chevron-down Created with Sketch Beta.

The Tax Lawyer

The Tax Lawyer: Spring 2020

Burden of Proof in Tax Cases: Valuation and Ranges—An Update

John A Townsend

Summary

  • Understanding the interplay of the burden-of-proof rules and ranges in valuations can help the parties and the courts focus more crisply on litigation needs.
  • Sometimes when ranges are identified, arbitrary conventions (such as the midpoint in trades as in the case of publicly-traded stock) can be used to determine the value in tax litigation. But when there is no such convention that should be applied, the burden of persuasion can resolve the valuation issue by identifying the range.
  • Careful analysis may not only identify a preliminary appropriate range but also further assist the court in finding the “sweet spot” with the requisite degree of persuasion.
Burden of Proof in Tax Cases: Valuation and Ranges—An Update
boonchai wedmakawand via Getty Images

Jump to:

Abstract

In this Article, the author discusses the difficulty in many valuation cases of determining a definite valuation point by the required degree of persuasion (more likely than not in most civil cases). This point was made cogently in Cede & Co. v. Technicolor, Inc., a frequently cited opinion by the Delaware Court of Chancery, a forum for significant litigation involving corporate valuations:

[I]t is one of the conceits of our law that we purport to declare something as elusive as the fair value of an entity on a given date. . . . [V]aluation decisions are impossible to make with anything approaching complete confidence. Valuing an entity is a difficult intellectual exercise, especially when business and financial experts are able to organize data in support of wildly divergent valuations for the same entity. For a judge who is not an expert in corporate finance, one can do little more than try to detect gross distortions in the experts’ opinions. This effort should, therefore, not be understood, as a matter of intellectual honesty, as resulting in the fair value of a corporation on a given date. The value of a corporation is not a point on a line, but a range of reasonable values, and the judge’s task is to assign one particular value within this range as the most reasonable value in light of all the relevant evidence and based on considerations of fairness.

Corporate valuations for estate tax are just one context of tax litigation, but there are many other contexts. A prominent example for some time now has been transfer pricing.

Sometimes when ranges are identified, arbitrary conventions (such as the midpoint in trades as in the case of publicly-traded stock) can be used to determine the value in tax litigation. But when there is no such convention that should be applied, the burden of persuasion can resolve the valuation issue by identifying the range. The party bearing the burden of persuasion (or risk of nonpersuasion) then has persuaded only as to the end of the range that does not favor that party; the value, based on persuasion, is determined accordingly.

The party bearing the burden of persuasion in tax cases is usually the taxpayer. In this Article, the author discusses interesting features of the burden and how, at least before the Tax Court, the burden of persuasion might shift to the Service under the Supreme Court’s decision in Helvering v. Taylor, which the author urges is often misunderstood.

Another benefit of identifying a range of values is that, if it is determined on appeal that the trier of fact misapplied the burden of persuasion but did identify the range, the court of appeals can resolve the case by picking the other end of the range (unless a successful attack is made on the trial court’s choice for the ends of the range).

I. Introduction

In 2019, Tax Court Judge Gustafson delivered the fourth chapter in a long-running, highly episodic gift-tax valuation case involving the Cavallaros, taxpayers who are husband and wife (Cavallaro IV). The Service started its examination of the transaction in 1998. The litigation history beginning during the examination is as follows: The Cavallaros unsuccessfully opposed a third-party summons (Cavallaro I). After the Service issued notices of deficiency and the Cavallaros unsuccessfully litigated in the Tax Court (Cavallaro II), the Cavallaros then, mostly unsuccessfully, litigated in the Court of Appeals (but with a partial limited remand) (Cavallaro III). On the remand and over 20 years after the examination started, the Tax Court in 2019 acted for what may be the Tax Court’s last time in Cavallaro IV, although there may be an appeal bringing us Cavallaro V.

Although a detailed discussion of the Cavallaro decisions appears later in this Article, the following high-level summary will provide the necessary context. The issue was whether, in a merger involving family-owned corporations, the younger generation received disproportionately more stock in the merged corporation than its contributions so that the older generation made disguised gifts through the merger. Variations on this theme are not uncommon as a way of shifting wealth to a younger generation free of transfer tax. The key to the planning is to adopt and then, if necessary, sustain a valuation that supports the desired transfer tax outcome (i.e., no gift).

The valuation issue presented in the Cavallaro cases has overtones of the issue I discussed in an article many years ago. The Cavallaro family saga, with its most recent episode, Cavallaro IV, offers an opportunity to update that article. My original article was inspired by Professor Leandra Lederman’s then-recent article on arbitrary notices of deficiency and burden of proof in valuation cases. At the time of the article and since, the burden-of-proof issue has rumbled around in my mind. In the early 1970s, while gainfully, sometimes even productively, employed by the Appellate Section of the Department of Justice Tax Division, I first visited the burden-of-proof issue. I had cases requiring that I research and brief burden-of-proof concepts. I perceived that there was much confusion in the courts’ discussions of burden of proof in tax cases. At the encouragement of the Tax Division, I prepared a memorandum on the burden of proof in tax cases. The memorandum was supposed to guide Tax Division attorneys in asserting positions before the courts with the hope that, over time, the courts’ discussions might be more consistent and precise. The memorandum was circulated to all Tax Division attorneys and, I am told, was read by many of the attorneys but then ignored by most. The continued confusion in this area as addressed by Professor Lederman confirmed that my memorandum had had no effect upon the quality of the discussions by the courts in the intervening years. So, Professor Lederman’s article inspired me almost 20 years ago to rethink the burden-of-proof issue in the context of valuation.

The argument in this Article is that, in valuation cases, a trier of fact will often be unable to set a definite value but will be able to establish a range of values based on persuasion. This is consistent with real-world valuations, in which many appraisals provide ranges rather than definite values. The trier of fact thus may be persuaded to the required degree (preponderance of the evidence in the general civil case) that the value cannot be less than $X nor more than $Y. Depending upon the quality of the evidence, that range may be large or small. If the finder of fact can rationally find some persuasive value point, there is no range of nonpersuasion (often called equipoise in the language of burden of proof) and burden of proof is irrelevant. If, however, the trier of fact cannot rationally find some definite point within a range by a preponderance of the evidence, the evidence regarding value is in a state of equipoise within the range. Nevertheless, valuation ranges can be deployed to reach correct results in tax cases based on the allocation of the burden of persuasion between the parties. The argument is that, when the trier of fact has not been persuaded as to a definite value and is left with a range of reasonable, but not persuasive, values, the trier of fact should select the value at one of the ends of the range based on which of the parties bears the burden of persuasion. The party bearing the burden of persuasion should suffer a valuation at the end of the range that favors the party not bearing the burden of persuasion.

II. Valuation and Ranges

Valuation often determines the outcome of litigation. In the tax context, valuations are involved in such commonly encountered areas as estate and gift taxation, transfer pricing, and charitable contribution deductions. In these contexts, “[t]he fair market value is the price at which the property would change hands between a willing buyer and a willing seller, neither being under any compulsion to buy or to sell and both having reasonable knowledge of relevant facts.” Note the use of “the,” a definite article that is commonly read in this context “as a function word to indicate that a following noun or noun equivalent is a unique or a particular member of its class.” The question is what this definition requires in litigation if, in truth, the fair market value is elusive as it often is.

Some valuation issues are easy if there is a readily accessible trading market for similar property (e.g., ExxonMobil shares), but even there “the” value uses ranges with arbitrary assumptions as to where within such ranges “the” value should be determined. Valuations of other assets are more problematic. Specifically, for present purposes, how are corporate businesses valued? How are shares in closely held, nonpublicly-traded corporations valued? The same questions may be asked for other types of assets (such as real estate) for which no easily accessible proxies exist to determine “the” fair market value.

For purposes of discussion, unless stated otherwise, I will assume that the target of the valuation is a business in a closely held corporation that requires that the business or the shares in the corporation be valued. In this context (but in others also), valuation is difficult. Here is an iconic statement used often by Delaware courts and others in valuing businesses and corporations:

[I]t is one of the conceits of our law that we purport to declare something as elusive as the fair value of an entity on a given date . . . . [V]aluation decisions are impossible to make with anything approaching complete confidence. Valuing an entity is a difficult intellectual exercise, especially when business and financial experts are able to organize data in support of wildly divergent valuations for the same entity. For a judge who is not an expert in corporate finance, one can do little more than try to detect gross distortions in the experts’ opinions. This effort should, therefore, not be understood, as a matter of intellectual honesty, as resulting in the fair value of a corporation on a given date. The value of a corporation is not a point on a line, but a range of reasonable values, and the judge’s task is to assign one particular value within this range as the most reasonable value in light of all the relevant evidence and based on considerations of fairness.

