The R&D Tax Credit · Guide · Working level
The research credit for architecture and engineering firms: design phases, contracts, and the funded-research minefield
How AEC work maps to Section 41: design development versus construction documents, performance-based and energy-modeling work, why client contract terms decide most claims after Populous Holdings, and a project-by-project table of qualifying and non-qualifying work.
Architecture and engineering firms sit in an odd position under Section 41: their daily work is engineering — explicitly a qualifying technological discipline — yet their claims fail more often on contract language than on science. The credit reaches design work that resolves genuine uncertainty about the capability, method, or appropriate design of a building system through evaluation of alternatives. Whether the firm or its client owns that research for credit purposes depends almost entirely on how the engagement agreement allocates payment risk and rights to the design — the funded-research question that Populous Holdings v. Commissioner (T.C. 2019) resolved in favor of a fixed-fee architect.
Where the four-part test bites in design work
The four-part test applies to A&E work the same as anywhere: a permitted purpose (new or improved product or process — a building design is a product of the firm's business), technological in nature (engineering and physical science, easily met), elimination of uncertainty, and a process of experimentation. The last two do the sorting.
Uncertainty must be technical — capability, method, or appropriate design — not aesthetic or economic. A firm choosing among three proven curtain-wall systems on cost has no Section 41 uncertainty. A firm that cannot know at the outset whether a long-span roof can meet deflection limits within the architectural envelope, and that models candidate systems to find out, does. Treas. Reg. §1.41-4(a)(5) frames appropriate-design uncertainty exactly this way, and the regulations are at eCFR Title 26. The process of experimentation is the systematic part: identifying alternatives, modeling or calculating each, and eliminating candidates against defined criteria. BIM clash iterations, structural option studies, CFD and energy models, and mock-up testing all fit; producing a code-minimum design by lookup does not.
One exclusion deserves early mention: Section 41(d)(4) removes research relating to style, taste, cosmetic, or seasonal design factors. Massing studies driven by appearance, material palettes, and interior aesthetics are out even though architects experiment with them constantly. The line is function versus appearance — a façade study to hit a solar heat gain target is technical; the same study to hit a look is not.
Design phases: where qualification concentrates
Typical qualification by project phase, assuming an unfunded contract:
| Phase | Share of typical fee | Qualification pattern |
|---|---|---|
| Pre-design / programming | 5–10% | Rarely — requirements gathering, not experimentation |
| Schematic design | 15% | Often — alternatives generated and screened against uncertain constraints |
| Design development | 20% | The core — systems selection, performance modeling, iterative resolution |
| Construction documents | 40% | Mostly not — recording decided designs; discrete technical problems can qualify |
| Bidding / negotiation | 5% | No — administrative |
| Construction administration | 20% | Rarely — RFI responses that require new engineering analysis can, routine review does not |
The concentration matters for claim size: even a strong project usually yields qualified time from perhaps a third of the fee, not the whole engagement. Claims that sweep entire project fees into QREs — every hour from kickoff to closeout — are the A&E equivalent of the manufacturer claiming every machinist, and they draw the same adjustment. Time records that distinguish option studies and analysis from production drafting are what separate a defensible claim from an estimate; wage QRE mechanics are covered in the QRE guide.
Performance-based design and energy modeling
Two genres of A&E work qualify unusually cleanly. Performance-based design — seismic performance-based engineering, fire engineering by computational modeling rather than prescriptive code, wind engineering of irregular structures — replaces code tables with analysis precisely because the prescriptive path cannot answer the design question. The uncertainty is definitional and the modeling is the experiment. Iterative energy modeling qualifies when the model drives design: a team testing envelope, orientation, glazing ratio, and systems alternatives against an aggressive EUI target, discarding schemes the model kills, is running a process of experimentation.
The contrast case: a single energy model run on a completed design to certify compliance — including modeling performed to substantiate a Section 179D deduction — is documentation, not research. The same software, opposite answer. What distinguishes them is whether the model's output could still change the design.
The funded-research minefield: Populous and the contract file
Section 41(d)(4)(H) excludes research funded by any grantee, contract, or otherwise — and every A&E project is performed under a client contract, so every project must survive the analysis in Treas. Reg. §1.41-4A(d). Two independent prongs, both of which the firm must win:
- Economic risk. Is payment contingent on the success of the research? A fixed fee means the firm eats redesign, failed schemes, and abortive analysis — risk stays with the firm. Payment "regardless of outcome" makes the research funded.
- Substantial rights. Does the firm retain the right to use the research results — the design approaches, details, and engineering knowledge — in its business? A firm that keeps its instruments of service or at least a license to reuse its methods retains substantial rights; a contract assigning all work product and know-how exclusively to the client does not.
Populous Holdings v. Commissioner is the touchstone. The Tax Court held that the architecture firm's five fixed-price design contracts did not fund its research: the firm was paid for a completed design deliverable, not for research as such, bore the cost of its own failed and reworked design effort, and retained rights to reuse the knowledge gained. The Populous case brief walks the reasoning; the funded research exclusion guide covers the doctrine generally, including Fairchild's inspection-and-acceptance analysis that fixed-fee firms borrow.
