The Foundation of Research Incentives in the Internal Revenue Code

The federal research and development (R&D) tax credit under Internal Revenue Code Section 41 encourages investment in qualified research. It is separate from the rules governing deductions for research and experimental expenditures. Historically those deduction rules were contained in Section 174; for tax years beginning after December 31, 2024, Section 174A generally permits immediate deduction of domestic research and experimental expenditures, while Section 174 continues to require amortization of foreign research expenditures. Credit eligibility depends on the statute, Treasury regulations, and the facts established in relevant judicial decisions.

The credit, formally titled the Credit for Increasing Research Activities, originated in the Economic Recovery Tax Act of 1981 and was made permanent by the Protecting Americans from Tax Hikes Act of 2015. The research-expenditure threshold is only one element of Section 41 eligibility. Harris v. Commissioner, 16 F.3d 75 (5th Cir. 1994), addressed the historical Section 174 deduction and applied an existing realistic-prospect analysis; it did not originate that analysis or decide entitlement to the Section 41 credit. Section 41 also has its own trade-or-business requirements, including a special rule for certain startup in-house research expenses under Section 41(b)(4).

Research credit claims require substantiation of both qualified activities and eligible costs. The IRS formerly treated certain research credit issues as Tier I issues, but discontinued its tiered issue management process in 2012. That designation should not be described as current policy. Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, illustrates the risks of failing to connect technical activities with the statutory requirements. It does not establish a universal contemporaneous-timesheet requirement or abolish reasonable estimation. Maintaining reliable records during development remains a practical way to support a claim.

The Realistic Prospect Test: Analyzing Harris v. Commissioner (16 F.3d 75)

Harris involved Research One Limited Partnership and CemCom Research Associates, Inc., which developed cementitious-composite technology for aerospace tooling and shipping pallets. Research One paid CemCom to undertake the research and arranged for commercialization through licensing. Harris, a limited partner, deducted his share of the partnership’s research expenditures. The courts denied the deduction because the expenditure lacked the required connection to the partnership’s own trade or business. The electronic-device project involving LDL Research & Development II and Larson-Davis belongs to a different case, LDL Research & Development II, Ltd. v. Commissioner, 124 F.3d 1338 (10th Cir. 1997).

The Fifth Circuit considered the historical Section 174 phrase “in connection with” a trade or business, which is broader in timing than Section 162’s requirement of carrying on a business. Research may precede commercial operations. Nevertheless, funding another entity’s research does not by itself establish the taxpayer’s own business connection. In Harris, the question was whether the partnership realistically would conduct a business exploiting the research results, rather than merely hold a passive investment.

Harris examined the economic substance of the research and licensing arrangements. A profit motive alone did not establish the required business nexus. The analysis considered whether Research One realistically would market the technology itself and whether its prearranged licensing transaction amounted to a regularly conducted business. The court relied on earlier decisions, including Spellman, rather than announcing a new two-pronged credit test.

The Fifth Circuit affirmed the denial of Harris’s deduction. Research One’s prearranged license back to CemCom did not establish a regularly conducted licensing business. This historical Section 174 holding must be distinguished from the separate substantial-rights and financial-risk requirements applicable to funded research under Section 41.

For startups and special-purpose entities, the practical lesson is to substantiate how research relates to the entity’s own intended business. Commercial plans, contractual rights, resources, and actual operations may all be relevant. Harris does not impose a universal requirement that every claimant hire engineers, build manufacturing facilities, or perform all research internally. Outsourced research can qualify when the applicable deduction or credit requirements are met.

Defining the Four-Part Test: Section 41 and Qualified Research Activities

Qualified research under Section 41 must satisfy four requirements at the business-component level, subject to the shrinking-back rule. Expenses also must fall within eligible statutory categories and avoid the exclusions in Section 41(d)(4). A failure at the overall component level does not necessarily eliminate qualifying research within an identifiable subset.

