Wicor, Inc. v. United States

The litigation in Wicor, Inc. v. United States remains a fundamental case study for tax professionals, legal scholars, and corporate researchers seeking to navigate the complexities of the Internal Revenue Code (IRC) Section 41. Decided during a period of intense regulatory flux in the late 1990s and early 2000s, the case crystallized the "Discovery Test" and the "High Threshold of Innovation" standard for internal use software (IUS), providing a definitive interpretation of what constituted "qualified research" at the turn of the millennium. By examining the Seventh Circuit's reasoning alongside contemporary and subsequent regulations, such as Treasury Decision (TD) 9104, one can trace the trajectory of the R&D credit from a restrictive, "pioneering" standard toward a more inclusive, "uncertainty-based" framework.

Statutory Framework and the Genesis of the Wicor Dispute

To comprehend the significance of Wicor, one must first understand the statutory environment of the mid-1980s. Before 1981, research and development costs were generally deductible under Section 174, but the introduction of the Section 41 credit was intended to provide an incremental incentive for companies to increase their investment in high-technology research. The 1986 amendments to the Code sought to narrow the scope of the credit, as Congress became concerned that taxpayers were claiming the incentive for routine product development rather than true technological advancement. This concern led to the implementation of the four-part test for qualified research, which serves as the foundational hurdle for any Section 41 claim.

Wicor, Inc., the parent company of Wisconsin Gas, entered this fray in the early 1990s when it attempted to modernize its utility operations through the development of a centralized Customer Information System (CIS). The project aimed to integrate several core business functions—meter reading, billing, service orders, and accounts receivable—into a unified digital architecture. Wisconsin Gas initially attempted to utilize third-party vendor software but found that commercial solutions, such as those offered by CS&A, were unable to accommodate the specific regulatory and operational requirements of the utility at a reasonable cost. Consequently, Wicor engaged Andersen Consulting to assist in an internal, custom development effort.

The resulting conflict with the Internal Revenue Service (IRS) arose when the company claimed a research tax credit for the millions of dollars expended on this integration effort. The IRS disallowed the claim, asserting that the project was a standard business implementation of existing software technology rather than a discovery of new computer science principles. The case proceeded to the District Court for the Eastern District of Wisconsin and subsequently to the Seventh Circuit, where the central question was whether Wicor had "discovered" information that was technological in nature.

The Discovery Test and the Innovation in Underlying Principle

The "Discovery Test" became the focal point of the Wicor litigation. At the time, the IRS and several courts interpreted the phrase "discovering information" to mean that the taxpayer had to find information that was new to the field of science or engineering, rather than merely new to the taxpayer. This standard, often referred to as the "knowledge of the field" test, was supported by the 1986 legislative history, which suggested that qualifying research must pass a high threshold of innovation and have a broad effect on the industry.

The Seventh Circuit, citing United Stationers, Inc. v. United States, held that "discovery demands something more than mere superficial newness; it connotes innovation in underlying principle". For software specifically, the court relied on the House and Senate Conference Reports which clarified that research does not rely on computer science principles merely because a computer is used; rather, the research must expand or refine the existing principles of computer science.

In applying this to Wicor, the court scrutinized the expert testimony of Donald O'Neill, who argued that the CIS project applied software engineering principles to complex integration problems. The court was ultimately unpersuaded, finding that the project was "merely adapting existing computer technology to the special needs of the gas company". This distinction between "adaptation" and "discovery" remains one of the most significant implications of the Wicor case. The court reasoned that while the project involved significant complexity and business risk, the engineers were essentially using known methodologies to achieve a predictable, albeit difficult, result.

Comparative Analysis of Early Discovery Jurisprudence

To understand the weight of Wicor, it is helpful to compare its outcomes with contemporaneous cases that shaped the "Discovery Test" landscape.

The consensus among these cases was a restrictive "pioneering" view of Section 41. The courts were wary of allowing the credit for activities that appeared to be routine software engineering, even when those activities involved significant trial-and-error. The Wicor court specifically noted that fine-tuning or debugging computer programs does not satisfy the "Process of Experimentation" test if the underlying design is certain from the outset.

