Garnett v. Zeilinger

Full title: SHONICE G. GARNETT, et al., Plaintiffs, v. LAURA ZEILINGER, Defendant.

Court: UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA

Date published: Sep 9, 2020

485 F. Supp. 3d 206 (D.D.C. 2020)

Opinion

Case No. 17-cv-1757 (CRC)

2020-09-09

Shonice G. GARNETT, et al., Plaintiffs, v. Laura ZEILINGER, Defendant.

Chinh Q. Le, Pro Hac Vice, Chelsea C. Sharon, Jennifer F. Mezey, Legal Aid Society of the District of Columbia, Peter R. Bisio, Susan Musser, Pro Hac Vice, Lance Murashige, Hogan Lovells US LLP, Emily Goldman, Washington, DC, Kaitlin Welborn, Pro Hac Vice, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, Marc Cohan, Pro Hac Vice, Travis England, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiffs Shonice G. Garnett, Bread for the City. Emily Goldman, Susan Musser, Pro Hac Vice, Lance Murashige, Hogan Lovells US LLP, Jennifer F. Mezey, Chinh Q. Le, Pro Hac Vice, Legal Aid Society of the District of Columbia, Washington, DC, Kaitlin Welborn, Pro Hac Vice, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, Marc Cohan, Pro Hac Vice, Travis England, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiffs James Stanley, Kathryn Harris. Chinh Q. Le, Pro Hac Vice, Jennifer F. Mezey, Legal Aid Society of the District of Columbia, Emily Goldman, Lance Murashige, Hogan Lovells US LLP, Washington, DC, Kaitlin Welborn, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiff Darroll Green. Conrad Z. Risher, Fernando Amarillas, Mateya Beth Kelley, Office of the Attorney General, Gregory Martin Cumming, U.S. Department of Justice, Washington, DC, for Defendant.

CHRISTOPHER R. COOPER, United States District Judge

Chinh Q. Le, Pro Hac Vice, Chelsea C. Sharon, Jennifer F. Mezey, Legal Aid Society of the District of Columbia, Peter R. Bisio, Susan Musser, Pro Hac Vice, Lance Murashige, Hogan Lovells US LLP, Emily Goldman, Washington, DC, Kaitlin Welborn, Pro Hac Vice, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, Marc Cohan, Pro Hac Vice, Travis England, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiffs Shonice G. Garnett, Bread for the City.

Emily Goldman, Susan Musser, Pro Hac Vice, Lance Murashige, Hogan Lovells US LLP, Jennifer F. Mezey, Chinh Q. Le, Pro Hac Vice, Legal Aid Society of the District of Columbia, Washington, DC, Kaitlin Welborn, Pro Hac Vice, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, Marc Cohan, Pro Hac Vice, Travis England, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiffs James Stanley, Kathryn Harris.

Chinh Q. Le, Pro Hac Vice, Jennifer F. Mezey, Legal Aid Society of the District of Columbia, Emily Goldman, Lance Murashige, Hogan Lovells US LLP, Washington, DC, Kaitlin Welborn, ACLU of Alabama, Montgomery, AL, Katharine Deabler-Meadows, Pro Hac Vice, National Center for Law and Economic Justice, New York, NY, for Plaintiff Darroll Green.

Conrad Z. Risher, Fernando Amarillas, Mateya Beth Kelley, Office of the Attorney General, Gregory Martin Cumming, U.S. Department of Justice, Washington, DC, for Defendant.

MEMORANDUM OPINION

CHRISTOPHER R. COOPER, United States District Judge Not being a State often places the District of Columbia at a disadvantage. In this case, however, it works to its benefit.

The Supplemental Nutrition Assistance Program ("SNAP") Act requires participating States to abide by strict deadlines for processing benefit applications and periodically recertifying benefit eligibility. The statute treats the District as if it were a State. In 2017, a group of D.C. residents and the non-profit organization Bread for the City brought suit under 42 U.S.C. § 1983 against the director of the agency that administers the District's SNAP program over problems the city was having in meeting these statutory deadlines. In May 2018, the Court entered a partial preliminary injunction compelling the District to comply with the SNAP Act's deadlines for recertifying benefit eligibility. Following extensive discovery, both parties now move for summary judgment, and Plaintiffs move for a permanent injunction.

Ordinarily, plaintiffs challenging a State's compliance with SNAP Act processing timelines would be entitled to summary judgment if they could show that the State has fallen short of absolute compliance (or something very close to it). But, because D.C. is a municipality, not a State, Plaintiffs here shoulder a heavier burden. They must show not only that the District has failed to strictly comply with the Act's processing deadlines; they must also establish that the failure resulted from a policy or practice adopted by District officials. See Monell v. Dep't of Soc. Servs. of City of New York, 436 U.S. 658, 694, 98 S.Ct. 2018, 56 L.Ed.2d 611 (1978). Plaintiffs have not satisfied this second element of liability. Finding that the record reflects substantial efforts by the District to improve its timeliness performance, the Court concludes that there is insufficient evidence that a policy or practice of the District caused violations of the Plaintiffs’ statutory rights to timely processing of their SNAP applications. The Court will therefore vacate its earlier injunction and grant summary judgment to the District.

I. Background

The Court has addressed the relevant background at length in several prior opinions in this case. See Garnett v. Zeilinger ("Garnett I"), 301 F. Supp. 3d 199, 203–04 (D.D.C. 2018) (class certification order); Garnett v. Zeilinger ("Garnett II"), 313 F. Supp. 3d 147, 150–54 (D.D.C. 2018) (preliminary injunction opinion); Garnett v. Zeilinger ("Garnett III"), 323 F. Supp. 3d 58, 62–63 (D.D.C. 2018) (ruling on motion to dismiss). It therefore will provide only an abbreviated background discussion here, with a focus on new developments since the Court's prior rulings.

A. Regulatory Background

The Supplemental Nutrition Assistance Program ("SNAP") provides benefits to low-income households to help them purchase food. 7 U.S.C. §§ 2011 – 2036d. To receive assistance, a household must file an initial application for benefits, participate in an interview, and verify certain eligibility information. Id. § 2014(a); 7 C.F.R. § 273.2(a)(2). If approved, the household is certified to receive benefits for a specific timeframe, known as the "certification period." 7 U.S.C. § 2020(e)(4) ; 7 C.F.R. § 273.10(f). Prior to the end of its certification period, a household must submit a recertification application to continue receiving benefits. 7 U.S.C. § 2020(e)(4).

Responsibility for administering SNAP is divided between the federal government and state governments. States that elect to participate in the program receive funding for benefits and 50% of administrative costs from the federal government. Id. §§ 2013(a), 2025. In return, States must administer their programs in accordance with federal statutory and regulatory requirements. See 7 U.S.C. § 2020(e) ; 7 C.F.R. § 273.2. As relevant here, those requirements include strict deadlines within which States must process their residents’ SNAP applications.

With respect to initial applications, the certification process must be completed, and benefits provided, within 30 days of an application's filing. 7 U.S.C. § 2020(e)(3). The deadline is even more stringent—no later than seven days after an application is filed—for so-called "expedited" applications from households with extremely low income. Id. § 2020(e)(9)(A) ; 7 C.F.R. § 273.2(i)(2), (3), (4). As for recertification applications, the agency must provide each household with a notice of expiration and of the need to recertify before the start of the last month of their certification period. 7 U.S.C. § 2020(e)(4) ; 7 C.F.R. § 273.14(b)(1). If a household submits its recertification application at least 15 days prior to the expiration of its certification period, the State must provide benefits—if the household remains eligible—without interruption. 7 U.S.C. § 2020(e)(4) ; 7 C.F.R. § 273.14(c), (d). If a household submits its recertification application within a 45-day grace period after the 15th day of the last certification month, the State must process the application within 30 days of the application filing date. 7 C.F.R. § 273.14(e)(1).

The certification or recertification process may be held up if the customer is missing verification or does not complete an interview. In the former scenario, the agency is required to provide households with a notice of required verification and at least ten days to provide the missing verification. 7 C.F.R. §§ 273.2(f) (initial application), 273.14(b)(4) (recertification). In the latter situation, the agency must "promptly" schedule an interview within the 30 days of the application and, if a household misses an interview, provide a notice of missed interview. Id. §§ 273.2(e)(3) (initial application); 273.14(b)(4) (recertification).

The federal Food and Nutrition Service ("FNS" or the "Service"), an agency within the Department of Agriculture, oversees the States’ administration of SNAP. 7 U.S.C. § 2013(c) ; 7 C.F.R. § 271.3. As part of its extensive oversight duties, FNS monitors the States’ SNAP application processing using three timeliness metrics. See Memorandum from Lizbeth Silbermann on the Three Ways Initial SNAP Application Processing Timeliness is Measured to All SNAP Regional Directors ("FNS June 2017 Memo") 1 (June 2, 2017), Gov. MSJ, Exh. Q. First , the Application Processing Timeliness Rate ("QC APT Rate") is based on a sample of approximately 90 active cases (households currently receiving benefits) pulled each month. Id. at 1–2. The State agency's Quality Control ("QC") team collects information on whether initial applications sampled were processed on a timely basis, and the number of applications approved timely is divided by the total applications in the sample to arrive at the QC APT Rate. Id. Second , the State Timeliness Rate is calculated from the entire universe of State SNAP cases by dividing the total number of initial applications approved timely by the total number of applications. Id. at 3–4. Finally , the Certification Section of the FNS Program and Budget Summary Statement, Part B-Program Activity Statement ("FNS-366B") tracks the number of initial and recertification applications approved or denied by the State agency in the specified reporting quarter, including the number of application decisions that were overdue by 1–30 days, 31–60 days, 61–90 days, and 91 days or more. Id. at 4–5.¹ In addition to these quantitative measures, FNS regularly monitors States’ SNAP administration through management evaluation reviews, advocate and client complaints, and other methods of information gathering. Id. at 1; see, e.g., Pl. MSJ, Exh. G.

