The federal Credit for Increasing Research Activities, codified in Section 41 of the Internal Revenue Code (IRC), requires taxpayers to establish both qualifying research activities and qualifying expenditures. Fudim v. Commissioner, T.C. Memo. 1994-235, illustrates how credible testimony and other evidence can support research-time allocations despite imperfect records. Later decisions clarify the limits of estimation without eliminating it. This study examines Fudim, the development of the research-credit rules, and the practical implications for contemporary claims.

The Judicial Genesis of Fudim v. Commissioner

The narrative of Fudim v. Commissioner begins with Efrem and Margarita Fudim, two highly educated immigrants from the Soviet Union whose technical backgrounds were central to the court’s ultimate determination. Efrem Fudim held a doctorate from the Institute of Control Sciences of the U.S.S.R. Academy of Sciences and, by the mid-1980s, had published nearly 40 scientific papers and held a similar number of patents. Margarita Fudim possessed a degree in mechanical engineering from the Moscow Institute of Chemical Machine Building. This concentration of technical expertise established a baseline of credibility that would later prove decisive when the court evaluated their research claims.

The historical context of the mid-1980s is essential for understanding the technological significance of the case. During this era, the field of “rapid modeling”—now commonly referred to as 3D printing or additive manufacturing—was in its infancy. Efrem Fudim established Light Sculpting Co. in 1985 specifically to innovate within this nascent sector. The process utilized ultraviolet light and light-sensitive liquid polymers to fabricate plastic objects directly from instructions provided by a computer-aided design system. This was not merely an incremental improvement on existing manufacturing; it represented a fundamental shift from subtractive or formative processes to additive manufacturing, promising to eliminate the expensive and time-consuming steps of machining, casting, or molding.

The legal dispute arose when the Commissioner of Internal Revenue determined deficiencies in the Fudims’ federal income taxes for the years 1986, 1987, and 1988. These deficiencies, totaling several thousand dollars across the three years, were primarily based on the disallowance of research and development tax credits and losses related to time-share condominiums. The Fudims challenged these determinations in the United States Tax Court, leading to a trial before Special Trial Judge Pate. The case served as a vehicle for the court to examine the intersection of high-level scientific research and the practical requirements of tax substantiation.

Technical and Foundational Findings of Fact

Light Sculpting Co. worked on technical problems involving radiation and uncured photopolymers. Thick polymer layers presented difficulties for light penetration, and the research considered multilayer formation and alternative ways of solidifying material, including heat and ultrasonic energy. These subjects involve physical science and engineering. The statutory technological-information requirement was already part of the post-1985 research-credit framework; it was not first created after Fudim.

The rapid-modeling work involved transmitting polymer-solidifying radiation through a radiation-transmittent material in contact with uncured photopolymer. The objective was to leave a surface capable of further cross-linking so that additional layers could adhere into a three-dimensional object. Two patents associated with the process supported the account of technical development. Patents do not, however, establish every requirement for a research credit or substantiate particular expense amounts.

The Fudims’ operation was a small, family-run business where the boundaries between professional and personal life were often blurred. Efrem Fudim was the primary researcher, but he also provided consulting services to clients at an average fee of $1,500 per day. Margarita Fudim worked as a computer programmer during the years in issue but spent significant part-time hours supporting the research activities. Their daughter, Natalia, was also paid wages for various tasks within the company. This organizational structure is typical of early-stage startups but creates significant challenges for tax substantiation, as formal time-tracking systems are rarely a priority in the pursuit of scientific breakthroughs.

The financial data presented during the trial showed a company in a state of rapid growth. Gross receipts rose from $17,810 in 1986 to $84,302 in 1988. During this time, the Fudims claimed research credits based on supply costs, wages paid to family members, and Efrem’s own self-employment income. The supplies used in the research included photopolymers, solvents like alcohol, photo masks, light sources (bulbs, starters), fans for cooling, photometers to measure light intensity, and timers. The direct link between these supplies and the physical process of experimentation was a critical factor in the court’s eventual allowance of some expenses.

Procedural Battles and Statutory Limitations

The case also involved a limitations dispute concerning 1986. The source study’s explanation that an unresolved tax liability alone keeps the assessment period open is incorrect. Section 6501 generally imposes a three-year assessment period, subject to statutory exceptions, extensions, and suspension rules. Under Section 6503(a)(1), mailing a qualifying notice of deficiency suspends the running of that period while assessment is prohibited and for the additional statutory period. The particular filing, notice, and litigation dates matter; an unresolved liability does not itself create an indefinite assessment period.

