The FedEx Ruling: Which Rules Applied to the R&D Credit?

FedEx spent years building an internal billing system meant to fix revenue tracking across a transaction volume no existing system could handle. By 2001, the technological challenges had proved too great to overcome, and FedEx abandoned the project. It still claimed research credits on the work, for tax years 1997 through 2000. The IRS disallowed the claim, and FedEx went to court seeking a refund of more than $11 million.

The dispute that followed wasn’t about whether FedEx’s engineers did real work. It was about a regulatory mess Treasury created for itself.

The Regulatory Problem Behind the Case

In January 2001, Treasury issued final regulations under Section 41 that included a stricter version of the discovery requirement. The rules also contained provisions addressing internal-use software, meaning software developed primarily for a company’s own operations rather than software developed for sale or external use.

After reviewing comments, the statute and its legislative history, Treasury and the IRS concluded that the 2001 discovery requirement did not fully address Congress’s concerns and proposed a revised standard.

The December 2001 proposed regulations eliminated the discovery test that had created the controversy and changed other aspects of the research credit rules. However, the internal-use software provisions did not move forward in the same way.

When Treasury issued final regulations in 2003, those regulations did not finalize the internal-use software provisions. Treasury later issued additional guidance addressing the status of the earlier IUS rules while it continued working on the issue.
That left FedEx with a difficult regulatory question.

Could it rely on the later general research rules while also relying on the earlier internal-use software provisions? Or, as the IRS argued, did using the earlier IUS rules also require FedEx to accept the older discovery standard?

What FedEx and the IRS Argued

The IRS’s position: pick one complete package. Either the full 2001 regulations, tough discovery test and IUS rules together, or the full 2003 regulations, easier discovery test with no usable IUS provisions.

FedEx argued it should get to combine the best of both: the lenient 2003 discovery test with the 2001 IUS rules, since the 2003 regulations never actually revised those IUS provisions.

What the Court Actually Decided

In FedEx Corporation v. United States, the Court sided with FedEx on the regulatory question. The court held that FedEx could rely on the later general research rules while also relying on the applicable internal-use software provisions from the earlier regulatory framework.

The court also rejected the IRS’s attempt to use later administrative guidance to bring the older discovery requirement back into the analysis for FedEx’s IUS claim.

In other words, the court did not accept the IRS’s position that FedEx had to take the older discovery test simply because it was relying on the earlier IUS rules.

But there is an important limitation here. The court did not decide that FedEx’s software automatically qualified for the federal R&D tax credit. It decided which legal standards should be used to evaluate the claim.

Winning the Standard Is Not the Same as Winning the Credit

Here’s what this ruling didn’t do: It didn’t decide whether FedEx’s abandoned billing project actually met the discovery test or the IUS standard. It only decided which standard applied.

FedEx still had to establish that the work satisfied the applicable requirements for qualified research and, because the software involved internal use, the applicable internal-use software requirements.

The distinction matters because a taxpayer can win an argument about which test applies and still have to prove that its activities satisfy that test. FedEx’s billing project had already been abandoned years before the litigation reached this stage. The research itself had been performed even earlier.

So the court’s ruling resolved one part of the dispute, not the entire question of whether the credit was ultimately allowable. Winning the standard is not the same as winning the credit

Why the FedEx Case Still Matters

The specific regulatory gap here is gone. IUS rules were rebuilt from scratch in 2016 under a new high-threshold-of-innovation test. But the pattern that created FedEx’s problem still shows up whenever Section 41 regulations change:

A claim can span a period where old and new rules overlap in ways nobody worked out in advance.
The IRS may take the position that you’re locked into the least favorable combination of standards.
None of this touches whether your actual research qualifies. It just determines what test gets applied to it.

If your R&D credit claims involve software development, or cover years that straddle a regulatory change, know which framework governs your specific tax years before that question gets raised in an audit.

The Tax Year Matters

R&D tax credit rules have changed over time. That means a business looking at an older software development project cannot necessarily apply today’s rules without first considering the tax year involved.

This becomes especially important when a claim involves:

Internal-use software
Older tax years
Software developed during a period of regulatory change
Claims that are being examined or defended years after the research took place

The question is not simply, “Does this software qualify under today’s rules?”

The better question is:

What rules apply to the research performed in the specific tax year being claimed?

FedEx shows why that distinction can matter.

What Businesses Should Take From FedEx

There are three practical lessons from the case.

Identify the applicable rules before evaluating the claim.

A software project should be evaluated under the rules that apply to the relevant tax year and facts. Current guidance should not automatically be treated as though it governed every historical claim.

Internal-use software requires additional attention.

Software developed primarily for a company’s own internal operations can be subject to additional requirements under Section 41. Businesses should determine whether the internal-use software rules apply before assuming ordinary software development rules are enough.

Meeting the right legal standard still requires evidence.

Even after determining which test applies, the taxpayer still needs to establish that the underlying research satisfies the applicable requirements.

That is ultimately what makes FedEx useful as a case study.

The court’s ruling was about the standard.

The credit still depended on the research.

Getting the Rules Right Before You File

TaxDrone.AI helps businesses organize and evaluate their federal R&D tax credit claims with attention to the tax years, software use, and qualification requirements relevant to the claim.

For software development, that means looking at the nature of the software, how it was used, the tax year involved, and whether internal-use software rules apply.

The goal is not to assume that an older case controls a current claim.

It is to make sure the claim is being evaluated under the framework that actually applies to it.

Before You Assume the Current Rules Tell the Whole Story

If your business has federal R&D tax credit claims involving software development, internal-use software, or older tax years, it is worth confirming which rules apply before the issue comes up during an IRS examination.

Schedule a free consultation with our R&D tax specialists to discuss your software development activities and evaluate your federal R&D tax credit opportunities.