Wage Changes and Unit Labour Cost: Build a Fair Productivity Bridge

Three rising figures can tell different stories

A factory's labour bill rises by 15.5%. The average hourly labour cost rises by 10%. Good output rises by 10% too. Has the factory kept labour cost per garment unchanged?

If labour hours also rose, unit cost has not stayed unchanged, because the calculation depends on accepted output per hour. A pay increase, a larger wage bill and a change in unit cost are related, but they are different measures.

That distinction matters when managers discuss wage increases and garment productivity. A weak comparison can blame workers for a product change, hide extra overtime or promise savings that never reach the accounts.

The worked factory below is fictional. Its currency units, pay costs, hours and output counts are teaching figures. They are not wage benchmarks, legal pay rates or results from an actual business.

Settle the pay obligation before modelling the cost

A factory must establish the pay rules that apply to its workers, taking account of the jurisdiction, sector, job grade, agreement and effective date. A historical wage table does not establish the rate due today.

The ILO's minimum-wage guidance describes a binding pay floor that cannot be lowered by an individual contract or collective agreement. It also stresses that the pay components and hourly or monthly basis need clear definitions.

A productivity model does not change those rules. Nor does it decide what agreed wages can be withheld. Finance can estimate the cost of a change while the responsible payroll team confirms its lawful implementation.

Keep that distinction clear in meetings with workers. “We are checking how the new pay costs affect each garment” has a different meaning from “Your pay depends on meeting this model.” The calculation is a business planning tool, not a replacement for pay rights or fair wage setting.

Decide what belongs in labour cost

Start with a named group of workers and a defined production boundary: when cost includes sewing, material handling and inspection staff, the hours must cover those same people. The output must belong to the work covered by that cost.

Gross wages, take-home pay and the employer's total cost are not interchangeable. Employee deductions can change take-home pay without reducing the employer's gross wage expense. Employer contributions add costs that may not appear in a worker's net payment.

ILOSTAT distinguishes earnings from labour cost. Earnings concern workers' remuneration; labour cost concerns the employer's spending on employment. Its definitions also separate hours paid from hours actually worked.

For this example, cost means the labour compensation charged to one production unit, including gross pay and the employer contributions within its defined account set. The same items are included in both periods. Rent, fabric, utilities and equipment costs sit outside this measure.

The OECD's hourly compensation definition includes gross wages, salaries and employer social contributions. That helps explain why a wage rise and a rise in total hourly compensation need not be equal.

Finance should reconcile the accounts before comparing periods, since a bonus posted late, a new benefit or a changed cost-centre allocation can move the total. Those changes need an explanation of their own.

Match the period, hours and good-output count

The fictional unit makes the same garment type in both periods, with unchanged quality rules and the same production boundary. Output means accepted complete garments from that unit, counted once after its required checks.

The comparison assumes no material change in unfinished work between the start and end of either period. Otherwise, current labour may have produced work that is not yet in the output count. A real review must reconcile that work before treating the monthly ratio as a performance result.

All hours in this example are paid and actually worked. There is no paid leave or unpaid work in the figures. This keeps the teaching arithmetic simple; real payroll and time records often need a bridge between the two hour totals.

Do not define “worked hours” as needle-running time alone. ILOSTAT's definition includes related work, certain downtime and short rest periods. Excluding every interruption would make the denominator tell a different story from total labour input.

The cost and hour records need the same cut-off because different payroll and production reporting dates can create an apparent gain or loss. Align the periods or show the adjustment openly.

Read the first comparison in full

Use CU to mean the fictional factory's currency units. The figures cover all staff included within the unit's agreed labour boundary.

The later bill is CU 4.40 × 10,500 hours = CU 46,200. Both hourly cost and hours increased, so their effects multiply. The bill rose by 15.5%, not by the simple sum of 10% and 5%.

Although output rose by 10%, using 5% more hours meant that the improvement in output per hour was only about 4.76%. Calling the whole output increase “productivity growth” would overstate the gain.

Unit labour cost is CU 46,200 ÷ 22,000 = CU 2.10. The extra good output absorbed part of the hourly cost increase. It did not absorb all of it.

This result does not establish why productivity improved, since better material supply, a more stable method or fewer faults could all contribute. The table measures the change; an investigation must explain it.

Use the relationship that keeps the units consistent

Let C be labour compensation, H the matching labour hours and Q accepted output. For positive hours and output:

Unit labour cost = C ÷ Q.

Hourly labour cost = C ÷ H.

Labour productivity = Q ÷ H.

Therefore, unit labour cost = (C ÷ H) ÷ (Q ÷ H).

The hours cancel, leaving currency per accepted unit. The OECD definition of unit labour cost uses this relationship between hourly compensation and output per hour. Our factory calculation applies the relationship to its defined physical output.

The exact growth relationship is:

Unit-cost growth = [(1 + hourly-cost growth) ÷ (1 + productivity growth)] − 1.

