Yellow Freight System, Inc. v. United States illustrates the importance of distinguishing a court’s actual ruling from broader developments in research-credit law. This study separates the company’s investment-credit litigation from its software research-credit dispute and explains the limits of applying historical decisions to present-day claims.
Foundations of the Controversy: Investment Credit and Research Credit
The investment-credit dispute concerned terminal facilities and taxes for 1963–1966. The district court ruled for Yellow Freight in 413 F. Supp. 357 (W.D. Mo. 1975), but the Eighth Circuit reversed as to docks and inspection lanes in 538 F.2d 790 (8th Cir. 1976). It treated those structures as buildings, considering both appearance and function. This was not a research-credit decision.
| Structure Type | Function Identified by the Court | Eligibility Rationale |
|---|---|---|
| Docks/Additions | Freight transfer and employee workspace. | The appellate court held these were buildings excluded from the investment credit. |
| Inspection Lanes | Vehicle inspection and maintenance workspace. | The appellate court likewise treated these facilities as buildings. |
| Fences | Terminal protection. | The district court allowed the credit; the government did not challenge that aspect on appeal. |
The appellate court separately remanded the treatment of heated cargo-protection rooms for further findings. The investment-credit classification rules do not establish software eligibility under Section 41.
The 1991 Research-Credit Decision: A Dispute Requiring Trial
In Yellow Freight System, Inc. v. United States, 24 Cl. Ct. 804 (1991), the United States Claims Court considered software expenditures for 1983 and 1984 under former Section 44F. It examined nine representative systems. The parties disputed whether the work constituted research and produced new or significantly improved software. The court denied both motions for partial summary judgment because material facts required trial. That disposition was not a final denial of the credits.
Technical Complexity and the Limits of the SYSNET Narrative
The cited opinion does not establish the source’s claimed SYSNET rejection, 20% productivity finding, or origin of a field-wide discovery test. It discusses projects including freight selection, dispatch timing, electronic billing, and customer records. Competing expert assertions must not be presented as resolved judicial findings.
| Project Attribute | Yellow Freight’s Position | Court’s Determination |
|---|---|---|
| Technological Nature | Asserted novel or improved software. | Material factual disputes remained. |
| Uncertainty | Asserted uncertainty for certain projects. | No blanket finding rejecting all technical uncertainty. |
| Experimentation | Claimed qualifying software development. | The record required factual resolution at trial. |
| Outcome | Sought partial summary judgment. | Both sides’ motions were denied. |
The High Threshold of Innovation and Internal Use Software
Modern internal-use software rules must be analyzed separately from the 1991 ruling. Treasury Decision 9786, issued in 2016, addresses software developed primarily for general and administrative functions. Internal deployment alone does not decide the classification; third-party interaction and dual-function software have specific rules.
Where the high-threshold-of-innovation test applies, software must be innovative, development must involve significant economic risk, and a commercially available solution must not be usable for the intended purpose without modifications satisfying the other requirements. Innovation can involve a substantial, economically significant cost reduction, speed improvement, or other measurable improvement. Economic risk requires substantial resource commitment and substantial uncertainty, because of technical risk, about recovery within a reasonable period.
The ordinary research requirements and exclusions also apply. Payroll or bookkeeping software is not automatically incapable of qualifying, and cloud hosting does not itself create an exception. The source’s Norwest project-level outcome and claimed reliance on a Yellow Freight doctrine are not established here and should not guide eligibility decisions.
The 2003–2004 Regulatory Changes and Later Litigation
Treasury Decision 9104, published in January 2004, rejected a requirement to exceed, expand, or refine the knowledge of skilled professionals. The regulations focus on uncertainty about capability, method, or appropriate design and the use of technological principles. Applying known science can therefore satisfy the technological-information requirement, but unfamiliarity to the taxpayer alone is insufficient.
United States v. McFerrin, 570 F.3d 672 (5th Cir. 2009), came later. The Fifth Circuit vacated and remanded a judgment that had applied an incorrect research standard. It did not precede or cause the 2004 regulations.
| Criterion | Earlier Restrictive Discovery Interpretation | T.D. 9104 and Modern Qualification |
|---|---|---|
| Discovery Requirement | Some interpretations required advancement beyond professional common knowledge. | Such advancement is not required. |
| Definition of Uncertainty | The source incorrectly attributes a universal scientific-breakthrough requirement to Yellow Freight. | Capability, method, or appropriate design may be uncertain. |
| Experimentation Method | Historical rules and authorities must be matched to the tax year. | Systematic evaluation can include modeling, simulation, and trial and error. |
| Internal Use Software | A separate statutory and regulatory issue. | Additional software rules remain applicable; later clarification includes the 2016 regulations. |
These changes do not establish that a historical SYSNET claim would qualify today. That conclusion would require the project’s actual activities, costs, dates, and software classification.
The Four-Part Test in Modern Application
Under current Section 41, qualified research generally must satisfy four requirements: expenditures treated as domestic research or experimental expenditures under Section 174A; technological information; intended usefulness in developing a new or improved business component; and substantially all activities constituting a process of experimentation for improved function, performance, reliability, or quality. Earlier years require their applicable statutory rules.
The requirements apply separately to each business component. A transportation company’s development project is not eligible merely because it uses algorithms or improves efficiency.
