This policy brief is based on a forthcoming paper. The views expressed are those of the authors and not necessarily those of the institutions the authors are affiliated with.
Abstract
European countries plan to boost their defence capabilities to deter external security threats. This implies a substantial change in the volume and composition of fiscal expenditure, as well as the structure of the industrial base. We analyse the macroeconomic implications of alternative policy choices to boost national security. We augment a global dynamic general equilibrium model to include a public good role for defence capabilities, a defence industry with R&D externalities, trade in military equipment, and public investment in dual-use goods. Enhancing European defence capabilities boosts aggregate output (GDP) through increased defence industry production and government value added. However, this reorientation leads to a permanent crowding out of private consumption and a minimal effect of CPI inflation. If the shock is temporary, inflation increases on impact and persistently.
Russia’s invasion of Ukraine has prompted a re-assessment of European defence capabilities necessary to serve as an effective deterrent to security threats. At the same time, the United States have pushed for greater burden sharing of deterrence provision. NATO alliance members have committed to a substantial increase in “core defence requirements and defence- and security-related spending” (NATO, 2025). Achieving an annual target of 5% of GDP by 2035 will have significant macroeconomic implications, even if there is a phased approach to meeting this goal.1
While the high-level aims are clear, there are scant details regarding implementation plans. Given that only a limited number of countries have meaningful defence industries, where will these goods be produced? How will militaries increase the number of available personnel? Through competitive wages, or through conscription? And what are the implications of increased security risks for European firms and households?
In Clancy and Lozej (2026), we examine some of these issues by adding several novel features to an otherwise standard global dynamic general equilibrium model. We motivate a need for defence capabilities to serve as a deterrent to external security threats by including a function that maps risks to the probability that private capital remains safe, and therefore to private investment decisions (Figure 1). A sufficient level of deterrence keeps private capital safe and thus facilitates a smooth functioning of the economy. In this case, security concerns play no role in decision making (point A). However, if defence capabilities fall, an increased probability of damage to the existing and future capital stock in the event of an armed conflict reduces incentives to invest (point B).2
Our function captures two important aspects. First, the size of this effect is non-linear. Firms adjust their behaviour more rapidly the further defence capabilities fall from the level that makes capital safe. Second, there is an asymmetry. Defence capabilities above the level where capital is safe provide almost no additional benefit.
We model defence capabilities (and therefore the size of the deterrent) as a combination of the stock of military equipment (i.e. arms) and dual-use goods, as well as the number of military personnel.3 Importantly, in Europe, the size of the deterrence is a combination of both the regions’ own and the US’ defence capabilities. This feature allows us to capture international interdependencies in security provision, such as amongst NATO members.
A defence sector, utilising domestic labour and (private and public) capital produces defence equipment. Military equipment is tradable, but only purchasable by governments. This means that its price is not part of the consumer price index (CPI) targeted by the Central Banks. Defence firms have some market power due to the differentiated nature of their products and therefore can set prices for their goods in domestic and foreign markets. The production of military equipment generates an R&D externality that potentially benefits the other sectors of the economy. Crucially, this externality is linked to the (domestic) production of military equipment and does not apply to expenditure on arms imports.
Figure 1. Effect of level of deterrence on investment decisions
Governments can invest in dual-use goods. For example, maritime surveillance technologies are useful for national defence, but also search and rescue, environmental protection, fisheries and vessel traffic management. The stock of such goods is available to all sectors of the economy, which generates a small, but positive externality in production. Unlike for defence equipment, the externality from dual-use goods depends on the location of expenditure rather than the location of production.
Finally, governments also decide on the size of the military personnel that depends on either hiring professional soldiers or via conscription. We distinguish between these two cases through the wages they receive: professional soldiers receive the wage linked to the average wage of the economy (plus, potentially, a premium), while conscripts only receive a subsistence payment set by the government.
