United States of America, Plaintiff-appellee, v. Kenneth P. Kontny and Joann L. Kontny, Defendants-appellants, 238 F.3d 815 (7th Cir. 2001)

Annotate this Case
U.S. Court of Appeals for the Seventh Circuit - 238 F.3d 815 (7th Cir. 2001) Argued December 5, 2000Decided January 4, 2001

Appeals from the United States District Court for the Western District of Wisconsin. No. 00-CR-4--John C. Shabaz, Chief Judge.

Before Posner, Easterbrook, and Evans, Circuit Judges.

Posner, Circuit Judge.

The Kontnys were convicted of fraudulent nonpayment of federal payroll taxes and sentenced to prison. Their appeal complains about the denial of their motion to suppress documents and statements that they gave to an Internal Revenue Agent and about a sentencing increase that they received by virtue of the "sophisticated" character of their fraud.

The Fair Labor Standards Act requires employers to pay their hourly employees time and a half for overtime (that is, hours worked above 40 hours a week), but, of course, the overtime wage is taxable income to the employee. To defeat both the overtime and tax laws, the Kontnys, who own an equipment-supply business that employs 25 to 30 workers, concocted the following scheme. They would pay the workers normal wages rather than time and a half for overtime work but not report the overtime wages to the government as taxable income, thus making it easy (or easier) for the workers to avoid detection if they did not report this income on their tax returns. The employees benefited from this scheme by obtaining a greater after-tax income and the Kontnys by not paying either overtime wages at the rate of 1.5 times regular wages or payroll taxes on the overtime wages.

The scheme continued for at least a decade until the Kontnys became embroiled in a bitter labor dispute with their workers. One of them decided to tattle to the government. He visited an office of the IRS and was interviewed by Special Agent Babbitt, a criminal investigator. The matter was turned over to Revenue Agent Furnas to investigate. Revenue agents, unlike special agents, conduct civil rather than criminal investigations. Furnas interviewed a number of employees of the Kontnys' company and concluded that despite their disgruntlement over the labor dispute, they might well be telling the truth. In that event the Kontnys had committed a fraud; and tax fraud is criminal, though more often handled on a civil than on a criminal basis.

Furnas requested an interview with the Kontnys. They agreed. At the interview he explained that he was investigating allegations that they had failed to withhold payroll taxes from overtime payments to their employees. Before Mr. Kontny arrived for the interview, Mrs. Kontny asked Furnas whether she needed to have a lawyer present for the interrogation. He replied that this was "a civil exam" and it was up to her to decide whether she needed to have a lawyer present. But he added that if he discovered fraud he would refer the matter for a criminal investigation. He asked her for various business records, which she gave him, and she made some statements that were later used against her at trial, for example that she realized that payroll taxes have to be withheld from overtime wages. In a follow-up phone call from Furnas a few days later she mentioned that she had shredded some checks that Furnas had inquired about. At the mention of the shredded checks his suspicions crystallized and he decided that he now had firm indications that the Kontnys had committed tax fraud and he turned the case over to the criminal investigatory arm of the IRS and had no further contact with the Kontnys.

As an original matter it is extremely difficult to see what possible basis there could be for a motion to suppress in this case. Confessions or other admissions obtained in the course of an interrogation are deemed involuntary and therefore inadmissible only if they are procured by threats or promises. Bram v. United States, 168 U.S. 532, 542-43 (1897); Johnson v. Trigg, 28 F.3d 639, 641-42 (7th Cir. 1994); United States v. Glover, 104 F.3d 1570, 1579 (10th Cir. 1997); United States v. Guerrero, 847 F.2d 1363, 1366 (9th Cir. 1988). The Miranda rule is not in play here since the interrogation of the Kontnys by agent Furnas was not custodial. Beckwith v. United States, 425 U.S. 341 (1976); compare Mathis v. United States, 391 U.S. 1 (1968). But the fact that the Kontnys were not in custody has a broader significance. Virtually all cases involving coerced confessions involve the questioning of a suspect who is in police custody, an inherently intimidating situation in which people find it difficult to stand up for their rights or even to think straight. The situation is different when a person who does not even know that he is a criminal suspect (that is a premise of the Kontnys' appeal) is being interviewed in his home, and by a civil rather than a criminal investigator to boot. Furnas was unarmed, un-uniformed, unaccompanied. The Kontnys were at no disadvantage in dealing with him. They were under no pressure to answer his questions. Any answers they gave were voluntary.

