The Statutory Framework and the Evolution of the Four-Part Test

Section 41 imposes separate requirements for qualifying research activities and eligible expense categories. The regular credit generally measures qualifying expenses against a historical base amount; alternative calculation methods have different rules. The four-part test must be applied separately to each business component, subject to the statutory exclusions.

For the years litigated in Trinity, the first threshold referred to expenses eligible for treatment under Section 174. Research in the experimental or laboratory sense seeks to resolve uncertainty about capability, method, or appropriate design in connection with the taxpayer’s trade or business. For taxable years beginning after December 31, 2024, the amended Section 41(d)(1)(A) instead refers to Section 174A, governing domestic research or experimental expenditures. Deductibility and credit eligibility remain distinct questions.

The second threshold is the technological information test, which requires that the research be undertaken for the purpose of discovering information that is technological in nature. This means the process must fundamentally rely on principles of the physical or biological sciences, engineering, or computer science. Historically, the “discovery” of this information was subject to a more restrictive “discovery test” which required the information to exceed, expand, or refine the common knowledge of skilled professionals in the field. However, Treasury Decision 9104 in 2004 clarified that research does not require the taxpayer to expand the common knowledge of the field; it merely requires the intent to eliminate uncertainty within the taxpayer’s own development process.

The third threshold is the business component test, where the application of the discovered information must be intended to be useful in the development of a new or improved business component. A business component is broadly defined as 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 fourth threshold requires substantially all of the relevant research activities to constitute elements of a process of experimentation for improved function, performance, reliability, or quality. Treasury Regulation Section 1.41-4(a)(6) sets an 80% threshold, measured on cost or another consistently applied reasonable basis. The analysis concerns research activities, not simply the proportion of a product that is new. Experimentation requires identifying uncertainty and alternatives and evaluating those alternatives through a process such as modeling, simulation, or systematic trial and error. Novelty or uncertainty alone is insufficient.

Factual Underpinnings of Trinity Industries, Inc. v. United States

Trinity Industries sought research-credit refunds for tax years ending in March 1994 and March 1995. The disputed expenditures arose from Trinity Marine Group’s shipbuilding operations. The six projects considered in the district court’s first phase ranged from substantially new designs to modifications of existing designs. First-in-class vessels served as prototypes, but their status did not independently establish qualification.

Substantiation was difficult because the expenditures were remote in time, relevant computer systems were no longer used, Trinity had spun off the business and no longer held the records, and Hurricane Katrina had destroyed many records. Trinity relied on available accounting evidence and testimony. It did not supply evidence permitting meaningful allocation of costs to smaller vessel components, leaving the court to decide qualification at the whole-vessel level.

Analysis of the Six Shipbuilding Projects and the 80% Threshold

The district court’s 2010 decision, 691 F. Supp. 2d 688, evaluated six projects on the evidence Trinity presented. Without a basis for estimating qualifying costs of smaller components, Trinity’s whole-vessel approach made the 80% threshold decisive. The following table describes that phase of the litigation, rather than every project considered later.

The Mark V involved demanding speed, stealth, transport, and operational requirements. Trinity built two prototypes, using aluminum and Kevlar hulls. The court found that more than 80% of the overall prototype costs were incurred in experimentation and qualified research. The hull design drew on the experimental P050; describing it as wholly without any design antecedent would overstate the finding.

The Dirty Oil Barge required extensive redesign to address double-hull and stability requirements associated with the Oil Pollution Act of 1990. Although double hulls were already known, their use changed the barge’s structure, weight distribution, piping, and heating arrangements. The court found that the relevant development costs satisfied its 80% analysis. A regulatory requirement itself does not establish qualified research.

The court recognized qualifying research within the four unsuccessful projects, but Trinity did not establish qualification for each vessel as a whole. For XFPB and T-AGS 60, the court found the evidence insufficient to establish the 80% threshold; it did not calculate an exact routine-work percentage. For the Crew Rescue Boat and Hurley Dredge, the court found the experimental work below that threshold. Hurley also failed on the alternative ground that it adapted an existing component to a customer’s needs. Missing component-level cost evidence prevented a narrower award.

Legal Precedents: Custom Manufacturing and the Integration of Systems

Trinity provides a fact-specific example of how custom manufacturing and system integration can satisfy the research-credit rules.

The government argued that special-order ships were not held for sale because they were built for particular customers. The district court rejected that restriction and treated the vessels as business components. Custom manufacture therefore does not itself prevent qualification, but all other requirements and exclusions still apply.

The court also rejected the suggestion that integrating existing systems necessarily amounts to routine selection from a menu. Systems may interact in ways that create design uncertainty and require experimentation. Conversely, installing a familiar component in a first-in-class ship does not automatically constitute experimentation. Suder v. Commissioner later discussed the importance of evaluating integration work on its facts.

The Shrink-Back Rule and the Substantiation Crisis

Treasury Regulation Section 1.41-4(b)(2) provides for shrinking back when the overall business component fails the qualification requirements. The analysis proceeds to the most significant subset of elements and continues until a qualifying subset is found or the most basic element fails. The rule did not fail as a legal mechanism in Trinity; the evidence did not permit its application.

