United States Court of Appeals FOR THE DISTRICT OF COLUMBIA CIRCUIT
Argued November 3, 2023 Decided June 7, 2024
No. 23-5025
AFGHAN AND IRAQI ALLIES, UNDER SERIOUS THREAT BECAUSE OF THEIR FAITHFUL SERVICE TO THE UNITED STATES, ON THEIR OWN AND ON BEHALF OF OTHERS SIMILARLY SITUATED, APPELLEE
v.
ANTONY J. BLINKEN, ET AL., APPELLANTS
Appeal from the United States District Court for the District of Columbia (No. 1:18-cv-01388)
Steven A. Platt, Senior Litigation Counsel, U.S. Department of Justice, argued the cause for appellants. With him on the briefs were Brian M. Boynton, Principal Deputy Assistant Attorney General, Yamileth G. Davila, Assistant Director, and Ruth Ann Mueller, Sean L. King, and David J. Byerley, Trial Attorneys.
Mariko Hirose argued the cause for appellee. With her on the brief were Deepa Alagesan, Melissa S. Keaney, Linda H. Martin, David Y. Livshiz, Rebecca C. Kerr, and Justin C. Simeone. Anika Havaldar entered an appearance. 2 Travis L. Gray and Stephanie F. Cagniart were on the brief for amicus curiae Ambassador Ryan C. Crocker in support of appellee.
Robert Reyes Landicho, Jeremy C. Marwell, Meghan Natenson, and Robert H. Wu were on the brief for amicus curiae U.S. Representative Earl Blumenauer in support of appellee.
Christopher W. Dempsey was on the brief for amicus curiae Association of Wartime Allies in support of appellee.
Before: SRINIVASAN, Chief Judge, PILLARD and KATSAS, Circuit Judges.
Opinion for the Court filed by Circuit Judge KATSAS.
KATSAS, Circuit Judge: Congress has authorized the Secretary of State to give special-immigrant visas to certain Iraqi and Afghan nationals who face serious threats because of their faithful service to the United States during recent armed conflicts. After initial applications for these visas languished, Congress further provided that the government “shall” improve its efficiency so that it “should” process the applications within nine months, except in cases involving unusual national- security risks.
The plaintiffs here represent a class of individuals who have had applications for such visas pending for more than nine months. In 2019, the district court held that the government had unreasonably delayed processing these applications. In 2020, the court approved a plan requiring the prompt adjudication of applications filed by class members and pending for more than nine months as of May 21, 2020. In 2022, the Secretary moved to terminate or modify the plan based on changed circumstances in the two years since 2020. 3 The district court recognized that changed circumstances warrant modifying the plan, but it refused to terminate the plan. The government appeals the refusal to terminate. We affirm.
I
For much of the last few decades, the United States has engaged in armed conflict in Iraq and Afghanistan. During this time, many Iraqi and Afghan nationals helped the United States—often at great personal risk. To aid such individuals, Congress enacted the Refugee Crisis in Iraq Act of 2007, Pub. L. No. 110-181, §§ 1241–49, 122 Stat. 395 (2008) (RCIA), and the Afghan Allies Protection Act of 2009, Pub. L. No. 111-8, §§ 601–02, 123 Stat. 807 (AAPA).1 These statutes authorize the Secretary of State, in consultation with the Secretary of Homeland Security, to confer immigration benefits on Iraqi and Afghan nationals who have worked for the United States Government or the International Security Assistance Force for at least one year, provided “faithful and valuable service,” and experienced “an ongoing serious threat” as a result. RCIA § 1244(b)(1); AAPA § 602(b)(2)(A). Specifically, the Secretary of State may confer “the status of a special immigrant” on aliens who satisfy these criteria, apply for special-immigrant visas, establish eligibility for such visas, and clear a background check. RCIA § 1244(a); AAPA § 602(b)(1). The Secretary also may confer the same status on immediate family members of such aliens. RCIA § 1244(b)(2)(A)–(B); AAPA § 602(b)(2)(B)(i)–(ii). With such status, these individuals may receive visas reserved for “special immigrants.” See 8 U.S.C. §§ 1101(a)(27), 1153(b)(4).