Conventions can be used to determine the fair market value within a range. An oft used convention is to split the difference. But the split and the result are arbitrary, not an actual finding of the fair market value. It is simply a convention to determine what serves, at best, as an approximation that might be deemed the fair market value. That or some other convention could be adopted for judicial determinations of value, but, as far as I am aware, no such convention has been adopted to resolve tax cases (other than in some contexts such as publicly-traded stock noted above). Rather, in tax cases (and I suspect nontax cases) the quest is to determine the value based on persuasion, rather than using a convention to force the value. Ranges do not express the value, but, instead, provide the high and low values based on persuasion.

When a jury is the trier of fact as to value, the jury may pick the value from within a range of values based on the evidence introduced and need not explain its choice. If the jury determines a value outside the range the trial court or appellate court believes a reasonable jury could find, the jury’s determination will and should be rejected, but the jury can pick any value within the range for any reason without explanation or risk of reversal.

When judges are triers of fact, however, they must explain their determinations of value. But, the concept of a range necessarily means that the judge cannot determine and explain a persuasive reason for selecting a particular value inside that range. In such a situation, the judge should select the value at one of the ends of the range based on the allocation of the burden of persuasion between the parties. Consequently, the judge need only determine the appropriate range for valuation. The case can then be resolved simply by deciding which party bears the burden of persuasion.

A careful reader may ask why the judge does not simply assign the burden of persuasion and then only address the evidence’s persuasiveness to establish the relevant value, thus requiring only the finding of a single value at the appropriate end of the range of nonpersuasion rather than the finding of two values at each of the ends of the range. For example, if the taxpayer has the burden of persuasion in a tax case to establish the highest value (as in, for example, the deduction of a charitable contribution), the trier of fact need only determine, in a single step, the lowest value from the range of persuasive valuations. Why bother with the extra step (and work) to establish the other (i.e., irrelevant) end of the range? I think there are two reasons. First, because the evidence for difficult-to-value assets will be meaningful only in terms of ranges, the trier of fact will have assistance in identifying possible ranges and should be able to calibrate each end of the range without much more difficulty than determining value at only one end of the range. Second, a one-step approach of determining the value relevant only to the party upon whom the burden of persuasion rests will be meaningless if the trier of fact has misallocated the burden of persuasion. Importantly, if, on appeal, the trial court erred regarding the allocation of the burden of persuasion, the appellate court can resolve the case without further trial-level proceedings if the trial judge determined both ends of the range.

Fights concerning the allocation of the burden of persuasion are common in valuation cases in the Tax Court (the venue for the great majority of tax cases) because the Service is likely to adjust the valuation it used in issuing the notice of deficiency or the Tax Court will find a value less than the Service used in issuing the notice of deficiency. These fights often are based on the application of the Supreme Court’s decision in Helvering v. Taylor that suggests that the burden of persuasion is shifted to the Service if the evidence shows the Service’s determination (in this context valuation) underlying the notice of deficiency is arbitrary and excessive. This Article will explore some aspects of Helvering v. Taylor and the allocation of the burden of persuasion in tax cases.

As part of this discussion, I will use certain examples with assumed ranges bounded by high and low values. I often use significant ranges for illustration purposes. In real life, in well-litigated cases with reasonable parties, the ranges are likely to be more constricted.

III. General Burden-of-Proof Concepts

A. Context

Burden-of-proof concepts assist in finding facts at trial. The classic discussion model to highlight the role of burden-of-proof concepts is the jury trial model. In a jury trial, the jury is the ultimate trier of fact. The judge, however, determines whether the evidence is of sufficient quality that a reasonable jury could find either way on the fact issue. If the evidence is of sufficient quality that a reasonable juror could reach only one result, the judge will decide the case on a motion for directed verdict without submitting the case to the jury. In this discussion, I use the jury trial model because it best teaches the function of burden-of-proof concepts, but the concepts are equally applicable (although often blurred) in bench trials.

Burden of proof is divided into two burdens—the burden of persuasion and the burden of production—that serve different functions in the trial. A significant related concept, called a presumption, can affect at least the burden of production.

B. Burden of Persuasion

The ultimate burden of proof is the burden of persuasion. Most often when burden of proof is mentioned, burden of persuasion is meant. The party bearing the burden of persuasion loses if the evidence is not persuasive to the degree required in order to prevail. In risk-analysis terms, the burden of persuasion is often referred to as the risk of nonpersuasion. A litigant with a burden of persuasion on a fact issue bears the risk that the evidence will not persuade the jury of a fact that is critical to its case.

Normally, in a civil trial, in order to meet the burden of persuasion, the evidence must persuade the trier of fact that the fact in question is more likely than not true. Many authorities quantify or at least illustrate this burden as a likelihood greater than 50% (50+%), and I adopt this rough and ready (albeit not perfect) model for the normal civil burden of persuasion. In certain contexts, a higher quality of evidence is required to persuade and thereby to prevail, but the higher burdens are not important to the issues discussed in this Article. In all events, at whatever level of persuasion is required, one party bears the burden of persuasion and the corresponding risk of nonpersuasion.

In the jury trial paradigm, the jury makes the determination whether the evidence is persuasive to the extent required—50+% for present purposes. You can see, using the 50+% benchmark for persuasiveness, if the jury is 51% persuaded as to the existence of the fact, it will find the existence of that fact. Similarly, if the jury is 51% persuaded as to the nonexistence of the fact, it will find that that fact does not exist. At least in theory, the same is true if the jury is 50.1% persuaded of the existence or nonexistence of the fact. If the jury is thus persuaded (either for or against the existence of the fact), which party actually bears the burden of persuasion is irrelevant. The burden of persuasion operates to determine which party wins if the jury is unpersuaded that the fact does or does not exist—a state of equipoise. In the percentage model, this phenomenon occurs when the evidence is 50% for the existence of the fact and 50% for the nonexistence of the fact or, stated otherwise, the trier of fact just is not persuaded as to a value of more or less than 50%. This Article deals with the resolution of cases when the evidence is in a state of equipoise, in which case the allocation of the burden of persuasion determines the party who wins on that fact.

I think that situations in which the evidence is in equipoise are not generally encountered. Usually, triers of fact are persuaded as to the existence or nonexistence of the fact and that is the end of the matter. Only when the evidence is in a state of equipoise does the burden of proof determine the finding on that fact. Indeed, most trial observers believe that outcome-determinative states of equipoise are rare, but that they can occur. Specifically, in the context of valuation, equipoise happens when the trier of fact concludes that the evidence introduced results in a range of values which, by definition, exists only if the points inside the range are not persuasive (i.e., the evidence is in equipoise).

The following examples illustrate these burden-of-persuasion concepts.

Example 1: A sues B, a tax lawyer, for malpractice. A claims in the complaint that the damages from B’s alleged negligence are $1 million. At trial, who has the burden of persuasion? Of course, A does. A seeks damages from B and asks the court to enter a judgment quantifying the amount of money A is entitled to collect. A, therefore, must prove not only liability but also the amount of her damages. Assume that A proves negligence or even better that B, being the honest fellow that he is (consistent with the standards of his profession), admits liability. A’s recovery thus turns only on proof of the amount of damages. Assume the jury is persuaded that $800,000 is the right number. A is awarded damages in that amount. The jury bases its verdict on persuasion and does not need or apply the rules regarding the burden of persuasion to determine the amount of damages.
Example 2: Same situation, except the jury is persuaded that damages are at least $800,000 but not more than $900,000. In burden-of-persuasion terms, the jury is not affirmatively persuaded to the extent required (50+%) that the amount exceeds $800,000. The evidence regarding the amount of damages is in equipoise between $800,000 and $900,000. The jury should award damages at $800,000. A has not borne her burden of persuasion as to any greater amount. So, in this respect, this situation is like that of Example 1 except that all the jury has to determine is the low end of the range of damages—that point beyond which it is not persuaded. Although the jury picked the value, my experience of group dynamics on such determinations is that the group (the jury here) starts out with a range (perhaps as to individual members’ ranges within the larger group range), which the group then refines to a definite number the group agrees upon. So, although there may have been no formal determination as to the top end of the range, the range was in play in the discussion. I will return to this concept later in refining the discussion.

