United States v. Brennick

BOUDIN, Circuit Judge. John Brennick was convicted of various offenses centered around his failure to pay over to the Treasury income and social security taxes withheld from his employees’ paychecks. The district court calculated the range of imprisonment fixed by the sentencing guidelines at 41 to 51 months but then departed downward and imposed a sentence of 13 months’ imprisonment. The government now appeals, arguing that the downward departure was error. I. John Brennick was the president and sole proprietor of a number of head injury treatment centers in Massachusetts, Pennsylvania, Delaware and Maryland. He also operated one head trauma center in New Jersey as a limited partnership, Brennick being the general partner. Some of the centers provided sophisticated medical treatment; others appear to have been supported living centers for head injured patients. Taken as a whole, the companies were a large and successful business venture. Employers like Brennick are required to withhold income taxes and social security taxes from employee paychecks on a periodic basis and to pay those amounts over to the Treasury. The Internal Revenue Service specifies the periods for which such withholding is required. Employers are required by law to deposit the withheld taxes into the Treasury within three days after the end of each such period. Regular returns, specifying the amounts withheld and paid over, are also required on a quarterly basis. From 1986 to 1992, Brennick followed a regular pattern of withholding the taxes from his employees’ pay but delaying payment of the monies into the Treasury for a substantial period beyond the time due. Normally his payments to the government were between two and six months after the due dates. Brennick routinely filed returns accurately describing the amounts withheld, and when he ultimately made the delayed payments to the Treasury, he also paid the interest and penalties prescribed by law for late payments. During this period, Brennick frequently withdrew money from his businesses by means that avoided bank reports to the IRS that are required when a person withdraws more than $10,000 from an individual bank on a single banking day. Brennick told various of his employees and family members to cash cheeks drawn on Brennick’s various business accounts and to turn the money over to him. The individual checks were for less than $10,000 each; but the total withdrawn from his company accounts was often well over $10,000 a day. There is no claim that Brennick was forbidden to withdraw the monies from the companies’ accounts; in fact, for most of them he was the sole proprietor, and for the remaining one he was the general partner. The charge later brought against him was that the withdrawals were structured to avoid the filing of currency transaction reports and to deflect the attention of the tax authorities. It is said that Brennick took much or all the money he withdrew and lost it in gambling: he claims to have lost more than $1 million a year. During the second half of 1992, Brennick’s businesses began to suffer financial prob *12 lems. Changes were occurring in the health care industry adversely affecting providers like Brennick. Insurance reimbursements came more slowly and for lower amounts, while the costs of providing service increased. In December 1992, one of the banks that had been lending money to Bren-nick failed and Brennick could not find another lender to replace it. At the same time, the IRS began to investigate Brennick’s pattern of chronically late payments. In a meeting with an IRS agent on October 30, 1992, Brennick agreed to a payment plan, including a commitment to keep current on future payments. He promised that his businesses would seek to expedite payments to the IRS and would cut his own pay and the pay of other executives in order to pay back taxes. Instead, Brennick removed another $80,000 cash from the businesses in November 1992 and almost twice that amount in December. In addition, Brennick now began to file false quarterly withholding tax returns for many of the companies. Returns filed in the third and fourth quarter of 1992 incorrectly stated that Brennick had paid over to the government virtually all of the withheld taxes; in truth, the companies in question had paid none of the taxes over to the IRS. In two cases Brennick signed the false returns himself; in other cases they were signed by employees, but Brennick was the person responsible for the withholding of the taxes. In February 1993, Brennick filed for reorganization of his businesses under chapter 11 of the Bankruptcy Code, and later the case was transformed into a chapter 7 liquidation. At the initial filing, Brennick owed the Treasury over $1.4 million in withheld taxes that should have been, but had not been, paid over to the government. During reorganization, Brennick took additional funds out of the businesses for himself while failing to pay over the full amount of taxes withheld during the same period. In 1995, a grand jury indicted Brennick. In a superseding indictment, Brennick was charged with 22 counts of willful failure to account for, and pay over quarterly, specified withholding taxes, 26 