Skip to main content

Types of government spending explained

Government spending is not a single bucket; it comprises dozens of distinct programs and categories, each with different purposes, constituencies, and economic effects. Defense spending supports military personnel and weapons. Healthcare spending (Medicare, Medicaid) pays for medical care for seniors and the poor. Education spending funds schools and universities. Infrastructure spending builds roads, bridges, and utilities. Social Security sends checks to retirees and the disabled. Understanding the composition of government spending is essential because not all spending has equal economic impact, and spending priorities reflect deep choices about what a government values and what it believes will most effectively support growth.

Quick definition: Government spending consists of expenditures on defense, healthcare, education, infrastructure, social security, interest on debt, and other programs. These fall into mandatory (required by law) and discretionary (appropriated annually) categories.

Key takeaways​

  • The U.S. federal budget totals roughly $6.3 trillion (as of 2024), split between mandatory and discretionary spending.
  • Mandatory spending (Social Security, Medicare, Medicaid) is controlled by law and grows automatically; it now accounts for <60% of the budget.
  • Discretionary spending (defense, education, infrastructure, non-defense agencies) requires annual appropriation by Congress; it accounts for roughly 30% of the budget.
  • Interest on the national debt is a growing share, now <15% of the budget and rising as debt levels grow.
  • Spending composition differs sharply between countries and reflects different policy priorities.
  • The economic multiplier of spending varies by type: spending on lower-income programs has higher multipliers than spending on defense.

Mandatory spending and entitlements​

Mandatory spending refers to programs controlled by permanent law rather than annual appropriations. Once Congress enacts a law establishing a program (e.g., Social Security), the spending is "mandatory"—it continues until Congress changes the law. No new vote is needed each year. The four largest mandatory programs are Social Security, Medicare, Medicaid, and CHIP (Children's Health Insurance Program).

Social Security is the largest single spending program, consuming roughly 21% of the federal budget (approximately $1.3 trillion in 2024). It provides retirement checks to 68 million Americans age 65 and older, survivor benefits to their families, and disability benefits. Workers and employers each pay a 6.2% payroll tax (12.4% combined) into the system, and benefits are funded from incoming payroll taxes plus a shrinking trust fund. Social Security is politically sacrosanct; any attempt to raise taxes or cut benefits triggers fierce opposition, so the program's future solvency problem (the trust fund is projected to deplete in the early 2030s) has not been resolved.

Medicare is the second-largest mandatory program, providing health insurance to 66 million Americans age 65 and older plus some younger disabled beneficiaries. Medicare consists of Part A (hospital insurance), Part B (doctor and outpatient), Part D (prescription drugs), and Parts C and E (Medicare Advantage and other managed care). Total Medicare spending is roughly 15% of the federal budget (about $848 billion in 2024). Medicare is financed partly by payroll taxes (2.9% combined employer-employee), partly by beneficiary premiums, and partly by general revenue. Like Social Security, Medicare faces long-term solvency challenges because the aging population is growing while the working-age population is shrinking.

Medicaid is a joint federal-state program that provides healthcare to lower-income Americans—children, pregnant women, elderly, disabled, and other groups. Unlike Medicare (a federal program), Medicaid is administered by states, which set eligibility and payment rates, but the federal government matches a portion of spending. Total Medicaid spending is roughly 10% of the federal budget (about $616 billion in 2024), and it has been the fastest-growing major spending program over the past 20 years. During the COVID-19 pandemic, Medicaid enrollment surged, and during the subsequent unwinding, millions lost coverage as the government ended the pandemic emergency declaration.

CHIP, the Children's Health Insurance Program, covers non-Medicaid-eligible children in lower-income families. It is much smaller than Medicare or Medicaid, consuming roughly 1% of the budget (about $17 billion in 2024). CHIP is jointly funded and administered by states, similar to Medicaid, and it has achieved high enrollment in most states, reducing the uninsured rate among children significantly.

All four programs face similar pressures: aging populations increase per-person costs, new treatments and drugs are expensive, and political pressure to expand benefits is persistent. Yet reform is politically difficult. Any benefit cut is seen as breaking a promise to beneficiaries; any tax increase is seen as burden on workers.

Discretionary spending​

Discretionary spending refers to programs that require annual appropriation by Congress. Congress must vote each year to fund these programs; if Congress does not appropriate, the spending stops. The largest discretionary programs are defense, non-defense federal agencies (State Department, EPA, FBI, etc.), and various grants to states for education, transportation, and other purposes.

