The Internal Revenue Service’s Large Business and International Division (LB&I) has recently turned its focus back to the research tax credit by announcing new campaigns and directives. LB&I’s efforts aim to centralize review and assessment of issues, promote increased compliance, and provide consistency across the nation with respect to examinations. Most taxpayers have experienced challenges during IRS examinations when they claim the research tax credit, and the changes that will soon be implemented may offer insight into what might be in store for future audits as well as help to arm taxpayers with the latest tax controversy techniques. Understanding the outcomes of substantiation-related court cases provides insight into potential exposures that the IRS will try to exploit as a path to disallowing or denying claims. Taxpayers should consider using tools and techniques, such as automated data collection, to support the qualifying nature of their activities.
LB&I Newfound Focus on Research Tax Credits
Research Issues Campaign
In February 2020, LB&I announced a new compliance campaign on research issues, indicating that taxpayers can expect an increase in examinations of their claims. Section 41 (Credit for Increasing Research Activities) and Section 174 (Research and Experimental Expenditures) are the source of some of the most common tax issues in LB&I audits and are often time-consuming for both the taxpayer and the IRS. The campaign’s objectives are to promote voluntary compliance, focus resources on the highest-risk research issues, and increase consistency of examinations.
Directive on the Centralized Risking of Research Issues
Following the campaign announcement, LB&I then issued a directive in March 2020 that describes the requirements and processes for evaluating, on a centralized basis, the risks associated with potential issues related to research credits.1 This directive identifies the research risk review team (RT) as a “national strategy to improve the identification of the highest risk research issues under IRC Sections 41 and 174.” The RT is not new; it has already been helping examiners to assess research credit issues in compliance assurance process cases. However, it appears that this team will be used more often, formally, to help LB&I to identify high-risk research issues and presumably offer updated forms and guidance documents.
Concurrently, in March 2020 LB&I also announced that it is adding 500 new employees. The agency’s workforce has significantly and steadily declined over the past five years, so this hiring announcement may be aimed at rebuilding staffing levels. However, it is more likely tied to the upcoming and increased scrutiny of campaign issues and the directive.
What Taxpayers Should Prepare For
The IRS campaign designation and the new LB&I directive are important reminders regarding the preparation of all research tax credit claims, and audits, including:
- establishing that all four-part tests (and additional three-part tests for internal-use software) for qualifying research have been met;
- establishing a nexus or connection between claimed qualified research expenses (QREs) and qualified research activities (QRAs); and
- having detailed work papers identifying the amounts and types of all QREs and the overall credit calculation.
It is premature to conclude that the examinations will be better for taxpayers. Historically, the use of a more targeted approach by the IRS has resulted in more efficient examinations. More focused and targeted examinations often result in similar savings for taxpayers. However, the addition of a significant workforce, who will need to be trained and exposed to various issues, will take time, and the new team is bound to make mistakes—at the taxpayer’s expense.
Recent Research Credit Cases
Taxpayers should be aware of two recent substantiation-related court cases when preparing their studies and especially during examinations. Exams have been focused on using these cases to deny or disallow research credit claims.
In Siemer Milling Co. v. Commissioner,2 the U.S. Tax Court ruled that a Midwestern wheat flour miller failed to prove that projects carried out at its mills were qualified research. Siemer had claimed relatively modest research credits (less than $125,000 for each of the two years at issue) based on studies prepared by a reputable certified public accounting firm. Although Siemer had no engineers on staff, it employed a research and development team and lab technicians and supervisors. Furthermore, the projects at issue had the “look and feel” of typical qualified research activities in an industrial setting. For example, the “ozone project” involved introducing ozone into different parts of the milling process to produce low-microorganism flour. The “flour heat-treatment project” involved process development to make flour without chlorine and low-bacteria flour without chemicals. At least on the surface, all of the projects had elements of qualified research, such as clearly defined objectives, new production inputs and methods, and lab testing.
Nevertheless, the IRS contended that Siemer’s projects satisfied none of the statutory qualified research criteria, and the Tax Court, while rejecting some of the IRS’ specific arguments, agreed that all of Siemer’s projects failed the “process of experimentation test.” The court stated that this test requires “a methodical plan involving a series of trials to test a hypothesis, analyze the data, refine the hypothesis, and retest the hypothesis so that it constitutes a process of experimentation in the scientific sense.”3 Even if the projects possessed some characteristics of qualified research, the Tax Court reasoned that Siemer had not shown that it conducted a process of experimentation in any project. For example, for the flour heat-treatment project, the court stated, “While the record establishes that Siemer set out to develop three new products . . . it is not clear how it set out to do that and whether the process was a true process of experimentation.”4 The other projects failed for similar reasons.
