When government spending rises, aggregate demand rises with it, first by the exact dollar amount of the new spending and then by an additional amount as that money gets re-spent through the economy. The final size of the boost is where the interesting work happens: the Congressional Budget Office puts the real-world multiplier for federal purchases between 0.5 and 2.5, far below the figures that classroom formulas produce.1Congressional Budget Office. The Fiscal Multiplier and Economic Policy Analysis in the United States Taxes, imports, saving, interest rates, and the state of the economy when the money arrives all pull the effect downward.
The Immediate Dollar-for-Dollar Shift
Aggregate demand is the total spending on all final goods and services in an economy, and it adds up four categories: household consumption, business investment, government purchases, and net exports. Government purchases sit inside that sum, so any increase in them pushes aggregate demand higher by the same dollar amount before anything else happens.
If the federal government awards 10 billion dollars in new highway contracts, the aggregate demand curve shifts to the right by 10 billion dollars at the first stage. That money then enters the circular flow of income. The federal government spent roughly 7 trillion dollars in fiscal year 2025, equal to about 23 percent of GDP, so even small percentage changes in that figure represent enormous sums.2U.S. Treasury Fiscal Data. Federal Spending
How the Multiplier Builds on the First Dollar
The initial shift is only the beginning. When the government pays a construction firm, that payment becomes income for the firm’s workers and suppliers. Those workers spend part of it at grocery stores, restaurants, and other businesses, creating income for a second group. The second group spends part of what it receives. The chain continues in diminishing rounds.
The fraction of each new dollar of disposable income that households spend rather than save is the marginal propensity to consume, or MPC. If the MPC is 0.75, the textbook multiplier equals 1 divided by (1 minus 0.75), which comes to 4. Under that formula, every dollar of new government spending would eventually generate four dollars of total activity. If the MPC falls to 0.5, the multiplier drops to 2. The textbook formula is useful for seeing the mechanism, but it assumes away several forces that cut the real-world figure substantially.
Why Real Multipliers Are Much Smaller
The CBO’s estimated range of 0.5 to 2.5 for direct federal purchases is well below the textbook figures of 4 or 5. A multiplier of 1.0 means each dollar of spending produces exactly one dollar of additional output. Anything below 1.0 means the spending partially displaces private economic activity rather than adding to it on top.
The gap between theory and reality comes from taxes, imports, interest rate effects, and how much slack the economy has when the spending lands. Each force bleeds money out of the re-spending chain.
The Type of Spending Changes the Result
Not every dollar enters the economy the same way. Direct purchases, like building a bridge or buying military equipment, put the full dollar into the market for goods and services immediately because the government is buying something real.
Transfer payments work differently. The government hands money to individuals, who then choose how much to spend and how much to save. Because recipients save some portion, not all of the dollar enters the spending stream. The CBO puts the multiplier for transfer payments to individuals at 0.4 to 2.1, slightly below the 0.5 to 2.5 range for direct purchases.1Congressional Budget Office. The Fiscal Multiplier and Economic Policy Analysis in the United States
The difference widens further for other categories. One-time payments to retirees carry a multiplier of just 0.2 to 1.0, and tax cuts aimed at higher-income individuals come in at 0.1 to 0.6. Higher-income households save a larger share of windfalls, so less money recirculates. A dollar spent hiring workers to repair infrastructure moves aggregate demand more than a dollar sent as a rebate to someone who parks it in a savings account.
Leakages That Drain Each Round
Economists call the forces that pull money out of the domestic spending chain leakages. Three main ones matter.
- Taxation. When a worker earns new income, a share goes to federal and state income taxes before any spending happens. Higher marginal rates shrink the effective MPC and the resulting multiplier.
- Imports. When households buy goods made abroad, that money leaves the domestic economy. A country with a high propensity to import sees a weaker domestic multiplier because each round sends more dollars overseas.
- Saving. Income set aside in savings accounts, retirement funds, or debt repayment exits the chain. Empirical estimates of the U.S. marginal propensity to consume vary widely depending on the population studied, running anywhere from 0.05 to 0.9, which explains much of the range in multiplier estimates.
