The United States Tax Court’s December 23, 2024 decision in Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, illustrates the evidence needed to support engineering-related research tax credits. Judge Travis A. Greaves concluded that the taxpayer had not established qualified research in three projects selected for trial. The decision concerns Phoenix Design Group; attributing this engineering litigation to Kozma v. Commissioner is incorrect.

This study examines the distinction between complex professional design work and qualified research under Internal Revenue Code Section 41. The decision applies existing statutory and regulatory requirements to the evidence presented. It does not establish an industry-wide prohibition on engineering credits or require an advance beyond the knowledge of the engineering profession.

Foundations of the Section 41 Research Credit

The federal research credit was introduced in 1981 and made permanent by the Protecting Americans from Tax Hikes Act of 2015. Qualification depends on the activities performed, the expenses incurred, and applicable exclusions. Permanence of the credit does not imply automatic eligibility, and disallowance does not automatically establish liability for penalties.

The regular credit’s principal incremental component is generally calculated as follows, before applicable adjustments:

Regular incremental credit = 20% × the excess of current-year qualified research expenses over the base amount.

The base amount is subject to statutory rules, including a minimum equal to 50% of current-year qualified research expenses. The alternative simplified credit is a separate method, generally using 14% of the excess over 50% of average qualified research expenses for the preceding three years, with a special rule where a preceding year has no such expenses. Section 280C coordination and other limitations affect the ultimate tax benefit.

Qualified research expenses can include employee wages for conducting, directly supervising, or directly supporting qualified research; qualifying supplies; certain payments for computer use; and eligible contract research expenses. The usual contract research inclusion rate is 65%, although special statutory categories have different percentages. General administration and all engineering payroll do not automatically qualify.

The Phoenix Design Group litigation concerned historical research years, for which the opinion applied the Section 174 test then in effect. For tax years beginning after December 31, 2024, legislation added Section 174A for domestic research and experimental expenditures and amended Section 41’s expenditure cross-reference. Current claims must use the law applicable to their tax year. Deductibility of research expenditures and eligibility for the research credit remain separate questions.

The Four-Part Test

The requirements operate together and generally apply separately to each business component. Failure at the overall component level does not necessarily disqualify every activity in the project: the shrinking-back rule may permit evaluation of a qualifying subset.

The regulations do not require discovery of information that exceeds the common knowledge of skilled professionals. Existing technologies and established scientific principles may be used. Conversely, a project being new to a firm does not establish qualification without the required uncertainty, experimentation, and other statutory conditions.

Factual Analysis of Phoenix Design Group, Inc. v. Commissioner

Phoenix Design Group was a Tennessee-based engineering consulting firm specializing in mechanical, electrical, plumbing, and fire protection systems, particularly for hospitals and laboratories. The disputed research activities related to 2013 through 2016 and more than 200 projects. The opinion distinguishes those research years from the 2015 through 2019 credit years affected by the claims.

The parties selected three projects for a nonbinding trial sample. The court expressly stated that its conclusions on those projects would not bind the remaining projects; it expected the opinion to help the parties resolve them. The trial addressed whether the selected projects involved qualified research, rather than the amounts of qualified research expenses.

PDG characterized the integrated building systems as the relevant business components and relied on its design process, employee activities, and project records. The dispute required more than proof that the designs were complicated or involved multiple engineering disciplines.

The Project Sample Framework

The actual trial sample consisted of healthcare and university projects, not the retail, warehouse, and healthcare examples originally attributed to the case. The descriptions below identify the work and claimed issues; they do not imply that the court found qualifying experimentation.

The court held that none of the three trial projects entailed qualified research on the evidence presented. Its analysis examined both the overall systems and whether qualification could be established for smaller components.

The Judicial Analysis of Uncertainty

Qualifying uncertainty can concern capability, method, or appropriate design. Design uncertainty is therefore a recognized category; it should not be categorically contrasted with technological uncertainty as though design questions can never qualify. The inquiry considers what information was available when the relevant research activities began.