1 An initial application is overdue if the approval determination is made more than 30 days from the date of a regular application and more than seven days from the date of an expedited application. Id. at 5. A recertification application is overdue if the approval decision is made after the date of the household's normal issuance date. Memorandum from Lizbeth Silbermann on Clarifications for Reporting on the Certification Section of the FNS-366B to All SNAP Regional Directors ("FNS Jan. 2017 Memo") 3 (Jan. 11, 2017), https://fns-prod.azureedge.net/sites/default/files/snap/366B-Certifications-Question-and-Answer.pdf.

Based on this monitoring, FNS enforces the SNAP Act's timeliness requirements through a standard escalation protocol. See Memorandum from Lizbeth Silbermann on Guidance for Improving State Timeliness Rates & Standardizing the Escalation Process to All SNAP Regional Directors ("FNS March 2016 Memo") (Mar. 18, 2016), Gov. MSJ, Exh. S, ECF No. 138-21. The protocol is based primarily on the QC APT Rate: FNS considers a rolling six-month average QC APT Rate of 95% or higher to be acceptable performance and subjects States to escalation procedures only if their rate dips below 90%. Id. at 2.² The first step in the escalation protocol is to subject States to a Corrective Action Plan ("CAP"). Id. at 3–4; 7 C.F.R. § 275.17. If States do not improve their timeliness within a specified amount of time, FNS sends an Advance Warning Letter of possible suspension of federal funds, see FNS March 2016 Memo 4–5; 7 C.F.R. § 276.4(d)(1), which may be followed by a Formal Warning Letter, see FNS March 2016 Memo 5–6; 7 C.F.R. § 276.4(d)(2), and eventually, suspension or disallowance of federal SNAP funding, see FNS March 2016 Memo 5–6; 7 C.F.R. § 276.4(e). FNS will assist poor performing States by engaging with their senior management officials, expanding technical assistance, sharing information on effective practices, procedures, and policies of States with recent improvements or sustained high performance, and providing advice on business process reengineering. FNS March 2016 Memo 6–7.

2 FNS may also subject States to escalation procedures if their State Timeliness Rate (calculated from the entire universe of SNAP cases rather than a QC sample) falls below 90%. Id. at 3.

B. The District's Administration of SNAP

The District of Columbia, which is treated as a State under the SNAP Act, see 7 U.S.C. § 2012(r), has delegated its administration of SNAP to the Economic Security Administration ("ESA") of its Department of Health Services ("DHS" or the "Department"). Gov. Stmt. Mat. Facts ¶ 2. Each month, the District distributes SNAP benefits to about 65,000 households. FNS Monthly Report from DC for January 2020 ("Jan. 2020 FNS Report") 2, Gov. MSJ, Exh. U, ECF No. 138-23. ESA operates five neighborhood service centers throughout the District, as well as a centralized Call Center. Deposition of Garlinda Bryant-Rollins ("Bryant-Rollins Dep.") 18:7–14. The vast majority (approximately 85%) of SNAP applications are processed in person at a service center by what the District refers to as the "lobby unit," with the remaining 15% handled by the "non-lobby unit" over the phone. Id. at 21:14–21, 129:20–130:14; Dep. of Laura Zeilinger ("Zeilinger Dep.") 49:14–50:8, 56:4–17, 57:3–12; Dep. of Anthea Seymour ("Seymour Dep.") 129:30–131:8.

The District tracks and reports statistics on its timeliness in processing SNAP applications in several ways. A team within the Department's Division of Data Analytics Research and Evaluation ("DARE") pulls raw data from the agency's internal database to create statistical reports. See Dep. of Won-Ok Kim ("Kim Dep.") 19:16–18, 43:3–9, 54:4–13, 269:1–7; Dep. of Yogi Tripurneni 8:18–22, 12:14–17, 14:2–4, 15:6–11. These reports include the three FNS metrics described above—the QC APT Rate, the State Timeliness Rate, and the FNS-366B. See, e.g., Jan. 2020 FNS Report 2 (reporting the QC APT rate and the State Timeliness Rate); District's FNS-366B for Fiscal Year 2020, Quarter 1 ("FNS-366B for FY 2020 Q1"), Pl. MSJ Reply, App. 2, Exh. 8, ECF No. 146-35. As will be explained further, the District has also created a new metric called the "Adjusted State APT Rate," which it has been reporting monthly to the Court and FNS since the issuance of the Court's partial preliminary injunction in May 2018. See, e.g., SNAP Application Processing Timeliness Data for Jan. 2020, ECF No. 134 (court report); Jan. 2020 FNS Report 2 (FNS report).

C. Procedural History

In October 2016, the District rolled out a new operating system for SNAP administration known as the District of Columbia Access System ("DCAS"). Kim Dep. 52:9–11; Gov. Opp. to Prelim. Injunction, Exh. E. Difficulties associated with implementing the new system caused problems in the District's administration of SNAP, including increased delays in processing benefits applications. See, e.g., Pl. Prelim. Injunction Reply, Attach. 1, Exh. A at 1 (noting "widespread technological problems we are seeing in activating Food Stamps benefits following the transition to DCAS"). Indeed, the District's rolling-average QC APT rate for October 2016 through March 2017 was estimated at 88.45%.³ Based on that metric, FNS informed the District in October 2017 that it was required to develop a CAP because its QC APT rate was below 90%. Letter from Eric Ratchford to Anthea Seymour ("FNS Oct. 2017 Letter") 1 (Oct. 23, 2017), Gov. MSJ, Exh. R.

3 This figure is based on a calculation of the upper bound of a 95% confidence interval (an estimated range of values of which FNS is 95% confident includes the true value) surrounding the FNS APT rate point estimate. The upper bound of the interval, rather than the point estimate, is used to identify and monitor poor timeliness. FNS June 2017 Memo 2.

Suffering the adverse effects of DHS's SNAP administration problems, a group of D.C. residents and non-profit organization Bread for the City ("Bread") brought suit in August 2017 against Laura Zeilinger in her official capacity as Director of DHS.⁴ See Am. Compl. ¶ 15. They sought injunctive and declaratory relief under 42 U.S.C. § 1983 on the grounds that the District was violating their federal rights by (1) failing to process applications for benefits within the SNAP Act's mandatory timeframes; (2) failing to issue recertification notices on a timely basis; and (3) failing to provide notice to households of application processing delays and their right to a hearing. Am. Compl. ¶¶ 171–173. Plaintiffs filed a motion for class certification contemporaneous with the complaint and moved for a preliminary injunction shortly thereafter. The Court granted Plaintiffs’ motion for class certification on March 28, 2018, see Garnett I, 301 F. Supp. 3d at 211–12 (certifying three classes), and partially granted their motion for a preliminary injunction on May 31, 2018, see Garnett II, 313 F. Supp. 3d at 150–51. The District followed with a motion to dismiss, which the Court granted on August 23, 2018 only as to Plaintiffs’ third claim concerning notices of delays and hearings and denied as to the remaining two claims.⁵ The Court also entered the preliminary injunction that same day. See Preliminary Injunction Order (Aug. 23, 2018). The injunction required the District to "expeditiously enact any changes to policy statements, procedure manuals, and internal directives" and to "conduct such training" as is necessary to ensure the District's full compliance (95% timeliness) with the SNAP Act's deadlines for processing of recertification applications within 18 months. Id. ¶¶ 2–3, 5. It also required the District to file monthly reports containing information on its timeliness in processing regular initial, expedited initial, and recertification applications as well as the status of its CAPs with FNS. Id. ¶¶ 6–7.

4 While Director Zeilinger is currently the sole defendant in this case, consistent with its prior opinions the Court will refer to the defendant here as "the District."

5 Plaintiffs ask the Court to consider the two remaining claims together. Tr. of Hearing on Summ. J. Mots. 4:22–5:8 (July 28, 2020).

Following extensive discovery, the parties have cross moved for summary judgment. The District also moves to dismiss Bread for the City as a plaintiff for lack of standing.

II. Legal Standards

Article III standing goes to a court's subject-matter jurisdiction and is properly raised under Federal Rule of Civil Procedure 12(b)(1). See Harbury v. Hayden, 444 F. Supp. 2d 19, 26 (D.D.C. 2006), aff'd, 522 F.3d 413 (D.C. Cir. 2008). "The objection that a federal court lacks subject-matter jurisdiction may be raised by a party, or by a court on its own initiative, at any stage in the litigation ...." Arbaugh v. Y & H Corp., 546 U.S. 500, 506, 126 S.Ct. 1235, 163 L.Ed.2d 1097 (2006) (citation omitted); see Fed. R. Civ. P. 12(h)(3). Because the case is now at summary judgment, "the plaintiff can no longer rest on ... mere allegations, but must set forth by affidavit or other evidence specific facts [that establish its standing]." Lujan v. Defs. of Wildlife, 504 U.S. 555, 560–61, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992).

As to liability, the Court may grant summary judgment if "the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a). Where, as here, "both parties file cross-motions for summary judgment, each must carry its own burden under the applicable legal standard." Fay v. Perles, 59 F. Supp. 3d 128, 132 (D.D.C. 2014) (Cooper, J.) (quoting Ehrman v. United States, 429 F. Supp. 2d 61, 67 (D.D.C. 2006) ). In other words, each party "bears the ‘initial responsibility of informing the district court of the basis for its motion, and identifying those portions of the pleadings, depositions[,] ... admissions on files, [and] affidavits ... which it believes demonstrate the absence of a genuine issue of material fact.’ " Hodes v. Dep't of Treasury, 967 F. Supp. 2d 369, 372 (D.D.C. 2013) (quoting Celotex Corp. v. Catrett, 477 U.S. 317, 323, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986) ). To defeat a motion for summary judgment, the non-moving party must provide "specific facts showing that there is a genuine issue for trial." Celotex Corp., 477 U.S. at 324, 106 S.Ct. 2548. "[A] material fact is ‘genuine’ ... if the evidence is such that a reasonable jury could return a verdict for the nonmoving party" on a particular claim. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986).

As a remedy, Plaintiffs seek a permanent injunction compelling the District's absolute compliance with the SNAP Act's timeliness requirements. "According to well-established principles of equity, a plaintiff seeking a permanent injunction must satisfy a four-factor test before a court may grant such relief." eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006). A plaintiff must demonstrate: (1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for that injury; (3) that, considering the balance of hardships between the plaintiff and defendant, a remedy in equity is warranted; and (4) that the public interest would not be disserved by a permanent injunction. Id.

