Separate growth in output from growth per person

Separate growth in output from growth per person

Question and evidence

Can total production rise while output per resident falls?

A fictional economy produces 1000 units with 100 residents and 50 workers in year 1. In year 2 it produces 1100 units with 120 residents and 55 workers. Work hours per worker are unchanged.

This packet uses fictional data to isolate a mechanism. It is not an empirical estimate.

Mechanism and worked example

Total output, output per person and output per worker use different denominators. Growth in productive capacity can shift a production possibilities frontier or long-run aggregate supply. Diminishing marginal returns to one input hold other conditions fixed; they do not imply innovation is impossible. GDP per person is still an incomplete wellbeing measure.

Output grows 10%. Output per resident falls from 10 to 1100/120≈9.17. Output per worker stays 20. In the technology scenario it rises to 1210/55=22, a 10% productivity increase.

Complication

A technology scenario raises output to 1210 with the same 55 workers, but requires training and equipment. Physical capital, human capital, institutions and environmental constraints can affect feasible production; more inputs alone do not prove technological improvement.

Your investigation

Construct three indexed series with year 1=100, label every denominator, and sketch how a capacity improvement differs from movement toward an existing frontier.

Revise your answer to “Can total production rise while output per resident falls?” using both source IDs. Separate the checked calculation from a judgment, and identify evidence that could reverse your judgment.

Vocabulary

Productivity: Output per unit of input, with the input and time period stated.

Assumption: A condition accepted for a model or argument; it needs justification before applying the result elsewhere.

Independent extension

Independently extend the investigation “Separate growth in output from growth per person” with a documented real dataset or scholarly source. Preserve definitions, recreate one calculation or evidence table, compare a rival explanation and report why the teaching example cannot establish the real-world conclusion.

Investment, replacement and production

Replace worn capital before counting new capacity

Fictional economy with 50 workers, 100 residents and constant hours and skills. Opening capital is 4 equipment units per worker, or 200 total. Investment is installed at year end; the table reports the annual output capacity supported by that end-of-year stock, not actual production during the installation year.

Annual gross investment per worker0 equipment units0.4 equipment units0.8 equipment units Annual depreciation fraction10% of opening capital20% of opening capital Technology factor11.2

End capital = opening capital − retired capital + gross investment. Net investment is gross investment minus retired capital. Annual output per worker = 10 × technology × square root of capital per worker. A square root is the nonnegative number that squares to the input: √4 = 2. The coefficient 10 carries this teaching model's unit conversion; it is not an empirical estimate.

The selected technology applies from the starting observation onward. Changing it compares two economies with different productivity, not a measured technological transition. Saving, investment financing, costs, worker skills and population dynamics are outside this model.

Solid curve: technology 1. Dashed curve: selected technology. Square: starting stock. Circle: year-five stock. Identical curves or points overlap. Capital moves along a fixed-technology curve; a technology change shifts the curve. Axes are fixed, and this is a production relationship, not a time-axis graph.

Year 5 capital per worker 4.00; annual output capacity per worker 20.00; per resident 10.00. Stationary capital per worker under these fixed inputs: 4.00.

At the stationary stock, depreciation exactly equals gross investment. This is a conditional accounting benchmark, not a forecast of when an economy reaches it or proof that investment is freely available.

Per-worker capital ledger and end-stock annual production capacity; unrounded calculations

Year	Opening capital	Retired	Gross investment	Net investment	End capital	Output / worker	Total output	Output / resident
1	4.00	0.40	0.40	0.00	4.00	20.00	1000.00	10.00
2	4.00	0.40	0.40	0.00	4.00	20.00	1000.00	10.00
3	4.00	0.40	0.40	0.00	4.00	20.00	1000.00	10.00
4	4.00	0.40	0.40	0.00	4.00	20.00	1000.00	10.00
5	4.00	0.40	0.40	0.00	4.00	20.00	1000.00	10.00

Curve anchors (capital per worker, annual output per worker): at technology 1, (0,0), (1,10), (4,20), (9,30); at selected technology 1, (0,0), (1,10), (4,20), (9,30).

