Chapter 20 — Key Takeaways

A one-page reference card. Self-contained: if you come back to this in two years having forgotten the chapter, this page should re-ground you.

Key Takeaways

  • Labor is the only major project cost you genuinely control. In unit cost = quantity × productivity × rate, quantity is fixed by the drawings and rate is fixed by the market, the wage determination, and your insurance carrier. Productivity is the term that moves, and it can swing 40 percent for reasons that appear nowhere in the contract documents.

  • Productivity is a management variable before it is a worker variable. Sequence, access, material staging, crew size, hours, and supervision ratio are all decided in an office. Crew effort is on the list, and it is nowhere near the top.

  • Measure in man-hours per unit, not units per day. A production rate is entangled with crew size; a unit rate is not. Unit rates are what you can compare across crews, jobs, and years — and what you can build a database from. The database is the real asset.

  • The productivity factor is the field's single most useful number. PF = earned MH ÷ actual MH, where earned MH = quantity installed × budgeted unit rate. It is self-normalizing: 0.86 means the same thing in week 3 and week 30. PF is the labor cost performance index expressed in hours, and hours are the unit a foreman can actually see.

  • Forecast, don't report. Forecast MH = budget ÷ PF. Cross-check with actual MH ÷ percent complete. A forecast that survives two derivations is one you can put in front of an executive.

  • Wage is not rate. A fully burdened rate adds paid time off, health and welfare, retirement, FICA, unemployment insurance, workers' compensation by class code times your experience mod, general liability on payroll, small tools, and overhead. Multipliers typically run 1.5–1.8 — higher for low-wage classifications (flat per-person costs) and higher again for high-hazard trades. An ironworker's comp can be nearly three times a carpenter's.

  • Sustained overtime buys production at roughly 1.8× straight time, and by week four at about 2.5×. The efficiency loss exceeds the premium. Short bursts work; ten-week schedules are a way of spending money to go slower. Compressing a week is not extending it — three 12-hour days is 36 hours and usually carries no weekly premium at all.

  • Every work face has a ceiling. Past the optimum crew, added people deliver production above your budgeted unit cost while looking exactly as busy as everyone else. The loss is invisible to observation and completely visible in arithmetic.

  • Prevailing wage is a field discipline, not a payroll one. Classification follows what the hands are doing, hours split by classification need contemporaneous records, and apprentice ratios have to be checked weekly. Requirements vary by jurisdiction and change — verify, date the verification, and re-verify.

  • The workforce problem is a business problem with a business answer: real apprenticeship, visible career paths, predictable schedules, safety, respect expressed as process, and honest pay. The barriers keeping under-recruited groups out of the trades are mostly structural and mostly cheap to fix.

  • Schedule pressure is a hazard. Fatigue, new workers, heat, and an unwritten expectation to "make it up" are the mechanism connecting every productivity lever in this chapter to your incident rate.

Action Items — this week, on your job

  1. Pick one self-performed or heavily-tracked cost code and compute its PF. If you cannot, you have found your real problem: either no counted quantity or no clean hours.
  2. Read the Notes field on yesterday's daily quantity reports and ask a foreman about one of them. Do this tomorrow morning and keep doing it. A report nobody reads becomes fiction within a month.
  3. Pull your overtime hours as a percentage of total hours, by cost code, for the last four weeks. Anything over ten percent without a written reason is a question.
  4. Ask one foreman what the optimum crew is for their operation and why. They know. Almost nobody asks.
  5. Walk one work face at 9:30 a.m. and count how many trades are in it. Then count how many people are walking versus installing. That is a five-minute rating, and it is free.
  6. If you are on a prevailing-wage job: verify the wage determination in your file is current, write the verification date on it, and check one crew's apprentice ratio.
  7. Compare one subcontractor's actual Friday head count against their planned manpower curve. If the gap has widened three updates in a row, make the call now, not after the schedule update reports it.

