5.1. Introduction
5.2 Expanding educational provision
5.3 Type of school provision and organisation
5.4 School inputs
5.5. Community involvement and awareness
5.6. Improving girls' health and nutrition
5.7 Recruiting more female teachers
5.8 Reducing direct costs
5.9 Reducing indirect costs
Strategies to redress gender inequalities are usually constructed on the basis of "supply" and "demand factors". Supply side strategies to expand access are necessary to increase girls' enrolment but they are rarely sufficient. International experience has shown that simply expanding education programmes does not automatically result in greater female enrolment. In some South Asian countries and Yemen, school expansion policies have only been effective when accompanied by other policies that lower the direct or opportunity cost of education or raise the benefits (King and Hill. 1993).
Numerous gender specific education policies have been tried by governments, donors and NGOs in a variety of combinations in each of the three countries under scrutiny. Because most of them have only been introduced over the last five years or so, any systematic and comprehensive impact evaluation is not possible. Nonetheless, some data are already available about short term outputs and impacts. The World Bank has undertaken or directly supported much of the research that has informed and shaped policy initiatives of all kinds in support of female education (mostly in South Asia and some in Africa). Early Bank funded projects tended to focus on single interventions. However, most were not successful because they could not address the multiple constraints affecting girls. In contrast, projects that implemented 'package' approaches have had better results (Herz et al, 1991). There is, however, some disagreement among World Bank programme officers regarding the desirability and feasibility of the 'package approach'. Some believe that compound strategies are more effective than single approaches while others consider that the multiple approaches may overburden government bureaucracies in some countries and resulting in poor implementation. Despite these reservations, the weight to research evidence and theoretical considerations would suggest that policy 'packages' are likely to be the most effective (Stromquist, 1994).
Two main types of gender strategies have been tried: (i) providing more educational facilities for girls and; (ii) reducing the direct and indirect costs of schooling. The following policies have been developed by governments and aid donors.
It is obvious that educational provision must be tailored to the specific needs of particular countries. More than half of existing World Bank projects (in 1991) emphasised physical access to schools and 46% of the projects included having more female teachers and removing sexual stereotypes from textbooks (Ibid). Various strategies have been designed to improve girl's access to schooling. In some countries like Malawi and Zambia there is an urgent need to construct more schools in both rural and urban areas. In terms of physical provision, it is essential that any new schools or converted schools provide adequate sanitary facilities and privacy for girls as these factors often affect the attendance of girls as we have seen.
Quota systems for girls entering secondary school have been introduced at various times in Malawi, Zambia, Zimbabwe as well as other SSA countries. In Tanzania, a quota system involving the separate selection of boys and girls and admission of girls with lower grades into Form I of secondary school has been quite successful in increasing female enrolments of girls although a 50:50 policy should be the goal (Mbilinyi et al, 1991). Quotas for secondary schooling are probably necessary but again they are not sufficient to improve girls' participation if they are not accompanied by financial support. In Malawi the 33% quota system for girls' entry into secondary school does not seem to have significantly raised the proportion of girls over the last decade. Girls' access to secondary education in Zambia has been hampered by the practice of reserving approximately one-third of places in co-educational state schools for girls (Kelly, 1994).
Organisation issues are of paramount importance when considering any expansion of school enrolments. The size and spacial distribution of schools is particularly critical. The choice is often between one large centralised school or smaller schools at shorter distances from children's' homes. Research suggests that smaller schools with closer community ties are more effective for boys as well as girls (Herz et al, 1991). The distance problem has been approached in two ways: either the schools are brought closer to homes or boarding facilities are provided. There is a strong move towards decentralisation of school structures in many African countries which has the potential of bringing schooling closer to the people. Different modes of educational delivery have been experimented with, including multigrade classrooms, double shifting with feeder and satellite schools at the primary level, radio and correspondence courses at the post-primary levels, and literacy programmes for adults (King and Hill, 1993). Distance education is particularly suited to countries like Zambia with widely dispersed rural communities.
The relative merits of single sex day and boarding schools have already been discussed at some length. Although single sex boarding schools are relatively costly, they do seem to offer a safe and secure atmosphere for girls. Until mixed schools are able to provide an atmosphere that is supportive to girls, single sex schooling, in theory at least, remains a desirable, although costly, option.
