Blavatnik School of Government, University of Oxford
The RISE Programme is investigating how education systems can become more coherent for learning.1 The program has produced substantial empirical evidence on the current state of education systems and children's learning, and on approaches that do and do not improve children's learning. It has also generated theoretical work on the characteristics of a coherent system. What has been missing, but is slowly emerging from this body of work, is an articulation of and empirical evidence on how systems shift from being incoherent for learning to being coherent for learning.
The nominal purpose of an education system is to equip the children of a country (or region, or district) with a high-quality education. However, in practice, education systems serve many purposes, some explicit (such as goals of social cohesion) and some implicit (such as providing rents to political or bureaucratic actors). When these purposes stray too far from the purpose of providing a high-quality education, low learning outcomes often result.
This blog highlights a short but growing list of examples of education systems that have successfully shifted to being (more) coherent for learning, often with large learning gains resulting. A common theme across these examples is the restoration of the core purpose of the education system to equip children with learning. This often has taken the form of strong political will and the commitment and dedication of leaders to explicit learning goals, and the clear communication and delegation of these goals to the rest of the system. The commitment to and delegation of learning goals then allowed other elements in the system, which in the RISE systems framework we think of as information, motivation, support, and finance, to also align around these goals.2 Such political will and commitment have often been spurred by new and widely available information on low learning levels, and by grassroots movements demanding higher quality education.
There is a growing, though still limited, number of examples of education systems that have shifted from incoherence for learning to greater coherence for learning.3
In just 12 years, Sobral, Brazil went from the 1,366th municipality for learning to being the top performer in 2017 in the national basic education ranking. This is despite high levels of poverty; its scores are 80 percent higher than would be expected for its level of education expenditure. A learning assessment in 2000-2001, conducted by the municipality, had revealed that 40 percent of primary school students could not read. In response to these findings and others, Sobral’s Mayor delegated seven education goals, the top two priorities of which focused on literacy for primary school aged children. Under the banner of “alphabetization at the right age,” the goals were sharp and easy to communicate: 1) ensure all 6- and 7-year-olds learn to read and 2) ensure all students in Grades 2 through 6 who could not read receive remedial support and learn to read. Initial policies, however, focused on one of the other goals, reducing grade repetition, which did not achieve intended learning gains.
In 2006 the literacy goals were reemphasized, driving a set of policies that brought information (learning-based monitoring systems with feedback loops for teachers and schools to implement assessment informed instruction, and all children assessed twice per year), support (scripted lessons, well-aligned teaching materials, and professional development based on materials), and motivation (good communication, bonuses, awards, and public recognition events) into line with learning goals. Large improvements in early literacy resulted. The clear establishment of learning goals and alignment of other elements of the system with the goals were possible largely because of the championship of learning by political leaders.4
In 2009-2012 multiple learning assessments, including the Uwezo assessment and the baseline assessment for the Primary Math and Reading (PRIMR) pilot program, showed very low student learning levels, serving as a “wake up call” to education leaders. Subsequent success of PRIMR led to a desire among government authorities for a nation-wide scale up to improve learning, which resulted in the Tusome program. Tusome established national benchmarks for learning outcomes and these expectations were communicated down the system to schools. These clearly delegated benchmark expectations enabled other elements in the system to be aligned with the benchmarks, including support (well-structured year-long curriculum, teachers’ guides, structured lessons, formative assessments for classroom use, ongoing coaching for teachers, textbooks for students) and information (use of EGRA and EGMA to track learning, accountability mechanisms for coaches and supervisors), with large learning improvements resulting.5
Puebla, Mexico has the fifth highest poverty level out of Mexico’s 32 states, yet ranks among the highest in national education assessments. In 2011, a new administration came into office, and the new state secretary of public education established three education objectives: universal school attendance, universal persistence and completion through upper secondary school, and universal learning achievement above minimum level on national assessment. From these clearly established and delegated goals, a program was implemented initially in select lower-performing schools, and later more widely, that brought support (well designed and structured materials distributed to schools, teacher training focused on quality not quantity, locally recruited training facilitators from among the best teachers/principals/supervisors), information (new mid-year Grade 2 assessment allowing time to course correct, assessments in Grade 5, identification of at-risk children for targeted attention, school visits by decision-makers), and motivation (facilitators/supervisors given new honorary positions with status and prestige) in line with the targets.6
Between 2006 and 2012, pass rates on Tanzania’s primary school leaving exams more than halved, with only 31 percent passing in 2012. This, combined with poor results in foundational skills on Uwezo assessments and subsequent EGRA and EGMA assessments, spurred the government to take action to improve learning outcomes. In response, the government instituted the 3Rs curriculum reform with the clearly delegated purpose to improve reading, writing, and arithmetic (3Rs) in the early grades, specifically Grade 1 and 2.
