Bilateral support for Ukraine, 2022–2026

Kiel Institute Ukraine Support Tracker (Release 30v2), 24 Jan 2022 – 30 Jun 2026, in euros. Sources last checked 28 Sep 2026.

Measure

Period

Show

Russia

All data — drill down

Notes

Allocated = support delivered or designated for delivery; Committed = publicly announced. Annual figures follow the tracker's dating, so commitments and allocations for the same package can fall in different years.
GDP: IMF World Economic Outlook nominal GDP (USD, current prices; 2025–2026 are IMF estimates/projections), converted to euros at ECB annual average USD/EUR rates (2026: Jan–Jun average).
2026 is a half year (to 30 June). Its % of GDP is measured against half of annual GDP so it is comparable with full years. Period totals use the same weighting.
Groups: EU institutions (Commission & Council, EIB, European Peace Facility) are compared against EU-27 GDP. EU member states are the 27 members' own bilateral aid. Other Europe: UK, Norway, Switzerland, Iceland. Rest of world: Canada, Japan, Australia, South Korea, New Zealand, Taiwan, Türkiye, China, India. China's and India's GDP is left out of the Rest of world and all-donor GDP totals, because their very small aid set against their very large economies would drag the group's % of GDP down. Their aid (about €5m) is still counted, and their own % of GDP still appears in the country rows.
EU views (top buttons): Separate shows EU institutions and member states as in the source data. EU as one merges institutions and all 27 member states into one donor, "European Union", measured against EU-27 GDP. Countries only hands the EU institutions' aid (Commission & Council, EIB, Peace Facility) back to the member states in proportion to their net contribution to the EU budget that year. Only net payers receive a share (Germany ≈ 45–50%, France ≈ 21%, Italy ≈ 7–10%, then the Netherlands, Sweden, Austria, Ireland, Denmark, Finland, and Luxembourg in 2023). Net contribution = national contributions (excluding customs duties) minus EU spending in that country (excluding administration). Source: European Commission, EU spending and revenue 2000–2024; 2025–2026 reuse the 2024 key.
Military aid includes in-kind equipment as valued by the tracker; it is not the same as government cash outlays.