The research-expenditure requirement addresses uncertainty about capability, method, or appropriate design. The inquiry considers the information available to the taxpayer at the outset. Historically described as the Section 174 test, this requirement must be read with the statutory version applicable to the claim year; current Section 41(d)(1)(A) refers to expenditures eligible for treatment under Section 174A. Technical uncertainty differs from uncertainty about demand, financing, or customer preferences.

The second prong, known as the Technological in Nature Test, requires that the research fundamentally rely on principles of the physical or biological sciences, engineering, or computer science. This distinguishes qualified research from activities based on the social sciences, economics, or humanities, which are explicitly excluded from the credit. Documentation for this test must describe how the specific scientific or engineering principles were applied to address the technical challenges identified in the first prong.

The third prong is the Business Component Test, or the Permitted Purpose Test. The research must be intended to be useful in the development of a new or improved “business component” held for sale, lease, or license, or used in the taxpayer’s trade or business. The research must relate to a new or improved function, performance, reliability, or quality. This requirement ensures that the credit is focused on activities that enhance the utility of products or processes rather than purely aesthetic or “routine” modifications.

The Process of Experimentation Test requires substantially all of the relevant research activities to constitute elements of a process of experimentation for a permitted purpose. Treasury Regulation Section 1.41-4 generally sets the threshold at 80%, measured on a cost or another consistently applied reasonable basis. The process identifies uncertainty and one or more alternatives, then systematically evaluates them through methods such as modeling, simulation, or trial and error. No particular research label or number of failed trials establishes qualification by itself.

Little Sandy Coal Co. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), affirming T.C. Memo. 2021-15, illustrates the difficulty of proving the substantially-all requirement in shipbuilding. Novelty and prototype status did not establish what proportion of activities constituted experimentation. Taxpayers should identify the relevant activities and substantiate a reasonable measurement rather than assume that all work on a new vessel qualifies.

The Modern Evidentiary Standard: Phoenix Design Group, Inc. v. Commissioner

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, applied the qualified-research requirements to mechanical, electrical, plumbing, and fire-protection engineering. The dispute included more than 200 projects; the court examined three selected trial projects. It is a fact-specific illustration of the statutory tests, not a claim that this was the latest decision or that a new engineering-specific standard replaced the statute.

The court found no qualified research in the three trial projects. Those findings did not automatically determine qualification for every other project in the dispute. Accuracy-related penalties followed under the parties’ stipulation when none of the trial projects qualified; the decision should not be presented as imposing an automatic penalty whenever an engineering credit is denied.

The court distinguished unresolved project inputs and ordinary design choices from qualifying technical uncertainty. An unfinished design or the possibility of later revisions does not necessarily establish uncertainty about capability, method, or appropriate design. Qualification depends on the actual information available and the activities undertaken to resolve the identified unknowns. The decision applied the existing uncertainty requirement rather than narrowing the statutory definition for all taxpayers.

PDG’s general description of its six-stage design process did not establish a qualifying process of experimentation for the trial projects. Its activity records did not adequately show the systematic evaluation needed to resolve technical uncertainty. A standard engineering workflow may contain qualifying research, but the workflow’s name or sequence alone is insufficient. The regulations do not require a fixed number of iterations or a document explicitly labeled a hypothesis.

The technological-in-nature requirement remains separate from the uncertainty and experimentation requirements. Employing engineers, performing difficult calculations, or designing complex facilities does not independently prove that the four-part test is satisfied. The practical focus is the substance of the work and the supporting evidence. This principle should not be described as a new holding that engineering principles were absent from every PDG activity.

For architectural and engineering firms, Phoenix Design Group reinforces the need to identify actual technical uncertainties and explain how alternatives were systematically evaluated. Routine design and qualified research may occur within the same engagement. The distinction requires evidence about particular activities, rather than an assumption based on the firm’s industry.

Documentation and the Burden of Proof: From Reconstruction to Contemporaneous Tracking

Taxpayers generally bear the burden of substantiating their claimed research credits, subject to applicable burden-shifting rules. Treasury Regulation Section 1.41-4(d) requires records in sufficiently usable form and detail. Contemporaneous business records are valuable, but the law does not categorically prohibit retrospective interviews, reasonable allocations, or credible testimony. Their reliability and connection to actual qualified activities determine their usefulness.

