Skip to main content

What is the Keynesian multiplier math?

The Keynesian multiplier is the mathematical relationship showing how an initial injection of government spending ripples through an economy, creating more total economic activity than the original amount spent. When the government spends $1, that dollar doesn't simply disappear—it becomes someone's income, which they spend again, creating further income and spending in a chain reaction. The multiplier quantifies this amplification effect, turning a $10 billion spending boost into $20 billion, $30 billion, or more in total economic output, depending on consumption behavior and economic conditions.

Quick definition: The Keynesian multiplier measures how much total economic output increases for every dollar of initial government spending, calculated as 1 divided by the marginal propensity to save (or 1 divided by the sum of the marginal propensity to save plus the tax rate plus the import propensity).

Key takeaways​

  • The multiplier formula: Multiplier = 1 ÷ (1 − Marginal Propensity to Consume) or equivalently 1 ÷ MPS (Marginal Propensity to Save)
  • A multiplier of 2.0 means $1 of government spending yields $2 of total economic growth
  • Real-world multipliers typically range from 0.8 to 1.5, lower than simple textbook models predict
  • Leakages—savings, taxes, and imports—reduce the multiplier's size and speed
  • The multiplier works in both directions: spending cuts shrink the economy by a multiple amount
  • Multiplier size depends on economic slack, the interest rate environment, and consumer confidence
  • Empirical estimates vary widely; recent research suggests smaller multipliers during normal times, larger during recessions

How the Keynesian multiplier works: the basic chain reaction​

Imagine the government hires workers to build roads and pays them $100,000 in total wages. These workers now have $100,000 in new disposable income. If they spend 80% of it ($80,000) on groceries, restaurants, and services, those businesses receive $80,000 in new revenue. The owners and workers at those businesses now have additional income; they spend 80% of their share, generating another $64,000 in spending, and so on.

This endless chain of re-spending is the multiplier effect. The original $100,000 government spending has, theoretically, triggered $500,000 in total economic activity (because 1 ÷ 0.2 = 5). The multiplier is 5.

The formula for this simple, open-economy multiplier is:

Multiplier = 1 ÷ (1 − MPC)
         or = 1 ÷ MPS

where MPC = Marginal Propensity to Consume
      MPS = Marginal Propensity to Save = (1 − MPC)

If consumers spend 80% of each extra dollar (MPC = 0.80), then they save 20% (MPS = 0.20). The multiplier is 1 ÷ 0.20 = 5. Every dollar of initial spending grows to $5 of total income and output.

The role of the marginal propensity to consume​

The marginal propensity to consume (MPC) is the fraction of an additional dollar of income that people spend rather than save. It is the engine of the multiplier. Higher MPC means more spending in each round, a longer chain reaction, and a larger multiplier.

Consider two economies:

  • Economy A: MPC = 0.90 (90% of extra income is spent)
    • Multiplier = 1 ÷ (1 − 0.90) = 1 ÷ 0.10 = 10
  • Economy B: MPC = 0.60 (60% of extra income is spent)
    • Multiplier = 1 ÷ (1 − 0.60) = 1 ÷ 0.40 = 2.5

In Economy A, a $1 billion government spending increase generates $10 billion in total output growth. In Economy B, the same $1 billion boost produces only $2.5 billion in total output. The difference lies entirely in household spending behavior.

The MPC is not fixed. It depends on income level (poorer households spend a higher fraction of extra money), wealth, confidence about the future, and interest rates. During recessions, when households are pessimistic and saving aggressively, the MPC falls, the multiplier shrinks, and government spending becomes less potent. During booms, when confidence is high, the MPC rises and the multiplier expands.

Real-world leakages reduce the multiplier​

The textbook formula assumes a closed economy with no taxes, no saving beyond consumption, and no imports. Real economies have three major leakages that drain spending before the full multiplier can work:

1. Saving (and financial flows) Not all extra income is spent immediately. Some households buy bonds, invest in stocks, or save in bank accounts. These savings are "leaked" out of the spending stream. This is already captured in the MPS term of the formula.

2. Taxes Governments collect income tax and sales tax on every round of spending. If the tax rate is t, then only (1 − t) of each new dollar reaches households as spendable income. A 25% tax rate cuts the effective MPC by 25%.

The adjusted multiplier becomes:

Multiplier = 1 ÷ (MPS + t)

Example: MPS = 0.20, tax rate = 0.25

  • Multiplier = 1 ÷ (0.20 + 0.25) = 1 ÷ 0.45 = 2.22

Without taxes, the multiplier would be 5. Taxes reduce it to 2.22.

