Separate recovery, prior peaks and potential output
Use dated GDP, potential-output and payroll records to test three different recovery claims.
Your task: Reproduce a quarterly payroll mean, compare output with its prior level and estimated potential, and qualify the interpretation using source vintages.
A growth rate compares periods; a level describes one period. A percentage gap needs an explicit benchmark. Revised historical data are not the information originally available.
Output gap
Actual output minus estimated potential output, often expressed as a percentage of potential output.
Trough
A turning point from contraction toward expansion; its identification depends on the indicators and time frequency used.
Data vintage
The version of an estimate available at a specified release or retrieval date.
01 · TRY
Can GDP recover its earlier level while the output gap and payroll shortfall remain?
Audit all three series definitions and update dates. Recreate one three-month payroll average, identify the lowest GDP quarter in the packet, and compare 2021 Q1 with both 2019 Q4 output and estimated potential. Report the effect of one hypothetical potential revision.
Use the supplied table to calculate one payroll mean and distinguish rising output from a return to the earlier level. Explain the two denominators in words.
Make a provisional claim
Loading saved evidence…
Inspect evidence and test an explanation
The investigation controls appear when your local evidence is ready.
04 · UNDERSTAND
Connect evidence and mechanism
A recovery can pass one benchmark and miss another
A newspaper sees real GDP above its pre-contraction level and declares that every part of the economy has recovered. Before accepting that claim, separate four questions: Is output growing? Has it regained the earlier level? Is it at estimated potential? Have payroll jobs regained their earlier level? Each question uses different evidence.
Inspect the records and their dates
This packet stores seven U.S. quarters, 2019 Q4 through 2021 Q2. The source tables were retrieved on 25 October 2026. The GDP and employment observations are revised historical estimates; potential output is an estimated path. These are not forecasts made by this course.
Monthly, seasonally adjusted, thousands; updated 2 October 2026. This lesson computes quarterly means.
FRED quarter-start dates such as 2020-04-01 label 2020 Q2, not a one-day observation. For payroll, use April, May and June together to form that quarter’s mean. Keep release/update dates separate from observation periods.
Potential is a modeled sustainable-output benchmark, not another observed GDP series. The CBO potential path has no ordinary seasonal fluctuations to remove; retain its source adjustment label instead of silently relabeling it seasonally adjusted. The output levels use the same chained-dollar basis and annual output scale. We compare the stored levels directly and do not divide only one series by four. Changes in these annual-rate levels give quarter-to-quarter growth; annualizing that growth is a separate operation.
Scope limit: update dates differ across the three series. This is an explicitly mixed-vintage retrospective comparison, not an official CBO output-gap release and not an estimate of what a policymaker could have known in any 2020 month. A real-time analysis requires the vintages actually available then. Never use later information without labeling it.
Predict, then compare
Start with 2020 Q2 and predict what the next quarter must do for each of the four recovery claims to become true. Select 2020 Q3, 2021 Q1 and 2021 Q2 in turn. Write a claim using the displayed levels and denominators, not just the direction of a line.
Compare recovery with the output gap
Local snapshot retrieved 25 October 2026. BEA output, CBO estimated potential and BLS payroll data have different update dates. These are revised historical records, not a reconstruction of what was known in 2020. No live download is needed.
Predict whether recovery from the lowest output quarter must restore the prior output level, close the output gap and restore payroll employment at the same time. Test 2020 Q3, 2021 Q1 and 2021 Q2.
Solid line and circles: real GDP. Dashed line and squares: potential output. Dotted line and triangles: payroll. GDP and potential share the 2019 Q4 actual GDP denominator; potential is not forced to 100. Payroll uses its own 2019 Q4 quarterly mean as 100. Equal indices do not mean equal dollars and jobs. The fixed 80–110 scale highlights change and does not start at zero.
2020 Q2: real GDP index 90.91, payroll index 88.18, output gap -8.83%. Gap uses the published potential estimate.
Output gap = 100 × (actual − potential) ÷ potential = 100 × (19077.992 − 20925.856) ÷ 20925.856 = -8.83%. GDP growth from the previous quarter: -7.88% (not annualized).
Dollar levels: billions of chained 2017 dollars. Payroll: quarterly mean in thousands. Display rounded; calculations use stored precision.
Quarter
Actual GDP
Potential used
Payroll mean
GDP index
Payroll index
Gap %
2019 Q4
20985.448
20722.820
151640.333
100.00
100.00
1.27
2020 Q1
20709.212
20825.890
151739.667
98.68
100.07
-0.56
2020 Q2
19077.992
20925.856
133712.333
90.91
88.18
-8.83
2020 Q3
20558.879
21020.539
140615.333
97.97
92.73
-2.20
2020 Q4
20791.917
21116.617
142580.333
99.08
94.03
-1.54
2021 Q1
21106.619
21216.865
143491.667
100.58
94.63
-0.52
2021 Q2
21474.077
21322.289
145157.333
102.33
95.72
0.71
Inspect the three monthly payroll inputs per quarter
Take the arithmetic mean of the three seasonally adjusted monthly observations. Adding them would count repeated payroll positions across months; the mean is still not a count of unique people.
