Aggregate Supply
Aggregate Supply
FREE RESOURCES
EXAM RESOURCES
PREMIUM
Every firm in an economy decides how much to produce; added together, their plans show how much the whole economy will supply.
Vandria is a middle-income country that imports almost all of its oil and exports electronics and garments. Each of its rice farms, garment factories, electronics assemblers and ports decides how much to produce this year.
Economists add these plans together, using two measures: one for prices and one for output. The first is the price level, the average of all prices in the economy, shown as an index, where 100 stands for the prices of a chosen base year. The second is real national output: the goods and services an economy produces, valued at constant prices, which means the prices of one fixed year are used for every year. So a rise in real output means more was produced, not that the same things cost more. The money value of output, what the output sells for, can rise just because prices rose. So whenever the quantity produced changes, real output changes, whatever happens to prices.
Aggregate supply (AS): the total real output that all the firms in an economy plan to produce at each possible price level over a given period.
Put simply, aggregate supply asks how much all the country's firms would produce in total if prices across the economy were higher or lower. In Vandria, firms plan to produce $800bn (800 billion dollars) of real output a year at a price level of 100.
An AS diagram looks like the supply diagram for one market, but it works differently. Its vertical axis shows the price level, and its horizontal axis shows real national output. In one market, only that good's price changes. On an AS diagram, a change in the price level moves all prices together. Over time, this includes input prices, the prices firms pay for labour and materials. How long input prices take to change is why time matters.
In the short run, some input prices are fixed, such as wages fixed in workers' contracts and rents fixed in agreements for buildings. In the long run, every input price has had time to adjust.
So economists draw two AS curves. The curve for the short run is called short-run aggregate supply (SRAS). The curve for the long run is called long-run aggregate supply (LRAS). In the long run, every input price has adjusted, so costs rise as much as prices and a higher price level gives firms no reason to produce more. Then only the economy's resources limit how much it can produce. So LRAS is set by productive capacity (the most the economy can produce with its workers, machines and land used at a normal rate).
While some costs are fixed in the short run, a higher price level raises the profit on each unit, so firms produce more.
This year, prices of finished goods across Vandria rise by 4%. Most workers are paid money wages: a set amount of money, whatever that money can buy. Their contracts fix these wages until next year.
What matters to a firm is its unit cost (the cost of producing one unit). Materials may cost more this year too, but money wages are part of the cost of every unit, and they are fixed. So even if materials cost 4% more, the unit cost rises by less than 4%, the rise in the selling price.
Short-run aggregate supply (SRAS): the total real output that firms plan to supply at each price level in the short run, when the prices of some factors of production, such as money wages, are fixed.
Put simply, SRAS shows how much firms will produce at each price level while some of their costs, mainly wages, cannot change yet.
Why do firms produce more? The gap between the selling price and the unit cost is the profit per unit (what a firm earns on one unit after paying that unit's cost). Because selling prices rise by more than unit costs, profit per unit rises. At the old prices, some extra units would have earned too little to be worth making. Now they earn enough, so firms produce more, for example by working extra hours. This is why the SRAS curve slopes upward: a higher price level goes with more real output.
The extra profit per unit lasts only until contracts are renewed. Then money wages rise to match the price level, so unit costs rise as much as prices, and the extra profit per unit disappears. So the reason to produce more disappears too.
Exam Matters: When you explain the slope, name the cost that is fixed, such as money wages, and say why that raises profit per unit. Without the fixed cost, your chain gives no reason for firms to produce more.
A change in the price level moves the economy along the SRAS curve, but a change in firms' costs moves the curve itself.
Two kinds of change alter the real output Vandria's firms plan to supply. One is the price level, on the vertical axis. The other is anything not shown on either axis, such as firms' costs.
A movement along the AS curve: a change in the real output firms plan to supply that is caused by a change in the price level, with every other influence on supply unchanged.
