Norwest Corp. v. Commissioner

The historical trajectory of the Research and Development (R&D) tax credit in the United States is fundamentally defined by the tension between legislative intent and administrative enforcement. At the center of this tension lies the landmark case of Norwest Corp. v. Commissioner, 110 T.C. 454 (1998), a decision that crystallized the rigorous standards for internal use software and ignited a decade-long debate over the definition of technological discovery. As the Internal Revenue Code (IRC) Section 41 credit has evolved from a temporary incentive to a permanent fixture of the tax code, the precedents established in Norwest continue to inform the Internal Revenue Service (IRS) Audit Technique Guides, Treasury Regulations, and the most recent legislative updates under the One Big Beautiful Bill Act (OBBBA) of 2025.

The Genesis of Section 41 and the Internal Use Software Dilemma

The federal research credit was originally introduced in 1981 to bolster domestic innovation and economic competitiveness. In its nascent form, the credit was broadly applicable, but the rapid digitization of the American economy necessitated more precise definitions of what constituted "qualified research." By the mid-1980s, the proliferation of internal software development for administrative and management purposes led Congress to implement stricter barriers to entry. The Tax Reform Act of 1986 (TRA 1986) introduced the "four-part test" for all research and a specific "high threshold of innovation" test for software developed primarily for a taxpayer's internal use.

The four-part test, which remains the cornerstone of Section 41 today, requires that research be deductible under Section 174, be undertaken to discover information that is technological in nature, be intended for use in developing a new or improved business component, and involve a process of experimentation. For internal use software (IUS), the additional three-part test requires the software to be innovative, involve significant economic risk, and not be commercially available.

Fact-Finding and Judicial Reasoning in Norwest Corp. v. Commissioner

In Norwest Corp. v. Commissioner, 110 T.C. 454 (1998), the United States Tax Court was tasked with evaluating 67 software development projects undertaken by Norwest Corporation, a bank holding company, and its subsidiaries between 1986 and 1991. To manage the litigation effectively, the parties selected eight representative projects to serve as the basis for the court's determination of eligibility for the research credit.

Detailed Review of the Sample Software Projects

The projects evaluated in Norwest represented a broad spectrum of banking operations, ranging from routine payroll administration to complex, integrated data systems. The court conducted a granular analysis of each project against the statutory and regulatory requirements, emphasizing that the burden of proof rests with the taxpayer to demonstrate that each activity satisfies all seven components of the IUS test.

The Success of the Strategic Banking System (SBS)

The SBS customer module was the only project to satisfy the court's rigorous interpretation of the IUS requirements. The court recognized that Norwest, in collaboration with Electronic Data Systems Corp (EDS), undertook a project of unprecedented scale and complexity for the banking industry at the time. The project involved significant "technical uncertainty" because it required the integration of customer data in a way that standard software products could not achieve.

Crucially, the court found that Norwest bore the "significant economic risk" associated with the project. Norwest committed substantial resources, and there was a high degree of uncertainty regarding whether the technical challenges could be overcome to recover those resources. The court also noted that Norwest retained substantial rights in the research results through a perpetual license, further supporting the claim that the research was not "funded" by a third party in a manner that would disqualify the credit.

The Failure of Routine Software Development

For the remaining seven projects, the court arrived at a conclusion that continues to serve as a warning to taxpayers. These projects were categorized as "routine software development," maintenance, or the simple adaptation of existing technology. The court emphasized that a "cookbook" approach—using known methodologies and skilled practice to achieve a predictable result—does not constitute a process of experimentation. For example, the "Cyborg Payroll" project failed because it primarily involved the installation and customization of a commercially available package, which did not satisfy the "discovery" or "innovation" requirements.

The Discovery Test Controversy: From Norwest to McFerrin

Perhaps the most significant and controversial legacy of the Norwest decision was the court's narrow interpretation of the "discovery" requirement under Section 41(d)(1)(B)(i). The court held that to satisfy the "technological in nature" test, the research must be undertaken to discover information that is not already known in the field of science or engineering.

The "New to the World" Standard

The Norwest court argued that the knowledge gained from research must exceed, expand, or refine the common knowledge of skilled professionals in the particular field—in this case, computer science. The discovery of information that was new to the taxpayer, but already known to the industry, was deemed insufficient. This interpretation set an exceptionally high bar, effectively suggesting that the research credit was reserved for breakthroughs that advanced the boundaries of human knowledge rather than the development of proprietary business technology.

