Executive Summary
UK official development assistance (ODA), the internationally agreed measure of aid, was £13.0 billion in 2025, 0.43% of national income, down from £15.4 billion and 0.58% in 2023.12 The UK spent 0.7% of national income on aid every year from 2013 to 2020, a level written into law by the International Development (Official Development Assistance Target) Act 2015. The government cut it to 0.5% from 2021, citing the pandemic, and in February 2025 decided to reduce it to 0.3% by 2027 to fund higher defence spending, while stating its commitment to return to 0.7% "as soon as fiscal circumstances allow".1 Since July 2021 that return has depended on two published tests, which the House of Commons was asked to approve: the OBR must forecast that, on a sustainable basis, the government is not borrowing to finance day-to-day spending and underlying debt is falling. The October 2026 Budget says they are not due to be met within this Parliament.3
Aid is counted under international rules set by the OECD, which allow donors to include the cost of supporting asylum seekers and refugees in their own country during their first year. The UK has used this heavily. These "in-donor refugee costs" rose from £628 million in 2020 to £4.3 billion in 2023, 27.9% of all UK aid, and were £2.4 billion in 2025.41 The Independent Commission for Aid Impact (ICAI) estimated in March 2026 that nearly £10 billion of aid between 2020 and 2024 went on these costs, and that the UK's cost per refugee was more than $25,600, against an average of $6,100 among other donors.4
ICAI also found that aid was managed "mainly" to hit the annual spending target rather than to deliver agreed results, with no consistent overarching strategy and repeated late-year cuts to programmes.4 The government accepted all five of its recommendations in April 2026.5 The cut to 0.3% now falls hardest on bilateral aid, the money the UK spends directly in other countries, which the Foreign Office's own assessment puts at about 37% lower over the next three years.6
Key Proposals
Pay for refugee costs in the UK from the Home Office budget, outside the aid target. Stop counting the cost of supporting asylum seekers in the UK towards the aid target, fund it from the department responsible for asylum, and report it separately, so that the aid figure measures what is spent overseas.
Put the tests for restoring 0.7% into law, and report against them at every Budget. The two fiscal tests set in 2021 for returning to 0.7% rest on a ministerial statement and a Commons vote, and can be changed without legislation. Write them into the 2015 Act and require the OBR to say at every Budget whether they are met and, if not, when they are forecast to be.See how Budget 2026 relates to this
Allocate the smaller budget by published evidence of results. Publish, for every aid programme above £50 million, its expected and achieved results and cost per result, and use that evidence, reviewed by ICAI, to decide which programmes are protected when the budget falls.
1. What Counts as Aid
Official development assistance is defined by the OECD's Development Assistance Committee, the club of donor countries. It covers government spending aimed mainly at the economic development and welfare of developing countries, whether given directly to those countries (bilateral aid) or through bodies such as the World Bank and United Nations agencies (multilateral aid).1 Aid is spent by many government departments, not only the Foreign, Commonwealth and Development Office, which is why the UK's aid figure differs from any one department's budget and from the Treasury's own figure for "international services", £10.5 billion in 2024-25.7
The rules also allow donors to count some spending in their own country, including the cost of supporting asylum seekers and refugees from developing countries during their first twelve months: transport, food, shelter and training.1 This is legitimate under the rules, and many donors do it. The question is how much of the budget it takes, and at what cost.
