DFID’s work on education: Leaving no-one behind
Written evidence submitted by Camfed International
Camfed is an international non-profit organization tackling poverty and inequality by supporting girls to go to school and succeed, and empowering young women to step up as leaders of change. Since 1993, Camfed’s education programmes in Zimbabwe, Zambia, Ghana, Tanzania and Malawi have directly supported over 1,419,000 students to attend primary and secondary school, and over 3.5 million children have benefited from an improved learning environment.
- Does DFID currently apportion enough priority and dedicate enough funding to its work on global education? Has it got the balance of spending on early years, primary, secondary and tertiary right?
1.1. The launch of the Sustainable Development Goals heralds a new commitment to achieving quality secondary education for all, and sets a challenge in how this commitment will be delivered and financed. There is growing recognition of the need for scalable solutions, to include:
1.1.1. Needs-based financing mechanisms that target the most marginalised to ensure no child is left behind in the drive for secondary education;
1.1.2. Innovative approaches to push up learning outcomes and provide young people with a meaningful pathway to transition from school to employment.
1.2. DFID is recognized as a global leader for its prioritization of investment in education (including by USAID and by major foundations, e.g. MasterCard). In particular, DFID has been at the forefront of recognizing the importance of investment throughout the pipeline of education. While some donors (e.g. USAID) have extended the pipeline downward to early childhood development, DFID has pioneered the extension of education investment upward to secondary level. This is particularly critical given the recent success in increasing primary school access, meaning that there are large cohorts of children now at the point of transitioning to secondary education which necessitates substantial investment at this level.
1.3. Education is the bedrock of stable societies and equality of opportunity. It is vitally important that this value is universally acknowledged and upheld, and that investment in education is not restricted to fragile and conflict affected states but is recognized as the precursor to development universally. The maximum returns to development are achieved when investment is targeted to ensure that those who are most marginalized within societies secure their right to education, ensuring that no child is left behind in the wider drive for education and economic development.
1.4. Camfed's direct experience of the wider impact of education provides a powerful reference point for the value of this universal investment in education in Less Development Countries (LDCs), particularly at the secondary level:
1.4.1. Marginalised girls who receive support through education are delaying marriage and motherhood;
1.4.2. Educated young women are increasing their civic engagement and being elected to office;
1.4.3. Local authorities are held increasingly to account in delivering services.
1.4.4. Educated young women are stepping up as the main breadwinners in their families;
1.4.5. Educated young women are substantially increasing philanthropic contributions within their communities. By the end of 2015, 147,128 children had been supported to attend secondary school by young women who had originally been supported to attend school by Camfed.
1.5. Through DFID’s investment since 2012, Camfed has supported 238,068 girls to progress through secondary school.
“I want to study and become a medical doctor, earn money for me and my sister, but beyond the money I just want to save lives. Moreover, I want to reach out to those who will be in my current situation. I am convinced that with the support I have in school through the bursary I will complete secondary school and pursue my interests. I believe I can make it. I tell my younger sister that we should concentrate on education and all else will follow.”
Mphatso, Mwanza District, Malawi
- Is DFID's education funding being targeted appropriately? How effective are its interventions through multilateral organisations (e.g. Global Partnership for Education, the World Bank), implementing partners, centrally managed programmes and direct support to government education budgets?
2.1. In Camfed’s experience of bilateral and centrally managed programmes, funding targeted through implementing partners is highly effective when positioned to enhance government provision of education and build the capacity of the most vulnerable and marginalized communities to hold the government to account. In this respect, DFID adds significant value as a partner in national education working groups, in reinforcing coordination among education stakeholders. Camfed’s programme – in particular its model of engagement with Ministries to extend educational access to marginalized groups, scaled through DFID investment – has won recognition by the OECD for ‘best practice in taking development innovation to scale’, demonstrating the value of DFID’s approach in working with and through implementing partners.
“This [Camfed’s] is an innovative governance model that engages Ministries, tracks progress publicly and hence forces accountability, thereby circumventing bureaucracy to achieve results.” Director of Science, Technology and Industry, OECD.
- Are there any best practice examples of DFID interventions in global education programmes? What is working well and can this be replicated? Are there any innovative approaches that could be scaled up?
