines the number of student places installed. Enrolment will typically be less, perhaps substantially less, than the number of places installed due to dropouts and repetitions leaving upper-grade classrooms only partially filled; similarly, student absences on any given day will leave classrooms only partially utilized. The lightly shaded area in figure 2 indicates system capacity in student-years per year; the darker area indicates system utilization in the same units. The ratio of utilization to capacity provides, we suggest, the appropriate measure of the quantitative efficiency of a school system. The concept of qualitative efficiency, which relates the rate of learning to the expenditure per student-year of actual attendance, is more frequently discussed in the literature [26]. A major impact of health and nutrition intervention will be to increase, albeit at some cost, the quantitative efficiency of school systems. Although the effects of these interventions on learning rate (and thus on qualitative efficiency) may also be significant, we expect that the most economically important impact will be through effects on quantitative efficiency. (The links between intervention and increased enrolment are not discussed here, but are indicated in figure 1 and have been thoroughly developed statistically in another context [28] Further modelling along these lines may be found elsewhere [25].)

FIG. 2. Utilization of school system capacity

The discussion around figure 2 brings us to the point of being able to discuss costs more fully. In terms of the diagram, there are essentially three marginal-cost regimes. Increasing average daily attendance (ADA) by reducing student illness will increase system utilization (and quantitative efficiency) at virtually zero marginal cost in terms of student time, student places on learning materials Increasing enrolment up to installed capacity has a marginal cost consisting of the cost of student time plus that of learning materials, but not including the major teacher time and construction costs associated with installing places. Finally, increasing enrolment beyond the current number of installed places requires full coverage of average costs.

The capacity to increase the quantitative output of school systems at low marginal cost depends, obviously, on current levels of quantitative efficiency. Unfortunately, as a general rule, data are not presented in a way that make it easy to assess quantitative efficiency; a reasonable guess would be that numbers in the range 0.5 to 0.7 would be common. In Nicaragua, for example, about 10 years ago students attended on average 14 out of 20 days per month, for an ADA of 0.7; class sizes in the sixth grade of about 60% of those of first grade suggest a ratio of enrolment to student places of about 0.8, for a quantitative efficiency of 0.56 [28]. Obviously, only part of this quantitative inefficiency results from poor health and malnutrition of students or of teachers, but a quarter or a third of it could easily be due to these factors and thus be potentially remediable through health and nutrition interventions.

An example using reasonable estimates may provide a useful way to illustrate the importance of the potential effects of health and nutrition interventions on quantitative efficiency of schools. Assume, for example, a lower-middle-income developing country with an ADA of 0.7, a ratio of enrolments to student places installed of 0.8, average annual earnings among the economically active of $1,50O, and per student expenditures on elementary education of $250. Assume that the intervention is a relatively expensive one, school feeding, that costs $25 per student per year and that has the effect of increasing ADA from 0.7 to 0.75 and the enrolment ratio from 0.80 to 0.85, thereby producing approximately 0.11 additional school years of system output per beneficiary-year of intervention. Given that an additional year of education might reasonably be associated with an 8% increase in annual earnings, this school feeding investment of $25 might well generate an annual stream of future productivity increase of as much as $13 (0.08 x 0.11 x $1,500). Even assuming that this stream does not begin for ten years or so, and discounting back at a high discount rate, benefits will far exceed costs. It has been argued that school feeding programmes that provide take-home food could be attractive, despite their high cost, precisely because of the high rate of return to drawing children into school [29].

This example is hypothetical, particularly in its assumptions about the quantitative impact of school feeding on ADA and enrolment. Yet the figures are well within plausible ranges, and any additional benefits associated with improving the learning rate have been entirely ignored. In large part because of the possible impact on quantitative efficiency, the potential exists for benefits greatly to exceed costs for health and nutrition interventions in the schools. The mechanisms for assessing this effect empirically will require more attention to data on enrolment, attendance, and dropout rates than is typical, but these data are relatively easily obtainable, and evaluations should soon provide sharper case studies on which to base benefit-cost generalizations.