Government debt and interest

What is shown: three columns in the table, and the hover box of each dot in the chart Government debt against support, for the 27 EU member states, Norway, Iceland and the United Kingdom. The EU member states row (and European Union in the EU as one view) shows the EU-27 figure. For the EU, the source is Eurostat: the figures countries report under the EU's budget rules (tables gov_10dd_edpt1, gov_10q_ggdebt, gov_10q_ggnfa). There are no figures here for the other donors, and none for the Other Europe group as a whole.
Norway, Iceland, United Kingdom use the same definitions, from these sources:
- Norway: Eurostat. Its debt figure jumps at each year end (41.8% in September 2025, 54.4% in December), so the year-end level overstates the level during the year. Eurostat's figures do not show the cause. Norway's government also holds far more assets (its oil fund) than debt.
- Iceland: interest from Eurostat. Eurostat has no debt figure for Iceland, so debt comes from Statistics Iceland (government securities and loans as % of GDP). That exists per year only, so there is no 2026 figure yet; over all years the table shows the 2025 level.
- United Kingdom: the Office for National Statistics (ONS), which still publishes debt and interest on the definitions the UK used when it reported to Eurostat. About a quarter of UK debt is linked to inflation, and the yearly inflation uplift counts as interest. That is why UK interest peaked at 4.3% of GDP in 2022.
Gov. interest % GDP is what the government paid in interest on its debt in the period, as a share of GDP. It is a yearly cost, so it compares directly with support as % of GDP. Over several years it is the average, weighted by GDP.
Gov. debt % GDP is everything the government owes in total, as a share of one year's GDP. It is a level at one moment, not a yearly cost: the end of the selected year, or the latest figure when all years are selected.
- What counts (the EU's "Maastricht" definition): government bonds and short-term paper, which are by far the largest part; loans, for example from banks, other countries or the European rescue funds; and deposits held with the government plus coins in circulation, usually a small item.
- How it is counted: at face value, meaning the amount to be repaid at the end, not the bond's market price. Gross, so assets are not subtracted: Norway shows 49% debt while its oil fund is worth many times that. For the whole government (central, regional, local and social security funds), leaving out what one part owes another.
- What is left out: future pensions of government staff and other pension promises, unpaid bills to suppliers, guarantees that have not been called, and the debt of state-owned companies outside the government sector.
- Decimals: the sources publish the debt ratio to one decimal. For the EU's full years (2022–2025) the second decimal is worked out from Eurostat's own amounts of debt and GDP, which reproduce the published figure. For 2026, and for Norway, Iceland and the United Kingdom, only one decimal exists, so the second is shown as 0.
- The gap can be large. Iceland's government had total liabilities of 88% of GDP in 2025. Of that, 57% was bonds and loans (the figure shown here), 25% pension obligations and 6% unpaid bills.
of which allocated / committed support (pp) answers one question: if every euro of support had been borrowed, how much of the debt in the column beside it is support? It is in percentage points of GDP, so the two columns subtract directly. Denmark over 2022–2026: debt 26.8% of GDP, of which support 2.77 points, so debt without the support would be 24.0%. It takes the support in the selected period and the debt at the end of that period, and follows the Measure and EU switches; in Countries only it includes each country's share of the EU institutions' aid.
- Relative to the debt itself, the same Danish figure is an increase of 11.5% (2.77 on 24.0). Countries that owe little see the largest relative increase: Denmark and Estonia. France and Italy stay close to zero partly because their debt is so large.
- It is an upper limit. Part of military aid is equipment from existing stocks, counted at the tracker's valuation but not paid for with new borrowing. Part of financial aid is a loan to Ukraine, so the government holds a claim against it. In reality support is also paid from taxes.
- It leaves out the interest on the borrowed amount, which works the other way but is small over four years.
- Iceland has no 2026 debt figure, so over all years its support is set against the debt at the end of 2025 (56.7% of GDP).
The chart "Government debt against support" puts those two columns against each other, one dot per country, and follows the Measure, Period and EU switches.
- Across is in effect the support as % of GDP: debt % GDP × support / debt leaves support / GDP. So the chart asks whether countries with a high debt give less.
- The thin line at 60% is the ceiling for government debt that EU countries agreed in the Maastricht Treaty (with 3% of GDP for the yearly deficit). It binds the 27 member states; Norway, Iceland and the United Kingdom are shown against it only for comparison.
- In Countries only, part of each net payer's support is its share of the EU institutions' aid. That part is drawn as a darker slice of the dot, like a piece of pie, starting at the top and running anticlockwise. The Netherlands over 2022–2026: 0.95 points of its own, 1.49 with its EU share, so (1.49 − 0.95) / 1.49, a little over a third of the dot, is darker (37% before rounding). Countries that are not net payers get no slice.
- Dot size follows the country's GDP in the last year of the selected period, on a logarithmic scale: each doubling of GDP adds the same amount to the dot's radius. So size shows the order of the economies, not their proportions. Germany's GDP is about 180 times Malta's, but its dot is less than four times as wide. With areas in true proportion, the small countries would be invisible.
- No average line is drawn, on purpose. A straight line through the dots accounts for only about a fifth of the differences in debt between countries (2022–2026), so it would suggest an average relation that is hardly there. Countries that give much tend to owe less, but Finland, Belgium and the United Kingdom give a fair amount with a high debt, and Malta, Ireland and Bulgaria give little with a low one.
- What pattern there is, is not a cause. The largest givers are the Nordic and Baltic countries, which are close to Russia and also had low debt before 2022.
2026: Eurostat has published so far. Debt is the level at the end of that quarter. Interest is what was paid over the four quarters ending with that quarter, against the GDP of the same four quarters. So the 2026 interest figure still consists largely of 2025, and it moves slowly.
- Why not the 2026 quarters alone? Interest payments fall unevenly over the year in several countries. France paid 1.8% of GDP in the first quarter of 2025 and 2.6% in the second; Hungary 5.0% in the first and 3.2% in the fourth. One quarter set beside a full year would show a fall or a rise that is only the season.
- Norway, Iceland and the UK are shown up to the same quarter.
Repayments are not included. Governments rarely pay debt down: loans that fall due are replaced by new ones. Interest is the part that actually takes room in the budget.
The two do not move together: interest ≈ debt × the average rate a country pays. Hungary has about half Greece's debt but pays more interest, because it borrows at much higher rates.

