Three Felonies a Day

by Harvey Silverglate

  Temple’s advice, while it later played a substantial role in getting Andersen convicted of obstruction of justice, was “the kind of thing lawyers do all the time,” observed Stephen Bokat, vice-president and general counsel of the U.S. Chamber of Commerce, which later filed a friend-of-the-court brief on behalf of Andersen with the Supreme Court. “There but for the grace of God go I,” he added in an interview by Tony Mauro in The National Law Journal.8

  On October 23, Andersen employee David B. Duncan, who had for years headed the firm’s engagement with Enron, directed compliance with the document policy. All of these steps, by various Andersen personnel, seeking to assure compliance with the firm’s document-retention policy, the Supreme Court noted, were followed by substantial destruction of paper and electronic documents. This, of course, was entirely predictable, since the policy, like all such policies that are ubiquitous across all organizations that generate paper documents, directed not only retention, but also destruction of documents after a certain period of time, in the absence of a likely, threatened, or actual proceeding to which they might be relevant.

  It was not until October 30 that the SEC opened its formal investigation and sent Enron—not Andersen—a letter requesting accounting documents. Document destruction at Andersen continued, noted the Supreme Court, “despite reservations by some of [Andersen’s] managers.” It was not until November 8 that the SEC served Andersen itself with subpoenas for records. The next day, Duncan had his secretary send an email that stated: “Per Dave [Duncan]—No more shredding…. We have been officially served for our documents.”

  Andersen was indicted in March 2002. The charge was that the firm “did knowingly, intentionally and corruptly persuade…[its] employees” to withhold and alter documents for use in “official proceedings.”

  The two sides had argued vigorously at the trial over how the trial judge should instruct the jury on the meaning of the word “corruptly.”9 In the end, the jury was told by the judge that, “even if [Andersen] honestly and sincerely believed that its conduct was lawful, you may find [Andersen] guilty.” The trial judge agreed with the DOJ’s position, and instructed the jury that it could convict “if it found [Andersen] intended to ‘subvert, undermine, or impede’ governmental fact finding by suggesting to its employees that they enforce the document retention policy.” The jury convicted, and Andersen was fined $500,000, though the fine was the least of its problems by that time, since the indictment, not to mention the conviction, had already sounded Andersen’s death knell.

  The Supreme Court noted that these instructions called for the conviction of Andersen merely because of the destruction of documents pursuant to the firm’s document-retention policy. “No longer was any type of ‘dishonest[y]’necessary to a finding of guilt,” wrote the high court, “and it was enough for [Andersen] to have simply ‘impede[d]’ the Government’s factfinding ability.” In the end, the Court vindicated Andersen, and in May 2005 unanimously overturned the firm’s conviction, owing to problems with the trial judge’s jury instructions. The law was intended, said the Court, to punish the act of “knowingly… corruptly persuad [ing] another to impede an official investigation.” At the crux of the dispute was the notion that there could have been perfectly lawful reasons for Andersen, or any other party in a similar situation, to follow its document-retention policy before receiving a formal notice from the government that an investigation was underway. Not every act that makes the government’s work harder can be considered a criminal obstruction of justice. The high court noted that persuading a person to “withhold” testimony or documents from a government proceeding “is not inherently malign.” It gave some readily comprehended examples, such as “a mother who suggests to her son that he invoke his right against compelled self-incrimination…or a wife who persuades her husband not to disclose marital confidences.” Likewise, it is not corrupt for a lawyer to persuade or advise a client to “withhold” documents from the government in the course of an investigation where, for example, those documents might be covered by a legal privilege. If there is a legal basis for the client’s refusing to turn over documents to the government, the lawyer who gives such advice is surely “obstructing” the investigation and even intending to do so, but is not doing so with corrupt intent.

