- Conflicts between macroeconomic objectives
- Market-based supply-side policies · coming soon
- Interventionist supply-side policies · coming soon
When spending is not the problem
Picture a country (invented) where almost everyone who wants a job has one. Factories run every shift. Builders have waiting lists. Inflation is 5% and rising. The government wants faster growth, so it cuts taxes and gives people more to spend.
People do spend more. But the factories are already full, and there are no spare builders. So firms raise prices instead. Most of the extra spending becomes inflation, not extra output.
The last two lectures were about fiscal and monetary policy. Both work on demand: how much people, firms and the government spend. When an economy has spare capacity, more spending raises output. When it has none, more spending mostly raises prices. The monetary policy lecture ended by naming a third tool, which works on the other side: the economy's capacity itself. This lecture is about that tool.
What supply-side policy is
Supply-side policy is any government policy that aims to raise the economy's productive capacity (potential output), shifting long-run aggregate supply (LRAS) to the right.
The growth lecture showed where capacity comes from. Output depends on the number of workers and the hours they work, and on how much each hour produces. Each hour produces more when workers have more skills (human capital), better machines and roads (physical capital), and better methods (technology, measured as total factor productivity). Good institutions hold all of this together.
So every supply-side policy must work through at least one of these. When you meet a new policy, ask one question: which source of capacity does it raise, and how? If you cannot answer, it is probably not a supply-side policy.
Look carefully at the figure. Real GDP rises and the price level does not, but only because AD has grown too. The growth lecture showed what happens if capacity grows on its own: prices fall, and output rises only part of the way, because nobody is spending enough to buy the extra production. Supply-side policy does not cause growth by itself. It makes room for spending to grow without inflation.
In the invented country from the start, the government switches to supply-side policy. Will inflation fall next month?
Show the answer
Almost certainly not. Capacity grows slowly. New skills take years to learn. A new railway takes years to build. Even a change in rules takes time before firms invest and hire because of it. In the short run, inflation is a demand-side problem, and it needs a demand-side answer, such as higher interest rates. Supply-side policy is a long-run answer. It decides how fast the economy can grow without inflation over the next ten years, not next month.
Two families of policy
Economists agree that capacity matters. They disagree about how a government should raise it. There are two broad families of supply-side policy, and they rest on different beliefs about markets.
Market-oriented policies: remove the barriers
The first family says that markets already know how to use resources well, if the government stops getting in the way. The policies remove barriers and sharpen incentives.
- Deregulation. Remove rules whose costs are larger than their benefits. Firms spend less time and money on paperwork, and new firms can enter more easily. In the UK's "Big Bang" of 27 October 1986, the London Stock Exchange abolished fixed commissions and opened its member firms to outside owners, including foreign banks. London grew into one of the world's largest financial centres. The risk is that some rules protect safety, workers or the environment, and removing them creates costs for other people.
- Privatisation. Sell state-owned firms to private owners, as the UK did with British Telecom in 1984 and British Gas in 1986. Private owners face the pressure of profit. But the evidence suggests that the pressure of competition is what matters most. A privatised firm that is still the only seller, such as a water company, needs strong regulation.
- Lower income tax. A lower tax rate lets people keep more of each extra hour's pay, which may encourage more work and enterprise. How strong this effect is is the question in the next section.
- Benefit reform. If people gain little from moving from benefits into work, some will not move. Reforms make sure that work always pays more.
- Flexible labour markets. Make it easier to hire, to change jobs and to move workers to growing firms. Supporters say this lowers structural unemployment. Critics say workers lose security.
Interventionist policies: fix what the market under-provides
The second family says that markets, left alone, provide too little of some things that raise capacity. So the government should provide them, or pay for them.
- Education and training. A firm that trains a worker may lose that worker to a rival, so firms train too little. Public schools, colleges and support for apprenticeships raise human capital.
- Infrastructure. Roads, railways, power networks and broadband help many firms at once, and it is hard to charge each one for the benefit. So the market builds too little.
- Research and development (R&D). New knowledge spreads to other firms. The inventor cannot collect all of its value, so the market does too little research. Grants, tax credits and public universities correct this.
