Journal Issue

Improving Primary Schools in Developing Countries

International Monetary Fund. External Relations Dept.
Published Date:
January 1990
  • ShareShare
Show Summary Details

Investments that accelerate learning

Ninety percent of the nearly three million primary schools in the world are in low- and middle-income developing countries; in these schools some 480 million children struggle to learn. Unlike their peers in developed countries, who attend modern, well-equipped, well-organized, and well-funded primary schools, students in poorer countries face great odds. They often attend classes in poorly constructed and equipped schools that may operate only 500 hours a year (900 hours in the developed world). This schooling is supported by less than $2 of educational materials per child ($52 in developed countries), a poorly designed curriculum, and a teacher with only ten years of education (16 years in developed countries). Classes often have more than 50 children (compared with 20 in developed countries), many of whom are chronically undernourished, parasite ridden, and often hungry.

As a result, schools in developing countries are less productive than those in developed countries. Only about one half of all school-age children in developing countries acquire a complete primary education, with nearly half of even those who are enrolled dropping out before the end of the primary school cycle in low-income developing countries (see chart). Even those who do complete primary school have learned relatively little (see table); they may know how to read and do simple arithmetic, but they appear to lack the capacity to apply their knowledge to unfamiliar problems. Many cannot comprehend simple written material, such as instructions on a fertilizer package or directions for administering medicine to children.

This article is based on a book. Improving Primary Education in Developing Countries, by Marlaine E. Lockheed, Adriaan Verspoor, et al (1990, forthcoming).

One reason that primary schools are unproductive in developing countries is that education systems are very poorly funded. Median public recurrent expenditures per primary student, as a percentage of GNP per capita, fell nearly 50 percent from 1965 to 1985, dropping in real terms from $41 to $31 in low-income countries. In addition, primary education has often suffered from efforts to cut expenditures as developing countries undertook painful but necessary adjustment measures to cope with global and domestic economic difficulties. A study of countries undertaking structural adjustment over 1979-83 found that 68 percent reduced spending on education.

Improving the conditions of schools in developing countries requires careful attention to the allocation and management of resources, to ensure the selection and use of inputs that help student learning. By concentrating resources on inputs that work, more students will be able to learn what is taught and complete the primary cycle in fewer years. In many countries, management improvements are necessary preconditions for the implementation of programs to enhance learning. Three key management improvements are: (1) management training for administrators, to increase the likelihood that essential inputs reach schools; (2) organizational restructuring, to ensure that inputs are well utilized; and (3) systematic information collection, to provide managers with information about the impact of their programs on student achievement and other education indicators.

Effective strategies

Theory, research, and experience suggest five areas where efforts to accelerate learning in primary schools are likely to succeed. These efforts require well-managed interventions, with an important role for school principals and administrators.

Improving the curriculum. In primary schools, the official curriculum is remarkably similar worldwide. Not only are the same subjects taught, but the same relative importance is given to them: approximately 35 percent of available time is spent on literacy and 18 percent is spent on numeracy. While there are no clear differences in curricular emphasis between developed and developing countries, many developing countries divide instructional time for literacy between teaching a national language and a nonindigenous official language, a practice that may contribute to generally lower levels of language achievement in developing countries (see table).

Unfortunately, curricula in developing countries often suffer from poor instructional design. For example, a recent study of curricular scope and sequence in mathematics and reading textbooks for grades one, three, and five in 15 developing countries found that the material in both subjects was too difficult at the earlier grades. In the upper grades, the mathematics curriculum was too difficult, but the reading curriculum was too easy and failed to develop problem-solving skills appropriately.

To counter this situation, some 60 percent of Bank-financed primary education projects over 1970-83 included a curriculum reform component. One Bank-assisted project in Colombia, “Escuela Nueva,” developed an integrated curriculum for early primary grades, which both increased the achievement and lowered the repetition of rural children significantly.

Instruction materials. Well-designed instructional materials, such as textbooks, can enhance student achievement significantly. In Nicaragua, for example, students in classes with adequate textbooks scored significantly higher on tests of mathematics achievement than students in classes with no supplementary materials. Another evaluation in the Philippines found that when the student-to-textbook ratio was reduced from 10:1 to 2:1, learning improved, on average, by one third. The availability and use of libraries has generally been found to be effective too.