Trickery, deceit, even impersonation do not render a confession inadmissible, certainly in noncustodial situations and usually in custodial ones as well, unless government agents make threats or promises. Frazier v. Cupp, 394 U.S. 731, 739 (1969); Holland v. McGinnis, 963 F.2d 1044, 1051 (7th Cir. 1992); United States v. Rutledge, 900 F.2d 1127, 1131 (7th Cir. 1990) ("far from making the police a fiduciary of the suspect, the law permits the police to pressure and cajole, conceal material facts, and actively mislead"); United States v. Byram, 145 F.3d 405, 408 (1st Cir. 1998) ("trickery is not automatically coercion. Indeed, the police commonly engage in such ruses as suggesting to a suspect that a confederate has just confessed or that police have or will secure physical evidence against the suspect. While the line between ruse and coercion is sometimes blurred, confessions procured by deceits have been held voluntary in a number of situations"). And these were custodial cases. Nothing is more common in the noncustodial setting of police investigations than for an undercover police officer to extract a damaging admission from a criminal suspect simply by pretending to be another criminal. The admission is usable in evidence against the suspect even though he would never have spilled the beans to the officer had he known the officer's status. Planting informers is not an unconstitutional method of collecting evidence for use in criminal trials. Illinois v. Perkins, 496 U.S. 292, 298-99 (1990); Hoffa v. United States, 385 U.S. 293, 303-04 (1966). There is no right to require secrecy of the people whom one confides in.

So even if Furnas was pretending to be conducting a civil investigation but was really, as the appeal argues, conducting a criminal one, this would not, under the rules that govern the admissibility of incriminating statements (written or oral) made to government officers even by a suspect who is in custody, make the statements inadmissible. The circumstances did not remotely prevent the Kontnys from making a rational decision about whether to play ball with Furnas. United States v. Lawal, 231 F.3d 1045, 1048 (7th Cir. 2000) ("a confession is voluntary if the totality of the circumstances demonstrates that it was the product of rational intellect and not the result of physical abuse, psychological intimidation, or deceptive interrogation tactics calculated to overcome the defendant's free will"); United States v. Westbrook, 125 F.3d 996, 1006 (7th Cir. 1997) ("nothing in this record leads us to believe the agents misled him or exploited Mr. Westbrook's anxiety to the point that he was unable to make a rational decision about whether to confess"); Sprosty v. Buchler, 79 F.3d 635, 647 (7th Cir. 1996) ("the police did not magnify or exploit Sprosty's fears, anxieties and uncertainties to the point where he was unable to make a rational decision about whether to confess"); United States v. Doucette, 979 F.2d 1042, 1045 (5th Cir. 1992) ("a confession is voluntary if, under the 'totality of the circumstances,' the statement is the product of the accused's 'free and rational choice'"); United States v. Velasquez, 885 F.2d 1076, 1089 (3d Cir. 1989) ("although the deception [by the police] may have been a partial cause of Velasquez's statements, we do not think that her will was overcome or her capacity for self- control vitiated"); United States v. Guerrero, supra, 847 F.2d at 1365 ("an inculpatory statement is voluntary only when it is the product of a rational intellect and a free will"). We might have a more difficult case had Furnas gone further and promised the Kontnys they would not be prosecuted if they played ball with him, for that conceivably is the kind of false promise that might induce a rational person to rely. United States v. Baldwin, 60 F.3d 363, 365 (7th Cir. 1995), vacated and remanded on other grounds, 517 U.S. 1231 (1996); United States v. Rutledge, supra, 900 F.2d at 1130. Furnas did not do that. On the contrary, he as much as warned the Kontnys that any evidence they provided of fraud would lead to a criminal investigation. As we have said, he didn't have to go further and give them Miranda warnings.

