The dominant narrative right now is that IRS workforce cuts mean lower audit risk. If you own real estate and commission cost segregation studies, that narrative may be leading you somewhere you do not want to go.
I interviewed James C. Peacock last week. He spent 38.5 years at the IRS as a General Engineer. He was among the first IRS engineers to examine cost segregation. He contributed to the original Cost Segregation Audit Techniques Guide, developed 2000 through 2002 and released publicly in 2004, and every major update through the February 2025 edition. He served as the IRS's primary technical expert on Section 179D from 2014 until he retired in September 2025.
According to James, he personally trained approximately 200 new-hire IRS engineers on cost segregation and Section 179D methodology before retiring. This is his own account, consistent with his role as the designated Section 179D Subject Matter Expert for the entire IRS, where training new engineers was a core part of his job.
Quick Answer: Per James Peacock, he trained approximately 200 IRS engineers on cost segregation and Section 179D before retiring in September 2025. Business audit rates are still around 0.07%, or roughly 0.078% with a cost segregation study, but specialist review capacity means weak studies remain exposed.
That is not the story most tax professionals are telling their clients right now. Here is what James told me.
Why "IRS Cuts = Less Audit Risk" Is the Wrong Read for Cost Segregation
The argument goes like this: IRS staffing is shrinking, enforcement budgets are down, audit rates are at historic lows. Therefore, the odds of getting examined on your cost segregation study are negligible. Therefore, study quality matters less than it used to.
Every part of that chain is wrong, at least for cost segregation at any meaningful property value.
Cost segregation examinations are not handled by regular revenue agents. They require specialized IRS engineers. The LB&I Division, Large Business and International, is where cost seg audits with real dollar figures actually land. These are not the same agents who audit a Schedule C for a freelancer. They are engineers trained specifically in construction cost estimating, property classification, and the Cost Segregation Audit Techniques Guide.
The pool of IRS engineers with that specific training did not get smaller before 2026. Per James, he trained roughly 200 of them.
"The audit rate might not go down," he told me. "If you get caught in a trap, it's still a trap."
That sentence is worth sitting with. The question is not whether the IRS has fewer total agents. The question is whether the engineers who know cost segregation are still there, and whether they know what to look for. Both answers are yes.
The 200 Engineers: What James Actually Trained Them On
James was not running a general IRS onboarding program. He was training engineers specifically on cost segregation classification methodology and Section 179D energy efficiency deductions, the two areas he spent most of his career on.
The training covered the Audit Techniques Guide in detail, including the LUQ framework (Large, Unusual, Questionable items), RS Means cost code specificity standards, land allocation requirements, property classification rules under the Whiteco test, and how to read a cost segregation study for the flags that generate Information Document Requests.
These 200 engineers entered the IRS workforce before James retired in September 2025. They are working cases now.
General workforce cuts at the IRS reduce the number of people reviewing straightforward individual returns. They do not eliminate a specialized cohort trained on a highly technical examination area. The IRS has invested significantly in this specific capacity. That investment does not disappear because headline staffing numbers went down.
The broader IRS enforcement picture is also more nuanced than the cuts narrative suggests. The IRS has been deploying AI for return selection through what James described as its "agent force" contractor program. AI tools are now being used, in limited form, for identifying returns that warrant closer examination. The IRS is not replacing engineers with AI, it is using AI to flag which returns those engineers should look at first.
That is a different threat model than "the IRS has fewer people." It means a smaller enforcement workforce may be more precisely directed at the cases with the highest return on examination.
How Cost Segregation Studies Actually Get Examined
This is the part of James's interview that changes how you should think about study quality.
Cost segregation is almost never the reason an audit opens. James was direct on this: "Cost seg studies don't get discovered until there is already an audit open." The IRS uses DIF scores, Discriminant Inventory Function, a proprietary statistical model, and targeted examination campaigns to select returns. Once an audit opens on a business or real estate entity for any reason, the engineer assigned to the case looks at depreciation schedules. That is when the cost segregation study gets reviewed.
