Kosinski v. Commissioner

541 F.3d 671, 102 A.F.T.R.2d (RIA) 5955, 2008 U.S. App. LEXIS 18617, 2008 WL 3981220

• Court: Court of Appeals for the Sixth Circuit
• Decided: August 29, 2008
• Docket: 07-2136
• Status: Published

Opinion

OPINION

SUTTON, Circuit Judge.

Timothy and Barbara Kosinski challenge a decision by the Tax Court upholding a tax deficiency and a fraud penalty. They raise two issues: that earlier findings made by a federal district court in the course of imposing a sentence on Timothy Kosinski in a criminal case precluded the Tax Court’s deficiency findings, and that the government failed to show that the deficiency resulted from fraud. Because the district court’s sentencing findings lack issue-preclusive effect and because the government carried its burden of proving fraud, we affirm.

I.

This case arises from a complex, multi-year tax-evasion scheme, for which the government successfully prosecuted the Kosinskis and for which it now seeks to collect one year’s worth of unpaid taxes and penalties. In 1991, Timothy Kosinski founded T.J. Construction, a wholly owned S corporation, to perform construction contracts for a single customer, Thyssen Steel. T.J. Construction generally farmed out the work to subcontractors — principally Melvin Phillips and his own wholly owned company — while it focused on acquiring the contracts and handling the paperwork. Thyssen paid T.J. Construction directly for completed projects, and T.J. Construction in turn paid Phillips, deducting the payments to Phillips from its gross income as part of its cost of goods sold.

To the end of minimizing their taxes, the Kosinskis began processing the payments differently in 1996. Phillips would endorse the checks from T.J. Construction over to the Kosinskis, who would then deposit them in their personal bank accounts. In 1997, the year at issue in this case, the endorsed-back checks totaled $2,919,974. The Kosinskis left some of the money in their accounts but withdrew much of it (nearly $2 million in 1997) in cash through hundreds of less-than-$10,000 transactions.

No one knows exactly where all of the money went — save initially to the Kosinsk-is, who “regularly destroyed” what records they kept. JA 51. All agree that some large amount (the parties dispute how much) went back to Phillips. The Ko-sinskis characterize these payments as cash advances that Phillips repaid with more endorsed-back checks; the government characterizes these payments as under-the-table exchanges that allowed Phillips to handle his payroll in cash and to evade federal employment taxes and withholding requirements. The government also points out that this scheme allowed the Kosinskis to duck significant tax liability, because neither they nor T.J. Construction (their wholly owned flow-through S corporation) reported these amounts on their annual returns.

The government filed criminal charges against the Kosinskis (and Phillips) for a number of tax-related offenses. Barbara Kosinski pleaded guilty to structuring currency transactions, and a jury convicted Timothy Kosinski of several counts of filing false tax returns, one count of structuring currency transactions and one count of conspiring to defraud the IRS and structure currency transactions.

The district court sentenced Timothy Kosinski in 2003. As directed by the then-mandatory sentencing guidelines, see U.S.S.G. §§ 2T1.1, 2T4.1 (1995), the court based the sentence on its determination of the “tax loss” attributable to Kosinski’s conduct, namely the amount of taxes Ko-sinski and others avoided paying due to the conspiracy. The government told the court that the scheme resulted in an aggregate tax loss of $2.3 million, while Ko-sinski argued that the loss was less than $200,000. The district court in effect split the difference, finding a loss of $973,176, after which it imposed two concurrent, 30-month (within-guidelines) sentences. We vacated those sentences in light of United States v. Booker, 543 U.S. 220, 125 S.Ct. 738, 160 L.Ed.2d 621 (2005), see United States v. Kosinski (Kosinski I), 127 Fed.Appx. 742, 751 (6th Cir. Mar.22, 2005), and we likewise vacated (for Booker-related reasons) the sentence the district court imposed on remand, see United States v. Kosinski (Kosinski II), 480 F.3d 769, 777-78 (6th Cir.2007).

In the meantime, the government in 2004 sent the Kosinskis a deficiency notice for 1997, alleging a tax underpayment of $1,205,548 and imposing a $904,161 fraud penalty. The Kosinskis filed a petition for redetermination of their deficiency with the Tax Court, in which they claimed they “owed no tax” for 1997. JA 8. After a trial, the Tax Court upheld the government’s original calculation of the deficiency, as modified by the government’s concession that $1 million of the alleged understatement was legitimate, as well as the modified fraud penalty.