Implications for education planners

By the year 2000 there will be over 1.4 billion school-age children in the world, approximately 87% of whom will be living in developing countries. Our definition of school age encompasses the age range from 5 through 14 years and corresponds approximately to the school period from kindergarten through lower secondary school; it begins after the period of high mortality risk in the pre-school years and continues through the adolescent growth spurt and sexual maturation to the beginning of adulthood. Children in this age group in developing countries are 14 times more likely to die between their fifth and fifteenth birthdays than are their age-mates in the industrialized market-economy countries; it is likely, but not empirically demonstrated, that they suffer an even greater relative burden of morbidity and malnutrition. Illness and its consequences are ever present in most of these children's lives.

Attending school is (or should be) the principal occupation of these children. The main adverse consequence of their ill health, then, beyond immediate welfare losses and the cost of care, is in learning opportunities forgone. Conversely, an important potential instrument for affecting their health and nutrition status will be school-based policies, particularly since the health problems of school-age children are often relatively neglected by the formal health services.

In the preceding paper [1], drawing substantially on a recent review of the literature prepared for Unesco [7], we summarized what is known about the implications of ill heath for aggravating three of the most significant problems currently facing education planners in developing countries: children who are unprepared to begin school at the usual age for school entry, the failure of students to learn adequately during their school-age years, and unequal participation in schooling by girls. The implications prove to be substantial. SO, too, is the range of relatively effective yet low-cost interventions that school systems could implement to prevent or ameliorate the relevant health problems.

The analysis-reported here clusters the options open to education planners into a number of "intervention vehicles," each constituting a set of logistically and programmatically related activities capable of carrying interventions that might address a number of diverse conditions. Several of the interventions are directed principally toward pre-school-age children to better prepare them for entering school. The main alternatives for this purpose are:

Interventions directed to school-age children can be reasonably carried by five vehicles. For many of them, even though the school may be the focus for delivery, the target group will be the entire age cohort in the community and, where relevant, the school's teachers and staff. (For some interventions, however, such as school feeding, restricting provision of the service to those actually in attendance on a given day is important to the objective of motivating attendance.) Vehicles oriented directly to the school-age group are:

In many environments, establishing a programme for mobile school health teams will be a particular priority, in part because of the high prevalence and educational importance of the conditions they can address, and in part because of the low cost and high efficacy of available interventions. School health workers, health and nutrition education, and selected school improvements will also often appear attractive. For several dollars per affected student per year, much can be done. School feeding programmes must be considered somewhat separately, as the costs for such efforts are much higher-US$10 to US$50 per student per year. Nonetheless, school feeding may be a very good use of food aid, and, even if there is no food aid, appropriate targeting could make feeding programmes into useful tools for poverty alleviation, objectives that go beyond improving school attendance and learning.

In summary, given what is known about the probable effect of health and nutrition interventions for learning and attendance, and given the relatively modest cost of a carefully designed, carefully targeted programme, the implication for education planners is clear: More investment in child health and nutrition will pay off well for education. Illustrative cost-benefit analyses suggest that the mechanism for realizing this payoff may lie principally in enrolment and attendance effects, with their concomitant implications for fuller and more efficient utilization of existing school system capacity. Under a broad range of assumptions, these cost-benefit analyses suggest that appropriate health and nutrition interventions in the schools are likely to prove to be very high-yield investments.



The United Nations Administrative Committee on Co-ordination, Sub-committee on Nutrition (ACC/ SCN), provided financial support for the preparation of this paper.

The authors are indebted to Alan Berg, Marlaine Lockheed, John Nkinyangi, François Orivel, Ernesto Pollitt, and Susan Van der Vynckt for valuable discussions concerning the substance of the paper.



1. Leslie J, Jamison DT. Health and nutrition considerations in education planning: 1. Educational consequences of health problems among school-age children. Food Nutr Bull 1990;12(3):191-203.