Use growth rates as decimals. In the example, 1.10 ÷ 1.047619 − 1 is 0.05, or 5%. Subtracting the two percentage changes gives about 5.24 percentage points. That shortcut is an approximation, not the exact unit-cost change.

Keep full precision while calculating and round only the displayed result. Otherwise, a rounded productivity rate can make a correct cost bridge appear not to balance.

Show where the extra ten cents came from

A currency bridge can make the result easier to discuss than two growth rates. Begin at CU 2.00 per garment, change hourly cost first, then change the hours used per good garment.

At the old output rate, each garment used 0.5 labour hour. Charging that time at CU 4.40 gives CU 2.20. The hourly-cost change adds CU 0.20 per garment on that basis.

Actual later time per garment is 10,500 ÷ 22,000 = 0.477273 hour. At CU 4.40, that costs CU 2.10. The lower hours per good garment remove CU 0.10 from the intermediate figure.

The bridge balances at CU 2.00 + CU 0.20 − CU 0.10 = CU 2.10, with the hourly-cost change calculated first. Reversing the sequence changes how the interaction is assigned to each step, although the final total stays the same.

This is an accounting explanation, not proof that workers personally caused either movement. The average hourly cost can change with staffing mix, overtime premiums, benefits and other included items.

Find out why the average hourly cost changed

A higher average does not prove that each worker received the same increase, because the mix of paid hours can change while every rate stays fixed.

Consider another small example with two job groups. Group A costs CU 3 per hour and Group B CU 6. These are total hourly compensation figures on a common basis. They are invented, not recommended pay differences.

In the first period, A supplies 3,000 hours and B supplies 1,000. Total compensation is (3,000 × 3) + (1,000 × 6) = CU 15,000. The average across 4,000 hours is CU 3.75.

In the next period, each group supplies 2,000 hours. Total compensation becomes CU 18,000 and the average is CU 4.50. The average has risen by 20%, although neither hourly rate changed.

A different job mix can explain this change without implying that the factory paid someone incorrectly or that any worker became less productive. The output and reason for the staffing change still need review.

To separate rate and mix, compare like jobs on a consistent pay basis. Weight rates by the relevant hours, not just the number of people. A part-time worker and a full-time worker do not necessarily contribute equal hours to the average.

Keep personal pay records restricted. The management bridge can often use grouped figures while the payroll team checks individual entitlements. The grouped average should never replace that individual check.

Allow for a change that begins within the period

An effective date can also explain a smaller reported increase than the announced rate. Suppose the same group works 5,000 hours at a total cost rate of CU 4.00. It then works another 5,000 hours after that rate becomes CU 4.40.

The period cost is CU 20,000 + CU 22,000 = CU 42,000. Across 10,000 hours, the average is CU 4.20: 5% above CU 4.00. A whole later period at CU 4.40 would show the full 10% change, if the other conditions stayed the same.

This illustration assumes equal hours before and after the change. Different hour weights would give a different average. A real forecast must also consider benefits, premiums and any other included cost components.

Keep the actual period result beside the full-period forecast, so readers can distinguish the recorded CU 42,000 from the cost expected in a later period. Conversely, using the partial-period average for the next budget could understate the ongoing cost.

Ask what would hold unit cost steady

The US Bureau of Labor Statistics explanation shows how higher productivity can offset higher hourly compensation. The amount needed depends on both rates, with the same cost and output definitions.

For the example, hourly cost rises by 10%. To hold unit cost at CU 2.00, productivity would also need to rise by 10%: from 2.0 to 2.2 accepted garments per hour.

At 10,500 hours, that would mean 23,100 accepted garments. The actual 22,000 falls short of that cost-neutral scenario by 1,100 garments. Before using that quantity in a production plan, the team would need evidence that it is feasible, can be produced safely and has a customer.

These alternative scenarios all start from the same CU 2.00 base cost:

The range lets a planning team test its assumptions without turning the cost-neutral row into an automatic target for each operator. Capacity, materials, quality, fatigue and customer demand still need evidence.

Separate a harder product mix from poorer performance

Raw garment counts work best when the products are comparable. A shift toward more complex work can lower the piece count even when the same amount of useful work is completed.

Consider a separate fictional example with two styles. Style A has an approved standard of five minutes within the measured scope. Style B has ten. The standards, methods and acceptance rules remain unchanged throughout this comparison.

Earlier standard work content is (8,000 × 5) + (2,000 × 10) = 60,000 minutes. Later work content is (4,000 × 5) + (4,000 × 10) = 60,000 minutes too.

Suppose labour compensation and hours are unchanged in this separate case. Cost per physical garment rises by 25%, since the same cost is divided by 8,000 rather than 10,000 pieces. Cost per standard minute stays unchanged. So does standard work content per labour hour.

The work-content measure depends on valid, stable standards for the same scope; its calculation cannot establish whether those standards are sound. A changed standard can move the measure even when the physical process has not improved.

ApparelDiary's SMV calculation guide explains the basis for a defensible standard. Keep the physical counts alongside the weighted measure so readers can see what changed.