C. Burden of Production

The burden of production, also referred to as the burden of going forward, is related to the burden of persuasion but is addressed to the role of the judge in the jury trial paradigm. The burden of production serves the role of assuring that the evidence is of a certain minimum level of quality in order for the case to be submitted to the jury to determine whether the evidence is persuasive as to the critical fact. Basically, the burden of production asks the question whether, based on the evidence introduced, reasonable minds could differ as to the existence or nonexistence of the fact.

The quantum of evidence to meet the burden of production is not that the fact must be found in favor of the party bearing the burden; rather it is only that a reasonable juror could find in favor of that party. Stated alternatively, a directed verdict will be rendered if no reasonable juror could find in favor of the party bearing the burden of production on that fact. For example, if the judge determines that a reasonable juror could not find fact A, he will take that decision away from the jury. At any given point in a trial, the party who bears the risk of having the case taken away from the jury can be said to bear a burden of production (or burden of going forward), and, while the burden of production may be with a party at the beginning of trial, it can shift depending upon the quality of the evidence as perceived by the judge. In risk-analysis terms, the burden is described as the risk of nonproduction.

The function of the production burden is less crisp in valuation cases. I think, however, that its role could be illustrated in the context of Examples 1 and 2 above, involving damages which are a form of valuation.

Example 3: A has the burden of production because A has the burden of persuasion. Assume that, at the end of A’s case in chief, the evidence compels damages in the amount of $850,000 (the judge believes no reasonable juror could find otherwise). If B then rests his case without the introduction of any evidence, the judge will render a directed verdict for A. B had a burden of production (risk of nonproduction) that he failed to meet. If, however, the evidence is not sufficiently compelling (as is typically the case for valuations for a definite value), the judge may be able (at least in the judge’s mind) to determine that the evidence will permit a reasonable juror to determine damages between $600,000 and $1 million. The judge will submit to the jury the issue of the proper quantum of damages based on the jury’s view of the persuasiveness of the evidence (although the judge is not likely to advise the jury of his view of the high and low ends of reasonableness). If the jury nevertheless determines damages outside the range viewed as reasonable by the judge, the judge can grant a new trial or even, depending on the quality of evidence, a damage amount notwithstanding the verdict.

In these examples, A has the burden of persuasion and necessarily has the burden of production. A must produce evidence (or assume the risk if evidence is not produced) in order to persuade a jury. If A introduces no evidence in her case in chief, B’s lawyer will move for a directed verdict that should be granted. Without evidence, there is no factual dispute to submit to the jury to resolve based on the persuasiveness of the evidence. Thus, in these examples, A must introduce sufficient evidence that would permit a reasonable juror to find some amount of damages.

D. Presumptions.

A related concept is the presumption. The role of presumptions in civil trials is a large and nuanced subject, so I offer here only a general summary of what is called the “traditionalist” view” of presumptions, incorporated in Rule 301 of the Federal Rules of Evidence. Rule 301 provides:

In a civil case, unless a federal statute or these rules provide otherwise, the party against whom a presumption is directed has the burden of producing evidence to rebut the presumption. But this rule does not shift the burden of persuasion, which remains on the party who had it originally.

If the opposing party to whom the burden of production has shifted (the party not having the burden of persuasion) introduces the minimum evidence sufficient to meet the burden of production as to the nonexistence of the presumed fact, the presumption has no further effect.

Presumptions must be distinguished from inferences. A presumption requires that, if fact A is proved, fact B must be presumed unless the presumption is rebutted; an inference permits the trier of fact “to deduce the existence of fact B from fact A by ordinary rules of reasoning and logic.” In the jury trial paradigm, presumptions are directed to the burden of production managed by the trial judge without involving the jury while inferences are directed to the trier of fact, the jury, which finds facts based on inferences from the evidence. Although a presumption may be rebutted by evidence sufficient to meet the opposing party’s burden of production, the trier of fact (the jury) may still consider reasonable inferences that in its judgment are permitted by the facts that raised the presumption in the first place.

Example 4: C, the plaintiff, sues D, the defendant, on a contract. The contract provided that, if fact X exists and C timely notified D in writing of its existence, C would be entitled to a $1 million payment from D. The parties agree that fact X exists. The issue for trial is whether D received timely notice from C. C can introduce evidence only that C deposited written notice in the U.S. mail, properly addressed and with sufficient postage prepaid, in sufficient time for D to have received the notice. C cannot, however, prove that D actually timely received the notice, and the contract requires D’s timely receipt of the notice. The issue the jury must ultimately resolve is whether D timely received the notice. C may use a rebuttable presumption, the so-called “mailbox rule,” to survive D’s motion for a directed verdict at the conclusion of C’s case in chief. The presumption provides that, upon proving timely and proper mailing of the notice, the fact of timely receipt may be inferred by the jury. The presumption will at least meet C’s burden of production on the issue of timely receipt and, under Rule 301, will shift the burden of production to D. If the case then goes to the jury, the judge will instruct the jury that it may infer that D received the notice timely if it believes the evidence of timely mailing (and, correspondingly, disbelieves D’s opposing evidence (such as D’s testimony)). Note that the jury will not be instructed that it must find D’s timely receipt of notice merely from proof of timely mailing in time to permit timely delivery and thus receipt.

A presumption of correctness is said to attach to the Service’s determinations in a notice of deficiency. The presumption of correctness attaching to the notice of deficiency is a subset of the general presumption of regularity attaching to government actions. Under traditionalist theory, as noted above, a presumption does nothing for a party not having a burden of persuasion or production, for the most a presumption can do is shift a burden of production to the other party. In tax cases, the taxpayer already has the burden of production because the taxpayer has the burden of persuasion; the taxpayer must produce evidence in order to persuade. Consequently, the imposition of the presumption of correctness favoring the Service accomplishes nothing; as one court pungently noted, such a presumption covers with a handkerchief something already covered by a blanket. For that reason, in my mind, the presumption of correctness is meaningless in resolving tax cases when the taxpayer has the burden of persuasion and, as a result, the burden of production; when mentioned, courts and readers should be aware that the presumption qua presumption has no or limited meaning in terms of affecting the outcome of a case when the taxpayer already bears the burden of persuasion.

E. An Example Relating Ranges to Burden of Persuasion

Example 5: Return to Example 2, in which A sues B for malpractice and the evidence shows that As damages are at least $800,000 but not more than $900,000. In this situation, the jury was persuaded to the relevant degree (more likely than not or, its equivalent in this Article, 50+%) that A suffered $800,000 damages and, although the jury believes that the damages could be as high as $900,000, it is not persuaded to the relevant degree as to any amount over $800,000. The evidence of damages can be said to be in a state of equipoise between $800,000 and $900,000. Because A bears the burden of persuasion, the jury should award damages of only $800,000. That was Example 2. Now vary Example 2 by assuming that, under the applicable state law, once A establishes liability for malpractice, B bears the burden of persuasion with respect to damages. With this change in the burden of persuasion, the jury should award damages of $900,000.

For this reason, it is very important as to which party is assigned the burden of persuasion. Equipoise in the range is resolved against the party bearing the burden of persuasion. There really is nothing startling in that proposition other than stating it as a range of equipoise concept.

IV. Burden of Proof in Tax Valuation Cases

A. Introduction and Assumptions

The cases are all over the lot in discussing burden of proof in tax cases and in valuation cases specifically. I will not try to reconcile those cases, for to do so would unduly prolong the agony of my writing and your reading of this Article, and likely would serve no beneficial purpose. Rather, I deal first with the Supreme Court cases that set the parameters for discussion of the issue and then deal with some illustrative cases.

At the outset, let me state my assumptions. First, a trial in a federal tax case is like any other trial. Facts are determined by the trier of fact, and the judge determines the law that applies to the facts. I am not aware of any reason that the burden-of-proof rules applicable in general federal civil trials should not also apply in federal civil tax trials. So, I assume that the discussion above regarding the burden of proof generally in federal civil trials applies in federal civil tax trials unless there is some reason to conclude otherwise.

Second, and flowing from the first assumption, I will attempt to reconcile the canonical statements—those by the Supreme Court—on burden of proof in tax cases with generally recognized burden-of-proof principles under general federal civil procedure. The Supreme Court has made two principal pronouncements on burden of proof in tax cases that we must accept. My bias is to try to reconcile those pronouncements with general federal burden-of-proof concepts. I do not have any authority for that bias, other than a belief that any other bias or even neutrality on the subject would be nonsensical. I believe, in this regard, that general procedure and burden-of-proof principles formed the background with which the Supreme Court was familiar and, therefore, that the Supreme Court would likely not have used familiar and known burden-of-proof concepts to mean something different in tax cases without some explanation.

