From Nanyuki to New York: Co-Designing the Whole Child Model

4 days ago

8 min read

Reimagining Education Through Trust, Collaboration and Collective Action to Create an Education Ecosystem Where Every Child Can Thrive

Kenya is a nation of runners. The world knows it, and we are proud of it. Running rewards personal discipline, personal grit, and personal glory, and that instinct shows up in how we build. Every founder runs their own race, and Kibera shows what that looks like at scale. Some reports put it among the highest NGO-to-population ratios in Africa, with organisation after organisation running in its own lane, and yet homes are still crammed, youth still sit idle by the roadside, and people still live on a dollar or two a day. We are good at founding organisations and tackling a problem alone. We are much less practised at passing the baton, the one moment in a relay where nobody wins alone. When good work happens in parallel instead of together, someone pays for it, usually the people the work was meant to serve.

Education has the same problem, and the stakes are high. The 2024 national assessment of Grade 3 learners found that 73.2% had not reached minimum proficiency in numeracy and 58.9% had not reached it in literacy. As Romana writes in Weaving Wholeness, education is "firstly and fundamentally a systemic and structural challenge of inequity and oppression," and that calls for an intersectional, interdisciplinary approach. No single organisation can carry that alone. Kania and Kramer made the same point about social change in general: large-scale change needs broad cross-sector coordination, yet the sector keeps funding isolated interventions. At Metis, we call the alternative Go Further Together.

Trust comes before collaboration

Our partnership with Transcend Education and Tenacre on the Whole Child Model is that belief in practice, and it rests on something that takes years to build. When people come together only because funding is available, the collaboration can work against the very work it is meant to serve. We saw this during the Atlassian Foundation call for proposals, when organisations scrambled to find partners as the deadline approached, all hoping to make a big impact in education in the age of AI. Research on nonprofit partnerships says the same. Nonprofit Quarterly's analysis of why collaborations fail finds that funder-driven collaborations, and those formed mainly to raise income, often lack a purpose strong enough to hold the group together. A 2014 study by Sibbald, Tetroe and Graham in Implementation Science looked at partnerships required by a Canadian health research funder, and some participants described the process as a game, not meaningful collaboration.

Trust has to be built long before the funding call. Bryk and Schneider's study of twelve Chicago elementary schools found that schools with high relational trust were far more likely to improve in reading and mathematics than low-trust schools, roughly a one in two chance against one in seven. Trust is a working condition for results. Transcend, Tenacre, and Metis have built ours over months, through dinners, games, and figuring out logistics together. Along the way, six words emerged to hold the partnership together: Compass, Consensus, Curiosity, Calibrate, Critique, and Collective.

The six Cs are how that trust gets practised day to day, and each one answers a way collaboration usually breaks. Compass gives partners one shared north star, the flourishing of every child, so nobody is quietly pursuing a separate agenda. Kania and Kramer name a common agenda as the first condition of collective impact. Consensus names the norms before conflict arrives, so disagreement has somewhere to go. Curiosity and Calibrate matter because this is an adaptive challenge in Heifetz's sense: nobody holds the answer, so partners have to say what they don't know and change course together. Critique keeps the trust honest. Bryk and Schneider found that relational trust rests on respect, competence, integrity, and personal regard, and integrity includes saying what could be better. Collective is where it lands, many voices working one puzzle, the opposite of running alone.

The same conditions hold a classroom together

If trust is what holds a partnership together, we wanted to see whether it does the same work in a school. In September, Muthoni and Naftaly visited six schools in New Jersey, Washington DC, and Brooklyn, four of them implementing Transcend's Whole Child Model, to see it in practice rather than on paper. What they found lines up with a large body of research. The Learning Policy Institute's review of the science of learning and development concludes that human relationships are the essential ingredient for healthy development and learning, and that children need a sense of safety and belonging to thrive in school. We saw that in practice, and it sounded a lot like our six Cs: a shared purpose, clear norms, honest feedback, and many people carrying the work.

Whole-child learning runs on people, trust, and routines, and budgets matter less.

A US classroom carries 30 learners and two teachers. A Kenyan one can hold 40 to 150 with a single teacher. But handwritten visuals, daily routines, and a classroom arranged for belonging cost almost nothing, and our task is to flip the deficit script into the strengths already in our communities. It is easy to look at large classes as a constraint rather than an asset, especially in our context where a classroom's capacity could stretch up to 150 students, until they're reframed as an abundance of co-teachers. The furthest-along schools showed where that leads: learners who've lived the model for three or four years start leading parts of it themselves.

