What an Hour of Flagger Time Actually Costs: Building a Loaded Labor Rate

Ask a traffic control owner what a flagger costs and you will get an hourly wage. Ask what an hour of flagger time costs the company and the conversation slows down, because the honest answer requires numbers most operations do not track — and the gap between the two figures is where thin-margin companies quietly lose money on jobs they thought they won.

This is a method for building a loaded hourly labor cost you can defend in a bid, in a dispute, and to yourself. It deliberately does not hand you a national average to copy. Averages are the reason so many bids are wrong.

A word on sources before the math. Wage figures move annually and vary enormously by market, and prevailing wage figures are project-specific and legally binding. Rather than quote numbers that will be stale before you read them, this article points at the live, authoritative lookups and shows you what to do with what you find. Verify everything against your own payroll and the determination that governs your specific project.

Why the wage is the smallest interesting number

Start with what the wage does not include.

For every field hour you pay, you also carry employer payroll taxes, unemployment insurance, workers' compensation at a construction-class rate, general liability, auto and equipment insurance allocated to the crew that uses it, any benefits you offer, paid time that produces no revenue, and the administrative work required to keep the person legally deployable — certification, drug testing, onboarding, and in many cases certified payroll reporting.

Workers' compensation deserves particular attention in this trade because the class code is not a clerical detail. Roadway work carries genuine hazard, and the loss experience behind the rates is not theoretical. Using Fatality Analysis Reporting System data compiled by the Work Zone Safety Information Clearinghouse, there were 850 work zone fatalities in 763 fatal crashes in 2024, after 905 fatalities in 824 fatal crashes in 2023 and 903 in 833 in 2022. For the workers specifically, FHWA's work zone facts and statistics, drawing on the Bureau of Labor Statistics Census of Fatal Occupational Injuries, records 94 highway construction worker occupational fatalities in 2022 and 108 in 2021; the underlying series is published in BLS's CFOI tables.

That is the actuarial reality your comp rate is priced against. It also means safety performance is a cost lever, not just a moral one — experience modification moves your loaded rate on every hour you bill, in both directions.

Step one: get your real base wage, from your own payroll

Not from an article. Not from what a competitor allegedly pays. From your payroll register, by classification, over a trailing twelve months — including the overtime you actually paid, not the overtime you planned.

For market context, the Bureau of Labor Statistics publishes Occupational Employment and Wage Statistics for Crossing Guards and Flaggers under SOC 33-9091, with national, state, and metropolitan-area detail. Use it to sanity-check whether you are competitive in your metro, not as your cost input. Your cost input is what you paid.

Two adjustments people skip:

Overtime is not an exception; it is a rate. If your crews reliably run past forty hours in season, the premium portion is a structural cost of your delivery model. Averaged across the year it belongs in the rate. Treating every overtime hour as an unplanned event means your standard rate is wrong for most of the summer.

Shift differentials for night work are a traffic control specialty problem. A large amount of lane closure work happens at night precisely because that is when closures are permitted. If night work is a meaningful share of your volume and you pay a differential, you either need a separate night rate or a blended rate that admits it.

Step two: build the burden honestly

Burden is everything above the wage that scales with employing the person. The categories are not exotic; the discipline is in refusing to guess at any of them.

• Employer payroll taxes — the employer side of FICA, plus federal and state unemployment. Your accountant can give you an effective percentage from last year's filings in about ten minutes.
• Workers' compensation — your actual rate for the applicable class code, adjusted by your experience modification. Not the manual rate; yours.
• General liability and auto — allocated per field hour, which requires knowing your field hours (see step four, where this bites).
• Benefits — health, retirement, per diem policies, anything you actually pay.
• Paid non-working time — holidays, PTO, training days, safety meetings, certification time.

The output of step two is a burden multiplier: burdened cost divided by base wage. Whatever it comes out to for your company is the number to use. Do not adopt a rule-of-thumb multiplier from another company or another trade. Workers' compensation class rates alone vary enough between states and experience mods to make a borrowed multiplier meaningless.

Step three: decide where non-productive paid time lives

This is the step that most often produces double-counting or a silent omission, and it is a decision rather than a calculation.