3. Imports When domestic consumers spend extra income on imports (cars, electronics, food from abroad), that spending leaves the country and does not generate domestic income. If the import propensity is m (the fraction of each dollar spent on imports), the multiplier formula expands again:

Multiplier = 1 ÷ (MPS + t + m)

Example: MPS = 0.20, tax rate = 0.25, import propensity = 0.15

  • Multiplier = 1 ÷ (0.20 + 0.25 + 0.15) = 1 ÷ 0.60 = 1.67

The multiplier has fallen from 5 (textbook closed economy) to 1.67 (realistic open economy with taxes). A 25% leakage occurs in each round of spending, greatly reducing the amplification.

Worked example: the US fiscal stimulus of 2021​

In early 2021, the US government enacted the American Rescue Plan (ARP), an approximately $1.9 trillion fiscal stimulus during a partial economic recovery from the COVID-19 recession. Most households received $1,400 stimulus checks; businesses received grants and loans; and state/local governments received transfers.

Assumptions for the multiplier calculation:

  • Average US MPC for stimulus recipients: ~0.6–0.75 (lower-income households receiving checks spend roughly 60–75% of the stimulus)
  • US average tax rate: ~20%
  • US import propensity: ~15%
  • Effective leakage rate: 0.20 + 0.15 = 0.35 (already incorporated into the MPC through after-tax spending)

Simplified calculation: If we use an empirical multiplier estimate of 1.0–1.5 (reasonable for a partial-recovery environment, accounting for slack labor markets but offsetting Federal Reserve tightening expectations), the $1.9 trillion spending injection would yield:

  • Low estimate: $1.9 trillion × 1.0 = $1.9 trillion additional output
  • Mid estimate: $1.9 trillion × 1.2 = $2.28 trillion additional output
  • High estimate: $1.9 trillion × 1.5 = $2.85 trillion additional output

In 2021, US GDP was roughly $23 trillion, so the stimulus would add roughly 0.8% to 1.2% to the annual growth rate—a material, but not overwhelming, boost. The actual 2021 GDP growth was 5.9%, driven also by pent-up demand from COVID reopening and monetary accommodation from the Federal Reserve.

Why real-world multipliers are smaller than textbooks predict​

Empirical macroeconomists have found that actual multipliers are typically 0.8 to 1.5 in normal times, far smaller than textbook predictions of 2–5 because of additional real-world frictions:

Inflation expectations and monetary policy If the Federal Reserve views the stimulus as overheating the economy, it may tighten monetary policy—raising interest rates to cool spending. Higher borrowing costs offset some of the stimulus's demand boost, shrinking the multiplier. During the 2021 stimulus, the Fed kept rates near zero, which supported a higher multiplier.

Finite resources and supply constraints If the economy is already at full employment and factories are running flat-out, extra government spending cannot increase output without hitting supply bottlenecks. Prices rise instead of quantities. The multiplier applies to real output; inflation absorbs part of the stimulus. In the 2021 example, supply-chain disruptions meant that some stimulus money chased limited goods, driving inflation rather than output growth.

Crowding out Large government spending often requires increased borrowing, which pushes up interest rates and crowds out private investment. This partially offsets the stimulus's multiplier effect.

Forward-looking behavior Rational households anticipate that larger government deficits today mean higher taxes (or inflation) tomorrow. They may save rather than spend the stimulus to offset future tax burdens. This Ricardian equivalence hypothesis, though debated, suggests that multipliers decline when households are forward-looking.

Dependency on slack The multiplier is larger during recessions with high unemployment and idle capacity, and smaller during booms with tight labor markets. The US economy in 2021 was recovering but still had pandemic-related slack, so multipliers were moderate.

Multipliers during recessions vs. booms​

Research by economists including those at the International Monetary Fund has found that multipliers are state-dependent:

  • Recession multipliers: 1.5 to 2.0+

    • High unemployment and idle factories mean extra government spending directly increases output without hitting supply limits.
    • Monetary policy is typically accommodative (low rates, low inflation).
    • Household and business confidence are low, so the multiplier is stable.

  • Boom multipliers: 0.5 to 1.0

    • The economy is near full capacity; extra spending raises prices instead of output.
    • The central bank may tighten to prevent overheating, crowding out private investment.
    • Forward-looking behavior intensifies as households fear future taxation.

