The Statutory and Regulatory Foundation of the Research Credit

The Research and Development Tax Credit is an incremental credit designed to reward taxpayers for increasing their investment in qualified research over a base amount. To qualify for the credit under Section 41(a), a taxpayer’s activities must satisfy a rigorous four-part test applied at the level of the “business component,” which is defined as any product, process, computer software, technique, formula, or invention intended to be held for sale, lease, or license, or used in the taxpayer’s trade or business. The four requirements are cumulative, meaning the failure of any single test disqualifies the activity from credit eligibility.

For the 2014 tax year at issue, the first requirement was that the expenditures be eligible for treatment under Section 174. This generally requires research in the experimental or laboratory sense directed at uncertainty about capability, method, or appropriate design. The technological information requirement calls for reliance on principles of physical or biological sciences, engineering, or computer science. The business component requirement asks whether the information is intended to be useful in developing a new or improved product, process, software, technique, formula, or invention. The historical Section 174 discussion in this study must be distinguished from the domestic research rules under Section 174A effective for tax years beginning after 2024.

The process of experimentation requirement was decisive in Little Sandy Coal. Treasury Regulation Section 1.-4(a)(6) requires 80% or more of the relevant research activities, measured on a cost or other consistently applied reasonable basis, to constitute elements of a process of experimentation. The process must evaluate alternatives to resolve uncertainty about capability, method, or appropriate design and relate to a permitted purpose: function, performance, reliability, or quality. Meeting this threshold does not override the other statutory requirements or exclusions.

Factual Underpinnings of the Little Sandy Coal Controversy

Little Sandy Coal Company, Inc., acting as the parent entity for its shipbuilding subsidiary, Corn Island Shipyard, Inc. (CIS), filed for research credits for the tax year ended June 30, 2014. The claim centered on the design and construction of 11 “first-in-class” vessels, which the taxpayer asserted were unique prototypes that had never been previously constructed by the yard. These vessels were built under contracts for third-party customers, and the taxpayer claimed significant qualified research expenditures (QREs) related to wages, supplies, and contract research.

The IRS disallowed the claimed credit and assessed a deficiency and an accuracy-related penalty. The Tax Court sustained both; only the credit issue was pursued on appeal. For trial, the parties treated two vessels as representative: the Apex 720 Tanker, Project 720, and the Detyens Dry Dock, Project 730. The central issue was whether the taxpayer substantiated that substantially all of the relevant research activities constituted elements of a process of experimentation.

The Apex Tanker was a tank barge designed based on a previous model, the Penn 80, but incorporated a unique stern notch and towing bridle intended to improve performance. The dry dock was an entirely new design for the shipyard, requiring complex engineering to ensure stability during the raising and lowering of vessels. The taxpayer’s primary argument was that because the vessels were novel and complex engineering projects, the vast majority of the time spent by engineers and production workers was inherently experimental. This “novelty heuristic” relied on the assumption that if the end product is new, the activities leading to its creation must satisfy the Section 41 requirements.

The Tax Court’s Approach to the Activity Test

On February 11, 2021, Judge James S. Halpern issued T.C. Memo. 2021-15. The Tax Court applied the regulatory 80% threshold and rejected the taxpayer’s reliance on the proportion of new physical features in the vessels. Its treatment of support activities later drew disagreement from the Seventh Circuit.

The court emphasized that the “substantially all” test applies to activities, not to the physical elements of the product being developed. Consequently, the taxpayer’s argument that because 80% or more of the vessel’s physical components were new, 80% of the activities must be experimental, was categorically rejected. The Tax Court construed the “substantially all” fraction as follows:

Substantially all ratio = research activities constituting elements of a process of experimentation ÷ relevant research activities. For the law applicable to this case, the denominator comprised research activities whose expenses were eligible under Section 174 and that were not excluded under Section 41(d)(4). Both parts must use a consistent measurement basis. The denominator is not automatically all project costs or all manufacturing labor.

Section 41(b)(2)(B) identifies qualified services as engaging in qualified research or directly supervising or supporting that research. The Tax Court used this expense classification to distinguish experimentation from supporting activities in the activity fraction. In particular, its analysis could include pilot-model production in the denominator while categorically excluding it from the numerator. The table describes that historical approach; the Seventh Circuit rejected the categorical exclusion.

The courts distinguished supply costs from research activities when discussing the fraction. Excluding supplies from this activity calculation does not categorically exclude supplies from qualified research expenses under Section 41(b). Supplies used in qualified research may qualify if the statutory conditions are met. In Little Sandy Coal, the failure to establish qualified research prevented recovery of the claimed expenses; the outcome was not a general prohibition on supply credits.

