August 1998

Distinguishing Civil and Criminal Examinations: The Point They Turn Ugly

       In United States v. Peters, ___ F.3d ___ (7th Cir., 8/17/98), the Court commenced its decision with the following:

This case is largely concerned with the distinction between a civil tax audit, which is conducted by revenue agents in the Examination Division of the Internal Revenue Service, and a criminal tax investigation, which is conducted by "special agents" in the IRS's Criminal Investigative Division (CID)

Since practitioners in tax controversy matters must understand this distinction, we deal at some length with this case and the differences between the majority and concurring opinions.

    As always, the facts are critical to frame the discussion.  The facts, highly summarized are:  An apparently disgruntled ex-husband called a CID agent to report that his ex-wife, a podiatrist, was cheating on taxes.  The ex-husband gave substantial documentation to the CID agent, but the CID agent did not commence a formal investigation.  Rather, over a period of 18 months, he would occasionally review the file which he kept "at the bottom of his desk" and, on one occasion, he drove by the taxpayer's residence to see what he might glean about her life style.  The CID agent then "walked" the information to the Examination Division ("Exam") for its consideration and thereafter did no further investigation.  Exam assigned the case to a rookie agent who treated the case as a routine civil audit because, in her experience (apparently limited), nothing serious came from informants' information.  The agent began the examination with the routine correspondence to the taxpayer and thereafter conducted the case as routine examination.   At no time did the agent advise the taxpayer of the origin of the investigation or that the case might turn ugly (ugly being criminal).  Specifically, she did not advise the taxpayer of her rights.  Factually, that was explainable because the agent had not yet determined that the case was anything other than a civil audit.  But, at some point in the investigation, the agent reached the conclusion that the case should be referred to CID and promptly prepared a referral report.  Her new supervisor, however, determined that the civil agent's referral report had too many errors to substantiate a fraud referral and that, although there were "indicia of fraud" the taxpayer should be given an opportunity to explain.  The supervisor assigned the case to two other civil agents.  The new agents performed some work, including obtaining a consent to extend the statute of limitations.  Thereafter, they submitted a fraud referral.  The supervisor approved the referral indicating that there was then "firm indications of fraud."  The case processed criminally from CID all the way through conviction of the taxpayer.

    The taxpayer moved at her criminal trial to suppress the documents and the information she had provided the agents during the civil investigation, alleging that they were obtained by the IRS in violation of the Fourth and Fifth Amendments to the Constitution.

    Using these facts and arguments as a fulcrum, the majority opinion launched into a discussion of the civil and criminal tax enforcement regimes.  The general parameters are (or should be) known by all tax practitioners -- the civil examination is concerned exclusively with determining the correct tax liability and asserting any civil penalties that might apply; the CID or criminal examination is concerned with applying the criminal penalties of the Code.  The Court then summarized:

Under IRS regulations, a revenue agent who uncovers a "firm indication of fraud on the part of the taxpayer" must immediately suspend her audit and refer the case to the CID. See Internal Revenue Manual 4565.21(1). At that point, the CID enters the case and the IRS' efforts become focused on the possibility of criminal prosecution.

    The taxpayer argued that the civil agents (first the "rookie" and then the others) were conducting a criminal investigation in the guise of a civil investigation, without the constitutional safeguards that normally attend a criminal investigation (including e.g., reading the rights, etc.).  The civil investigation thus, she urged, constituted an unreasonable search and seizure and was violative of due process because the investigation was conducted under false pretenses.   The Court concluded that both constitutional claims rested upon the premise that the taxpayer's cooperation in the civil investigation was induced by "fraud, deceit, trickery or misrepresentation by the revenue agent" and addressed whether the facts showed that level of misconduct.

    Based on the authority in the area, the simple failure to advise the taxpayer of the potential criminal focus of the inquiry is not alone sufficient to rise to the level of a constitutional violation.  Rather, the majority reasoned, the words or actions of the civil agents must rise to the level of an affirmative misrepresentation as to the nature of the investigation.  For example, if the taxpayer or her representative had expressly inquired whether the investigation had any criminal ramifications and was provided a misleading response (either directly or by failure to respond in circumstances designed to mislead), there might be the type of affirmative misrepresentation violating constitutional rights.  Thus, viewed the issue in the case turned upon whether the civil agents were conducting a civil or a criminal investigation, and the trial court had found that they were conducting a civil investigation until the end when they developed the firm indication of fraud that resulted in the referral to CID.

