The Presumption of Correctness
A pesky problem in state tax law

print this article

A common question tax directors ask when considering whether to litigate a state tax case is, What is the likelihood we will win? It’s an obvious question to ask, but an impossible one to answer. In our own experience, we have won cases we’d thought were stacked against us and lost cases we felt strongly about. Too many factors go into the judicial decision-making process to predict a specific outcome accurately.

This article addresses one of those peskier factors: the presumption of correctness that attaches to department-of-revenue determinations in most states. The presumption of correctness is why, even if the facts and case law are otherwise stacked equally in favor of the taxpayer and the department of revenue, the department of revenue is more likely to win the case because its position is presumed to be correct.

But does it have to be that way?

This article takes a deep dive into what the presumption means and why it even exists. Armed with a true understanding of what the presumption is intended to do, you may conclude that the department of revenue didn’t earn its litigation advantage in a particular case. Under the right facts, you can attack the applicability of the presumption itself. If successful, at a minimum you can level the playing field and—as shown in several cases discussed below—may even convince the court to cancel an assessment of tax liability altogether.

What Is a Presumption of Correctness?

It’s easy to blow past some of the boilerplate language you seem to find at the beginning of almost every tax case decision—language along the lines of “the determination of the department of revenue is presumed to be correct, and the burden rests on the taxpayer to prove otherwise.” It’s easy to read that language quickly and to keep going.

But next time, stop there for a second. What is the actual impact of that language? Does it create a burden of proof or a standard of proof? A burden of production? A burden of persuasion? Is it just language that sits there with no real impact? Does it mean that the department of revenue’s interpretations of law are presumed correct or that its findings of fact are presumed correct, or both?

The answers to these questions will vary depending on how the issue has evolved in each state court. And, oftentimes, the same court may describe the presumption and its effects slightly differently from case to case. So let’s start at the very beginning.

Why Is There a Presumption of Correctness?

The earliest recitations of some version of the presumption of correctness in case law come from United States Tax Court cases in 1902 and 1906. But, as with many older cases, the language can be arcane and abstract, so it may be helpful to consider more recent court decisions that derive support from those older decisions. Because a series of New Jersey tax cases walk through this history quite nicely, the remainder of this article focuses on New Jersey as a case study, though these issues are by no means limited to the Garden State.

The New Jersey Supreme Court, relying in part on a United States Supreme Court decision from the early 1900s, first started discussing a presumption of correctness in property tax matters. In 1933, the New Jersey Court of Errors and Appeals (at the time, the state’s highest court) described the presumption as being rooted in the “assumption that these local officers [i.e., property tax assessors] will act from a sense of duty” in reasonably assessing property, and therefore presumed their assessments to be accurate.1 In the same case, the court recognized that “[t]here is always a possibility of misconduct on the part of officials, but legislation would be seriously hindered if it may not proceed upon the assumption of a proper discharge of their duties by the various officials.”

Just a few years later, the court repeated this theme, stating that “[i]n the absence of a contrary showing, it is to be assumed that the provisions of the statute were followed” in the issuance of a proposed assessment.2 Soon after, the court repeated that “[t]he good faith of taxing officials and the validity of their actions are presumed.”3

More recent decisions from the New Jersey Supreme Court continue to reflect this justification for the presumption. For example, in the 1985 Pantasote case, the court ruled:

It is clear that the presumption is not simply an evidentiary presumption serving only as a mechanism to allocate the burden of proof. It is, rather, a construct that expresses the view that in tax matters it is to be presumed that governmental authority has been exercised correctly and in accordance with law.4

This presumption that government representatives will correctly and legally exercise their authority is, in theory, balanced by their related obligation to “turn square corners”—in other words, to act fairly—when dealing with the public.5

The Pantasote court went on to explain that the presumption of correctness could be properly invoked in a matter where the methodology used by the assessor was “incorrect” so long as any shortcomings did not “manifest an arbitrary or capricious discharge of the assessor’s responsibility.”6 Pantasote’s clear implication is that if audit deficiencies do reflect an “arbitrary and capricious discharge” of the government’s responsibilities, the presumption of correctness should not attach at all.

No discussion of New Jersey’s presumption of correctness is complete without mention of Yilmaz, Inc. v. Director, Division of Taxation.7 While Yilmaz does not explicitly state a requirement that a proposed assessment must be reasonably supported by facts and law for the presumption to attach in the first place, the Yilmaz court upheld the validity of the presumption based on the auditor “reasonably” using an established markup method, having given “reasonable consideration” to the taxpayer’s positions, the auditor’s “reasonable” use of invoices from a later period, the director’s “reasonable” use of data, and the adjustments being “reasonably arrived at.” The Yilmaz court determined that the auditor’s reasonableness (referred to no fewer than five times) in reaching his assessment justified the presumption of correctness, particularly where the taxpayer attempted to overcome the presumption through “naked assertions” regarding the legitimacy of the director’s method.

