3.2 Aggregate demand and aggregate supply

Aggregate demand

In microeconomics, demand is the quantity of one good that consumers are willing and able to buy at each price. Aggregate demand applies the same idea to the whole economy: it is total planned spending on all the final goods and services produced in a country. Because it covers every good and service, its “price” is the average price level, and its “quantity” is real GDP.

The average price level is a weighted average of the prices of everything produced in the economy, measured by a price index such as the GDP deflator or the consumer price index. A rise in the average price level means that prices in general are higher, even though some individual prices may have fallen.

The four components of aggregate demand

Total spending comes from four groups of buyers, matching the sectors of the circular flow of income: households, firms, the government and the rest of the world. Together they make up the expenditure approach to measuring gross domestic product.

Consumption is the largest component of AD in almost every economy, so changes in household spending have a large effect on total output. Investment is much smaller but far more volatile, because firms cut it sharply when they expect sales to fall.

In economics, investment means adding to the stock of real capital. Buying shares or putting money in a bank account is saving, not investment, because it does not create new capital goods.

Government spending in AD excludes transfer payments such as pensions and unemployment benefits. These payments move income from taxpayers to recipients without buying any output, so they only affect AD when the people who receive them spend the money, which then counts as consumption.

Exports are added because they are spending by foreigners on domestic output. Imports are subtracted because consumption, investment and government spending all include some spending on foreign goods, which is not part of domestic output. Net exports can be negative: a country that imports more than it exports has a negative X − M.

The size of each component differs a great deal between economies. Consumption is about two-thirds of spending in the United States but well under half in China, where investment is unusually high. Germany’s exports and its imports of goods and services are each worth about 40% of its GDP, against 11–14% in the United States, so changes in world demand affect German AD far more.

Components of expenditure as a percentage of GDP, 2023. Investment is gross capital formation; X − M is exports minus imports of goods and services. Rows may not sum to exactly 100 because of rounding. Source: World Bank, World Development Indicators (2026).

Why the AD curve slopes downwards

The AD curve shows the relationship between the average price level and the quantity of real output that households, firms, the government and foreigners plan to buy, with everything else held constant (ceteris paribus). It slopes downwards: a lower price level is associated with a larger quantity of real GDP demanded.

The diagram shows a movement along the curve from A to B. The reasons for the slope differ from those for a single good’s demand curve. When the price of one good falls, buyers switch to it from other goods, but when all prices fall together there is nothing cheaper to switch to. Three effects explain the slope instead.

1. The wealth effect. Many assets, such as savings in bank accounts, have a fixed money value. When the price level falls, the real value (purchasing power) of this wealth rises, so households feel better off and consume more. A rise in the price level reduces real wealth and consumption.

2. The interest rate effect. At a lower price level, households and firms need less money for their transactions, so they have more to save and lend. The greater supply of loanable funds pushes interest rates down, which encourages borrowing for consumption and investment.

3. The international trade effect. If the domestic price level falls while prices abroad do not, domestic goods become cheaper relative to foreign goods. Exports rise and imports fall, so net exports increase.

Movements along AD and shifts of AD

A change in the average price level causes a movement along the AD curve. A change in any other factor that affects C, I, G or X − M causes the whole curve to shift. An increase in AD shifts the curve to the right, so more real output is demanded at every price level; a decrease shifts it to the left.

A sudden, unexpected change in one of the components of AD is called a demand shock. A shift in AD can also set off further rounds of spending, a process known as the Keynesian multiplier.

The determinants of aggregate demand

Each component of AD depends on its own set of factors. Knowing these determinants makes it possible to analyse almost any macroeconomic event, from a stock market crash to a change in a trading partner’s tariffs, as a shift in AD.

Determinants of consumption

The main determinant of consumption is household income. Households spend part of each extra unit of income and save the rest, so when real incomes rise, consumption rises. The factors below explain why consumption changes even when current income has not.

