Unemployment is the one macroeconomic problem that arrives with a name and an address. Inflation is a rate, growth is a percentage, the current account is a balance. Unemployment is a person who wants to work and cannot, and the damage compounds: skills decay, confidence goes, and the longer it lasts the harder it becomes to reverse. That is why economists treat it as more than a number, and why the type of unemployment matters more than the level.
Unemployment
People of working age who are without work, are available for work and are actively seeking work.
Employment
The use of labour in producing goods and services: people of working age who are in paid work, full time or part time, or self-employed.
Full employment
A situation where everyone of working age who wants a job at the current wage can find one, so the only unemployment is short-term frictional or seasonal unemployment.
All three conditions matter and dropping one loses the mark. Someone who has retired is not unemployed, because they are not seeking work. Nor is a full-time student, nor someone who has given up looking. 'People without a job' is not the definition and is marked as incomplete.
There are two common ways of measuring it and they rarely agree. The claimant count records those receiving unemployment-related benefits. It is cheap and quick, because the government already holds the data, but it misses anyone who is genuinely seeking work yet ineligible to claim, and it changes whenever the eligibility rules change rather than when the economy does. The labour force survey asks a large sample whether they are working and whether they are looking, following an internationally agreed definition. It is more accurate and internationally comparable, but slower and more expensive. The survey figure is normally the higher of the two.
Percentage points against percentages
A country's unemployment rate rises from 5% to 6% over a year. A newspaper reports this as 'a 1% rise in unemployment'. Calculate the change correctly in both forms and explain why the newspaper's phrasing is misleading.
The types
Frictional unemployment is the short-term unemployment of people moving between jobs. It exists even in a healthy economy and is not, in moderation, a problem. Structural unemployment arises when the skills or the location of workers no longer match the jobs available, typically because an industry has declined. It is long-lasting and the hardest to cure. Cyclical unemployment, sometimes called demand-deficient unemployment, occurs when total demand in the economy falls and firms need fewer workers; it rises in recession and falls in recovery. Seasonal unemployment follows predictable patterns in the year, affecting agriculture, construction and tourism.
Technological unemployment is a form of structural unemployment caused by new technology replacing workers, as when self-service checkouts reduce the number of cashiers a supermarket needs or automated machines replace workers on a production line. New technology usually creates jobs elsewhere, in designing, making and maintaining the machines, but the workers who lose their jobs may not have the skills to do those new jobs.
Questions usually give you a scenario and expect the right type named. A coal miner unemployed after the pit closes is structural, not cyclical, even though it happened during a downturn. A hotel worker without shifts in January is seasonal. A graduate spending two months choosing between offers is frictional. Naming the type and then giving the reason from the scenario is what secures both marks.
The consequences run in three directions at once. For the individual there is lost income, and beyond that a loss of skills and often of health and self-esteem. For the government there is a double hit: benefit payments rise while income tax and spending tax receipts fall, so the budget position worsens from both sides simultaneously. For the economy as a whole, output that could have been produced is simply lost forever, and unemployed workers spend less, which reduces demand for other firms and can deepen the downturn that caused the unemployment.
Firms feel unemployment too. Falling demand means lower sales and profits, but a large pool of unemployed workers also means firms can recruit more easily and may pay lower wages. Over a long period, though, the skills of the unemployed decline, so firms find it harder to hire the trained workers they need when demand recovers.
Governments use several policies to reduce unemployment, and each suits a different type. Expansionary fiscal policy, cutting taxes or raising government spending, and expansionary monetary policy, cutting interest rates, raise total demand, so firms produce more and hire more workers: they work against cyclical unemployment. Supply-side policies tackle structural and frictional unemployment: education and retraining match workers' skills to the jobs available, better job centres and information help people find vacancies faster, and help with moving home or subsidies for firms that set up in areas of high unemployment tackle the geographical mismatch. Demand policies work quickly but may raise inflation; supply-side policies take years but tackle the cause.
Which policy is most likely to reduce structural unemployment?