Restore capital comparison

Controls are temporary. Save your replacement calculation, graph comparison and one assumption that would change the interpretation in the notebook.

Gross investment can leave capacity unchanged

At the default settings, predict net investment before opening the ledger. Then raise gross investment per worker to 0.8 while keeping depreciation at 10%. Compare the first-year changes in total capital and output capacity. Finally set gross investment to zero and depreciation to 20%; explain why a positive opening stock does not preserve itself.

Check replacement and accumulation

Initially 10% of four capital units per worker is 0.4. Gross investment 0.4 exactly replaces retirement, so net investment is zero, capital stays four and annual output capacity stays twenty per worker. Across fifty workers, replacement is twenty capital units per year.

With investment 0.8, first-year end capital is 4−0.4+0.8=4.4 per worker, or 220 total. Annual capacity is 10√4.4≈20.98 per worker, about 1,048.81 total. Capital rises 10%, but capacity rises about 4.88% under the stipulated square-root relationship. With zero investment and 20% depreciation, year-five capital per worker is 4×0.8⁵=1.31072. These are production possibilities under the model, not realized sales or utilization.

More capital and better technology are different changes

Hold technology at one. Compare adding one capital unit per worker when starting at one versus starting at four. Use 10√k, where k is capital per worker. Then restore capital to four and raise technology to 1.2. Explain whether the higher productivity means a continuing 20% growth rate each year.

Check diminishing returns and the technology shift

Moving from one to two capital units raises capacity from 10 to about 14.14, a gain of 4.14. Moving from four to five raises it from 20 to about 22.36, a gain of 2.36. Equal extra capital produces a smaller extra output gain when other inputs and technology are fixed. Total output still rises: diminishing marginal returns are not negative returns.

At capital four, technology 1.2 gives output per worker 24 instead of 20. If capital and technology then stay fixed, output stays 24; the level is 20% higher, but subsequent growth is zero. Sustained improvements need a continuing mechanism. In the widget, fixed gross investment and depreciation lead toward a stationary stock equal to investment divided by depreciation; a higher stationary level is not perpetual growth at a fixed positive rate.

Hold capital fixed when changing employment

The equivalent total-production rule is Y=10A√(K×L), where Y is annual capacity, A technology, K total equipment and L employed workers. The widget fixes L at fifty. For this separate comparison, fix K=200 and A=1, then compare L=50 with L=200. Calculate total capacity and capacity per worker before describing the result.

Check employment and the denominator

With fifty workers, capacity is 10√10,000=1,000 and capacity per worker is twenty. With two hundred workers sharing the same capital, capacity is 10√40,000=2,000 and capacity per worker is ten. Total production doubles while employment quadruples. The fixed equipment is spread across more workers; this calculation does not show that workers became less skilled. If both capital and employment quadrupled, capacity would quadruple and capacity per worker would stay twenty under this particular rule.

Output per resident also depends on the employed share of the population. For example, fifty workers producing twenty each among one hundred residents give ten per resident. Sixty workers producing twenty each among the same population give twelve per resident. Holding productivity constant in this second arithmetic example is a separate assumption, not the fixed-capital case above. Hours, unemployment and participation need their own denominators before attributing a real change to productivity.

Connect financing with capital without erasing public investment

Use a separate two-year capital ledger. Opening private capital is 100, annual depreciation 10%, and baseline private gross investment 20 each year. A financing scenario lowers private investment to 16 each year, holding all other stated conditions fixed. Calculate both paths. Now add a separate public project installing six capital units at the end of year one; it depreciates 10% in year two and receives no further investment. Explain why adding public and private equipment counts still does not establish their combined productive value.

Check the private and public capital comparison

Baseline private stocks are 110 and 119. Lower private investment produces 106 and 111.4, leaving year-two private capital 7.6 below baseline. The public stock is six after year one and 5.4 after year two. Simply adding stocks gives 116.8 in the financing-plus-project scenario versus baseline private 119, but the two types of capital may not be interchangeable. A road, classroom and machine can have different services, users, costs and complementary effects. These records supply no public-project production function or net-benefit estimate. They show a possible accumulation channel, not that public borrowing always reduces total productive capacity by a known amount.

