The United States research and development (R&D) tax credit under Internal Revenue Code (IRC) Section 41 and the deductions for research or experimental (R&E) expenditures are related but separate tax benefits. Section 41 imposes activity and expenditure requirements beyond those applicable to a research deduction. Public Law 119-21, enacted July 4, 2025 and commonly called the One Big Beautiful Bill Act (OBBBA), introduced Section 174A for domestic R&E expenditures. The 2024 decisions in Phoenix Design Group and Meyer, Borgman & Johnson illustrate why deduction eligibility, credit qualification, substantiation, and contractual funding must be analyzed separately.

The Historical Genesis of Section 174: The Stanton Precedent and the Trade or Business Threshold

Section 174 was enacted in 1954 to encourage research by permitting qualifying expenditures connected with a taxpayer’s trade or business to be deducted currently. Historically, taxpayers could alternatively elect amortization over at least 60 months, subject to statutory conditions. These historical rules should not be confused with the mandatory capitalization rules that began in 2022 or the domestic expensing rules enacted in 2025.

The Facts and Findings of Stanton v. Commissioner (1968)

Stanton v. Commissioner, 399 F.2d 326 (5th Cir. 1968), involved Jack P. Stanton and his wife, Virginia G. Stanton, and expenditures on a storm-proof boat. The court accepted that the expenditures were research or experimental in character but affirmed their disallowance because they were not connected with a trade or business of the taxpayers. His inventive work as an employee did not establish a separate personal inventing business, and his independent activities lacked sufficient continuity, regularity, time, and effort.

Stanton, Snow, and the “In Connection With” Standard

The Supreme Court’s later decision in Snow v. Commissioner, 416 U.S. 500 (1974), rejected an interpretation that required an operating, sales-producing business before Section 174 deductions could begin. Snow involved Edwin A. Snow’s investment in a partnership developing an incinerator. The Court distinguished Section 174’s “in connection with” language from Section 162’s narrower “carrying on” language.

Stanton must therefore be read in light of Snow rather than used to impose a blanket existing-business requirement on startups. Later cases still examine whether research relates to the taxpayer’s own prospective business rather than merely an investment. In Zink v. United States, 929 F.2d 1015 (5th Cir. 1991), the court concluded that the taxpayers’ aircraft-component arrangements were investment activities insufficient to establish their own trade or business. These cases concern the commercial connection of expenditures; they do not establish a modern Section 41 documentation doctrine.

The Mechanics of the Research Credit: The Four-Part Test Under Section 41(d)

Credit eligibility is generally evaluated separately for each business component. Passing the four-part test does not establish the amount of the credit: the taxpayer must also identify eligible wages, supplies, computer-use costs, and contract research expenses, apply the relevant computation, and consider statutory exclusions. Not every deductible research expenditure is a qualified research expense.

The Research Expenditure Test: Resolution of Uncertainty

For the historical years considered in Phoenix Design Group, Section 41 referred to Section 174. For taxable years beginning after December 31, 2024, the amended statute refers to Section 174A. The underlying inquiry concerns research intended to eliminate uncertainty about capability, method, or appropriate design. Routine calculations using information already available do not automatically establish such research.

The Technological in Nature Test

The investigative process must fundamentally rely on physical or biological sciences, engineering, or computer science. Research in the social sciences, arts, or humanities is excluded. A project’s industry label or the professional credentials of its staff do not, by themselves, establish qualification.

The Business Component and Qualified Purpose Tests

The information sought must be intended to help develop a new or improved product, process, software, technique, formula, or invention held for sale, lease, license, or business use. The qualifying improvement concerns function, performance, reliability, or quality, rather than style, taste, cosmetics, or seasonal design.

The Process of Experimentation Test

At least 80% of the relevant research activities, measured by cost or another consistently applied reasonable basis, must constitute elements of a process of experimentation for a qualified purpose. This generally involves identifying uncertainty and alternatives and evaluating them through methods such as modeling, simulation, or systematic trial and error. An iterative workflow can support the claim, but a fixed number of iterations, a failed prototype, or a particular project-management method is not universally required.

