Foreign aid cuts reach beyond government budgets

The number looks almost reassuring: 0.9% of GDP. That is the fiscal deficit IMF staff expects Somalia to run in 2026, according to mission conclusions published on 5 October. Spending discipline has helped contain the deficit, and domestically financed social spending has proceeded in line with expectations. Yet the same assessment describes an economy under pressure from declining aid, climate shocks and external tensions. Real growth is projected at 2.3%. These are forecasts, rather than the completed accounts for the year. ^[1]

An IMF report published in December 2025 helps explain the apparent disconnect. In its attached policy memorandum, the Somali authorities identified off-budget funding through humanitarian agencies and NGOs as the main channel exposed to aid cuts. The debt sustainability analysis prepared by IMF and IDA staff describes households losing transfers and businesses supplying programmes scaling back activity. Economic activity can weaken well before the effects reach the government’s accounts. ^[2]

This is where the retreat of aid becomes a macroeconomic issue. A funding decision in a donor country can reduce a government’s resources, a haulier’s orders, a clinic’s supplies or the foreign currency available to an economy. Public finances capture one part of that transmission. Import capacity and functioning services reveal others.

One contraction measured, another still projected

On 9 April 2026, the OECD published preliminary figures for 2025 showing a 23.1% real decline in official development assistance, or ODA. Aid amounted to USD 174.3 billion in current dollars across the Development Assistance Committee members and associates covered by the release. The real change adjusts for prices and exchange rates. Dividing two current-dollar totals would produce a different measure. ^[3]

The headline uses grant equivalents, a measure of donor effort. An outright grant counts in full. For a loan, the calculation values the generosity embedded in its financing terms instead of treating the entire principal as a gift. This makes different instruments more comparable, but it answers a different question from how much cash was disbursed and repaid during the year. ^[4]

The OECD’s forward-looking report, published on 19 June, uses net flows, deducting principal repayments from disbursements under the relevant statistical convention. In that series, the 2025 decline is 23.3%. A further 6.9% fall is projected for 2026. The 23.1% and 23.3% figures therefore reflect different measurements, rather than a discrepancy to be averaged away. ^[5]

Setting net ODA in 2024 equal to 100 gives 76.7 in 2025 and approximately 71.4 in 2026, provided the June projection materialises. That would imply a cumulative fall of roughly 28.6% over two years. This l0g calculation uses the OECD’s rounded published rates on a consistent basis. It describes the volume of aid. Translating it into a loss of recipient-country GDP would require a separate analysis.

Calculation and scope

l0g index, 2024 = 100. Net ODA for the donor scope in the OECD’s 19 June 2026 policy brief, measured at constant 2024 prices and exchange rates. 100 × 0.767 × 0.931 = 71.4077, displayed as 71.4. The published rates are themselves rounded. This series is separate from the USD 174.3 billion grant-equivalent headline. The 2026 point remains a dated projection.

The June outlook combines available budget plans, donor survey responses and assumptions where information is missing. Much of the sectoral allocation is inferred from shares observed in 2024. As of 11 October, it should be read as an evidence-based risk assessment with an explicit vintage. The OECD’s normal publication cycle releases final detailed annual figures in December for the preceding year. The aggregate contraction is better established than its complete and final distribution across programmes. ^{[5] [6]}

Follow the resources to their use

ODA combines transactions that take very different routes through the economy. Some eligible expenditure occurs inside the donor country, including certain refugee-hosting costs. Other funding goes to a partner government’s budget. An organisation may pay a local team directly, purchase medicines abroad for delivery, or channel resources through a multilateral institution. Each route supplies something different. ^{[4] [6] [8]}

That distinction changes the meaning of a cut. Lower refugee-hosting expenditure in a European country reduces recorded ODA without automatically withdrawing an equal amount from a recipient economy. Ending a budget-support grant directly affects the partner’s treasury. Cancelling a humanitarian procurement contract first affects a supplier or the people receiving the goods. An identical change in the headline can conceal markedly different local shocks.

This statistical caution should not obscure the scale of the withdrawal. The April figures show that core bilateral development programmes fell by 26.3% in 2025, even after excluding in-donor refugee costs, humanitarian aid and debt relief. The contraction extends well beyond changes in those particular categories. ^[3]

Reading the routes

This is a qualitative mechanism diagram. In-donor costs remain outside the partner economy. Budget support, organisation funding and aid in kind can reach that economy through different channels. Multilateral contributions are examined in Figure 6. An in-kind delivery supplies goods without requiring a foreign-currency deposit in the recipient treasury.