Russia's military spending: what the line means and how reliable it is

War extra (default; switch War extra / Total at the top): the line shows only what Russia spends above its pre-war level, which is a better measure of what the war costs Russia than the whole military budget. The pre-war level is 2021's share of GDP (3.6%) applied to each year's GDP, so a year counts as "extra" only for spending beyond what Russia would have spent anyway at pre-war intensity.
- Why not simply subtract the 2021 amount (4.9 trillion roubles)? Russian prices rose about 45% between 2021 and 2025, so a fixed rouble amount would count inflation as war cost. Keeping the 2021 share of GDP adjusts for both inflation and economic growth. Adjusting the 2021 amount for inflation only gives similar results.
- The extra is still an approximation: without the war, Russia's spending might have drifted up or down anyway, and part of the extra goes to rearmament that Russia will keep after the war.
- The extra rises from about 1.1% of GDP in 2022 to 3.9% in 2025, falling to 2.7% in the 2026 budget.
What is shown: Russia's military spending (switch at the top: exchange rate, purchasing power, or hidden) as estimated by SIPRI (Stockholm International Peace Research Institute; Cooper, A budget for a fifth year of war, SIPRI Insights 2026/01, table 8). SIPRI rebuilds it from budget laws, Finance Ministry reports and the hidden budget lines. It includes the "National defence" budget chapter plus military pensions, social support for troops, the National Guard (Rosgvardiya) and Border Service, and military housing, health and education. Converted to euros at the Bank of Russia's annual average rate (the ECB stopped quoting the rouble in 2022).
2022–2026 H1: Russia spent about on its military at market exchange rates, or about at purchasing power. About ( at purchasing power) of that is above its pre-war level of 3.6% of GDP, which is a rough measure of what the war itself added. Allied allocated aid over the same period: . Hover the dots for each year.
Fairly reliable: the order of magnitude and the trend. Spending rose from 3.6% of GDP in 2021 to 7.5% in 2025, the highest since the Soviet era. The totals come from Russia's own budget laws and Treasury reports, and several independent analysts reach similar figures.
Less reliable:
- Secrecy: since 2023 roughly a third of the federal budget is classified, and SIPRI marks its 2023–2024 figures as the author's estimates.
- 2025 is the amended budget, not final spending.
- 2026 is the budget law as passed (14.9 trillion roubles, 6.3% of GDP). It was already amended twice in 2025 and may well be again. It is shown here as half the annual budget (dashed line) to match the Jan–Jun aid data.
Not included, so the real cost is higher:
- payments by Russia's regional governments (sign-up bonuses for recruits, compensation to the wounded and families of soldiers killed)
- state-directed cheap bank loans to arms makers, outside the budget
- equipment losses
- the wider economic cost: about $300bn of central-bank reserves frozen abroad, lost trade and energy income under sanctions, workers drawn into the army, high inflation and interest rates
Exchange rate versus purchasing power (switch at the top): a rouble buys far more soldiers' wages, fuel and domestically made equipment inside Russia than the exchange rate suggests. The Purchasing power line converts roubles with the IMF's PPP rate (about 26–35 roubles per international dollar, against 72–100 roubles per euro in the market), so it shows what Russia's spending would cost at US prices. It is then converted to euros. The result is roughly 2.5–3 times the exchange-rate figure, in line with the IISS (International Institute for Strategic Studies), which put 2024 at about $146bn at exchange rates and about $462bn at purchasing power.
- Treat the exchange-rate line as a lower bound and the purchasing-power line as closer to Russia's real effort, but probably on the high side. PPP rates are built from a general basket of consumer and investment goods, not weapons. Imported components, microchips and sanctions-busting purchases cost Russia world prices or more.
- The rouble was also unusually strong in 2022 because of capital controls, which inflates that year's exchange-rate figure.
- % of GDP is the same either way: spending and GDP are both in roubles, so any conversion changes both by the same factor. Purchasing power changes the amount, not the share (4.8–7.5% of GDP in 2022–2025).
Russia's government debt is low: of GDP at the end of 2021 and in 2026 (IMF World Economic Outlook, general government gross debt; the latest years are projections). Russia is left out of the chart Government debt against support, because its dot would mislead:
- Its place across would be the war extra over 2022–2026 H1, about percentage points of GDP (total military spending: ). That is about four times the largest donor (Denmark, 2.77), and it is Russia's own war spending, not support to another country.
- "If all of it was borrowed" plainly does not hold. The war extra is more than half the size of the entire government debt, yet that debt rose by only points. Russia cannot borrow abroad and pays for the war in other ways: oil and gas income, higher taxes, and its National Wealth Fund, whose liquid part fell from $113.5bn before the invasion to about $44bn in October 2026 (Finance Ministry figures, as reported by The Moscow Times and TASS).
- Part of the borrowing is kept off the government's books. A law of February 2022 lets the state order banks to lend to war-related companies on terms it sets. One estimate, by Craig Kennedy (Russia's Hidden War Debt, 2025), puts those loans at $210–250bn. They count as company debt, not government debt.
- The debt figure itself is fairly reliable: the Finance Ministry publishes it and the IMF uses it. It is close to, but not the same as, the Maastricht definition used for the other countries.
Not like for like: Russia's figure is a whole national military budget, including forces not fighting in Ukraine and the nuclear arsenal. The aid figures are what donors allocated to Ukraine, and military aid includes in-kind equipment at the tracker's valuation. Ukraine's own defence budget is not included either. Read the comparison as scale, not as a balance sheet.