3.1. Investments in girls’ education through DFID’s bilateral programme and through the Girls’ Education Challenge, have yielded many examples of best practice that have a significant effect on both retention and learning outcomes. Camfed’s model for supporting the education of the most marginalized girls, which has been scaled up through investment by DFID, has been recognized as best practice by the OECD (see above), and highlighted as an approach from which other countries can learn in the 2014 Education for All Global Monitoring Report and the newly launched Millions Learning report[1]. The REAL Centre at the University of Education focused on Camfed as a case study for The International Commission on Financing Global Education Opportunity (see appendix).
- Should DFID support low-fee schools, including private schools, in developing countries? If so, what should this support look like? If not, how can universal access be achieved?
4.1. At secondary level, school-going costs are a reality and will be so for the foreseeable future – in both government and private schools. Even in those countries where school fees have been abolished (the most recent example being Tanzania where school fees at secondary level have been removed), there are additional costs that prevent the poorest and most marginalized children from progressing through the school system. The issue lies in finding effective measures to lower the cost barriers, and in targeted support mechanisms to address the exclusion of the most marginalized children.
4.2. The reality is that there is an ecosystem of education provision, and DFID needs to work with this ecosystem, including to ensure that there is an appropriate regulatory framework to oversee this ecosystem. Private and public sector provision should not be seen as mutually exclusive; there are significant examples where there has been successful private sector engagement with public school systems. Indeed, innovation in education should also not be seen as exclusive to the private sector (as it sometimes is) but collaboration to incubate and scale innovation through the government school system has huge potential in improving learning outcomes.
4.3. It is also important to recognise that a drive for universal access should not be considered separately from improving learning outcomes. Many countries have made substantial gains in increasing access to education, particularly at primary school level, but this has not been coupled with a similar investment in the learning environment. For example, in rural areas of Zambia, teacher pupil ratios can be as high as 1:100. A poor learning environment leads to high rates of pupil drop out and poor learning outcomes. Those students who do complete school, are still not sufficiently equipped to transition to a life of economic independence, therefore the poverty cycle persists.
- How can DFID most effectively support efforts to increase learning outcomes, particularly ensuring that those children in school are gaining proficiency in literacy and numeracy?
5.1. Our experience and evidence based research demonstrates the importance of ensuring holistic support for children’s education that addresses the whole learning environment, rather than isolated activities to push up learning. The importance of considering the wider educational environment to improve learning outcomes is laid out in this Camfed article in the Center for Universal Education, which also cites the results achieved through DFID investment.
- Should DFID be doing more to support countries in developing more effective education systems, including high quality teacher education and leadership training?
6.1. It is important to adopt a long term view towards systemic change rather than a short term approach to hot-housing of student cohorts to achieve a push up in learning outcomes. In this regard, it is crucial that financing mechanisms are positioned to incentivise a longer term systemic approach (see below for reference to results based financing mechanisms). The Learning Metrics Task Force (on which DFID has participated) provides further important insights on this area. See here for a link to their report.
- Does DFID need to develop a greater focus on tertiary education, including technical and practical education? How can it support efforts to ensure that the growing youth population in many of its bilateral partner countries are given the life skills they need?
7.1. A greater focus on technical and practical education is of particular importance in view of the burgeoning youth population who are completing school into a dearth of employment opportunity. While support for tertiary education is important, there is a dramatic bottleneck in places at this level and a need for larger scale solutions, including appropriate preparation for young people while in school for the challenging pathway ahead. Education journalist Jordan Shapiro picks up this issue in a recent article.
- How best can DFID support efforts to eliminate inequalities in access and quality of education on the basis of gender, disability, indigeneity, and for children in vulnerable situations, including emergencies?
8.1. This is where DFID needs to be (or continue to be) particularly innovative, in prioritising targeted financing mechanisms to secure education for the most marginalized groups. Furthermore, data should be disaggregated to be able to track the results of particular groups. DFID’s investment to date has shown that new and innovative solutions are possible, including community led scale of education for the most marginalised children, through partnership with government school systems. The REAL Centre paper, included below as an appendix to this submission, provides further insight into what is possible.
Other issues for consideration:
- Results Based Financing linked to learning outcomes
9.1. In addition to providing an evidence-base of what works to promote education of marginalised girls at secondary education, Camfed’s experience of Results Based Financing (RBF), otherwise known as Payment by Results (PbR), provides insight to support the exploration of appropriate incentives for improved educational outcomes. RBF is a relatively under-developed science in the context of education, particularly where this is attached to learning outcomes. Given that DFID is committed to the further roll out of Results Based Financing in its investment in education, this is particularly timely to review this approach.
