Research and development tax incentives require a distinction between qualifying experimental work, ordinary production, and the accounting treatment of related costs. Kollsman Instrument Corp. v. Commissioner, T.C. Memo. 1986-66, affirmed at 870 F.2d 89 (2d Cir. 1989), provides historical context for that distinction. It concerned tax years before the federal research credit was introduced in 1981. Its historical deduction issues should therefore be distinguished from the additional requirements imposed by modern IRC Section 41.

The Foundational Conflict in Kollsman Instrument Corp. v. Commissioner

In 1962, Kollsman contracted with the United States Air Force to make prototypes for a photo mapping system and a ground data subsystem for a fixed price of $6,718,480. Air Force changes to engineering, specifications, design, and production increased costs. Kollsman subsequently requested an equitable price adjustment of $5,528,205.

For 1963 through 1967, the company used a cost percentage-of-completion accounting method. As described in the appellate opinion, it compared annual costs with projected total project costs, applied that percentage to estimated contract revenue, and deducted the annual costs in calculating taxable income. Its anticipated contract price exceeded the amount ultimately agreed.

The Accounting and Deduction Dispute

After the contracting officer allowed only a relatively small adjustment in 1966, Kollsman claimed a $4,359,000 deduction that year to reverse previously accrued income. The published appellate opinion gives inconsistent figures for the officer’s initial adjustment, so a single precise amount should not be treated as settled. In September 1967, the parties agreed to a $2,000,000 increase, establishing a final contract price of $8,718,480.

The Second Circuit agreed that Treasury Regulation Section 1.451-1(a) placed the adjustment in 1967, when the exact amount was determined. It also rejected an inventory write-down: costs had already been deducted under the accounting method and were no longer in year-end inventory. Kollsman could not obtain another deduction by recharacterizing those costs.

The Section 174 R&D Controversy

The appellate opinion also describes the Commissioner’s treatment of $868,000 of claimed research and development expenses. Instead of allowing the full amount in 1966, the Commissioner allocated it across three years: $289,000 in 1966, $289,000 in 1967, and $290,000 in 1968. The resulting loss carryback produced a separate-year overassessment. The court rejected the proposed offset against the deficiency under the jurisdictional rules then applicable.

Later IRS Field Service Advice 200006023 cites the Tax Court’s Kollsman decision for the proposition that the contracts did not require the taxpayer to invent or design a product. That advice is nonprecedential. The language about investigative expenditures developing a product concept is attributed there to Mayrath v. Commissioner, rather than quoted as a holding of the Kollsman appellate opinion.

Kollsman is consequently useful historical context for distinguishing research from manufacturing to supplied specifications. It did not establish the modern four-part research credit test or a general rule that advanced manufacturing cannot qualify. Ordinary production expenses may receive tax treatment under other provisions even when they do not qualify for a research deduction or credit.

Evolution of the R&D Tax Credit Framework

The Economic Recovery Tax Act of 1981 introduced the research credit as Section 44F. It was redesignated Section 30 in 1984 and Section 41 in 1986. The 1986 legislation added requirements beyond the historical Section 174 expense standard. Research deductions and research credits remain separate provisions that can operate alongside each other, subject to coordination rules.

The applicable expense rules depend on the tax year. Before 2022, Section 174 generally allowed an election to deduct qualifying expenditures currently. For tax years beginning in 2022 through 2024, capitalization and amortization generally applied over five years for domestic research and 15 years for foreign research. Public Law 119-21 added Section 174A, permitting current deduction of domestic research or experimental expenditures for tax years beginning after December 31, 2024. An alternative election permits amortization over at least 60 months beginning when benefits are first realized. Transition provisions address earlier domestic balances and certain eligible small businesses.

The Four-Part Test for Qualified Research

Current Section 41(d)(1) refers to domestic research or experimental expenditures under Section 174A. Historical cases commonly describe the corresponding requirement as the Section 174 test. Each business component must meet the applicable requirements:

• Research expenditure test: The expenditures must meet the applicable research or experimental expense standard. Relevant uncertainty concerns the capability, method, or appropriate design for developing or improving the product or process.
• Business component test: The information sought must be intended to help develop or improve a product, process, computer software, technique, formula, or invention held for sale, lease, or license, or used in the taxpayer’s trade or business.
• Technological information test: The experimentation must fundamentally rely on physical or biological science, engineering, or computer science. Current regulations do not require discovery that advances the knowledge of the industry as a whole.
• Process of experimentation test: Substantially all relevant research activities must involve evaluating alternatives to resolve uncertainty for improved function, performance, reliability, or quality. Cosmetic changes alone do not meet the permitted-purpose requirement.

