United States v. McFerrin

The Historical and Legislative Landscape of Section 41

The federal tax credit for increasing research activities, established under Section 41 of the Internal Revenue Code, represents a critical intersection of economic policy, scientific innovation, and administrative law. Originally introduced as a temporary incentive in 1981, the credit has undergone numerous extensions and modifications, ultimately becoming a permanent fixture of the United States tax code through the Protecting Americans from Tax Hikes (PATH) Act of 2015. The overarching objective of the credit is to stimulate domestic investment in research and development by reducing the after-tax cost of innovation, thereby fostering long-term economic growth and competitiveness. However, the administration of the credit has historically been a source of significant friction between taxpayers and the Internal Revenue Service (IRS). This friction arises primarily from the complexity of the statutory requirements and the often-exacting documentation standards imposed by the Service during examinations.

At the heart of this complexity is the "four-part test" mandated by Section 41(d), which defines what constitutes "qualified research." To qualify for the credit, an activity must meet four distinct criteria: it must involve expenses deductible under Section 174, be undertaken to discover information that is technological in nature, intend for that information to be useful in developing a new or improved business component, and involve a process of experimentation. The interpretation of these requirements, particularly the "discovery" of information and the nature of "experimentation," has been the subject of intense litigation for decades. The case of United States v. McFerrin stands as a watershed moment in this history, challenging the government’s restrictive interpretations and re-establishing the role of reasonable estimation in tax substantiation.

The Factual Genesis of the McFerrin Dispute

The litigation concerning Arthur R. McFerrin and his affiliated corporations began with a series of amended tax returns for the 1999 tax year. Arthur McFerrin was a prominent chemical engineer who founded KMCO, Inc., a Subchapter-S corporation that manufactured specialty and commodity chemicals for the petrochemical industry. McFerrin’s contributions to the field were so significant that Texas A&M University renamed its chemical engineering department in his honor in 2005. In addition to KMCO, McFerrin owned several related entities, including KMCO Port Arthur, Inc., and KMTEX, Inc., all of which operated as pass-through entities whose income and credits flowed directly to his personal tax returns.

In 2000, McFerrin and his corporations filed their original 1999 tax returns without claiming any research tax credits. However, in May 2003, KMCO engaged alliantgroup, L.P., a specialized tax consulting firm, to conduct a comprehensive study evaluating whether the company’s chemical manufacturing activities met the criteria for the research credit. Based on the results of this study, McFerrin and his corporations filed amended 1999 returns in September 2003, claiming an aggregate credit of $472,092. Due to a clerical error, the IRS issued a refund check for the full amount plus interest, totaling $601,228.40, within less than a month of the filing.

The government eventually realized the error and initiated a lawsuit in October 2005 under Section 7405 to recover the refund. The United States’ complaint alleged that the amended returns included no supporting documents and that the taxpayer had failed to substantiate the claimed credits with sufficient evidence. This initiated a legal battle that would eventually reach the United States Court of Appeals for the Fifth Circuit and reshape the standards for research credit substantiation across the country.

Procedural Nuances and Burdens of Proof

A critical element of the McFerrin litigation was the procedural posture of the case. Because the government was suing to recover a refund it had already paid, it bore the ultimate burden of proof to show that the refund was erroneous. This differs from a standard tax deficiency case where the taxpayer bears the burden of proving they are entitled to a deduction or credit. The Ninth Circuit in Soltermann v. United States and the Seventh Circuit in United States v. Commercial Nat'l Bank of Peoria established that in erroneous refund suits, the government must prove not only that an amount was erroneously refunded but also the exact amount of that error.

McFerrin also challenged the government's pleading standards, arguing that because the IRS claimed he "misrepresented facts," it was required to meet the heightened particularity requirements of Federal Rule of Civil Procedure 9(b), which applies to allegations of fraud or mistake. The Fifth Circuit rejected this argument, noting that the government's complaint only alleged a failure to document, which falls under the more liberal notice standards of Rule 8(a)(2). This ruling clarified that the IRS does not need to allege fraud to seek the recovery of a refund; a simple lack of substantiation is a sufficient legal basis for the claim.