Consumer confidence

When households are optimistic about their jobs and future incomes, they spend more of their current income and borrow more. When they are worried about losing their jobs, they save more as a precaution and delay large purchases such as cars and furniture.

Confidence is measured by surveys. The University of Michigan’s index of US consumer sentiment fell from 101.0 in February 2020 to 71.8 in April 2020 as the COVID-19 pandemic spread, and dropped to 50.0 in June 2022, during a period of rapidly rising prices. Falls like these usually come before falls in spending on durable goods.

Sources: University of Michigan, Surveys of Consumers, via FRED series UMCSENT (2026).

Interest rates

An interest rate is the price of borrowing money and the reward for saving it. A rise in interest rates makes credit cards, car loans and mortgages more expensive, so households borrow and spend less. It also makes saving more attractive. For households with variable-rate mortgages, higher interest rates also raise monthly repayments, which leaves less income for other spending.

Wealth

Wealth is the value of the assets households own, such as houses, shares and savings, minus their debts. It is a stock, measured at a point in time, while income is a flow per period. When house prices or share prices rise, households feel richer and consume more, even though their incomes have not changed; when asset prices fall, they cut spending.

China’s property slump shows this effect working in reverse. The property developer Evergrande defaulted on its US-dollar bonds in December 2021, and house prices then fell for several years, by about 30% between 2021 and 2026 on economists’ estimates. Falling house prices reduced the wealth of the many households that own property, weakening their confidence and holding back consumption.

Sources: Wikipedia, “Chinese property sector crisis (2020–present)”, citing Fitch, S&P Global and the National Bureau of Statistics of China (accessed 2026).

Income taxes

Households can only spend the income left after income tax has been paid and transfer payments received. A cut in income tax raises disposable income, so consumption rises; an increase in income tax reduces it. Income taxes are one of the tools of fiscal policy.

Household indebtedness

Households that owe a great deal relative to their incomes have to spend more of their income on interest and repayments, and lenders are less willing to give them new loans. High household indebtedness therefore limits consumption, and it makes consumption more sensitive to interest rates, because each rise in rates raises the cost of a large existing debt.

Household debt varies widely between countries. In early 2026 household debt was about 109% of GDP in Australia and about 96% in Canada, compared with about 49% in Germany in late 2025. Interest rate changes therefore tend to have a larger effect on consumer spending in the first two countries.

Sources: Bank for International Settlements, household debt as a percentage of GDP, via FRED series HDTGPDAUQ163N, HDTGPDCAQ163N and HDTGPDDEQ163N (2026).

Expectations of the future price level

If households expect prices to rise soon, they bring purchases forward to beat the price increases, so consumption rises now. If they expect prices to fall, as in a period of deflation, they delay purchases, especially of expensive durable goods, and consumption falls now. Expectations of future income work in a similar way, which links this determinant to consumer confidence.

Determinants of investment

Firms invest when they expect the extra revenue from new capital to be greater than its cost. The determinants of investment are the factors that change either the expected return or the cost of the project.

• Interest rates. Most investment is financed by borrowing, and even firms that use their own profits give up the interest they could earn on them (an opportunity cost). Higher interest rates raise the cost of investment projects, so fewer are profitable and investment falls.

• Business confidence. Investment pays off over many years, so it depends on what firms expect future sales and profits to be. Optimism about future demand raises investment; pessimism or uncertainty, for example about future trade policy, makes firms postpone projects.

• Technology. New technologies, such as artificial intelligence, robotics or renewable energy, create profitable opportunities and push firms to invest so that they are not left behind by competitors.

• Business taxes. A lower corporate (profits) tax raises the after-tax return on investment, and tax allowances for spending on capital goods lower its cost. Both encourage investment.

• Corporate indebtedness. Firms that already carry large debts find it harder and more expensive to borrow more, and they may use their profits to pay down debt instead of investing.

Determinants of government spending

Government spending does not depend on the price level or on interest rates in the way that private spending does. It depends on political and economic priorities: the services a government chooses to provide, its plans for infrastructure, defence and welfare, and its aims for the economy as a whole.

