18.3 Cost Drivers, Change Order Impact & Resource Utilization Analysis
Key Takeaways
Labor cost variance decomposes into rate variance (actual minus budget rate times actual hours), efficiency variance (actual minus earned hours times budget rate), and quantity variance (actual minus budget quantity times budget unit cost).
The three variance components demand different responses — rate points at wage agreements and crew mix, efficiency at supervision, sequencing, access and engineering completeness, and quantity at design growth or takeoff error.
Change order impact has three layers: direct cost, indirect and time-related cost for any added duration, and disruption to unchanged work, of which only the first is routinely priced.
Approving a change and incorporating it are separate acts; a change approved but never moved into the baseline and commitment register produces a permanent administrative variance that discredits real ones.
Utilization must be reported alongside productivity factor, because high utilization on work that earns little value consumes budget while creating the appearance of activity.
18.3 Cost Drivers, Change Order Impact & Resource Utilization Analysis
Three analytical tasks close out Domain 1: 1.V identify/evaluate cost drivers, 1.W evaluate impact of change orders, and 1.X evaluate resource allocation/utilization. They are what a cost engineer does when the reports show a problem and someone asks why.
1. Cost Drivers (Task 1.V)
A cost driver is a variable whose movement causes a disproportionate movement in cost. Identifying drivers converts a cost report into a management instrument, because drivers are where intervention actually works.
| Driver category | Examples | Why it dominates |
|---|---|---|
| Quantity | Installed tonnes, linear metres, cubic metres | Multiplies through every unit rate |
| Productivity | Labor-hours per unit installed | Compounds across the whole labour base |
| Rate | Wage agreements, commodity indices, currency | Applies to every hour or unit |
| Scope definition | Design maturity, late design change | Drives rework and quantity growth |
| Schedule | Duration, sequence, overtime, acceleration | Drives indirects, premium time, disruption |
| Site conditions | Access, congestion, weather, subsurface | Depresses productivity across many activities |
Separating price, quantity, and productivity variance
When a labour cost account overruns, the useful decomposition is:
- Rate (price) variance = (Actual rate − Budget rate) x Actual hours
- Efficiency (productivity) variance = (Actual hours − Earned hours) x Budget rate
- Quantity variance = (Actual quantity − Budget quantity) x Budget unit cost
Worked example. A structural steel erection account was budgeted at 1,200 tonnes at 22 labor-hours per tonne and $68 per hour — a budget of 26,400 hours and $1,795,200. Actuals: 1,260 tonnes installed, 31,500 hours expended, average rate $71 per hour, actual cost $2,236,500.
| Component | Calculation | Value |
|---|---|---|
| Earned hours | 1,260 t x 22 h/t | 27,720 h |
| Efficiency variance | (31,500 − 27,720) x $68 | −$257,040 unfavourable |
| Rate variance | ($71 − $68) x 31,500 | −$94,500 unfavourable |
| Quantity variance | (1,260 − 1,200) x 22 x $68 | −$89,760 unfavourable |
| Total | sum | −$441,300 |
Check: $2,236,500 − $1,795,200 = $441,300. The decomposition matters because the three components require completely different responses. Efficiency (58% of the overrun) points at supervision, sequencing, access, or engineering completeness. Rate (21%) points at the wage agreement or crew mix. Quantity (20%) points at design growth or takeoff error. A report that says only "steel is $441,300 over" invites the wrong intervention.
2. Evaluating Change Order Impact (Task 1.W)
A change order's price is rarely its cost. Blueprint task 1.W asks for the impact, which has three layers.
| Layer | Content | Typically priced? |
|---|---|---|
| Direct | Labor, material, equipment, and subcontract for the changed work itself | Usually |
| Indirect / time-related | Extended supervision, site facilities, equipment hire, staff, insurance and bonding for any added duration | Often omitted |
| Impact / disruption | Lost productivity on unchanged work caused by the change — rework, out-of-sequence work, trade stacking, learning-curve reset, acceleration | Rarely, and hardest to prove |
The cumulative impact problem
Individually trivial changes can combine into a severe productivity loss. Twenty small changes, each priced at direct cost only, can leave a project with a large unrecovered disruption cost that no single change order captured. This is why change frequency and change volume are tracked as leading indicators, not merely change value.
The evaluation checklist
- Entitlement and scope. Is this genuinely a change against the contract baseline, or work already included?
- Direct pricing. Are the quantities, rates, and markups consistent with the contract's change-pricing mechanism?