Finally, in my view, one of the phenomena that inhibits courts’ crisp thinking on the subject of burden of proof in tax cases is that they are almost always bench trials. Trials in the Tax Court where most tax cases are brought are bench trials. Trials in the Court of Federal Claims, where some tax cases are tried, are bench trials. Trials in the district courts, where some tax cases are tried, may be jury trials but are often bench trials or resolved without the involvement of a jury. So, tax trials are generally not jury trials where the classic notions of burden of persuasion, burden of production, and presumptions can be more easily recognized, articulated, and applied in order to separate the functions they serve. All of those concepts still apply in a bench trial; the problem is that the judge is both judge and jury and may merge the concepts or at least not crisply separate them. For the balance of the discussion, unless I specifically state otherwise, I will assume that the trier of fact is the court rather than the jury.

B. The Governing Authority

1. Introduction

The governing authorities—the canon—are two Supreme Court opinions: Helvering v. Taylor and Lewis v. Reynolds. Helvering v. Taylor dealt with trials in the Tax Court (then the Board of Tax Appeals) in a deficiency redetermination proceeding, a prepayment remedy; Lewis v. Reynolds dealt with trials in the district court in a refund suit, a post-payment remedy.

2. Tax Refund Cases

The tax refund suit was the historical method for contesting a tax liability. So I start with refund suits. When a taxpayer brings a tax refund suit against the United States, the taxpayer seeks a judgment against the United States so that the taxpayer can recover on that judgment. Using general civil procedure concepts, the Supreme Court in Lewis v. Reynolds held that, in a refund suit, the taxpayer must prove that the government had collected too much tax and must quantify the amount in excess of the tax that was properly due so that a judgment can be rendered accordingly. There is nothing startling here—simply traditional Anglo-American procedure to which the normal burden-of-proof rules apply. Because one party seeks to recover money from the other, the party seeking recovery is required to show not only that he is entitled to some money but also the amount of money to which he is entitled. Let us use an example to illustrate.

Example 6: An estate reports its estate-tax liability by valuing closely held stock at $50 per share. Upon audit, the Service determines a deficiency based upon a valuation at $100 per share. The estate pays the tax and sues for a refund. The estate proves at trial that the $100 value is excessive. The evidence is thus in equipoise in the range from $50 to $99.99. In this simple example, Lewis v. Reynolds holds that the taxpayer is entitled to a tax refund based on a $99.99 valuation. To the extent that the taxpayer may be entitled to more, the taxpayer has failed to persuade the trier of fact, and therefore, a judgment can be entered against the government only in the amount that the evidence affirmatively persuaded the trier of fact—i.e., a refund based on one cent per share. The doubt—the equipoise—is resolved against the taxpayer because he bore the burden of persuasion as to the amount of the refund and failed to meet it. This example illustrates the fundamental interaction between equipoise and the burden of persuasion. Of course, this example is highly unrealistic because valuation ranges do not function this way in the real world. If the evidence shows that the top end of the range is $99.99, it will surely also show that the bottom end of the range is above, perhaps well above, $50.00. So, we turn to a more realistic example.
Example 7: Same case, except that, based on all the evidence, the court is persuaded that the value cannot be less than $90 or more than $100. The range of equipoise is between $90 and $100. The issue will be resolved based on the burden of persuasion. Remember that the estate has to establish that it is entitled to a refund and has not done so because the evidence is not persuasive as to a value of less than $100. Again, equipoise of the evidence is resolved against the party with the burden of persuasion.

Notice the importance of the range in this example. Because the Service’s determination of a value of $100 per share is at, but still within, the upper end of the range and the evidence is in equipoise with respect to the values within the range, the estate has not shown that the Service’s value is excessive and thus has not shown that it is entitled to a refund.

Example 8: Same case, except that, based on all the evidence, the court first makes a preliminary determination that the value range is at least $70 but not more than $80. The court then finds some rational (and articulable) reason to be persuaded (more likely than not or 50+%) that the actual value is $75. Although the taxpayer bears the burden of persuasion to show the correct amount of the refund, the allocation of the burden does not resolve the case because the evidence is not in equipoise as to value. The court is persuaded as to the correct value which can then be used to calculate the amount of the refund, and judgment is entered accordingly. In this example, although the court initially finds a range, its subsequent determination of a value based on actual persuasion means that the range has been constricted to the point of persuasion.
Example 9: Same case, except that the court is not persuaded as to a point within the $70–$80 range as to the value. The evidence is in equipoise within the range. In this situation, the estate has only established that it is entitled to the refund determined by a value of $80 per share. Again, equipoise of the evidence is resolved against the party with the burden of persuasion.

3. Tax Court Deficiency Cases

a. Background. I now turn to Tax Court litigation and start with a bit of history. The refund suit was the historical method for contesting tax liability. Refund suits require prepayment. Moreover, under trial procedures before the advent of the modern rules of civil procedure in the 1930s, refund suits were fraught with traps for the unwary. With the enactment of the modern income tax, Congress provided an alternative forum in the Board of Tax Appeals (BTA) (since renamed the Tax Court) which would offer a prepayment forum and would provide substantial justice to citizens without untoward traps for the unwary.

In establishing the Tax Court, however, Congress did not address the issue of the burden of proof. Congress soon was on notice that the BTA (now the Tax Court) placed the burden of proof (meaning persuasion) on the taxpayer as virtually a first order of business but without in any way distinguishing a deficiency redetermination from a refund suit vis-a-vis the burden of proof. Further, I doubt that Congress intended to provide the taxpayer with a resolution on the substantive merits of a case in a deficiency redetermination that the taxpayer could not have achieved in a refund suit (setting aside the effect of a jury trial). Congress, I think, viewed the BTA (now the Tax Court) as both a prepayment and a more convenient forum than the district courts, but there is no indication that it intended that the BTA would offer different justice on the merits.

Let us first discuss the Tax Court proceeding in the background of Anglo-American jurisprudence. In a Tax Court proceeding, although the taxpayer is formally the moving party, the Service seeks to obtain a judgment (called a decision in the Tax Court) permitting it to assess and collect the tax from the taxpayer. The taxpayer is only nominally the moving party; the Service started the ball rolling by alleging additional tax in the notice of deficiency, which is the jurisdictional prerequisite to a deficiency redetermination proceeding in the Tax Court. More importantly, the Service seeks to collect an amount from the taxpayer through assessment and collection based on the decision in the Tax Court. Setting aside the nominal roles of the parties in the Tax Court proceeding (the taxpayer as the petitioner and the Service as the respondent), what do you think normal Anglo-American jurisprudence would say about a party in court asking the court to enter a judgment requiring the other party to pay a stated amount to the first party? Normally, the party seeking a court’s judgment in his favor bears the burden of persuasion to provide the court with a basis for awarding the party the relief sought and specifically incorporating an amount of damages in any judgment. This would suggest that, setting aside the nominal roles of the parties in a Tax Court proceeding, the Service should establish that the taxpayer is liable for the tax and the amount thereof. Obviously, this would be a dramatic change from the tax refund suit. I do not think any reading of the history of the legislation establishing the Tax Court would suggest Congress affirmatively intended that the Service bear a burden of persuasion (with the consequent burden of production) in the Tax Court that was different from that in refund suits in the district court.

For policy reasons having to do principally with the protection of the fisc and the taxpayer’s better access to information, the BTA (now the Tax Court) early assigned the burden of proof to the taxpayer. If the taxpayer files a petition in the Tax Court and then does nothing, the Service prevails without further ado. The Tax Court never gets to the issue of whether it was persuaded as to the correctness of the tax asserted in the notice of deficiency because the taxpayer did not meet his burden of production.