Keep rejected and repeated work visible

A higher piece count is of limited use if more pieces need repair. The unit-cost denominator should follow the stated acceptance boundary, while the numerator retains the labour spent on rejected work and repair within that scope.

Count an accepted garment once. A garment that passes through the same repair station twice has consumed more labour; it has not become two garments. Keeping repair movements separate prevents an apparent output gain from being created by repeated counting.

The time to inspect and repair work must also remain within the chosen labour boundary. Moving repair to another cost centre can improve a sewing report while leaving the factory's total burden unchanged.

A factory may have good reasons to show both direct-sewing cost and a broader unit view, provided each is clearly labelled. If support labour grows, the broader measure helps reveal whether a local improvement simply shifted work elsewhere.

Quality standards must stay comparable as well. Accepting work that previously failed may raise reported good output without improving the product. Review any specification change before treating the new count as a productivity gain.

Look beyond effort when tracing the cause

The OECD's discussion of output per hour explains that labour productivity depends on capital, technology, inputs and organisation. It is not a direct measure of each worker's effort or personal ability.

In a factory, this points the investigation toward the work system. Did cut panels arrive on time? Did a machine fault interrupt the line? Was the method clear? Did the team have to stop for a late approval?

Ask the people doing the work before choosing a remedy, so training addresses a real skill gap rather than a problem elsewhere in the process. It will not make missing materials arrive or repair a damaged guide. Those problems need owners in the relevant functions.

A better labour ratio can also come from purchased equipment whose depreciation, maintenance, energy and funding costs sit outside this labour-only example. A reduction in unit labour cost therefore does not, by itself, prove that total unit cost fell.

Similarly, sending work to a subcontractor can reduce recorded internal hours. The purchased service and the changed production boundary must be considered before calling the result an internal productivity gain.

Distinguish nominal pay from purchasing power

Managers see a payroll cost; workers experience what their pay can buy. These views answer different questions. A rise in money wages does not translate one-for-one into a rise in purchasing power when consumer prices change.

The BLS technical note on real earnings explains its use of consumer price indexes to adjust earnings for price changes. A factory discussion should use the relevant official index and period for its own context.

For a separate hypothetical example, suppose a worker's comparable gross earnings rise by 10% and the relevant consumer price index rises by 8%. The index-adjusted change is:

(1.10 ÷ 1.08 − 1) × 100 = 1.85%, rounded.

That is not a statement about any country's current inflation. Nor does an average index describe every household's spending. It shows why a nominal pay increase and a real earnings increase should be labelled separately.

Avoid mixing this worker-income calculation into the factory's physical output ratio. The factory may need a separate inflation-adjusted cost analysis, but it must state the chosen price basis. A garment count does not become more productive because currency prices change.

Check whether the apparent improvement reaches the business

Producing more accepted garments can spread labour cost across more units. Yet stock that has no customer can consume cash and storage space. When a cost model is used to guide production, the plan must still reflect demand instead of adding output solely to improve the ratio.

For the first example, investigate whether demand exists for the 23,100-unit scenario. Check the next process, packing and dispatch capacity too. Extra work waiting at a blocked stage is not the same as completed output at the agreed boundary.

Lower unit labour cost also differs from a lower total labour bill. A factory can keep its staff, pay more in total and still reduce labour cost per garment through useful output growth. Describe the actual result instead of calling every ratio improvement a cash saving.

If a proposal claims a cash saving, identify the payment that would change and when, since a better allocation rate on an unchanged wage bill creates no cash by itself. The business may instead gain capacity, delivery reliability or a chance to earn additional sales.

Turn the review into decisions people can follow

Bring payroll, finance, production, quality and worker input into the review. Each holds part of the evidence. A single efficiency number cannot explain a pay structure, output quality and product mix at once.

The ILO's 2022 Asian garment-sector report found an association between wage and productivity growth in its historical data. It also warned about measurement and interpretation. That broad finding cannot predict a particular factory's response to a pay rise.

For this factory, the first action is to confirm the CU 0.10 unit-cost increase. Then the team can trace the 500 additional hours, check the good-output count and examine the causes of the productivity change.

Give each proposed improvement a named owner and a measure linked to its purpose; a materials action, for example, might track interruptions caused by late panels. For a method change, a controlled trial should preserve quality and safe work while gathering evidence that the next review can compare with the original assumption.

ApparelDiary's factory KPI dashboard guide can help present the measures clearly. Show hourly cost, labour hours, good output and unit cost together. Add the product-mix note and explain any change in scope.

The useful outcome is a shared explanation of what changed and what the factory can improve. Workers receive the pay due to them. Managers gain a cost model that reflects real work, clear limits and actions they can actually test.

References

• ILO: how to define a minimum wage.

• ILOSTAT: wages and working-time concepts.

• OECD: labour compensation per hour worked.

• OECD: unit labour costs.

• US BLS: the relationship between compensation, productivity and unit labour cost.

• OECD: output per hour and the role of other production inputs.

• US BLS: real earnings technical note.

• ILO: employment, wages and productivity in the Asian garment sector, 2022.