U.S.C. § 7202; nine counts of structuring currency transactions, 31 U.S.C. §§ 5313, 5322 and 5324; and one count of corruptly endeavoring to obstruct and impede the IRS, 26 U.S.C. § 7212(a). There was an additional single charge of bankruptcy fraud, 18 U.S.C. § 152, but the jury later deadlocked on that issue. In December 1995, Brennick went on trial. The government, in addition to offering evidence of the events already described, called several of Brennick’s former employees who testified that Brennick had known the deadlines for paying over the withheld taxes but had deliberately chosen to ignore them even though his employees had sought to get him to pay over the taxes on a timely basis. The bankruptcy fraud count aside, the jury convicted Brennick on all remaining counts. The district court held a two-day proceeding to determine Brennick’s sentence and after sentencing, issued a memorandum and order explaining the court’s analysis. United States v. Brennick, 949 F.Supp. 32 (D.Mass.1996). After briefly setting out the background facts, the memorandum calculated the normal guideline range, referring (as we do) to the 1992 version of the guidelines. Then, at length, it set out the framework for departures and the court’s reasons for departing in this case. Brennick was convicted of violating three different statutes — failure to pay over withheld taxes, structuring, and obstructing the IRS — but the conduct was arguably related. In any event, the district court chose to treat the offenses as closely related counts to be grouped under U.S.S.G. § 3D1.2, and its choice is not disputed on this appeal. 1 Where counts are so grouped, the court selects the offense level for the violation among the group that has the highest offense level. U.S.S.G. § 3D1.3(b). The district court ruled that the highest offense level was generated by the offense of corruptly impeding tax officials under 26 U.S.C. § 7212(a). Although no specific *13 guideline exists for this offense (unless force is used), see U.S.S.G., appendix A, the court is directed to use the guideline for the offense most analogous to the criminal conduct of which the defendant was convicted. U.S.S.G. § 1B1.2. Here, the district court concluded that the closest analogy for the obstructive conduct was the offense of tax evasion, a violation of 26 U.S.C. § 7201, for which a specific tax evasion guideline is set forth, U.S.S.G. § 2T1.1. Although Brennick was not charged with tax evasion, this choice of analogy is not challenged by either side, and we accept it as reasonable for purposes of this appeal. The government’s obstruction charge embraced all of Brennick’s behavior (deliberate underpayments, structuring, and other acts of falsity or concealment) and that conduct includes withholding revenues from the government combined with elements of conscious wrongdoing and personal gain. The base offense level for the tax evasion guideline is driven by the tax loss inflicted on the government, and in this case the undisputed level of the government’s loss — “more than $1,500,000” — corresponds to offense level 18. U.S.S.G. §§ 2Tl.l(a), 2T4.1(M). The district court added two levels on the ground that Brennick had used “sophisticated means” to impede discovery of the offense, see U.S.S.G. § 2Tl.l(b)(2), and two more levels for obstruction of justice because of untruthful testimony by Brennick at trial, see U.S.S.G. § 3C1.1. Given a total offense level of 22 (and a criminal history category I), the guideline range for Brennick was a term of imprisonment of 41 to 51 months. From this range, the district court departed downward to level 13, for which the prescribed range for a defendant in criminal history category I is 12 to 18 months’ imprisonment. The court imposed a sentence of 13 months, as well as a fine of $6,000 and the statutory special assessment, noting that Brennick remained personally liable to the government for tax losses he had caused, 26 U.S.C. § 6672. The court’s reasons for the departure were set forth in some detail but reflect two central themes: first, that Brennick’s intent was not as wicked as that of the typical tax evader because, despite some conscious wrongdoing, he did not intend permanently to deprive the government of the funds he failed to pay over; and second, the ultimate losses to the government were due not merely to Brennick’s conduct but to contributing causes as well, including failure of his business’s main bank and adverse developments in the health care market. The government has now appealed to challenge the sentence. It argues that the departure was based on a misconstruction of the guidelines and that even if a ground for departure exists in theory (which the government denies), the district court’s decision to depart and degree of departure were unreasonable on the present facts. We take the issues in that order. IÍ. Departures from the guideline range are allowed where “the court finds that there exists an aggravating or mitigating circumstance of a kind, or to a degree, not adequately taken into consideration by the Sentencing Commission in formulating the guidelines that should result in a sentence different from that described.” 