Opinion
CASE NO. 1:16CR52
04-17-2018
UNITED STATES OF AMERICA, Plaintiff, v. MOHAMMAD H. MOHAMMAD, OMAR H. MOHAMMAD, Defendants.
OPINION AND ORDER CHRISTOPHER A. BOYKO , J. :
On October 20, 2017, Plaintiff United States of America ("the Government") filed its original Sentencing Memorandum as to Defendants Mohammad H. Mohammad and Omar H. Mohammad. (ECF DKT #71). The Court conducted an Oral Hearing as to the reasonableness of the loss calculation in this matter on January 25, 2018 and January 29, 2018. On February 23, 2018, the Government filed its Supplemental Sentencing Memorandum. (ECF DKT #91). Also on February 23, 2018, Defendants filed their Supplemental Brief Regarding Loss Calculation. (ECF DKT #92). For the following reasons, the Court rejects the Government's methods for calculating loss, adopts Defendants' loss calculation in part and includes in the amount attributable to Defendant Mohammad Mohammad the tax loss amount of $498,189.00.
BACKGROUND
In or around 2009, the Ohio Department of Public Safety collected statements from individuals relating to "exchanging food stamp benefits for cash and other non-food items at Holy Land [Holy Land Imported Foods, Inc., Defendants' store]." (ECF DKT #91 at 2). The United States Department of Agriculture Office of the Inspector General ("USDA OIG") determined a pattern where the Wahdan brothers and their employees would buy food stamp cards from customers at their convenience store at 50 cents on the dollar and use those cards at Holy Land to purchase stock to use at the Wahdan's restaurant. (ECF DKT #91 at 2-3). The USDA OIG identified 155 cards, totaling $51,000 of food stamp benefits used at Holy Land within 24 hours after being used at a Wahdan convenience store. (ECF DKT #91 at 3). Also, the USDA OIG sent cooperating witnesses with audio and video recording devices to Holy Land to engage in unauthorized food stamp transactions. (ECF DKT #91 at 4). As a result of this investigation, which lasted for about a year, it was determined that the loss was approximately $11,100. (ECF DKT #91 at 5). However, Defendants calculated the loss, resulting from the cooperating witnesses' observed purchases, at $10,101.68. (ECF DKT #92 at 6).
In estimating the total fraudulent loss total, the Government provided two methodologies: (1) Comparable Analysis Method - comparing Holy Land to other grocery stores in the same area and (2) Holy Land to Holy Land Analysis Method - examining Holy Land's Supplemental Nutrition Assistance Program ("SNAP") redemptions in three different time periods. (ECF DKT #91 at 11-14). Under the Comparable Analysis Method, the food stamp fraud loss was determined to be approximately $5.2 million. (ECF DKT #91 at 13). Under the Holy Land to Holy Land method, the food stamp fraud loss was determined to be approximately $2.6 million. (ECF DKT #91 at 14). Defendants contend that the total fraud loss is $10,101.68. (ECF DKT #92 at 7).
LAW AND ANALYSIS
1. Loss Calculation Standard
The Sixth Circuit has stated that the district court is to "determine the amount of loss by a preponderance of the evidence, and the district court's findings are not to be overturned unless they are clearly erroneous." United States v. Triana, 468 F.3d 308, 321 (6th Cir. 2006) (citing United States v. Guthrie, 144 F.3d 1006, 1011 (6th Cir. 1998)); see also United States v. Bracciale, 374 F.3d 998, 1003 (11th Cir. 2004) (noting that "the district court cannot merely 'speculate' as to the proper amount of loss."). "In situations where the losses occasioned by financial frauds are not easy to quantify, the district court need only make a reasonable estimate of the loss, given the available information." Triana, 468 F.3d at 320. Yet, the estimates "need not be determined with precision." Id. (quoting United States v. Miller, 316 F.3d 495, 503 (4th Cir. 2003)). Further, where the district court has multiple reasonable methods of calculating the amount of loss, and if the methods yield a similar amount of loss , this serves as evidence that the methods are accurate. United States v. Jarjis, 551 F. App'x 261 (6th Cir. 2014). (Emphasis added).