We calibrate the model to four blocs: Germany, the rest of the euro area, the United States and the rest of the world. While we primarily rely on the existing literature for the calibration, a novel contribution is our extension of work conducted by the Stockholm International Peace Research Institute to provide a time series of trade in military equipment for the period 2001-2020 in financial rather than unit terms. This allows us to assess the extent of bilateral arms trade flows, which are important in determining where the R&D benefits from military equipment spending will materialise.
We further back out, based on our assumption that past military expenditure indicates a revealed preference, what constitutes a “sufficient” level of European defence capabilities. Using the average defence spending to GDP ratios between 2001-2024, we can compute the share of each region in provided deterrence as the weighted sum of their capabilities. The US has by far the largest share, at 78.3%, with Germany providing 5.3% and the rest of the euro area accounting for 16.4% of total European defence capabilities (Figure 2).
We start our analysis with what we see as the impetus for the recent focus on this issue: a reduced contribution of the US to European deterrence. We calculate that compliance with NATO targets, whereby military spending in the US and European NATO members will reach 5% of GDP per year by 2035, would reduce the US share to around 60%. This is the black line in Figure 2.
Figure 2. Share of European deterrence provided by the US
To illustrate how the reduction of defence capabilities works, we assume Europe is at point A in Figure 1 and conduct a counterfactual whereby we assume that the US reduces its share of provision to European defence by 20 percentage points. Crucially, in this case there is no response by European policymakers to actively boost their own contribution to improved defence capabilities. Therefore, there is a rise in security risk. This makes capital in Europe riskier, which reduces investment and induces a reduction in all forms of economic activity. The effect of a reduction in US defence provision is therefore negative, albeit not very large (output falls by about 0.1% in the long run).
We next examine the effects of a permanent increase in spending on defence equipment, equivalent to the NATO-agreed increase in defence spending to 3.5% of GDP in Europe. Europe has historically spent about half of its defence budget on equipment and related expenditure (we include here intermediate goods and materials). Therefore, this change amounts to an increase from the current levels of defence spending that is allocated to equipment equivalent of roughly 1 p.p. of ex-ante GDP. The remainder is allocated to spending on personnel, which we analyse below. We present the results in Figure 3 using the black line.
This change boosts aggregate output through increased defence industry output, necessitating a reallocation of resources from other sectors. Moreover, it also boosts measured GDP through the depreciation of the growing equipment stock, which counts towards output in the national accounts. In the long run, the accumulation of R&D from arms production drives further output growth. Inflation falls initially due to an exchange rate appreciation but later returns to initial levels due to wage increases. This effect, however, is somewhat attenuated by the positive effect on supply capacity from knowledge spillovers due to the increase in R&D. Investment increases, but private consumption falls, as additional resources created by arms production are not available to households.
Defence equipment is not useful without (well-trained) military personnel to operate it. While the initial focus of the policy discussions largely emphasised expenditure targets, NATO members have been less explicit regarding the potential size of increases in their armed forces. We next examine the macroeconomic implications of adjusting the number of military personnel. We scale this increase so that it accounts for the remainder of the increase of defence spending that is missing after increasing defence equipment spending, to bring the overall level to 3.5% of GDP. After accounting for the wages paid to armed forces, this implies an increase in personnel by 38.2% in the REA and by 73.5% in Germany. We plot the results in Figure 3 using the blue line.
An increase in military personnel through professional recruitment leads to greater government value added (wages in the defence sector count towards GDP). However, this increase results in a long-term decline in private sector output due to labour shortages, pushing up wages and reducing productive capacity. While CPI inflation may fall slightly in the short run due to an exchange rate appreciation, the GDP deflator rises in the medium run due to increased government spending on military wages.
Investing in technology and infrastructure that has both a civilian and a military use is a key component of the updated NATO expenditure targets. We simulate an increase in dual-use investment increase that is equivalent to the NATO-agreed targets, which amounts to a permanent increase in dual-use investment of 0.9 p.p. of ex-ante GDP in the euro area. We show the results in Figure 3 using the red line.