Defense spending (officially "National Defense" within the Pentagon and related military programs) is the largest single discretionary item, consuming roughly 13% of the federal budget (about $821 billion in 2024). This includes salaries for 2.9 million military personnel and civilians, purchases of weapons and equipment, maintenance of bases, and military operations. The U.S. defense budget is larger than the next 10 countries' defense budgets combined, though it is small relative to GDP (roughly 3.5% of GDP in 2024). Defense spending is cyclical; it rises during wars and periods of strategic concern, and falls during periods of relative peace. The wars in Iraq and Afghanistan (2001–2021) drove it to <4% of GDP at their peak; the end of active major combat and the pivot toward great-power competition with China has kept it elevated.

Education and training includes federal spending on K–12 schools (mostly through grants to low-income districts), colleges and universities (research funding, student loans, Pell Grants), vocational training, and other educational programs. Total education spending is roughly 2% of the federal budget (about $114 billion in 2024), though most K–12 education is funded by states and localities. Federal spending has grown as a share of education costs because states have shifted burden to the federal government.

Transportation and infrastructure includes highways, public transit, airports, and ports. Federal spending here is roughly 1% of the budget (about $64 billion in 2024), though recent years have seen increases. The Infrastructure Investment and Jobs Act (2021) allocated roughly $110 billion over five years for roads, public transit, and broadband, but much of that was new authorization, not immediate spending. Like education, most infrastructure in the U.S. is funded by states and localities, so federal spending is supplementary.

Other non-defense discretionary programs include the EPA, FBI, State Department, National Institutes of Health, NASA, small-business loans, food assistance (SNAP), housing assistance, and hundreds of other agencies and programs. Collectively, these account for roughly 6% of the federal budget (about $380 billion in 2024). Many of these programs are popular with specific constituencies but lack broad political priority, so they remain small.

Interest on the national debt​

Interest on the national debt is a growing but often overlooked budget item. The U.S. national debt is now roughly $34 trillion (as of 2024), and the federal government pays interest on this debt to bondholders. Interest payments have climbed as debt has grown and interest rates have risen. In 2024, interest payments are roughly 13% of the budget (about $659 billion) and rising rapidly. Within a decade, interest payments are projected to exceed defense spending and rival Social Security as the largest single budget item.

High interest payments are problematic because they do not fund any services or infrastructure; they are pure transfer to bondholders. Every dollar spent on interest is a dollar not available for schools, roads, or medical research. When interest payments dominate the budget, fiscal flexibility shrinks. If a recession hits and policymakers want to run a deficit to stimulate the economy, they must borrow more, which raises interest rates further and creates a vicious cycle.

Comparing spending across countries​

Government spending as a share of GDP varies sharply across developed countries. Denmark spends roughly 50% of GDP as government spending; the Nordic countries average around 48%. The OECD average is roughly 39%. The U.S. spends roughly 38% of GDP as government spending (federal, state, and local combined). Germany spends roughly 45%, and the UK roughly 42%. Japan, despite having long pursued fiscal stimulus, spends roughly 39%.

These differences reflect different values. Scandinavian countries prioritize universal healthcare, generous pensions, and strong social safety nets, so they tax and spend heavily. The U.S. relies more on private healthcare and pensions, so it spends less on government. But the U.S. also spends much more on defense (as a share of GDP) than most other developed countries, offsetting some savings elsewhere.

Differences in spending composition also matter for economic impact. A country that spends heavily on education and healthcare likely builds more human capital and has greater long-term growth potential. A country that spends heavily on pensions (transfer to retirees who spend little) may have lower immediate stimulus impact than one spending on education (investment that pays off later) or infrastructure (immediate demand stimulus).

Spending by fiscal year and supplemental appropriations​

The federal budget is set annually, typically for the fiscal year (October 1 through September 30). Congress debates a budget resolution, then passes appropriations bills for the major spending categories (defense, non-defense, etc.). The entire process ideally wraps up before the fiscal year begins, but in practice, Congress often passes continuing resolutions (short-term funding bills) that keep the government running at previous spending levels while debates continue.

In emergency situations—recessions, wars, pandemics—Congress passes supplemental appropriations outside the regular budget process. The CARES Act (2020), American Recovery and Reinvestment Act (2009), and various emergency spending bills for wars and disasters are supplements. These allow policymakers to respond quickly to crises without waiting for the full budget process.

Supplemental spending can be problematic because it lacks the scrutiny of the regular budget process. Money can be allocated to favored projects or industries without full vetting. On the other hand, waiting for the regular budget process during a crisis means people suffer in the meantime.