Transfer: capacity versus utilization

An unfamiliar workshop has equipment per worker nine, technology one, ten workers and twenty residents. It uses only 80% of its annual capacity. Calculate capacity, actual output and actual output per resident. Then compare full utilization with a technology increase to 1.2 at the original 80% utilization. Draw a production-frontier sketch and an LRAS sketch showing which changes expand capacity. Their axes differ: the frontier compares two goods; LRAS places aggregate real output against the price level. Do not invent a numerical conversion between those axes.

Check the unfamiliar capacity case

Capacity per worker is thirty, total capacity 300, actual output 240 and actual output per resident twelve. Full utilization yields actual output 300 and fifteen per resident without changing capacity. With technology 1.2, total capacity is 360; at 80% utilization actual output is 288 and output per resident 14.4. Better use of the original resources moves toward an existing frontier; improved technology can move the feasible boundary outward and the potential-output benchmark rightward. Actual output need not equal potential, and higher measured output alone does not identify which change occurred.

Save the capital ledger and unfamiliar diagram before opening worked feedback. High-school practice should explain replacement and denominators; AP practice should distinguish movements and shifts with labeled diagrams; college extension should use attributable investment, depreciation and labor data and defend any aggregation assumptions. These fictional controlled examples do not establish the causes of a country’s growth or independently assessed proficiency.

Skills, hours and policy: identify the channel

More output per worker can mean longer hours

These are separate fictional annual workplace records. Output units have identical quality in this exercise. Calculate total hours, output per worker and output per hour before ranking productivity.

Record	Workers	Hours per worker	Total output
Original	50	1,000	1,000
Longer hours	50	1,200	1,200
Improved process	50	1,000	1,200

Predict which comparison supports greater output per hour. Name two reasons why even that comparison would not establish the effect of a training program without further evidence.

Check worker and hour denominators

Original total hours are 50,000, output per worker twenty and output per hour 0.02. Longer hours produce 60,000 total hours, output per worker twenty-four and unchanged output per hour 0.02. The improved-process record has 50,000 hours, output per worker twenty-four and output per hour 0.024, a 20% hourly increase. Higher output per worker does not alone show improved hourly productivity.

Equipment, task mix, demand, worker selection, quality measurement and utilization could differ between observed workplaces. The label “improved process” states this fictional scenario’s assumption; it is not evidence that training caused an actual difference. A real comparison needs consistently measured hours and output, dates, relevant inputs and a credible comparison group.

Give human capital its own mechanism

Human capital means acquired knowledge and skills that can affect productive work. Training time and credentials are imperfect measures of those skills. For this separate teaching model, replace worker count L in the production rule with effective labor h×L. The skill factor h represents productive services per worker relative to a stated baseline; it is not a measure of a person’s worth. Capacity is Y=10A√(K×h×L). Keep K=200, L=50 and A=1. Compare h=1 with h=1.44. Then compare the same baseline with h=1 and A=1.2 instead.

Check skills versus technology

Baseline capacity is 1,000. With h=1.44, capacity is 10√14,400=1,200; equipment and headcount are unchanged. With h=1 and A=1.2, capacity is also 1,200. These two mechanisms can give the same output. Output alone therefore cannot identify skill acquisition versus a technology improvement. Independent skill assessments, process records, capital measures and comparable work could help distinguish them. The numerical factor 1.44 is supplied, not inferred from attendance or a certificate. Training costs and time are excluded from this capacity calculation and must be evaluated separately.

Participation is not the same as employment or productivity

A fictional community has 125 residents, including 100 working-age adults. Initially 60 adults participate in the labor force, 54 are employed and six are unemployed. A later scenario has 70 participants, 63 employed and seven unemployed; population and working-age population stay fixed. Each employed worker supplies 1,000 hours annually at 0.02 output units per hour in both records. This holds equipment access and hourly productivity sufficient for the added workers by assumption.