In Siemer Milling Co. v. Commissioner, T.C. Memo. 2019-37, the taxpayer failed to establish qualification, including the required process of experimentation. The sound lesson is to document how particular activities satisfy the tests, rather than treat the absence of one specific timekeeping format as automatically fatal. In evaluating litigation such as the Kyocera disputes, a government allegation or motion must be distinguished from an adjudicated holding; it does not establish a categorical prohibition on estimates.

Practical recordkeeping measures include the following. These are recommendations for substantiation, not a statutory requirement to use a particular software system or template:

Project-Based Timesheets: Tracking employee time at the project and activity level, with clear descriptions that distinguish between qualifying research and non-qualifying tasks like marketing or routine production.

Technical Repositories: Maintaining records of design iterations, modeling results, testing protocols, and laboratory notes that document the “failures” and “alternatives” explored.

Narrative Descriptions: Developing contemporaneous narratives for each business component that explicitly map the activities to the Four-Part Test, identifying the specific uncertainty being addressed.

Electronic Records: Retaining e-mails, meeting minutes, and code repositories (for software) that evidence the evaluative process and the collaboration between researchers.

Form 6765 now requests additional claim information. Tax and technical teams should reconcile the form with retained business records and follow the instructions applicable to the claim year. More detailed disclosures do not replace the requirement to establish eligibility, and the form does not mandate a single accounting or timekeeping system.

Comparative Analysis: Successful and Unsuccessful R&D Claims

Suder v. Commissioner, T.C. Memo. 2014-201, illustrates that a taxpayer may establish qualified research and employee participation through credible evidence. Senior executives’ activities can qualify when they actually conduct, directly supervise, or directly support qualified research. General management is insufficient. Suder was not an unqualified taxpayer victory: the court also limited the wages taken into account because part of the chief executive’s compensation was unreasonable.

Suder demonstrates the potential role of substantiated estimates. Eustace v. Commissioner, T.C. Memo. 2001-66, affirmed, 312 F.3d 905 (7th Cir. 2002), illustrates that a court need not estimate research credits without an adequate evidentiary foundation. The Cohan principle is not a substitute for proving entitlement or supplying a rational basis for an estimate. Neither case supports claiming that reasonable estimation has been abolished.

Little Sandy Coal likewise cautions against using a product’s novelty as a substitute for evidence about activities. Where the overall business component does not satisfy the requirements, the shrinking-back rule may permit consideration of an identifiable subset supported by the record.

The Shrinking-Back Rule and Granular Substantiation

The “Shrinking-Back Rule,” found in Treasury Regulation Section 1.41-4(b)(2), is a critical safety valve for taxpayers. It provides that if a business component fails the Four-Part Test as a whole, the test may be applied to a subset of that component. This process continues until a subcomponent satisfies the test or the most basic level of the component is reached.

For example, research on a ship’s propulsion system might qualify even if the overall vessel does not satisfy the tests. The taxpayer must identify that subset and establish its research activities and associated expenses. The 80% test is not simply a comparison of experimental work with every dollar of total construction cost; its scope and measurement follow the regulation and applicable case law.

Phoenix Design Group also illustrates that shrinking back depends on evidence identifying a qualifying subset. The rule cannot cure a record that fails to establish experimentation at any supported level. Taxpayers can improve their substantiation by preserving records for potentially qualifying subsystems while the work is underway.

Contractual Rights and the Funded Research Doctrine: Section 41(d)(4)(H)

The “Funded Research” exclusion is one of the most complex areas of Section 41. It prevents a taxpayer from claiming the credit for research to the extent it is funded by another person or governmental entity. For companies performing research on behalf of clients—such as engineering firms, software houses, or CROs—the eligibility for the credit depends on two key factors: financial risk and substantial rights.