The District Court and the Resurgence of the Discovery Test

The initial bench trial lasted six days and culminated in a victory for the United States. The district court ordered McFerrin to repay the refund with interest, concluding that the taxpayer had failed to properly support the credits. The court's decision was rooted in a highly restrictive interpretation of the four-part test for qualified research. Specifically, the district court relied on the "Discovery Test," which posits that research is only qualified if it expands or refines existing principles in the field, has a high threshold of innovation, and possesses a broad effect on the industry or economy.

Furthermore, the district court held that a "process of experimentation" required the formal forming and testing of hypotheses. It explicitly rejected "trial and error" testing as a qualifying activity, arguing that such systematic evaluation did not meet the rigorous scientific standards intended by the statute. Applying these definitions, the court found that while some of the KMCO chemical projects may have involved research, they did not constitute "qualified research". Additionally, the court refused to accept the rough estimates provided by KMCO employees years after the research occurred, effectively demanding contemporaneous, project-specific records.

The Fifth Circuit Appeal: Correcting Legal Misinterpretations

The Fifth Circuit's reversal of the district court’s decision was predicated on the finding that the lower court had applied the wrong legal standards. This reversal touched upon three foundational aspects of Section 41: the meaning of "discovering information," the scope of the "process of experimentation," and the evidentiary standards for substantiation.

The Shift from Discovery to Uncertainty

The Fifth Circuit clarified that the "Discovery Test" relied upon by the district court—the idea that research must result in information that is new to the industry—was no longer the law. While the IRS had initially introduced a discovery requirement in proposed regulations in 2001 (T.D. 8930), these were superseded by final regulations in 2004 (T.D. 9104). These newer regulations established that research is undertaken for the purpose of discovering information if it is intended to eliminate "technical uncertainty" concerning the development or improvement of a business component.

Crucially, the 2004 regulations clarified that the information discovered does not need to be new to the world or the industry; it only needs to be "new to the taxpayer". The Fifth Circuit noted that even the Treasury Department had acknowledged that the older discovery standard did not reflect Congressional intent. This shift significantly broadened the eligibility for the credit, particularly for manufacturers and engineers whose work involves solving technical problems that may have been addressed elsewhere but remain uncertain within the context of their specific production environments.

Redefining the Process of Experimentation

The appellate court also addressed the definition of the "process of experimentation." It ruled that the district court's requirement for a formal hypothesis-testing framework was overly rigid. The Fifth Circuit held that a process of experimentation can involve systematic trial and error or other approaches aimed at evaluating alternatives to achieve a desired result. This recognition is vital in industrial settings like McFerrin’s chemical manufacturing plants, where the development of a more efficient chemical reaction or a more stable specialty compound often requires iterative testing of variables rather than the rigid scientific method used in academic laboratories.

The Cohan Rule and the Mandate of Reasonable Estimation

The most enduring legacy of the McFerrin decision is its instruction to apply the "Cohan Rule" to research credit substantiation. The Cohan Rule, derived from the 1930 case Cohan v. Commissioner, establishes that if a taxpayer can prove they incurred a deductible expense but cannot establish the exact amount due to a lack of contemporaneous records, the court must make "as close an approximation as it can".

"Allowing Something Rather Than Nothing"

The Fifth Circuit explicitly criticized the district court for refusing to consider employee estimates simply because they were provided years after the fact. The appellate court stated that if the taxpayer demonstrates that "qualified research" activities actually occurred, the court is obligated to estimate the related expenses rather than granting no credit at all. While the court should "bear heavily" against the taxpayer whose records are inexact, it is legally inconsistent to allow nothing when the evidence confirms that research took place.

This mandate for estimation directly contradicts the IRS's long-standing preference for perfect, project-based contemporaneous records. It acknowledges that many innovative companies—especially small to mid-sized firms—may not have the administrative infrastructure to track every employee hour to a specific business component in real-time. By allowing for "fair estimates" based on testimony and "institutional knowledge," McFerrin created a more equitable environment for domestic innovators.

