Ensuring education policy translates into practice: the role of the middle tier in government

Teaching effectively is hard craft. To make it possible, teachers need instructional support and feedback. One of the biggest barriers to improvement is that in many systems teachers do not receive enough support. Where there are enough district-level coaches or pedagogical advisors, teachers may receive regular visits and support. In places where that coaching layer is thin or nonexistent, the responsibility to guide and improve instruction often falls on school-based staff themselves.

The frequency and quality of teacher visits depend not only on the commitment of coaches but also on how they are managed and incentivised by the system above them.

We refer to this management layer – typically at the district, sub-district, or block level – as the middle tier of education systems. These officials can play a decisive role in translating policy into practice, closing large and persistent gaps. Through their management of coaches and mentors, they determine whether foundational literacy and numeracy programmes succeed at scale. They ensure that instructional support reaches classrooms, that mentors focus on the right pedagogical priorities, and that reforms stay on track. Yet there is surprisingly little research on how best to strengthen this layer to manage for learning outcomes – by motivating coaches to visit schools more often, prioritise struggling schools, and focus their efforts where they are needed most.

With this in mind, we are delighted to launch the Middle Tier Implementation Research Initiative, a partnership between Brink, the What Works Hub for Global Education, and the Gates Foundation which aims to bring the middle tier into focus. 16 organisations across Sub-Saharan Africa and India proposed new ideas for research and experimentation. Their collective message was clear: engaging this middle tier is essential for sustainability and scale in the government system.

Our goal with this initiative is to support and study how to enable the middle tier of government systems. We aim to bring the best of implementation research to the topic. This includes tweaks and tests, with the goal of better supporting the middle tier and systematically codifying lessons learned for the sector.

Beyond coaching: a wider lens on the middle tier

Much of the research on instruction has focused on externally managed programmes rather than on understanding the change management required within government systems to sustain a focus on learning. We believe it is important to study what enables effective management of coaching and instructional support within these systems.

For coaching to be effective, it must be grounded in timely, actionable data and embedded in systems that motivate new behaviours. Sustained improvement requires shifting officials’ roles from administrative oversight to instructional leadership, while acknowledging the complexity of their mandates.

With this in mind, we have identified three priority areas for implementation research:

1. Delivering instructional support
2. Data that drives action
3. Norms, behaviours and incentives that stick.

Delivering instructional support

Four of the sixteen proposals are exploring ways of improving the delivery of instructional support through the middle tier. Scaling coaching has long been challenging: effective, but costly. Instead of replicating gold-standard models, these ideas adapt coaching functions to be more feasible and cost-effective. Organisations proposed hybrid blends of in-person and remote mentoring, WhatsApp nudges to keep momentum going, narrowing pedagogical focus so coaches are not spread too thin, or adding routines like teacher debriefs after observations. Together, these approaches are asking: what is the sweet spot between cost-effectiveness, quality, and sustainability in coaching?

Data that drives action

Another important theme was actionable data, which we saw across five of the sixteen proposals. Collection and use of data is being reimagined — not as reports filed upwards, but as tools for problem-solving and smart targeting. Some teams proposed testing dashboards that flag struggling teachers and schools, tools that sharpen supervisors’ ability to tailor support, and systems that guide where limited coaching resources should go. Others are asking what truly counts as ‘actionable data’ and designing job aids that make insights usable in real time. A few explored simplified guidance documents or even visual data (photos, voice notes, videos) to sharpen feedback loops.

Norms, behaviours and incentives that stick

Nearly half of the proposals focused on norms, behaviours, and incentives, probing what could shift middle-tier officials from ticking compliance boxes to leading instructional improvement. Organisations proposed experimenting with approaches such as job descriptions that clarify expectations, routines that make feedback part of the day-to-day, practical tools like observation templates, and support systems that embed practice even under pressure. Think of it as redesigning the scaffolding around their role so it nudges them toward becoming mentors rather than monitors. Some proposals aimed to simplify complex ‘management cycles’ into realistic routines that officials can sustain. Others planned to co-design planning tools directly with officials. A few suggested testing tweaks like structured review meetings to strengthen officials’ instructional leadership.

India and Sub-Saharan Africa: different priorities, shared principles

This call for proposals covered India and Sub-Saharan Africa, which provided an opportunity to see how the role of the middle tier is being framed across a wide range of contexts. Applicants in India leaned mostly into questions of norms, behaviours and incentives, probing how institutional routines might shift officials beyond compliance toward genuine instructional leadership. In Sub-Saharan Africa, proposals more often centered on actionable data, reflecting the urgency of timely, usable insights to guide decision making.

Yet across both contexts, a set of deeper principles emerged — the real ‘secret sauce’ of middle tier reform. The strongest proposals went beyond co-creation to reflect genuine government leadership and alignment. They were anchored in district, state or national priorities, backed by letters of support, and in all cases, brought government officials into the selection process to participate in deeper dives on the proposed research. These proposals signaled government demand, not just ideas designed for them.

The power of simplicity was salient. In systems overloaded with competing demands, lightweight routines and easy-to-use tools were seen as far more likely to take root.

And finally, there was a shared recognition that middle tier officials do not operate in isolation, and that the context they operate in shapes their behaviour. The institutional conditions around them, from incentive and accountability systems to administrative routines and resource flows, can both enable and constrain their work, shaping what is possible and determining whether solutions endure.

None of this is neat. But in that messiness lies the opportunity. Implementation research gives us a way to test, adapt and learn in real time, grounding evidence in what middle-tier officials can actually do within their daily constraints.

As the first cohort of five grantees begins its journey, we are eager to learn in real time and share insights as they emerge. The big questions are clear: Will new routines stick? Do they strengthen support for schools? Can they scale affordably? And will they embed into systems for the long term?  If these proposals are any indication, the middle tier is finally shifting from the margins to the centre of the conversation on how to deliver foundational learning at scale.