II. .

The Kosinskis first challenge the Tax Court’s deficiency determination on issue-preclusion grounds, maintaining that the district court’s findings of fact at Timothy’s criminal sentencing hearing barred the Tax Court in his civil tax-deficiency proceeding from imposing a $812,182 deficiency. To invoke issue preclusion successfully, a litigant must show four things:

(1) the precise issue raised in the present case must have been raised and actually litigated in the prior proceeding; (2) determination of the issue must have been necessary to the outcome of the prior proceeding; (3) the prior proceeding must have resulted in a final judgment on the merits; and (4) the party against whom estoppel is sought must have had a full and fair opportunity to litigate the issue in the prior proceeding.

United States v. Cinemark USA, Inc., 348 F.3d 569, 583 (6th Cir.2003) (internal quotation marks omitted).

These requirements demand more than the Kosinskis can supply. First, their claim stumbles over the initial demand that they identify the “precise issue” decided by the sentencing court that purportedly estops the government here. While they maintain that the government “is collaterally estopped from using other numbers that were previously determined in the criminal conviction in the Tax Court,” Br. at 17, they never clarify what “other numbers” they are talking about. As best we can tell, they contend that the district court’s assessment of the “tax loss” attributable to Timothy Kosinski’s crimes foreclosed the Tax Court’s determination of the amount of their tax underpayment. Yet how the district court’s decision could do so remains a mystery, not least because it made only aggregate findings for several years combined, while the Tax Court case concerned just 1997. In the absence of a finding by the sentencing court on the “precise issue” before the Tax Court, preclusion is a guessing game.

Second, the Kosinskis have not shown that the sentencing court’s relevant fact findings — whatever they were — were “necessary” to its judgment. As we explained in rejecting a similar request to tie a Tax Court’s hands after a related criminal proceeding, the sentencing court’s determination of the underpayment “was not essential to the district court’s judgment because it was not an element of the crime of conviction.” Hickman v. Comm’r, 183 F.3d 535, 538 (6th Cir.1999). The same thing happened here. See Kosinski I, 127 Fed.Appx. at 751. Kosinski’s conviction cannot preclude the Tax Court from making findings on an issue the jury never had any reason to decide.

It is true that the district judge estimated the tax loss in determining Kosinski’s guidelines range. And it is true that the guidelines direct the district judge to gauge the tax loss in ascertaining a criminal’s base-offense level. See U.S.S.G. § 2T1.1. But the broad tax-loss bands of the guidelines diminish the contention that a given tax-loss finding was necessary to the sentence. Here, for example, the district court could have reached the same within-guidelines, 30-month sentence so long as the tax loss fell anywhere between $550,000 and $2,500,000. See id. §§ 2T4.1, 5A. The sentencing judge himself appeared to appreciate this fact, adopting Kosinski’s calculation of the tax loss for one charge, instead of the government’s much higher figure, apparently because the difference would not have affected Kosinski’s base-offense level.

But that is only half the problem. In the aftermath of Booker, the guidelines could not constitutionally cause his sentence to turn on the district court’s tax-loss finding. That indeed is why we vacated Kosinski’s initial sentence, explaining that the use of non-jury-found facts in applying mandatory guidelines presented the same constitutional problem at issue in Booker. Kosinski I, 127 Fed.Appx. at 750-51. And when the district court on remand resentenced Kosinski without making any findings whatsoever regarding the tax-loss amount — laboring “under the misapprehension that it simply could not do so” after Booker — we vacated the sentence again, clarifying that the district court “should recognize and exercise its discretion to consider — or not to consider — [Kosinski’s] tax loss.” Kosinski II, 480 F.3d at 777. That discretion undermines the Kosinskis’ contention that the district court’s tax-loss finding was necessary to the final judgment. Cf. Hickman, 183 F.3d at 538 (holding that a district court’s discretionary determination of the amount of restitution a criminal tax-evasion defendant had to pay was unnecessary to the judgment and thus could not be given preclusive effect); Morse v. Comm’r, 419 F.3d 829, 834 (8th Cir.2005) (same).

Third, even if the district court’s determination of Timothy Kosinski’s criminal sentence had hinged entirely on the district court’s tax-loss determination, that finding still would not entitle the Kosinskis to preclusion because no final judgment existed in the criminal proceeding when the Tax Court issued its decision. Only final judgments, not surprisingly, possess issue-preclusive power, Cinemark USA, 348 F.3d at 583, and “[a] judgment that has been vacated, reversed, or set aside on appeal is thereby deprived of all conclusive effect, both as res judicata and as collateral estoppel,” Erebia v. Chrysler Plastic Prods. Corp., 891 F.2d 1212, 1215 (6th Cir.1989); see also Dykstra v. Wayland Ford, Inc., 134 Fed.Appx. 911, 917 (6th Cir.2005); cf. Durning v. Citibank, NA, 950 F.2d 1419, 1424 n. 2 (9th Cir.1991) (noting that “[a] decision may be reversed on other grounds, but a decision that has been vacated has no precedential authority whatsoever”).

