Middle Tier Initiative
Middle Tier Initiative: Backing the middle tier to make foundational learning reforms stick
The layer of the system that turns national reform into classroom practice is one of the most overlooked levers for learning at scale. The Middle Tier Initiative funds government-requested research to back it.
The challenge
District and sub-district education officials sit at the heart of how school systems actually function day to day, yet they remain one of the most overlooked levers for improving learning outcomes at scale. The Middle Tier Initiative is a multi-country implementation research fund backing this layer of the system across South Africa, Nigeria, Zambia, Namibia, and India. Seed-funded by the Bill & Melinda Gates Foundation, it funds five implementer-led research projects embedded within large-scale foundational literacy and numeracy (FLN) programmes, including Teaching at the Right Level (TaRL) in Zambia and structured pedagogy reforms in South Africa, Nigeria, and India. Each project was co-designed with and formally requested by government district counterparts, so the research questions are owned by the system from the outset.
Foundational learning reforms often falter not at the policy level but in the middle of the system, where district and sub-district officials translate national ambitions into classroom practice. Without investing in this layer, reforms risk being short-lived or uneven in their impact. The question the Middle Tier Initiative set out to answer is this: what tools, coaching models, data systems, and management routines do middle-tier officials actually need to drive reform at scale, and how do you generate that evidence without creating a parallel research process that sits outside their live programmes?
What was achieved
- Five implementer-led research projects funded and embedded across South Africa, Nigeria, Zambia, Namibia, and India, each formally requested by government district counterparts and sitting alongside large-scale FLN programmes (TaRL in Zambia; structured pedagogy in South Africa, Nigeria, and India).
- Every project co-designed with government from the outset, ensuring research questions are owned by the system rather than imposed on it.
- Brink has consistently received a Net Promoter Score of 9/10 from grantees where it has acted as grant agent, indicating 90% of the portfolio would recommend us.
- Milestone payments issued in 10 to 15 days on average, matching the pace implementation research requires.
- Learn and Adapt sessions and payment-contingent learning have shifted incentives from output delivery to honest reflection across the portfolio.
How we did it
Brink co-leads the Middle Tier Initiative in partnership with the What Works Hub for Global Education. Brink runs the grant-giving and technical assistance function; the What Works Hub brings research frameworks, implementation measurement tools, and global thought leadership.
- Designed the fund's RFP and application process to be government-facing from the start, requiring applicants to submit a letter of approval from their district government counterpart and using an early conversation to understand how each partner worked with government stakeholders.
- Used proposition testing with potential grantees to validate optimal grant size, grant duration, the model of technical assistance, and priority research topics before finalising fund design.
- Ran a request-for-proposals process with invite-only communications, virtual coffee chats, office hours, and a two-stage investment committee review to identify the five grantees.
- Delivered demand-led, market-driven technical assistance designed to sit alongside each implementer's live programme rather than create a parallel research process.
- Supported grantees to use methods including A/B testing, implementation fidelity measurement, mixed-methods qualitative research, and classroom observation.
- Built Learn and Adapt sessions into the workplan as structured reflection points at the end of milestone periods, creating space for honest learning across the portfolio.
- Designed milestone payments to be contingent on demonstrated learning rather than delivery of pre-specified outputs, creating a financial incentive for honest reflection.
- Collected direct grantee feedback through a Net Promoter Score mechanism, reviewed at portfolio level to improve fund management in real time.