III. Analysis

A. Motion to Dismiss Claims by Bread for the City

To sustain subject-matter jurisdiction, "[a]t least one plaintiff must have standing to seek each form of relief requested in the complaint." Town of Chester, N.Y. v. Laroe Estates, Inc., ––– U.S. ––––, 137 S. Ct. 1645, 1651, 198 L.Ed.2d 64 (2017). At the threshold, Plaintiffs contend that the Court need not decide whether Bread has standing for the suit to proceed because it is undisputed that the individual named Plaintiffs have standing. Pl. Opp. 9–10; see, e.g., Bowsher v. Synar, 478 U.S. 714, 721, 106 S.Ct. 3181, 92 L.Ed.2d 583 (1986) (declining to address the standing of other plaintiffs where at least one plaintiff had standing). The Court previously declined to address Bread's standing for that reason at the motion-to-dismiss stage. Garnett III, 323 F. Supp. 3d at 68 n.7 (citing Ark Initiative v. Tidwell, 64 F. Supp. 3d 81, 92 (D.D.C. 2014) ). The Government responds that the Court must now decide Bread's standing because it seeks relief distinct from the individual named Plaintiffs.

The Government is correct that the dispositive question is whether Bread is "pursu[ing] relief that is broader than or different" than that sought by the individual named Plaintiffs. Little Sisters of the Poor Saints Peter & Paul Home v. Pennsylvania, No. 19-431, ––– U.S. ––––, –––– n.6, 140 S.Ct. 2367, 207 L.Ed.2d 819, 2020 WL 3808424, at *8 n.6 (U.S. July 8, 2020). Bread's request for declaratory and injunctive relief is identical to that sought by the individual named Plaintiffs. Am. Compl., Request for Relief ¶¶ C–D; see also O.A. v. Trump, 404 F. Supp. 3d 109, 138 (D.D.C. 2019) ("Where multiple plaintiffs assert claims seeking precisely the same declaratory or injunctive relief, ... and where the court has subject matter jurisdiction to consider the claims of at least one of those plaintiffs, the court need not address its jurisdiction to consider the claims of the remaining plaintiffs."). As the District points out, however, Bread also seeks attorney's fees and costs under 42 U.S.C. § 1988(b). Am. Compl., Request for Relief ¶ E. Other courts have held that each plaintiff must have standing in order to recover attorney's fees. See, e.g., Shaw v. Hunt, 154 F.3d 161, 166 (4th Cir. 1998) (noting that "a plaintiff without standing will not be able to recover fees" under § 1988 "because the possession of Article III standing is interwoven into the very concept of plaintiff status"). Therefore, out of an abundance of caution and because the parties have briefed the issue extensively, the Court will assess Bread's standing. See Women's Med. Ctr. of Providence, Inc. v. Roberts, 512 F. Supp. 316, 320 (D.R.I. 1981) ("[T]he potential availability of attorney's fees turns the standing question raised by defendants’ motions to dismiss into a situation in which a great deal ‘may be gained or lost (depending upon) the presence or absence of’ multiple plaintiffs." (quoting Doe v. Bolton, 410 U.S. 179, 189, 93 S.Ct. 739, 35 L.Ed.2d 201 (1973) ))

The requirements of Article III standing are: (1) an "actual or threatened injury in fact" (2) "that is fairly traceable to the alleged illegal action" and (3) is "likely to be redressed by a favorable court decision." Spann v. Colonial Vill., Inc., 899 F.2d 24, 27 (D.C. Cir. 1990). Furthermore, because Bread seeks injunctive and declaratory relief, it must show that it "is suffering an ongoing injury or faces an immediate threat of injury." Dearth v. Holder, 641 F.3d 499, 501 (D.C. Cir. 2011). The plaintiff, as the party invoking federal jurisdiction, bears the burden of establishing standing " ‘throughout all stages of litigation.’ " Va. House of Delegates v. Bethune-Hill, ––– U.S. ––––, 139 S. Ct. 1945, 1950–51, 204 L.Ed.2d 305 (2019) (quoting Hollingsworth v. Perry, 570 U.S. 693, 705, 133 S.Ct. 2652, 186 L.Ed.2d 768 (2013) ) (emphasis added). At summary judgment, the plaintiff's burden is to "set forth by affidavit or other evidence specific facts" that support its standing. Lujan, 504 U.S. at 560–61, 112 S.Ct. 2130. Bread has met that burden.

1. Injury-in-Fact

An organization such as Bread may establish standing by showing that the challenged practices have caused a "concrete and demonstratable injury to the organization's activities." Havens Realty Corp. v. Coleman, 455 U.S. 363, 379, 102 S.Ct. 1114, 71 L.Ed.2d 214 (1982) (holding that an organization had standing to challenge racial steering practices that perceptibly impaired the organization's ability to provide housing counseling and referral services and drained the organization's resources). The D.C. Circuit "has applied Havens ... to justify organizational standing in a wide range of circumstances," so long as two conditions are met. Abigail All. for Better Access to Developmental Drugs v. Eschenbach, 469 F.3d 129, 133 (D.C. Cir. 2006) ; see Nat'l Fair Hous. All. v. Carson, 330 F. Supp. 3d 14, 41 (D.D.C. 2018) (recognizing the Circuit's "two-prong inquiry"). First , the plaintiff must show that the challenged governmental action or inaction " ‘perceptibly impaired’ a non-abstract interest" of the organization. Nat'l Ass'n of Home Builders v. EPA, 667 F.3d 6, 12 (D.C. Cir. 2011) (quoting Havens, 455 U.S. at 379, 102 S.Ct. 1114 ). "Second , the plaintiff must show that it ‘used its resources to counteract that harm.’ " Elec. Privacy Info. Ctr. v. Presidential Advisory Comm'n on Election Integrity, 878 F.3d 371, 378 (D.C. Cir. 2017) (quoting People for the Ethical Treatment of Animals v. USDA ("PETA"), 797 F.3d 1087, 1094 (D.C. Cir. 2015) ).

Bread for the City's mission is to provide basic staples (food, clothing, and medical care) and legal and social services to low-income D.C. residents to reduce the burden of poverty. Bread for the City, Our Mission (last visited July 20, 2020), https://breadforthecity.org/about/#mission-vision. Here, Bread has established that the District's SNAP administration deficiencies "[i]nhibit[ ] ... [its] daily operations" by increasing its clients’ demand for food, SNAP-related legal services and counseling, and assistance with management of Social Security Disability Insurance ("SSDI") and Social Security Insurance ("SSI") benefits. Action All. of Senior Citizens of Greater Philadelphia v. Heckler, 789 F.2d 931, 938 (D.C. Cir. 1986). Specifically, the record contains evidence that beginning in October 2016 (when the District's administration of SNAP began experiencing widespread problems due to the new computer system), Bread's Legal Clinic saw more clients coming in with claims concerning improper food stamp terminations and application delays, Bread's Food Program experienced an increased demand for food packages, and Bread's Representative Payee Program's clients made more requests for additional funds for food from their own SSI and SSDI benefits. Pl. 3d Supp. Interrog. Resp. ¶ 10, Def. MTD, Exh. A; Dep. of George Jones ("G. Jones Dep.") 119:9–120:20; 121:3–122:7, Pl. MTD Opp., Exhs. A, E & F. The impairment of these non-abstract interests satisfies the first prong of the Havens analysis. Nat'l Ass'n of Home Builders, 667 F.3d at 12.

As to Havens’ second prong, Bread has shown that it has taken concrete steps in response to the District's SNAP administration deficiencies. For one, Bread dedicated more staff time and resources to meet the increased demand for the Legal Clinic's assistance with SNAP-related matters, including hiring a public benefits paralegal (who is paid $40,000 a year) and a law-school volunteer (who had to be trained by paid employees) and diverting time of existing staff members to SNAP-related matters. Pl. 3d Supp. Interrog. Resp. 25–26, 28. Bread also invested significant staff time and resources into developing and implementing an internal referral system to ensure that clients that initially engage with other SNAP programs receive help from the Legal Clinic with SNAP-related matters. Id. at 28–29, 31. The District counters that "an organization's diversion of resources to litigation or to investigation in anticipation of litigation" cannot create standing. Gov. MTD Reply 4 (quoting PETA, 797 F.3d at 1093 ); see also Nat'l Taxpayers Union, Inc. v. United States, 68 F.3d 1428, 1434 (D.C. Cir. 1995) ("An organization cannot ... manufacture the injury necessary to maintain a suit from its expenditure of resources on that very suit."). True. But that principle does not apply where the "purportedly illegal action increases the resources the group must devote to programs independent of its suit challenging the action. " Spann, 899 F.2d at 27 (emphasis added). Here, the record establishes that Bread undertook these additional expenditures on SNAP-related legal services and counseling "in response to, and to counteract, the effects of the defendants’ alleged [statutory violations] rather than in anticipation of litigation." Equal Rights Ctr. v. Post Properties, Inc., 633 F.3d 1136, 1140 (D.C. Cir. 2011).

Nor is there credence to the District's suggestion that Bread's expenditures somehow constitute self-inflicted injury under Fair Emp. Council of Greater Washington, Inc. v. BMC Mktg. Corp. ("BMC"), 28 F.3d 1268, 1276 (D.C. Cir. 1994). Gov. MTD Reply 4. In BMC, the D.C. Circuit held that the organizational plaintiff's "diversion of resources to testing [BMC's housing discrimination]" was "self-inflicted; it result[ed] not from any actions taken by BMC, but rather from the [plaintiff's] own budgetary choices." BMC, 28 F.3d at 1276. But, the Circuit's holding in that regard "assume[d] that BMC's actions did not have any other effect on the [plaintiff's] programs independent of its efforts to increase legal pressure on possible open housing violators" such that the organization's "programs would have been totally unaffected if it had simply refrained from making the re-allocation." Id. at 1276–77. Bread's injuries, by contrast, are more analogous to the "increase [in] the number of people in need of counseling" directly caused by BMC's discrimination, which the Circuit found to be injury-in-fact. Id. at 1276.

