The consolidated cases of Mark Betz and Christine Betz v. Commissioner and Dennis Lincoln and Julia Lincoln v. Commissioner, decided as Betz v. Commissioner, T.C. Memo. 2023-84, concerned research credits claimed through Catalytic Products International, Inc. (CPI). Judge Nega issued the Tax Court’s opinion on July 6, 2023. CPI, an Illinois S corporation that designed and supplied air pollution control systems, claimed a $501,531 research credit for 2014 in connection with nineteen projects. The shareholders’ cases also involved credit carryforwards and accuracy-related penalties for 2015 and 2016.

The court denied the claimed credit. The taxpayers did not establish that the systems were pilot models or that the claimed employee wages related to qualified services. For five projects, the court also found that CPI had not retained substantial rights in the research results. This study examines those findings and their implications for documenting research credit claims. The decision applies existing statutory and regulatory requirements to CPI’s evidence; it does not establish that custom engineering is categorically ineligible or that all retrospective estimates are prohibited.

Historical and Statutory Context of the Research Credit

The federal research credit originated in 1981 and became permanent under the Protecting Americans from Tax Hikes Act of 2015. Its eligibility requirements distinguish qualified research from general product development, routine engineering, and other excluded activities. Reliance on engineering or other qualifying sciences does not, by itself, establish eligibility. Conversely, the credit does not require an advance in the knowledge of an entire industry.

For the 2014 tax year examined in Betz, Section 41 required research expenditures to be eligible for treatment under the then-applicable Section 174. The uncertainty inquiry concerned whether information available to the taxpayer established the capability, method, or appropriate design of the product. Meeting that threshold did not, by itself, satisfy the remaining research credit requirements.

The statutory framework must be applied for the relevant tax year. Section 174A, enacted in 2025, permits deductions for domestic research and experimental expenditures in tax years beginning after December 31, 2024, subject to applicable rules and elections. The historical Section 174 discussion in Betz should not be treated as a complete statement of present-day deduction rules.

The Factual Matrix of Catalytic Products International

CPI was founded in 1969. Beginning in 1987, it moved away from manufacturing catalysts and toward designing and supplying custom air pollution control systems, primarily catalytic and thermal oxidizers. In 2014, Mark Betz and Julia Lincoln each owned half of the corporation.

CPI’s systems treated hazardous airborne manufacturing byproducts. Catalytic oxidizers used a catalyst, while thermal oxidizers relied on high temperatures. Customer requirements included airflow volumes, pollutant composition, destruction efficiency, energy consumption, and installation constraints.

CPI typically developed proposals using customer specifications and established calculations. Its personnel used spreadsheets and hand calculations to size components, prepared drawings, and sought subcontractor bids. Fabrication, installation, and performance testing followed. The taxpayers argued that unresolved design questions and testing supported research treatment, including treatment of complete systems as pilot models. The court required evidence connecting those assertions to the particular activities and costs claimed.

Judicial Analysis of the Section 174 Uncertainty Test

The taxpayers emphasized uncertainty about the appropriate design of each system. However, the possibility that a completed unit might need adjustments did not establish that the entire unit had been produced to resolve developmental uncertainty.

The court considered CPI’s accumulated experience, established design tools, procurement decisions, and testing practices. Those facts undermined the claim that the systems as a whole were experimental models. Describing a project as complex or customized was insufficient to establish the necessary uncertainty.

Calculations and engineering tools are not inherently disqualifying. Their significance depends on whether they implement an established design or evaluate alternatives to resolve an identified uncertainty. Similarly, knowing that a product can be built does not necessarily resolve a separate uncertainty about its appropriate design.

The Limits of Post-Production Testing as Evidence

The court did not accept final performance testing as proof that the design of each complete system remained uncertain until testing ended. Testing that checks compliance with customer specifications may be quality control rather than experimental development. The purpose of the testing and the uncertainty it addresses therefore matter.

CPI’s commitment to expensive, specifically sized components before final testing also informed the court’s assessment. That factual inference does not create a general rule that procurement always ends research. Evidence may establish that uncertainty remains during construction or that a particular component requires further experimental development.

The Process of Experimentation and the 80% Rule

Section 41 requires substantially all of the relevant research activities to constitute elements of a process of experimentation for a permitted purpose. The regulatory threshold is 80%, measured on a cost basis or another consistently applied reasonable basis, and applied separately to each business component.

A taxpayer should be able to explain the uncertainty, the alternatives evaluated, the evaluation methods, and the results. Modeling, simulation, and systematic trial and error may qualify. Technical work and repeated design revisions do not establish experimentation without evidence of what the activities were intended to evaluate.

In Betz, the taxpayers failed to establish that the claimed wages related to qualified services. The decision supports careful activity-level substantiation, but it should not be presented as a universal requirement for a formal laboratory protocol or a particular document titled “hypothesis.” General descriptions of engineering challenges cannot substitute for evidence of qualifying work.

Direct Supervision and the Numerator of Experimentation

In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the Seventh Circuit affirmed the denial of research credits. It disagreed with parts of the Tax Court’s reasoning, including categorical restrictions affecting direct supervision and support, but did not reverse the judgment in the taxpayer’s favor.

Supervision and support may constitute elements of experimentation when the underlying activities satisfy the applicable requirements. Job titles do not determine the answer. The taxpayer still needs a defensible allocation of activities to each business component and a consistent basis for calculating the fraction.

This research activity test is separate from the employee wage rule that can treat all of an employee’s wages as qualified when substantially all of that employee’s services are qualified services. Passing one test does not automatically satisfy the other.

