The following is a summary of the article.

In January 2004, Andrew and Lea Fastow reached plea agreements with the Enron Task Force. In the plea agreements, they plead guilty to certain counts in their earlier indictments related to the Enron debacle. Andrew’s indictment charged 109 counts related principally to certain of the off the books financing arrangements. Andrew was charged with multiple counts of conspiracy to commit wire fraud, conspiracy to falsify records, wire fraud, money laundering conspiracy; money laundering and tax perjury with respect to four years’ income tax returns. Lea was charged with conspiracy to commit wire fraud, a related Klein conspiracy, money laundering conspiracy and tax perjury. The tax perjury counts in both indictments related to the same omitted income for four years derived from Enron off the books shenanigans. The plea agreements, if accepted, will resolve the criminal charges against Andrew and Lea.

If the sentencing judges follow the scripts served up to them by the parties, the basic deal is that Andrew will serve 10 years incarceration (with a possible 15% reduction for good behavior) and Lea will serve 5 months incarceration and 5 months home confinement. In addition, each will incur a significant (but perhaps not material in a financial accounting sense) fine, and Andrew will incur substantial (material in any sense) forfeitures, to which Lea consents with respect to any interest she may have in the assets forfeited.

This article addresses the structuring of the plea agreement to achieve the parties’ goals stated above. Before discussing the Fastows’ specific plea agreements, the article provides an introduction to the criminal laws charged in the Fastows’ respective indictments, the Sentencing Guidelines (including the effect of the loss numbers, relevant conduct, and other sentencing factors), and how plea bargains work under DOJ’s policies. Much of the discussion is directed to the Sentencing Guidelines which have been compared in complexity to the Internal Revenue Code. The Sentencing Guidelines’ complexity, like the Code’s complexity, invites aficionados to game the system by carefully picking out the goodies and avoiding the landmines, thus achieving results that were not contemplated by the drafters.

The article then discusses the specifics of the Fastows’ plea agreements, including their interrelationship. The article concludes that the parties skillfully manipulated the Sentencing Guidelines to confine the sentencing courts’ ability to avoid the basic deal. In Andrew’s case, the parties have assured the result they desire. In Lea’s case, it is unclear that they have locked the judge in to their deal, but they have done the best they could. By considering only some of the sentencing factors (the factors served up to them by the parties), Lea’s sentencing judge might well get to the agreed-upon sentence. However, by considering other sentencing factors including the Sentencing Guidelines’ policy of real offense punishment (relevant conduct), Lea’s sentencing judge might well determine qualitatively that the punishment does not fit the real offenses.

Tax lawyers should have a basic understanding of the Sentencing Guidelines even if they do not represent clients in criminal tax investigations or prosecutions. While the article goes deeper than the basics for the pleas considered, the article discusses the key concepts that play out in most criminal tax prosecutions. And, in the author’s view, it is always a joy to read about skillful lawyers at work, and the Fastows’ counsel and the Enron Task Force lawyers are skillful lawyers indeed.

3/16/04 Update:  According a Houston Chronicle article on 3/16/04, the Probation Office in its PSR recommends a sentencing range of 10-16 months, which is what the parties wanted because the lower end of the range can achieve their 10 month goal.  The PSR, however, also "includes ways that the judge might make the sentence stiffer."  The Enron Task Force and Lea filed objections to the PSR.  Those objections are in the public record but the underlying PSR is not.  According to the article, the parties urge that there is no tax loss, a critical factor in setting the guideline range, because, although the taxable income was omitted from the Fastows' returns, it was included and taxed on another taxpayer's returns.  I guess the tax theory is that, having structured what would otherwise be taxable income as gifts, the putative donor of the gift included it in his income without an offsetting deduction and thus there was no tax loss to the Government considering both the Fastows' and the donor's returns.  In other words, the theory is that the quantum of income gets taxed once.  That is an interesting theory, but may not wash.  Section 162(c)(2) provides that illegal bribes, kickbacks or other payments may not be deducted by the payor (here the putative donor).  Hence, the putative donor's payments in the guise of gifts could not be deducted, thus requiring a tax to the putative donor and would still be income to the Fastows.  In effect, the quantum of income is properly taxed twice -- once to the payor and once to the payee (the Fastows) -- so the payment of only one tax by the putative donor and none by the Fastows does result in a real tax loss.  Moreover, even if the law permitted the putative donor not to pay a tax on that quantum of income and he did so "voluntarily," it seems a bit of a stretch to give the Fastows' credit in their TLN calculation for incorrect treatment on another taxpayer's return.  While there is certainly an argument to be made that real economic loss to the Government ought to be the driver and, if the Government is made whole by taxes paid improperly by another taxpayer, then there has been no harm.  But, we think that, in other cases, DOJ Tax, the Probation Office and the courts would likely resist a free ranging inquiry to determine whether some third party taxpayer touching the transaction overpaid his tax liability in order to reduce the TLN for the defendants' returns.   In the final analysis, according to the article, the parties' briefs appeal to the importance of Lea's contribution to resolution of Andrew's criminal case, including his cooperation and substantial restitution.  They thus in effect urge that, whatever the correct sentencing guideline calculation, the sentencing judge should impose the agreed-upon sentence based on factors that are relevant to departures if the guideline ranges do not otherwise permit.