Bread's expenditures related to helping their clients deal with missing or late SNAP payments are therefore cognizable harm under Havens. See, e.g., PETA, 797 F.3d at 1096 (recognizing organizational standing where plaintiff spent "more than $10,000 on staff attorney time" "as a direct result of the USDA's failure to regulate birds").

Even if the record establishes that Bread suffered cognizable harm under Havens, the District contends that Bread cannot rely upon "stale evidence" of past harm in order to establish ongoing or future injury. Gov. MTD Reply 1–3. It invokes California Cattlemen's Association v. U.S. Fish & Wildlife Service, 369 F. Supp. 3d 141 (D.D.C. 2019), where the court concluded that the plaintiffs had not shown ongoing or imminent injury from the agency processes that they challenged because the processes were already complete, and there was "only conjecture and hypothesis" that the agency would reinitiate the challenged processes in the future. Id. at 147. Here, however, at least some of Bread's increased expenditures on legal resources and counseling related to SNAP administration are ongoing, see, e.g., Pl. 3d Supp. Interrog. Resp. 27 (averring that public benefits paralegal is paid $40,000 a year and "continues to work on food stamps matters"); id. at 29 (averring that "Legal Clinic staff estimate that they spend approximately two hours per month re-training staff, sending email reminders, and maintaining the online referral system"), and there is no indication in the record that Bread will cease these expenditures in the future. Bread has therefore shown that it suffers "ongoing injury" due to the District's alleged SNAP administration deficiencies. Dearth, 641 F.3d at 501.

2. Causation and Redressability

"An organizational plaintiff must, in addition to ... establish[ing] cognizable harm under Havens, satisfy the requirements for ... the second and third elements of Article III standing—causation and redressability." Nat'l Fair Hous. All., 330 F. Supp. 3d at 42. "[C]ausation and redressability ‘are closely related’ like ‘two sides of a ... coin.’ " West v. Lynch, 845 F.3d 1228, 1235 (D.C. Cir. 2017) (quoting Dynalantic Corp. v. Dep't of Defense, 115 F.3d 1012, 1017 (D.C. Cir. 1997) ). Although the two concepts are distinct—"causation focuses on the ‘connection between the assertedly unlawful conduct and the alleged injury’ whereas redressability focuses on the ‘connection between the alleged injury and the judicial relief requested,’ " id. at 1235–36 (quoting Allen v. Wright, 468 U.S. 737, 753 n.19, 104 S.Ct. 3315, 82 L.Ed.2d 556 (1984) )—the Court will address them together here because the judicial relief requested is an injunction against the District's allegedly unlawful conduct. The causation standard for Article III standing is not particularly demanding. It "does not require that the defendant[’s] [conduct] be the most immediate cause, or even a proximate cause, of the plaintiffs’ injuries; it requires only that those injuries be ‘fairly traceable’ to the defendant." Attias v. Carefirst, Inc., 865 F.3d 620, 629 (D.C. Cir. 2017). The Court assumes for the purposes of the standing analysis that Bread will secure the relief that it seeks. See Fla. Audubon Soc. v. Bentsen, 94 F.3d 658, 664 n.1 (D.C. Cir. 1996).

The District contends that Bread's injuries are not fairly traceable to the District's deficiencies in administering SNAP because numerous other factors—such as a change in the law affecting cash benefits, the rising cost of food, Bread's creation of an internal referral system among its programs, and the need to serve furloughed government workers and homeless individuals—also have affected D.C. residents’ demand for Bread's services. Gov. MTD 12, 16–17, 19. The fact that it may be difficult to quantify exactly how much of Bread's increased expenditures is attributable to the District's SNAP administration deficiencies as opposed to these other factors does not defeat causation. See generally Competitive Enter. Inst. v. Nat'l Highway Traffic Safety Admin., 901 F.2d 107, 113 (D.C. Cir. 1990) ("For standing purposes, petitioners need not prove a cause-and-effect relationship with absolute certainty; substantial likelihood of the alleged causality meets the test."). For purposes of Article III standing, "it may be enough that the defendant's conduct is one among multiple causes." Orangeburg, S.C. v. Fed. Energy Regulatory Comm'n, 862 F.3d 1071, 1080 (D.C. Cir. 2017) (quoting 13A Charles A. Wright, Arthur R. Miller & Edward H. Cooper, Fed. Practice & Procedure § 3521.5 (3d ed. 2008)). And here, "[c]ommon sense and basic economics tell us that" the District's failure to get SNAP benefits to Bread's clients on time or at all will likely cause them to seek more food, legal services, and counseling from Bread, which in turn will affect Bread's resources. Carpenters Indus. Council v. Zinke, 854 F.3d 1, 6 (D.C. Cir. 2017).

The Court therefore concludes that Bread has Article III standing to challenge deficiencies in the District's administration of SNAP and will proceed to analyzing the merits of Plaintiffs’ claims.

B. Cross-Motions for Summary Judgment

As the Court held at the motion to dismiss stage, § 1983 provides a cause of action for Plaintiffs to enforce the District's compliance with the SNAP Act's timeliness requirements. Garnett III, 323 F. Supp. 3d at 71–75. At that stage, the Court declined to decide whether the "policy or practice" standard for municipal liability set forth in Monell, 436 U.S. 658, 98 S.Ct. 2018, applied to this suit and invited the parties to address the issue in future briefing. Garnett III, 323 F. Supp. 3d at 75. The parties now agree that Monell’s "policy or practice" standard applies to the Plaintiffs’ claims against Director Zeilinger. Gov. Reply 2–4; Tr. of Hearing on Mots. for Summ. J. ("MSJ Tr.") 7:2–11. Although the SNAP Act defines the District as a "State," 7 U.S.C. § 2012(r), the District of Columbia is considered a municipality for purposes of § 1983, see Act Now to Stop War & End Racism Coal. & Muslim Am. Soc'y Freedom Found. v. District of Columbia ("Act Now"), 846 F.3d 391, 413 (D.C. Cir. 2017) ("Both [S]tates and cities can be sued under section 1983, and for that purpose the District of Columbia is treated as a city." (internal citations omitted)); Salazar v. District of Columbia, 954 F. Supp. 278, 324 (D.D.C. 1996) (applying Monell standard to the District's compliance with Medicaid processing requirements even though the statute defines D.C. as a "State"). And, Plaintiffs’ suit against Ms. Zeilinger in her official capacity as a municipal officer "is equivalent to a suit against the municipality itself." Atchinson v. District of Columbia, 73 F.3d 418, 424 (D.C. Cir. 1996) ; see Kentucky v. Graham, 473 U.S. 159, 165, 105 S.Ct. 3099, 87 L.Ed.2d 114 (1985) (holding that suits brought against individuals in their official capacities "generally represent only another way of pleading an action against an entity of which an officer is an agent" (quoting Monell, 436 U.S. at 690 n.55, 98 S.Ct. 2018 )).

Therefore, even "in an official-capacity suit" like this one, Plaintiffs must show that "the entity's ‘policy or custom’ ... played a part in the violation of federal law." Graham, 473 U.S. at 166, 105 S.Ct. 3099 (quoting Monell, 436 U.S. at 694, 98 S.Ct. 2018 ); see also Los Angeles Cty., Cal. v. Humphries, 562 U.S. 29, 31, 131 S.Ct. 447, 178 L.Ed.2d 460 (2010) (holding that "the ‘policy or custom’ requirement also applies when plaintiffs seek prospective relief [under § 1983 ], such as an injunction or a declaratory judgment"). Courts conduct a two-step inquiry to assess municipal liability under § 1983. First, they must determine whether the plaintiffs have established a predicate federal constitutional or statutory violation. If so, they must decide whether a custom or policy of the municipality caused the violation. Baker v. District Columbia, 326 F.3d 1302, 1306–07 (D.C. Cir. 2003).

Plaintiffs’ theory is that the District's protracted failure to correct its deficiencies in administering SNAP amounts to "deliberate indifference" to the risk of violating Plaintiffs’ federal rights. See, e.g., MSJ Tr. 7:17–23 ("[t]he plaintiffs have established that the defendant had a custom and practice of being deliberately indifferent to ... failure[s] to timely process SNAP ... applications"); Baker, 326 F.3d at 1306–07 (noting that municipal liability can lie where the municipality fails "to respond to a need (for example, training of employees) in such a manner as to show ‘deliberate indifference’ to the risk that not addressing the need will result in ... violations [of federal law]"); Bd. of Cty. Comm'rs of Bryan Cty., Okl. v. Brown, 520 U.S. 397, 407, 117 S.Ct. 1382, 137 L.Ed.2d 626 (1997) ("[A] plaintiff seeking to establish municipal liability on the theory that a facially lawful municipal action has led an employee to violate a plaintiff's rights must demonstrate that the municipal action was taken with ‘deliberate indifference’ as to its known or obvious consequences."). According to Plaintiffs, the District was put on notice of its deficiencies in timely processing SNAP applications when FNS sent the District a letter in October 2017 subjecting it to a corrective action plan due to its poor APT rate. MSJ Tr. 8:8–15; FNS Oct. 2017 Letter 1. Plaintiffs contend that the lack of improvement in the District's timeliness figures, after being made aware of the problem, suffices to establish the District's deliberate indifference. Contrary to Plaintiffs’ assertion, however, the record shows that the District has undertaken significant efforts to improve SNAP application processing timeliness and that, at least by some timeliness metrics, those efforts have been successful.