A shift of the AS curve: a change in the real output firms plan to supply at every price level, caused by a change in something other than the price level, such as firms' costs.
Put simply, a movement is a slide from one point to another on the same curve. A shift means a new curve is drawn, because firms now plan a different output at every price level.
The key difference is what changed. If the price level changed, the economy moves to another point on the same curve. If something not shown on either axis changed, such as costs, the whole curve moves.
A fall in firms' costs reverses each of these changes. If costs fell by about 7% instead, firms would need a price level about 7 points lower to keep the same profit per unit, so the curve moves down by 7 points. To supply $800bn, they would now need only 93. So at 100 they supply the output they used to supply at 107 on the old curve: 7 × $5bn = $35bn more, $835bn. At every price level, firms now plan more real output, so the SRAS curve shifts to the right.
Exam Matters: Draw a movement as an arrow along one curve between two points, each read off to both axes. Draw a shift as a second curve, labelled SRAS₁, with an arrow showing its direction.
When the fuel and materials firms buy cost more, each unit costs more to make, so the SRAS curve shifts left.
Vandria imports almost all of its oil, and this year world oil and gas prices rise sharply. Almost every firm uses energy (the fuel and electricity that run its machines), and many use raw materials (the basic materials they turn into goods, such as cotton).
More expensive energy raises each firm's unit cost (the cost of producing one unit) at every level of output, because every unit needs fuel and power. So at each price level, each unit earns less profit, and some units no longer earn enough to be worth making. Firms plan less real output at every price level.
This is a shift of the whole curve, called a decrease in SRAS. The curve moves left, because firms plan less output at every price level. It also moves up, because firms need a higher price level to supply the same output. Its opposite, an increase in SRAS, moves the curve right and down.
How far the curve moves depends on how large a share of firms' costs the input is. For the same rise in its price, an input that is a large share of costs raises unit costs more than one that is a small share, so the curve shifts further.
When energy prices fall, each step reverses. If unit costs fell by about 5%, the curve would move down 5 points, and at a price level of 100 firms would supply 5 × $5bn = $25bn more, or $825bn: an increase in SRAS.
The value of a country's currency decides what its firms pay, in their own money, for the inputs they buy abroad.
Vandria's firms buy fuel, machine parts and cloth abroad, in US dollars. What these cost in Vandria's own money depends on the exchange rate (the price of one currency in terms of another). In many countries the exchange rate floats: the demand for and supply of the currency set it. Vandria's exchange rate floats.
Depreciation: a fall in the value of a currency against other currencies in a floating exchange rate system.
Appreciation: a rise in the value of a currency against other currencies in a floating exchange rate system.
Put simply, after a depreciation each unit of Vandria's currency buys fewer dollars, so a good priced in dollars costs more in Vandria's money. After an appreciation each unit buys more dollars, so the same good costs less.
The key difference is what each does to imported inputs, the materials, parts and fuel firms buy from abroad. Suppose their price in dollars does not change. After a depreciation, firms pay more of their own money for them; after an appreciation, they pay less. Take a depreciation that makes imported inputs 10% more expensive in Vandria's money, and an appreciation that makes them about 10% cheaper.
Why do unit costs rise by only about 3% when imported inputs cost 10% more? Imported inputs are only part of each unit's cost. Firms also pay wages, rent and local suppliers, and the exchange rate does not change these. So the more of their costs firms spend on imported inputs, the larger the shift.
Unit costs change at every level of output, so the curve shifts. At a price level of 100, a rise of about 3% in unit costs moves the curve up 3 points, and each point goes with $5bn of output, so firms plan 3 × $5bn = $15bn less, $785bn. After the appreciation, unit costs fall by about 3%, so the curve moves down 3 points, and firms plan $15bn more, $815bn.
The same depreciation also makes Vandria's electronics cheaper for foreign buyers, which tends to raise net exports (exports minus imports), part of aggregate demand (AD, total spending on the economy's output). So it acts on two curves: SRAS shifts left through costs, and AD shifts right through exports.
