The birthing centre in Gamgadhi, the small district headquarters of Mugu in upper Karnali, is reached by a Twin Otter that flies only when the weather behaves. On the morning we visited in March, three women were waiting for postnatal checks in a corrugated-roof annex that had been added the previous year using funds from a programme almost none of them had heard of. The annex cost about £18,000. The programme that paid for it is run from a four-floor office building in Kathmandu’s Sanepa neighbourhood. The money originated in a budget line approved in Whitehall.
This is one project, in one district, paying for one room. The number of organisations that touch the cash between London and Gamgadhi is seven.
The donor: FCDO, London
The starting point is the Foreign, Commonwealth and Development Office’s Nepal Health Sector Programme III, a £42 million bilateral commitment running 2022 through 2027. It is one of the largest single-donor health investments in the country. The programme document, lodged in the IATI registry under activity identifier GB-GOV-1-301077, lists three outputs: improved maternal and neonatal health services in underserved districts, strengthened federal-to-local health planning, and a results-based financing pilot covering twelve districts.
FCDO does not implement. It contracts. The bulk of the programme is delivered through a managing agent — in this case, a consortium led by a London-headquartered consultancy that won the competitive tender in late 2021. The managing agent’s job is to translate the high-level programme logic into actual contracts with implementers on the ground.
The managing agent: a London consultancy
The managing agent runs a programme management unit out of Sanepa with roughly twenty-eight staff, most of them Nepali, a handful of expatriate technical leads. Its monthly burn is a non-trivial slice of the headline £42 million — overheads in this kind of programme generally run between fifteen and twenty-two per cent depending on how aggressively the donor presses on value for money. The remainder flows out as sub-grants and service contracts.
For maternal health in Karnali, the unit issued three downstream contracts. One went to UNICEF for cold-chain and equipment procurement. One went to a national NGO consortium for community demand-generation work. One — the one funding our birthing centre annex — went to a Nepal-registered INGO for direct service-strengthening in five Karnali districts.
The implementing INGO
The INGO in question has worked in Nepal since the late 1990s and maintains a Karnali field office in Surkhet, three days’ drive from Kathmandu if the highway is open and the rivers are behaving. Its Karnali team is fourteen people, including two midwifery technical advisors, four district coordinators, and a small finance unit. The contract with the managing agent is worth roughly £2.1 million over three years.
The INGO does not, itself, run the birthing centre. The birthing centre is a government facility, operated by the Mugu District Hospital under the Ministry of Health and Population’s federal structure, with day-to-day oversight from the Mugu municipality health section. The INGO’s role is to fund specific gaps — equipment, training, the occasional infrastructure top-up like our £18,000 annex — that the government budget chronically does not cover.
The local partner
Below the INGO is one more layer: a Karnali-registered NGO that the INGO uses as its formal local partner for community-level work. The use of a local partner is a contractual requirement under the FCDO programme’s localisation commitments — at least thirty per cent of programme spend is meant to flow through Nepali organisations. The local partner runs the female community health volunteer engagement, the radio messaging in local languages, and the household-level outreach that produces, eventually, the women who arrive at the birthing centre.
The local partner’s contract for Mugu alone is worth about NPR 7.2 million annually — roughly £41,000. That sum supports a team of six and underwrites a network of about sixty volunteers paid a modest stipend per outreach event.
The government counterpart
The annex itself sits on land owned by the federal government, on the compound of a facility run by the provincial health ministry, with operational responsibility devolved to the municipality. Three levels of Nepali state are involved before the question of who actually staffs the room can even be answered. In practice, the auxiliary nurse-midwife running postnatal care that morning is on the municipal payroll; her in-service training was paid for by the INGO; the cold chain in the next room was procured by UNICEF; the building she stood in was built with FCDO money.
This is not unusual. It is the standard structure of a Nepal health intervention in 2026.
What gets lost in the handoffs
Every one of the seven handoffs — donor, managing agent, three sub-contractors, INGO, local partner, government — is governed by its own reporting cycle, its own audit schedule, its own results framework. The friction is real. A senior INGO finance officer estimated, in conversation, that roughly seven to nine per cent of contract value disappears into reconciliation work that produces no actual programmatic output. The donor’s auditors would call that compliance; the local partner’s staff call it Tuesday.
What does not get lost, when the chain functions, is the room. The annex is real, the cold chain is stocked, and the auxiliary nurse-midwife showed up to work that morning. That this requires seven institutions, three currencies, and a six-step contractual relay to achieve is a feature of the system, not a failure of it. The question that the localisation conversation in 2026 is meant to ask — and largely is not asking — is whether the same room could be delivered with three institutions instead of seven, and what would have to change in the donor’s risk appetite for that to be possible.
What to watch
FCDO’s Nepal Health Sector Programme is up for mid-term review later this year. The review terms of reference, circulated quietly to managing agents in April, ask explicitly about the cost of the intermediary layer and whether direct district funding — bypassing both the managing agent and the INGO — could be piloted in one or two of the twelve focus districts. If it is, Mugu is on the shortlist. The next annex may, in that case, be built with substantially fewer handoffs. Whether the room it produces is any better is the question worth holding open.