1. Predicate Violations of the SNAP Act's Timeliness Requirements

As to the first step of the Monell analysis, the Court has previously held that the SNAP Act requires absolute compliance with its timeliness requirements. See Garnett I, 301 F. Supp. 3d at 207–08 ; Garnett II, 313 F. Supp. 3d at 155. The Court is also cognizant, however, that "with respect to the statutory timelines, ‘absolutely perfect compliance is unattainable.’ " Garnett II, 313 F. Supp. 3d at 159 (quoting Withrow v. Concannon, 942 F.2d 1385, 1388 (9th Cir. 1991) ). In other words, "[t]here is ... doubtless a point at which any failure of total compliance is truly de minimis , where the [S]tate has come to comply ‘as strictly as is humanly possible.’ " Id. (quoting Withrow, 942 F.2d at 1388 ). The Court therefore defined "full compliance" in the preliminary injunction order as 95% application processing timeliness, see Preliminary Injunction Order 1, consistent with other courts that have issued injunctive relief in SNAP Act compliance challenges, see, e.g., Preliminary Injunction Order 4 (defining "full compliance" to "mean Defendant the federal processing time-frame in all cases except those individual and isolated instances of delay that occur inevitably in an agency such as Defendant's," which the Court set at 95%), Booth v. McManaman, No. 10-CV-680 (RAR) (D. Haw. Jan. 23, 2012); Preliminary Injunction Order 5, Briggs v. Bremby, No. 3:12-CV-324 (VLB), 2013 WL 1987237 (D. Conn. May 13, 2013) (defining "full compliance" as "compliance in all cases except those individual and isolated instances of delay that occur inevitably in an agency such as Defendant's," which the Court set at 97%).

In assessing whether the District violated the SNAP Act, the Court will consider its performance from June 2016 through present, as the three certified classes include SNAP applicants since June 1, 2016, see Garnett I, 301 F. Supp. 3d at 212–13.⁶ The undisputed record for that time period shows that the District has not achieved 100% timeliness in processing initial regular, initial expedited, or recertification SNAP applications. The parties disagree, however, as to the extent of the District's noncompliance. As an initial matter, they dispute which metrics the Court should use to measure the District's timeliness; the primary source of disagreement is whether the timeliness measures should exclude overdue applications whose processing has been delayed beyond the applicable deadline due to the fault of the applicant.

6 The Court lacks data since the beginning of the Covid-19 pandemic when, with the parties’ consent, the Court suspended the District's reporting requirements under the preliminary injunction. See Min. Order (Mar. 23, 2020); Min. Order (April 16, 2020); Min. Order (Aug. 20, 2020).

A bit of background is needed to understand the nature of this dispute. As explained, there are four timeliness metrics in the record: the three FNS metrics—the QC APT Rate, State Timeliness Rate, and FNS-366B—as well as the Adjusted State APT Rate created by the District. Two of these metrics—the State Timeliness Rate and the FNS-366B—include all overdue applications, even where the reason for untimely processing is client delay. See FNS June 2017 Memo 3, 5. The remaining two metrics—the QC APT Rate and the Adjusted State APT Rate—exclude overdue applications that are attributed to client fault. See id. at 2 (explaining that the QC APT Rate excludes cases pended for missing verification); Tripurneni Dep. 73:5–13 (explaining that the Adjusted State APT Rate excludes overdue applications attributed to client delay). Plaintiffs urge the Court to rely only on the State Timeliness Rate and the FNS-366B, while the District contends that the QC APT Rate and Adjusted State APT Rate are more accurate measures of timeliness under the circumstances here.

Under FNS regulations, an overdue application may be attributed to client fault "if the household failed to complete the application process"—i.e. , by not providing required verification or completing the required interview. 7 C.F.R. § 273.2(h)(1)(i). However, the State agency may attribute overdue applications to client fault only after taking certain mitigation actions, such as offering assistance with completing the application and obtaining the required verification. Id. Plaintiffs contend that the Court should not rely on the two timeliness metrics that exclude client-delayed applications (QC APT Rate and Adjusted State APT Rate) because the record does not show whether the District first took these mitigation steps before attributing the delay to each individual applicant.

As to the QC APT Rate, the record belies Plaintiffs’ contention. As previously explained, the QC APT Rate is based on a sample of active cases pulled by an agency's QC team each month.⁷ Under FNS regulations, the District's QC team is required to review individual case files to "accurately determine if an application was properly pended due to a client's delay in providing verification." FNS March 2016 Memo 7; 7 C.F.R. § 273.2(h)(1)(i). The District has presented ample evidence that the agency arrived at its internal QC APT numbers in line with this FNS protocol. As required by FNS regulations, DHS has a dedicated QC team that remains "arms length" from the operations team. Dep. of Ellen Wells ("Wells Dep.") 19:16–20:1, 45:9–22, Gov. MSJ, Exh. E. And, no fewer than four agency officials testified that QC reviewers "pull cases from the SNAP population" using "a sampling methodology approved by FNS" and "review them according to the specifications in FNS Handbook 310."⁸ Id. at 20:7–21, 90:1–13; see also Zeilinger Dep. 97:2–98:3 (testifying that QC reviewers "pull[ ] applications for sample review and report[ ] directly to FNS whether those applications met the standards for timeliness"); Seymour Dep. 121:8–16 (testifying that "QC works directly with FNS and FNS identifies concern criteria that would be included in a sample, ... and then based on that, QC pulls the sample based on that criteria and that sample is reviewed by QC."); Kim Dep. 169:1–9 (testifying that the QC "sample review goes through the entire system, including all the scanned documents"); see generally Sept. 2018 CAP, ECF No. 130-6 at 1 ("[W]e closely monitor the QC APT rate as it uses exactly the same methodology applied to calculate the FNS APT rate ..."). Plaintiffs point to nothing in the record that would rebut this testimony, which is entitled to a presumption of good faith. See SafeCard Servs., Inc. v. SEC, 926 F.2d 1197, 1200 (D.C. Cir. 1991). And while the record is silent on whether FNS independently reviewed the District's most recent internal QC APT Rate figures, see MSJ Tr. 68:11–18; Pl. Stmt. Facts ¶ 54, nothing suggests that FNS rejected or questioned the numbers, which it received monthly.⁹ The Court will therefore consider the District's internal QC APT data in assessing the District's SNAP application processing timeliness.

7 FNS regulations require State agencies to examine a sample of active cases (those households currently receiving SNAP benefits) each month for quality control review. 7 C.F.R. § 275.12(a) ; FNS June 2017 Memo 1. State agencies must follow certain FNS procedures, handbooks, worksheets, and schedules in conducting these QC reviews. 7 C.F.R. § 275.14.

8 FNS Handbook 310 provides comprehensive procedures for States’ conducting QC reviews of their SNAP programs.

9 FNS requires States to submit QC APT data to enable its "reviewers to conduct a thorough and independent assessment of the case results reported to FNS." SNAP Quality Control Policy Memo: 16-02 (Jan. 20, 2016), https://fns-prod.azureedge.net/sites/default/files/snap/QCPolicyMemo16-02.pdf; see also Kim Dep. 35:3–7. It is unclear from the record how often FNS conducts these independent reviews of QC results or publishes State QC APT rates. See MSJ Tr. 22:24–3, 49:24–50:1, 50:15–51:10. The most recent FNS-published QC APT rate data in the record precedes the Court's entry of the preliminary injunction. See Pl. Reply in Support of Prelim. Injunction, Attach. 2, Exh. B, ECF No. 45-29 (QC APT rates for October 2016–March 2017).

The District-created Adjusted State APT Rate is similar to the QC APT Rate except that it is based on the agency's entire universe of SNAP applications rather than a QC sample. The Adjusted State APT Rate is calculated in the same manner as the FNS State Timeliness Rate (the total number of applications approved timely divided by the total number of applications), except that the denominator for the Adjusted State APT Rate excludes overdue applications that are pended for missing verification. See Jan. 2020 FNS Report 2; Tripurneni Dep. 73:5–13; D.C. Application Processing Timeliness Corrective Action Plan ("Sept. 2018 APT CAP") 3 (Sept. 1, 2018), Pl. MSJ, Exh. C; Gov. Resp. to Pl. Stmt. Facts ¶¶ 61–62. FNS has explained that the State Timeliness Rate "will not adjust for properly pended applications" because "this protocol uses a universe of certification data pulled from an eligibility system where all factors needed to determine properly pended [applications] are unlikely to be available." FNS March 2016 Memo 7. Nonetheless, "recognizing the aforementioned limitations, the District [has] attempted to identify properly pended applications using the date when the customer provided the last remaining verification item." Sept. 2018 APT CAP 3. Specifically, the DCAS system deems an overdue application as attributed to client fault if the system shows that the customer has verification pending past the SNAP Act's due date. Id.; Tripurneni Dep. 101:1–103:22.

Plaintiffs contend that the Adjusted State APT Rate is an unreliable measure of timeliness because the DCAS system does not track whether the agency took the mitigation steps required under 7 C.F.R. § 273.2(h)(1)(i) for each excluded application. The District concedes that DCAS does not track this information. See Tripuneni Dep. 77:8–13, 78:6–79:5; Zeilinger Dep. 50:19, 69:3–10; Bryant-Rollins Dep. 70:9–20; Kim Dep. 109:1–10. But, the District counters, there is no need for DCAS to confirm that the agency took the necessary mitigation steps for each individual excluded application because the agency's standard business processes ensure that such steps were taken in the vast majority of cases.

The District has the better argument. As will be explained further, the undisputed record shows that approximately 85% of the District's SNAP applications are processed through a "one-and-done" system. See Zeilinger Dep. 49:2–13; Seymour Dep. 129:20–130:14. Under this system, Social Service Representatives ("SSRs") interacting with customers coming into a service center are trained to address all potential issues with the customer's application on the spot so that the agency can make an eligibility determination before the applicant leaves the service center that day. See Decl. of Laura Zeilinger ("Zeilinger Decl.") ¶¶ 69, 73–74, Gov. Opp. to Prelim. Injunction, Exh. A; Zeilinger Dep. 49:2–21; Bryant-Rollins Dep. 61:3–62:6; Seymour Dep. 130:2–131:8; Dep. of Carole Jones ("Jones Dep.") 96:3–4. That includes performing many of the mitigation steps required by FNS regulations. 7 C.F.R. § 273.2(h)(1)(i). For one, the SSR reviews a manual checklist of required verifications with the customer. If documents are missing, the SSR informs the customer what documents they need and how to obtain them. Zeilinger Dep. 49:2–21; Bryant-Rollins Dep. 61:3–62:6; 103:18–104:5. As one supervisor explained, if a customer has "had a fire or flood in their home" and verification "document[s] are not available," the agency will "offer additional assistance ... with obtaining those documents." Bryant-Rollins Dep. 69:19–70:1. Moreover, DCAS is programmed to automatically ping other databases (e.g. , Social Security) for applicant information such as income and immigration status. Id. 69:11–14; Jones Dep. 104:9–105:19. If verifications are still missing, the customer is sent home with a checklist of outstanding verification requirements, which indicates the types of documents that can satisfy each requirement. Gov. MSJ, Exh. K; Jones Dep. 106:5–107:9. DCAS will also generate a formal notice of missing verification, which is later mailed to the customer. Gov. MSJ, Exh. V at 3–5; Jones Dep. 142:7–22; Kim Dep. 106:2–5. This evidence suffices to establish that the necessary mitigation steps are usually taken before an application is tagged as "pending verification" in DCAS, at least for the vast majority of cases that are processed on a "one-and-done" basis.