Exam Matters: If a question asks about the effect of an exchange-rate change on SRAS, follow the cost of imported inputs through to the curve. The effect on exports is real, but it acts on AD, so it is not part of the answer about SRAS.
A tax a firm pays on each unit it sells, each worker it employs or each input it buys is part of its costs, so it shifts SRAS.
When a firm in Vandria sells a litre of fuel, part of the price goes to the government. This is an indirect tax (a tax on spending, paid by the seller), such as a sales tax or an excise duty (a tax on one kind of good, such as fuel). The firm pays it on every unit it sells.
Many firms also pay a payroll tax (a tax employers pay on each worker they employ). It is usually a percentage of each wage, so a higher rate raises what every worker costs the firm.
A tax on an input the firm buys, such as a duty on imported steel, works the same way: the firm pays it on every tonne it uses. The payroll tax is paid on the workers who make every unit. So each of these taxes, called a tax on production because firms pay it as a cost of producing, is part of each unit's cost, and a change in any of them works like a change in energy prices: unit costs change at every level of output, and SRAS shifts.
Say a tax on production rises and unit costs rise by about 4%. At a price level of 100, 4% is 4 points, so the curve moves up 4 points, and each point goes with $5bn of output: firms plan 4 × $5bn = $20bn less at 100. A tax cut that lowers unit costs by about 2% moves the curve down 2 points: 2 × $5bn = $10bn more.
Income tax is different. It is a tax on the income people earn, and workers pay it out of their wages, so it is not part of a firm's unit cost. A cut in income tax gives households (people as buyers of goods and services) more to spend. AD is total spending on the economy's output, so more spending by households shifts AD to the right, not SRAS. A rise in income tax leaves them less to spend, so AD shifts left. Over years, a lower tax on income or profits can also make working or investing more rewarding, so more people may work or firms may buy more machines. A higher tax can do the opposite. This is a long-run effect on productive capacity, not on SRAS.
Energy prices, the exchange rate and taxes on production all change the same thing, unit costs, and none of them changes productive capacity at the time the cost changes. So their effects add or offset. Suppose an appreciation lowers Vandria's unit costs by about 3% in the same year as the tax rise adds about 4%. Unit costs rise by only about 1%, so SRAS shifts up 1 point. One point goes with $5bn, so at a price level of 100 firms plan about $795bn instead of $800bn: a small shift to the left. When one cause reverses, for example when the tax returns to its old rate, its part of the shift reverses, and the curve moves back if nothing else has changed.
Real Example: On 1 July 2020 Saudi Arabia raised its VAT, a sales tax on goods and services, from 5% to 15%. A shop that wanted to keep the same amount on each item after paying the tax had to charge buyers more. This is how a tax on sales works like a rise in costs: firms need a higher price to supply the same output.
Exam Matters: When a tax that firms pay rises, draw a second SRAS curve, labelled SRAS₁, to the left of SRAS, and read output at one price level on both curves. The lower output on SRAS₁ shows a shift of the curve, not a movement along one curve.
Once every wage and input price has risen to match the price level, only the economy's resources limit what it can produce.
Look at Vandria over several years. Every wage contract is renewed, and every input price rises to match the price level. Firms' costs now rise as much as their selling prices, so a higher price level no longer raises profit per unit, and the reason SRAS slopes upward is gone. What limits output now is Vandria's resources: its workers, machines, land and enterprise (the people who start and run firms), and how well they are used.
Productive capacity: the maximum real output that an economy can produce when all its resources are employed at a normal, sustainable rate.
Put simply, productive capacity is the most an economy can keep producing, year after year, without wearing out its workers or machines. Vandria's productive capacity is $800bn of real output a year.
Long-run aggregate supply (LRAS): the real output an economy can supply when the prices of all factors of production have fully adjusted to the price level.