Research has established that once schools have been built, further investment in physical infrastructure is less important for student learning than incremental expenditures on curriculum design, textbooks and other inputs to improve school quality. Parents are more willing to shoulder the costs of educating children if they consider the curriculum relevant and the quality of schooling adequate
Appleton's study of Kenyan primary schools showed that students from homes with books and from schools with more textbooks performed better in the primary school leaving exams (Appleton, 1994). It is now widely accepted that textbooks have a definite positive impact on children's educational progress. Providing more textbooks per child at the same time as making the curriculum and textbooks gender sensitive and promoting positive images of women and girls, should work to both improve the quality of education and reduce gender bias in the long term. World Bank assisted projects in Bangladesh and The Gambia support the development of unbiased educational materials and teacher training programmes designed to eliminate gender bias in the classroom (Herz et al, 1991). The impact of textbooks and other school learning materials is particularly important where these constitute the sum of written materials in childrens' lives. It is likely in the African context that girls will derive more benefit than boys from improved curricula and textbooks as they receive less attention from teachers and tend to be left more on their own (Kelly, 1994).
Anna Obura in a study of school textbooks in Kenya found evidence of severe gender bias and distortions in the roles and images of women portrayed in textbooks across all subjects. The passive and domestic roles of women found in these texts was sharply at odds with reality (Obura, 1991). In countries like Zambia and Zimbabwe there have been some efforts to improve the general quality of textbooks, but due to general shortages, many schools are using older textbooks which are still gender biased. Less progress has been made in changing content of curriculum from a gender perspective than in textbook development although clearly these should be closely interconnected.
In many parts of the world, girls study less science and maths than boys. Research shows that this reflects a subtle interaction between traditional attitudes on what women 'should study' and the options for study offered to women and girls. Single sex schools often lack the same offerings in maths and science as boys schools and in Africa girls often avoid these subjects because they perceive them to be 'hard'. It is undoubtedly the case that science and maths provision and support for girls needs to be improved. If girls are encouraged to take more maths and science options the benefits could be considerable in terms of access to non-traditional and traditional occupations. In Kenya, for example, where proportionately fewer qualified girls than boys gain admission to government schools, parents spend more to send girls to private schools (Herz et al, 1991). Three times as many of these girls study science there as do girls in the less costly government schools, which suggests that parents are more willing to pay if girls can study science (World Bank, 1989). The Bank also assists a project in Tanzania which promotes science courses in new secondary schools, 60% of the facilities added are for girls' schools. The Ghana Education Services Offer a two week science clinic for 150 secondary school girls on an annual basis. Furthermore, the Association of Women in Science and Technology in Ghana also organise 'science days' for girls, workshops for women science teachers, and career guidance talks. These initiatives have had a positive impact (Odaga and Heneveld, 1995). Such measures can help in a small way to increase the relevance of the school curriculum and improve educational outcomes for girls.
Negative parental attitudes concerning the education of girls may change over time, but in the meantime community awareness campaigns stressing the benefits of educating girls and involving the community in the management of schools need to be explored. If a dialogue can be developed between the community and the school then policy makers are better able to respond to the particular demands of communities. Satellite schools in Bangladesh were popular partly because schools were located closer to communities and partly because the community itself chose opening hours to fit in with their work schedules. In China, India, Morocco, Nepal and Papua New Guinea, community input, from awareness campaigns to educating the community to manage the education projects themselves, has been key to the success of various programmes designed to raise enrolments (Herz, 1991). Due to the tight parental control over girls in Africa, it is essential to involve and educate communities about the benefits of girls education. However, this must not merely be a top down process: the community must be genuinely involved in educational decision making.
Children's' health and nutritional status has consequences for drop out and school performance. Research in such diverse social settings such as Guatemala and Kenya has established that malnutrition inhibits children's learning even as early as the age of two (Herz et al, 1991). School feeding programmes have been adopted by some governments usually with the assistance of donors (Most notably the World Bank and the World Food programme). Some have had positive effects in improving girls' participation. Recent research in northern Ghana has also shown that schools can be important centres for the delivery of health to teenage children (particularly girls) whose health status is low. However, the same research also shows that cultural constraints affect the drop out of girls from school which is of great concern because they do not benefit from school health services (Fentiman, 1995). In Zimbabwe, both primary and secondary schools under the new UNICEF Family Life programme will become the main vehicle for the delivery of HIV/AIDS education to high risk groups of young people.
