By David Nkrumah Boateng, Country Programme Lead, PEAS Ghana
What does it actually take to improve education systems at scale?
I kept hearing versions of this question at CIES 2026 in San Francisco – in panels, over coffee, and in the corridors between sessions. The sector is converging on something transformative, even though the pathways look different. One theme stood out consistently: if we want to improve learning outcomes at scale, we have to start with the system.
Working with the middle-tier
At CIES, I joined partners from Dignitas and Leadership for Equity, with Gates Foundation as a discussant. We explored how middle-tier actors – inspectors, district officials, coaches – can shift from traditional compliance driven roles to becoming genuine drivers of change.
These roles are often overlooked, but they sit at the centre of how education systems actually function. They are the link between policy and practice: the people who support schools, guide teachers, and hold the line on standards.
It no longer feels like a question of the middle-tier’s importance: that feels broadly understood. The real focus now is on how we support these actors to fulfil their roles effectively.
One of the clearest lessons from our work at PEAS is that we cannot do this in isolation. This means working alongside government from the outset: identifying problems, refining approaches, and building routines together with middle-tier actors. Too many interventions fail not because the idea was wrong, but because sustainability was an afterthought. We know from decades of evidence that bolt-on sustainability doesn’t work. In our experience, this means designing for adoption and endurance from the outset, rather than trying to retrofit it later.
In Uganda, PEAS has worked closely with the Directorate of Education Standards through the Inspect and Improve programme. Together, we co-designed inspection tools, trained inspectors, and changed how data feeds into school improvement. A digitised inspection prototype tool developed by PEAS reduced analysis time from one to five months to just a few days (NFER, 2022). The Ministry has since rolled out a digitised inspection tool nationally, informed by this approach, and PEAS is now supporting the development of complementary digital school improvement planning tools. Together, these shifts have the potential to transform how quickly decisions are made, how precisely support is targeted, and how fast schools can improve.
But tools alone do not get you there. Sustainable change depends on whether these approaches become part of how the system operates, not a project running alongside it. That takes alignment with government priorities, clarity on roles, and supporting people across the system to deliver their existing responsibilities more effectively – not adding parallel processes or additional layers of work.
Shifting how the sector thinks about scale
These lessons are not unique to Uganda. Beyond the panel, conversations at CIES pointed to a broader shift in how the sector thinks about what counts.
System delivery is getting more attention. Not just what happens inside individual schools, but the structures around them that determine whether good practice spreads or stays local.
There is also sharper scrutiny on what actually works. AI came up constantly, and the interest was real. But so were the concerns: data protection, equity, and whether any of these tools make sense in classrooms where electricity is unreliable and devices are shared between many students, especially in more remote and deprived areas.
On financing, the picture is familiar but intensifying. Tighter budgets, higher expectations, and growing demand for approaches that link funding directly to outcomes.
The thread running through all of it: improving education at scale demands approaches that hold up over time, across contexts, and within existing systems; not beside them.
The importance of skills, knowledge and mindset
Another takeaway that kept surfacing: none of this works without investing in skills, knowledge and mindset across the system – from school leaders and teachers to the middle-tier actors who support them.
This goes well beyond giving people better tools. It means creating the ongoing structures to support them to develop the skills and knowledge to do the core work successfully – whether that is lesson observations, coaching, mentoring, or giving teachers feedback that actually supports positive change in the classroom. These shifts in mindset and behaviour take time, but they need to be reinforced on an ongoing basis.
That is a fundamentally different proposition from compliance. It means moving from checking boxes to building practice. And it requires a shift in how these roles are understood – not just for middle-tier actors, but across the structures around schools that drive continued improvement.
Looking ahead
CIES brought together implementers, funders, and policy actors from across the education system. The approaches vary, but the framing is converging: more focus on systems and actors, more emphasis on sustainability. There are no simple answers. But there is growing clarity on what it takes, and growing willingness to do the harder, slower, more embedded work that real change requires.
For PEAS, these conversations confirm the direction we are already heading. Improving education outcomes at scale means working with governments, strengthening the systems that support schools, and building approaches designed to outlast us.