Put simply, LRAS shows what the economy can produce once no cost is held fixed, which is its productive capacity. So Vandria's LRAS is drawn at $800bn.
Why a normal, sustainable rate? For a while, firms can produce more. At a price level of 104, Vandria's firms supply $820bn on SRAS, $20bn above capacity, with overtime and extra hours. But workers tire, machines that run all night need repair, and costs rise as contracts are renewed, so this output does not last.
How large is capacity? That depends on two things: how many people of working age (old enough to work and below retirement age) work or look for work, and how much each worker produces.
Labour force: the people of working age who are in work or actively looking for work.
Put simply, the labour force is everyone of working age who is working or looking for work now. Children, retired people and full-time students who are not looking for work are not in it. The second thing capacity depends on is output per worker: the real output one worker produces in a year.
So only two changes move LRAS: a change in the labour force, or a change in output per worker. A change in capacity shifts the curve; a change in the price level alone changes neither term, so it does not. Economists draw LRAS vertical at capacity, which means the same real output at every price level. But they disagree about its shape below capacity, because they disagree about how long output can stay below capacity.
When many workers and machines stand idle, firms can produce more without raising their costs.
Imagine Vandria in a deep downturn, when spending and output have fallen sharply. Many workers are unemployed, and many factories use only half their machines. Workers and machines that could produce but stand idle are called spare capacity.
Keynesian economists (economists who follow John Maynard Keynes) argue that spare capacity changes the shape of the LRAS curve. They draw it in three ranges, each with its own reason:
Why can an economy stay on the flat range for years? Keynesians argue that money wages are sticky downwards: they do not fall easily. Workers resist pay cuts and employers avoid them, so wages fall slowly, if at all, even when unemployment is high. If wages fell, firms' costs would fall and they would supply more. But while wages stay where they are, costs stay where they are too, and nothing pushes output back towards capacity.
Where Vandria sits on the curve depends on aggregate demand (AD, total spending on the economy's output): the economy is where AD meets the curve. A rise in AD moves the economy to the right along the curve. On the flat range, output rises and the price level stays at 90. On the rising range, output and the price level both rise. On the vertical range, only the price level rises. A fall in AD reverses each of these changes. With sticky wages, Vandria could stay well below $800bn until AD rises.
Exam Matters: When you draw it, show all three ranges and label the vertical range at productive capacity. A single smooth upward line loses the flat range, which is where the Keynesian argument starts.
If money wages fall when unemployment is high, output returns to capacity after a fall in demand.
Start where Vandria's aggregate demand (AD), SRAS and LRAS curves all meet: real output of $800bn, its productive capacity, at a price level of 100. This is a long-run equilibrium (where AD meets SRAS on the LRAS curve).
Classical economists (economists who argue that markets correct themselves) say money wages are flexible: they fall when unemployment is high and rise when labour is scarce. Follow what this means when Vandria's AD falls by $60bn. At the old price level of 100, the real output demanded is now $740bn, on the new curve AD₁.
At the old price level of 100, firms still plan to supply $800bn, but only $740bn is demanded, so unsold goods push the price level down. Today, firms' costs are fixed by contracts, so as the price level falls, profit per unit falls and firms produce less: the economy moves down along SRAS. At the same time, a lower price level means the same money buys more, so the real output demanded rises along AD₁. In this case, each point the price level falls adds $5bn to the real output demanded, the same amount that each point takes off SRAS.
So the two curves meet at $770bn and a price level of 94, a short-run equilibrium (where AD meets SRAS). Only $30bn of the $60bn fall shows in output, because the lower price level has raised the real output demanded by the other $30bn.
