The taxpayer in Little Sandy Coal Co. v. Commissioner, T.C. Memo 2021-15 (Feb. 11, 2021) has appealed an unfavorable United States Tax Court decision to the United States Court of Appeals for the Seventh Circuit. The decision likely misinterprets the statutory and regulatory requirements a taxpayer must satisfy to claim federal income tax credits for qualified research expenses. Miller Canfield's March 4, 2021 alert summarized the Tax Court decision.  

The case presents the issue of whether, and the extent to which, a taxpayer's activities constitute elements of a process of experimentation. The Commissioner's Tax Court briefs run hundreds of pages but do not include an argument that the definition of qualified research expenses informs the definition of activities that constitute elements of a process of experimentation. Nonetheless, the Tax Court itself developed that rationale to decide that the taxpayer failed the process-of-experimentation test. The court further concluded that its decision was dispositive of the case, and that other issues in the case need not be decided. The court's disposition of the case is problematic, which this alert discusses. 

Research is qualified and tax-credit eligible if (i) research activities are intended to discover information that would eliminate uncertainty in the development of a product, (ii) the purpose of the research is to discover information that is technological, (iii) application of the information is intended to be useful in development of the product, and (iv) substantially all of the research activities constitute elements of a process of experimentation for the purpose of developing a product's function, performance, reliability, or quality. (IRC §41(d))

The process-of-experimentation test: The elements of a process of experimentation test are (i) identification of technological uncertainty concerning development of a product, (ii) identification of a design alternative intended to eliminate the technological uncertainty, (iii) identification of an evaluative process intended to eliminate the technological uncertainty, (iv) conduct of that process, (v) evaluation of the results, and (vi) design modifications to address design failures, if any. (Treas. Reg §1.41-4(a)(5))

Technological uncertainty must relate to development of a qualified purpose, for example, a new function for the product. Substantially all the activities must constitute elements of a process of experimentation relating to development of that new function. Substantiality is satisfied if at least 80% of the research activities constitute elements of the experimental process. The research activities may be measured by cost of other consistently applied reasonable basis. For example, to determine if at least 80% of the research activities constitute elements of the experimental process, the cost of the research activities constituting elements of the experimental process to develop the new function is divided by the cost of all research activities conducted to develop the product even if one of more of the activities do not constitute an element of the experimental process to develop the product function. If the resulting quotient is 80% or more, the process-of-experimentation test is passed.   

We pause for a moment to ask why activities that do not constitute elements of the experimental process for a qualified purpose are even considered. The reason is that if substantially all the activities are elements of an experimental process for a qualified purpose, like developing function, then other activities, for example, product styling, which is not a qualified purpose, are included in the definition of qualified research. Product styling then becomes a qualified research activity, and the amount of the credit increases for the expenses incurred for styling. Styling activity never disqualifies research performed for developing function. The substantially-all test is intended to help taxpayers and not sink their credit claims. If there is too much styling activity, it should be eliminated from consideration and the credit allowed just for product function activities.

Direct Support of Research: The court's analysis in Little Sandy erred in determining that substantially all the taxpayer's activities did not constitute elements of an experimental process conducted for a qualified purpose. The court mistakenly read the definition of qualified research expenses into the definition of activities that constitute elements of an experimental process. The definition of qualified research expenses (IRC §41(b)) does not inform the measure of activities that are elements of an experimental process. (Treas. Reg. §1.41-4(a)(6))

At issue in Little Sandy is whether activities of production workers who fabricated nautical vessels – a tanker barge and a drydock – were activities that constitute elements of the experimental process. If they are, their cost is included in the numerator of the fraction computing substantiality, and if not, they are excluded. The taxpayer argued that fabrication activity is an element of the experimental process. The court rejected the argument.

The court's analysis began with a reference to the statutory provision (IRC §41(b)(2)(B)(ii)) that treats wages to an employee who "direct[ly] supports" qualified research as qualified research expenses. The next analytical step was that an employee who fabricates an experimental model only "directly supports" qualified research and is not, itself, "qualified research," which is correct. The final analytical step was that an activity that only directly supports qualified research and is not itself qualified research cannot be an activity constituting an element of an experimental process, which is the linchpin of the decision and is incorrect. 

The court confuses activities that come within the definition of qualified research with the definition of expenses that are allocated to that qualified research. Qualified research is determined first. (IRC §41(d)(1); Treas. Reg. §1-41-4(a)(5)(i)). Expenses allocable to that qualified research are determined thereafter. ((IRC §41(b)(2)(B)(ii))

Two reasons show why the Tax Court's analysis is incorrect. First, the research credit has two different substantially-all tests. The measure of wages of an employee who engages in direct support of qualified research is treated as 100% of the employee's annual wages if at least 80% the employee's wages are allocable to qualified research activities. This is the "substantially-all" test for wage expense performed by an employee engaging in direct support of qualified research. (The same "gross-up" rule applies to wages of an employee who directly performs qualified research.) The substantially-all test for research activities constituting elements of an experimental process, which is a different test, provides that wages of an employee engaging in direct support do not equal 100% of the employee's wage expense for the entire year even if at least 80% of the annual wages are allocable to activities constituting elements of an experimental process. Only the specific wages allocable to the activities of the employee performing direct support are counted. For this substantially-all test, there is no gross-up. To put a fine point on the court's error, it decided that wages for direct support are never included as a measure of a research activity constituting an element of an experimental process even though express regulatory direction is given (Treas. Reg. §1.41-4(a)(6)) to exclude only a gross-up amount, but not the actual amount, for such wages. The court's decision violates the regulation.

Second, "modeling" is expressly included as an activity that constitutes an element of an experimental process. (Treas. Reg. §1-41-4(a)(5)(i)) The vessels arguably were "models" used to evaluate design alternatives. A modeler who fabricates a model to be used experimentally is, literally, "modeling." The subsequent vessel sales should not disqualify the modelers' activity as an activity constituting an element of a process of an experimental process because the subsequent sales do not appear to be fundamentally inconsistent with the premise on which the credit was based. (Hillsboro National Bank v. Commissioner, 460 U.S. 370 (1983))   

Supplies used for modeling: The court makes a similar mistake in rejecting the cost of supplies used to fabricate an experimental model. The court reasoned that a supply is not an "activity" because only people perform activities. Consequently, according to the court, the cost of supplies for an experimental model cannot be considered in measuring the substantiality of activities constituting elements of an experimental process. The court's conclusion conflicts with regulatory guidance (Treas. Reg. §1.41-4(a)(5)(i)) because "modeling" is expressly defined as the "conduct" of an evaluative experimental process. 

Moreover, the Tax Court decision has the perverse effect of causing a taxpayer to fail the process-of-experimentation test if the nature of the taxpayer's business requires considerable expenditures to fabricate experimental models. The cost of modeling supplies is deductible (IRC §174(a)) and therefore includible in the denominator of the fraction determining whether substantially all activities are elements of an experimental process. The court's reasoning excludes these supply costs from the numerator. If the expenses are considerable, the taxpayer will fail the substantially-all test because the percentage will be less than 80% and will not be entitled to research credits. Those who have only modest supply expenses will be in a better position to claim the credits because they will pass the substantially-all test. This is an indefensible difference resulting from excluding supplies for experimental models from the process-of-experimentation test. Supplies used for modeling are part of the cost of modeling and a measure of the activities that constitute elements of an experimental process.

Statutory Design Issues: Statutory interpretative issues for the research credit are plentiful. Proof issues are burdensome. A redesign of the statute incentivizing research effectively may well be in order and save government and taxpayer resources that can be put to more productive use.