    The Court of Appeals noted the tough nature of this inquiry into whether the civil investigation was continued after it should have been forwarded for criminal investigation.  It requires the court to delve into the good faith of the agents, and can be bogged down in a turbulent sea of subjectivity which the courts should be loath to second-guess except in the most egregious of circumstances.   The extreme cases (such as the leading case United States v. Tweel, 550 F.2d 297, 299 (5th Cir. 1977)) must be dealt with, as even the IRS Manual requires, so that civil examinations are not used as a guise to criminal investigations.  In Tweel the civil investigation arose in the first instance at the request of the DOJ's Organized Crime and Racketeering Section and clearly had an unspoken, deceptively so, criminal focus that the Fifth Circuit was inconsistent with the government's obligations to the citizen.   There, the taxpayer's representative asked if a CID agent was involved.  The agent answered no, a technically correct answer which failed to disclose the criminal investigation by Organized Crime and Racketeering.  [No, Bill Clinton was not the agent.]  The Court found that the answer was misleading since the gravamen of the request was whether there was a criminal investigation involved.

     The majority then began to distill and lay out certain indications that the civil examination may have become criminal.  First, if the CID agent were to become involved in the civil investigation prior to formal referral.   In recognition of this concern, the Manual forbids such involvement.  Second, continuing of active civil audit after the civil agent begins preparation of the fraud referral "also may be indicative of an agency attempt to gather information for a criminal prosecution while keeping the taxpayer in the dark as to the true nature of its investigation."  In recognition of this concern, the Manual directs the civil agent to suspend her investigation at the earliest opportunity after developing firm indications of fraud.

    The inquiry therefore is into when the civil agent has developed firm indications of fraud.  There is no litmus test, but the Court distilled the case law as follows:

    The case law suggests that a revenue agent has developed a firm indication of fraud when she has established that the taxpayer has engaged in a consistent pattern of substantial underreporting of income and/or overstatement of deductions such that an intent to evade taxes can be inferred.

    The case law recognizes that an assessment of the taxpayer's intent is the most critical element in a revenue agent's determination of whether "firm indications of fraud" exist in any particular case. * * * * Indeed, it is the taxpayer's intent to evade taxes that differentiates a criminal violation from a civil case. For this reason, a revenue agent who discovers potential violations of the revenue laws will almost always give the taxpayer an opportunity to explain the violations before determining the appropriateness of a criminal referral.    This step is viewed as a necessity because, given the complexity of the tax code, many innocent explanations exist for unreported income. **** Once the revenue agent has interviewed the taxpayer, he must assess the taxpayer's explanation. During that process, the revenue agent may consider the proffered explanation in light of that particular taxpayer's knowledge of taxes and business practices. **** In fact, one court has held that a revenue agent engages in constructive dishonesty when she continues a civil audit after receiving an incredible explanation from a taxpayer with extensive knowledge of the tax laws. See United States v. Toussaint, 456 F. Supp. 1069, 1071-72 (S.D. Tex. 1978).

    Finally, the majority distinguished between the firm indication of fraud and the first indication of fraud, quoting the revenue manual (4565.21(1)) as follows:

A firm indication of fraud must be distinguished from a first indication of fraud. A first indication of fraud can be described as a mere suspicion of fraud. Examiners are legally permitted and should endeavor to ask the taxpayer, the preparer, the representative, or any other involved party for an explanation of the "discrepancies" which are the basis of the examiners['] suspicion of fraud and any other question(s) which will resolve the question of the taxpayer's intent.

    Returning to the governing issue, the majority reasoned that the inquiry is whether the taxpayer was misled and that issue turns upon when the civil agents had firm indications of fraud.  Focusing that inquiry on the facts at hand, the majority concluded that the district court's finding that there were not firm indications of fraud at each critical step -- from the CID agent's walking the file to Exam up to the point of Exam's formal criminal referral.  The majority found comfort in the lack of record evidence that the CID agent remained involved after he had thus walked the file.  (In this regard, the mere fact that thereafter the CID agent received information and passed it on to Exam, was viewed by the majority as passive conduit type of activity, rather than active involvement in the civil examination.)   The majority further found support in the way the initial agent had consistently treated it as a routine audit up to the point of her abortive referral report.  The majority further found comfort in the finding that the supervisor's rejection of the initial fraud referral recommendation and assignment of two new civil agents was a good faith determination that the case was not yet ripe for fraud referral.  Since a fraud referral is serious business, it is best that the case be properly developed to show the firm indications of fraud before referring.  The majority further found comfort in the civil agents' seeking of a consent to extend the statute of limitations, a consent that would be unnecessary if the agents were convinced of fraud.  Thereafter, when the civil agents began to suspicion fraud, it was not so firm that the taxpayer should not be given the opportunity to explain.  The Court specifically noted in this regard that, until there were firm indications of fraud, the IRS had no duty to warn the taxpayer that the audit could lead to a criminal investigation or that the audit was based upon an informant's tip.  On the basis of the foregoing, the majority concluded that the district court's findings were not clearly erroneous.