How common A&E contract terms score on the two prongs:
| Contract feature | Risk prong | Rights prong |
|---|---|---|
| Lump-sum / fixed fee for defined deliverables | Firm bears risk — favorable | Neutral; check IP clause |
| Hourly, billed monthly, no acceptance conditions | Client bears risk — unfavorable | Neutral |
| Hourly with client acceptance/approval gates and rework at firm's cost | Mixed; acceptance terms can restore firm risk | Neutral |
| AIA-style ownership: architect retains instruments of service, client gets license | — | Firm retains rights — favorable |
| Full assignment of all work product, methods, and know-how to client | — | Unfavorable; may fail rights prong |
| Guaranteed maximum price on design-build | Firm bears risk above GMP — favorable | Check flow-down IP terms |
| Errors-and-omissions rework obligations | Supports firm risk | — |
Fixed-fee versus hourly: the economics of the same project
The funding analysis makes contract form worth real money. Consider a $1.5M structural engineering engagement on a long-span venue, with $450,000 of the fee attributable to qualifying design-development work (wages plus overheadless QRE wages of, say, $300,000):
Illustrative. Assumes $300,000 of qualifying wage QREs before the funding screen; hourly-with-acceptance outcomes depend on the specific rework and acceptance terms.
The middle bar is deliberately uncomfortable: hourly contracts are not automatically fatal — Fairchild Industries v. United States (Fed. Cir. 1995) located risk in inspection-and-acceptance terms rather than billing cadence — but they put the firm in a clause-by-clause fight rather than a Populous safe harbor. Firms that care about the credit increasingly negotiate fee structure and IP terms with Section 41 in mind, which is legitimate planning so long as the terms are real.
Qualifying and non-qualifying project work: examples
Representative A&E activities sorted by how they typically fare under Section 41(d):
| Activity | Typically qualifies? | Why |
|---|---|---|
| Structural option studies for a 180-foot clear-span roof with untested geometry | Yes | Appropriate-design uncertainty; systematic evaluation of alternatives |
| Iterative energy modeling to reach a net-zero target, redesigning envelope between runs | Yes | Process of experimentation driving design |
| Performance-based seismic design using nonlinear response-history analysis | Yes | Prescriptive path unavailable; analysis is the experiment |
| Foundation design alternatives on a site with unprecedented soil conditions | Yes | Method uncertainty resolved through modeling and testing |
| BIM development of a novel prefabrication approach the firm has never executed | Yes | Process business component; capability uncertainty |
| Producing construction documents from an approved design development set | No | Recording decisions; no remaining uncertainty |
| Code-prescriptive HVAC sizing from ASHRAE tables | No | Known method, certain outcome |
| Façade material selection driven by appearance | No | Style/taste exclusion, §41(d)(4) |
| Confirmatory energy model for a 179D certification on a finished design | No | Compliance documentation, not experimentation |
| Construction administration site visits and routine RFI responses | No | Post-design activity |
| Adapting the firm's standard school prototype to a new site with routine changes | No | Adaptation exclusion, §41(d)(4)(B) |
Building the A&E claim
The order of operations mirrors any industry claim, with the contract screen moved to the front:
- Pull every active project's agreement and score it on risk and rights. Projects that fail the funding screen exit before anyone reads a timesheet.
- For surviving projects, identify the business components — the building, and often discrete systems (structure, envelope, MEP) analyzed separately under the shrink-back rule.
- Isolate qualifying phases and activities; apply the style/taste, adaptation, and routine-engineering screens honestly, and memo the projects excluded.
- Build wage QREs from time records at the activity level; capture supplies (physical models and mock-ups consumed in testing) and any subconsultant work that passes its own rights-and-risk read at 65%.
- Reconcile to the ledger and report at the business-component level on Form 6765 Section G.
The IRS research credit overview frames the general rules, but for design firms the claim is won or lost in two files: the contract and the timesheet. Firms with neither in order should fix the records before claiming — an A&E claim built on firm-wide percentages and unread agreements is the profile examiners are trained to find.
Frequently asked questions
- Can architecture and engineering firms claim the R&D tax credit?
- Yes. Design firms can claim the Section 41 credit for work resolving genuine technical uncertainty — novel structural systems, performance-based design, energy modeling iterations, unusual site or code constraints solved through analysis of alternatives. The two recurring obstacles are the funded-research exclusion, which turns on client contract terms, and separating conceptual and technical design from routine document production.
- Does a fixed-fee contract help an A&E firm's research credit claim?
- Generally yes. Under the funded-research rules a firm paid a fixed fee bears the economic risk of its design work — if the design fails or needs rework, the firm absorbs the cost. In Populous Holdings v. Commissioner the Tax Court found fixed-price architectural contracts left risk with the designer, so the research was not funded. Hourly contracts billed regardless of outcome point the other way, though inspection-and-acceptance and rework clauses can restore risk.
- Which design phases qualify for the R&D credit?
- Qualification concentrates in schematic design and design development, where alternatives are generated and evaluated against uncertain structural, mechanical, or performance requirements. Construction documents largely record decisions already made and mostly do not qualify, though discrete technical problems solved during CDs can. Construction administration, bidding, and permitting are generally excluded as post-design or administrative activity.
- Is energy modeling qualified research?
- Iterative energy modeling to resolve design uncertainty — testing envelope, glazing, and systems alternatives against a performance target where the appropriate design is unknown at the outset — is a process of experimentation and can qualify. Running a single confirmatory model on a finished design to document code compliance or support a 179D deduction is routine engineering and does not.