  The Court’s criticism of the Department of Justice, of the trial judge who followed the department’s lead in crafting instructions to the jury, and of the Congress that wrote such an obtuse statute, was even more stinging. “Indeed, it is striking how little culpability the instructions required,” the justices wrote. The Supreme Court was particularly taken aback by the judge’s instruction to the jury that “even if [Andersen] honestly and sincerely believed that its conduct was lawful, you may find [Andersen] guilty.” “The instructions,” concluded the high court, “also diluted the meaning of ‘corruptly’ so that it covered innocent conduct.”10

  What the Court was trying to explain to the government is that citizens, and those who advise them, have no obligation to make the government’s job easy. There are certain statutes and regulations that govern when cooperation must be forthcoming and that define the nature of that cooperation. In the absence of a law clearly stating the citizen’s obligation, the citizen is free to go about his or her business. It did not matter that the Andersen firm’s document-retention (and destruction) policy made the government’s job harder. Making the government’s job harder simply was “not inherently malign,” except when in violation of a specific law.

  The underlying problem in the Andersen prosecution is in fact present in a very large (and growing) number of federal prosecutions, but it arose in this case in a form sufficiently stark that it caught the attention of the high court, which noted caustically that both Congress and the DOJ shared the blame for the destruction of a presumptively innocent firm. Justice Anthony Kennedy told the government’s oralist, Deputy Solicitor General Michael Dreeben, that the DOJ’s definition of crime in this case amounted to a “sweeping position that will cause problems for every corporation or small business in the country.”11

  The executed, if exonerated, corpse of the Arthur Andersen firm—left unceremoniously rotting in the public square—caught the attention of another of the Big Five (now the Big Four) national accounting firms, KPMG. The firm came under investigation for its involvement, beginning around 1997, in the creation and promotion of particularly aggressive tax shelters, an effort that produced some $124 million in fees for the firm by 2001. The Internal Revenue Service (IRS), however, was not impressed and, when it got wind of the nature of these esoteric tax-avoidance mechanisms, declared them abusive and refused to allow tax deductions to filers who had used the schemes.

  Typically, the IRS’s playbook would have had the agency deny deductions taken by tax-filers who used the shelters, whereupon taxpayerinitiated civil litigation in the federal courts would commence over the disallowance, with substantial interest and penalties assessed if and when the taxpayer lost. But the government used a different playbook this time, one developed in its tussles with other sectors of civil society since the mid-1980s, the technique that put Andersen out of business.

  The U.S. attorney in Manhattan opened a criminal investigation focusing on KPMG, several of its partners and employees who were involved in the tax-shelter program, and even one of the outside attorneys at the law firm of Sidley & Austin who had approved the shelters. By February 2004, the government told some 32 KPMG partners and employees, past and present, that they were subjects of a grand jury investigation. (A subject is a person or entity deemed by prosecutors within the orbit of the matter being investigated, but against whom there is not yet sufficient evidence to indict. It is somewhere between a mere witness, and a “target” against whom the prosecutors feel they already have sufficient evidence to indict.)

  At first, KPMG fought to defend its handiwork and its profits. It resisted the investigation and sent three of its partners to testify and defend the firm and the shelters before
a U.S. Senate Permanent Subcommittee on Investigations hearing in November 2003. Meanwhile, competitor national firms PricewaterhouseCoopers and Ernst & Young, which had likewise been caught up in the shelter investigation, quickly settled. KPMG persevered in defending not only the legality of the shelters, but also its several roles as creators and sellers of the shelters, auditors and tax preparers for the firm’s clients, and authors of “opinion letters” on the validity of the devices (i.e., formal written advice regarding the legality of the anticipated tax benefits from a sheltered investment).12

  KPMG’s initial decision to fight the investigation was hardly irrational. The tax shelters it sold and helped create were of a genre characterized by complex schemes designed to help high-income taxpayers reduce their taxes by obtaining outsized deductible write-offs against income. The history of the Internal Revenue Code can be written, from one perspective, as a constant cat-and-mouse game between the IRS on the one hand, and wealthy taxpayers and their lawyers and accountants on the other. Congress could eliminate the textual basis for these devices by amending the Code in order to simplify it and remove the loopholes. It has instead made a veritable hobby out of writing loopholes into the tax code, often in response to lobbying efforts, campaign contributions, or other pressures—unseemly, perhaps, but common in our politics. For whatever reasons, the Code and the even more massive volume of regulations seeking to explicate the Code’s provisions remain a tangle of sometimes impenetrable text that virtually invites tax avoidance. (Lawful “avoidance” should not be confused with illegal “evasion,” which consists of steps to avoid a known legal duty by, for example, simply not reporting earned income such as wages, tips, or commissions.) No surprise, then, that lawyers and accountants have built a profitable industry around helping wealthy individuals reduce their tax burdens through these lawful (if constructed within the technical rules) means.