Draw a production possibility curve and mark a point inside it. Label that point as an economy with cyclical unemployment. Show the movement that would occur if demand recovered, and separately show what would be needed to reduce structural unemployment once the economy is back on the curve.
This separates the two problems visually. Cyclical unemployment is a point inside the curve, so recovery is a movement towards it using capacity that already exists. Structural unemployment persists even on the curve, because the mismatch is between the skills people have and the jobs available, so curing it requires retraining rather than more demand.
A shipyard closes and its welders remain out of work for three years despite vacancies existing in software and healthcare. This is best described as:
Which consequence of higher unemployment affects the government budget from both directions at once?
Cyclical unemployment and the output gap
At AS, cyclical unemployment is analysed through aggregate demand and the output gap. The output gap is the difference between actual output and the economy's potential output. A negative gap means the economy is producing below capacity, which is the same condition as cyclical unemployment seen from the output side rather than the labour side. Because labour is a derived demand, wanted not for itself but for what it produces, a fall in demand for goods translates directly into a fall in demand for the workers who make them.
How a demand shock becomes unemployment
- Consumer and business confidence falls, so households postpone purchases and firms delay investment.
- Consumption and investment fall, so aggregate demand shifts left.
- Firms find inventories building up as sales fall short of production.
- They cut output in response, so real GDP falls and a negative output gap opens.
- Labour is a derived demand, so producing less requires fewer workers and firms reduce hours, stop recruiting, then make redundancies.
- Cyclical unemployment rises.
- Unemployed households have lower incomes, so consumption falls further, and the multiplier works in reverse.
- Aggregate demand falls again, deepening the contraction unless policy intervenes.
The last two steps are what makes recessions self-reinforcing and are the analytical basis for the case for intervention. AO2 is 40% of the written papers, and the reverse multiplier is a step most candidates omit.
Practise this chain →Draw an AD/AS diagram showing an economy in equilibrium below full employment. Mark the full employment level of output, the current equilibrium, and the negative output gap between them. Then show the effect of expansionary policy closing the gap.
Label the gap explicitly rather than leaving it implied. The distance between the current equilibrium and full employment output is the cyclical unemployment, and showing it as a measurable horizontal distance is what turns a description into analysis. Note also that closing the gap raises the price level only slightly while AS is flat, which is the argument for acting early.
Policy must be matched to the type, and the commonest failure in exam answers is proposing a remedy that cannot work on the diagnosis given. Cyclical unemployment responds to demand-side policy, whether fiscal or monetary, because the binding constraint is a shortage of demand. Structural unemployment does not, because the jobs already exist; it requires retraining, relocation assistance and improved information. Frictional unemployment responds to better job matching and information, and to benefit design that does not make prolonged search attractive. Seasonal unemployment responds to diversification of the local economy. Naming the type before naming the remedy is a habit that pays across the whole topic.
An economy has a negative output gap and rising unemployment following a fall in consumer confidence. Which policy addresses the binding constraint most directly?
Applied context
The pandemic labour market of 2020 to 2022 produced an unusually clean natural experiment. Job retention and furlough schemes across many countries deliberately kept workers attached to their employers rather than letting them become unemployed, on the argument that preserving the match between worker and firm would allow a faster recovery. Measured unemployment consequently rose far less than the fall in output would historically have implied, and employment recovered quickly once restrictions lifted. The counter-argument is that the schemes also supported firms that were not viable and slowed the reallocation of labour to growing sectors, so some of the cost appeared later as reduced productivity rather than immediately as unemployment.
This example works well against questions on the consequences of unemployment, because it shows a government treating the loss of the worker-firm match as the central cost rather than the lost income alone.
The natural rate and hysteresis
At A Level the key concept is the natural rate of unemployment, the rate that persists when the labour market is in equilibrium and inflation is stable. It is composed of frictional and structural unemployment, so it is not zero and cannot be reduced by demand management. Attempts to push unemployment below it raise inflation without lasting employment gains, which is the vertical long-run Phillips curve. The natural rate is not fixed, though: it rises with skills mismatch, with barriers to labour mobility, with generous and poorly designed benefits, and with strong insider power in wage bargaining. Reducing it is the job of supply-side policy.
