How Fiscal Policy Works

Idea

Fiscal policy works by moving the components of aggregate demand. Government expenditure ($G$) is itself a component; taxes and transfers work on consumption ($C$) and investment ($I$) by altering households’ disposable income and firms’ post-tax profits. A change in any of these shifts the aggregate-demand curve, and the change in real national income is amplified by the multiplier.

Mechanism

Expansionary fiscal policy. The government raises $G$, raises transfer payments (lifting disposable income, and so $C$ and $I$), and cuts taxes (again lifting $C$ and $I$). As $C$, $I$ and $G$ rise, aggregate demand shifts right. The initial injection raises national income; households re-spend part of the additional income, raising $C$ further, and the cycle of spending and re-spending continues with diminishing increments — the multiplier effect — so national income rises by a multiple of the initial injection.

Contractionary fiscal policy. The reverse: $G$ is cut, transfers are reduced, and taxes are raised, lowering $C$, $I$ and $G$. Aggregate demand shifts left, and the reverse multiplier contracts national income by a multiple of the initial withdrawal.

Automatic stabilisers achieve a milder version of the same thing without any deliberate act. Because income tax is progressive, a rising-income boom pushes households into higher tax brackets while unemployment-benefit transfers fall, so $G$ falls and $C$ rises more slowly — damping demand-pull inflation. In a downturn the mechanism runs the other way: falling incomes attract lower tax rates and rising transfers, slowing the fall in aggregate demand.

Diagram

Limitations

• Size of the multiplier. The smaller the multiplier $k$, the smaller the eventual change in national income from a given injection, and the weaker the policy.

• Size of $G$ relative to AD. If government expenditure is a small share of total aggregate demand, $G$ must rise more than proportionately to move AD appreciably — and not every government has the budget for that.

• Size of external demand. In a small, open economy external demand can dwarf domestic demand, and fiscal policy does not act on foreign demand for exports.

• Time lags. Large projects take time to plan and execute, so the effect of an expansion may arrive only after the economy has already recovered, risking overheating.

• Tax (in)sensitivity. When confidence is low in a downturn (or high in a boom), $C$ and $I$ respond less than proportionately to a tax change.

• Crowding out. Financing a deficit by borrowing from financial markets reduces the funds available to households and firms, raising interest rates and depressing private $C$ and $I$ — government demand crowds out private demand. (Its size is disputed; in a deep slump private agents may be too pessimistic to borrow anyway.)

• Government debt. Persistent deficits accumulate debt that must be serviced with interest, carrying an opportunity cost and, at the extreme, the risk of sovereign-debt default — the risk austerity measures are designed to head off.

• Fiscal drag. Automatic stabilisers are not selective: the same mechanism that cools an overheating economy can also slow a recovery, and they merely dampen fluctuations rather than reverse them.

Context

Singapore illustrates several of these limits at once. Government expenditure was around 11–12% of GDP in 2022, below the world average, and external demand is large relative to domestic demand — so fiscal policy has a comparatively small grip on Singapore’s aggregate demand, which is one reason the authorities lean on supply-side and exchange-rate instruments instead.