How African Governments Responded to the 2025 Aid Shock
It has been just over a year since the world of international development as we knew it began to unravel. The abrupt suspension and restructuring of US assistance—including the dismantling of USAID—was followed by a wider wave of announced reductions across several major donors, reflecting tighter fiscal space and shifting geopolitical priorities. As the region receiving the most official development assistance (ODA) relative to national income, Africa has been hit hardest by these cuts. The question for African governments, therefore, is not just how large the cuts are, but whether and how they can respond in ways that protect essential services and enable a longer-run shift toward self-reliance.
This post takes a first, fast-cut look at what African governments actually did or said in response to current or impending aid cuts over the course of 2025. The goal is to map the type, scope, and intensity of government responses during 2025, using the previous year as a baseline.
Tracking sovereign responses
I compiled a dataset of 442 government-involved events that could be directly or indirectly linked to the aid cuts across 54 African countries in 2024 and 2025. An “event” is defined as a discrete, dateable instance in which a government actor (e.g., president/prime minister, parliament, or a key ministry) takes an action, announces a plan, or makes an official statement related to aid cuts or associated shifts in financing and geopolitical alignment. Events were identified through targeted web searches and structured scraping of government websites and mainstream media, supported by multiple Large Language Models, and then verified manually based on the summary included in the tracker. The tracker is not exhaustive: countries with stronger online communications and media ecosystems are likely overrepresented. To reduce this bias, I treat 2024 as a baseline period of “normal” foreign aid and geopolitics-related activity, and focus on the change in 2025 relative to that baseline. The raw data and summary table of events by country can be accessed here.
Each event is coded along two dimensions. The first is its link to aid cuts: events are tagged as “direct” only when the communication explicitly mentions aid cuts and includes an action, plan, or call for action; “indirect” when aid cuts are not explicitly mentioned but the response plausibly reflects a link to shifting aid; and “baseline” otherwise. Baseline, indirect, and direct events are assigned values of 0, 1, and 2, respectively. The second dimension is response strength, which captures the intensity of the government response, ranging from rhetorical signaling (value 1), to formal decisions including tapping existing financing sources (value 2), to new material commitments (value 3). I multiply the aid-cut link score by the strength score to obtain an event-level response intensity, then aggregate these to a country-level response score. To estimate 2025 response intensity, I subtract the 2024 baseline and then normalise the resulting values to a 0–1 index, where higher scores indicate stronger and more direct responses.
South Africa scores highest on the response index (Figure 1), driven largely by direct responses to US funding cuts affecting health programmes, particularly HIV/AIDS treatment and prevention services. National and provincial authorities moved quickly from statements to budgetary measures and transition planning. By contrast, the high scores of countries such as Morocco and Algeria reflect a different pattern: a series of multilateral financing engagements or partnership agreements that are only indirectly linked to the aid-cut shock but matter for the fiscal envelope in which governments will now operate. At the other end of the distribution, several countries show little or no increase in observable government-led response activity in 2025 relative to the previous year.
An interesting pattern emerges when the response index is compared with aid dependency (measured as net ODA received as a share of GNI, using the most recent available data). Figure 2 shows that a great majority of countries above the median aid-dependency threshold sit in the low-response half of the chart (the lower-right quadrant). In other words, the countries that appear most exposed to an aid cliff are not the ones mounting the most visible, government-led response (contrast the densely populated upper-left quadrant for the less aid-dependent countries with the sparsely populated upper-right quadrant for the more aid-dependent countries). That is consistent with a simple constraint story: governments with limited fiscal space and weaker administrative capacity have less room to cushion abrupt external shocks. What is slightly surprising is that they were not only doing little about the aid cuts—they are also saying little about them.
What governments are responding to, and how
Roughly a quarter of direct responses from African governments focus on keeping specific services running or securing resources for frontline delivery: for example, maintaining HIV and TB diagnostic and treatment services, covering commodity shortfalls, or announcing stopgap measures for affected programmes. A smaller but important share of direct responses pertains to actions at the macro level (for example, budget reprioritisation, domestic revenue measures, or explicit calls for partners to step in).
