Estate of Burton W. Kanter, Deceased, Joshua S. Kanter, Executor, and Naomi Kanter, Petitioners-appellants, v. Commissioner of Internal Revenue, Respondent-appellee, 337 F.3d 833 (7th Cir. 2003)

The Estate of Burton Kanter and Naomi Kanter appeal a decision of the Tax Court. This consolidated appeal deals with six out of forty-one separate issues decided by the Tax Court with respect to alleged deficiencies of the late Burton W. Kanter, his wife, Naomi Kanter, and related entities, as well as two additional post-trial issues. We affirm in part and reverse in part.

INTRODUCTION

The late Burton W. Kanter,¹  and various family entities associated with him, have been audited by the Internal Revenue Service virtually, if not literally, every year since Richard Nixon was President. Kanter was a well-known and accomplished tax and estate lawyer. He graduated from the University of Chicago Law School. He had a very successful law practice beginning in 1956, founding what would eventually become the law firm of Neal, Gerber & Eisenberg. Among Kanter's clients was the Pritzker family of Hyatt Corporation fame. Kanter was also an accomplished businessman, with "extensive exposure" to a "good many public.... [and] private companies." (Tr. at 5278.)²  Kanter wrote extensively on tax-related subjects (originating a "Shop Talk" column in the Journal of Taxation), and was an expert on the subject of trusts and estate planning. See, e.g., Burton W. Kanter & Michael J. Legamaro, The Grantor Trust: Handmaiden to the IRS and Servant to the Taxpayer, 75 TAXES 706 (1997); Sheldon I. Banoff & Burton W. Kanter, LLC Announcements: Damage Control, 80 J. TAX'N 255 (1994); Burton W. Kanter & Sheldon I. Banoff, Tax Planning for the Elderly, 70 J. TAX'N 191 (1989); Burton W. Kanter, AARP— Asset Accumulation, Retention and Protection: Prelude to Transmission, 69 TAXES 717 (1991); Burton W. Kanter, Cash in a "B" Reorganization: Effect of Cash Purchases on "Creeping" Reorganization, 19 TAX L. REV. 441 (1964). In the 1960s and 1970s, Kanter helped Hollywood finance movies through tax shelter arrangements, and was involved in the production of many major Hollywood films, including "One Flew Over the Cuckoo's Nest." The IRS's extraordinary attention to Kanter is understandable given that from 1979 to 1989 Kanter, the highly successful tax attorney, who hobnobbed with Pritzkers and Hollywood producers and who participated in countless extremely large and lucrative business ventures, reported a negative adjusted gross income each year on his federal tax return and paid no federal income taxes. (Tr. at 5290-91.)

This consolidated appeal involves Kanter's petitions for review of deficiencies assessed during the years from 1978 to 1986, which is itself only a portion of the original consolidated case tried by the Tax Court in 1994—a trial that generated almost 5500 pages of transcript, more than 4600 pages of briefs and thousands of exhibits consuming hundreds of thousands of pages, and was eventually, five years later, distilled into a 606 page opinion covering forty-one separate issues. A thorough description of the entire factual background to this case can be found in the Tax Court's opinion. Investment Research Associates, Ltd. v. Comm'r, 78 T.C.M. (CCH) 951 (1999) [hereinafter IRA]. The trial was conducted by Special Trial Judge Couvillion, to whom the Tax Court had assigned the case under 26 U.S.C. § 7443A(b) (4). See also Tax Court Rule 180.³  Under the Tax Court's rules, the Special Trial Judge (STJ) then submitted a report containing findings of fact and opinion to the Tax Court's Chief Judge, who then assigned the case to Tax Court Judge Dawson. See Tax Court Rule 183(b). Judge Dawson subsequently issued his opinion, which stated that the Tax Court "agrees with and adopts the opinion of the Special Trial Judge, which is set forth below." IRA, 78 T.C.M. (CCH) at 963. Of the forty-one issues decided by the Tax Court, six were appealed to this court:⁴ 

1. Fraud: The Tax Court determined that Kanter (and two colleagues) helped individuals obtain business opportunities in exchange for payments that later were fraudulently diverted through a series of Kanter-controlled entities in order to disguise the payments' origins and lower the tax assessed on the income (by dividing it up and assigning parts of it to various entities claiming losses). Kanter concedes that there was an underpayment of taxes but disputes that the Commissioner was able to prove by clear and convincing evidence that the underpayment was due to fraud.

2. Bea Ritch Trusts: Kanter challenges the Tax Court's determination that capital gains reported in 1986 by the Bea Ritch Trusts (BRT) were properly taxable to Kanter under the grantor trust provisions of the Internal Revenue Code (IRC).

3. Washington Painting: Kanter challenges the Tax Court's refusal to allow him to deduct expenses he incurred during an aborted sale of a painting.

4. 1982 Bank Deposits: Kanter challenges the Tax Court's determination of a deficiency for the year 1982 based upon an analysis of his bank deposits. He argues that the Commissioner failed to meet his burden to prove a deficiency, that the Tax Court erred in presuming the Commissioner's deficiency determination to be correct, and that in any event the evidence Kanter presented at trial was sufficient to overcome any presumption of correctness.

5. Equitable Leasing: Kanter challenges the Tax Court's determination that payments from Equitable Leasing Company to Kanter entities were taxable commissions and not loans.

6. Cashmere: The Tax Court disregarded a series of transactions involving (a) the contribution of certain partnership interests to a shelf corporation (Cashmere) and (b) the subsequent installment sale of Cashmere's stock, the result of which was an immediate recognition of capital gain to Kanter on the partnership interests. Kanter argues that the transactions had economic substance and should not have been disregarded as an attempt to avoid the payment of federal income tax.

The remaining two issues in this appeal concern events after the conclusion of the Tax Court's trial. Beginning in April 2000, Kanter sought repeatedly to have the original report filed by the STJ placed in the record, or in the alternative made available for this Court's review in camera. Kanter alleged that informal conversations with two Tax Court judges had revealed that the issued opinion had undergone significant alterations from the original report filed by STJ Couvillion. The Tax Court denied all of Kanter's motions. Kanter appeals the Tax Court's refusal to produce the STJ's original report. The final issue concerns Kanter's wife's (Naomi's) efforts to seek innocent-spouse relief from the deficiencies levied against her husband's estate.

To make this very complex case easier to understand we have placed a brief statement of facts relevant to each issue immediately preceding that issue's analysis. We begin with the broadest (and least fact dependent) issue: whether the STJ's original report should have been made part of the record on appeal. Then we address the six transactional issues from the Tax Court's decision. Finally, we address Naomi Kanter's post-trial motions.

The United States Courts of Appeals have exclusive jurisdiction to review decisions of the United States Tax Court. 26 U.S.C. § 7482(a) (1); Seggerman Farms, Inc. v. Comm'r, 308 F.3d 803, 805 (7th Cir. 2002).

Kanter's first argument is that the STJ's original report must be made a part of the record on appeal so that this court can determine whether the appropriate degree of deference had been paid to it by the Tax Court judge, whose opinion is before us. Kanter claims that informal conversations between his attorney and other Tax Court judges revealed that the STJ who presided over the trial of this case submitted a report that found Kanter credible and recommended rejection of much of the Commissioner's assessed deficiencies, specifically the fraud deficiency. Kanter argues that the STJ's report cannot be rejected by the Tax Court unless clearly erroneous, and that, without the STJ's report in the record, there is no way for this court to determine if proper deference was accorded it. Moreover, this secret and unaccountable process of review allegedly violates Kanter's due process rights. Kanter, relying on a Supreme Court case examining the relationship between U.S. district court judges and magistrate judges, argues that this quasi-collaborative process affords a Tax Court the opportunity to reverse an STJ's credibility findings without first hearing or seeing the witnesses itself—thus offending due process. See United States v. Raddatz, 447 U.S. 667, 681 n. 7, 100 S. Ct. 2406, 65 L. Ed. 2d 424 (1980) (observing in dicta that in the criminal context a district court judge's reversal of a magistrate judge's credibility findings without the district judge hearing or seeing the witnesses would raise "serious questions"). Kanter argues that in addition, our review of the Tax Court's decision is unconstitutionally impaired by the omission of the STJ's report from the record. Kanter's challenge of the Tax Court's refusal to include the "original" STJ report presents questions of law that we review de novo. Pittman v. Comm'r, 100 F.3d 1308, 1312 (7th Cir. 1996).

Of course, Kanter's arguments are immaterial if the Tax Court's final opinion is the STJ's report. See Ballard, 321 F.3d at 1042-43. The Tax Court's final opinion clearly states that it "agrees with and adopts the opinion of the Special Trial Judge." IRA, 78 T.C.M. (CCH) at 963. The Chief Judge of the Tax Court, Judge Dawson, and Special Trial Judge Couvillion himself all signed that final opinion, and we take their statement at face value. Therefore, notwithstanding Kanter's attorney's declaration, we accept as true the Tax Court's statement that the underlying report adopted by the Tax Court was in fact Special Trial Judge Couvillion's. See Ballard, 321 F.3d at 1042-43. This renders moot all of Kanter's arguments.

But even if, as the dissent suggests, the phrase "agrees with and adopts" masks what is in fact a quasi-collaborative judicial deliberation in which an STJ's initial findings are malleable, neither the Tax Court's own rules of procedure, the Federal Rules of Appellate Procedure, nor Congress' scheme for appeals from Tax Court decisions would require the Commissioner to include the STJ's preliminary report as part of the appellate record. Furthermore, this purportedly quasi-collaborative process would not offend our notions of fundamental fairness, nor would due process require the inclusion of the report in the appellate record to preserve the fairness of our review.

First, it is clear that the Tax Court's own rules do not require the report to be disclosed to the parties or made part of the appellate record. To the contrary, its rules specifically preclude the report's disclosure. That the Tax Court has the power to prescribe its own rules of procedure is undisputed. See 26 U.S.C. § 7453; Stone v. Comm'r, 865 F.2d 342, 347 (D.C. Cir. 1989) ("The Tax Court is of course free to make its own rules determining the relation between it and its Special Trial Judges."). Having exercised that rulemaking authority, the Tax Court no longer requires an STJ's report to be made available to the parties and, by extension, no longer allows those parties to file objections to it. Compare Tax Court Rule 182(b), (c), 60 T.C. 1149, (1973) (providing for service of the STJ's report on each party and allowing each party to file objections to the report's findings), with Tax Court Rule 183, 81 T.C. 1070, (adopted 1983) (noting that " [t]he prior provisions for service of the [STJ's] report on each party and for the filing of exceptions to that report have been deleted").

Neither do the Tax Court procedures prescribe any particular level of deference due the STJ's report. Under the current rule, the Tax Court maintains sole authority to decide cases assigned to an STJ. Tax Court Rule 183(c) ("The Judge to whom or the Division to which the case is assigned may adopt the [STJ's report] or may modify it or may reject it in whole or in part...."); see also Freytag v. Comm'r, 501 U.S. 868, 875 n. 3, 111 S. Ct. 2631, 115 L. Ed. 2d 764 (1991) (" [A] special trial judge has no authority to decide a case assigned under [§ 7443(b) (4)]."). The Tax Court thus acts as the original finder of fact. Conversely, the STJ's inability to decide cases limits the amount of deference that the Tax Court, as the original factfinder, must pay to those preliminary findings. Although the Rule requires that "due regard" be given to the STJ's opportunity to evaluate the credibility of witnesses and that those findings be presumed correct, see Tax Court Rule 183(c), to impose the further requirement that the Tax Court review an STJ's findings for clear error, as Kanter urges, would all but abdicate the Tax Court's original decisionmaking authority. Instead, we believe Rule 183's due-regard language merely instructs the Tax Court to be cognizant that the STJ had the opportunity to evaluate the credibility of witnesses, and allows the Tax Court to overcome the presumption of correctness it prescribes should it find that the evidence suggests those findings were incorrect. Consequently, secreting the report does not offend any rule-mandated check on the Tax Court's power to decide cases assigned to an STJ.

Second, Congress has by statute precluded direct appellate review of STJ reports. We lack jurisdiction to review anything but "decisions of the Tax Court." 26 U.S.C. § 7482(a) (1). We have repeatedly held that the use of the term "decisions" in § 7482 means that the appellate courts can review only (i) dismissals (e.g., for lack of jurisdiction) or (ii) formal determinations of deficiency (or lack thereof). See, e.g., Kreider v. Comm'r, 762 F.2d 580, 584 (7th Cir. 1985). In other words, a "decision" of the Tax Court is the final formal ruling of the Tax Court; a preliminary "report" is not a decision. See 26 U.S.C. § 7459(a) ("A report upon any proceeding instituted before the Tax Court and a decision thereon shall be made as quickly as practicable. The decision shall be made by a judge in accordance with the report of the Tax Court, and such decision so made shall, when entered, be the decision of the Tax Court." (emphasis added)). An STJ report, therefore, is not reviewable, see Estate of Smith v. Comm'r, 638 F.2d 665, 670 (3d Cir. 1981), which lends credence to the Commissioner's argument that Congress intended STJ reports to be treated as preliminary findings comprising part of the Tax Court's internal deliberative process.

Third, the Federal Rules of Appellate Procedure do not require that STJ reports be made part of the statutorily required record on appeal of a "decision" of the Tax Court. Federal Rule of Appellate Procedure 14 excepts appeals of Tax Court decisions from certain rules of appellate procedure.⁵  Among those rules not applicable to Tax Court review is Rule 16, which provides that the record on review of an administrative order shall include "any findings or report on which [the order] is based." Fed. R. App. P. 16(a) (2). Unlike other administrative actions, the Federal Rules of Appellate Procedure thus do not require that Tax Court decisions be reviewed in light of the preliminary findings upon which the decision was based. Instead, Rule 13 notes that Rule 10 governs the contents of a Tax Court appellate record, and that rule does not require the record to include any preliminary findings or reports. Fed. R. App. P. 13 & 10(a)⁶ 

With these considerations in mind, we find that the relationship between the preliminary reports of STJs and the final, reviewable "decisions" of the Tax Court bears striking resemblance to the relationship between reports of "divisions" of the Tax Court and final decisions of the Tax Court itself. A "division" is a subset of the Tax Court (often a single judge) that is designated to hear a single case and is empowered to make determinations with respect to disputes before the Tax Court. 26 U.S.C. §§ 7444(c), 7460(a). Under § 7460, the division's decision generally becomes the decision of the Tax Court, but the Tax Court retains the power to review the division's decision and render its own. Those preliminary recommendations of the division which do not become part of the final decision of the Tax Court also do not become part of the record for any additional future review. Id. § 7460(b) ("The report of a division shall not be a part of the record in any case in which the chief judge directs that such report shall be reviewed by the Tax Court.").

Thus, the two circuits to have interpreted § 7460 have ruled that a division's preliminary report is not part of the appellate record and not available for review by a federal appeals court once the Tax Court has undertaken an internal review. See Estate of Varian v. Comm'r, 396 F.2d 753, 755 n. 2 (9th Cir. 1968); Heim v. Comm'r, 251 F.2d 44, 48 (8th Cir. 1958). The Ninth Circuit noted a congressional intent in § 7460 to preclude a two-tier appellate relationship between the division and the full court and determined that such a mandate did not offend the court's notions of fundamental fairness. Varian, 396 F.2d at 755 n. 2.

Given the similarities between these two relationships, we are led to the same conclusion that the Ninth Circuit reached in Varian: there is no two-tier appellate relationship between STJs and the Tax Court. Instead, STJ reports are treated by the Tax Court as preliminary findings only and, in accordance with applicable rules and statutes, are not required to be made part of the record on appeal.

The dissent takes issue with this comparison, noting that divisions are comprised of one or more Tax Court judges, who are all presidentially appointed and serve a fifteen-year, statutorily mandated term of office from which they can be removed only for "inefficiency, neglect of duty, or malfeasance in office but for no other cause." 26 U.S.C. § 7443(f). STJs, however, serve at the discretion of the Chief Judge and have no statutorily mandated term of office. Id. § 7443A. If Tax Court Rule 183 provides for a quasi-collaborative process, the dissent fears this distinction between the participants impairs the judicial independence of the STJ. As a result, the dissent questions whether an STJ can participate meaningfully in this internal deliberative process. See post at 87.

We respectfully disagree. We are hesitant to suggest that members of the Tax Court would either expressly coerce or by nature of their office exert undue influence over an STJ. Nor will we discredit an STJ's express statement that the final Tax Court opinion "agrees with" his own based solely on this observed distinction.