- Industrial policy. The government supports chosen industries. Supporters point to South Korea's growth, which the growth lecture described. Critics point to the many cases where governments supported the wrong industries. Whether governments can "pick winners" is still argued about.
Now test the difference between the two families, and the difference between supply-side and demand-side.
Put each policy in a box. Ask first: does it change what the economy is able to produce? If yes, ask: does it remove a barrier, or does the government pay for something the market under-provides?
Nine policies, sorted by how they work.
- Market-oriented supply-side: removing a rule that forces shops to close at 6 pm · selling a state-owned airline and letting rivals fly the same routes · cutting the top rate of income tax · making it easier to hire on short contracts.
- Interventionist supply-side: paying half the cost of training apprentices · building a fast rail line between two cities · a tax credit for research spending.
- Not supply-side (demand): cutting Bank Rate from 4% to 3% · a one-off payment of $500 to every household.
Supply-side policies change what the economy is able to produce. A policy that only changes spending on today's output is demand-side.
Two cards belong to two stories at once. The rail line raises demand while it is built and capacity once it is finished. The tax cut raises spending at once, and may raise capacity through incentives. The skill is to say which effect comes first, and which lasts.
Can a tax cut raise more tax?
Tax revenue is the tax rate multiplied by the income that is taxed. At a rate of 0%, revenue is zero. At a rate of 100%, nobody gains anything from earning taxable income, so it disappears, and revenue is zero again. Somewhere in between, revenue must be highest. A curve that rises from zero, reaches a peak and falls back to zero is called the Laffer curve, after the American economist Arthur Laffer.
If a tax rate is above the peak, cutting it raises revenue. If it is below the peak, cutting it loses revenue. Everything depends on where the peak is. And that depends on how strongly taxable income responds to the rate. People can work fewer hours, but they can also delay income, move it into another tax year, or hide it.
An invented country taxes a slice of high incomes worth $100 billion at 40%. Choose how strongly that income responds to the tax rate. Then move the rate and watch revenue. Find the peak for each response.
Response of taxable income:
Three Laffer curves for an invented slice of income worth $100 billion at a 40% rate. The only difference between them is how strongly taxable income responds to the rate.
| Tax rate | Revenue, weak ($ bn) | Revenue, medium ($ bn) | Revenue, strong ($ bn) |
|---|---|---|---|
| 30% | ≈ 31.2 | ≈ 32.2 | 35.0 |
| 40% | 40.0 | 40.0 | 40.0 |
| 50% | ≈ 47.8 | ≈ 46.1 | ≈ 41.7 |
| 60% | ≈ 54.2 | ≈ 50.0 | 40.0 |
| 70% | ≈ 58.9 | ≈ 51.2 | 35.0 |
With a strong response, moving between 45% and 50% hardly changes revenue: the curve is flat near its top. With a weak response, the same rise brings in billions more. The model is invented and simple. It ignores other taxes paid by the same people, and it does not separate real extra work from income that is only moved or hidden.
A real test: the UK's 50% rate. The UK raised its top income tax rate from 40% to 50% in April 2010, for income above £150 000. The government had expected it to raise about £2.7 billion in 2012–13. Before the rise, top earners moved about £18 billion of income into the year before, to pay 40% on it. In 2012, HMRC (the UK tax authority) estimated that the underlying gain was only about £0.6 billion, and that cutting the rate to 45% would cost only about £100 million a year. The rate was cut to 45% in April 2013. The Institute for Fiscal Studies warned that the estimate was very uncertain.
So, for this small group, the UK seems to have been near the flat top of its Laffer curve. That is a result about very high earners, who can move income more easily than most people. It is not evidence that cutting ordinary tax rates pays for itself.
Is a cut in income tax a demand-side policy or a supply-side policy?
Show the answer
It can be both, and the two effects work on different timescales. People have more money after tax, so they spend more almost at once: that raises AD. If the lower rate also makes people work more hours, or start businesses, that raises capacity: that shifts LRAS, slowly. A government that calls a tax cut "supply-side" is claiming the second effect. An economist asks how large it really is, and notices that the first effect arrives first.
A real case: Germany, 2003–05
The problem. In the early 2000s, Germany was often called "the sick man of Europe". Its unemployment rate had risen from below 4% in the 1970s to almost 9% on average between 1995 and 2005. Much of it was structural: long-term unemployment among people whose skills no longer matched the jobs available.