It is true that the Internal Revenue Service by regulation requires that a civil investigation cease when the investigator develops firm indications of fraud, Internal Revenue Manual sec.sec. 4565.21(1), 9311.83(1), which the Kontnys argue happened before the fatal interview and the check-shredding phone conversation. But the federal exclusionary rule, which forbids the use of evidence obtained in violation of the Fourth or Fifth Amendments, does not extend to violations of statutes and regulations. The Supreme Court so held in United States v. Caceres, 440 U.S. 741, 755 (1979), with specific reference to a regulation of the IRS. See also United States v. Peters, 153 F.3d 445, 456 (7th Cir. 1998); United States v. Michaud, 860 F.2d 495, 498-99 (1st Cir. 1988); Groder v. United States, 816 F.2d 139, 142 (4th Cir. 1987), and, for application of the principle outside the tax area, United States v. Chaparro-Alcantara, 226 F.3d 616, 621 (7th Cir. 2000); United States v. Page, 232 F.3d 536, 2000 WL 1682523, at *3 (6th Cir. Nov. 9, 2000); United States v. Hinton, 222 F.3d 664, 674-75 (9th Cir. 2000); United States v. Felipe, 148 F.3d 101, 109 (2d Cir. 1998); United States v. Hensel, 699 F.2d 18, 29-30 (1st Cir. 1983). The Kontnys do not claim to have relied, reasonably or unreasonably, on the existence of the regulation that required Furnas to back off as soon as he obtained firm indications of fraud. United States v. Caceres, supra, 440 U.S. at 752- 53; United States v. Ani, 138 F.3d 390, 392 (9th Cir. 1998); United States v. Pipes, 87 F.3d 840, 842 (6th Cir. 1996). But this means, as Pipes makes clear, that there was no causal relation between Furnas's alleged violation of the regulation and the Kontnys' decision to make incriminating statements. "The defendant obviously did not know that the officers were violating [the statute]. Thus, the officers' failure to comply with [it] had no impact on defendant's decision to commit the offense." Id. Nor is there any suggestion that Furnas violated the prohibition against enforcement of a summons for tax records after a matter has been referred to the Justice Department for possible criminal prosecution. 26 U.S.C. § 7602(c) (1); United States v. Michaud, 907 F.2d 750 (7th Cir. 1990) (en banc).

A number of decisions explore at length the nebulous distinction in the IRS regulation between "first" and "firm" indications of fraud; Furnas only admitted that he had the former sort before he interviewed the Kontnys. But the cases generally and we think rightly do not treat the distinction as an independent basis for determining whether evidence obtained in an IRS investigation is admissible. They treat it merely as a factor to be considered in evaluating the defendant's constitutional claim. The defendant must prove that "the IRS's conduct resulted in prejudice to defendant's constitutional rights." United States v. Peters, supra, 153 F.3d at 452 n.10; see also United States v. Grunewald, 987 F.2d 531, 534 (8th Cir. 1993); United States v. Knight, 898 F.3d 436, 438 (5th Cir. 1990), and the concurring opinion in Peters, 153 F.3d at 462-64.

There are some outliers, such as United States v. McKee, 192 F.3d 535, 541 (6th Cir. 1999), which states (in dicta, as the concurring judge pointed out, id. at 545), reflecting a common but perhaps excessive hostility to the Internal Revenue Service, that section 4565.21(1) of the IRS manual is "mandated by the Constitution." It is true as we have noted that Caceres left the door slightly ajar by indicating that it might be a denial of due process to induce reasonable reliance on the regulation and then pull the rug out from under the defendant; but nothing of that kind is involved in this case. United States v. Tweel, 550 F.2d 297, 299 (5th Cir. 1977), contains broad McKee-like language, and has been cited frequently. But besides having been decided before Caceres, it was a case, unlike ours, involving the issue of consent to a search, and the defendant in giving his consent was held to have relied reasonably on the agent's promise that the investigation was purely civil. United States v. Powell, 835 F.2d 1095, 1098 (5th Cir. 1988). The government had broken its promise, and we know that admissions extracted by false promises are sometimes excluded as being involuntary. There were no promises in the present case.