Political priorities change with elections, wars and demographic change: an ageing population raises spending on health care, for example. Economic priorities matter too. A government may raise spending to fight a recession, as many did in 2020, or cut it to reduce a budget deficit or government debt. The deliberate use of government spending and taxes to influence AD is fiscal policy.

Determinants of net exports

• Incomes of trading partners. Exports are the consumption and investment of people in other countries. When trading partners grow quickly, they buy more of a country’s exports; when they fall into recession, exports fall.

• Exchange rates. A depreciation of the currency makes exports cheaper for foreigners and imports more expensive at home, so X − M tends to rise. An appreciation has the opposite effect.

• Trade policies. Trade protection by trading partners, such as a tariff on a country’s exports, reduces its exports. Trade agreements that lower barriers can raise both exports and imports.

Net exports also depend on domestic income. When the domestic economy grows, households and firms buy more imports, which reduces X − M. This is one reason a country’s net exports often fall during a boom.

Summary of the determinants of AD

Short-run aggregate supply

Aggregate supply is the other half of the model. It shows how much output the firms in an economy plan to produce at each average price level. Its shape depends on the time period, because the prices that firms pay for their inputs, especially money wages, take time to adjust.

The short run in macroeconomics is the period in which the prices of the factors of production stay fixed. Money wages are agreed in contracts that last a year or more, rents are set in leases, and many raw material prices are agreed in advance. During this period, a change in the price level changes the prices firms receive for their output but not the prices they pay for their inputs.

Why the SRAS curve slopes upwards

Suppose the average price level rises while money wages stay the same. Each firm sells its output at a higher price but pays the same wages, so its profit per unit rises, and producing more becomes worthwhile. Firms raise output by asking staff to work overtime, hiring more workers and using machines more intensively. Real GDP supplied therefore rises.

The SRAS curve also becomes steeper at high levels of output. As firms approach full use of their capacity, extra output requires overtime pay, less-skilled workers and overused machines, so the cost of each extra unit rises. For simplicity, SRAS is often drawn as a straight upward-sloping line, which is acceptable in most diagrams.

What shifts the SRAS curve

Any change in the costs of production that is not caused by a change in the price level shifts the SRAS curve. A rise in costs shifts SRAS to the left (upwards): firms supply less at each price level, or need a higher price level to supply the same output. A fall in costs shifts it to the right.

• Costs of factors of production. Higher money wages, rents or raw material prices raise firms’ costs. Energy and oil are especially important because almost every firm uses them, directly or through transport. Imported inputs also become dearer when the currency depreciates.

• Indirect taxes. A higher indirect tax, such as a rise in value-added tax or in fuel duty, adds to firms’ costs across the economy and shifts SRAS left.

• Subsidies. A subsidy to producers lowers their costs and shifts SRAS right. Energy subsidies for firms are an example.

• Supply shocks. Sudden events such as wars, pandemics, droughts and earthquakes can disrupt supply chains or destroy capacity, raising costs for many firms at once.

The war in the Middle East in 2026 was an adverse supply shock for oil-importing economies. The Brent crude oil price rose from about $71 a barrel at the end of February 2026 to $118 on 18 March 2026, and stood at $135.51 on 2 October 2026, raising transport and production costs for firms around the world.

Sources: US Energy Information Administration, Brent spot price, via FRED series DCOILBRENTEU (2026); World Bank, Commodity Markets Outlook, April 2026.

Short-run macroeconomic equilibrium

Short-run equilibrium occurs where the AD curve crosses the SRAS curve. At that price level, the real output that buyers plan to purchase equals the real output that firms plan to produce, so there is no pressure for the price level or output to change.

The diagram shows why the economy moves towards E. If the price level were P₂, firms would produce more than households, firms, the government and foreigners plan to buy. Unsold goods would pile up in warehouses, so firms would cut prices and output until the economy returned to P₁ and Y₁. Below P₁, the opposite happens: planned spending exceeds output, stocks run down, and firms raise prices and output.

