1The idea in one paragraph

Take every buyer in the economy and add up what they plan to spend on domestic output: that is aggregate demand. Take every firm and add up what they plan to produce: that is aggregate supply. Where the two meet, the economy settles at a level of real output and an average price level. Everything that happens in macroeconomics — a boom, a recession, inflation, mass unemployment — is one of those two curves moving. The one genuine argument in the subtopic is about the shape of the supply curve, and that argument decides whether a government should act in a downturn or wait.

2Aggregate demand and its curve

Aggregate demand (AD) is the total quantity of real output that all buyers plan to buy at each average price level, over a given period of time. Three words in that definition do work. Aggregate means everything added together, so there is one curve for the whole economy. Plan means intended purchases, not what actually happened. And average price level means the price of everything at once, not the price of one good.

The axes change, and the new ones must be labelled correctly or the diagram earns nothing. The horizontal axis is real output, real GDP, the quantity of goods and services. The vertical axis is the average price level, an index of all prices in the economy rather than a price in marcs.

AD slopes downward, as in Figure 1, but not for the reason a single demand curve does. When the price level falls, there is no cheaper alternative to switch to, because everything fell together. Three other things happen instead.

• The purchasing power of savings rises. Households holding money and savings find that their stock of wealth buys more, so they feel richer and consumption rises.
• Interest rates tend to fall. With a lower price level, households and firms need less money to make their usual transactions, and a lower demand for money pushes interest rates down. Cheaper borrowing raises consumption and investment.
• Net exports rise. Domestic goods are now cheaper relative to foreign goods, so foreigners buy more of our exports and we buy fewer imports.

Each of the three raises a component of AD, which is why a lower price level goes with a higher planned real output.

3The four components of aggregate demand

AD = C + I + G + (X − M)

• Consumption (C) is household spending on goods and services.
• Investment (I) is spending by firms on capital goods: machines, buildings, vehicles, and additions to stock. In economics investment never means buying shares.
• Government spending (G) is government purchases of goods and services — teachers' salaries, roads, defence. Transfer payments such as pensions and benefits are not G, because nothing is produced in exchange. They change C when the recipient spends them.
• Net exports (X − M) is total exports minus total imports. Imports are subtracted because spending on them is demand for another country's output, not ours.

That is the same expression that measured GDP by the expenditure approach in 3.1, and for the same reason: planned spending on domestic output and the value of domestic output are two views of one thing.

Size matters when you choose which determinant to write about. Marovia's 2025 figures, in billions of marcs, were C 640, I 190, G 245 and X − M of −25, giving 1,050 in total.

Consumption is the biggest component in almost every economy, so anything that moves consumer confidence moves AD more than anything else does. Investment is the smallest of the three positive components and by far the most volatile, which is why it drives the business cycle out of proportion to its size. Net exports can be negative without anything being wrong; it means Marovia bought more from abroad than it sold.

4What shifts aggregate demand

A shift of AD means a different level of planned spending at every price level. The guide lists the determinants component by component, and an exam answer is expected to name the component, name the determinant, and say which way the curve goes.

Figure 2 shows the two directions. An increase in AD shifts the curve to the right: more real output is demanded at each price level. A decrease in AD shifts it left.

Determinants of consumption.

Determinants of investment.

Determinants of government spending. The guide names one: political and economic priorities. A government elected on a promise of new hospitals raises G. A government that decides its deficit is too large cuts G. A government trying to pull the economy out of a recession raises G deliberately, which is fiscal policy and belongs to 3.6.

Determinants of net exports.

One caution about this subtopic's boundaries. A change in AD of 30 billion marcs shifts the curve by 30 billion marcs here. The idea that the final change is larger than the first one is the multiplier, and it is HL material in 3.6 Fiscal policy. Do not bring it into a 3.2 answer.

5Movement along AD versus a shift of AD

This is 2.1 Demand again, at the level of the whole economy, and it is examined just as often here as it was there.

In the left panel of Figure 3 the average price level falls from PL₁ to PL₂ and we slide down the same curve from A to B. The curve has not moved. In the right panel, consumer confidence rises, the price level is unchanged at PL₁, and the whole curve moves right from AD₁ to AD₂, so more output is demanded at every price level including the one we started at.

If the average price level changed, move along AD. If anything else changed, shift AD.

The wording follows the picture, exactly as it did for one good. A slide along the curve is a change in the quantity of real output demanded. A move of the curve is a change in aggregate demand.

Test yourself on two sentences that look alike. "Prices across the economy fell by 2% and output rose": movement along AD. "Firms expect a strong year and order new machines": shift of AD, because investment is a determinant and the price level did not change. When you are unsure, ask which axis the change belongs to. If it is the vertical axis, you move along; if it is not on either axis, the curve moves.

6Short-run aggregate supply

Short-run aggregate supply (SRAS) is the total real output that all firms plan to produce at each average price level, in a period when the prices of the factors of production are fixed.

