A new UNESCO report published this week finds that 113 countries — home to 6.1 billion people — spent more on servicing foreign debt in 2025 than on education. Among these, 18 countries spent more than five times as much on debt repayments as on their schools. Sub-Saharan Africa, the region with the world’s fastest-growing youth population, spent 3.6 times more on debt than on education. The UNESCO Global Education Monitoring Report simultaneously warns that international aid to education is projected to decline by up to 30 percent between 2023 and 2027, compounding the fiscal pressure on governments already struggling to service obligations accumulated during the pandemic borrowing surge. These figures arrive in a week when the World Bank and IMF are convening discussions on debt relief frameworks, and as the United States continues to reduce its contributions to multilateral development institutions.
The received wisdom
The progressive analysis of this data is powerful and largely correct as far as it goes. Here is the argument: the global debt architecture systematically transfers resources from the poorest populations to the wealthiest creditors. Loans extended to developing governments during crises — often at unfavourable terms, frequently to governments with poor accountability records — have compounded into obligations that now crowd out the most basic public goods. Education is not merely a consumption item; it is the primary mechanism by which poor countries generate the human capital needed to escape the debt trap in the first place. The irony is structural: the more a country spends servicing debt, the less it invests in the productivity that would eventually generate the revenue to retire that debt sustainably. Aid cuts from Western governments, driven by domestic austerity politics, make the situation worse by reducing the grant financing that could have replaced or supplemented public education budgets. On this reading, debt relief is not charity; it is rational policy for a global economy that needs functioning states rather than perpetually insolvent ones.
As I said, this analysis is largely correct. The numbers are genuine, the mechanism is real, and the human cost is measurable in school enrolment rates, learning outcomes, and lifetime earnings.
A different read
Where the mainstream framing goes wrong is in its diagnosis of cause and its implied prescriptions. The story of developing-country debt is not only a story of rapacious creditors imposing unjust terms on innocent governments. It is also, in many cases, a story of governance failure, elite capture, and corruption that diverted borrowed resources away from productive investment and toward rent extraction, patronage networks, and capital flight. The Zambian debt crisis, to take one well-documented example, involved Chinese infrastructure loans whose terms were partly opaque and whose projects often employed imported rather than local labour — but it also involved Zambian officials who accumulated substantial personal wealth while public services deteriorated. Blaming the creditors is easier, and sometimes warranted; it is also incomplete.
The UNESCO numbers are particularly striking when disaggregated by creditor type. A significant share of the most distressed debt in Sub-Saharan Africa and South Asia is owed not to Western multilateral institutions — which typically offer concessional terms and have established debt relief mechanisms — but to bilateral creditors, including China’s Export-Import Bank and Chinese commercial banks. The Common Framework for debt restructuring, established by the G20 after the pandemic, has been slow precisely because China has been reluctant to participate on terms comparable to those the Paris Club creditors accept. This is not a minor operational detail; it is the central obstacle to making debt relief work at scale.
The proposed solution — more aid, more grants, more debt cancellation — runs up against a fiscal and political reality that the UNESCO report does not adequately engage with. Western publics, facing their own strained budgets and sceptical of aid effectiveness after decades of mixed results, are not in a mood to endorse large-scale transfers. The USAID budget cuts under the current US administration are politically unpopular among development advocates, but they reflect genuine scepticism — not only among Republicans — about whether money sent through government channels actually reaches children in classrooms. That scepticism is sometimes wrong and sometimes right; it cannot be dismissed by pointing at UNESCO charts alone.
The more tractable path, and one that has stronger empirical support, involves domestic revenue mobilisation in developing countries — improving tax collection, reducing evasion, expanding the formal economy — combined with targeted debt relief from multilateral institutions on terms that include conditionality around educational spending floors. Countries that receive debt relief without governance reform simply accumulate new obligations. The structural problem is not only that debt is too high; it is that the institutional capacity to spend productively on education is often as constrained as the fiscal space.
What to watch
— G20 Common Framework negotiations: The next round of restructuring talks will reveal whether China is willing to accept haircuts comparable to those Western creditors accept. Any breakthrough, or breakdown, will be a leading indicator of whether the 113-country debt trap can be addressed at scale.
— USAID supplemental appropriations: Congressional negotiations over development assistance in the next budget cycle will determine whether the projected 30-percent aid-to-education decline is moderated or accelerates.
— World Bank’s IDA replenishment: The International Development Association’s next replenishment round will set grant and concessional lending levels for the poorest countries through 2027. Watch for the US and European contribution levels.
— Domestic education spending tracking: The UNESCO report notes that several high-debt countries have nonetheless maintained or increased education spending as a share of national income. These cases — typically correlated with stronger civil society — are worth studying as models.
— J