Compute participation, unemployment, total output and output per resident for both records. A memo says: “Training reduced unemployment and raised worker productivity.” Identify what the records actually support. Propose a participation policy mechanism and one rival explanation without claiming the mechanism has been observed.

Check participation and the policy claim

Participation rises from 60% to 70%. Unemployment remains 10%: six divided by sixty, then seven divided by seventy. Employment rises by nine. Output rises from 1,080 to 1,260, and output per resident from 8.64 to 10.08. Output per worker remains twenty, and output per hour remains 0.02. The unemployment count rises while its rate stays fixed; neither training nor a productivity change is demonstrated.

A childcare or transport policy might make participation feasible for some adults. A rival explanation could be changed household finances or unrelated employer demand. Evidence should measure eligibility, take-up, employment, hours, costs and affected groups over time. Additional participants do not automatically find jobs. The unchanged 10% unemployment rate and sufficient equipment here are supplied conditions, not universal policy effects.

Compare feasible projects and their timing

A separate council can choose one of two projects, each costing twelve resource units at the start of year one. It has no capacity to run both. The table gives stipulated additional output over a four-year horizon, after operating costs but before the initial project cost. These are two possible scenarios for each project, not probability estimates. Assume zero discounting and no residual asset value solely for this comparison. Monetary transfers are excluded from the resource costs.

Project	Year 1 output	Years 2, 3 and 4 output, each year	Lower-output scenario
Infrastructure	0	6	2 in each of years 2–4
Training	4	4	2 in each of years 1–4

Calculate each project’s total additional output minus the initial resource cost in both scenarios. Recommend a conditional choice, state what uncertainty could reverse it, and identify an affected group whose experience is hidden by the totals. Do not call these net-output figures a complete welfare analysis.

Check project timing and uncertainty

Infrastructure produces eighteen in the first scenario, net of the initial cost six; in the lower-output scenario it produces six, net minus six. Training produces sixteen, net four; its lower-output scenario produces eight, net minus four. Infrastructure has the larger net amount in the first scenario, while training loses less in the lower-output scenario and delivers earlier. No probabilities are supplied, so an expected-value ranking is not available. A choice can depend on the scenario judgment, timing needs and tolerance for losses.

The totals omit distribution, environmental effects, nonmarket benefits, financing constraints and outcomes after year four. Construction disruption may burden particular neighborhoods; training access may exclude workers with inflexible schedules. Maintenance, completion and skill-transfer evidence would test different weak links. Changing the horizon, discounting or residual-value assumptions could change the ranking; do not quietly import those changes into the supplied totals.

Separate spending today from capacity tomorrow

A different, explicitly stipulated aggregate scenario starts with AD: P=200−Y, SRAS: P=40+Y and potential output 80. Here P is a price-level index and Y is real output in teaching units. A policy package initially shifts AD to P=220−Y while SRAS and potential stay fixed. Later, a successful capacity improvement raises potential to 100 and changes SRAS to P=20+Y; AD remains P=220−Y. These changes are assumptions for graph practice, not effects calculated from the twelve-unit project cost above.

Calculate the three intersections and draw the original and changed potential-output lines. Then consider a failed implementation that never changes capacity. Explain why a tax credit encouraging training or investment does not by itself prove either the assumed take-up or the later capacity increase.

Check demand, capacity and implementation

Initially output is eighty and the price index 120. The initial demand shift gives output ninety and price 130, a positive output gap of ten relative to unchanged potential. Under the later successful capacity assumptions, output is one hundred and price 120, equal to the new potential. If capacity never improves, the supplied later SRAS/LRAS shifts are unjustified; the initial demand intersection is not proof of permanent growth. A longer-run failure path would need an explicit price/wage adjustment rule rather than holding a positive gap forever by omission.

A credit can lower an eligible private cost, but take-up, additionality, learning, complementary equipment and implementation time determine whether productive services change. Some subsidized activity might have happened anyway. Financing also matters for near-term demand: taxes, borrowing or displaced purchases can change other spending. The task therefore supplies a net AD shift rather than treating the project’s cost or credit amount as an automatic aggregate-demand increase. Compare the incentive, behavior, measured skill or equipment change, and output evidence as separate links.