This procedure, while admittedly unusual vis a vis typical judicial procedure, does not offend our notions of fundamental fairness. See Varian, 396 F.2d at 755. In this respect, we agree with the dissent's conclusion that due process neither requires the Tax Court to be constrained by a formal degree of deference to the STJ nor requires the Tax Court to rehear witnesses whether or not it ultimately reverses the STJ's findings. See post at 89-94 (discussing United States v. Raddatz, 447 U.S. 667, 100 S. Ct. 2406, 65 L. Ed. 2d 424 (1980) and Universal Camera Corp. v. NLRB, 340 U.S. 474, 492-94, 71 S. Ct. 456, 95 L. Ed. 456 (1951)). Although dictum in the Supreme Court's decision in Raddatz suggests that "serious questions" may arise from a district court's reversal of a magistrate judge's credibility findings without the opportunity to personally see or hear the witnesses, see 447 U.S. at 681 n. 7, 100 S. Ct. 2406, that observation was made within the context of a criminal appeal. In a civil tax proceeding, a party's interests are more akin to those at stake in a typical administrative adjudication, and we agree with the dissent that the risk of erroneous deprivation in a civil proceeding is neither "as high, nor the costs as great, as would be the case in a criminal milieu." See post at 93 (discussing Tax Court Rule 183 within the procedural-due-process framework established by Mathews v. Eldridge, 424 U.S. 319, 333, 96 S. Ct. 893, 47 L. Ed. 2d 18 (1976)). Moreover, we agree that the "only fully responsive remedy" to Kanter's complaint would prove unworkable, as it would require the enormous burden and prohibitive cost of rehearing witnesses, which would ultimately prove to add little value given the continued input that the STJ retains under Tax Court Rule 183's purportedly quasi-collaborative process. See post at 883-884.

But the recognition that little value would be added to the process given the STJ's continued involvement—signified by his adoption of the statement that the final Tax Court opinion reflects his own opinion—is ultimately why we take issue with the dissent's conclusion that our review of the Tax Court's opinion is unconstitutionally impaired by this procedure. The conclusion rests on the premise that STJs do not enjoy an equal voice in this purportedly quasi-collaborative process. See post at 886-887. As discussed above, we reject this premise. In as much as the final Tax Court opinion purports to agree with and adopt the opinion of the STJ, we therefore believe that the final opinion reflects the true legal opinions and findings of the STJ. Any differing preliminary recommendations—if they ever existed—would no longer be constitutionally relevant because the STJ has abandoned them. Should he feel otherwise, we would expect him—or the Tax Court—to say so.

As the Eleventh Circuit recently noted, "there is nothing unusual about judges conferring with one another about cases assigned to them." Ballard, 321 F.3d at 1043. If Tax Court Rule 183 in fact provides the opportunity for STJs and Tax Court judges to conference regarding the STJ's preliminary findings, then we have every reason to believe that Tax Court judges would duly regard the input of the STJ and that he, in turn, would participate meaningfully in the exchange. Like the Ninth and the Eleventh Circuits, we too are loath to interfere with another court's deliberative process. See id.

In any event, the issue is academic since the Tax Court's opinion in this case purports to "agree with and adopt" Special Trial Judge Couvillion's opinion. We will take the Tax Court at its word and, thus, move on to a discussion on the merits of Kanter's appeal.

The bulk of the Tax Court's opinion involves its determination that certain income tax underpayments were made with fraudulent intent. The facts surrounding the fraud issue are the most complex of all and will be described as succinctly as possible.

The Tax Court's decision tells a story of Kanter and two business associates—Robert Lisle and Claude Ballard. Lisle and Ballard worked in real estate, principally as managers in the real estate division of Prudential Insurance Company, where they had significant authority and influence over the conduct of Prudential's business. Kanter, Lisle, and Ballard concocted a plan whereby they would use Lisle and Ballard's positions along with Kanter's legal skills to collect and hide payments from people who wanted to do real estate business with Prudential.⁷ 

The Tax Court focused on five individuals (referred to collectively as the Five) who paid Kanter, Lisle, and Ballard for their influence.⁸  For the Five, Lisle and Ballard would help secure business opportunities with Prudential, and in return the Five made payments, including cash and partnership interests, to Kanter.

It was at this point that the IRS became interested. According to the Commissioner these payments were diverted (the Commissioner and Tax Court use the more highly charged word "laundered") to entities created by Kanter or someone acting at Kanter's direction. The dominant method was to divert payments to a Kanter-controlled corporation called Investment Research Associates, Ltd (IRA).⁹ 

The payments to IRA were deposited in the accounts IRA kept with another Kanter-controlled entity called the Administration Company. The Administration Co. was exactly what its name indicates: a company that administered the records and funds of its clients. It maintained books and records for its clients, and often prepared client tax returns. Additionally, it administered client funds, collecting receivables and paying bills. Money for clients was kept in pooled bank accounts listed under Administration Co.'s name. Two of these accounts were called the Special E account and the TACI Special Account. Client funds were pooled in order to get the higher rate of return on aggregated funds and to avoid a multiplicity of account maintenance charges on many, smaller client bank accounts kept in the clients' individual names. Administration Co. itself (and not the bank where the accounts were kept) tracked individual client balances within the pooled accounts on its internal records for each client. IRA and many of its subsidiaries were clients of Administration Co.¹⁰  In fact, it does not appear that Administration Co. had any significant clients not directly associated with (or controlled by) Kanter.

The payments by the Five to IRA that were deposited in the accounts of the Administration Co. were commingled with other non-Five-related funds of IRA as well as the funds of other Administration Co. clients. This commingling and the inadequate record keeping of Administration Co. made it difficult to track or account for the payments from the Five.

It was IRA—and not Kanter, Lisle, or Ballard—that reported the income from the Five's payments on its tax return (where it was off-set by aggregated losses). Subsequently, IRA distributed the income into subsidiary "sham" corporations, BWK, Carlco, and TMT, controlled by Kanter, Lisle, and Ballard, respectively, in a 10/45/45 ratio, respectively, that represented the division of the payments allegedly agreed to by Kanter, Lisle, and Ballard.¹¹  These sham corporations were used, in effect, as the personal bank accounts of Kanter, Lisle, and Ballard, and these men could withdraw or use the funds contained in them at their leisure.¹²  The funds withdrawn were never repaid and were eventually written off as bad debts. Alternatively, the controlled corporations' funds were used to fund the personal expenditures of Kanter, Lisle, and Ballard, as well as their families.

Kanter, on the other hand, argues that the payments to the corporations were merely investments by the entities, "ceded" to them by Kanter. The underlying scheme of securing business opportunities for the Five interested individuals in return for payments to Kanter (through Kanter entities) is disputed in its entirety by Kanter.

The Commissioner argued, and the Tax Court agreed, that this income was properly taxable to the individuals, and a deficiency was assessed. The Tax Court held that the diversion of these funds from the individuals to the sham corporations was undertaken with the fraudulent intent to evade the payment of taxes. According to the Tax Court, this arrangement not only disguised income earned by the individuals, but also hid their activities from Prudential.

In the Tax Court, Kanter presented an impressive list of witnesses associated with Kanter, Lisle, Ballard, the Five, and the various relevant entities—all of whom expressly denied that the Five paid "kickbacks" to Kanter for steering business the Five's way. But the Tax Court found that these witnesses had testified that the Five had, in fact, entered into payment arrangements with Kanter, Lisle, and Ballard in exchange for their influence in obtaining business. IRA, 78 T.C.M. (CCH) at 1065. This testimony established that Kanter had not reported income on which he was taxable and had therefore underpaid his federal-income-tax obligations. The Tax Court found that these underpayments were undertaken with intent to evade taxes based on, inter alia, the following factors: the lack of credibility of Kanter's testimony; Kanter's legal education and experience indicating an awareness of his obligation to report income accurately and pay taxes; Kanter's substantial underreporting of income for many years; Kanter's creation of a complex laundering network of sham corporations and other entities that made it difficult to trace the flow of funds; Kanter's lack of cooperation with the IRS—namely, the withholding of documents and the destruction of documents subject to summonses; Kanter's commingling of "kickback" monies with other monies in the Administration Co.'s accounts; Kanter's movement of monies through conduits that had no legitimate business purpose showing an intent to disguise sources of funds; Kanter's reporting of his personal income on the tax returns of IRA in order to create the appearance that income was earned by IRA; and Kanter's use of phony loans to disguise distributions to himself, later written off as bad debts. IRA, 78 T.C.M. (CCH) at 1083-85.

A finding of fraud by the Tax Court is a factual finding that we review for clear error. Toushin v. Comm'r, 223 F.3d 642, 647 (7th Cir. 2000). As with all findings of fact of the Tax Court, we must view the evidence in the light most favorable to the Tax Court's findings and reverse only if we are left with the definite and firm conviction that a mistake has been made. Pittman, 100 F.3d at 1312-13.

To establish fraud, the Commissioner must prove by clear and convincing evidence that Kanter's underpayment of taxes was done with intent to evade taxes that he knew or believed were owed. Toushin, 223 F.3d at 647; Pittman, 100 F.3d at 1319. Fraud may be proven through circumstantial evidence. Pittman, 100 F.3d at 1319. There are a variety of commonly recognized indicia of fraud, "such as keeping a double set of books, making false entries or alterations, [using] false invoices or documents, [destroying] books or records, [concealing] assets or covering up sources of income, handling ... one's affairs to avoid making the records usual in transactions of the kind, and [undertaking] any conduct, the likely effect of which would be to mislead or to conceal." Spies v. United States, 317 U.S. 492, 499, 63 S. Ct. 364, 87 L. Ed. 418 (1943). Additionally, fraud can be shown by significant and repeated understatement of income, Pittman, 100 F.3d at 1320, by a lack of cooperation with investigating agents, Korecky v. Comm'r, 781 F.2d 1566, 1568-69 (11th Cir. 1986); Zell v. Comm'r, 763 F.2d 1139, 1146 (10th Cir. 1985), by destruction of records, Powell v. Granquist, 252 F.2d 56, 59 (9th Cir. 1958), or by considering the legal experience and education of the taxpayer, Plunkett v. Comm'r, 465 F.2d 299, 303 (7th Cir. 1972).

Kanter's principal argument is based on the representation he makes above concerning the STJ's report: Kanter and his witnesses were found credible, and his story was believed by the STJ. Therefore, there can be no finding of intent to defraud. It was, Kanter alleges, only the rejection of the STJ's findings by the Tax Court judge that led to the finding that Kanter was not credible and that the underpayment of taxes was accomplished with fraudulent intent. As discussed above, we refuse to credit Kanter's allegations in light of the Tax Court's statement that it "agrees with and adopts" the opinion of the STJ. Therefore, Kanter's principal argument is without merit. The Tax Court found that Kanter's testimony was not credible and also found significant circumstantial indicia of fraud. Kanter's arguments now do not demonstrate that this is clearly erroneous.

Kanter's remaining arguments are also meritless. He argues that it is a common estate planning technique to cede personal investments to family trusts and corporations. But he does nothing to undermine the considerable evidence cited by the Tax Court that demonstrates that the Five paid for Kanter, Lisle, and Ballard's help in securing business opportunities, and that those payments went through IRA to Carlco, TMT, and BWK for division of the proceeds in a 10/45/45 split. And inasmuch as this argument seeks to show a lack of fraudulent intent by attacking the existence of a deficiency, this argument was expressly waived by Kanter.

Additionally, Kanter argues that fraud was raised by the respondent late in the process by the Commissioner's amendment to his Answer, that none of the Commissioner's notices of deficiency indicated fraud and that IRS agents' reports had found no intent to defraud. While the Tax Court might have found this persuasive when, as the trier of fact, it considered the issue of fraud, this argument is unpersuasive at this stage. Kanter does not argue that the Commissioner's amendment was improperly allowed or otherwise defective. Nor is the Commissioner's delayed amendment particularly probative of the sufficiency of the fraud evidence, given the difficulties facing the Commissioner in unraveling the relevant transactions during the time leading up to (and during) the trial.

Finally, Kanter's attacks on the circumstantial indicia of fraud are little more than efforts to recharacterize the evidence in a manner favorable to Kanter's version of events. But a review of the record in the light most favorable to the Tax Court's findings makes clear that it was not clear error to find significant indicia of fraud.

To review just a few of the relevant indicia: first and foremost, even Kanter concedes that he significantly underreported his income from the transactions with the Five over many years. (Pet. Br. at 33.)

Second, the money earned by Kanter through the transactions with the Five was diverted to IRA and commingled with other monies, where the funds were reported as income earned by IRA, before ultimately becoming available to Kanter personally—a complicated money trail that made it difficult to trace the flow of the funds and gave the appearance that the money had been earned by IRA.

Third, there was evidence that Kanter and those acting at his behest did not cooperate with investigating agents and even destroyed records that the IRS had requested formally through its summons power. See 26 U.S.C. § 7602. James Lunk, an IRS Case Manager, was asked during his testimony at trial what records Kanter and his representatives had provided voluntarily, to which he responded, "Generally they were records that really wouldn't do us a lot of good in the examination in terms of really getting at the type of information we needed to examine the transactions that we were most interested in." (Tr. at 1055-56.) IRS Agent Paul Dion testified similarly:

I have some very direct recollection [s] of the meeting, because it was the first time I met Mr. Kanter. And at the very beginning of the meeting, we discussed some—had a very nice conversation relating to very general type [s] of topics. I remember art was one of them, and a few other generalities. And as soon as we started discussing the required documentation, it was almost as if a different person appeared. And that person became, I guess, very demonstrative in terms of not wanting to provide us with the information that we asked for; basically saying that he was going to frame the issue and not us.

(Tr. at 831-32.) The Commissioner had to seek the assistance of the district court in order to enforce summonses against Kanter seeking documents relevant to the present case. The district court, in two separate rulings, indicated that Kanter had not cooperated and that documents had been destroyed even after the IRS summonses for those documents had been issued.¹³  The court stated that

[Solomon] Weisgal testified that since [receiving a summons to compel production of documents], some of the BK documents, including summoned documents relating to the Kanters, had been turned over to the Administration Company and that some have been discarded as part of a three-year record retention and discard policy.... [Linda] Gallenberger testified that she disposed of some documents [related to Kanter] after receipt of the IRS summons.

United States v. Administration Co., 74 A.F.T.R.2d 94-5252, 1994 WL 240518, at *2 (N.D. Ill. May 31, 1994). The district court went on to note that

The facts are that the Kanters first promised to produce the documents sought from [entities involved in the present case] and then, in early February 1994, notified government counsel that the Entities were third-parties over whom they had no control. The eleventh-hour change of position by the Kanters is indicative of bad faith on the part of the Kanters.

Id. at *3. Less than a month later, the same judge held Gallenberger in contempt for a continuing failure to comply fully with IRS summonses. United States v. Administration Co., 74 A.F.T.R.2d 94-5256, 1994 WL 285064, at *2 (N.D. Ill. June 23, 1994) ("Gallenberger misconceives her duty to use all reasonable efforts to comply with the summonses. A half-hearted request and cursory further search are insufficient.") (internal quotations and citations omitted).

These and the other indicia cited by the Tax Court, when considered in a light favorable to the Tax Court's findings, paint a clear and convincing picture of an intent by Kanter to evade the payment of taxes he knew or believed that he owed. It was not clearly erroneous for the Tax Court to find that Kanter's underpayment of taxes involved fraud.

The Bea Ritch Trusts (BRT) are a group of twenty-five trusts established in 1969 under one trust document. The trust document named Beatrice Ritch, Kanter's mother, as the grantor of BRT, Kanter's friend and business associate Solomon Weisgal as the trustee and the members of Kanter's family as the beneficiaries. According to the trust document, Beatrice Ritch funded each of the separate BRT trusts with a $100 check. No evidence (beyond the trust document itself) was presented to substantiate that this actually happened. Kanter himself was originally named as a beneficiary of twenty-four of the twenty-five trusts and he alone was granted a power of appointment over the beneficial interest of BRT. Kanter allegedly renounced his interest in BRT in 1971, 1977, and 1978, thereby purportedly eliminating his power of appointment over BRT. At sometime during or before 1987, however, sixty new beneficiaries were added to BRT.

Before 1987, Kanter borrowed money from BRT, and as of January 1, 1987, Kanter still owed the trust $287,030. This debt does not appear to have been secured, nor is there any indication in the record that a reasonable interest rate (if any) was charged on the loans.

In the early 1970s Kanter participated in business ventures in Long Island, New York, involving the then nascent cable television industry. He and other members of his law firm helped an entity that eventually became Cablevision Co. raise funds, secure financing and find investors for its cable business. In return, Kanter (along with others) was to receive interests in partnerships that themselves owned interests in Cablevision.¹⁴  Kanter arranged to have BRT receive the Cablevision partnership interests to which he was entitled—in essence he contributed the Cablevision partnership interests to BRT. The Tax Court found that the contribution of Kanter's Cablevision partnership interests (along with other income and assets) to BRT was the principal source of funding for BRT. Kanter, therefore, was the true grantor of BRT. IRA, 78 T.C.M. (CCH) at 1098-99.