2019 Q4, months in chronological order: 151460, 151667, 151794 thousand; mean = 151640.33 thousand.
2020 Q1, months in chronological order: 152031, 152293, 150895 thousand; mean = 151739.67 thousand.
2020 Q2, months in chronological order: 130426, 133040, 137671 thousand; mean = 133712.33 thousand.
2020 Q3, months in chronological order: 139255, 140821, 141770 thousand; mean = 140615.33 thousand.
2020 Q4, months in chronological order: 142460, 142733, 142548 thousand; mean = 142580.33 thousand.
2021 Q1, months in chronological order: 142863, 143380, 144232 thousand; mean = 143491.67 thousand.
2021 Q2, months in chronological order: 144587, 145065, 145820 thousand; mean = 145157.33 thousand.
Try both potential adjustments with actual output unchanged. Explain why an estimated gap can change without any change in measured production. The ±1% choices are invented sensitivity checks, not a confidence interval or a CBO revision. Potential is sustainable capacity in a model, not a physical ceiling that actual output can never exceed.
Record quarter, data vintages, potential assumption and reasoning in the source notebook. Controls reset on reload and are not saved in captures. This comparison does not identify a policy’s causal effect or award independent mastery.
Reconstruct one result
The monthly payroll inputs for 2020 Q2 are 130,426, 133,040 and 137,671 thousand. Their mean is 401,137 ÷ 3 = 133,712.333 thousand. Do not present 401,137 thousand as that quarter’s employment. These are repeated monthly payroll positions, not three disjoint groups of workers. The payroll series is also different from the household-survey employment measure in the labor-status lesson.
For 2021 Q1, output 21,106.619 exceeds the 2019 Q4 output 20,985.448, but falls short of the stored potential 21,216.864736890268. The gap is approximately −0.52%. The corresponding payroll mean remains below its 2019 Q4 mean. Restoring one level therefore does not establish all four claims.
A positive gap means actual output exceeds the selected estimated benchmark. It does not mean every person benefits, every sector is above capacity, or production has crossed an impossible physical limit. A negative gap can coexist with positive growth while recovery continues.
Date a turning point with the right evidence
In this packet, GDP falls through 2020 Q2 and rises in Q3. Q2 is the lowest quarterly output observation in this limited window. A plotted minimum alone does not establish a complete national business-cycle chronology or explain the downturn’s causes.
The NBER announced the April 2020 monthly trough on 19 July 2021, following the February 2020 monthly peak. Its dating announcement and chronology distinguish monthly and quarterly dating. A recession ending means the contraction has ended; it does not mean output or employment has regained every earlier benchmark. Quarterly averages cannot locate a turning month by themselves. Two quarters of falling GDP are not a sufficient universal dating procedure.
Your evidence brief: identify the packet’s output minimum, compare the next quarter’s growth with its remaining gap, explain the differing payroll path, and keep the NBER event date separate from its later announcement date. Name additional indicators you would inspect before making a causal explanation. The pandemic period involves more than one possible demand and supply mechanism; this chart cannot isolate their effects.
Apply the reasoning to an unfamiliar economy
A separate fictional economy reports output 100, then 96, then 102. Its potential in the final period is 106, and its employment index is 98 against an initial 100. Draw the three output points and mark the final potential. Calculate final output growth, the final gap, and the change relative to the first period. Explain which recovery claims hold before opening feedback.
Check the unfamiliar recovery claim
Final-period growth is 100×(102−96)/96 = 6.25%. Relative to the first period, output is 2% higher. The final output gap is 100×(102−106)/106 ≈ −3.77%. Employment is 2% below its initial index. Output has grown and surpassed its starting level, while remaining below estimated potential and with employment still below its base.
If only the estimated final potential changes to 101, the gap becomes about +0.99%. None of the actual-output or employment facts changes. That is sensitivity to a benchmark, not evidence that a policy suddenly raised production.
Use the table for guided high-school work. AP practice should reconstruct the labeled comparison and explain its mechanism without treating a graph’s direction as the whole answer. Introductory-college work should retrieve an earlier vintage, document its availability date and reproduce the comparison with consistent definitions. The open workbench and answer are practice; a reviewer must still assess an unfamiliar independent brief.
Worked example
For 2021 Q1, real GDP 21,106.619 exceeds 2019 Q4 GDP 20,985.448 but remains below estimated potential 21,216.864736890268: the calculated gap is approximately −0.52%. Its payroll quarterly mean remains below the 2019 Q4 mean.
