Ninth Circuit Requires Relevant Conduct Financial Loss Determinations in NonTax Case to be Made by Clear and Convincing Evidence (3/30/15)

Last night, in our Tax Fraud and Money Laundering Class at University of Houston Law School, we covered the U.S. Sentencing Guidelines.  The U.S. Sentencing Guidelines Manual may be viewed here.  We covered the following points:

• The financial loss for financial crimes is the principal driver for the Guidelines.  In the tax Guidelines, the financial loss is called the tax loss.  See S.G. 2T1.1, here (for the major tax crimes); see S.G. 2T4.1 (Tax Table), here.  
• The tax loss from the count(s) of conviction and the tax loss from related unconvicted conduct -- called "relevant conduct" -- can be included the the tax loss for purpose of the determinations.  For the  concept of relevant conduct, see SG 1B1.3, here.  
• These sentencing determinations are generally made by a preponderance of the evidence.  We did note that the Ninth Circuit may require some sentencing determinations to be made by clear and convincing evidence.

This morning, while going through my list of recent cases that I had not yet read, I came across United States v. Hymas, ___ F.3d ___, 2015 U.S. App. LEXIS 4822, 1-21 (9th Cir. Idaho 2015), here.  In Hymas, the Ninth Circuit held that, in the facts of the case, relevant conduct financial loss determinations for nontax financial crimes must be made under by clear and convincing evidence.  Although sentencing determinations are normally made by a preponderance of the evidence, the Ninth Circuit requires those determinations to be made by clear and convincing evidence "when a sentencing factor has an extremely disproportionate effect on the sentence relative to the offense of conviction." United States v. Mezas de Jesus, 217 F.3d 638, 642 (9th Cir. 2000) (citing United States v. Restrepo, 946 F.2d 654, 659 (9th Cir. 1991) (en banc)).  The NInth Circuit has established no "bright-line rule for the disproportionate impact test," but requires it to be made in the context of the totality of the facts under guidance first given in United States v. Valensia, 222 F.3d 1173 (9th Cir. 2000).  The Ninth Circuit said.

Under the Valensia totality of the circumstances test, six factors, none of which is dispositive, guide the determination of whether a sentencing factor has a disproportionate impact on the sentence:(1) whether the enhanced sentence falls within the maximum sentence for the crime alleged in the indictment; (2) whether the enhanced sentence negates the presumption of innocence or the prosecution's burden of proof for the crime alleged in the indictment; (3) whether the facts offered in support of the enhancement create new offenses requiring separate punishment; (4) whether the increase in sentence is based on the extent of a conspiracy; (5) whether an increase in the number of offense levels is less than or equal to four; and (6) whether the length of the enhanced sentence more than doubles the length of the sentence authorized by the initial sentencing guideline range in a case where the defendant would otherwise have received a relatively short sentence.Treadwell, 593 F.3d at 1000.

In Hymas, the Ninth Circuit reviewed the relevant facts as follows:

A presentence report ("PSR") was prepared for each defendant. For Aaron, the PSR calculated the total loss as $3,689,953.73. The loss attributed to the Count Four loan was $162,758.79. The rest represented losses allegedly suffered by lenders on other loans, including loans that were the subject of counts that were dismissed. Losses from these loans were included because other "relevant conduct," separate from the specific activity that is the subject of the criminal conviction, may be considered in imposing a sentence. See U.S.S.G. § 1B1.3. 

Adding the losses from other loans substantially increased the proposed Guidelines sentencing range calculated in the PSR. The base offense level for Aaron's conviction under the Sentencing Guidelines was 7, but the loss amount as determined in the PSR increased that level by 18, to a total of 25. Following a reduction of 3 levels for acceptance of responsibility, the PSR determined that Aaron's total offense level was 22. With a criminal history category of I, Aaron's Guidelines imprisonment range was 41 to 51 months. 

Aaron filed objections to the PSR loss calculation. He contested the relevant conduct, the proper burden of proof, the number and identification of victims, and the loss amount for sentencing. He also contested the loss amount and the proper victims for restitution. The district court held a three-day evidentiary hearing to resolve the factual issues. 

Following the hearing, the district court issued a written order. Although Aaron argued that the clear and convincing evidence standard applied, the court explicitly held that the burden of proof that applied was preponderance of the evidence. The court applied that standard to determine the total loss amount for the purpose of calculating Aaron's sentence, including losses from other loans as relevant conduct. The court found that Aaron Hymas had committed fraud in the nineteen other loan applications and that the total loss amount was $3,416,337.97, slightly less than the amount calculated in the PSR. The district court agreed with the PSR's calculation of the Guidelines imprisonment range as 41 to 51 months. The district court subsequently sentenced Aaron to 24 months in prison.

The Ninth Circuit divided its consideration into the loss for the count of conviction (referred to as Count 4) and the loss from the relevant conduct.  As to the court of conviction, the Court concluded that, although there was some concern about the enhancement for the size of the loss:

But the size of a loss enhancement, standing alone, does not compel the use of the clear and convincing standard. Treadwell, 593 F.3d at 1001—02. In this instance, we conclude that there were no serious due process concerns that required application of a heightened standard, even considering the Valensia factors, to the extent that the sentence was based on the loan that was the subject of the conviction. Aaron admitted the facts of the fraud that caused the loan to be made. He also knew the size of the loan, and that defined the potential extent of the loss. In that situation it was not necessary to apply a heightened standard to protect against a violation of due process.