PDG’s argument that designs could change before construction ended did not, by itself, establish uncertainty about the basic design of an entire system. Project revisions may introduce particular technical questions, but the taxpayer must explain their scope and show the investigative activities undertaken to resolve them.

Calculations and Investigative Activities

Routine calculations, equipment selection, and responses to customer decisions do not establish experimentation merely because they are technically demanding or repeated. The evidence must explain how the work evaluated alternatives to resolve an identified uncertainty, rather than simply applying known inputs to obtain a design answer.

This distinction does not create a categorical ban on calculations, commercially available software, or engineering models. Modeling and simulation are recognized ways to evaluate alternatives. A defensible analysis explains the role of the calculation or model in the investigation, the alternatives assessed, and what the results established.

The practical question is whether the record demonstrates an evaluative process. A completed drawing may show what was built while leaving unanswered why alternatives were investigated or how uncertainty was resolved.

Integrated Systems and the Shrinking-Back Rule

The interdependence of mechanical, electrical, plumbing, and fire protection systems can be relevant, but it does not automatically extend uncertainty in one element to the entire design. The taxpayer needs evidence connecting the claimed uncertainty to the component for which qualification is asserted.

Treasury Regulation § 1.41-4(b)(2) provides for applying the qualification tests to the most significant subset of elements when the overall business component fails. The inquiry continues to smaller subsets where appropriate. Organizing records by actual systems and technical issues helps support that analysis; merely dividing work by employee does not necessarily identify the relevant subset.

The Process of Experimentation

A process of experimentation evaluates one or more alternatives where capability, method, or appropriate design is uncertain at the beginning of the research. It must rely on the specified scientific or engineering principles and relate to a qualified purpose. A successful end result alone does not show that the required process occurred, and an unsuccessful experiment is not automatically ineligible.

PDG’s general description of a six-stage design process was insufficient to establish that employees actually performed qualifying experimentation. Project-stage terminology and the possibility of revisiting earlier design decisions cannot substitute for evidence of the underlying evaluative activities.

The Scientific Method Standard

The opinion discussed an evaluative process resembling the scientific method. This should not be converted into an invented judicial quotation or an inflexible requirement for a particular hypothesis form, laboratory notebook, or testing template. Useful records explain the actual technical investigation in a form suited to the business.

• Identify uncertainty: Describe the capability, method, or design question and the relevant information available at the outset.
• Identify alternatives: Explain the candidate designs or approaches intended to resolve that question.
• Evaluate alternatives: Preserve models, simulations, trial results, and the technical criteria used to compare options.
• Record findings: Explain what was learned, why an alternative was rejected or refined, and whether uncertainty remained.

Emails, meeting notes, drawings, calculations, and time records can all be relevant evidence. Their usefulness depends on whether they establish the claimed investigation and its connection to the expenses. Generic descriptions of work performed may leave that connection unproven.

Comparative Jurisprudence: The Funded Research Exclusion

Section 41(d)(4)(H) separately excludes research to the extent it is funded by another person or governmental entity. Under the implementing regulations, contractual payment risk and retention of substantial rights require analysis. A fixed price, a warranty, or a professional duty to deliver acceptable work is not conclusive on its own.

In Meyer, Borgman & Johnson, Inc. v. Commissioner, 100 F.4th 986 (8th Cir. 2024), the Eighth Circuit affirmed the denial of credits because payment under the contracts was not contingent on successful research. The decision distinguished ordinary professional performance obligations from research-success risk. It did not rest on the asserted additional finding that the firm surrendered its intellectual property rights.

In Geosyntec Consultants, Inc. v. United States, 776 F.3d 1330 (11th Cir. 2015), the Eleventh Circuit likewise affirmed an adverse funded-research determination concerning the two capped contracts at issue on appeal. Describing that appellate decision as a taxpayer win based on experimental documentation is incorrect.