9.2. Emerging issues from Camfed’s experience for consideration include:
9.2.1. The challenges of setting targets linked to RBF are exacerbated by the fact that there are relatively few education interventions that have been evaluated at scale with respect to learning outcomes in developing country contexts. For RBF to be effective as a mechanism, it needs to be linked to deliverables that are known to be achievable within a particular context.
9.2.2. There is a need to calibrate targets for learning outcomes to ensure these are appropriate both for the age group and within the associated timeframe for the intervention.
9.2.3. There is a potential risk that RBF can lead to ‘hot-housing’ of cohorts of students and/or distortion of a set of narrow learning outcomes to demonstrate results, to the detriment of the opportunity for systemic change.
9.2.4. There is a risk that RBF attached to learning outcomes may indirectly drive the exclusion of children who are most marginalised, as they have greater hurdles to overcome to achieve the same level of push-up in their learning compared to their less marginalised peers, and thus a potential disincentive for their results to be included.
9.2.5. Related to the above, measurement of learning in the context of RBF should also be linked to measurement of school retention, particularly for those who are most marginalised, recognising that those who are most marginalised may be more vulnerable to dropping out.
9.2.6. In delivering improved learning outcomes linked to RBF, it is important that interventions target all children, rather than setting targets linked to pushing up the learning outcomes of one group relative to another which may otherwise have a detrimental impact on absolute levels of learning for all groups. At the same time, there is value in ensuring data can be disaggregated to explore the effect on learning across different groups.
9.2.7. There is value in ensuring a more rounded exploration of learning outcomes, again to mitigate the potential for a reductionist approach driven by RBF; for example, to include indicators that reflect improvements in the learning environment alongside individual academic assessment, to ensure a focus on longer term system change.
Appendix: Draft paper submitted to the Financing Commission (not for circulation).
Targeted, multidimensional approaches to overcome inequalities in secondary education: Case study of Camfed in Tanzania
Research for Equitable Access and Learning (REAL) Centre, University of Cambridge
Paper Commissioned by the International Commission on Financing Global Education Opportunity
DRAFT – NOT FOR CIRCULATION
Introduction
Interventions aimed at tackling multiple dimensions of disadvantage are important to improve both access and learning for the most marginalised children. These interventions aim to lift or ameliorate a range of barriers for educational access, while also boosting learning outcomes, thus reducing the chances that children leave schooling early or without basic literacy and numeracy skills.
Over the period of the Millennium Development Goals, strategies have been adopted to address both demand- and supply-side barriers in particular to children’s access to school. Notably, the abolition of official primary school fees has contributed to a significant increase in enrolment in countries that were furthest from achieving universal access to primary schooling. This is leading to pressure on secondary school systems. Even so, it remains the case that the most disadvantaged girls are not completing primary school, let alone making it to secondary school. This raises a question about the most appropriate support for those who do make it to secondary school.
The targeted support offered by the Campaign for Female Education (Camfed) provides an innovative example of interventions tackling the multiple dimensions of disadvantage for adolescent girls who make it to secondary school in Tanzania, as well as other countries in sub-Saharan Africa. Camfed’s support targets a range of barriers to girls’ secondary education at an age when girls are at great risk of dropping out due to factors such as poverty, early marriage and teenage pregnancy. In particular, Camfed’s programme (Camfed International, 2015):
- Removes financial barriers by covering direct and indirect costs of schooling for girls who are identified as needing support
- Provides support to community-led initiatives to improve schooling
- Trains teacher mentors and staff and parents to improve educational quality
- Develops and distributes low cost educational resources
- Enables young women school graduates to take on a leadership role as ‘Learner Guides’ in their local schools to provide mentoring and deliver a life skills curriculum
As part of DFID’s Girls’ Education Challenge programme, to date Camfed has provided support to 40,219 marginalised girls (in addition to 24,650 less marginalised girls, 48,584 marginalised boys and 29,778 less marginalised boys) across 201 secondary schools in Tanzania. In the following analyses, we focus on a subsample for which Camfed collected detailed data for evaluation purposes. For this subsample, we analyse enrolment and learning outcomes for 1,640 girls from six districts who received Camfed support in relation to 849 girls in comparison to schools who did not receive support (Camfed International, 2016). More specifically, we compare changes in these outcomes between the evaluation’s baseline (when girls were in Grade 2) and midline (when girls were in Grade 4) points. We use logistic regression models to assess the impact of Camfed support on access, and difference-in-difference models to assess its impact on learning outcomes. The learning outcome on which we focus is a maths assessment designed by the National Examination Council of Tanzania to be age and curriculum appropriate. The assessment took the form of one-hour exam scored between 0 and 100, and was taken by girls at both baseline and midline (with different questions of an equivalent level at each time point). Even though based on curricular standards, scores were very low at the baseline level, at around 12 points out of 100 for girls and 18 points for boys. While surveyed children were also assessed in English, we focus on learning outcomes in maths as they are less likely to be influenced by distortions in home factors.