How shifts in AD and SRAS change the equilibrium

Every macroeconomic event in the short run can be analysed by asking which curve shifts and in which direction. The new equilibrium then shows the effect on the price level, on real GDP and, through real GDP, on employment.

An increase in AD (left panel) moves the economy up along SRAS: real GDP rises from Y₁ to Y₂, firms hire more workers so unemployment falls, and the price level rises from P₁ to P₂. A rise in the price level caused by higher AD is demand-pull inflation. A decrease in AD has the opposite effects: lower output, higher cyclical unemployment and a lower price level, or at least lower inflation.

A decrease in SRAS (right panel) is more damaging, because the price level rises from P₁ to P₂ while real GDP falls from Y₁ to Y₂. The combination of rising prices and falling output is called stagflation, and the rise in the price level is cost-push inflation. An increase in SRAS, for example from cheaper energy, is the best case: output rises and the price level falls.

When AD and SRAS shift together

Some events shift both curves at once. The final effect on real GDP and the price level then depends on which shift is larger. The COVID-19 pandemic in 2020 is an example: lockdowns and fear of infection cut consumption and investment sharply (AD left), while closed workplaces and broken supply chains raised costs (SRAS left).

Because the fall in AD was much larger than the fall in SRAS in 2020, output fell steeply while inflation stayed low in most economies. When demand recovered in 2021 and supply problems continued, the balance changed and inflation rose. When both curves shift left, real GDP falls in every case, and the direction of the change in the price level depends on which shift is larger.

Two views of aggregate supply in the long run

Economists agree about the short run but disagree about what happens over a longer period, once wages and other input prices can adjust. The disagreement goes back to the 1930s. Classical economics held that markets clear on their own; John Maynard Keynes argued in 1936 that an economy could remain in a deep slump for years. Later, monetarist and new classical economists revived and extended the classical view.

The two views lead to two different aggregate supply curves. The monetarist/new classical model adds a vertical long-run aggregate supply (LRAS) curve to the AD and SRAS curves. The Keynesian model replaces both with a single Keynesian AS curve with three sections.

The monetarist/new classical LRAS curve

Monetarism and new classical economics assume that wages and prices are flexible over time. In the short run, contracts fix money wages, but when contracts are renegotiated, wages adjust to the price level. Once all input prices have adjusted, a higher price level no longer makes extra output more profitable, because costs have risen by the same proportion as prices.

The quantity of output an economy can produce in the long run therefore does not depend on the price level at all. It depends on the quantity and quality of the factors of production and on the technology and institutions that combine them. The LRAS curve is vertical at this level of output, which is the economy’s potential output or full employment level of output, labelled Yf.

Full employment does not mean that everyone has a job. Some unemployment always exists: people moving between jobs (frictional unemployment), people whose skills or location do not match the available jobs (structural unemployment) and people in seasonal work (seasonal unemployment). Together these make up the natural rate of unemployment. At full employment there is no cyclical unemployment.

The diagram shows long-run equilibrium, where AD, SRAS and LRAS all meet at E. Output is at Yf and the price level is Pe. In this model, AD determines only the price level in the long run: a permanent rise in AD raises prices but cannot keep output above Yf.

The Keynesian AS curve

Keynes argued that money wages are “sticky downwards”: workers and trade unions resist cuts in money wages, and minimum wage laws and long contracts prevent them. Firms also avoid cutting wages because it harms morale. With wages unable to fall, an economy with low AD can stay at a level of output well below full employment, with many unemployed workers and idle machines.

Each section of the curve reflects how much spare capacity the economy has:

• Section I (horizontal). In a deep recession there are many unemployed workers and idle machines. Firms can raise output by hiring workers at the existing wage and using unused capacity, so output rises without any increase in costs or prices.

• Section II (upward sloping). As output approaches full capacity, bottlenecks appear. Some industries run short of skilled labour, raw materials or machines before others, so firms have to pay higher wages and input prices to raise output, and the price level starts to rise.