Thus, the presumption of correctness is effectively an assumption that government officials knew the rules they were applying and applied them reasonably. As a result of this assumption, the court will uphold the government action absent some “cogent” evidence to the contrary or a showing that the government’s action was arbitrary or capricious, unreasonable, or an abuse of discretion.

But surely not every government official exercises their authority to the letter of the law. So, what if the department of revenue fails to act reasonably during its audit, including ignoring its own audit protocols? What happens if an audit team goes rogue?

What If the Department of Revenue Fails to Perform a Complete and Reasonable Audit?

Sticking with New Jersey, in 2009 the Tax Court determined that the department of revenue should “reverse” an assessment if “it was arbitrary, capricious, or unreasonable . . . or . . . the findings on which it was based were not supported by substantial, credible evidence in the record.”8

Would the Tax Court actually go that far? Would it invalidate an assessment because the Division of Taxation failed to fulfill its obligations during audit?

Yes, it would, and did so in Morgan Stanley & Co. v. Director, Division of Taxation (2014).9 During the audit in Morgan Stanley, the taxpayer “went to great lengths to provide” the department of revenue with “substantial documentation” supporting its position. Instead of reviewing the documentation, the audit team reviewed a few select items and then rejected the taxpayer’s position outright.

The court determined that the department of revenue acted “unreasonably” in failing to review the taxpayer’s documentation (even though the court acknowledged that “it may very well be that a determination may have been made that those facts and circumstances did not support [the] deduction [at issue]”). Ultimately, the court concluded that because “the Director has acted unreasonably in the application of the unreasonable exception [during the audit], the court will not independently evaluate the merits of the underlying transactions.”10 Specifically, even though the taxpayer provided “substantial documentary evidence” to show that its facts aligned with those discussed in the division’s administrative guidance, the division instead relied entirely on one adverse fact, which was not determinative under either the statute or the division’s own guidance.11

Thus, the Tax Court actually cancelled a proposed assessment based on audit failures.

Similarly, the Tax Court applied these concepts more recently in a sales tax case, Saulwil, Inc. v. Director, Division of Taxation (2020), ultimately determining that one aspect of the department of revenue’s assessment would be reversed based on audit failures.12 In Saulwil, the division applied an estimated markup on the taxpayer’s sales after determining that the taxpayer’s books and records were inadequate for accurately computing the proper tax due to the absence of daily cash register tapes.13 The court, however, found that the taxpayer had maintained adequate books and records and that the division’s finding otherwise was unreasonable, since the taxpayer had provided credible documentation of its sales, including daily, monthly, and annual summaries based on its point-of-sale system, and because the division relied on mere speculation and unsupportable assumptions to determine that the taxpayer’s books and records were inadequate.14 Because the presumption of correctness afforded the division’s assessment was “based on the expectation that Taxation, as a governmental agency, will be reasonable in its initial determination that a taxpayer’s books and records are inadequate,” the court’s conclusion that this initial determination was unreasonable nullified the presumption in this case.15 The court then concluded, “Therefore, the court does not need to examine the reasonableness of the markup methodology used by Taxation.”16

Does This Result Make Sense?

Although the New Jersey courts did not frame these cases as being grounded in due process concerns, they are consistent with the constitutional requirement of fundamental fairness. The due process clauses of the United States Constitution and the New Jersey Constitution17 prevent a state from acting by “impermissible means.”18 Furthermore, government action must “reasonably relate to a legitimate legislative purpose and [not be] arbitrary.”19 Recently, the New Jersey Supreme Court confirmed that the state constitution “serves to protect [taxpayers] generally against unjust and arbitrary government action.”20

So, yes, it is absolutely appropriate to cancel an assessment if the taxpayer can demonstrate that the department of revenue did not fulfill its obligations to audit appropriately.

When Is This Fact Pattern Likely to Arise?

Three common scenarios immediately come to mind where there would be a strong argument to detach the presumption of correctness, and another tempting scenario comes to mind to make the argument, but we suspect it will likely fail.

First, as the examples above show, a department of revenue’s refusal to consider documentation provided by the taxpayer is problematic. This is not to say that departments must accept taxpayer documentation without questioning it. But if the department simply refuses to consider information outright, you can fairly question whether it is doing its job appropriately, as the Morgan Stanley court did above.

The second scenario arises when a department of revenue audit team follows its gut instead of drawing on actual information. Many of us have had auditors or supervisors tell us their version of how they “know” something works. Sometimes they’re dead wrong, and you’ve got testimonial or documentary evidence to show it. If the department of revenue can point to something other than its gut to support its position, the department may have enough to survive a presumption challenge. But if it can’t refute your evidence with anything meaningful yet continues to assess tax based on its gut, there may be a good case for a presumption argument, as in Saulwil above.