The OECD’s country programmable aid indicator brings the analysis closer to programmes that can be planned with a partner. It is useful for examining a project pipeline. Its exclusions, however, include humanitarian aid, which can deliver substantial real resources and incomes. No single aggregate removes the need to identify the funding route behind the activity being studied. ^[7]

The relevant observation then becomes the programme as executed: when payments arrive, what is purchased locally or abroad, who is employed and what service is delivered. This also helps prevent double counting. A donor’s contribution to an institution and the institution’s subsequent spending are two stages in a financing chain. Adding both as separate new resources received by the beneficiary would overstate the support. ^[5]

Somalia’s demand shock extends outside the treasury

The December 2025 report identifies the funding chain. The Somali authorities’ memorandum describes programmes delivered through major UN agencies and NGOs. IMF and IDA staff trace the expected loss of household transfers and activity among businesses supplying those operations. This was a forward-looking assessment at the time of publication, rather than an audit of the losses eventually realised. ^[2]

Consider a distribution operation as an illustration, not as a reconstruction of a specific contract. An organisation rents a warehouse, employs staff and hires a transport company. A programme reduction can cut all three revenue streams together. The haulier reduces purchases; some employees cut consumption. Shops serving those employees experience weaker demand. How far the shock travels depends on alternative customers, cash buffers and the ease of replacing the lost activity.

Economists use the term “multiplier” for these subsequent rounds of effects. Attaching a numerical value here would require country- and programme-specific research. A consignment purchased mainly overseas, a team spending its wages locally and a direct household transfer have different economic linkages. The initial amount of aid does not reveal the resulting change in domestic output on its own.

The public budget may be affected later through weaker tax receipts, pressure to replace a service or additional social spending. The sequence matters. Surveillance focused on the fiscal deficit may recognise private-income losses late. Conversely, an aid-funded foreign purchase can supply an essential benefit without generating an equally large amount of domestic business revenue.

The IMF’s 5 October 2026 conclusions also document buffers. Remittances remain resilient, supporting household incomes and foreign-exchange resources. Domestically financed social expenditure is being implemented as expected. Those sources of resilience belong in the assessment. Private transfers reach their recipients, however, whereas disease surveillance or shared infrastructure requires collective decisions and organisation. Whether funding can substitute for aid depends on its use as well as its amount. ^[1]

Imports disappear while the balance holds

The balance of payments records an economy’s transactions with the rest of the world. Its current account includes goods, services, income and current transfers. A gift of medicines from abroad has two corresponding entries: imported goods and a transfer received. The IMF’s accounting framework provides for this treatment of current aid in kind. ^[8]

Take a deliberately simple example, with no country data. An identical consignment of medicines is valued at 100 units. Received as a current grant, it produces a goods debit of −100 and a transfer credit of +100. The operation’s net contribution to the current-account balance is zero. The recipient still has the medicines, without having to arrange 100 units of external payment itself.

Now the grant ends and nobody replaces the shipment. The goods entry falls to zero, as does the transfer. The net effect remains zero, but the medicines have disappeared. If the government or households instead buy the same consignment overseas, the goods debit returns to −100 without the matching transfer. Holding all other transactions constant, 100 units of external financing must be found, or another expenditure must adjust. Maintaining a service can make the funding constraint more visible than abandoning it.

Assumptions

100 arbitrary value units for an identical consignment. Current goods transfer, excluding investment, transport and tax. Imports are shown as negative debits; transfers received as positive credits. The sum covers this operation alone, never an economy’s entire current account. Replacement can be financed in different ways; no source of foreign currency is assumed to be available.

Malawi illustrates a different version of the same analytical trap. In its 2026 country outlook, the African Development Bank estimates that the current-account deficit narrowed from 21.7% of GDP in 2024 to 19.0% in 2025. It attributes that narrowing to imports constrained by foreign-currency shortages. The improvement in the balance can therefore accompany a restriction in supplies. The source does not identify this as the isolated effect of aid cuts. ^[10]

Reserve coverage expressed in months of imports requires similar care. By construction, it compares a foreign-exchange stock with a monthly spending flow. If imports collapse, the number of months covered can rise without any additional reserves. The ratio becomes more comfortable while firms find it harder to obtain inputs. Assessing external resilience requires looking jointly at usable reserves, import volumes and payments waiting to be settled.

