Statutory Foundations and the Evolution of the Four-Part Test

To understand the impact of the Phoenix Design Group decision, one must first examine the statutory framework governing the research credit. Section 41(d) defines “qualified research” through a rigorous four-part test, each element of which must be satisfied for every individual “business component” claimed by the taxpayer. A business component is broadly defined as any product, process, computer software, technique, formula, or invention to be held for sale, lease, or license, or used by the taxpayer in a trade or business.

Treasury Regulation § 1.41-4 does not require discovery of information that exceeds the common knowledge of skilled professionals in the relevant field. It does require the identification of uncertainty and a process designed to evaluate alternatives. Systematic trial and error can qualify. Neither Phoenix Design Group nor the regulation makes novelty to the industry a prerequisite, and routine application of known engineering information is not automatically experimentation.

Factual Background of the Phoenix Design Group Litigation

Phoenix Design Group (PDG) was a multidisciplinary engineering firm taxed as a C corporation. Its research credit claims arose from design work during 2013 through 2016, with credits used in 2015 through 2019. The study underlying the claims considered more than 200 potentially eligible projects. The parties tried three sample projects to establish a framework for resolving the dispute; the sample findings were not automatically binding on every remaining project.

The trial projects were the Gerald Champion Military Psychiatric Unit, Baptist Memorial Health North Mississippi Oxford, and Vanderbilt University Engineering and Science Building. PDG maintained that designing integrated MEPF systems for these facilities involved uncertainty and research. The projects required specialized engineering, but the final design being unknown at inception did not itself establish eligibility.

The IRS challenged the research credits. The Tax Court examined whether the evidence established qualifying uncertainty and investigative activity under the applicable Section 174 standard and a process of experimentation under Section 41. The opinion found that the three trial projects did not involve qualified research; the parties’ arrangements governed the consequences for the remaining disputed projects and penalties.

Uncertainty and Investigative Activity under Section 174

A central pillar of the Phoenix Design Group decision was the court’s interpretation of “uncertainty” under Section 174. To satisfy this test, a taxpayer must show that the information available at the outset does not establish the capability, method, or appropriate design of the business component. PDG argued that because they did not know the exact configuration of ductwork, electrical loads, or plumbing layouts at the start of a hospital project, they faced technological uncertainty.

The relevant inquiry is what information was available to the taxpayer at the outset and whether that information established capability, method, or appropriate design. It is misleading to define qualifying uncertainty as something no engineer could resolve using established principles. In Phoenix Design Group, the court distinguished investigative work from performing calculations using information already available to the firm.

For engineering and architectural firms, an unfinished design is therefore insufficient evidence by itself. A claim should identify what technical information was lacking and the activities undertaken to resolve that gap. Established engineering principles may support qualified research; the taxpayer must still prove that the particular activities meet the statutory tests.

The Failure of the Process of Experimentation Test

The process of experimentation requirement asks whether substantially all of the relevant research activities constitute elements of a process designed to evaluate alternatives for resolving uncertainty. Modeling, simulation, or systematic trial and error may satisfy that requirement when directed to a qualified purpose. A successful final design alone does not demonstrate the process used to reach it.

PDG used a six-stage design process covering basis of design, schematic design, design development, construction documents, bidding, and construction administration. Naming these phases does not establish experimentation. The evidence must connect particular alternatives and evaluations to identified technical uncertainty; a standard project workflow is insufficient by itself.

Records and testimony must explain what work was experimental and how claimed expenses relate to that work. Generic descriptions such as design work or coordination may be too vague to establish that connection. These examples are practical illustrations, not verified quotations from PDG’s time records. Treasury Regulation § 1.41-4(d) requires records in sufficiently usable form and detail to substantiate eligibility, without prescribing one mandatory documentation system.

Comparative Jurisprudence: Trends in R&D Tax Litigation

Phoenix Design Group can be compared with Little Sandy Coal Co. v. Commissioner and Siemer Milling Co. v. Commissioner. These decisions illustrate how weaknesses in proof of qualifying activities can defeat a research credit claim even where the taxpayer develops a new product or improves a process.

Little Sandy Coal and the “Novelty Argument”

In Little Sandy Coal, the taxpayer, a shipbuilder, claimed credits for the design and construction of “first-in-class” vessels. The taxpayer argued that because the ships were prototypes (pilot models), all activities involved in their development should qualify under the “substantially all” rule.

Both the Tax Court and the Seventh Circuit rejected this “novelty argument,” emphasizing that the test must be applied to activities, not to the physical product being developed. Like the taxpayer in Phoenix Design Group, Little Sandy Coal failed because it could not offer a “principled way” to determine which portion of employee activities constituted experimentation as opposed to routine fabrication or project management.

Siemer Milling and the Limits of Trial and Error

In Siemer Milling Co. v. Commissioner, T.C. Memo. 2019-37, the Tax Court found insufficient evidence that the claimed milling projects satisfied the qualified research requirements. Describing improvements and using technical terminology did not establish a methodical evaluation of alternatives. The lesson is not that trial and error is prohibited: systematic trial and error can qualify when the evidence establishes the required process.

Funded Research: Smith and System Technologies

The early rulings in Smith v. Commissioner and System Technologies, Inc. v. Commissioner concerned IRS motions for summary judgment on funded research. Denying those motions should not be characterized as a final award of credits or a categorical holding that standard professional-service contracts qualify.