The court could not estimate qualifying expenses below the vessel level because Trinity supplied no meaningful evidentiary basis for doing so. A project that falls below 80% is not automatically entitled to a credit equal to its estimated experimental percentage. A smaller component must independently satisfy the applicable tests, and its eligible expenses must be substantiated. Failure of proof also does not establish that no research occurred.

Comparative Jurisprudence: Trinity vs. Union Carbide, Suder, and Little Sandy Coal

The Trinity precedent must be synthesized with other landmark rulings to understand the full scope of modern R&D tax credit standards.

Supply Cost Qualification and the Union Carbide Standard

Trinity’s treatment of qualifying vessel costs should not be read as a universal full-cost prototype rule. Expense eligibility remains a separate inquiry under Section 41(b), including limitations on wages, supplies, and contract research. In Union Carbide Corp. v. Commissioner, 697 F.3d 104 (2d Cir. 2012), the court affirmed denial of ordinary production-input costs claimed alongside process experiments. The case distinguishes research-related additional supplies from costs that would have arisen in routine production; it does not make all materials used during an experimental production run eligible.

Wage Substantiation and the Suder Reasonableness Test

In Suder v. Commissioner, T.C. Memo. 2014-201, the Tax Court considered testimony and supporting evidence in determining research activities and employee time allocations. It also reduced the compensation amount eligible for the credit under the applicable reasonableness requirement. The lesson is to substantiate both the research allocation and the eligible expense amount. Neither testimony nor estimation eliminates the need for a credible factual foundation.

The Little Sandy Coal Refinement and the Numerator Dispute

Little Sandy Coal Co. v. Commissioner, T.C. Memo. 2021-15, affirmed, 62 F.4th 287 (7th Cir. 2023), also concerned shipbuilding. It rejected reliance on product novelty or first-in-class status as a substitute for proof of experimentation. The Tax Court’s treatment of direct-support activities raised a separate question about the numerator of the substantially-all fraction.

The Seventh Circuit rejected categorical exclusion of direct-support and supervision activities from the numerator, but affirmed the denial of the credit. Such activities must themselves satisfy the research-activity requirements to enter the fraction and constitute elements of experimentation to enter its numerator. Inclusion is not automatic for every support or supervisory activity. The taxpayer still lacked a principled basis for distinguishing experimentation from routine work.

The Consistency Rule and the Fifth Circuit Appeal

Section 41(c)(6) requires consistent determination of credit-year and base-period QREs. Trinity’s appeal concerned whether the district court applied that requirement consistently when it treated credit-year vessels on a whole-vessel basis but used a different approach for base-period vessels.

In Trinity Industries, Inc. v. United States, 757 F.3d 400 (5th Cir. 2014), the Fifth Circuit vacated the judgment and remanded for further proceedings. The ruling did not simply approve Trinity’s proposed base-period reduction. Consistency and the taxpayer’s burden of proof still had to be addressed.

The figures below distinguish the amended-return percentages from the expert’s calculation. The proposed $26,706,987 reduction figure in the source cannot be paired with a 1.2847% fixed-base percentage and unchanged gross receipts: those numbers are mathematically inconsistent. A lower base can affect the credit, but a proposed adjustment is not an adjudicated refund entitlement.

Administrative Implications: Form 6765 and Section G

Form 6765 now requests more business-component detail. This should not be described as a proven direct response to Trinity or Little Sandy Coal. Under the December 2025 instructions, Section G is optional for tax years beginning before 2026 and required for years beginning after 2025, subject to exceptions.

Exceptions include qualified small businesses electing the payroll tax credit and original-return filers meeting both the $1. million QRE ceiling and the $50 million average-gross-receipts ceiling under the prescribed rules.

Business-component detail: Required filers apply the 80%/Top 50 selection rule and aggregate remaining components. This filing threshold is separate from the experimentation test and does not replace the shrink-back analysis.

Expense detail: Section G separates wage categories, supplies, and contract research. The information-sought description in column 49(f) currently applies to amended returns.

Amended returns: Separate research-credit refund-claim requirements must also be considered; optional Section G treatment does not eliminate those requirements.

Officer wages: Section E separately asks for officer wages included in QREs.

These requirements increase the value of component-level records but do not mean that every filer must attach all supporting evidence to an original return.

Future Outlook: Implications for Large-Scale Innovation

Trinity shows that engineering development may qualify even when a project uses established technologies. It also shows how insufficient cost segregation can prevent a credit for experimental work within a larger project.

Shipbuilding, aerospace, construction, and other engineering businesses should evaluate their own technical uncertainties, alternatives, and experimental activities. Custom design and integration may support qualification, but neither establishes eligibility by itself.

Contemporaneous records are valuable because they connect technical work to the expenses claimed. Later reconstruction and reasonable estimates are not categorically prohibited; their acceptability depends on the governing law and the evidence available. Project records, design changes, testing results, payroll allocations, and supply records can provide a more defensible basis than broad retrospective assertions.

Final Thoughts

Trinity’s enduring lesson is to evaluate research at an appropriate business-component level and preserve evidence supporting both qualification and cost allocation. Shrink-back analysis can identify narrower qualifying work when a larger project fails, but it cannot supply missing proof. Consistent base-period calculations and attention to applicable filing requirements complete that analysis.