Wicor, Inc. v. United States, 263 F.3d 659 (7th Cir. 2001), illustrates the restrictive interpretation of qualified research applied to software development in several early research-credit cases. The Seventh Circuit affirmed the denial of Wisconsin Gas Company’s research credit because it found no clear error in the trial court’s determination that the project failed the discovery test then applied in that circuit. Later Treasury regulations rejected the requirement to advance knowledge beyond that of skilled professionals in the field. This study distinguishes the historical decision from the standards applicable to later research activities.
Statutory Framework and the Genesis of the Wicor Dispute
The federal research credit originated in 1981, and the 1986 amendments established additional requirements for qualified research. During the years at issue in Wicor, the expenditure requirement referred to Section 174. The following table summarizes the statutory framework relevant to the litigation.
| Requirement | Statutory Basis | Description and Criteria |
|---|---|---|
| Section 174 Test | IRC § 41(d)(1)(A), as then in effect | The research expenditures had to qualify for treatment under the historical Section 174 research and experimental expenditure rules. |
| Technological Information Test | IRC § 41(d)(1)(B)(i) | The research had to be undertaken to discover information technological in nature. The interpretation of discovery was central to Wicor. |
| Business Component Test | IRC § 41(d)(1)(B)(ii) | The information’s application had to be intended to be useful in developing a new or improved business component of the taxpayer. |
| Process of Experimentation Test | IRC § 41(d)(1)(C) and (d)(3) | Substantially all research activities had to constitute elements of a process of experimentation relating to function, performance, reliability, or quality. |
For current applications, the statute must be read as amended. Legislation enacted in 2025 changed Section 41(d)(1)(A) to refer to domestic research or experimental expenditures under Section 174A, generally for amounts paid or incurred in taxable years beginning after December 31, 2024. Treatment of software development costs under Section 174A does not establish that the separate research-credit requirements are satisfied.
Wisconsin Gas was a wholly owned Wicor subsidiary. Its customer information system project began with an internal planning team in 1984, rather than first beginning in the early 1990s. The project covered billing, service orders, meter reading, cash processing, and related customer-service functions. After work with CS&A, Wicor engaged Andersen Consulting to help complete the system. Much of the software came from outside suppliers, while Andersen and Wisconsin Gas jointly developed integration software.
The research-credit dispute concerned tax years 1990 and 1991. The district court denied the credit after trial in Wicor, Inc. v. United States, 116 F. Supp. 2d 1028 (E.D. Wis. 2000). It found failures of discovery, experimentation, innovativeness, and significant economic risk. The Seventh Circuit affirmed on August 14, 2001, relying on the discovery issue alone as sufficient to resolve the credit claim.
The Discovery Test and Innovation in Underlying Principle
Under the Seventh Circuit’s approach at the time, developing something new to a taxpayer did not necessarily satisfy the statutory discovery requirement. Wicor followed United Stationers, Inc. v. United States and distinguished technological discovery from implementation of existing research. The trial court considered testimony about software engineering and integration, but concluded that the project did not meet the historical standard.
The appellate court reviewed the factual determination for clear error. It regarded the work as adapting existing computer technology to the utility’s particular needs. That holding should not be restated as a present-day rule that every qualifying software project must invent a new computer-science principle or produce technology marketable to other companies.
Comparative Analysis of Early Discovery Jurisprudence
| Case Name | Outcome | Key Legal Reasoning |
|---|---|---|
| United Stationers, Inc. v. United States | Credit denied; judgment affirmed in 1998. | The Seventh Circuit applied restrictive discovery and experimentation standards to internal software development, alongside the internal-use software requirements. |
| Norwest Corp. v. Commissioner | Mixed outcome in 1998. | The Tax Court found qualifying research in development of the Strategic Banking System customer module to the extent specified in its opinion; the other seven sample projects failed. It applied a restrictive discovery standard but recognized that existing technology could be used in new ways. |
| Wicor, Inc. v. United States | Denial affirmed in 2001. | The Seventh Circuit upheld the finding that the project failed the historical discovery test. It did not need to decide the remaining disputed credit requirements. |
| Tax and Accounting Software Corp. v. United States | Taxpayer’s favorable summary judgment reversed and case remanded in 2002. | The Tenth Circuit required discovery of technological information and qualifying experimentation, but disagreed with aspects of the Seventh Circuit’s more demanding approach. It did not simply adopt an identical pioneering-research test. |
These decisions demonstrate both restrictive interpretations and differences among courts. Norwest was not a blanket denial of all software research, and the courts did not apply a single uniform test. Later regulations are essential when assessing what these historical opinions mean for subsequent taxable years.