The Shrink-Back Rule and the Substantially All Requirement
Treas. Reg. § 1.41-4 provides the 80% substantially-all standard, measured on a cost or other consistently applied reasonable basis. If a business component fails the requirements, the shrink-back rule applies the tests to its most significant subset of elements, continuing until a qualifying subset or the most basic element is reached.
Little Sandy Coal Co., Inc. v. Commissioner, 62 F.4th 287 (7th Cir. 2023), affirmed a 2021 Tax Court decision because the taxpayer failed to substantiate qualifying experimentation adequately. Novel vessels and unsupported wage percentages were insufficient. The decision should not be reduced to a rule that routine fabrication necessarily dilutes a whole-project fraction: the proper numerator, denominator, and role of the activities matter.
| Level of Analysis | Qualified Research Expenditures (QRE) Calculation | Strategic Implication |
|---|---|---|
| Total Project | Identify qualifying costs; do not include all construction or development costs automatically. | A project label does not prove qualification. |
| Sub-Component | Identify costs attributable to a qualifying subset under the shrink-back rule. | A narrower claim still requires evidence and satisfaction of the tests. |
There is no established basis for concluding that shrink-back would have changed Yellow Freight’s outcome.
Documentation and Direct Supervision
Treas. Reg. § 1.41-2 recognizes conducting qualified research, directly supervising it, and directly supporting it as qualified services. Direct supervision generally means immediate, first-line supervision; it excludes higher-level management of a first-line supervisor even when the manager is a scientist. This regulatory rule was not newly created by the source’s asserted Moore ruling.
Executive titles neither qualify nor disqualify wages automatically. An executive’s actual research, qualifying supervision, or direct support must be established, with an appropriate allocation of wages. Receiving progress summaries or exercising general management is insufficient.
Records should substantiate activities and expenditures in usable detail. Contemporaneous evidence is valuable, but neither a universal daily timesheet requirement nor automatic rejection of every retrospective study follows from these rules.
Technical Evolution in Logistics: Autonomous and Connected Systems
Potential areas for investigation include dispatch algorithms, vehicle-control systems, intelligent transportation infrastructure, and electric-truck technologies. These are candidate subjects for factual review, not automatic credit categories. Teams should distinguish technical development from purchasing equipment, routine configuration, business optimization, and ordinary implementation. Record the uncertainty, alternatives, evaluation, and resulting design decisions for the activities actually claimed.
Current Implications: Section 174A, Form 6765, and ASC 730
The source’s statement that domestic research must always be amortized over five years is outdated. The TCJA capitalization regime applied beginning in 2022; Public Law 119-21 added Section 174A for domestic expenditures in tax years beginning after December 31, 2024. It permits current deductions, with an alternative capitalization election. Foreign research remains subject to 15-year amortization under Section 174. Revenue Procedure 2025-28 addresses implementation and transition options, including recovery of certain previously capitalized domestic costs. Deduction treatment and research-credit eligibility remain distinct.
The December 2025 Form 6765 instructions make Section G optional for tax years beginning before 2026 and required afterward subject to stated exceptions and completion rules. The form includes business-component and wage-category information. Amended refund claims have separate information requirements; the source’s blanket 2024–2025 requirement is inaccurate.
The IRS ASC 730 directive provides an examination approach for eligible large taxpayers using adjusted financial-statement research costs. It requires the prescribed adjustments and supporting information; financial-statement classification alone does not make every cost a qualified expense.
| Documentation Requirement | Practitioner Responsibility | Risk of Failure |
|---|---|---|
| Contemporaneous Records | Retain technical notes, testing evidence, and reliable cost allocations. | Insufficient support may prevent substantiation. |
| Business Component Detail | Identify components and apply the appropriate tax-year filing rules. | Unsupported aggregation can obscure qualification. |
| Wage Support | Document actual research, direct supervision, or direct support. | General management allocations may be disallowed. |
Corporate Restructuring and Preservation of Tax Records
Yellow Corporation and affiliated debtors entered Chapter 11 proceedings in August 2023. Bankruptcy does not itself establish the existence or value of an R&D refund claim. The source’s predictions about final returns and One Yellow research credits lack support. A practical implication is to preserve technical and accounting records through restructuring; an absent pre-bankruptcy study does not automatically extinguish otherwise valid tax rights.
Mathematical Analysis of the Research Credit
For the regular research-expense component, the credit is generally 20% of qualified research expenses exceeding the base amount, subject to the statutory base rules. It is not a flat 20% of all R&D spending.
Regular research-expense component = 20% × max(0, current-year QREs − base amount).
The alternative simplified credit generally equals 14% of current QREs above half the average QREs for the preceding three years. If any of those years has no QREs, the statutory alternative generally uses 6% of current-year QREs.
ASC = 14% × max(0, current-year QREs − 50% × preceding-three-year average QREs).
These simplified expressions omit other credit components and tax adjustments. Method selection requires actual data; the source does not support its claim that ASC would have been preferable for Yellow Freight’s earlier growth years.
Final Thoughts
Yellow Freight is useful chiefly as a reminder to read the applicable opinion and tax-year law precisely. Its 1991 summary-judgment ruling cannot support the source’s narrative of a definitive SYSNET defeat or a Yellow Freight discovery doctrine. Modern claims require project-specific evidence and application of the current statutory and regulatory requirements.
Recommendations for Practitioners
Preserve development evidence and cost records; identify business components before applying shrink-back; assess software purpose and applicable exceptions; and support wage allocations with actual services. Avoid promising eligibility from project labels, hosting arrangements, executive titles, or technological complexity alone.
