ECB staff estimate a defence spending multiplier near one and higher inflation
Victoria Shinder2 min readAn article in the European Central Bank's Economic Bulletin, published on 24 September, looks at what the European rearmament now under way means for the euro area economy. Prepared by ECB economists Cristina Checherita-Westphal, Marta Rodríguez-Vives, Tibor Lalinský and Miles Parker, it starts from the commitments: NATO members agreed in June 2025 to spend 5% of GDP a year by 2035, 3.5% on core defence and 1.5% on related security, and reaffirmed that in July 2026.
Actual spending is far below that. Euro area defence spending rose slightly in 2025 to almost 1.5% of GDP by the national-accounts measure, and only four euro area countries were above 2%. According to the JANES budget database cited in the article, aggregate euro area spending could rise by almost one percentage point of GDP over the next decade to 2.5% in 2035, driven first by procurement and later by personnel, operations and maintenance. The additional spending already announced since early 2025 is in the Eurosystem's June 2026 projections and amounts to 1.5% of GDP cumulatively over 2025-28.
The macroeconomic estimates are the part most relevant to monetary policy. An empirical analysis across EU countries from 1999 to 2025, summarised in the article, finds a cumulative GDP multiplier of around one at its peak. Price effects are positive and significant: a 1% of GDP rise in defence spending lifts HICP inflation cumulatively by up to 1.3 percentage points in the third year in the authors' preferred specification. The authors stress that this depends heavily on what is bought, that the GDP effects carry wide confidence bands, and that delivery delays for military equipment - a median of two years and up to four in their data - would push GDP effects later and inflation earlier.
What it means
For the ECB the question is not whether defence spending is desirable but how it shows up in prices and output, and the article gives a staff estimate to work with. A multiplier around one means spending adds to output roughly one for one at the peak; an inflation effect of that size means a sustained build-up is not neutral for price stability, particularly if orders run into supply constraints.
Two cautions from the article itself are worth keeping in view. Spending that goes to imported equipment supports demand abroad rather than in the euro area, so the domestic effect depends on where contracts land. And the numbers are averages over past episodes with wide uncertainty bands; they describe what has happened before, not a forecast of this build-up.