4/30/04 Update:  We have not previously commented on Judge Hittner's refusal to accept Lea Fastow's plea agreement with the Government which caused her to withdraw her plea agreement from concern/fear that Judge Hittner would substantially increase her incarceration period.  The withdrawal of that plea sent Lea and the Enron Task Force back to the negotiating table to avoid a trial that could potentially hold great risk for Lea.  The Houston Chronicle today reports a new deal between Lea and the Task Force designed to further constrain Judge Hittner's discretion to an acceptable range of sentences.  We have had to read between the lines because the article was not written for white collar criminal attorneys, but we think we have this right.  Basically, the Government makes a superseding indictment withdrawing the earlier indictment which contained several felony counts as discussed in the earlier article and replaces the indictment with an indictment charging a single count under Section 7207 of the Internal Revenue Code which is a misdemeanor count having a maximum incarceration period of one year.  The Judge thus may sentence only to a maximum of one year.  (Those who read the article will recall that Andrew Fastow and the Enron Task Force used this type of technique to limit the Judge's discretion by entering a plea to two five year felony counts, thus making the maximum incarceration period ten years.)  Lea and the Enron Task Force thus did not insure the 5 and 5 deal they originally sought, but they did insure that Lea gets no more than one year incarceration.

We should note that resolution of felony tax charges by falling back to a misdemeanor charge is not usually done.  The following is the pertinent portion of the DOJ Tax's Criminal Tax Manual, with comments indicated in bold:

16.02 POLICY LIMITING THE USE OF SECTION 7207

The Tax Division's policy regarding the use of section 7207 in criminal prosecutions has undergone dramatic changes. For a number of years, prosecutions were not authorized under this section. (3) On September 17, 1976, the Attorney General approved a Tax Division proposal to modify this policy. The modification in policy approved by the Attorney General allowed for the use of section 7207 in a limited number of criminal tax cases, commonly referred to as altered document cases. Such cases routinely involve the submission to the Internal Revenue Service of checks containing amounts which had been altered (usually, increased) after the checks had cleared the bank ("raised cancelled checks"), altered invoices, and similar altered documents as support for overstated deductions.

Section 7207 prosecutions, however, were not authorized where the altered document was a federal tax return. Otherwise stated, prosecution under this section was limited to cases involving the falsification of documents other than federal tax returns.  [T&J Comment: the prior felony charges involved false federal income tax returns.]

Furthermore, the use of section 7207 was restricted to those cases where the Tax Division determined that the circumstances did not warrant a felony prosecution. Under this policy, the Tax Division would consider the following factors in determining whether prosecution under the misdemeanor section was justified:

(1) the computed tax deficiencies were such as to be considered de minimis in relation to the circumstances of the particular case under consideration and (2) the means and methods utilized in committing the offense were commensurate with charging a misdemeanor rather than a felony.  [T&J Comment: as to (1), the parties represented to the Government in the prior plea agreement that no tax loss was involved; for a discussion of that issue, see the note above; as to (2), the publicly disclosed facts suggest no mitigating factor.];