Ultimately, Siemer Milling hammers home the need for research credit claimants and study providers to accumulate detailed factual support for claims and be ready during audit to present such detail and explain how it satisfies the process of experimentation test and other criteria for QRAs. A reading of the Tax Court’s opinion suggests that Siemer made a greatly scaled-back trial presentation, most likely due to the small credit amounts at stake. For instance, the company called only one witness who actually participated in the projects, a high-level company officer who was likely unfamiliar with project details. However, because taxpayers shoulder the burden of proof with respect to research credits, economizing on the facts is a precarious strategy, as this case clearly illustrates.
One silver lining in Siemer Milling is the Tax Court’s rejection of accuracy-related penalties. The court found that the study provider had sufficient expertise in the research credit and that Siemer had provided necessary and accurate information to the provider and reasonably relied on the study. These facts established that Siemer had acted with reasonable cause and in good faith, thus providing a complete defense against penalties. This should give some measure of comfort that taxpayers who engage and rely on competent advisors will be protected from penalties for research credits not ultimately sustained.
Harper v. United States5 is a tax refund suit in which husband-and-wife taxpayers claimed approximately $400,000 in research credits for each of 2008 and 2010 based on research performed by the husband’s construction company in design-build projects. Before initiating litigation, the Harpers filed timely refund claims on amended returns, with Forms 6765 itemizing the claimed QREs. The IRS audited the claims over a period of years before formally disallowing them, after which the Harpers filed suit in U.S. district court.
Early in the litigation, the Justice Department filed a motion urging the court to dismiss the Harpers’ case for “lack of jurisdiction” on the ground that their administrative refund claims were fatally defective. Citing the Treasury Regulation generally requiring a refund claim to set forth “in detail” the grounds for the claimed refund and facts sufficient “to apprise the Commissioner of the exact basis thereof,”6 the court agreed with the government that the Harpers’ claims were deficient because they did not identify any specific activities performed during the tax years or apply the statutory qualified research activities criteria to such activities. Noting also that the Harpers had based their claims “simply on estimates,” the court dismissed their case.
The Harpers have appealed the dismissal to the Ninth Circuit. It is difficult to reconcile the district court’s ruling that the Harpers provided inadequate notice with the IRS’ extensive audit and disallowance of the claims on their merits. Moreover, in practice, most research credit refund claims provide minimal factual detail apart from identifying the claimed expenses. Nonetheless, if the district court’s ruling is affirmed on appeal, Harper should cause taxpayers and their advisors to carefully consider the level of factual detail to include in refund claims claiming the research credit.
Automation of Data Collection
Given the announced developments from LB&I, and the outcomes of these substantiation-related cases, documentation and record-keeping are now more important than ever before. Taxpayers’ ability to locate and analyze the data underlying a research credit claim can be considerably enhanced using automated collection tools. Algorithms can collect emails, technical reports, test data, and project information, often with more accuracy than when humans perform those tasks. The data can be used as a starting point for the study. For example, it could direct the focus of the study to material areas for investigation by identifying business units with a high relative probability of having qualified research expenses. If done correctly, automated data collection is also less intrusive for employees, particularly the engineers and scientists who are often confused and annoyed by requests for documentation of their research activities from previous years.
Artificial intelligence and machine learning increase the sophistication of this process, expanding capabilities beyond data collection into analysis. It is possible, for instance, for software to analyze the text of thousands of contracts to determine whether the contractual terms constitute “funded research.”
Of course, completeness and quality of the data inputs are just as important as the quality of the automation software. No algorithm, no matter how sophisticated, can make up for inaccurate or incomplete data. Further, the algorithm may be overinclusive or underinclusive. It may not be able to adequately address edge cases, such as executive-level employees who also perform substantive research not usually associated with their job titles. There are additional considerations, such as cybersecurity risks. There may also be data privacy concerns in certain jurisdictions, such as with the General Data Protection Regulation in the European Union or the California Consumer Privacy Act.
As useful as these automated tools can be, they cannot substitute for the involvement of technical subject matter experts and other key employees. After all, a taxpayer can’t put a robot on the stand at trial. You will still need to support your position with human testimony about the nature of research conducted. If anything, it is more important to have employees actively engaged as the champions of the project from execution to potential examination. Those employees should be educated or well-versed in the types of qualifying activities and work done during the claim (and base) years. As the Tax Court’s Suder decision shows,7 the credibility of taxpayer employee testimony can make or break a research credit claim.
