Reform efforts can achieve stronger results in teaching and learning when implementation is treated as a process of learning, adaptation and shared problem solving.

Authors: Neelofar Javaid and Emma Cameron
This blog was originally published on the GPE Education for All blog on July 7, 2026, here.

Education systems don’t struggle because they lack plans or policies. They struggle because those plans don’t translate into real change in classrooms.

Across many countries, reforms set ambitious goals but stall at the point of implementation. This gap between policy and practice is especially stark in low- and middle-income countries where complex systems, limited resources and diverse local contexts make change difficult to deliver.

Increasingly, the challenge is not just identifying what works but understanding how to make it work in practice.

But this challenge is not insurmountable.

A growing body of evidence, including insights from the Learning Generation Initiative’s Problem-Solving in Education Reform: Evidence and Insights for Low- and Middle-Income Countries, points to a broader shift: rather than asking, “which reform should we implement?” the more important question is: “how do systems learn what works, for whom and under which conditions?”

Here, the term ‘systems’ refers to the wider education ecosystem: the actors, institutions, processes, routines and relationships that shape reform.

This systems focus is already at the center of how partners such as GPE support reform, strengthening the capacity of education systems to use data, empowering middle-tier actors and continuously learning from and adapting implementation.

In this framing, problem solving is not an alternative to delivery, but a way to strengthen effective implementation at scale—helping systems translate reforms into real improvements in teaching and learning, and building the foundations for more adaptive and resilient education systems over time.

Why problem solving works for education policy and reform

Traditional education reforms often rely on delivery- and accountability-driven approaches: setting targets, monitoring performance and using incentives or sanctions to drive results. These approaches assume the right solutions are already known and the task is to implement them effectively.

In some cases, this can improve effort and short-term performance. In complex systems however, problems are not always fully understood and solutions rarely work uniformly across local contexts.

In contrast, problem-solving approaches to policy and reform focus on helping education system actors identify challenges, test solutions, learn from implementation and adapt over time. They start from a different premise than delivery approaches:

Solutions must often be discovered through implementation. This requires shifting from fixed plans to adaptation, from enforcing compliance to enabling learning and from centralized control to greater, local problem solving.

In practice, this requires moving away from top-down practices focused on implementing policy as designed to an approach that answers the question: how do we support those closest to the problem to solve it?

It reframes reform as something dynamic and iterative—a continuous process of learning and adaptation. And in doing so, it strengthens how reforms are implemented, helping systems to adapt to local challenges and to translate policy into improvements in teaching and learning at scale.

Problem solving in practice

Successful approaches to problem solving in education tend to share a few core features:

  • Shared ownership of problems where governments, districts, schools and communities work together to identify and address challenges. In Ghana, stakeholder forums created space for district officials, school leaders and communities to identify local implementation challenges—from school performance gaps to teacher absenteeism and resource constraints—to then develop solutions tailored to their specific contexts.
  • Use of data for learning to guide action in improving education, going beyond performance monitoring. The municipality of Sobral, Brazil transformed its education system within a decade, rising from being ranked 1,366th nationally in terms of education quality to first place in the country. What changed was how policies were implemented. Leaders used data to identify problems, engaged stakeholders across the system and continuously adapted their approach at municipal, school and classroom levels based on feedback from assessments of student learning (student learning increased from 34% and 49% to 89% and 92% for six- and seven-year-olds, respectively). Sobral’s experience highlights a broader lesson: transformation is possible when implementation matters as much as design.
  • Iteration and adaptation in approach so that systems can continuously refine how they work based on feedback and real-world experience. In South Africa’s Funda Wande program, teacher training materials were modified based on early implementation feedback from teachers which then led to training videos being reshot so that they reflected real classroom conditions and hence better support for teachers.
  • Empowered middle-tier and frontline actors so that those closest to schools and classrooms are equipped and supported to respond to challenges that arise, meaning they have the capability, agency and autonomy to respond effectively to local challenges. In Kenya’s Tusome program, teachers were provided structured materials alongside the flexibility to adjust them, strengthening both ownership and effectiveness. In Pakistan, a national delivery reform initially struggled to translate high-level goals into improvements at the school level. It was only when sustained engagement from senior leadership helped reinforce ownership and problem solving across the system that the reform gained traction.

These elements reflect a broader shift towards building systems that can learn and adapt. Education systems are complex and adaptive, shaped by human behaviors, incentives and local context.

In such environments, solutions cannot always be designed in advance or applied uniformly. Problem-solving approaches in education work because they align reform with how education systems actually function, creating the conditions for continuous learning and adaptation, making reform more responsive, effective and sustainable.

What can change depends on a willingness to learn

Education reform is often treated as a technical challenge: design the right policy and results will follow. But the evidence suggests something different. Reform is how problem solving is built into systems as an integral part of implementation.

While the evidence on problem solving is promising, important gaps remain tied to:

  • Sustainability: Can problem-solving approaches be sustained over time, especially through political change?
  • Cost and scalability: How can reforms be effectively scaled in resource-constrained systems?
  • How collaboration works in practice: Which forms of collaboration and leadership are most critical for success?
  • Inclusion and equity: How do gender, inclusion and diverse perspectives shape problem-solving processes and outcomes?

For policy makers and practitioners, the implication is clear: improving education systems requires investing in the capacity to implement reforms effectively, adapt to local challenges and continuously improve over time.

Systems that succeed are not those with the best plans, but those that can evolve as challenges emerge.

Read also the previous blog on delivery by the Learning Generation Initiative.

Photo: Students from Class 8 take part in a maths class at Marble Quarry Primary School in Kajiado Central on the outskirts of Nairobi, Kenya.
Credit: GPE/Luis Tato