While the district court issued serial judgments in Kosinski’s criminal case, this court serially vacated them, and not one of those judgments could be characterized as final before the Tax Court’s decision. Thus, the district court issued a judgment in imposing Kosinski’s initial October 2003 sentence, but this court vacated that judgment on March 22, 2005. Kosinski I, 127 Fed.Appx. at 750-51. The district court issued a second judgment in imposing Ko-sinski’s second sentence in September 2005, but this court vacated that judgment on March 22, 2007. See Kosinski II, 480 F.3d at 778. The district court did not hold another resentencing hearing until January 15, 2008 — months after the Tax Court issued its memorandum opinion on July 2, 2007, and its August 23, 2007, final decision. When the Tax Court decided this case, in short, there was no valid final judgment to which it could give preclusive effect.

Fourth, Timothy Kosinski’s sentencing proceeding did not give the government a full and fair opportunity to litigate the tax-loss issue. As the Supreme Court recognized when it first approved the offensive use of issue preclusion, allowing a litigant to invoke the doctrine “might be unfair ... where the second action affords the [opposing party] procedural opportunities unavailable in the first action that could readily cause a different result” or where the opposing party “ha[d] little incentive to defend vigorously” in the first action. Parklane Hosiery Co. v. Shore, 439 U.S. 322, 330-31, 99 S.Ct. 645, 58 L.Ed.2d 552 (1979). Efforts to bind adjudicators presiding over civil cases to fact findings made in prior sentencing proceedings, as other circuits have recognized and as this case illustrates, raise this precise problem. See Maciel v. Comm’r, 489 F.3d 1018, 1023-26 (9th Cir.2007); SEC v. Monarch Funding Corp., 192 F.3d 295, 305 (2d Cir.1999); see also United States v. U.S. Currency in the Amount of $119,984, 304 F.3d 165, 175-78 (2d Cir.2002).

The procedural ground rules for criminal-sentencing proceedings differ considerably from the ground rules that govern civil actions. At sentencing, the defendant may have a limited opportunity to take discovery and “has no absolute right either to present his own witnesses or to receive a full-blown evidentiary hearing,” Monarch, 192 F.3d at 305. And he does not enjoy the protection of the Federal Rules of Evidence, where the “judge is largely unlimited either as to the kind of information he may consider, or the source from which it may come, so long as the information has sufficient indicia of reliability to support its probable accuracy.” Id. (internal quotation marks, citations and brackets omitted).

The prosecution, too, faces disparities between the two settings. It may lack “procedural mechanisms crucial to [its] ability to gather probative evidence” relevant to the subsequent civil matter. $119,981, '304 F.3d at 176-77 (noting that in a later civil action, unlike at sentencing, the government “could rely on [the rules of evidence] to challenge the admissibility of the financial records” offered by the defendant at sentencing). In contrast to “the full array of civil discovery procedures against the defendant” available in a civil action, “including interrogatories, requests for admissions, document requests, depositions, and so forth,” the Federal Rules of Criminal Procedure applicable at sentencing allow the government to seek “discovery of documents and tangible objects from the defendant only if the defendant seeks reciprocal discovery from the Government, and only of such material ‘which the defendant intends to introduce as evidence in chief at the trial.’ ” Id. at 177 (quoting Fed.R.Crim.P. 16(b)(1)(A)).