Among indirect responses, multilateral financing agreements are especially prominent. Many of these are IMF programmes and World Bank operations. These financing agreements are not always framed as a reaction to aid cuts, but they will shape development financing in practice.
Diplomacy and geopolitical positioning also feature in both direct and indirect responses, echoing the broader shift toward more explicitly strategic partnerships.
The sectoral distribution closely mirrors the type pattern. Health dominates direct responses, accounting for well over two-thirds of them, underscoring how abruptly the US disruption translated into immediate risks to service delivery. Several cross-sector responses (for example, fiscal packages or transition committees) are also implicitly anchored in health services continuity. In contrast, the most common sectoral category for indirect responses is macro-fiscal / PFM, which is consistent with governments searching for domestic mechanisms to absorb a sustained decline in concessional flows.
Two absences are notable. First, humanitarian and social protection responses appear relatively limited, despite the large humanitarian footprint of US assistance in many contexts. This likely reflects the most sobering reality in the data: countries most reliant on humanitarian financing may have the least capacity to respond visibly at sovereign level. Second, education barely registers in either rhetoric or action. Some of this reflects the political economy of prioritizing responses: health shocks are immediate and life-threatening, whereas education cuts are slower-burning. Moreover, many of the countries that are capable of responding in the short term do not rely on foreign aid for key education programmes. But this is also a warning sign: if education falls out of the response frame, the long-run costs to human capital could be high.
Timing of responses: most direct responses cluster around the US aid suspension
The timing of responses (Figure 5) reinforces the interpretation that the most visible direct reactions were triggered by the initial US suspension and programme disruptions in early 2025. There is comparatively limited direct government response in the tracker to aid reductions announced by other donors later in 2025. A benign interpretation is that many European cuts will bite with a lag, as multi-year commitments unwind. A less benign interpretation is that governments are not yet anticipating the true impacts of a second wave of reductions that could materialise over the next budget cycles. The phased approach adopted by European donors may allow for more gradual adaptation but could also lead to complacency, particularly at the expense of programmes in such sectors as education that are less visible to the public.
What the language of responses reveals
The word clouds below (Figure 6) offer a more vivid picture of the focus and tone of the responses, while also providing a simple cross-check on the coding. In the direct-response word cloud, the most prominent terms are clearly operational and time-sensitive such as “health”, “funding” and “suspension”. This vocabulary is consistent with governments focusing on immediate continuity and risk management in the face of disrupted external financing for essential services.
In contrast, the indirect-response word cloud is dominated by terms associated with longer-horizon partnerships and project finance such as “cooperation”, “approved” and perhaps most interestingly, “China”.
Taken together, the two clouds suggest a split between short-run mitigation in health-related shocks and a broader, more structural response channeled through multilateral programmes and cooperation agreements with partners such as China.
What this implies for outcomes in aid receiving countries and the next phase of the aid cuts
This is an indicative, early mapping exercise rather than a definitive audit of country responses. But even with that caveat, two implications stand out. First, the countries that rely most on aid appear least able to mount timely, visible sovereign responses. Second, the response signal is heavily skewed toward health, while education and other long-run investments are largely missing from the reaction frame. If the coming years bring a broader and more sustained contraction in ODA, the risk is not only a short-run service disruption but a long-run human capital cliff.
It is not obvious how effective or sustainable even the most concrete responses by relatively proactive governments will be. Countries like South Africa (where external support was concentrated in specific programmes that can be integrated into the national budget frameworks relatively easily) may be able to sustain their response, but countries where aid has been critical across multiple sectors will be less able to do so, even when governments have taken initial concrete steps to offset the cuts. The response process may also be complicated, or slowed, by the fact that many aid-funded programmes have rarely been integrated to national budgets. At the same time, the disruption could create an opportunity to test whether the same services can be delivered more efficiently by fully integrating these programmes into government systems.
Based on these findings, donors that are planning to reduce or restructure their aid budgets and recipient governments likely to face declining ODA flows should consider the following recommendations:
Image: Map of Africa showing response intensity index by country (2025 relative to 2024 baseline), with illustrative country examples. Source: CGD.