Plaintiffs complain that evidence of the District's usual protocol is insufficient to establish that mitigation steps were taken in each individual case. But, as the District points out, Plaintiffs have not pointed to a single case in which the District did not follow its established protocol. To the contrary, the named plaintiffs testified that they had not experienced many issues with the application or verification process. See, e.g., Harris Dep. 35:3–36:6 (explaining that she had not had "difficulty verifying the information in [her] application or reporting any changes"); Garnett Dep. 62:6–11 (explaining that she "normally do[esn't] have a problem" with getting her SNAP benefits). This case can therefore be distinguished from other cases where the record contained evidence that specific plaintiffs’ applications were processed outside of statutory requirements. See, e.g., Robertson v. Jackson, 766 F. Supp. 470, 474 (E.D. Va. 1991), aff'd, 972 F.2d 529 (4th Cir. 1992), as amended (Aug. 12, 1992) (noting that the named plaintiff "had been informed two weeks subsequent to her filing, in fact, that it would be 45–60 days before her application would be processed" and that "other class members have had to wait as long as, and in excess of, 60 days after their initial request for any assistance").

The Court therefore finds the District's internal QC APT Rate and Adjusted State APT Rate data to be competent evidence for assessing the District's compliance with the SNAP Act. Briggs v. Bremby, No. 3:12-CV-324 (VLB), 2012 WL 6026167 (D. Conn. Dec. 4, 2012), aff'd, 792 F.3d 239 (2d Cir. 2015), on which Plaintiffs heavily rely, is not the contrary. There, the court declined to consider data that excluded overdue applications that were attributed to client fault. Id. at *16–17. However, that was not because the court considered the data to be inherently misleading or unreliable. See id. at *16 ("The Court agrees with Defendant that it would inappropriate to find [the agency] in violation of Sections 2020(e)(3) and (9) where the delay was truly caused by the applicant."). Rather, the court found the specific dataset at issue to be insufficient because the agency "fail[ed] to provide any evidence indicating what percentage of the delay is attributable to applicants under the rubric of FSA regulations." Id. Here, by contrast, the District has presented sufficient evidence from which the Court can roughly estimate the percentage of overdue applications that is properly attributable to client delay. These evidentiary issues resolved, the Court will proceed to evaluating the District's timeliness in processing initial and recertification applications for benefits.

a. Initial Application Processing Timeliness

Following FNS's lead, see, e.g., FNS Oct. 2017 Letter 1 (relying upon the District's QC APT data to impose a CAP), the Court will focus on the QC APT rate to measure the District's timeliness in processing initial SNAP applications. See generally Tripoli Rocketry Ass'n, Inc. v. Bureau of Alcohol, Tobacco, Firearms, & Explosives, 437 F.3d 75, 77 (D.C. Cir. 2006) ("This court routinely defers to administrative agencies on matters relating to their areas of technical expertise."). Because the QC APT rate is based on an individualized review of case files conducted pursuant to a uniform protocol, it is most likely to accurately capture an agency's true performance in timely processing applications. See FNS March 2016 Memo 7 (noting that "a more thoroughly determined APT rate results from the QC process"). For October 2016 through March 2017, the District's six-month rolling average QC APT rate was 88.45%. Gov. MSJ, Exh. R at 1. The District's most recent internal data shows a rolling average QC APT rate of approximately 95.05% for March through September 2019.¹⁰ Gov. MSJ, Exh. U at 2. That change reflects a significant improvement in the District's initial SNAP application processing timeliness. Indeed, FNS considers 95% QC APT or higher to be acceptable performance. FNS June 2017 Memo 2.¹¹ The rolling average Adjusted State APT Rate for March through September 2019 was 95.65%. Jan. 2020 FNS Report 2.¹² Although the Adjusted State APT Rate is not based on an individualized review of case files in a sample, its logic mimics that of the QC APT Rate. The fact that the District's Adjusted State APT Rate is very close to the QC APT Rate for the same time period therefore corroborates that the District is performing at around 95% timeliness in initial application processing.

10 The Court approximated a six-month rolling average by taking the average of the monthly QC APT rates reported for March through September 2019. Jan. 2020 FNS Report 2. March through September 2019 is the most recent six-month period for which data on all relevant metrics is available.

11 Plaintiffs point out that FNS still has not lifted the District's CAP, which it is supposed to do if a State's APT rates reach 95%. They suggest that FNS has not done so because it finds the District's internal APT rates to be unreliable. However, there is nothing in the record that supports this explanation or provides any other explanation. The Court therefore declines to speculate as to why FNS has not yet lifted the CAP. As mentioned, FNS's most recent correspondence with the District indicates its satisfaction that the District's "timeliness rate has improved and is stable." FNS Mar. 2020 Letter 1.

12 Again, the Court approximated a six-month rolling average by taking the average of the monthly Adjusted State APT rates reported for March through September 2019. Jan. 2020 FNS Report 2.

The metrics that include pended applications for client delay reflect less improvement. The District's rolling average State Timeliness Rate for March through September 2019 was 88.07%, compared to a rolling average of 87.1% for April 2017 through September 2017. Sept. 2018 CAP 1. The most recent FNS-366B before the Court when it entered the preliminary injunction (for the first quarter of fiscal year 2018) showed that the District had approved 94.88% of initial applications on time. Garnett II, 313 F. Supp. 3d at 156–57 (citing Gov. Surreply, Exh. B).¹³ According to the most recent FNS-366B in the record now (for the first quarter of fiscal year 2020), the District approved 87.18% of initial applications on time. Pl. Reply, App. 2 & Exh. 8.¹⁴ Notably, only 3% of the overdue initial applications were pending for more than 30 days, id., compared to 7% in Fiscal Year 2018 Quarter 1.

13 This figure includes both regular and expedited applications. To calculate this rate from the FNS-366B report, the Court first determined the total number of applications approved overdue by adding together the applications approved overdue 1–30 days (Column D in the form), approved overdue 31–60 days (Column E), approved overdue 61–90 days (Column F) and approved overdue 91+ days (Column G) for each category of applications ("total" applications, which encompasses all types; initial; recertification; or expedited). The Court then divided this sum by the total number of approved applications (Column A) for each category of applications and multiplied the result by 100 to arrive at a percent. Garnett II, 313 F. Supp. 3d at 157 n.6.

14 The District's timeliness for subsequent fiscal quarters has varied from 85.57% to 89.97%. See Pl. MSJ Reply, App. 2 & Exhs. 1–7.

As the Court has explained, however, these metrics include applications pended due to missing verification, and the Court finds its reasonable to assume that many, if not most, of the included applications were overdue because the client did not submit the required verification. See Briggs, 2012 WL 6026167, at *16 ("[I]t would inappropriate to find [the agency] in violation of Sections 2020(e)(3) and (9) where the delay was truly caused by the applicant."). Making that assumption, the District's true initial application processing rate is very likely in the ballpark of 90% and 95%.¹⁵ Although that is certainly close to de minimis noncompliance, a reasonable juror might conclude otherwise. Plaintiffs have therefore created a genuine issue of material fact as to the predicate SNAP Act violations for initial applications.

15 At the preliminary injunction stage, the Court relied on both the QC APT rate and the raw FNS-366B data to assess the District's initial application processing timeliness. See Garnett II, 313 F. Supp. 3d at 156. At that stage, however, neither party raised the issue of whether the proper metrics should include applications pended due to client delay. Based on the evidence in the present record, the Court finds that the raw FNS-366B data is less likely than the QC APT Rate data to reflect the District's true application processing timeliness because it includes all overdue applications, even where the reasons for the delay in processing is missing verification.

b. Recertification Application Processing Timeliness

The record also shows that the District has made progress with recertification application processing timelines. As the Court has previously noted, the QC APT rate only captures initial application processing timeliness; it does not capture any data pertaining to recertification applications. Garnett II, 313 F. Supp. 3d at 158. When the Court entered the preliminary injunction, it relied primarily on the FNS-366B to measure recertification application processing timeliness. Based on the District's FNS-366B for Fiscal Year 2017 Quarter 1, the Court determined that District was only processing 59.28% of recertification applications on time. Garnett II, 313 F. Supp. 3d at 158 (citing Gov. MTD Surreply, Exh. B). The District's most recent FNS-366B (for Fiscal Year 2020 Quarter 1) shows that it has increased its recertification application timeliness to 84.20%. Pl. MSJ Reply, App. 2 & Exh. 8. This apples-to-apples comparison shows a dramatic improvement in the District's timely processing of recertification applications. Moreover, the most recent FNS-366B shows that only a single overdue application was pending for more than 30 days, see id., which is a marked improvement from Fiscal Year 2018 Quarter 1, where 4% of overdue recertification applications were pending for that long, see Gov. MTD Surreply, Exh. B.