The controversy over burden of proof arises when the taxpayer in a Tax Court proceeding does something—introduces some evidence. What then is the importance of the Service’s original determination in the notice of deficiency? This is the context of the decision in Helvering v. Taylor.

b. Helvering v. Taylor. In Helvering v. Taylor, the Service determined a deficiency of $9,156.69 and issued a notice of deficiency accordingly. The alleged deficiency arose from the sale of certain stock the taxpayer acquired in a reorganization that required that his basis in the stock be allocated. The issue was how much of the basis to allocate to the stock sold so that gain could be computed. During the trial before the BTA, the taxpayer introduced evidence tending to show that the Service allocated too little basis, thus overstating the deficiency. The Service introduced no persuasive evidence in support of its basis allocation or any other basis allocation that would justify any deficiency amount. In a summary disposition, the BTA affirmed the Service’s determination in the notice of deficiency by invoking the burden of proof, which, as described by Judge Learned Hand on appeal before the Second Circuit, is “a rubric which has saved the Treasury many a doubtful case, but which can easily be pushed to deny taxpayers privileges plainly theirs.” The Court of Appeals qualitatively assessed the taxpayer’s evidence to establish that too little basis had been allocated and thus that the amount of tax asserted in the deficiency was excessive. The Court of Appeals remanded to give the Service an opportunity to introduce some evidence as to the correct allocation that might justify a deficiency less than the excessive amount originally asserted.

The Service (in the name of Helvering, the Commissioner) petitioned for certiorari, seeking to sustain the deficiency in the excessive amount. The Supreme Court quoted the issue as presented in the petition for certiorari:

Whether the Circuit Court of Appeals erred in remanding this case to the Board of Tax Appeals for a new hearing on the ground that the Commissioner’s determination of the amount of income was incorrect, although the taxpayer had failed to prove facts from which a correct determination could be made.

The Supreme Court further noted that “[t]he only question for consideration is that stated in the petition for the writ of certiorari.”

The Service urged “that in this case the burden on the taxpayer was not only to prove that the commissioner’s determination is erroneous but to show the correct amount of the tax.” Clearly put out with the Service’s attempt to collect a tax in an amount obviously not due through burden-of-proof incantations, the Supreme Court concluded its opinion with the following key language, some of which has become sound bites (or write bites) for courts in the burden-of-proof area:

Unquestionably the burden of proof is on the taxpayer to show that the Commissioner’s determination is invalid. Frequently, if not quite generally, evidence adequate to overthrow the Commissioner’s finding is also sufficient to show the correct amount, if any, that is due. But, where as in this case the taxpayer’s evidence shows the Commissioner’s determination to be arbitrary and excessive, it may not reasonably be held that he is bound to pay a tax that confessedly he does not owe, unless his evidence was sufficient also to establish the correct amount that lawfully might be charged against him. On the facts shown by the taxpayer in this case, the Board should have held the apportionment arbitrary and the Commissioner’s determination invalid. Then, upon appropriate application that further hearing be had, it should have heard evidence to show whether a fair apportionment might be made and, if so, the correct amount of the tax. The rule for which the Commissioner here contends is not consonant with the great remedial purposes of the legislation creating the Board of Tax Appeals. The Circuit Court of Appeals rightly reversed and remanded the case for further proceedings in accordance with its opinion.

The Supreme Court’s holding is wholly unstartling and is totally consistent with Anglo-American jurisprudence generally and with the holding in Lewis v. Reynolds, that a court will not enter a judgment allowing collection or retention of money in an amount shown to be excessive.

Before moving on however, note that the two-stage procedure blessed by the Supreme Court in Helvering v. Taylor is not generally followed. The Court said, in effect, that the Service would prevail at trial without further ado if its notice of deficiency was not found arbitrary or excessive. But, if the Tax Court found that the notice of deficiency was arbitrary or excessive, the Service could apply to have further proceedings to introduce evidence it might then introduce to show the tax due. The Supreme Court noted that usually two-step trials will not be needed because, in most cases, the quality of the evidence will permit the Tax Court to determine the correct amount. Two-step trials are thus rare.

Let us parse the Court’s language briefly. The Court technically dealt only with the issue presented in the petition for certiorari—i.e., whether the Court of Appeals should have remanded for a determination of the correct tax liability after the notice was shown to be excessive. On burden of proof, the Court only said that the statute creating the BTA did not “prescribe any rule of evidence or [rule of procedure].” Courts and authorities have subsequently read into the following statement a holding on burden of persuasion after the notice is shown to be excessive:

On the facts shown by the taxpayer in this case, the Board should have held the apportionment arbitrary and the Commissioner’s determination invalid. Then, upon appropriate application that further hearing be had, it should have heard evidence to show whether a fair apportionment might be made and, if so, the correct amount of the tax.

Technically, all the Court said was that, upon application, there should be a second step to allow for evidence to show whether the correct amount of tax can be determined. The Court did not expressly say that, in that second-step hearing, the Service bears the burden of persuasion. Nevertheless, I think a fair reading of the Court’s action on the issue presented is that, since the notice of deficiency is no longer relevant, the Service bears any risks that might be inherent in the BTA (now Tax Court) not being persuaded as to a correct tax liability so that it could enter a decision accordingly.

There is, nevertheless, one nuance in Helvering v. Taylor’s use of the phrase “arbitrary and excessive” in the context of valuation. I think the arbitrary prong means that a taxpayer can shift the burden to the Service by showing the Service’s determination to be wholly arbitrary even if the taxpayer does not necessarily show that it was excessive. Of course, courts generally will not go behind the notice of deficiency, except in cases involving unsupported allegations of income. So, a showing of arbitrariness of a valuation based on the contents of a notice of deficiency is unlikely to occur.

What can be shown in valuation cases, and is shown regularly, is that the amount of the deficiency based on the Service’s value determinations is excessive. In valuation cases, excessiveness in the deficiency amount is often conceded by the Service before trial by submitting an expert report adopting a valuation that is different from the valuation in the deficiency notice. And, of course, if the court actually finds a value or a range that favors the taxpayer more than the amount or range supporting the notice of deficiency, the deficiency amount has been shown to be excessive. So, the excessive amount branch of the Helvering v. Taylor analysis is the one most often encountered in valuation cases. As discussed below, the Service’s concessions result in taxpayer-favorable adjustments or judicial findings that result in taxpayer-favorable adjustments, although clearly showing that the deficiency amount was excessive, do not invoke Helvering v. Taylor.

At this point, it might be helpful to consider the effect of Helvering v. Taylor when the Service, in a multi-issue Tax Court redetermination case, concedes one or more issues. The concession shows that the aggregate amount in the notice of deficiency is excessive and, depending upon the nature of the position conceded, might even show the aggregate amount to be arbitrary. Does that establish that the portions of the deficiency for other non-conceded determinations were excessive, thus requiring that the Service bear the burden of persuasion and the burden of production on the unconceded issues or amounts? No. Should that answer be different as to a valuation that the Service adjusts in a case before the Tax Court?

Finally, we must distinguish transfer pricing cases from the run of the mill valuation cases to which Helvering v. Taylor applies and to which this Article is addressed. As noted above, transfer pricing involves valuations and is susceptible to the types of ranges of equipoise discussed in this Article; the regulations specifically incorporate these concepts. Transfer pricing is authorized by section 482, which gives the Service broad authority to reallocate income, generally, within parameters allowed by the arm’s-length and commensurate-with-income standards. It is not necessary here to drill deeper into those standards and the relationship between them. But, the broad authority and discretion of section 482 has been held to impose a “heavier than normal burden of proving arbitrariness,” suggesting some differentiation from the Helvering v. Taylor formula. As the Tax Court recently summarized:

When the Commissioner has determined deficiencies pursuant to a section 482 allocation, the taxpayer bears the burden of showing that the allocations are arbitrary, capricious, or unreasonable. If we hold that the adjustments set forth in the notice of deficiency are arbitrary, capricious, or unreasonable, then the taxpayer must next show that the allocations it proposes satisfy the arm’s-length standard.

Thus, contrary to Helvering v. Taylor, a finding that the Service’s valuation determinations in transfer pricing cases are “arbitrary, capricious, or unreasonable” does not appear to impose the burden of persuasion on the Service. I have not attempted to research the potential for discrepancy between Helvering v. Taylor and the transfer pricing cases. I will confine my consideration to run of the mill valuation cases susceptible to analysis under Helvering v. Taylor.

C. Examples to Highlight the Issues

The following examples illustrate the issues raised and develop them in a one-step trial setting. Assume unless stated otherwise that the case is before the Tax Court.

