Thoughts on Estate of Elkins and Valuations (10/21/14)

I write today on the Fifth Circuit's opinion in Estate of Elkins v. Commissioner, ___ F.3d ___, 2014 U.S. App. LEXIS 17882 (5th Cir. 2014), here.  In that case, the Fifth Circuit reversed the Tax Court in a valuation case involving discounts for fractional interests in valuable art.  The Fifth Circuit opens with a summary of its reasoning, such as it is (footnote omitted):

In the Tax Court, the Commissioner steadfastly maintained that absolutely no fractional-ownership discount was allowable. This presumably accounts for his failure to adduce any affirmative evidence—either factual or expert opinion—as to the quantum of such discounts in the event they were found applicable by the court.
The Tax Court rejected the Commissioner's zero-discount position, but also rejected the quantums of the various fractional-ownership discounts adduced by the Estate through the reports, exhibits, and testimony of its three expert witnesses—the only substantive evidence of discount quantum presented to the court.1 Instead, the Tax Court concluded that a "nominal" fractional-ownership discount of 10 percent should apply across the board to Decedent's ratable share of the stipulated FMV of each of the works of art; this despite the absence of any record evidence whatsoever on which to base the quantum of its self-labeled nominal discount.
We agree in large part with the Tax Court's underlying analysis and discrete factual determinations, including its rejection of the Commissioner's zero-discount position (which holding we affirm). We disagree, however, with the ultimate step in the court's analysis that led it not only to reject the quantums of the Estate's proffered fractional-ownership discounts but also to adopt and apply one of its own without any supporting evidence. We therefore affirm in part, reverse in part, and render judgment in favor of Petitioners, holding that the taxable values of Decedent's fractional interests in the works of art are the net amounts reflected for each on Exhibit B of the Tax Court's opinion. This, in turn, produces an aggregate refund owed to the Estate of $14,359,508.21, plus statutory interest.

Just a few paragraphs down, the Fifth Circuit continues:

This entire appeal thus begins and ends with the question of the taxable value of Decedent's fractional interests in those 64 items of non-business, tangible, personal property that were jointly owned in varying percentages by Decedent and his three adult children at the instant of his death. And, the answer to that one question begins and ends with the proper administration of the ubiquitous willing buyer/willing seller test for fair market value: "Fair market value is defined as '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 sell and both having reasonable knowledge of relevant facts.'"

The context for the Fifth Circuit’s opinion is the Tax Court’s opinion below.   In Estate of Elkins v. Commissioner, 140 T.C. No. 5, 2013 U.S. Tax Ct. LEXIS 6 (T.C. 2013), here, the issue was the familiar one of the appropriate discounts for fractional interests.  The IRS generally disfavors fractional interests, due in no small part to taxpayers’ frequent – perhaps even common – use of aggressive discounts which will either prevail because they win the audit lottery or, if caught, will be recognized as improperly inflated and reduced accordingly.  (Most practitioners would say that it is entirely proper to assert aggressive discounts -- but not so aggressive that serious penalties would apply in the full expectation that, if contested, there will likely be some adjustment; that is the way the game is played.) Essentially, as I read the opinion, the Tax Court judge, Judge Halperin, found the estate's proffered too aggressive and found an alternative discount.  The estate, of course, had "experts" to testify as to the discounts.  The IRS essentially had no "experts" to testify that no discount or any discount less than testified by the estate's experts was appropriate.  So, on the record presented, Judge Halperin found that the proper discount was 10%, substantially below the discounts claimed by the estate.  He based that on the entire record before him.


Among the facts that the Tax Court considered was that, in the initial meetings with the members of the Elkins family, the expert was "convinced that any buyer of decedent's interests in the art would have to take into account the fact that the children (whom he refers to as 'the other shareholders') are 'committed to retaining the art in the family until the last shareholder dies.'"  In effect, although the test of fair market value is the hypothetical willing buyer-willing seller test, that test is not applied in a vacuum, untethered to any natural market for the property being valued.  The Tax Court found that there was such a natural market in the Elkins family (the owners of other fractional interests) who valued the art and would be likely to provide a market close to the actual pro rata fair market value for the art, with a much smaller discount than if that market and the family dynamics were not considered.  In the final appraisals, the Tax Court found, the estate’s experts did not properly consider the family dynamics that would make them the natural market that the hypothetical buyer would exploit.

Now, going back to the Fifth Circuit's opinion, there are several issues I want to discuss.

1. The Fifth Circuit held that Section 7491, here, imposed the burden of proof on the IRS and that the IRS failed to meet that burden because it had not introduced evidentiary proof in support of the findings of the Tax Court.  The Fifth Circuit reasoned that, since the taxpayer introduced expert evidence of a discount and the IRS  introduced none, Section 7491 gives the victory to the taxpayer in the full amount claimed.  The court said "Under a proper administration of § 7491's burden of proof rule, this case should have ended at that point with a judgment for the Estate."  The Court repeated this theme later:

The Tax Court failed to require the Commissioner to bear that burden of proof, even though 26 U.S.C. § 7491 mandates that when, as here, the petitioning taxpayer adduces sufficient evidence to establish the material facts—in this case, the amounts of the discounts—the Commissioner has the burden of refuting such facts and proving different ones. Yet he chose not to adduce any evidence of discount quantum whatsoever, sticking instead to his no-discount position. 

