Corporate taxpayers that benefit from cost of performance sourcing face the unwelcome prospect of filing correctly under a state’s statute, yet later learning that a state revenue department is imposing an alternative market-based sourcing methodology that produces more tax. It was hoped that the Tennessee Supreme Court, in Vodafone v. Roberts, would provide some concrete limits on the Tennessee Department of Revenue’s ability to exercise its alternative apportionment power. However, the court’s decision describes no such clear boundaries.

This alert provides brief background on the case, summarizes the majority and dissenting opinions, and discusses the Tennessee and national implications of the case.

Background

For all relevant periods, the taxpayer owned an interest in a partnership that provided wireless communication and data services (collectively, “wireless services”) to customers nationwide, including Tennessee customers. On its Tennessee franchise and excise tax returns, the taxpayer sourced receipts from the wireless services based on “primary place of use,” meaning customer location. Therefore, receipts from Tennessee-based customers were sourced to Tennessee.

After filing its Tennessee franchise and excise tax returns, the taxpayer was informed by an advisor that the relevant Tennessee statutes require sourcing services based on costs of performance. In other words, the taxpayer should have sourced its receipts from the wireless services based on where the costs of providing the wireless services were highest–outside of Tennessee–because the costs of operating the wireless network were highest outside Tennessee. On that basis, the taxpayer filed refund claims reflecting the costs of performance sourcing method and removing receipts from the wireless services from its Tennessee receipts factor numerator.

The Tennessee Commissioner denied the refund claims and, by way of explanation, sent the taxpayer a letter stating that the claims were being denied because a “variance” was being imposed to require receipts factor sourcing based on primary place of use (the way the returns had initially been filed). “Variance” is the Tennessee term for alternative apportionment, so significantly the Tennessee Commissioner was not arguing that the costs of performance were highest in Tennessee, but rather that the ordinary statutory method of sourcing receipts was being disregarded in this case in favor of an alternative method. 

Both at the trial court and on appeal, the taxpayer was unsuccessful in contesting the denial of its refund claim. In its argument before the Tennessee Supreme Court, the taxpayer pointed to language in the Tennessee variance regulation stating that a variance is only appropriate “in limited and specific cases” that present “unusual fact situations which ordinarily will be unique and nonrecurring” and where the statutory apportionment and allocation provisions “produce incongruous results.” (Tenn. Comp. R. & Regs. 1320-06-01.35). The taxpayer argued that this was not an unusual case warranting imposition of a variance and, therefore, it must be permitted to use the ordinary cost of performance statutory sourcing method and its refund claims must be granted.

The Majority Opinion

The majority concluded that the variance was properly imposed and, in so doing, interpreted the regulatory phrase “unusual fact situation” broadly. The majority implied that there is an “unusual fact situation” whenever the statutory formula does not source sufficient receipts to Tennessee. In the majority’s words, “…the statutory apportionment formula causes millions of dollars in receipts from Vodafone’s Tennessee customers to vanish, for tax purposes. This qualifies as an ‘unusual fact situation’ that produces an ‘incongruous result.’”

In the majority’s understanding, receipts from Tennessee customers vanish when the cost of performance test is applied. However, this is not a completely accurate assessment since receipts do not vanish, but rather are reflected in the Tennessee receipts factor denominator and not the numerator.

According to the majority, a second basis for concluding that there is an “unusual fact situation” is that the Tennessee Commissioner is “unable, as a practical matter, to verify Vodafone’s representations regarding the situs for the greater proportion of its costs.” However, the cost of performance test is inherently difficult to apply, and, for most service providers, verifying the costs is challenging. Pointing to the difficulty in applying the test as a basis for imposing a variance in this particular case seems strained, as this is a general problem with the cost of performance test, not something unique to Vodafone’s business model or facts.

The language in the majority opinion is also perhaps subject to criticism on historical accuracy grounds. The majority wrote that “the drafters of UDITPA could not have envisioned a multistate telecommunications service provider such as Vodafone.” However, UDITPA was drafted by the Uniform Law Commission in 1957. At that time, the Bell System and General Telephone & Electric Corporation were operating in multiple states, so it seems likely that the UDITPA drafters could have envisioned multistate telecommunications companies and perhaps even multistate wireless service providers such as Vodaone.

Even if the majority is right that UDITPA’s drafters could not have envisioned Vodafone’s business model, this seems a curious argument for imposing a variance. UDITPA was promulgated in 1957. There have been many new business models since that time. If the cost of performance statute produces unacceptable results for post-1957 business models, that is an indication that the statute is problematic, not that variances should be imposed on companies with modern business models that are merely filing returns in accordance with the statute.

The Dissent

Justice Jeffrey S. Bivins filed a nuanced dissent concurring in part and dissenting in part. First, he agreed with the majority that the cost of performance test produces distorted results that do not fairly reflect the taxpayer’s Tennessee business activity. However, he was unconvinced that the Tennessee Commissioner carried his burden of showing that the facts in the taxpayer’s particular case were “specific” or “unique” enough to warrant imposing a variance. 

Implications

Effective July 1, 2016, Tennessee will abandon costs of performance in favor of market-based sourcing for sales other than sales of tangible personal property. Hopefully, this will result in less “all or nothing” receipts factor sourcing and variances will be uncommon. Nevertheless, some taxpayers may find that the market-based sourcing model is difficult to apply to their business or perhaps produces less tax than did cost of performance sourcing. In such cases, one can only hope that the Tennessee Commissioner will resist the urge to impose a variance. Interestingly, one area in which cost of performance sourcing will continue to apply, at least in part, is for telecommunications companies. Under the new statutory regime, telecommunications companies will source receipts from sales other than sales of tangible personal property using the arithmetical average of: (1) cost of performance; and (2) market-based sourcing.

Vodafone v. Roberts is part of a nationwide trend of state revenue departments imposing market-based sourcing through alternative apportionment when the statutory cost of performance method produces unacceptable results from the state’s perspective. A difficulty in this is that taxpayers effectively are punished for following the statute. Results become unpredictable when state revenue departments can change statutory results to increase a taxpayer’s tax. With states increasingly abandoning cost of performance in favor of market-based sourcing, hopefully this trend will stop and the battleground will instead shift to determining where a taxpayer’s market is located.