Opinion
No. C 10-02488 LB
10-18-2011
YUEN HAN LI, et al., Plaintiffs, v. UNITED STATES, Defendant.
JUDGEMENT AND ORDER AFFIRMING DISQUALIFICATION
I. INTRODUCTION
Plaintiffs Yuen Han (Nancy) Li and Judy Yuen Li, doing business as Golden Well (collectively, "Plaintiffs"), seek judicial review of the United States Department of Agriculture's Food and Nutrition Service's ("FNS") decision to permanently disqualify their retail market, Golden Well, from the Supplemental Nutrition and Assistance Program ("SNAP"), formerly known as the Food Stamp Program. Complaint, ECF No. 1 at 2, 14-15, ¶¶ 1-4, 64. Golden Well was permanently disqualified for allegedly having trafficked food stamp benefits (i.e., exchanged food stamp benefits for cash). Id. at 10-12, ¶¶ 45, 54. All parties consented to the court's jurisdiction, and the court conducted a two-day bench trial that started on September 26, 2011, and concluded the following day. ECF Nos. 5, 9, 73, and 74.
Citations are to the docket numbers in the Electronic Case File (ECF) with pin cites to the electronically-stamped pages at the top of the document (as opposed to numbers at the bottom).
In accord with Rule 52(a) of the Federal Rules of Civil Procedure, the court now sets forth its findings of fact and conclusions of law, affirming the agency's decision to permanently disqualify Golden Well from SNAP. The court finds that Plaintiffs did not demonstrate by a preponderance of the evidence that they did not engage in trafficking. The parties stipulated that if the court reached this result, then the agency's permanent disqualification of Plaintiffs from SNAP was not arbitrary and capricious. See Stipulation, ECF No. 48-4 at 1-2.
II. LEGAL STANDARDS
Benefits received through SNAP may "be used only to purchase food from retail food stores which have been approved for participation in the supplemental nutrition assistance program." 7 U.S.C. § 2013(a); 7 C.F.R. § 278.2(a). "[T]he buying or selling of coupons, ATP cards or other benefit instruments for cash or consideration other than eligible food" is "trafficking" under the SNAP regulations. 7 C.F.R. § 271.2. Stores are permanently disqualified on the first occasion of a trafficking violation. 7 U.S.C. § 2021(b)(3)(B); 7 C.F.R. § 278.6(e)(1)(i).
Any grocery store permanently disqualified from participating in SNAP may bring an action for judicial review by filing a complaint against the United States in federal district court. Plaid Pantry Stores, Inc. v. U.S., 799 F.2d 560, 561 (1986) (citing 7 U.S.C. § 2023(a)(13)). In reviewing a suspension of food stamp redemption privileges, the court must determine (1) whether the alleged violation occurred and (2) whether the penalty is valid. Id. at 563.
The court reviews de novo the first issue. Id.; see also Wong v. U.S., 859 F.2d 129, 132 (9th Cir. 1988). The store owner bears the burden "to prove by a preponderance of the evidence that the violations did not occur." Kim v. United States, 121 F.3d 1269, 1271 (9th Cir. 1997). At trial, the plaintiff may offer any relevant evidence available to support his or her case, whether or not it has been previously submitted to the agency. Id. at 1272. A finding of a violation may be made on the basis of "facts established through on-site investigations, inconsistent redemption data, [or] evidence obtained through a transaction report under an electronic benefit transfer system." 7 C.F.R. § 278.6(a). The court reviews the FNS's sanction under the arbitrary and capricious standard, determining "whether the sanction is unwarranted in law or without justification in fact." Wong, 859 F.2d at 132.
III. FACTS
A. Stipulated Facts
The parties stipulated to the following facts:
1. Congress designated SNAP to alleviate hunger and malnutrition among low-income households by augmenting their ability to purchase food. 7 U.S.C. §§ 2011, 2013(a). SNAP benefits may be redeemed only in exchange for food items from retail food stores which have been approved for participation in SNAP. 7 U.S.C. § 2013(a).Joint Proposed Final Pretrial Order, ECF No. 48 at 3-6. B. Court's Findings of Fact
2. Only those stores whose participation will effectuate the purposes of SNAP are authorized to accept SNAP benefits in exchange for eligible food items. 7 U.S.C. § 2018; 7 C.F.R. § 278.2(a). This statutory and regulatory scheme provides for the disqualification or, in exceptional circumstances, a civil money penalty, of participants if they accept or use SNAP benefits in violation of the program. 7 U.S.C §2021(a); 7 C.F.R. § 278.6(a). Trafficking is "the buying or selling of coupons ... or other benefit instruments for cash or consideration other than eligible food." 7 C.F.R. § 271.2.