That last clause is the definition of "short run" in macroeconomics. It is not a length of time. It is the period during which wage contracts, rents and supply agreements have not yet been renegotiated, so factor prices stay where they are while output prices move.

SRAS slopes upward because of exactly that. If the price level rises while wages and other input costs are unchanged, the gap between what a firm receives and what it pays widens. Higher profit per unit makes it worth producing more, so firms take on overtime, run extra shifts and hire from the pool of unemployed workers. Real output rises with the price level.

What shifts SRAS. The guide names two determinants, and only two.

• The costs of the factors of production. Wage rises above productivity, dearer oil, dearer imported components, higher rents. Costs up, SRAS shifts left; costs down, SRAS shifts right.
• Indirect taxes. A tax on production or sales is a cost to the firm at every level of output, so a rise shifts SRAS left and a cut shifts it right.

Figure 5 shows one of each. Wages rise, so SRAS₁ moves left to SRAS₂: at every price level firms now produce less. Indirect taxes are cut, so SRAS moves right. Note which way "left" points on this diagram: a leftward shift of SRAS means less output at the same price level, or the same output only at a higher price level. That is cost-push inflation, which 3.3 draws on exactly this pair of curves.

7Two views of aggregate supply

Here is the part of the subtopic that students get wrong most often, and it is not because the diagrams are hard.

The monetarist/new classical long-run aggregate supply curve and the Keynesian aggregate supply curve are two competing models of the same economy. They are not two facts, and they are not two curves belonging on one diagram. Each is a complete picture of how an economy behaves, drawn by economists who disagree about a single question: what happens to wages and other factor prices when demand falls?

The monetarist/new classical view. In the long run, all factor prices are fully flexible. If the price level doubles, wages eventually double too, so firms have no reason to produce any more than before. Output in the long run depends only on the quantity and quality of the factors of production and the state of technology, not on prices. So long-run aggregate supply (LRAS) is a vertical line at potential output, written Yf: the full employment level of output. In this model there are two supply curves — an upward-sloping SRAS for the period while factor prices lag, and a vertical LRAS for when they have caught up.

The Keynesian view. There is one AS curve, and it has three sections.

• A perfectly elastic (horizontal) section at low levels of output. With mass unemployment and idle factories, firms can produce more without bidding up wages or input prices, so extra output costs no more per unit and the price level does not move.
• An upward-sloping section as output rises. Spare capacity runs out unevenly: skilled workers become scarce, older machines are brought back into use, bottlenecks appear in some industries before others. Costs rise, so more output comes with a higher price level.
• A vertical section at Yf. At full employment no more can be produced whatever buyers offer.

The Keynesian curve makes no distinction between a short-run and a long-run supply curve, because its central claim is that money wages are sticky downwards: they rise readily and fall very slowly, if at all. Workers resist wage cuts, contracts fix wages for a year or more, and firms that cut pay lose their best staff.

Say which model you are using, and draw only that one. A diagram with SRAS, LRAS and a Keynesian curve on it says the candidate has memorised three shapes and understood none.

8What raises aggregate supply in the long run

Both models agree about this part. The economy's capacity to produce grows for the same four reasons whichever curve you drew, and the guide names all four.

• Changes in the quantity and/or quality of factors of production. More workers through population growth or migration; more capital through investment; the discovery of new land or resources. Quality counts as much as quantity: a better-educated and healthier workforce produces more from the same hours.
• Improvements in technology. New methods produce more output from the same inputs.
• Increases in efficiency. The same factors organised better — deregulation that removes a bottleneck, better management, competition that forces firms to cut waste.
• Changes in institutions. The rules an economy runs on: property rights that make investment safe, courts that enforce contracts, a banking system that lends, a stable government, less corruption. Weak institutions hold output below what a country's factors could produce.

Figure 7 draws the same improvement in both models. In the monetarist/new classical model, LRAS₁ moves right to LRAS₂ and potential output rises from Yf₁ to Yf₂. In the Keynesian model the whole AS curve moves right, so the vertical section now stands at a higher Yf₂. This is the diagram 3.3 calls potential economic growth, and it is the only kind of growth that raises output without raising the price level.

A change in these four is a change in the long run. A change in wages or indirect taxes is a change in the short run. Moving the wrong curve is the most expensive single mistake in a macro diagram.

9Macroeconomic equilibrium

Short-run equilibrium is where AD crosses SRAS.

At PL₁ in Figure 8, planned spending equals planned output and nothing pushes the price level either way. Above PL₁ firms cannot sell everything they produce, stocks build up and they cut prices and output. Below PL₁ buyers want more than is being made, shortages appear and prices rise. The economy is pushed back to PL₁ and Y₁ from either side.

Notice what Figure 8 does not say: nothing in it promises that Y₁ is the full employment level of output. Whether the economy gets there on its own is the whole disagreement.

Long-run equilibrium in the monetarist/new classical model. Long-run equilibrium is where AD, SRAS and LRAS all cross, at Yf. The model's claim is that the economy returns there by itself.

