September 15, 2010.
On Appeal from the United States District Court for the Western District of Kentucky.
A jury convicted Timothy Mitts of preparing fraudulent tax returns. See 26 U.S.C. § 7206(2). On appeal, he raises a number of challenges to his conviction and sentence. We reject his arguments and affirm.
Beginning in 1989, Mitts worked as tax preparer in New York. In April 2007, a federal grand jury returned an indictment charging him with 17 counts of violating 26 U.S.C. § 7206(2). The indictment alleged that, between 2001 and 2004, Mitts had assisted in the preparation of fraudulent individual and partnership tax returns on behalf of two married couples and a single taxpayer. A superseding indictment followed a month later, charging Mitts with an additional count relating to a fourth client.
Four of the taxpayers whose returns were charged in the indictment testified at Mitts's trial, telling a relatively consistent story. In general, Mitts agreed to review the client's tax returns and financial records for the previous three years, in exchange for a share (usually one third) of any refund he generated by filing amended returns. Then, for each year at issue, he prepared a partnership tax return for a business supposedly owned by the client. He claimed enough deductions on the partnership return to generate a net loss, and then filed an amended individual return passing through the partnership's loss as a deduction. The clients testified, however, that the partnerships never existed and did not have any expenses or losses. They further testified that Mitts did not discuss with them whether the partnerships actually existed and did not explain the details of how he was able to produce the refunds for them.
There were some variations on the theme. For one client, Mitts prepared an original tax return one year, rather than an amended one. Another client testified that, rather than creating a fictitious partnership, Mitts had claimed a deduction for a loss on the sale of a business that did not actually occur.
Over Mitts's objection, the government also presented testimony from five other clients whose returns were not charged in the indictment. Four of them echoed the general scheme outlined above, testifying that Mitts had attributed deductions to non-existent partnerships. The fifth testified that Mitts had helped him fabricate various business deductions, but had not set up a fictitious partnership.
In addition, and also over Mitts's objection, the government introduced testimony from Dewayne Howell, a tax accountant who had hired Mitts out of college in 1989. Howell testified pursuant to a cooperation obligation in his plea agreement on separate tax-fraud charges. At trial, he said that he had taught Mitts how to prepare fraudulent tax returns when the two worked together during the early 1990s. In particular, he described how he had taught Mitts to hide fabricated deductions on separate returns for fictitious partnerships, because the IRS was less likely to notice them there than on an individual's tax return. He also explained that, in light of their discussions about their clients, Mitts would have known that the deductions were false and that the partnerships did not exist.
Mitts's primary defense was that he had a good-faith belief that the partnerships for which he had filed returns did exist, at least for tax purposes. According to his testimony, he had determined that his clients were operating businesses through "de facto" partnerships after examining their financial records and consulting with them. In addition, Mitts testified that he treated his clients' willingness to sign returns including partnership deductions as an indication that the partnerships did exist. The jury rejected Mitts's defenses, and convicted him.
The case proceeded to sentencing. In calculating Mitts's advisory Sentencing Guidelines range, the district court focused on the charged tax returns (and not the other ones placed into evidence at trial, which the government had sought to include as relevant conduct). The court found that Mitts was responsible for roughly $316,000 in "tax loss," corresponding to a base offense level of 18. See U.S.S.G. § 2T1.4(a)(1); id. § 2T4.1(G). After a two-point increase because Mitts was in the business of preparing tax returns, see id. § 2T1.4(b)(1)(B), and another two-point increase because his offense involved sophisticated means, see id. § 2T1.4(b)(2), Mitts's adjusted offense level was 22. After accounting for his criminal history category of I, Mitts's advisory Guidelines range was 41 to 51 months' imprisonment. The district judge selected the high end of the range and sentenced Mitts to 51 months' imprisonment.
This appeal followed.
Mitts first argues that the prosecution constructively amended the indictment by presenting at trial a theory of guilt different from the one set out in the indictment. A constructive amendment occurs "when the terms of an indictment are in effect altered by the presentation of evidence and jury instructions which modify essential elements of the offense charged," resulting in "a substantial likelihood that the defendant may have been convicted of an offense other than the one charged in the indictment." United States v. Martinez, 430 F.3d 317, 338 (6th Cir. 2005) (quotation marks omitted). We review the question de novo. United States v. Budd, 496 F.3d 517, 521 (6th Cir. 2007).