A second procedural challenge involved the allegation of an improper “second examination” of the Fudims’ books and records under Section 7605(b). This section is designed to protect taxpayers from repeated, harassing inspections by the IRS. The Fudims claimed that the IRS had already completed an audit of their 1988 records and was now reopening the case without following proper notice procedures. The court disagreed, finding that the initial adjustments were based on “mathematical or clerical errors” found on the face of the return itself, rather than a full inspection of books. Furthermore, because the Fudims had refused to provide their records during the subsequent attempt to audit, a second inspection never actually occurred.

The mathematical-error issue concerned the assessment of excess research credits for 1988 without the ordinary deficiency procedure. Section 6213(b) permits specified mathematical or clerical errors to be assessed through a separate notice process, subject to the taxpayer’s statutory opportunity to request abatement. This authority is limited to qualifying errors; it is not a general power to bypass deficiency procedures whenever records are missing.

Substantive Analysis of the Research and Development Tax Credit

The core of the dispute centered on Section 41, which provides a credit for qualified research expenses (QREs). To be eligible for the credit, research must meet four independent requirements, often referred to as the “four-part test,” though the terminology and specific nuances of this test have evolved since the Fudim decision.

The Four-Part Test of Qualified Research

The following framework distinguishes the research requirements from expense substantiation. Fudim concerned historical tax years, so its analysis must be read under the law applicable to those years. For current claims, the four requirements are commonly expressed as follows:

Technological in Nature: Based on principles of the physical or biological sciences, engineering, or computer science.

Permitted Purpose and Business Component: The information must be intended to help develop or improve a taxpayer’s business component, with experimentation relating to function, performance, reliability, or quality. Changes confined to style, taste, cosmetics, or seasonal design do not satisfy this permitted-purpose requirement.

Research or Experimental Expenditures: The work must address uncertainty about capability, method, or appropriate design in the experimental or laboratory sense. Historically Section 41(d)(1)(A) referred to Section 174; for tax years beginning after December 31, 2024, the statute refers to domestic research or experimental expenditures under Section 174A. Deductibility under that provision alone does not establish credit eligibility.

Process of Experimentation: Substantially all of the relevant research activities must be elements of a process for evaluating alternatives to resolve technical uncertainty. The requirements apply separately to each business component, subject to the shrink-back rule and statutory exclusions.

The Fudim decision accepted that qualified research occurred. Its practical importance lies in the court’s evaluation of testimony and the record when deciding how much time particular individuals devoted to qualified services. A patent can support technical credibility, but a patent is neither required for the credit nor definitive proof of all four tests. Current regulations contain a limited patent safe harbor for the discovering-technological-information requirement, not for the entire credit.

The Substantially All Rule and Wage Allocation

Two different 80% rules must be kept separate. The employee wage rule in Treasury Regulation § 1.41-2(d)(2) permits all otherwise eligible wages to be treated as paid for qualified services when at least 80% of that employee’s services for the taxpayer during the year are qualified services. These include conducting qualified research and its direct supervision or direct support. Below that threshold, wages are allocated to qualified services. Separately, Treasury Regulation § 1.41-4(a)(6) applies the substantially-all process-of-experimentation test to research activities for a business component, measured on a cost or other consistently applied reasonable basis. Satisfying one rule does not automatically satisfy the other.

The court accepted that Efrem and Margarita devoted at least 80% of their relevant services to qualified work, while finding the proof for Natalia insufficient. Technical education and experience helped the court assess the testimony, but a degree is not a statutory prerequisite. The decisive questions concern actual services, the underlying qualified research, and credible allocations. Direct-support personnel may qualify without scientific credentials; general administrative services do not qualify merely because they assist a research business.

The Cohan Rule and the Threshold of Credibility

Fudim is frequently discussed alongside the estimation principle associated with Cohan v. Commissioner, 39 F.2d 540 (2d Cir. 1930). That principle allows a reasonable approximation when qualifying expenditure is established but the exact amount is uncertain. In United States v. McFerrin, 570 F.3d 672 (5th Cir. 2009), the Fifth Circuit expressly cited Fudim when explaining that testimony and other evidence can support an estimate of research time and expenses. Estimation cannot substitute for establishing that qualified research and eligible expenses existed.

The source describes the Fudims as having discarded contemporaneous records and relied on summaries. Whatever the evidentiary shortcomings, Fudim should not be read as permission to discard records. Nor is it sound to predict automatic total disallowance under current law simply because formal time sheets are unavailable. Courts assess the available evidence, its credibility, and whether it supports a reasonable connection between particular activities and claimed expenses.