In cultural settings where male/female contact is limited and tightly defined, it is generally the case that the recruitment of female teachers enhances girls enrolments. Several studies using international cross sectional data have found a positive correlation between the proportion of women teachers and size of female school enrolment (Herz et al 1991, Rockerfeller Foundation, 1995). However, this broad statistical picture needs to be unpacked at the national and regional level. As we have seen in Malawi, women teachers do not necessarily have better pedagogical practices than men nor are they more supportive of girl students even though they have more subjective understanding of the 'double burden' of school and domestic work.
In Bangladesh, World Bank funded projects have been part of successful packages to raise the enrolment of girls. Women teachers are popular in schools because of the very strict separation of the sexes required by religious and cultural norms. These conditions are not exactly the same in Africa where girls do have more relative freedom of movement. However, as we have seen, the level of sexual harassment surrounding schooling and risks involved in unwanted pregnancies and HIV are of obvious concern to many African parents. Although the reasons for drop out in our three countries have been by no means exhaustively researched, fear for girls' safety and security is in many cases a very important factor preventing particularly rural households from sending girls to school and keeping them there. The security issue also causes delays in sending girls to school if they are required to travel long distances.
It is undoubtedly the case that in a predominantly female environment, the girls and their parents feel more secure about their education and the girls perform better academically. At the present time, single sex schools, run mainly by women teachers and women heads, are able to boost performance of girls but this must be seen as a short term solution to the problem.
The positive impact of women teachers depends on the environment in which they are working and whether they have been effectively trained. However, the training of all teachers, from a gender perspective is crucial to changing attitudes in the school and community and is likely to have a greater impact than community awareness campaigns in the short and long term. It could well be that the positive impact of having women heads might be even greater than women teachers in terms of providing a secure environment for girls' learning. Experience in Africa has shown that when these conditions are present and there are not financial constraints, parents do not hesitate to send their girls to school.
Scholarship programmes aimed at reducing direct costs of secondary schooling for girls and their parents may involve the provision of fees as well as meeting other costs associated with schooling such as uniforms, equipment, textbooks etc. Scholarship schemes for girls in particular have proved to be both practical and effective. While the high costs of such schemes make them difficult to replicate country wide, they have, nonetheless, been shown to be extremely helpful in raising female enrolment in specific localities and producing positive female role models. A World Bank funded project in Bangladesh that dates from 1977, gives 18,000 girls annually in grades 6-10 of secondary school scholarships to cover school fees. Female enrolment doubled from 27.3% to 43.5% during the course of the eighteen year project. Female drop outs also fell dramatically from 14.7% to 3.5% (Herz et al, 1991). Overall, the project met its main goals: female graduates 'are marrying later, bearing fewer children and finding employment outside the home' (ibid: 44). A number of important lessons have been learned from this project, notably the necessity of establishing realistic selection criteria for students receiving scholarships and the need to build these into the project from the beginning.
A similar World Bank assisted project was begun in 1987 in a rural Indian community in Guatemala to support female primary students. The programme is supervised by local women and selection carried out by parents' committees. The programme has shown remarkable success in enroling and retaining girls: over 90% of the scholarship girls completed the first year. The involvement of community groups in the targeting and monitoring process has been instrumental in its success. An unanticipated problem has arisen, however, in that girls completing their education have been encouraged to continue with secondary education yet existing post-primary the schools are located far from the communities. Building on the experience of past programmes, another Bank financed project is currently expanding the scholarship effort in Bangladesh and incorporating other measures to reduce the direct costs of girls' schooling.
In Zimbabwe, The Cambridge Female Education Trust (CAMFED), which was founded in 1992, assists girls from disadvantaged social and economic backgrounds in three rural districts to continue their education at the upper primary and secondary levels in day and boarding schools. Teachers, parents and the wider community are actively involved with CAMFED in administering and developing all aspects of the project. As we have seen, many girls in Zimbabwe drop out during puberty for various reasons including poverty, lack of separate bathroom facilities and running water in school, sexual harassment, pregnancy and early marriage. This initiative has its origins in a research project undertaken by Ann Cotton during 1991 in the village of Mola in Nyaminyami district, which found that the main constraint on girls' secondary education was poverty. The project seeks not only to solve the immediate financial needs of girls who want secondary schooling but it operates an integrated approach to enhance the social environment in which girls grow and learn.
