Why the IRS Lost in Harper — And What It Means for Your Architecture or Engineering Firm's R&D Tax Credit

If you’re an architect, engineer, or design-build firm owner who has ever been told “your work probably doesn’t qualify for the R&D tax credit,” the Harper Construction case is worth your attention.

The IRS went after a design-build contractor’s Section 41 research credit claim — and the Tax Court stopped them cold at summary judgment. Here’s why that matters, what the court actually said, and how firms like yours can use this ruling to claim credits confidently.

What Happened in Harper?

Harper Construction Company claimed the federal R&D tax credit (Section 41) for technical work done on its design-build projects. The IRS pushed back hard, arguing that Harper’s work failed the very first test — the “business component” requirement — and moved to toss the case before trial.

The Tax Court denied the IRS’s motion. In doing so, the court made two points that are critical for any A/E or design-build firm:

Designs don’t have to be a “product” to qualify. Under Section 41, a business component can be a process, technique, or invention — not just something you sell. That opens the door for the work architects and engineers do every day.
The IRS can’t invent extra tests. The court rejected the IRS’s attempt to layer a special “meaningful use” standard onto the statute. The plain language — “used in a trade or business” — is enough.

In short: the court told the IRS that design and construction work can involve real technical development, and that “business component” analysis can fit design-build work when framed correctly.

That’s good news. But winning a court case and winning an audit are two different things.

Why Most A/E Firms Still Lose — Or Never Claim at All

The Harper decision doesn’t mean every engineering project automatically qualifies. The court said the IRS’s attack failed — not that Harper’s documentation was perfect.

This is where most firms fall short. The three most common failure points:

Vague business component definitions. “We did engineering” isn’t a business component. A qualifying component needs to be specific: a design methodology for solving a constructability constraint, an engineering process for meeting structural or MEP performance requirements, a repeatable technique for value engineering or sequencing. Without that specificity, the IRS finds an easy foothold.
No trail of technical uncertainty or iteration. The Tax Court in Harper noted the project lifecycle — conceptual design through construction documentation — as evidence of iterative development. That same narrative needs to live in your credit study. What was uncertain? What alternatives did you evaluate? What did you revise and why?
Cost allocations that don’t hold up. The most defensible credits aren’t lost on the technical argument — they’re lost because the taxpayer can’t explain who did what, on which projects, and how wages were allocated. “Our engineers spend about 40% of their time on qualified work” isn’t an allocation methodology. It’s an invitation for an audit adjustment.

How NTG + TaxDrone.ai Solves Each of These Problems

We built TaxDrone.ai specifically for the A/E and design-build space, because the generic R&D credit process doesn’t fit the way these firms actually work.

We Map the Right Business Components

Rather than floating a claim around broad descriptions, we force every credit into a business-component framework the IRS can’t easily attack:

Design methodologies for solving constructability, performance, or code-compliance constraints
Engineering processes for structural, MEP, civil, or energy performance requirements
Repeatable techniques for BIM coordination, value engineering, cost modeling, or field sequencing
Internal tools — scripts, calculation templates, custom workflows — developed for the firm’s own use

This directly mirrors the Tax Court’s logic in Harper: designs may qualify as a process, technique, or invention.

We Capture Iteration and Uncertainty the Way the IRS Expects to See It

For every business component, we document:

The technical uncertainty at the outset (what capability or approach was in doubt)
The alternatives evaluated (schematic options, modeling runs, VE scenarios, structural systems considered)
The evaluation and testing performed (analysis, simulation, mock-ups, field validation)
The iteration trail (revisions, RFIs, submittal cycles, redlines, and what drove the changes)

This translates normal project work — the stuff your team does on every job — into the “attempt to eliminate uncertainty” narrative that survives scrutiny.

We Anchor the Claim to Your Firm's Asset, Not the Owner's Building

One of the IRS’s arguments in Harper was that the contractor didn’t “own” the building being improved. While the court didn’t let that win at summary judgment, the lesson is clear: don’t anchor your credit to the owner’s asset when you can anchor it to your own business component.

We structure claims so they’re tied to the firm’s design and delivery process — not the client’s structure. That directly neutralizes the “held for sale / owned” framing the IRS likes to use.

We Produce Filing-Ready Documentation

The IRS has published explicit expectations for research credit refund claims — business component lists, activity descriptions, QRE totals, and cost methodology. TaxDrone.ai produces a substantiation package that meets those requirements on day one:

A business-component inventory by project and component type
Activity descriptions per component (what was done and why it qualifies)
QRE totals with a documented, defensible cost methodology
Supporting artifacts: drawings, calculations, models, RFIs, revision histories

We Build Allocations That Actually Hold Up

This is where most credit studies fall apart. Our approach uses:

Smart sampling and workflow evidence — not back-of-envelope estimates
Role-based activity mapping — PM, PE, designer, BIM coordinator, energy modeler, each mapped to qualified activities
Clear cost nexus — wages tied to specific components and activities
Year-over-year consistency — so your methodology is repeatable, not reinvented each filing season

The Bottom Line for A/E and Design-Build Firms

Harper is an encouraging case because it confirms what we’ve always known: the technical work architects and engineers perform every day — working through uncertain design problems, evaluating alternatives, iterating toward a solution — is exactly the kind of activity Congress intended to incentivize with Section 41.

The IRS knows this too. That’s why they’ll keep challenging claims that aren’t properly structured, documented, and defended.

NTG + TaxDrone.ai operationalizes the Harper framework: we define the right components, build the uncertainty and iteration narrative, and produce the documentation and allocations the IRS expects to see.

Your firm is likely doing work that qualifies. The question is whether your credit study can prove it.

Want to see how this applies to your specific project types?

Share the 3–5 project categories your firm handles most — site civil, structural retrofits, MEP redesigns, energy modeling, prefab detailing, or others — and we’ll build out a component framework with example qualified activities you can use right away.

Contact TaxDrone.AI to Get Started →