Three principles
Based on this analysis, we identify three principles that provide lessons for supporting disadvantaged girls in secondary school in other related contexts.
1. Targeted interventions should support access and learning for those most in need
It is both important for interventions to support access amongst the most marginalised, as well as to make sure that this is accompanied by gains in their learning. Accounting for household wealth, marginalised girls in Camfed-supported schools were 18% less likely to dropout compared with marginalised girls from comparison districts. This improvement is primarily driven by Camfed’s bursary scheme: marginalised girls receiving a bursary are 31% less likely to dropout than are marginalised girls from comparison districts. This improvement was similar across different degrees of household disadvantage and levels of disability.
It is important to recognise that the chances of a girl dropping out depends on initial learning levels. Amongst weaker learners who receive a bursary, dropout reduces from 20% to 16%. However, their chances of dropping out are still far higher than amongst stronger learners with a bursary: amongst this group, dropout falls from 8% to 3%. It is evident that girls who are initially weaker learners will require additional forms of support to further increase their opportunities to remain in school. These could include whether the amount of the bursary needs to be further increased, targeted at weaker learners, or whether complementary support is needed beyond the bursary itself.
The improvement in access was also reflected in considerable gains in learning outcomes among Camfed-supported girls: on average, marginalised girls who received Camfed support almost tripled their scores on the learning assessment, from 11 to 28 points. This is in stark contrast to equivalent marginalised girls who did not receive support, who made no significant improvement in their assessment scores.[2] This is all the more striking given that comparison schools were selected in part on the basis that they had similar prior Grade 4 assessment scores to those schools receiving Camfed support (Camfed International, 2015). In addition, while Camfed support also boosted learning among marginalised boys (from 17 to 31 points), benefits were greatest amongst marginalised girls, thus helping to close the gender inequality in learning.
These improvements contrast with programmes that use singular conditional-cash transfer interventions, rather than the multidimensional forms of support that Camfed provides. Whereas cash transfers have often been shown to improve access to education, there is limited evidence on their impact on learning.[3] In contrast, it appears that Camfed’s multidimensional approach leads to modest improvements in access that are accompanied by sizeable gains in learning. This implies that combining financial incentives with pedagogical and other forms of in-school support is important for learning gains to be achieved. Given though that the assessment was scored out of 100 and based on curricular expectations, it is also important to note that there is still a way to go before all these children are learning at the appropriate level for their grade.
2. The most marginalised girls (including girls with disabilities) can benefit equally from targeted interventions
Using Camfed’s marginality tools in the survey, we are able to identify those girls facing the greatest degree of marginalisation, even compared with the other marginalised girls also supported by the programme. For example, while some girls may be targeted due to one condition of need (such as low income) others may have multiple sources of disadvantage (low income and having a long term illness or disability). Figure 3.1 shows girls’ absolute improvement in mathematics between Grade 2 and Grade 4 by levels of disadvantage, as measured by household characteristics and asset ownership.
First, it is important to note that all children had low levels of learning: no group averaged more than 20% on the baseline assessment. But, while girls who did not receive Camfed support did not make any progress between Grades 2 and 4, those who did receive support made substantial learning gains regardless of degree of disadvantage. For example, while the most advantaged girls receiving support improved their maths scores by 15 points (from 12 to 27), the most disadvantaged girls improved their maths scores by 17 points (from 18 to 35).