Increased productivity in the defence and other industries leads to a rise in aggregate output with no inflationary costs. While consumption decreases, the medium-run decline is somewhat less severe compared to investments solely in defence equipment. Private investment increases significantly due to the broader applicability of capital from dual-use investments. Wages rise as public capital (and the associated productivity benefits) accumulates, and the R&D stock increases due to positive spillovers from defence sector production. Although inflation initially falls due to an exchange rate appreciation and lower import prices, it remains stable in the long run due to the benefits of increased public capital and R&D. This increased activity leads to a mild rise in the GDP deflator.
Figure 3. Permanent increase in European defence capabilities, euro area
(pp and % deviation from initial steady state)
Our analysis above assumes that the defence spending increase is permanent and fully credible. This is different from standard government spending shocks, not only in terms of the composition of spending (usually focused on public consumption or investment), but also in terms of the duration (typically focused on a temporary change). We therefore conduct an experiment where we shock government spending on defence components for five years, after which the spending components return to their initial levels. Figure 4 shows the results.
The results for a temporary shock are more in line with what one would expect for a conventional government spending shock – an increase in output and (CPI) inflation, with crowding-out of private consumption and investment. This crowding-out is the weakest in the case of arms production spending due to the R&D externality, which is also responsible for a small, but long-lasting increase in output. The difference from the permanent shock comes from the response of consumption, whereby it transitions to a new (lower) steady state level and therefore there is a permanent reduction in private demand.4
Figure 4. Temporary increase in European defence capabilities on euro area economy
(pp and % deviation from initial steady state)
For decades, European firms and households did not have to factor external security risks into their decision making. This situation has now changed. NATO members have agreed on several measures to improve defence capabilities and boost deterrence to security threats. These measures involve changes to government spending targets, which are not necessarily comparable with those typically analysed in standard macroeconomic models. In Clancy and Lozej (2026), we develop a modelling framework that allows us to explore the effects of different policy measures agreed by the NATO members and their macroeconomic implications. We find that in all cases aggregate output increases, and that a substantial part of this increase is due to government value added (especially when military personnel increases). Private consumption is crowded out, as defence goods are not available to households. These expenditure increases are not inflationary, if the shift is permanent and credible. However, when the shock is temporary, the responses of output, consumption and inflation are closer to a standard government spending shock, i.e. there is an increase in output and crowding-out of private consumption and investment, while inflation increases.
Clancy, D and M. Lozej (2026), “Spending on the frontline: The macroeconomic implications of rising defence expenditure”, Central Bank of Ireland Research Technical Paper, forthcoming.
Antonova, A., R. Luetticke, and G. J. Muller (2025). The military multiplier. CESifo Working Paper.
Hartley, K. and B. Solomon (2015). Measuring defense output: An economics perspective. In: Military cost-benefit analysis, pp. 70–107. Routledge.
Ilzetzki, E. (2025). Guns and growth: The economic consequences of defense buildups. Technical report, Kiel Report.
NATO (2025). Defence expenditures and NATO’s 5% commitment. https://www.nato.int/cps/en/natohq/topics_49198.htm. Accessed: 2025-09-14.
Members are to allocate 3.5% of GDP to an “agreed definition” of defence expenditure, and the remaining 1.5% of GDP to “protect critical infrastructure, defend networks, ensure civil preparedness and resilience, innovate, and strengthen the defence industrial base.” They are also to submit annual plans to NATO showing a “credible, incremental path” to reaching this target.
Using a similar function, we also include a channel where a less secure environment inhibits private consumption.
Although increased expenditure does not automatically translate into greater military capability (Antonova et al. 2025), our assumption mimics NATO “capability” targets that focus on defence expenditure to GDP ratios. Given the difficulty in mapping defence “inputs” into “outputs” (Hartley and Solomon, 2015), we believe this is a reasonable approach. See Ilzetzki (2025) for a comprehensive discussion of the economics of defence buildups.
A permanent reduction in private consumption also implies a permanent increase in labour supply (at a given wage), and the need to supplement that labour in production with more capital and therefore more investment. This ultimately affects the cost of capital, which increases in the case of a temporary shock and decreases in the case of a permanent shock. This mechanism is the main reason for the different responses of inflation to permanent and temporary changes in defence spending.