"When a defendant challenges one of the factual bases of his sentence . . . the Government has the burden of establishing the disputed fact by a preponderance of the evidence." United States v. Lawrence, 47 F.3d 1559, 1566 (11th Cir. 1995) (citing United States v. Ismond, 993 F.2d 1498, 1499 (11th Cir. 1993)). (emphasis added).
2. Comparable Analysis Method
The Government undertook a comparable analysis of Holy Land's SNAP redemptions versus other area grocery store redemptions based on the following features: proximity, size, type and category. (ECF DKT #91 at 11). Plaintiff compared Holy Land to Sapell's Food Basket due to its similar size; Assad's Bakery because of its proximity to Holy Land and because it was a Middle Eastern Market; and Giant Eagle because it had the same proximity to highway I-90 as Holy Land. (ECF DKT #91 at 12).
The Court finds this analysis problematic and faulty for many reasons. First, the SNAP redemptions of the three comparable stores was determined to be 16.33%, but the Government rounded up to 17%. (ECF DKT #91 at 12). Not only does that math go against elementary teachings on rounding up and down, but also Giant Eagle is a massive supermarket conglomerate whose nearest store is nearly ten times the size of Holy Land and not remotely similar in size and type. Like issues apply to Sapell's and Assad's Bakery. Sapell's does not cater to Middle Eastern consumers and sells products such as alcohol and pork, which are not found at Holy Land. Assad's is a bakery and take-out restaurant that offers only a few shelves of grocery items and is considerably smaller than Holy Land. Defendants point out that the 37% redemption rate attributed to Holy Land is erroneous; and in fact, should be 35.5%, by taking out a 468% redemption percent that was admittedly incorrectly included in the analysis by special agent Springer of the USDA OIG. (ECF DKT #91 at 12). The Comparable Analysis method offered by the Government is severely flawed, the $5.2 million is not supported by a preponderance of evidence and thus, is rejected.
3. Holy Land to Holy Land Analysis Method
The Government's Holy Land to Holy Land Analysis Method compared SNAP redemptions at Holy Land over three time periods: (1) "pre-fraud period," from January 2000 to the end of June 2003; (2) "fraud period," from July 2003 to October 2014; and (3) "post-fraud period," from November 2014 to August 2016. (ECF DKT #91 at 13). What is troubling is the asymmetry of the time frame, comparing eleven years and three months to twenty-two months. The lack of comparable duration in time periods does not allow the Court to make a determination with sufficient data to represent the "post-fraud period." Morever, the Government's brief creates more confusion surrounding the time frames provided because "[f]rom his previous experience with food stamp fraud cases, SA [Special Agent] Springer knew that fraud tends to cease once the defendants are aware they are being investigated." (ECF DKT #91 at 13). Yet, the Government's fraud period includes several years during which Defendants were previously investigated and indicted for food stamp fraud. The Court questions the entire alleged fraud period because the Government's own justification for creating a "post-fraud" period calls into question the very fraud period it seeks to establish. The prior conviction of Mohammad Mohammad, based on the Government's own evidence, occurred in 2004; thus, whatever fraud that was occurring at that time would have ceased because Defendants would have been aware they were investigated.
Regardless of the questionable time periods, the Government did not take into account several factors in its loss calculation. The Government did not consider the economy and the recession that occurred in 2008. This, coupled with the fact that the food stamp program expanded during this time, could explain the increase in Holy Land's percentage of food stamp benefits. The Government also did not take into account the change in Holy Land's store size, going from a 3,500 to 9,500 square foot store (increased by a factor of 2.7). Other factors were changes due to an influx of Holy Land's demographic base and the effects of possible negative publicity. Moreover, as a matter of consistency, the Government's first brief calculated the fraud under this analysis to be around $2.8 million, (ECF DKT #71 at 4), yet, in the Supplemental Brief it decreased to $2.6 million, without any explanation. Given these unreasonable time periods and internally inconsistent results, the Holy Land to Holy Land Analysis yields an unreliable outcome. Thus, the Court rejects the proposed $2.6 million amount as too speculative.