3. SNAP benefits are distributed through an Electronic Benefit Transfer ("EBT") system. Benefit recipient receive plastic cards, similar to credit cards, that have magnetic strips encoded with a card number that is linked to the recipient's date. In a typical transaction, the cashier rings up the purchase and swipes the card through a Point of Sale ("POS") device. The POS communicates with a host computer, which approves or denies the purchase. The date and time is recorded for various accounting and reporting purposes, and enables the FNS to identify transactions which may indicate trafficking.
4. When a person makes a purchase with an EBT card, he is required to use a personal identification number ("PIN") in order for the cashier to be able to complete the transaction. In the absence of any suspicious circumstances, as long as the card user inputs the correct PIN, the cashier is not required to ask for any identification in order to confirm the identity of the individual making the purchase with the EBT card.
5. Plaintiff Judy Yuen Li and Yuen Han ("Nancy") Li are the co-owners of a food retail store named Golden Well, located at 918 Clay Street, in the City of San Francisco, California. They acquired Golden Well on October 24, 2007.
6. Their brother Donald Li also works at Golden Well. Donald worked at Golden Well in 2008 on a full-time basis, usually from Monday through Saturday.
7. Golden Well was a participant in SNAP from around December 2007 until its disqualification by FNS on February 23, 2009. SNAP is administered by FNS.
8. On January 16, 2009, FNS sent a charge letter to Plaintiffs, detailing the reasons FNS believed Golden Well had violated SNAP regulations. The charge letter stated that during the months of May 2008 through December 2008, FNS found a number of EBT transactions that according to FNS, indicated repetitive patterns of unusual, irregular, and inexplicable SNAP activity. FNS identified a number of transactions or sets of transactions that exhibited such activity and included a list of those transactions.
9. According to the January 16, 2009 charge letter, the listed transactions indicated the following four patterns of what FNS considered to be unusual, irregular, and inexplicable SNAP activity that FNS found to be highly suspicious and indicative of food stamp trafficking:
(a) Numerous transactions ended in even dollar values. More specifically, the listed transactions ended in a cent value of $.00. FNS considered this to consist in an unusual number of even dollar EBT transactions.10. In response to the January 16, 2009 charge letter, on January 31, 2009 Nancy and Donald Li met and spoke with Teresa Toups, the officer in charge at the FNS' Sacramento field office, and provided her with various documents and explanations for the questionable transactions. On February 23, 2009, FNS made a finding that the violations cited in its January 16, 2009 charge letter did in fact occur at Golden Well and thus permanently disqualified Golden Well from the SNAP program.
(b) Forty-five sets of consecutive EBT transactions in which the transactions in each set were made within one to three minutes of each other. FNS considered these EBT transactions to be too rapid to be credible.
(c) Twelve sets of multiple EBT transactions in which the transactions in each set were made from the same SNAP account and took place within a few minutes to a couple hours of each other. FNS considered these EBT transactions to have occurred within unusually short time periods.
(d) Thirty-four SNAP customers whose monthly benefits were depleted 93% to 100% in a single EBT transaction, and seven SNAP customers whose monthly benefits were depleted within two to 29 minutes of each other.
11. On March 2, 2009, plaintiffs requested an administrative review of the FNS' finding that Golden Well was involved in trafficking and its decision to permanently disqualify Golden Well from participating in SNAP. Plaintiffs submitted a brief to the administrative review officer along with various evidence and a report from an accountant.
12. On May 6, 2010, plaintiffs received a final administrative review decision from the USDA, upholding the initial decision to permanently disqualify Golden Well from SNAP.