  Some tax shelters are fairly common and widely used. Certain real estate investments, for example, where property is purchased largely with mortgages, on which interest payments are currently deductible and where improvement costs are amortized over a period of years, can yield substantial deductions against current income, with a large payoff at the end when the appreciated property is sold, with the profit taxed at a lower capital gains rate. These devices are seen by many as unfairly favoring the wealthy, but this is a political and not a legal criticism. Indeed, the very respectable American Bar Association publishes a manual on “asset protection strategies.” Tax shelters are one such strategy to help the rich keep their money.

  The shelters at the center of the KPMG investigation proved to be considerably more esoteric and controversial. For one thing, the IRS considers a tax shelter legitimate only if it is primarily a real investment, with the tax benefits secondary. A shelter, to be legal in the eyes of the IRS, must serve a “legitimate economic or business purpose,” rather than serving merely as a device to avoid taxes.

  The “business purpose” doctrine is a creature of the IRS’s interpretation of the congressional statute that is the Internal Revenue Code. The Code itself is exceedingly difficult to understand, and this complexity has created a huge industry that produces a rather high standard of living for armies of accountants and lawyers. The IRS, according to two highly reputed white collar crime experts at Stanford, uses this “business purpose” doctrine, which has “no explicit statutory support,” to “disallow tax results which are otherwise consistent with the technical provisions of the tax law.” In other words, the IRS follows the spirit of the law, as it divines it, rather than the letter. The whole notion that the IRS somehow knows how much “real gain” (in contrast to tax savings) has to be produced before a shelter is deemed legitimate is, according to these experts, “the moral equivalent of that much-loved definition of pornography” articulated by the late Supreme Court Justice Potter Stewart: “I know it when I see it.”13 When it comes to a legitimate versus a bogus tax shelter, the Stanford experts tell us, “the IRS knows it when it sees it.”14

  Senator Carl Levin of Michigan, however, who was present at the subcommittee hearing attended by the three KPMG partners, took a different view of the matter. He told the Senate in the summer of 2005 that there was a fairly clear line between “abusive tax shelters” and the “legitimate” variety. “Interest paid on a home mortgage or Congressionally approved tax deductions for building affordable housing,” said the Senator, were examples of the legitimate variety. The “abusive” shelters, said Senator Levin, “are complicated transactions promoted to provide large tax benefits unintended by the tax code.”15

  The litmus test offered by Senator Levin clarified nothing. Citizens obviously have a legal obligation to follow the letter of the law. But if the line between legal “tax avoidance” and criminal “tax evasion” is drawn according to someone’s view of what Congress in its heart of hearts “intended,” then the law indeed becomes a trap for the unwary (or even for the wary). As The Wall Street Journal editorialized in criticizing the government’s pursuit of the KPMG accountants via a criminal proceeding, “the finger-waggers in Congress might acknowledge their role in creating the 6,000-page, 2.8 million-word, tax code Frankenstein that facilitates the tax-avoidance industry.”16

  Congress could, of course, enact into law the informal IRS test for the validity of a tax shelter—that it serve a legitimate economic or business purpose, not just that of creating tax deductions. But even this standard would prove elusive, leaving the creators of shelters very little guidance. The problem, as the Journal put it, is that the standard for tax shelters is that “it’s OK to avoid taxes in any of the myriad ways Congress approves of,” but “abusive if Congress didn’t intend it—assuming anyone can ever figure out what Congress really intends.”17