The Court now has an additional data source to measure the District's recertification timeliness. As mentioned, the preliminary injunction ordered the District to provide monthly reports with data on processing of initial, recertification, and expedited applications. The data from those monthly reports show a rolling average recertification timeliness rate of 95.57% for September 2019 through February 2020. Pl. Reply, App. 1, tbl. 3, ECF No. 146-26. Again, Plaintiffs complain that the timeliness rates calculated in these reports are inaccurate because they improperly exclude applications due to client fault. But even adding all of the excluded recertification applications back in, the District's rolling average recertification processing timeliness rate from September 2019 through February 2020 was 90.93%. Id. Considering all of the relevant evidence, the District's true recertification application timeliness rate is likely in the ballpark of high 80s to low 90s. Based on the record as a whole, then, a reasonable juror could conclude that the District has not been fully compliant with the SNAP Act's deadlines for processing both initial and recertification applications.

2. Whether a Policy or Practice of the District Caused the SNAP Act Violations

Assuming that Plaintiffs have established a genuine issue of material fact as to the District's predicate violations of the SNAP Act, the more difficult question is whether "any District custom or policy was the moving force" behind the violations. Act Now, 846 F.3d at 413 (internal quotation marks omitted). As explained, Plaintiffs’ theory is one of deliberate indifference. To prevail on such a theory, Plaintiffs must establish that municipal decisionmakers continued to adhere to "an approach that they know or should know has failed to prevent tortious conduct by employees." Brown, 520 U.S. at 407, 117 S.Ct. 1382. But here, the record shows that the District took numerous steps to improve its SNAP application processing timeliness once it was made aware of the system's deficiencies.

Following its receipt of the October 2017 FNS Letter calling for a CAP, the agency conducted an extensive "root cause" diagnosis of SNAP administration. Issues were fielded from customer complaints (both in-person and through the help desk), community partners and advocates, and internal analysis and escalated to high-level agency management. Seymour Dep. 81:6–84:11. A significant product of this root cause analysis was the District's implementation of a "one-and-done" system for customer interactions. Id. 129:20–130:1. For applicants coming into a service center, the agency now endeavors to make an eligibility determination before the applicant leaves the service center that day. See Decl. of Laura Zeilinger ("Zeilinger Decl.") ¶¶ 69, 73–74.¹⁶ The District therefore has trained SSRs to address all potential issues with a customer's application on the spot, including conducting the interview and helping them find missing verification. Zeilinger Dep. 49:2–21; Bryant-Rollins Dep. 61:3–62:6; Seymour Dep. 130:2–131:8. If the clients’ applications were fully processed all in one setting, the District reasoned, customers would not have to "come back to us with this piece of paper or that piece of paper," and the District's APT rate would improve. Seymour Dep. 130:15–131:8. Since its initial rollout, the new system has been "in a constant state of improvement" and "adjusted based on HR and union feedback." Long Dep. 276:15–22.

16 Upon arrival at a service center, customers are given two options: either they can wait for a face-to-face interview that day or they can submit their applications and documents and someone will call them from the lobby within 24 hours to conduct their interview. Bryant-Rollins Dep. 45:6–9.

DHS has also overhauled its recertification application process. In the past, the agency sent customers "a letter that says you must come on this day [to do your recertification interview] and that's it," meaning that a client's only option for filing for recertification would be to arrive on the specified day, fill out the long recertification application form, and potentially wait all day and not be seen. Deposition of Edward Wilkins Long III ("Long Dep.") 40:4–9, 40:20–41:7. To relieve customers of the burden of returning to the service center every year, the agency now sends the recertification application form to the customer with the notice of expiration and has obtained a waiver from FNS that permits recertification interviews to be conducted over the phone. Long Dep. 39:6–41:12.

The Department also brought on a technical expert to fix the problems that it was having with notice generation. Platt Dep. 64:19–21. That expert established metrics to monitor notice generation to test whether technical fixes worked and a reconciliation process for ensuring that DHS's print vendor was actually printing and sending the notices it was provided. Platt Dep. 67:7–68:6. The new system "went through refinement" and "evolved over time" with input from municipal decisionmakers. Id. at 248:13–21, 270:15–21 (explaining that "timeliness was always a discussion item as a problem that needed to be addressed" at the agency's twice weekly meetings). These reforms appear to have improved customers’ experiences applying for recertification. One of the named Plaintiffs, for example, testified that the agency "did send [him] a notice" about his most recent recertification and that he "ha[sn't] had any problems lately" with filing for recertification. Dep. of James Stanley ("Stanley Dep.") 27:8–22, 34:10–11, 35:1–3; see also Dep. of Kathryn Harris ("Harris Dep.") 31:8–35:8 (testifying that she prefers completing the recertification application over the phone rather than having to go to a service center and that she has not had difficulty verifying information for her recertification applications).

Moreover, the record shows that the District has given SSRs ongoing "specific training about how to" process applications timely. Robinson v. Pezzat, 818 F.3d 1, 13 (D.C. Cir. 2016) (distinguishing Smith v. District of Columbia, 413 F.3d 86, 99 (D.C. Cir. 2005), "where the District had provided employees with no relevant training"); see generally Long Dep. 300:4–15 (explaining that "the caseworkers [were the agency's] second most important stakeholder group after District residents who are receiving benefits" and that "they have very hard jobs" and the agency's "objective [was] to make their jobs less painful"). A general refresher training was given to DHS employees to "[e]nsure an understanding of the policies that dictate timely processing to increase timeliness" followed by targeted trainings to "ensure specific deficiencies are addressed." Pl. Reply, Exh. J at 1; see also Wells Dep. 194:10–195:5 (explaining that the agency provided "a quick shot workshop refresher on SNAP timeliness requirements" to all staff in February 2018). More generally, the record reveals that when the QC team sees a high number of errors in a particular activity, that information is provided to the training team and incorporated into "refresher trainings" and built into the agency's biannual corrective action plan. Wells Dep. 41:16–42:1, 48:15–49:3. Similarly, the agency "immediately roll[s] out a training" when it "get[s] a new policy directive from FNS." Id. at 42:8–12.

Plaintiffs present little evidence to rebut the District's extensive testimonial and documentary evidence of its efforts to improve timeliness, which is entitled to a presumption of good faith. See Kretchmar v. FBI, 882 F. Supp. 2d 52, 56 (D.D.C. 2012) ("Agency declarations are accorded ‘a presumption of good faith,’ and to rebut them plaintiff ‘must point to evidence sufficient to put the Agency's good faith into doubt.’ " (quoting Long v. DOJ, 450 F. Supp. 2d 42, 54 (D.D.C. 2006) ; Ground Saucer Watch, Inc. v. CIA, 692 F.2d 770, 771 (D.C. Cir. 1981) )). Notably, Plaintiffs do not cite any deposition testimony from the individual named plaintiffs that documents existing problems with the District's administration of SNAP. To the contrary, the portions of the plaintiffs’ deposition testimony cited by the District suggest that D.C. residents are seeing improvements in their experience applying for SNAP benefits. See, e.g., Harris Dep. 31:8–35:8 (testifying that she is happy that the District switched to allowing her to apply for recertification over the phone rather than having to go to a service center and that she has not had difficulty verifying information for her recertification applications); Stanley Dep. 27:8–22, 34:10–11, 35:1–3 (testifying that the agency "did send [him] a notice" about his most recent recertification and that he "ha[sn't] had any problems lately" since March 2018 with filing for recertification).

The only affirmative evidence that Plaintiffs point to are past letters from FNS citing the District for failing to provide notices of required verification. Pl. MSJ 23. But, that evidence is outdated—some of the letters are from as early as May 2018. See Pl. MSJ, Exh. G. Although FNS has not officially removed the District from the CAP, the record indicates that FNS is presently satisfied with the District's efforts to address application processing timeliness. After extensive back-and-forth with the District, FNS accepted the District's proposed CAP in January 2020 and specifically noted that "the District's plan to correct errors related to the timel[y] issuance of correct notices is acceptable." See Letter from Eric Ratchford to Anthea Seymour ("FNS Jan. 2020 Letter") 1–2 (Jan. 29, 2020), Gov. MSJ, Exh. P. And, in FNS's most recent correspondence to the District in March 2020, FNS commented that the District's "timeliness rate has improved and is stable" and "encourage[d] the District to balance the focus on timeliness with an increased attention on payment accuracy and being stewards of public funds."¹⁷ Letter from Eric Ratchford to Anthea Seymour ("FNS Mar. 2020 Letter") 1 (Mar. 6, 2020), Gov. Notice, ECF No. 152; see also Seymour Dep. 241:13–242:3 (noting that FNS is "pleased" with and "cognizant of [the District's] strategies to continue to work on our APT and to maintain the level").

17 As FNS recognizes, there is a trade-off between the amount of time that agency employees can spend processing each application and the number of customers that the District can serve in one day. Compare Jones Dep. 103:7–12 (testifying that the average transaction time is about two hours at the Taylor Street Center), with Harris Dep. 42:10–14 (suggesting that the District "try to streamline the application process so that the participants as well as the social workers would not have to spend so much time with one client"). Plaintiffs point out that the wait times at some of the District's service centers can be very long and that many customers may be turned away. See, e.g., Harris Dep. 42:4–9 (testifying that "participants have to wait so very, very long is all kinds of weather and they have to wait sometimes inside and outside" of service centers); Pl. MSJ, Exh. N; Bryant-Rollins Dep. 70:2–72:20, 79:5–80:16. But, while no doubt frustrating, long wait times for customers does not bear on whether the District was deliberately indifferent to Plaintiffs's rights to timely processing of benefits applications. The District's choice to have SSRs spend more time with each application so as to ensure that they can be processed "one-and-done" (and therefore, timely) does not reflect deliberate indifference as to the Plaintiff's statutory rights.

In light of the record as a whole, the Court finds that the District's approach is reasonably designed to achieve full compliance with the SNAP Act's timeliness requirements. SNAP is a complex program to manage, and there is no perfect blueprint for ensuring perfect compliance.¹⁸ This is hardly a case in which "in light of the duties assigned to specific officers or employees the need for more or different training is so obvious, and the inadequacy so likely to result in the violation of [federal] rights, that the policymakers of the city can reasonably be said to have been deliberately indifferent to the need." City of Canton, Ohio v. Harris, 489 U.S. 378, 390, 109 S.Ct. 1197, 103 L.Ed.2d 412 (1989) ; compare with, e.g., Salazar, 954 F. Supp. at 331 (finding an agency's practices so inadequate as to establish deliberate indifference where the agency "deliberately assigned only one person ... to run the entire [assistance] program" and completely failed to monitor compliance).