This standard was further reinforced in United Stationers, Inc. v. United States, 163 F.3d 444 (7th Cir. 1998), where the court explicitly aligned with Norwest. The United Stationers court held that "discovery" demands more than "mere superficial newness" and must connote "innovation in underlying principle". Because United Stationers was merely modifying a commercially available software program (DCS) to improve its internal inventory control, the court ruled the activities did not constitute qualified research.

The Regulatory Reversal and the "Elimination of Uncertainty"

The Discovery Test triggered a decade of administrative friction. In 2001, the Treasury Department issued T.D. 8930, which codified the strict discovery standard. However, after intense public criticism and a change in administration, the Treasury suspended these rules within 30 days. By 2004, the Treasury issued T.D. 9104, which definitively abandoned the requirement that research must expand the "common knowledge" of a field.

The new standard, which remains in effect, clarifies that research is undertaken to discover information if it is intended to "eliminate uncertainty" regarding the development or improvement of a business component. This uncertainty can relate to the capability of achieving a result, the method of achieving it, or the appropriate design of the component. The case of United States v. McFerrin, 570 F.3d 672 (5th Cir. 2009), finally brought judicial closure to this era. The Fifth Circuit vacated a lower court's decision that had relied on the outdated Norwest discovery standard, ruling that the lower court had applied "overly strict definitions". McFerrin confirmed that research only needs to solve a technical uncertainty for the taxpayer’s own business, not the entire industry.

Evolution of the Internal Use Software (IUS) Standard: T.D. 9786

While the "discovery" hurdle was lowered, the "internal use software" hurdle remained complex. In 2016, the Treasury issued T.D. 9786, which provided much-needed clarity on what constitutes IUS and how the high threshold of innovation (HTI) test should be applied in a modern software environment.

Redefining General and Administrative Functions

The 2016 regulations significantly narrowed the scope of the IUS exclusion. Software is now considered IUS only if it is developed for use in "general and administrative functions" that support the conduct of the taxpayer’s trade or business. These functions are strictly limited to three categories:

Crucially, software developed to enable a taxpayer to interact with third parties (such as customers, vendors, or the general public) is not considered internal use software. This includes mobile apps, online banking portals, and e-commerce platforms. This distinction is a direct response to the integration challenges highlighted in Norwest, recognizing that in the modern economy, software integration often serves a dual purpose of internal efficiency and external service delivery.

Dual-Function Software and the 25% Safe Harbor

Recognizing the complexity of modern software architecture, the 2016 regulations introduced the concept of "dual-function software"—software that serves both internal administrative functions and external interaction functions. The regulations provide a safe harbor for such software: if the taxpayer can reasonably anticipate that the third-party use of the software will account for at least 10% of its total use, the taxpayer may include 25% of the qualified research expenditures (QREs) for the dual-function portion without having to satisfy the HTI test.

The Modern HTI Test

For software that remains classified as IUS, the 2016 regulations refined the three-part HTI test originally derived from the Norwest era. The taxpayer must demonstrate:

• Innovation: The software must be intended to result in a reduction in cost or improvement in speed that is substantial and economically significant.
• Significant Economic Risk: The taxpayer commits substantial resources and there is "substantial uncertainty," because of technical risk, that the resources will be recovered within a reasonable period.
• Commercial Availability: The software cannot be purchased, leased, or licensed and used for the intended purpose without modifications that meet the first two requirements.

The IRS Audit Landscape and High-Risk Software Categories

In recent years, the IRS has moved from broad legal challenges like the "Discovery Test" to a more tactical, risk-based approach to auditing R&D claims. The Large Business & International (LB&I) division has issued specific directives and Audit Technique Guides (ATGs) that rank software development activities based on their likelihood of being disallowed.

The 17 High-Risk Software Development Activities

The IRS has identified 17 specific activities that are considered "high risk" of not constituting qualified research. These categories are used by examiners to focus limited audit resources on claims that appear to be routine or administrative in nature.

Legislative Changes and the Return of Immediate Expensing: OBBBA 2025

The most significant shift in the R&D landscape since the 1986 Tax Reform Act occurred in the early 2020s. Under the Tax Cuts and Jobs Act (TCJA) of 2017, the long-standing ability to immediately deduct research and experimental (R&E) expenditures under Section 174 was eliminated. Starting in 2022, taxpayers were required to capitalize and amortize domestic R&E over five years and foreign R&E over 15 years. This change also officially classified all software development as a Section 174 activity, making the capitalization requirement mandatory regardless of whether a credit was claimed.

The Restoration of Section 174A

The One Big Beautiful Bill Act (OBBBA), signed into law on July 4, 2025, has dramatically restored the incentive structure for domestic research. The OBBBA introduced Section 174A, which permanently allows for the immediate expensing of domestic R&E expenditures for tax years beginning after December 31, 2024.