Domestic money reaches an external constraint

A government can try to preserve services through taxes, spending reallocations or domestic borrowing. Those decisions redistribute resources within the economy. They do not guarantee access to the foreign currency needed for an overseas purchase. A bank financing local-currency expenditure still has to obtain the currency accepted by the foreign supplier.

As the constraint tightens, adjustment can take several forms: a less favourable exchange rate, a longer wait to settle an import, the use of reserves or reduced purchases. The route depends on the monetary regime, controls and banking conditions. The mechanism discussed here concerns an economy with a domestic currency and constrained foreign-exchange access. It is not a description of Somalia’s particular monetary arrangements.

Scope

Conceptual diagram for an economy with a domestic currency and an external constraint. Arrows represent neither monetary amounts nor estimated causal effects. Malawi documents foreign-exchange scarcity and long-standing difficulties; the figure attributes no national loss entirely to aid. Arrears are an analytical possibility, not a quantified finding established here.

Malawi also warns against an easy causal story. The World Bank’s economic assessment released on 24 February 2026 describes accumulated imbalances, barriers to exports and domestic financing pressures on the private sector. Lost external resources arrive in that setting. Aid cuts can tighten an existing constraint, while domestic policy choices continue to shape the economy’s ability to respond. ^[9]

Payment arrears offer another possible adjustment mechanism. Instead of cancelling an order, a government may postpone paying for it. The supplier still has wages and other obligations to meet. Its cash resources become part of the transition financing. This is an economic possibility, not a claim that aid cuts have caused a particular measured stock of arrears in the countries examined.

Tax collection and credit also operate on different timelines. A lasting improvement in revenue collection can strengthen fiscal autonomy. An immediate increase in the burden on already weakened activity may also reduce demand. A workable transition has to align revenues, foreign-exchange needs and useful expenditure, rather than assuming that every financing source can replace another at once.

Health services depend on complementary functions

Healthcare delivery depends on complementary activities. Products need to be ordered, transported, stored and used by an available team. Information identifies changing needs and guides procurement. A cut concentrated in one function may affect the service by more than that function’s share of its budget. This complementarity is why operational continuity needs to be monitored separately from total funding.

The WHO had already documented disruption in spring 2025. Its rapid March–April stocktake covered 108 country offices, mainly in low- and lower-middle-income countries. Among offices answering each question, 66% reported disruption to public-health surveillance (105 responses), and 54% to the health and care workforce (106 responses). These are reporting frequencies for a defined set of respondents. They are neither estimates of the percentage of capacity destroyed nor a global assessment for October 2026. ^[11]

Data and interpretation

108 country offices surveyed in March–April 2025; release dated 10 April. Public-health surveillance: 66% of 105 responses; health and care workforce: 54% of 106 responses. Question completion varied, and the reports were not quantitatively verified. Categories may overlap and cannot be added. The bars do not measure percentages of service lost. The operational diagram is qualitative and assigns no estimated output or delay to its links.

Losing information can make the damage harder to observe. A reduced surveillance team may be less able to detect needs that should trigger an intervention. The absence of a warning becomes ambiguous: stable conditions, or weakened observation? The risk extends beyond today’s service volume to the quality of future decisions.

A programme closure can also leave assets underused: premises, equipment or a transport network without regular orders. Restarting requires bringing the relevant functions together again. No restart duration is estimated here. The mechanism explains why an interruption and a planned handover can have different effects even when cumulative funding is comparable.

