Reinforcing the Section 41 “Process of Experimentation”: What Recent Court Cases Teach the Technical Tax Community and Why This Matters Now
Recent decisions—Little Sandy Coal (7th Cir., Mar 7, 2023), Betz (Tax Court, July 6, 2023), and Phoenix Design Group (Tax Court, Dec 23, 2024) — have crystallized how IRS and courts now evaluate Section 41 claims. The focus is clear: employee time claimed must directly map to rigorous, scientific experimentation — mere design or novelty won’t suffice. For highly technical taxpayers, these cases demand a strategic re-tooling of how R&D credit studies are prepared and defended.
Section 41 Refresher – The Four-Part Test & “Substantially All” Requirement
Qualified research under Section 41 requires:
- Permitted purpose (function, performance, reliability/quality)
- Technological uncertainty
- Process of Experimentation (PoE)
- Technological nature
Plus, under § 41(d)(1)(C), at least 80% of research activities must be “elements of a process of experimentation” — measured by time or cost — known as the “substantially all” threshold. How courts interpret that 80% fraction is pivotal.
Case Walkthroughs & Key Dialogues
Little Sandy Coal Co. v. Commissioner (7th Cir., March 2023)
- IRS auditors/Tax Court treated direct support and supervision as automatically excluded from the PoE numerator.
- 7th Circuit ruled: support/supervision can be included — if actually engaged in experimentation tasks.
- The court emphasized that “shortcut estimates” and departmental allocations are inadequate — entity failed to provide a principled breakdown of employee time by experimentation activity.
- Shrink-back strategy must be applied at subcomponent level; treating an entire vessel as one business component prevented benefit of subcomponent-level qualifying activities.
Dialogue Highlight: The court asked, implicitly, “Which engineers did what, on which subcomponent, to resolve which uncertainty?” General claims weren’t enough.
Betz v. Commissioner (Tax Court, July 2023)
- Taxpayer submitted opaque testimony about hydrogen/oxidizer units — but failed to show technological uncertainty or a systematic process of experimentation.
- The pilot-model exception was rejected — the constructed units were end-products, not prototypes used to test design alternatives.
- Funded research exclusion (Section 41(d)(4)(H)) applied in several contracts where the taxpayer surrendered IP rights — those costs were disqualified.
Dialogue Highlight: The court essentially said, “Show us your hypothesis, your testing plan, and your results — just explaining the end-design is insufficient.”
Phoenix Design Group, Inc. v. Commissioner (Tax Court, December 2024)
- PDG’s standard MEPF engineering phases (AIA schematic, basis of design, shop drawings) were treated as routine design — not experimentation. No true hypothesis-testing or iteration documented.
- Sampling of 238 projects reduced to a few example projects — but courts refused to limit discovery, requiring documentation for all claimed projects.
Dialogue Highlight: Courts demanded concrete evidence, phrased as: “How did engineers test uncertainties? How did they document variations and outcomes?”
Dissecting “Process of Experimentation” vs “Elements of PoE”
- Process of Experimentation = the structured methodology: uncertainty → alternatives → experimentation → evaluation → redesign.
- Elements of a PoE = discrete tasks that support that process: prototyping, simulation iterations, testing logs, analysis, hypothesis revision, supervisory review.
Little Sandy clarifies that support/supervision can count if those roles are directly involved in experimentation. This opens opportunities, but only with proper documentation.
Implications for Practitioners – Best Practices
Documentation is foundational
- Break out activities and employee time at project + sub-component level
- Track alternatives tested, iteration cycles, hypotheses tested, evaluation results
- Avoid based on vague testimony or estimates
Time tracking must be granular
- Support and supervision roles must be documented showing involvement in testing, reviewing experimental results, etc., to be included in the PoE numerator
Component/sub-component structure and shrink-back
- Structure claims at the sub-component level to allow shrink-back where only parts qualify
Funded research clauses matter
- Review agreements to ensure IP rights retained; avoid disqualification under §41(d)(4)(H)
Don’t rely on sampling alone
- Courts reject limiting discovery to sample projects unless all underlying records and methodology are robust and represent the full population.
Process of Experimentation Flow
Case Comparison:
| Case | Key PoE Issue | Support/Supervision | Shrink-Back Application | Documentation Failure |
|---|---|---|---|---|
| Little Sandy Coal | Arbitrary allocations; unsupported numerator | Allowed if tied to PoE | No – entire vessel claimed, blocking shrink-back | Lack of granular time/activity data |
| Betz | No experimentation; pilot-model misuse | Not material | Not addressed | Vague testimony, no hypothesis/test logs |
| Phoenix Design | Standard engineering ≠ PoE | Limited – routine oversight | Projects treated generically, no subcomponent analysis | Broad sampling, no project-level narratives |
DST’s Approach: Technical Credibility Meets Compliance
At DST, our method aligns directly with emerging judicial expectations:
- We capture uncertainty and alternatives testing through SME interviews and engineering documentation
- We map individual tasks to PoE steps and tie time codes accordingly
- We review client contracts to avoid funded research pitfalls
- Our deliverables include audit-ready narratives, timelines, and flowcharts that reflect a scientific methodology
By partnering with your engineers, DST ensures claims reflect real technical work — not retrospective spin.
What Technical Tax Leaders Should Do Now
Courts are asking sharper questions: “How do you know your team was experimenting, not just building?” The answers must be rooted in technical process, structured testing, and concrete documentation.
These rulings provide a playbook:
- Use elements-of-PoE language to claim support/supervision time when appropriately documented
- Avoid generic or high-level categorization of research activities
- Build modular, iterative proof trails at the sub-component level
DST stands ready to elevate your audit defensibility and technical credibility.