Two headlines, same day: "Unemployment falls to 4.1%" and "Millions have stopped looking for work." Both true. The unemployment rate is a fraction, and people who give up job-hunting silently exit the denominator — the rate can improve while the labor market worsens. Governments rise and fall on this number, and the AP exam demands you know exactly how it's built, what the three species of joblessness are, and why an economy "at full employment" still has millions of people between jobs. The definitions here are precise, testable, and — once you've got them — free points.
The adult (working-age, 16+, non-institutionalized) population splits into:
Labor force = employed + unemployed
Unemployment rate = (unemployed ÷ labor force) × 100
Labor force participation rate (LFPR) = (labor force ÷ adult population) × 100
Apply It: Adult population 200M; employed 120M; unemployed 10M. UE rate? LFPR? → Labor force = 130M. UE = 10/130 ≈ 7.7%. LFPR = 130/200 = 65%.
The discouraged-worker distortion: when a discouraged worker quits searching, the unemployed count and the labor force both fall — the unemployment rate can drop on bad news. Conversely, a recovering economy can pull discouraged workers back into the search, raising the measured rate on good news. Also hidden: underemployment (part-timers wanting full-time; PhDs driving rideshares count as fully employed).
| Type | Cause | Example | Cure |
|---|---|---|---|
| Frictional | Normal search time between jobs; new entrants | New graduate interviewing; coder who quit to find better fit | Better job-matching (can't/shouldn't hit zero) |
| Structural | Skills mismatch; technology or geography shifts demand for labor permanently | Coal miner displaced by automation; skills obsolete | Retraining, relocation |
| Cyclical | Recession — deficient overall demand | Hotel staff laid off in a downturn | Recovery; expansionary policy (Lessons 8, 10) |
(Seasonal unemployment — lifeguards in January — exists but is statistically smoothed away and rarely tested.)
Natural rate of unemployment (NRU) = frictional + structural
Full employment means unemployment equals the NRU — cyclical unemployment is zero. People are still between jobs (frictional) and still mismatched (structural); that's a healthy, churning economy, not a failure. The U.S. NRU is conventionally around 5% (use the number a problem gives you).
When the economy is at full employment, it's producing potential output — on the PPC (Lesson 1), at the trend line (Lesson 2), and, soon, where LRAS stands (Lesson 7):
Apply It: Frictional = 2%, structural = 3%, actual unemployment = 9%. Cyclical? → NRU = 5%; cyclical = 9 − 5 = 4% → the economy is in a recessionary gap.
Cyclical unemployment is the policy target: fiscal and monetary stimulus exist to close it. Frictional and structural unemployment don't respond to demand stimulus — pumping up spending won't reteach a displaced miner to code. Matching the type of unemployment to the right response is a classic FRQ move.
[GRAPH: Unemployment and the Business Cycle
X-axis: Time (years)
Y-axis: Real GDP
Curve 1: actual real GDP — waves around trend
Curve 2: potential output — smooth upward trend
Annotation: where actual GDP dips below potential, cyclical unemployment > 0
and actual UE rate > NRU; where actual GDP exceeds potential, UE rate < NRU
Reading: cyclical unemployment is the labor-market shadow of the output gap]
AP labeling requirements: this is the Lesson 2 business-cycle diagram with the unemployment interpretation layered on. On FRQs you'll more often state the relationship in words: "actual output below potential implies unemployment above the natural rate." Practice writing that sentence both directions — output language ↔ unemployment language. The AD-AS graph (Lesson 7) will let you draw it.
1. D. Unemployed = jobless + actively searching (past four weeks). A: no duration threshold exists for the basic count. B: benefits are irrelevant to the classification. C: willingness at a wage isn't the criterion — active search is. E: new labor-market entrants can be unemployed too.
Fix: The unemployment test is two boxes: no job ✓, actively looking ✓.
2. A. No search in the past four weeks → out of the labor force (a discouraged worker). B: fails the active-search test. C: clearly not working. D/E: both types presume labor-force membership, which Rios lost.
Fix: Stopped searching = stopped being counted — discouraged workers vanish from the rate.
