The federal research credit under Internal Revenue Code Section 41 rewards qualifying research activities, but eligibility depends on the taxpayer’s business, the work performed, the costs claimed, and the available evidence. This study examines Lewin v. Commissioner alongside later research-credit decisions and the statutory rules applicable to domestic research expenditures.
Foundational Jurisprudence: The Trade or Business Threshold in Lewin and Early R&D Litigation
Two different Lewin decisions must be distinguished. Adolph and Joyce Lewin v. Commissioner, 569 F.2d 444 (7th Cir. 1978), concerned an untimely Tax Court petition and the resulting dismissal. It did not establish an R&D trade-or-business test and predates the research credit’s introduction in 1981.
The relevant research-expenditure decision is Nathan Lewin v. Commissioner, 335 F.3d 345 (4th Cir. 2003). It concerned I-Tech R&D Limited Partnership’s Section 174 deductions for 1984–1986. I-Tech financed research by five Israeli companies. The court affirmed disallowance because the partnership lacked a realistic prospect of commercially exploiting the technology in its own trade or business. The arrangement functioned as an investment rather than the partnership’s operating business.
Lewin therefore concerns historical research deductions, not a direct adjudication of Section 41 credits. Its lesson is that financing research and holding contractual rights do not, by themselves, establish the required business connection. It does not impose a universal requirement that every investor personally conduct experiments.
In Snow v. Commissioner, 416 U.S. 500 (1974), the Supreme Court recognized that the historical Section 174 standard could cover research before sales began. Its “in connection with” wording was broader than Section 162’s “carrying on” language. Snow supports qualifying development-stage businesses; it does not make every research investment deductible or creditable.
| Case Reference | Court / Jurisdiction | Primary Legal Significance for R&D |
|---|---|---|
| Lewin v. Commissioner (1978), 569 F.2d 444 | Seventh Circuit | Tax Court petition timeliness; not substantive R&D eligibility precedent. |
| Lewin v. Commissioner (2003), 335 F.3d 345 | Fourth Circuit | Historical Section 174 deductions denied where the partnership lacked a realistic prospect of conducting its own relevant business. |
| Snow v. Commissioner, 416 U.S. 500 (1974) | Supreme Court | Research deductions could be available before an enterprise began sales. |
| Kellett v. Commissioner, T.C. Memo. 2022-62 | Tax Court | Distinguished website start-up costs from operating expenses and addressed inadequate proof of qualifying research expenditures. |
These authorities require separate attention to the taxpayer’s actual business and the particular tax benefit claimed. A research study should explain the intended commercial use of the work and identify the relevant business components without treating historical deduction decisions as automatic authority for a credit.
The Statutory Evolution of Research and Experimental Expenditures: Interplay between Sections 174, 174A, and 41
The deduction for research expenditures and the research credit are separate benefits with different eligibility rules. An expenditure’s research-related character does not establish that it belongs in the narrower categories of qualified research expenses under Section 41.
For tax years beginning in 2022–2024, the Tax Cuts and Jobs Act generally required specified research expenditures to be capitalized and amortized over five years for domestic research or fifteen years for foreign research, using a midpoint convention. The original draft’s assertion that mandatory five-year domestic amortization continues indefinitely is outdated.
Public Law 119-21 added Section 174A. For tax years beginning after December 31, 2024, it generally allows an immediate deduction for domestic research or experimental expenditures. Taxpayers may instead elect capitalization and amortization over at least 60 months under the statutory conditions. Foreign research expenditures remain subject to fifteen-year amortization under Section 174. Transition provisions address previously capitalized domestic costs; eligibility, elections, and filing deadlines require separate review.
For illustration, under the five-year midpoint rule, the first-year deduction for a full twelve-month tax year equals domestic research costs divided by five, multiplied by one-half. Thus, $10,000,000 of costs produces a $1,000,000 first-year deduction. At an assumed 21% corporate rate, the deduction reduces tax by $210,000, compared with $2,100,000 for an immediate $10,000,000 deduction. The $1,890,000 difference illustrates timing only, before credits, losses, Section 280C adjustments, and other limitations. It is not a universal cash-tax result and should not be presented as the default treatment for domestic costs incurred in 2025 or later.
Current Section 41 refers to domestic research expenditures under Section 174A. Taxpayers should reconcile their research-expense schedules with their credit calculations while separately identifying excluded activities and costs. Section 280C coordinates deductions with the credit and provides a reduced-credit election.
Technical Rigor and the Four-Part Test: Qualitative Requirements for Qualified Research
Section 41(d) applies its requirements separately to each business component. The four-part framework examines qualifying research expenditures, technological information, a new or improved business component, and a process of experimentation directed toward an eligible improvement. Exclusions and the shrinking-back rule must also be considered.