Before listing the details of the tax returns at issue in the case, the indictment charged that Mitts had violated 26 U.S.C. § 7206(2) by assisting in the preparation of tax returns
which were false and fraudulent as to material matters, in that the tax returns misrepresented and under-reported taxes owed by [the] taxpayers, resulting in whole or in part from certain business deductions and expenses, including but not limited to partnership losses, business expenses, and other ordinary losses; whereas the defendant then and there knew and believed the taxpayers whose names appear on the returns set forth below were not entitled to those business deductions and expenses, and, therefore, owed substantially more taxes for the years specified.
Initially, it appeared that the government intended to prove its case by showing that Mitts had fabricated individual line items on the charged returns — i.e., by claiming business expenses that did not exist at all. But the district judge expressed skepticism that this line-by-line approach was compatible with the government's plan to introduce evidence relating to the uncharged tax returns. The judge worried that getting into the details of the uncharged returns could distract the jury's attention from the charged ones. So the government changed tack. Specifically, the government suggested that Mitts's fraud was in attributing business expenses and losses to non-existent partnerships — whether or not the deductions could have been claimed legitimately in some other way.
Mitts does not argue that the government's theory of criminal liability was invalid on its own terms. He contends, instead, that the government's theory was so different from the one charged in the indictment as to result in a constructive amendment. The indictment charged that Mitts knew that the taxpayers "were not entitled to [the] business deductions and expenses" he had claimed in their returns. In Mitts's view, this language required the government to show that the business expenses were themselves entirely fictitious. But a taxpayer is "not entitled" to a deduction or expense that has been fraudulently attributed to a non-existent entity. The words used in the indictment were thus broad enough that Mitts cannot show that the government's theory was a departure from the charged offense.
The same reason disposes of Mitts's related claim that the government's theory amounted to a prejudicial variance from the indictment. Because the government's theory at trial did not depart from the indictment, there was no variance.
Mitts next challenges the district court's decision to admit two categories of prior-acts evidence: Dewayne Howell's testimony, and the testimony of five taxpayers relating to uncharged tax returns. Mitts contends that the evidence should have been excluded under Fed.R.Evid. 404(b), which provides in pertinent part:
Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show action in conformity therewith. It may, however, be admissible for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake or accident[.]
We review the district court's decision to admit prior-acts evidence for an abuse of discretion. See United States v. Jenkins, 593 F.3d 480, 484 (6th Cir. 2010).
The district court concluded that both categories of evidence were probative of "intent, plan, knowledge, or absence of mistake or accident." In this context, all of those options amount to roughly the same thing: that Mitts acted willfully in preparing fraudulent tax returns, instead of operating under a good-faith belief in the lawfulness of his conduct or having been misled by his clients. The district court's conclusion was correct. Howell's testimony tended to show that Mitts knew that preparing a return for a fictitious partnership was a useful strategy for evading the IRS's attention. It also tended to show that Mitts understood that the practice was fraudulent, rather than just aggressive accounting. The testimony of the uncharged clients likewise showed that Mitts had come to same conclusion as to the existence of a "de facto" partnership time after time, in return after return — which belied his defense that the returns reflected his good-faith understanding of each taxpayer's situation, as well as his suggestion that the taxpayers had somehow misled him.
Mitts responds that Howell's testimony pertained to conduct that was too far in the past and too dissimilar from the charged offenses to be relevant. But we have rejected the notion that there is any "absolute maximum number of years that may separate a prior act and the offense charged." United States v. Jones, 403 F.3d 817, 821 (6th Cir. 2005) (quotation marks omitted). And although Mitts is correct that the conduct Howell recalled was different in some respects from the charged offenses (for instance, Howell's testimony did not focus on amended tax returns), we think the similarities were more salient. In particular, the testimony that Mitts and Howell used fabricated partnerships to deflect the IRS's attention — and that Mitts must have understood the practice to be fraudulent — was highly probative.