    Judge Easterbrook concurred in the result but raised some troubling points about the majority's analysis.  Judge Easterbrook asked pungently:  "Why should a solid basis for believing that the suspect has committed a crime require the government to curtail its investigation?" Judge Easterbrook viewed the inquiry into whether there was firm indications of fraud as opposed to first indications of fraud merely a "snipe hunt," distracting and irrelevant.   The inquiry, he reasoned, was whether the taxpayer's cooperation in providing information was voluntary, a decision that is based upon the taxpayer's state of mind and assessment of risks.  Thus, for example, an examination that is purely civil under any test can turn criminal if the taxpayer supplies records that incriminate him or her.   Yet, when the agent asks for the records (regardless of the state of the agent's mind), the taxpayer must make the decision whether to turn them over, with the risk of prosecution that that necessarily entails, and the taxpayer's voluntariness in doing so cannot be determined based upon the state of mind of the agent.  In this regard Judge Easterbrook notes that there are all sorts of contexts in which the government conducts investigations which, intentionally or not, may mislead the taxpayer as to the focus of the investigation.  A "sting" investigation is the classic case, and they are not hobbled with any notion that the government must suffer suppression for not disclosing to the stingee that he or she is under investigation.  Accordingly, Judge Easterbrook would have dealt with the case at hand simply by concluding that the IRS had not deceived the taxpayer period, without the digressions that the majority took to develop a rationale that he did not agree with.

    What do we learn from the case?  At least the following:

1.     Assuming that the majority view prevails, as it Tweel seems to suggest it does now for those of us lucky enough to practice in the Fifth Circuit, a key consideration in any tax audit that has fraud potential is to specifically ask the question and get a specific response that the case is not criminal.  Yet, merely asking the question may suggest to the agent that the case, while not then criminal, may have criminal potential as to which he must devote time and resources to see if there are firm indications of fraud.  Accordingly, the drill in most sensitive cases will be to ask the question in a way that does not not generate that response.

2.     If you have determined to take the risk to ask the question, you must do so in a way that, if push really does come to shove, you can prove the question and the response.  This too is sensitive, for too much formality into the contemporaneous proof is likely to cause the agent's suspicion and be conterproductive.

3.     As in all audits, you have to be careful what documents and information you provide the agent.  Listen carefully to the agent's specific description of the information and/or documents he or she requires; preferably make the agent make the requests in writing (do this as a routine practice so as to not draw attention to any specific request).  Respond only to the request the agent makes, not the one you (or the agent for that matter) think he intended to make.   (This may sound like too much of a Bill Clinton type technicality.)  Then, if he makes the requests in clear and unequivocal language (preferably written) for information and documents that, if produced, will be incriminating, you must consider your alternatives seriously.  Do you provide the agent the information or documents surrounded by other information or documents that may detract from drawing attention to the incriminating information?  You obviously cannot mislead the agent (remember, it is your client's problem, not yours, so don't make it your problem).  But you do not have to draw the agent's attention to the incriminating information.  And, if the risks are too high, do you recommend that your client simply not respond and suffer the risk (likelihood) that the agent will use the summons to get the information or documents subject to any privileges your client may assert (a whole new topic beyond the scope of these comments).