- Credits. Is deleted work credited at the same basis as added work is priced?
- Time impact. Does the change affect the critical path? Task 5.C — translating schedule change into cost — is the analysis, covered in Section 21.2.
- Indirect and time-related cost. If duration extends, the time-related cost must be priced with it.
- Disruption. Is there a measurable productivity effect on unchanged work?
- Baseline incorporation. Once approved, the change must move the baseline (task 1.J) and the commitment (task 1.S). A change approved but never incorporated produces a permanent, unexplainable variance.
Important
A change order and a baseline revision are two separate acts. Approving the change authorizes the work and the money. Incorporating it moves BAC, the time-phased baseline, and the commitment register. Projects that do the first and forget the second report cost overruns that are purely administrative, and lose credibility for the real ones.
3. Resource Allocation and Utilization (Task 1.X)
Resource analysis asks whether the people and equipment the project is paying for are actually producing value.
| Metric | Definition | What it exposes |
|---|---|---|
| Utilization | Hours charged to productive work / hours paid | Idle and standby time |
| Direct work rate | From work sampling: % of observed time in direct value-adding work | Access, materials, and supervision problems |
| Productivity factor (PF) | Budget hours / actual hours for the same earned quantity | Overall labour efficiency (PF below 1.0 is unfavourable) |
| Composite rate variance | Actual blended rate vs. planned crew mix rate | Crew mix drift toward higher-cost trades |
| Overtime ratio | Premium hours / total hours | Cost premium plus the productivity decay of sustained overtime |
| Peak-to-average manning | Peak headcount / average headcount | Congestion risk and hiring/demobilization cost |
Two analytical points the exam favours:
- Sustained overtime degrades productivity. Extended scheduled overtime — the classic 6 x 10 pattern held for months — produces cumulative efficiency loss, so the true cost of an overtime acceleration is the premium plus the productivity decay, not the premium alone.
- Overmanning is self-defeating. Adding crew beyond the work face's physical capacity produces trade stacking and congestion, so the marginal crew's productivity can approach zero while its cost is fully incurred. Resource analysis that reports headcount without reporting earned hours per head cannot detect this.
Tip
Report utilization alongside earned value, never instead of it. High utilization on work that earns little value is worse than idle time, because it consumes budget while creating the appearance of activity. The diagnostic pair is utilization (are they working?) and productivity factor (is the work producing earned value?).
A structural steel erection account was budgeted at 1,200 tonnes, 22 labor-hours per tonne, and $68 per hour. Actuals are 1,260 tonnes installed, 31,500 hours expended, an average rate of $71 per hour, and $2,236,500 of cost. Which decomposition of the $441,300 overrun is correct?
Rate variance $94,500, efficiency variance $257,040, quantity variance $89,760 — with efficiency the largest component, pointing at supervision, sequencing, access, or engineering completeness.
Rate variance $257,040, efficiency variance $94,500, quantity variance $89,760 — with rate the largest component, pointing at the wage agreement.
Rate variance $189,000, efficiency variance $163,200, quantity variance $89,100 — with the overrun split roughly evenly across all three causes.
The overrun is entirely a quantity variance of $441,300, because 60 additional tonnes were installed beyond the budgeted quantity.
A project has processed 34 approved change orders in six months, each priced at direct cost only and each individually small. Craft productivity has fallen from a factor of 1.02 to 0.79 over the same period with no other identifiable cause. What is the most likely explanation, and what should the cost professional do?
The productivity fall is unrelated to the changes, since each change was individually small and separately priced at fair rates.
The changes have produced a cumulative impact — rework, out-of-sequence work, trade stacking, and learning-curve resets on unchanged work — which no individual change order captured; track change frequency and volume as leading indicators and evaluate the disruption cost.
The correct response is to reprice each of the 34 change orders at a higher unit rate to recover the productivity shortfall retroactively.
The productivity factor should be rebaselined to 0.79 so future reporting reflects the achievable rate.
A site reports 94% craft utilization, defined as hours charged to productive cost codes divided by hours paid, and management concludes that labour is performing well. What additional metric is essential, and why?
Peak-to-average manning ratio, because utilization is only meaningful when the workforce is at steady-state headcount.
Overtime ratio, because utilization above 90% can only be achieved through scheduled overtime.
Total headcount trend, because utilization percentages are unreliable unless the denominator is stable across periods.
Productivity factor (budget hours divided by actual hours for the earned quantity), because high utilization on work that earns little value consumes budget while creating the appearance of activity.
Sections you finish are checked off in the contents.