Time for learning. Research from a variety of countries has shown that the amount of time available for studies is consistently related to how much children learn in school. In general, the more teachers teach, the more students learn. While learning time is valuable for all students, it is especially important for poor students, whose out-of-school time and opportunities for learning are limited.

The annual number of hours children are allotted to study any subject in school is determined by three factors: the length of the official school year in hours, the proportion of these hours assigned to the subject, and the amount of time lost through school closings, teacher absences, student absences, and miscellaneous interruptions. Worldwide, the official academic year for primary schools (grades one through six) averages 880 instructional hours, or 180 days. In some developing countries, the academic year is substantially shorter (e.g., 610 hours in Ghana), in others it is longer (e.g., 1,070 hours in Morocco).

Proportion of students completing primary school, by GNP per capita, 1970-85

Source: Unesco.

Note: For definitions of country groupings, see note on table.

Actual instructional time, however, is much less than official time in developing countries. In Haiti, for example, the school year in 1984 had 162 days, 18 short of the international standard, but it was made significantly shorter by unofficial closings and delayed openings: the school day often began late, teachers frequently were absent on Tuesday and Friday market days, and 48 public holidays were celebrated, instead of the 28 holidays built into the school year.

Teacher absences due to administrative problems are also common in developing countries; for example, many teachers need to travel considerable distances to be paid, while others are assigned to teach far from their homes. Both situations can be improved with better management of the sector. Moreover, policy reforms, such as encouraging flexible scheduling of school hours, can reduce student absences by accommodating rural children’s work schedules.

Effective teaching. Teachers are central to the delivery of primary education. The academic and professional training of teachers has a direct and positive bearing on the quality of teaching performance, and, consequently, on the achievement of students.

Evidence from developed countries shows a strong positive effect of teacher knowledge on student achievement, but only a few studies from developing countries have examined this question. The results, however, are consistent. For example, the English-language proficiency of teachers in Uganda had a positive effect on their students’ achievement in both language and mathematics. Teachers’ achievement on a secondary school leaving examination was correlated with their second grade students’ achievement in Iran.

Effective teaching also depends upon teachers’ pedagogical practice. In many developing countries, teachers employ practices that do not facilitate learning. These include whole class instruction emphasizing lectures by teachers, student rote memorization, few opportunities for students to actively participate in learning, and little ongoing monitoring of student learning. This type of instruction tends to produce students who are unable to apply their knowledge in real-life situations and makes for a far less interesting learning environment.

Cross-national comparisons of achievement in mathematics, reading, and science
Average percentage of correct answers
Countries, by income group

per capita

grade 6


grade 8


grade 6


grades 4-6

Low and lower middle49394936
Upper middle55544353
Source: Improving Primary Eduction in Developing Countries by Marlaine E. Lockheed. Adnaan Verspoar et al, 1990.Note: Low income - 1986 GNP per capita of $425 or lessLower middle income - 1956 GNP per capita of $426-1,725.Upper middle income - 1986 GNP per capita of $1,726-7,425.High income - 1986 GNP per capita of above 57,425. excluding the high-income oil exporters.
Source: Improving Primary Eduction in Developing Countries by Marlaine E. Lockheed. Adnaan Verspoar et al, 1990.Note: Low income - 1986 GNP per capita of $425 or lessLower middle income - 1956 GNP per capita of $426-1,725.Upper middle income - 1986 GNP per capita of $1,726-7,425.High income - 1986 GNP per capita of above 57,425. excluding the high-income oil exporters.

Improving teacher knowledge and pedagogical practice is not easy, but in-service (often “distance”) training programs with suitable follow-up measures (through continuous “training and visit” programs) have proven cost-effective. Distance education takes a number of forms, ranging from educational television and radio programming to correspondence courses.

More educational materials = less cost

In rural Northeast Brazil in the mid-1980s, it cost $90 to “produce” a fourth grade student. But in schools supplied with an extra $3 per student in grades 1-3 for educational materials, improved facilities, and teacher training, it cost less. In fact, for every $1 invested in educational materials, a savings of $4 was realized. This apparent anomaly results from the fact that the addition of educational materials to classrooms in rural Northeast Brazil significantly increased the amount that children learned, and therefore the probability that they would be promoted to the next grade on time, saving resources otherwise spent on repeaters. Improved learning was the central feature of this intervention.

