Why does the composition of government spending matter?

The short answer

Public spending is not a homogenous economic input. A dollar spent on infrastructure repair, basic research, public health or skills training enters the economy through different transmission channels than a dollar spent on debt service, transfers to high-savings households or pure consumption subsidies. The effect on growth, productivity and long-run debt dynamics differs accordingly.

The Konstantinou (2024) panel study of 33 OECD economies systematised this by separating spending into COFOG functional categories and estimating multipliers for each. The findings confirm what earlier sectoral studies suggested: composition is a first-order determinant of the macro impact, not a detail.

This has policy implications, but also analytical ones — aggregate fiscal indicators alone cannot tell us much about the growth-productivity outlook unless we know what is being financed.

→ New to fiscal analysis? Financial education hub

What the data shows

The Konstantinou OECD panel work, alongside earlier studies by the IMF and ECB, reveals a clear dispersion of multipliers across spending categories.

Key figures (Konstantinou 2024 OECD panel, IMF Fiscal Monitor):

• Health expenditure multiplier: ~1.0 to 1.5, with longer-horizon productivity effects
• Education expenditure multiplier: ~0.8 to 1.4 in the medium term, larger in long-run growth-accounting frameworks
• Defense and public-order multipliers: typically above 1 in panel estimates, with high country variation
• General-purpose transfers and social protection: typically 0.3 to 0.7, depending on targeting
• Government wage bills (excluding investment): generally lower multipliers, often near 0.5

The exception worth noting is that these are average panel estimates: country-specific institutional capacity strongly affects whether spending in a given category translates into productive output. A health expenditure dollar in a country with weak procurement systems behaves very differently from one in a country with strong institutional capacity.

→ Dataset: US Federal Debt to GDP

Why it happens — the macro mechanism

Composition matters because public spending categories differ along three dimensions that determine their economic effect.

The productive-capacity channel. Spending that augments physical capital (infrastructure), human capital (education, training, basic health) or knowledge capital (research) raises the supply side of the economy as well as the demand side. Over a multi-year horizon, this category of spending shifts the production possibilities frontier. Spending that funds current consumption — by recipients or by the government itself — supports demand without expanding capacity, generating short-run multipliers but limited long-run effects.

The propensity-to-spend channel. Transfers reach the economy via household behaviour, and household behaviour is heterogeneous. Transfers to liquidity-constrained households are spent quickly and locally, generating high short-run multipliers; transfers to high-saving households are partially saved or used to deleverage, with smaller short-run macro impact. This is the angle most underappreciated outside academic circles: untargeted transfers can have surprisingly small multipliers because they reach savers as much as spenders.

Note that these channels can reinforce each other. Targeted social spending in a recession can produce both high short-run multipliers and meaningful longer-run productive effects.

The institutional-leakage channel. All spending categories are subject to leakages — to imports, to rents, to inefficient procurement. The size of these leakages varies systematically by category. Defense procurement is typically import-intensive in small open economies; infrastructure spending depends on the rigour of project appraisal and execution; transfer programs depend on targeting accuracy. Konstantinou’s panel results partly reflect average institutional quality across the 33 OECD countries; country-by-country, the magnitudes shift substantially.

Synthesis by regime: in advanced economies near full employment with high public debt, the case for composition discipline is at its strongest — spending categories with high productive-capacity content (infrastructure, R&D, education) can offset the multiplier compression from a fiscal-dominance regime. In recessions with output gaps and accommodative monetary policy, the relative case for composition weakens because slack-channel multipliers are high across most categories. The implication is that the right composition is itself state-dependent, not a universal optimum.

The fiscal headline tells you what was spent. The composition tells you what was actually built.

→ Framework: Macro-financial regimes

What it means for different economic actors

Sovereign-debt investors face a question that goes beyond the deficit headline: is the additional debt being used to finance assets that raise future tax capacity, or current consumption that does not? The 1990s East-Asian and Irish experiences with infrastructure-led debt accumulation contrast with episodes of consumption-financed debt that produced weaker growth response — though disentangling these effects econometrically is genuinely difficult.

Equity investors in cyclical sectors see different exposures by composition. Infrastructure spending benefits materials and capital-goods producers; defense spending benefits a narrow set of contractors; education and health spending have more diffuse beneficiaries. Sectoral allocation responses to fiscal announcements typically reflect these compositional channels more than aggregate spending levels.

Macro analysts should resist treating “deficit reduction” or “fiscal expansion” as homogenous policy levers. Two packages of equal size can have substantially different macro effects depending on what is cut or added.

A common error is to evaluate fiscal packages by headline magnitude alone. The composition often dominates the magnitude in determining the macro outcome.

Practical observation

Go deeper

Related questions

• What is the multiplier effect of government spending?
• Why do wealthy nations have persistent deficits?
• How do automatic stabilizers work during recessions?
• How do primary deficits differ from total deficits?

Frequently asked questions

Is investment spending always more productive than consumption spending?

Not always. Poorly executed investment — bridges to nowhere, white-elephant infrastructure, low-quality construction — can have lower multipliers than well-targeted transfers to liquidity-constrained households in a recession. The category labels matter less than the project-level economics. The Konstantinou and IMF results are panel averages that can mask significant within-category dispersion.

How can investors observe composition in real time?

Most national statistical offices publish quarterly or monthly breakdowns of government spending by economic category (compensation of employees, intermediate consumption, investment, transfers). The OECD and Eurostat publish COFOG functional breakdowns annually with some lag. For the US, the BEA NIPA tables and the Treasury’s Monthly Treasury Statement provide more granular data.

Does composition matter equally for fiscal expansions and consolidations?

The literature suggests yes, often with mirror-image effects. Spending-based consolidations, particularly those that protect investment categories while cutting current spending, have been associated in some IMF and ECB studies with smaller GDP costs than tax-based consolidations of equal size. Alesina, Favero and Giavazzi’s work has been particularly influential — though contested — on this point.

Last updated — 21 July 2026