Spending multipliers and economic impact​

Not all government spending has the same economic impact per dollar. The fiscal multiplier varies depending on the type of spending and the state of the economy.

Spending on transfer programs (Social Security, unemployment benefits, SNAP) that go to lower-income households has high multipliers, often <2.0 or greater. Why? Because lower-income households have high marginal propensity to consume (MPC)—they spend nearly all of any additional income. A $100 payment to a low-income household results in roughly $80–90 spent immediately, stimulating demand. Higher-income households have lower MPC; they might spend only $30–40 of a $100 payment, saving the rest.

Spending on infrastructure (roads, bridges, schools) has multipliers typically between 1.5 and 2.0 if the projects are well-designed. The government contracts with construction firms, which buy materials and hire workers. Those workers spend their paychecks, creating demand for other goods. The multiplier is substantial but takes time to materialize (projects must be planned, approved, and constructed).

Spending on defense is complex. Military salaries go to workers who spend them (high MPC), but weapons purchases go to large corporations that save a larger fraction. Studies estimate defense spending multipliers between 0.8 and 1.5 depending on circumstances.

Tax cuts have multipliers typically between 0.8 and 1.5, lower than spending on transfers or infrastructure. Why? Because not all of a tax cut is spent; some is saved, particularly if households are uncertain about future income.

These multipliers assume that the economy has slack (unemployment above natural rate, capacity underutilized). When the economy is already at full capacity, multipliers are close to zero or even negative; fiscal stimulus just bids up prices without increasing real output.

Real-world examples​

U.S. federal spending during COVID-19 (2020–2021). In response to the pandemic, Congress passed five major relief bills totaling over $5 trillion. The CARES Act ($2.2T) included stimulus checks ($1,200–1,800 per person), enhanced unemployment ($600/week), Paycheck Protection Program for businesses, and aid to states. The American Rescue Plan ($1.9T) in 2021 added more stimulus checks, extended unemployment, and aid to schools and states. The rapid spending injection (totaling roughly 25% of 2020 GDP) averted a depression, but it also contributed to a post-pandemic inflation spike as demand bounced back faster than supply could adjust. The composition mattered: direct payments to households had high multipliers because households spent them immediately; the fiscal multiplier of this spending was estimated at 1.5–2.0.

Japan's Lost Decade (1990s). After a property bubble burst, Japan ran large budget deficits to stimulate growth. Spending went heavily into public works (roads, bridges, dams) with mixed results. Much of the spending went to rural areas with low returns; the multiplier was estimated at 1.0–1.5, moderate at best. The spending did prevent depression but did not generate strong growth. By the 2000s, critics argued that Japan had wasted trillions on white-elephant projects.

German austerity (2010s). Germany passed a balanced-budget constitutional amendment and cut spending as a share of GDP during the 2010s. The result was sluggish growth and weak demand, especially relative to core European growth. German austerity may have prolonged the eurozone crisis by reducing a major source of demand for southern European exports. This illustrates how spending composition and timing interact: austerity in a weak economy raises unemployment and reduces tax revenue, making the fiscal problem worse, not better.

The New Deal (1930s). President Franklin D. Roosevelt's spending programs (Public Works Administration, Civilian Conservation Corps, Works Progress Administration) employed millions and undertook major infrastructure projects. The New Deal did not end the Depression (unemployment remained near 10% in 1939), but it provided relief, redistributed income to lower-income groups, and left lasting infrastructure (dams, national parks, schools). The fiscal multiplier of New Deal spending is debated; estimates range from 0.7 to 2.5 depending on methodology.

Common mistakes​

Assuming all spending is equally productive. Spending on bridges that reduce commute time is productive; spending on a bridge to nowhere is not. The economic impact depends on the return on investment, which varies sharply by project. Policymakers often struggle to distinguish between the two, leading to wasteful spending during booms and insufficient infrastructure investment during normal times.

Confusing government spending with government size. A government that spends heavily may be large or small depending on what it spends on. A government that spends 40% of GDP on pensions and healthcare (transfers to retirees) is not necessarily larger in productive capacity than one that spends 35% of GDP on education and infrastructure. The size of the productive apparatus (teachers, researchers, engineers, public facilities) is what matters for long-term growth, not the share of GDP spent.

Forgetting about crowding out. When government borrows to finance spending, it absorbs credit from the market, raising interest rates. Higher rates can discourage private investment in factories and equipment. Large government spending can thus crowd out private investment, offsetting some of the fiscal stimulus. This effect is larger when interest rates are already high and credit is tight.

