Aggregate demand and aggregate supply

Lessons 5 and 6 explained the long run, in which output is determined by productivity, capital and labor and prices adjust fully. But in the short run the economy can produce more or less than its potential, as recessions and booms show. To analyze these fluctuations, economists use the model of aggregate demand and aggregate supply (AD-AS). It looks like the supply and demand diagram of ECON-101, but the axes measure the economy as a whole: real GDP on the horizontal axis and the overall price level on the vertical axis. The curves have different explanations from those of a single market, and confusing the two is a common mistake. This lesson builds the model, uses it to analyze demand shocks and supply shocks, and introduces the spending multiplier.

Aggregate demand

Aggregate demand (AD) is the total quantity of final goods and services that households, firms, governments and foreigners want to buy at each price level: \( C + I + G + NX \). The AD curve slopes downward, but not for the reason a single demand curve does. In a single market, a higher price makes buyers switch to other goods; for the whole economy there is nothing else to switch to. Instead, three effects explain the slope. The wealth effect: a higher price level reduces the real value of money and other nominal assets, so households feel poorer and spend less. The interest rate effect: a higher price level raises the demand for money to make transactions, which tends to push up interest rates and reduce investment and interest-sensitive consumption; in modern practice, the central bank also tends to raise interest rates when prices rise faster. The exchange rate effect: higher domestic prices make domestic goods more expensive relative to foreign goods, reducing net exports.

The AD curve shifts when spending changes at any given price level. It shifts right when consumer or business confidence rises, when household wealth rises because of higher stock or house prices, when the government increases purchases or cuts taxes, when the central bank lowers interest rates, when foreign incomes rise and boost exports, or when the currency depreciates. It shifts left in the opposite cases. Many of the most important events in macroeconomics, such as the collapse of investment and consumption in 2008–2009, or the surge in spending supported by fiscal programs and low rates in 2021, are shifts in aggregate demand.

Aggregate supply in the long run

The long-run aggregate supply (LRAS) curve is vertical at potential output. In the long run, the quantity of output the economy can produce depends on labor, capital, natural resources and technology, as lesson 5 explained, not on the price level. If all prices and wages doubled, real output would be unchanged. The LRAS curve shifts right when the labor force grows, when capital accumulates, when education improves or when technology advances, and it shifts left if productive capacity is destroyed, for example by a natural disaster or a war.

Aggregate supply in the short run

The short-run aggregate supply (SRAS) curve slopes upward: in the short run, a higher price level is associated with more output. The reason is that some prices and costs are sticky. Wages are often set by contracts or adjusted only once a year; many firms change prices infrequently because of menu costs or concern about customer relations; and workers and firms may not immediately recognize that a rise in their own prices reflects a general rise in prices. If the price level rises unexpectedly while wages are fixed, firms' profit margins improve and they expand output. If prices fall unexpectedly, margins shrink and firms cut production and employment.

The SRAS curve shifts when production costs change at any given price level. It shifts left (up) when wages rise, when the prices of key inputs such as oil rise, when expected inflation rises (because workers and firms build expected price increases into wages and prices), or when supply chains are disrupted. It shifts right when input costs fall or productivity improves. A leftward shift of SRAS caused by a sudden rise in costs is called an adverse supply shock.

The economy is in long-run equilibrium when AD intersects SRAS at a point on the LRAS curve: output equals potential, and the price level is what people expected, so there is no pressure on wages and prices to change faster than expected. In the short run, AD and SRAS can intersect to the left of LRAS (a recessionary gap) or to the right (an inflationary gap). Over time, wages and prices adjust and the economy moves back toward potential, though this adjustment may be slow, which is why policy can play a stabilizing role.

A demand shock

Suppose a fall in business confidence reduces investment, shifting AD to the left. In the short run, the economy moves down along SRAS: output falls below potential, unemployment rises, and inflation falls. The economy has a recessionary gap. What happens next? With output below potential and unemployment high, wage growth slows and firms cut prices or raise them more slowly. Lower costs gradually shift SRAS to the right, returning output to potential at a lower price level (or, with ongoing inflation, a lower inflation rate). This self-correcting process may take years, however, and in the meantime unemployment imposes real costs. Alternatively, the central bank can cut interest rates or the government can raise spending to shift AD back to the right, speeding recovery. A positive demand shock works in reverse: output rises above potential, unemployment falls below the natural rate, and inflation rises, until rising costs shift SRAS left or policy tightens.

