Irobe v. US Dept. of Agriculture
890 F.3d 371
1st Cir.
2018
Check Treatment
Docket

MAHDI IROBE and SUUQA BAKARO GROCERY, Plaintiffs, Appellants, v. UNITED STATES DEPARTMENT OF AGRICULTURE, Defendant, Appellee.

No. 17-1943

United States Court of Appeals For the First Circuit

May 21, 2018

Before Torruella, Selya and Kayatta, Circuit Judges.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MAINE [Hon. D. Brock Hornby, U.S. District Judge]

Sarah A. Churchill, with whom Nichols & Churchill, P.A. was on brief, for appellants.

John G. Osborn, Assistant United States Attorney, with whom Halsey B. Frank, United States Attorney, and Sheila W. Sawyer, Assistant United States Attorney, were on brief, for appellee.

SELYA, Circuit Judge. This appeal challenges a finding by the United States Department of Agriculture (USDA), echoed on de novo review by the district court, that a grocery store unlawfully trafficked in Supplemental Nutrition Assistance Program (SNAP) benefits. See 7 U.S.C. § 2023(a)(13), (15). Our task requires us to decide, among other things, the allocation of the burden of proof in a civil action brought pursuant to 7 U.S.C. § 2023(a)(13) — a question of first impression in this circuit. After careful consideration, we hold that the district court properly placed the burden of proof on the grocer. See . We further hold that the court, acting at the summary judgment stage, supportably determined that the grocer had failed to carry this burden. See Consequently, we affirm the judgment below.

I. BACKGROUND

The plaintiffs are Mahdi Irobe and Suuqa Bakaro Grocery (a grocery store in Lewiston, Maine, catering principally to that community‘s sizeable Somali immigrant population). For ease in exposition, we refer to the plaintiffs, collectively, as the “Store.”

Since we are tasked with reviewing the district court‘s entry of summary judgment, we take the facts in the light most congenial to the nonmovant (the Store). See , . The Store is diminutive: it is only about 800 square feet in size, lacks shopping baskets or carts, and contains a single 2.5-by-1.5-foot-long checkout counter. It carries minimal amounts of fresh produce and frozen foods and does not offer many of the staples commonly found in markets (such as baby food, eggs, and fresh bread). In lieu of such staples, the Store offers Somali delicacies like goat and camel meat, along with certain nonperishables like sugar, flour, rice, pasta, and cooking oil. The Store operates in what might be called a “no frills” fashion: it does not have any optical scanning equipment, and it does not use a cash register in processing SNAP transactions. Instead, Irobe — the Store‘s owner and lone full-time employee — ordinarily computes each customer‘s purchases using a calculator. When Irobe cannot be at the Store, his brother-in-law pinch-hits for him.

On June 20, 2015, the USDA authorized the Store to deal in SNAP benefits (commonly known as “food stamps“). Because this authorization proved to be the first step down the road that led to this litigation, we pause to acquaint the reader with the SNAP framework.

Congress established SNAP “to safeguard the health and well-being of the Nation‘s population by raising levels of nutrition among low-income households.” 7 U.S.C. § 2011; see 7 C.F.R. § 271.1. Authorized merchants may accept SNAP benefits in payment for certain food items. See 7 U.S.C. § 2013(a). The USDA then redeems those benefits (as described below). See id.

SNAP-qualified households receive electronic benefit transfer cards (EBT cards), which are similar to debit cards and may be used to purchase eligible foodstuffs at authorized stores. In a typical SNAP transaction, a cashier rings up the total food purchases, a household member pays using her EBT card through a point-of-sale device, and the funds in the household‘s SNAP account are electronically transferred to the store‘s bank account.

Households may use their monthly SNAP allotments to procure food items that are suitable for “home consumption.” 7 U.S.C. § 2012(k). They may not use their allotments to procure cash, hot foods, or non-food items, even though such items may frequently be available at grocery stores. See id. These proscribed items include, for example, lottery tickets, alcoholic beverages, tobacco, vitamins, toothpaste, and cosmetics. See 7 C.F.R. §§ 271.2, 278.2(a).

