The Statutory Architecture of Innovation Incentives

For the historical research years addressed in Phoenix Design Group, Section 41 incorporated the research and experimental expenditure requirements of Section 174. Satisfying that threshold did not establish entitlement to the credit: the remaining requirements and exclusions of Section 41 also applied.

The statutory wording has since changed. For amounts paid or incurred in taxable years beginning after December 31, 2024, Section 41(d)(1)(A) refers to domestic research or experimental expenditures under Section 174A. Section 174A generally permits current deductions for domestic research expenditures, subject to applicable elections and transition provisions. The historical Section 174 terminology used in this study describes the law applicable to the cited disputes and should not be mistaken for the current deduction framework.

Technical uncertainty concerns whether the available information establishes the capability or method of developing or improving a product, or its appropriate design. Unknown customer preferences, project coordination issues, and commercial risk do not automatically establish this type of uncertainty. Conversely, uncertainty about appropriate design can qualify even when capability and method are established.

The credit also requires eligible expense categories, proper calculations, and consideration of exclusions, including funded research. Passing a single element of the four-part test cannot compensate for failing another.

Investigatory Activity and the Limits of Iterative Design

PDG relied on its engineering design process and the challenges encountered when developing building systems. The court required evidence of what information was unavailable and what the engineers actually did to resolve the identified uncertainties. Describing a design problem and its eventual solution left an evidentiary gap.

The distinction is between investigating technical uncertainty and applying available information to a known problem. Basic calculations using available data may be insufficient. However, using established scientific principles does not itself disqualify research: engineering experimentation ordinarily relies on such principles. Complexity, repeated calculations, and successive drawings must be assessed in context.

Investigatory activity under the historical Section 174 test is also distinct from the Section 41 process of experimentation. An email or meeting that obtains missing information may support the former without establishing the latter. A design revision can reflect experimental evaluation, a client preference, or ordinary coordination; its label alone does not resolve eligibility.

The table retains the source’s column heading for continuity; its case discussion concerns Phoenix Design Group, not a Peterson engineering ruling. Capability, method, and design are alternative forms of uncertainty, rather than three separate hurdles that every project must fail to resolve at its outset.

Documentation, Business Components, and the Shrinking-Back Rule

Useful substantiation explains the business component, the uncertainty, the alternatives considered, the evaluation performed, and the connection between that work and claimed expenses. Drawings and time records are more persuasive when witnesses explain their technical significance. Contemporaneous records are valuable, but there is no universal requirement to use one particular project-accounting system or timekeeping format.

In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the claim concerned eleven vessels, with a tanker barge and a dry dock selected as representative trial projects. It did not treat an entire group of barges as one business component. The Seventh Circuit affirmed disallowance because the taxpayer did not provide a principled basis for determining the relevant activity proportions, although it disagreed with aspects of the Tax Court’s reasoning.

The substantially-all test generally requires at least 80% of the relevant research activities for each business component, measured by cost or another consistently applied reasonable basis, to constitute elements of experimentation. It is not a test of how much of the finished product is new. Nor does reaching 80% establish eligibility if other requirements or exclusions are unsatisfied.

The source’s table heading references Peterson, but the relevant engineering case is Phoenix Design Group. Betz v. Commissioner, T.C. Memo. 2023-84, is a separate research-credit case cited in the PDG opinion.

Under Treasury Regulation § 1.41-4(b)(2), a component that fails the requirements as a whole must be examined at progressively smaller significant subsets where appropriate. This shrinking-back rule can identify qualifying work within a broader project. It does not dispense with proof of the activities at the smaller level, and it does not require every subcomponent to have been uncertain merely because one was.

Contractual Risk: The Funded Research Obstacle

Section 41(d)(4)(H) excludes research to the extent another party funds it. The analysis considers the full contractual relationship, including whether payment depends on successful research and whether the performer retains substantial rights in the results. Contract labels alone do not decide these questions.

The funding discussion attributed to Peterson in the source instead concerns separate authorities. In the December 18, 2024 summary-judgment order in Smith v. Commissioner, the Tax Court declined to resolve the IRS’s funding challenge to architectural projects on the existing record. The order addressed contractual interpretation and governing foreign law. It was an interlocutory ruling, not a final determination that the research credit was allowable.

A fixed price, milestone schedule, or obligation to correct errors may be relevant to financial risk, but none provides an automatic safe harbor. The contracts and applicable law must establish the parties’ actual obligations if the research fails. Likewise, ownership of drawings or copyrights should be distinguished from rights to use the underlying research results.