Recognizing that its policy created obstacles for the government in negotiating with lower-echelon individuals in a wider scheme who expressed a desire to cooperate in ongoing or future investigations in return for leniency, the Tax Division reconsidered its long-standing position that an income tax return could not form the basis of a section 7207 prosecution. On March 21, 1989, the Tax Division issued Directive No. 75 which modified its original position, permitting misdemeanor prosecution under section 7207 where the false document was a federal income tax return under very limited circumstances. The guidelines set forth the following provisions:

The Department of Justice, Tax Division, agrees to consider approving plea agreements with charges brought under 26 U.S.C., Section 7207 for witnesses cooperating in Title 18 and Title 26 grand jury investigations and in no other circumstances under the following conditions:

1. Approval for Section 7207 charges will not be given in any case in which the Tax Division has previously authorized charges against the subject under Section 7206(1), Section 7201, or a tax (Klein) conspiracy.   [T&J Comment:  DOJ Tax had almost certainly previously approved the charges under Section 7206(1); hence this condition of the policy is surely not present; however, this condition almost certainly did not have the unique circumstances of Lea's case in mind where the initial charges were not part of an independent tax crimes investigation and were brought by the financial crimes task force as a squeeze play directed to someone else (Lea's husband, Andrew); read on.]

2. The Tax Division must be provided with a prosecution statement or letter describing the outlines of the Title 26 and/or Title 18 investigation, the involvement of the cooperating witness who will plead, and the anticipated cooperation that the witness is expected to provide in the investigation.   [T&J Comment:  We assume that this had already been done in obtaining DOJ Tax approval for the preceding Section 7206(1) indictments.]

3. The subject must have agreed to be a cooperating witness in a Title 18 or Title 26 investigation to which the witness' proposed income tax violation related.  [T&J Comment:  The plea contemplates that Lea's husband and not Lea will be the cooperating witness; again these particular circumstances were not contemplated and thus a deviation from the condition seems warranted.]

4. In addition to his cooperation in the ongoing criminal investigation and prosecution, the subject must agree to cooperate fully and truthfully with the Internal Revenue Service in any civil audit or adjustment of the tax liability arising out of the circumstances of the criminal case.   [T&J Comment:  This is almost surely an implied if not explicit; however, on a relative basis, the amount of tax dollars involved are de minimis even if they are not zero as earlier alleged by parties, so there will be no reason for her not to cooperate.]

5. The subject must be informed that any plea agreement to tax misdemeanors under 26 U.S.C. § 7207 is subject to the approval of the Tax Division, Department of Justice. No such plea agreement is to be executed until authorized by the Tax Division or, if executed, unless it contains a provision that the plea agreement is subject to the approval of the Tax Division.  [T&J Comment: DOJ Tax approval was surely cleared in advance.]

6. Approval for use of Section 7207 will not be given, hence should not be requested, if the underpayment of taxes resulting from the false statements in the return exceeds $2500 in any of the years. In such cases the plea must be to a tax felony.  [T&J Comment:  Here's the rub.  Although the parties earlier represented to the Court that there was no tax loss, in fact there almost surely was; as noted above, that representation apparently was based upon the notion that, on the quantum of income omitted from the Fastows' tax returns, the payor of the income had paid tax, so that considering two unrelated taxpayers together, there was no tax loss; that notion is suspect for the reasons noted above; hence, we question whether this condition was met.]

7. The IRS must make a referral pursuant to 26 U.S.C. § 6103(h)(3)(A). The United States Attorney must have obtained tax disclosure confirming the filing of the return(s). The Tax Division should be provided with an abbreviated SAR, a computation of the taxes due, the tax return(s) involved, and a copy of the plea agreement or a statement of its terms. Section 7207 approval will not be given if the tax disclosure material suggests that a tax misdemeanor would be an inappropriate disposition of the case.   [T&J Comment:  This condition could have been easily met and probably had already been met when the original approval under Section 7206(1) was obtained; We assume therefore that this condition was met.]

8. The subject must sign a statement reflecting the amount of the unreported income or fraudulent deductions and the circumstances involved in all the years under investigation.  [T&J Comment:  We assume that because the taxes are at most relatively de minimis, this condition could be easily met and was.]

Tax Division Directive No. 75, dated March 21, 1989.

Bottom line, the Lea Fastow's unique circumstances required deviation from the policy.

