The federal research and development tax credit under Internal Revenue Code Section 41 provides an incentive for qualifying research. In Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, issued December 23, 2024, the Tax Court examined whether an engineering firm’s activities on three selected projects met the statutory requirements. This study considers that decision, the separate funded-research ruling in Meyer, Borgman & Johnson, Inc. v. Commissioner, and their implications for architecture, engineering, and construction businesses.
Phoenix Design Group should not be identified as Pegasus Energy Group. The decisions discussed here apply existing statutory and regulatory requirements to particular facts; they do not establish a categorical prohibition on engineering research credits.
The Evolutionary Context of Section 41 Litigation
The research credit was made permanent in 2015, but qualification continues to depend on the activities performed and the evidence supporting the claim. Technical complexity, professional expertise, and a successful design do not independently establish eligibility. A credit study must connect qualifying work to an identified business component and eligible expenditures.
The distinction between ordinary design work and qualified research is fact dependent. Applying established information to produce a design may fail to establish research, while engineering activities that address unresolved technical questions through a qualifying evaluation of alternatives may satisfy the requirements. Phoenix Design Group illustrates the application of those existing rules, rather than the creation of a new four-part test.
Procedural History and the Fact Pattern of Phoenix Design Group
Phoenix Design Group (PDG) designed mechanical, electrical, plumbing, and fire protection (MEPF) systems. The dispute involved credits generated for 2013 through 2016 and used in tax years 2015 through 2019. The broader consultant engagement also addressed 2012.
The consultant considered 409 projects, identified 238 as potentially qualifying, and selected 24 for detailed review. It treated 20 of those sampled projects as qualifying and extrapolated its analysis. For trial, the parties selected three projects: the Gerald Champion Military Psychiatric Unit, Baptist Memorial Hospital–North Mississippi Oxford, and Vanderbilt University Engineering and Science Building.
The court found no qualified research within the three trial projects. The parties’ agreement provided that the findings would not bind the remaining projects, although they would supply a framework for resolving them. The agreement also linked the accuracy-related penalty outcome to whether any of the three projects contained qualified research. It is therefore misleading to describe this memorandum opinion as automatically disallowing every project in PDG’s claim.
| Project Identified in Sampling | Primary Engineering Focus | Outcome of Court Review |
|---|---|---|
| Gerald Champion Military Psychiatric Unit | MEPF systems for a psychiatric facility. | No qualified research established in the trial project. |
| Baptist Memorial Hospital (BHNM Oxford) | Hospital MEPF systems, including HVAC and electrical design. | No qualified research established in the trial project. |
| Vanderbilt University (VU ESB) | MEPF systems for an engineering and science building. | No qualified research established, including through consideration of narrower components. |
Deconstructing the Four-Part Test: The Judicial Standard of Proof
Section 41 applies several requirements commonly described as the four-part test: qualifying research expenditures, technological information, development of a new or improved business component for a permitted purpose, and a qualifying process of experimentation. Statutory exclusions and expense substantiation must also be considered.
The Section 174 Test: Defining Objective Uncertainty
For PDG’s historical tax years, Section 41 incorporated the research-expenditure standard under Section 174. Treasury Regulation Section 1.174-2 describes uncertainty by reference to whether available information establishes the capability, method, or appropriate design of a product. The standard does not require discovery that advances knowledge across an entire industry, and it does not require proof that a design is globally optimal.
- Capability uncertainty: Available information does not establish whether the technical objective can be achieved.
- Method uncertainty: Available information does not establish how to achieve the technical objective.
- Design uncertainty: Available information does not establish the appropriate design.
PDG’s position that designs might change before construction ended did not, by itself, demonstrate qualifying uncertainty. A taxpayer must explain what technical information was missing and what investigative work it undertook. Ordinary calculations using information already available do not establish an investigation merely because the calculation produces a previously unwritten answer.
The Process of Experimentation Test: Design Stages and Evaluation of Alternatives
PDG’s six-stage design workflow did not itself prove a qualifying process of experimentation. Describing a professional workflow is different from demonstrating how alternatives were evaluated to resolve a technical question. The analysis requires evidence of the actual activities, rather than reliance on the names of project phases.
A qualifying process may use modeling, simulation, or systematic trial and error. Under the regulations, at least 80% of the relevant research activities must constitute elements of experimentation for a qualified purpose, measured on cost or another consistently applied reasonable basis. This threshold is not automatically an 80% test of every activity in an entire construction project. Failed experiments are not mandatory, and iteration alone does not prove qualification.
The Technological in Nature and Permitted Purpose Tests
In PDG, the parties stipulated that the technological-in-nature and permitted-purpose requirements were satisfied. The opinion should not be presented as rejecting those stipulated elements or as independently denying the sample projects under the adaptation and quality-control exclusions.
Those exclusions nevertheless remain relevant to other claims. Customer-specific work requires a separate analysis of the adaptation exclusion; neither customization nor the existence of uncertainty, standing alone, settles that issue. Research must also serve a qualifying improvement in function, performance, reliability, or quality.
Project Post-Mortems: VU ESB and BHNM Oxford
The sample projects illustrate the need to link the claimed uncertainty, investigative activities, business component, and expenditures. A problem affecting one part of a system does not necessarily establish uncertainty across the entire system.
Vanderbilt University Engineering and Science Building (VU ESB)
For VU ESB, the relevant inquiry concerned the MEPF design and the work performed on its components. The original account’s attribution of a hybrid operating room to this university project should not be used to support its analysis. A facility’s scientific purpose does not establish that the engineering firm itself conducted qualified research.
