Types of unemployment — frictional structural cyclical — Economics

Before this

This leads to

Four people, one number

Meet four people who are all looking for work this month.

Priya is a nurse. She moved to a new city to live near her parents, and she has two interviews next week. Kofi was a welder in a shipyard for twenty years. The shipyard closed, and nobody in his town needs welders now. Mateus made car parts. During a recession, people stopped buying cars, and his factory laid off a third of its workers. Hana works at a beach hotel that closes every winter.

In the official figures, each of them is simply one unemployed person. But their problems are completely different. Priya needs a few weeks. Kofi needs new skills, or a move to another region. Mateus needs people to start buying cars again. Hana needs the summer.

Here is the surprising part. Even in a healthy economy, people like Priya and Hana always exist, so unemployment never falls to zero. So the useful question is not "how high is unemployment?" It is "what kind of unemployment is it?" The cause decides the cure. A cure that helps one kind can do nothing for another, or even make things worse.

Who counts as unemployed

First, a careful definition. A person is unemployed if they have no job, are available to start work, and are actively looking for work. A retired grandmother is not unemployed, and neither is a full-time student who is not looking. They are outside the labour force.

The labour force is everyone who has a job plus everyone who is unemployed. The unemployment rate is the share of the labour force that is unemployed:

unemployment rate = (number unemployed \div labour force) \times 100

The next lecture shows how countries count them. This one asks why people are unemployed. There are three main types, plus two smaller ones.

Frictional unemployment: the time it takes to find a match

Frictional unemployment is unemployment while people move between jobs or search for their first job. Priya is frictionally unemployed. She has skills that employers want, and there are jobs for nurses in her new city. The only thing missing is time. She must find the vacancies, apply, wait for interviews and agree a start date.

Some frictional unemployment is a good sign. It means that people are free to leave a job for a better one. An economy where nobody ever changed jobs would waste a lot of talent.

The size of frictional unemployment depends on two things: how many people start searching each week, and how long a search lasts. Multiply them together. If 0.3% of the labour force starts searching every week, and each search lasts 10 weeks, then at any moment about $0.3 \times 10 = 3%$ of the labour force is searching. This is why better information about vacancies matters. It does not stop people moving. It shortens each search.

Structural unemployment: the jobs exist, but not for these workers

Structural unemployment happens when the skills or the location of workers do not match the jobs that are available. Kofi is structurally unemployed. There may be many vacancies in his country, but they are for care workers, engineers or programmers, often in other cities. His welding skills, built over twenty years, are no longer wanted where he lives.

The structure of the economy has changed. The last lecture described how growth replaces old industries with new ones, a process the economist Joseph Schumpeter called creative destruction. It promised that this has a cost. Structural unemployment is that cost. The economy as a whole gains, but particular workers, in particular towns, lose.

There are three usual causes. New technology replaces some tasks, such as routine office work. Trade moves production to where it costs less. Demand moves away from a product, as it has moved away from coal.

Britain's coal industry shows how deep and how long this can be. In 1984 there were about 173 coal mines still working in Britain. Over the next three decades almost all of them closed. The last deep coal mine, Kellingley Colliery in Yorkshire, closed on 18 December 2015, and its last 450 miners lost their jobs. Many mining towns had few other large employers. A miner's skills did not transfer easily to the new jobs, and those jobs were often far away.

That is why structural unemployment can last for years. Retraining takes time. Moving to another region is expensive, and it means leaving family and a home.

Stop and think

A country has 800 000 unemployed people and 800 000 job vacancies. A politician says: "There is a job for everyone, so anyone without one is not trying." Is the conclusion right?

Show the answer

No. Equal numbers do not mean the jobs and the people fit each other. The vacancies may need skills the unemployed do not have, or they may be in a different part of the country. That is structural unemployment. Some of the unemployed will also be in the middle of a normal search, which is frictional unemployment. Many of them may be trying hard. The numbers can match while the people and the jobs do not.

Cyclical unemployment: not enough spending

Cyclical unemployment is caused by too little total spending in the economy. It is also called demand-deficient unemployment. Mateus is cyclically unemployed. His skills are fine, and his factory is in the right place. But in a recession, households and firms spend less. Firms sell less, so they produce less, so they need fewer workers.

The AD-AS diagram from the growth lectures shows this. Aggregate demand (AD) is total planned spending on the economy's output. When it falls, the AD curve shifts left, from AD₁ to AD₂. Real GDP falls from potential output, $Y^{*}$ , to $Y_{2}$ .

A fall in aggregate demand from AD₁ to AD₂. Real GDP falls from potential output

Y^{*}

, where the long-run aggregate supply (LRAS) line stands, to

Y_{2}

. The difference is the output gap. Producing less output needs fewer workers, so the output gap shows up as cyclical unemployment.