Implementing the reform included providing support to teachers and schools (developing and supplying new curriculum, textbooks, and teacher instructional materials; in-service teacher training; school-based continuous professional development modules, and training for head teachers to support school leadership), new information was provided in the system through the introduction of an annual Grade 2 assessment, and the government received financial support from GPE, UNICEF, and USAID that was aligned with the delegated goals. At least nine actors were involved in the design and implementation of programs related to the 3Rs reform, ranging from the Ministry of Education Science and Technology (MoEST), the Tanzania Institute of Education (TIE), GPE, UNICEF, USAID, and others. While the number of actors involved presented coordination challenges, the clear, common, delegated goal of improving foundational literacy and numeracy enabled many actors to undertake separate tasks and still achieve coherence to a common goal, and ultimately improve learning outcomes. An external evaluation of the reform estimates that it achieved large, positive increases in children’s learning in both subjects.7
Pratham, arguably one of the largest and most successful education NGOs, launched its ASER assessments in 2005 under a theory of change that information on (low) learning could empower individuals to hold governments to account and spur political dedication to learning, delegation of learning goals, and ultimately improvements to learning outcomes. Indeed, 15 years later, ASER results have been heavily cited by education authorities in India and informed education policies at both national and state level.
In part because of low learning results revealed in ASER assessments, multiple state governments have delegated learning improvements to schools, and 15 have partnered with Pratham to support and train teachers on instructional techniques to ensure children master foundational literacy and numeracy, and on using ASER-like assessments in the classroom for ongoing information on learning levels and improvements. Further government interest and partnerships have emerged as well, such as Uttarakhand state’s work with Room to Read on foundational learning, with large results. And ASER assessment questions have been incorporated into some state-level assessments to provide ongoing information on learning improvements.
Findings from ASER assessments over many years also contributed to the delegation of foundational learning goals from the national government in the new National Education Policy 2020 and associated national mission on Foundational Literacy and Numeracy. It is too early to gauge the effects of these new learning goals, but the political attention to foundational skills is itself an achievement.
The Breakthrough to Setswana program is a nationwide program in Botswana, started in the 1980s, that to date ensures most children achieve foundational literacy in the early primary years. Initially piloted in 1983, the program expanded nationwide over the next eight years, with the clear dedication and delegation from the government to ensure lower primary students master foundational reading comprehension, writing, and listening skills through an integrated teaching approach. The Ministry of Education funded the expansion of the program, which involved large scale support to teachers to implement the approach, including providing teacher training, “teacher advisors,” and teaching and learning materials (teacher manuals, children’s books, charts, cards and sentence makers). RISE is currently undertaking a case study of the program to better understand the origins and drivers, development and deployment, and impressive longevity of the program.8
Many education actors, including national education leaders, non-governmental actors, and international donors and funders, want to know what to do to bring education systems into coherence for learning and improve learning outcomes. There are actions that can be taken to facilitate dedication to, and delegation of, clear learning goals, that allow other elements in the system to align for learning. Many involve longer time frames than typical development “projects,” but by facilitating the reorientation of a system (whether at the national, regional, or district level) they have the potential to produce large gains. These are just a few proposals.