Field labor in traffic control carries a lot of paid time that generates no billable output: driving to and from the yard, loading and unloading devices, standby while another trade finishes, waiting for an agency inspector, weather holds, and the time between a call-out and arrival.

You have two coherent options.

Bill it separately. If your contracts allow mobilization, standby, or travel line items, bill them and keep those hours out of the loaded rate calculation. This is cleaner and it makes the cost visible to the customer, which is where it belongs.

Absorb it in the rate. If your contracts are structured as a device-day or a crew-day price with no separate mobilization, that time is real cost with no revenue line, and it must be inside the loaded rate — otherwise every single job silently subsidizes it and you will never find out which jobs are actually profitable.

What you cannot do is bill it and leave it in the rate, or leave it out of both. Both errors are common and both are invisible until year-end.

Step four: divide by billable hours, not 2,080

Here is the change that moves most companies' numbers the furthest.

The instinct is to take annual burdened labor cost and divide by 2,080 hours. Nobody is billable 2,080 hours a year. Subtract holidays, PTO, training, weather days, and — the big one in this trade — the seasonal and weather-driven troughs when crews are on payroll and not on jobs.

Divide by the hours you can actually bill. The resulting rate is higher, sometimes uncomfortably so, and it is the true one. Companies that skip this step are systematically underpricing, and they typically discover it during the slow season when the underpricing has nothing left to hide behind.

This is also the moment utilization stops being an abstract metric and becomes the dominant variable in your economics. A crew paid ten percent above market at high utilization beats a cheap crew idling in the yard, every time. Most traffic control companies can tell you their wage rates to the penny and their utilization not at all — which is exactly backwards, because utilization has the larger effect on the outcome and is the one management can actually move.

If you cannot currently answer "what percentage of paid field hours last month were on a billable job," that is the first reporting gap to close. It is also the reason a GPS time clock tied to jobs, rather than a paper timesheet reconciled weekly, pays for itself as a measurement instrument before it pays for itself as a payroll tool. See how Traffic OS ties time clock entries to jobs.

Step five: handle prevailing wage as a separate regime

On federally funded work subject to Davis-Bacon, and on state or local projects with their own prevailing wage laws, an applicable wage determination sets minimum rates and fringe benefit amounts by classification and locality. Those govern for that project regardless of your normal pay scale.

Three rules, and the third is where companies get hurt:

1. Pull the determination for the specific project. Determinations are published and searchable through the federal wage determinations lookup on SAM.gov. They are revised. Reusing last year's determination, or one from an adjacent county, is not a shortcut — it is an exposure.
2. Classification is a judgment with consequences. Which classification a flagger falls under on a given determination is not always obvious, and the answer changes the required rate. When in doubt, ask the contracting agency in writing, and keep the answer.
3. Certified payroll is a real administrative cost and it belongs in your overhead allocation. Weekly reporting, correct classifications, fringe accounting, and the reconciliation work when something does not tie — that labor is not free, and companies that bid prevailing-wage work at their private-work overhead rate are absorbing it. We cover the mechanics in certified payroll for traffic control contractors.

Step six: price the compliance tail

There is a category of cost that sits between labor and overhead and gets allocated to neither, which means it gets recovered from nothing.

Certification and requalification. Flagger certification has to be obtained and maintained, and what counts as qualified is jurisdiction-dependent — Chapter 6E of MUTCD Part 6 covers flagger control including qualifications and high-visibility apparel, in the MUTCD 11th Edition published December 2023 and now carrying Revision 1 dated December 2025. The training hours, the trainer, and the productive time lost are all real. See flagger certification requirements explained.

Turnover. Every departure costs recruiting time, onboarding, certification, and a productivity ramp, and this trade's turnover is not low. If you have never estimated your cost per replacement, do it once — the number tends to change retention policy faster than any argument about culture. See hiring and retaining flaggers.

PPE and apparel. High-visibility apparel is not optional and it wears out. TxDOT's work zone guidance, for example, requires that all contractor employees in the work zone have proper high-visibility safety apparel — not only the flagger. Garment class selection is covered in high-visibility apparel classes explained.