The name comes from the business cycle, the rise and fall of spending through booms and recessions.

In the recession that followed the financial crisis of 2007 to 2008, the United States shows how fast this can happen. The unemployment rate was 5.0% in December 2007. It peaked at 10.0% in October 2009 (US Bureau of Labor Statistics). Workers did not suddenly lose their skills. Spending collapsed, and firms cut jobs.

Two smaller types: seasonal and real-wage

Seasonal unemployment happens in jobs where demand changes with the time of year. Hana's hotel closes in winter. Farm workers may have no harvest to bring in for some months. Tourism, farming and building are the usual examples. This unemployment is predictable, and many workers in these jobs plan for it.

Real-wage unemployment, also called classical unemployment, is a different idea. It happens if wages are held above the level where the labour market clears. A legal minimum wage is a price floor in the market for labour. You met price floors in the lecture on price controls. If the floor is above the equilibrium wage, firms want to hire fewer workers than want to work.

A minimum wage above the equilibrium wage

W^{*}

. At the higher wage, firms want to hire

Q_{d}

workers, but

Q_{s}

people want to work. In this simple model, the gap is real-wage unemployment.

This model is contested, and you should say so whenever you use it. Where a few large employers set wages, a minimum wage can raise pay without cutting jobs. A famous 1994 study by David Card and Alan Krueger compared fast-food restaurants in two neighbouring US states after one of them raised its minimum wage. It found no fall in employment. Other studies have found small job losses. The honest summary is that the effect depends on how high the minimum wage is, compared with typical wages.

Now practise telling the types apart. The question to ask is always: what is missing?

Sort — which type of unemployment?

Put each person in the box for their type of unemployment. Ask what is missing for each person: time, the right skills or place, enough spending, or the right season.

Eight people, sorted by the type of their unemployment.

Person	Type	What is missing
Aiko finished university in June and has had three interviews.	Frictional	Time to find the right match.
Dmitri left his sales job to find one nearer home.	Frictional	Time: his skills are wanted.
Sunil processed paper insurance claims, which software now does.	Structural	Skills that employers now want.
Grace's shipyard closed; new jobs are in health care, in another region.	Structural	The right skills and the right place.
Rafael built houses until a recession stopped house buying.	Cyclical	Enough total spending.
Tomasz lost his factory job when spending fell; the factory is now hiring again.	Cyclical	Spending, which has returned.
Ines picks fruit, and there is no harvest in winter.	Seasonal	The right time of year.
Arjun guides walkers at a mountain resort that closes for the rainy months.	Seasonal	The right time of year.

What decides the type is not the job, but the reason it disappeared. Rafael's and Tomasz's jobs come back when spending returns. Grace's job will not come back.

The natural rate of unemployment

Suppose the economy is producing at its potential output. Total spending is just enough to employ the workers that the economy can use. Then cyclical unemployment is zero. But frictional and structural unemployment are still there. People are still changing jobs, and some still have the wrong skills or live in the wrong place.

The unemployment that remains is called the natural rate of unemployment:

natural rate = frictional unemployment + structural unemployment

The word "natural" does not mean good or unchangeable. It means the rate that remains when spending is at the right level. Better job matching and retraining can lower it.

Why not use extra spending to push unemployment below the natural rate? Because the workers who remain are not the kind that extra spending reaches. Firms that want more workers must compete for the few suitable ones, so wages rise. Firms pass the higher costs on in prices, so inflation rises. If spending keeps pushing, inflation keeps rising. For this reason, many economists use a second name, the NAIRU: the non-accelerating inflation rate of unemployment. Strictly, the two names come from slightly different theories, but at this level they describe the same floor.

No one can observe the natural rate directly. It is estimated from data, the estimates differ, and the rate itself changes as the structure of the economy changes.

Matching the cure to the cause

Now the reason for all this sorting becomes clear. Each type has a different cause, so each needs a different cure.

Type	What is missing	Policies that help
Frictional	Time to match	Better information about vacancies; job centres and job websites
Structural	Skills or location	Retraining, education, help to move, investment in new industries in struggling regions
Cyclical	Total spending	Raising aggregate demand: more government spending, lower taxes, lower interest rates
Seasonal	The season	Little policy needed; sometimes work that fits other seasons
Real-wage	A wage the market clears at	Debated: depends on the labour market, see above

The hard part is that all the types exist at once, and nobody knows the mix exactly. A government must diagnose before it prescribes. Try it in the model below.

Model — diagnose, then prescribe

This is an invented model with invented numbers. Choose an economy and a policy, then choose the size of the policy. The bars show unemployment before and three years after. Watch what each policy changes, and what happens to inflation.

Economy:

Policy:

Size of the policy: 2 (0 to 4)

Three invented economies, each given one of three policies of size 2. Unemployment is shown three years later, with inflation beside it. Every economy starts with inflation of 2%.

