The Procedural Foundation: Pena and the Finality of Tax Litigation

Pena v. United States, 883 F. Supp. 154 (S.D. Tex. 1994), concerned the Penas’ 1971 income tax liability, not an R&D credit claim. The opinion describes an earlier Tax Court proceeding, Pena v. Commissioner, No. 2302-76, dismissed for lack of prosecution in 1983 after a petition filed in 1976.

The Penas later sought a refund exceeding $73,994.22. The district court dismissed their action because it lacked jurisdiction and alternatively held that claim preclusion barred relitigation. Section 6512(a), subject to statutory exceptions, generally restricts subsequent refund litigation for a year placed before the Tax Court. The opinion also addressed the full-payment requirement for income-tax refund jurisdiction.

The general lesson is to prosecute a tax case and preserve appeal rights diligently. Pena did not decide Section 41 eligibility, reject research-hour estimates, or establish an R&D documentation standard. No connection between these taxpayers and a technology entrepreneur is established by this decision.

The Statutory Architecture of the Research and Development Tax Credit

Research must satisfy the four-part test separately for each business component, subject to the shrinking-back rule. A business component includes a product, process, software, technique, formula, or invention held for sale, lease, or license, or used in the taxpayer’s trade or business. Commercial sale is therefore not essential.

The Four-Part Test Framework

The ordinary incremental portion of the regular credit equals 20% of current-year qualified research expenses exceeding the statutory base amount. The base amount generally cannot be less than 50% of current-year qualified research expenses. The alternative simplified credit generally uses 14% of expenses exceeding half the preceding three-year average; a special 6% rule applies where there were no qualified expenses in one or more of those years.

Qualified expenses can include qualifying wages, supplies, computer-use costs, and eligible contract research payments, generally at 65% for ordinary contract research. Separate rules, exclusions, elections, and Section 280C coordination affect the result; neither formula represents a credit on every dollar labeled R&D.

Decoding Technological Uncertainty: The Research Expenditure Requirement

Uncertainty concerns what the information available at the outset establishes about capability, method, or appropriate design. The research need not discover something new to an entire industry. Existing scientific principles can support qualifying experimentation; their availability does not itself disqualify the work.

In Phoenix Design Group, Inc. v. Commissioner, T.C. Memo. 2024-113, the Tax Court rejected research credits claimed by an engineering consulting firm. The case illustrates why broad descriptions of design challenges and revisions are insufficient when they do not establish qualifying research in the activities examined.

Routine calculations and compliance with established design requirements do not automatically prove technical uncertainty. Conversely, it would overstate the decision to say that engineering using established formulas can never qualify. The analysis must address the actual uncertainty and work undertaken for the component.

The Process of Experimentation: Moving Beyond Routine Design

A development workflow must be evaluated on its substance. Iteration alone does not establish experimentation, because revisions can arise from customer preferences, coordination, or routine corrections.

The Systematic Evaluation of Alternatives

Useful evidence explains the following:

• The technical uncertainty identified at the beginning.
• The alternatives considered to resolve it.
• The modeling, simulation, systematic trial and error, or other evaluation performed.
• The results and how they informed subsequent development decisions.

A formally labeled hypothesis document is not a universal statutory requirement. Under Treasury Regulation Section 1.41-4, the substantially-all threshold is 80% or more, measured on a cost basis or another consistently applied reasonable basis. It is not simply a count of projects, drawings, or design changes.

The remaining activities must satisfy the applicable research-expenditure requirement for the regulatory rule to apply. A successful product, extensive novelty, or a large development budget does not substitute for showing which activities were experimental.

The Limits of Traditional Design Phases

Phoenix Design Group illustrates why standard design stages are not automatic proof of qualifying experimentation. The workflow discussed included:

• Pre-design.
• Schematic design.
• Design development.
• Construction documents.
• Bidding and negotiation.
• Construction administration.

These labels describe project organization. The eligibility analysis still needs evidence of the uncertainty, alternatives, and evaluation within the activities claimed. This is a practical lesson about proof, rather than a new rule requiring every engineering firm to adopt a particular recordkeeping system.

Economic Risk and the Funded Research Exclusion

Section 41(d)(4)(H) excludes research to the extent funded by another person. Customer payments do not automatically make every contractor’s research ineligible. The contractual allocation of financial risk and substantial rights must be examined under the funding regulations.

• Economic risk: Determine whether payment depends on successful research, considering the actual agreement and applicable law.
• Substantial rights: Determine whether the researcher retains meaningful rights in the research results.

The Importance of Contractual Language

In Meyer, Borgman & Johnson, Inc. v. Commissioner, the Eighth Circuit affirmed the Tax Court in 2024. The contracts did not expressly or by clear implication make payment contingent on successful research. Professional standards, general acceptance provisions, and fixed prices did not establish the required contingency in those agreements.

The January 3, 2025 order in System Technologies, Inc. v. Commissioner, No. 12211-21, denied the IRS’s motion for partial summary judgment on funding. The court considered Indiana law requiring repayment if the contractor failed to deliver a functioning product. The order addressed the funding issue presented; it was not a determination that every expense met all research-credit requirements.