• Section III (vertical). At Yf all resources are fully employed. Output cannot rise further, so any extra spending only bids up prices.

The effect of a rise in AD therefore depends on where the economy is on the Keynesian AS curve. In section I it raises output and employment, in section II it raises both output and the price level, and in section III it raises only the price level.

This is the central policy message of the Keynesian model: when the economy has a lot of spare capacity, raising AD increases real GDP and employment at little cost in inflation. In the monetarist/new classical model, by contrast, a rise in AD raises output only temporarily.

Output gaps

An output gap is the difference between actual real GDP and potential output. Output gaps link the AD/AS model to the business cycle: during a recession the economy produces less than it could, and during a boom it produces more than it can sustain.

Output gaps can be negative or positive, and each has its own name and its own consequences for unemployment and prices.

A deflationary gap is also called a recessionary gap or a negative output gap. It shows up as rising unemployment, falling or slow-growing prices and firms working below capacity. The term “deflationary” refers to the downward pressure on prices: the price level may fall, but more often inflation simply slows.

An inflationary gap (a positive output gap) is possible in the short run because firms can produce more than their normal capacity for a while, through overtime, extra shifts and postponed maintenance. It cannot last, because labour shortages push up wages and other costs.

Output gaps in the monetarist/new classical model

In the left panel, AD and SRAS cross at Y₁, below Yf: there is a deflationary gap equal to the distance Y₁–Yf, and cyclical unemployment. In the right panel, they cross at Y₁, beyond Yf: there is an inflationary gap, unemployment is below its natural rate and the labour market is tight. In both cases the gap is measured along the real GDP axis, between Y₁ and Yf.

Output gaps in the Keynesian model

In the Keynesian model, a deflationary gap is shown by AD crossing the AS curve in its horizontal or rising section, to the left of Yf. Because the economy cannot produce beyond Yf, an inflationary gap is shown differently: AD rises so far that it crosses the vertical section, and the extra spending raises only the price level.

Japan after its asset-price bubble burst in the early 1990s is often used as an example of a long-lasting deflationary gap. Real GDP grew slowly for years, contracting in 1993, 1998 and 1999, and consumer prices fell in most years between 1999 and 2012 as weak demand held prices down.

Sources: IMF, World Economic Outlook database (2026).

Long-run equilibrium: the monetarist/new classical view

In the monetarist/new classical model, the economy returns to full employment on its own after any shock to AD. The process works through the labour market: when there is unemployment, money wages fall; when labour is scarce, they rise. Changes in wages shift SRAS until output is back at Yf.

Adjustment from a deflationary gap

The economy starts in long-run equilibrium at A. A fall in AD, from AD₁ to AD₂, moves it along SRAS₁ to B: output falls to Y₂, below Yf, and a deflationary gap opens. Unemployment is now above the natural rate. Unemployed workers compete for jobs and accept lower money wages, and when contracts are renegotiated, wages and other input prices fall.

Lower costs shift SRAS to the right, from SRAS₁ to SRAS₂. The economy moves down along AD₂ to C, where output is back at Yf and unemployment is back at its natural rate. The only lasting effect of the fall in AD is a lower price level, P₃. In this view, the recession corrects itself without any government action.

Adjustment from an inflationary gap

When AD rises from AD₁ to AD₂, the economy moves from A to B: output rises above Yf to Y₂ and unemployment falls below the natural rate. Firms compete for scarce workers, and workers, seeing prices rise, demand higher money wages to protect their real wages. As wages and other input prices rise, SRAS shifts left to SRAS₂.

The economy moves up along AD₂ to C. Output returns to Yf and unemployment to the natural rate, but the price level is permanently higher at P₃. The model’s conclusion is that policies which raise AD cannot keep output above Yf: in the long run, they only cause inflation.

