This post concludes a four-part series prompted by the Supreme Court’s stay of the Eighth Circuit’s mandate in Food Marketing Institute v. Argus Leader Media, 889 F.3d 914 (2018). The first four posts (available here, here, here, and here) addressed the Food Marketing Institute’ frontal assault on the National Parks test for identifying “confidential” commercial and financial information under Freedom of Information Act Exemption 4. The Food Marketing Institute seeks to protect store-level aggregate data regarding the amounts SNAP recipients spend on foodstuffs (i.e., store-level aggregate redemption data).
Taking a different tack, on August 6, the Texas Retailers Association filed suit in the Western District of Texas to enjoin the U.S. Department of Agriculture (USDA) from releasing SNAP redemption data for grocery retailers within that state. It asserted that online retailers’ imminent entry into the SNAP market created different circumstances than those litigated in Argus Leader, necessitating a renewed competitive impact analysis. This post discusses the SNAP program’s impending expansion to include online retailers and the Texas Retailers’ suit.
Expansion of SNAP to Include Online Retailers
Online grocery shopping has become increasingly popular. One industry study predicts that the percentage of U.S. households that purchase some groceries online will increase from 23% in 2016 to 70% by 2024 at the latest. NIELSON & FOOD MARKETING INSTITUTE, THE DIGITALLY ENGAGED FOOD SHOPPER, DEVELOPING YOUR OMNICHANNEL COLLABORATION MODEL 5 (2018)(available for download here); see Complaint, Texas Retailers Association v. U.S. Department of Agriculture, Dkt No. 18 Civ. 659, ¶31 (August 6, 2018)(available in advance.lexis.com, Law360 database, “Texas Retailers Sue USDA to Keep SNAP Data Confidential” (August 6, 2018)). According to the USDA, online retailers had long overlooked the potential market provided by SNAP beneficiaries, who received their benefits in the form of electronic benefits transfers (EBT). USDA Food & Nutrition Service, Electronic Benefits Transfer Online Purchasing Pilot Request for Volunteers at 4 (Sept. 15, 2016) (“Request for Volunteers”). USDA theorized that online retailers assumed that SNAP beneficiaries could not afford their delivery fees or lacked digital access. Id. Over time, however, online retailers’ interest in serving SNAP beneficiaries increased.
Section 4011(b) of the Agricultural Act of 2014, Pub. L. 113-79, §4011(b), 128 Stat. 649, 791, adding subsection (k) to 7 U.S.C. §2016, directed the Secretary of Agriculture to authorize online SNAP transactions, subject to the results of a pilot project testing the idea’s feasibility. On September 15, 2016, the Food and Nutrition Service (“FNS”) requested volunteers willing to participate in the pilot program. See Request for Volunteers. On January 5, 2017, USDA announced its selection of nine retailers that would participate in a pilot program covering eight states. The list included companies that largely engaged in online retailing, like Amazon and FreshDirect, brick-and-mortar supermarket chains that had created an online presence, such as Safeway and Shoprite, and at least one brick-and-mortar megastore chain offering a variety of consumer goods beyond foodstuffs, namely Walmart. Press Release, USDA Announces Retailer Volunteers for SNAP Online Purchasing Pilot (Jan. 5, 2017). USDA expects to launch the pilot project in 2019. Id. The pilot program does not include the State of Texas. Thus, Texas retailers face little prospect of competition from online retailers under the pilot program.
The Texas Retailers’ Complaint
In its suit, the Texas Retailers Association asserts that USDA’s impending expansion of the SNAP program alters the calculus relevant to the release of store-level aggregate SNAP redemption data. See, Complaint, ¶30. The Association alleged that the Argus Leader litigation was conducted under the assumption that food retailers faced competition only from physically proximate retail outlets. Id. ¶¶27, 30. In the Texas Retailers’ view, the District Court had concluded that “the competitive harm of releasing” aggregate store level data “was attenuated by the various other factors retailers would consider in opening rival brick-and-mortar locations.” Id. at ¶27.