Mitts also contends that the probative value of Howell's testimony was substantially outweighed by a danger of unfair prejudice. See Fed.R.Evid. 403. In light of the district court's broad discretion in making that balancing call, however, we cannot say that the court erred. For the reasons already stated, Howell's testimony had important probative value. And it went to an issue — willfulness — that was the keystone of Mitts's defense. Nor do we think that the introduction of Howell's testimony risked tarring Mitts by association with a convicted criminal. The district court thoughtfully prevented the government from inquiring into the details of Howell's convictions; and the prosecutor addressed Howell's convictions during closing argument only in the course of acknowledging that those convictions diminished his credibility as a witness.
With respect to the testimony of the uncharged clients, Mitts argues that permitting so many prior-acts witnesses to testify was unfairly prejudicial because it risked confusing the jury. That objection goes to the "confusion of the issues" concern in Rule 403. That concern was present here, given the number of prior-acts witnesses at trial. But with the exception of only one such witness, their testimony was quick and straightforward, and thus unlikely to derail the trial or distract the jury's attention. The district court did not abuse its discretion.
Mitts finally argues that the district court failed to give an appropriate limiting instruction, because it failed to explain to the jury how the prior acts evidence was probative of one of the permissible purposes under Rule 404(b). But the instruction correctly identified the purposes for which the evidence was admissible: "intent, plan, knowledge, or absence of mistake or accident." And it stressed that the jury should "not consider [the evidence] for any other purpose." Nothing more was required, especially in light of Mitts's failure to object to the instruction. See United States v. Newsom, 452 F.3d 593, 606 (6th Cir. 2006) (holding that a limiting instruction is not plainly erroneous so long as it correctly identifies at least one permissible purpose for the prior acts evidence).
Mitts raises two additional evidentiary issues. First, he argues that the district court erred by preventing him from "going line by line" through the charged tax returns in order to explain how, in light of the information he received from his clients, he had a good faith belief that the deductions he claimed on their behalf were in fact legitimate business expenses. Mitts does not contend that the district court actually excluded any such evidence, much less that he made the offer of proof required by Fed.R.Evid. 103(a)(2). He instead points to a statement by the district court, during a pre-trial conference, suggesting that evidence tying particular deductions to legitimate business expenses would be irrelevant under the government's theory of the case. Even so, Mitts testified in detail at trial about the specifics of fraudulent tax returns charged in the indictment and those of the prior-acts witnesses, without any objection by the government or adverse ruling by the court. Moreover, a judge's observations differ from rulings; and Rule 103 presupposes "a ruling which admits or excludes evidence." Fed.R.Evid. 103(a). To the extent that Mitts did not offer additional evidence based on the district court's pre-trial comments, we do not see how the court could have erred by excluding evidence that was never offered at trial. See Coe v. Yellow Freight Sys., Inc., 646 F.2d 444, 454 (10th Cir. 1981).
Mitts also argues that the district court erred by allowing the government to introduce, without sufficient authentication, a document that purported to be an advertisement for Mitts's accounting business. The dispute arose during the government's cross-examination of Mitts, when the prosecutor asked whether he had ever held himself out to his clients as a certified public accountant (CPA). Mitts had already testified on direct examination that he had never been a CPA and never held himself out as one; one of his clients had testified that Mitts claimed to be CPA, and several others testified that they understood him to be one. After Mitts again denied representing himself as a CPA, the prosecutor confronted him with a document purporting to be an internet advertisement for his accounting business, which invited potential clients to contact "Timothy Mitts, CPA." In response, Mitts denied placing the advertisement and said that he did not recognize it.
When the prosecutor moved to introduce the advertisement into evidence, Mitts's counsel objected that the government had not laid a proper foundation for it. The district judge did not rule squarely on the matter, instead stating "I'll admit it provisionally" and "[w]e will talk about it." The parties never returned to the question of admissibility, even though the prosecutor emphasized the advertisement during his closing argument to show that Mitts had lied on the stand. Here too it is unclear whether Mitts's objection produced "a ruling which admits or excludes evidence." Fed.R.Evid. 103(a). But in its appellate brief the government bypassed this potential forfeiture and addressed Mitts's argument on the merits. So we will do the same.
