The federal research credit under Section 41 rewards qualifying research, but technical complexity alone does not establish eligibility. This study examines substantiation, experimentation, cost allocation, and customer contracts using identifiable authorities, including Betz v. Commissioner, Little Sandy Coal Co. v. Commissioner, and Phoenix Design Group, Inc. v. Commissioner.
The Legal Framework of Section 41 and the Four-Part Test
Qualified research must satisfy each statutory requirement, and claimed costs must separately fall within eligible expense categories. A project’s commercial value or engineering difficulty does not, by itself, establish entitlement to the credit. The statutory exclusions must also be considered.
The Structural Requirements of Qualified Research
| Test Component | Statutory Basis | Legal Standard and Objective |
|---|---|---|
| Research expenditure test | IRC § 41(d)(1)(A) | Research must satisfy the applicable research-expenditure cross-reference. Current law refers to Section 174A; historical claims require the law applicable to their tax years. |
| Technological in Nature Test | IRC § 41(d)(1)(B)(i) | The information sought must be technological in nature. |
| Process of Experimentation Test | IRC § 41(d)(1)(C) | Substantially all research activities must be elements of a process of experimentation for a qualified purpose. |
| Permitted Purpose Test | IRC § 41(d)(3) | The purpose must concern improved or new function, performance, reliability, or quality. |
The business component is the unit of analysis. Failure at the broadest component level does not automatically eliminate every smaller component; the shrinking-back rule may apply.
Correcting the Case Attribution
T.C. Memo. 2023-47 is Gage v. Commissioner, a decision concerning a settlement-payment deduction and related penalties. It is not the Yeh research-credit decision described in the source text. The alleged Yeh holdings about hydrogen units, employee time records, and experimentation could not be verified and should not be relied upon.
Betz v. Commissioner, T.C. Memo. 2023-84, concerns Catalytic Products International, an S corporation that designed and supplied air-pollution-control systems, including catalytic and thermal oxidizers. The disputed 2014 research credit was $501,531. The court found that the taxpayers failed to establish pilot-model treatment for production expenditures and failed to substantiate certain employee wages as qualified services across 19 projects. Five projects also failed because the company lacked substantial rights under customer contracts. The court sustained accuracy-related penalties. These findings should not be relabeled as holdings from Yeh or expanded into a general prohibition on engineering research.
Research Expenditures and Technical Uncertainty
The older cases applied the Section 174 framework then in force. Legislation enacted in 2025 added Section 174A for domestic research or experimental expenditures and amended Section 41’s cross-reference. Section 174A generally permits a current deduction for domestic expenditures for tax years beginning after 2024, with an alternative capitalization election. Deduction treatment and research-credit eligibility remain separate questions.
Capability, Method, and Design Uncertainties
| Category of Uncertainty | Judicial Requirement | Evidence Failure in Yeh |
|---|---|---|
| Capability Uncertainty | Whether the taxpayer can achieve the intended technical result. | No verified Yeh finding; identify the actual capability question in the project records. |
| Method Uncertainty | How the intended technical result can be achieved. | No verified Yeh finding; explain the methods evaluated and why. |
| Design Uncertainty | Which design is appropriate for the intended technical result. | No verified Yeh finding; connect design alternatives to technical evaluation. |
Under Treasury Regulation § 1.41-4, design uncertainty can exist even when capability and method are known. Professional experience does not automatically eliminate uncertainty, and a subsequent design change does not automatically prove it existed. The relevant question concerns the information available when the research began.
The Process of Experimentation
Beyond Routine Engineering and Calculations
Treasury Regulation § 1.41-4 permits evaluation through modeling, simulation, or systematic trial and error. It does not prescribe a universally required formal hypothesis document. Records should establish the uncertainty, alternatives, and evaluation actually performed. Calculations and meetings may support that explanation, but naming them does not establish experimentation.
The Separate Substantially-All Tests
Two 80% rules must be distinguished. The business-component experimentation test addresses the proportion of research activities constituting elements of experimentation, measured by cost or another consistently applied reasonable basis. The employee wage rule in Treasury Regulation § 1.41-2(d)(2) addresses qualified services over the taxable year. Qualified services include performing qualified research and qualifying direct supervision or direct support. Meeting the component test does not automatically qualify every employee’s wages.
Under the employee rule, wages may receive full treatment when at least 80% of the employee’s services are qualified services. Otherwise, only the properly allocable portion qualifies. Time records are useful, but the regulation does not impose a universal requirement for daily timesheets or mathematically exact reconstructions.
Documentation Standards and the Cohan Rule
The source text’s assertion that Yeh abolished estimation under Cohan is unsupported. Section 274(d) is not a blanket substantiation rule for research credits. In Little Sandy Coal, the Seventh Circuit explained that estimation requires an adequate evidentiary foundation; a court need not invent an allocation from unsupported assumptions. Establishing that some technical work occurred does not prove all statutory requirements or the amount of eligible expenses.