In 1987, BRT reported capital gains of $2,033,368 for the fiscal year ending September 30, 1987, from the sale of the Cablevision partnership interests owned by BRT.¹⁵  Because Kanter was the true grantor of BRT and because he had a power of appointment to name beneficiaries of BRT, the Commissioner assessed a deficiency for the year 1987 against Kanter with respect to BRT's income from the Cablevision sale. At trial, the evidence made clear that the income at issue had been earned during the calender year 1986, and the Commissioner sought to amend his pleadings and reallocate the BRT deficiency from 1987 to 1986. The Tax Court allowed the amendment, and denied Kanter's request to shift the burden of proof to the Commissioner for having raised a "new matter." See Tax Court Rule 142(a) (1).

The Tax Court found that Kanter had failed to prove that he did not fund BRT and, despite his three renunciations of any beneficial interest in BRT, that he had not appointed the 60 new beneficiaries. As a result, the Tax Court held Kanter liable for tax on BRT's income for 1986 and 1987 under the IRC's grantor trust provisions. 26 U.S.C. §§ 671, 674. The Tax Court also found that, in the alternative, Kanter's borrowing of trust funds made him taxable on BRT's income. See 26 U.S.C. § 675(3).

1. Was the allowance of the amendment proper?

Kanter's first argument concerns the Tax Court's allowance of an amendment of the pleadings to conform to the proof that the alleged deficiency relating to BRT applied to 1986 rather than to 1987. Kanter argues that it was error to allow this amendment at all, and that, once allowed, the Tax Court further erred by not treating this as a "new matter" under Tax Court Rule 142(a), which would have shifted the burden of proof on the issue to the Commissioner.

The decision to allow or deny an amendment of the pleadings under Tax Court Rule 41 is reviewed for abuse of discretion. See Estate of Ashman v. Comm'r, 231 F.3d 541, 542 n. 2 (9th Cir. 2000); LeFever v. Comm'r, 100 F.3d 778, 786 (10th Cir. 1996); Braude v. Comm'r, 808 F.2d 1037, 1039 (4th Cir. 1986). It was not an abuse of discretion for the Tax Court to allow the amendment of the Commissioner's pleadings to reallocate the BRT deficiency to 1986. Kanter was on notice as to the specific partnership income of BRT that was the subject of the assessed deficiency. (Tr. at 4483-86.) There was no prejudice to Kanter in the adjustment of the date, and he had a fair opportunity to defend with respect to the amended claim. In fact, it was Kanter, and not the government, who elicited the testimony (from Kanter's accountant, Gallenberger) that revealed the timing error and prompted the government's request for amendment.¹⁶  The government apparently did not realize its error until Kanter pointed it out at trial. Further, Kanter points to no additional evidence that he was prevented from introducing by virtue of the late amendment. The amendment of the pleadings to conform to the evidence was proper.

2. Should the amendment have shifted the burden of proof?

Kanter argues next that the amendment to the pleadings was a new matter that shifted the burden of proof on this issue to the Commissioner. See Tax Court Rule 142(a) (1) ("The burden of proof shall be upon the petitioner, except ... that, in respect of any new matter, ... it shall be upon the respondent."). The distribution of burdens is a question of law that we review de novo. This argument fails as well. The Commissioner is allowed the latitude to amend his pleadings and even adopt entirely new theories supporting assessed deficiencies without triggering Rule 142's shift in burden, so long as the new theory is not inconsistent with the original allegation, does not require new evidence in its support, nor increases the amount of the deficiency. See, e.g., Friedman v. Comm'r, 216 F.3d 537, 543 (6th Cir. 2000) ("A new position taken by Commissioner is not necessarily a `new matter' if it merely clarifies or develops Commissioner's original determination without requiring the presentation of different evidence, being inconsistent with Commissioner's original determination, or increasing the amount of the deficiency."); Abatti v. Comm'r, 644 F.2d 1385, 1390 (9th Cir. 1981) (same); Achiro v. Comm'r, 77 T.C. 881, 890, 1981 WL 11333 (1981) (same).

The problem for Kanter is that the Commissioner's amendment did not offer a new theory for the alleged deficiency. The theory under which the Commissioner proceeded at all times was that specific transactions produced taxable income that was reported by BRT but which was properly taxable to Kanter by virtue of the grantor trust provisions of the tax code.¹⁷  The case relied upon by Kanter here supports our conclusion. Achiro involved an amendment under which the Commissioner alleged for the first time that a deduction was improper under a section of the tax code completely different from the section originally argued. 77 T.C. at 889-90; see also Shea v. Comm'r, 112 T.C. 183, 190-92, 1999 WL 177471 (1999). The Commissioner's amendment here presented a much more modest change, more akin to a clarification of the original allegation. Nor is Kanter correct when he alleges that the amendment increased the assessed deficiency. The stated deficiency remained constant, and involved the same income from the same transactions, but was re-allocated from one disputed year to another disputed year to correct a calender error that did not prejudice Kanter's case.¹⁸  The Commissioner did not propose additional income nor require Kanter to defend against a larger deficiency than had been assessed before the amendment. Therefore, it is incorrect to claim that the amendment resulted in "an entirely new and increased deficiency in a different year." (Pet. Br. at 46 (emphasis in original).) The Tax Court was correct in determining that the amendment did not shift the burden of proof to the Commissioner.

Although Kanter argues principally that the Commissioner should have borne the burden of proving that Kanter was the grantor and owner of BRT, and that the Commissioner could not meet that burden, Kanter maintains that, regardless of burden, he is not taxable on BRT's income under the IRC's grantor trust provisions. A finding of grantor status is a factual finding that we review for clear error. Scott v. Comm'r, 226 F.3d 871, 874 (7th Cir. 2000). Kanter has failed to show that the Tax Court's decision was clearly erroneous.

Despite the trust document's showing that Beatrice Ritch was the grantor of BRT, the familiar principle of substance over form views as the true grantor the one who principally funded the trusts. Schulz v. Comm'r, 686 F.2d 490, 496 (7th Cir. 1982); United States v. Buttorff, 761 F.2d 1056, 1060-61 (5th Cir. 1985). In Schulz, for example, the petitioner's wife was considered the grantor of a family trust because both the Commissioner and this court disregarded the conveyance of the wife's assets, which were subsequently used to fund the trust in question, to the petitioner. 686 F.2d at 496. The wife, in substance, provided the funds for the trust, and the form of the funding, by routing the funds through the petitioner, was ignored. Similarly, although Kanter was not listed in the trust document as a grantor, and, even if Beatrice Ritch actually did contribute $100 towards the funding of each BRT trust, there is evidence that Kanter contributed to BRT income and assets he earned, to substantially fund the twenty-five trusts.

The Tax Court found that Kanter transferred to BRT income and assets associated with his services in the Cablevision transactions as well as others. IRA, 78 T.C.M. (CCH) at 1098. Kanter provided no evidence (other than the trust document itself) to show that Beatrice Ritch actually funded the trusts nor any evidence to show from whence the substantial assets owned by BRT originated. Kanter's assertion that he merely ceded investment opportunities in Cablevision to BRT does nothing to undermine the Tax Court's findings and, further, depends on his credibility, which the court found to be lacking. Because Kanter was the principal source of the funding of BRT, he is deemed a grantor of BRT. See 26 C.F.R. § 1.671-2(e) (1) (" [A] grantor includes any person to the extent such person either creates a trust, or directly or indirectly makes a gratuitous transfer ... of property to a trust.... If a person creates or funds a trust on behalf of another person, both persons are treated as grantors of the trust.").

Kanter argues that the Tax Court's finding that he was the grantor is directly contrary to an Illinois trial court decision, in which the Cook County Circuit Court determined that the Cablevision partnership interests did not constitute property of Kanter's law firm because they did not result from payment of fees for legal services. Statland v. Levenfeld, No. 84 CH 6494 (Ill.Cir.Ct., Ch. Jan. 28, 1988). Kanter argues that (a) the state court ruling stands for the proposition that the Cablevision partnership interests were not in payment for any services (not just legal services) but were rather investments and (b) the state court ruling is determinative and binding on the matter. Kanter overstates the holding of the cited case. The case involved a former law partner of Kanter, who alleged that the Cablevision partnership interests were the property of the partnership because they were provided in payment for legal services. The Illinois court disagreed and held that the Cablevision partnership interests were not the product of legal fees paid to Kanter's law firm. The court did not hold, however, that the Cablevision partnership interests were not in payment of fees for services furnished by individual law firm partners in securing financing and investors for Cablevision, which is what the Commissioner alleges here. The trial court did, at various times, refer to the Cablevision partnership interests as individual investments, but it also spoke of the activities of Kanter and his partners in working to obtain financing and investors for Cablevision. None of these dicta contradicts the position of the Commissioner nor renders the factual findings of the Tax Court clearly erroneous.

Generally, if a grantor of a trust has the power to dispose of the beneficial enjoyment of that trust through a power of appointment, then the grantor is treated as the owner of the trust and the income of the trust must be included in the income of the grantor. 26 U.S.C. §§ 671, 674(a); 26 C.F.R. § 1.674(a)-1. Kanter argues that he three times renounced his beneficial interest in BRT and, by those renunciations, lost the power of appointment that he had under the trust document. But the Tax Court notes that, after the third of Kanter's renunciations, sixty new beneficiaries were added to BRT.¹⁹  Kanter was the only person who ever had the power under the trust document to appoint new beneficiaries. The most logical inference (which the Tax Court drew) is that Kanter himself appointed the new beneficiaries and that his earlier renunciations were shams. Kanter fails to rebut this. Kanter's only argument directly on this issue appears in his reply brief, in which he argues that the Tax Court improperly refused to reopen the record to admit evidence that Kanter's children were the grantors of the various JSK Trusts that had been added as beneficiaries of BRT.²⁰  (Reply Br. at 18.) This purported evidence would not be probative of the issue whether Kanter had exercised the power of appointment under BRT. The Tax Court's findings that Kanter was the grantor of BRT and that he held a power of appointment of beneficiaries of BRT were not clearly erroneous. Therefore, the Tax Court's finding that Kanter is taxable on BRT's income in 1986 (and 1987) is also not clearly erroneous.

As an alternative, the Tax Court also found that Kanter's borrowing from BRT without repayment and without adequate security subjected Kanter to liability for BRT's income. See 26 U.S.C. § 675(3) ("The grantor shall be treated as the owner of any portion of a trust in respect of which.... [t]he grantor has directly or indirectly borrowed the corpus or income and has not completely repaid the loan, including any interest, before the beginning of the taxable year."). It appears undisputed that Kanter had borrowed money from BRT in a way that subjects him to liability under § 675(3), and that at the beginning of 1987 he still owed BRT $287,030. This would be sufficient for Kanter to incur liability for 1987.²¹  There are, however, no findings by the Tax Court as to any amount Kanter owed to BRT as of January 1, 1986. Therefore, there could be no tax liability for Kanter in 1986 based on § 675(3). In the end this conclusion does not affect ultimate tax liability because under § 674 Kanter remains liable for both 1986 and 1987.

In conclusion, we affirm the Tax Court's determination that Kanter was the grantor of BRT in 1986 and 1987 and is taxable on BRT's income for those years.

In 1980, Kanter was approached by Richard Feigen, a client, who wanted help in securing funding to buy a painting of George Washington located in England that Feigen believed to be by John Trumbull.²²  Kanter put Feigen in touch with another client, Mr. Rappaport,²³  in Switzerland, who agreed to finance the purchase of the painting. The transaction closed. After the purchase, the painting was discovered not to be a Trumbull, and the seller agreed to rescind the sale. As a result of exchange rate and interest rate changes, Rappaport, who had advanced the funds, lost money. Rappaport wanted to be made whole, and when rebuffed, threatened to sue Feigen and Kanter for his claimed loss. Kanter, in convincing Feigen to reimburse Rappaport, agreed to provide $94,231 of his own funds. Kanter also paid $10,000 in legal fees incurred throughout the course of the transaction.

On his 1980 tax return, Kanter claimed the sum of $104,231 as a deduction on Schedule C, stating that his main business activity was "buying and selling paintings." The IRS disallowed the claimed loss. The Tax Court found, based on Kanter's testimony, that Kanter's main occupation was as an attorney and that he did not hold himself out as an art expert or art dealer. This made the claimed deduction unrecognizable under 26 U.S.C. § 162 as an ordinary and necessary business expense. Additionally, the Tax Court did not allow this deduction as a § 212 expense for production of income "because Kanter received no fees and no contract existed therefor, [and so] the expenditure did not bear a reasonable and proximate relationship to the production of income." IRA, 78 T.C.M. (CCH) at 1121. " [Kanter] merely served as an intermediary to introduce Feigan to another friend, Rappaport. There was no written contract between Feigan and Kanter for the sharing of profits." Id.

The Commissioner's determination that Kanter's expenses associated with the failed sale of the Washington painting were not legitimate deductions under 26 U.S.C. §§ 162 and 212 is a finding of fact that we review for clear error. Reynolds v. Comm'r, 296 F.3d 607, 615 (7th Cir. 2002); Buelow v. Comm'r, 970 F.2d 412, 415 (7th Cir. 1992).

The Tax Court found, essentially, that Kanter could not deduct the expenses under §§ 162 and 212 because Kanter's business was the practice of law, not art, and there was no evidence that Kanter had an agreement with Feigen to receive fees or share profits from the purchase and planned resale of the painting. The Tax Court simply did not believe the "self-serving and uncorroborated" testimony of Kanter. IRA, 78 T.C.M. (CCH) at 1121. Given the totality of the Tax Court's findings that Kanter systematically engaged in extensive (non-law related) business dealings whereby he "entered into arrangements pursuant to which he would use his... contacts ... to assist individuals ... in obtaining business opportunities or in raising capital for business ventures [,]" we find it implausible for the Tax Court to conclude that Kanter was not engaged in such dealings when he helped Feigen secure financing from Rappaport. IRA, 78 T.C.M. (CCH) at 970.

Without excessively detailing the transactions relating to the Five that we summarized in Part II, we note that the Tax Court's opinion went to considerable lengths to confirm deficiencies against Kanter based on a characterization of Kanter's "business" as ranging far beyond the simple practice of law. In each of the transactions involving the Five, the foundation for the Tax Court's determination that Kanter consistently understated his income was its findings that time after time Kanter provided non-legal business services in facilitating business opportunities for the Five with Prudential, and that he was paid handsomely for that facilitation. It is difficult for us to read the first few hundred pages of the Tax Court's opinion and not be left with the distinct impression that the Tax Court believes Kanter was involved in this business "activity with continuity and regularity and... [his] primary purpose for engaging in the activity [was] for income or profit." Comm'r v. Groetzinger, 480 U.S. 23, 35, 107 S. Ct. 980, 94 L. Ed. 2d 25 (1987). At the very least, Kanter has shown a distinct proclivity to seek income and profit through activities similar to the failed sale of the painting.

Determining whether an activity was engaged in for profit requires an examination of the relevant factors outlined in Treasury Regulation 1.183-2. See 26 C.F.R. § 1.183-2; Burger v. Comm'r, 809 F.2d 355, 358 (7th Cir. 1987). The nine factors for determining whether Kanter's involvement in the sale of the George Washington painting was carried on for profit include: (1) the manner in which the taxpayer carries on the activity; (2) the expertise of the taxpayer; (3) the time and effort expended by the taxpayer in carrying on the activity; (4) the expectation that assets used in the activity may appreciate in value; (5) the success of the taxpayer in carrying on other similar or dissimilar activities; (6) the taxpayer's history of income or losses with respect to the activity; (7) the amount of occasional profits, if any, which are earned; (8) the financial status of the taxpayer; and (9) elements of personal recreation or pleasure in carrying on the activity. 26 C.F.R. 1.183-2(b).

We can, based on Kanter's testimony and considering the Tax Court's other findings, construct a highly probable scenario under which Kanter was to find financing for Feigen's purchase in exchange for part of the anticipated profits. Under factor (1), this manner of doing business, while not well documented, is consistent with the method pursued in other transactions detailed by the Tax Court. Under factor (2), Kanter's apparent expertise in facilitating financing transactions and putting people together for deals in order to generate income was so evident to the Tax Court that it found him liable for large underpayments of income tax. Additionally, Kanter has indicated that the art project was attractive because of Feigen's expertise in generating sufficient publicity and excitement about artwork to likely increase the resale price, thus satisfying factor (4). (Tr. at 4416-30.) Regarding factor (5), Kanter's obvious success in other facilitation-of-business ventures was what landed him in trouble with the Commissioner in the first place. An appealing tale of the Washington painting scheme as a failed for-profit venture emerges from this evidence. In fact, while there is significant evidence that Kanter undertook similar, though not identical, transactions as the failed painting purchase for profit, there is no evidence that Kanter facilitated any other deals of this kind for personal reasons (i.e., gratuitously). And the Commissioner presents no evidence contradicting Kanter's version of events. In fact, by asserting that, in all of his other disputed transactions in which income was produced, Kanter had participated in the manner alleged here and with the kind of profit motive he alleges here, but arguing that in this single transaction, where losses were generated, Kanter participated for only personal reasons, the Commissioner appears to want to have his cake and eat it too.

