Construction project manager compensation is set as a market-matching exercise and functions as a behavioral control. A bonus tied only to job margin pays the person who forecasts the margin, which delays bad news. Read pay regionally and design the bonus so honest forecasting is not expensive.

Project manager compensation plans routinely ask a good person to choose between accuracy and money. Forecast the job at break-even when the bonus depends on that job’s margin, and the choice is worth about eighteen thousand dollars. Most people are honest. Almost nobody is fast about it.

That delay is the real cost of a badly designed bonus plan, and it does not appear in any compensation benchmark. The benchmark says whether you are paying market. It says nothing about what your pay structure is quietly instructing people to do.

Construction project manager compensation gets treated as a benchmarking problem because that is the part with data attached. The structural half is where the money actually leaks.

What does construction project manager compensation look like regionally?

Wide spreads between metros for the same role, driven by local backlog rather than cost of living. Read the metropolitan wage tables directly. A national band produces offers that lose consistently in the tightest markets and overpay in the softest, and averages the difference into a number that fits nowhere.

The primary sources are worth going to directly rather than relying on aggregated survey summaries.

The BLS OEWS construction manager wage tables publish the distribution, and the metropolitan area breakdowns split it by market. Texas has its own state table, which is useful given how much of the current demand sits in Dallas, Houston and Austin. The BLS occupational profile describes the scope of the function being priced.

The thing those tables cannot tell you is direction of travel, which is set by local backlog. A metro where three large programs started in the same quarter will price project management above its own published median within a year, because the competing bidders are hiring from the same pool at the same time. Census Bureau construction spending data by sector is the leading indicator worth watching.

Getting the number right still leaves the harder question of how it is structured.

Why does a job-margin bonus create a forecasting problem?

Because the person who reports the forecast is the person paid on it. That is a control weakness, not a character question. A project manager who suspects a job is fading has a financial reason to wait a month for better information, and a month is how a fade becomes a surprise.

The consequence lands somewhere the compensation committee never looks.

Under percentage-of-completion accounting, reported profit is downstream of cost-to-complete estimates supplied by project teams. A systematic bias toward optimism, even a mild one produced by ordinary human incentive rather than deception, means the firm books profit it has not earned and discovers the shortfall at closeout. The CFMA financial management literature treats forecast integrity as a governance control for exactly this reason.

Put a number on it. On a US$30 million job carrying a 4 percent target margin, the margin at stake is US$1,200,000. A bonus paid at 1.5 percent of that margin is US$18,000 to the project manager. A fade caught in month three might cost the firm US$150,000 to recover; the same fade found in month nine can cost US$600,000, because the recovery options have closed. The plan is offering US$18,000 to the person best placed to prevent a US$450,000 difference, and pointing the incentive the wrong way.

Cost of a fade by the month it is identified Bar chart, Cost of a fade by the month it is identified: Found in month three $150k, Found in month nine $600k, Bonus at stake $18k. Cost of a fade by the month it is identified The bonus that discourages early reporting is small next to what late reporting costs. $0k $250k $500k $750k $1,000k Found in month three $150k Found in month nine $600k Bonus at stake $18k Source: Worked example in this section
The bonus that discourages early reporting is small next to what late reporting costs.

Sound plans separate the incentive from the estimate. Three structures that do it:

  • Split the award. A portion on job margin, a portion on forecast accuracy measured at closeout, and a portion on something firm-wide the individual does not control alone.
  • Pay on the closeout number, not the interim forecast. Slower, and it removes the incentive to protect a monthly report.
  • Reward early bad news explicitly. A fade identified in month three costs far less than the same fade found in month nine, and a plan that recognizes that changes behavior quickly.

The third is unusual and the most effective. It is also the hardest to get approved, because it looks like paying for failure until someone models what the late version costs.

What actually drives retention in this market?

Authority and workload, more reliably than pay. Compensation gets someone to take the call. What decides whether they stay is whether they run their jobs or administer them, and how many jobs they are carrying. A project manager on four active projects is not retained by a raise.

That is not an argument for underpaying, and the market makes underpaying expensive.

In the 2025 AGC and NCCER workforce survey, 91.7 percent of the 1,041 contractors answering the salaried-hiring question reported difficulty filling salaried positions, and the AGC release reports 45 percent of firms saying shortages are delaying projects.

The BLS JOLTS series tracks construction hires and separations, and the separations line is the one to watch: in a tight market, voluntary departures cost more than the wage differential that would have prevented them. Whether repeated departures from one seat are a pay problem at all is the question our note on PM turnover works through.

The comparison worth having internally:

What the review usually adjustsWhat actually decides retention
Base salary against a national bandBase against the specific metro’s tables
Bonus percentage of baseWhat the bonus rewards reporting-wise
Vehicle, allowance and benefitsNumber of concurrent active projects
Title progression on a scheduleDecision authority that matches the title
Annual review cycleResponse when they raise a problem early
Counteroffer when they resignFixing the cause before the resignation

The last row is where most firms spend the most and get the least. A counteroffer resets the number and leaves the reason intact, which is examined in the construction counteroffer trap. During a sale or a backlog surge the instrument is a stay bonus, which holds the seat through a window rather than fixing the reason.

What are the compliance limits on manager compensation?

Salaried project managers are usually exempt, but the classification is a test rather than an assumption, and the consequences of getting it wrong are federal. Where public funding is involved, prevailing wage obligations apply to covered classifications regardless of how the firm structures its own pay.

The Department of Labor executive exemption guidance sets out the actual test, and it turns on duties rather than job title or salary alone. On publicly funded construction the prevailing wage rules govern covered work, with state contacts administering state-level equivalents.

Nothing about engaging a search firm changes any of those obligations, and no compensation structure should be designed around avoiding them. Where field supervisory duties blend into a salaried role, the classification question deserves an actual answer from counsel rather than an inherited assumption.

Selah works with contractors and construction consultancies across the United States, on preconstruction and estimating, project and construction management, and cost and commercial management. Candidates are never charged a fee at any stage.

Because the most accurate read on what a market pays comes from people who are not applying to anything. Published tables are lagging and survey data is self-reported. Live conversations with employed project managers give current numbers, and those conversations only happen off-market.

This is adjacent to why the searches themselves run the way they do. Selah Talent Partners runs confidential executive search for mandates that cannot be publicly posted, and the same channel that reaches a passive candidate is the one that produces honest compensation intelligence. The mechanics are set out in how a confidential construction search runs.

This article is not written for a firm setting pay for a first project management hire against a straightforward portfolio. There the published metro tables are sufficient and a simple structure is correct. It is written for firms whose bonus plans have accumulated over a decade and are now quietly shaping what gets reported.

Pay is a set of instructions

Every compensation structure tells people what the firm values, and it tells them more clearly than any stated priority. A plan paying entirely on job margin says: protect the margin number. People are good at following instructions.

The intended message was probably about accountability for outcomes. The received message is about the monthly report, and there is a gap between those two that shows up at closeout on the jobs that went wrong.

Benchmarking fixes the level. It does not fix the instruction.

So read the metro tables, get the number right, and then ask the harder question: what does this plan pay someone for telling us bad news in month three?

If you are reworking a compensation structure and want to compare notes, get in touch.

Sources and further reading