[Return to Top]

When May the Taxpayer Insist Upon Substance Over Form -- Tales of Woe

    In Nestle Holdings, Inc. v. Commissioner, ___ F.3d ___ (2d Cir. 7/31/98), the taxpayer acquired Carnation Company. It elected Section 338, using the residual method to step up the basis in the intangibles. Shortly after the purchase, the taxpayer sold the intangibles to its foreign parent corporation for the then appraised value and claimed a slight loss. Upon audit, the IRS determined that the intangibles were worth far less than taxpayer claimed, both upon the taxpayer’s original acquisition and upon the ensuing sale to the parent corporation. That factual determination, if correct, of course, that (1) the taxpayer had a much lower basis for the intangibles, (2) the parent had substantially overpaid for the intangibles and (3) as a consequence the taxpayer had a whopping capital gain on a transaction that it had structured to produce a small loss. The taxpayer was not pleased and went to the Tax Court. It lost. On appeal, the taxpayer urged that (1) any excess of the amount paid to a wholly owned subsidiary over the value of the asset purchased was a capital contribution that is not included in the subsidiary’s amount realized upon the sale and (2) in any event, the intangibles were worth more than the Tax Court had held.

    Addressing the first issue, the Court held the taxpayer to the transaction it had structured -- i.e., a sale. Although it is standard analysis in the tax law that payment by a related party of an amount exceeding or less than the fair value of the property purchased can be recharacterized as being a purchase for the fair market value with the difference then being accounted for as the circumstances suggest. Thus, in the case of a parent’s purchase of property from a subsidiary, a purchase for less than the fair market value could be recharacterized as a dividend from the subsidiary. Correspondingly, a purchase for more than the fair market value can be recharacterized as a contribution to the capital of the subsidiary. The question is who can recharacterize. The Court in a cursory analysis held the taxpayer to the form of the transaction it had structured -- i.e., a sale in which the entire payment amount enters the amount realized calculation in computing gain.

    The Court then moved to the valuation issue -- had the subsidiary (the U.S. taxpayer) really sold for more than the fair market value and, if so, what was the fair market value. The Tax Court had sustained the IRS in asserting that the intangibles could be value under the "relief from royalty" method of valuation. That method values intangibles based upon the royalties that would have been paid for the use of the intangibles -- i.e., "the net present value of the stream of royalty payments from which the purchaser of a mark is relieved. This stream is calculated by (i) determining if the trademarks are profitable, or capable of being licensed, (ii) picking a royalty rate for each trademark, and (iii) multiplying this rate by the estimated revenue stream of the product associated with the mark."

    The Court rejected the relief from royalty method because, in its view, it "necessarily undervalues trademarks." The Court reasoned that the drill is the willing-buyer, willing-seller standard. Ownership of intangibles may be worth more than the relief from royalties because:

Additional Comments on these Cases:

• The rejection of the relief from royalty method as a method for valuing   intangibles, while a loss in Nestle, will undoubtedly be redound to the IRS's benefit in future cases where the taxpayer desires the lowest possible valuation.
• Note that, in Nestle, the IRS could have insisted upon substance over form to treat the excess purchase price as a contribution to capital.  Even more certainly, it could have invoked Section 482 to tax the sale at fair market value.  This points out again that Section 482 is a one-way street and, while Courts sometimes mouth the notion that a taxpayer too can rely upon substance the same as the Commissioner, the various rules noted above put substantial barriers in the way of success.
• Obviously, as noted, the taxpayer would have prevailed if the IRS had invoked Section 482.  Although apparently that case offered no window of opportunity, taxpayers and their representatives should remember the lesson of Pikeville Coal Co., et al. v. United States, 37 Fed. Cl. 304 (1997), unofficially reported at 97 TNT 15-8 (1/23/96) and 79 AFTR2d Par. 97-353, where, the IRS having invoked Section 482, successfully insisted that the fair market price standard in Section 482 should be taken all the way (farther than the IRS wanted to go).

[Return to Top]

Tidbits

• Estate of Bessie I. Mueller v. Commissioner (6th Cir. 8/20/98).   In this case, the Court held that the Tax Court lacks equitable jurisdiction to the right of equitable recoupment.  For years, nobody believed that the Tax Court had equitable jursidiction.  But, in this case, the Tax Court, in a reviewed decision, finally decided that, indeed it did, at least to the limited extent that it could apply the doctrine of equitable recoupment just as the District Courts could in refund suits.   The Sixth Circuit viewed this as an unwarranted extension of the Tax Court's jurisdiction, saying that the Tax Court could only determine the validity of the deficiency before it and that whether taxpayer may be due money from the IRS for some other reason was simply not relevant to the issue of the amount of the deficiency or the refund for the year before the Court.

[Return to Top]

Return to Townsend & Jones, L.L.P. HomePage