Equilibrium in the Keynesian model: gaps that persist

Keynes accepted that wages might eventually fall in a recession, but argued that the process could be very slow and might not restore full employment even when it happened. His best-known remark on the point was aimed at economists who relied on the long run to fix short-run problems.

But this long run is a misleading guide to current affairs. In the long run we are all dead.

— John Maynard Keynes, A Tract on Monetary Reform (1923), chapter 3

The Keynesian model gives several reasons why a deflationary gap can last:

• Sticky wages and prices. Money wages are fixed by contracts, minimum wage laws and union agreements, and workers resist cuts. Many firms also keep their prices unchanged for long periods. If wages do not fall, SRAS does not shift right and the economy stays at Y₁.

• Falling wages reduce demand. Wages are workers’ income as well as firms’ costs. Even if wages did fall, workers would have less to spend, so AD could fall further and cancel out the effect on SRAS.

• Falling prices make things worse. Expected deflation encourages households to delay purchases, and it raises the real value of debts, forcing indebted households and firms to cut spending. Both reduce AD further.

• Low confidence. In a deep recession, firms do not invest however low interest rates go, because they do not expect to sell more output. Low confidence keeps AD low.

In the Keynesian model, equilibrium can therefore settle well below Yf, at a level of output where AD meets the horizontal section of the AS curve. Unemployment stays high because total spending is too low to employ everyone who wants to work. The solution Keynes proposed was for the government to raise AD directly, through higher government spending or lower taxes.

The Great Depression of the 1930s is the classic example. In the United States, unemployment reached an estimated 24.9% in 1933, and output and employment took the rest of the decade to recover.

Sources: Wikipedia, “Great Depression in the United States”, citing Stanley Lebergott’s unemployment estimates (accessed 2026).

Shifts of LRAS and of the Keynesian AS curve

Over the long term, an economy’s capacity to produce grows. In both models this is shown by the full employment level of output, Yf, moving to the right: the LRAS curve shifts right in the monetarist/new classical model, and the vertical section of the Keynesian AS curve shifts right in the Keynesian model. An outward shift of the production possibilities curve shows the same thing.

The diagram shows the same increase in capacity in both models. Four groups of factors cause these shifts:

• Changes in the quantity and/or quality of factors of production. More labour (from population growth, immigration or higher participation), more capital from past investment, or newly discovered natural resources raise the quantity of factors. Better education, training and health raise the quality of labour, which is the economy’s human capital.

• Improvements in technology. New production methods and equipment let the same factors produce more output, as computers, the internet and automation have done.

• Increases in efficiency. Greater competition, better management and the movement of resources from low-productivity to high-productivity uses raise output per unit of input, which is productivity.

• Changes in institutions. Secure property rights, a reliable legal system, low corruption, stable government and well-functioning banks encourage investment and enterprise. Better institutions raise potential output; weak ones hold it back.

Shifts in LRAS can also be to the left. A war that destroys factories and infrastructure, mass emigration of skilled workers, or a lasting rise in the cost of a key input such as energy can all reduce the economy’s productive capacity.

Growth without inflation

A rightward shift of LRAS is the only way to raise real GDP permanently in the monetarist/new classical model. When LRAS and AD grow together, real GDP rises while the price level stays stable, which is the ideal combination of sustained economic growth and low inflation.

This diagram shows why governments care about the supply side of the economy. If AD grows faster than LRAS, the extra spending creates inflation; if LRAS grows faster than AD, a deflationary gap opens. Policies designed to shift LRAS to the right are supply-side policies.

Evaluating the Keynesian and monetarist/new classical models

The two models are simplifications built on different assumptions, and each leads to different conclusions about policy. Evaluating them means examining those assumptions, comparing their predictions with the evidence, and judging when each is more useful.

Assumptions of the two models

Implications for policy

The monetarist/new classical model implies that governments should not try to manage AD. Because the economy returns to Yf on its own, policies to raise AD in a recession are unnecessary and, with time lags, may add to AD just as the economy recovers, causing inflation. The best policies, in this view, are a stable environment with low inflation and market-based supply-side policies that make labour and product markets more flexible and raise LRAS.

