Construction Law, Transactions and Disputes

The United States Tax Court released its opinion today on Smith v. Commissioner, T.C. Memo. 2026-50, and it is a significant R&D credit win for taxpayers that perform sophisticated technical work for clients. I am particularly satisfied with the outcome, having had the privilege of litigating this case.

This matter involved research credits claimed by Adrian Smith + Gordon Gill Architecture, LLP, or AS+GG, for research activities performed in connection with large scale architectural projects during the 2008, 2009, and 2010 tax years. AS+GG is known for ambitious, complex, and highly sustainable architectural design. The projects at issue involved supertall towers, zero carbon and low energy design concepts, wind studies, solar orientation analysis, structural and environmental performance issues, and other technical design challenges.

This context matters, as the case was not about routine design work. The opinion describes a firm working at the edge of what architecture, engineering, and sustainable design could accomplish. In that respect, the case is a useful reminder that R&D credits are not limited to laboratories, software companies, or manufacturing floors. Innovation also happens in design studios, engineering meetings, building models, wind studies, and iterative technical problem solving.

The IRS Conceded the Four-Part Test

One of the most important features of the case is what was not in dispute by the time of trial.

The IRS ultimately conceded that AS+GG’s claimed business components satisfied the four-part test for qualified research under section 41(d). That concession mattered, and it did not happen by accident.

A significant part of the case was devoted to developing the factual record on the nature of AS+GG’s work. Through extensive discovery, the taxpayers built the record showing that the projects involved real technical uncertainty, a process of experimentation, and design work directed at performance, sustainability, structure, energy usage, wind behavior, solar orientation, and other technical objectives.

By the time of trial, the government no longer tried the case on the theory that the work failed the four-part test. Instead, the IRS conceded that issue, and the trial focused on the funded research exclusion under section 41(d)(4)(H).

That is a major point. For taxpayers in architecture, engineering, design, and other technical service industries, the concession reinforces something important: sophisticated client work can involve qualified research. The remaining fight was not whether AS+GG’s work was research. It was whether the funded research rules limited the credit.

The Funded Research Issue

The funded research exclusion is one of the most important limitations on R&D credits for companies that perform technical work under customer contracts. Section 41(d)(4)(H) excludes research “to the extent funded” by another person.

The regulations generally ask two questions:

• First, was payment contingent on the success of the research
• Second, did the taxpayer retain substantial rights in the research

The Court applied that familiar framework. The taxpayers argued that, after Loper Bright, the Court should not simply accept the regulatory test and should instead adopt a narrower reading of “funded.” The Court rejected that argument and held that the funded research regulations remain valid.

That part of the opinion is important, but it should not obscure the practical taxpayer win. Even applying the regulatory framework urged by the government, the taxpayers prevailed on the central issue that mattered most for four of the six sample projects: substantial rights.

The Taxpayer Win: Substantial Rights

The heart of the opinion is the Court’s substantial rights analysis.

The Court held that AS+GG retained substantial rights in the research performed on four of the six sample projects: Atrium City Tower, Masdar HQ, Atrium City Masterplan, and Plot R2.

That holding matters because a taxpayer that retains substantial rights in the research may still claim R&D credits on a reduced basis, even if the Court concludes that payments were not contingent on success. The practical result is that AS+GG was not treated as having simply performed fully funded research for a customer on those projects with no credit available. Instead, the Court recognized that AS+GG retained meaningful rights to use the results of its research in its business.

That is a significant result in a funded research case.

It is also a lesson in contract drafting. The Court did not treat intellectual property provisions, copyright language, licenses, confidentiality clauses, settlement agreements, and use restrictions as boilerplate. It parsed the language project by project. On some projects, the language preserved enough rights for the taxpayer. On others, it did not.

That project-by-project analysis is one of the most useful aspects of the opinion. It shows that small differences in contract language can produce very different R&D credit outcomes.

Why this Matters for R&D Credit Taxpayers

Smith is especially useful for architecture, engineering, construction, consulting, design, and other technical service businesses.

Many of these companies assume that R&D credits are unavailable because they perform work for clients. That assumption is too broad. A customer contract does not automatically eliminate the credit. The funded research rules are more nuanced. The key questions are who bears the relevant risk and whether the taxpayer retains meaningful rights in the research.

The opinion reinforces several important points.

First, design and architecture can involve qualified research. The Court’s factual findings describe technical work involving sustainability, wind, energy, structure, geometry, performance, and environmental constraints. Those are exactly the kinds of uncertainties that can support R&D credit claims when properly documented.

Second, substantial rights matter. A taxpayer does not necessarily lose the credit merely because the customer receives project rights, licenses, or deliverables. The question is whether the taxpayer retained meaningful rights to use the research results in its own business.

Third, contracts matter. The funded research analysis is driven heavily by the agreement between the taxpayer and the customer. Taxpayers who wait until audit to think about substantial rights are often too late.

Fourth, documentation still matters. The Court left the precise credit amounts to be determined through computation. That is a reminder that winning the legal issue is not enough. Taxpayers also need project level documentation and expense substantiation that allow the allowable credit to be calculated.

A Strong, Reasonable Compensation Win

The opinion also includes an important win on reasonable compensation, an issue I personally handled at trial, and one of the most satisfying parts of the case.

AS+GG’s R&D credit included wage related qualified research expenditures attributable to the partners. The government challenged the partners’ 2008 compensation under section 174(e), arguing for a much lower reasonable compensation amount.

At trial, a central part of the taxpayers’ case revealed that the government’s reasonable compensation theory did not fit the economics of the business. That point came through clearly in the expert testimony. The IRS expert advanced a much lower compensation number under a multifactor analysis, but his own independent investor analysis showed that, even after his adjustments, AS+GG generated a 939% return on equity. On that math, he concluded that the partners’ total compensation was not unreasonable.

That was a powerful trial point. The government’s expert opinion, when tested against the controlling Seventh Circuit standard, supported the taxpayers.

The Court agreed that the independent investor test applied because the case was appealable to the Seventh Circuit. Once that standard governed, the result followed: the partners’ 2008 compensation was reasonable under section 174(e).

That holding is meaningful for owner-operated businesses where the people driving the research are also principals of the business. In those cases, the R&D credit often depends not only on whether the work qualifies, but also on whether compensation paid to key technical leaders can be included in wage QREs. Here, the Court accepted the taxpayers’ position and preserved an important component of the credit.

Practical Takeaways

The biggest takeaway from Smith is that R&D credit planning should begin before the contract is signed.

Taxpayers performing technical work for customers should review their agreements for at least three things:

• First, what happens if the research fails
• Second, who owns the work product, technical information, drawings, designs, models, methods, and other research results
• Third, does the taxpayer retain the right to use what it learned and developed in future work

Those questions should be addressed in the contract itself. They should not be left to implication, course of dealing, or after-the-fact argument.

This case also highlights the importance of project-level documentation. Taxpayers should identify the business components, the technical uncertainties, the process of experimentation, the employees or principals performing qualified services, and the expenses tied to the research.

Conclusion

Smith v. Commissioner is a significant R&D credit decision and a gratifying taxpayer win.

The IRS conceded that the work satisfied the four-part test. The Court held that AS+GG retained substantial rights in four of the six sample projects and allowed the taxpayers to claim research credits to the extent permitted by the funded research rules. The Court also accepted the taxpayers’ position on reasonable compensation, preserving an important wage QRE issue.

For taxpayers in architecture, engineering, design, consulting, and other technical service businesses, the message is encouraging: customer contract work does not automatically defeat the R&D credit. But the contract language matters, the retained rights matter, and the record matters.