Then, turning to the relevant conduct based on loans not included in the count of conviction, the Ninth Circuit concluded that it should be determined based on clear and convincing evidence:

The sentence imposed by the district court was not entirely based on the loan that was the subject of the conviction, however. The district court also used losses from other loans to calculate Aaron's total offense level, increasing the total offense figure by an additional 8 levels. Based on the principles articulated above, the clear and convincing standard of proof should have been applied to determine the amount of the losses from the other loans. 

Aaron did not plead guilty to fraud for the other loans. He was not charged with a conspiracy, nor did he admit in his plea agreement that he had participated in a scheme to defraud involving multiple transactions. Losses from these loans were based on conduct for which he was not convicted. Aaron did not have a guilt-phase trial where the government was required to prove beyond a reasonable doubt that he committed fraud on the other loan applications. 

Similar to the analysis of Count Four, the fifth and sixth Valensia factors support the use of the heightened standard for the other loans, even though the first four factors may not require that result. Inclusion of the losses from the other loans ultimately resulted in an increase of 8 offense levels, from 10 (based on the loss from the Count Four loan by itself) to 18. This additional 8-level increase more than doubled the Guidelines imprisonment range. Under our precedents, we conclude that the district court should have employed a heightened clear and convincing standard of proof with regard to the losses from those other loans. 

In United States v. Munoz, for example, we held that the district court was required to use the clear and convincing evidence standard when calculating losses from uncharged conduct. 233 F.3d 1117, 1127 (9th Cir. 2000), superseded on other grounds by statute as stated in United States v. Van Alstyne, 584 F.3d 803, 817—18 (9th Cir. 2009). Two defendants were indicted on ten counts but convicted of only two counts of fraud in connection with specific sales of bus shelters as part of a Ponzi scheme. Id. at 1123. The district court, applying a preponderance of evidence standard, included within the loss calculation the losses from hundreds of sales made to other investors not included in the two counts of conviction. Id. at 1124. We vacated the sentence and remanded for resentencing, holding that, while the sales to the other investors were relevant conduct, the heightened standard of proof should have been used because the enhancement had a "disproportionate effect on the sentence." Id. at 1127. The same is true here. 

The district court in this case concluded that the appropriate standard of proof was preponderance of evidence based upon a line of cases that applied that lower standard when determining the extent of losses from a conspiracy. The decisions cited by the district court — Treadwell, 593 F.3d at 1001; Berger, 587 F.3d at 1048—49; and United States v. Armstead, 552 F.3d 769, 777 (9th Cir. 2008) — hold that where losses are based on the extent of a criminal conspiracy, those losses need not be proven by clear and convincing evidence because the defendants had the opportunity at trial to challenge evidence of the extent of the fraud conspiracy. 

In this case, however, the government did not charge Aaron with a conspiracy to defraud that included the other acts of fraud alleged in the indictment. He only pled guilty to one count of fraud regarding a specific loan transaction. He had neither need nor opportunity to contest the alleged conspiracy, and he cannot be sentenced using the lower standard as if he had challenged a conspiracy charge. 

To be sure, the allegations against the Hymases bore similarities to a conspiracy, and the multiple counts resembled each other by alleging similar misrepresentations in similar loan applications. But the representations were not identical in all applications. Most of the losses included in the district court's calculation were based on loans that did not contain the specific false representations in Count Four acknowledged in the Plea Agreement. Even as to the loan applications that included the same statements, Aaron argues with justification that his admission that the statements were false in 2007 did not mean he admitted that those same statements were false in earlier years, before the family's financial circumstances deteriorated. 

The government argues that, even if the district court applied the wrong standard, its decision to impose a 24-month sentence was harmless beyond a reasonable doubt because the sentence varied significantly below the guidelines range. We decline to engage in such guesswork. It is true that the district court might have made the same loss calculation applying the clear and convincing standard, but it might not have, either. The court made a point of stating that it was applying the preponderance standard, and the court's emphasis on the standard could imply that a higher standard would have resulted in a different loss calculation. Sometimes a district court says in finding a loss amount that it would reach the same result under either standard, but the court in this instance did not. 

Similarly, the district court might have imposed the same sentence even if it had calculated a lower loss figure under the clear and convincing standard and, as a result, a lower sentencing Guidelines range, but that possible outcome is too uncertain for us to rely upon it. It is also inconsistent with our normal approach to sentencing. "[T]he district court must correctly calculate the recommended Guidelines sentence and use that recommendation as the 'starting point and initial benchmark.'" United States v. Munoz-Camarena, 631 F.3d 1028, 1030 (9th Cir. 2011) (per curiam) (quoting Kimbrough v. United States, 552 U.S. 85, 108 (2007)) (internal quotation marks omitted). We cannot say on this record that the failure to calculate the correct recommended Guidelines sentence was harmless error because the district court's analysis for the extent of the variance was not based on the correct range. Id. at 1030—31. Accordingly, we vacate Aaron Hymas's sentence and remand the matter to the district court. On remand, the court should apply the clear and convincing standard in calculating losses attributable to the other loans. 

JAT Comment:

1. Although this was not a tax case, the same type of analysis could apply in a tax sentencing.
2. The sentencing judge probably calibrated the sentence based on Booker variance discretion, so that the precise Guidelines calculation did not really matter -- even if none of the relevant conduct were included.  Still, it is critically important to the process to make good Guidelines calculations and the Ninth Circuit was not sure that significantly lower offense level calculations that might result from the clear and convincing standard would not have affected the ultimate sentence.