Common Mistakes and the Fix

Mistake What it costs The fix
Concluding "the crew is slow" from a bad cost code The real cause repeats every week Run the diagnostic order: measurement → environment → crew → rework → effort
Accepting a foreman's "percent complete" as the quantity A fictional PF, in the optimistic direction Count something physical: CY, SF, LF, EA, TON
Cost codes with no unit of measure You can never compute a unit rate, so the code only ever tells you what you spent Every code gets a unit; keep the list small enough to be used correctly
Pricing an overtime hour at 1.5 × the burdened rate Overstates each OT hour by roughly 8 percent, and hides the far larger efficiency loss Half-time premium on base wage, grossed up for statutory burden
Pricing acceleration as premium only On Northgate the unpriced disruption was 57 percent of the priced premium Present three numbers: premium · disruption allowance · residual exposure
Adding bodies to a crowded work face Production above budgeted unit cost, while everyone looks busy Do the marginal arithmetic: extra cost per day ÷ extra production per day
Splitting a crew to cover two locations Loses specialization and rhythm faster than headcount — Thursday's 33 percent Set a minimum crew per position; a split is a priced decision, so price it
Applying one burden multiplier to an average wage Misprices every mixed crew Price each seat at its own burdened rate and blend
Stacking five adjustment factors multiplicatively Arithmetic nonsense that no analyst will defend Take the dominant factor, add a judgment increment, and say in writing that you did
Using a factor table as a disruption claim The weakest form of proof there is Contemporaneous records and a measured mile (Ch 33)
Treating misclassification as a payroll department issue Back wages, penalties, withheld payment, possible debarment It is a field discipline: classification follows the hands, recorded at the time
Incentivizing zero recordables Buys silence, not safety Incentivize leading indicators: near-misses reported, hazards corrected
Telling a crew to "make it up" Finding #3 of the scaffold investigation If you need a date, provide a plan and resources. If you cannot, you need a different date.

Decision Framework

When a labor cost code goes bad — the diagnostic order

1. IS THE MEASUREMENT REAL?
   Recount the quantity physically.  Pull the time cards.
   |  Wrong inputs make every downstream conclusion wrong.
   v
2. DID THE WORK ENVIRONMENT CHANGE?
   Access - sequence - trade stacking - material at the face - information
   |  This is where the answer usually is.
   v
3. DID THE CREW CHANGE?
   Size - mix - foreman - overtime - absenteeism - turnover
   v
4. IS THERE REWORK HIDING IN THE CODE?
   Check the nonconformance log and the RFI log.
   v
5. ONLY THEN, AND RARELY:  effort.

Reading a productivity factor

PF Meaning What you do
≥ 1.05 Beating the estimate Find out why and write it down — this is database gold
0.95 – 1.05 Performing Keep measuring
0.85 – 0.95 Losing quietly Walk the work face this week; check access, sequence, crew mix
< 0.85 Losing loudly Diagnose before adding anything; forecast to completion and tell somebody today

Before you buy schedule with labor — in this order

  1. Sequence. Can the work be re-ordered so the crew is not waiting? Free.
  2. Access and staging. Is material at the face and is the layout shot? Nearly free.
  3. Crew size up to the optimum. Cheap — often below budgeted unit cost.
  4. Compressed hours (e.g., three 12-hour days). No weekly premium in most jurisdictions — check state daily-overtime rules and your labor agreement.
  5. A second crew or a second shift. Real money, real coordination cost.
  6. Sustained overtime. Last. It is the most expensive per day recovered and the one that moves your incident rate.

The five questions before you approve added headcount

  1. What will they actually do? (If the answer is "help," you have an access problem, not a headcount problem.)
  2. What is daily production now, and what will it be with them?
  3. What classification — and does adding a laborer free a journeyman from material handling?
  4. Will the work face physically hold them?
  5. Marginal cost per unit versus budgeted unit cost — and if it is above budget, am I deliberately buying schedule at a known price, in writing?

The prevailing-wage checklist

Current wage determination in the file with a verification date · classification assigned by what the hands are doing · split hours recorded contemporaneously · fringe paid for all hours worked · overtime premium computed on the basic rate per the applicable act · apprentice registration verified and ratio checked weekly · certified payroll submitted on schedule with the statement of compliance signed by an officer · subcontractor payrolls collected and spot-audited against the site roster.

The Three Sentences

  1. Labor is the only major cost you control, and productivity is the term that moves.
  2. Productivity is a management variable before it is a worker variable.
  3. What you do not measure in man-hours per unit, you will explain in dollars per month — and by then it is a post-mortem.