Supply & DemandChapter 8

Aggregate Supply and Demand

The aggregate supply and demand (AD-AS) model is the basic framework of macroeconomics as taught in textbooks. The American economist Gregory Mankiw, a Harvard professor and the author of standard textbooks, presents it as the fundamental tool for understanding short-run economic fluctuations: how the general price level and total output are determined, and how economic shocks and policies affect them.

Last updated: 16 October 2026

Definition

The AD-AS (Aggregate Demand - Aggregate Supply) model relates the general price level to the total quantity of goods and services produced in the economy (real GDP).

The aggregate demand curve (AD) represents the inverse relationship between the price level and the total quantity of goods and services demanded: when prices fall, households consume more, firms invest more and the country exports more, through three combined mechanisms — the wealth effect, the interest rate effect, and the exchange rate effect. The aggregate supply curve (AS) represents the relationship between the price level and the quantity firms are willing to produce.

There are two supply curves. Short-run aggregate supply (SRAS) slopes upward: firms produce more when prices rise more than expected, because some wages and prices are set in advance, or because for a while they mistake a rise in the general price level for a rise in their own price. Long-run aggregate supply (LRAS) is vertical, at the level of potential GDP.

How it works

The textbook model, as Gregory Mankiw presents it, explains a country's total output and the general price level by two forces: what all buyers want to buy (aggregate demand) and what firms want to produce (aggregate supply). It describes what happens when a shock hits one or the other, assuming two things: in the short run, some prices and wages take time to adjust; in the long run, output depends only on labor, capital, natural resources and technology, not on demand.

A shock hits demand or supply.
1. A shock hits demand or supply
Buyers suddenly want to buy more or less at unchanged prices (demand shock), or producing becomes more or less costly, for instance when oil prices soar (supply shock).
In the short run, prices and wages follow slowly.
2. In the short run, prices and wages follow slowly
Some are set in advance, by contract or in a catalogue: faced with a change in demand, firms first produce more or less, and their prices move only gradually.
Output and prices move together or in opposite directions. After a demand shock, they go the same way: more output and higher prices, or less output and, in the model, lower prices (in practice, mostly lower inflation).
3. Output and prices move together or in opposite directions
After a demand shock, they go the same way: more output and higher prices, or less output and, in the model, lower prices (in practice, mostly lower inflation). After an adverse supply shock, output falls while prices rise: when the shock is large, this is stagflation.
In the long run, prices and wages finish adjusting. This can take several years: the change in prices gradually brings output back to its potential level (the model assumes purchases stay equal to output).
4. In the long run, prices and wages finish adjusting
This can take several years: the change in prices gradually brings output back to its potential level (the model assumes purchases stay equal to output). In the model, a demand shock leaves only a different price level; after a temporary supply shock, if demand is not supported, prices come back down and output returns to its potential level. A lasting supply shock changes potential itself.

What shifts long-run supply: the LRAS curve moves when what the country can produce sustainably changes. More workers (for example through immigration), more machines and buildings, new natural resources or better techniques shift it to the right; their lasting loss shifts it to the left.

Why demand falls when prices rise: Mankiw's introductory textbook puts forward three effects, detailed in the vocabulary (wealth, interest rate, exchange rate). In more recent versions of the model, which track inflation rather than the price level, it is mainly the central bank that makes the link: when inflation rises, it raises its rates by more than inflation increases; borrowing then costs more, even after deducting the rise in prices, which holds back credit and spending.

Policies can cushion a shock: faced with a demand shock, the central bank (through its rates) and the government (through its spending and taxes) can support or restrain spending, which stabilizes both output and prices. Faced with an adverse supply shock, with these demand tools, they have to choose: support output at the cost of more inflation, or restrain inflation at the cost of more unemployment.

A supply shock can also be favorable: output rises and prices rise more slowly, or even fall. In the United States, hourly labor productivity in the nonfarm business sector accelerated from 1.4% a year (1973-1995) to nearly 3% (1995-2000); output grew by more than 4% a year on average and unemployment fell to 4% without prices taking off. In early 1999, the Federal Reserve attributed the low inflation of recent years to two kinds of causes: temporary factors (in 1998, the fall in oil and import prices) and more lasting ones, including faster productivity growth and stronger competition.