The Keynesian model implies that governments should act when there is a deflationary gap. Since the gap may not close on its own, or only after years of high unemployment, expansionary fiscal policy and expansionary monetary policy can raise AD and restore employment. Near full employment, the same model warns that raising AD will only cause inflation, so policy should then aim to restrain AD.

What the evidence suggests

Each model fits some episodes better than others. The deep and long recessions of the 1930s, and the slow recovery after the 2008 financial crisis, are hard to explain if wages and prices adjust quickly, which supports the Keynesian view. The stagflation of the 1970s, when higher AD brought rising inflation without lower unemployment, supported the monetarist/new classical argument that AD policies cannot keep output above its natural level.

Most economists today combine elements of both. They accept that in the short run, sticky wages and prices give AD a large effect on output, so demand management can reduce the depth of recessions. They also accept that over long periods output is determined by supply-side factors, and that excessive demand leads to inflation. The disagreement is mainly about how long the short run lasts and how much governments can safely do.

Judgment: which model is more useful?

The usefulness of each model depends on the situation. In a deep recession with high unemployment and low inflation, the Keynesian model is more helpful, because the evidence suggests self-correction is slow and costly. In an economy close to full employment with rising inflation, the monetarist/new classical model is a better guide, because further increases in AD are likely to raise prices rather than output.

Other factors also affect the judgment: how flexible the labour market is (strong trade unions or a high minimum wage make wages stickier), whether people expect inflation to stay low, and how large government debt already is. A reasoned judgment states which model fits the particular economy and time period, and why.

Case study: Germany’s energy shock and stagnation, 2022–2025

Germany, the largest economy in Europe, entered 2022 heavily dependent on natural gas imported from Russia by pipeline. After Russia invaded Ukraine in February 2022, Gazprom cut flows through the Nord Stream 1 pipeline to 40% of capacity in June 2022, and from early September no gas flowed through it at all; explosions damaged both of its pipes later that month. European gas prices rose to levels never seen before.

A supply shock: SRAS shifts left

Gas is used to generate electricity, to heat buildings and as a raw material in the chemical industry. The rise in its price, shown in the chart, was a large increase in the costs of factors of production for almost every German firm, shifting SRAS to the left. Energy-intensive industries such as chemicals and metals cut output sharply in 2022 and again in 2023, and the effect spread through the economy as firms passed on higher costs.

As the AD/AS model predicts for a fall in SRAS, the price level rose while output stagnated. Consumer price inflation reached 8.7% in 2022 and 6.0% in 2023, compared with 1.4% in 2019, while real GDP fell by 1.0% in 2023 (revised down from a first estimate of −0.3%) and did not grow in 2024.

Demand also fell: AD shifts left

AD weakened at the same time. Household consumption fell by 0.8% in real terms in 2023, with spending on durable goods down 6.2%, as high prices reduced households’ real incomes. To fight inflation, the European Central Bank raised its deposit rate from −0.50% in July 2022 to 4.00% in September 2023, which raised borrowing costs and hit construction and investment. Exports of goods and services fell by 1.8% in 2023 and by 0.8% in 2024, as German carmakers and machinery producers faced stronger competition in key export markets.

The diagram combines the two shifts. Because SRAS shifted further than AD, the price level rose while real GDP fell. Unemployment rose only slowly, from 3.1% in 2023 to 3.4% in 2024, which suggests that firms facing shortages of skilled labour preferred to keep their workers.

Two readings of a long stagnation

By 2024 energy prices had fallen from their peak and inflation was back to 2.5%, yet output was still not growing. A Keynesian reading stresses weak AD: high interest rates, low consumer and business confidence, and falling exports left a deflationary gap, which supports a case for raising government investment, especially as Germany’s fiscal rule (the “debt brake”) had limited government borrowing.

