The federal credit for increasing research activities under Section 41 of the Internal Revenue Code (IRC) encourages qualifying research through tax credits for eligible expenditures. This study examines substantiation issues affecting engineering, architectural, and manufacturing businesses. The source draft attributes engineering research-credit findings to Gantt v. Commissioner, but that attribution could not be substantiated. The described engineering decision is Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113. The analysis below corrects that attribution and compares the engineering case with Little Sandy Coal Co. v. Commissioner. These decisions apply existing eligibility and proof requirements; they do not establish a universal prohibition on estimates or oral testimony.

Legislative and Regulatory Framework of the Section 41 Credit

Section 41 limits the credit to qualified research and eligible expense categories. The four-part test generally applies separately to each business component. Historical cases discussed here apply the Section 174 rules in force for their tax years. For amounts paid or incurred in taxable years beginning after December 31, 2024, the amended Section 41(d)(1)(A) instead refers to domestic research or experimental expenditures under Section 174A. Deduction or capitalization treatment and research-credit eligibility are distinct questions; satisfying the expenditure test alone does not establish credit eligibility.

The Business Component and the Four-Part Test

A business component is any product, process, computer software, technique, formula, or invention which is to be held for sale, lease, or license, or used by the taxpayer in a trade or business. The eligibility of expenses associated with these components depends on satisfying the following cumulative criteria:

The “substantially all” requirement within the Process of Experimentation (PoE) test is perhaps the most frequent point of failure in modern litigation. It requires that at least 80% of the taxpayer’s research activities, measured by cost or another reasonable basis, involve elements of a process of experimentation for a qualified purpose—such as improving function, performance, reliability, or quality.

Correcting the Case Attribution: Phoenix Design Group v. Commissioner

The source does not provide a verifiable research-credit decision supporting its account of Gantt v. Commissioner. The engineering facts and holdings described instead correspond to Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, decided December 23, 2024. Phoenix Design Group performed mechanical, electrical, plumbing, and fire-protection engineering design. The court found the evidence insufficient to establish qualified research for the sampled projects. The discussion should therefore be understood as an analysis of Phoenix Design Group, not a verified R&D holding in Gantt.

The Section 174 Challenge: Uncertainty vs. Calculation

In Phoenix Design Group, the taxpayer relied on its engineering design work and calculations to support research-credit eligibility. The relevant inquiry was whether the work addressed uncertainty concerning capability, method, or appropriate design and whether the activities satisfied the separate process-of-experimentation requirement.

Routine calculations and communication of design results do not, by themselves, establish a process of experimentation. However, the use of existing engineering principles does not automatically disqualify research: the regulations expressly allow research relying on existing technologies and do not require an advance in the state of scientific knowledge. The question is what uncertainty existed and how the taxpayer evaluated alternatives to resolve it. A qualified process can use modeling, simulation, or systematic trial and error; a conventional laboratory experiment is not mandatory.

The Failure of the Process of Experimentation (PoE) Test

Phoenix Design Group illustrates the need for evidence connecting actual project activities with the process-of-experimentation test. The regulatory framework calls for:

1. Identifying the technological uncertainty at the outset.
2. Identifying one or more alternatives intended to eliminate the uncertainty.
3. Conducting a systematic evaluation of those alternatives (e.g., through modeling, simulation, or systematic trial and error).

Final designs alone do not explain whether or how the taxpayer evaluated alternatives to resolve technical uncertainty. A persuasive evidentiary record connects the uncertainty, alternatives considered, evaluation methods, and results. Engineering records and credible testimony should be evaluated together; the law does not prescribe one mandatory hypothesis worksheet or time-log format.

Comparative Analysis of Little Sandy Coal and Phoenix Design Group

Little Sandy Coal and Phoenix Design Group address different factual records but illustrate the distinction between performing technically demanding work and proving that the statutory research requirements were satisfied. Their outcomes should not be described as concurrent rulings or a newly enacted documentation standard.

The “All or Nothing” Strategy and the Shrinking-Back Rule

In Little Sandy Coal, a shipbuilding company claimed expenses for eleven “first-in-class” vessels. The taxpayer argued that because the vessels were novel, “substantially all” of the activities involved in their design and construction must have been experimental. The court dismissed this “novelty argument,” ruling that the 80% test must be applied to activities, not the physical novelty of the product.

Treasury Regulation § 1.41-4(b)(2) provides the shrinking-back rule: if the requirements are not met for the overall business component, they are applied to the most significant subset of elements and then to progressively smaller subsets until a qualifying subset is reached or the smallest subset is considered. Little Sandy Coal lacked a sufficient evidentiary basis to apply that rule to potentially experimental subcomponents. The rule does not automatically preserve part of a credit, and the source does not establish a corresponding holding in Gantt.

The “Design Spiral” and Systematic Evaluation

In Little Sandy Coal, experts referred to the shipbuilding process as a “design spiral,” where interdependent components require iterative adjustments. While this spiral resembles experimentation, the Seventh Circuit held that the taxpayer must still offer a “principled way” to determine which portion of those iterations constitutes qualified research as opposed to routine engineering. The court warned that “shortcut estimates” and “arbitrary allocations” of employee wages would not suffice.