In 2013, Congress amended the RCIA and AAPA to address the slow pace at which the government had been
1 The RCIA and AAPA are codified as amended in notes to 8 U.S.C. §§ 1157 and 1101, respectively. 4 processing applications. See National Defense Authorization Act for Fiscal Year 2014, Pub. L. No. 113-66, §§ 1218–19, 127 Stat. 672, 910–15 (2013). As amended, the statutes direct that the Secretaries of State and Homeland Security, in consultation with the Secretary of Defense, “shall improve the efficiency by which applications for special immigrant visas [under both statutes] are processed, so that all steps under the control of the respective departments incidental to the issuance of such visas, including required screenings and background checks, should be completed not later than 9 months after the date on which an eligible alien submits all required materials to complete an application for such visa.” RCIA § 1242(c)(1); AAPA § 602(b)(4)(A).2 At the same time, Congress qualified this directive by providing that “[n]othing” in it “shall be construed” to limit the Secretaries’ ability to “take longer than 9 months” to consider visa applications in “high-risk cases for which satisfaction of national security concerns requires additional time.” RCIA § 1242(c)(2); AAPA § 602(b)(4)(B). Congress also required the Secretaries to make quarterly reports about the processing of these visa applications, which must explain “the reasons for the failure to process any applications that have been pending for longer than 9 months.” RCIA § 1248(f)–(g); AAPA § 602(b)(11)–(12).
In 2018, five applicants for special-immigrant visas under the RCIA and AAPA—four from Afghanistan and one from Iraq—filed this lawsuit. These plaintiffs sought to represent a class of individuals who have had such applications awaiting government action for more than nine months. The plaintiffs alleged that the government has unreasonably delayed its
2 In 2021, Congress amended this language in the AAPA, but not the RCIA, to elaborate that the steps under control of the government include “Chief of Mission approval.” Emergency Security Supplemental Appropriations Act, Pub. L. No. 117-31, § 401(a)(3), 135 Stat. 309, 316 (2021). 5 processing and adjudication of these applications. Counts one and two of the complaint sought declaratory and injunctive relief with respect to the alleged unreasonable delay. Counts three through five of the complaint raised other claims. The district court provisionally certified the class, denied a motion to dismiss, permitted discovery, and consolidated a preliminary-injunction hearing with a merits trial on the first two counts of the complaint.
In September 2019, the district court granted partial summary judgment to the plaintiffs on their first two claims. To assess unreasonable delay, the court applied our decision in Telecommunications Research & Action Center v. FCC, 750 F.2d 70 (D.C. Cir. 1984) (TRAC), which requires consideration of how long the agency has taken to act, the interests affected by its delay, and the effect of expedition on other agency priorities. See id. at 80. As to the first consideration, the court stressed the substantial difference between the “9-month statutory benchmark” imposed by Congress and the “5-year average wait time” endured by class members. J.A. 10. As to the second, the court concluded that many class members face a substantial risk of retribution for assisting the United States and that the risk has not dissipated over time. Id. at 14–15. As to the third, the court concluded that the government had not identified any “specific agency activities” that would be harmed if visa applications under the RCIA and AAPA were processed more quickly. Id. at 16. Based on these considerations, the court declared that the government had unreasonably delayed its processing and adjudication of class members’ visa applications. Id. at 20. As a remedy, the court ordered the government to develop and submit “a plan for promptly processing and adjudicating the applications of current class members.” Id. at 19. In later orders, the court denied reconsideration and finalized its class certification. 6 In June 2020, the district court entered an injunction approving a plan with specific timelines for adjudicating special-visa applications that had been pending for more than nine months as of May 21, 2020. The adjudication plan recognized and took account of challenges arising from the COVID-19 pandemic, from closure of the Visa Unit at the United States embassy in Afghanistan, and from severe limits on the availability of consular services in Iraq. J.A. 68. The plan also accommodated cases “requir[ing] additional processing time to reconcile any national security concerns.” Id. at 69. The plan required the government to submit progress reports every 90 days. Id. at 70–75. The government appealed the injunction but then voluntarily dismissed its appeal. See Afghan & Iraqi Allies v. Blinken, No. 20-5251, 2021 WL 4765441 (D.C. Cir. Sept. 16, 2021) (mem).