Example 10: Assume the same basic fact pattern as assumed above for refund suits (Examples 6 through 9) but in a Tax Court setting. The taxpayer, an estate, reported its estate tax valuing common stock owned by the decedent at $50 per share. The notice of deficiency is based on a value of $100 per share. The sole issue at trial is the per-share value of the common stock. At the close of trial, the evidence establishes that $100 is excessive but does not establish any other value. Indeed, the evidence is so inconclusive that the Court cannot find that the value is at least $50. Who wins? In this circumstance, Helvering v. Taylor awards the victory to the taxpayer, because the deficiency amount is excessive and there is no basis to enter a deficiency decision for the Service. In the refund suit counterpart, discussed in Example 6, the taxpayer would prevail but only to the extent of the tax on one cent per share. Having said that, as Example 6 explains, it is almost impossible to imagine a real-world case in which this state of equipoise between $0.00 and $99.99 would exist.
Example 11: Same facts as in Example 10, except that, at the close of trial, the quality of the evidence is that the $100 might be excessive, but the quality of the evidence is that the Tax Court is not persuaded (50+%) that $100 is excessive. This could occur when the value range is $90 to $100, provided that the evidence is in equipoise within this range and the court is unable to find a value of less than $100. In this circumstance, Helvering v. Taylor awards the victory to the Service. As discussed above in Example 7, in a refund suit, the government wins in this situation because the finding of a range that includes the value as determined by the Service means that the estate has not established that the value determined by the Service is arbitrary and excessive.
Example 12: Same facts as in Example 11, except as varied herein. At trial, the Service introduced expert testimony valuing the stock at $90, and the taxpayer introduced expert testimony valuing the stock at $50, just as originally reported on the estate-tax return. At that point, the evidence shows that the notice of deficiency is excessive. As is typical, the Tax Court is not persuaded that either party’s asserted valuation is correct and proceeds to slice and dice the expert testimony to arrive at a preliminary conclusion based on the evidence introduced that the valuation range for the stock is from $70 to $80. And, after further analysis, the Court is able to zero in on a definite value point within the range—say $75—which it is persuaded is the value. (In effect, the range of equipoise is eliminated.) In this case, of course, the Court has been persuaded, and the burden of persuasion has not affected the outcome. Helvering v. Taylor is irrelevant in this example. The same result is achieved in a refund suit with the same phenomenon—affirmative persuasion.
Example 13: Same facts as in Example 12, except that the Court cannot determine the sweet spot in the $70 to $80 range—i.e., as to any point in that range—because the evidence is in a state of equipoise. Now the burden of persuasion becomes important. Conceptually, if the estate bears the burden of persuasion, equipoise is resolved against the estate, and the Court should find a value of $80 per share. But, if the Service bears the burden of persuasion, equipoise is resolved against the Service, and the Court should find a value of $70 per share. For the reasons noted below in the discussion of Cavallaro III, the Service’s concession should not shift the burden of persuasion to the Service either under Helvering v. Taylor or as new matter. Consequently, the Court should find a value of $80 per share because the taxpayer has not persuaded the court that $80 is excessive, the same result that is obtained in a refund litigation. But, even if a court were persuaded that the Service’s concession did invoke Helvering v. Taylor, thereby shifting the burden of persuasion to the Service, it could still find a value of $70 at the bottom of the range because it is persuaded.

The key in aligning the result before the Tax Court with that in a refund litigation is, of course, based on an interpretation of Helvering v. Taylor that a concession or even a court finding of a value lower than that in the notice of deficiency does not make the value asserted in the notice of deficiency arbitrary and excessive. I return to this issue in discussing Estate of Mitchell and the Cavallaro cases below.

D. Legislative Grace and Burden of Proof

Courts frequently pen the notion, which the Supreme Court has characterized as a “familiar rule,” that “an income tax deduction is a matter of legislative grace and that the burden of clearly showing the right to the claimed deduction is on the taxpayer.” Does this notion affect the rule in Helvering v. Taylor?

The effect of a deduction was at issue in Helvering v. Taylor—the issue concerned the proper amount of basis for the stock sold. Because basis reduces the amount realized in determining gain, basis has the effect of a deduction. In Helvering v. Taylor, the Service allowed an amount of basis less than that claimed by the taxpayer and issued a notice of deficiency. The taxpayer showed that the Service allowed too little basis and, thus, that the basis asserted in the notice of deficiency was excessive. Does this suggest that there is some inherent tension between the legislative grace rule and Helvering v. Taylor? Perhaps.

Example 14: The taxpayer claims a deduction for property donated to charity. The taxpayer values the property at $100. The Service issues a notice of deficiency based on a value of $50. The taxpayer petitions the Tax Court. At trial, the Service revises the value to $60, with appropriate expert testimony, thus in effect conceding that the deficiency based on the $50 value was excessive. The taxpayer has expert testimony supporting $80 in value. Assume that the Court determines that the value falls between $70 and $78 but cannot on the basis of persuasion determine a definite value within that range. The taxpayer’s deduction should be $70 because the taxpayer has not established a value exceeding $70, entitling the taxpayer to a higher deduction. Helvering v. Taylor is irrelevant because of the taxpayer’s burden to prove deductions.

E. Some Thoughts on Shifting Burden of Persuasion

1. The Context

In my original article, I discussed three cases that Professor Lederman also discussed in her article. For this update and revision, I discuss only one of those cases, Estate of Mitchell v. Commissioner, and omit the two earlier cases since they do not add materially to this updated Article. I also discuss the subsequent Tax Court decision on remand in Estate of Mitchell and two subsequent cases, Kohler v. United States and the Cavallaro line of cases, that raise analogous issues. These decisions deal with the burden of persuasion when the valuation supporting the notice of deficiency is revised, thus making the notice of deficiency excessive.

Admittedly, this discussion is a bit of a detour from the main point of this Article—the use of the ends of the range of values to determine the value based on the burden of persuasion. That argument, if valid, remains true regardless of whether the taxpayer or the Service bears the burden of persuasion. Shifting the burden of persuasion from the taxpayer to the Service simply shifts the risk nonpersuasion and affects only which end of the range is applied. Still, I think the discussion of these cases echoes some of the issues.

2. Estate of Mitchell v. Commissioner

In Estate of Mitchell v. Commissioner, the Tax Court considered an estate-tax valuation dispute over the stock of a closely held corporation based on competing expert testimony introduced at trial. The value asserted by the Service during the trial was significantly less than the value asserted in the notice of deficiency. The stock at issue had been held by a deceased shareholder owning slightly less than 50% of the stock. An unrelated shareholder owned 50% of the stock, and other unrelated shareholders owned the remaining stock. Ignoring the other unrelated shareholders’ 50+% ownership, the Service valued the decedent’s shares as a controlling interest. Curiously, the Service’s own expert initially had not valued the shares as a controlling interest, but for some unexplained reason, the Service directed the expert to do so and issued the notice of deficiency accordingly. Valuing the less than 50% interest in these circumstances as a controlling interest was simply ill-advised—in the language of Helvering v. Taylor, the resulting valuation was arbitrary and excessive. The Tax Court nevertheless held that the taxpayer still bore the burden of persuasion, but concluded that the taxpayer had satisfied the burden of persuasion. Consequently, the assignment of the burden of persuasion was irrelevant as the Tax Court explicitly noted.

On appeal, the Ninth Circuit (1) found that the circumstances of Helvering v. Taylor were present, meaning that the Service bore the burden of persuasion as to an amount justifying a decision in its favor and (2) although recognizing that the assignment of the burden of persuasion was irrelevant if the Tax Court properly determined value on the basis of the evidence, found fault with the Tax Court’s determination of a 35% discount because it felt the Tax Court had not adequately explained the discount. Accordingly, without such an explanation, the Ninth Circuit panel concluded that it could not meaningfully review the valuation determined by the Tax Court and remanded for the Tax Court to explain its decision so that the Ninth Circuit could review it on appeal.

On remand, after analyzing the evidence and stating the basis for its conclusion, the Tax Court reaffirmed its 35% discount previously determined. As an affirmative finding based on persuasion, this time with explanation, the assignment of the burden of persuasion did not matter, but the Court was careful to say that it recognized that the burden of persuasion was on the Service rather than the taxpayer.