Further, under Jarjis, the Court concludes that both the Comparable Analysis and the Holy Land to Holy Land Analysis are inaccurate as they yield widely disparate results. Jarjis, 551 F. App'x. at 261. Under the Comparable Analysis, the Government alleges a loss of approximately $5.2 million, whereas under the Holy Land to Holy Land Analysis, the Government alleges a loss of either approximately $2.6 million or $2.8 million. Arguably, if the difference in loss amounts between two methodologies is more than two million dollars, the results under those methodologies could not be considered "similar" under Jarjis, id. Therefore, the Court concludes that the Government's loss estimates under both methods are neither accurate nor reliable.
4. Wahdan's Fraud Analysis
To accept the $51,000 fraud loss, the Court must find that shopping at one store and then, within a period of twenty-four hours, shopping at another store is too implausible. By virtue of that implausibility alone, the Government contends that it has to be fraud. The Court finds that line of argument erroneous. The Government asks the Court to find it reasonable to assume that every purchase made at the Wahdan store and then at Holy Land was a fraudulent transaction. That is far too speculative and at best a 50/50 proposition, which fails the preponderance of the evidence standard. People routinely shop at one store, do not find the exact items they are looking for, and then shop at another store within hours, let alone a day. Thus, given the speculative nature of the Government's line of reasoning, the Court also rejects the $51,000 fraud loss figure.
5. Actual Fraud Loss Calculation
Given that the above-stated methodologies are too speculative to adopt as reasonable fraud calculations, the Court can only consider the actual fraud loss investigation conducted by the USDA OIG. As for the issue of whether the actual fraud loss was $11,100 as calculated by the Government or $10,101.68 per Defendants, the Court adopts Defendants' calculation. The discrepancy appears to be that Government invalidates the entire transaction if any single item is not a permitted item for purchase with the food stamp card. For their part, Defendants excised out the impermissible food items, but retained the rest of the transaction. The Court adopts the latter's view that impermissible items do not invalidate entire transactions. In contract law, for example, when a court determines a provision is unconscionable, the court does not invalidate the entire contract, but rather strikes out the unconscionable provision and upholds the rest of the contract. Here too, the Court will not invalidate entire transactions just because a few items were impermissible, instead, the Court invalidates those purchases and upholds the rest of the transaction. Thus, the Court accepts Defendants' itemized determination and adopts the $10,101.68 actual fraud loss calculation.
6. Tax Loss
The Government presented evidence that Defendant Mohammad Mohammad underreported Holy Land's income for the 2012 and 2013 tax years, resulting in tax due and owing in the amount of $498,189.00. The Government argues that the Court should include the tax loss amount of $498,189.00 as relevant conduct for calculating Defendant Mohammad Mohammad's sentence.
Defendant Mohammad Mohammad contends that the presentence report offers no indication of the Government's intention of pursuing relevant conduct. Thus, introducing evidence of the tax loss amount is improper under Fed.R.Crim.P. 32(d).
First, under the heading "The Offense Conduct," Paragraph 24 of the Presentence Report (ECF DKT #62 *SEALED*) recites: "As to Mohammad Mohammad, although not charged, he committed conduct related to Tax Fraud in the amount of $498,189." Second, in the Written Guilty Plea of Mohammad H. Mohammad (ECF DKT #39), Paragraph 14 Relevant Conduct recites: "Defendant understands that the government intends to present evidence to the Court at sentencing, or at a separate hearing as designated by the Court, in the form of relevant conduct related to Defendant's tax liability." Defendant showed his assent by initialing at the bottom of the page of his written plea.
Therefore, the Court finds that the tax loss amount of $498,189.00 may appropriately be included in the total loss amount attributable to Defendant Mohammad Mohammad.
CONCLUSION
For the foregoing reasons, the Court rejects all of the Government's methodologies for calculating fraud loss, adopts Defendants' actual fraud loss calculation in part as to the amount of $10,101.68 for Omar Mohammad, includes the tax loss amount of $498,189.00 for Mohammad Mohammad, and finds the latter's total fraud loss to be $508,290.68.
IT IS SO ORDERED.
s/ Christopher A. Boyko
CHRISTOPHER A. BOYKO
United States District Judge Dated: April 17, 2018