This is why macroeconomists argue that countercyclical fiscal policy—spending more in downturns, less in booms—is effective. Stimulus in a recession works; stimulus in a boom is wasteful or inflation-inducing.

The multiplier and crowding out: competing forces​

The Keynesian multiplier assumes that the initial government spending does not displace private spending. But in reality, if the government borrows to finance the spending, it competes with private borrowers for credit, driving up interest rates and reducing private investment. This crowding-out effect can partially or fully offset the multiplier.

If the government spends $100 billion by borrowing, it might:

  • Create a $150 billion multiplier boost (multiplier = 1.5) in Keynesian demand.
  • But simultaneously crowd out $80 billion in private investment due to higher interest rates.
  • Net effect: $150 billion − $80 billion = $70 billion in additional output.

The true net multiplier is roughly 0.7, not 1.5. The relative strength of the Keynesian multiplier and crowding-out effects determines the final outcome. Crowding out is weaker when interest rates are already low (less room for further increases) and stronger when rates are high (more room to rise).

Time lag: when does the multiplier kick in?​

The multiplier chain reaction does not happen instantly. Policymakers must first propose and pass legislation, then funds must be disbursed to contractors and workers, who then spend the income. Economic statisticians must collect data and calculate GDP.

Typical lags:

  1. Recognition lag: 2–3 months (time to recognize a recession and propose fiscal stimulus)
  2. Implementation lag: 2–6 months (time for Congress to debate and pass legislation)
  3. Spending lag: 3–12 months (funds flow to contractors, workers, or transfer recipients over time)
  4. Statistical lag: 1–3 months (GDP data is published with a delay)

By the time the multiplier reaches its peak effect, the economic situation may have changed. A stimulus passed during a deep recession might be hitting the economy at full force just as the recovery is underway, at which point it risks overheating the economy rather than stabilizing it. This timing problem is a major criticism of discretionary fiscal policy.

Automatic stabilizers and built-in multipliers​

Not all government spending requires legislative action. Automatic stabilizers—like unemployment benefits, food stamps, and income tax brackets—adjust spending automatically during economic downturns without new laws.

When unemployment rises, more people receive benefits and pay less income tax, boosting disposable income immediately. The multiplier for automatic stabilizers is typically 0.5 to 1.0 but kicks in faster than discretionary stimulus because there is no implementation lag. Some economists prefer automatic stabilizers to discretionary stimulus precisely because of timing reliability.

The paradox of thrift through the multiplier lens​

The multiplier reveals a famous Keynesian paradox. If all households try to increase their savings during a recession (reducing MPC), they collectively reduce their spending. This multiplies downward through the economy, shrinking income further and defeating the original savings goal. Households that save more see their incomes fall, leaving them unable to save as much as they intended.

This paradox suggests that during recessions, government spending is a solution: by maintaining demand, it prevents the downward multiplier spiral and allows households to actually save more.

Recession triggered by loss of confidence
↓ Households try to save more (reduce spending)
↓ Lower spending = lower incomes everywhere
↓ Paradoxically, savings decline despite higher savings rate
↓ Government stimulus maintains demand, stabilizing incomes
↓ Households can then save without triggering a spiral

Real-world examples of multiplier effects​

1. New Deal spending (1933–1941) The US government hired millions of workers through programs like the Works Progress Administration. Estimates suggest multipliers of 1.5 to 2.0. The ratio of output growth to spending was smaller than textbook formulas predicted, partly because monetary policy was still tight (the gold standard constrained the Federal Reserve's ability to keep rates low).

2. Japanese stimulus, 1990s Japan spent trillions of yen on public works during its "Lost Decade" after the 1990 real estate bust. Empirical studies found multipliers of 0.8 to 1.0, lower than expected, partly because fiscal stimulus competed with private debt deleveraging—households and firms were saving to repair balance sheets, not spending.

3. 2008 financial crisis stimulus The US enacted the American Recovery and Reinvestment Act (ARRA) in early 2009, a $787 billion stimulus. Estimates of the multiplier ranged from 0.5 to 1.5, with most studies settling on 0.9 to 1.2. The lower end reflected supply-chain disruptions and tight credit conditions; the higher end came from studies of specific programs, like infrastructure spending or aid to states.

4. COVID-19 stimulus payments Households that received stimulus checks spent roughly 30–40% of the payment in the short term, implying a near-term MPC of 0.3–0.4. Over longer horizons, accumulated MPC was higher, 0.5–0.75. This suggests that one-time stimulus checks have moderate, lagged multiplier effects.