These are analytical indicators, not conclusive classifications. Increased experience alone is not a substantial research right under Treasury Regulation § 1.41-4A(d).

In the January 3, 2025 order in System Technologies, Inc. v. Commissioner, Docket No. 12211-21, the court denied the IRS’s motion for partial summary judgment on funded research. Indiana law preserved a refund remedy for nondelivery notwithstanding warranty limitations. That contract-specific reasoning did not establish that every refund remedy makes research unfunded or that all other credit requirements were satisfied.

The Danielson Rule and Contractual Characterization

Peterson v. Commissioner, 827 F.3d 968 (11th Cir. 2016), concerned Mary Kay retirement payments and self-employment tax. It did not involve PDG or its owners. Its discussion of the Danielson rule belongs to a separate line of authority about taxpayers challenging the tax consequences of their agreements.

Where applicable, the rule restricts attempts to disavow an unambiguous agreement without evidence that could alter its construction or establish unenforceability, such as fraud, duress, or mistake. Its application depends on the jurisdiction and circumstances. The Eleventh Circuit had applied this doctrine before Peterson; the case did not introduce it for research-credit contracts.

A contract describing work as consulting does not by itself preclude the research credit. The actual activities and substantive rights and obligations remain central. The practical lesson is to evaluate agreements accurately, especially their payment and research-rights provisions, without assuming that a service label resolves the tax treatment.

Penalty Defense and Issues of First Impression

The separate decision at 148 T.C. 463 (2017), also designated 148 T.C. No. 22, is Petersen v. Commissioner. It concerned Section 267, accrued expenses, and employees participating in an employee stock ownership plan. It was not a research-credit case. The court declined to impose accuracy-related penalties in light of the taxpayers’ good-faith efforts and the unresolved legal issue.

This does not create an automatic first-impression defense for R&D claims. Section 6664(c) and Treasury Regulation § 1.6664-4 generally require reasonable cause and good faith, assessed from all relevant circumstances. A professional study or an adviser’s involvement does not automatically meet that standard; the reasonableness of reliance and the information supplied matter.

In Phoenix Design Group, penalties followed an express stipulation tied to the outcome of the three trial projects. That result should not be portrayed as a successful penalty defense or as a ruling that every unsuccessful research-credit claim attracts a penalty. Firms should assess substantive eligibility and possible penalties separately.

Statistical Sampling and the Burden of Discovery

Kapur v. Commissioner, T.C. Memo. 2024-28, addressed a motion to limit discovery and trial to two projects in a research-credit dispute. The court denied the requested protective order. The taxpayers could not unilaterally narrow the inquiry to their selected projects while claiming credits across a much larger project population.

The decision did not prohibit statistical sampling or decide the merits of every underlying research activity. It illustrates the importance of agreement on sampling scope and the continuing burden to substantiate the credit claimed. A sample used to calculate a claim does not necessarily control the evidence the IRS may seek in litigation.

Where a firm intends to use sampling, it should retain information about the underlying population and discuss the methodology and evidentiary scope early. In the absence of agreement, broader discovery may be required. The appropriate approach depends on the claim, available records, procedural rules, and the court’s directions.

Practical Insights from Research-Credit Litigation

Professional Qualifications and Activity Content

Engineering qualifications help establish expertise, but credit eligibility depends on the activities performed. Phoenix Design Group reinforces the need to explain technical work in terms that show how the statutory requirements are met. It does not mark a change from a former rule under which professional titles were sufficient.

Novelty and Experimentation Are Distinct

A unique building, vessel, or system can result from established methods without qualifying experimentation. Conversely, an improvement need not be new to the industry. Claims should explain the uncertainty and evaluative process rather than rely on novelty, sophistication, or the commercial importance of the finished product.

Evidence of an Evaluative Process

Modeling, simulation, and systematic trial and error may demonstrate experimentation when connected to an identified uncertainty. The regulations do not prescribe a universal laboratory protocol or require every project to include a documented failure. Useful evidence can include:

• Computational fluid dynamics models explaining the alternatives evaluated.
• Design comparisons recording technical objectives, assumptions, and results.
• Testing records and revisions explaining why a proposed solution was accepted or rejected.

Peer review and meetings can contribute evidence, but their occurrence alone does not prove experimentation. The study must explain what was evaluated and how that evaluation addressed the uncertainty.

Contracts and Research Rights

Contract review should focus on enforceable payment conditions, risk allocation,