A monetarist/new classical reading stresses the supply side. If gas stays permanently more expensive than before 2021, some energy-intensive production is no longer profitable in Germany at all, so the country’s potential output has fallen: LRAS has shifted left. On this reading, Germany is not far below its full employment output, and higher AD would mainly raise prices; the answer lies in cheaper energy, faster approval of investment and other supply-side reforms. Low unemployment throughout the period is consistent with this view.

The case shows how one diagram can organise a complex episode: a supply shock, a policy response that cut demand, and a weak world market for exports. It also shows the limits of the model. Separating short-run shifts in SRAS from long-run shifts in LRAS is difficult while the events are happening, and estimates of potential output are themselves uncertain, which is why economists drawing the same diagram reached different policy conclusions.

Sources: IMF, World Economic Outlook database (2026); Federal Statistical Office of Germany (Destatis), GDP press releases of January 2024 and January 2025; European Central Bank deposit facility rate via FRED series ECBDFR (2026); Wikipedia, “2022–2023 Russia–European Union gas dispute” (accessed 2026).

Exam practice

AD/AS analysis appears in almost every macroeconomics question, in Paper 1 and Paper 2, because it is the diagram used to explain growth, unemployment, inflation and the effects of policy. Questions on AD and AS themselves usually ask you to explain a shift in AD or SRAS, to compare the two views of AS, or to evaluate whether an economy will return to full employment on its own.

When a question asks about the “Keynesian” or “monetarist/new classical” view, draw that model’s AS curve: the three-section Keynesian AS curve, or SRAS plus a vertical LRAS. A question that only says “using an AD/AS diagram” can be answered with either, but the monetarist/new classical version with SRAS and LRAS makes it easier to show output gaps and adjustment.

Paper 1-style question

Part (b) of the same question widens the theory into an evaluation and needs real-world examples.

Paper 2-style questions

Diagram questions in Paper 2 usually give a short extract and ask you to apply the AD/AS model to one event in it.

The final part of a Paper 2 question asks for an extended evaluation using the text and your own knowledge.

Summary

• Aggregate demand is total planned spending, AD = C + I + G + (X − M). The AD curve slopes down because of the wealth, interest rate and international trade effects.

• AD shifts when a determinant of a component changes: confidence, interest rates, wealth, taxes, debt and price expectations (C); interest rates, confidence, technology, taxes and debt (I); government priorities (G); trading partners’ incomes, exchange rates and trade policy (X − M).

• SRAS slopes up because input prices are fixed in the short run; it shifts with the costs of factors of production, indirect taxes, subsidies and supply shocks.

• Short-run equilibrium is where AD = SRAS. AD shifts move the price level and real GDP in the same direction; SRAS shifts move them in opposite directions.

• The monetarist/new classical LRAS is vertical at Yf; flexible wages shift SRAS until the economy returns to full employment, with unemployment at the natural rate.

• The Keynesian AS curve is horizontal, then rising, then vertical at Yf; sticky wages mean a deflationary gap can persist, which justifies raising AD.

• LRAS and the Keynesian AS curve shift right when the quantity or quality of factors, technology, efficiency or institutions improve.

• Which model is more useful depends on the size of the output gap, the flexibility of wages and the time frame considered.

Self-check

• To do:

  I can explain why the AD curve slopes downwards using the wealth, interest rate and international trade effects.

• To do:

  I can explain how each determinant of C, I, G and X − M shifts the AD curve.

• To do:

  I can explain why SRAS slopes upwards and what shifts it.

• To do:

  I can draw the monetarist/new classical LRAS curve and the Keynesian AS curve and explain their shapes.

• To do:

  I can draw and explain inflationary and deflationary gaps in both models.

• To do:

  I can explain the automatic adjustment to full employment in the monetarist/new classical model with a fully labelled diagram.

• To do:

  I can explain why a deflationary gap can persist in the Keynesian model.

• To do:

  I can explain the four causes of shifts in LRAS and the Keynesian AS curve.

• To do:

  I can evaluate the two models’ assumptions and policy implications with real-world examples and reach a judgment.