In addition, the District Judge had found no competitive harm in a context of informational symmetry. Id. at ¶32. In particular, South Dakota retailers were similarly situated because “most of the retailers whose information [was] subject to release [would be] getting access to similar information about their competitors.” Id. But the introduction of online retailers would create informational asymmetries. Online retailers would have access to years of brick-and-mortar stores’ sales data, while retailers selling only through brick-and-mortar stores would lack access to any historical data on the online competitors entering the market. The asymmetry could even continue after online retailers had become established in the market, because they might serve people in a broad region from a single distribution point. As a result, the precision of the aggregate sales data for online retailers and brick-and-mortar retailers would differ. Id. Physical stores would generate neighborhood-level data, reflecting their customer base; online retailers would generate only regional-level data given the breadth of the geographic area they could serve. Id. In short, the District Judge’s analysis in the Argus Leader litigation relied upon dated evidence relating to a “fundamentally different” SNAP program. Id. at ¶33.
The Complaint requests declaratory and injunctive relief. The Association sought a ruling that the release of store-level redemption data: (1) is not required because the data is exempt from disclosure under FOIA by virtue of Exemption 4, (2) is unlawful under the Administrative Procedure Act (presumably because the decision to release the information would be “arbitrary and capricious” in violation of 5 U.S.C. §706(2)(A)), and (3) is prohibited by that the Trade Secrets Act, 18 U.S.C. §1905. Complaint, ¶¶34-37. Injunctive relief was to be limited to aggregate store-level data for Texas retailers. Id. at ¶41.
The Texas Retailers’ complaint raises at least two procedural issues, issue preclusion and ripeness, and well as substantive issues.
The Texas Retailers’ case will obviously raise questions of issue preclusion (traditionally referred to as collateral estoppel). For a general discussion of issue preclusion, see 18 CHARLES ALAN WRIGHT, ARTHUR R. MILLER, ET AL., FEDERAL PRACTICE & PROCEDURE §4416 (3d ed.)(“FED. PRAC. & PROC.”)(available in westlaw); RESTATEMENT (SECOND) OF JUDGMENTS § 27 (1982)(available in westlaw). First, the Supreme Court could grant certiorari in Food Marketing Institute, and reverse the Argus Leader judgment. Though trial court judgments pending appeal can be treated as “final” so as to be accorded issue preclusion effect, the sounder course is to stay the later litigation pending final disposition of the appeal. See, 18 FED. PRAC. & PROC. §§4432, 4433.
Second, the Texas Retailers Association was not a party in the Argus Leader litigation, and is probably not sufficiently related to any of the parties in that litigation to be bound by any final judgment. See, 18 FED. PRAC. & PROC. §§4416, 4449. The Texas Retailers’ have no relationship with the South Dakota retail outlets whose information was sought (except perhaps for any chain that encompasses locations in both states). The Association might have a sufficient relationship with the USDA, which litigated the case on behalf of all SNAP participating stores. But it is not at all clear that the relationship between participating stores and the USDA suffices for issue preclusion purposes, particularly given that the FOIA request litigated covered redemption data for only South Dakota retailers. The Food Marketing Institute, which represents over 1200 food retailers, Food Marketing Institute v. Argus Leader Media, Application To Recall And Stay Mandate, etc. at 2 (Aug 7, 2017), intervened in the Argus Leader litigation, but that relationship may well not justify binding the Texas retailers, particularly those who are not members of the Food Marketing Institute.
Third, a judgment’s issue preclusion effect is limited to the issue resolved in the prior case. RESTATEMENT (SECOND) OF JUDGMENTS § 27, comment c; 18 FED. PRAC. & PROC. §4417. The Texas Retailers Association clearly seeks to distinguish the competitive harm issue tried in the Argus Leader litigation from the competitive harm issue it seeks to raise, namely whether release of store-level data will produce substantial competitive harm in the context of online retailers’ participation in the SNAP market. But the Texas Retailers’ effort to do so relies on a faulty description of the Argus Leader Court’s factual findings.
The finding of no “substantial competitive harm” in Argus Leader turned on the existence of extensive publicly-available data about retail outlets, which led the Court to conclude that adding SNAP redemption data would only marginally impact any retailer’s competitive position. The Texas Retailers’ Complaint does not suggest that plaintiff will contest the Argus Leader Court’s conclusion regarding the wealth of data already available. And, of course, the same informational asymmetries the Texas Retailers emphasize often occur in online retailers’ absence, if for instance, a brick-and-mortar chain expands into a new territory. In such circumstances, existing stores would lack relevant historical redemption data for the new entrant in the relevant geographic market. (Indeed, at least one study suggests that competition from existing retailers, rather than new entrants, have posed the greatest threats to a retailer’s existence. Daniel Hanner, Daniel Hosken, et al., Dynamics in a Mature Industry: Entry, Exit and Growth of Big-Box Grocery Retailers, JOURNAL OF ECONOMICS AND MANAGEMENT STRATEGY 22, at §§4.5, 4.6 (Spring 2015)(available at Wiley Online Library).