1. Parties' Arguments and Evidence
a. Plaintiffs
Plaintiffs asserted that they did not traffic in SNAP benefits and that all of their EBT transactions were legitimate. They presented testimony that they did not engage in trafficking, and they submitted evidence that their inventory purchases matched their actual sales of goods. The short version of their argument is that Plaintiffs legally could not ask SNAP card users for identification, their customers with SNAP EBT cards had working PINs and bought qualifying food items, they never exchanged anything but eligible food for any SNAP transaction, their direct testimony on these points means that they carried their burden of showing by a preponderance of the evidence that the violations did not occur, and at worst, Defendant demonstrated only that people other than the actual SNAP beneficiaries used EBT cards to buy eligible food and did not demonstrate that Plaintiffs exchanged cash for cards.
In support of their case, Plaintiffs called the following witnesses.
Patricia McHugh, a longtime customer, testified about the following: her own purchases ("junk food" like soda and ice cream and not rice or Chinese speciality products); the store's inventory (including rice in big bags); the store's sales (including her observations of rice sales and the fact that customers never left the store with rice, leading to her assumption that the rice was delivered); her occasional observations of customers using EBT cards to buy rice; and she never saw EBT cards swiped for cash. RT 21:10, 22:7-10, 24-25, 26:1-6. On cross-examination, she testified that she never saw rice being delivered or cashiers rounding transactions to whole dollar amounts. RT 31-32.
Citations to the Reporter's Transcript (RT) are to page numbers (e.g., RT 22) and sometimes to page and line numbers (e.g., 22:9-17).
Nancy Li, co-owner and cashier, testified about the following: Plaintiffs' purchase of the store in 2007; the store's inventory (including convenience store items, large bags of rice, and Chinese speciality food items such as sausage, fish belly, oyster sauce, and sea cucumber); the store's storage capacity; the store's sales (including the sale of rice and Chinese speciality items to customers using SNAP cards); the store's location in Chinatown; the store's clientele (90% or more Asian origin generally and almost all Asian SNAP clientele); the store's acceptance of only cash and SNAP EBT cards; the practice of rounding up prices to a full dollar amount; the inability to ask SNAP customers using EBT cards for identification; how SNAP EBT users would make a purchase and make a subsequent purchase right after if they had an additional balance left on their card; the short time (less than a minute) for a SNAP EBT transaction (attributable to already knowing the price, quickly swiping the card, getting authorization for the transaction, printing a cash receipt and an EBT receipt, and stapling them together); the pricing policies for goods (generally 10 to 15 percent above cost); inventory receipts matched sales (showing that the store stocked the goods it claims to have sold); SNAP EBT card users always used the cards only to buy eligible goods; and the store never gave cash for EBT swipes. RT 36:14-25, 37-55, 59-60, 95, 97, 100-109. She also testified about the tax returns and the information reflected on it in support of the conclusion that the store bought inventory goods in a quantity that matched their claimed retail sales. RT 64-69, 78-94, 97. On cross-examination, she testified that her regular full-time job was at PG & E, 99.5% of the SNAP clientele were Asian, and the store never made a profit even when part of the food stamp program. RT 119-21.
Donald Li, Nancy's brother and the de facto store manager, testified about the following: the family's purchase of the store in 2007; his general workday (8 a.m. to 7 or 7:30 p.m., six and half days a week); his parents' participation at the store (stocking and helping to translate); the short time (less than a minute) for a SNAP EBT transaction (including inputting the price and quantity manually and generating separate EBT and cash register receipts); the store's inventory (including large bags of rice and the Chinese speciality food items described previously); his role in buying inventory personally and stocking the store; and how goods were priced (taking into consideration what customers would pay, competitive pricing with other Chinatown stores, profit margins, the propensity of customers to bargain, and his wholesale costs, which fluctuated). RT 132-33, 140-165. He also described the store's practice of rounding up to a dollar amount, in part because customers bargained. RT 151. Also, he described his EBT customers' purchase patterns of buying rice and Chinese speciality products in bulk (for example, multiple bags of rice or a couple of pounds at a time of sea cucumber), and their practice of making purchases early in the day and returning later with their receipts to pick the purchases up. (He ascribed this practice in part to wives or women shopping and husbands or men later returning with a car to pick the goods up, particularly when the goods were bulk and heavy.) RT 154, 155, 161, 181-186. He described how he would deliver for customers and how the store's location (which allowed easy double parking nearby) helped sales, even when his prices were slightly higher than other Chinatown markets. RT 165-178. He contrasted his products and sales with other nearby, specific stores, which did not stock Chinese speciality items and/or were more difficult to shop in and/or had more difficult parking even if they stocked rice. Id.; RT 221. He testified specifically that EBT customers did not buy convenience store-type items (like ice cream or sodas or junk food, which mostly tourists bought) but instead bought rice and the Chinese speciality items in bulk. RT 155, 160-61. Rice was one of the most popular purchases, and a significant number of SNAP EBT customers bought multiple bags of rice. It was not unusual, for example, for an SNAP EBT customer to buy two or three 50-pound bags of rice at one time. RT 185.
