DCAA Compensation Compliance: The Mid-Market Contractor's Guide to Surviving an Incurred Cost Audit
The audit finding that should scare you more than the cap
Most GovCon finance and HR leaders know the FAR 31.205-6(p) compensation cap number. The 2025 cap is $671,000. The 2026 cap has not yet been officially published. Every GovCon CFO can cite that figure on demand.
That is not where the audit finding usually comes from.
The finding comes from FAR 31.205-6(b): the reasonableness test. DCAA compares your actual compensation against survey data, builds a "range of reasonableness" that is typically 10 percent above survey midpoints, and questions every dollar above the line. The cap is a ceiling. Reasonableness is the floor you are actually fighting on.
If your company has cost-reimbursable contracts, an incurred cost submission due under FAR 52.216-7, and anyone earning more than the survey midpoint for their role, this post is for you.
What DCAA compensation compliance actually requires
DCAA compensation compliance means satisfying two separate tests under FAR 31.205-6. The first is the statutory cap in paragraph (p), which limits allowable compensation for covered employees to a specific dollar amount set annually. The second is the reasonableness test in paragraph (b), which requires that compensation be comparable to what similar companies pay for similar work. A compensation cost is allowable only if it passes both tests.
Miss either one, and the excess becomes unallowable. Unallowable costs get excluded from indirect cost pools. That reduces your billable overhead and G&A rates. Which reduces what you can bill the government. Which compresses your margin in a business model that already runs thin.
The contract award date rule most contractors still misread
Here is the rule that consistently generates audit surprises.
For contracts awarded before June 24, 2014 (or before December 31, 2011 for DoD, NASA, and Coast Guard contracts), the cap applies only to the top few senior executives. The exact definition varies by contract year, but it was roughly "the CEO plus the four most highly compensated employees in management positions."
For contracts awarded on or after June 24, 2014, the cap applies to all employees. Not just executives. Not just the top five. Every employee in a management position whose total compensation exceeds the cap.
Most mid-market contractors understand the cap applies to executives. Many have not updated their cost accounting practices to reflect that post-2014 contracts expanded the scope to include senior engineers, program managers, and technical specialists. If your contract portfolio is a mix of pre- and post-2014 awards, your compliance work is not uniform across the book. The cap applies differently to different contracts in the same incurred cost submission.
This is where I have seen clean incurred-cost proposals get messy. A company treats the cap as an executive-only issue, staffs a 750,000 dollar principal engineer on three post-2014 contracts, and discovers during an audit that 79,000 dollars of that employee's compensation is unallowable on the government portion of the work.
Why the reasonableness test generates more findings than the cap
The cap is mechanical. Total compensation above the statutory number is unallowable. No judgment involved.
Reasonableness is judgment all the way down.
DCAA's approach, codified in Chapter 6 of the Contract Audit Manual, is to compare your executive and high-earner compensation against multiple third-party compensation surveys. They build a composite market figure, apply the 10 percent range of reasonableness, and question everything above the line.
The approach has been challenged successfully. I have some direct experience with this.
In 2012, I was a second-year senior associate at Mercer when I was staffed on the Metron Aviation engagement. DCAA had disallowed approximately $1.1 million of Metron's executive compensation as unreasonable. The Mid-Atlantic Compensation Review Team had built their challenge by averaging four surveys and matching Metron's Senior Engineers against benchmark positions that did not reflect the actual responsibilities of those roles. Digging into the comparables, I found that DCAA was benchmarking against irrelevant position-level matches. They were anchoring their reasonableness analysis to survey cuts that did not correspond to the work Metron's executives actually performed. I re-benchmarked the data against the correct comparables and wrote the analytical brief that our Partner, Christopher McGee, used as the foundation for his expert witness testimony at the Armed Services Board of Contract Appeals.
The Board sustained Metron's appeal. In Metron, Inc., ASBCA Nos. 56624, 56751, 56752 (June 4, 2012), the ASBCA found DCAA's methodology "fatally flawed" and ruled that all of the disputed compensation costs were allowable. Alongside the J.F. Taylor, Inc. decision earlier that year, Metron established that DCAA's statistical approach to reasonableness was contestable, and that position-level job matching was one of the methodology's weakest links.
Fourteen years later, I still see DCAA auditors making the same position-matching errors against mid-market contractors who do not have the resources to push back. Metron and J.F. Taylor did not change DCAA's approach. They did establish that contractors with proper analytical support can win.
The practical problem is this: most mid-market contractors do not have the compensation documentation DCAA expects to see. They have a salary band. They may have a market study from three years ago. They do not have a current-year, role-by-role justification tied to reputable surveys, with fringe benefit adjustments per the Techplan Corporation (98-2 BCA) precedent, and a written rationale for every compensation decision above the survey median.
When DCAA asks "how did you determine this salary was reasonable?" and the answer is "our HR director benchmarked it," that is not a process. That is a description of a person. The contractor has already lost the frame.
What a DCAA-ready compensation file actually contains
Based on FAR 31.205-6(b)(2) requirements and DCAA audit practice, a defensible compensation file for each covered employee includes:
Role match documentation. The specific survey job description used for benchmarking, with a written rationale for why this survey role matches the actual duties. If your "Vice President of Engineering" is really a hands-on technical lead, you match to a technical role, not a VP role. Survey position semantics matter and DCAA will exploit mismatches.
