The federal research and development (R&D) tax credit under Internal Revenue Code section 41 requires taxpayers to establish both qualifying research activities and eligible expenses. Architectural and engineering work can qualify, but technical complexity, professional expertise, or repeated design revisions alone do not establish eligibility. This study examines contractual funding, technical substantiation, and changes to research-cost deductions and filing requirements.
Two distinct questions recur in these disputes: whether the activities satisfy the statutory research tests, and whether customer funding excludes otherwise qualifying research. These questions must be evaluated separately. The governing rules do not require academic research or an advance beyond knowledge available throughout an industry.
The Statutory Nexus of Section 41 and Sections 174 and 174A
The cases discussed here applied the research-expenditure provisions applicable to their historical tax years. Research uncertainty concerns capability, method, or appropriate design, judged against information available to the taxpayer. A business component may be a product, process, computer software, technique, formula, or invention held for sale, lease, or license, or used in the taxpayer’s business.
Qualified research must satisfy the expenditure test, be technological in nature, seek information useful in developing or improving a business component, and involve the required process of experimentation for an eligible purpose. The permitted purposes concern function, performance, reliability, or quality. Passing the expenditure test alone does not establish entitlement to the credit.
The Tax Cuts and Jobs Act generally required capitalization of research expenditures for tax years beginning in 2022 through 2024, with five-year amortization for domestic research and fifteen-year amortization for foreign research. Public Law 119-21, enacted July 4, 2025, added section 174A, generally allowing current deductions for domestic research expenditures in tax years beginning after 2024. Foreign research remains subject to fifteen-year amortization under section 174. The legislation also provided elections concerning previously capitalized domestic expenditures; eligibility and procedural requirements must be considered separately. Revenue Procedure 2025-28 provides implementation guidance.
| Feature | Section 174 Expenditure | Section 41 Research Credit |
|---|---|---|
| Primary Requirement | Research or experimental expenditures under the applicable deduction rules; domestic expenditures now generally fall under section 174A. | Satisfaction of the four-part test and the expense rules. |
| Timing of Benefit | Historical domestic capitalization rules must be distinguished from current section 174A deductions; foreign research generally remains amortizable over fifteen years. | Credit utilization depends on applicable tax limitations, carryover rules, and any eligible payroll tax election. |
| Documentation Focus | Nature, amount, timing, and location of research expenditures. | Qualifying activities and their connection to claimed expenses. |
| Statutory Exclusions | Separate limitations and exclusions apply; deductible research expenditures are broader than credit-eligible expenses. | Section 41(d)(4) exclusions include funded research and research conducted outside the permitted geographic area. |
The Smith Litigation: Navigating the Funded Research Doctrine
The source incorrectly identifies Sarah and Frank Weinstein as parties to the consolidated AS+GG cases. Smith v. Commissioner, T.C. Memo. 2026-50, concerns partners Adrian Smith, Carlisle Gill, and Robert Forest and their spouses, in Docket Nos. 13382-17, 13385-17, and 13387-17.
The June 16, 2026 opinion found that payments under all six trial contracts were not contingent on research success. AS+GG retained substantial rights under four contracts, potentially permitting credits to the extent research expenses exceeded funding. It retained no substantial rights under the other two. The four-part test had been conceded; the court did not decide it through a contested technical analysis. Earlier procedural success therefore cannot be presented as an unrestricted substantive victory.
Contractual Risk and the Contingent-on-Success Standard
Under Treasury Regulation section 1.41-4A(d), incorporated into the funding rules by section 1.41-4(c)(9), the analysis considers all relevant agreements. Payments contingent on research success are not treated as funding. Where substantial rights are retained but payment is not contingent, the regulations govern the extent of the funding reduction. A fixed fee, milestone schedule, or exposure to cost overruns does not independently establish that payment depends on successful research.