Unlike the defendant, moreover, the government at sentencing faces constitutional restrictions that either do not apply to civil cases or that as a practical matter will not limit it in a civil trial. The Fifth Amendment’s Self-Incrimination Clause bars the prosecution from compelling the defendant’s testimony and from advancing an adverse inference from his decision not to testify- — -a limitation that is generally absent in civil proceedings, see Baxter v. Palmigiano, 425 U.S. 308, 318-19, 96 S.Ct. 1551, 47 L.Ed.2d 810 (1976), and one that a defendant cannot invoke in a Tax Court case to satisfy his burden of proving that the government miscalculated his tax deficiency. After all, “[t]he Fifth Amendment privilege cannot be used by a taxpayer to meet his burden of proof in a proceeding which he himself has instituted.” Tweeddale v. Comm’r, 841 F.2d 643, 645 (5th Cir.1988). A defendant’s privilege not to testify likewise limits the government’s ability to show the sentencing court the implausibility of the defendant’s version of events, a limitation that might well have made a difference in this case: The Tax Court found Timothy Kosinski’s testimony “inconsistent and implausible,” JA 52, a finding that surely hurt the Kosinskis’ cause, while at sentencing (where he did not testify) Kosinski escaped such scrutiny. Cf. Monarch, 192 F.3d at 305 (noting that “a defendant, though uniquely knowledgeable about underlying events, may be reluctant to testify during sentencing” for “a number of reasons ... not the least of which is that if the- defendant is disbelieved, his sentence may be enhanced under Guideline § 3C1.1”). And although giving issue-preclusive effect to sentencing findings undoubtedly would increase the parties’ incentives to litigate issues in the sentencing court, it likely would do so to a perverse degree, potentially transforming sentencing hearings into “mini-trials,” undercutting the very efficiency goals preclusion is designed to serve. See Monarch, 192 F.3d at 306.

Perhaps most importantly, the burden of persuasion differs in each setting. In a sentencing hearing, the government carries the burden of persuasion when it comes to proving relevant sentencing facts. See United States v. Silverman, 889 F.2d 1531, 1535 (6th Cir.1989); United States v. Snipe, 515 F.3d 947, 955 (9th Cir.2008). In a civil tax case, the taxpayer carries the burden of persuasion when it comes to proving that the deficiency is incorrect. See Indmar Prods. Co. v. Comm’r, 444 F.3d 771, 776 (6th Cir.2006).

The parties’ incentives to litigate an issue also may differ between a sentencing hearing and a later civil proceeding. A defendant “will often choose not to challenge sensitive issues during sentencing for any number of reasons, including a belief, or at least a hope, that the sentencing court will grant a prosecutorial downward departure motion or other recommendation.” Monarch, 192 F.3d at 305. And while, “[i]n some cases, the government’s obligation to seek a sentence consonant with a criminal defendant’s culpability will be incentive enough to ensure that relevant issues are litigated vigorously,” that will not always be the case. Maciel, 489 F.3d at 1025. Where a district court’s finding on an issue likely will have little effect on a defendant’s sentence — for example, where any conclusion the district court reached would have led to the same guidelines range — the government’s motivation to litigate the issue before the sentencing court reflects a fraction of its incentive to contest the issue in later civil proceedings. See id. (declining to extend preclusive effect in á subsequent Tax Court proceeding to a sentencing court’s pre-Booker finding that the defendant had not intended to defraud the government because “the government had virtually no incentive to litigate” the issue, given that the finding made no difference to the defendant’s mandatory guidelines range).

Here the district court’s tax-loss findings had little if any effect on Kosinski’s base-offense level — a fact the district court itself appreciated — much less his ultimate sentence given the overlapping guidelines ranges. The Kosinskis brought this case in the Tax Court, by contrast, precisely to challenge the amount of (and associated penalties for) their tax deficiency as calculated by the government. We need not say that the government had no incentive to litigate the issue in the sentencing proceeding; it suffices for issue-preclusion purposes to establish that the government lacked sufficient incentives and procedural opportunities to litigate the issue vigorously in the district court.

After reading all of this, one might question how a determination reached in a criminal-sentencing proceeding could ever satisfy this issue-preclusion requirement— whether an individual or the government seeks to invoke the defense. And, to be sure, we know of no case (and the parties have cited none) where a federal court has ascribed preclusive effect to a sentencing court’s findings of fact, and two other circuits have held issue preclusion presumptively inapplicable to sentencing findings. See Maciel, 489 F.3d at 1025; Monarch, 192 F.3d at 306. But to resolve this case we need not, and therefore do not, decide whether sentencing determinations categorically or even presumptively lack pre-clusive power. We simply conclude, for the reasons given, that the Kosinskis’ claim falls far short of the mark and therefore presents no ground for overturning the Tax Court’s conclusion concerning the amount of their tax deficiency.

III.

The Kosinskis next challenge the Tax Court’s finding that they fraudulently understated their 1997 taxes, a finding that subjected them to a 75% penalty. See 26 U.S.C. § 6663(a). In seeking to impose this penalty, the government undertook the burden of proving fraudulent intent by clear and convincing evidence. Richardson v. Comm’r, 509 F.3d 736, 743 (6th Cir.2007); 26 U.S.C. § 7454(a); Tax Ct. R. 142(b). In seeking to overturn this finding on appeal, the Kosinskis undertook the burden of establishing that the Tax Court’s fraud finding suffers from clear error. See Richardson, 509 F.3d at 740.