For us ordinary lawyers who practice in the area without such compelling and unique circumstances, the above policy sets the constraints for trying to move a felony tax charge to a misdemeanor.

Preliminary Observations on Martha Stewart's Sentencing

On March 5, 2004, Martha Stewart and Peter Bacanovic were convicted on multiple counts related to their cover up of activities related to Stewart's sale of ImClone stock.  They now face sentencing after the Presentence Investigation Report is prepared.  I shall comment here on matters related to the Stewart's eventual sentencing.  I deal only with Stewart's conviction and do not address Bacanovic's conviction (although the same comments may apply to his conviction).

Many of the considerations that will arise in Stewart's sentencing are addressed in my  downloadable article on the Fastow plea agreements.  (See above.)  Therefore, I shall here only cover some major points and summarize the ones that are adequately developed in the Fastow article.

First, the counts of conviction determine the maximum permissible sentence.   For purposes of determining the maximum, counts may be stacked.  Stewart was convicted of four counts (Count One Conspiracy to Obstruct Justice, Make False Statements, and Commit Perjury, Count Three False Statements, Count Four False Statements, and Count Eight Obstruction of Justice).  Each of these counts has a five year maximum sentence, permitting a 20 year maximum aggregate sentence for Stewart's four counts of conviction.

Second, actual sentencing is determined under the Sentencing Guidelines ("Guidelines").  Except in the rarest of cases, the Guidelines produce a sentencing range substantially below the maximum that can be imposed.  If you read the article on Andrew Fastow's plea, you will see that Andrew's plea provides an instance where the Guidelines would produce a range in excess of the maximum permitted by the counts of conviction (the counts to which he pled) and thus the maximum permitted by the counts of conviction actually determines the sentence.  That is not usually the case.  I have not actually attempted a projected Guideline calculation for Stewart, but most reported comments have indicated that she might actually be sentenced in the range of 1 year or slight more.  That sounds about right considering only the counts of conviction.

Third, Martha Stewart was not convicted on a significant count.  Prior to submitting the case to the jury, Judge Cederbaum dismissed Count Nine, the securities fraud count.  Judge Cederbaum's opinion may be reviewed by clicking here.  The facts giving rise to the securities count were Stewart's public denials of guilt and misstatements of fact in an alleged effort to affect the market price of MSLO stock as the stock began eroding due to publicity as to the investigation.  Stewart argued that it was a novel application of the law that mere denials of guilt could be viewed as manipulation of the market.  Focusing on the actual evidence in the case, Judge Cederbaum concluded that "that a reasonable juror could not, without resorting to speculation and surmise, find beyond a reasonable doubt that Stewart’s purpose was to influence the market in MSLO securities."  Does that mean that the securities count is out forever?  No.  At sentencing, the court may consider nonconviction criminal conduct -- called "relevant conduct" -- in determining the sentence so long as not in excess of the maximum allowed by the counts of conviction.  Criminal conduct underlying dismissed counts (as well as even counts of acquittal as well as uncharged conduct) may be considered as relevant conduct if shown by a preponderance of the evidence (rather than beyond a reasonable doubt).  I develop this consideration in the Fastow article.  Suffice it to say here that Judge Cederbaum's conclusion that no reasonable juror could find guilt beyond a reasonable doubt does not foreclose Judge Cederbaum from finding by a preponderance of the evidence that this crime was committed.  This would then permit Judge Cederbaum to consider the securities count either to affect the range or where, inside the range otherwise indicated, she will sentence.  In this regard, in sentencing, a court is not bound by rules of evidence so that the Government may able to introduce evidence it would not have been able to introduce at trial.  So this issue may be revisited in order to potentially impose a stiffer sentence than would otherwise be indicated.  And, if it is considered, the loss presumably would be measured by the market loss which would be very large indeed.

To Testify or Not to Testify -- That is the Question

Martha Stewart's criminal trial attorney has been subjected to Monday morning quarterbacking on the decision that Martha Stewart would not testify in her criminal case.   The Fifth Amendment guarantees the citizen's right not to have to testify against himself/herself.  The decision as to whether or not to testify was, in the final analysis, Martha's decision.  All her attorney could do was to advise her as to the pros and cons of testifying and make a recommendation based upon his insight as to the strength of the Government's case and the jury's disposition at the moment in the trial where she would testify if she chose to do so.  Her attorney is a good lawyer, so presumably he gave her good legal advice.  Since we don't know what recommendation he made to her based upon nonlegal factors (the jury's disposition at the time), we won't be able to legitimately second-guess his recommendation.  We discuss here some of the factors that go into this type of decision.