The relevant distinction is between an evidence-based approximation and an unsupported guess. Later cases continue to recognize estimation when there is a sufficient factual basis. McFerrin also explains that a court need not accept a taxpayer’s retrospective reconstruction merely because it is offered. Contemporaneous technical and financial records make reconstruction more reliable, but the source’s claimed categorical end to estimates is unsupported.

Regulatory Metamorphosis: From Discovery to Information

The interpretation of the statutory discovery requirement changed over time. Some older decisions and the 2001 final regulations required information beyond the common knowledge of skilled professionals. It is misleading to treat all years before 2004 as a single regime or to attribute the exact 2001 regulatory formulation to the Fudim court in 1994. The breadth-of-knowledge question also differs from whether uncertainty concerns capability, method, or design.

The Transition to the Information Test

T.D. 9104, published in January 2004, finalized regulations generally applicable to taxable years ending on or after December 31, 2003. They explain that discovering information does not require expanding the common knowledge of skilled professionals. The research must instead seek to eliminate uncertainty concerning development or improvement of a business component. The phrase “Information Test” is a descriptive label, not a replacement statutory credit. The remaining requirements, including a qualifying process of experimentation, still apply.

Research credits remain subject to IRS examination and substantiation requirements. Older commentary describing the credit as a Tier I issue reflects a historical administrative framework and should not be presented as the current audit designation. Examination procedures do not amend the statute or establish that a particular claim will necessarily be audited.

The Modern Documentation Standard: Evidence and Substantiation

A hypothetical present-day Fudim outcome cannot be predicted from the date of the original decision alone. Section 6001 and Treasury Regulation § 1.41-4(d) require adequate records supporting eligibility and expenses. Project-level contemporaneous documentation is valuable, but there is no universal rule that every taxpayer must use a particular time-tracking system or that oral testimony is inadmissible. The quality and specificity of the evidence remain central.

The Limits of Unsupported Estimates

Siemer Milling Co. v. Commissioner, T.C. Memo. 2019-37, illustrates the consequences of failing to prove the claimed activities met the research tests, especially the process-of-experimentation requirement. Describing a project, a difficulty, and an eventual improvement is insufficient without evidence of the qualifying investigative and experimental work. The decision should not be reduced to a blanket prohibition on accountant-prepared studies or retrospective interviews; those materials must establish the required facts.

Little Sandy Coal Co. v. Commissioner, T.C. Memo. 2021-15, was affirmed in 62 F.4th 287 (7th Cir. 2023), although the appellate court disagreed with aspects of the Tax Court’s reasoning. The taxpayer failed to supply a principled allocation showing that substantially all relevant activities constituted elements of experimentation. The newness of a vessel and arbitrary employee-time percentages did not establish that fraction. The decision applies an existing evidentiary standard; it does not categorically prohibit reasonable estimates or impose a single mandatory record format.

The Role of Subject Matter Experts (SMEs)

People with firsthand knowledge can explain technical uncertainties, alternatives evaluated, tests performed, and the services provided by particular employees. Their accounts are stronger when connected to design revisions, test results, correspondence, and financial records. An outside subject matter expert is not universally mandatory. The source’s discussion of Moore concerns the distinction between direct research supervision and higher-level management: under Treasury Regulation § 1.41-2(c), immediate supervision can qualify, while supervision through intervening managers does not qualify on that basis. A job title alone neither establishes nor defeats eligibility for separately proven qualified services.

Case Studies in the Modern Disallowance Era

Later cases show that technically demanding work does not automatically qualify for the research credit. Some claims fail because activities do not satisfy substantive tests; others fail because proof is insufficient. These are related but distinct problems, and neither should be described simply as a new rule against estimates.

Phoenix Design Group and the Uncertainty Threshold

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, concerned research credits associated with more than 200 projects. The parties tried a nonbinding sample of three projects, rather than individually trying every project. The court found insufficient proof of the relevant uncertainty, investigative activities, and qualifying experimentation in the work examined. Complexity and later design changes did not by themselves establish qualification. A practical response is to record what information was unavailable, what alternatives were considered, and what work resolved the uncertainty. This is a documentation recommendation, not a universal requirement to prepare a separate narrative before any project work begins.

Betz v. Commissioner and Funded Research

Betz v. Commissioner, T.C. Memo. 2023-84, involved shareholders of Catalytic Products International and resulted in disallowed research credits and accuracy-related penalties. The issues included failure to establish qualifying research and funded research for certain projects. Section 41(d)(4)(H) and Treasury Regulation § 1.41-4A(d) exclude research to the extent funded. A performer retaining no substantial rights