* * * * 

In most trials, "[a] determination of where the preponderance lies requires a measuring and weighing of all the evidence, pro and con." But, when, as here, the only evidence on an issue is that presented by but one party—and by the one that did not have the burden of proof, at that—there is no "preponderance": It takes two to tango. As with its misapplication of the burden of proof, however, the Tax Court's error in announcing its use of the preponderance standard to determine the amounts of the discounts ultimately makes no difference. This is because, having put all of his eggs in the one, no-discount basket at trial, the Commissioner cannot be heard on appeal to question the quantity, quality, or sufficiency of the evidence adduced by the Estate to prove the quantum of the fractional-ownership discounts to be applied. Likewise, given the total absence of substantive evidence from the Commissioner on the issue of quantum, the Tax Court should have accepted and applied the uncontradicted quantums of the partial-ownership discounts that the Estate proved with much more than substantial evidence.

To see the error in that simplistic analysis, it is necessary to be clear about the burden of proof in in issue and its meaning.  The burden of proof is the burden of persuasion, better described as the risk of nonpersuasion.  See  9 John Henry Wigmore, Evidence In Trials At Common Law § 2485 (1981), cited in John A. Lynch, Jr., Burden of Proof in Tax Litigation under I.R.C. § 7491 - Chicken Little was Wrong, 5 Pitt. Tax Rev. 1, 16 n. 112 (1999); and Steve R. Johnson, The Dangers of Symbolic Legislation: Preceptions and Realities of the New Burden of Proof Rules, 84 Iowa L. Rev. 413 (1988).  All Section 7491(a) does it to give the party not bearing the risk of nonpersuasion when the trier of fact -- here the Tax Court (not the Court of Appeals) -- is not persuaded on the basis of all the evidence.  The Tax Court was persuaded, so Section 7491(a) had no application.

Moreover, another related point apparently missed by the Court of Appeals is that the burden of proof -- the risk of nonpersuasion -- does not shift an affirmative evidentiary duty to the IRS to proffer evidence supporting the finding required; rather, it merely requires that the bearer of the burden bear the risk of nonpersuasion – i.e., bear the risk that the evidence in the record will not persuade the trier of fact.  It does not matter how the evidence gets into the record – the evidence can get there entirely through the plaintiff’s exhibits or witnesses (whether on direct or cross examination or by questioning from the trial judge).  The only issue is whether the evidence supports the trier of fact’s finding of persuasion.

2.  I think it is helpful to reiterate precisely what the Tax Court found.  First, the Tax Court was persuaded that the taxpayer's claimed discounts were wrong.  Second, it found that the correct discount was 10%.  The first finding is subsumed in the latter, but is in no way dependent upon the latter.  Specifically, a finding that a claimed discount is too aggressive simply because a Court of Appeals finds that a finding of a lesser discount is clearly erroneous.  But the Court of Appeals dealt only with the latter.  The Tax Court's holding was not an either or binary choice -- either the taxpayer’s proffered discounts are right or 10% is right.  It found the taxpayer's discount wrong and, even if one were to quibble (as the Fifth Circuit did) with the second holding -- 10% -- that does not mean that, on the evidence presented, the only persuasive discount would be the estate's claimed discounts.

3.  Bottom line, even if on the record, the Fifth Circuit could conclude that the 10% discount was clearly erroneous, the Fifth Circuit does not deal at all with the first key finding that the estate’s proffered discount was not persuasive full bore.  In other words, I think the proper resolution would have been to remand the case to the Tax Court to reconsider on the record presented with the the risk of nonpersuasion assigned to the IRS but eliminating any erroneous consideration identified by the Fifth Circuit.  Then, if without that consideration, there were no persuasive evidence as to any discount other than claimed by the estate, the estate would win.  But, if on the records, there is persuasive some discount between 10% and the discount proffered by the estate, that persuasive discount should prevail.

4.  The key is that valuations and the discounts that underlie many valuations are really ranges. John A. Townsend, Burden of Proof in Tax Cases: Valuations and Ranges, 2001 TNT 187-37 (2001).  In a simple fact situation where the key fact is X, the spectrum will at one end persuasively establish Fact X's existence and at the other persuasively establish that Fact X does not exist.  Conceptually, there is a point somewhere in between where the evidence is not persuasive -- said by burden of proof enthusiasts to be in a state of equipoise.  Where the evidence is not persuasive as to the existence or nonexistence of fact X, then the party bearing the burden of persuasion (risk of nonpersuasion) loses. Now, putting this concept in the case of valuations and ranges, say that the parties have the following positions in a case where the stipulated true value of the 100% interest is $200: the taxpayer claims a value of 100 because of discounts and the IRS claims a value of 200 with no discounts.  Consider these alternatives none of which consider how evidence gets into the trial record (burden of persuasion principles do not consider how evidence gets into the record but only risk of nonpersuasion is the record does not contain persuasive evidence):

a.  If the evidence is persuasive that the taxpayer's claimed discounts and resulting value of $100 is correct, the taxpayer wins altogether.