18 U.S.C. § 3553(b). Sometimes, the guidelines identify a “circumstance” that is a permissible or forbidden basis for departure, sometimes further indicating that departure is encouraged or discouraged. Absent such explicit guidance, the Commission itself has told courts that they should treat each guideline as carving out a “heartland” representing “a set of typical cases embodying the conduct that each guideline describes.” U.S.S.G. eh. 1, pt. A intro, comment 4(b). “When a court finds an atypical case, one to which a particular guideline linguistically applies but where conduct significantly differs from the norm, the court may consider whether a departure is warranted.” Id. If the characteristic is “atypical” and aggravates or mitigates the typical conduct, it may provide a basis for departure. Where a district court does depart, an aggrieved party may appeal from both the decision to depart and the extent of the departure. 18 U.S.C. § 3742. The standard *14 of review varies with the nature of the issue involved, deference being limited or absent on abstract issues of law but more generous as to questions of law application and fact-finding. United States v. Black, 78 F.3d 1, 8 (1st Cir.), cert. denied, — U.S. -, 117 S.Ct. 254, 136 L.Ed.2d 180 (1996). The present case presents issues of all three kinds. We start with the district court’s determination that Brennick, although he had deliberately failed to pay the government the withheld wage and social security taxes at the time they were due, genuinely intended to pay them in due course. In the district court’s view, Brennick’s main aim was to use the IRS as a bank. It is not clear that the government directly challenges this finding, but in any case we think the finding is not clearly erroneous, the standard ordinarily applied to determinations of fact made at sentencing. United States v. Pineda, 981 F.2d 569, 572 (1st Cir.1992). Brennick’s pattern before financial difficulties engulfed him was to retain the use of the funds in question for periods of four to six months and then to pay over the funds, adding penalties and interests. The likelihood that he would be able to make this repayment obviously declined as troubles loomed in late 1992, but he continued to scramble for resources to continue payment. Whether an intent to repay can be ascribed to all of the delays in payment is a more difficult issue. See part III below. In the district judge’s sentencing memorandum and order, she relied heavily upon this intention to repay to carve the present case out of the “heartland” of typical tax evasion eases. The government says this rationale was a belated attempt to bolster a departure earlier premised on a different ground, namely, that there were multiple causes for the loss to the government. Our own reading of the sentencing transcript suggests that the benign view of Brennick’s intent was always an element in the district court’s reasoning. The nature of the scienter element in a tax evasion ease is complicated to summarize given that different requirements may apply on different issues. Still, the taxpayer usually is attempting to deprive the government permanently of taxes owed to it. Typically, the instruction requires the government to prove that the defendant “willfully evaded, or attempted to evade, income taxes with the intention of defrauding the government of taxes owed.” Leonard B. Sand, et al., Modem Federal Jury Instructions: Criminal ¶ 59.01, Instruction 59-8 (1992) (emphasis added). See, e.g., United States v. Aitken, 755 F.2d 188 (1st Cir.1985). Admittedly, it would do a defendant no good to say that he deliberately understated his income but sincerely intended to pay the money back to the government in five years’ time. But neither is it easy to imagine a fraud conviction where a defendant files an accurate return, intends shortly to pay in full, but remits the funds with interest shortly after the April 15 deadline. Indeed, the guideline covering the failure to pay over payroll taxes notes in the commentary that “[t]he offense is a felony that is infrequently prosecuted.” U.S.S.G. § 2T1.6, commentary. In all events, we are inclined on the basis of the information we have and our common sense to think that such a temporary delay in payment — where the defendant expected to pay — is not a “typical” or “heartland” case of tax evasion. Thus, even if the evasion statute and guideline might “linguistically” be extended to embrace such temporary delay cases, the intent to delay payment only briefly could take the case out of the heartland. And, as already noted, the district court made such a finding in this case, sustainable at least as to much of the losses driving the guideline sentence. The district court had another theme in its departure analysis. It said that the $1.5 million loss suffered by the government overstated the seriousness of Brennick’s offense, partly because the losses were due to multiple causes, some of which were not Brennick’s fault or within his control (failure of his bank, the changes in health care reimbursement). The government says that these concepts are part of the fraud guidelines and applying them to the tax crime guidelines is an error of law. The fraud guidelines, like the tax guidelines, set