There are certain difficulties in reversing the Tax Court here. We would have to reverse the Tax Court's credibility finding on the matter. In that connection, neither Feigen's testimony, nor Rappaport's testimony, nor any piece of specific evidence exists in the record to support Kanter's testimony. And we have found elsewhere that, while a taxpayer's uncontradicted testimony was sufficient to demonstrate that the Commissioner's determination was erroneous, the Tax Court can disregard that testimony if it is not credible. Lerch v. Comm'r, 877 F.2d 624, 631 (7th Cir. 1989). The Tax Court did not here believe Kanter's claim of a profit motive. While finding Kanter not credible in this instance is consistent with the Tax Court's refusal to find him credible on other issues, it results in a completely inconsistent and implausible factual scenario — Kanter's incurring more than $100,000 in expenses gratuitously to facilitate a business transaction where huge potential profits lie. But without any indication why this transaction is unique among all the other Kanter activities, we think the Tax Court was clearly erroneous in rejecting Kanter's claim that he was pursuing commercial objectives here as elsewhere.

In conclusion, we find that there must have been a profit motive in Kanter's involvement in the aborted purchase of the George Washington painting; the Tax Court's finding to the contrary is clearly erroneous and we reverse. Kanter's deduction for the $104,231 of expenses incurred in facilitating the failed sale of the George Washington painting is allowed.

In examining Kanter's 1982 tax return, the IRS determined that inadequately documented deposits into three of Kanter's bank accounts during that year greatly exceeded the income Kanter reported on his federal tax return. Based on a comparison of the deposit amounts to Kanter's tax return for 1982, the Commissioner determined that Kanter had unreported gross income that year in an amount of $2,084,017, and issued an assessment of deficiency. After Kanter produced sufficient records documenting some of the deposits in question, the Commissioner reduced this deficiency to $1,303,207, which comprised the sum of four specific deposits:

------------------------------------------------- DEPOSIT AMOUNT PAYOR OR SOURCE ------------------------------------------------- $787,129 The Holding Co. (THC)  ----------------------------------------------- $ 40,000 Computer Placement Services, Inc. (CPS)  ------------------------------------------------- $190,078 Administration Co. (Special E Account) ------------------------------------------------- $286,000 Administration Co. (Special Account) -------------------------------------------------

The Tax Court found that the Commissioner had produced sufficient evidence to create a presumption of correctness in the indicated deficiency. The bank deposit analysis method was a reasonable method of reconstructing income, and deposits from THC, CPS and the Administration Co. accounts often included taxable income. Therefore it was proper, the Tax Court ruled, for the Commissioner to determine gross income from an analysis of total deposits to Kanter's bank account minus any deposits properly excluded from income and to assess a deficiency for the excess of that income over the reported income. The Tax Court found that Kanter had not adequately documented that the four deposits in question were loans, and they were properly included in his gross income under the deposit analysis. Moreover, the Court did not find the testimony of Kanter's accountant, Gallenberger, credible on the issue.

The Tax Court's determination that a taxpayer has unreported income is a finding of fact that we review for clear error. Reynolds, 296 F.3d at 612. The Commissioner's assessment of a deficiency is presumed correct, but can be overcome by rebuttal evidence presented by the taxpayer. Pittman, 100 F.3d at 1313.

Kanter has three arguments with respect to the deficiency determined by the bank deposit analysis: first, the change in the amount of the deficiency for 1982 under the bank deposit analysis constituted a "new matter" under Tax Court Rule 142 that required the burden of proof to shift to the Commissioner, and with that burden the Commissioner cannot prevail. Second, the lack of evidence on this issue made the deficiency a "naked assessment," and it was, therefore, improper to give the Commissioner's assessment of a deficiency the usual presumption of correctness. See Weimerskirch v. Comm'r, 596 F.2d 358, 360 (9th Cir. 1979) (finding no presumption of correctness when Commissioner's assessed deficiency lacks any reasonable foundation in the evidence). Finally, Kanter argues that the Tax Court clearly erred in finding a deficiency in light of the evidence presented. All of these arguments fail.

First, there is no merit in Kanter's contention that the Commissioner's reduction in the amount of deficiency assessed in 1982 under the bank deposit analysis is a new matter requiring a shift in the burden of proof. The Commissioner's concession that part of the original deficiency was properly excluded from Kanter's income is not a new theory, is not inconsistent with the original assessment of deficiency, does not require new or different evidence from Kanter and does not increase the assessed deficiency. See discussion supra Part III. The same theory of bank deposit analysis was being applied before and after the Commissioner's concessions, and the resulting deficiency was reduced, not increased, when Kanter produced documentation demonstrating to the Commissioner's satisfaction that certain deposits were not income. The remaining, insufficiently documented deposits, therefore, do not constitute a new matter.

Nor can Kanter escape the presumption of correctness by relying on the "naked assessment" exception of Weimerskirch. "The general rule is that a presumption of correctness attaches to the Commissioner's deficiency determination; the taxpayer has the burden of disproving it. A narrow exception exists where the determination is arbitrary and erroneous or without rational foundation." Pfluger v. Comm'r, 840 F.2d 1379, 1382 (7th Cir. 1988) (citations omitted). It is not disputed that "before the Commissioner can rely on [the] presumption of correctness, the Commissioner must offer some substantive evidence showing that the taxpayer received income from the charged activity." Weimerskirch, 596 F.2d at 360. But the threshold for properly invoking the presumption is not as high as Kanter would have us believe, nor is the evidence presented by the Commissioner as thin as that presented in Weimerskirch.

In Weimerskirch, the Commissioner assessed a deficiency based on more than $24,000 of unreported income from the taxpayer's alleged sale of heroin. Id. at 359. The Commissioner, however, presented no evidence from which one could even infer that Weimerskirch was involved in heroin sales in any way. There were, for example, no records of bank deposits, net worth, or cash expenditures. Id. at 361-62. The deficiency was calculated based on an IRS agent's purely hypothetical calculation of what income would be realized from a given level of heroin sales each week over a given period of time. Id. at 359. Not only was the deficiency calculation completely divorced from any evidence of actual or inferrable heroin sales by Weimerskirch, but the entire method of calculation used was unsupported by evidence. Id. The Commissioner's attempt to attach the presumption of correctness to the assessed deficiency in Weimerskirch was, in every sense of the word, naked.

In contrast, the presumption of correctness in the present case appears far more modestly appareled. The Commissioner's use of bank deposits as circumstantial evidence of gross income is an accepted methodology. See, e.g., United States v. Ludwig, 897 F.2d 875, 878 (7th Cir. 1990). The Tax Court found evidence that Kanter was engaged in income producing businesses. And the court found that actual deposits were made that had the appearance of income. IRA, 78 T.C.M. (CCH) at 1104; see also United States v. Esser, 520 F.2d 213, 217 (7th Cir. 1975) (describing test for implementing deposit analysis method of reconstructing income). There is no dispute about the evidence of deposits. Consequently, there was sufficient evidence to support the presumption of correctness.

Lastly, Kanter argues that the uncontradicted evidence at trial shows that the four deposits in question were loans and were not income. By uncontradicted, we can only assume Kanter means uncontradicted by any evidence beyond the evidence we have outlined, establishing, prima facie, that there was a deficiency during 1982. Because we have found that the Tax Court properly gave the Commissioner's assessed deficiency a presumption of correctness, we must interpret Kanter's argument as claiming that the evidence presented overcame the presumption. In that respect, Kanter's argument also fails.

In considering whether Kanter met his burden of proof, the Tax Court found Gallenberger's summary and Kanter's assertions unconvincing. "No credible evidence was introduced to support Kanter's assertion that the deposits were loans.... We find the testimony of the accountant, Gallenberger, unreliable and her analysis fatally flawed...." IRA, 78 T.C.M. (CCH) at 1104. The Tax Court also recalled "Gallenberger's regular practice of record destruction" and the general lack of any documentation other than her summary and testimony. Id. at 1104-05. The evidence that Kanter asserts was "uncontradicted" was considered by the Tax Court and found wanting. Exhibit 9172PK, which, according to Kanter, demonstrates that the deposits from CPS, THC, and Administration Co. were loans, reflects a summary analysis by Linda Gallenberger that the Tax Court found not credible. Nor is it correct for Kanter to say that the Tax Court had determined previously that one of the deposits of money from Administration Co. was money already found to belong to Kanter. (Pet. Br. at 58.) Kanter cites to a sentence in the Tax Court's opinion as to issue 21 (allowable capital gains and losses in 1987) wherein the court states, "As indicated previously, funds from the Administration Co. special E and PSAC special E accounts were Kanter's funds." IRA, 78 T.C.M. (CCH) at 1126. This sentence, in context, refers only to the specific funds from the accounts cited and used in the specific transactions at issue (five years after the deposits at issue here) in that section of the Tax Court opinion. These specific funds were Kanter's own funds, and he was entitled, therefore, to the basis he had claimed for the assets he had purchased with those funds. But Kanter seeks to generalize this statement about his ownership of specific money to mean that all the money in the Special E accounts, at all times, already belonged to Kanter. The Tax Court's opinion does not say this.

In conclusion, the Tax Court did not clearly err in its findings regarding Kanter's unreported income in 1982 as determined by bank deposit analysis. These findings are therefore affirmed.

VI. Equitable Leasing's Payments to THC & Zion

Equitable Leasing Co., Inc. (Equitable) was the wholly owned company of Joel Mallin, a friend and former law partner of Kanter. Mallin used Equitable (among other companies) to promote and structure equipment leasing transactions to make money for investors. The investors' capital would be highly leveraged to purchase the equipment to be leased, and the residual value of the equipment at the end of the lease would provide the ostensible paper profit on the investors' capital. A principal motivation behind the investment in the leasing transactions, however, was the opportunity for tax benefits associated with the venture. Kanter acted as a middleman, introducing Mallin to investors for his leasing transactions. In return, Mallin paid commissions to Kanter entities. The Commissioner issued a notice of deficiency for 1983 that included a deficiency of $635,250 with respect to four payments by Equitable to Kanter entities, Zion, and THC.²⁶ 

------------------------------------------ FORM OF DATE PAYMENT PAYEE AMOUNT ------------------------------------------ Jan. 4, Bank Zion $317,250 1983  Transfer ------------------------------------------ Jan. 24, Check THC $ 9500 1983 ------------------------------------------ June 1, Check Zion $ 6500 1983 ------------------------------------------ June 30, Bank THC $302,000 1983 Transfer ========================================== Total $635,250 ------------------------------------------

The Commissioner alleged that these payments were commissions paid for Kanter's procurement of investors in Mallin's enterprises. Kanter argued that Zion and THC were loaned money by Equitable that was then turned around and invested in Equitable as an accommodation to Equitable to enable it to close certain transactions.

The Tax Court found that there was sufficient evidence to indicate that Equitable had paid commissions to Zion and THC for Kanter's services in providing investors to Mallin. Two of the four payments, for $9500 and for $6500, were labeled (on the check in payment and in THC's adjusting journal entries, respectively) as commissions. The Tax Court determined that this evidence was sufficient to support the presumption that all of these payments to Zion and THC were income to Kanter. In response to this presumption, the Tax Court found that Kanter had provided no evidence other than his "uncorroborated, self-serving testimony" concerning the "accommodation" arrangement with Equitable. IRA, 78 T.C.M. (CCH) at 1103. The court found, as previously noted, that Kanter was not credible and that the four payments were income attributable to Kanter.

Whether or not monies are taxable income to a taxpayer is a finding of fact that we review for clear error. Reynolds, 296 F.3d at 612. The Commissioner's assessment of a deficiency is presumed correct, but can be overcome by rebuttal evidence presented by the taxpayer. Pittman, 100 F.3d at 1313.

Kanter argues first that the Commissioner's assessment of a deficiency should not be given the benefit of the presumption of correctness because the Commissioner "failed to provide any evidence connecting Kanter to this income." (Pet. Br. at 59.) This argument fares no better here than it did when we considered it with respect to the Bank Deposit issue. There was evidence that Mallin and Kanter were partners in various investments, and that Kanter's law firm had procured investors for Mallin's Equitable Leasing transactions. (Tr. at 5213-15.) There was also evidence that Mallin had paid commissions for these services to Kanter entities. (Id.) Additionally, at least two of the payments in question were documented as commissions. This was sufficient evidence to provide a rational foundation for the Commissioner's assessment. The Commissioner's assessment appropriately received its presumption of correctness. See Pittman, 100 F.3d at 1313.

Kanter also argues that he has rebutted the Commissioner's deficiency with evidence that only the $16,000 that was labeled as commissions (comprised of the $9500 and $6500 checks) should be considered commission payments, and the remainder consisted of loans made as an accommodation to Equitable. Even accepting Kanter's concession that the $9500 and $6500 checks from Equitable Leasing to THC and Zion, respectively, were taxable commission income, we do not believe he has provided sufficient evidence to show that the two bank transfers of $317,350 and $302,000 were loans.

Regarding the $317,250 transfer, there is clearly evidence supporting the Tax Court's finding that the transfer was a commission paid to Zion, and Kanter has provided no evidence to rebut this finding despite his control of the entities involved and his access to the records of those entities. First, as noted, there was circumstantial evidence that Equitable Leasing was making payments to THC and Zion as commissions. Additionally, THC's adjusting journal for 1983 has an entry "32" indicating an amount of $317,250 that was labeled, "Due from Zion — Commission Income to adj. for commiss. from Eq. Leasing, loaned to Zion on 1/4/83." (Ex. 146, THC adjusting journal at 6.) Finally, Kanter appears to concede in the facts section of his brief that this amount was received as commission. In his brief on appeal, Kanter breaks down the $635,250 as including "$317,250 as commission income from Equitable (see Ex. 146, THC adjusting journal, p. 6, adjusting journal entry 32, reflecting commission income of $317,250 which was paid to THC's subsidiary Zion as a loan from THC ...)." (Pet. Br. at 16.) The Tax Court's finding that this payment was commission income was not clearly erroneous.

What about the June 30, 1983 transfer from Equitable to THC for $302,000? Against the general circumstantial evidence of commission payments to THC by Equitable and the presumption of correctness of the Commissioner's assessment of deficiency, there is an entry in THC's general ledger for June 30, 1983, for $302,000 labeled "Loan from Equitable Leasg." (Ex. 148, THC general ledger at 12.) Together with this entry there is an entry indicating an $8000 transfer three days earlier also labeled "Loan from Equitable Leasg." (Id.) This $8000 transfer is corroborated by a check from Equitable to THC on June 24, 1983, for $8000 that has a memo line reading, "loan." (Ex. 9203PK, check # 2391.) The corroboration of the general ledger entry for the $8000 transfer is indirectly probative of the likelihood that the roughly contemporaneous and similarly labeled ledger entry for the $302,000 transfer may also be correct. These circumstances make it more likely that the ledger is accurate in its indication that the $302,000 transfer was a loan and not a commission payment.

But there is one additional piece of evidence in the record that the Commissioner and Kanter have not mentioned. In the THC adjusting journal, entry 59 is a barely legible entry for $310,000 that appears to read, "N/P — Equitable Leasing, Commission Income, to reclassify funds from Eq. Leasing." (Ex. 146, entry 59.) This entry is significant because we believe it may also represent the $302,000 and $8000 transfers and show that those transfers were commissions. These two transfers were roughly contemporaneous, and Kanter himself aggregates them and describes them as a cumulative amount in his brief. (See Pet. Br. at 16 ("$310,000 as loans from Equitable ... reflecting a $302,000 loan and an $8,000 loan....").) The $310,000 in entry 59 would appear to be the same as the $310,000 in the general ledger and as the $310,000 described by Kanter, comprising the total of the $8000 and $302,000 transfers. Entry 59 in the THC adjusting journal weighs against the general ledger's notations for both the $302,000 and the $8000 transfer and weighs further in favor of the Tax Court's finding that the $302,000 was a commission payment. In any event, this evidence does, at the very least, make the characterization of the $302,000 transfer unclear, and because Kanter bears the burden of rebutting the Commissioner's assessment of deficiency and controlled the entities whose records could have cleared this matter up definitively, we cannot say that the Tax Court clearly erred in finding that the $302,000 was a commission payment to THC for Kanter's services.