Figure 3.1. Absolute gains in Mathematics test scores for girls according to extent of marginalisation (low to high marginalisation)
Source: Authors’ analysis based on Camfed Tanzania data
Camfed’s targeted support has also enabled learning benefits for marginalised girls with disabilities, even though this is not a central aim of the programme. The Camfed programme for Tanzania shows that all girls, regardless of whether they reported difficulties in seeing, hearing, concentration or walking, even when these difficulties were moderate to severe, improved their attainment in English and mathematics compared with girls who were not supported by the programme.
Marginalised girls who were identified as having mild, moderate or severe difficulties in seeing, hearing, walking, caring, understanding or remembering performed at similar learning levels in mathematics to marginalised girls without difficulties (Figure 3.2). Mathematics attainment before the programme started was 12 points for marginalised girls without difficulties or disabilities, whereas it was 11.5 points for girls with mild difficulties and 11 points for those with moderate to severe difficulties. One likely reason for this is that Camfed’s work with the Girls’ Education Challenge focuses on supporting marginalised girls who are still in the second year of secondary education.[4] This means that marginalised girls affected by any of the above difficulties have managed to receive up to 8 years of education, and so are likely to be atypical of children with disabilities. Even though this might be a selected group of girls affected by difficulties, the results demonstrate that removing barriers to access and facilitating perseverance in education enables learning.
The targeted support benefitted girls’ learning across the spectrum of mild, moderate or severe difficulties. Once supported by the programme, we find that marginalised girls significantly increased their English and mathematics test scores over the period of two years, regardless of physical or mental learning difficulties. For example, on average girls with moderate/severe difficulties improved their maths scores from 11 to 28 points out of 100. By comparison, equivalent marginalised girls not supported by the programme did not improve. In addition, it is worth noting that marginalised girls with physical or mental difficulties supported by Camfed improved their learning at the same rate as marginalised girls who did not report any difficulties.
Figure 3.2. Absolute gains in Mathematics test scores for girls according to extent of disability
Source: Authors’ analysis based on Camfed Tanzania data
3. Pedagogical reforms need to support girls’ self-esteem as a means to boost learning
Children’s positive self-perceptions are essential to their learning, but this topic remains relatively unexplored in the context of marginalised children in low-income countries. Camfed’s focus on promoting more inclusive education environments together with its support of Learner Guide (who are female secondary school graduates from marginalised backgrounds, many of whom had previously been supported by Camfed (Camfed International, 2016)) could be expected to support girls with low self-esteem or aspirations. In particular, the programme’s Learner Guides aim to expand the presence of female role models, and lead students through a broad life skills curriculum, organise academic study groups, and provide counselling and follow up on students in danger of dropping out.
During Camfed’s baseline survey, girls reported on their confidence in a range of skills, including working in groups, working with other people, helping others, managing or solving problems, doing homework, and future planning.[5] From this scale we divided girls into quartiles, from low to high self-esteem.
Figure 3.3. Absolute gains in Mathematics test scores for girls according to self-esteem
Source: Authors’ analysis based on Camfed Tanzania data
Figure 3.3 shows that, for those girls identified with low self-esteem, Camfed support is helping them to improve their learning at a similar rate to those girls receiving support who are identified as having higher levels of self-esteem. At baseline, girls’ learning differed according to level of self-esteem: girls with low self-esteem achieved, on average, 11 points in mathematics test scores, whereas girls with high self-esteem achieved, on average, 13 points. Yet, learning benefits resulting from Camfed support were seen across the whole self-esteem scale, with increasing in mathematics of around 15 to 16 points on average regardless of whether girls had shown high or low levels of self-esteem. As before, this suggests that the benefits of programme support also reached those girls likely to be most at risk of failing to complete secondary school.
Conclusion
There is still much more to do to raise learning outcomes for all. This is clear given how far children’s assessment scores are from national curricular expectations, even amongst those who were able to improve. Nonetheless, Camfed support shows that, under appropriate conditions and with good targeting, it is possible to level the playing field.
Experience from Camfed’s programme offers principles to help support disadvantaged girls in secondary school in other related contexts:
1. Targeted interventions should support access and learning for those most in need
2. The benefits of targeted interventions can be shared equally for more disadvantaged children, as well as for children with disabilities
3. With pedagogical reform, it is also important to enable attitudes and behaviour that boost learning