Guidelines Calculations Games -- More on the Chernick Plea (8/3/09)

In my recent blog on the Jeff Chernick Guilty plea here, I said, in effect, that Chernick and the Government were playing games with Chernick’s Sentencing Guidelines Calculations. I would like to explore further that issue further by stepping through the relevant Guidelines calculations. Under Booker, of course, the Guidelines are advisory rather than controlling; they must be considered, however, and, in most sentencings since Booker, the Guidelines have been the principal determinant in setting sentences. (Tomko may be seen as an aberration – an important one – in the pattern of still hewing to the Guidelines’ advisory range). So, with that said, let’s calculate the Guidelines range for Chernick. 

1. I start first with the plea agreement to a single 3 year count of conviction for § 7206(1). Plea to a single 3 year count caps the possible Guidelines sentence at 3 years. It simply makes no difference whether the Guidelines calculations produces a greater sentence than 3 years; it is capped at three years. I discussed this phenomenon in an earlier article involving the plea by the infamous Andy Fastow of Enron fame back when the Guidelines were mandatory and would have generated a sentence well exceeding 20 years; Fastow capped his possible sentence at 10 years by pleading to only 2 five year counts. John A. Townsend, Analysis of the Fastow Plea Agreements, 2004 TNT 44-46 (3/5/2004). 

2. Since the plea was for one § 7206(1) (tax perjury count), the beginning point for the Guidelines calculations is determining the base offense level under SG §2T1.1(a)(2). The base offense level is determined by the tax loss table in § 2T4.1. That table requires that the tax loss be determined. The tax loss is determined by the court by a preponderance of the evidence after receiving recommendations by the probation office and objections by the parties (the Government or the defendant). Normally, in tax cases, the initial calculation of the tax loss is made by the Government in the Special Agent’s Report that is available to the probation office. In addition, in the plea discussions, the parties may negotiate over the proper amount of the tax loss. By that time, the defendant’s counsel will negotiate over the amount that is a proper amount under the Guidelines. If they can reach agreement, they will state their agreement as to the tax loss in the plea agreement. That statement of agreement is not binding on the probation office, and defendant’s counsel must be careful to advise his client that the probation office can determine a larger tax loss which could dramatically affect the sentence. 

In the Chernick plea agreement, although the parties did not agree as to the tax loss, the parties did try to close some parameters on the issue by agreeing: 

 a. Tax Loss: The relevant amount of actual, probable, or intended tax loss under Section 2T1.1 of the Sentencing Guidelines resulting from the offense committed in this case and all relevant conduct is the tax loss associated with accounts at UBS that were disclosed to the Government pursuant to the Deferred Prosecution Agreement with UBS, and of which the defendant was the beneficial owner for tax years 2001 through 2007. 

Note that this agreement constricts possible tax loss in two ways: (1) it limits the tax loss to the tax loss associated with the UBS accounts disclosed to the Government and (2) it limits the tax loss to the years 2001 through 2007. As I noted in my earlier discussion of the Chernick plea, both of these limitations are improper. The tax loss is “the total amount of loss that was the object of the offense.” “‘Offense’ means the offense of conviction and all relevant conduct under §1B1.3.” SG § 1B1.1, Application Notes (H). Relevant conduct includes conduct which would require grouping (suffice it now to say that all tax crimes would require grouping) “that were part of the same course of conduct or common scheme or plan as the offense of conviction.” In this case, under the Statement of Facts entered along with the guilty plea, Chernick admitted a course of conduct from 1981 forward involving (a) depositing income into foreign bank accounts which he failed to report and pay U.S. tax and (b) earning income on or with respect to those deposits which he failed to report and pay tax. That is precisely the conduct he admitted guilt for the year of conviction. The time frame for the relevant conduct was thus 1981 through 2007 and includes all banks. Nevertheless, the Government and Chernick attempt to limit the time frame to 2001-2007 and then only include the tax on the deposits into and earnings from the UBS accounts. 

Now, in my experience, this type of limitation will sometimes occur in a plea agreement where the IRS has not investigated earlier years and there are serious limitations on ability to reconstruct the tax loss for the earlier years. All parties – the Government, the defense, the probation office and the Court – know that there is earlier relevant conduct tax loss that could be included, but it is not because it has not been and cannot reasonably be quantified. One may ask whether that is what happened here which might account for the limitations in the agreement noted above? That could hardly explain omitting the amounts deposited in other banks, at least one of which was stipulated to exist in this time frame. 