3. E. LF = 150M; LFPR = 150/250 = 60%. A: uses employed only (140/250). B: that's the unemployment rate. C: unemployed over population. D: inverted subgroup arithmetic.
Fix: LFPR = (employed + unemployed) ÷ adult population — build the labor force first.
4. B. 10/150 ≈ 6.7%. A: divides by the population (10/250). C: divides by employed only (10/140). D/E: rounding/denominator slips.
Fix: The unemployment rate's denominator is the labor force — never the population, never just the employed.
5. C. Technology permanently eliminated the job and her skills no longer match demand: structural. A: frictional is temporary search, with skills still marketable. B: no downturn caused this. D: nothing seasonal. E: she didn't choose it.
Fix: "Technology/permanent shift/skills mismatch" → structural, every time.
6. A. Full employment = zero cyclical; frictional + structural (the natural rate) remain. B: measurement quirks aren't the reason. C: contradicts the definition. D: LFPR is a different statistic entirely. E: part-timers count as employed.
Fix: Full employment = UE at the natural rate, not at zero.
7. E. NRU = 3 + 2 = 5% = actual → cyclical = 0 → full employment, output at potential. A/D: would require actual > 5%. B: would require actual < 5%. C: confuses actual with cyclical.
Fix: Compare actual UE to (frictional + structural); equality means the economy is exactly at potential.
8. D. Deficient demand in a recession = cyclical. A: no permanent industry decline is described. B: the cause is the downturn, not normal search. C: no seasonal pattern. E: "natural" covers frictional + structural — not layoffs from recession.
Fix: Ask what caused the job loss; recession-caused = cyclical by definition.
9. B. They re-enter the labor force (LFPR ↑) as unemployed (UE rate ↑) — statistics worsen on economic good news. A/D: falls are backwards. C: both counts change. E: LFPR rises, not falls.
Fix: Re-entry inflates both the numerator and the labor force — expect the rate to tick up early in recoveries.
10. C. Structural unemployment is a skills/location mismatch — retraining attacks the mismatch itself. A/B: demand stimulus cures cyclical unemployment; it can't update anyone's skills. D: benefits cushion income but don't fix mismatch (and may slow search). E: freezes protect insiders without helping the displaced.
Fix: Match the cure to the cause — retraining for structural, demand policy for cyclical.
11. D. Actual below natural → cyclical "negative" in the arithmetic sense → economy beyond potential, which typically brings inflationary pressure. A: cyclical is actual − natural = −1.5, not +1.5. B: that's actual above natural. C: structural can't be negative — the gap is what's negative. E: the NRU doesn't move to match the data.
Fix: Actual < natural = overheating (inflationary gap); actual > natural = recessionary gap.
12. E. UE above NRU = cyclical unemployment = output below potential = actual GDP under the trend line. A: outside the PPC is unattainable, not recessionary. B: above trend = below-natural unemployment. C: a vertical line is LRAS, not a state of the economy. D: a labor supply curve alone says nothing about gaps.
Fix: Below trend ↔ above-natural unemployment; the output gap and the unemployment gap are the same fact in two languages.
13. B. If unemployed workers exit the labor force, the numerator falls with employment unchanged — the rate improves cosmetically. A: "negative cyclical" describes overheating, which would show hiring. C: population isn't in the formula's numerator or denominator (labor force is). D: part-time/full-time shifts don't change either count. E: type conversion doesn't change the total.
Fix: When the rate improves, check the labor force — shrinkage can fake a recovery.
The republic of Anders has an adult population of 100 million: 58 million employed and 6 million unemployed. Anders' natural rate of unemployment is 5%.
(a) Calculate Anders' labor force and unemployment rate. Show your work.
(b) Calculate Anders' labor force participation rate. Show your work.
(c) Is Anders producing at, above, or below its potential output? Explain using the relationship between the actual and natural unemployment rates.
(d) Two million discouraged workers in Anders begin actively searching for jobs but are not yet hired. Calculate the new unemployment rate.
1. D. Unemployed = jobless + actively searching (past four weeks). A: no duration threshold exists for the basic count. B: benefits are irrelevant to the classification. C: willingness at a wage isn't the criterion — active search is. E: new labor-market entrants can be unemployed too. Fix: The unemployment test is two boxes: no job ✓, actively looking ✓.