In conclusion, it was not clear error for the Tax Court to find that the transfers from Equitable Leasing were commissions and were taxable income.

VII. Cashmere Investment Assoc.'s transactions

During the 1970s, Kanter was involved in a series of real estate developments with developer Sam Zell. The development properties were owned by partnerships (real estate partnerships). Kanter's interests in the real estate partnerships were held through a series of Kanter-controlled trusts, including the BWK Revocable Trust, the Everglades Trusts 1-5, and the BWK Family Trusts (referred to collectively as the grantor trusts). Each of these trusts was a grantor trust whose income was attributable to Kanter personally. A detailed breakdown of the real estate partnership interests and their associated financial attributes is provided in the Tax Court's opinion. IRA, 78 T.C.M. (CCH) at 1107-11. The interests in these real estate partnerships had, as of 1982-83, zero basis.

During 1982, Zell expressed his desire to purchase all of the real estate partnership interests held by the grantor trusts. Kanter was interested in selling to Zell, but was concerned that significant gains would be triggered by an unqualified sale of the real estate partnership interests held by his grantor trusts. The Tax Court found that twenty one of the real estate partnership interests held by the grantor trusts had negative capital accounts; that is, their liabilities exceeded their bases, by an amount, in total, of $476,889. IRA, 78 T.C.M. (CCH) at 1108. Therefore, the unqualified transfer of those interests would involve the assumption of liabilities in excess of the bases, and would trigger capital gains to the seller — Kanter, who was the owner by virtue of the IRC's grantor trust provisions. See 26 U.S.C. §§ 671-677.

In 1983, a series of transactions took place that ultimately resulted in the transfer of the real estate partnership interests from the grantor trusts to a Zell-controlled entity. The transfer took place in three stages.

1. Transfer of real estate partnership interests to Cashmere

Cashmere Investment Associates, Inc. (Cashmere) was an inactive "shelf" corporation incorporated in Delaware in 1982 and controlled by Kanter. On or about May 15, 1983, Kanter directed the grantor trusts to transfer their real estate partnership interests to Cashmere in what was intended as a nontaxable exchange under 26 U.S.C. § 351 in return for Cashmere common and preferred stock. Concurrently with the § 351 transfer, the grantor trusts also transferred to Cashmere eight promissory notes payable to, and held by, the grantor trusts with a face value, in total, of $498,500. The promissory notes were all dated May 1, 1983, and were payable on August 1, 1983. These promissory notes, according to Kanter, had a basis equal to face value and increased the total aggregate basis of the transferred property so as to eliminate the gain that would otherwise have been realized under 26 U.S.C. § 357(c) through the assumption by Cashmere of the negative capital accounts in the transfer of the real estate partnership interests. See 26 U.S.C. § 357(c).²⁸ 

On July 12, 1983, in the next stage of the transfers to Zell-controlled entities, Kanter directed the grantor trusts to sell their common and preferred stock in Cashmere to Waco Capital Corp. in return for promissory installment notes totaling approximately $1.5 million. The grantor trusts reported the income from this sale under the installment method of 26 U.S.C. § 453.

Waco was a Delaware corporation whose sole shareholder was BRT. BRT's beneficiaries were the members of Kanter's family, who were also the beneficiaries of the grantor trusts. The Tax Court found that, under 26 U.S.C. § 453(f), Waco and the grantor trusts were "related persons." See 26 U.S.C. § 318(a) (2), (3).

On August 31, 1983, the promissory notes held by Cashmere were paid off by eight checks drawn on the Administration Co.'s Special E account. As noted, the Special E account contained the commingled funds of a variety of entities, mostly (if not entirely) Kanter entities, and the Administration Co. accounted internally for the specific source of a given disbursement. The Administration Co.'s general ledger, however, did not specify the source of the eight checks paying the notes. The checks themselves indicate, in part, what entity is the payor of the notes. Three of the promissory notes were ostensibly paid with funds from BRT, an entity that was not the maker of the notes in question, and there was no documentary evidence that BRT was advancing the funds to the makers of the promissory notes or had, in some way, assumed the obligations on the notes. IRA, 78 T.C.M. (CCH) at 1110-11.

As of September 1, Waco held all of the stock in Cashmere, and Cashmere's assets included the real estate partnership interests and $498,500 in cash (from the note payments).

Kanter subsequently negotiated the sale of Waco's Cashmere stock to Equity Financial Management Co. (Equity), an entity controlled by Zell. On September 2, 1983, Waco sold the Cashmere stock to Equity for $1,647,500 payable by check.²⁹  Immediately after the purchase, Zell liquidated Cashmere.

The Tax Court found that this convoluted series of transactions taking place within a three-and-one-half month period that ultimately resulted in the transfer of the real estate partnership interests to Zell was arranged principally to avoid immediate recognition of gain on the transfer and otherwise lacked any bona fide business purpose. IRA, 78 T.C.M. (CCH) at 1113; see 26 U.S.C. § 357(b) (1) (A), (B). Therefore, the assumption of liabilities by Cashmere was recognized as money received by Kanter (through his grantor trusts), and he was required to be taxed on his gain resulting from the transfer, up to the full amount of the liabilities. 26 U.S.C. § 357(b). Additionally, the Tax Court found that the transfer of the notes with the real estate partnership interests also lacked any bona fide business purpose and were principally intended as a means to avoid income tax. The Tax Court found that the notes represented mere transfers between and among Kanter-controlled entities, and did not represent true indebtedness. Therefore, Kanter had no basis in the notes, and the notes did not, therefore, increase the aggregate basis in the property transferred to Cashmere. Under § 357(c), the excess of liabilities over basis in the transferred property was therefore recognized as capital gain to Kanter. Because the basis in the partnership interests was zero, under both § 357(b) and (c), the full amount of the transferred liabilities was taxable to Kanter.

Additionally, the Tax Court found that the subsequent sale to Waco using the installment method was a "dispos [ition] of property to a related person." 26 U.S.C. § 453(e) (1) (A). When Waco then subsequently transferred, within two years, the Cashmere stock to Equity, the installment method of reporting was no longer available. 26 U.S.C. § 453(e) (1), (2). The entire amount realized by Waco in the second disposition was treated as received (at the time of the second disposition) by the seller-grantor trusts in the first disposition. Id.

Whether the transactions involving Cashmere had economic substance for federal income tax purposes is a factual question reviewed for clear error. N. Ind. Pub. Serv. Co. v. Comm'r, 115 F.3d 506, 510 (7th Cir. 1997). Similarly, whether the installment method of reporting is available is a factual question that we review for clear error. Applegate v. Comm'r, 980 F.2d 1125, 1128 (7th Cir. 1992).

The tax-free transfer of property to a controlled corporation solely in exchange for the transferee corporation's stock is permitted under 26 U.S.C. § 351. The IRC also provides a means for preserving § 351 eligibility in circumstances where the transferee corporation assumes liabilities of the transferor together with the property transferred (an event that is economically equivalent to a transfer of money (boot) from the transferee corporation to the property transferor). 26 U.S.C. § 357. There are two limitations, however, to tax-free treatment under § 351. First, the taxpayer-transferor must prove by a preponderance of the evidence that the liabilities were not transferred for the principal purpose of tax avoidance and that the transfer had a bona fide business purpose. § 357(b). If the taxpayer fails to clear this hurdle, the assumption of the liability is treated as money received by the taxpayer, and the taxpayer recognizes any gain to the full amount of the liability. 26 U.S.C. §§ 357(b) (1), 351(b) (1) (A). Second, if the amount of the liability transferred is greater than the basis of the property transferred, then, in general, the transferor-taxpayer recognizes a capital gain on the amount by which the liability exceeds the basis of the transferred property. 26 U.S.C. § 357(c) (1).

a. 26 U.S.C. § 357(b)

There can be little question that the Tax Court was correct when it viewed the totality of the convoluted Cashmere transactions and found that their only purpose was the avoidance of federal income tax. Kanter's argument that § 357(b) (1) does not reach this transaction is meritless. Within a four-month period of time an inactive shelf corporation controlled by Kanter (Cashmere) had its stock transferred three times, twice between Kanter-controlled entities. Within that same time span, promissory notes, virtually equal in total value to the total negative capital account balances, were made by entities controlled by Kanter, transferred (along with the real estate partnership interests holding the negative balances) to entities controlled by Kanter and then satisfied by entities controlled by Kanter. Both Cashmere and the promissory notes completed their entire useful life-cycle within the span of the larger, intended transaction — transferring the real estate partnership interests to Zell. Yet neither the promissory notes nor Cashmere functioned to facilitate the transfer of the real estate partnership interests to Zell: the promissory notes were created, transferred, and satisfied before the consummation of the sale to Zell, and Cashmere was liquidated as soon as its stock came into Zell's possession leaving Zell with only the desired asset — the real estate partnership interests. The entire transaction was constructed to avoid the recognition of the gain realized on the assumption of the real estate partnerships' liabilities (by, ultimately, Zell). Therefore, the assumption of those liabilities by Cashmere in the initial stages of the process cannot be described as other than having as its principal purpose the avoidance of federal income tax. See 26 U.S.C. § 357(b).

Kanter's remaining argument with respect to § 357(b) appears to be that the intent of that statutory section was to prevent taxpayers from creating additional liabilities (like personal loans or debts) to be packaged with assets contributed in a § 351 exchange in order to extract the gains contained within the contributed property without recognition for tax purposes. This kind of intent to avoid taxes, argues Kanter, is far removed from the present case where the liabilities were a substantial aspect of the real estate partnership interests to be contributed. This argument fails as well. The fact that the liabilities being contributed were "ordinary business liabilities of the partnerships" does nothing to save this transaction. As noted, the entire business of contributing the partnership interests and their associated liabilities to Cashmere was a transaction whose only function was the avoidance of federal tax. Therefore, "taking into consideration the nature of the liability and the circumstances in the light of which the arrangement for the assumption was made," it is clear that the principal purpose of the assumption of liabilities was the avoidance of tax. 26 U.S.C. § 357(b) (1).

It was not clearly erroneous for the Tax Court to have found that the assumption of the real estate partnership interests' liabilities by Cashmere had as its principal purpose the avoidance of federal income tax.

b. 26 U.S.C. § 357(c)

The Tax Court also found, alternatively, that § 357(c) would apply to the Cashmere transactions.³⁰  Section § 357(c) is triggered when the amount of the liabilities assumed exceeds the basis of the property contributed in the § 351 exchange. The taxpayer is taxed on the amount by which the liabilities exceed the basis. The Tax Court found that the contributed promissory notes did not represent genuine indebtedness and had, therefore, a basis of zero. Without any additional basis from the promissory notes, the basis of the contributed real estate partnership interests was zero, and the amount of the liabilities assumed (in the form of the oft-mentioned negative capital accounts) necessarily exceeded that zero basis. Therefore, the gain realized in the § 351 transfer must be recognized to the full extent of the assumed liabilities.

Kanter vigorously argues that the substance of the promissory notes should be respected and not disregarded as mere form, and the notes should be given a basis equal to their face value. For this argument Kanter relies heavily on the Ninth Circuit's determinations in Peracchi v. Comm'r, 143 F.3d 487 (9th Cir. 1998), where a taxpayer made and contributed notes to his closely held corporation. The Ninth Circuit found substance in Peracchi's promises to pay himself, and increased the basis of property contributed (in a § 351 exchange) to a figure greater than the liabilities assumed by the closely-held transferee corporation. Id.; see also Lessinger v. Comm'r, 872 F.2d 519 (2d Cir. 1989) (finding also that personal note had face-value basis in hands of transferee corporation). The hypothetical risk that this note might somehow become the property of a creditor (through bankruptcy of the corporation) and that a creditor might then be able to enforce the note was sufficient to make the note more than an empty promise. Peracchi, 143 F.3d at 493. The Ninth Circuit panel realized the potential for abuse of this holding, however, and cabined it.

We confine our holding to a case such as this where the note is contributed to an operating business which is subject to a non-trivial risk of bankruptcy or receivership. [The closely held company] is not, for example, a shell corporation or a passive investment company.

Id. at 493 n. 14.

Even if we were to accept Peracchi's underlying premise that a taxpayer's self-made obligations to his own closely held corporation should be respected for tax purposes, in light of the language limiting Peracchi's holding, we would not extend that premise to the present case. Cashmere was a passive investment company that, before the transactions at issue, was a shelf corporation. And after the transfer of the real estate partnership interests to Zell, Cashmere was liquidated. Cashmere was not an operating business, and, unlike Peracchi, where an insurance corporation needed more assets to meet a minimum premium-to-asset ratio for regulatory purposes, there was no underlying business purpose here for these transitory promissory notes. Nor did Cashmere face a "non-trivial risk of bankruptcy." Peracchi is, therefore, quite distinguishable from the present case. The promissory notes here did not represent genuine indebtedness, and the Tax Court did not clearly err in finding that the promissory notes had no basis. Therefore, the Tax Court's conclusion that the liabilities assumed exceeded the basis of the contributed property by the full amount of the liabilities and were taxable to that extent was not clearly erroneous.

2. Disallowance of § 453 Installment Method

The next steps in the sale of the real estate partnership interests to Zell involved, first, the sale of the Cashmere stock from Kanter's grantor trusts to Waco and, second, the subsequent sale within two months of the Cashmere stock from Waco to Equity (the Zell-controlled entity). The grantor trusts reported their gains from the sale to Waco under the installment method allowed by 26 U.S.C. § 453.³¹  But § 453(e) limits the use of the installment method in circumstances where there is a second disposition of the property within two years that is effected by a person "related" to the original seller. In other words, if A sells to B, and then B sells to C (within two years), A cannot report the income from its sale under the installment method if A and B are "related persons" under 26 U.S.C. §§ 267(b) or 318(a). The Tax Court found that the grantor trusts and Waco were "related persons," and that the entire amount of income attributable to the sale of the Cashmere stock to Waco must be recognized in 1983. We affirm this finding as well.

Section § 453(f) (1) defines a "related person" for purposes of determining the reach of § 453(e):

(A) a person whose stock would be attributed under section 318(a) (other than paragraph (4) thereof) to the person first disposing of the property, or

(B) a person who bears a relationship described in section 267(b) to the person first disposing of the property.

26 U.S.C. § 453(f) (1). Waco's stock is owned entirely by BRT, whose beneficiaries were undisputedly the members of Kanter's family. Similarly, Kanter does not dispute that these same family members were the beneficiaries of the grantor trusts. Under 26 U.S.C. § 318(a) (2) (B) (i), the stock owned by BRT is considered owned by its beneficiaries. Therefore, the BRT beneficiaries are treated as owning the Waco stock. Similarly, those same BRT beneficiaries are also beneficiaries of the grantor trusts, and their ownership of the Waco stock is imputed "upstream," under § 318(a) (3) (B) (i), to the grantor trusts. Therefore, the stock of Waco would be attributed to the grantor trusts under § 318(a), making Waco a person "related" to the grantor trusts under § 453(f) (1) (A). The attribution, therefore, makes the sale of the Cashmere stock to Waco ineligible for the installment method under § 453(e).³² 

Similarly, as noted in Part III, Kanter was the substantial owner of BRT under the grantor trust provisions of 26 U.S.C. §§ 671 et seq. Therefore, there is an alternative means by which the grantor trusts and Waco are "related persons."³³  Under § 318(a) (2) (B) (ii), stock owned by a grantor trust is attributed to its grantor; BRT's stock, therefore, is attributed to Kanter. That is, Kanter himself is considered to own the Waco stock owned by BRT. Following a similar statutory path as before, Kanter's status as the substantial owner of the grantor trusts means that any stock he owns personally is attributed to the grantor trusts under § 318(a) (3) (B) (ii). Therefore, Kanter's attributed ownership of the Waco stock means that the grantor trusts also own the Waco stock. Because the grantor trusts are attributed ownership of the Waco stock under § 318(a), Waco and the grantor trusts are "related persons" under § 453(f) (1) (A), and the sale of the Cashmere stock from the grantor trusts to Waco is ineligible for reporting under the installment method by operation of § 453(e).

Whether one, the other, or both methods of analysis are followed, the grantor trusts' sale of Cashmere stock to Waco was ineligible for the installment method of reporting for federal income tax purposes. The Tax Court's findings on this issue were not clearly erroneous.

