Fiscal multiplier

We have treated the primary balance as a dial the government turns. But turning it has consequences for g, and how large those consequences are is the subject of the fiscal multiplier, one of the most contested numbers in economics and one where an honest teacher must show you both the theory and the wide empirical uncertainty.

Definition. The fiscal multiplier is the change in real GDP produced by a one-unit change in a fiscal instrument:

Fiscal multiplier = ΔY / ΔG      (for a change in government spending G)

A multiplier of 1.5 means a ₹100 crore increase in government spending raises GDP by ₹150 crore; a multiplier of 0.6 means it raises GDP by only ₹60 crore.

How is the fiscal multiplier calculated?

The intuition: why there is a multiplier at all. Recall M7.01's Fact 2: one person's spending is another's income. When the government builds a road, the ₹100 crore it pays becomes income for the construction firm and its workers. Those workers re-spend a fraction of it, the marginal propensity to consume (MPC), on groceries and phones, which becomes income for shopkeepers, who re-spend a fraction again, and so on down a geometric chain. Summing the chain gives the textbook closed-economy multiplier:

Simple multiplier = 1 / (1 − MPC)      e.g. MPC = 0.6 → 1 / 0.4 = 2.5

Every rupee leaks a bit at each round into savings, taxes, and imports (spending on foreign goods creates foreign income, not domestic), and the more it leaks the smaller the multiplier. A fuller version with a tax rate t and an import propensity m is 1 / [1 − MPC·(1 − t) + m], which shrinks the 2.5 toward, say, 1.2 for a leaky open economy. But the static formula is only the skeleton. The real determinant of whether the multiplier is big or small is not algebra. It is the regime the economy is in.

When is the fiscal multiplier greater than 1?

When the multiplier is greater than 1 (fiscal policy is potent). Spending pulls extra output into existence rather than just reshuffling it:

• Deep recession with slack. When factories sit idle and workers are unemployed, government spending mobilizes otherwise-wasted resources: the road gets built using labor and cement that would otherwise have produced nothing. No private activity is displaced, so the full multiplier chain runs.
• The zero lower bound (ZLB). When the central bank has already cut rates to near zero and cannot offset the stimulus by hiking (M7.02), the boost to demand is not neutralized. It compounds. Multiplier estimates at the ZLB are commonly 1.5 to 2.5 (Ramey's survey; Auerbach–Gorodnichenko's state-dependent estimates).
• High-MPC recipients. A rupee to a poor household or an unemployed worker is almost entirely re-spent (MPC near 1); a rupee to a rich household or a corporation is largely saved. So targeted transfers (unemployment benefits, food support) have far higher multipliers than untargeted tax cuts.
• Productive public investment. Infrastructure that raises the economy's productive capacity (power, ports, roads) can have a multiplier well above 1 over a multi-year horizon, because it lifts private-sector productivity too. This is the analytical heart of India's "capex push": RBI and IMF studies estimate India's capital-expenditure multiplier at roughly 2.4 to 4.5 over several years, versus a revenue-expenditure (subsidy/salary) multiplier below 1 (illustrative ranges from RBI research). If those numbers are even roughly right, shifting a rupee from subsidy to capex is close to a free lunch for growth, and that is exactly the bet Indian fiscal policy has been making.

When is the fiscal multiplier less than 1?

When the multiplier is less than 1 (fiscal policy is weak or self-defeating). Spending mostly displaces other activity rather than adding to it:

• Full employment. With no slack, the government can only get resources by bidding them away from the private sector, so public spending crowds out private spending (the mechanism of crowding out), and the extra demand shows up as inflation, not real output.
• Monetary offset. If the central bank is targeting inflation and not at the zero bound, it will raise rates to neutralize a fiscal stimulus it deems inflationary. In the limit, the central bank fully offsets fiscal policy and the multiplier collapses toward zero, which is the "monetary offset" critique. This is why fiscal stimulus is potent at the ZLB (where the central bank can't offset) and feeble at full employment (where it will).
• Import leakage. In a small, open economy, much of the stimulus leaks abroad, since spending on imported goods boosts foreign output. India's import intensity (especially oil) is a real leakage.
• Ricardian equivalence (the saver's response). If households are forward-looking and believe today's deficit means tomorrow's higher taxes, they save a tax cut rather than spend it, blunting the multiplier. Pure Ricardian equivalence (multiplier ≈ 0 for tax cuts) is an extreme that rarely holds fully, but it captures why tax-cut multipliers are smaller than spending multipliers: the first rupee of government spending is 100% spent by definition, while the first rupee of a tax cut is only partly spent (the rest saved). The tax multiplier is roughly −MPC / (1 − MPC), smaller in magnitude than the spending multiplier 1 / (1 − MPC).

How large is the fiscal multiplier in practice?

The empirical bottom line, stated honestly. Credible estimates of fiscal multipliers span a wide range, roughly 0.3 to 2.5, and the spread is not measurement failure; it is genuine state-dependence. The multiplier is high in slumps and at the zero bound, low at full employment and under an active inflation-targeting central bank; higher for spending than tax cuts, for investment than consumption, for the poor than the rich. The single most consequential real-world episode was the Eurozone austerity of 2010–2013: policymakers assumed multipliers around 0.5 and were stunned when output fell far more than predicted. Blanchard and Leigh's 2013 IMF study concluded the true multipliers had been closer to 0.9–1.7, which is why austerity into a zero-bound recession backfired, deepening the recessions it was meant to cure and (via the doom loop) raising the debt ratios it was meant to lower. An analyst who remembers one thing about multipliers should remember this: their size depends on the regime, and getting the regime wrong is how well-intentioned fiscal policy destroys economies.