Pilot Model Analysis and the Supply QRE Disallowance

CPI sought to treat production costs for complete systems as research expenditures on the theory that the systems were pilot models. The taxpayers failed to prove that characterization for all nineteen projects, so the claimed production costs did not satisfy the threshold research expenditure requirement.

A pilot model is produced to evaluate and resolve uncertainty concerning a product. It need not be a small prototype, discarded after testing, or withheld from sale. A commercially delivered product can qualify if the relevant requirements are established. Customer orders and eventual sale therefore do not independently disqualify a model.

The difficulty in Betz was evidentiary: the taxpayers did not establish that producing the systems as a whole served the required experimental purpose. Customization and final acceptance testing did not bridge that gap. Even where pilot-model treatment is established under the research expenditure rules, the separate requirements of Section 41 must still be met.

The Application of the Shrinking-Back Rule

The shrinking-back rule permits evaluation of the most significant subset of elements of a business component when the component as a whole does not meet the qualified research requirements. The analysis continues at narrower levels as appropriate.

For example, experimental work on a burner or control subsystem might require a different analysis from production of an entire oxidizer. In Betz, the evidence did not support a qualifying narrower claim. Businesses should preserve records that allow specific uncertainties, activities, and costs to be evaluated at the relevant component or subcomponent level.

The Funded Research Exclusion and Substantial Rights

Section 41(d)(4)(H) excludes research to the extent another person funds it. Contract analysis considers both whether payment is contingent on successful research and whether the taxpayer retains substantial rights in the results. Retaining rights does not, by itself, resolve the separate payment-risk inquiry.

The court examined eight CPI projects for the funded research issue and found that CPI lacked substantial rights in five. Contractual ownership and restrictions on using research results were central to that analysis. Incidental gains in experience did not establish substantial rights.

Exclusive ownership is not necessary to retain substantial rights, and a confidentiality clause does not automatically eliminate them. The full agreement and the taxpayer’s legally retained ability to use the results must be examined. Generic labels such as “fixed price” or “work for hire” cannot replace that review.

Risk of Failure vs. Economic Risk

In Meyer, Borgman & Johnson, Inc. v. Commissioner, decided by the Eighth Circuit on May 6, 2024, the court affirmed a funded research determination. The contracts did not expressly or by clear implication condition payment on research success. Professional standards, inspection provisions, and obligations to correct defective work did not establish the required contingency.

The practical distinction is between ordinary risks of providing professional services and the contractual risk of unsuccessful research. Neither performing engineering for a customer nor receiving payment for a design automatically determines the outcome. The contractual terms and surrounding facts control.

Substantiation, Documentation, and the Cohan Rule

CPI did not maintain employee time records that established the qualifying activities claimed. Its research credit study relied on retrospective estimates, and the evidence did not establish that the claimed wages were paid for qualified services.

The Cohan rule can permit reasonable estimation where a taxpayer first proves entitlement to a tax benefit and provides a basis for an estimate. It does not allow a court to assume that research qualified or to accept unsupported percentages. Betz should therefore be understood as a failure of proof, rather than a categorical prohibition on estimates or employee interviews.

The Value of Contemporaneous Tracking

Records created while development work is occurring can strengthen a claim. Useful records include:

• Project and activity records linking employee effort to identified business components.
• Design alternatives, calculations, test plans, results, and explanations of revisions.
• Technical correspondence showing what was uncertain and how alternatives were evaluated.
• Payroll, purchasing, and accounting records supporting the amounts claimed.
• Contracts addressing payment contingencies and rights in research results.

No single timekeeping platform or document format is universally mandated by Betz. The requirement is to retain sufficiently usable and detailed evidence to substantiate eligibility and costs. Interviews and reasonable allocations are more persuasive when supported by reliable underlying records.

The Adaptation Exclusion: Defining the Limits of Engineering

Section 41(d)(4)(B) excludes research related to adapting an existing business component to a particular customer’s requirements. The M&W Ireland project in Betz illustrated the problem of claiming research for a system closely based on a previously supplied design, with limited changes for its installation.

That analysis does not establish that all work performed to customer specifications is excluded. A taxpayer must distinguish adaptation of existing technology from qualifying development of a new or improved business component. The business component may be a product, process, software, technique, formula, or invention within the statutory definition; it is not automatically the designer’s “design process.”

Applying established calculations to implement familiar specifications can support an adaptation finding. By contrast, a substantiated development effort that meets the statutory tests requires its own factual analysis. Industry labels and project novelty are insufficient on their own.

Accuracy-Related Penalties and Professional Reliance

The court sustained accuracy-related penalties under Section 6662(a) for 2014, 2015, and 2016. The taxpayers did not establish the reasonable-cause and good-faith defense. Engaging a credit consultant and a return preparer did not automatically establish reasonable reliance on professional advice.

A reliance defense depends on the adviser’s competence, the information provided, and actual good-faith reliance on appropriate advice. An unsupported credit study cannot cure missing evidence about the research activities. A consultant’s compensation arrangement may be relevant to evaluating independence, but it does not alone establish that a claim is invalid or that a penalty necessarily applies.

Implications for Future R&D Tax Credit Applications

Betz and Little Sandy Coal illustrate the importance of proving the relationship between technical activities and claimed expenditures. Their practical lessons include identifying the correct business component, distinguishing experimentation from routine implementation, substantiating allocations, and reviewing customer contracts.

Later IRS filing requirements are separate administrative developments. They should not be described as a court-ordered consequence of Betz or as proof that every suggested documentation practice has become mandatory.

The Impact of Revised Form 6765

The IRS instructions revised in December 2025 make Section G optional for tax years beginning before 2026 and required for years beginning aft