2. The Aid That Never Left the UK
In-donor refugee costs were £628 million in 2020. They rose to £4.3 billion in 2023, 27.9% of all UK aid, as the number of asylum seekers in hotels and other contingency accommodation grew, and fell back to £2.8 billion in 2024 and £2.4 billion in 2025.41 ICAI estimates that nearly £10 billion, 14.5% of all UK aid, was spent this way between 2020 and 2024.4
Two features of how the UK did this stand out in ICAI's review. The first is cost. In 2022-23 the UK's in-donor refugee cost per head was over $25,600, against an average of $6,100 among other donors, and the UK takes the most expansive reporting options available in most of the areas ICAI compared.4 The second is the effect on aid overseas. Because these costs were deducted from the Foreign Office's allocation, every pound spent on asylum accommodation in the UK displaced a pound of aid abroad. Home Office forecasts varied widely during each year, so the Foreign Office had to hold back money and cut programmes late in the year to make room. In 2024 it gave up £991 million in July, received £400 million back in December, and had to re-plan another £135 million at short notice.4
3. Managed to a Target, Not to Results
ICAI's March 2026 review of how the aid budget was managed between 2021 and 2025 found that cross-government coordination was mainly aimed at hitting the annual spending target rather than at delivering agreed results or long-term value for money. The UK had no consistent overarching aid strategy over the period, and allocations often reflected historical patterns and short-term pressures. Treating the target as both a floor and a ceiling forced the Foreign Office, as "spender and saver of last resort", into repeated year-end changes and spending pauses, including a pause on non-essential aid in July 2022.4
The government accepted all five of ICAI's recommendations in April 2026. It has set aid budgets in cash, so they no longer move automatically with forecasts of national income, ended the Foreign Office's role as spender and saver of last resort, and set three-year allocations.5 These are real improvements in predictability. They do not, on their own, change how money is allocated between programmes.
4. Where the Cut to 0.3% Falls
The Foreign Office published its aid allocations for 2026-27 to 2028-29 in March 2026, with an assessment of their effects. Against 2025-26, its total aid programme falls by about 31%. Bilateral aid falls by about 37% and multilateral aid by about 22%. Centrally managed programmes fall by about 44%, and aid-funded research and development by 44%.6
Some areas were protected: the multilateral development banks, including the World Bank's fund for the poorest countries, in cash terms; support for Gavi, the vaccine alliance, to immunise up to 100 million children between 2026 and 2030; spending on violence against women and girls held at 2025-26 levels; and funding for Palestine, Sudan and Ukraine. Others were cut: core humanitarian contributions to multilateral agencies by about 25%, global health spending through multilateral bodies by about 23%, with no further funding for polio eradication or the Pandemic Fund, and health and social protection programmes in countries including Malawi and Sierra Leone. The share of country spending going to sub-Saharan Africa falls by eight percentage points.6
The Foreign Office's own assessment concludes that multilateral cuts are not expected to fall disproportionately on any protected group, but that safeguarding spending "will disproportionately reduce, with women and children particularly affected", that some country-level gender programmes may close, and that cuts to social protection will affect poor households vulnerable to shocks.6
The UK is not alone. The OECD's preliminary figures for 2025 show aid from its donor members falling by about 23% in real terms, the largest annual fall on record, driven mostly by the United States. The UK's aid fell by about 11%, and Germany became the largest donor for the first time.8
ICAI found UK aid managed mainly to hit a spending target, with no consistent strategy and allocations driven by history and short-term pressures, and estimated that nearly £10 billion went on refugee costs in the UK in 2020-2024 at more than four times the average cost per person of other donors.4 The cut to 0.3% reduces bilateral aid by about 37%.6
This connecting argument is this project's own. A budget that is shrinking, partly spent at home and allocated by habit rather than by measured results is likely to cut programmes that work as readily as those that do not. The smaller the budget, the more each misallocated pound costs in results forgone.
One option is to accept the current approach, now made more predictable by cash budgets and three-year allocations. A second is to reverse the cut, which is a decision about the overall level of aid. A third, which this pillar proposes and which works at any level, is to take refugee costs out of the aid target, put the existing conditions for restoring 0.7% into law, and allocate the remaining budget by published evidence of results.
Moving refugee costs out of the target would reduce the aid that can be reported against it, and some will see that as a backdoor cut or, from the other side, a backdoor increase. This pillar's judgement is that the aid figure should measure what is spent on development overseas, whatever level is chosen. A reader who sees support for refugees as part of the UK's contribution to development is making a defensible case under the OECD rules.
Fund refugee costs from the Home Office outside the aid target, put the tests for restoring 0.7% into law, and publish results and cost per result for every major programme. The test below sets out what would show this is not working.
5. The Steel Man
There are two strong cases against this pillar, from opposite directions.
The first holds that the UK should not be spending on aid at the current level at all. Defence faces real and rising threats, public services at home are under severe strain, and borrowing is high. Governments of both main parties have now cut it. On this view the question is not how to manage aid better but why the UK spends billions abroad when it cannot fill potholes or clear hospital waiting lists at home, and counting refugee costs as aid at least recognises a real cost that falls on UK taxpayers because of crises abroad.