Ignoring the composition of entitlements growth. Medicare and Medicaid are growing faster than Social Security, driven by rising per-person healthcare costs. Policymakers sometimes speak as if all entitlements are the same, but the drivers are different. Social Security's problem is demographic (more retirees, fewer workers); healthcare programs' problem is rising unit costs (the price of care per person is climbing faster than inflation). Fixing the two requires different approaches.

Assuming deficit spending always stimulates. Deficit spending can stimulate demand, but only if it injects purchasing power into the private sector. If the government borrows $100 billion and spends it on long-term infrastructure projects that take years to complete, the immediate stimulus is small. If the government borrows $100 billion and distributes it as checks to households, the stimulus is immediate and large. The timing matters.

FAQ​

How much of the federal budget goes to "welfare" or "foreign aid"?​

A common misconception is that welfare and foreign aid are large budget items. Social Security and Medicare (transfer programs) together account for roughly 36% of the budget. Means-tested programs (Medicaid, SNAP, EITC, housing assistance) together account for roughly 15% of the budget. Foreign aid accounts for less than 1% of the budget (roughly $50–60 billion). When Americans are polled on the budget, many overestimate both welfare and foreign aid, and underestimate healthcare and Social Security.

Why can't we just cut "waste" and balance the budget?​

Every year, politicians promise to eliminate "waste" and balance the budget without raising taxes or cutting major programs. The reason they fail is that waste is real but small, while major programs (Social Security, Medicare) are large. To balance the budget without raising taxes, the government would need to cut Social Security benefits by roughly 30%, or Medicare by 35%, or defense by 60%, or some combination. These are politically extremely difficult.

Is deficit spending a burden on future generations?​

Deficit spending today means the government borrows money that must be repaid from future tax revenue. To the extent that borrowed money is invested in productive assets (roads, schools, research), future generations benefit. To the extent that it is spent on current services or transfers, future generations bear the burden of repaying without gaining the benefit. Interest payments on debt are a pure transfer from future taxpayers to current bondholders.

How do spending multipliers work when the economy is at full capacity?​

In a fully employed economy with no slack, fiscal stimulus does not increase real output; it increases prices. A government spending increase bids up wages and materials prices. Instead of expanding production, firms pass costs to consumers. The multiplier is close to zero or negative (inflation reduces real purchasing power, offsetting stimulus). This is why fiscal stimulus is most effective during recessions when there is unemployment and underutilized capacity.

What happened to discretionary spending over time?​

As a share of the budget, discretionary spending has fallen sharply as mandatory spending and interest payments have grown. In 1970, discretionary spending was roughly 60% of the budget; by 2024, it was roughly 30%. This means policymakers have less flexibility to fund new priorities (education, infrastructure, research) because existing programs (Social Security, Medicare) have grown. The squeeze is projected to worsen as the debt grows and interest payments climb.

Why doesn't the government spend more on education if it boosts long-term growth?​

Education has a high social return (each dollar spent generates multiple dollars of future output) but a low political payoff (benefits accrue over decades, not years). Politicians are rewarded for programs with immediate visible results. Additionally, most education is funded by states and localities, not the federal government, so federal officials have limited ability to increase education spending without constitutional concerns or budget pressure.

How do tax cuts compare to spending increases as economic stimulus?​

Tax cuts have lower multipliers than spending increases because some of a tax cut is saved, not spent (especially by higher-income households). A $1 billion tax cut might increase demand by $800 million–$1 billion; a $1 billion government spending increase might increase demand by $1.2 billion–$2 billion. The composition of the tax cut matters: tax cuts for lower-income households have higher multipliers than cuts for higher-income households.

Summary​

Government spending encompasses hundreds of programs in four main categories: mandatory spending (Social Security, Medicare, Medicaid, CHIP), discretionary spending (defense, education, infrastructure, agencies), interest on debt, and other smaller programs. The composition of spending varies sharply across countries and reflects different policy priorities; Nordic countries spend heavily on social programs, the U.S. spends more on defense, and all developed countries face growing pressure from healthcare and aging-population costs. The economic multiplier of spending varies by type: transfers to lower-income households have high multipliers (1.5–2.5), infrastructure has moderate multipliers (1.0–2.0), and other spending has lower multipliers (0.8–1.5). Understanding spending composition is essential for fiscal policy design because it determines both the immediate stimulus impact and the long-term growth implications.

Next​

→ Discretionary vs mandatory spending