Common mistakes​

1. Confusing multiplier size with multiplier existence. Critics sometimes claim that because the multiplier is 1.2 instead of 3, stimulus spending is ineffective. This is wrong. A 1.2 multiplier means that $1 billion in spending generates $1.2 billion in total output—a 20% boost, which is economically significant and politically valuable.

2. Assuming the multiplier is constant. The multiplier changes with economic conditions, interest rates, and expectations. Using a 2008 recession multiplier estimate to evaluate 2021 stimulus is misleading. The context matters enormously.

3. Ignoring the direction of the arrow. The multiplier works both ways. A $100 billion spending cut shrinks output by $100 billion to $200 billion, depending on the state of the economy. Austerity policies in weak economies have high negative multipliers and can trigger downward spirals.

4. Forgetting about crowding out. The Keynesian multiplier is the gross effect on demand. Crowding out reduces the net effect on output. Observing that stimulus spending increased total spending doesn't prove the multiplier is large if private investment fell by an equal amount.

5. Mixing up correlation and causation in real data. When government spending increases and output grows, it might seem like the multiplier is working. But if the spending increase occurred during a boom when private demand was already strong, the correlation is weak evidence of the multiplier's strength.

External resources​

FAQ​

What happens to the multiplier if the Federal Reserve raises interest rates?​

The multiplier shrinks. Higher interest rates crowd out private investment and consumer borrowing, reducing the net boost to aggregate demand. The gross Keynesian multiplier might still be 1.5, but the crowding-out effect might offset 0.5, leaving a net multiplier of 1.0 or less.

Is the multiplier bigger for government purchases or for tax cuts?​

Generally, government purchases have larger multipliers than tax cuts. When the government buys goods and services directly, 100% of the spending enters the economy immediately. When taxes are cut, households must choose to spend the extra income—and they save some of it. If the tax-cut MPC is 0.75, a $100 billion tax cut only generates $75 billion in immediate spending, weakening the multiplier.

Can the multiplier be less than 1?​

Yes. If leakages are very large (high savings, high taxes, high imports), or if crowding out is severe, the multiplier can fall below 1. In this case, each dollar of government spending increases total output by less than $1. This is more likely in highly open economies, during boom times, or when government borrowing is seen as a threat to future growth.

How does the multiplier work during hyperinflation?​

It doesn't work the conventional way. When inflation is extremely high, the central bank might not accommodate fiscal stimulus, and the government might struggle to borrow (lenders fear currency depreciation). The multiplier can collapse toward zero or even be negative if stimulus spending triggers a currency crisis that reduces real output.

Does the multiplier apply equally to all types of spending?​

No. Infrastructure spending and hiring workers often have higher multipliers (1.5+) because the money directly enters working households' incomes, with high MPCs. Subsidies to large corporations or tax cuts for the wealthy often have lower multipliers (0.5–1.0) because recipients have lower MPCs and may save or spend abroad.

  • Learn about the crowding-out effect and how government borrowing can offset fiscal stimulus: ../chapter-08-fiscal-policy/08-crowding-out-effect
  • Understand the crowding-in effect, where government spending can boost private investment: ../chapter-08-fiscal-policy/09-crowding-in-effect
  • Explore budget deficits and how persistent spending gaps shape national fiscal policy: ../chapter-08-fiscal-policy/10-budget-deficit-explained
  • Examine unemployment and how fiscal stimulus specifically targets job creation: ../chapter-06-business-cycle/05-unemployment
  • Learn how monetary policy and the Federal Reserve interact with the multiplier: ../chapter-07-monetary-policy/02-interest-rates-inflation
  • Discover the economy's short-term vs. long-term behavior in the business cycle: ../chapter-06-business-cycle/01-introduction-business-cycle

Summary​

The Keynesian multiplier is a mathematical framework showing how an initial injection of government spending multiplies through the economy via repeated rounds of income and spending. The simple formula, Multiplier = 1 ÷ (1 − MPC), reveals that economies with higher consumption propensities experience larger multiplier effects. However, real-world multipliers are dampened by taxes, savings, imports, and crowding-out effects, typically ranging from 0.8 to 1.5 in normal times. The multiplier is state-dependent, stronger during recessions and weaker during booms. Timing lags and forward-looking behavior further complicate the effect, which is why the debate over fiscal stimulus's efficacy remains unsettled in modern economics.

Next​

→ The crowding-out effect