A supply shock

Now suppose a sharp rise in oil prices raises production costs across the economy, shifting SRAS to the left. In the short run, output falls and the price level rises: the economy suffers stagflation, a combination of stagnation and inflation. Supply shocks are much harder for policymakers than demand shocks, because output and inflation move in opposite directions. If the central bank cuts interest rates to support output, it adds to inflation; if it raises rates to fight inflation, it deepens the slowdown. If it does nothing, the economy may adjust as the shock fades or as wages adjust, but inflation will be above target for a time.

The 1970s provide the classic example. Oil prices quadrupled in 1973–1974 and rose sharply again in 1979–1980, contributing to high inflation and recessions in Canada, the United States and most other oil-importing countries. Because central banks had accommodated earlier inflation, expectations had become unanchored, and the shocks set off a wage-price spiral. Bringing inflation down required very high interest rates in the early 1980s, which caused deep recessions. The lesson that central banks drew was that anchored inflation expectations make supply shocks much less costly, because a one-time rise in prices does not turn into persistent inflation.

The 2026 oil shock posed this dilemma again. Higher gasoline prices pushed headline CPI inflation to 3.7 % in September 2026, while payroll growth slowed sharply. Because inflation had been above the Fed's 2 % target for more than five years, the FOMC worried that expectations could drift upward and raised its target range in September. The United States is now a net exporter of petroleum, so higher oil prices raise incomes in producing states such as Texas and North Dakota, but most consumers and many businesses lose, and the overall effect on growth is still negative in the short run.

Inflation expectations and the shifting SRAS

Expected inflation shifts the SRAS curve. If workers and firms expect 2 % inflation, wages and prices rise by about 2 % a year even with output at potential. If they come to expect 5 %, they build 5 % into contracts, and the SRAS curve shifts up faster. This is why central banks watch surveys of expected inflation and financial market measures closely. When expectations are anchored, temporary shocks fade; when they are not, inflation becomes self-perpetuating. Lesson 12 develops this idea with the Phillips curve.

The spending multiplier

Changes in spending have effects larger than their initial size, because one person's spending is another's income. If the government spends 100 million dollars more on infrastructure, construction workers and firms earn 100 million more. They spend part of it, say 60 %, on consumer goods, which becomes income for others, who spend 60 % of that, and so on. The marginal propensity to consume (MPC) is the fraction of an extra dollar of disposable income that households spend. In the simplest model, with no taxes or imports, the total increase in spending is \( 1 / (1 - MPC) \) times the initial change.

In real economies, part of each extra dollar of income goes to taxes and part is spent on imports, which leak out of the domestic spending stream. If \( t \) is the marginal tax rate and \( m \) the marginal propensity to import (the share of each extra dollar of income spent on imports), the multiplier becomes smaller.

# Multiplier rounds: initial spending, then successive rounds of induced consumption
mpc, t, m = 0.8, 0.3, 0.25
induced = mpc * (1 - t) - m        # extra domestic spending per extra dollar of income
dG = 10.0                          # billions
round_spend, total = dG, 0.0
for r in range(1, 31):
    total += round_spend
    if r in (1, 2, 3, 5, 30):
        print(f"after round {r:2d}: cumulative GDP change {total:6.2f} billion")
    round_spend *= induced
print(f"formula: 1 / (1 - MPC(1 - t) + m) = {1 / (1 - mpc * (1 - t) + m):.3f}")

after round  1: cumulative GDP change  10.00 billion
after round  2: cumulative GDP change  13.10 billion
after round  3: cumulative GDP change  14.06 billion
after round  5: cumulative GDP change  14.45 billion
after round 30: cumulative GDP change  14.49 billion
formula: 1 / (1 - MPC(1 - t) + m) = 1.449

Real-world multipliers are smaller and less certain than this formula suggests, for several reasons. When the economy is near full employment, extra spending mainly raises prices rather than output. The central bank may raise interest rates to offset fiscal stimulus, crowding out investment. Households may save temporary tax cuts rather than spend them, especially if they expect future tax increases. And the multiplier depends on the type of spending: transfers to low-income households, who spend most of what they receive, tend to have larger effects than tax cuts for high-income households. Empirical estimates of government spending multipliers range widely, from below 0.5 to above 2, with larger values found in deep recessions when interest rates are at their lower bound. Lesson 10 discusses these debates.