Economy (start)	Raise spending	Retrain and invest	Job matching
Northland, a recession (7%: 2 + 1.5 + 3.5)	5.0%, inflation 2%	6.1%, inflation 2%	6.5%, inflation 2%
Eastport, shipyards closed (7%: 2 + 5 + 0)	6.5%, inflation 5%	6.0%, inflation 2%	6.5%, inflation 2%
Southvale, at the natural rate (3.5%: 2 + 1.5 + 0)	3.0%, inflation 5%	2.6%, inflation 2%	3.0%, inflation 2%

The three numbers in brackets are frictional + structural + cyclical. Northland and Eastport start with the same unemployment rate, but the best policy is different. Extra spending works in the recession. In Eastport and Southvale it mainly raises inflation, and the small fall in unemployment does not last.

What should you have seen? Northland and Eastport both start at 7%. In Northland, raising spending works best, because half of the unemployment is cyclical. In Eastport, the same policy barely moves unemployment and pushes inflation up, because the unemployment is structural. In Southvale, there is nothing for spending to fix. The model leaves out how long policies take, what they cost, and how a long recession can turn cyclical unemployment into structural. But the main lesson holds: the same number can need opposite treatments.

Real case: keeping jobs alive in the United Kingdom, 2020

The situation. In March 2020, the UK government closed large parts of the economy to slow the spread of COVID-19. Restaurants, shops and airlines lost most of their customers almost overnight. This was a sudden, huge fall in spending.

The policy. The government launched the Coronavirus Job Retention Scheme, usually called furlough. Firms could keep workers on their payroll without work, and the government paid up to 80% of their wages. The scheme ran until 30 September 2021.

The prediction. The reasoning was that the fall in spending was cyclical and temporary. The jobs would be needed again when the economy reopened. Keeping workers attached to their jobs would avoid a long, slow search to rebuild those matches later. Without such help, forecasts were very gloomy. In July 2020, the UK's official forecaster, the Office for Budget Responsibility, published a central scenario in which unemployment rose to about 12%.

What happened. Over its life, the scheme supported 11.7 million jobs at 1.3 million employers, at a cost of about £70 billion (HM Revenue and Customs, final evaluation, 2023). The unemployment rate rose, but far less than feared. The Office for National Statistics' first estimate of its highest point was 5.1%, for October to December 2020.

Verdict. The policy fitted the type of unemployment, and this is the main reason it worked. Part of the gap between the forecast and the outcome has other causes: the scheme was extended beyond its first planned end, and forecasts made in a crisis are very uncertain. The scheme was also expensive. And it carried a risk that economists discussed at the time. Where a job would never come back, because the shock had made it structural, paying to keep it alive only delayed the move to new work. A policy for cyclical unemployment helps less when the problem is really structural.

Worked examples

Worked Example 1 — Splitting an unemployment rate by type

An invented country has a labour force of 40 million. 2.8 million people are unemployed. Economists estimate that 0.8 million are frictionally unemployed, 1.0 million are structurally unemployed, and the rest are cyclically unemployed. Find the unemployment rate, the natural rate and the cyclical rate.

Unemployment rate: $\frac{2.8}{40} \times 100 = 7%$ .

Cyclical unemployment: $2.8 - 0.8 - 1.0 = 1.0$ million, which is $\frac{1.0}{40} \times 100 = 2.5%$ .

Natural rate: frictional plus structural, $\frac{0.8 + 1.0}{40} \times 100 = 4.5%$ .

Check: $4.5% + 2.5% = 7%$ . ✓ Meaning: raising spending could, at best, bring unemployment down to about 4.5%. Getting below that needs policies for skills and job matching.

Worked Example 2 — Shortening the search

In the same country, 0.3% of the labour force starts looking for a new job each week, and the average search lasts 10 weeks. A new national job website cuts the average search to 6 weeks. How does frictional unemployment change?

Before: $0.3% \times 10 = 3%$ of the labour force is searching at any moment, which is $0.03 \times 40 = 1.2$ million people.

After: $0.3% \times 6 = 1.8%$ , which is $0.018 \times 40 = 0.72$ million people.

So frictional unemployment falls by 1.2 percentage points, or 480 000 people, even though exactly the same number of people change jobs. Check: the search is 40% shorter, and $3% \times 0.6 = 1.8%$ . ✓

Three classic mistakes

Trap 1 — One cure for all unemployment

"The government should cut interest rates to reduce unemployment" is only right if the unemployment is cyclical. Good answers start with "It depends on the type." Raising spending cannot give a welder the skills of a care worker, and it may raise inflation instead.

Trap 2 — Full employment means zero unemployment

At potential output, cyclical unemployment is zero, but frictional and structural unemployment remain. "Full employment" means unemployment at the natural rate, not at 0%. Trying to push it to zero with spending raises inflation.