At $770bn, Vandria produces $30bn less than its capacity. Unemployment rises at once, and over the following years, as contracts are renewed, the adjustment runs in four steps:
Money wages keep falling while there is unemployment, so SRAS keeps shifting right. As it does, the price level keeps falling, and the real output demanded keeps rising along AD₁. The shifts stop when unemployment has gone: at 88, twelve points below 100, the real output demanded is $740bn + 12 × $5bn = $800bn, Vandria's capacity. So in the long run, the fall in AD has changed only the price level, from 100 to 88, the new long-run equilibrium. Real output is back at $800bn. This is what a vertical LRAS shows. Only a change in productive capacity moves the curve itself. A rise in AD works the same way in reverse. If AD rises by $60bn, the economy first moves up along SRAS to $830bn at a price level of 106. Labour is now scarce, so as contracts are renewed, money wages and costs rise, SRAS shifts left, and output returns to $800bn at a price level of 112.
Exam Matters: On a classical diagram, draw AD, SRAS and a vertical LRAS, and mark both the short-run and the long-run equilibrium. Label the shifted curves, such as AD₁ and SRAS₁, so each step of the adjustment can be seen.
The two shapes rest on one disagreement: how quickly money wages adjust when demand falls.
Classical and Keynesian economists look at the same economy and draw different LRAS curves. The difference comes from one assumption: how quickly money wages adjust when aggregate demand changes.
The second and third rows follow from the first. The last row is where the two views agree. If wages fall quickly, unemployment pushes costs down and output returns to capacity, so in the long run only the price level changes. If wages are sticky, costs do not fall, so nothing pulls output back, and the economy can stay on the flat range.
Which view fits? It depends on the economy. The classical view fits better than the Keynesian view over long periods, because wages then have time to adjust, and where wages are renegotiated (agreed again) often, because costs then follow unemployment more quickly. The Keynesian view fits better in a deep downturn, when many resources are idle and workers resist pay cuts. Neither is simply right. The same economy can fit the Keynesian view in a deep downturn and the classical view a few years later.
Exam Matters: When a question asks which shape fits an economy, use its evidence on wages and spare capacity: how deep the downturn is and how often wages are renegotiated. Base your judgement on that evidence.
A better way of producing lets the same workers produce more, so the economy can produce more.
A factory installs an automated line. A farmer plants a seed that survives drought. A delivery firm uses software that plans shorter routes. Each is a change in the state of technology: the knowledge and methods an economy has for turning inputs (labour, materials and machines) into output.
Capacity has two terms: the labour force, and output per worker. A better method leaves the labour force the same, but each worker now produces more. So technology acts through output per worker.
Two conditions matter. First, a new method usually has to be built into new machines, so it arrives as firms buy those machines: gradually, not overnight. Second, unlike a fall in energy prices, it is not undone when a price changes back. Once a better method is known, the capacity it adds usually stays.
Real Example: In Kenya, M-Pesa lets people send and receive money by mobile phone, so a trader need not travel to a bank to pay a supplier. A 2016 study estimated that the spread of M-Pesa helped about 194,000 households, around 2% of Kenya's households, leave extreme poverty, and that many women moved from farming into business. A trader who no longer travels to a bank can spend that time selling: the kind of change that can raise output per worker, though the study measured poverty and jobs, not output.
Output can rise because more people work, or because each person produces more; only the second raises output per worker.
Vandria's output rises this year. Is that because more people are working, or because each worker produces more? Total output rises in both cases, but the two changes are different.
Productivity: output per unit of input in a given period, usually measured as output per worker or per hour worked.
Put simply, productivity is how much each worker, or each hour of work, produces. Suppose a firm hires more people, and each new worker produces the same as the others. Output rises, but productivity does not.
Productivity is one of the two terms of productive capacity: output per worker. When it rises, capacity rises with the same labour force.
What raises output per worker? Four things, each giving workers more to produce with or a better way of working:
Productivity can also fall. Suppose machinery grows old and is not kept in good repair, and output per worker falls by 1%, from $20,000 to $19,800. With the same 40 million workers, capacity falls from $800bn to 40 million × $19,800 = $792bn, so LRAS shifts left.