When it does not go as planned

A fall in prices does not always revive demand: when it lasts, households and firms may put off their purchases while waiting for lower prices and, as the economist Irving Fisher (1933) described, debts already taken on weigh more heavily. In Japan, consumer prices fell or stagnated in 12 years out of 14 between 1999 and 2012; the Bank of Japan sees in this, among other things, persistently insufficient demand: its rates, already close to zero, could no longer fall enough to revive it.

Inflation sometimes falls less than expected in a recession: after the 2008 crisis, US unemployment reached 10% (October 2009), but inflation excluding food and energy, about 2% before the crisis, only fell to 0.6% year on year at its lowest, well above what the relationships observed before 2008 suggested: inflation lower still, or even negative. The explanations are still debated: households expecting more inflation because of oil prices (Coibion and Gorodnichenko, 2015), wages that are rarely cut.

Output does not always return to its earlier path: out of 122 recessions between the 1970s and the 2000s in 23 advanced countries, about two out of three were followed by output that remained below the pre-recession path three to seven years later (Blanchard, Cerutti and Summers, 2015). The authors warn that this link does not prove that the recession is the cause: in many cases, the realization that future growth would be weaker may have caused both the recession and the lower output that followed. But even after a recession linked to a deliberate reduction in inflation, output remains below its earlier path in nearly two cases out of three (63%), which suggests that a recession can also leave lasting scars.

Stagflation does not necessarily come from a supply shock: for the 1970s stagflation, Blinder and Rudd (2013) mainly point to supply shocks (oil, food, the end of price controls), Barsky and Kilian (2001) mainly to overly loose monetary policy, which would also have pushed up oil prices; the debate remains open.

Supply and demand intertwine: in June 2022, US consumer prices rose 9.1% year on year, a level unseen since 1981. According to Bernanke and Blanchard (2023), excessive demand set off this surge, mainly in two ways: it contributed to the rise in energy and food prices, which other causes such as the war in Ukraine also fed, and to shortages of goods whose supply could not keep up. Labor market tightness initially counted for little, but its effect lasts longer.

To learn more, read the full explanation, then follow the guided notions at the end of the explanation.

Why it matters

The AD-AS model is the textbooks' basic tool for analyzing the causes of recessions, the mechanisms of inflation, and the effectiveness of stabilization policies; economists also use more recent and more detailed versions, which track inflation rather than the price level. Mankiw introduces and develops it by showing how demand and supply shocks explain the major economic crises: the Great Depression of 1929 (massive negative demand shock), the oil shocks of 1973 and 1979 (negative supply shocks in the textbook reading, a reading discussed above). The model is also used to read the Covid-19 pandemic, as a simultaneous shock on supply and demand.

An equation sums up short-run supply: output moves away from its potential when prices surprise, that is, when they differ from what was expected. Mankiw's introductory textbook gives three explanations, which all lead to this equation: sticky wages (set by contract), sticky prices (set in advance, adjusted with a lag), and misperception (mistaking a rise in the general price level for a rise in one's own price). The equation reads Y = Ȳ + α(P − Pᵉ), where Y is output, Ȳ natural output (another name for potential GDP), P the actual price level, Pᵉ the price level that was expected, and α a positive number that says how much output reacts to a price surprise.

It also explains the return to potential: as long as P differs from Pᵉ, output stays away from Ȳ. But households and firms gradually revise what they expect, and wages follow: Pᵉ and P converge, which shifts the SRAS curve, and output returns to Ȳ. After a recession, for example, expected prices and wages fall: at each price level, firms then produce more, and the SRAS curve shifts to the right.

A worked example, step by step (figures chosen for the example, with α = 2):

1. At the start, the economy produces 100, its potential, and its prices stand at 100, as everyone expected.

2. Households lose confidence: at each price level, they buy the equivalent of 6 less. Their response to prices does not change: each one-point fall in prices still makes them buy 1 more.