The Andersen Source Code and the Portability Evidence
The contract assigned ownership of the integrated system’s source code to Andersen. The original remained at Wisconsin Gas, and Andersen did not take a copy when the project ended. The Seventh Circuit treated this conduct as evidence that the work was specific customization rather than a technological innovation useful to other customers.
This was an inference from the record, not an independent statutory requirement that research results be portable, patented, or commercially marketed. A taxpayer can develop a business component for use in its own trade or business. Contractual rights and funding arrangements may require separate analysis, but Wicor’s source-code discussion did not decide a funded-research or substantial-rights issue.
Internal Use Software and the High Threshold of Innovation
The CIS was treated as internal-use software in the litigation. Historical internal-use software analysis involved additional requirements concerning innovation, significant economic risk, and commercial availability. The government did not dispute that the CIS was not commercially available. The source document’s assertion that the Seventh Circuit found failure of all three requirements is therefore incorrect.
The table separates the modern regulatory requirements from the historical findings. Under Treasury Regulation § 1.41-4(c)(6), software subject to the high threshold must also satisfy the general qualified-research requirements and other applicable exclusions. Certain software used in qualified research or qualifying production processes, and specified integrated hardware-software products, is excepted from the additional internal-use software rule.
The Three Prongs of the HTI Test
| HTI Test Prong | Requirement Detail | Application in Wicor |
|---|---|---|
| Innovation | Under the modern regulation, anticipated cost reduction, speed improvement, or another measurable improvement must be substantial and economically significant. Novelty alone is not the measure. | The trial court found insufficient innovation under its historical analysis. The Seventh Circuit did not separately decide this issue. |
| Economic Risk | The taxpayer must commit substantial resources with substantial uncertainty, because of technical risk, that those resources will be recovered within a reasonable period. | The trial court found insufficient technical risk, considering Andersen’s experience and testimony distinguishing scheduling concerns from technological uncertainty. |
| Commercial Availability | The software cannot be purchased, leased, or licensed and used for its intended purpose without modifications that themselves satisfy the innovation and significant-economic-risk requirements. | The government conceded that the CIS software was not commercially available; this requirement was not a basis for denial. |
The modern economic-risk test does not require uncertainty about whether the end result could ever be achieved. Technological uncertainty affecting whether resources can be recovered within a reasonable period can matter. Ordinary budget or scheduling concerns alone do not establish this requirement, but timing should not be dismissed when the uncertainty arises from technical risk.
Process of Experimentation and the Elimination of Uncertainty
The Wicor trial court separately found that the project did not constitute a qualifying process of experimentation under the interpretation it applied. It was unpersuaded that the evidence demonstrated research-oriented evaluation of competing hypotheses. The Seventh Circuit did not independently resolve that ground because failure of discovery was sufficient.
Eustace v. Commissioner, 312 F.3d 905 (7th Cir. 2002), subsequently applied the circuit’s historical discovery and experimentation precedents to software developed for sale. Its appellate decision should not be characterized as a holding about denial of expense estimates under the Cohan rule. Its central issues were the applicable definitions of discovery and experimentation.
Under Treasury Regulation § 1.41-4(a)(5), the evaluative process may involve modeling, simulation, or systematic trial and error. It must address uncertainty about capability, method, or appropriate design and fundamentally rely on physical or biological science, engineering, or computer science. Known technical principles can be used. Routine implementation or debugging does not establish qualification simply because it involves testing; equally, systematic trial and error is not categorically disqualified.
Useful evidence connects the initial uncertainty, alternatives considered, evaluations performed, and resulting technical decisions. Contemporaneous records can support that account, but the regulation does not impose a universal requirement for a particular hypothesis document created at project inception. Records must substantiate both eligibility and the expenditures claimed.
Regulatory Pivot: TD 9104 and the Revised Discovery Standard
Treasury Decision 9104, published on January 2, 2004, finalized regulations generally applicable to taxable years ending on or after December 31, 2003. The regulations define discovery in terms of eliminating uncertainty concerning development or improvement of a business component. They expressly reject a requirement to exceed, expand, or refine the common knowledge of skilled professionals in the relevant field.
The statutory discovery requirement remains; its regulatory interpretation changed. Merely learning something new to the taxpayer is not enough without the required technological uncertainty and the other elements of qualified research. The change did not erase the separate experimentation requirement or retroactively establish that Wicor’s claimed costs qualified.