The policy. Between 2003 and 2005, the government of Gerhard Schröder passed four laws named after Peter Hartz, who led the commission that designed them. Hartz I to III reformed the job agency, to match job-seekers with vacancies faster, and created new kinds of small jobs. Hartz IV, in January 2005, cut the benefits paid to most long-term unemployed people.
The prediction. Market-oriented theory predicted that faster matching and a stronger reward for taking a job would lower structural unemployment. In AD-AS terms, more people in work would shift LRAS to the right.
What happened. The number of registered unemployed peaked at over 5 million in March 2005. The unemployment rate then fell from almost 11% in 2005 to 7.5% in 2008, hardly rose in the global recession of 2008–09, and reached 5.5% at the end of 2012.
The verdict. Unemployment clearly fell. How much of the fall the reforms caused is still argued about, because other things changed at the same time. One study (Krebs and Scheffel, 2013) estimated that Hartz IV lowered the long-run unemployment rate by about 1.4 percentage points, and Hartz I–III by about 1.5 points more. Another (Launov and Wälde) found that the benefit cut explained less than 0.1 points, and that the faster job agency did most of the work. The reforms also had costs. Krebs and Scheffel found that the long-term unemployed lost out, and that real wages fell. The reforms remain unpopular in Germany. A fair verdict: the reforms probably raised capacity, but nobody knows exactly how much, and the gains and the costs fell on different people.
Numbers to work with
The UK is the standard example of a country that wants more capacity. Its output per hour worked has grown slowly since 2008, a pattern known as the "productivity puzzle", and it is lower than in France and Germany.
The Office for National Statistics (ONS) estimated output per hour worked in 2021, using one method of measuring hours: UK £43.59, Germany £47.57, France £51.45 (all converted at purchasing power parity). How much more productive was France? How far behind was the UK?
France produced about 18% more per hour than the UK. But the UK was about 15% below France. The two numbers differ because they are percentages of different starting values. For Germany, the same method gives about 9% more per hour. The ONS warns that the German gap is between about 9% and 19%, depending on how hours are measured.
Check: and . ✓
An invented country has potential output of $800 billion. Hours worked grow 0.5% a year and output per hour grows 1% a year. A supply-side policy raises productivity growth to 1.5% a year. Compare potential output after ten years.
Growth factors multiply. Without the policy, potential output grows by each year, about 1.5%. With it, by , about 2.0%. After ten years:
The difference is about $47 billion a year, and it keeps growing. That is why economists care so much about small changes in productivity growth.
Research spending tells a similar story, with one surprise. The OECD reports that its members spent on average 2.7% of GDP on R&D in 2023, Israel 6.3% and South Korea 5.0%. For years the UK figure was thought to be about 1.7%, and in 2017 the government set a target of 2.4% by 2027. Then, in 2022, the ONS changed how it measured business research, and found far more of it. On the new method, UK R&D was 2.64% of GDP in 2023. The target was met by better measurement, not by more research. Always ask how a figure was measured before you build an argument on it.
Why supply-side policy is slow, and who pays
Every evaluation of supply-side policy comes back to three questions.
- How long? A school reform takes ten or more years to reach the workforce. A railway takes years to plan and build. Even deregulation works only when firms decide to invest because of it.
- How big? Many other things change over the same years, as the German case showed.
- Who gains, and who pays? Benefit cuts and flexible labour markets place costs on the people with the least. Interventionist policies cost public money, which must come from taxes or borrowing.
Three traps
A cut in interest rates helps firms, but it works on spending, not on capacity. Use the test from the start of the lecture: name the source of capacity the policy raises. If you cannot, it is a demand-side policy.
A rightward shift of LRAS makes room for growth without inflation. It does not create the spending to fill that room. Growth with steady prices needs capacity and demand to grow together.
The curve only says that a peak exists. Whether a tax cut raises revenue depends on which side of the peak the rate is on, and that is an empirical question. The UK evidence points to a flat top for the highest earners only.
Work each one on paper first. Open a hint only when you are stuck.