Mr. Li reviewed receipts establishing the store's inventory, including purchases of rice and Chinese speciality items like Chinese sausage, fish belly, oyster sauce, and sea cucumber. He reviewed the EBT and cash register receipts (which contain only dollar amounts and do not specify the actual good sold), and he explained how based on the dollar amounts on the receipts, he could tie them to his inventory items. For example, a certain dollar amount might reflect three bags of rice, and another dollar amount around the holiday season would correspond to the sale of Chinese sausage. He admitted that this process (summarized in Exhibit 1, which is both a summary of voluminous business records in columns one and two and a demonstrative aid in column three)could contain errors because he generated it in response to the audit by comparing and reconciling receipts showing EBT sales with the receipts for his inventory purchases. RT 240. Still, based on his knowledge of his inventory, his actual purchase prices for his inventory (shown on the inventory purchase records), and his pricing practices, he generally could match the sales receipts to his inventory purchases. Those records, he testified, showed that his sale of goods matched his inventory purchases of goods. See RT 187-219. He testified that every receipt for a SNAP transaction had a legitimate explanation. RT 217.
The court admitted columns one and two as a summary of voluminous records under Fed. R. Evid. 1006, and with the government's agreement, allowed Plaintiffs to use column three as essentially a demonstrative aid. Column three was not admitted into evidence, at least as an exhibit for Plaintiffs, although it is a party admission and admissible against the Plaintiffs to the extent that it was used by the government. See Order, ECF 73 at 4.
Mr. Li also testified about events on January 8, 2009, a date that FNS personnel visited the store covertly. He remembered that day, testifying specifically about his schedule: arriving early in the morning, a mini-rush from 8 to 9 a.m., and a slow down until the afternoon. RT 179-80. He remembered a specific regular customer who regularly picks up rice purchased by his wife. RT 180-81. That man came in a van later that day to pick up rice purchased by his wife earlier that morning, bringing the receipt. Id. This was consistent with the general purchasing patterns of SNAP EBT customers, as summarized previously. RT 179-186, RT 220-222
On cross-examination, Mr. Li testified that 99% of his customers were Asian and 100% of his EBT customers were Asian. RT 218-19. He also testified that he called FNS once in 2008 to ask if rice was an eligible good for the SNAP program. RT 223-224.
Ray Chan, an accounting expert, testified about the store's business records, including receipts for inventory purchases, receipts for the purported EBT sales, and tax records, and he concluded that the inventory purchases match the purported sales on the EBT receipts and that the flow of income on the financial records also was consistent with purchased inventory and purported retail sails. RT 250-70. Put another way, Golden Well bought and stocked goods in an amount consistent with their revenues reflected on their sales receipts and tax returns. Mr. Chan also testified specifically that based on the records, he could not conclude that no trafficking had occurred, and he could not determine whether legitimate sales had taken place. RT 270.
b. Defendant
Defendant counters that the irregular activity is indicative of trafficking by the store and that Plaintiffs' explanations are not credible.
Defendant called the following witnesses.
Judy Li, co-owner, testified about the following. For every EBT transaction, the cashiers at Golden Well printed a cash register receipt and stapled it to the EBT receipt. RT 318. EBT customers bought different items than cash customers. Id. Most bought bulk rice. Id. Cash customers bought smaller items like oranges, bananas, and ginger. RT 318-19. She agreed that EBT sales increased significantly from January 2008 to December 2008. RT 320; see Defendant's Exhibit 133. The reason is that they had more sales of rice over the year. RT 321. She called EBT to verify that rice was an "eligible good" to sell to SNAP EBT customers. RT 323.