Even if the tools for automated data collection and analysis exist, a crucial threshold question remains: Will the IRS accept automated collection methods as sufficient documentation? Exam will need to gain a level of trust in any such method before it will be recognized as adequate substantiation. Human review will still be necessary, at least at first, to ensure the accuracy of the output. Automated systems must take changes in the law into account. There may also be a tension between Exam’s need to verify and proprietary interests in the underlying software.
Recent legislative changes from the Tax Cuts and Jobs Act, as well as the more recent Coronavirus Aid, Relief, and Economic Security (CARES) Act, create additional considerations for taxpayers who claim the research credit. Taxpayers should evaluate how these changes reduce the incentive to invest in research and increase the compliance burden of accounting for research expenditures. Taxpayers should also be aware of the indirect implications to research credits resulting from net operating loss carryback made available by the CARES Act.
The Ever-Closer New Section 174
When the Tax Cuts and Jobs Act prospectively amended Section 174 to replace the current deduction for research or experimental expenditures with mandatory amortization over five years (fifteen years for foreign research) for amounts paid or incurred after December 31, 2021, the effective date of this change seemed far in the future, and there was a general sense that Congress would never allow “new Section 174” to take effect. Over the past two years, proposals to head off this change have failed to gain momentum. Life under new Section 174 beginning in 2022 is seemingly becoming a reality. Absent a legislative solution, this will have stark implications for many taxpayers.
From a tax standpoint, businesses will have less incentive to invest in “core” research and experimentation or will at least be motivated to account for such costs differently. Moreover, under present law taxpayers may claim research credits for “noncore” research activities, such as projects conducted at manufacturing plants, so long as they are eligible for Section 174 treatment, even if they are not accounted for as such. This may very well change under new Section 174. Since the new regime will require taxpayers to amortize their research and experimental expenditures, the IRS could argue that taxpayers cannot claim QREs that were accounted for as Section 174 costs. Since legislative amendment is the only way to avoid such fallout, negatively impacted businesses should consider contacting their representatives on Capitol Hill.
CARES Act Impact on Research Credit
The CARES Act signed into law on March 27, 2020, includes a significant number of tax and other provisions designed to support businesses and individuals affected by the novel coronavirus pandemic. A provision of the CARES Act that indirectly impacts the research credit is the five-year net operating loss (NOL) carryback.
To understand the impact of the NOL carryback on the research credit, we must first consider the ordering rules of NOLs and general business credits. NOLs are used to reduce taxable income prior to considering Internal Revenue Code (IRC) Section 41 research credits (or any IRC Section 38 general business credits) to offset remaining tax liability.8 In certain situations, the NOL carryback may limit the previous use of a research credit. When this happens, tax credits are displaced from the original year they were used. When a research credit is not fully used in the year it was generated, taxpayers must first carry the credit back to the preceding tax year prior to carrying the credit forward for twenty subsequent years, even if the preceding tax year is a closed year.9
This “freeing up” of previously used credits has an interesting impact on research credit controversy. For example, if a taxpayer utilizes the new NOL carryback rules, research credits that were previously associated with a tax year with an expired statute of limitations (SOL) may now be “open” and subject to IRS audits and scrutiny. As a result, taxpayers should revisit their documentation and support for research credits displaced due to NOL carryback. Also, taxpayers should consider revisiting uncertain tax position calculations in light of previously “closed” credits now potentially being “open” and subject to IRS scrutiny.
For many organizations, the research credit represents a substantial tax position and source of value. It is important to understand what is on the horizon with respect to upcoming legislation, case law developments, best practices, and associated IRS campaigns.
Diane Stogiannes is president of the DST Advisory Group. Adam Quattlebaum is a partner at Dixon Hughes Goodman (DHG). Rob Kovacev is a partner at Norton Rose Fulbright. Alex Sadler is a partner at Morgan Lewis.
- IRS Directive, Centralized Risking of Research Issues under IRC §§ 41 & 174, LB&I-04-0320-0060, March 10, 2020, www.irs.gov/businesses/lbi-directive-on-the-centralized-risking-of-research-issues-under-irc-sections-41-and-174.
- Siemer Milling Co. v. Commissioner of Internal Revenue, United States Tax Court Memo 2019-37, April 15, 2019, www.ustaxcourt.gov/ustcinop/opinionviewer.aspx?ID=11930.
- Id. at *11.
- Case No. 18cv2110 DMS (LL) (S.D. Cal. April 25, 2019).
- Treasury Regulations Section 301.6402-2.
- Suder v. Commissioner of Internal Revenue, United States Tax Court Memo 2014-201, October 1, 2014, www.ustaxcourt.gov/UstcInOp/OpinionViewer.aspx?ID=744.
- IRC Section 172, IRC Section 38(c).
- IRC Section 6511(d)(4)(A).