In October 2021, the parties jointly sought a temporary stay of the adjudication plan to pursue settlement negotiations. The district court granted the stay and, after the negotiations broke down, extended it over the plaintiffs’ objections. The plaintiffs appealed the district court’s refusal to lift the stay, but we dismissed that interlocutory appeal for lack of jurisdiction. See Afghan & Iraqi Allies v. Blinken, No. 22-5183, Doc. 1973117 (D.C. Cir. Nov. 10, 2022). The stay remains in effect.
In May 2022, the government moved for relief from the adjudication plan based on changed circumstances. Among other things, the government cited a surge in visa applications after the United States withdrew from Afghanistan, increased difficulty in processing applications following closure of the embassy in Kabul, worsening security risks in Iraq, ongoing complications from the pandemic, a drain on agency resources due to needs in Ukraine, and a significant improvement in the average times for processing visa applications under the RCIA and AAPA. 7 The district court granted relief in part. It held that these factual developments were “significant” and “warrant[ed] allowing the government to propose modified timing benchmarks, but not to abandon its reporting or explanation obligations.” J.A. 994. In reaching this conclusion, the court reassessed the TRAC factors in light of the changed circumstances. Id. at 995. The court acknowledged that the government’s job of adjudicating visa applications had “undoubtably become more difficult,” but it noted that many applications remained pending for more than nine months. Id. at 995–96. The court also concluded that additional applicants had entered the class since 2020 and that continuing delays had put all the plaintiffs even more at risk. Id. at 997–98. The court referred the case to a magistrate judge to oversee the development of a new adjudication plan that accommodated the various changed circumstances and covered newly added class members. Id. at 999–1001.
The government appealed the court’s refusal to terminate the injunction in its entirety. We have jurisdiction under 28 U.S.C. § 1292(a)(1), which authorizes review of interlocutory orders “refusing to dissolve or modify injunctions.”
II
We begin with preliminary points about the governing procedural rule and our standard of review.
A
In the district court, the government moved to terminate the adjudication plan under Federal Rules of Civil Procedure 54(b) and 60(b). These two rules afford different degrees of latitude to the district court to modify its rulings in ongoing cases. Rule 54(b) provides for the disposition of cases involving multiple claims or parties. It permits district courts 8 to enter “final judgment” as to a subset of claims or parties “only if the court expressly determines that there is no just reason for delay.” Otherwise, any order “that adjudicates fewer than all the claims or the rights and liabilities of fewer than all the parties does not end the action as to any of the claims or parties and may be revised at any time before the entry of a judgment adjudicating all the claims and all the parties’ rights and liabilities.” Rule 60(b) sets forth six narrow grounds on which parties may seek relief “from a final judgment, order, or proceeding.” The government contends that Rule 54(b), which allows district courts to revisit non-final orders, governed its motion. The plaintiffs counter that Rule 60(b), which significantly restricts the ability of district courts to revisit final orders, alone governed the motion.
On this point, we agree with the government. The 2020 order addressed counts one and two of the complaint. But the complaint raised three other counts, and the district court has neither ruled on them nor entered a partial final judgment after finding that there is no just reason for delay. Because the district court did not enter final judgment on any counts in the action, it could revise its orders “as justice requires.” Capitol Sprinkler Inspection, Inc. v. Guest Servs., Inc., 630 F.3d 217, 227 (D.C. Cir. 2011) (cleaned up).
The plaintiffs object that Rule 54(b) does not apply to permanent injunctions because they are immediately appealable without any finding that there is “no just reason for delay.” But a permanent injunction addressing a subset of counts is neither “final” under Rule 54(b) nor immediately appealable as a “final” decision under 28 U.S.C. § 1291. Such injunctive orders are immediately appealable, but only because 28 U.S.C. § 1292(a)(1) permits review of interlocutory orders granting, denying, modifying, or refusing to modify injunctions. In sum, injunctions are no more “final” than any 9 other orders for purposes of revision consistent with Rule 54(b). We thus evaluate the district court’s refusal to terminate the injunction as a partial denial of a Rule 54(b) motion to reconsider a non-final order.