3. Kohler v. United States

In Kohler v. United States, a tax refund suit, the parties fought over a valuation issue. The government’s valuation proffered at trial—$19.5 million—was simplistic and clearly excessive, at least according to Judge Posner writing for the panel in his inimitable fashion of bringing pure logic to the task. The taxpayer’s valuation proffered at trial—$11.1 million—was clearly too low. The record offered no persuasive evidence as to a point in between these two erroneous extremes nor did the record indicate whether, if there were a range of equipoise, what that range might have been. In traditional refund suit theory, requiring the taxpayer to show not only that the Service erred but also the amount of the refund to which the taxpayer is entitled, this lacuna should have required a judgment of no refund. Nonetheless, apparently perceiving the government’s erroneous position as more outrageous than the taxpayer’s erroneous position, Judge Posner side-stepped the traditional refund theory by declaring the assessment to be a “naked assessment . . . without any foundation whatsoever.” Where the Service’s valuation is plainly excessive even in a refund suit, the taxpayer has no burden beyond showing that the Service’s claim is excessive. The Service loses. Judge Posner concluded his opinion: “The Service could have justified a more modest estimate yet one well above $11.1 million, but clinging stubbornly to its untenable valuation it suggested no alternative to $19.5 million. It played all or nothing, lost all, so gets nothing.” So, the taxpayer won—even though the taxpayer’s affirmative proof at trial was not persuasive or even credible—simply because the government was more off base than the taxpayer.

But, let us test what Judge Posner said. Let us say that the Service asserted an $18 million valuation in Kohler, with at least some modicum of evidence for that amount. Then, at trial, the trier of fact finds that the taxpayer’s proffered valuation of $11.1 million is too low, that the Service’s proffered valuation of $18 million is too high, that the real valuation is somewhere in between, but that the evidence is so inconclusive that it does not permit the trier of fact to pick the in-between point (or even range) by a preponderance of the evidence. (This is an extreme and highly unlikely example, but assume it for purposes of analysis.) Would or could Judge Posner have applied the naked assessment side-step to shut the Service out? Would not the Service then have prevailed under the standard formulation of the burden of proof in a refund suit—no basis to quantify an amount of refund beyond the amount based on the Service’s reasonable value of $18 million if the evidence does not show that the refund is due? Is not Kohler just a specific adaptation in a litigation context of the adage that “bulls make money, bears make money, pigs get slaughtered?”

Kohler does help in discussing the warp and woof of the burden-of-proof theory in tax cases, but the circumstances will rarely be present in the real world. At a trial on a valuation issue, even if the Service’s original assessment is excessive, the Service is unlikely to rest on an excessive valuation and will propose a reasonable—or at least a somewhat reasonable—valuation in order to maintain credibility before the court. So, in the above example, even if the Service’s original assessment was based on a $19.5 million value, if at trial the Service admitted that the value did not exceed $18 million and offered some basis for $18 million, the court would not reject the Service’s proposed valuation simply because of its admission that the original $19.5 million valuation was excessive. Rather, the $18 million would become the relevant valuation, and the above analysis would apply because, even if wrong, the Service’s position was not arbitrary. In burden-of-proof terms, if $18 million were at the upper end of the range, the taxpayer would bear the risk of nonpersuasion as to any amount less than $18 million. In other words, it appears that the Tax Division’s attorney in Kohler simply botched it and suffered the consequences of irritating Judge Posner in the process.

4. The Cavallaro Saga

a. Introduction. As previously discussed, the key decisions in the Cavallaro saga are Cavallaro I, Cavallaro II, Cavallaro III, and Cavallaro IV. Cavallaro IV is the decision that inspired this updated Article. Because this Article is about valuation, I do not discuss Cavallaro I as it deals with a summons enforcement action. Although the Cavallaro trajectory (II–IV) does not deal specifically with valuation ranges, it does deal with valuation and Helvering v. Taylor in ways that could be applicable in cases involving ranges.

b. Cavallaro II. In Cavallaro II, the Tax Court (Judge Gustafson) held that the parents who owned a corporation that merged with a corporation owned by their children had not received sufficient value in the merged corporation and had thereby made gifts to their children, resulting in a very substantial gift tax. The court also held that the parents had reasonable cause to rely on their tax advisors and thus avoided penalties.

The law is easy to state. If a party (parents here) transfer property to the objects of their bounty (children here) for less than fair consideration, they have transferred value to the children and have thus made a gift to them. The key issue is one of valuation: what was the value of the gift? In Cavallaro II, the Tax Court resolved the value issue on the basis of the burden of proof. In most cases, disputed valuation issues are decided after each side has proffered expert testimony with the Court finding a value somewhere in between. Even when the parties are reasonable in their valuations, the valuations are usually reasonably aggressive with the actual value resting somewhere between. The court found the value the Service claimed because it found that the taxpayers had failed to meet the burden of proof imposed upon them.

The key issue on which the court focused was whether the parents’ corporation or the children’s corporation owned certain valuable technology. The potential for a shift in value from the parents to the children occurred only if the parents’ corporation owned the technology. The parents claimed that, because the children’s corporation owned the technology, the parents did not receive less than they contributed to the merger and, thus, did not make a gift to the children. The parents’ experts assumed that the parent’s corporation did not own the technology. The court found, however, that the parents’ corporation did own the technology, thereby rendering the parents’ expert witness reports irrelevant. All the court then had regarding value was the claim by the Service (which had been reduced from the amount originally asserted in the notice of deficiency), supported by a report from the Service’s expert Mr. Bello. The court found the Bello valuation credible. The court thus felt it had no basis for doing anything other than sustaining the Service’s claim based on the taxpayers’ failure to meet the burden of proof. In sustaining the Service’s position, however, the court did not consider the Cavallaros’ arguments regarding flaws in the Bello report.

The components of the holding with respect to valuation included the following:

(1) The court first held that the taxpayers had the burden of proof even though the Service had substantially reduced its valuation from the amount originally asserted in the notice of deficiency. The court reasoned:

In general, the IRS’s notice of deficiency is presumed correct, “and the petitioner has the burden of proving it to be wrong”. Welch v. Helvering, 290 U.S. 111, 115 (1933); see also Rule 142(a). The Commissioner has conceded that the taxable gifts totaled not $46.1 million (as in the notices of deficiency) but instead $29.6 million (as yielded by Mr. Bello’s analysis). Where the Commissioner has made a partial concession of the determination in the notice of deficiency, the petitioner has the burden to prove that remaining determination wrong.

(2) The court then rejected the argument that the Service’s valuation reduction shifted the burden of proof under Tax Court Rule 142(a)(1) which imposes the burden of proof on the Commissioner “in respect of any new matter.”

(3) The court rejected the argument that the original deficiency notice was arbitrary and without foundation because the Service had “abandoned” the original claim in favor of the reduced one at trial.

(4) In finding that the Cavallaros had failed to meet their burden of proof, the court held:

Petitioners’ criticisms might have greater significance on the next sub-issue—i.e., determining what portion of that value is properly attributable to each of the two companies—but we need not resolve those criticisms or attempt to correct the Commissioner’s figures. It is the Cavallaros who have the burden of proof to show the proper amount of their tax liability, and neither of the expert valuations they provided comports with our fundamental finding that Knight owned the valuable CAM/A LOT technology before its merger with Camelot. We are thus left with the Commissioner’s concession, effectively unrebutted by the party with the burden of proof. The Cavallaros risked their cases on the proposition that Camelot had owned the CAM/A LOT technology (and on a valuation that assumed that proposition), but they failed to prove that proposition (and the evidence showed it to be false). That being so, it would serve no useful purpose to review our agreement or disagreement with each and every aspect of the experts’ opinions. We conclude that Mr. and Mrs. Cavallaro made gifts totaling $29.6 million on December 31, 1995.

(5) The original valuation used in computing the deficiency in the notice of deficiency ($46.1 million) was significantly reduced by the Service’s concession to $29.6 million. That figure represented a 35.7% ($16.5 million) reduction in value. Nevertheless, that reduction did not mean that the notice of deficiency was arbitrary and excessive (as confirmed by the Court of Appeals in Cavallaro III).

c. Cavallaro III. In Cavallaro III, the First Circuit addressed the Cavallaros’ argument that the Tax Court erred in not shifting the burden of proof to the Service. The court started by stating the background proposition that “[a] rebuttable presumption of correctness cloaks an IRS notice of deficiency,” thereby imposing on the taxpayer “the burden of proving by a preponderance of the evidence that the Commissioners tax assessment is erroneous.” The court then addressed the Cavallaros’ argument that the Service’s deficiency notice was arbitrary and excessive:

Burden-shifting for an excessive-and-arbitrary deficiency notice is a fairly narrow doctrine. See United States v. Janis, 428 U.S. 433, 441–42 (1976). It involves “a challenge to the deficiency assessment itself on the basis that it bears no factual relationship to the taxpayer’s liability, not a challenge to any proof offered by the Commissioner at trial before the Tax Court.” Where an assessment is shown to be “naked” or utterly without foundation, we remand the case for further action to determine the amount that might lawfully be taxed. See Janis, 428 U.S. at 442 . . . . In this limited circumstance, the presumption of correctness is overcome, and the burden shifts to the Commissioner.