The shrinking-back rule permits examination of the most significant subset of a business component when the whole component fails the requirements, followed by progressively smaller subsets where necessary. It does not allow arbitrary selection of an employee’s work or a design phase as a substitute for a component. A narrower claim still needs evidence supporting both the qualifying activities and associated costs.
BHNM Oxford: Distinguishing Revisions from Research
For BHNM Oxford, design issues and resulting solutions did not establish the necessary investigative process. Revisions can arise from equipment selections, client decisions, coordination, or the application of existing information. Such changes require analysis before being counted as experimentation.
Calculations may be part of qualified research, but they do not qualify solely because they are technically demanding. The practical question is whether the activities evaluated alternatives to resolve uncertainty, rather than simply applying available information to complete a design.
The Relationship Between PDG and Meyer, Borgman & Johnson (MBJ)
These cases address different obstacles to a research-credit claim. PDG concerned the nature and proof of the sample activities. The Eighth Circuit’s May 6, 2024, MBJ decision, No. 23-1523, affirmed denial of approximately $190,000 in credits for 2010, 2011, and 2013 on funded-research grounds.
The Risk Standard and Funded Research
Section 41 excludes research to the extent another person funds it. MBJ’s contracts did not expressly or by clear implication make payment contingent on the success of its research. Duties to meet codes and professional standards did not establish that contingency. The court distinguished ordinary performance obligations from contractual exposure to research failure.
The decision does not automatically disqualify every fixed-price or AIA-based agreement. Funding requires analysis of the actual agreements and circumstances. Substantial rights in the research results are a separate regulatory consideration, rather than a substitute for the payment-risk analysis.
| Factor | MBJ (2024) Judicial Finding | Implication for Future Applications |
|---|---|---|
| Contractual Contingency | The contracts did not make payment contingent on research success, expressly or by clear implication. | Examine the actual payment and acceptance obligations; fixed pricing alone is insufficient. |
| Standard of Care | Professional standards and code compliance did not establish the required research-success contingency. | Distinguish ordinary performance duties from risk of nonpayment for unsuccessful research. |
| Rights Retention | The appellate decision turned on payment contingency, rather than a holding that MBJ lacked substantial rights. | Assess substantial rights separately under the funded-research regulations. |
Implications for Future R&D Tax Credit Applications
Read together, the cases support careful activity analysis and contract review. They do not replace the statute with a new industry-specific eligibility standard.
The Limits of Generic Documentation
Time entries labeled design, drafting, or coordination may establish that work occurred but provide little explanation of why it qualifies. They should be considered with drawings, calculations, models, correspondence, meeting records, and reliable testimony. Contemporaneous records are valuable; no universal rule requires a particular uncertainty-log format.
- Identify the specific technical question and the relevant business component.
- Explain the alternatives considered and how they were evaluated.
- Connect supporting records and eligible employee activities to claimed expenditures.
The Accuracy-Related Penalty
PDG’s penalty outcome must be understood in light of the parties’ agreement concerning the three projects. Section 6662’s 20% accuracy-related penalty was not newly created by the case, and a disallowed research credit does not automatically produce a penalty. Applicable statutory grounds, procedural requirements, and reasonable-cause and good-faith rules matter. A consultant’s involvement does not, by itself, resolve those questions.
The Legislative Landscape: Sections 174 and 174A and the 2025 OBBBA
The One Big Beautiful Bill Act, enacted July 4, 2025, added Section 174A, allowing immediate deductions for domestic research or experimental expenditures for tax years beginning after December 31, 2024. Foreign research expenditures remain subject to 15-year amortization under Section 174. Deductibility does not establish Section 41 credit eligibility.
For 2022–2024, the prior rules generally required five-year domestic and 15-year foreign amortization. Eligible small businesses could elect retroactive domestic treatment for those years, subject to procedural and refund limitations. Revenue Procedure 2025-28 set a July 6, 2026, election deadline, with some refund deadlines earlier. That general election window has passed. Separate transition provisions address recovery of remaining domestic balances.
The IRS instructions for Form 6765 make Section G optional for tax years beginning before 2026 and required thereafter, subject to exceptions. It requests business-component information; it was not universally mandatory for 2024. Return disclosures remain separate from the evidence needed to substantiate eligibility.
Analytical Perspective: Expertise and Technical Uncertainty
PDG does not create an expertise penalty. Experienced engineers can perform qualified research, and the use of established scientific principles is consistent with the credit’s technological requirement. The challenge is evidentiary: the taxpayer must explain why available knowledge did not establish the capability, method, or appropriate design and how the work addressed that gap.
A successful completed facility shows the result of an engagement. It does not reveal which activities were experimental. Descriptions should therefore explain the investigation and distinguish it from ordinary delivery work without assuming that every difficult problem qualifies.
Strategic Recommendations for Compliance
Contractual Risk Assessment
Review agreements for payment contingencies, acceptance provisions, termination rights, ownership, and retained use rights. Milestone payments and rights-retention language should reflect the real commercial arrangement. They are not automatic safe harbors, and adding a technical guarantee solely for tax purposes does not establish eligibility.
Implementation of a Technical Uncertainty Log
A practical log can record the initial information gap, alternatives, evaluation methods, results, and related records. Update it as the work develops and connect it to employee activities and costs. The log is an organizational tool, not a substitute for evidence or a legally mandated document.
Rigorous Application of the Shrinking-Back Rule
Identify the appropriate business component and apply the requirements at that level. If it fails, examine the most significant subset and continue as the regulation provides. Avoid unsupported allocations or invented percentages of experimental work. A smaller component qualifies only if its own activities satisfy the requirements and its expenditures are supported.