Training lets each worker produce more, but its effect on capacity arrives years after the money is spent.
A worker who can read a technical manual or repair a machine produces more with the same equipment than one who cannot. Education and skills (schooling, apprenticeships, which teach a trade while the learner works, and training at work) raise output per worker, so they raise productive capacity.
In Vandria, a national training programme raises output per worker by 2%, from $20,000 to $20,400 a year. With the same 40 million workers, capacity rises from $800bn to $816bn. But this happens years after the programme starts:
This delay is a time lag (the delay between a cause and its effect). In the year it is spent, government spending on education is part of aggregate demand (total spending on the economy's output): it pays teachers and builds colleges, so it acts on AD at once. Its effect on LRAS needs trained workers at work, which can be years later. So it cannot raise capacity quickly.
Skills must also match what firms need. Training in skills that no employer hires for leaves output per worker where it was.
Real Example: Singapore's SkillsFuture scheme gives every citizen aged 25 and over an opening credit of S$500 to spend on approved training courses. Workers can add skills in the middle of their careers, not only at school. This is one way a government can try to keep the skills of its labour force matched to what firms need.
Exam Matters: When you evaluate education as a way to raise LRAS, use the time lag: the effect is real but slow, so it cannot raise capacity this year. Then ask whether the skills match what firms need.
Rules and taxes change the reward for working and investing, so over years they can raise or hold back capacity.
Think about what it takes to start a firm, hire a worker or build a factory. Government regulations (rules firms must follow) decide how long each step takes and what it costs.
Some rules add cost and no output. A licence that takes eight months, or the same form filed with three ministries, uses time that could have gone into production. Fewer new firms start and less capital is built, so each worker has less capital to use. Output per worker, one of the two terms of capacity, is then lower, and so is capacity.
Other rules raise capacity. Courts that enforce contracts, clear property rights (rules on who owns what) and product standards (rules on quality and safety) that buyers trust let firms invest with confidence, because they know they will keep the return, the profit their investment earns. So the test of a rule is whether its cost is repaid in output.
Taxes act on LRAS through incentives (rewards that make an action worth taking), and the effect builds up over years:
So one tax can act on both AS curves. Take a rise in a payroll tax, paid on each worker employed. Now, it is part of the cost of each unit, so it shifts SRAS left. Over the following years, it also lowers the reward for employing people, so firms may build fewer new workplaces, and output per worker grows more slowly: it can shift LRAS too.
Real Example: Rwanda lets a new business register online, usually within about six hours and at no charge, with one application that also gives it a tax number. Time that people starting a firm would spend on paperwork can go into production instead. Rules like this cut the time and cost of starting a firm, which is the kind of change that can raise capacity over time.
Exam Matters: When a question names a tax, decide first which effect it asks about: the cost of supplying each unit now, which moves SRAS, or the reward for working and investing over years, which moves LRAS.
How many people can work depends on the population's ages, not only on its size.
Vandria has far more people than workers. Children, retired people and full-time students are not in its labour force. So the number who can work depends on people's ages, not only on how many there are.
Demography: the study of the size, structure and changes of a population.
Put simply, demography looks at how many people a country has, how old they are, and how this changes as people are born, grow older, die and move. For capacity, the key question is how many are of working age (old enough to work and below retirement age, the age at which most people stop working).
Participation rate: the percentage of the working-age population that is in the labour force.
Put simply, the participation rate is the share of people who could work who are working or looking for work. The rest are studying, caring for family or have stopped looking.
An ageing population (one in which the share of older people grows) changes the working-age population, the first number in this sum. More people pass retirement age, and fewer young people replace them. Suppose Vandria's working-age population falls by 1 million, to 49 million. At the same participation rate of 80%, its labour force is 49 million × 80% = 39.2 million. With output per worker unchanged at $20,000, capacity is 39.2 million × $20,000 = $784bn. So LRAS shifts left; if other changes are raising capacity at the same time, it shifts right more slowly than it would have.