Before viewing the checks, retain one denominator table and a policy brief with a mechanism, alternative explanation, timing assumption and distributional concern. For AP practice add the three labeled intersections and explain the output gaps. For college practice propose data and a comparison strategy for one weak causal link; do not represent the fictional project scenarios as an empirical policy evaluation.

Evidence workshop: what does measured productivity establish?

Use the BLS March 21, 2025 archived release, Tables A and B. These are historical estimates for annual 2024 growth in the private nonfarm business sector, not current estimates or the whole economy. The latest year in that release was preliminary. Keep the publication date, reference year and sector in your evidence notes.

Measure	Published annual growth
Real output	2.9%
Hours worked	0.2%
Labor productivity	2.7%
Combined capital and labor inputs	1.6%
Total factor productivity (TFP)	1.3%

Reconstruct the denominator

Before opening feedback, draw two index strips starting at 100: output and hours. Calculate their later values. Output per hour changes by the ratio of their growth factors: 100 × (1.029 / 1.002 − 1) percent. Explain why dividing 2.9 by 0.2 answers a different question. Repeat with combined inputs in the denominator. Treat these calculations as reconstructions from rounded figures.

Check the productivity denominators

The normalized output index becomes 102.9 and the hours index 100.2. Their ratio is approximately 1.026946, giving hourly productivity growth of 2.695%, consistent with the published 2.7% after rounding. The quotient 2.9/0.2 compares two growth rates; it does not calculate output per hour. The percentage change in a ratio requires growth factors, including the original level.

Using combined inputs gives 100 × (1.029/1.016 − 1), approximately 1.280%, consistent with 1.3% after rounding. These denominators describe different questions. Hourly productivity can rise when workers have more productive capital available; TFP accounts for combined inputs. Rounded published rates cannot reconstruct every underlying index exactly.

Repair a misleading contribution chart

The release gives approximate contributions to the 2.7% labor-productivity increase: TFP 1.3 percentage points, capital intensity 1.1 points and labor composition 0.3 points. On paper, draw a horizontal bar from zero to 2.7 with those three segments and label its units. A draft caption says: “Worker training grew 0.3%, while technology alone caused all the remaining growth.” Rewrite it using only what this record supports.

Check contributions and causal claims

The segments add to 2.7 percentage points. A contribution to an aggregate growth rate is not the growth rate of training, spending or each worker’s skill. This accounting breakdown does not identify the effect of a particular training program or technology purchase. A defensible caption is: “The reported labor-productivity increase has positive contributions from all three accounting components; these figures alone do not identify an intervention’s causal effect.”

For a training claim, design a comparison that records who received training, baseline skills, subsequent skills, hours, equipment and comparable tasks. Explain how you would address selection into training and other simultaneous changes. A before/after gain without such a comparison leaves rival explanations open. Your design is a proposed investigation, not an observed result.

Transfer: audit a proposed investment decision

A manager proposes buying new software because the national figures “prove our team will produce 2.7% more.” Write a short decision brief with four parts: what the source measures, why the inference exceeds that evidence, a small local pilot, and an outcome that would make you revise the purchase recommendation. Include implementation time and a quality measure alongside output and hours. Do not invent a treatment effect or a return on investment from these national figures.

Check the investment evidence plan

A defensible pilot compares similar work with and without the proposed change over a specified period, records baseline differences and checks accuracy as well as time. It accounts for setup, training and maintenance rather than counting only faster completion. A quality decline, no sustained time improvement or costs exceeding the measured benefit can reverse a recommendation. Assignment and spillovers between teams need attention. Other reasonable designs are acceptable if they explain how the comparison addresses the claim and what uncertainty remains.

Retain your calculations, labeled bar and brief before opening feedback. High-school practice should explain the denominator and reject the unsupported headline. AP practice should connect productive inputs to potential output while distinguishing this accounting evidence from an identified supply shift. Introductory-college practice should specify a comparison design, measurement limitations and a rival explanation. These are public guided tasks; opening the checks does not establish independent mastery.