The Seventh Circuit’s Analysis: Disentangling Expenses from Activities

The Seventh Circuit affirmed the denial of the research credit in 2023 but disagreed with the Tax Court’s categorical treatment of certain support and supervision activities. The opinion is appellate precedent within the Seventh Circuit and may be persuasive elsewhere; it did not establish a single nationwide evidentiary rule for every court.

Judge Michael Brennan’s opinion explained that the distinction among expense categories under Section 41(b) should not be imported mechanically into the activity test under Section 41(d). Activities involving direct supervision or support may enter the numerator when they themselves constitute elements of the experimental process. Their inclusion depends on the activity and supporting evidence, rather than the employee’s title or expense label.

Pilot-model production may therefore contribute to the numerator when it forms part of a systematic evaluation of alternatives. Simply constructing a first-of-its-kind vessel is insufficient. The Seventh Circuit did not decide whether the two vessels actually were pilot models: it examined the evidence under both assumptions and found the taxpayer’s proof inadequate in either event.

The Fatal Flaw: Failure of Proof and the “Principled Way”

Little Sandy Coal lost because it did not establish a principled allocation of activities to qualifying experimentation for each vessel. The court required usable evidence of the activities and their relationship to the claimed business components. This was a failure of proof under the existing rules, not a new requirement that every taxpayer use a particular timekeeping system.

The taxpayer claimed $609,276 of estimated nonproduction wages across the 11 vessels without a vessel-by-vessel breakdown. A 60% allocation was used for lead engineer Bud Johnson, and testimony characterized various estimates as reasonable or fair. The court found that such assurances did not demonstrate the proportion of research activities constituting experimentation. Production time records also failed to establish the necessary distinction because the taxpayer relied heavily on whether vessel features were new.

Treasury Regulation Section 1.-4(d) requires records in sufficiently usable form and detail, without prescribing one exclusive format. Contemporaneous time records, design revisions, engineering analyses, testing records, and project studies can help support a consistent allocation. Retrospective evidence is not automatically barred, but unsupported percentages cannot establish eligibility. Estimating the amount of qualified expenses becomes relevant only after the taxpayer has established that qualified research occurred.

Rejection of the Novelty Argument and Trinity Industries

The Seventh Circuit criticized the novelty-based reasoning used by the district court in Trinity Industries, Inc. v. United States, 691 F. Supp. 2d 688 (N.D. Tex. 2010), affirmed at 757 F.3d 400 (5th Cir. 2014). It rejected the use of newness as a shortcut for measuring the proportion of experimental activities. This should not be described as the Seventh Circuit overturning Fifth Circuit precedent or as a rule that Trinity automatically allowed every activity associated with a new vessel.

New or complex products may involve ordinary fabrication and engineering as well as qualifying experimentation. The taxpayer must show which activities evaluated alternatives to resolve technological uncertainty. Conversely, welding, assembly, or model construction is not automatically ineligible when the evidence connects the activity to that experimental process. Integrated-system uncertainty can be relevant, but it must be demonstrated rather than inferred solely from the product’s novelty.

The Strategic Importance of the Shrink-Back Rule

The taxpayer’s vessel-level claim lacked evidence sufficient to establish the 80% requirement. That does not mean the courts calculated a proven experimentation percentage below 80% for each vessel. The evidentiary shortfall also prevented a supported alternative claim for smaller portions of the vessels.

Treasury Regulation Section 1.-4(b)(2) first applies the requirements to the discrete business component. If those requirements are not met, the analysis proceeds to its most significant subset of elements, continuing until the requirements are met or the most basic element is reached. Shrinking back does not guarantee that any subset qualifies, and it does not dispense with substantiation. The examples below distinguish the actual outcome from possibilities that would require further evidence.

The Seventh Circuit recognized that activities concerning the tanker’s stern notch and towing bridle and the dry dock’s outboard side plate might involve experimentation. It nevertheless found that the documentation lacked enough detail to establish eligibility for those subsets. Maintaining evidence at an appropriate subcomponent level can preserve a supported alternative analysis.

Implications for Specific Industries and Professional Practices

The activity-based reasoning is relevant to other businesses undertaking complex development, including aerospace, manufacturing, construction, and architectural and engineering practices. Its practical significance depends on the facts, the governing jurisdiction, and the particular activities claimed. The decision alone does not establish that claims in these industries are eligible or ineligible.

For architectural, engineering, and construction firms, records should identify the claimed business component and the uncertainty, alternatives, and evaluation process associated with it. A building system or structural element may be an appropriate subset where supported by the facts. Direct-supervision and direct-support expense rules remain separate considerations under Section 41(b) and Treasury Regulation Section 1.-2(c); Little Sandy Coal did not create a new “one-up, one-down” supervision rule. The following table presents documentation practices rather than a mandatory checklist imposed by the decision.

The Evolving Landscape: Form 6765 and Research Expense Deductions

The December 2025 Instructions for Form 6765 state that Section G business-component information is o