Shift in Definition of Discovery
| Era | Source | Standard of Discovery |
|---|---|---|
| Historical Wicor litigation | Wicor and the Seventh Circuit’s United Stationers precedent | Required more than taxpayer-specific newness and treated innovation in underlying principle as central to discovery. |
| General framework under TD 9104 | Treasury Regulation § 1.41-4(a) | Research intended to eliminate uncertainty about capability, method, or appropriate design; no requirement to advance the field’s common knowledge. All other credit requirements remain applicable. |
A comparable integration project might warrant a different analysis under later regulations, but its outcome would depend on the actual activities, software classification, relevant tax year, and supporting evidence.
Evolution of Internal Use Software Regulations
Regulations proposed in 2015 and finalized in 2016 clarified the treatment of internal-use software. The final rules generally apply to taxable years beginning on or after October 4, 2016, with specified transition rules for earlier years. They define internal use by reference to general and administrative functions, including financial management, human resources management, and support services.
Software developed to enable interaction with third parties, or to let third parties initiate functions or review data on the taxpayer’s system, can fall outside that definition. The intended functions at the beginning of development matter. Software with both internal and third-party functions is subject to additional classification rules, including identification of a subset used only for third-party interaction.
A safe harbor can permit inclusion of 25 percent of otherwise qualified research expenditures for remaining dual-function software or a dual-function subset when anticipated third-party interaction accounts for at least 10 percent of its use and the other conditions are satisfied. This is not an automatic credit for customer-facing software.
The Wicor record describes utility information processing, meter-reading terminals, and integration work. It does not establish that Wisconsin Gas was building modern mobile apps or customer web portals. Applying the later third-party rules to such hypothetical features is therefore illustrative only; it cannot support a conclusion that part of Wicor’s actual claim would necessarily qualify.
The Section 1341 Claim: A Parallel Legal Battle
Wicor also pursued an unrelated claim under Section 1341, concerning income previously received under an apparent unrestricted right. The utility’s regulated rates had reflected anticipated tax costs that were subsequently reduced by a change in tax law. Regulators required lower future rates to pass the benefit to customers.
The Seventh Circuit rejected relief because the future rate reduction reduced income rather than creating an allowable deduction. Section 1341 requires an otherwise allowable deduction; it does not independently create one. Wisconsin Gas had not allocated refunds to the particular customers who made the earlier payments.
The court expressly declined to decide whether the company’s initial right to the income was merely apparent. Its decision should not be described as resolving a separate “same circumstances” test. The key distinction was between lower future prices and a deductible repayment.
Substantiation and the Shrinking-Back Rule
Treasury Regulation § 1.41-4(b)(2) applies the qualified-research requirements first to the business component. If they are not satisfied at that level, the analysis proceeds to the most significant subset of its elements and continues to smaller subsets until a qualifying subset is reached or the most basic element fails. Shrinking back does not override other statutory exclusions or prove that any particular expense qualifies.
| Level of Analysis | Definition | Application Post-Wicor |
|---|---|---|
| Business Component | A product, process, software, technique, formula, or invention held for sale, lease, or license, or used in the taxpayer’s business. | Apply the qualification requirements separately to each component. |
| Sub-Component | A subset of the component’s elements, such as a particular software module. | Evaluate its activities and evidence when the component as a whole does not satisfy the requirements. |
| Shrinking-Back | Successive application of the requirements to appropriate subsets. | May identify qualifying research within a broader implementation, but requires substantiation and does not guarantee eligibility. |
The regulation’s ERP integration example provides a concrete illustration: ordinary configuration does not qualify, while systematic testing to resolve uncertain design of specialized data-caching and synchronization software can satisfy the research requirements for a subset. This is more precise than claiming that Wicor lost solely because it presented an entire system, or that shrinking back necessarily would have saved its claim.
Final Thoughts
Wicor is useful for understanding historical software research-credit litigation and the need to distinguish a court’s actual holding from other trial-level findings. Its pioneering-discovery approach should not be treated as the current regulatory standard. Complexity, spending, and business benefits do not themselves prove qualified research, but an industry-wide technological breakthrough is not required under the later general regulations.
Strategic Recommendations for Taxpayers
- Identify the applicable tax year and governing law before relying on historical cases.
- Explain technological uncertainty in capability, method, or design and connect it to the alternatives evaluated.
- Retain technical and cost records sufficient to substantiate the claimed activities and expenditures.
- Analyze applicable exclusions, including adaptation of an existing business component to a customer’s requirements, without assuming that all custom development is excluded.
- Classify software functions and apply the internal-use and dual-function rules where relevant.
- Review contractual rights, payment obligations, and funding arrangements separately; source-code ownership alone does not resolve credit eligibility.
- Apply the shrinking-back rule where appropriate and substantiate the activities and costs of any qualifying subset.
A sound software research-credit analysis connects the governing requirements to the project’s actual technical work. Wicor’s history and the later regulatory changes make careful attention to dates, facts, and the scope of each holding essential.