Trafficking in SNAP benefits is unlawful, see 7 C.F.R. § 278.2(a); see also 7 U.S.C. § 2021(a)(1), (b)(3)(B), and a store engages in trafficking by accepting SNAP benefits in exchange for cash or other proscribed items, see 7 C.F.R. § 271.2. For instance, a store trafficks when it “accept[s] food stamps for sales that never took place,” allowing its customers to receive “cash rather than merchandise.” .

The Food and Nutrition Service (FNS) is the bureau within the USDA charged with administering the SNAP regime. This bureau maintains a searchable database containing the household, store, date, time, and amount involved in each and every SNAP transaction. If the FNS detects a statistically unusual pattern of SNAP transactions at a SNAP-authorized store, it typically refers the matter to a program specialist who arranges for a contractor to visit the store and conduct an on-site investigation. After completing her review of the relevant EBT data and whatever reports emerge from the on-site investigation, the program specialist makes a recommendation to the FNS section chief. If this recommendation is for further action, the section chief sends a charge letter detailing the allegations to the store and affords the store an opportunity to respond. See 7 C.F.R. § 278.6(b). Thereafter, the FNS issues its determination. See id. § 278.6(c).

Once the FNS has issued its determination, an aggrieved store may prosecute an appeal to an administrative review officer. See 7 U.S.C. § 2023(a)(3); 7 C.F.R. §§ 279.1(a)(2), 279.5. Upon completion of his work, the review officer issues the final agency decision. See 7 U.S.C. § 2023(a)(5); 7 C.F.R. § 279.5. The governing statute empowers the USDA to impose a lifetime program-participation ban on “the first occasion or any subsequent occasion” of trafficking, but such a ban is not an automatic response to a program violation; rather, the USDA has discretion, in lieu of such a ban, to levy civil monetary penalties under certain circumstances. 7 U.S.C. § 2021(b)(3)(B); see 7 C.F.R. § 278.6.

With this backdrop in place, we return to the case at hand. As of June 2015 (when the Store was first authorized to participate in SNAP), there were approximately forty-five other shops within a one-mile radius of the Store that accepted SNAP benefits, including several larger ethnic Somali markets, a Walmart Supercenter, and two chain supermarkets. In short order, the FNS detected a suspicious pattern of transactions in the Store‘s EBT database. This red flag sparked an investigation by a program specialist, which included two on-site visits in the fall of 2015. On December 17, the FNS sent a charge letter detailing hundreds of sets of suspicious transactions that, in its view, evinced trafficking.1 After receiving the Store‘s response, the FNS made a determination, dated January 12, 2016, in which it concluded that the Store had engaged in trafficking and permanently disqualified the Store from SNAP participation.

The Store seasonably requested an administrative review of the FNS‘s determination. On April 22, 2016, an administrative review officer upheld both the FNS‘s finding that the Store had violated the SNAP guidelines and the order for permanent disqualification.

The matter did not end there. The Store commenced an action in Maine‘s federal district court, challenging the agency‘s final decision. Following the close of discovery, the USDA moved for summary judgment. The Store opposed the motion. The district court heard oral argument, took the matter under advisement, and subsequently wrote a thoughtful rescript explaining why it would grant the motion for summary judgment. See . This timely appeal ensued.

II. ANALYSIS

A party aggrieved by the USDA‘s final determination may seek judicial review through “a trial de novo . . . in which the court shall determine the validity of the questioned administrative action in issue.” 7 U.S.C. § 2023(a)(13), (15); see 7 C.F.R. § 279.7. This de novo review is wider in scope than that available under the Administrative Procedure Act. See ; . When a court carries out such a review, it must reexamine “the entire matter” instead of simply determining “whether the administrative findings are supported by substantial evidence.” (quoting ). The section 2023 inquiry is not restricted to the record compiled before the agency but, rather, extends to the augmented record compiled before the district court. See ; (per curiam).

The de novo review standard applies only to the agency‘s liability determination, not to its choice of a sanction. See . A reviewing court may disturb the agency‘s choice of a sanction only if it finds that choice to be “arbitrary, capricious, or contrary to law.” ; see .