In System Technologies, Inc. v. Commissioner, the Tax Court’s January 3, 2025 order denied the IRS’s motion for partial summary judgment. Indiana law supplied a refund remedy if the manufacturer failed to deliver the promised product, despite a limited repair-or-replacement warranty. The court treated that remedy as making payment contingent on success. The order concerned the funding issue and did not establish satisfaction of every requirement for the credit.
The Retention of Substantial Rights
Retaining experience or general professional knowledge is insufficient. The taxpayer must retain substantial rights to use the research results without paying for those rights. Shared rights may qualify, but ownership, licenses, confidentiality restrictions, and reuse limitations must be evaluated together. Copyright ownership or silence in a contract does not automatically resolve this inquiry.
| Case Name | Taxpayer Industry | Key Issue | Result | Judicial Reasoning |
|---|---|---|---|---|
| Smith / AS+GG | Architecture | Funded Research | 2026 ruling limited potential eligibility. | Payments were noncontingent; substantial rights survived under four of six contracts. |
| System Technologies | Industrial finishing systems | Funded Research | IRS partial-summary-judgment motion denied. | Indiana law provided a refund remedy for nondelivery. |
| Meyer, Borgman & Johnson | Structural engineering | Funded Research | Taxpayer loss, affirmed in 2024. | Contractual obligations did not make payment contingent on successful research. |
| Populous Holdings | Architecture | Funded Research | Taxpayer prevailed on funding. | The particular fixed-fee agreements allocated relevant risk to the firm; fixed pricing is not a universal safe harbor. |
The Phoenix Design Group Case: Technical Substantiation
Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, addressed mechanical, electrical, plumbing, and fire-protection engineering. The December 23, 2024 opinion found no qualified research in three trial projects. The parties expressly agreed that those findings were not binding on the remaining projects. Accordingly, describing this opinion as denying every claimed credit in full overstates its scope.
The penalty consequence followed a stipulation: if none of the trial projects qualified, accuracy-related penalties would apply for credit years with deficiencies determined by the court or agreed by the parties. This should not be described as a general holding that engaging a research-credit consultant necessarily fails a reasonable-cause defense.
Identifying Technical Uncertainty
The court distinguished unresolved technical questions from information supplied by clients, ordinary design choices, and calculations using available data. The taxpayer needed to connect the claimed uncertainty to investigative work. Uncertainty about one subsystem did not automatically establish uncertainty throughout the entire project.
The practical inquiry is what information was missing, why it mattered to capability, method, or design, and how the taxpayer investigated it. The use of established engineering principles does not itself disqualify research: qualifying experimentation ordinarily relies on those principles. Likewise, software simulations may qualify when used to evaluate alternatives to resolve technical uncertainty.
Design Iteration and the Process of Experimentation Test
A standard project workflow does not by itself prove experimentation. Revisions may result from a client’s preferences, coordination, cost constraints, or experimental findings. The record should explain the reason for a revision and the evaluation that produced it.
A useful project record captures:
- The technical uncertainty concerning capability, method, or appropriate design.
- The alternative or alternatives evaluated.
- The modeling, simulation, testing, or systematic trial-and-error activities performed.
- The results and their effect on the next design decision.
These are practical documentation categories, not a requirement to create an academic paper or use a particular form. Treasury Regulation section 1.41-4 recognizes evaluation of one or more alternatives; multiple prototypes or a failed experiment are not invariably required.
The Reliability of Testimony and Documentation
Technical testimony is more persuasive when supported by project records that explain what employees actually did. Generic entries such as design or coordination may leave a gap between the identified problem and the final solution.
Relevant support may include design histories, emails, calculations, test results, meeting notes, and credible explanations from the people who performed the work. Maintaining records during the project is prudent, but the regulations do not prescribe one exclusive record format. The evidence must be sufficiently usable and detailed to substantiate eligibility and expenses.
The Substantially All Test and the Shrinking-Back Rule
The experimentation test generally requires at least 80% of the relevant research activities, measured on a cost or another consistently applied reasonable basis, to constitute elements of a process of experimentation. This is an activity test, not a claim that 80% of a product must be new. It is also distinct from the business-component disclosure threshold on Form 6765.