Multiple reputable surveys. DCAA uses multiple surveys for a reason. A single-source justification is always weaker than a three-source justification. Radford, Mercer, Willis Towers Watson, Culpepper, and industry-specific surveys all qualify depending on the role.
Size and geography adjustments. Compensation varies meaningfully by company revenue band, headcount, and metro area. A national average is not a defensible benchmark for a $200M contractor in Huntsville, Alabama.
Fringe benefit offset analysis. Per Techplan, if your fringe benefits are below market, you can adjust cash compensation upward in the comparison. This is one of the most commonly missed offsets in contractor compensation files. If your health plan is thin and your 401(k) match is 3 percent, those data points help you, not the auditor.
Written compensation policy and process. The HR policy that governs how compensation decisions are made. Board or compensation committee minutes where applicable. Documentation that compensation was set through a defined process, not by intuition.
Annual reasonableness review. A documented, pre-determination review showing that compensation was set with reference to market data before the year began, not reconstructed in response to an audit.
The goal is to shift the audit posture. When the file is complete, DCAA has to prove your analysis is wrong. When the file is thin, you have to prove DCAA's analysis is right. The asymmetry is significant.
The ICE Model and Schedule B reality
When you submit your annual incurred cost proposal under FAR 52.216-7, the executive compensation data goes into the ICE Model's Supplemental Schedule B. That schedule is where DCAA verifies cap calculations. Every named covered employee, every compensation element, the cap calculation, the order in which compensation elements are disallowed.
The order matters. If you have a CEO at $400,000 salary, $300,000 bonus, and $175,000 deferred comp, total compensation is $875,000. The cap excess is $204,000 against the 2025 cap. You have a choice about which compensation elements to disallow first. Bonuses have different downstream implications than salary. Deferred compensation has different implications than both. The downstream impact on your indirect cost pools can differ materially based on which elements you treat as unallowable.
If your finance team is filling out Schedule B for the first time in April under audit pressure, that optimization decision is being made badly.
Four patterns I see in mid-market GovCon compliance gaps
Every mid-market contractor I have worked with shows some version of these four patterns.
Pattern 1: Cap-focused, reasonableness-neglected. The finance team tracks the statutory cap meticulously. The compensation file for anyone earning under the cap is almost nonexistent. Then DCAA questions the $385,000 director of programs and there is no file to defend.
Pattern 2: Out-of-date market data. The last formal compensation benchmark was done three years ago, before the last two rounds of inflation and wage compression. Current salaries have drifted from the reference data. No one has refreshed the baseline.
Pattern 3: No contract-award-date segmentation. The company treats every contract the same under the cap rules. Post-2014 contracts get the executive-only treatment instead of the all-employees treatment. This creates a systematic compliance gap that compounds every year the contract continues.
Pattern 4: The M&A integration gap. A company acquires another contractor. The target's compensation structure was built under different assumptions, possibly in a different industry, possibly with different cap logic. Integration focuses on payroll, benefits, and HRIS. The FAR reasonableness documentation, role matching, and compensation policy alignment gets skipped. Three years later the first post-deal incurred cost audit finds the mess.
The fourth pattern is common enough that I have come to treat M&A compensation due diligence as a distinct service line. Most buyers do not scope a FAR-specific compensation review into transaction diligence. The findings land after close.
How to fix this before your next audit
The work is not glamorous. None of it requires new software or new headcount. It requires a compensation infrastructure that works the same way every year.
Segment your contract portfolio by award date. Know which contracts are pre-2014 and which are post-2014. The cap scope differs. Your compliance work differs accordingly.
Identify your covered population. For post-2014 contracts, this is everyone in management positions whose total compensation approaches or exceeds the cap. Build the list. Refresh it annually.
Build a current-year compensation file for each covered employee. Multiple survey sources. Role match rationale. Size and geography adjustments. Fringe benefit offset where applicable. Written narrative.
Document the process, not just the outcome. Compensation policy. Review cycle timing. Approval authority. Survey selection criteria. The process documentation is often more valuable than the individual employee files.
Fill out Schedule B before you need to. Run the cap math mid-year. Identify which elements to disallow and in what order. Do not make that decision under audit pressure.
The effort is front-loaded. The annual maintenance is small. The payoff is every audit you walk into with a file that makes DCAA's work harder than theirs makes yours.
When the audit comes
A DCAA compensation reasonableness audit is winnable. The Metron and J.F. Taylor cases established that the methodology is contestable. Contractors who push back, with documentation, win more often than the folklore suggests.
Contractors who do not push back, or who cannot because the file is thin, settle. The settlement is always expensive.
The mid-market contractors who struggle most are the ones whose compensation function was built for payroll accuracy and benefits administration, not for regulatory defensibility. These are different jobs. A good payroll function is table stakes. A compliance-ready compensation infrastructure is the moat.
Most mid-market contractors do not have it. Building it is the work.
The Barksdale Group builds compensation infrastructure for mid-market companies and private equity portfolio companies, with specific GovCon compliance expertise. If you are preparing for an incurred cost audit, restructuring post-acquisition, or working through a compensation policy gap, the blog has additional resources. You can also reach Stephen directly through barksdalegroup.co.