The “Substantially All” Requirement

The 80% requirement applies separately to each business component. The remaining activities must satisfy the applicable research expenditure standard and must not be otherwise excluded. It is not a rule that 80% of all company spending or all project construction costs must be experimental.

A qualifying process identifies uncertainty and alternatives and evaluates those alternatives. Modeling, simulation, and systematic trial and error can qualify. A formal laboratory or a universally prescribed document format is unnecessary, but merely encountering uncertainty does not establish experimentation. If a whole business component fails the test, the regulatory shrinking-back rule may permit evaluation of a qualifying subset.

Case Law Analysis and Comparative Precedents

Later decisions illustrate how research eligibility, expense classification, and evidentiary requirements interact. Their holdings must be read in light of the tax years and legal standards at issue, rather than treated as a single line of decisions uniformly tightening the law.

Union Carbide and the Production vs. Research Boundary

In Union Carbide Corporation & Subsidiaries v. Commissioner, T.C. Memo. 2009-50, the Tax Court found qualifying research in the anticoking and UCAT-J projects among five representative projects. These activities involved chemical manufacturing processes, including experimentation at commercial plants. The decision demonstrates that work at production scale can qualify when the relevant activity meets the statutory tests.

The court nevertheless rejected other projects and denied the additional production-material costs claimed as research supplies. A qualifying process experiment does not automatically convert all raw materials used in concurrent commercial production into qualified research expenses. Product and process business components, and their associated costs, require separate analysis.

Suder v. Commissioner and the Recognition of Incremental Innovation

Suder v. Commissioner, T.C. Memo. 2014-201, concerned telephone systems and related technology developed by Estech Systems, Inc. The Tax Court found that 11 of 12 representative projects satisfied the four-part test. The unsuccessful project involved cosmetic interface changes rather than a qualifying improvement.

The decision supports the eligibility of incremental technological development when uncertainty and experimentation are established. It does not require invention of entirely new scientific principles. The court also found that the CEO’s compensation exceeded the reasonable amount eligible for the claimed research expenses and reduced the wage amounts accordingly.

Little Sandy Coal and the Burden of Proof for Support Activities

In Little Sandy Coal Company, Inc. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), the Seventh Circuit affirmed the denial of credits following the Tax Court’s 2021 decision. Two vessels served as representative projects. The taxpayer did not provide a principled allocation demonstrating which employee activities were research and which constituted experimentation. Newness of the vessels and arbitrary estimates were insufficient.

The court also rejected the Tax Court’s categorical exclusion of direct support and supervision from the numerator of the substantially-all calculation. Such activities, including qualifying pilot-model production, can enter both numerator and denominator when the relevant conditions are met. However, merely labeling a vessel a pilot model does not establish experimentation. The dry dock’s submergence test was treated as verification that it operated as designed, rather than proof of an experimental evaluation of alternatives.

Implications for Defense and Aerospace Research

Section 41(d)(4)(H) excludes research to the extent it is funded by another person or governmental entity. A government customer does not automatically make every activity funded, and a technically demanding contract does not automatically make its costs eligible.

The Funded Research Doctrine

Treasury Regulation Section 1.41-4A(d), applied through Section 1.41-4(c)(9), requires examination of the relevant agreements and circumstances. For a contractor claiming its own research costs, the analysis generally concerns whether it retains substantial rights in the research results and bears the relevant financial risk. Payments contingent on successful research are treated differently from reimbursement payable regardless of success. Substantial rights need not be exclusive, but retaining only incidental experience is insufficient.

The late-2024 summary-judgment proceedings in Smith v. Commissioner and System Technologies, Inc. v. Commissioner illustrate the importance of governing law. In Smith, disputed contractual and foreign-law questions prevented summary judgment for the IRS. In System Technologies, remedies under Indiana law, including potential refunds upon failure, informed the court’s conclusion about payment contingency. These procedural rulings should not be characterized as blanket approval of every research expense or as automatic qualification of all contracts governed by similar law.