This means the relevant question is not "will the IRS specifically target my cost seg study." It is "if my return gets examined for any reason, will my cost seg study hold up."
James described exactly what he looked for when he reviewed a study. He went to the back of the report first. He looked for RS Means codes, the 12- or 16-digit codes that correspond to specific construction cost items. "RS Means mechanical" with no code after it does not tell an IRS engineer anything. It generates an Information Document Request. If the codes were there and specific, the study was defensible. If they were vague or absent, the study was a problem.
He then looked at land allocation. "If a property is purchased for a million dollars and the study adds up to a million dollars, we go, where's the land?" Land is not depreciable. A study that accounts for 100% of the purchase price in depreciable assets, with nothing allocated to land, is a flag that generates an automatic adjustment, not a negotiation.
For larger properties, he described reviewing the structural components of 1250 property claims. He told me about a high-rise study where the 1245 property classification looked reasonable on the surface, but the 1250 column had no steel framework and no concrete slabs allocated. "Tens of millions of dollars just gone," he said. "That was an automatic adjustment right there."
These are not edge-case scenarios. These are the checks IRS engineers run on every study they examine. The 200 engineers James trained know these checks. They will run them too.
The LUQ Framework: Why Thresholds Were Never Published
The IRS uses a framework called LUQ, Large, Unusual, Questionable, to flag items inside a cost segregation study that warrant closer scrutiny. James spent considerable time on this in our interview.
The framework is deliberately non-specific. The IRS never published percentage thresholds for what constitutes a "large" allocation to 1245 property because doing so would immediately give providers a number to game. James explained it directly: "As soon as we say X percent is acceptable, everyone's going to claim X minus 1 percent."
The example he used was a standard warehouse. If 50% of the total construction cost is allocated to HVAC in a non-refrigerated warehouse, that is an LUQ item. Not because 50% crosses a published threshold, there is no published threshold, but because HVAC at that proportion is implausible for the property type. An IRS engineer familiar with construction costs knows what HVAC typically runs in a warehouse. A study that deviates significantly from that range will draw scrutiny.
This is why square footage models fail under examination. A study that allocates 1245 property based on square footage averages rather than property-specific RS Means codes tells the examiner the engineer never looked at the building. It generates more IDRs, more scrutiny, and a worse outcome.
The contractors who build your cost segregation study need to show their work. The IRS engineers reviewing that work know construction costs. They were trained, in many cases, by someone like James Peacock.
What "Support, Not Report" Actually Means
There was a phrase James came back to several times during our interview. He said it was a mantra among IRS engineers throughout his career: "It's the support, not the report."
A cost segregation report can look professional, run 200 pages, and include every section an investor expects to see. None of that matters if the underlying support, the contractor invoices, the plans and specs, the RS Means codes tied to specific property components, is not there.
The first IDR James issued in most cost segregation audits asked for two things: the purchase agreement and the cost segregation study. If the answers to those documents raised questions, the second IDR asked for the contractor's final application for payment and the construction plans and specifications. The study summary is not the support. The documentation behind the estimate is the support.
James told me that one of his consistent findings was studies where so much analytical focus went into identifying 1245 personal property that the 1250 structural component allocation became an afterthought. Incomplete 1250 classifications, where non-residential or residential structural elements were not properly broken out, created problems even when the 1245 property work was solid. The study had to be complete in both directions, not just aggressive on the 1245 side.
This is also where AI-generated studies fail at a structural level, not because the technology is unsophisticated, but because the IRS requires physical inspection of the property. A study produced without a site visit "goes against every IRS rule," James said. AI tools can serve a useful function as completeness checks, making sure the 1250 property side of a study is not underdeveloped, but classification decisions require an engineer who has been to the property. See our related piece on what the IRS says about AI-generated cost segregation studies for more on this.