offense levels primarily based upon *15 loss. But the fraud guidelines alone refer in comment to the possibility of a departure where computed losses under- or overstate the seriousness of the offense; likewise, the fraud guidelines alone at one time referred to multiple causes as a possible example of an overstatement and while that language has been deleted, they retain that concept in one of the examples. Compare U.S.S.G. § 2F1.1, application note 11 (1990) with application note 10 (1991). See generally United States v. Rostoff, 53 F.3d 398, 406 (1st Cir.1995). We agree with the government that provisions in one set of guidelines cannot normally be transferred to another separate set of guidelines. See United States v. Smallwood, 920 F.2d 1231, 1238 (5th Cir.1991); United States v. Anders, 899 F.2d 570, 580 (6th Cir.1990). The guidelines for each offense or set of offenses tend to function as an integrated unit, containing their own tradeoffs and specifications. Thus, without laying down an iron rule, we view skeptically any importation of language from another offense guideline, absent an explicit cross-reference. Yet this does not take the government very far. The notion in the fraud guideline that the loss table may under- or overstate the seriousness of the offense is little more than another way of saying that departures from the loss table may be warranted for good cause. Even if we treat the fraud guideline’s language as generously inviting a search for such causes, the fact remains that the all-purpose departure provision remains available for tax eases whenever the case falls outside the heartland. See 18 U.S.C. § 3553(b); U.S.S.G. § 5K2.0. The fraud guidelines’ multiple-cause language is a more complicated matter. The government says that the fraud guidelines may need such flexibility because of the diverse situations to which they must apply. By contrast, it says, “loss” for tax purposes is based on calculations, set forth in the guidelines, that (in words of the brief) “focus upon the amount due and owing at the time of the offense.” If a tax evader repays what was stolen, says the government, he merely deserves a few levels off for acceptance of responsibility. Tax loss seems to be a somewhat more protean concept than the government implies, 2 but we think that the argument is beside the point. We are here concerned not with computing the loss — the parties have agreed that it should be treated as “more than 1.5 million” — but rather with whether a departure is proper. And we are dealing not with a tax evader who stole the government’s money and later had a change of heart but with someone who (accepting the district court’s finding) never intended to steal the money at all (or at least most of it). Further, regardless of the fraud guideline, the facts mentioned by the district court in its causation analysis are obviously relevant even if the analysis is not. To distinguish Brennick from the ordinary tax evader, it is essential to show that he did intend to pay over what was owed and was merely deferring payment. This premise would be hard to sustain unless some other cause had contributed to his later failure to pay over the funds. This said, we think that it merely invites confusion to treat “multiple causation” as an independent basis for a departure. And we think that to do so would be inconsistent with the normal presumption that provisions in one guideline are not to be read into the guideline for a different offense — absent an explicit cross reference or some other reason to believe that the Commission so intended. We doubt that this emendation would alter the district court’s desire to depart, but as a remand is required for other reasons, it is free to decide the point for itself. III. While a departure could be justified in theory in this case, we do not think that either the decision to depart or the amount of the departure has been adequately explained. Our reasons are not the usual ones — that the departure is based on an impermissible ground or that there has been no effort to *16 explain the degree of departure. Rather, we think that factors weighing against any departure, and certainly one of this degree, received inadequate attention. In this case the guideline range for Bren-nick was 41 to 51 months; and the 13-month sentence imposed was less than a third of the minimum and just over a quarter of the maximum. A 13-month sentence would be the midpoint in the range for a first time offender who evaded or sought to evade $40,-000 or more in taxes but had no other adjustment. Brennick, of course, caused the government a tax loss of over $1,500,000. It would be easy enough to understand the sentence if Brennick had merely withheld a large payment, reasonably expecting to pay the money shortly but using it in the meantime for business purposes which then unexpectedly collapsed. Absent loss to the government, there would probably not even be a prosecution in such a case; and certainly the intent would be less culpable than in ordinary tax evasion. But Brennick’s actions and intentions were more serious than this abstraction allows. First, Brennick may in some sense have intended repayment, but the reasonableness, and