Assessing non-formal education provision through a system-wide lens on coherence: Evidence from the Open School programme in Tanzania

Only around one-third of young people in Tanzania complete formal secondary schooling, lower than the sub-Saharan regional average (UNICEF, 2018). Not only is access a factor, but so is education quality. Retention rates show that less than eight percent of students transition from lower to upper secondary and a mere six percent complete the full education cycle (URT, 2025). For girls in poor and rural areas, completing secondary education is even more of a challenge (Ciampi et al., 2024). These figures underscore significant structural failures in Tanzania’s secondary education system and, in response, the government has turned to non-formal education to help improve access and quality at this level of schooling. Specifically, it has turned to the Open School programme, a non-formal education alternative to gaining secondary education qualifications for out-of-school youth. The Open School programme has existed informally for decades in Tanzania and is run primarily by non-state actors out of community schools. In 2013, the programme was formalized under national reforms to help expand access to secondary education. In 2022, with support from the World Bank, further reforms targeted the programme as part of the Secondary Education Quality Improvement Project (SEQUIP). SEQUIP is a large-scale government effort to strengthen quality in the whole of the secondary education system; for the Open School programme, its aim is to boost access to accelerated learning in order to facilitate the mainstreaming of young girls previously banned from secondary education back into formal schools. The use of non-formal education programmes like Open School has been a common response to the learning crisis, where governments and aid partners alike have looked to the potential and flexibility of these interventions to help students who remain marginalised from learning. However, research on these programmes rarely includes system-wide assessments of coherence, where evidence can point to how well the programmes are “tuned to deliver” learning (Komba & Shukia, 2019). To fill this gap, the purpose of this study is to assess coherence in the Open School programme; to understand how its system components interact in the provision of learning and to determine how, if in any way, SEQUIP reforms have affected coherence. The study adapts the Levy-Walton (2013) Domains of Power Framework and the RISE Accountability Matrix (Pritchett, 2015) to produce a combined Domains of Accountability Framework that helps identify and diagnose coherence in Open School provision. It employs a qualitative approach to engage with coherence across three levels of governance: the first level examines policy, unpacking the political and sectoral dynamics that inform learning objectives and service rules; the second examines institutional arrangements, honing in on the delegation of actor roles and responsibilities for the programme’s provision; the third level looks at implementation, assessing how provision is managed and delivered in schools. Data collection includes a review of 22 policy documents related to the Open School programme published between 1995–2025. These, coupled with a stakeholder mapping exercise, help identify the formal distribution of provision rules. In order to compare formal rules to their enactment on the ground, 70 key informant interviews were also conducted. Key informants are middle-tier implementation actors including education officials, school managers, and teachers from 20 open schools in two regions in Tanzania. Key informants were selected for their ability to provide in-depth, ground-level insights into Open School practices. The regions and schools were selected using purposive sampling; the former due to their high concentration of open schools and the latter to include a variation of the different types of open schools in Tanzania: government-owned open schools, private open schools, and unregistered (i.e. legally unrecognised) open schools. The study argues that misalignments in policy aims, actor arrangements, and implementation practices create incoherence at all levels of Open School provision. At the policy level, it finds that objectives are inconsistent due to elite interests that are not well-oriented towards non-formal education priorities; the programme has historically been subordinate to formal education needs, inheriting deep institutional instability and poorly defined goals for the benefit of marginalised learners. At the institutional level, the study finds that roles are poorly enacted despite many explicitly defined directives. This stems from delegation that ignores bureaucratic capacity and from the weak integration of the formal and non-formal sectors. At the implementation level, management and accountability are also weak as a result of uneven information and support mechanisms, as well as from the poor monitoring of learning. On the ground, Open School provision relies less on the coherent execution of policy rules and more on the resources and capacity of school actors who often mitigate the system’s incoherence with innovative adaptations that compensate misalignments to support student learning. The study also argues that SEQUIP reforms have aggravated provision incoherences. At the policy level, they have further cemented formal education concerns in programme objectives, repackaging it as a tool to mainstream students who have dropped out of the formal system. It has enforced accelerated stages for learning without adjusting the programme’s curriculum or assessments—diminishing the programme’s long-standing flexibility and ignoring the learning deficits with which students arrive to the programme. At the institutional level, SEQUIP reforms have layered new formal education actors onto the chain of Open School provision without considering existing bureaucratic misalignments, diluting the programme’s enforceability and accountability on the ground. At the implementation level, SEQUIP subsidies for girls’ have not included gender-specific programme support, placing additional service burdens on schools who often address learning gaps with their own resources. The contributions of this thesis are significant, producing a multidimensional understanding of how policymaking, governance, and accountability interact in the provision of the Open School programme in Tanzania. It contributes with a unique research design that allows for a deep understanding of coherence in non-formal education provision, and of how formal and informal actor interactions contribute to that coherence. It also illustrates a need to assess non-formal education programmes using a political economy lens, where the objectives of their design and implementation can be measured against politically calibrated elite interests. Lastly, it highlights the importance of middle-tier actors whose knowledge, expertise, and capacity are often better aligned to learning than what state reforms account for. This is particularly useful with regard to scaling non-formal education programmes through aid partnerships, where resources can be better optimised by recognising, supporting, and catalysing localised strategies that help deliver learning in more meaningful ways.

Name dc:type.qualificationname

Doctor of Philosophy (PhD)

Level dc:type.qualificationlevel

Doctoral

Grantor dc:publisher.institution

University of Cambridge

Year dc:date.issued

2026

Author dc:creator

• Guzman Correa, Regina

Advisor dc:contributor.advisor

• Rose, Pauline

• Non-formal education
• Alternative education
• System coherence
• Systems thinking
• Political economy analysis
• Policy analysis
• The politics of education
• Tanzania
• Aid
• Middle-tier capacity
• Policy implementation
• Implementation accountability
• Out-of-school children
• Learning crisis

Licence

• https://www.repository.cam.ac.uk/bitstreams/96a1f66f-70aa-48b1-97c6-40807fda6564/download
• http://purl.org/NET/rdflicense/allrightsreserved

Language dc:language

eng

DOI dc:identifier.doi

https://doi.org/10.17863/CAM.132268

OAI identifier oai:identifier

oai:www.repository.cam.ac.uk:1810/406138