3. In the short run, in the model, prices only fall to 98, because many are set in advance. This fall brings back 2 in purchases: firms sell and produce 96, not 94. With expected prices of 100, the equation indeed gives 100 + 2 × (98 − 100) = 96. The output gap is −4.

4. Gradually, expected prices and wages fall, the SRAS curve shifts to the right, prices keep falling and, with lower prices, purchases rise again.

5. In the long run, prices stand at 94, as people now expect, and output is back at 100: in the model, only a lower price level remains.

Mankiw stresses the fundamental distinction between short-run and long-run effects: in the short run, demand policies (monetary and fiscal) influence output and prices, because prices, wages and expected prices do not adjust right away. In the long run, output returns to its potential level and only prices have changed, in line with the verticality of the LRAS curve.

Key points

A positive demand shock (renewed household confidence, higher foreign demand) shifts the AD curve to the right, as a stimulus policy would (rise in public spending, rate cut), which raises both output and prices in the short run. In this model, in the long run, output returns to its potential level and only prices have risen

A negative supply shock (oil price rise, supply chain disruptions) shifts the SRAS curve to the left, simultaneously triggering a fall in output and a rise in prices (stagflation, if the shock is large). This is a difficult scenario because, in this model, a demand-side policy cannot correct both problems at the same time. If the shock is temporary and demand is not supported, output eventually returns to its potential level as prices come back down, which can take a long time; a supply-side policy (lower production costs, productivity gains) can also, in theory, fix both

Automatic long-run adjustment works through the flexibility of prices and wages: in the long run, after a recession, the adjustment of prices and wages can reduce real costs, shift the SRAS curve to the right and allow output to return toward its potential level. This adjustment can, however, take several years

With demand tools alone, the stagflationary dilemma imposes a difficult trade-off: an expansionary policy reduces the recession but worsens inflation, a restrictive policy reduces inflation but worsens the recession. If the shock is temporary, two paths are then open: support demand to offset the shock (this is called "accommodating" the shock), and output quickly returns to its potential level, but prices stay lastingly higher; or do not support it, and prices eventually come back down, but after a longer recession

Concrete example

Examples

The 1973 oil crisis illustrates the negative supply shock: the embargo declared by the Organization of Arab Petroleum Exporting Countries (OAPEC), combined with the price increases decided by OPEC, quadrupled in a few months the official price of a barrel of Saudi oil, which served as the benchmark, from $2.90 a barrel in June 1973 to $11.65 in January 1974, sharply raising production costs across the world economy. In the textbook reading, the aggregate supply curve shifted to the left, causing both a recession and high inflation in most major industrialized countries. Among the seven that would form the G7 in 1976 (the United States, Japan, West Germany, France, the United Kingdom, Italy, Canada), GDP fell in 1974 or 1975 in all but Canada, and inflation exceeded 10% in 1974 in all but West Germany (7%). The Covid-19 pandemic in 2020 was a dual shock: lockdowns reduced supply (factory closures, logistical disruptions) and demand (mobility restrictions, uncertainty) simultaneously, triggering the deepest global recession since World War II. The 2021-2022 episode provides the most recent textbook case: as economies reopened, excess demand (accumulated savings, massive fiscal stimulus) met constrained supply (clogged supply chains, then the 2022 surge in energy prices), pushing inflation to a forty-year high in the United States and to its highest level since the euro's creation in the euro area, and reviving stagflation fears. The sharp monetary tightening that followed drove equity and bond markets down simultaneously in 2022: in periods of high inflation, higher-than-expected inflation makes markets expect rate hikes, which weighs on the price of both bonds and equities (Lombardi and Sushko, Bank for International Settlements, 2023).

Market impact

Markets

Negative supply shocks are particularly feared by financial markets, because they can produce stagflation, a situation in which a central bank cannot cut its rates without fueling inflation, nor raise them without worsening the recession: it has to choose, like the Federal Reserve, which from 1979 sharply raised its rates at the cost of two recessions, in 1980 and then in 1981-82. Equity and bond markets often suffer together when high inflation pushes central banks to raise their rates sharply, as in 2022. Stimulus plans, which push demand up, are often welcomed at first by equity markets, but can push bond yields up in anticipation of future inflation.

Official source