perhaps even the possibility, of such a belief must have lessened over time. To the eve of bankruptcy and apparently beyond, Brennick appears to have deferred payment to the government while withdrawing very substantial sums for his own use. Without more findings, it would be hard to give Bren-nick the benefit of a bona fide intention to repay the entire loss, even if much of it may be encompassed. Second, even apart from an intention to repay, Brennick’s good faith is marred by dishonesty in at least two respects, (even apart from his falsehoods at trial which were the subject of a separate adjustment). On a number of the later returns, Brennick falsely stated or had others misstate that the amounts due to the government had been paid when he knew that they had not. And his elaborate structuring of withdrawals was effectively an effort to mislead and conceal, as perhaps also was his use of multiple employer identification numbers. Third, Brennick committed the crime of structuring and the government points out that the structure counts alone, if no other offense had been committed, could easily have produced an adjusted offense level of 17, 3 and a guideline sentence of 24 to 30 months. The minimum is almost twice the amount of Brennick’s actual sentence after departure. The government has not argued that this makes a departure impermissible as a matter of law, but it certainly bears on the reasonableness and degree of departure. We appreciate that where a ground for departure exists, the district court’s discretion is at its zenith deciding both whether and how far to depart. United States v. Diaz-Villafane, 874 F.2d 43, 49-50 (1st Cir.1989). But the quid pro quo for departures is reviewability, including review for abuse of discretion, 18 U.S.C. § 3742(b)(3); and even if review is hedged by deference, Koon v. United States, 518 U.S. 81, -, 116 S.Ct. 2035, 2046, 135 L.Ed.2d 392 (1996), it has to mean something. In this case, we fail to see how a departure to 13 months can be justified as reasonable on this record in light of the three considerations set forth above, all of which appear to us relevant. We have put to one side Bren-nick’s gambling, the significance of which is a matter of reasonable dispute, and the government’s claim that he deprived his employees of health care, which was neither a charged offense nor clearly relevant conduct. Possibly, even after these factors are considered and weighed in full, there is still warrant for a substantial departure, but we *17 think that some further explanation is essential. Indeed, while the district court takes note of Brennick’s false filings, the government says that the discussion understates them; 4 and the district court’s decision does not squarely address our concerns about Brenniek’s good faith on the later losses or the import of the structuring guideline. The sentence was not imposed casually: the district court conducted a lengthy sentencing and wrote at length, addressing itself primarily to the government’s objections— which we think are overstated. The area is complicated; there is little helpful precedent; and Brennick’s circumstances are unusual. If it takes one more round to fine-tune the sentence, this is a price worth paying. On remand, the district court is free to consider whether its inclination to depart is affected by our conclusion that the fraud guideline should be put to one side. Assuming not, we expect that in resentencing the district court will address the considerations that we have outlined. While expressing doubt that a sentence of 13 months is justified, we impose no mechanical downward limit. What procedure to follow on remand is entirely for the district court to decide. The sentence imposed by the district court is vacated in its entirety and the case is remanded to the district court for further proceedings consistent with this opinion. It is so ordered. Footnotes [1] The government says for the record that the structuring counts should have been grouped separately from the tax counts, which would have resulted in a one-level increase. U.S.S.G. § 3D 1.4. But this caveat was not raised in the district court and is not pursued here. [2] Tax loss is defined somewhat differently for the different tax offenses, compare U.S.S.G. §§ 2T1.1 (a), 2T1.2(a), 2T1.3(a), and 2T1.6(a), and the tax table at 2T4.1 has changed over time. [3] That level might have been anywhere between 15 and 21. Under the 1992 guidelines, the structuring counts generated a base offense level of 13, U.S.S.G. § 2S1.3(a)(l), and would have been adjusted upward two levels for the amount of money involved. U.S.S.G. § 2S1.3(b)(2). Brennick’s two-level adjustment for obstruction of justice would presumably also have applied, generating a level of 17. A further increase of four levels would have resulted if the court determined that "the defendant knew or believed that the funds were criminally derived property.” U.S.S.G. § 2S1.3(b)(l). [4] The district court mentioned two returns filed by Brennick falsely claiming that the amount indicated was paid in full. The government notes that although two false returns were actually signed by Brennick, an additional fourteen false returns were signed by his employees.