The restoration of immediate expensing under Section 174A serves to maximize the synergy with the Section 41 credit. While Section 280C still requires a reduction in the deduction equal to the amount of the credit (or an election to reduce the credit itself), the ability to avoid long-term amortization provides a massive cash-flow benefit for innovative companies.

Procedural Rigor: The New Section G and Documentation Standards

As the financial benefits of the R&D credit have been restored, the IRS has implemented stricter procedural requirements to ensure that only legitimate research is credited. The 2024 and 2025 updates to Form 6765 (Credit for Increasing Research Activities) represent a significant shift toward upfront disclosure.

Mandatory Disclosure of Business Components

For tax years beginning after 2024, the IRS has introduced "Section G," which is mandatory for businesses with over $1.5 million in QREs. Taxpayers are now required to break down their research expenses by specific "business component" rather than reporting them in aggregate.

For each business component, the taxpayer must:

• Identify whether it is a product, process, software, technique, formula, or invention.
• Describe the specific technical uncertainty encountered at the outset.
• Provide a narrative of the "information sought to be discovered" through the research activities.
• Identify the total qualified employee wages, supply costs, and contract research expenses attributable to that specific component.

This requirement is a direct institutionalization of the Norwest project-based evaluation. By requiring this information at the time of filing, the IRS is forcing taxpayers to perform the granular analysis that was previously only required during a multi-year audit.

Contemporary Jurisprudence on Substantiation: Eustace and Betz

Two recent cases highlight the continued importance of the "process of experimentation" and the danger of failing to provide contemporary documentation. In Betz v. Commissioner, the Tax Court issued a decisive ruling disallowing R&D credits for a company that designed custom air pollution control systems. The court found that the taxpayer failed to establish that its products were "pilot models" or that at least 80% of its research activities followed a structured process of experimentation.

Similarly, in Eustace v. Commissioner, T.C. Memo. 2001-66, the court emphasized that oral testimony alone, without supporting project logs, emails, or design iteration notes, is often insufficient to meet the burden of proof. These cases underscore that while the McFerrin decision allows for the "estimation" of expenses, the taxpayer must first prove that qualified research occurred through credible, contemporaneous records.

Strategic Implications for Future R&D Applications

The evolution of the R&D tax credit from Norwest to the OBBBA of 2025 provides several strategic lessons for taxpayers seeking to claim the credit in the modern regulatory environment.

The Integration Threshold

The Norwest SBS project succeeded because it was a "high-tech" integration that went beyond the routine application of software principles. Future applications should focus on projects that involve "new data constructs," "architectural breakthroughs," or the resolution of "design spirals" where the integration of multiple components creates systemic uncertainty. Projects that primarily involve the "configuration" of third-party platforms remain the highest risk for disallowance.

The Domestic Incentive

The bifurcated treatment of Section 174 costs under the OBBBA creates a powerful incentive for domestic research. Foreign-based research activities now face a double penalty: they are excluded from the Section 41 credit computation (the "Foreign Research Exclusion") and they must be amortized over 15 years. Companies should meticulously track the location of their software development activities to ensure that U.S.-based costs are immediately expensed under Section 174A.

Documentation as a Compliance Shield

The introduction of Section G on Form 6765 means that "prepackaged R&D studies" that use general estimates or cost-center allocations are no longer sufficient. Taxpayers must implement project-tracking systems that capture:

• The technical goals of each sprint or development phase.
• The "alternatives evaluated" during the process of experimentation.
• The specific role of each employee in resolving technical uncertainties.

Conclusion: The Enduring Legacy of Norwest Corp.

The Norwest Corp. v. Commissioner case remains the definitive benchmark for the "Permitted Purpose" and "Process of Experimentation" tests in the United States. While the legal standard for "discovery" has been liberalized to focus on uncertainty from the taxpayer's perspective, the administrative requirement for "innovation" in internal use software remains a significant barrier.

The current landscape, shaped by the 2016 IUS regulations, the IRS’s 17 high-risk activity clusters, and the 2025 OBBBA legislation, represents a more mature and certain environment for R&D tax planning. Taxpayers who move beyond the "routine maintenance" of their software systems and engage in structured, technological experimentation can secure significant tax benefits. However, they must do so with the understanding that the IRS now possesses the tools—through Form 6765 and specialized Audit Technique Guides—to rigorously scrutinize the nexus between expenditures and innovation. The legacy of Norwest is not the disallowance of seven projects, but the creation of a blueprint for the one project that succeeded: a project characterized by substantial risk, innovative architecture, and the successful resolution of complex technical uncertainties.