Moreover, the Guidelines provide expressly for estimates where more precise calculations cannot or have not been made. How can they be made in the Chernick case Chernick admits in his Statement of Facts that, by 2005, the amount in his UBS account was $8 million. Since, under the pattern admitted, tax amounts would not have been deposited into the accounts, we may presume that these are untaxed amounts. Chernick admits further that, not all of his offshore assets connected with this scheme were in UBS. Around 2003, Chernick moved some portion of his UBS assets to another smaller Swiss Bank with no U.S. presence in order to avoid detection in the U.S. How much was moved is not stated, but that should be easily quantifiable given that this conduct was relatively recent. Chernick also admitted that he typically withdrew some $300,000 + per year that was then deposited into his U.S. corporate account. It is unclear whether he reported this amount on the U.S. corporate return, so it is unclear whether these amounts have been taxed. I thus, cannot assume for present purposes that these withdrawn amounts should enter the tax loss calculation. (Timing differences are effectively included by including the earnings on the previously untaxed amounts.) Finally, he did send $700,000 in the U.S. by sham loan in 2004, so this amount would not be on deposit in UBS or any foreign bank, so there is another $700,000 in untaxed income (as well as the relatively minor amount of taxes due on the interest deductions he claimed on the sham loan. So it appears that, on the amounts stipulated, there was at least $8.7 million of untaxed income over all the years. Even if some precise quantification year by year would be impossible for all of the years involved, it is admitted that all of this cumulative amount was untaxed and the Guidelines permits an estimate as follows (SG § 2T1.1 Notes (A):

If the offense involved filing a tax return in which gross income was underreported, the tax loss shall be treated as equal to 28% of the unreported gross income (34% if the taxpayer is a corporation) plus 100% of any false credits claimed against tax, unless a more accurate determination of the tax loss can be made.

So, based on what we know from the court filings, the tax loss can be calculated as.28 times $8.6 million, for a tax loss of $2.4 million (rounded down). The base offense level under the Tax Table is thus at least 22. I should note that some portion of the $8.7 million might well have been taxable at lesser capital gains rates, but under the foregoing presumption a court would require that the defendant show which portion in order to make that more accurate determination but it looks like the parties are now unwilling to delve into the pre-2001 years (that may change in the sentencing process). Moreover, under the Tax Table, the tax amount would have to drop by over $1,400,000 to lower the base offense level which I don’t think would be likely on the difference between the capital gains and presumed 28% ordinary income rate. So, in sum, we have a base offense level of 22. 

3. An additional 2 levels are added because the parties stipulated that the offense involved sophisticated means. (That stipulation may also not be binding on the probation office or the court, but it is unimaginable that this 2 level increase does not apply.) So, we are now at level 24. 

4. The parties stipulate to the 2 or 3 level acceptance of responsibility reduction under S.G. 3E1.1, so Chernick gets a 3 level reduction. So, we are now at level 21. 

5. Chernick admits that he paid – or at least intended to pay – a bribe to a Swiss official for information about whether his accounts would be turned over pursuant to the U.S. Government’s full court press in 2007. This might be considered obstruction of justice which could justify a 2 level increase, but there are not enough facts known now to add that. Leave the offense level at 21. 6. Moving to the sentencing table (Ch. 5, Part A), the indicated sentencing range for offense level 21 at criminal history I is 37-46 months. Note that this indicated level could go up depending on whether additional amounts (deposits in other banks and cash brought back into the United States) are properly included in the tax loss amount, which would ripple down to a higher offense level for sentencing. 7. The low end of the indicated range thus exceeds the maximum incarceration for the sole 3 year count of conviction. These calculations assume, of course, that the tax loss calculations include all relevant conduct. In the plea agreement, the Government agreed “to recommend that the defendant be sentenced at the low end of the guideline range, as that range is determined by the court.” Either the Government was making the defendant a meaningless promise or, more likely as indicated above, it was willing to ignore tax loss amounts clearly and relatively easily includible and even required by the Guidelines. Maybe the Government is playing that game. However one may view the propriety of the Guidelines, I don’t think that type of game playing speaks well for the Government. At a minimum, the Government should tell the probation office and court when it is playing that game.