Funded Research and the Allocation of Financial Risk

The funded-research exclusion under Section 41(d)(4)(H) is a separate limitation for contract research. It excludes research to the extent funded by another person. The source’s assertion that this was a secondary holding in Gantt could not be verified and should not be relied upon. Contract funding must be evaluated independently of whether the underlying activities pass the four-part test.

The Two-Pronged Funding Test

Treasury Regulation § 1.41-4A(d), incorporated by § 1.41-4(c)(9), requires analysis of payment arrangements and the taxpayer’s retained rights:

Financial Risk: Payments contingent on the success of research are not treated as funding under the regulation. Noncontingent payments can constitute funding, but the exclusion applies to the extent funded. Where substantial rights are retained, otherwise qualified expenses exceeding allocable funding may remain eligible. Fixed-price terms or the risk of ordinary cost overruns alone do not settle the question.

Substantial Rights: The researcher must retain substantial rights in the research results. Exclusive ownership is not required. If the researcher retains no substantial rights, its research is treated as fully funded for this purpose. That does not automatically give the customer a credit; the customer must independently satisfy the applicable contract-research and other Section 41 requirements.

Milestone billing alone does not establish that payment depends on successful research, and no verified Gantt summary-judgment ruling supports that proposition. Contract terms, acceptance conditions, remedies, and governing law must be examined together. A ruling denying summary judgment because facts remain disputed is not a final determination that the research qualifies.

Intellectual Property and Local Law

Intellectual-property ownership and rights to use research results depend on the contract and applicable law. Delivering drawings or granting a client a license does not necessarily eliminate the designer’s substantial rights, but neither does default copyright ownership automatically establish eligibility. The source’s broad description of Smith should not be treated as a rule that architectural contracts need no review. Transfer provisions, confidentiality duties, reuse restrictions, and remedies require project-specific analysis.

Substantiation and the Limits of the Cohan Rule

For decades, many taxpayers and R&D consultants relied on the Cohan rule to salvage credits when documentation was sparse. The Cohan rule, derived from the 1930 case Cohan v. Commissioner, allows a court to estimate a deduction if the taxpayer can prove an expense was incurred but lacks exact records.

When Estimates Can Be Used

Little Sandy Coal did not abolish the Cohan rule or all reasonable estimates. It distinguished estimating the amount of qualified research expenses from proving that qualified research occurred. A taxpayer must first establish eligibility and provide a rational evidentiary basis for any amount the court is asked to estimate.

An estimate cannot replace proof of the four-part test. Unsupported departmental percentages are vulnerable when they do not identify qualifying activities or reliably connect expenses to them. Records need not take a prescribed form, and credible testimony can contribute to substantiation. Contemporaneous technical and financial records generally make the evidentiary foundation stronger.

Implications for Future R&D Tax Credit Applications

These decisions reinforce the need for research-credit studies that connect technical activities, business components, and expenses. Separately, the IRS screens amended-return research-credit refund claims for sufficient information. That administrative validity review is distinct from a substantive determination that research qualifies. It should not be described as a new judicial test or an automatic denial system created by Gantt.

The Shift from “What” to “How”

A description such as “We built a new bridge with unique structural requirements” does not, by itself, establish qualified research. The study should explain what technical uncertainty existed and how engineering principles were used to evaluate alternatives—for example, alternative structural designs addressing uncertain load-bearing performance. This illustrates application of the existing test, rather than a replacement of an earlier novelty-only standard.

The “Elements of PoE” Standard

Activities that may form part of a qualifying process, depending on their purpose and supporting evidence, include:

• Prototyping and pilot model production.
• Simulation iterations and software testing.
• Testing logs and analysis of failure.
• Hypothesis revision and redesign cycles.
• Supervisory review of experimental results.

Little Sandy Coal explains that direct support and direct supervision are not categorically excluded from the substantially-all calculation. Inclusion depends on whether the particular activities meet the relevant research requirements and constitute elements of the experimental process. General management or routine production does not qualify merely because it occurs in a research department. This is separate from the employee-level wage rule under Section 41(b).

Strategic Compliance Recommendations for the AEC and Manufacturing Sectors

Architectural, engineering, construction, and manufacturing firms should assess whether their existing project and accounting records adequately support the claims they make. Practical improvements should focus on the actual gaps in evidence.

Documenting Technological Uncertainty at the Outset

Record the uncertainty concerning capability, method, or appropriate design as research begins whenever practicable. Explain why available information did not resolve it and distinguish technical questions from aesthetic preferences or commercial uncertainty. Contemporaneous documentation is useful evidence, but no single prescribed pre-project form is universally required.

Adopting a Modular, Iterative Proof Trail

Identify the business component first, and preserve sufficient detail about relevant subsets to apply the shrinking-back rule if nec