On May 8, 2001, in the midst of the Tax Court's post-trial Rule 155 computations, attorney Karen Hawkins entered an appearance in the Tax Court on behalf of Kanter's wife, Naomi Kanter (Naomi), for the purpose of claiming that Naomi should not be jointly and severally liable for the Tax Court's determined deficiencies against Kanter. Shortly thereafter, Randall Dick, attorney of record for both Kanter and Naomi, moved to withdraw his representation of Naomi. By her new counsel, Naomi first filed a series of objections to the Commissioner's computations because they held Naomi liable for fraud penalties. The Tax Court upheld these objections. (Order, IRA, 6/20/01; App. at 0346.) Naomi then filed seven motions asking the Tax Court to find that she had not meaningfully participated in the litigation as provided in 26 U.S.C. § 6015(g) (2). In the alternative, Naomi asked that the Tax Court reopen the record in order to hear additional evidence that Naomi was an "innocent spouse" under 26 U.S.C. § 6015(b) and § 6015(f), and not jointly and severally liable with her husband. In response to Naomi's motion, the Commissioner stated that "respondent has no objection to petitioner's first request for relief [a finding that Naomi had not meaningfully participated in the litigation]." (See App. at 0364.) Therefore, the Commissioner felt the alternative request to reopen to hear additional evidence that Naomi was an innocent spouse was moot.

In its September 20, 2001 Order, the Tax Court denied Naomi's seven motions. (See App. at 0224.) The Tax Court found that the language of § 6015(g) (2) required that any finding that a spouse had not "meaningfully participated" in the litigation must occur in a subsequent, separate proceeding that could properly consider the matter now before us as a "prior proceeding." 26 U.S.C. § 6015(g) (2). Therefore, the Tax Court ruled, Naomi would have to wait for a separate proceeding before the Tax Court to have the res judicata effect of the present case adjudicated. Additionally, the Tax Court exercised its discretion not to reopen the record and take additional evidence on the merits of Naomi's claim that she was an "innocent spouse." The Tax Court found that, even though the innocent-spouse provisions had been significantly amended in 1998, and clarified in 2000, the underlying availability of innocent-spouse relief had been unchanged since the origination of the present case in the early 1990s. The court refused to reopen a case whose trial had concluded more than five years earlier, especially when there had been no mention of innocent-spouse relief during the period since the trial. Given the possibility of later relief for Naomi in a subsequent proceeding, the Tax Court did not feel the burden on the Commissioner to reopen the present case and litigate the issue was justified.

This court reviews the denial of motions to reopen the Tax Court's record for abuse of discretion. Coleman v. Comm'r, 16 F.3d 821, 829 (7th Cir. 1994). Whether or not § 6015(g) (2) requires a subsequent proceeding to determine whether Naomi had "meaningfully participated" in the matter now before us as a "prior proceeding" is a question of statutory interpretation that we review de novo. Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 467 (7th Cir. 2000).

The substance of current 26 U.S.C. § 6015 was enacted as part of the Internal Revenue Service Restructuring and Reform Act of 1998, Pub. L. No. 105-206, 112 Stat 685, § 3201 (IRRRA). Technical corrections to the IRRRA were enacted in 2000, leaving us with the statute as it presently appears. See Community Renewal Tax Relief Act of 2000, Pub. L. No. 106-554, 114 Stat. 2763, App. G. Section 6015 contains the so-called "innocent spouse" provisions that allow a spouse to avoid joint and several liability for a tax deficiency assessed against both husband and wife based on a joint tax return filing. To avoid joint and several liability, the innocent spouse must show, generally, that (1) a joint return was filed, (2) the return understated the tax owed based on the "erroneous items" of the other joint filer, (3) she did not know, and had no reason to know, that there was an understatement, (4) it is inequitable to hold her liable for the understatement and (5) she has applied for innocent-spouse protection no later than two years after the Commissioner begins "collection activities." See 26 U.S.C. § 6015(b). Section 6015 expanded previous innocent-spouse provisions by removing the requirement that the understatement be "substantial" and that the return be "grossly erroneous" in order to receive protection. Additionally, § 6015 provided increased protection for divorced or separated spouses by holding such a spouse liable for only those items on which she would have been liable had she filed a separate return. Finally, the modifications to § 6015 provided for equitable relief for an innocent spouse. 26 U.S.C. § 6015; see also John B. Harper, Federal Tax Relief for Innocent Spouses: New Opportunities Under the IRS Restructuring and Reform Act of 1998, 61 ALA. LAW. 204 (2000).

Section 6015 also contemplates the possibility that a court will adjudicate a joint tax liability to completion before an innocent spouse invokes the section's protection. Under § 6015, res judicata will attach to the decision of a court if the innocent spouse "participated meaningfully in [the] prior proceeding," even if the innocent-spouse issue was never presented to the adjudicating court. 26 U.S.C. § 6015(g) (2).

Naomi first argues that the Tax Court should have "determined," under § 6015(g) (2), that she did not meaningfully participate in the Tax Court litigation and that the Tax Court's IRA decision against her husband would have had no preclusive effect with respect to her potential innocent-spouse defense to joint liability. Naomi argues that the record and the Tax Court's decision clearly show that she was not involved in the present case. See IRA, 78 T.C.M.(CCH) at 969 ("Petitioner Naomi R. Kanter, Kanter's wife, was not involved in any of the activities giving rise to this litigation. However, she filed joint Federal income tax returns with Kanter for the years at issue."). She further notes that there would likely not be any delay in the course of the present case with such a determination, that there would be no need to reopen the record in order to make such a determination, and that the Commissioner expressly noticed no objection to such a determination.

Unfortunately for Naomi, the general principles of res judicata and the language of the statute deny her this relief at this time. Section 6015(g) (2) is only properly invoked in a subsequent judicial proceeding to avoid the preclusive effect of a prior judicial determination, and has no application during the pendency of the initial judicial proceeding whose preclusive effect she wishes to avoid. First, we start with the plain language of the statute. Lara-Ruiz v. INS, 241 F.3d 934, 940 (7th Cir. 2001). As the Tax Court observed, the plain language of § 6015(g) (2), which is labeled "Res judicata," limits its effect to consideration of a "decision of a court in any prior proceeding," to determine if an "individual participated meaningfully in such prior proceeding." To us, this language is clear: a decision from a prior proceeding means that this statute is only applicable when the original court proceeding determining tax liability has concluded. We can only conclude that this section is designed to assist innocent spouses in avoiding the preclusive effect of the other spouse's prior, completely adjudicated court case, but that the section has no application internal to the prior judicial proceeding.

This reading gains support from the general principle of res judicata. Res judicata prevents parties from relitigating claims that have already been adjudicated by a court to a final judgment on the merits. Of necessity, res judicata requires two proceedings: an original proceeding wherein a final judgment on the merits is rendered, and a subsequent proceeding wherein a party seeks to litigate again a claim decided in the original proceeding. See N.H. v. Me., 532 U.S. 742, 748, 121 S. Ct. 1808, 149 L. Ed. 2d 968 (2001) ("Claim preclusion [res judicata] generally refers to the effect of a prior judgment in foreclosing successive litigation of the very same claim, whether or not relitigation of the claim raises the same issues as the earlier suit.") (emphasis added); see also BLACK'S LAW DICTIONARY 1305 (6th ed.1990) (defining res judicata as the " [r]ule that a final judgment rendered by a court ... constitutes an absolute bar to a subsequent action involving the same claim.") (emphasis added). Therefore, Naomi cannot have the level of her meaningful participation (or lack thereof) in the present case determined until res judicata becomes relevant in a subsequent proceeding.

In the alternative, Naomi argues that, if she must wait for a subsequent proceeding before § 6015(g) (2) becomes relevant, she wants to have her innocent-spouse defense to joint liability (under 26 U.S.C. § 6015(b) & (f)) adjudicated on the merits during the pendency of the present case. This would require, she argues, reopening the record and allowing her to introduce the necessary evidence to support her innocent-spouse claim, and she further claims that it was an abuse of discretion for the Tax Court not to allow her to do so.

Naomi seeks support in prior Tax Court decisions that, she claims, have "bifurcated proceedings" in order to allow an innocent spouse to have issues of joint liability tried proximately to the general issues of tax liability. (Pet. Br. at 78.) Yet in none of the cases upon which Naomi relies did the innocent spouse remain silent for the duration of the Tax Court proceedings (with over ten years elapsing from Kanter's petition to Naomi's first motion regarding this issue) and, only after the case was all but closed, raise, for the first time, an innocent-spouse defense to joint liability. In Vetrano v. Comm'r, 116 T.C. 272, 2001 WL 423012 (2001), the innocent-spouse defense was asserted in the original petition to the Tax Court. Id. at 274. Likewise, both spouses in Charlton v. Comm'r, 114 T.C. 333, 2000 WL 626760 (2000), asserted in their original petitions that they qualified for innocent-spouse relief. Id. at 338. In King v. Comm'r, 116 T.C. 198, 2001 WL 356124 (2001), the original petition was, in its entirety, an innocent-spouse defense, and the so-called not-innocent spouse was involved as an intervenor. The one consistent thread running through all of these cases relied upon by Naomi is that the request for innocent-spouse relief was squarely before the Tax Court well before the resolution of the case. Naomi, for whatever reason, never put the Tax Court on notice that she had an innocent-spouse defense to joint liability.

Naomi claims her silence was due, in part, to a conflict of interest with respect to the joint representation by counsel of her and her husband. This conflict of interest, she alleges, is the kind of extraordinary circumstance that makes refusing to reopen the record an abuse of discretion. What Naomi fails to argue adequately, however, is that there was an actual conflict of interest. Joint representation, by itself, is not a conflict of interest; the representation of one client must actually conflict with the representation of the other. See United States v. Fox, 613 F.2d 99, 102 (5th Cir. 1980) ("However, an actual, not merely hypothetical or speculative, conflict must be demonstrated before it can be said that an accused has been deprived of effective assistance of counsel."); cf. Dorchester Indus., Inc. v. Comm'r, 108 T.C. 320, 339, 1997 WL 210795 (1997) ("Certainly, one spouse's claim that she (he) is an innocent spouse can present a conflict of interest to counsel trying to represent both spouses. If, indeed, the spouses do have differing interests with respect to any issue in a case, our rules provide that counsel must secure informed consent of the client, withdraw from the case, or take whatever other steps are necessary to obviate the conflict of interest."). Naomi shows no actual conflict in the joint representation of her and her husband during the present case. If Naomi could point us to an innocent-spouse defense that she had at any time during the trial, before a decision by the Tax Court, that would have relied upon arguments in conflict with Kanter's defenses, then she is correct that Tax Court Rule 24(g) may have required separate representation. See Dorchester Indus., 108 T.C. at 339. She has not yet alleged such an argument — in other words, she does not show us that she ever had a viable innocent-spouse defense that she was prevented from raising because it conflicted with Kanter's defense strategy. There does not appear to have ever been a time when she could say that the joint representation faced conflicting interests between Naomi and Kanter. Now, with the Tax Court's having entered a final decision, Naomi would have us engage in hindsight and find that she was inevitably prejudiced by the joint representation because she is now liable for deficiencies for which she is the "innocent spouse." Had Kanter's arguments succeeded in the Tax Court, she would not be facing joint liability nor would she be alleging conflict of interest. We cannot, and will not, automatically conclude that her silence was helpless ignorance and not a strategic decision.

Naomi is not prejudiced by the Tax Court's refusal to reopen the record. The opportunity to assert innocent-spouse defenses to joint and several liability remains fully available to her for up to two years after the Commissioner's first "collection activity." 26 U.S.C. § 6015(b) (1) (E).³⁴  She will have her asserted defense administratively reviewed by the IRS, and, if necessary and appropriate, judicially reviewed by the Tax Court.³⁵ 

The Tax Court's decision not to determine that Naomi had not "participated meaningfully" in the present litigation was proper under § 6015(g) (2), and it was not an abuse of discretion for the court to refuse to reopen the record in order to receive evidence concerning an innocent-spouse defense.

CONCLUSION

In summary:

Issue I — STJ Report. Because we take the Tax Court at its word that in rendering its final opinion it agreed with and adopted the opinion of Special Trial Judge Couvillion, we find Kanter's arguments challenging the Tax Court's refusal to disclose the STJ's "original" report moot. We AFFIRM the Tax Court's denial of Kanter's motions for access to the STJ's report.

Issue II — Fraud. There is significant circumstantial evidence of fraudulent intent. It was not clearly erroneous for the Tax Court to find fraud. We AFFIRM the Tax Court's findings on this issue.

Issue III — BRT. It was not clearly erroneous for the Tax Court to find that Kanter was the grantor of BRT and to affirm deficiencies against him for BRT's income during the years at issue. We AFFIRM the Tax Court's findings on this issue.

Issue IV — Washington Painting. There can be no question that Kanter sought to facilitate the sale of the George Washington painting for profit. The Tax Court found that, in every instance where the potential for profit was involved, Kanter engaged in significant non-law-related business activities with the purpose of facilitating the business opportunities of others for his own profit. Given the number of evidentiary indicators supporting that this venture was also for profit, we believe it was clearly erroneous to find otherwise. We REVERSE the Tax Court's findings on this issue.

Issue V — Bank Deposits. The Tax Court did not clearly err in its findings regarding Kanter's unreported income in 1982 as determined by bank deposit analysis. We AFFIRM the Tax Court's findings on this issue.

Issue VI — Equitable. It was not clear error for the Tax Court to find that the transfers from Equitable Leasing were commissions and were taxable income. We AFFIRM the Tax Court's findings on this issue.

Issue VII — Cashmere. It was not clearly erroneous for the Tax Court to have found that the assumption of the real estate partnership interests' liabilities by Cashmere had as its principal purpose the avoidance of federal income tax. Alternatively, it was not clearly erroneous for the Tax Court to conclude that the promissory notes did not represent genuine indebtedness and had no basis, and, therefore, that the liabilities assumed exceeded the basis of the contributed property by the full amount of the liabilities and were taxable to that extent. Finally, the Tax Court did not clearly err in finding the grantor trusts' sale of Cashmere stock to Waco ineligible for the installment method of reporting for federal income tax purposes. We AFFIRM the Tax Court's findings on this issue.

Issue VIII — Naomi Kanter. We believe the Tax Court's interpretation of § 6015(g) was correct, and that Naomi must wait for a subsequent proceeding before she can have the level of her participation in the present case determined. We also do not believe that it was an abuse of discretion for the Tax Court to refuse to reopen the present case to allow Naomi to litigate her innocent-spouse defense. We AFFIRM the Tax Court on this issue.

For the foregoing reasons we AFFIRM in part and REVERSE in part the decision of the Tax Court.

CUDAHY, Circuit Judge, concurring in part, dissenting in part.

I concur with the majority's opinion as to Naomi Kanter's issues on appeal, but dissent otherwise. I write separately to address Kanter's threshold argument that the Special Trial Judge's ("STJ's") original report must be made a part of the record on appeal so that this court can determine whether its contents have been adequately considered by the Tax Court judge, whose opinion is before us.

Before I begin the legal analysis that, I believe, demonstrates why the withholding of the report is improper, I want to take a few lines to address the policy concerns that leap to mind when first encountering the suppression of the report. For the Tax Court is "not" merely "unusual;" it is, I believe, unique among all the institutions in the law where one official conducts a trial (and thus hears the witnesses) and prepares a report or other document containing her findings or recommendations based on the trial, and another official or group of officials subsequently makes the operative decision. Even the Commissioner, at oral argument, acknowledged that in every milieu except that of the Tax Court, the document containing the findings or recommendations of the official conducting the trial are available to a court reviewing the operative decision. This includes, for example, the report of a federal magistrate judge to a district court responsible for a decision. This is also the practice under the Administrative Procedure Act which governs practically all federal administrative proceedings and where the hearing officer (usually an administrative law judge) must file a recommended decision which is distributed to both parties, any appellate court conducting a review and to the public at large. Transparency is the universal practice of agencies and courts employing these decisional practices. The question then becomes, if there are policy reasons that dictate transparency for everyone else, why do these reasons not apply to the Tax Court? The Tax Court has not denied that a document containing the original findings of the STJ exists, yet it refuses to include this document in the record on appeal. It is no answer at all to claim that the report of the STJ is like a law clerk's memorandum to a judge or the memorandum of a fellow jurist on a panel — an internal privileged decisional document. The document here is of an official who presided over the trial and heard the witnesses, and it is directed towards an official who has no first hand knowledge of any aspect of that same trial. I am not impugning the integrity of the Tax Court judges here, or at any point in this dissent; I am merely questioning the propriety of their denial of procedural transparency in a circumstance where every other like process known to the law is transparent. If we approve the Tax Court's practice here, are we not suggesting to the whole administrative array of the federal government that it may seek by available means when statutes permit to deny transparency when engaging in like decisional processes? I believe that the legal analysis of the majority as well as my own must be examined in the context of the larger implications of allowing an administrative body (technically, an Article I court) like the Tax Court to flout the otherwise ubiquitous principle of transparency in its proceedings. That said, I believe there also exist sound constitutional grounds demanding transparency in this instance.