Peters goes on to state that the defendant must show "affirmative misrepresentations," "affirmative deceit," or "affirmative misleading" (153 F.3d at 456-57) (these terms are synonymous), but it would be a mistake to infer that such a showing without more requires exclusion of incriminating statements. Proof of deceit must be linked up to the constitutional standard of threat or promise. Deceit by itself is neither, though it can be the basis of either- -if Furnas had pretended to be a representative of the Mob and told the Kontnys that they would be killed if they didn't turn over their business records to him, or pretended to be an Assistant U.S. Attorney and assured them they would not be prosecuted if they cooperated with him, the Kontnys might have a sound ground for exclusion. Cf. Arizona v. Fulminante, 499 U.S. 279, 287-88 (1991). They showed nothing of the sort, and must therefore lose even if the district court clearly erred (the applicable standard of appellate review, United States v. Peters, supra, 153 F.3d at 459; United States v. McKee, supra, 192 F.3d at 543; United States v. Wadena, 152 F.3d 831, 851 (8th Cir. 1998))--which, incidentally, it did not--in finding that Furnas did not have firm indications of fraud before he interviewed the Kontnys. That issue is not determinative. A failure to terminate a civil investigation when the revenue agent has obtained firm indications of fraud does not without more establish the inadmissibility of evidence obtained by him in continuing to pursue the investigation. There is nothing more here.

Moving to the sentencing issue, we confront the argument that the efforts the Kontnys made to conceal their scheme of tax evasion did not amount to the "sophisticated concealment" that requires a two-level sentencing bonus under U.S.S.G. sec. 2T1.4(b) (2). That they did make such efforts is not in question. They wrote separate checks to the employees, one for regular wages and one for overtime, and sometimes the overtime checks would include reimbursement for expense items to disguise the fact that the checks were for wages. The Kontnys programmed their computer so that the amount of the overtime checks was classified in nonwage expense categories. The stubs for the overtime checks, which they gave their accountant, likewise placed the expense in nonwage categories.

But did these efforts amount to "sophisticated concealment"? They were not very sophisticated in the lay sense of the word, especially in context. By creating a fraud that involved the knowing participation of more than two dozen employees, they not only armed the employees to blackmail them but greatly increased the risk of eventual detection, though it is true that the fraud persisted for at least a decade before the inevitable occurred. The Kontnys' efforts at concealment were sophisticated in relation to a case in which the owner of a shop evades taxes by emptying the drawer of the cash register before counting the day's cash receipts and puts the cash thus skimmed into a shoebox and slides it under his bed, but unsophisticated in relation to a scheme of evasion that does not depend on the continuing goodwill of one's entire workforce and that creates a paper trail that is more difficult to follow to its guilty conclusion than the one the Kontnys created.

The existence of a statutory sentencing range reflects the fact that criminal acts that involve the same statutory elements (in the case of criminal tax fraud they are essentially that the defendant knowingly made a materially false return, 26 U.S.C. § 7206(1); United States v. Pirro, 212 F.3d 86, 89 (2d Cir. 2000)) may differ in circumstances that are pertinent to the appropriate penalty. One criminal act may be much more lucrative for the offender because it involves a very large amount of money relative to the cost of committing the offense, and so a heavier punishment will be necessary to deter. Another may be more lucrative than the average not because it involves a larger take but because the probability of detection is lower; an economist would say in such a case that the "expected" profit of the crime was greater. The existence of a sentencing range as opposed to a sentencing point allows these differences to be reflected in sentencing. The federal sentencing guidelines guide and discipline the judge's choice of the sentence within the range. They do this by fixing a sentencing range (narrower than the statutory range) for the average offense within the offense category (here, criminal tax fraud) and by prescribing bonuses and discounts to adjust for relevant differences between the average and the particular offender's offense.