Inspection and documentation overhead. Agencies inspect on a cadence. TxDOT's construction contract administration manual directs the district's responsible person to perform formal inspections of all traffic control devices twice a month at approximately two-week intervals, with at least one at night after initial setup on projects with overnight traffic control. Somebody attends those. Somebody produces the records. That is paid time.

Regulatory floor obligations. 29 CFR 1926.201(a) states that "Signaling by flaggers and the use of flaggers, including warning garments worn by flaggers, shall conform to Part 6 of the MUTCD." Conformance is not free either — it is training, equipment, and supervision.

A worked structure you can fill in

Rather than a fake example with invented numbers, here is the arithmetic skeleton. Fill each line from your own books; the discipline is in refusing to leave any line blank because it is hard to get.

Line 1 — Base hourly wage. Trailing twelve months, by classification, straight time only, from payroll.

Line 2 — Overtime premium, annualized. Total premium dollars paid ÷ total field hours. This converts overtime from an event into a rate.

Line 3 — Night or shift differential, annualized. Same method. Skip only if you genuinely pay none.

Line 4 — Burden multiplier. (Employer payroll taxes + unemployment + workers' comp at your experience-modified rate + GL and auto allocation + benefits + paid non-working time) ÷ base wages. Ask your accountant for the first three from last year's filings and your broker for the next two.

Line 5 — Non-productive paid time. Only if you are not billing mobilization, travel, or standby separately. If you are billing it, this line is zero and it must stay zero.

Line 6 — Realistic billable hours per employee per year. Not 2,080. Start at 2,080 and subtract holidays, PTO, training and safety meetings, weather days, and the seasonal trough. Be pessimistic; you will still be optimistic.

Line 7 — Compliance tail per field hour. Annual certification, requalification, PPE replacement, turnover cost, inspection attendance, and certified payroll administration ÷ annual billable hours.

The rate is (Line 1 + 2 + 3) × Line 4, scaled by the ratio of paid hours to Line 6 billable hours, plus Line 5 and Line 7 where applicable. Then a separate variant for prevailing-wage projects, because that regime replaces Line 1 and changes Line 4's fringe component.

Two sanity checks when you finish. Does the rate multiplied by your actual billable hours reconcile to your actual annual labor cost in the general ledger? If it does not, one of the lines is wrong, and it is almost always Line 6. And is your bid rate above this number by a margin you consciously chose, rather than by whatever the market happened to bear on the last job you priced?

What to do with the answer

The rate is not the deliverable. The decisions it enables are.

Re-rank your job types. Sort last year's completed work by margin using the new rate. The order will change. Short-duration private work with heavy mobilization usually falls; long-duration agency work with stable crews usually rises. Some companies discover their most prestigious contract is their worst.

Set a walk-away number and actually use it. A loaded rate you will not defend in a negotiation is a spreadsheet exercise. The number's value is that it tells you which bids to lose on purpose.

Watch utilization monthly, not annually. Line 6 is the most volatile input and the one management can move. A month where billable hours drop ten percent moves your true cost per hour more than any wage negotiation you are likely to have.

Revisit after any comp-rate or experience-mod change. Workers' compensation is the burden component most sensitive to your own safety record, which means an improving record is a pricing advantage you can only capture if you recalculate.

Putting it together, and the software footnote

The assembled rate looks like this: base wage, times your burden multiplier, plus your allocation of non-productive paid time if you are not billing it separately, divided by realistic billable hours, plus a compliance-tail allocation, adjusted upward for prevailing-wage projects and night differentials where they apply.

Run it once and two things usually happen. Some job types you thought were marginal turn out to be your best work, and some you have been proud of turn out to be break-even. That reordering is the entire value of the exercise. It is also, unglamorously, why bidding and estimating fundamentals is worth revisiting after you build the rate rather than before.

One last thing worth saying plainly, because it connects directly to the arithmetic above. When your software is priced per user, every field hire raises your fixed software cost at exactly the moment you are already absorbing recruiting, certification, PPE, and a productivity ramp. That is a per-head tax landing on the least profitable part of the employment cycle. Traffic OS is priced in flat monthly tiers rather than per seat for that reason — the pricing is published here, and the argument in full is in per-user vs flat-tier pricing.

If you want to test your own numbers against how the system computes job labor cost, bring a real job to a walkthrough — ideally one you suspect you lost money on.