As a threshold matter, everyone agrees that Kanter's arguments are immaterial if the Tax Court's opinion is the verbatim reproduction of the STJ's report, which the majority and the Eleventh Circuit, Ballard v. Commissioner, 321 F.3d 1037 (11th Cir. 2003), appear to believe is the case, and which the Tax Court's opinion superficially purports to be. The Tax Court's opinion clearly states that it adopts and agrees with the "opinion" of the STJ. If that recital is to be interpreted as meaning that the STJ's initial report lies before us already, then there is no issue of the Tax Court judge's according due regard or a presumption of correctness to the STJ's findings. Deference would not be an issue if there has been outright adoption of the STJ's findings. This state of affairs would also appear to moot Kanter's due process argument as well as his Rule 183 argument. Kanter relies on United States v. Raddatz, 447 U.S. 667, 100 S. Ct. 2406, 65 L. Ed. 2d 424 (1980), to argue that an STJ should be treated the same as a magistrate judge. Using Raddatz, Kanter argues that when the Tax Court reverses an STJ's credibility findings without having heard the witnesses personally, due process is violated. However, the verbatim adoption of the STJ's findings by the Tax Court would fully comport with even Kanter's interpretation of Raddatz's due process requirements.

But I agree with Kanter that when the Tax Court "agrees with and adopts the opinion of the Special Trial Judge," it does not mean that the Tax Court opinion is the verbatim reproduction of the original STJ's report. Kanter argues that the Tax Court routinely reviews and alters STJ reports through an internal process that is concealed in published Tax Court opinions by the language, "agrees with and adopts." Kanter presented two pieces of evidence to support his claim: 1) Kanter's attorney allegedly was told informally by Tax Court Judge Julian Jacobs and Chief Special Trial Judge Peter J. Panuthos that the credibility findings of Special Trial Judge Couvillion on fraud were reversed by Tax Court Judge Dawson;¹  and 2) there exists not a single Tax Court decision since the adoption of current Rule 183 where a Tax Court Judge has purported to modify or reverse a finding of a Special Trial Judge.

What Kanter alleges happened in the present case, and what commonly occurs in the Tax Court, is that a Tax Court judge takes the STJ's report (which had been filed with the Chief Judge pursuant to Rule 183) and works together with the STJ to edit it. From this process emerges a final report that may or may not bear any resemblance to the original report, but that still may be called the STJ's "opinion" (but not the STJ's "report") if the STJ agrees to subscribe to it. This modified report is then "adopted" by the Tax Court judge and filed as the Tax Court opinion. This is the reason, Kanter argues, that, in the 880-plus Tax Court decisions since 1983 that I could find that involved an STJ report, the Tax Court judge purported to agree with and adopt the opinion of the STJ in every instance. Never, in any instance since the adoption of the current Rule 183 that I could find, has a Tax Court judge not agreed with and adopted the STJ's opinion.

I find this extraordinary unanimity telling. It is difficult to believe that over the course of nineteen years (since the amendments giving rise to current Rule 183), not a single Tax Court judge (and there are 19 of them, 26 U.S.C. § 7443(a)) has ever disagreed with a single original finding of any STJ (and there are about 20 of them). I say with confidence that this degree of unanimity is not only unusual, but impossible in a system of arms-length appellate style review involving 39 independent individuals. I believe that it is highly likely, therefore, that there is some kind of collaborative process involved in the path from STJ report to Tax Court decision. I draw support in this conclusion from the fact that neither the Commissioner nor the Tax Court has ever settled this issue by unambiguously stating otherwise, despite opportunities to do so. Notably, at oral arguments on the present case the Commissioner did not dispute Kanter's contention that the STJ's report undergoes some kind of revision during the process of "adoption" by the Tax Court. And in his brief the Commissioner is very careful in stating that the Tax Court adopted the "opinion of the [STJ]." Resp. Br. at 111 (emphasis added). And the Commissioner is just as careful in never stating that the Tax Court adopted the STJ's "report." This care mirrors the Tax Court's own language in its refusal to release the STJ's report to Kanter.

I believe that the record supports the notion that the Tax Court engages in a quasi-collaborative process of review of the STJ's report from which a new and frequently different STJ's opinion emerges to be adopted and agreed with by the Tax Court. If my understanding is correct, there are two "STJ's reports" in many, if not most (or even all), Tax Court cases — the original "report" filed under Rule 183 with the Chief Judge of the Tax Court, which is solely the work product of the STJ (and which represented the STJ's views at the end of trial) and the later "opinion" of the STJ, which is a collaborative effort, but which the Tax Court then "agrees with and adopts" as the opinion of the Tax Court. In any event, I do not claim to know the degree to which the STJ's original filed report in the present case was altered, and I do not take as determinative of that fact the declaration of Kanter's attorney regarding his conversations with Tax Court personnel. So, given this interpretation of the Tax Court's procedure, I want to take a closer look at Kanter's argument.

Kanter's argument wraps a number of different issues into one request for the STJ's report. First, whether or not the Tax Court's procedure denying access to the STJ's report violates its own Rule 183. Second, whether the Tax Court's procedure violates other law, including the Internal Revenue Code ("IRC"). Third, whether the Tax Court's procedure violates due process protections. Finally, and most importantly, whether Kanter's due process rights on appellate review by this court are violated when we undertake review of the Tax Court's decision without the context of the STJ's original report with respect to credibility findings.

1. Does the Tax Court's procedure violate Tax Court Rule 183?

I agree with the majority's determination that Rule 183 imposes no requirement of disclosure or of clearly erroneous deference upon the Tax Court. However, I think some additional discussion of the evolution of Rule 183 into its current form, and why the current rule does not compel production of the STJ's report nor require any particular deference to the STJ's report, would be very informative. Kanter points us to Stone v. Commissioner, 865 F.2d 342 (D.C. Cir. 1989), where the Court of Appeals for the District of Columbia Circuit found that an STJ's findings should be reviewed by the Tax Court under a clearly erroneous standard. Id. at 347. Under a prior version of the Tax Court's rules (pre-1983 and which governed the case before the Stone court), the STJ's report was served on each party and each party had an opportunity to file objections to the report's findings. See Tax Court Rule 182(b), (c), 60 T.C. 1149, (1973). The Stone court's finding that the STJ's report was owed deference followed from the Tax Court rules, Rule 182(d) at the time, which stated that " [d]ue regard shall be given to the circumstance that the [STJ] had the opportunity to evaluate the credibility of witnesses; and the findings of fact recommended by the [STJ] shall be presumed to be correct." Tax Court Rule 182(d), 60 T.C. 1150, (1973). The explanatory notes to this rule prescribed that, in regard to the "special weight" to be given the STJ's findings, one should look to Court of Claims Rule 147(b). Id. The Stone court found this prescription particularly instructive because Rule 182(d)'s language was lifted practically verbatim from Court of Claims Rule 147(b). Stone, 865 F.2d at 345. At the time that the language of Court of Claims Rule 147(b) had been adopted for the Tax Court's rules, the Court of Claims interpreted that language to require review of the findings of its version of an STJ's report under a clearly erroneous standard. See, e.g., Elmers v. United States, 172 Ct. Cl. 226, 232 (1965). The Stone court found that the language of the rule, the rule's command to look to the Court of Claims and the Court of Claims' use of a clearly erroneous standard of review required the use of such a standard in the Tax Court's review of STJ findings. Stone, 865 F.2d at 347. However, the Stone court went on to note that

The Tax Court is of course free to make its own rules determining the relation between it and its Special Trial Judges. Moreover, we assume that the Tax Court's construction of its own rules enjoys the deference, on review in this court, enjoyed by an administrative agency interpreting its own regulations.

Id. The Tax Court did exactly that — changed its rules.²  In 1983, the Tax Court amended and redesignated the rule in question, adopting its current form as Rule 183,³  and noted that " [t]he prior provisions for service of the [STJ]'s report on each party and for the filing of exceptions to that report have been deleted."⁴  81 T.C. 1070, (1983). The Tax Court has never documented any explanation of why this amendment was undertaken. The Tax Court's power to prescribe its own rules of procedure is undisputed. 26 U.S.C. § 7453. And although the Tax Court is no longer an executive agency, see Freytag v. Commissioner, 501 U.S. 868, 887-88, 111 S. Ct. 2631, 115 L. Ed. 2d 764 (1991) (noting that in 1969 Congress removed the Tax Court from the realm of executive agencies and made it an Article I court), it is clear, and Kanter does not dispute, that the Tax Court's interpretation of its own rules of procedure receives a great deal of deference. Therefore, the 1983 amendment to the Tax Court rules had the effect of no longer requiring that parties (or the general public or a reviewing court, for that matter) have access to the STJ's report.

Kanter notes, however, that the language of the earlier rule that prompted the clearly erroneous standard in Stone remains unchanged. The Tax Court judge still "shall" give " [d]ue regard" to the fact that the STJ had the opportunity to hear and evaluate the credibility of the witnesses, and the STJ's recommended findings of fact still "shall be presumed to be correct." Tax Court Rule 183(c). This language, according to Kanter, still commands the Tax Court judge to whom the STJ's report is submitted to adopt the STJ's findings unless the findings are clearly erroneous.⁵  And there certainly appears to be some consensus in the literature that the Rule still embodies a clear error standard. See, e.g., 35 Am.Jur.2d Fed. Tax Enforcement § 905 (2002) ("The Tax Court is required to review a special trial judge's factual findings according to the clearly erroneous standard, and cannot overturn a special trial judge's ruling on the basis that the Tax Court finds the testimony credited by the trial judge to be unbelievable."); 20A Federal Procedure, L. Ed., Internal Revenue § 48:1274 (2000) (same); but see Tax Court Litigation, 630-2nd Tax Mgmt. Portfolio at A-49 n. 599 (1997) ("The D.C. Circuit (but not the Tax Court) has taken the position that the level of deference is to review the Special Trial Judge's draft opinion on a `clearly erroneous' standard."). However, I agree with the majority's conclusion that this is no longer the case, and I have nothing additional to add to its reasoning on the matter.

Therefore, like the majority, I also do not believe that Tax Court Rule 183 requires an STJ's report to be reviewed under a clearly erroneous standard, nor that Rule 183 is violated by a quasi-collaborative process of revision of an STJ's report, nor that the Rule requires the production of the report. In spite of this I must again note how remarkable it is that not only is the Tax Court unique in the opacity of its process, but it has arrived at this opaque process by abandoning a transparent process — an evolution completely counter to the trend towards transparency in analogous areas of the law. See, e.g., Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2331-32 (2001) (describing new theory of administrative control in which author notes that transparency is a core value of administrative procedure). There is no public indication why the Tax Court rules and procedures were changed in 1983 to the current system that disadvantages those appealing Tax Court decisions. And, of course, an appellate-style procedure such as that typical in all other areas of federal administrative adjudication would facilitate challenges, whether made by the taxpayer or by the Commissioner. The previous procedure may well have been abrogated for exactly this reason.

2. Is the Tax Court's procedure otherwise unlawful?

The majority quite ably and clearly outlines why existing rules and statutes do not appear to compel inclusion of the STJ's report in the record on appeal. In doing so, the majority places significant weight on the Commissioner's argument analogizing the STJ-Tax Court judge relationship to the division-Tax Court relationship governed by 26 U.S.C. § 7460, in which a division's preliminary report is never made public if the Tax Court reviews the case and issues its own opinion. Before moving on to the meatier due process issues, I want to note that this analogy, however, overlooks some important considerations. First, a division whose report is reviewed by the Tax Court has the opportunity to file a dissent from the Tax Court's final decision and place in that dissent any of its objections — objections that could, in theory, include the division's overruled findings that were contained in its original report. Thus, those original findings can be made public, albeit in a roundabout manner, if the division wants them to be. An STJ's original report is never made public.

Second, § 7460 differs significantly from Rule 183 in that it does not require "due regard" for the fact that the division has heard witnesses and evaluated credibility (perhaps because the division may not have been the adjudicator who heard the witnesses), nor does § 7460 require any presumption that the division's report is correct. In contrast, Rule 183 requires both due regard and a presumption of correctness for the STJ's report. Therefore, to the extent that the Tax Court's final opinion must, under Rule 183, accord some kind of respect to the STJ's original findings, the STJ's original report has some ongoing significance, whereas the division's report is of no consequence in the formulation (or appellate review) of the Tax Court's opinion.

Third, since the division may not have conducted the trial nor heard the witnesses, a policy of deference to such an adjudicator would not necessarily be appropriate. By contrast, an STJ is always the person who hears witnesses and, for that reason, is an adjudicator entitled to deference. Again, this raises the significance of the original STJ's report to the ultimate adjudication and review of the case in a way not present with a division's report.

And finally, the relationship between a division and the Tax Court is far different from the relationship between an STJ and the Tax Court. Tax Court judges are all, essentially, equal. They are presidentially appointed for statutorily mandated 15 year terms, and they each have an equal vote in the business of the Tax Court. 26 U.S.C. §§ 7443, 7444. An STJ is appointed at the discretion of the Chief Judge and has no statutorily mandated term of office. 26 U.S.C. § 7443A. Congress has authorized specific and limited means for removing a Tax Court judge from office: "Judges of the Tax Court may be removed by the President, after notice and opportunity for public hearing, for inefficiency, neglect of duty, or malfeasance in office, but for no other cause." 26 U.S.C. § 7443(f). There is no such statutory protection for STJs. Ultimately, keeping a division's preliminary report secret appears less problematic given the division's almost unfettered ability to make its wishes clearly known in the final opinion without concern for job security or reprisal. However, an STJ serves at the discretion of the Tax Court, and his or her judicial independence is therefore quite circumscribed. Only by allowing access to the original STJ's report can the Tax Court insulate itself from the perception that an STJ's "findings" are arbitrarily malleable at the discretion of the Tax Court.⁶ 

In addition, there are other provisions of the Internal Revenue Code that, at the very least, show that Congress did not demonstrate a clear intent to keep STJ reports secret — unlike the clear intent to keep division reports private that is shown in § 7460. 26 U.S.C. § 7459(b) states that the "Tax Court shall report in writing all its findings of fact, opinions, and memorandum opinions." Additionally, 26 U.S.C. § 7461(a) states that "all reports of the Tax Court ... shall be public records," and 26 U.S.C. § 7462 states that the "Tax Court shall provide for the publication of its reports at the Government Printing Office in such form and manner as may be best adapted for public information and use...." While the majority is correct that there are no rules or statutory sections that specifically require that the initial report of the STJ be made public, there are, by the same token, no sections that forbid that the report be made public — in contrast to § 7460's clear intent to keep private a division report that is reviewed by the full Tax Court. And §§ 7459, 7461 and 7462 appear to establish a strong presumption in the IRC in favor of public dissemination of Tax Court documents that, apparently, could easily apply to the STJ's report. Only a formalistic interpretation by the Tax Court that the STJ's report is not a "report of the Tax Court" allows it to avoid such a presumption.

3. Does the Tax Court's procedure violate due process?

Due process requires that Kanter have been afforded a fair hearing before he is deprived of property. Mathews v. Eldridge, 424 U.S. 319, 333, 96 S. Ct. 893, 47 L. Ed. 2d 18 (1976). Notice and an opportunity to be heard are the hallmarks of a fair hearing. Mullane v. Central Hanover Bank & Trust Co., 339 U.S. 306, 313, 70 S. Ct. 652, 94 L. Ed. 865 (1950). The Supreme Court in Raddatz reiterated the three part test announced in Mathews for evaluating due process protections:

[T]hree factors should be considered in determining whether the flexible concepts of due process have been satisfied: (a) the private interests implicated; (b) the risk of an erroneous determination by reason of the process accorded and the probable value of added procedural safeguards; and (c) the public interest and administrative burdens, including costs that the additional procedures would involve.

Raddatz, 447 U.S. at 677, 100 S. Ct. 2406 (citing Mathews, 424 U.S. at 335, 96 S. Ct. 893). In the context of the present case, one would need to determine whether a quasi-collaborative process wherein the ultimate finder of fact, who has not heard the witnesses herself, can amend, revise or reverse the preliminary findings of the person who actually heard the witnesses (and never reveal those preliminary findings) without running afoul of the Fifth Amendment. This comprises two separate questions: (1) must the Tax Court review the STJ's findings with a formal degree of deference (such as clear error); and (2) must the Tax Court itself hear witnesses to determine issues of credibility before reversing the STJ?⁷ 

As Raddatz and Universal Camera make clear, at one end of the due process spectrum — the general administrative law context — due process does not require that the ultimate fact finder be constrained by a formal degree of deference to the original hearing officer. Nor must that fact finder rehear witnesses before making findings, whether or not the fact finder reverses the original hearing officer's findings. See id. at 680, 100 S. Ct. 2406 ("Generally, the ultimate factfinder in administrative proceedings is a commission or board, and such trier has not heard the witnesses testify.... While the commission or board ... may defer to the findings of a hearing officer, that is not compelled."); Universal Camera Corp. v. NLRB, 340 U.S. 474, 492-94, 71 S. Ct. 456, 95 L. Ed. 456 (1951); see also 5 U.S.C. § 557(b) ("On appeal from or review of the initial decision, the agency has all the powers which it would have in making the initial decision...."); Kenneth Culp Davis & Richard J. Pierce, Jr., Administrative Law § 8.6, at 396 (3d ed.1994) (noting that the command of Morgan v. United States, 298 U.S. 468, 481, 56 S. Ct. 906, 80 L. Ed. 1288 (1936), that "the one who decides must hear," "did not mean that an agency head who decides must listen to the witnesses testify").