In System Technologies, Inc. v. Commissioner, Docket No. 12211-21, the Tax Court’s January 3, 2025 order denied the Commissioner’s motion for partial summary judgment. The court applied Indiana law and reasoned that the taxpayer would have to refund payments if it failed to deliver the product, making ultimate payment contingent on success. The order did not concern Dubai or UAE law and should not be treated as a final determination that every requirement for the credit was satisfied.

These decisions address distinct issues. Passing the technical qualification tests does not resolve the funded-research exclusion, and prevailing on a funding issue does not prove experimentation. The table describes the identified opinions and order, rather than asserting the current docket status of unresolved matters.

Documentation and the Substantially All Requirement

The experimentation threshold generally requires at least 80% of the relevant research activities, measured by cost or another consistently applied reasonable basis, to constitute elements of experimentation for a qualified purpose. It applies separately to each business component. It is not a requirement that 80% of all company operations be research, nor is it interchangeable with the separate employee-wage substantially-all rule.

Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), emphasizes the need for a supported allocation between experimental and nonexperimental activities. Novelty of the finished vessel did not establish that the activities used to develop it met the threshold. A taxpayer must support the numerator and denominator of its allocation with evidence rather than assume all work on a new product is experimental.

Detailed contemporaneous records are useful, but the governing recordkeeping standard does not universally mandate task-by-task time sheets in one prescribed format. Credible testimony and other records may matter. Estimates require a reliable factual foundation and do not cure failure to establish that qualified research occurred.

The Impact of Accuracy-Related Penalties

In Phoenix Design Group, the parties stipulated the penalty consequences if none of the trial projects qualified. The court’s penalty holding followed that stipulation. It does not establish that penalties are almost mandatory whenever an R&D credit is disallowed.

Section 6662 penalties require an applicable statutory basis, and Section 6664 provides a reasonable-cause and good-faith exception where its requirements are satisfied. A taxpayer should assess both the substantive credit position and the factual basis for any penalty defense. Reliance on an adviser is not automatically sufficient.

Implications for Future R&D Tax Credit Applications

The practical lesson is to connect technical activities, business components, and expenses before making a claim. The opinion illustrates evidentiary weaknesses that can undermine a case; it does not enact a new four-part test or prove that all engineering claims face a newly heightened legal standard.

Strengthening R&D Studies

A retrospective study can organize and explain existing evidence, but it should not invent uncertainties, experiments, or employee allocations. Businesses can improve the underlying record through the following practices:

• Capture initial technical questions: Record the uncertainty when research begins and distinguish it from budget, scheduling, or customer-preference questions.
• Preserve the evaluation: Retain alternative designs, test or simulation outputs, technical discussion, and reasons for changes.
• Connect labor and expenses: Use sufficiently descriptive records and a consistent allocation method to identify qualified services and exclude unrelated work.
• Review contracts: Evaluate payment contingencies, refund obligations, governing law, and substantial rights alongside the technical analysis.

IRS Review of Research Credit Refund Claims

The IRS has established information requirements for research-credit refund claims. These concern claim validity and screening and are distinct from proving ultimate entitlement to the credit. It is misleading to characterize the process as a newly introduced automated gatekeeper that necessarily rejects claims before any human review.

Under the IRS’s updated guidance effective for claims postmarked on or after June 18, 2024, taxpayers are no longer required to include the names of individuals performing each activity or the information each individual sought to discover as initial claim items, although the IRS may request them during examination. The remaining required information includes the business components, the research activities for each component, and the specified totals of qualified expenses. Taxpayers should follow the instructions applicable when filing, including Form 6765 requirements.

The Future of Engineering-Based R&D Credits

Engineering firms can qualify where their actual activities satisfy the statutory tests and avoid the exclusions. They do not have to demonstrate that they are pushing the boundaries of the entire technical field. At the same time, complexity, customization, regulatory compliance, and professional reputation do not independently establish qualification.

Learning and Technical Investigation

Keeping a clear record of what a team learned can help explain experimentation, but an analogy to educational problem-solving is not legal authority. Educational research by an author sharing a name with a litigant does not establish a connection to this tax case or alter the Section 41 requirements.

Technological Trends a