The Texas Retailer Association’s assertions regarding competitive harm assume a static marketplace, in which brick-and-mortar chains decline to enter the online retail market. Absent such a questionable assumption, access to information regarding redemptions at online retailers may well be just as useful to brick-and-mortar stores as the brick-and-mortar stores’ data is to online retailers. Moreover, store-level data may not be particularly helpful to online retailers that market regionally. Unless, online retailers price products differently by neighborhood, neighborhood-level data may have limited usefulness to a regional food retailer. Nor is SNAP redemption data likely help competitors formulate their marketing strategy toward non-SNAP customers, the bulk of the retail grocery customer base. The purchasing patterns of SNAP recipients do not necessarily mirror those of non-SNAP recipients.
Nevertheless, perhaps the presence of the issues outlined above distinguishes the competitive harm issue in the context of online retailers from that involved in the Argus Leader case.
The Texas Retailers’ suit may fail for lack of ripeness. Ripeness turns on an assessment of “the fitness of the issues for judicial decision and the hardship to the parties of withholding court consideration.” Abbott Laboratories v. Gardner, 387 U.S. 136, 149 (1967).
Texas retailers will suffer little harm if judicial consideration of their claims is delayed. Only information related to actual transactions reimbursed by the federal government is at issue. Were financial or commercial information unrelated to SNAP program redemptions at issue, the retailers could argue that they need to know whether USDA can hold their information confidentially before deciding whether to forego participation in the program to protect their corporate information. See, Application to Recall and Stay Mandate, supra, at 21. But it hardly makes sense to refrain from participating in the SNAP program, the only means to sell to a segment of customers, because by doing so competitors could determine one’s aggregate sales to SNAP recipients.
The issue the Texas Retailers seek to litigate is also unfit for resolution at this stage. First, the Texas Retailers’ own description of the critical issue suggests that it is a fact-intensive one rather than as a pure question of law. See, Abbott Labs, 387 U.S. at 149. Second, no one has sought aggregate store-level redemption data for Texas retailers. Third, online retailers are not yet in a position to compete with Texas brick-and-mortar retailers because Texas is not included in USDA’s pilot project. Fourth, were a FOIA request for Texas retailers’ data submitted, the USDA would assess whether the data is protected under Exemption 4, in consultation with the affected retailers. (Note, non-mutual collateral estoppel does not apply against the federal government, U.S. v. Mendoza, 464 U.S. 154, 160-62 (1984).) And, if the retailers were dissatisfied with the USDA’s determination, they could then initiate a “reverse FOIA” action.
Indeed, if the retailers were permitted proceed now, no entity may have a sufficient interest in conducting expensive litigation to oppose them. USDA probably has little interest in vigorously contesting the Texas Retailers’ claims; indeed, in the Argus Leader litigation the USDA argued that aggregate store-level data could and should be withheld. Nor will online retailers likely possess a sufficient stake to oppose efforts to protect aggregate store-level data (which may well be of limited use to regional retailers). And no media outlet or public interest group may be willing to bear the cost of trying a case to secure access to data that they may never seek.
Trying the competitiveness issue under the National Parks test could raise troubling issues, namely the prospect of inconsistent judgments and inconsistent treatment of aggregate store level data based on different geographic markets. But two substantive grounds may allow the court to avoid a detailed, fact-specific National Parks inquiry.
Simple Resolutions that May Not Involve an Evidentiary Hearing
First, the court could focus on another requirement of Exemption 4, namely that the information be obtained from the retailers. As I argued in Part II, the interest in the government’s fiscal transparency suggests that government reimbursement to commercial entities providing a service pursuant to a government aid program should never be considered as the commercial entity’s information. The information regards a transaction between the commercial entity and the government, or between a government beneficiary and a commercial entity; thus the redemption data is generated by coordinated action between government and the commercial entity. (Courts in the D.C. Circuit have discussed this issue in the context of access to the pricing provisions of government contracts. See, AT&T Information Systems v. General Services Adm., 627 F. Supp. 1396, 1403 (D.D.C. 1986); Racal-Milgo Government Systems v. Small Business Adm., 559 F. Supp. 4, 6 (D.D.C. 1981); but see, Hodes v. U.S. Department of the Treasury, — F. Supp. 2d —, 2018 WL 4680276 *4-*6 (D.D.C. Sept. 28, 2018).) Under such an approach, the issue of competitive harm would become irrelevant.