The Role of Employee Testimony as Evidence

The Fifth Circuit emphasized that the "institutional knowledge of employees" is a valid form of evidence in determining a fair estimate of research time and expenses. This was further supported by the Tax Court’s decision in Union Carbide Corp. v. Commissioner, which allowed employee testimony to substantiate qualified wages even without detailed time logs. These cases collectively establish that while documents are preferable, oral evidence from the individuals who performed the research is both credible and legally sufficient to support a claim.

The Bonus Dispute: McFerrin's $6.4 Million Performance Incentive

A specific factual dispute in McFerrin involved a $6.4 million bonus paid to McFerrin in 2004, which he sought to include as a qualified research expense (QRE) on the grounds that it compensated him for his technical work. The district court had disqualified the bonus, finding that it was based on KMCO’s overall profits and cash flow rather than on specific research performed during the year.

The Fifth Circuit did not make a final determination on the eligibility of the bonus but sent the issue back to the lower court for further examination. The appellate court noted that if any portion of the bonus was reasonable compensation for research services performed, it could potentially count as a QRE. This suggests that executive compensation in innovation-heavy firms can be included in the credit calculation, provided the taxpayer can document a nexus between the payment and specific research activities.

IRS Reaction: Strategic Entrenchment and Tier I Classification

The IRS reaction to the McFerrin decision was characterized by a push for higher administrative standards to counteract the judicial flexibility granted to taxpayers. Shortly after the ruling, the IRS’s Large and Mid-Size Business (LMSB) Division designated research tax credit refunds as a "Tier I" examination issue. This designation signaled that the IRS would treat these claims with the highest level of scrutiny, often requiring centralized audit management and a proactive approach toward asserting penalties.

The Conflict over Documentation Standards

IRS examiners have routinely applied "exacting documentation standards" that are not explicitly found in the Section 41 regulations. The Service argues that without project-level records that tie specific expenses to defined business components, it cannot properly audit a claim. Taxpayers, however, argue that these standards are an attempt to circumvent the McFerrin and Cohan precedents, which allow for reasonable substantiation through broader evidence.

The IRS Audit Techniques Guide (ATG) remains critical of the use of retrospective studies prepared by outside consultants, which it labels "prepackaged submissions". The ATG emphasizes that estimation is only a "fallback" and should not be used if the taxpayer's failure to maintain records was an "inexactitude of their own making". This creates a high hurdle for taxpayers who engage consultants years after the research has concluded.

The Nexus Requirement and Project-Based Accounting

In the wake of McFerrin, the concept of "nexus" has become the primary battleground in R&D audits. The IRS uses the term to describe the evidentiary link between individual qualified research expenses (QREs) and specific business components. While McFerrin allows for the estimation of the amount of expense, the taxpayer still bears the burden of establishing that the activity itself qualified under the four-part test.

Evaluating Methods of Allocation

Taxpayers traditionally use one of three methods to capture QREs: project-based, cost-center, or hybrid. The IRS strongly prefers the project-based approach, which requires contemporaneous accounting records tracking every hour of every engineer to a specific business component. However, the McFerrin decision provided institutional support for the "hybrid" approach, where costs are initially captured by department (cost center) and then allocated to projects through employee testimony and subject matter expert interviews.

This hybrid approach effectively uses employee interviews to approximate the "qualified percentage" of an employee’s time. As the McFerrin court noted, if it is clear that an engineer spent a significant portion of their time solving technical problems, it is unreasonable for the government to deny the credit entirely simply because they did not use a stopwatch to record their minutes.

State-Level Conformity and the Impact on Local Research Credits

The legal principles established in United States v. McFerrin have direct implications for state-level research credits, as many states conform their credit calculations to the federal standards set forth in Section 41.

Indiana: A Focus on Contemporaneous Verification

The Indiana Department of Revenue (IDOR) frequently cites McFerrin in its Letters of Findings, particularly regarding the "legislative grace" doctrine. Indiana requires that taxpayers maintain and produce "contemporaneous records sufficient to verify" the credit. In cases where Indiana manufacturers have relied solely on retrospective studies without supporting technical documents, the IDOR has used the narrow construction principle of McFerrin to deny the state REC.