When it commenced its civil action, the Store included in its complaint a boilerplate allegation that the USDA‘s chosen sanction (a lifetime program-participation ban) was arbitrary and capricious. In proceedings before the district court, however, the Store abandoned this allegation and challenged only the agency‘s liability finding. The administrative record was submitted to the district court, and the parties engaged in a modicum of pretrial discovery. After the close of discovery, the USDA moved for summary judgment. See Fed. R. Civ. P. 56(a). As said, the district court granted that motion.

We review an order granting summary judgment de novo. See . A court may grant summary judgment only if the record, construed in the light most amiable to the nonmovant, presents no “genuine issue as to any material fact and reflects the movant‘s entitlement to judgment as a matter of law.” ; see Fed. R. Civ. P. 56(a). A fact is “material” if it “has the capacity to change the outcome of the [factfinder‘s] determination.” . An issue is “genuine” if the evidence would enable a reasonable factfinder to decide the issue in favor of either party. See

A party seeking summary judgment must, at the outset, inform the court “of the basis for [its] motion and identif[y] the portions of the pleadings, depositions, answers to interrogatories, admissions, and affidavits, if any, that demonstrate the absence of any genuine issue of material fact.” (citing ). So long as the movant crosses this modest threshold, the nonmoving party “must, with respect to each issue on which [it] would bear the burden of proof at trial, demonstrate that a trier of fact could reasonably resolve that issue in [its] favor.” “Such a showing ‘requires more than the frenzied brandishing of a cardboard sword.‘” (quoting ). The nonmovant must point to materials of evidentiary quality, see , and such materials must frame an issue of fact that is “more than ‘merely colorable,‘” (quoting ). Put another way, summary judgment is warranted if a nonmovant who bears the burden on a dispositive issue fails to identify “significantly probative” evidence favoring his position. .

Of course, the summary judgment standard cannot be applied in a vacuum. The resolution of such a motion may depend on which party bears the burden of proof on a particular issue. See, e.g., ; . Section 2023 does not indicate which party is to bear the burden of proof. Thus, we must allocate this burden before endeavoring to apply the summary judgment standard.

Faced with a statute that is silent about the burden of proof, we start by recognizing that Congress legislates “against a background of common-law adjudicatory principles.” . With this in mind, we presume — unless a contrary statutory purpose is apparent — that Congress has crafted a statute with the expectation that settled common-law principles will apply. See

In the American legal system, it is a settled principle that the risk of failing to prove a claim ordinarily falls on the claimant. See . It follows that when Congress authorizes a private right of action without specifying which party bears the burden of proof, the plaintiff bears that burden unless a statutory purpose to the contrary is apparent. See . Thus, we must impose the burden of proof on the claimant (here, the Store), unless there is sufficient evidence of a contrary legislative intent.

We discern congressional intent by examining “the language, structure, purpose, and history of the statute.” (quoting ). That task is simplified in this instance, as the Store has failed to muster even a sliver of persuasive evidence that Congress intended to depart from the traditional rule. This failure is unsurprising: Congress has deployed an elaborate remedial regime to ensure that SNAP benefits “are used only to purchase eligible food items, and are not exchanged for cash or other things of value.” . A store is in the best position to show what actually happens on its premises. Allocating the burden of proof to the store tacitly recognizes this reality and, in the bargain, incentivizes shopkeepers to maintain accurate records of all SNAP transactions. A contrary rule would have the perverse effect of rewarding businesses for shoddy record-keeping. See generally (explaining, in analogous context, that holding otherwise would “place a premium on an employer‘s failure to keep records“); (explaining, in labor-law context, advantage of imposing evidentiary burden on party in the “best position to maintain records“).

Precedent bears out this intuition. All of the courts of appeals that have addressed the burden-of-proof issue under Section 2023 have placed the burden of proof on the party challenging the USDA‘s finding of liability. See ; ; ; . We join those courts and hold that when a store challenges the USDA‘s determination that the store trafficked in SNAP benefits, the store bears the burden of proving by a preponderance of the evidence that its conduct was lawful.

In this case, the USDA submits that no genuine issue of material fact exists with respect to the Store‘s liability for trafficking. In support, it relies primarily on transaction reports derived from the EBT database, which analyzed all available statistical information concerning the Store‘s handling of SNAP benefits during the four-month period from July through October of 2015.