The threshold question, then, is whether the Cavallaros have carried their burden of producing evidence from which it can be concluded that their deficiency assessments utterly lacked rational foundation. The Cavallaros’ challenge falls short of the mark. . . .

Without more, however, the fact that the Commissioner later conceded a portion of the original deficiency does not compel a conclusion that the initial assessments lacked a rational foundation.

The First Circuit distinguished Estate of Mitchell as a case in which the erroneous valuation report supporting the notice of deficiency during the examination was arbitrary and excessive.

In Estate of Mitchell v. [Commissioner], the estate tax deficiency notice rested on a stock valuation that the appraiser had altered according to the Commissioner’s instructions and that the Service’s expert disavowed. The circumstances in the instant case are not analogous.

Thus, a mere reduction in value—and perhaps even a substantial reduction in value as in Cavallaro II—does not alone establish that a Service valuation supporting a notice of deficiency is arbitrary or excessive within the meaning of Helvering v. Taylor.

The court then addressed the taxpayer’s argument that the burden shifted because the Service’s new, albeit reduced, valuation was a “new matter.” The court rejected this argument:

The Commissioner’s subsequent adoption of the Bello report was simply a refinement of that original theory (i.e., a clarification of the extent to which Knight was undervalued). We have previously said that “if a deficiency notice is broadly worded and the Commissioner later advances a theory not inconsistent with that language, the theory does not constitute new matter, and the burden of proof remains with the taxpayer.” The original deficiency notices were more than adequate to put the Cavallaros on notice that the Commissioner was challenging the value of Knight as transferred within the merger. . . . Indeed, the Cavallaros’ Petition for Re-determination makes it clear that they were aware all along that the value of Knight, to the extent that it exceeded the value of the stock they received at the time of the merger, would be integral to determining their tax liability.

For these reasons, the First Circuit held that the burden of proof did not shift to the Service.

Regarding the Tax Court’s determination of the value of the gift, the First Circuit rejected the Cavallaros’ argument that the critical component of value, the valuable technology, was owned by the children’s corporation rather than the parents’ corporation. The court then held that, although the Cavallaros’ expert valuations were fatally flawed because they failed to account for the valuable technology ownership by the parents’ corporation rather than the children’s, the Tax Court had erred in applying the Helvering v. Taylor analysis by declining to consider the Cavallaros’ attack on the Service’s expert report:

The Cavallaros should have had the opportunity to rebut the Bello report and to show that the Commissioner’s assessment was “arbitrary and excessive.” If they succeeded in doing so, the Tax Court should have then determined for itself the correct amount of tax liability rather than simply adopting the Commissioner’s position. See Taylor, 293 U.S. at 515–16 (stating that upon determining the Commissioner’s valuation to be arbitrary, the Board of Tax Appeals should have conducted a “further hearing” in which it “heard evidence to show whether a fair apportionment might be made and, if so, the correct amount of the tax”) . . . .

In accordance with those cases, we remand so that the Tax Court can evaluate the Cavallaros’ arguments that the Bello valuation had methodological flaws that made it arbitrary and excessive. If the Tax Court determines that the Commissioner’s assessment was arbitrary, then it must determine the proper amount of tax liability for itself. “The court need not, in making this determination, be able to precisely establish the correct figures; reasonable approximations may be employed, provided the findings disclose the method used in calculating the deficiency.” Miller v. United States, 296 F.2d 457, 460 (7th Cir. 1961). The court is free to accept in whole or in part, or reject entirely, the expert opinions presented by the parties on the subject. . . . Further, the court may take new evidence, including a new expert valuation.

In other words, based on the record, the Cavallaros should have been allowed to show that the Bello valuation was flawed in some respects. The Tax Court just did not consider the Cavallaros’ arguments and, thus, had not considered whether the Bello valuation (which was the new target for the Helvering v. Taylor analysis) was itself arbitrary and excessive, so that adjustments could be made accordingly.

d. Cavallaro IV. The Tax Court (Judge Gustafson) summarized its job on remand in the following manner: “The question before us on this remand is whether the Commissioner’s expert’s valuation is ‘arbitrary and excessive.’ If it is, then we are tasked with determining the proper amounts of the Cavallaros’ tax liabilities.” As articulated, the scope of the remand was fairly limited. For its factual consideration, the Court assumed familiarity with the facts from Cavallaro II and III.

Addressing value, the court focused on the Service’s expert valuation, the Bello valuation, that the Tax Court had accepted in Cavallaro II in order to consider, as directed by the First Circuit, the Cavallaros’ claims of error in the Bello valuation. The court first rejected the Cavallaros’ argument that the Bello report should be rejected in its entirety because Mr. Bello was biased. The Court found Mr. Bello credible, with any errors the result of mistakes and not bias.

Moving to the Cavallaros’ claims of specific errors in the implementation of Mr. Bello’s valuation methodology, the court rejected all of the claims except one, the so-called “90th percentile profit margin calculation.” A detailed review of the error asserted is not important for this discussion, but in general the court found that Mr. Bello erred in certain assumptions and that, when those assumptions were corrected and revised calculations made pursuant to his methodology, the value of the disguised gift should have been reduced by $6.9 million, resulting in total gifts from the parents of $22.8 million. The court rejected the Cavallaros’ argument that the error was “deceptive” and rendered the balance of the valuation methodology suspect. According to the court, the valuation, as adjusted, was persuasive, without any alternative more persuasive as to a lesser value. The court summarized:

This one error does not make us unable to use Mr. Bello’s valuation, and it does not require a new trial or necessitate receiving additional evidence. Rather, we “accept . . . in part . . . the expert opinion [] presented by . . . [the Commissioner] on the subject.” Cavallaro III, 842 F.3d at 27. On the basis of the trial record, the Cavallaros’ identification of the error and proposal of a correction, and the Commissioner’s acceptance of the Cavallaros’ calculation, we are able to correct for this error and determine the proper amount of the Cavallaros’ gift. The parties agree that if Mr. Bello had used the correct 90th percentile figure, 9.66% rather than the 7.5% incorrect value, the value of the disguised gifts would be reduced from approximately $29.7 million to $22.8 million, a difference of about $6.9 million. We conclude that the $22.8 million value is the correct value of the disguised gifts made by the Cavallaros to their sons.

The final determined value of $22.8 million was less than 50% of the original valuation in the notice of deficiency. And yet that drastic reduction, most of which resulted from the Service’s concession in Cavallaro II, did not mean that the original valuation and resulting notice of deficiency was arbitrary and excessive for purposes of Helvering v. Taylor.

e. Summary of the Cavallaro Saga. One issue discussed in the Cavallaro decisions is whether a reduction in value from that in the notice of deficiency by the Service at trial invokes the Helvering v. Taylor requirement that the Service prove the amount of the deficiency. There are statements in decisions that can be read both ways on that issue, but I think that the better position is that in Cavallaro III which held that, barring some egregious circumstances other than mere reduction at trial, the original valuation supporting the notice of deficiency is not arbitrary and excessive within the meaning of Helvering v. Taylor.

Regardless of which party bears the burden of persuasion, the analysis that I argue for here—that the party with the burden of persuasion must bear the risk in the range of equipoise—is still correct. Persuasion is persuasion and not some guesstimate beyond the point of persuasion. A shift of the burden of persuasion to the Service would simply shift the risk in the range of equipoise from the taxpayer to the Service. But that is simply to say that, in using ranges and the burden of persuasion, the Service is subject to the same rules as is the taxpayer.

V. Conclusion

Understanding the interplay of the burden-of-proof rules and ranges in valuations can help the parties and the courts focus more crisply on litigation needs. The parties at trial should pay more attention to ranges and, unless an obvious persuasive definite valuation presents itself (not often, at least initially), conform their trial strategies to achieve the best ranges. A court must carefully analyze the evidence to determine the appropriate range and to constrict the range to the extent the evidence permits. Careful analysis may not only identify a preliminary appropriate range but also further assist the court in finding the “sweet spot” with the requisite degree of persuasion. Such an approach will permit the trial court to articulate the basis for its finding so that an appeals court can meaningfully review its decision. And, if the trial court’s decision is overturned because of a misallocation of the burden of proof, a finding of value in terms of a range will provide the result on appeal without the need for a remand.

    Author