At the other end of the due process spectrum lies the criminal procedure context, where due process protections are the most demanding. First, regarding deference, the Supreme Court in Raddatz did not directly address the issue, but the Federal Magistrates Act under review required the district court to conduct a de novo review of the magistrate judge's findings, and the Court took no issue with that standard of review, even for credibility findings. This parallels the administrative context, and demonstrates, I believe, that along the full continuum of due process concerns framed by Raddatz and Universal Camera, there is no per se due process violation when the ultimate finder of fact reviews preliminary findings de novo. Therefore, I agree with the majority that the Fifth Amendment does not require that the Tax Court review STJ findings using any particular degree of deference. This means also that there is no constitutional requirement that the Tax Court use an appellate-style review of its STJs' reports. In this respect, the quasi-collaborative model adopted by the Tax Court is permissible.

Second, what about the rehearing of witnesses in a criminal procedure context? While, under the Mathews analysis, the interests of a criminal defendant in a suppression hearing are not as significant as they may be in a full criminal trial, they are significant enough that the Supreme Court issued a warning that "serious questions" existed in the situation wherein a district court judge reversed a magistrate judge's dispositive credibility findings without hearing the witness herself. Raddatz, 447 U.S. at 681 n. 7, 100 S. Ct. 2406.⁸  Several of our sister circuits have found that these "serious questions" have a single answer: a district court judge cannot reverse a magistrate's credibility findings without hearing the witness at issue. See, e.g., United States v. Cofield, 272 F.3d 1303, 1305-06 (11th Cir. 2001); Hill v. Beyer, 62 F.3d 474, 482 (3d Cir. 1995); United States v. Rosa, 11 F.3d 315, 328-29 (2d Cir. 1993); In re Hipp, Inc., 895 F.2d 1503, 1519-21 (5th Cir. 1990); see also United States v. Mejia, 69 F.3d 309, 316-20 (9th Cir. 1995) (finding that footnote 7 of Raddatz applied to a suppression hearing where the judge making the ruling received no findings on credibility from the judge who heard the witnesses' testimony; the court ruled it was a due process violation to make such a ruling without having heard the witnesses).

However, as the Commissioner points out, the interests in Raddatz were more significant because that case involved an aspect of a criminal trial, and the proceedings before the Tax Court are eminently civil. Under the Mathews framework, the private concerns involved in a civil proceeding are not entitled to the same level of due process protection as the concerns in a criminal proceeding, just as Raddatz notes that within a criminal proceeding a suppression hearing embodies a lower interest than other aspects of a criminal trial. See Bristol-Myers Squibb Co. v. McNeil-P.P.C., Inc., 973 F.2d 1033, 1045 (2d Cir. 1992) ("Moreover, we have indicated that the Raddatz dicta may be inapplicable outside the criminal context....").

While it is not an easy issue, I believe that the interests at stake in a civil tax court proceeding do not rise to the level addressed in Raddatz, and are more analogous to the interests involved in an administrative adjudication. There certainly is some cause for concern that a finding of credibility on a "cold record" as voluminous as the one before us here increases the chance of an erroneous determination, but I do not believe that that possibility in this kind of civil proceeding is as high, nor the costs as great, as would be the case in a criminal milieu. Additionally, I would note that the only fully responsive remedy would be to require the Tax Court itself to rehear the witnesses whose credibility was at issue. Under the third prong of the Mathews test this added procedure would probably be an enormous burden and impose a prohibitive cost. The added value of such procedure under the second prong of Mathews seems insubstantial, especially given that the quasi-collaborative model can provide the Tax Court with ongoing access to the thoughts and impressions of the STJ who actually heard the witnesses. Additionally, I believe that one possible advantage of the quasi-collaborative process (over standard appellate-style review) might be an opportunity for the STJ to have additional input into the decision making process beyond her original report.⁹  This opportunity should do something to minimize the risk of an erroneous determination.

Hence, the dictum in Raddatz does not persuade me that the reversal of an STJ's findings by the Tax Court following a quasi-collaborative procedure violates due process. In a real sense, my writing to this point has been more concurrence than dissent. But I feel it very important to navigate these areas of agreement in order to properly prepare for the pivotal area of disagreement, where I part ways with the majority.

4. Does appellate review of the Tax Court's findings without access to the STJ's report violate due process?

There is another stage of procedure involved here (and a key one from my perspective) that requires due process analysis — appellate review of the Tax Court's decision. Whether or not the STJ's report is made available to the parties for comment before the Tax Court issues its findings, and whether or not the Tax Court can reverse the STJ's purportedly dispositive credibility findings without having heard the relevant witnesses, the question still remains whether or not the due process rights of the parties before this court are violated when we have no opportunity to review the Tax Court's factual findings for clear error in light of the STJ's initial report. See Evitts v. Lucey, 469 U.S. 387, 393, 105 S. Ct. 830, 83 L. Ed. 2d 821 (1985) (there is no constitutional right to appeal, but once a right of appeal is created, it must comport with due process to be meaningful and effective). This question distinguishes the issue whether the Tax Court's procedures are intrinsically unfair (which neither I nor the majority believe is true) from the issue whether the Tax Court's procedures are unreviewable (on which the majority and I disagree).¹⁰  The essential difference is that the Tax Court has the report; we do not.

Our review of the Tax Court is governed by the same standards as those governing our review of a district court's civil bench trial; this means that legal conclusions are reviewed de novo and findings of fact are reviewed for clear error. See 26 U.S.C § 7482(a); Pittman v. Commissioner, 100 F.3d 1308, 1312-13 (1996). Obviously, we do not need access to the STJ's report to conduct meaningful de novo review of the Tax Court's legal conclusions. But clearly erroneous review involves deference to the conclusions of the fact finder — the Tax Court in the present case. This is a deference that the Supreme Court has attributed, in the case of credibility, to the fact finder's first-hand observations of the witnesses in question.

[A] finding is "clearly erroneous" when although there is evidence to support it, the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.

....

When findings are based on determinations regarding the credibility of witnesses, [the clearly erroneous standard] demands even greater deference to the trial court's findings; for only the trial judge can be aware of the variations in demeanor and tone of voice that bear so heavily on the listener's understanding of and belief in what is said.

Anderson v. City of Bessemer City, N.C., 470 U.S. 564, 573, 575, 105 S. Ct. 1504, 84 L. Ed. 2d 518 (1985) (citations omitted). Thus, it is integral to the standard of clear error review that there be deference to the credibility findings of the official who has actually heard the witnesses. Although the Supreme Court in Anderson was discussing the deference due a finder of fact who has, himself, heard the witnesses, I think the Court's command is also instructive in two ways. First, on its face Anderson instructs that on issues of credibility the opportunity to hear witnesses is significant in a clear error context. Second, Anderson informs that context by impliedly undermining the reliability of findings that reverse the credibility determinations of an official who has actually heard the witnesses. If we are to give "even greater deference" to the findings of a judge who has heard the witness whose credibility is at stake, we must inevitably give less deference to the judge who subsequently reverses those findings.

I find major support for this line of thinking in the administrative law arena.¹¹  In the administrative context, the Administrative Procedure Act requires that a reviewing court examine an agency determination based on a record that includes any preliminary findings from the hearing officer (like an Administrative Law Judge (ALJ)).¹²  5 U.S.C. §§ 557(b), 706. When a reviewing court reviews agency findings on credibility for substantial evidence, it is strongly influenced by the preliminary findings of the ALJ who actually heard the witnesses — influence that becomes even more significant when an agency has reversed those preliminary credibility findings. See Kopack v. NLRB, 668 F.2d 946, 953 (7th Cir. 1982) ("One must attribute significant weight to an ALJ's findings based on demeanor because neither the [NLRB] nor the reviewing court has the opportunity similarly to observe the testifying witnesses."); Moore v. Ross, 687 F.2d 604, 609 (2d Cir. 1982) ("Accordingly, reviewing courts have often found federal decisions unsupported by substantial evidence when they hinge on assessments of credibility contrary to those made by the ALJ who heard the witnesses."); Ward v. NLRB, 462 F.2d 8, 12 (5th Cir. 1972) ("The preeminence of the Examiner's conclusions regarding testimonial probity does not amount to an inflexible rule that either the Board or a reviewing court must invariably defer to his decision, thereby effectively nullifying either administrative or judicial review. But when the Board second-guesses the Examiner and gives credence to testimony which he has found — either expressly or by implication — to be inherently untrustworthy, the substantiality of that evidence is tenuous at best.").

The Supreme Court in Universal Camera best summarized the philosophy behind this process: "We intend only to recognize that evidence supporting a conclusion may be less substantial when an impartial, experienced examiner who has observed the witnesses and lived with the case has drawn conclusions different from the Board's than when he has reached the same conclusion." 340 U.S. at 496, 71 S. Ct. 456. This is never more the case than when the issue is one of credibility. As the Supreme Court said in Raddatz, it is within the context of credibility that findings based on a "cold record" are most suspect if they differ from the findings of the one who actually heard the witnesses in question.

To be sure, courts must always be sensitive to the problems of making credibility determinations on the cold record. More than 100 years ago, Lord Coleridge stated the view of the Privy Counsel that a retrial should not be conducted by reading the notes of the witnesses' prior testimony:

"The most careful note must often fail to convey the evidence fully in some of its most important elements.... It cannot give the look or manner of the witness: his hesitation, his doubts, his variations of language, his confidence or precipitancy, his calmness or consideration;... the dead body of the evidence, without its spirit; which is supplied, when given openly and orally, by the ear and eye of those who receive it." Queen v. Bertrand, 4 Moo.P.C.N.S. 460, 481, 16 Eng.Rep. 391, 399 (1867).

Raddatz, 447 U.S. at 679-80, 100 S. Ct. 2406. And unlike Raddatz, in the present case we are dealing with a full trial by a judge on the merits, albeit a civil trial (on the "quasi-criminal" issue of fraud), and not an ancillary motion to suppress. Additionally, the present case was inordinately long and complicated, and the resolution of issues required the synthesis of multiple witnesses' testimony that was separated by days or even weeks (and by hundreds or thousands of pages in the transcript). Whatever advantage is to be gained by a first-hand observation of witnesses is multiplied exponentially when the trial is so long and the transcript so voluminous. The detailed interconnection of the credibility of different witnesses on different factual issues makes the accumulated impressions of the presiding officer irreplaceable. I can think of no single item of more significance in evaluating a Tax Court's decision on fraud than the unfiltered findings of the STJ who stood watch over the trial.

The difficulty comes in determining how and when this concern rises to the constitutional level of due process. No court of which I am aware has ever considered the ramifications of an agency's swallowing and refusing to regurgitate a preliminary factual finding in the manner done by the Tax Court here. The reason for this is clear: the Administrative Procedure Act requires the publication of such findings for executive agencies. And no court that I could find has ever discussed the availability of preliminary findings as being related to due process protections. Only Universal Camera's dicta on the value of the preliminary findings comes close, and the Supreme Court there clearly based its decision in the language of the Administrative Procedure Act.

However, under the three-part test of Mathews and Raddatz, I think it not unreasonable to invoke due process in this context at the appellate court level of review. Under the first prong of Mathews, I note again that the quasi-criminal nature of fraud is a more significant private interest than a simple civil determination, but not as weighty an interest as in the criminal suppression hearing in Raddatz. Under the second prong, however, I believe the risk of error here is greater and the value of the added procedural safeguard (the STJ's report) is higher in this context. Without the STJ's report we would be reviewing deferentially a credibility finding made by the Tax Court based on a cold record (albeit with the theoretical collaboration of the STJ who actually heard the witnesses) based on our own analysis of that same cold record. The precedents noted in the administrative context (as well as Universal Camera) clearly demonstrate how valuable preliminary findings are for review in cases like the present one. Under the Mathews test's third prong, the added cost and administrative burden in this instance is de minimis — publishing the STJ's report. On balance, I believe strongly that the absence of the STJ's report in the record for our consideration of the Tax Court's decision creates legitimate due process concerns with respect to our review.

What throws a real analytical monkey wrench into all of this is that the result of the collaborative process in the Tax Court is an opinion that, allegedly, represents the "opinion" of the STJ (but not, I repeat, the "report" of the STJ). What this could mean is that the collaboration of the Tax Court with the STJ has produced an opinion that represents the revised and true legal opinions and findings of the STJ. The original report, by necessity therefore, would no longer be a valid statement of the STJ's findings inasmuch as it differed from the final opinion. Therefore, to the extent that due process is concerned with the changes that the Tax Court makes to the findings contained in the STJ's report (which would represent points of disagreement between the Tax Court and the STJ), those concerns are balanced by the fact that the STJ does not actually disagree with those changes and, in fact, has certified to their correctness by signing off on the opinion. I could, in the best of all possible worlds, liken this to a learning process whereby the original impressions of the STJ are tempered through the collaborative process with the Tax Court, and, I would assume, the Tax Court's opinions would be molded and informed by the first-hand impressions of the STJ. Whatever limitations on review this process entails would be balanced by the fact that the STJ does not still hold opinions or findings in conflict with those represented in the opinion. In such a world, it no longer seems so strange that every Tax Court case involving an STJ resulted in an opinion that agreed with and adopted the opinion of the STJ — the final opinion represents a compromise between the positions both of the Tax Court judge and of the STJ. (This view of things would still remain somewhat disingenuous because the Tax Court opinion's language clearly seeks to imply that the opinion represents the original report of the STJ.) I believe that this is, more or less, the position taken by the majority.

The obvious rejoinder to this utopian view of the Tax Court's process is to point out what I have already noted: the STJs are not equal to the Tax Court judges and it might be naive to assume that the STJs have an equal voice in the collaborative process that results in Tax Court opinions. The result is a system whereby the Tax Court maintains total discretionary control over the function of STJs but expects a reviewing court to simply accept at face value the declaration that the opinion is the opinion of the STJ. On the one hand, I certainly do not believe that the Tax Court is prevaricating and forcing STJs to cooperate under the express threat of unemployment. However, judicial independence in the context of due process is not a principle to be taken lightly, and its absence has consequences. The fact is that this entire process of the Tax Court appears designed to extract the efficiencies involved in designating cases to be heard by STJs without having to bear any of the procedural costs traditionally associated with this kind of adjunct decision-making (e.g., transparency). I do not believe that the STJ's ultimate assent to the final opinion of the Tax Court is protection enough to the parties. The majority does.

Because I dissent, I do not have to articulate a final outcome, but merely note that I find a due process violation. However, the solution is simple in theory. The due process violation is avoided by interpreting 26 U.S.C. §§ 7459, 7461 and 7462 so as to require publication of the STJ's original report as a report of the Tax Court. See Edward J. DeBartolo Corp. v. Florida Gulf Coast Bldg. & Const. Trades Council, 485 U.S. 568, 575, 108 S. Ct. 1392, 99 L. Ed. 2d 645 (1988). This appeal would be stayed to allow the STJ's original report to be made part of the record. It is significant to summarize what I do not say here. I do not believe that due process requires that the parties be allowed to file objections to the STJ's report before the issuance of the Tax Court's opinion. Nor do I necessarily believe due process requires that the STJ's report be made public before the Tax Court issues its ultimate opinion. But by eventually making the STJ's report public (and part of the available record of the Tax Court on appeal), this court will have an opportunity to conduct meaningful appellate review. Further, this is a procedural result that may benefit all parties, including the Commissioner, not just petitioners like Kanter — Tax Court decisions can very easily reverse findings (credibility — related and otherwise) of STJs in a manner that is detrimental to the Commissioner as well as to a petitioner.

After all of these pages, what I find most interesting is that I believe the majority and I are in complete agreement on the central issue here — that the views of the STJ matter. When I say that, however, I look at the structure of the process under which the STJ's views can be discarded without leaving a trace and I find the glass half-empty. The majority sees the verbal formula ("agrees with and adopts") and finds the glass half-full. I do not believe that the concealment of the Tax Court's revision process behind that verbal formula allows this court to conduct meaningful appellate review. I appear, at the moment at least, to be alone in that belief. Therefore, I respectfully dissent.¹³