This does not assume that the earlier organisation of aid was perfectly efficient. WHO guidance published on 3 November 2025 recommends improving procurement, protecting essential services and integrating externally funded programmes into primary healthcare. Less fragmented delivery can improve the use of resources. Achieving that depends on preparation, domestic capabilities and actual implementation. ^[12]

Multilateral finance follows its commitments

Multilateral resources provide a material counterweight to a story of uniform withdrawal. In December 2024, the World Bank announced USD 23.7 billion in donor pledges for the IDA21 replenishment, supporting USD 100 billion in financing capacity. The International Development Association provides grants and credits to eligible countries. This cycle covers July 2025 to June 2028. The USD 100 billion is an announced three-year capacity; the USD 23.7 billion helps underpin it. ^{[13] [14]}

The mechanism works through a balance sheet, drawing on donor resources, repayments on earlier financing and access to capital markets. It can support commitments across several years. A pledge, an approved operation and a disbursement reaching a country occur at different stages. This structure can absorb some changes in donor payment schedules. It retains its own constraints, including eligibility rules and the institution’s capacity to sustain its financing. ^{[15] [17]}

IDA’s financing page, updated on 5 October 2026, provides an initial implementation marker. In the fiscal year ending 30 June 2026, commitments reached USD 30.53 billion, including USD 8.45 billion in grants. Commitments differ from disbursements already received. The grants are included in the total, rather than added to it. These figures do not establish that bilateral aid cuts have been fully offset. ^[15]

Units and limitations

US dollars, billions; amounts announced in December 2024. USD 23.7 billion in donor pledges helps support the USD 100 billion envelope for July 2025–June 2028. These figures are neither additive nor annual disbursements. No quantitative split by financing source, grants, credits or year is reconstructed. The announcement does not establish that bilateral aid cuts have been offset.

Replacement nevertheless requires more than a global envelope. A repayable investment credit and a grant funding a team each month serve different purposes. In December 2025, Somalia obtained increased access to its IMF financing programme. That illustrates a possible source of additional support through a particular instrument. It does not establish that lost programme funding was fully restored. ^[18]

The OECD’s June note also projects falling ODA contributions to multilateral organisations. Its donor-imputed multilateral aid calculations trace resources back to providers; they differ from institutions’ own disbursements to countries. Testing whether funding has genuinely been replaced requires following those disbursements through to their use. ^[5]

A transition depends on replacements that actually work

Funding concentration makes coordination important. Programmes with different operators and logos can depend on the same national budget decision. Apparent diversity in delivery may conceal a common source of resources. The OECD note highlights exposure to a limited number of major providers. A simultaneous cut can then remove several potential sources of replacement at once. ^[5]

ICAI’s scrutiny of British spending plans, published on 16 July 2026, identifies a lasting shift in many bilateral programmes. It also recognises an improvement: three-year allocations provide greater predictability. The resource reduction and the quality of its timetable are separate dimensions. A known end date can give partners time to arrange a team transfer, stocks or domestic funding. A late decision may leave those replacements unfinished. These remain plans whose implementation must be followed. ^[16]

Lasting autonomy requires revenue and export capacity, skills and functioning institutions. With a credible transition, withdrawing a funding relationship may accelerate their development. It can also weaken the services and activity on which that development depends. The documents reviewed provide no basis for imposing a single answer on every recipient.

The most useful observations sit close to the constraints: disbursements actually received, supplier payments, foreign-exchange access, import volumes and staff availability. Declining ODA alongside a well-funded domestic handover has different implications from declining ODA followed by service interruptions and compressed imports. The IMF’s findings on Somalia’s domestically financed social spending show why evidence of resilience must remain part of the assessment. ^[1]

The macroeconomic risk lies in income losses, weaker import capacity and impaired services occurring together. A contained fiscal deficit or a narrower external deficit can coexist with an economy scaling down its activities to adjust. Understanding that adjustment requires following the functions preserved and the resources genuinely available, beyond the annual totals recorded by donors.

Further reading

The investigation into El Niño and emergency resources examines the capacity available during a climate shock. The guide to the dollar and foreign-currency funding explains exchange-rate and hedging mechanisms. These complement the assessment of services to preserve and currencies required for cross-border payments.

Sources, method and limits

This documentary investigation is closed as of 11 October 2026 and draws on accessible institutional publications. No original interviews, programme audits or confidential banking or contractual data were used. IMF mission conclusions, earlier observations and OECD projections retain their respective dates. The analysis estimates no isolated causal effect of aid on GDP, mortality or exchange rates.

The two ODA measures are kept separate. The volume index uses rounded published rates and its 2026 value remains conditional. The routes are qualitative mechanisms. The medicine example is hypothetical and fully specified. WHO percentages refer to country-office reporting. IDA21 is an announced multi-year envelope. The IMF reports and WHO technical report were reviewed in their complete official versions. The AfDB profile was examined in indexed text, with its access limitation recorded below.