The Permitted Purpose Test
The work must relate to a new or improved function, performance, reliability, or quality. Purely aesthetic changes do not qualify. A commercially valuable project can still fail if the evidence does not establish qualifying research. In Siemer Milling Co. v. Commissioner, T.C. Memo. 2019-37, descriptions of milling projects did not adequately establish the required experimental process.
The Technological in Nature Test
Experimentation must fundamentally rely on physical or biological science, engineering, or computer science. The rules do not generally require a scientific breakthrough or knowledge new to the industry. In Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, the central failures concerned proof of uncertainty and experimentation. The decision should not be described as finding that engineering calculations do not rely on engineering principles.
The Elimination of Uncertainty Test
The expenditure test asks whether the research seeks to resolve uncertainty about capability, method, or appropriate design. Explain what was unknown when the activity began. A demanding specification, a client’s changing preference, or an ordinary design choice does not alone establish research uncertainty.
The Process of Experimentation Test
The work must evaluate alternatives to resolve uncertainty through a qualifying process, such as modeling, simulation, or systematic testing. Neither repeated redesign nor the mere use of scientific tools establishes eligibility. Conversely, the statute does not require every project to follow a rigid laboratory sequence or to succeed. The evidence must connect the uncertainty, alternatives, evaluation, and resulting decisions.
Quantifying Innovation: The “Substantially All” Rule and the 80% Threshold
Treasury Regulation Section 1.41-4(a)(6) sets an 80% threshold for research activities constituting elements of experimentation, measured by cost or another consistently applied reasonable basis. This is an activity test, not a claim that 80% of a product’s physical features must be new. It is also distinct from the separate employee-wage rule and Form 6765’s business-component disclosure threshold.
The Impact of Little Sandy Coal Co. v. Commissioner
In Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the appellate court rejected categorical exclusion of direct support and supervision from the experimentation numerator. Relevant activities may count when their relationship to experimentation is established. Producing a pilot model does not automatically make every production activity experimental.
The court nevertheless affirmed disallowance because the taxpayer lacked a principled allocation of research activities. Its reasoning distinguishes qualified research expenses from the activities used in the fraction. Supply purchases are not themselves research activities, and broad estimates based on a vessel’s novelty do not establish experimental effort.
The Shrinking-Back Rule
Under Treasury Regulation Section 1.41-4(b)(2), a component that fails the requirements may be examined at its most significant subset of elements, continuing until the requirements are met or the most basic element is reached. Shrinking back requires evidence at that level. It does not cure missing proof or authorize choosing an unsupported favorable subset.
| Expenditure Category | Treatment in 80% Fraction (Numerator) | Treatment in 80% Fraction (Denominator) |
|---|---|---|
| Direct Research (Engineers) | Count the qualifying activities that constitute elements of experimentation. | Count relevant research activities within the applicable test. |
| Direct Supervision (Managers) | May count where the activity’s experimental connection is established. | May count where it is a relevant research activity; job title alone is insufficient. |
| Direct Support (Production) | May count for qualifying experimental support, including appropriate pilot-model work. | Include only relevant research activities, not all production effort. |
| General Administration | Ordinary administration is not experimentation. | Do not automatically include overhead merely because it is allocated to research costs; assess the underlying activity. |
The table summarizes activity classification. It is not a wage-allocation formula and should not be used to treat all engineering, supervisory, or production costs alike.
The Conflict over Funded Research: Analyzing Economic Risk and Substantial Rights
Section 41(d)(4)(H) and Treasury Regulation Section 1.41-4A(d) exclude funded research. For a contractor, two central questions are whether payment depends on successful research and whether the contractor retains substantial rights in the results. Review all relevant agreements and actual obligations; a contract label does not resolve either question.
Economic Risk and the Success Contingency
Fixed-price terms can place risk on the researcher, but they are not an automatic qualification rule. Ordinary cost-overrun risk and payment contingent on successful research must be distinguished. Examine acceptance conditions, termination rights, refunds, warranties, and enforceable remedies.
The decisions discussed in contemporary commentary concerning Smith v. Commissioner and System Technologies, Inc. v. Commissioner denied IRS motions for partial summary judgment on funded-research issues. They should not be characterized as final awards of the entire credits. The analyses considered contractual obligations and governing law; System Technologies included consideration of remedies under Indiana law.
Substantial Rights and IP Retention
Exclusive ownership is not always necessary, but the researcher must retain meaningful rights under the applicable rules. Incidental experience is insufficient. Nor should copyright ownership be assumed from authorship: assignments, licenses, and governing law can change the outcome. The Smith discussion illustrates why the agreement’s actual effect matters more than a general statement that documents belong to the customer.