In general, "[t]he requirement of authentication or identification as a condition precedent to admissibility is satisfied by evidence sufficient to support a finding that the matter in question is what its proponent claims." Fed.R.Evid. 901(a). Thus, the government was required to present sufficient evidence that the document was an advertisement that Mitts had placed. Invoking the principle that a piece of evidence can be authenticated through its distinctive characteristics, see Fed.R.Evid. 901(b)(4), the government contends that it met its burden here because the document listed Mitts's name, telephone number, and address. But there is nothing distinctive about that information. Anyone could have learned Mitts's phone number and address and then produced the document. The situation here is thus unlike the one in United States v. Jones, 107 F.3d 1147 (6th Cir. 1997), where we held that a letter was sufficiently authenticated because it included details about the recipient's family members that only the purported author could have known. See id. at 1150; see also Fed.R.Evid. 901(b)(4) advisory committee's note (observing that "a document or telephone conversation may be shown to have emanated from a particular person by virtue of its disclosing knowledge of facts known peculiarly to him)" (emphasis added).
Thus the document should not have been admitted. But for several reasons the error was harmless, a ground that we have discretion to invoke even though the government has failed to argue it on appeal. See, e.g., United States v. Achobe, 560 F.3d 259, 268 n. 29 (5th Cir. 2008); United States v. Davenport, 929 F.2d 1169, 1175 (7th Cir. 1991). First, the document was not the only evidence that Mitts had falsely held himself out as a CPA. One witness directly testified that Mitts had claimed to be a CPA, and several others testified that they had understood that he was one. Second, the evidence of Mitts's guilt was very strong. Witness after witness testified that Mitts had claimed deductions on behalf of partnerships that were entirely fictitious. And there was no evidence — other than Mitts's own vague, conclusory testimony — tending to show how he could have derived a good-faith belief that the partnerships existed from the records he claimed to have examined. Finally, Mitts's good-faith defense was further undermined by Howell's testimony that he had taught Mitts to use non-existent partnerships to shield deductions from the IRS's attention. Under these circumstances, we see no realistic chance that the error in admitting the document affected the jury's verdict.
Mitts raises two challenges to his sentence. The first is a claim that the district court started with the wrong "tax loss" figure when it calculated his advisory range under the Sentencing Guidelines. As is relevant here, the Guidelines define "tax loss" as "the total amount of loss that was the object of the offense (i.e., the loss that would have resulted had the offense been successfully completed)." U.S.S.G. § 2Tl.l(c)(1). The district court focused only on the deductions claimed on behalf of fictitious partnerships in the charged returns, and calculated a tax loss of roughly $316,000. According to Mitts, the court should have calculated the tax loss using a column labeled "Tax Loss Actual" in an attachment to one of the government's sentencing memoranda, which would have yielded a tax loss of roughly $183,000 and a two-point reduction of his base-offense level. See U.S.S.G. § 2T4.1(H), (G). The $183,000 figure differs from the $316,000 that the district court calculated because it accounts for reductions in the refunds paid to one taxpayer when the IRS audited two of his amended returns shortly after they were filed. Although the IRS did not discover during the audit that the taxpayer's partnership was fictitious, it did disallow several of the deductions Mitts had claimed. The result was that the taxpayer was paid a much smaller refund than the amended returns had claimed.
Mitts insists that, because this actual loss figure was readily ascertainable, it should have been used to calculate his Guidelines range. But the Guidelines are clear: tax loss means "the loss that would have resulted had the offense been successfully completed." U.S.S.G. § 2Tl.l(c)(1). Mitts was not entitled to benefit from the fact that the IRS uncovered part of the fraud before cutting the client's refund check.
Mitts also contends that the district court overlooked his argument that a lower sentence was justified by extraordinary family circumstances — namely, his status as sole provider for two sickly parents, two thirteen-year-old daughters, and a nineteen-year-old daughter whom he described as "troubled." Although Mitts now contends that the district court did not give sufficient reasons for rejecting his family-circumstances argument, he did not raise any objection to the court's explanation when offered the chance to do so at the close of sentencing. He therefore must show plain error in order to prevail on appeal. See United States v. Vonner, 516 F.3d 382, 385-86 (6th Cir. 2008) (en banc). He cannot, because the record makes clear that the district court carefully considered this argument. The court acknowledged that Mitts's family circumstances could play a role in mitigating his sentence, but it concluded that they were outweighed by other considerations, principally Mitts's refusal to take responsibility for his crime. This was not plain error.
The district court's judgment is affirmed.