Follow Figure 9. The economy starts at Yf. Aggregate demand falls from AD₁ to AD₂ — say a collapse in business confidence — and short-run equilibrium moves to Y₂, below Yf, at a lower price level PL₂. The distance between Y₂ and Yf is a deflationary gap, and workers are unemployed in it. Now the adjustment. With unemployed workers competing for jobs and idle machines earning nothing, money wages and other factor prices fall. Falling factor prices shift SRAS to the right, which raises output and lowers the price level further, and the process continues until output is back at Yf. The economy ends where it began, at potential output, with a lower price level PL₃. No government did anything.

Two consequences follow, and both are examined.

Output in the long run is set by the supply side alone. Demand decides the price level, not the level of output. That is the argument for supply-side policies in 3.7 and against demand management in 3.5 and 3.6.

Unemployment at the long-run equilibrium equals the natural rate of unemployment. Full employment does not mean nobody is unemployed. At Yf there are still people between jobs, people retraining and people waiting for a season to start, and 3.3 Macroeconomic objectives defines that sum as the natural rate of unemployment. What has gone at Yf is cyclical unemployment, the part caused by a shortage of demand.

Equilibrium in the Keynesian model. Keynes's objection is to the step where wages fall.

In Figure 10, AD crosses the AS curve at Ye, well below the full employment output Yf. Because money wages do not fall, AS does not shift. Nothing pulls output back up. The economy can sit at Ye for years, with a persistent deflationary gap and high unemployment, in equilibrium and nowhere near full employment. If the gap is to close, AD has to rise, and in a depressed economy the private components will not rise on their own: households with no job security do not spend, and firms facing empty order books do not invest. That leaves G, which is the case for government intervention and the reason intervention is a key concept for this subtopic.

Naming the gaps. Both models use the same two words for the distance between where the economy is and where full employment is.

• A deflationary gap, also called a recessionary gap, exists when equilibrium real output is below the full employment level. Unemployment is above the natural rate and there is downward pressure on prices.
• An inflationary gap exists when equilibrium real output is above the full employment level. The economy is producing beyond what it can sustain, factor markets are stretched, and the price level is pulled upward.

Both are measured horizontally, from the equilibrium output to Yf. A gap is always defined against full employment output, so a diagram with no Yf on it cannot show a gap at all.

10The same rise in AD, in each model

This is the question that separates a good answer from a memorised one: if aggregate demand rises, does the economy get more output, or only higher prices? Each model gives a different answer, and in the Keynesian model the answer depends on where the economy already is.

Figure 12 shifts AD right by the same amount three times.

• Deep in the flat section. The economy is in a slump with idle capacity everywhere. Real output rises from Y₁ to Y₂ and the price level does not move at all. Extra demand buys pure output. This is the strongest case for stimulating a depressed economy: there is no inflation to pay for it.
• On the rising section. Output rises and the price level rises. Some resources are getting scarce, so part of the extra spending goes into output and part into prices. Extra demand buys output at a cost.
• At full employment. Output cannot rise, so the entire increase goes into the price level. Extra demand buys nothing but inflation.

Figure 13 runs the same experiment in the other model, starting from full employment. In the short run, output rises above Yf to Y₂ and the price level rises to PL₂, because factor prices have not yet caught up: there is an inflationary gap. Then the adjustment. With labour and materials scarce, wages and input prices are bid up, SRAS shifts left, and output slides back down to Yf at a still higher price level PL₃. The output gain was temporary. The price rise was permanent.

Put the two figures side by side and the policy argument writes itself. In the Keynesian model, a government facing a recession should raise AD, because at that point on the curve it buys output almost for free. In the monetarist/new classical model, raising AD buys a short burst of output and permanent inflation, and the only way to raise output for good is to shift LRAS.

11The assumptions behind each model, and what follows from them

The guide asks for the assumptions and implications of the two models, which is an AO3 line: the marks are for judgement, not description.

The implications run straight into Unit 3's policy subtopics. If you accept the monetarist/new classical assumptions, demand management is at best useless and at worst inflationary, and the serious work is supply-side. If you accept the Keynesian assumptions, a government that waits for wages to fall is choosing unemployment that it could have prevented, and the cost of waiting is borne by the people least able to bear it.

Evaluating between them is not a matter of picking a winner. A strong part (b) answer says what the choice depends on.

• How much spare capacity there is. With mass unemployment, the flat section is a fair description of the economy, and extra demand really does raise output without inflation. Near full employment, the models agree anyway: both say extra demand is mostly inflation.
• How flexible the labour market is. Where wages are set by long contracts and strong unions, the downward stickiness Keynes described is real. Where hiring and pay adjust quickly, the adjustment in Figure 9 is more plausible.
• How long "the long run" takes. The monetarist/new classical model does not say the adjustment is instant, and a deep recession that corrects itself over five years still costs a generation of school-leavers their first job. A judgement that weighs the size of the eventual outcome against the time taken to reach it is doing exactly what an evaluation question asks.
• What else the policy costs. Raising G to close a gap adds to government debt, which 3.3 measures and 3.6 debates.