In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the court affirmed disallowance but rejected the categorical exclusion of pilot-model production activities from the experimentation numerator. Such activities can count when the evidence establishes their role in experimentation. The taxpayer nevertheless failed to substantiate the required activity allocations. Labels such as production or direct support do not determine the outcome without examining the actual work.
Application of the Shrinking-Back Rule
If the tests are not met for a whole business component, the analysis proceeds to its most significant subset and, when necessary, to smaller subsets. This can preserve eligibility for a qualifying subsystem even when a whole project does not qualify.
The taxpayer must identify the subset, demonstrate qualifying activities, and substantiate the related expenses. A particular timekeeping system is not legally mandatory, but unsupported allocations cannot fill evidentiary gaps. The rule does not simply reclassify an employee, a project phase, or a percentage of a project as a qualifying subcomponent.
Implications for R&D Tax Credit Applications
Contract review, technical substantiation, and return preparation should use consistent project information. Changes to a filing form do not themselves expand or narrow the statutory definition of qualified research.
Revisions to Form 6765 and Section G
The December 2025 instructions make Section G optional for tax years beginning before 2026 and generally required thereafter, with exceptions. Exemptions include qualifying small businesses making the payroll tax election and original-return filers meeting both the $1.5 million QRE and $50 million average-gross-receipts limits under the specified aggregation rules.
| Form 6765 Section | Data Requirement | Implementation Phase |
|---|---|---|
| Section E | Business-component count, officer wages, and other information. | Complete as directed for the applicable return. |
| Section G | Component identifiers, classifications, and expense details. | Generally required for tax years beginning after 2025, subject to exceptions. |
| Qualitative Info | Column 49(f) research-activity information. | Required for applicable amended claims; not required for timely original returns, including extensions. |
Where Section G applies, disclose components covering at least 80% of QREs, limited to 50 components, in descending expense order; aggregate the remainder. Follow the instructions for controlled groups and other special situations. Separate amended-claim requirements still apply.
IRS Classification and Refund-Claim Review
IRS guidance describes classifiers who review research-credit refund claims and request additional information through established procedures. It does not support the source’s claim that an autonomous Classifier system evaluates every claim before any human review.
For applicable claims postmarked on or after June 18, 2024, the IRS waived two previously required submission items: the names of individuals performing each activity and the information each individual sought to discover. Claims must still identify relevant business components, research activities for each component, and total qualified wage, supply, and contract-research expenses. The waived information may still be requested during examination.
The IRS’s January 2, 2026 procedural update extends the transition period for perfecting deficient claims through January 10, 2027. The specified procedures provide 45 days to respond to a request for missing information. Filing a complete claim initially remains important; an opportunity to perfect a claim does not establish substantive eligibility.
Practical Implications for Industry Practitioners
- Review actual contractual rights: Examine payment conditions, acceptance provisions, remedies, applicable law, and reuse rights. Milestone wording alone does not establish eligibility.
- Explain the investigative work: Record the technical question, alternatives considered, evaluations performed, and resulting decisions.
- Connect activities to expenses: Reconcile project evidence with payroll and other eligible costs using a supportable allocation method.
- Engage technical staff: Ask employees to explain their actual work and distinguish experimentation from routine implementation, client coordination, and other excluded activities.
- Apply the correct tax-year rules: Separate historical capitalization requirements, current domestic deductions, credit computation, and amended-claim procedures.
Final Thoughts
Research-credit eligibility depends on the work performed, the taxpayer’s rights and financial exposure, and the evidence supporting the claim. A favorable ruling on one issue does not establish entitlement on every other issue. Complex engineering can involve qualified research, but the connection must be demonstrated.
For architectural, engineering, and manufacturing firms, the practical lesson is to preserve clear project evidence and evaluate each requirement independently. Reliable records and careful contract analysis support defensible claims without treating every design revision as experimentation or every fixed-fee engagement as unfunded research.