Section 41(d)(4)(H) excludes research to the extent funded by another person. Treasury Regulation § 1.41-4A(d) examines payment contingencies and substantial rights retained in the research. Contract terms and applicable law matter. Fixed pricing, milestone billing, or a general obligation to provide competent professional services does not automatically establish that payment depends on successful research.

The later decision in Smith v. Commissioner, T.C. Memo. 2026-50, issued June 16, 2026, illustrates the distinction. Of six sample projects, two failed the substantial-rights requirement. For four others, rights were retained but payments were not contingent on research success; possible credit eligibility was limited to qualifying costs exceeding funding, rather than an automatic partial-credit award. The preliminary System Technologies ruling likewise should be understood in its procedural context. Neither decision establishes a universal safe harbor for milestone payments or fixed-price contracts.

Mathematical Rigor in the “Substantially All” Calculation

The substantially all requirement concerns whether at least 80% of the relevant research activities for a business component constitute elements of experimentation for a qualified purpose. Treasury Regulation § 1.41-4(a)(6) allows measurement on a cost or other consistently applied reasonable basis. This is an activity test, distinct from the computation of qualified research expenses and from Form 6765’s separate business-component disclosure threshold.

In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the court affirmed disallowance because the taxpayer did not substantiate the required proportion of qualifying activities. It also rejected a categorical exclusion of direct supervision and direct support from the experimentation analysis. The following table describes possible treatment under an appropriate cost-based activity measure; amounts do not enter either side automatically merely because they appear in a project ledger.

Direct supervision and direct support can be elements of experimentation when the evidence establishes that connection. Their treatment must be consistent with the research activities being measured. Including amounts in both numerator and denominator does not guarantee satisfaction of the threshold, and unsupported project-wide estimates are not cured by relabeling wages.

The Shrinking-Back Rule and Component-Level Evidence

Treasury Regulation § 1.41-4(b)(2) provides that, if a business component fails the qualified research requirements, the analysis can be applied to its most significant subset of elements and then progressively smaller subsets. This shrinking-back rule does not excuse missing evidence or make an entire project eligible because a small part involved experimentation.

In Phoenix Design Group, the court considered the possibility of analyzing the engineering disciplines within an MEPF system, such as mechanical, electrical, or plumbing work. It distinguished those subsets from merely selecting a design phase or the work of particular employees. The supporting evidence still had to establish eligibility at the appropriate component or subcomponent level.

A taxpayer claiming a large business component should retain evidence that permits analysis of meaningful subsets if the overall component fails. Where a subset qualifies, associated costs must be isolated on a supportable basis. This is a practical consequence of the shrinking-back rule, rather than a requirement to claim every possible subcomponent separately at the outset.

Administrative Shifts and the Impact of Form 6765 Revisions

Revisions to Form 6765 expand the information supplied with research credit claims. These filing requirements should be distinguished from substantive eligibility rules and court holdings. They do not establish that the IRS has adopted every rationale advanced in an individual case.

Key Changes in the New Form 6765

The December 2025 instructions make Section G optional for tax years beginning before 2026 and generally required for tax years beginning after 2025, subject to exceptions. Amended refund claims have separate information requirements.

Section E: Other Information. The form requests the number of business components generating QREs and officer wages included in QREs, among other information. These disclosures do not establish a documented causal connection to a particular judicial decision.

Section G: Business Component Information. Required filers generally provide detail for components covering at least 80% of total QREs, capped at 50 components, and aggregate the remainder. Wage categories distinguish conducting, supervising, and supporting research.

Disclosure Exceptions. The instructions provide exceptions for qualifying small businesses making the specified payroll tax election and, on original returns, taxpayers with controlled-group QREs of $1.5 million or less and prior-three-year average annual gross receipts of $50 million or less. An exception does not remove substantiation duties.

The filing disclosures supplement examination requirements. A credit remains subject to the four-part test regardless of whether Section G is required. The claim that these revisions created a new automated system that denies weak refund claims is not established by the form instructions.

Research-Cost Treatment and Accuracy-Related Penalties

Phoenix Design Group also involved accuracy-related penalties. The parties stipulated consequences that depended on whether any sample project involved qualified research; the court found none did. That case-specific outcome should not be converted into a rule that every unsuccessful research credit claim automatically carries a penalty. Section 6662 penalties depend on their statutory grounds, and reasonable-cause and good-faith relief under Section 6664(c) may apply.

The treatment of research expenditures depends on the tax year. The 2013–2016 activities in Phoenix Design Group predated mandatory amortization. The Tax Cuts and Jobs Act generally required capitalization and five-year domestic or fifteen-year foreign amortization for tax years beginning after 2021. Public Law 119-21 subsequently added Section 174A, permitting current deductions for domestic research or experimental expenditures in tax years beginning after December 31, 2024, with an elective amortization alternative and transition provisions. Foreign research generally remains subject to fifteen-year amortization under Section 174. Current Section 41 refers to Section 174A, and Section 280C coordinates the deduction and credit. Thus, claiming a credit does not invariably require five- or fifteen-year capitalization of the same expenditure.

The Future of R&D Tax Credit Applications: Strategic Recommendations

Little Sandy Coal and Phoenix Design Group show why eligibility conclusions should be grounded in identifiable activities and reliable expense allocations. A retrospective study is not automatically invalid, but high-level interviews and unsupported estimates may be insufficient. Contemporaneous records can make the factual connection substantially easier to demonstrate.

Implementation of Contemporaneous Activity Tracking

A useful practice is to capture information while technical work is underway and link it to projects, com