However, Indiana also follows the federal shift from the Discovery Test to the Uncertainty Test. Taxpayers in Indiana have successfully challenged audits that attempted to apply the older "innovation" threshold by pointing to the Fifth Circuit’s clarification that research only needs to resolve uncertainty for the taxpayer's own business.

Arizona: Applying Cohan in State Disputes

The Arizona Department of Revenue (ADOR) has also integrated the McFerrin interpretation of the Cohan Rule. Arizona law allows for research credits as determined under Section 41, and ADOR acknowledges that estimates can be used to establish a credit once it has been proved that qualified research activities occurred. Arizona’s judicial bodies have noted that while the absence of records may lead the court to "bear heavily" on the taxpayer, a complete denial of the credit is inappropriate if a reasonable basis for estimation exists.

Comparative Jurisprudence: Little Sandy Coal and the 80% Threshold

The flexibility established in McFerrin has been refined and, in some jurisdictions, restricted by subsequent decisions. The most significant modern counter-balance is Little Sandy Coal Co., Inc. v. Commissioner, decided by the Seventh Circuit in 2023.

The Limits of the Cohan Rule

In Little Sandy Coal, a shipbuilding company claimed research credits for the design and construction of "first-in-class" vessels. Similar to the taxpayer in McFerrin, Little Sandy Coal relied on employee interviews and "arbitrary estimates" to allocate time to experimentation. However, the Seventh Circuit affirmed the disallowance of the credit, finding that the taxpayer had failed to meet the "substantially all" threshold.

The "substantially all" test, codified in Treas. Reg. § 1.41-4(a)(6), requires that at least 80% of the activities for each business component constitute a "process of experimentation". The Seventh Circuit held that the taxpayer did not offer a "principled way" to determine what portion of activities met this threshold. Unlike the Fifth Circuit in McFerrin, the Seventh Circuit was unwilling to use the Cohan Rule to estimate a credit when the taxpayer could not prove that the primary nature of the activity was experimental.

Implications for Pilot Models and Production Wages

Little Sandy Coal also provided critical guidance on "pilot models"—representations of products built to resolve technical uncertainty. The court ruled that pilot model production expenses must be included in the denominator of the "substantially all" calculation. If the models are built to resolve uncertainty regarding the product's method or design, they may be included in the numerator. This case underscores that even under a more flexible substantiation regime, the taxpayer must still satisfy the rigorous statutory percentages required by Section 41.

Modern Administrative Evolution: Form 6765 and FAA 20214101F

The legacy of McFerrin continues to shape the IRS’s regulatory output, as the agency seeks to standardize the substantiation requirements that were at the heart of the litigation. The most significant recent development is the release of Field Attorney Advice (FAA) 20214101F, which established strict "specificity requirements" for all research credit refund claims postmarked after January 10, 2022.

The Five Items of Substantiation

Under the FAA 20214101F standards, a taxpayer filing an amended return for a research credit must provide five specific items of information:

• Identify all business components to which the Section 41 research credit claim relates for the year.
• For each business component, identify all research activities performed.
• Identify all individuals who performed each research activity.
• Identify all the information each individual sought to discover.
• Provide the total qualified employee wages, supply expenses, and contract research expenses.

This requirement effectively institutionalizes the project-based accounting that the IRS has long sought. While taxpayers initially have a 45-day "grace period" to perfect deficient claims, the failure to provide this data upfront can result in the summary rejection of the refund claim. This is a direct administrative counter-measure to the McFerrin decision, aimed at forcing taxpayers to do the work of documentation before filing, rather than relying on retrospective studies and the Cohan Rule during an audit.

Revisions to Form 6765 (2024-2025)

The IRS has also significantly revised Form 6765, Credit for Increasing Research Activities, to include "Section G," which mirrors the specificity requirements of FAA 20214101F. Following advocacy from the AICPA, the IRS adopted several simplification measures in the final 2024/2025 version of the form, including:

• The 80/50 Rule: Taxpayers are only required to report business components that account for at least 80% of total QREs, and this is capped at a maximum of 50 business components.
• Topic 730 Alignment: Taxpayers who expense R&D costs on their financial statements under FASB ASC Topic 730 can align their tax reporting with their financial statement amounts, simplifying the reconciliation process for large corporations.
• Simplified Terminology: Replacing complex descriptive requirements with a more straightforward "business component name" field.