The CAMFED project appears to have selected schools that are receptive to the idea of promoting girl's education and the aim is to encourage these schools to increase female enrolment overall so that there is near parity with boys (CAMFED, 1995). The Zimbabwean fieldworker is also in constant contact with the parents and stresses the importance of supporting girls in their school work. She also addresses practical constraints such as menstruation facing girls from poor families who attend school, by encouraging girls to make re-usable sanitary towels and she conducts discussion groups with the girls on key issues such as sexuality and HIV/AIDS. The communities themselves are gradually being integrated into CAMFED's planning process. Sewing cooperatives have been established alongside the project in order to provide employment opportunities for girls leaving school.
The basis of student selection is threefold; (1) the girls must want to go to school; (2) the parents should support her ongoing school attendance and; (3) parents lack adequate funds to support their daughter/s through secondary school. Girls are supported in groups from their communities into day or boarding schools within the districts of Nyaminyami, Chikomba and Nyanga. To date, drop out rates have been very low, although the first cohort of students has yet to receive their '0' level results. The integration of the project into the community helps to make a bridge between home and school which is the key to its apparent success. It is also an important element in encouraging progressive change in the schools themselves. Finally, CAMFED directly assists the schools by building up links between some secondary schools in Cambridge and project schools in Zimbabwe.
Only a comprehensive evaluation of these students' performance and progress at school will be able to determine the project's overall impact. An evaluation of the project is planned for 1997. In 1994, CAMFED supported 233 girls at secondary schools in Zimbabwe and it will maintain its existing programme of support and expand its development within the same three rural districts. The total budget between 1994-1997 amounts to £365,427 which will provide for 1,308 years of secondary schooling for girls and the costs of dissemination lessons learned from the project (CAMFED, 1995).
As with the World Bank's scholarship programmes, experience suggests that the cost of such programmes can be justified by the obvious health and economic benefits of educating girls in the long run. Both the World Bank and CAMFED type projects are making small contributions to what is a massive problem. CAMFED in particular is a 'flagship' project designed to provide role models in impoverished rural communities. This type of project is certainly replicable in Malawi but it cannot possibly be extended to meet the needs of all rural girls for secondary schooling on the grounds of cost. However, when there are so few female role models in rural communities it would seem to be a good idea to extend some kind of scholarship programme which could help the process of transforming secondary schools as male dominated institutions as well as assisting individual girls with their education.
The problems that are likely to arise from scholarship programmes are firstly the criteria for selection of students when there is great need and secondly the sustainability of such projects in the long run. An alternative strategy is to introduce fee waivers for girls entering secondary school. This has not been tried often in the three countries under consideration. The GABLE project financed fee waivers for primary school girls before October 1994 when primary schooling was made free in Malawi. This could well be a more equitable means than scholarship programmes of extending opportunities of secondary education to girls.
Providing more flexible school hours, establishing childcare facilities and improving the communities' supplies of fuel and water can all help decrease the opportunity cost of sending girls to school. Part-time and flexible scheduling have been incorporated in World Bank financed projects in Bangladesh, Morocco and Pakistan (Herz et al, 1991). They have already been tried with favourable results in China and India. In 1979-80, a non-formal programme offering night classes in 22 villages in Maharashtra State, India, addressed the problem of non-enrolment of working girls. The classes, aimed at drop outs and those who had never attended school, were held between 7 and 9 a.m, 300 days a year. Local volunteers were recruited as instructors and the curriculum was geared to the community situation. Over a five year period, more than 4,000 children enrolled, 75% of them female and all of them full time workers engaged in household or farm work. The 25% drop out rate in project schools was lower than the average 50-70% grade 1 drop out rate (ibid).
Among interventions to reduce women's work, an important objective of labour saving technologies is to free up girls' time for schooling. Examples are a fuel efficient woodburning stove in Nepal, accessible water wells and simple mechanised grain mills in Burkina Faso. It is hard, however, to assess the impacts of such measures, as they are usually part of a package of school based changes.
To reiterate, the World Bank has found that package approaches which combine a variety of different interventions have been the most effective in improving the access of girls to both primary and secondary schooling. The Fourth Primary Project (1980-86) in Bangladesh addressed both household-level and school-level constraints by improving school facilities, training more female teachers, providing free textbooks, and scholarships for girls. As a result, female enrolments improved 9% annually in project areas compared with only 2% in the country as a whole. In countries where direct costs are high for poor families, subsidies are required. Where poverty is regionally concentrated, careful targeting of such areas is necessary.