Implications for Contract Drafting
- Identify the technical deliverables and what constitutes successful performance.
- Determine who bears the cost of unsuccessful research and whether payments can be withheld or recovered.
- State the parties’ rights to use research results and review all related agreements together.
- Check governing law rather than relying on standard warranty or milestone language as a guarantee of eligibility.
Substantiation and Documentation: From Cohan to the Revised Form 6765
Taxpayers generally bear the burden of supporting the claimed credit. The useful evidence connects the technical work, eligible activities, and expense calculation. Contemporaneous records are valuable, but there is no universal rule requiring a particular timesheet system for every claim.
The Cohan Rule and Estimates
Cohan permits estimation in appropriate circumstances when a factual foundation exists. It does not establish that research qualified or authorize unsupported allocations. Little Sandy Coal demonstrates the weakness of estimates without a principled basis. References to Eustace and other substantiation cases should not be converted into a categorical prohibition on all estimates or credible testimony.
IRS Procedural Shifts and Form 6765
The IRS’s former Tier I designation is historical and should not be presented as its current research-credit examination framework.
The IRS research-credit refund-claim requirements took effect in January 2022. Effective June 18, 2024, the IRS waived the initial requirement to identify each researcher and the information each individual sought. Claims still require identification of business components, the research activities for each component, and total qualified wage, supply, and contract-research expenses. The IRS may request more information during examination.
Under the December 2025 Form 6765 instructions, Section G is optional for tax years beginning before 2026 and required afterward, subject to exceptions. Exceptions cover qualifying small businesses making the specified payroll-tax election and certain original-return filers with no more than $1.5 million in group-level QREs and no more than $50 million in three-year average gross receipts.
When required, list components in descending QRE order until at least 80% of QREs is covered, subject to a 50-component maximum; aggregate the remainder. Separate amended-return requirements continue to apply. These disclosures do not establish eligibility, and there is no basis here to claim that the form automatically determines whether experimentation occurred.
Recommended Documentation Standards
| Documentation Category | Specific Records Required | Purpose in Audit |
|---|---|---|
| Labor/Wages | Relevant payroll records, task records, calendars, meeting notes, and supported time allocations. | Connect wages to qualified services and explain allocation methods. |
| Supplies | Relevant invoices, purchase orders, usage records, and project identifiers. | Show what was used in qualifying research and exclude ineligible costs. |
| Contractors | Executed agreements, amendments, invoices, payment records, acceptance terms, and rights provisions. | Support expense treatment and analyze economic risk and retained rights. |
| Technical Process | Design versions, test plans and results, prototypes, simulations, issue logs, and technical decisions. | Explain uncertainty, alternatives, evaluation, and the scope of experimentation. |
The retained table heading “Specific Records Required” describes useful categories, not a universal statutory checklist. An invoice, tax form, questionnaire, or meeting minute by itself does not prove credit eligibility.
In Moore v. Commissioner, T.C. Memo. 2023-20, the court rejected inclusion of the president/COO’s wages because the evidence did not reliably establish qualifying services and time. The Seventh Circuit affirmed in 2024. The practical lesson is to substantiate executive participation rather than infer qualification from seniority or general familiarity with projects.
Strategic Implications for Future R&D Tax Credit Applications
The relevant cases address different issues: business connection, technical qualification, activity allocation, contractual funding, and proof. A defensible study should analyze each issue separately and then reconcile the results.
Granular Time Tracking and Activity Mapping
Use records sufficiently detailed to explain what employees did and how claimed percentages were calculated. Project and task tracking can help, especially where employees combine research with production or administration. Do not assert that every hour must be assigned to a single statutory test: the tests assess research activities together. Explain direct supervision and support with concrete examples.
The Integration of Sections 174, 174A, and 41 Compliance
Separate domestic and foreign research, apply the correct tax-year rules, and reconcile the deduction or amortization treatment with the credit. Preserve schedules for historical capitalized costs and relevant elections. An automated workflow can assist classification, but it cannot substitute for reviewing technical facts and statutory exclusions.
Re-evaluating Contractor Relationships and Governing Law
Review rights and financial obligations before adopting a funded-research position. International contracts may require analysis of the chosen law, but a foreign client or foreign governing law does not itself determine research location. Research performed outside the permitted geographic area remains excluded from the federal credit even if contractual risk and rights requirements are satisfied.
Avoiding the Routine Design Trap
Describe actual uncertainty and the evaluation undertaken to resolve it. Distinguish experimentation from adaptation, ordinary calculations, and routine implementation. Use project-specific facts rather than inserting statutory terminology into a generic narrative. Preserve unsuccessful tests as well as successful results, and consider subcomponents where the evidence supports shrinking back.