The Andersen Source Code and the Portability Evidence

One of the most insightful aspects of the Seventh Circuit's decision in Wicor was its reliance on the behavior of the parties involved in the development, specifically Andersen Consulting. The contract between Wisconsin Gas and Andersen stipulated that the consultant would own the rights to the source code for the integrated CIS. However, the court found it "telling" that Andersen did not take a copy of the source code when the project was finished, nor did they attempt to market the specific integration program to other utilities.

The court inferred from this that if the project had truly involved "inventing a new technology" or expanding computer science principles, the resulting source code would have been a valuable, portable asset for a global consulting firm. Because the software was so localized to the specific business rules and legacy systems of the Wisconsin Gas Company, it lacked the "broad effect" and "innovation in underlying principle" required for the credit. This highlights a critical lesson for future R&D credit applications: the intellectual property (IP) treatment and the commercial potential of the research results often serve as proxies for technological discovery in the eyes of the court.

Internal Use Software and the High Threshold of Innovation

Because the CIS was developed for the utility's internal administrative and bookkeeping functions, it was classified as Internal Use Software (IUS). Under Section 41(d)(4)(E), IUS is generally excluded from the research credit unless the taxpayer can satisfy an additional three-part "High Threshold of Innovation" (HTI) test. The Wicor decision reinforced the rigorous nature of this test, which was designed to prevent the credit from subsidizing ordinary back-office improvements.

The Seventh Circuit found that Wicor failed all three prongs of the HTI test. The most damaging finding was that the project lacked "significant economic risk" because there was no substantial uncertainty at the outset about whether the project was technically feasible. The court distinguished between "technical risk"—the possibility that the goal cannot be achieved—and "business risk"—the possibility that the goal will cost too much or take too long. For the R&D credit, only technical risk counts.

Process of Experimentation and the Elimination of Uncertainty

The Wicor decision also heavily criticized the taxpayer's failure to document a true "Process of Experimentation." Under Section 41(d)(1)(C), substantially all of the research activities must constitute elements of a process designed to evaluate one or more alternatives to achieve a result where the appropriate design is uncertain. The court noted that Wicor tested various hypotheses and algorithms, but it characterized these activities as "industrious development" and "fine-tuning" rather than an evaluative process aimed at resolving fundamental technological unknowns.

This finding was echoed in Eustace v. Commissioner, where the court declined to apply the "Cohan rule" (which allows for the estimation of expenses when some research is proven) because the taxpayer's own expert witness admitted that the changes being made were not "major" and did not constitute pioneering research. The implication for future applicants is clear: the process of experimentation must be more than just iterative software testing; it must be a structured investigation into a technical uncertainty that is documented from the project's inception.

Regulatory Pivot: TD 9104 and the Death of the Discovery Test

The restrictive "Discovery Test" used in Wicor was eventually deemed too high a bar by the Treasury and the public. In 2004, the Treasury Department issued TD 9104, which fundamentally changed the R&D credit landscape. These final regulations clarified that the "Discovery Test" was not an independent, high-level hurdle separate from the Section 174 test. Instead, research is undertaken to "discover information" if it is intended to eliminate uncertainty concerning the development or improvement of a business component.

Crucially, TD 9104 stated that discovery does not require the taxpayer to obtain information that exceeds the common knowledge of skilled professionals in the field. Information only needs to be new to the taxpayer. This was a direct rebuke to the logic used in Wicor and United Stationers. Under the post-2004 rules, Wicor might have had a stronger case if it could have documented that it was uncertain about the method or appropriate design of the integration, even if integration itself was technically possible.

This regulatory shift illustrates that while the Wicor court correctly applied the law as it was interpreted at the time, the Treasury ultimately chose a more taxpayer-friendly path to ensure the credit remained an effective incentive for a broader range of industrial and software development activities.

Evolution of Internal Use Software Regulations (2015–2016)

The implications of Wicor for software development continued to evolve with the release of new IUS regulations in 2015 and 2016. These regulations significantly narrowed the definition of internal use software, exempting many modern software applications from the high threshold of innovation test.

Under the new rules, software is considered IUS only if it is used for "general and administrative functions," which are limited to financial management, human resources, and support services like facilities management. Software that enables a taxpayer to interact with third parties—such as the customer-facing billing portals and mobile apps that Wisconsin Gas was trying to build—is now generally excluded from the IUS definition.