Under Treasury Regulation Section 1.41-4A(d), payments contingent on successful research are treated differently from payments for research regardless of outcome. Fixed-price, [{“@context”:”https://schema.org”,”@type”:”VideoObject”,”name”:”What is the R&D Tax Credit?”,”description”:”The research and experimentation tax credit, most frequently known as the R&D tax credit, is a dollar-for-dollar reduction of your tax liability.”,”thumbnailUrl”:[“https://i.ytimg.com/vi/mzGRiA_MUl4/sddefault.jpg”,”https://www.dropbox.com/s/n1iyfxaeo6rm5tg/Fed%20-%20US%20Flag.jpg?raw=1″],”uploadDate”:”2019-10-14T00:00:00+00:00″,”duration”:”PT3M54S”,”contentUrl”:”https://www.youtube.com/watch?v=mzGRiA_MUl4″,”embedUrl”:”https://www.youtube.com/embed/mzGRiA_MUl4″,”publisher”:{“@type”:”Organization”,”name”:”Swanson Reed”,”url”:”https://swansonreed.com”,”logo”:{“@type”:”ImageObject”,”url”:”https://swansonreed.com/logo.png”}},”transcript”:”the research and experimentation tax credit most frequently known as the r d tax credit is a dollar for dollar reduction of your tax liability it was established in 1981 as an incentive for companies to invent create and innovate within the united states here at swanson read the biggest problem we see as specialized r d tax advisors is self-censorship companies believing they are not eligible for the r d tax credit when in reality the irs has a very broad definition of what it considers r d does your company design engineer or manufacture its own products do you look to improve the functionality performance or reliability of these products do you create new or improved processes in order to make things better faster or cheaper do you develop prototypes or computer generated models or do you develop software technology or other intellectual property if you answered yes to any of the previous questions your company may qualify for the r d tax credit congress has created a four-part test to help you identify activities that would be considered qualified research your work must satisfy these four main requirements it must be technological in nature a process of experimentation there must be technical uncertainty and a permitted purpose let’s go through these one by one one technological in nature this means the process of experimentation used to discover such information fundamentally relies on principles of the physical or biological sciences engineering or computer science two process of experimentation this is defined as a systematic process designed to evaluate one or more alternatives to achieve a result where the capability or method of achieving that result or the design of that result is uncertain the beginning of the research three technical uncertainty as a taxpayer you must intend to discover information that would eliminate uncertainty concerning the development or improvement of the business component and four permitted purpose it is a qualified purpose if research relates to a new or improved function increased performance enhanced reliability or enhanced quality it is not a qualified purpose if research relates to aesthetics meaning style taste cosmetics or seasonal design companies that are benefiting from the credit are typically receiving a minimum in the tens of thousands of dollars of federal tax credits each year so don’t pass up this chance to significantly lower your tax liability and improve your cash flow call swanson read representative today for an assessment”},{“@context”:”https://schema.org”,”@type”:”AccountingService”,”name”:”Swanson Reed”,”description”:”One of the largest Specialist R&D Tax Credit advisory firms in the United States, exclusively providing R&D Tax Credit claim preparation and audit compliance solutions for over 30 