The Architecture Debate Is Missing Its Foundation
Everyone is talking about reforming global health. Almost no one is asking whether the states being handed responsibility can actually exercise it.
There is a lot of reform energy right now. The Lusaka Agenda. The Accra Reset. Five regional proposals commissioned by Wellcome. Africa CDC’s Health Security and Sovereignty Agenda (Africa CDC, 2025a; Africa CDC, 2025b). WHO’s Executive Board discussed architecture reform at its February 2026 meeting (UN Foundation, 2026). Geneva diplomats are meeting in informal groups. Frameworks multiply.
All of it carries the same refrain: sovereignty, self‑reliance, country ownership, domestic financing. The era of donor dependency is over. Countries must lead. The global level must support rather than dominate.
I’ve spent long enough inside health systems to know that sovereignty is not a declaration. It is a capability. The capability question — what it takes for a state to govern a health system — is almost entirely absent from the current debate.
How we got here
Development assistance for health fell 21% between 2024 and 2025, driven largely by a 67% collapse in US funding (IHME, 2024). IHME modelling suggests no rebound by 2030 (IHME, 2024). Countries like Gambia, Lesotho, and Mozambique have lost up to 16% of their total health spending at a stroke, and Nigeria lost more than 400 million dollars in absolute terms (IHME, 2024).
The World Health Assembly adopted its first resolution on health financing in fifteen years, co‑sponsored by more than twenty countries (UN Foundation, 2026). African heads of state convened in Addis Ababa, Accra, and New York to push health sovereignty and financing up the political agenda (Africa CDC, 2025b; Government of Ghana, 2025a). WHO Director‑General Tedros, speaking at the Africa Health Sovereignty Summit in August, put it plainly: Africa does not need charity. Africa needs fair terms (Government of Ghana, 2025a; Africa CDC, 2025b).
Fair terms and sovereignty require institutions to give them operational meaning. The reform debate has little to say about how those institutions get built.
What the debate has produced instead is a proliferation of principles. The Accra Reset calls for a new architecture anchored in sovereignty, workability, and shared value creation (Government of Ghana, 2025b; Spark Street Advisors, 2025). The Wellcome‑commissioned regional proposals converge on shifting power toward countries and regions most affected by health challenges (hera, 2025). Africa CDC’s sovereignty agenda is organised around a clear division of labour: countries lead, regions coordinate, the global level supports (Africa CDC, 2025b; Africa CDC, 2025a). These are the right principles. The question is what sits underneath them.
The capability gap
WHO’s own analysis has found that up to 13% of health budgets in low‑ and middle‑income countries go unspent each year because public financial management systems are too weak to deploy them (WHO, 2025). The problem is not a lack of money. The institutional machinery to use it does not exist.
If countries cannot spend what they already have, the transition to domestic financing is not primarily a revenue problem. It is a state capacity problem. No amount of global architecture will solve it.
The Centre for Global Development has pointed out for years that the current architecture dominates national institutions in ways that are sometimes subtle but often powerful — priorities set in Geneva, Washington, London, and Seattle; countries asked to co‑finance donor‑funded programmes rather than the other way around (Glassman & Kenny, 2024). Decades of vertical programmes running through parallel systems, analytics outsourced to donor‑funded consultants, regulatory decisions shaped by major funders — all of this filled gaps in the short term while systematically widening them over time (Glassman & Kenny, 2024; hera, 2025).
The Lusaka Agenda identified this problem. Its five shifts are the right framework (African Constituency Bureau, 2025; Frymus, 2026). But as its own analysts have acknowledged, the agenda is thin on country‑level mechanics (hera, 2025). How do you build public financial management systems, regulatory frameworks, strategic purchasing machinery, and analytical capabilities, without which the five shifts remain aspirations?
The Global Financing Facility, for all its alignment with national health priorities, failed in most cases to mobilise increased domestic resources — its structures substituted for public expenditure capacity rather than building it (Spark Street Advisors, 2025; hera, 2025). Many results‑based financing programmes collapsed after project closure because they were designed around donor cycles rather than government systems (Glassman & Kenny, 2024).
We have been here before.
The New Deal for Engagement in Fragile States made state‑building central to the donor agenda on paper. In practice, funding flows, fiduciary rules, and short political timelines all pushed in the opposite direction (Spark Street Advisors, 2025).
Commitments to use country systems ran into donor risk aversion. Pledges to share risk expired at the end of the funding cycle. Over time, the New Deal drifted toward what it had always been most comfortable with: matrices, indicators, and compacts — strong on process and weak on the deeper shifts in power and practice that real institution‑building requires (Spark Street Advisors, 2025).
The lesson for the current debate is direct. Adding state‑building language to reform frameworks changes nothing if the financing modalities, risk arrangements, and incentive structures all continue to work in the opposite direction. You cannot build state capacity by writing about it while the operational architecture continues to bypass the state.
Three transitions, one missing ingredient
Three pressures are hitting simultaneously, each demanding strong state capacity and confronting its absence in most of the countries that need it most.
The first is the domestic financing transition. Everyone agrees this is the direction of travel (IHME, 2024; African Constituency Bureau, 2025). But raising domestic revenue is only the first step. The rest — pooling resources equitably, purchasing services strategically, managing provider performance, enforcing contribution compliance, processing claims, auditing expenditure, and connecting all of it to health outcomes — requires fiscal and administrative capabilities that were neglected during the aid era (African Constituency Bureau, 2025; hera, 2025). The transition to domestic financing does not just need more money. It needs a different kind of state.