18 Indeed, a study conducted by FNS of all 50 States and the District of Columbia "did not find any specific management practice that led to a statistically significant difference in APT rate." Gov. MSJ, Exh. T at 2. Some strategies associated with high APT rates were: (1) establishing clear performance targets or goals for improving APT; (2) holding workers accountable for overdue cases in the worker's performance reviews or decisions about the worker's employment status; (3) training staff about new application processing procedures; and (4) monitoring APT performance either weekly or monthly. Id. The District appears to have implemented many of these strategies.

Moreover, the efficacy of the District's improvement efforts is borne out in its recent timeliness results. To be sure, there are no "numerical standard[s] control[ling] the determination of whether incidents of wrongful behavior cumulatively show a pattern amounting to a custom or policy." Carter v. District of Columbia, 795 F.2d 116, 124 (D.C.Cir. 1986). The Court finds it significant, however, that the District's current timeliness in processing initial and recertification applications—though not perfect—does not come close to non-compliance benchmarks in analogous cases. Take the level of noncompliance with regulatory statutes that other courts have found to constitute deliberate indifference. In Salazar v. District of Columbia, 954 F. Supp. 278 (D.D.C. 1996), retired Judge Kessler of this court confronted a similar provision in the Medicaid statute that required absolute compliance with deadlines for processing applications. Id. at 324. Applying the municipal liability standard, Judge Kessler found that the District was deliberately indifferent to the plaintiffs’ statutory rights where it had "repeatedly failed to process large numbers of Medicaid applications within" the statutory deadline—i.e. , timeliness of 50% for one category of applications and 67% for another category. Id. at 325 (emphasis added).

Nor is the District's noncompliance on par with that in other SNAP Act cases where courts granted injunctive relief. See, e.g., Briggs v. Bremby, No. 3:12-CV-324 VLB, 2012 WL 6026167, at *17 (D. Conn. Dec. 4, 2012), aff'd, 792 F.3d 239 (2d Cir. 2015) (granting preliminary injunction where Connecticut's SNAP application processing timeliness ranged from 59 to 81%); Booth v. McManaman, 830 F. Supp. 2d 1037 (D. Haw. 2011) (granting preliminary injunction where Hawaii's SNAP application processing timeliness ranged from 69 to 84%); Robertson, 766 F. Supp. at 474 (granting permanent injunction where Virginia's statewide SNAP application process timeliness was less than 80%, with some counties’ timeliness under 50%). Notably, none of these cases involved municipalities, so the courts directly applied the SNAP Act's absolute liability standard, rather than the deliberate indifference standard applicable here.¹⁹ Finally, a recent FNS research report shows that only ten States have initial QC APT processing rates of more than 95%; 19 States are between 90% and 95%; and 22 States are below 90%. Gov. MSJ, Exh. T at 1. Compared to other States, then, the District's timeliness completion rate does not stand out. All in all, even though the record reflects some noncompliance with the SNAP Act's timeliness requirements, it is not of such a magnitude to establish the District's deliberate indifference to the Plaintiffs’ statutory rights.

19 Robidoux v. Kitchel, 876 F. Supp. 575 (D. Vt. 1995), granted summary judgment to the plaintiffs on liability where Vermont's most recent SNAP application timeliness ranged from 90% to 93.5%. Id. at 578. In making that determination on liability, however, the court applied the SNAP Act's strict liability standard rather than Monell’s custom or policy standard.

Taking a different tack, Plaintiffs suggest that the District has been deliberately misleading FNS as to the extent of its untimeliness. They first point to the District's creation of the Adjusted State APT Rate, which excludes applications pended for missing verification, as an effort to inflate the District's timeliness numbers. But the record shows that the District has been transparent with FNS about how the Adjusted State APT Rate is calculated and its limitations. See, e.g., Sept. 2018 APT CAP 3 (explaining to FNS that "recognizing the aforementioned limitations, the District [has] attempted to identify properly pended applications using the date when the customer provided the last remaining verification item"); FNS Monthly Report from D.C. for Sept. 2018 ("FNS Sept. 2018 Report") 6, Gov. MTD Reply, Exh. A, ECF No. 135-1 (reporting the Adjusted State APT Rate and explaining that it "exclude[s] applications where the number of days from the application filing date to the latest verification clearance date exceeds the required timeframe" and "consider[s] them delayed due to customer's delayed submission of documentation"). Although FNS told the District in January 2020 that it "may" remove the Adjusted State APT Rate from its monthly reports, FNS. Jan. 2020 Letter 4, there is no indication that FNS considered the Adjusted State APT Rate to be deliberately misleading or inappropriate for the District to report.

Plaintiffs also point to an email chain from October 2018 in which Won-ok Kim, DHS's Deputy Administrator of Data Analytics Research and Evaluation, purportedly attempted to hide the extent of the District's document-processing backlog from FNS. In the email chain, Ms. Kim forwarded an email from FNS to a group of DHS employees seeking advice on how to respond to FNS's request for backlog data. Pl. MSJ Reply, Exh. P at 6–7. Ms. Kim stated, "I know that we have a bunch of backlog documents in DIMS ... but I cannot tell them we have thousands of backlog documents sitting in DIMS" and "I don't know how to describe here without revealing the problem we have. Appreciate if you can come up with some language to explain why we cannot provide DIMS backlog data." Id. at 7–8. Read in the light most favorable to Plaintiffs, a reasonable juror could infer that Ms. Kim was trying to conceal backlog information from FNS. But, a "single isolated incident" of misconduct (even if one could call it that) can establish municipal liability only if attributed to a municipal policymaker with final decision-making authority. City of Oklahoma City v. Tuttle, 471 U.S. 808, 821, 105 S.Ct. 2427, 85 L.Ed.2d 791 (1985) ; see also Triplett v. District of Columbia, 108 F.3d 1450, 1453 (D.C. Cir. 1997) ("The only acts that count ... are ones by a person or persons who have ‘final policymaking authority [under] [S]tate law.’ " (quoting Jett v. Dallas Independent School Dist., 491 U.S. 701, 737, 109 S.Ct. 2702, 105 L.Ed.2d 598 (1989) ) (second alteration in original)); Blue v. District of Columbia, 811 F.3d 14, 19 (D.C. Cir. 2015) ("[U]nder certain circumstances, a single decision by a municipal official with final policymaking authority can constitute a municipal policy."). Here, two employees who outrank Ms. Kim—the Chief Information Officer and Assistant Deputy Administrator of ESA, see D.C. Dep't of Human Servs. Senior Mgmt. (Dec. 2019), https://dccouncil.us/wp-content/uploads/2020/01/Attachment-1-DHS-Org-Chart.pdf—responded to Ms. Kim that the agency should be transparent with FNS about the backlog problem. See, e.g., Pl. MSJ, Exh. P at 5 (Chief Information Officer explaining that "I actually think we should just explain the problem as it is instead of trying to explain it without revealing the problem. FNS already know[s] it's a problem we are working to address"); id. at 4 (Assistant Deputy Administrator of ESA explaining that "we are on the same page ... with explaining the as-is state and our plans to address the backlog"). And, there is no evidence in the record that the backlog was actually concealed from FNS in the end or of how the backlog relates to the District's application processing timeliness. Moreover, unrebutted testimony from other agency officials supports the conclusion that the District generally has been transparent in its dealings with FNS. See, e.g., Long Dep. 264:10–17, 266:2–9 (testifying that "there's no situation [where] somebody said this data is inaccurate, let's give it to FNS anyway" and that no one discussed "strategies for altering the data logic to improve the rate of timely processing"). Thus, the October 2018 email chain does not establish the District's deliberate indifference.

IV. Conclusion

The evidence before the Court establishes that the District of Columbia currently processes somewhere between 90 and 95 percent of SNAP benefit applications within the deadlines mandated by the SNAP Act. And the vast majority of the delayed applications are resolved within 30 days of the deadlines. The Court appreciates that any delayed application threatens the loss of food and nutrition for needy families. At the same time, some minimal degree of untimeliness is inevitable in a sprawling public benefits program like SNAP.

Were the District a State, even a 90 to 95 percent timeliness rate likely would not be sufficient to avoid summary judgment for violating Act, which requires strict adherence to the statutory deadlines. Because it is not a State, however, Plaintiffs must show that the District's untimely processing of initial or recertification applications is attributable to a "persistent, pervasive practice, attributable to a course deliberately pursued by official policymakers." Salazar, 954 F. Supp. at 324 (quoting Carter, 795 F.2d at 125–26 ). They have not created a genuine dispute of fact on that question. Unlike in other cases finding deliberate indifference on the part of municipal officials, the record is clear that the District made extensive and sustained efforts at improvement once it was put on notice—by both Plaintiffs and FNS—of the deficiencies in its SNAP application processing timeliness, which stemmed in part from the roll-out of a new computer system. Compare id. at 326 (finding deliberate indifference based on the agency's failure to rectify problems in Medicaid application processing after "high-level DHS managers have known about [the agency's] recurrent backlogs of NPA Medicaid applications for several years"), and M.J. v. District of Columbia, 401 F. Supp. 3d 1 (D.D.C. 2019) (finding that a complaint adequately alleged deliberate indifference based on "numerous public reports throughout the complaint demonstrating that defendants were aware of the need for comprehensive community-based care, and the inadequacy of the services the District currently offers"), with Westfahl v. District of Columbia, 75 F. Supp. 3d 365, 378 (D.D.C. 2014) (Cooper, J.) ("[S]pecific inadequate practices do not necessarily evidence a deliberate indifference to constitutional violations if the city's overall policies demonstrate that it has undertaken efforts to reduce the illegal use of force[.]"). Plaintiffs therefore have failed to establish that the District's untimely processing of some SNAP applications is the product of a municipal policy or practice of deliberate indifference.

Accordingly, the Court will vacate its prior preliminary injunction and grant summary judgment to the District. The Court need not reach Plaintiffs’ request for a permanent injunction. A separate order follows.