Two further distinctions sharpen the analysis. Voluntary unemployment is where people choose not to take jobs available at the current wage, perhaps because benefits are close to the wage on offer or they are searching for a better match. Involuntary unemployment is where people are willing to work at the current wage but cannot find a job. Equilibrium unemployment is the unemployment that exists when the labour market clears, with the demand for labour equal to the number willing to work at the going wage: it is frictional and structural, and it is the natural rate. Disequilibrium unemployment arises when the labour market does not clear, because demand for labour falls in a recession, or because the wage is held above the market-clearing level by a minimum wage or union bargaining, so more people want to work than firms wish to employ.
Real wages are held above the market-clearing level by union bargaining. The unemployment that results is best described as:
Hysteresis is the mechanism that makes a recession leave permanent damage, and it is the single most valuable idea in this topic. A worker unemployed for a long spell loses current skills, misses new technology and working practices, and may lose the habits and networks that job search depends on. Employers use duration of unemployment as a screening signal, so the long-term unemployed are passed over even when vacancies appear. Some withdraw from the labour force entirely. The consequence is that unemployment which began as cyclical becomes structural, so the natural rate itself rises and the economy's potential output falls. A temporary shock has produced a permanent loss, which is the strongest argument in economics for intervening early and hard in a downturn rather than waiting for recovery.
Output gap (%) = ((actual output - potential output) / potential output) x 100A negative figure indicates spare capacity and cyclical unemployment; a positive figure indicates an economy operating above its sustainable level, which is inflationary. Formulae are not printed in any 9708 paper, so recall it.
Measuring the cost of a recession, and why the estimate is unreliable
An economy has potential output of $500 billion and actual output of $480 billion. Calculate the output gap as a percentage. Then explain why the true cost may be larger than this figure suggests, and why the figure itself is uncertain.
The headline unemployment rate conceals as much as it reveals, and A Level answers are expected to say so. Underemployment covers those working part time who want full-time hours and those in jobs well below their skill level; neither appears in the unemployment figure although both represent wasted capacity. The participation rate matters because a falling unemployment rate accompanied by falling participation usually means discouraged workers leaving the labour force rather than finding jobs. Averages hide sharp differences by age, region and ethnicity, and youth unemployment is typically several times the headline rate. In many developing economies the informal sector absorbs workers who would otherwise be counted as unemployed, so the measured rate can be low while underemployment is severe. Judging labour market health from a single number is exactly the error data-response questions are designed to catch.
Applied context
The regional legacy of industrial decline in older manufacturing areas across Europe and North America is the standard illustration of hysteresis operating over decades rather than years. Communities that lost a dominant industry in the 1980s often still show elevated unemployment, lower participation and lower average earnings a generation later, despite national recoveries in between. The pattern suggests that the loss was not simply of jobs but of the accumulated skills, networks and employer presence that make a local labour market function, and that these do not rebuild automatically when aggregate demand returns.
Use this to argue that structural unemployment has a geographical dimension that national policy tends to miss, and that labour immobility is as much about housing costs and family ties as about willingness to move.
Unemployment connects outward in four directions. It sits opposite inflation in the Phillips relationship, which is why it appears in almost every policy conflict question. It is the labour market's expression of the output gap, so it is inseparable from the growth topic and from the sustainability of any expansion. Reducing the natural rate is the central claim made for supply-side policy, so the two topics are argued together. And the microeconomics of the labour market, wage determination, trade unions, monopsony and the minimum wage, explains why real-wage unemployment persists at the level of the individual market rather than the whole economy.
9708-4.6 · 9708-5.4 · 9708-8.3 · 9708-9.2 · 9708-10.2 · 9708-11.4
Explain the difference between structural, cyclical and seasonal unemployment and discuss whether a general rise in income in a country would have the same effect on each of these types. (13 marks)
- Explain the three typesDefine each precisely and distinguish by cause rather than by description. Cyclical is caused by deficient aggregate demand and varies with the business cycle. Structural is caused by a mismatch of skills or location following a change in the pattern of production, and persists regardless of the cycle. Seasonal is caused by predictable variation in demand or production across the year. Use a brief example of each; the marks here are for the distinction, so do not over-write this section.