Trap 3 — Mixing up frictional and structural

Both involve workers looking for jobs. The difference is whether the worker's skills and location fit the jobs that exist. If they fit and only time is needed, it is frictional. If they do not fit, it is structural, however hard the person searches.

Try it yourself

Work each one on paper first. Open a hint only when you are stuck.

An invented country has 25 million people with jobs and 1.5 million unemployed people. Of the unemployed, 0.5 million are frictional, 0.4 million are structural and the rest are cyclical. Find the unemployment rate and the natural rate.

Hint 1
The labour force includes the unemployed. It is not 25 million.

Hint 2
Labour force = 25 + 1.5 = 26.5 million. The natural rate uses frictional plus structural.

Full answer
Labour force = 26.5 million. Unemployment rate = $\frac{1.5}{26.5} \times 100 \approx 5.7%$ . Natural rate = $\frac{0.5 + 0.4}{26.5} \times 100 \approx 3.4%$ . Cyclical = 0.6 million, $\approx 2.3%$ . Check: $3.4 + 2.3 = 5.7$ . ✓ A common error is dividing by 25 million, which gives 6%.
Each month, 1% of a country's labour force starts searching for a job, and the average search lasts 3 months. Find the frictional unemployment rate. A careers service cuts the average search by one month. Find the new rate.

Hint 1
Frictional unemployment = share starting each month × months each search lasts.

Hint 2
Before: $1% \times 3$ . After: $1% \times 2$ .

Full answer
Before: $1% \times 3 = 3%$ . After: $1% \times 2 = 2%$ . A one-month cut in search time removes a third of frictional unemployment. Check: $3% \times \frac{2}{3} = 2%$ . ✓ The number of people changing jobs has not changed. Only the time between jobs has.
An invented country has unemployment of 8%. In its capital city, firms cannot find software engineers and nurses. In its northern region, where three steelworks closed, unemployment is 15%. Inflation is already above its target. A minister proposes a large cut in interest rates. Explain what type of unemployment this mainly is, and evaluate the minister's plan.

Hint 1
Are there vacancies? Where are they, and what skills do they need?

Hint 2
Cutting interest rates raises total spending. What happens to spending when firms in the capital already cannot find workers?

Full answer
This is mainly structural unemployment: vacancies exist, but they need different skills and are in a different place from the unemployed steelworkers. Cutting interest rates raises total spending, which helps only cyclical unemployment. Firms in the capital would want even more workers that they cannot find, so wages there would rise faster, and inflation, already above target, would rise further. A better plan targets the mismatch: retraining for the jobs that exist, help to move, or investment that brings new firms to the north. These take years, and they cost money, so the evaluation should say that too.
Some economists argue that a long recession can raise the natural rate of unemployment. Explain how cyclical unemployment could turn into structural unemployment if it lasts for several years. What would this mean for policy at the start of a recession?

Hint 1
What happens to a worker's skills, habits and contacts after two or three years without a job? How might employers see a long gap in a worker's history?

Hint 2
If these workers no longer fit the jobs when spending recovers, which part of unemployment have they joined?

Full answer
A worker who is unemployed for years can lose skills, especially in jobs where methods change quickly. They lose contacts and the habit of work, and some give up searching. Employers may treat a long gap as a warning sign and choose other applicants. When spending recovers, these workers no longer fit the jobs, so they are now structurally unemployed. The natural rate has risen. Economists call this effect hysteresis, a word for a change that stays after its cause has gone. For policy, it means that acting quickly against cyclical unemployment matters. Unemployment that could have been cured with spending in the first year may need retraining a few years later. It is also part of the reasoning behind keeping job matches alive, as the UK's furlough scheme tried to do.

The bigger idea

A single number can hide very different problems. Two economies with 7% unemployment can need opposite treatments, just as two patients with the same temperature can have different illnesses. Good economics, like good medicine, diagnoses before it prescribes.

The types also connect the goals of policy. Pushing unemployment below the natural rate raises inflation, so low unemployment can conflict with stable prices, a conflict the policy lectures return to. First, the next lecture asks a more basic question: how do we count the unemployed at all?

Key concepts in this lesson

Frictional unemployment is the time people spend between jobs while they search for a good match.
Structural unemployment happens when workers' skills or locations do not match the jobs that exist, often after changes in technology, trade or demand.
Cyclical unemployment is caused by too little total spending, so it rises in recessions and falls in recoveries.
The natural rate of unemployment is frictional plus structural unemployment, the rate that remains when cyclical unemployment is zero.
Pushing unemployment below the natural rate with extra spending raises inflation, which is why it is also called the NAIRU.
The right policy depends on the type of unemployment, so diagnosis must come before the cure.