The third scenario arises when an audit team disregards specific instructions about how to audit a particular portion of the return. For example, some states have specific protocols for making a discretionary adjustment, including a transfer pricing adjustment. Did the department of revenue follow those requirements? It is often worth reviewing a state’s audit manual or audit guidelines or audit-related regulations to see if there are protocols in place. The failure to apply the department’s own guidance is at play in two matters we’re currently litigating, one in New Jersey and another in New York, and was an important consideration in Morgan Stanley, where the New Jersey Division of Taxation failed to address the taxpayer’s qualification under its own administrative guidance.

When will an attack on whether the presumption of correctness should apply in the first place most certainly fail? Such arguments should be saved for matters where there is demonstrable proof of an audit shortcoming. Simply asserting that the department of revenue didn’t understand your product or service or the legal question at issue is not enough. Taxpayers have some level of obligation to explain their facts and the application of law clearly and concisely. Arguing in litigation that the audit team didn’t bother to ask you the right questions is not going to garner any sympathy from the courts. In each example above, the taxpayer appears to have tried very hard to make things clear for the department. Unless you can show that you tried to show the department of revenue a copious set of facts or to explain your view of how the law works, your argument will likely fail.

Where’s the Proof?

Generally, there are two ways to document audit shortcomings. The first is to ask. Ask during your audit whether the audit team actually reviewed the materials you provided. Ask specific questions, ideally in writing, and ask for a written response. You can also ask about specific lines in audit work papers, something like “What documents have you reviewed during the audit or what legal authorities led you to conclude X?”

Of course, another school of thought is not to ask these questions and instead to receive a naked explanation of the department of revenue’s position, which may be easier to attack than one it worked harder to document. In this case, after the audit is finalized, you would want to ask similar questions in interrogatories or during a deposition or even on the stand if the matter has progressed all the way to trial.

The second way to determine if the department of revenue followed its own protocols is a careful review of the audit file. Many states allow a taxpayer to obtain a copy of the audit file through a public records request or an interrogatory. Most of the time, doing so results in your receipt of hundreds of pages of materials that don’t seem particularly revealing. But every so often you uncover a zinger in there.


Admittedly, it will be a rare occasion that a taxpayer can demonstrate sufficient failures on the part of the department of revenue to undermine a proposed assessment. But if a taxpayer manages to prove those failures, the assessment should be reversed, regardless of whether the quantum of the assessment is close to or far from the mark.

Leah Robinson is a partner in the New York office of Mayer Brown and leads the firm’s state and local tax group. Michael Kerman is counsel in Mayer Brown’s Washington, D.C., office and a member of the state and local tax group.


  1. Central R.R. Co. of N.J. v. State Tax Dep’t, 112 N.J.L. 5, 8 (1933), citing Michigan Central R.R. Co. v. Powers, 201 U.S. 245, 295 (1906).
  2. N.J. Bell Tele. Co. v. Newark, 118 N.J.L. 490, 495 (1937).
  3. N.J. Bell Tele. Co. v. Camden, 122 N.J.L. 270, 275 (1939).
  4. Pantasote Co. v. Passaic, 495 A.2d 1308, 1311 (N.J. 1985).
  5. For a recent example, see Saulwil, Inc. v. Director, Division of Taxation, 31 N.J. Tax 433 (2020), in which the ruling stated that “the square corners doctrine impacts Taxation’s determinations, and thus, balances with the presumption of correctness provided to those determinations.”
  6. Pantasote, 495 A.2d at 1311.
  7. Yilmaz, Inc. v. Director, Division of Taxation, 22 N.J. Tax 204 (2005), aff’d 390 N.J. Super. 435 (App. Div.), cert. denied 192 N.J. 69 (2007).
  8. AccuZip, Inc. v. Director, 25 N.J. Tax 158, 167 (2009).
  9. Morgan Stanley & Co. v. Director, Div. of Taxation, 28 N.J. Tax 197 (2014).
  10. Morgan Stanley at 226.
  11. Morgan Stanley at 216, 223.
  12. Saulwil, Inc. v. Director, Division of Taxation, 31 N.J. Tax 433 (Tax 2020).
  13. Saulwil at 444.
  14. Saulwil at 449, 453. In response to the division’s unsupported belief that the taxpayer’s point-of-sale system had the potential to be manipulated by a manager or owner, the court rightly noted “[t]hat something can happen is not credible proof that it happened.”
  15. Saulwil at 447.
  16. Saulwil at 458. Of course, the court suggested that the markup methodology was not in fact reasonable where the need for a markup was based on the auditor’s “lay opinion” that receipts were too low and the markup percentage used was based merely on the auditor’s “experience,” rather than estimates based on external indices.
  17. As framed in the Fourteenth Amendment of the US Constitution and in article 1, paragraph 1 of the New Jersey Constitution.
  18. Greenberg v. Kimmelman, 99 N.J. 652, 561–62 (1985).
  19. Greenberg, citing Nebbia v. New York, 291 U.S. 502 (1934).
  20. Oberhand v. Director, Division of Taxation, 193 N.J. 558, 577 (2008), with Justice Albin concurring.

Leave a Reply

Your email address will not be published. Required fields are marked *

XHTML: You can use these tags <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>