Everett Harry, the government's accounting expert, testified that EBT transactions at Golden Well grew substantially in 2008 from just around $1,000 in January 2008 to over $25,000 per month by the year's end. RT 334. That pattern was unusual and unexplained. Id. Nothing about the store itself explained the extra interest, there were no expenditures on advertising, and there was no price point advantage on the rice that they were selling that comprised the bulk of the sales. Id. Also, Plaintiffs took over an existing store. Id. And, in contrast to increase in SNAP transactions from May to December 2008, cash sales did not increase during this time period. See Plaintiffs' Exh. 37.
And in 2008, the losses for the year according to Plaintiffs' tax return were $36,000. RT 334. Unexplained losses are an indicia of fraud, he opined. Id. He observed that Plaintiff's expert merely confirmed that Plaintiffs' records demonstrated a consistency between purchased inventory and purported sales. RT 336. He did not dispute that point, including the point that the money moving into and out of the Plaintiffs' bank account was consistent with the transactions. Id. He pointed out, however, that trafficking could occur under this scenario: the store could sell the inventory on the side for cash and give the cash to EBT customers. Id. He also expressed skepticism about the proffered explanations of bulk purchases of rice. Monthly consumption of rice by three people would be about 50 pounds. RT 338. It would take 15 people to eat 250 pounds. Id. A typical FNS household usually has children, meaning, it would be hard for that household to exhaust 50 pounds of rice. RT338-39. The store itself is a mini market in a touristy part of the city. RT 341. SNAP EBT customers are poverty-level consumers. RT 342. Bulk purchases like those here are by definition suspicious for SNAP beneficiaries. Id.
Teresa Toups, a now-retired FNS investigator, testified about her investigation of Golden Well and gave expert testimony about patterns of SNAP EBT trafficking.
Based on her experience in investigations, she testified that food stamp fraud generally involves paying cash on the dollar to the SNAP beneficiaries. Normally it is 50 cents on the dollar. RT 362. A "runner" takes the card from the beneficiary and takes it to a store that traffics in food stamp benefits. If the store swipes the card for $100, then it pays the runner $50 in cash. Id. In her investigations, she never encountered a situation where runners purchased goods with EBT cards. Id.
She testified about a covert visit to the store on January 8, 2009. RT 367. On the store's application for the FNS program, it stated that it opened at 8 a.m. so investigators arrived at 8 a.m. RT 368. The store did not open until a little after 9:30. As soon as it opened, an African-American man went into the store. Id. Another man entered the store, and two Asian women did too. During this time period, Toups's real-time monitoring of SNAP transaction data showed five SNAP EBT transactions in rapid succession: $80.80, $140, $95.97, $4 and $100. But surveillance revealed only two customers leaving the store, one with a gallon of milk and the other a small bag with unknown contents. RT 371-372 & Exhibit 110.042. In the afternoon, a $63.98 EBT transaction registered, but no one was in the store. RT 372. The transaction also was a "keyed" transaction, meaning, a card was not swiped and instead the card number was entered manually. RT 372-73. A SNAP transaction without the cardholder's presence is indicative of trafficking. Id. The last names associated with the true beneficiaries of the EBT cards were all non-Asian. RT 373.
On January 8, 2009, the interior of the store showed dusty cans and jars of food, suggesting very little stock turnaround. RT 375. The store had very little "staple" food. RT 374.