b.  If the evidence is persuasive that the IRS"s claimed zero discount and resulting value of $200 is correct, the IRS wins altogether.

c.  But, what if, as is likely because both parties took aggressive positions, neither party's proffered value is correct and the correct value is somewhere in between.  In that state of evidence, is the trier of fact required to accept either party's proffered value?  No, and it should not.  To illustrate why, conceptualize the spectrum between 100 and 200 as the follows:  the trier is persuaded that the value is from 145 to 155 but in that range the trier is not persuaded as to any particular value.  That means, correspondingly, that the values immediately outside the range of nonpersuasion -- 144 and 156 -- are values that are persuasive.  In that case, the trier should find 154 if the IRS bears the risk of nonpersuasion or 156 if the taxpayer bears the burden of persuasion.  Why? because the finding that the range is 145 to 155 does not permit a preponderance, then the trier must find the value that does preponderate just outside that range of nonpersuasion.  See John A. Townsend, Burden of Proof in Tax Cases: Valuations and Ranges, 2001 TNT 187-37 (2001).   And, correspondingly, neither party is entitled to prevail on its aggressive claims of 100 and 200 regardless of which party bears the burden of persuasion.

5.  Another curious holding by the Court when the IRS attempted to support the 10% discount based on the record was the following:

   n17 For the first time on appeal, the Commissioner attempts to address the quantum of the fractional-ownership discounts as supporting the Tax Court's 10 percent discount. It is well settled, however, that we do not consider contentions raised for the first time on appeal. Crawford Prof'l Drugs, Inc. v. CVS Caremark Corp., 748 F.3d 249, 267 (5th Cir. 2014).

In effect, the Court of Appeals is holding that, because the IRS did not argue for some discounts and the 10% in particular below, the IRS may not, on appeal, support the discount affirmatively found by the trier of fact.  Stated differently, to show its absurdity, the Court is saying that had the trier of fact properly determined on the basis of the record that the fair market value was $X rather $X-50%, it is compelled as a matter of law to ignore the factual truth it found because the IRS did not argue for it below.  While it is generally true that issues not raised on presented below are not properly on appeal, the discount issue was raised below and was expressly decided by the Tax Court judge.  And the Tax Court judge did what judges often usually do in valuation cases where the norm is for each side to be aggressive – he came somewhere in between the aggressive discounts claimed by the taxpayer and the equally aggressive $0 discount claimed by the IRS.

6.  Looking at the substantive evidence of proper discount, the Fifth Circuit simply reinterprets what it calls “uncontradicted testimony” of a key witness.  With nothing more than the spare transcript, the Fifth Circuit interprets her testimony as being uncontradicted in support of the claimed valuations.  The Tax Court who heard her testimony live (not limited by a transcript) interpreted her words to mean something different than the Fifth Circuit, but certainly within a fair interpretation of the words the witness used – i.e., that the family owning other fractional interests had a keen interest both in the artwork itself and in avoiding the disruption of nonfamilial owners and that that keen interest would bear on the quantum of discounts under the hypothetical willing buyer - willing seller standard.  Thus, the Fifth Circuit improperly invoked behind the common statement that uncontradicted evidence may in some circumstances control the day.  (Even that notion is simply too broad, but moving past that here.)  That witness’ testimony was not contradicted.  The Tax Court did not find that she was not a truth teller; rather the Tax Court took her testimony at face and interpreted it consistent with the record before it.  That is uniquely the function of the trier of fact, not the Court of Appeals.

For other discussions of the case:

Paul Sullivan, A Potential Game Changer for Estate Taxes on Art (NYT 10/3/14), here.

JAT Note: The bottom line, of course, is that the rich really are different.  As the Ninth Circuit recently noted in Hawkins III v. Franchise Tax Board, ___ F.3d ___, 2014 U.S. App. LEXIS 17925 (9th Cir. 2014):

F. Scott Fitzgerald observed early in his career that the very rich "are different from you and me," n1 to which Ernest Hemingway later rejoined, "Yes, they have more money." n2
   n1 F. SCOTT FITZGERALD, The Rich Boy, in The Short Stories of F. Scott Fitzgerald: A New Collection 317 (Matthew J. Bruccoli ed., Scribner 1989) (1926).
   n2 ERNEST HEMINGWAY, The Snows of Kilimanjaro, in THE SNOWS OF KILIMANJARO AND OTHER STORIES 23 (Scribner 1961) (1936). (Hemingway, quoting the critic Mary Colum without attribution, used Fitzgerald's name in the original magazine version of the short story, but altered the name to "Julian" in the later published book. See Eddy Dow, Letter to the Editor, The Rich Are Different, N.Y. TIMES, Nov. 13, 1988, available at http://www.nytimes.com/1988/11/13/books/l-the-rich-are-different-907188.html.)