The second holds that this pillar is far too timid. The UK promised 0.7% in law and has now cut it twice, at a time when the world's largest donor has slashed its own aid. Supporters point to programmes that vaccinate children, prevent famine and treat HIV as among the most cost-effective things any government funds. Defining a fiscal test or publishing cost per result is administrative tinkering when the real issue is the scale of the cut. On this view the only defensible proposal is to restore 0.7%.
This pillar takes neither position on the level of aid, because that is a choice of values the evidence cannot settle. What it does say is that both sides should want the same three things: an aid figure that measures aid spent overseas, a commitment whose conditions are clear, and a budget that funds what works. Those questions arise at 0.3% and at 0.7% alike.
6. Counter-Arguments
'Taking refugee costs out of the aid budget is just a way to inflate aid'
It would make the aid figure smaller, not larger, unless the government chose to replace the money. The proposal is about what the number measures. At present, a rise in asylum costs in the UK automatically reduces aid overseas while the headline figure stays the same, which ICAI found distorted both policy and transparency.4
'The government has already accepted ICAI's recommendations, so this is done'
The government's acceptance moved the Foreign Office to fixed cash budgets and three-year allocations, which insulates it from swings in refugee costs.5 Refugee costs are still counted within the aid target, so a rise in them still reduces what is available overseas within the 0.3%. The proposal here goes further than the government's response.
'Cost per result can't be compared across humanitarian, health and economic programmes'
Not perfectly, and the proposal does not require a single league table. It requires each major programme to publish what it expected to achieve, what it did achieve, and what that cost, so that decisions about what to cut are made on evidence that can be challenged rather than on historical patterns.
'The government already has tests for restoring 0.7%, so this proposal is unnecessary'
It does, and an earlier version of this page wrongly said it did not. Since July 2021 the return to 0.7% has depended on two tests approved by the Commons: no borrowing to finance day-to-day spending and underlying debt falling, on a sustainable basis, as forecast by the OBR. The October 2026 Budget says they will not be met in this Parliament.3 But the tests rest on a ministerial statement, not the 2015 Act, so a future government can change them without legislation. Writing them into law costs nothing if the commitment is meant, and makes any change to them visible.
Cross-Pillar Dependencies
| Pillar | Connection |
|---|---|
| Defence | The cut to 0.3% was made explicitly to fund higher defence spending. The two budgets are now directly linked. |
| Immigration | In-donor refugee costs depend on the number of asylum seekers and how they are accommodated, which the Immigration pillar and the Channel Small Boats working paper examine. |
| Public Spending Overview | Aid falls within the Overview's "Other: foreign affairs, science, culture, admin" line. |
| Science | Aid-funded research and development falls by 44% under the new allocations, part of the wider funding picture for UK science. |
8. Proposals for Change
These are the evidence-based proposals of this pillar, put forward for public discussion and challenge.
-
P1 Fund the cost of supporting asylum seekers and refugees in the UK from the Home Office budget, report it separately, and stop counting it towards the aid target, so that the aid figure measures development spending overseas.
-
P2 Write the two fiscal tests set in 2021 for restoring 0.7% (on a sustainable basis, no borrowing to finance day-to-day spending and underlying debt falling) into the 2015 Act, and require the OBR to report against them at every Budget, so that the conditions are fixed in law and checkable.
-
P3 Publish expected and achieved results, and cost per result, for every aid programme above £50 million, reviewed by ICAI, and use this evidence to decide which programmes are protected when the budget falls.
-
Keep the multi-year cash allocations introduced in 2026, and extend them to a rolling basis so that programmes are not cut late in a year to meet a target.
Whatever level of aid the UK chooses, the money should measure what it says, the promises around it should be testable, and it should go where it does the most good. None of that has been reliably true over the last five years. The Generational Reset does not tell readers how much aid to give. It argues that whatever is given should be counted accurately and spent where it does the most good.
The Generational Reset is a non-partisan, public-interest project. It is not affiliated with any political party, does not accept corporate funding, and publishes all its work under open licence for public discussion and adaptation.
For public discussion. Not affiliated with any political party. | generationalreset.org
The Generational Reset | S1_14: Overseas Aid | For public discussion. Not affiliated with any political party. | generationalreset.org