- Analysis: the effect on cyclical unemploymentA general rise in income raises consumption and therefore aggregate demand. Firms increase output, and because labour is a derived demand they employ more workers. Cyclical unemployment falls, and the multiplier amplifies the effect. This is the type on which rising income works directly and strongly, and it should be argued with a chain, not asserted.
- Analysis: the effect on structural and seasonalStructural unemployment responds weakly at best. Higher demand does not give a redundant welder the skills for a software vacancy, and if the declining industry itself does not recover, the mismatch is untouched. Concede the partial effect: a stronger economy creates more vacancies overall and may draw some structurally unemployed workers into unrelated low-skilled work, and it makes employers likelier to train. Seasonal unemployment barely responds at all, since its cause is the calendar rather than the level of demand, though higher income may lengthen the tourist season or extend the construction year at the margin.
- EvaluationWeigh the differences rather than restating them. The effects are clearly not the same, and the reason is that rising income addresses a demand deficiency while two of the three types have non-demand causes. Qualify with time: over a long period sustained high demand may erode structural unemployment through employer-funded training and induced migration. Qualify with degree: hysteresis means cyclical unemployment left too long becomes structural, so the type is not fixed and the answer depends on how long the downturn lasted. Note that the size of the effect on cyclical unemployment also depends on the output gap, since near capacity extra income produces inflation rather than jobs.
- JudgementConclude that the effects differ sharply and rank them: large on cyclical, weak and indirect on structural, minimal on seasonal. Then add the condition that makes the answer sophisticated, that the boundary between cyclical and structural is not fixed, so a rise in income arriving early prevents structural unemployment from forming while the same rise arriving after years of slump finds the damage already done. Timing changes the answer, and saying so is what reaches the top band.
A real question from June 2015, Paper 4. Two command words, explain then discuss, so allocate roughly a third of the time to the explanation and two thirds to the discussion. Answers that spend half the time defining the three types run out of room where the marks are.
How costly is unemployment, and to whom?
- The lost output is real and permanent, since idle capacity in one year cannot be recovered in the next. Against that, measured output gaps rest on estimates of potential output that cannot be observed, so the size of the loss is genuinely uncertain rather than merely disputed.
- The costs fall very unevenly. National figures average across a labour market in which youth, regional and low-skill unemployment can run at several times the headline rate, so a modest national number can coexist with severe local distress. Policy judged on the average will under-respond to the concentration.
- The individual costs extend well beyond lost income, into skills, health and long-term earnings, and there is evidence these persist long after re-employment. This asymmetry, that the damage outlasts the episode, is the strongest argument for treating unemployment as more urgent than an equivalent-sized inflation problem.
- Hysteresis converts a temporary cost into a permanent one by raising the natural rate, which means the cost of delay is not linear. Acting late is disproportionately worse than acting slightly late, and this is what justifies early intervention even at the risk of overshooting.
- Some unemployment is unavoidable and even useful. Frictional unemployment reflects workers searching for a better match, and an economy with none would be one where nobody ever changed jobs, which would itself reduce productivity. The objective is the natural rate, not zero, and answers that treat any unemployment as failure have misunderstood the target.
- The strongest answers weigh the costs of unemployment against the costs of the policies used to reduce it, and note that the trade-off differs by type: cyclical unemployment can be reduced quickly at some inflation risk, structural only slowly and at fiscal cost, so a single judgement about 'the cost of unemployment' is less useful than one made type by type.
The examiner's dividing line on this topic is whether a candidate treats unemployment as one thing or four. Almost every high-band answer classifies by type in the opening paragraph and then keeps the classification alive throughout, because policy effectiveness, income responsiveness and duration all differ by type. Almost every mid-band answer describes unemployment in general and proposes remedies that would work for only one kind.