Contingency Fees: A Flag the IRS Looks For Specifically
The Cost Segregation Audit Techniques Guide includes a note that IRS examiners "should closely scrutinize" studies prepared on a contingency fee basis. James confirmed this in our interview.
His first IDR typically asked for the engagement agreement. A provider who is being paid a percentage of the tax benefit created by the study has a financial incentive to push classifications in ways that may not survive examination. That structural conflict is not disqualifying, contingency arrangements are legal, but it is a flag. It changes how the examiner reads the study.
James said he never personally encountered a contingency arrangement that directly led to an adjustment just because of the fee structure. But the pattern he described, a provider promising to "save you X and take a percentage", made him scrutinize the underlying classification work more carefully.
If your cost seg provider is working on contingency, that is worth knowing. It does not mean the study is bad. It means the study needs to be better documented, because that documentation will be read more carefully if you are ever examined.
The Enforcement Capacity Question Going Forward
The narrative that IRS capacity is shrinking applies most accurately to broad individual return audits. It does not apply cleanly to LB&I specialist examinations of complex real estate transactions.
James described LB&I as focused on larger cases, properties at the larger end of small business / self-employed division, and entities in the LB&I range. Revenue agents in those cases often lack cost segregation knowledge. Engineers like James are brought in specifically because the tax law and construction cost methodology require technical knowledge that most revenue agents do not have.
The roughly 200 engineers James says he trained before retiring are the bench for those examinations. They are in the workforce now. General IRS headcount cuts reduce the number of routine individual audits. They do not eliminate the specialist pipeline for complex real estate examinations.
The IRS is also adding AI-assisted return selection. This does not make the examinations more frequent for everyone. It makes them more targeted. A system that uses AI to flag statistical anomalies in depreciation schedules will, by design, direct the available engineers toward returns that look unusual. A study with vague RS Means codes, aggressive 1245 allocations, and a missing land component is going to look unusual to a statistical model.
"The audit rate might not go down," James told me. "If you get caught in a trap, it's still a trap."
The trap is a study that cannot survive examination. The IRS has spent years building the capacity to find it.
What This Means If You Have a Cost Segregation Study
The practical implication is not that you should avoid cost segregation. The strategy works. The tax code permits it. The audit rate, even with a study, remains well under 0.1% for most businesses.
The practical implication is that study quality is not optional. A study that passes the IRS's examination framework, specific RS Means codes, proper land allocation, complete 1250 property classification, documentation that supports the estimate, is a defensible position. A study that does not pass that framework is exposure waiting for an audit to open.
James put it plainly: the number of IDRs a study generates tells the story. Fewer IDRs means the study is well-documented and the classifications are supportable. More IDRs means the examiner found gaps. The audit outcome follows from that.
If you want to understand whether your existing study would hold up, or how to evaluate a provider before commissioning a new study, our analysis of cost segregation red flags and how to spot a legitimate study covers the specific documentation standards James described. For a broader comparison of providers and what separates the credible firms from the rest, see our cost segregation company comparison.
The IRS has been building capacity in this area for the better part of two decades. James Peacock is a significant part of why. Before he retired, he made sure that capacity did not leave with him.
About the Expert
James C. Peacock spent nearly 39 years at the IRS as a General Engineer and Subject Matter Expert in the LB&I Division. He was among the first IRS engineers to examine cost segregation, contributed to the Cost Segregation Audit Techniques Guide from 2004 through the 2025 update, and served as the IRS's primary technical expert on Section 179D from 2014 through his retirement in September 2025. Per James, he trained approximately 200 new-hire IRS engineers on cost segregation and Section 179D methodology before leaving federal service.
For a free cost segregation estimate on your property, visit FreeCostSeg.com. For background on whether cost segregation triggered the audit risk your advisor described, see Does Cost Segregation Trigger an IRS Audit?, James covers the actual 0.07% vs 0.078% numbers in detail there.