First, in a high profile financial crimes case involving a defendant of celebrity or at least notoriety status (the names Skilling and Fastow come to mind), the jury will want to hear from the defendant.  The jury will be instructed that the Fifth Amendment allows the defendant to forego testifying, and that no adverse inference should be drawn from the exercise of that right.  Still the jury wants to hear from him/her and it is naive to think that some or all of the jurors will not treat the exercise of the Fifth Amendment as a negative factor.  We got some feedback on this issue from the jurors in Martha's trial and perhaps will get some more in Kozlowski's trial which is drawing to a conclusion.

Second, and this perhaps is different from but relates to the exercise of the Fifth Amendment right not to testify, a minimalist affirmative case in the defendant's case in chief (after the Government has rested its case) is dicey indeed.  It is certainly true that, in appropriate cases, the defendant can present his or her case through cross examination and through arguments pointing out  weaknesses in the Government's case.  But, when the Government goes to pains to make a case that survives a motion for acquittal at the end of the Government's case for insufficiency of the evidence, the question a defense lawyer must answer is whether some affirmative presentation of the defendant's case is needed.  An affirmative presentation can in some cases be made without the defendant testifying, but even this is dicey where as is often the case that affirmative presentation leaves nagging doubts that the defendant could fill, one way or the other, by testifying.  This is dicey stuff, as Martha found out and as Jamie Olis found out in a predecessor white collar crime trial in Houston related to the Dynegy case.

Third, coming back to the issue of whether Martha's decision not to testify was a good one, what was at stake?  Some persons have said that, had she actually testified, the jury may not have liked her commanding/demanding personality and held that against her in determining guilt.  But they heard enough evidence of that personality so that her own "testimony" -- demeanor and otherwise -- on that issue would have been cumulative and, indeed, she might have been able through skillful woodshedding to reprieve her reputation on personality at least before the jury.  What else was at stake?

She might (and this is speculation because we will never really know) have been able to prove her innocence or at least cast sufficient doubt on the Government's case to raise a reasonable doubt in the jury's mind. 

But, had she testified and the jury still had concluded she was guilty, would she have been worse off for testifying?  The answer is that she could have been significantly worse off.  If the judge were convinced by a preponderance of the evidence that, in testifying, Martha had attempted to impede or obstruct the trial by giving false or materially misleading testimony, the judge could impose a 2 level adjustment to her offense level which could significantly increase her ultimate sentence.  A defendant's choice is to exercise the Fifth or take the stand and tell the truth; the defendant does not have the choice to lie or otherwise impede the trial.  Of course, the Government might be able to prosecute separately the blatant lie from the stand under the perjury statute, but perjury is often difficult to prove beyond a reasonable doubt and the Government's allocation of prosecutorial resources may not allow sufficient interest to mount a second prosecution.  Apart from a separate prosecution for perjury, however, in the original case the sentencing enhancement can apply if the sentencing court finds by a mere preponderance of the evidence that the taxpayer attempted to obstruct justice through his or her testimony.

So, not only was she at risk of losing or not helping her case if she testified, she was at risk of having an sentencing enhancement.  In this environment, it is hard to Monday morning quarterback her decision.

But it does raise an important threshold question for defendants being considered for indictment, particularly defendants of notoriety from whom the jury will almost inevitably want to hear at the criminal trial.  By plea bargaining on the front end, the defendant can get a substantial downward adjustment for acceptance of responsibility and will avoid the difficult and perhaps impossible choice of deciding whether to testify with the risks inherent in the defendant's testimony.  Simply put, had Martha accepted a plea bargain earlier, she could have substantially limited her exposure and avoided have to make the impossible choice of whether or not to testify.  This does not mean that every case should plea, but it does mean that every case must have realistic analysis of the risks all the way through the process so that the defendant can make a meaningful decision as to whether to accept a plea bargain.