In establishing fraud, the government need not establish direct evidence of the taxpayer’s untoward state of mind. Because “[i]t is the rare taxpayer who announces to the world his intent to defraud the Federal Government,” the government may prove fraudulent intent by circumstantial evidence, and we can infer fraud from “any conduct, the likely effect of which would be to mislead or to conceal.” Richardson, 509 F.3d at 743 (internal quotation marks omitted). While any effort to catalogue a list of evidence that satisfies this standard would be doomed to incompleteness, there are several telltale “badges of fraud”: where the individual fails to report income, fails to maintain and produce “adequate books and records” of financial activities, “conceal[s][his] income by dealing in cash” and, even though he has “business experience,” “give[s] implau sible explanations of conduct.” Id. (internal quotation marks omitted).

Nor need the government show that a taxpayer’s entire tax underpayment resulted from fraud. Once it “establishes that any portion of an underpayment is attributable to fraud, the entire underpayment shall be treated as attributable to fraud,” and the burden shifts to the taxpayer to prove (under a preponderance standard) that any part of the underpayment “is not attributable to fraud.” 26 U.S.C. § 6663(b) (emphasis added). To the extent the taxpayer fails to make that showing, the government may assess the 75% fraud penalty based on the entire amount of the tax underpayment. Id. § 6663(a).

The Tax Court did not clearly err in finding that the government met its burden of proving that at least part of the Kosinskis’ 1997 tax underpayment arose from fraud. The court permissibly identified several “badges of fraud” revealing that the Kosinskis’ large underpayment did not reflect an unfortunate but innocent “miscommunication between them and their [tax-]return preparers”: It relied on their similar understatement of income in 1996 and 1998, their practice of keeping detailed records for other matters but “regularly destroying]” records regarding their many cash transactions, their failure to apprise their accountants of their cash disbursements, their carefully designed and systematically executed scheme of cash withdrawals (utilizing Barbara Kosin-ski’s inside knowledge as a former bank teller) that enabled them to come just under federal reporting requirements (and thus just under the regulators’ radar), the large amount of cash (nearly $500,000) they had on hand when first interviewed by the government and their inconsistent and implausible explanations as to where the money came from and why they kept so much cash at the ready. JA 50-53. Even after accepting the Kosinskis’ argument that their criminal convictions and guilty pleas (and those of their coconspira-tors) did not preclude them from denying fraud in the civil deficiency case, the Tax Court also permissibly treated those convictions and pleas as probative evidence of fraud. See Morse, 419 F.3d at 833.

On appeal, the Kosinskis offer no cognizable reason why this evidence falls short of showing that at least some of their underpayment reflected fraud. They argue that the Tax Court should not have believed Melvin Phillips’s testimony at trial because he “is concealing assets from the government,” “owes the [IRS] money” and “has a history of demonstrating that honesty is not the best policy.” Br. at 21. And as they told the Tax Court, Phillips’s testimony did not deserve credence because the government “can (and did) use the threat of tax prosecution to control [his] testimony ... like the dance of a marionette.” JA 625. But these arguments do not offer a colorable reason for overturning the Tax Court’s finding because they amount to nothing more than attacks on the credibility of a witness the Tax Court observed. We have long given considerable deference to the Tax Court’s (and other fact finders’) first-hand assessment of the credibility of witnesses before them, see Indmar Prods., 444 F.3d at 778; Conti v. Comm’r, 39 F.3d 658, 664 (6th Cir.1994), and the Kosinskis offer no tenable basis for second guessing that assessment here.

As an apparent afterthought, the Kosinskis challenge the Tax Court’s finding that Barbara Kosinski was not entitled to innocent-spouse relief from the fraud penalty. See 26 U.S.C. § 6015. This argument runs into two problems. First, the Kosinskis failed to develop the argument beyond the most cursory men tion of the issue, Br. at 11; Reply Br. at 7, and thus they have waived this objection. See McPherson v. Kelsey, 125 F.3d 989, 995-96 (6th Cir.1997); cf. Kosinski I, 127 Fed.Appx., at 750. Second, their claim fails on the merits because Barbara, a former bank teller, plainly knew of many of the fraudulent transactions and played an active part in structuring hundreds of currency transactions to avoid regulators’ scrutiny. Of. Richardson, 509 F.3d at 745-46.

IV.

For these reasons, we affirm the Tax Court’s decision.