Findings from learning assessments appear in multiple examples above as a catalyst for spurring delegation of learning improvements. Such learning assessments are for driving attention and action; they are not for management accountability, nor are they high stakes for teachers or students. In Tanzania, Uwezo, EGRA, and EGMA assessments drew attention to low learning; in Kenya Uwezo results and baseline assessment for an education pilot project similarly achieved attention and action for learning. Drawing on other examples, an EGRA assessment conducted in partnership with the government in Nicaragua spurred immediate actions to improve learning, while an EGRA assessment in Senegal that was implemented mostly independently was used primarily by civil society to draw attention to low learning.
Of course, new information on learning does not always drive attention or action. The use of such assessments may be more politically feasible or desirable at the beginning of a new government than in the middle of an administration. In some places the effectiveness of such assessments to drive policy priorities may rely on the presence and prominence of civil society actors that can use the results to advocate for change, which relates to the next point.
Domestic think tanks and evidence-informed, action-oriented civil society actors can have an outsized influence on what politicians prioritize and delegate. These entities create domestically-relevant research and knowledge, develop and maintain ongoing relationships with government actors, and can advocate for reform from within a country.
Pratham, and their associated ASER assessments, have successfully driven attention and action on foundational learning in India. Central Square Foundation (also in India), in addition to implementing education programs, engages with education decision-makers and advocated for the focus on foundational learning in the new national education policy. The Centre for the Study of the Economies of Africa (CSEA), in Nigeria, provides not only research and analysis, but a forum for policy dialogue between government stakeholders, private sector, and civil society actors. SMERU, in Indonesia, similarly produces policy-relevant research and engages with the education ministry on improving learning. The INOVASI program works directly with Indonesia’s Education Ministry to implement education programs and has been partnering with and advocating for government officials to prioritize learning, and especially foundational skills, as schools reopen following the Covid-19 school closures. In Nicaragua, the success of EGRA results in bringing about political action to improve learning is in part attributed to a local NGO, CIASES, that RTI, the project lead, partnered with to implement the assessment. CIASES’ high-level connections with the ministry and ongoing presence and influence on the ground enabled the new information on learning to translate into government dedication and delegation of action.
A third way to support the dedication and delegation of education goals from high levels of an education system is to support future leaders today. Many of the successful examples above were driven by the dedication of a small set of leaders or bureaucrats who chose to prioritize and delegate learning goals.
Vanderbilt’s Graduate Program in Economic Development was established (in 1954) to provide students from developing countries training in economic development and has trained future finance ministers and heads of central banks. Developing or facilitating access to similar, high-quality programs for education could support future education leaders. Teach for All organizations “recruit promising leaders” and train and support them to teach in classrooms for at least two years. Through this experience in the classroom, and through cultivating ongoing connections for alumni, communities, and partner organizations, they seek to encourage young people to dedicate themselves to leadership in education. The Echidna Global Scholars Program, hosted at the Brookings Institution, selects leaders and scholars from developing countries, provides a residency at the Institution for them to conduct research on improving learning outcomes in their country of focus, and provides support for them to take action when they return to their country. The World Bank runs an Africa Fellowship Program for young African scholars, a model which could be tailored for those focused on education. The African Leadership Academy trains young people with the goal of “[transforming] Africa by developing a powerful network of young leaders who will work together to address Africa’s greatest challenges”. These types of efforts are undertaken with a long-term view, facilitating the dedication of current and future education leaders to children’s learning.
Improving education systems and learning outcomes is hard. There are no easy solutions, and there are few quick solutions. But change is possible, and the short list of success cases is slowly (too slowly) growing. There are steps international actors and funders can take to support not just “education interventions,” but shifts in education systems. These will take dedication and a willingness to invest in the long run. But such system shifts are necessary for the large-scale learning improvements that are needed to begin to ensure every child a quality education.
RISE blog posts and podcasts reflect the views of the authors and do not necessarily represent the views of the organisation or our funders.