A higher participation rate works the other way, because more of the working-age population then works or looks for work. If childcare and flexible hours raise the participation rate from 80% to 85% of a working-age population of 50 million, the labour force is 50 million × 85% = 42.5 million, and capacity is 42.5 million × $20,000 = $850bn.
Births work slowly. More births add to the labour force only when the children reach working age, fifteen or more years later.
Real Example: In 2024, 29.3% of Japan's people were aged 65 or over, the highest share among 200 countries and regions. Nigeria's population is young: half its people are younger than about 19. The same question, how many are of working age, matters for the labour forces of both countries in the years ahead.
People who move into or out of a country change how many work there, and often how much each produces.
Each year some people come to live in a country, called immigration, and some leave to live in another country, called emigration.
Net migration: the number of people who come to live in a country minus the number who leave to live in other countries, over a given period.
Put simply, net migration is how much a country's population changes because people move, once arrivals and departures are set against each other. A net inflow means more arrive than leave; a net outflow means more leave than arrive.
Many people move to find work, so migrants are more often of working age than the population as a whole. A net inflow therefore raises the labour force by a larger share than it raises the population. In Vandria, a net inflow of 1.5 million working-age people, all in the labour force, raises the labour force from 40 million to 41.5 million. If each worker still produces $20,000, capacity rises from $800bn to 41.5 million × $20,000 = $830bn, and LRAS shifts right.
Skills matter as well as numbers. An inflow of engineers or nurses can raise output per worker as well as the number of workers, if they produce more than the average worker. An outflow does the reverse. When trained doctors and engineers emigrate, the economy loses workers and some of its most productive ones, so both terms of capacity fall and LRAS shifts left.
So capacity can change even when net migration is zero. If the people leaving are more skilled than those arriving, and as many of them work, output per worker falls although the labour force is the same size.
Real Example: In February 2024, the head of Nigeria's nursing council (the body that registers nurses) said about 42,000 nurses had left the country in the previous three years. Each of those nurses took years to train. This is how emigration can take skills, not only numbers, out of an economy's labour force.
Exam Matters: When you examine migration, weigh the number effect against the skills effect. Net migration of zero can still change capacity, because the skills of the people who leave and the people who arrive can differ.
A firm that can lose customers to a cheaper rival must cut waste and adopt better methods.
Picture a market where the few firms secretly agree their prices. Firms that agree to fix prices instead of competing are called a cartel. Each member can survive with high costs and old methods, because no rival will offer buyers a lower price.
Other things protect firms from rivals too. A merger (two firms joining into one) can leave one firm controlling most of a market. Barriers to entry (things that make it hard for new firms to enter a market, such as a licence only existing firms can get) keep new rivals out.
Competition policy: government policy that aims to make markets more competitive, by stopping firms from fixing prices, controlling mergers and removing barriers that keep new firms out.
Put simply, competition policy makes firms face real rivals. A competition authority (the government body that applies these rules) can fine a cartel, block a merger and remove barriers to entry.
Why does this raise LRAS? A firm that could lose its customers to a cheaper rival must cut waste, adopt better methods and invest in new products. Less waste and better methods mean the same workers and machines produce more, so output per worker rises:
Competition also moves resources. Workers and capital leave firms that cannot compete and go to firms that use them more productively. The effect is slow and hard to see in any one year, and it depends on the authority making firms obey the rules.
Real Example: In 2009 the Competition Commission of Pakistan fined 20 cement makers about 6.35 billion Pakistani rupees for sharing out production to keep prices high. The firms appealed, and the case was still before the appeal court in 2025. This shows that enforcement can take many years, and the pressure on firms to cut costs depends on it.
Every firm in an economy decides how much to produce; added together, their plans show how much the whole economy will supply.
While some costs are fixed in the short run, a higher price level raises the profit on each unit, so firms produce more.