Ms. Toups described four specific sets of suspicious transactions at Golden Well that formed the basis for the administrative charges that led to FNS'S disqualification of Golden Well from the food stamp program: (i) 1,113 EBT transactions ending in a cent value of $.00; (ii) 45 sets of consecutive EBT transactions made too rapidly to be credible; (iii) 12 sets of multiple EBT transactions where the transactions in each set were from the same SNAP account in unusually close time frames; and (iv) 41 EBT transactions where the SNAP beneficiaries' benefits were exhausted in unusually short periods of time. See Joint Exh. 2 and attached tables showing transactions. She also described comparison data with other stores, summarized other beneficiary data from Golden Well sales, and opined about some of the Lis' behaviors.
i. 1,113 EBT Transactions Ending in Cent Value of $.00
1,133 SNAP EBT transactions from May through December 2008 ended in a same cents value. Joint Exh. 2. 569 of these were for $68. Rt 388. 77 percent of the transactions were made with EBT cards for SNAP beneficiaries who were homeless. Id. And 80% of the $68 transactions were for EBT cards for SNAP beneficiaries who were homeless. Id. Golden Well's explanation - that the $68 transaction represented two 50-pound bags of rice - was not credible for a homeless household. Id. Ms. Toups reviewed other patterns in the data, including that one homeless household registered the following transactions: (a) $68 (or two 50-pound bags of rice) on August 4; (b) $68 (same) on September 4; and (c) $68 (same) on October 6. RT 391. The SNAP benefits are disbursed on the fourth day of the month, and for this beneficiary, the benefit was $162 per month in August and September and $176 in October. Id. Right after receiving the benefit, the card was used to spend a large percentage of the total monthly benefit on rice. RT 392. That explanation makes no sense other than in a trafficking scenario involving runners getting cash for the cards. Id. A similar pattern was shown in another household. RT 392-393.
A total of 2,021 SAP transactions conducted by Golden Well from May to December 2008 were analyzed. 29% of these (or 569) were in the exact amount of $68. That means roughly 50% of the same dollar transactions were for the exact amount of $68.
ii. 45 Sets of Consecutive EBT Transactions Made Too Rapidly To Be Credible
In these transactions made from May through December 2008, multiple purchase transactions were made very rapidly. RT 393. A legitimate transaction requires different steps: ringing up the transaction on the cash register, swiping the EBT card, having the customer enter his or her PIN into machine, waiting for authorization/approval of the transaction, inputting the transaction amount, printing the cash register receipt, printing the EBT machine receipt, stapling the two receipts together, and handing the customer copy of the EBT receipt. Rt 393-94. Given these steps, the transactions here happened very rapidly. For example, one of the transactions took a minute and fifteen seconds. RT 394. Also, the three SNAP beneficiaries for specific transactions she reviewed were two homeless persons and a resident of a single room occupancy (SRO) hotel. RT 395.
iii. 12 Sets of Multiple EBT Transactions Where Transactions In Each Set Were From Same SNAP Account in Short Time Frames
In 12 sets of SNAP EBT transactions from May through December 2008, multiple transactions were made from individual benefit accounts in very short time frames. RT 395. This does not comport with standard purchasing behavior. Id. For example, one set of transactions for an EBT card for a homeless SNAP beneficiary showed a purchase for $68 (explained by Plaintiffs as two bags of rice) and then, less than 30 minutes later, another purchase of $68 (again purportedly two bags of rice). RT 396. Another example showed a "starter" transaction at one store for a small store called Pitco located outside Chinatown, which is how runners check balances. RT 409-10. Then the transactions began in Chinatown: large transactions at Golden Well and different stores within a few minutes of each other, including a store that subsequently was closed for trafficking. RT 410. Another example showed a series of transactions over a few days with the same card, including multiple transactions at different stores in rapid succession. RT 411. By contrast, other transactions are more consistent with typical beneficiary behavior, like meals at authorized restaurants. RT 411-12. The large transactions are typical runner transactions, and the small ones are normal use that one would expect from a homeless recipient. RT 412.
iv. 41 EBT Transactions Where Benefits Exhausted in Unusually Short Time Periods The evidence here showed 34 SNAP customers whose monthly benefits were depleted 93% to
100% in a single EBT transaction, and seven SNAP customers whose monthly benefits were exhausted within two to 29 minutes. See Joint Stipulated Fact 9(d). These patterns suggest trafficking. RT 396-97. This is unusual shopping behavior. RT 397. Typically, FNS studies show SNAP households spend 60 percent of benefits after week one and have 10 to 12 percent of benefits left by week three. RT 398. In one example, the SNAP beneficiary was homeless, and spent virtually his whole allotment of $162 immediately for (purportedly) large quantities of rice. Id.