On de novo review, we give no weight to the agency‘s finding that trafficking occurred. See . Even so, Congress has expressly authorized consideration of both transaction information gleaned from EBT databases and reports of on-site investigations as tools in the USDA‘s efforts to detect fraud. See 7 U.S.C. § 2021(a)(2); see also 7 C.F.R. § 278.6(a). As a general matter, such information and reports may be probative of trafficking: in appropriate cases, they may be sources of circumstantial evidence of fraud, sufficient to prove that a store is trafficking in SNAP benefits. See .

Of course, common-sense inferences will almost always play a major role in such cases. For instance, the factfinder may reasonably infer trafficking when the redemption data shows that a store regularly processes purported SNAP transactions for significantly higher per-transaction amounts than nearby stores offering similar wares. See . So, too, the factfinder may reasonably infer trafficking when the redemption data shows that a small store with a selective inventory and limited staffing regularly processes purported high-dollar SNAP transactions in rapid succession. See .

In the case at hand, the EBT database discloses more than 400 sets of suspicious transactions at the Store. A representative sampling suffices to illustrate the point:

  • On 51 separate occasions, households used up at least 90 percent of their monthly SNAP benefits in fewer than nine hours.2 Historical data indicates that it is markedly inconsistent with the normal shopping behavior of SNAP-qualified households to deplete all or most of a household‘s allotment in one fell swoop. According to an unchallenged government analysis of SNAP-related shopping patterns, it usually takes a minimum of two weeks for a SNAP-qualified household to deplete 80 percent of its monthly allotment and three weeks to deplete 90 percent of that allotment.
  • During the relevant period, the Store engaged in 205 high-dollar SNAP transactions, that is, transactions ranging from $174 to $1,050. Yet, historical data indicates that, during 2015, the average SNAP transaction in the Lewiston area was about $45.
  • With respect to multiple purchases in quick succession, 103 pairs of SNAP transactions were made on the Store‘s point-of-sale device during the relevant period in under nine minutes. These paired transactions included 21 pairs of transactions completed in 60 seconds or less and four pairs of transactions completed in under 39 seconds. Given normal shopping behavior, the practical realities of shopping at the Store, see, e.g., supra note 2, and the availability of only a single clerk, these paired transactions raise obvious concerns.

To be sure, all of this evidence is circumstantial, but its cumulative effect is powerful. Irregular patterns may emerge in virtually any retail operation, but a drumbeat of irregularities can be highly probative of unlawful conduct. See ; cf. (noting, in different context, that pattern of irregularities may support inference of fraud). The large number of aberrational transactions reflected in the Store‘s EBT database are adequate to ground a strong inference of trafficking, especially given the Store‘s characteristics. While that inference is rebuttable, the allocation of the burden of proof dictates that the Store must point to some significantly probative evidence to rebut it (and, thus, fend off summary judgment).3 See .

The Store has failed to carry this burden. In particular, it has failed to challenge in any meaningful way the agency‘s data-compilation methodology, the accuracy of the compiled EBT data concerning SNAP transactions at the Store, and the reliability of the agency‘s historical data.4 Nor has the Store attempted to establish the bona fides of so much as a single transaction pinpointed by the agency (even though each of those transactions is clearly linked to a specific household).

Indeed, the Store relies almost entirely on Irobe‘s deposition testimony, in which he offered generalized, non-specific observations about his customers’ shopping habits. He testified, for example, that his customers sometimes would purchase expensive items (such as goat or camel meat) or buy rice in bulk. This testimony, the Store argues, creates a genuine dispute about whether the 205 EBT transactions exceeding $174 were for SNAP-eligible foodstuffs. Similarly, Irobe testified that customers sometimes arrived in large groups, due to limited means of transportation. This testimony, the Store argues, is sufficient to create a genuine dispute about whether the 103 pairs of rapid-succession transactions were legitimate.

These arguments lack force in the face of the ample transactional data. It is common ground that “[t]he mere existence of a scintilla of evidence in support of the plaintiff‘s position” is not enough to ward off summary judgment. . Where the plaintiff has the burden of proof, “there must be evidence on which the [factfinder] could reasonably find for the plaintiff.” There is no such evidence here — and generalized conclusions, such as Irobe has proffered, cannot fill the void. See (requiring “[f]actual specificity” because “a conglomeration of ‘conclusory allegations, improbable inferences, and unsupported speculation’ is insufficient” to ward off summary judgment (quoting )).