The Limits of After-the-Fact Reconstructions
Contemporaneous evidence can make a claim easier to substantiate, but later testimony is not categorically inadmissible or worthless. Its usefulness depends on specificity, credibility, and its connection to the work and costs at issue. Treasury Regulation § 1.41-4(d) requires records sufficient to substantiate eligibility, rather than a single mandatory record format.
| Evidence Type | Substantiation Value post-Yeh | Judicial Reasoning |
|---|---|---|
| Contemporaneous Lab Notes | Potentially strong; no verified post-Yeh ranking. | Can identify testing, alternatives, observations, and dates. |
| Email Correspondence | Potentially useful. | Can explain technical decisions when the surrounding context is retained. |
| Time-Tracking Logs | Useful for supported allocations. | Descriptions must connect time to qualifying services; a project code alone may be insufficient. |
| After-the-fact Interviews | Depends on credibility and supporting evidence. | Can explain records but should not substitute unsupported percentages for factual analysis. |
| General Testimony | Limited when vague. | Specific explanations of activities are more informative than conclusions that all work was research. |
The table’s references to Yeh identify the original comparison labels, not a recognized judicial evidence-ranking system.
The Shrinking-Back Rule: Mechanics and Limitations
Treasury Regulation § 1.41-4(b)(2) permits successive examination of smaller components when the overall business component does not meet the requirements. The rule does not waive the tests or turn an arbitrary cost percentage into qualified research. A practical study should identify the smaller component, the qualifying activities, and a defensible allocation. There is no verified basis for attributing a failed hydrogen-unit shrinking-back analysis to Yeh.
Funded Research and Rights to the Results
Section 41(d)(4)(H) excludes funded research. Treasury Regulation § 1.41-4A(d) addresses entitlement to payment and retention of substantial rights. Contract labels alone do not decide eligibility. A fixed-price arrangement requires examination of the actual obligations and payment contingencies; exclusive ownership of intellectual property is not universally required for the contractor to retain substantial rights.
| Contractual Element | Impact on R&D Credit Eligibility | Judicial Interpretation |
|---|---|---|
| Payment Contingency | Relevant to financial risk. | Examine whether entitlement to payment depends on successful research. |
| IP Ownership | Relevant to substantial rights. | Analyze retained use rights; legal title and exclusivity are not the sole tests. |
| Milestone Payments | Depends on operative terms. | Review whether payment is earned independently of research success and any repayment obligations. |
Betz illustrates the importance of examining individual contracts: the substantial-rights finding applied to five projects. It should not be generalized to every client-funded engineering engagement.
Comparative Analysis of Verified Research-Credit Decisions
Little Sandy Coal and the 80% Rule
Little Sandy Coal Co. v. Commissioner, T.C. Memo. 2021-15, was affirmed by the Seventh Circuit in 62 F.4th 287 (2023). The dispute involved shipbuilding projects. The appellate court rejected reliance on the novelty of physical vessel features as a substitute for proving qualifying activities. It also disagreed with aspects of the Tax Court’s treatment of direct support and supervision while affirming the outcome. The decision therefore requires attention to both the result and the appellate reasoning.
Phoenix Design Group and Accuracy-Related Penalties
Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, addressed three agreed, nonbinding trial projects involving building engineering systems. The court held that none entailed qualified research and sustained penalties under the parties’ stipulations. The source’s description of a new, automatic negligence penalty for unsuccessful research claims overstates the decision. Penalty exposure requires analysis of the applicable law, facts, and defenses.
IRS Refund-Claim Review and Form 6765
Refund-Claim Screening
The IRS reviews research-credit refund claims for required information before considering their merits. Its published FAQ does not establish the automated risk-scoring system described in the source. Effective June 18, 2024, the IRS waived the initial submission of individual researcher names and the information each person sought to discover, although those details may be requested during examination. The transition period allowing 45 days to perfect deficient claims extends through January 10, 2027. Filing sufficiency does not establish substantive eligibility.
Changes to Form 6765
The December 2025 Form 6765 instructions make Section G optional for tax years beginning before 2026 and required for later tax years subject to exceptions. For filers required to complete it, the instructions generally call for business-component detail covering at least 80% of total qualified research expenses or a maximum of 50 components, with remaining components aggregated. This filing convention is separate from both substantive 80% tests. The form does not universally require detailed experimental narratives for every component; apply its year-specific instructions and separate amended-return requirements.
Implications for Future R&D Tax Credit Applications
Documenting Technical Evaluation
A useful project record explains what was uncertain, which alternatives were considered, what evaluation occurred, and what was learned. Preserve existing design revisions, test results, and decision records. Avoid inserting scientific terminology into a narrative unless it accurately describes the work.
Activity-Level Time Tracking
Use activity descriptions that permit reviewers to distinguish research, qualifying support, and ordinary production or administration. When estimates are necessary, retain the underlying records, assumptions, and method so that someone unfamiliar with the project can understand the allocation.
Integrating Technical and Tax Workflows
Engineers should explain technical facts while tax personnel map those facts to the eligibility rules and expense categories. Reconcile the resulting study to payroll and accounting records, and address contradictory evidence before filing. An interview summary should accurately reflect the participant’s knowledge and uncertainty.
Reviewing Contractual Rights and Risks
Review executed agreements, statements of work, change orders, and acceptance terms. Identify who bears unsuccessful-development costs and which research results the contractor can use. Do not assume that a later amendment changes the substance of an earlier arrangement.