v. Comparison SNAP Data With Other Stores
Compared to 10 other similar small to medium grocery stores in the area selling rice and similar products, all within a one-half mile radius, Golden Well had the largest average SNAP transaction amount ($57.99, with the next highest being $29.72), the second highest average monthly SNAP redemption ($16,470.47 compared to a high of $17,786.22), and the second highest total redemption ($98,825 compared to a high of $106,718 and the next highest of $59,839). Defense Exh.131; RT 400-403. This is unusually high. RT 402. Golden Well is higher than any of the comparison stores. RT 404-405; Defense Exh.131. 28% of 2,021 SNAP transactions (or 569 transactions) conducted by Golden Well during the eight-month period from May to December 2008 were for the same amount of $68. Defense Exh. 107. By contrast, its next closest business had only 15 transactions for the same amount, and that comprised only 14 percent of that business's total transactions. Defense Exh. 132. Defense Exhibit 132 shows SNAP transactions involving the four suspicious categories (cent value ending in .00, rapid transactions, repetitive transactions, and balance depletions). RT 404.
vi. Other Analysis of Beneficiary Data From Golden Well
As discussed previously, 2,021 SNAP transactions were conducted by Golden Well from May to December 2008. Fewer than seven percent of the EBT cards used at Golden Well were issued to beneficiaries who were Asian. RT 406; Defense Exhibit 110.039. Households with Asian surnames made only 16 percent of Golden Well's SNAP transactions. Defense Exh. 107. No households with Asian surnames made transactions for $68. Id. Only four households with Asian surnames made SNAP transactions in amounts of $34 and $102. Id. Over 63 percent of the four categories of suspicious transactions were conducted by households identified as homeless or living in SROs based on their addresses. Id.
Over 70 percent of the SNAP EBT cards used at Golden Well were issued to homeless beneficiaries. RT 406. For the non-homeless households, most lived in areas where they would travel passed larger, better-stocked grocery stores to transact at Golden Well. RT 406-07. Only eight non-homeless or non-SRO households lived within one-half mile of the store. Defense Exhs. 104 and 110. Travel like that here is indicative of trafficking. RT 407.
Other data showed that customers using EBT cards had different purchasing patterns than cash-paying customers. (Golden Well accepted only EBT cards and cash.) For example, data from October and November 2008 show that there were only 52 non-SNAP transactions of $20 or more. Defense Exh. 110. Not a single non-SNAP customer made a $68 transaction (corresponding to the rice sales that figure so prominently in the SNAP transactions). Id.
vii. Other Opinions
Ms. Toups opined that Donald Li and Judy Li called FNS to see if rice was an "eligible food" for SNAP transactions because they planned to use rice as a staple "sale" to cover up their EBT trafficking. RT 364. Similarly, she noted that they printed cash receipts up only for SNAP transactions (and stapled them to the EBT receipts), that approach made no sense because the EBT receipt showed the dollar amount of the transaction, and the likely explanation is that the Plaintiffs used this process to be able to show that cash receipts matched EBT transactions, thereby facilitating a cover-up of their trafficking activities. RT 378.
2. Court's Finding of Facts
The court finds that Golden Well's business records establish that the purchased inventory matched the store's purported retail sales, and the tax returns and transactions in the bank account were consistent with the purchased inventory and purported sales. The evidentiary support for this finding includes the business and financial records, the testimony of Nancy Li and Donald Li, and analysis of the records by plaintiffs' expert Ray Chan and defense expert Everett Harry.
The court also adopts the SNAP data admitted through Ms. Toups. That data, which Plaintiffs generally did not dispute, is set forth in the previous section under III, B, 1 (Parties' Arguments and Facts), b (Defendants), subsections i through vi, on pages 10 through 13, and the court adopts those sections in their entirety as fact findings. The court finds (and again, Plaintiffs did not dispute this) that this data establishes that people other than the EBT card beneficiaries used their EBT cards at Golden Well. The court finds that the SNAP EBT transaction analysis shows patterns of EBT transactions that are indicative of trafficking for the reasons summarized above.