Structure is the container for warmth.

When we set out to build school communities where learners feel safe, loved, valued, and successful, it is easy to assume that whole-child means soft, and that warmth comes at the cost of structure. In every school we visited, children knew exactly where the line was, and neither had to be sacrificed for the other. The strongest sites held both at once. Rituals were tight enough that "they keep time here" was the first thing visitors were told, and warmth was soft enough that a coach met each child at the door, centred them with a breath, and told them she loved them. The same work looked different from school to school. At Paul Robeson, Strong Start gives every day a greeting, movement, and a goal set before the day begins. At Garrison, a ball rolls from learner to learner with a name attached to every pass. Neither costs anything but time. Clear structure frees learners, it does not suppress them.

The adults matter as much as the model.

The model is only as strong as the people carrying it, and these schools invested in their adults the way they invested in their children. Coaching was affirming yet rigorous: coaches named bright spots, held each other accountable, and called feedback "love and a gift." Leadership was spread across the school. Teachers rotated through leading staff meetings, design teams and teacher champions trained their peers, and demo classrooms let teachers learn from teachers. Leaders stayed in the trenches, modelling practices and making the model easy to adopt.

Watching Racheal, a school coach, made the role tangible. She led with the science, so buy-in came from understanding, not instruction. She named the payoff before asking for the work, so teachers knew the effort would mean less workload, not more. She gathered practices across schools so teachers could learn from each other, not just from her, and paced her coaching to each teacher's progress, never running one script for everyone.

Set the timeline honestly.

"You are seeing Year 5, not Year 1," veteran teachers warned. Start with a few practices, learn them well, and tell schools, funders, and the ministry that culture takes years to shift.

Two alternative models sharpened our questions. Ember School dismantles power dynamics on purpose. Teachers are called "brothers" and "sisters," and learners are treated as contributing as validly as the adults in the room. To an observer trained in order, it looked chaotic at first glance, yet deep learning and real ownership were visible underneath. It raises a question for us: how do we loosen the fear-based hierarchies inherited from colonial education while keeping a classroom safe? Lead School pairs rigorous project-based learning with strong academic results, which points to a hybrid that honours community expectations of excellence while shifting the pedagogy underneath.

What the global convenings added

At the Global School Forum, Education House, and Remake Learning in Pittsburgh, one message kept resurfacing: change lasts when it is co-owned, measured honestly, and held by a system.

Technical versus adaptive challenges.

At Education House, a framework from Harvard's Ronald Heifetz gave us language for something we already knew. Technical problems have clear definitions and known solutions that experts can deliver, like engineering a bridge. Adaptive challenges, like tribalism, peacebuilding, and climate change, require the people involved to learn and change, because no existing answer fits. Kenyan education is built heavily for the first kind. The Whole Child Model prepares learners for the second.

The 20-year overnight success.

Remake Learning began in 2007 with a handful of people. It now counts hundreds of member organisations, and Education Week reports that it has inspired more than 40 similar ecosystems around the world. Ecosystem work takes decades of consistency, and it does not push a single model.

Measurement as a habit.

Speakers described countries that treat assessment as a routine tool for adjusting teacher support. In Sierra Leone, we heard, implementation research shaped teacher training and participation climbed from roughly 30% to 80–85%. Kenya has its own evidence. Tusome, the national early grade reading programme, gave coaches tablets so classroom observation data could guide support to teachers in real time. Independent evaluators tracked reading with the Early Grade Reading Assessment at baseline, midline, and endline, and found significant gains on every subtask in English and Kiswahili. Grade 2 English gains were roughly equal to an extra year of schooling. The model had first been tested in a randomised pilot, and the effect sizes held when it went national. Many learners were still reading below grade level, which is the point of measuring: it shows where support has to go next. Measurement works when it feeds straight back into what we do for teachers.

Islands of excellence in a sea of fragmentation.

Pilots show what is possible, and only systems reach every child.

Where we go from here

A Kenya where learners feel safe, loved, valued, and successful is within reach. The Competency-Based Curriculum (CBC) already names this vision, and the gap is in how teachers are equipped to live it every day. We have been doing this work since 2018. New York was a chance to learn from Transcend, which has codified its model in ways we want to adapt as we work to codify our own. Over the next year and a half, we will start small, rooted in real classroom challenges. We will build lighthouse schools where teachers learn from teachers, and partner with the Ministry, KICD, and TSC on trust, evidence, and patience.

The six Cs that hold our partnership together are the same conditions we hope schools will build for their own learners. It starts with us.