If Wicor were litigated under today's "Dual Function Software" rules, the portions of the CIS that allowed customers to review their data or initiate service would likely be subject only to the standard four-part test, rather than the HTI test. This regulatory easing suggests that the "all-or-nothing" denial seen in Wicor is less likely in the modern era, provided the taxpayer can accurately allocate costs between internal and third-party functions.

The Section 1341 Claim: A Parallel Legal Battle

A less-discussed but equally important aspect of the Wicor case was the company's separate claim under IRC Section 1341. This section provides relief to taxpayers who include an item in their gross income under a "claim of right" but are later forced to return it because it is determined they did not have an unrestricted right to the money.

Wisconsin Gas had collected rates from customers that were later deemed to be excessive due to a change in corporate tax rates. The public utility commission did not order a direct refund but instead required the company to reduce its future rates to return the windfall to the customers. Wicor argued that this rate reduction was functionally equivalent to a refund and should trigger a deduction under Section 1341, allowing them to recompute their taxes for the year the income was originally received.

The Seventh Circuit rejected this claim, holding that a reduction in future rates is not a "deduction" but a decrease in future income. The court emphasized the "same circumstances" test, which requires that the lack of an unrestricted right must arise out of the terms of the original payment. This ruling has long-standing implications for regulated utilities and any business facing government-mandated price adjustments, reinforcing the strict boundaries of Section 1341 relief.

Contemporary Audit Trends and the "Shrinking-Back" Rule

In the wake of Wicor and the subsequent regulatory changes, the IRS Large Business and International (LB&I) division has developed sophisticated audit techniques for R&D claims. Modern audits frequently focus on the "Shrinking-Back" rule found in Reg. § 1.41-4(b)(2). This rule requires that the four-part test be applied first at the level of the discrete business component (e.g., the whole CIS). If it fails at that level, the taxpayer must "shrink back" the claim to the sub-components that do qualify.

The Wicor case serves as a prime example of a failed "all-or-nothing" approach. Because the taxpayer presented the entire CIS as a single project, the court's finding that the overall goal was "technically feasible" led to the disallowance of the entire claim. Modern litigation, such as Trinity Industries, Inc. v. United States (dealing with ship prototypes), reinforces that taxpayers must be prepared to isolate and document the specific sub-systems—such as a novel propulsion system or a new integration protocol—where true uncertainty and experimentation occurred.

Implications for Future R&D Credit Applications

The legacy of Wicor, Inc. v. United States is a dual-edged sword for modern taxpayers. On one hand, the specific "Discovery Test" it championed has been officially retired by TD 9104, making it easier for evolutionary—rather than revolutionary—research to qualify. On the other hand, the court's skepticism toward "adaptation" and "routine engineering" remains a potent weapon in the IRS's audit arsenal.

Strategic Recommendations for Taxpayers

• Identify Uncertainty Early: Future applicants must move beyond "business risk" and document "technical uncertainty." This uncertainty can involve the "capability" to achieve the result, the "method" to achieve it, or the "appropriate design".
• Document the Hypothesis-Testing Cycle: The process of experimentation must be evidenced by more than just time-sheets. Contemporaneous records must show the identification of a problem, the formulation of alternatives, and the iterative testing of those alternatives.
• Address the Adaptation Exclusion: Taxpayers should explicitly document how their work goes beyond "adapting an existing business component to a particular customer's requirement". This is especially critical for consultants and software developers who build bespoke solutions.
• Leverage Dual-Function Software Rules: For software projects, taxpayers should categorize features based on whether they are internal-facing or third-party facing to take advantage of the safe harbors provided in the 2016 regulations.
• Secure Substantial Rights: Following the Andersen source code example, taxpayers must ensure their contracts grant them substantial rights to the research results, even if they do not have exclusive ownership.

In the final analysis, Wicor stands as a reminder that the R&D credit is an activity-based incentive, not a spending-based one. The Seventh Circuit’s refusal to credit Wicor’s millions in software costs, despite the clear difficulty of the project, underscores the necessity of aligning R&D claims with the rigorous, scientific-method-based requirements of the Code. While the legal standards for "discovery" have softened, the requirements for "experimentation" and "technical uncertainty" remain as stringent as ever, requiring a nuanced understanding of both tax law and the underlying technological processes.