The 2024 Decision: Phoenix Design Group, Inc. v. Commissioner

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, was issued on December 23, 2024. The taxpayer provided mechanical, electrical, plumbing, and fire-protection engineering services. The dispute concerned research credits for 2013–2016.

Analysis of the Trial Projects

The parties selected three projects for trial: the Gerald Champion Military Psychiatric Unit, Baptist Memorial Hospital–North Mississippi in Oxford, and the Vanderbilt University Engineering and Science Building. The court found no qualifying research in those projects. The trial findings were intended to provide a framework for resolving other projects, rather than automatically binding every remaining project.

The decision illustrates that a complex design, a revisable drawing, or a completed engineering solution does not alone prove technical uncertainty and experimentation. Phoenix Design Group’s description of its six-stage design process was insufficient without evidence of qualifying activities. General time descriptions also did not adequately explain how the actual work supported the claim.

The Accuracy-Related Penalty and Substantiation

The 20% accuracy-related penalty outcome reflected the parties’ stipulation concerning the trial projects. It should not be characterized as a new rule automatically penalizing missing contemporaneous timesheets. Penalty liability depends on the applicable statutory grounds, procedural requirements, and available reasonable-cause and good-faith defenses.

The Funded Research Exclusion: Meyer, Borgman & Johnson, Inc.

In Meyer, Borgman & Johnson, Inc. v. Commissioner, the Eighth Circuit affirmed the denial of research credits in May 2024 because the research was funded. Section 41(d)(4)(H) excludes research to the extent funded by another person or governmental entity. Treasury Regulation Section 1.41-4A(d) addresses payment risk and rights in the research results.

Economic Risk and Fixed-Price Contracts

A fixed price does not automatically establish that the contractor bears the relevant research risk. In this case, contractual duties to deliver design services and meet professional standards did not make payment contingent on research success, expressly or by clear implication. Ordinary cost-overrun risk differs from the risk of nonpayment for unsuccessful research.

The analysis must address the agreement as a whole, including acceptance, rejection, payment, and research-rights provisions. No single formula of words guarantees eligibility, and a contingency can arise by clear implication. Retention of substantial rights is a separate issue; satisfying it does not cure a failure of the payment-risk requirement.

Substantiation: Records, Document Production, and Summaries

Taxpayers must organize evidence so that the IRS and a court can connect claimed expenditures with qualifying activities. The volume of records does not establish their relevance. Anderson v. Commissioner, T.C. Memo. 2024-95, concerns general expense substantiation and illustrates the need to connect accounting summaries with underlying evidence. It is distinct from Stanton and does not establish a special research-credit standard.

Judicial Treatment of Document Dumps

United States v. Quebe, 321 F.R.D. 303 (S.D. Ohio 2017), addressed inadequate discovery responses in research-credit litigation despite a large document production. Its lesson concerns responsive, organized evidence and compliance with discovery obligations. It does not establish that large collections of records are inherently improper or that every taxpayer producing them is attempting to conceal a weak claim.

Practices for Audit-Ready R&D Studies

• Activity-level support: Maintain records connecting employee services and costs to specific research activities and business components. Detailed time records can help, but the law does not prescribe one universal timekeeping system.
• Evidence of investigation: Preserve relevant design alternatives, test data, simulation results, technical correspondence, and explanations of uncertainty. Summaries should trace back to underlying evidence.
• Shrinking-back analysis: If a business component fails the qualification requirements, apply the test to its most significant subset of elements and continue as appropriate. Identify actual components or subcomponents, rather than merely relabeling employee groups or design phases.

Legislative Transformation: From TCJA Capitalization to Domestic Expensing

The Tax Cuts and Jobs Act required specified R&E expenditures paid or incurred in taxable years beginning after December 31, 2021 to be capitalized and amortized over five years for domestic research and 15 years for foreign research, beginning at the midpoint of the taxable year. This delayed cost recovery relative to immediate deductions.