The second is the governance of mixed health systems. Private providers deliver a large share of services in most low‑ and middle‑income countries, often the majority in urban areas, with minimal effective regulation (hera, 2025). Setting the rules, enforcing them, and using purchasing power to shape private sector behaviour toward equity and quality requires legal authority, regulatory infrastructure, and teams with skills in contract management, market analysis, and enforcement (hera, 2025). Most health ministries have none of these in sufficient depth. As domestic financing grows and private sector participation expands, this gap widens. New insurance schemes without functioning regulators are not reforms. They are invitations to capture.
The third is AI and digital health. Technology companies are moving quickly into health systems across the Global South, promising to leapfrog infrastructure deficits (here, 2025; Africa CDC, 2025a). The concern is not the technology. It is who governs it. Without legal authority, technical skills, and regulatory frameworks, health ministries cannot negotiate fair contracts, manage data sovereignty, audit algorithmic decisions, or push back when commercial interests and public health goals diverge (Africa CDC, 2025a).
Consider what that asymmetry looks like in practice. A technology company arrives with a product, a data agreement, and a government counterpart that lacks a legal framework for AI procurement, a technical team capable of reviewing contract terms, and a regulatory body with jurisdiction over algorithmic decision‑making in health (Africa CDC, 2025a). The government signs because the product appears useful, and the alternative is nothing. Five years later, it has a dependency it cannot exit, data it does not control, and a vendor relationship it lacks the leverage to renegotiate. This is not a hypothetical. Versions of it are already happening (hera, 2025; Africa CDC, 2025a).
Africa CDC is right to call for country‑ and region‑owned digital intelligence ecosystems (Africa CDC, 2025a). But ownership of infrastructure and sovereignty over its governance are different things. The latter requires a state that can audit a system, modify it, and say no to it. Most cannot.
A harder question
There is a critiqué of this argument worth taking seriously. The state capacity frame can become a way of deferring accountability — translating political settlements into technical problems, making what are fundamentally questions about power and resources sound like questions about management and skills (Spark Street Advisors, 2025). It can shift responsibility onto countries while the structural conditions that produced weak states — aid dependency, debt, trade rules, intellectual property regimes — remain intact (Glassman & Kenny, 2024; Spark Street Advisors, 2025). When donors invoke capacity gaps to justify slow transitions or continued control, the frame is doing political work that has nothing to do with institution‑building.
This is not a marginal concern. The history of development cooperation is full of moments when the language of capacity and ownership was used to extend external influence under a different label. Technical assistance programmes that installed advisers with more authority than the ministers they were meant to support. Governance reform initiatives that set conditions for financing while leaving the power to define those conditions firmly with the funder (Spark Street Advisors, 2025). The current enthusiasm for state capacity in health could follow the same path if it is not held to a different standard.
The criticism does not dissolve the problem — it sharpens it. Building state capacity requires changes on both sides: in how external resources flow, how risk is shared, how partnerships are designed, and over what timeframe (Glassman & Kenny, 2024; African Constituency Bureau, 2025). Sovereignty without investment in the institutional conditions for sovereignty is a slogan. Investment in institutional conditions without genuine transfer of power and resources is a new form of conditionality. The architecture debate is walking toward both failures simultaneously.
What actually needs to change
Every new financing mechanism should be evaluated not only on what it delivers but on what institutional capability it leaves behind. Investment in regulatory agencies, public financial management, domestic data systems, and procurement capacity needs to be core programme design, not an optional line item that disappears when budgets shrink (African Constituency Bureau, 2025; Africa CDC, 2025b).
Every partnership — public‑private, donor co‑financing, AI collaboration — should be assessed on whether it builds joint management, transfers technical capability, and invests in regulatory oversight, or whether it contracts around government while calling the result country ownership (hera, 2025; Glassman & Kenny, 2024).
The timeline for what counts as success has to change. Institutional development is slower than programme delivery, harder to attribute, and less suited to the communications needs of donors and global health initiatives (Glassman & Kenny, 2024). These are the reasons it keeps getting dropped, and not sufficient reasons to keep accepting it.
The WHO Executive Board in February and the World Health Assembly in May are genuine opportunities to put these questions on the table with operational specificity — to ask of every proposed mechanism not just what it will finance or coordinate, but what governing capacity it will build in the states expected to lead (UN Foundation, 2026; WHO, 2025). If those meetings pass without that question being seriously addressed, the architecture debate will have produced another layer of frameworks resting on foundations nobody has agreed to build.
References
Africa CDC. (2025a). Africa’s Health Security and Sovereignty (AHSS) Agenda: A new vision for health security and sovereignty across the continent. Africa Centres for Disease Control and Prevention.
Africa CDC. (2025b). Africa’s Health Security and Sovereignty Agenda: A new way forward.
African Constituency Bureau. (2025). The Lusaka Agenda: Transforming health investments for Africa’s future. African Constituency Bureau for the Global Fund.
Frymus, D. (2026). The Lusaka Agenda shows the power of community voices. Think Global Health.
Glassman, A., & Kenny, C. (2024). The Lusaka Agenda Tracker: What gets measured gets done. Center for Global Development.
Government of Ghana. (2025a). African Health Sovereignty in a reimagined global health governance architecture: Statement at the Africa Health Sovereignty Summit.
Government of Ghana. (2025b). Media advisory: President Mahama to launch “The Accra Reset” at UNGA 2025.
hera. (2025). Mapping the future of global health governance: Insights from the Lusaka Agenda.
IHME. (2024). Global health financing 2024: Trends in development assistance for health. Institute for Health Metrics and Evaluation.
Spark Street Advisors. (2025). A summary of ongoing reform discussions, trends and perspectives on global health architecture. Global Health Diplomacy Support Unit.
UN Foundation. (2026). Aligning global health reforms in 2026. United Nations Foundation.
WHO. (2025). Executive Board, 158th session: Provisional agenda and supporting documentation. World Health Organization.
Views my own