Interactive radio instruction, which is used to supplement the knowledge and pedagogical skills of poorly trained teachers, is one of the most effective and efficient educational interventions. Once lessons have been developed, the cost per student per year is very low because the same lessons can be transmitted to thousands of new students at minimal cost.

Enhancing children’s “teachability.” Children’s capacity for learning is largely determined by their health and nutritional status, and their prior learning. Children in developing countries are likely to be undernourished or malnourished and unexposed to learning materials prior to entering school. To address these deficits, governments have undertaken both in-school feeding and health programs, and preschool education programs.

The evidence regarding the effectiveness of in-school feeding and health programs on improving learning is limited. School mid-day meal programs enhance school attendance, but available analyses fail to show a clear relationship between lunches and academic achievement. School breakfasts or snacks are more cost-effective alternatives and make their impact at the beginning of the day, when alleviating short-term hunger is most likely to improve learning. To counteract micronutritional deficiencies, iodine and iron supplements have been shown to be highly cost-effective, and deworming is an inexpensive approach to dealing with parasites. Visual and auditory screening can be conducted with the use of simple eye charts and “whisper” tests at negligible cost.

Preschool educational programs can compensate for academic deprivations in the home environment. While the provision of preschool experiences for all children may not be a feasible solution for most developing countries, programs aimed at families with particularly low incomes have been used effectively. Private provision of preschools, possibly through nongovernmental organizations, should be encouraged. In middle-income countries, an alternative to preschool is educational television aimed at young children.

Plans for the future

A major constraint on educational development, particularly in low-income countries, has been insufficient support for educational inputs that enhance student learning. However, a coordinated program of investment in these inputs, combined with reforms in teacher training and sector management, could generate substantial returns to the investment through improved productivity and efficiency. But such a program requires significant initial spending to be effective, and should include measures of achievement, such as measures of students’ progress in key curriculum areas (e.g., reading and mathematics).

Increased aid. During the 1980s, only 4.3 percent of all international aid flows for education ($181.3 million disbursed annually, or about 40 cents per child in school) was directed at primary education; of this, less than 35 percent covered recurrent expenditures. Sound primary education development programs will typically require more resources, perhaps ten times as much, sustained over a longer period and directed at those recurrent cost items that promote learning. Although countries will need to increase their own education funding, start-up expenditures are likely to be greater than can be mobilized internally, especially in the short and medium term; interim budget support from external sources will be required.

Bank landing for primary education

Recognizing education as a critical means of building human capital, the World Bank began lending to education in 1963. Since then, it has lent a total of nearly $9 billion for education and training, through 375 projects in 100 countries. This represents about 5 percent of total Bank lending. Bank lending for primary education commenced in 1970 and increased steadily to average more than 25 percent of total Bank lending to education in recent years. In fiscal year 1989 it amounted to $508.5 million.

A review of education sector lending over FY 1984-88 found that the largest proportion (34 percent) of policy measures to improve primary education addressed issues of effectiveness; half of these policy measures focused on improving teaching (primarily through in-service training) and providing instructional materials (primarily textbooks). Other important inputs for enhancing children’s achievements were accorded significantly less attention. Long-term improvements will require more balanced attention to the curriculum, learning time, and children’s teachability.

Increased aid, however, will have to be targeted more effectively than it has been in the past. Although some low-income countries will continue to require external support in their efforts to expand coverage, expansion should not preclude focusing on quality improvements.

Subsector development programs. For external assistance to be effective, it is important to put support to primary education in the context of broad subsector expenditure programs. Specific investment programs could be appraised by a national agency on the basis of agreed criteria and national spending priorities. External financing can be commingled with national resources or targeted toward specific activities.

Flexible lending. Priorities will differ among countries. Consequently, donors will have to be flexible in defining their lending priorities. While textbooks might be of the utmost urgency in some countries, other countries, which have long neglected the maintenance of their facilities because they lacked resources, will want to upgrade their physical plant.

Monitoring achievement. Finally, to determine whether these investments in education have yielded their expected short-term objectives—increased productivity and efficiency—systematic information on children’s learning is required.

Other Resources Citing This Publication