Second, the harms the Texas Retailers Association fears should not qualify as “competitive harm.” Both the Food Marketing Institute, in Argus Leader, and the Texas Retailers Association, in its Complaint, seem to express concern about a competitor using store-level aggregate data to aid in location selection decisions, Argus Leader Media v. United States Department of Agriculture, 224 F. Supp. 3d 827, 833-34 (D.S.D. 2016); Complaint, ¶19, or more broadly, competing for customers, Argus Leader, 224 F. Supp. 3d. at 834; Complaint, ¶¶13, 32. And given the limited nature of the data disclosed, which reveal only the aggregate redemptions at particular locations, the retailers really seek to protect information revealing the potential size of the SNAP beneficiary market. In effect, brick-and-mortar stores wish USDA to withhold information that would encourage other brick-and-mortar or online retailers to compete for SNAP recipients’ patronage. But such competition benefits SNAP recipients and does not unfairly harm retailers. Indeed, the Texas retailers essentially fear competition for SNAP customers from online retailers. Neither of these types of competitive effects should count as a competitive harm under the National Parks doctrine. Exemption 4 is not designed to function as a market barrier keeping new competitors out of the market, or, for that matter, to inhibit SNAP recipients’ access to the benefits of digital technology.
The National Parks Approach’s Anomalies
If the District Court cannot resolve the Texas Retailers’ case on the above bases, the case opens up the possibility of conflicting judgments from different federal courts. The Circuits differ in approaching the competitiveness issue. The Eighth Circuit in Argus Leader held that usefulness of information to competitors was insufficient to show competitive harm, requiring instead a showing of substantial competitive harm. Food Marketing Institute v. Argus Leader Media, 889 F.3d at 916-17. The D.C. Circuit has held that competitive harm can be established by showing that the requested data could be useful to competitors, even if that use is speculative and has not been quantified. See, e.g., State of Utah v. U.S. Dept. of the Interior, 256 F.3d 967, 970 (D.C. Cir. 2001); Public Citizen Health Research Group. v. FDA, 704 F.2d 1280, 1291 (D.C. Cir. 1983). Indeed, it has cautioned that a court “need not conduct a sophisticated economic analysis of the likely effects of disclosure” in determining whether “substantial competitive harm” would result from release of a business entity’s financial or commercial information. Id. Arguably, the First Circuit has found “substantial competitive harm” satisfied based on evidence regarding hypothetical harm from a “potential future competitor.” See, New Hampshire Right to Life v. HHS, 778 F.3d 43, 51 (1st Cir.), cert. denied, 136 S. Ct. 383 (2015). Moreover, the lower courts may well be inconsistent regarding the deference to accord agency affidavits on summary judgment, see generally, Margaret B. Kwoka, The Freedom of Information Act Trial, 61 AM. U. L. REV. 217, 244-50 (2011), and the Circuits do not agree on the standard of review for entry of summary judgment in FOIA cases where facts outside the documents in question are at issue, see, id. at 261-64. The Fifth Circuit’s last relevant Exemption 4 discussion appears in a cryptic 30-year-old decision, Calhoun v. Lyng, 864 F.2d 34 (5th Cir. 1988).
Even were the approach to recognizing substantial competitive harm perfectly consistent across the federal judiciary, the susceptibility of the information to release might well turn on the retail outlet’s geographical location. The retail grocery market is local, see, Dynamics in a Mature Industry, supra, at §3.1 & n.12, and is unlikely to transform itself into a national one in the near future. A true analysis of competitive harm argument would presumably turn on the geographic scope of the relevant local market and the attendant market conditions. Such inconsistency with regard to access to fairly basic data regarding retail outlets’ participation in a uniform national program seems troubling.
Both Food Marketing Institute’s impending certiorari petition and the Texas Retailers litigation raise important issues with regard to the scope of FOIA Exemption 4 in general and access to store-level aggregate SNAP redemptions in particular. And given the nature of Food Marketing Institute’s challenge and the textual trend in FOIA doctrine, the next year could see an upheaval in Exemption 4 law.