years.”,”url”:”https://www.swansonreed.com”,”logo”:”https://swansonreed.com/logo.png”,”image”:”https://www.swansonreed.com/wp-content/uploads/2025/03/Swanson-Reed-Specialist-RD-Tax-Credit-Advisors-is-the-largest-in-the-United-States.jpg”,”telephone”:”+1-800-986-4725″,”email”:”damian@swansonreed.org”,”priceRange”:”$195 – $395/hr”,”hasMap”:”https://www.google.com/maps?cid=10257056501192939853″,”address”:{“@type”:”PostalAddress”,”streetAddress”:”1120 South Freeway, Suite 123″,”addressLocality”:”Fort Worth”,”addressRegion”:”TX”,”postalCode”:”76104″,”addressCountry”:”US”},”areaServed”:[{“@type”:”City”,”name”:”New York City, NY”},{“@type”:”City”,”name”:”Los Angeles, CA”},{“@type”:”City”,”name”:”Chicago, IL”},{“@type”:”City”,”name”:”Houston, TX”},{“@type”:”City”,”name”:”Phoenix, AZ”},{“@type”:”City”,”name”:”Philadelphia, PA”},{“@type”:”City”,”name”:”San Antonio, TX”},{“@type”:”City”,”name”:”San Diego, CA”},{“@type”:”City”,”name”:”Dallas, TX”},{“@type”:”City”,”name”:”Austin, TX”},{“@type”:”City”,”name”:”Jacksonville, FL”},{“@type”:”City”,”name”:”San Jose, CA”},{“@type”:”City”,”name”:”Fort Worth, TX”},{“@type”:”City”,”name”:”Columbus, OH”},{“@type”:”City”,”name”:”Charlotte, NC”},{“@type”:”City”,”name”:”Indianapolis, IN”},{“@type”:”City”,”name”:”San Francisco, CA”},{“@type”:”City”,”name”:”Seattle, WA”},{“@type”:”City”,”name”:”Denver, CO”},{“@type”:”City”,”name”:”Washington, D.C.”}],”founder”:{“@type”:”Person”,”name”:”Joe William Norris”},”employee”:{“@type”:”Person”,”name”:”Damian Smyth”,”jobTitle”:”CEO”,”image”:”https://www.swansonreed.com/wp-content/uploads/2026/03/Damian-Smyth-Swanson-Reed.jpg”},”hasOfferCatalog”:{“@type”:”OfferCatalog”,”name”:”R&D Tax Credit Advisory Services”,”itemListElement”:[{“@type”:”Offer”,”url”:”https://www.swansonreed.com/services/our-fees/”,”itemOffered”:{“@type”:”Service”,”name”:”R&D Tax Credit Consulting & Preparation Services”,”description”:”State and federal R&D tax credit consulting and preparation services for all 50 states.”,”url”:”https://www.swansonreed.com/services/our-fees/”},”priceSpecification”:{“@type”:”UnitPriceSpecification”,”priceCurrency”:”USD”,”minPrice”:”195″,”maxPrice”:”395″,”referenceQuantity”:{“@type”:”QuantitativeValue”,”unitCode”:”HUR”,”value”:”1″}}},{“@type”:”Offer”,”itemOffered”:{“@type”:”Service”,”name”:”R&D Tax Credit Audit Advisory Services (creditARMOR)”,”description”:”A sophisticated R&D tax credit insurance and AI-driven risk management platform to mitigate audit exposure.”}}]}},{“@context”:”https://schema.org”,”@type”:”Service”,”name”:”R&D Tax Credit Training for CPAs”,”description”:”Free CPE credit training webinars and resources on the R&D Tax Credit for CPAs and Enrolled Agents.”,”provider”:{“@type”:”Organization”,”name”:”Swanson Reed”,”url”:”https://swansonreed.com”,”logo”:{“@type”:”ImageObject”,”url”:”https://swansonreed.com/logo.png”}},”url”:”https://www.swansonreed.com/research-tax-credit/webinars/free-cpe-credits/”},{“@context”:”https://schema.org”,”@type”:”Service”,”name”:”R&D Tax Credit Training for CFPs”,”description”:”Free CE credit training webinars and resources on the R&D Tax Credit for Certified Financial Planners.”,”provider”:{“@type”:”Organization”,”name”:”Swanson Reed”,”url”:”https://swansonreed.com”,”logo”:{“@type”:”ImageObject”,”url”:”https://swansonreed.com/logo.png”}},”url”:”https://www.swansonreed.com/research-tax-credit/webinars/free-ce-credits-for-cfps/”},{“@context”:”https://schema.org”,”@type”:”Service”,”name”:”R&D Tax Credit Training for SMBs”,”description”:”Training webinars and educational resources on the R&D Tax Credit for Small and Medium-sized Businesses.”,”provider”:{“@type”:”Organization”,”name”:”Swanson Reed”,”url”:”https://swansonreed.com”,”logo”:{“@type”:”ImageObject”,”url”:”https://swansonreed.com/logo.png”}},”url”:”https://www.swansonreed.com/research-tax-credit/webinars/small-business-guide/”}]