Tuesday, May 19, 2009

Sentencing Guidelines - The Tax Loss from Timing Issues

The Sentencing Guidelines calculation required even post-Booker is principally driven by the tax loss number -- the amount of tax the defendant intended to evade. This can often be calculated exactly. Where it cannot be reasonably calculated, the Guidelines permit a presumptive rate -- for individuals, 28% of omitted income or claimed improper deductions. The question addressed here is how the tax loss is, or should be, calculated if there is simply a timing difference. To take an extreme case, assume that a defendant improperly claimed a $1,000 deduction in year 1 that he is entitled to take in year 2 but does not claim in year 2. Assume further that the tax saved by claiming the deduction focusing only on year 1 is $250 (actual) or $280 (presumptive). But is that the real loss to the Government because the Government will make that up in year 2 when the taxpayer does not claim the deduction? (Note that there is a real sweet civil mitigation issues that could be addressed here if the defendant were to attempt to deduct the same item in year 2, but let's not get bogged down in noncriminal matters here because I assume that he did not claim the deduction in year 2.) In the bare facts given, what is the real tax loss? It is not $250 (actual) or $280 (presumptive), but rather (assuming constant or materially the same marginal rates), it is zero except for, perhaps, the time value of money for that short one year timing period. (Normally, except in collection evasion cases, the time value of money is not considered in calculating the tax loss.)

In United States v. Stadtmauer, 2009 U.S. Dist. LEXIS 9945 (D. N.J. 2009), the timing issue was in the context of depreciation where a current deduction was claimed for items that could be depreciated over future years. The Government wanted to apply the 28% presumptive rate to the entire deduction claimed without any mitigation for the future tax revenue the Government did or would collect. The defendant cried foul and the Court listened. The Court opined:

Mr. Stadtmauer argues that since the issue is only one of timing there is no tax loss associated with these deductions. The Government disagrees, arguing that there is at a least a loss due to the time value of money. However, the Government does not argue that the loss is the time value of money, rather the Government argues that because there is some loss due to the time value of money the 28% presumptive rate should be used. This Court agrees with the Government that the time value of money should be considered as a tax loss, but disagrees that using the 28% rate is appropriate; the 28% rate does not "fit the circumstances" for these deductions. See USSG 2T1.1, Application Note 1.

The Court recognizes, and the Government conceded at the sentencing hearing, that as a general matter, tax loss under the Guidelines for the crimes at issue here does not include interest and penalties. Id. The Court also recognizes that interest calculations are meant to account for the time value of money, so arguably any interest based calculation [*46] should not be included under the Guidelines. But, the Court also finds that recognizing a time value of money effect for these deductions is completely different than the general case of calculating and adding interest, as addressed in Application Note 1. Under § 2T1.1(c), "the tax loss is the amount of loss that was the object of the offense." Here, the purpose of taking the deductions in full in the year incurred was to receive the time value of money benefit from paying less taxes now rather than spread over time; it was not merely some ancillary benefit to the primary object of avoiding taxes by taking a deduction that was not permissible at all, the time value of money benefit was the object of the offense.

The question, then, is what is a reasonable way to estimate this loss. As noted above, to accept the presumptive 28% rate as the Government argues would be unfair and drastically overstate the tax loss. The Court finds that the most reasonable way to account for this loss is by using the Government's own method for compensating itself for the time value of money related to underpayments of tax. Interest on underpayments is calculated by the IRS pursuant to 26 U.S.C. § 6621(a)(2). [*47] This rate is determined quarterly. Id. at § 6621(b). For the years 1997 to 2001, the IRS rate for non-corporate underpayments varied between 7% and 9%, with the rate declining in the years after 2001. See Rev. Rul. 2008-54. This Court finds that using a rate of 8%, a rate in the middle of the range, is reasonable. This approach is not perfect. It does not account for compounding, but it also does not account for the exact timing of the deductions. However, exact precision is not required. This Court finds that the other methods suggested either understate or overstate the intended loss and that this method is the most fair and reasonable estimate of the loss intended for these items. See Gricco, 277 F.3d at 356

Nice indeed.

Hat tip to Caroline Rule of Kostelanetz & Fink for calling this to my attention. Caroline and her partner, Bob Fink, were involved in the case.

Addition on 5/26/2009: See Alan Ellis' Article on Intended Loss here.


  1. Jack,

    I was at the ABA Conference in Washington DC and enjoyed hearing about this case. However, I do not agree with the disposition of the case because taxpayers in similarly situated cases are not treated equally. Consider Taxpayer #1 and Taxpayer #2 below.

    Taxpayer #1: Files fraudulent return. Later feels remorse and files amended return paying tax, interest and penalties. Later is investigated, indicted and convicted for the originally fraudulent return. Government's position is that a later amended return does not cure the original fraud. Loss to government is zero but is sentenced to prison for originally intended tax loss of $100.

    Taxpayer #2: Files fraudulent return in Year 1 taking a full deduction for a depreciable item. Does not take any deduction in Year 2. Tax loss to government is zero. He intended the $100 tax savings in Year 1. Government decides that tax loss is only the AFR Interest of $10. Gets probation.

    Conclusion: In both cases tax loss is zero but Taxpayer #1 gets prison and Taxpayer #2 gets lower sentence because tax loss is treated as an interest calculation at a lower rate than say a 28% rate or top tax rate of 35%.

    Jack, I have read your entire book word for word and it is with the help of your book that I can more clearly think about these things.



  2. Anonymous / John:

    Thanks for your post. Your analysis is an excellent one. The argument I would make for Taxpayer #2 is that, given the fact that he is entitled to a future deduction that he foregoes by claiming it early, the intended loss to the Government really is the time value of money interest factor. By contrast, Taxpayer #1 really did intend in filing the original return to cheat the Government out of the entire amount. His contrition later, evidenced by filing an amended return, does not change his original intent.

    Of course, as you know, the likelihood of the Government prosecuting Taxpayer #1 is not great unless there are some really bad facts other than presented in the example. But, if the Government were to prosecute, the Sentencing Guidelines potential discrepancy exists (although I don't think it is a discrepancy for the reasons noted).

    Finally, keep in mind that, even if Taxpayer #1 were prosecuted and the entire amount included in his or her tax loss number for the Guidelines calculation, the district court could view the voluntary act of contrition by filing an amended return as a factor to consider in making a variance. If I were Taxpayer #1's lawyer, I would urge that and, again depending upon the facts, I think a court might be receptive and even a good prosecutor might signal to the court that a variance might be appropriate (I know that they are not supposed to do that.)

    Thanks again. I appreciate your comments about he blog and the book very much.


Please make sure that your comment is relevant to the blog entry. For those regular commenters on the blog who otherwise do not want to identify by name, readers would find it helpful if you would choose a unique anonymous indentifier other than just Anonymous. This will help readers identify other comments from a trusted source, so to speak.