Together, these decisions support examining enforceable obligations, acceptance conditions, remedies, and payment rights. Merely adding research terminology to a contract does not establish eligibility.

The Retention of Substantial Rights

Tangel v. Commissioner, T.C. Memo. 2021-1, illustrates the consequences of contractual restrictions that leave the researcher without substantial rights. Ownership and restrictions on reuse must be considered together. Rights need not be exclusive, and transferring patent ownership does not by itself resolve every substantial-rights question.

The funding regulations also contemplate circumstances in which neither party qualifies: for example, the researcher may bear the risk of failure but retain no substantial rights, while the customer has no obligation to pay for unsuccessful work. That possibility should not be presented as a finding that Tangel adjudicated the customer’s entitlement.

Substantiation and IRS Enforcement

A taxpayer must support both qualifying activities and the amounts claimed. Section 6662 accuracy-related penalties may apply when their requirements are met, but credit disallowance does not automatically establish a penalty. Relevant defenses, including reasonable cause and good faith under Section 6664, depend on the facts.

Estimates and the Limits of the Cohan Rule

The Cohan principle does not eliminate the need to prove entitlement or provide a reasonable basis for estimating expenses. Betz v. Commissioner, T.C. Memo. 2023-84, illustrates the weakness of allocations that lack adequate evidentiary support. It does not establish a universal ban on estimates in research-credit cases.

Contemporaneous evidence is valuable, but Treasury Regulation Section 1.41-4(d) requires sufficiently usable and detailed records rather than one prescribed time-tracking product. Depending on their content and reliability, project records, technical documents, testimony, and supported allocations may help establish the claim.

• Maintain payroll and cost records connected to the work claimed.
• Preserve test results, design versions, technical correspondence, and development records.
• Explain how allocation methods distinguish qualifying research from routine development and other work.
• Document why changes were made and which technical uncertainties they addressed.

The Redesigned Form 6765 and the Harper Precedent

The IRS’s December 2025 instructions make Section G business-component information optional for tax years beginning before 2026 and required for years beginning after 2025, subject to exceptions. Separate requirements apply to amended research-credit refund claims. Taxpayers should use the instructions applicable to the return being filed.

Harper v. United States, decided by the Ninth Circuit in 2021, reversed a dismissal after finding that the IRS had waived the refund claim’s regulatory specificity requirement through its substantive examination. It therefore does not support the proposition that audit-stage clarification can never matter. Nor does the decision establish that it caused the later Form 6765 redesign. Claimants should still provide the required information initially instead of assuming waiver will occur.

The Future of R&D Incentives: Policy Changes and Competitiveness

The credit and the deduction for research expenditure serve different functions. Deduction timing affects taxable income and cash flow; the credit has its own eligibility and calculation rules. A change to one does not automatically change entitlement to the other.

The TCJA and the 2025 Restoration of Domestic Expensing

The Tax Cuts and Jobs Act required capitalization and amortization for research expenditures paid or incurred in tax years beginning after 2021: five years for domestic research and fifteen years for foreign research, using a midpoint convention.

Public Law 119-21, enacted July 4, 2025, added Section 174A, generally allowing immediate deductions for domestic research or experimental expenditures in tax years beginning after December 31, 2024. Taxpayers may instead elect qualifying capitalization and amortization. Foreign research remains subject to fifteen-year amortization under Section 174. Thus, restoration of domestic expensing is enacted law, not merely a pending legislative proposal.

Transition provisions address previously capitalized domestic costs, including options for recovering remaining balances and time-limited retroactive relief for eligible small businesses. Eligibility, election procedures, and deadlines require separate review. Federal treatment also does not automatically determine state treatment.

Practical Implications for Future R&D Tax Credit Applications

A defensible study links the legal tests to the taxpayer’s actual technical work, expenditures, and agreements. The procedural lessons of Pena should remain distinct from the substantive eligibility lessons of research-credit decisions.

Activity-Level Tracking and Technical Documentation

For each component, maintain a clear account of:

• The technical objective and intended improvement.
• What was known at the outset and what remained uncertain.
• The alternatives evaluated and the methods used.
• The results, including unsuccessful approaches where relevant.
• The personnel, costs, and allocation basis associated with qualifying activities.

Existing operational records may supply much of this evidence. A later narrative should accurately explain those records rather than invent experiments or retrospectively label routine work as research.

Contract Review and Amendment Strategies

• Review the actual allocation of payment risk, acceptance obligations, and remedies for failure.
• Check intellectual-property ownership, licenses, confidentiality terms, and restrictions on using results.
• Evaluate the governing law with counsel where it affects enforceable rights or remedies.
• Make prospective amendments that reflect genuine commercial arrangements; do not assume wording changes establish historical eligibility.

Retaining a meaningful right to reuse results can be relevant, but no single license phrase guarantees substantial rights. Choosing a governing law also does not independently determine credit eligibility.

Applying the Shrinking-Back Rule

If an entire business component fails the test, examine the most significant subset of its elements, applying the rule successively until a qualifying subset is identified or no smaller subset remains. It is not permission to select arbitrary activities solely to reach a desired percentage.

For a mixed engineering project, a distinct technical subsystem might warrant separate evaluation. Its novelty alone is insufficient: the activities must satisfy the applicabl