-
For each policy, name the source of capacity it raises, and say which family it belongs to: (a) more places for engineering apprentices; (b) faster planning approval for new factories; (c) a tax credit for research spending; (d) cheaper childcare so that more parents can return to work.
Hint 1
The sources are: the quantity of labour, human capital (skills), physical capital, and technology (TFP).
Hint 2
For the family, ask whether the government removes a barrier or pays for something the market under-provides.
Full answer
(a) Human capital; interventionist, because the government pays for training firms under-provide. (b) Physical capital, because firms can build and equip factories sooner; market-oriented, because it removes a barrier (deregulation). (c) Technology (TFP); interventionist, correcting the spillover from research. (d) The quantity of labour, because more parents can work; usually counted as interventionist, because the government pays for it. Some books call it "making work pay", the market-oriented idea. Either answer can earn marks if the reason is clear.
-
An invented economy has potential output of $500 billion growing at 1% a year. A supply-side policy raises the growth rate to 1.5% a year. How much larger is potential output after 10 years with the policy than without it?
Hint 1
Use growth factors, as in Worked Example 2.
Hint 2
Work out and .
Full answer
Without the policy: billion. With it: billion. The difference is about $28.0 billion, or about 5% more output. Check: , and . ✓
-
An invented government taxes a $60 billion slice of income at 40%. It raises the rate to 45%, and the slice falls to $52 billion. Does revenue rise or fall? How small could the slice become before revenue falls below its old level?
Hint 1
Revenue = rate × taxed income. Work it out before and after.
Hint 2
For the second part, solve for the slice S.
Full answer
Before: billion. After: billion. Revenue falls by $0.6 billion, so this rise is above the peak of this group's Laffer curve. Revenue stays at $24 billion if , so billion. The slice can fall by about $6.7 billion (about 11%) before revenue drops. Here it fell by $8 billion, a little too much.
-
The invented country from the start of the lecture is at full capacity, with 5% inflation. One minister wants to cut income tax for everyone. Another wants to spend the same money on technical colleges. Using AD and LRAS, explain what each policy does to inflation in the next year and to growth over the next ten years. Which would you recommend, and what would make you change your mind?
Hint 1
Both policies raise AD in the first year. Ask which one raises LRAS, and how fast.
Hint 2
The tax cut puts money in people's pockets at once. The colleges take years to produce trained workers, but the spending on them happens now.
Full answer
Next year: both raise AD in an economy with no spare capacity, so both push inflation up. The tax cut probably does so faster, because households spend part of it at once. Any incentive effect on work is likely to be small and slow. The college spending also adds to demand, through wages for teachers and builders.
Over ten years: the colleges raise human capital, so LRAS shifts right as graduates join the workforce. This makes room for growth without inflation. The tax cut raises LRAS only if people work noticeably more, and the evidence for ordinary earners suggests a modest effect.
A reasonable recommendation is the colleges, with the central bank keeping demand in check in the short run, for example with higher interest rates. You might change your mind if there were clear evidence that high tax rates were stopping people from working, or if the colleges taught skills that firms do not need. A strong answer names both the short-run and the long-run effects, and the evidence it would want.
Over a few years, demand decides how close an economy gets to its capacity. Over decades, capacity decides almost everything: wages, living standards, what a government can afford.
The two families of supply-side policy are not really a choice between right and wrong. They are two answers to one question that has to be asked again for every problem: is this market failing, and if so, can the government do better? Training and research are good cases for government help. Rules that only protect existing firms are good cases for removal. And even a good supply-side policy can clash with other goals, such as low inflation now. The next lecture is about those clashes.
- Supply-side policy aims to raise the economy's productive capacity, shifting LRAS to the right.
- Every supply-side policy works through the quantity of labour, skills, physical capital or technology, and naming which one is the test of whether a policy is supply-side.
- Market-oriented policies remove barriers and sharpen incentives; interventionist policies pay for skills, infrastructure and research that markets under-provide.
- The Laffer curve shows that revenue peaks at some tax rate, but where the peak sits depends on how strongly taxable income responds, which must be measured.
- Supply-side policy works slowly, its effects are hard to measure, and its gains and costs often fall on different people.
- Growth without inflation needs capacity and demand to grow together.