The global health architecture is in a crisis of finance and legitimacy. High-income governments are cutting back support, and low- and middle-income countries—notably in Africa—are demanding more sovereignty and a ‘reset’. There has been a dizzying array of initiatives attempting to achieve this, yet real reform has hit an impasse. This may be because the initiatives tend to be over-ambitious in scope, and either lack legitimacy or the power to make the changes required. A new solution is needed: one that limits itself in scope to reforming financial support for low- and middle-income country (LMIC) systems, pragmatically (and regrettably) excludes the US, and is legitimate enough and powerful enough to deliver change. In this blog, I propose that an African Union-European Union-United Kingdom (AU–EU–UK) tripartite agreement could be exactly the solution needed.
Current reform initiatives lack either power or legitimacy
The range of current, active, or proposed initiatives on global health architecture reform have all had to tackle a core problem—the power to change lies predominantly with donor governments, yet the legitimacy lies with recipient governments, their populations, and civil society. None of them have this balance right.
- The first set of proposed options are those that have legitimacy but limited power. For example, the Accra Reset is primarily driven by Ghana, and lacks formal backing from both the AU and donor nations. The Africa CDC Lusaka Agenda continental secretariat intends to ensure reforms occur, but lacks a clear mechanism to do so. Others have suggested a World Health Assembly resolution and process, and/or a special UN commission. These are legitimate but they include too many nations that are not contributors nor recipients, diluting their legitimacy and slowing down the process. They also risk lacking power because of the current failure of multilateral governance. Of note, UN nations could not even agree whether to tax alcohol and tobacco.
- The second set have power but limited legitimacy. Take the European Commission’s consultation on global health architecture reform. This could collectively agree how to fund global health and establish new institutions, but is unlikely to have traction because it lacks legitimacy. The G20 is another example—while it could expand the number of potential donors from the current G7 dominance of global health, it too lacks legitimacy. Furthermore, it is paralysed and powerless by the incoming US chair.
- Finally, there is a group of initiatives with neither power nor legitimacy, which are more about idea generation and discussion, such as the Friends of the Lusaka Agenda, the Seville Platform for Action, and the current round of Wellcome Trust regional convenings.
Therefore, a new solution is needed: one that pragmatically reduces the scope and number of stakeholders to a manageable core group, and critically has both the power and the legitimacy to drive reform.
1) Focus on multilateral financial support for low- and middle-income health systems
The first step to pragmatic global health architecture reform is to acknowledge that we need two processes: one for reform of truly global functions (e.g., WHO’s core mandate, pandemic preparedness, etc) from which the whole world benefits, and one for reform of multilateral financial support for LMIC health systems. Reform of truly global functions requires all countries to govern it (e.g., through UN or the World Health Assembly) and all countries to contribute. In contrast, multilateral financial support for LMIC health systems has a precise and smaller set of stakeholders: governments (and philanthropies) that are providing resources, and governments (and populations) that are receiving them. This smaller group can proceed without global consensus.
2) For now—regrettably—exclude the US
The US has exited multilateral global health after decades of dominance. Its new global health strategy makes it clear that it is focused on bilateral compacts. It has left the World Health Organization, cut funding for Gavi, and it has undermined key global health agreements, including the political declaration of the UN High Level Meeting on Non- Communicable Diseases, which failed to win endorsement. The US is still considering supporting the Global Fund financially, although the conditions it may impose as part of this contribution—particularly on sexual and reproductive health and rights and engagement with progressive civil society—may make proponents of the US remaining involved come to regret it. Unfortunately, it seems the US is not—and does not wish to be—a reliable, good-faith partner in designing the future of multilateral financial support for LMIC health systems. The remaining donors and recipient governments must lead the reform, and look to engage the US in other ways.
3) A tripartite African Union–European Union–United Kingdom agreement
The Institute for Health Metrics and Evaluation (IHME) estimates that in 2025, sub-Saharan Africa will receive more than 50 percent of non-US development assistance for health (DAH), or over $11 billion. Sub-Saharan Africa will also receive 69 percent of Global Fund and 56 percent of Gavi expenditures in 2025. This makes it the most important and legitimate recipient constituency to decide the future of non-US multilateral finance assistance for health systems. And through the AU, Africa has—at least in theory—a mechanism to coordinate and express a single view.
Similarly, IHME estimates in that in 2025, EU nations and the UK make up 46.6 percent ($5.14 billion) of non-US DAH to sub-Saharan Africa, as well as 54 percent of Gavi pledges for 2026–2030 (Global Fund contributions TBC). It also has a solid coordination mechanism in the European Commission, as well as a loose but manageable Team Europe process of pulling into agreement the UK and others such as Norway.
If the AU, EU, and UK reach an agreement on global health architecture reform, it may not be fully legitimate, nor fully powerful. However, it may be just sufficiently legitimate and powerful to be a success. For example, if AU members, EU members, and the UK delivered a clear, single message of reform priorities to the boards of the Global Fund and Gavi—and this was backed by agreed conditionalities on replenishment commitments and willingness to withdraw resources if no progress was made—it could be decisive.
Making it happen
A three-phase approach over one year could be considered: launch, consultation, and negotiated agreement. The launch phase could begin with the EU-AU Summit on 24 November—a key moment for leaders to agree to this approach. A formal launch could follow at the 39th AU summit in February 2026, or at the proposed UK development summit in March 2026. The consultation phase would involve the AU and EU engaging member states, along with the UK. Other non-EU Team Europe members, such as Norway, may be brought into the consultation. A common voice emerging from the AU will be particularly critical to success, but may be challenging due to diverging interests between those highly reliant on aid and those less so. Finally, a formal negotiation phase would begin between the AU, EU, and UK, leading to a political declaration which could be published at a special meeting alongside the AU summit in February 2027.
A key challenge with this proposal is the exclusion of major funders such as the US, Canada, Japan, Australia, philanthropy, low-income countries in other regions, and civil society voices. The consultation phase should therefore include extensive and public consultations with these groups. For independence, parts of the consultation could be philanthropy-funded. Finally, wider support for the declaration could be sought by bringing a resolution to the World Health Assembly in May 2027, with the goal of tasking the WHO to monitor the reform, and to Global Fund and Gavi Boards throughout 2027.
There may be bigger visions for reform that justify bigger processes. But perhaps this is an imperfect solution for an imperfect world that might just work? Let me know your thoughts.
With thanks to Anastassia Demeshko for research support, and to Rachel Bonnifield and Anita Kappeli for comments on an early outline.
Pete Baker can be reached at [email protected].