The more sophisticated the efforts that an offender employs to conceal his offense, the less likely he is to be detected, and so he should be given a heavier sentence to maintain the same expected punishment, and hence the same deterrence, that confronts the average offender. Implementation of this rule requires both determining how much the average offense is concealed and relating the guideline concept of "sophistication" to deterrent needs. The complication in the first half of this inquiry is that fraud is by nature self-concealing--its success depends on its being hidden from the victim. The average criminal tax fraud thus involves some concealment; "sophisticated" tax fraud must require more. A parallel distinction has arisen in determining when statutes of limitations in fraud cases are tolled. If concealment were enough to toll such a statute of limitations, the statute would be tolled in almost every case, because fraud is inherently covert. So the courts distinguish between the initial fraud and any distinct efforts at cover up ("fraudulent concealment") and toll the statute only when the defendant has resorted to such efforts. Wolin v. Smith Barney Inc., 83 F.3d 847, 851 (7th Cir. 1996); Martin v. Consultants & Administrators, Inc., 966 F.2d 1078, 1093-95 (7th Cir. 1992). Likewise the concealment that is inherent in criminal tax fraud, as in our shoebox example, must be distinguished from efforts over and above that concealment to prevent detection. Only the latter permit the sentencing enhancement.

In light of its purpose and context, we think "sophistication" must refer not to the elegance, the "class," the "style" of the defrauder--the degree to which he approximates Cary Grant--but to the presence of efforts at concealment that go beyond (not necessarily far beyond, for it is only a two-level enhancement that is at issue, which in this case added roughly six months to the defendants' sentences) the concealment inherent in tax fraud. It is true that the guideline commentary illustrates with examples suggesting a higher level of financial sophistication: "'sophisticated concealment' means especially complex or especially intricate offense conduct in which deliberate steps are taken to make the offense, or its extent, difficult to detect. Conduct such as hiding assets or transactions, or both, through the use of fictitious entities, corporate shells, or offshore bank accounts ordinarily indicates sophisticated concealment." U.S.S.G. sec. 2T1.4, Application Note 3. But these are offered as examples, as emphasized in United States v. Friend, 104 F.3d 127, 130 (7th Cir. 1997), and United States v. Clements, 73 F.3d 1330, 1340 (5th Cir. 1996); the essence of the definition is merely "deliberate steps taken to make the offense . . . difficult to detect." When the term "sophisticated" is defined so, it becomes apparent that the district judge did not commit a clear error (the applicable standard of appellate review of this ruling too, e.g., United States v. Madoch, 108 F.3d 761, 765 (7th Cir. 1997); United States v. Aragbaye, 234 F.3d 1101, 1107 (9th Cir. 2000)) in enhancing the defendants' sentences.

The Kontnys point out that the government rarely prosecutes criminal tax fraud that is not "sophisticated" in the sense indicated by the facts of this case. Armed as it is with fearsome civil remedies involving huge penalties--for example the 75 percent penalty for taxes fraudulently not paid, 26 U.S.C. § 6663--the government brings few criminal tax cases (fewer than 700 a year) relative to the amount of tax fraud; and perhaps none against defendants less sophisticated than the Kontnys. We do not know this to be the case, but will assume it is for the sake of argument. No matter. The question is what the Sentencing Commission took to be the average criminal tax fraud when it promulgated the "sophisticated concealment" guideline back in 1987. That would be the benchmark for courts to use to decide whether the Commission would have wanted the sentences of the Kontnys increased by reason of the character or extent of their efforts at concealment. The government's lawyer told us without contradiction from his opponent that before the guidelines era the federal government prosecuted many unsophisticated criminal tax frauds, as illustrated by our shoe box case. The defendant would usually plead guilty and the judge impose a light sentence ("roughly half of all tax evaders were sentenced to probation without imprisonment, while the other half received sentences that required them to serve an average prison term of twelve months," U.S.S.C. sec. 2T1.1, Background Commentary), and sentences in those days were essentially unappealable unless they exceeded the statutory maximum. So these were easy cases for the government. When the guidelines came into force, limiting sentencing discretion, the government shifted its focus to the more serious cases, not wanting to become involved in trials of minor cases when under the guidelines defendants might be reluctant to plead guilty because they would be facing a heavier sentence and might, like so many other federal criminal defendants these days, appeal their sentences. So today the average criminal tax fraud that is prosecuted is more sophisticated than when the concept of sophistication was introduced into the guidelines. That is no reason for thinking the Commission would consider the enhancement imposed in this or like cases excessive even if they are the only type of criminal tax fraud being prosecuted nowadays.