The One Big Beautiful Bill Act of 2025

Public Law 119-21 added Section 174A without a scheduled sunset. It generally allows immediate deductions for domestic R&E expenditures paid or incurred in taxable years beginning after December 31, 2024. Taxpayers can instead elect qualifying capitalization and amortization under Section 174A(c). Foreign R&E expenditures remain subject to 15-year amortization under Section 174.

• Unamortized domestic costs: For expenditures from taxable years beginning in 2022–2024, taxpayers may elect to deduct the remaining domestic balance in the first taxable year beginning after December 31, 2024, or ratably over that year and the next.
• Small-business retroactive election: Eligible taxpayers meeting the Section 448(c) gross-receipts test for their first taxable year beginning after December 31, 2024, and not disqualified as tax shelters, could elect retroactive treatment. The 2025 threshold was $31 million, subject to aggregation and other applicable rules.
• Timing and procedures: Revenue Procedure 2025-28 sets out implementation methods and filing requirements. The general retroactive-election deadline was July 6, 2026, subject to earlier refund-limitation deadlines where applicable. That general deadline has passed as of September 2026; the relief should not be presented as an unrestricted election still available to every small business.

Sector-Specific Incentives and Foreign-Entity Restrictions

Domestic expensing under Section 174A is not limited to semiconductor design. Proposed semiconductor provisions in the Securing America’s R&D Advantage Act should not be treated as enacted Section 174A requirements. Semiconductor investment incentives and foreign-entity restrictions in other provisions require separate analysis and should not be presented as conditions governing every domestic research deduction.

Comparative Analysis of Section 174A and Section 41

A deduction reduces taxable income; a credit reduces tax liability, subject to limitations. Under the amended Section 280C(c)(1), domestic R&E expenditures otherwise deducted or capitalized are reduced by the Section 41(a) credit. Alternatively, Section 280C(c)(2) permits a timely election of a reduced credit, avoiding that expenditure reduction.

The reduced-credit calculation uses the maximum corporate tax rate under Section 11(b), rather than the claimant’s individual marginal rate:

Reduced credit = otherwise determined credit × (1 − 21%) = otherwise determined credit × 79%.

For example, an otherwise determined credit of $100,000 becomes $79,000 under the election. Whether the election produces a preferable result depends on the taxpayer’s circumstances, including deduction utilization and applicable tax rates. Historical years require the version of Section 280C and transition rules applicable to those years.

Strategic Implications for R&D Tax Credit Claims

A defensible claim separately addresses the commercial connection of the expenditure, research qualification, the amount claimed, contractual funding, and the rules for the relevant taxable year. The historical deduction cases and the 2024 credit decisions address different parts of that analysis.

Retrospective Studies and Filing Requirements

A retrospective R&D study is not categorically prohibited. Its reliability depends on whether evidence supports the activities, allocations, and legal analysis. Interviews and estimates should be corroborated where possible; unsupported percentages and generic narratives are vulnerable.

The IRS’s special research-credit refund-claim requirements began applying in January 2022 and should be distinguished from substantive credit eligibility. Effective June 18, 2024, the IRS waived the initial submission of researchers’ names and the information each individual sought to discover. Its published guidance still requires identification of business components, activities by component, and totals for qualifying wages, supplies, and contract research. Employee-specific information can still be requested during examination.

Contract Review for Risk and Rights

Professional service firms should assess how actual payment terms allocate the risk of unsuccessful research and what rights they retain in its results. Contract descriptions must reflect the real transaction. Adding tax-oriented wording without changing the underlying legal rights and obligations does not establish an unfunded research claim.

Documentation of Experimentation

• Version histories: Identify alternatives investigated and explain why designs changed.
• Simulation studies: Retain relevant computational fluid dynamics, finite element analysis, and other modeling outputs with the technical questions and results they address.
• Rejected alternatives: Preserve reasons for rejection where relevant. Failed experiments can be useful evidence, but failure itself is not a condition of credit eligibility.

Sector-Specific Analysis: Architecture, Engineering, and Software

Architecture, engineering, construction, and software businesses require activity-specific analysis. Professional complexity,