Aerospace R&D Complexity

Kollsman’s photo mapping and ground data prototypes illustrate the combination of technical work, changing customer specifications, and contractual cost exposure that can arise in aerospace projects. Those features warrant careful factual analysis; they do not independently establish credit eligibility.

Prototype materials and semiconductor fabrication costs also require attention to expense classification. Section 41 generally excludes land, improvements to land, and depreciable property from the definition of supplies. Whether a particular prototype, wafer, or mask qualifies depends on its use and tax classification. An IRS argument in a particular matter should not be presented as a universal rule for all prototype hardware.

Modern Documentation Standards and IRS Scrutiny

Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, illustrates the need to establish qualifying activities in engineering projects. The Tax Court denied the claimed credits where the evidence did not establish the statutory requirements for the projects examined. The decision applies existing qualification rules; it should not be described as independently creating a new universal documentation standard.

Identification of Technical Uncertainty

A useful record describes what was uncertain at the start of the relevant development activity, what information was available, and why alternative approaches required evaluation. Client preferences, budget constraints, scheduling problems, and ordinary compliance calculations do not by themselves establish technological uncertainty.

Emails, meeting records, calculations, design iterations, simulations, and test results can help establish the work performed. Their significance depends on their content and connection to the uncertainty being addressed. Calculations using known principles are not categorically excluded; they may form part of a qualifying evaluation of alternatives.

Substantiating the Process of Experimentation

Taxpayers should retain records that establish eligibility and expense amounts. Contemporaneous evidence is particularly useful, although the regulations do not prescribe one exclusive documentation format. A practical record should include:

• A clear identification of each business component and any relevant subset evaluated under the shrinking-back rule.
• The technical uncertainties, alternatives considered, evaluation methods, results, and resulting design decisions.
• Reliable links between qualifying activities and claimed wages, supplies, computer-use costs, and contract payments.
• A supportable allocation separating research, qualifying experimentation, and excluded activities.

Estimation principles associated with Cohan may assist in determining amounts when entitlement and a reasonable evidentiary basis are established. Estimates cannot supply missing proof that research qualifies. Little Sandy Coal cautions against unsupported percentages; it does not categorically exclude industries that use practical engineering methods rather than formal laboratory experiments.

Strategic Implications for Future R&D Tax Credit Applications

The practical lesson is to evaluate activities, costs, and agreements together. A sophisticated final product, an expensive prototype, or a project described internally as R&D is not enough. Each claimed business component and expense category needs support under the law applicable to its tax year.

Contractual Engineering for Credit Eligibility

Review contract terms and actual performance to determine whether the taxpayer:

• Retains substantial rights to use the research results.
• Bears relevant financial risk, considering payment conditions, acceptance provisions, refund remedies, termination rights, and governing law.
• Performs investigative development rather than solely manufacturing or adapting established designs to customer specifications.

A fixed-price label, milestone payment schedule, or intellectual-property clause is not conclusive in isolation. Contract drafting should accurately reflect the parties’ real rights and obligations. Meeting the funding rules still leaves the other Section 41 requirements to be established.

Managing Supply and Wage Expenditures

Supply claims must identify qualifying nondepreciable tangible property used in the conduct of qualified research. Commercial production materials should not be swept into the claim merely because an experiment occurred during the same run. Wage claims should distinguish actual research, direct supervision, and direct support from general administration, and substantiate reasonable compensation and activity allocations.

When measured on a cost basis, the substantially-all relationship can be expressed as:

Cost of research activities constituting elements of qualifying experimentation ÷ cost of relevant research activities ≥ 80%.

This activity calculation is separate from the computation of qualified research expenses. Little Sandy Coal distinguishes the activity fraction from supply costs and requires careful treatment of support and supervision. Passing the percentage test does not make every project expense eligible.

A simplified expense formula, where the ordinary contract-research percentage applies, is:

QREs = qualifying wages + qualifying supplies + qualifying computer-use payments + 65% of eligible payments for research performed by nonemployees.

The 65% rate is not universal: statutory exceptions can allow 75% or 100% for specified arrangements. “Contract research expenses” is itself a defined amount after applying the relevant percentage, so the percentage must not be applied twice. The resulting QRE total is an input to the credit computation, not the amount of the credit. The chosen computation method, base rules, exclusions, and [{“@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/”}]