These changes represent a modernization of the substantiation process, moving away from the "all-or-nothing" litigation of the McFerrin era toward a more structured, risk-based reporting system.

Mathematical Context: Incremental Incentives and the Base Amount

The financial benefit of the research credit is calculated as an incremental incentive, meaning it only rewards spending above a certain historical threshold or "base amount." The traditional credit calculation is as follows:

The "Base Amount" is generally calculated as the product of the taxpayer’s "fixed-base percentage" and the average annual gross receipts for the four preceding years. However, the base amount cannot be less than 50% of the current year’s QREs.

A major design concern cited by the GAO and other policy experts is that many taxpayers calculate their base amount using research spending data dating back to the 1984–1988 period. For companies like McFerrin’s, which may have changed ownership or production methods significantly since the 1980s, documenting this historical spending is an immense burden. The McFerrin instruction to use reasonable estimates is particularly vital here; if a taxpayer cannot reasonably estimate their base period spending, they may be unable to calculate a valid credit, even if their current research is perfectly documented.

Strategic Implications for Future R&D Tax Credit Claims

The synthesis of McFerrin, Union Carbide, and Little Sandy Coal provides a clear roadmap for tax professionals and businesses seeking to claim the research credit. The following strategies represent the modern best practices for substantiation in the current regulatory environment.

Leveraging the Hierarchy of Evidence

While oral testimony is legally sufficient under McFerrin, the hierarchy of evidence in an R&D audit clearly favors a multi-layered approach. Taxpayers should aim to provide:

• Contemporaneous Technical Documentation: Lab notebooks, schematics, test results, and emails that prove a process of experimentation occurred.
• Narrative Reports: Summaries for each business component that explicitly articulate the technical uncertainty being resolved.
• Corroborating Testimony: Structured interviews with subject matter experts that align with the provided documents.
• Principled Estimations: Allocation methodologies that are transparent, consistent, and grounded in the institutional knowledge of the researchers.

Navigating the "Shrinking-Back" Rule

If an entire project (business component) fails to meet the four-part test—perhaps because a large portion of the work was routine manufacturing—taxpayers should apply the "shrinking-back" rule under Treas. Reg. § 1.41-4(b)(2). This allows the taxpayer to apply the tests to a discrete sub-component or process within the larger project. McFerrin’s flexible approach to estimation is critical here, as it allows for the allocation of wages to just those specific experimental phases of a larger production run.

Management of External Consultants

The IRS remains highly critical of "study-driven" claims that lack direct taxpayer involvement. To avoid the pitfalls seen in the early stages of McFerrin, businesses should ensure that any outside consultants work closely with internal technical teams to capture contemporaneous data rather than relying solely on retrospective interviews. The value of the consultant lies in their ability to translate the company's technical work into the language of Section 41, not in manufacturing a credit where no technical uncertainty existed.

Conclusion: The Enduring Impact of United States v. McFerrin

The case of United States v. McFerrin remains one of the most important judicial checks on the IRS's administration of the research tax credit. By correcting the lower court’s reliance on the outdated "Discovery Test," the Fifth Circuit ensured that the credit remains accessible to the broad spectrum of American manufacturers and engineers who engage in incremental, practical innovation. Furthermore, by mandating the application of the Cohan Rule, the court established that tax administration must reflect the reality of business operations, where perfect recordkeeping is often a secondary concern to the work of discovery itself.

However, the subsequent decade of IRS administrative updates, including FAA 20214101F and the expansion of Form 6765, demonstrates that the government has not abandoned its desire for rigid documentation. The modern taxpayer must navigate a landscape where judicial flexibility (via McFerrin) exists alongside heightened administrative specificity requirements. The primary lesson for future R&D tax credit applications is that while "fair estimates" can save a legitimate claim from disallowance, the most defensible positions are those that provide a clear, principled, and documented "nexus" between innovation and expense. As domestic research and development remains a cornerstone of U.S. economic strategy, the legal framework established by McFerrin will continue to serve as the foundational standard for balancing the goals of tax compliance and technological progress.