  The shelters at the heart of the KPMG investigation and, later, of the indictment, were four creatures that were truly worthy adversaries of the “tax code Frankenstein.” Even their names had an other-worldly quality: FLIP (“Foreign Leveraged Investment Program”), OPIS (“Offshore Portfolio Investment Strategy”), BLIPS (“Bond Linked Issue Premium Structure”), SOS (“Short Option Strategy”), and variants of these. Here is how the eventual indictment described the FLIP and OPIS shelters (in a case, it must be noted, that ultimately was supposed to be decided by a jury of ordinary citizens):In all material respects, FLIP and OPIS were the same. FLIP and OPIS were generally marketed only to people who had capital gains in excess of $10 million for FLIP and $20 million for OPIS. These shelters were designed to generate substantial phony capital losses (i.e., in excess of $10 million for FLIP and in excess of $20 million for OPIS) through the use of an entity created in the Cayman Islands (a tax haven), for purposes of the tax shelter transaction. The client purportedly entered into an “investment” transaction with the Cayman Islands entity by purchasing a purported warrant or entering into a purported swap. The Cayman Islands entity then made a pre-arranged series of purported investments, including the purchase from either Bank A (which at the time was a KPMG audit client) or Bank D of either Bank A or Bank D stock using money purportedly loaned by Bank A or Bank D, followed by redemptions of those stock purchases by the pertinent bank. The purported investments were devised to eliminate economic risk to the client beyond the all-in cost and minimize the amount of the all-in cost used for the investment component. The purported investments were also devised to last for only approximately 16 to approximately 60 days.

  Because of the complexity of the regulations and the devices constructed to take advantage of the letter of the law, a system has evolved whereby taxpayers are able to test the correctness of their deductions only when the IRS balks and disallows them. The taxpayer may challenge the IRS’s disallowance in the federal courts. If the taxpayer loses, it costs him or her not only the deduction but interest and, often, steep penalties. In an extreme case, the IRS can even tag the taxpayer with civil fraud penalties. To be on the losing end of tax shelter litigation can be extraordinarily expensive for the
taxpayer, but at least it does not land him or her in prison for 20 years merely for guessing wrong or taking faulty advice.

  As the DOJ increased the pressure on KPMG to switch sides rather than continue the fight to protect itself and its targeted partners (and its client), the firm’s resistance weakened. It hardly matters, after all, if in the end KPMG could prove that its shelters were legitimate, or even sufficiently close to the line that their purveyors could not be said, beyond a reasonable doubt, to have violated a known legal duty. Saving the firm from Arthur Andersen’s fate became paramount. Once the firm caved in, the plight of the individual DOJ targets would become enormously more dreary.

  Why would KPMG’s capitulation so severely weaken the individual partners who, in theory, would be left free to fight for their own vindication? The answer lies partly in corporate sentencing guidelines enacted by Congress in 1991 that apply to executives and employees in white collar cases. That law severely limited judicial discretion in imposing sentences,18 while giving prosecutors enormous power to enhance sentences (by demonstrating certain aggravating factors) or reduce them (via mitigating factors). The most important mitigating provision, that effectively transferred sentencing authority from judges to the DOJ, allowed the prosecutor to inform the judge whether the defendant, after a guilty plea or conviction but before sentencing, had “cooperated” in aiding the government to prosecute and convict other miscreants. Since “cooperation” became one of the very few avenues to a lower sentence, corporate defendants faced enormous pressure to please prosecutors. The guidelines formalized and bolstered what previously had been an informal arrangement by which cooperation resulted in lower sentences.

  Recognizing the enormous advantage that the sentencing guidelines had handed federal prosecutors, the DOJ sought to expand and institutionalize this advantage within the department’s practice. Beginning with former Deputy Attorney General Eric H. Holder, Jr. (later the U.S. attorney general) in 1999, succeeding heads of the DOJ’s Criminal Division issued memoranda that expanded the definition and scope of what could be deemed adequate corporate cooperation to qualify for mercy.19 With regard to corporations, in contrast to somewhat different guidelines for individuals, former Deputy Attorney General Larry D. Thompson, then-head of the DOJ’s Criminal Division, issued in January 2003 a memorandum titled “Principles of Federal Prosecution of Business Organizations.”20 It laid out various criteria for judging whether, and to what extent, a business entity was truly cooperative. Corporations were expected to open their books and records to investigators, including records otherwise protected by legal privilege, such as the theretofore sacrosanct attorney-client privilege. They had to encourage employees to open up to and cooperate with the government, and to fire those who refused. Companies that had a policy of advancing funds for legal fees for corporate employees were also encouraged to refuse such fees to non-cooperating employees.