These leakages compound at every round. By the third or fourth pass, the amount being re-spent domestically is often a small fraction of the original injection.
Crowding Out From Government Borrowing
When the government finances new spending by borrowing rather than raising taxes, it competes with private borrowers for the available pool of savings. The Treasury issues bonds to cover the gap, and that extra demand for funds pushes interest rates higher.3U.S. Department of the Treasury. Financing the Government
Higher interest rates make borrowing more expensive for businesses weighing new factories, equipment, or hiring, so some projects that would have been profitable at lower rates get shelved. Households pull back on interest-sensitive purchases like homes and cars. The increase in government spending gets partially offset by a decrease in private investment and some consumption. This is crowding out.
How severe it gets depends on how sensitive private investment is to interest rate changes. If businesses have few alternative funding sources and operate on thin margins, even a small rate increase can kill projects. If capital markets are deep and firms are flush with cash, the effect is milder.
Do Households Cancel the Stimulus?
A more radical challenge to the multiplier comes from Ricardian equivalence, a theory arguing that deficit-financed spending may not boost aggregate demand at all. The logic: rational households know today’s borrowing must be repaid through higher future taxes, so when deficits grow, forward-looking households save more now to cover the eventual tax bill rather than spending their current income.
If the theory held perfectly, every dollar of deficit-financed spending would be matched by a dollar of reduced private consumption, and fiscal policy would be useless as a stimulus tool. In practice, Ricardian equivalence is more a theoretical benchmark than a description of reality. Most people do not calculate their lifetime tax liabilities and adjust their saving accordingly. Credit constraints, short planning horizons, and uncertainty about future tax policy all weaken the mechanism. Some households do save more when they see deficits grow, so the effect exists at the margin, but it does not come close to erasing the stimulus.
Time Lags Before the Effect Lands
Even when a spending increase will eventually boost aggregate demand, the gap between the decision and the economic impact can be long. Policymakers need time to recognize that stimulus is needed, Congress and the executive branch need time to pass legislation, and authorized money takes time to actually flow out the door.
The Infrastructure Investment and Jobs Act shows how long that last stage can take. As of early 2026, the Department of Transportation had obligated about 73 percent of the law’s funding under binding agreements with recipients, but only about 43 percent had actually been paid out.4US Department of Transportation. Infrastructure Investment and Jobs Act (IIJA) Funding Status The law was signed in November 2021. More than four years later, over half the money still had not reached the economy. Large infrastructure projects move through design, permitting, and phased construction, with payments trickling out across years.
These lags mean fiscal stimulus often arrives after the recession it was designed to fight has already ended. Spending that was meant to pull the economy out of a slump can land in a recovery, adding to inflation rather than output.
Why the Economy’s Starting Point Decides the Final Size
The same dollar of government spending produces very different results depending on whether the economy is in a slump or running near capacity. In a deep recession, unemployment is high, factories sit idle, and businesses have room to expand production without bidding up prices. New spending puts idle resources to work, generating real output gains with little inflationary pressure. The multiplier tends toward the higher end of its range.
The International Monetary Fund has found that fiscal stimulus using public spending is “particularly potent when there is economic slack” and interest rates are low.5International Monetary Fund. Countering Future Recessions in Advanced Economies When the central bank is already holding rates near zero, crowding out largely disappears because government borrowing cannot push rates much higher from an already-low baseline.
The picture reverses near full employment. With most workers already employed and factories running near capacity, new government demand competes with private demand for the same limited resources. The result is higher prices rather than higher output. The multiplier shrinks, and in extreme cases falls below 1.0, meaning the spending displaces more private activity than it creates.
Monetary policy decides much of the outcome. If the Federal Reserve accommodates fiscal expansion by keeping rates low, the multiplier is larger because crowding out stays muted. If the Fed responds to inflationary pressure by raising rates, it actively works against the fiscal expansion, shrinking the multiplier further. The interaction between fiscal and monetary policy often matters more than either policy in isolation, and that is where simple multiplier stories go wrong.