The court holds that the Plaintiffs' business and financial records alone do not sustain the Plaintiffs' burden of establishing that trafficking violations did not occur, particularly given that Plaintiffs' expert forthrightly admitted that he could not rule out trafficking or determine whether the purported sales actually happened, and given that the data. Instead, the Plaintiffs sustain their burden only if the court finds credible their testimony that they never gave cash for the EBT transactions. Nancy Li's observations were based only on her work on Sundays (a half day or full day). Her testimony arguably has corroborative value, but because her observations are based on limited exposure, she cannot establish that trafficking did not occur. Whether Plaintiffs meet their burden thus turns on whether the court accepts the explanations of Donald Li and Judy Li that they never paid cash for EBT transactions.
Customer Patricia McHugh testified only about occasional observations of customer behavior, and that does not affect the issue of trafficking one way or another.
--------
Both presented well at trial. And in the end, the court cannot conclude that it is impossible that Plaintiffs' explanation - that Defendants at best established that persons other than the EBT beneficiaries bought eligible food - could never have happened. But Plaintiffs bear the burden of establishing no trafficking, and the court does not credit their explanations based on the following.
One, the data is consistent with trafficking, meaning, paying cash for the EBT swipes, for the reasons discussed by Ms. Toups and Mr. Harry.
Two, the covert operation on January 8, 2009, revealed real-time EBT transactions without consumers being present, which is indicative of trafficking.
Three, Donald Li testified that he remembered January 8, 2009 so clearly that he remembered specific transactions involving picking up goods later. But his testimony that he arrived at work early that day, and had a mini rush at 8 a.m., is inconsistent with FNS's contemporaneously-documented surveillance that revealed he did not open the store until after 9:30 a.m. That inconsistency raises doubts about his testimony.
Four, multiple transactions at successive stores in the same time period is consistent with trafficking and not consistent with a narrative of runners actually buying goods, especially given Ms. Toups's testimony that she never encountered a situation where runners purchased goods with EBT cards.
Five, Golden Well's practice of printing both cash and EBT receipts makes no sense. The likely explanation is that it was an attempt to establish that the EBT transactions were actual sales.
Six, Plaintiffs' attempts to reconcile the EBT transactions to their actual inventory purchases was minimally probative. It establishes only that the Plaintiffs purchased inventory and does not establish that they sold that inventory in sales reflected on the EBT receipts.
Seven, Plaintiffs' testimony was that their EBT customers were all Asian, and they described a pattern of shopping in bulk and later pickup by (for example) a husband or delivery by Mr. Li. The clear arc of the narrative is that these are normal shopping patterns of an Asian customer base buying rice and Chinese speciality goods. It presupposes a secondary market for EBT cards by these Asian customers and acceptance of these customer shopping patterns. That narrative is not consistent with the actual data. For example, the customer patterns that the Lis describe are inconsistent with the "starter" purchases at one store and then large transactions at Golden Well and different nearby stores within minutes of each other. The customer patterns they describe are inconsistent with one customer's large bulk purchases of rice at the beginning of the month for three consecutive months in amounts that - according to Mr. Everett - are inconsistent with an ability to consume the rice. It also is not credible that only Asian customers with EBT cards (and not cash-paying Asian customers) bought bulk rice and speciality goods. (As summarized above, the data shows, and the Lis testified, that EBT cards were used for the bulk purchases here, and cash customers did not buy these items.)
Eight, as Mr. Everett testified, nothing explains the exponential growth of EBT transactions in 2008, starting at $1,000 in January to over $25,000 in December. Cash sales did not increase in this period.
Nine, the Lis calls to FNS to ensure that bulk rice was an "eligible food" seems suspicious under the circumstances and suggests an attempt to use a product as a staple sale for purported EBT transactions.
These findings of fact are sufficient to support the court's conclusion that Plaintiffs did not show that they did not engage in the trafficking of SNAP benefits. Not only are the Plaintiffs' explanations not credible, but also, the plausible explanation for the data is that it reveals trafficking in SNAP benefits.
IV. CONCLUSION
For the foregoing reasons, the court concludes that Plaintiffs failed to demonstrate by a preponderance of the evidence that they did not engage in trafficking and thus, pursuant to the parties' stipulation, affirms FNS's decision to permanently disqualify Golden Well from participating in SNAP.
IT IS SO ORDERED.
LAUREL BEELER
United States Magistrate Judge