Struggling to gain some traction, the Store cites another passage from Irobe‘s deposition. There, he suggested that “most” nearby ethnic Somali stores did not carry expensive goat or camel meat. In the Store‘s view, this testimony explains why it processed so many high-dollar transactions (each of which exceeded the average SNAP transaction in the Lewiston area by over $125).

This explanation does not hold water. After all, Irobe offered no foundation upon which his surmise might plausibly rest. Nor is this gap bridged by other evidence: the record is barren of any competent proof reflecting what inventory was carried by these other emporia. A court need not “take at face value” a party‘s “subjective beliefs,” even if offered in the form of testimony, if those subjective beliefs are “conclusory,” “self-serving,” and lack factual support in the record. . This is such an instance: Irobe‘s unsupported opinion about the limited availability of particular merchandise in the Lewiston area is simply not a game-changer in the summary judgment calculus.

Finally, the Store trumpets a series of receipts documenting its purchase of foodstuffs from vendors between May of 2015 and December of 2015. The Store asserts that these receipts establish a factual dispute about whether it “was legitimately selling groceries to its customers.” This assertion misses the mark: the Store has made no showing as to how the amount of inventory reflected by the receipts relates to the total volume of the Store‘s sales during the relevant period. What is more, the mere fact that the Store bought some SNAP-eligible foodstuffs and sold them to SNAP-qualified households does not insulate it from a finding of trafficking. Merchants may conduct legitimate business side-by-side with unlawful trafficking. Nothing about the purchases evidenced by the receipts impugns the agency‘s finding that, on many occasions, the Store trafficked in SNAP benefits.

We recognize that SNAP is an important part of the safety net woven by Congress for persons in need. The program‘s efficacy, though, depends in large measure on the good faith of both SNAP-authorized merchants and SNAP-qualified households. The USDA is charged with ensuring that merchants and food-stamp recipients alike color between the lines. When the evidence suggests that program rules are being flouted, agency action is appropriate.

So it is here: the USDA identified hundreds of sets of suspicious transactions, strongly indicative of trafficking. By means of this showing, it marshalled a robust (though circumstantial) case of trafficking. The Store has not meaningfully rebutted the compelling inferences suggested by the agency‘s mass of circumstantial evidence. Even when drawing all reasonable inferences in favor of the Store, no rational factfinder could conclude that the Store had demonstrated by a preponderance of the evidence that the finding of trafficking was improvident. It follows inexorably, as night follows day, that the district court did not err in granting summary judgment in favor of the USDA.

III. CONCLUSION

We need go no further. For the reasons elucidated above, the entry of summary judgment is

Affirmed.

Notes

1
We say “sets” because many (indeed, most) of the challenged transactions comprise several items ostensibly purchased with SNAP benefits.
2
Common sense suggests it is especially unlikely that SNAP-qualified households would exhaust their allotments so quickly at the Store, given the Store‘s limited inventory, the lack of either shopping carts or baskets, the absence of optical scanning equipment, and the tiny checkout counter.
3
The district court cited with approval an unpublished Sixth Circuit opinion stating that to thwart summary judgment, a store “must raise material issues of fact as to each alleged violation.” (quoting (emphasis in original)). We are skeptical of any interpretation of this statement that would always require a transaction-specific rebuttal of every transaction. One can easily imagine, for example, that a series of transactions labeled as suspicious for a certain reason (say, no other similar stores have transactions that are as large) could each and all be rebutted by proof that the reason for suspecting them is wrong (say, there are other stores that charge as much). In either case, though, the accused store would have to proffer competent evidence, not merely conclusory generalizations. Here, the Store‘s evidence is so unfocused and so weak that we need not delve more deeply into the Sixth Circuit‘s statement.
4
This is not to say that the Store goes down without a fight: it does argue in its appellate brief that the EBT transaction data is “suspect.” But this argument is wholly conclusory, and the Store fails to identify any evidence supporting its conclusion. Where, as here, the nonmoving party bears the burden of proof on a material issue, that party cannot forestall summary judgment simply by relying on its lawyer‘s unsupported arguments. See .
Read the detailed case summary