2. A. No search in the past four weeks → out of the labor force (a discouraged worker). B: fails the active-search test. C: clearly not working. D/E: both types presume labor-force membership, which Rios lost. Fix: Stopped searching = stopped being counted — discouraged workers vanish from the rate.
3. E. LF = 150M; LFPR = 150/250 = 60%. A: uses employed only (140/250). B: that's the unemployment rate. C: unemployed over population. D: inverted subgroup arithmetic. Fix: LFPR = (employed + unemployed) ÷ adult population — build the labor force first.
4. B. 10/150 ≈ 6.7%. A: divides by the population (10/250). C: divides by employed only (10/140). D/E: rounding/denominator slips. Fix: The unemployment rate's denominator is the labor force — never the population, never just the employed.
5. C. Technology permanently eliminated the job and her skills no longer match demand: structural. A: frictional is temporary search, with skills still marketable. B: no downturn caused this. D: nothing seasonal. E: she didn't choose it. Fix: "Technology/permanent shift/skills mismatch" → structural, every time.
6. A. Full employment = zero cyclical; frictional + structural (the natural rate) remain. B: measurement quirks aren't the reason. C: contradicts the definition. D: LFPR is a different statistic entirely. E: part-timers count as employed. Fix: Full employment = UE at the natural rate, not at zero.
7. E. NRU = 3 + 2 = 5% = actual → cyclical = 0 → full employment, output at potential. A/D: would require actual > 5%. B: would require actual < 5%. C: confuses actual with cyclical. Fix: Compare actual UE to (frictional + structural); equality means the economy is exactly at potential.
8. D. Deficient demand in a recession = cyclical. A: no permanent industry decline is described. B: the cause is the downturn, not normal search. C: no seasonal pattern. E: "natural" covers frictional + structural — not layoffs from recession. Fix: Ask what caused the job loss; recession-caused = cyclical by definition.
9. B. They re-enter the labor force (LFPR ↑) as unemployed (UE rate ↑) — statistics worsen on economic good news. A/D: falls are backwards. C: both counts change. E: LFPR rises, not falls. Fix: Re-entry inflates both the numerator and the labor force — expect the rate to tick up early in recoveries.
10. C. Structural unemployment is a skills/location mismatch — retraining attacks the mismatch itself. A/B: demand stimulus cures cyclical unemployment; it can't update anyone's skills. D: benefits cushion income but don't fix mismatch (and may slow search). E: freezes protect insiders without helping the displaced. Fix: Match the cure to the cause — retraining for structural, demand policy for cyclical.
11. D. Actual below natural → cyclical "negative" in the arithmetic sense → economy beyond potential, which typically brings inflationary pressure. A: cyclical is actual − natural = −1.5, not +1.5. B: that's actual above natural. C: structural can't be negative — the gap is what's negative. E: the NRU doesn't move to match the data. Fix: Actual < natural = overheating (inflationary gap); actual > natural = recessionary gap.
12. E. UE above NRU = cyclical unemployment = output below potential = actual GDP under the trend line. A: outside the PPC is unattainable, not recessionary. B: above trend = below-natural unemployment. C: a vertical line is LRAS, not a state of the economy. D: a labor supply curve alone says nothing about gaps. Fix: Below trend ↔ above-natural unemployment; the output gap and the unemployment gap are the same fact in two languages.
13. B. If unemployed workers exit the labor force, the numerator falls with employment unchanged — the rate improves cosmetically. A: "negative cyclical" describes overheating, which would show hiring. C: population isn't in the formula's numerator or denominator (labor force is). D: part-time/full-time shifts don't change either count. E: type conversion doesn't change the total. Fix: When the rate improves, check the labor force — shrinkage can fake a recovery.
Exam tip: Every unemployment MC is one of three moves: classify a person (employed/unemployed/not in LF), classify a cause (frictional/structural/cyclical), or run the ratios. Do the labor-force arithmetic in a fixed order — LF first, then rate — and the calculation traps disarm themselves. Next lesson: the other headline statistic, inflation, and the index that measures it.