2. UNICEF. The state of the world's children. New York: UNICEF, 1988.

3. McGuire JS, Austin JE. Beyond survival: children's growth for national development. New York: UNICEF, 1987.

4. Lockheed ME, Verspoor A. Improving primary education in developing countries: a review of policy options. Washington, DC: World Bank, Education and Employment Division, 1990.

5. Myers R. Preparing children for schools and schools for 20. children. Coordinator's notebook, no. 8. New York: Consultative Group on Early Childhood Care and Development, 1989.

6. World Health Organization. Guide for schoolteachers: training in the community for people with disabilities. Geneva: WHO, 1989.

7. Pollitt E. Malnutrition and infection in the classroom. Paris: UNESCO, 1990.

8. Davis JM, Svendsgaard DJ. Low-level lead exposure and child development. Nature 1987;329:297-300.

9. UNICEF. Essential drugs price list. New York, 23. Copenhagen: UNICEF, 1989.

10. Bureau for Food for Peace and Voluntary Assistance, USAID. P.L. 480 Title II evaluations, 19801985: the lessons of experience. Washington, DC: USAID, 1985.

11. Select Committee on Hunger, US House of Representatives. Alleviating world hunger: literacy and school feeding programs. Washington, DC: US Government Printing Office, 1987.

12. Levinger B. School feeding programmes: myth and potential. Prospects 1984;14(3):369-76.

13. Pfefferman GP, Griffin CC. Nutrition and health programs in Latin America. Washington, DC: World Bank, 27. 1989.

14. Parker L. The relationship between nutrition and learning: a school employee's guide to information and action. Washington, DC: National Education Association of the United States, 1989.

15. Powell C, Grantham-McGregor S, Elston M. An evaluation of giving the Jamaican government school meal to a class of children. Hum Nutr Clin Nutr 1983; 37C:381-88.

16. Favin M, Cebula D, Saidi R, Pryor D. Health education. Washington, DC: American Public Health Association, 1986.

17. National Research Council. Contraception and reproduction: health consequences for women and children in the developing world. Washington, DC: National Academy Press, 1989.

18. Lee VE, Lockheed ME. The effects of single sex schooling on student achievement and attitude in Nigeria. Comp Educ Rev 1990 (in press).

19. Germain A. Reproductive health and dignity: choices by third world women. Technical background paper prepared for the International Conference on Better Health for Women and Children through Family Planning, Nairobi, Kenya, Oct 1987.

20. Gupta MC, Hom K. Evaluation of P.L. 480 Title II school feeding program in India. Washington, DC: USAID, 1984.

21. Bautista A, Barker PA, Dunn JT, Sanchez M, Kaiser DL. The effects of oral iodized oil on intelligence, thyroid status, and somatic growth in school-age children from an area of endemic goiter. Am J Clin Nutr 1982;35(1):127-34.

22. Berg A. Malnutrition in Zimbabwe. Washington, DC: World Bank, Population and Human Resources Department, 1982.

23. World Bank. Education in sub-Saharan Africa: policies for adjustment, revitalization, and expansion. Washington, DC: World Bank, 1988.

24. Berg A. Malnutrition: what can be done? Iessons from World Bank experience. Washington, DC: World Bank, 1987.

25. Harbison RW, Hanushek EA. Educational performance of the poor. Washington, DC: World Bank, 1990 (draft).

26. Lockheed M, Hanushek E. Improving educational efficiency in developing countries: what do we know? Compare 1988;18(1):21-38.

27. Psacharopoulos G, Woodhall M. Education for development: an analysis of investment choices. Oxford, UK: Oxford University Press, 1985.

28. Jamison DT. Radio education and student repetition in Nicaragua. In: Suppes P, Searle B, Friend I, eds. The radio mathematics project: Nicaragua, 1916-77. Stanford, Calif, USA: Stanford University, Institute for Mathematical Studies in the Social Sciences, 1978:199221.

29. Schuh GE. Food aid and human capital formation. In: Nelson GO et al., eds. Food aid and development. New York: Agricultural Development Council, 1981:49-60.

Contents - Previous - Next