B
We review the denial of Rule 54(b) motions to reconsider for abuse of discretion. Capitol Sprinkler, 630 F.3d at 225–26. Such deferential review is particularly appropriate where reconsideration involves unreasonable-delay determinations. In the first instance, such a determination turns on “consideration of the particular facts and circumstances” regarding the specific agency action at issue. Mashpee Wampanoag Tribal Council, Inc. v. Norton, 336 F.3d 1094, 1100 (D.C. Cir. 2003) (citing TRAC, 750 F.2d at 80). So too does an assessment of how changed circumstances should or should not affect a TRAC order. In general, when a district court decides questions that are highly fact-intensive, deferential review is appropriate. See, e.g., Highmark Inc. v. Allcare Health Mgmt. Sys., Inc., 572 U.S. 559, 564 (2014) (determination whether a case is “exceptional”); Cooter & Gell v. Hartmarx Corp., 496 U.S. 384, 403–04 (1990) (determination whether Rule 11 sanctions are warranted); Pierce v. Underwood, 487 U.S. 552, 557–62 (1988) (determination whether an agency’s position was “substantially justified”). A conclusion that changed facts do not justify terminating an injunction to remedy past unreasonable delay fits comfortably within this line of cases.
The government responds that Cobell v. Norton, 240 F.3d 1081 (D.C. Cir. 2001), requires de novo review of a finding of unreasonable delay. But Cobell said only that “the legal standard used to determine whether agency delay is unreasonable is a question of law to be reviewed de novo.” Id. 10 at 1096. The parties here agree that TRAC establishes the governing legal standard, and Cobell does not speak to the standard for reviewing the district court’s TRAC-factor balancing in individual cases. Much less does it require de novo review of an assessment whether new developments require modifying an injunction to remedy past unreasonable delays. For that Rule 54(b) question, our review is deferential.
III
In TRAC, this Court set forth a framework for analyzing claims of unreasonable agency delay. We highlighted six considerations that, although neither “ironclad” nor exhaustive, still provide “useful guidance”:
(1) the time agencies take to make decisions must be governed by a rule of reason; (2) where Congress has provided a timetable or other indication of the speed with which it expects the agency to proceed in the enabling statute, that statutory scheme may supply content for this rule of reason; (3) delays that might be reasonable in the sphere of economic regulation are less tolerable when human health and welfare are at stake; (4) the court should consider the effect of expediting delayed action on agency activities of a higher or competing priority; (5) the court should also take into account the nature and extent of the interests prejudiced by delay; and (6) the court need not find any impropriety lurking behind agency lassitude in order to hold that agency action is unreasonably delayed.
750 F.2d at 80 (cleaned up). TRAC set forth this framework to assess whether agency delays are “so egregious as to warrant mandamus” under the All Writs Act, in cases where the delayed action would be directly reviewable in this Court. See 11 id. at 75, 79–80. But we have routinely applied the same framework to assess claims that agency action has been “unreasonably delayed” for purposes of the Administrative Procedure Act, 5 U.S.C. § 706(1). See, e.g., Da Costa v. Immigr. Inv. Program Off., 80 F.4th 330, 339–40 (D.C. Cir. 2023); Mashpee, 336 F.3d at 1099–1100. Upon finding such an unreasonable delay, the district court need not order immediate action, but may instead establish deadlines to ensure that the agency “is proceeding as diligently as possible with the resources available to it.” See Mashpee, 336 F.3d at 1102.
The parties here make sweeping arguments, but the question before us is narrow. In seeking termination of the injunction, the government did not contend that the district court impermissibly found unreasonable delay in 2019 or impermissibly entered the injunction in 2020. Instead, it sought relief based on changed circumstances between 2020 and 2022. So, we take the 2019 and 2020 orders as a given, and we consider only whether the district court reasonably responded to the changes. Moreover, the injunction governs the claims of only a subset of the class—individuals with applications for visas under the RCIA or AAPA that had been pending for more than nine months as of May 21, 2020. The district court contemplated future injunctions governing the claims of other class members, but the appropriateness of any such injunction is not presently before us. Finally, the district court granted the government considerable relief even as to class members covered by the 2020 injunction: The court held that changed circumstances warrant a new plan, stayed the 2020 plan, referred the case to a magistrate judge, and ordered the parties to develop a new plan. So, the government prevailed on its argument that intervening factual developments warrant changing the 2020 injunction. To win here, the government must show that no continued judicial involvement remains 12 appropriate, and that it was an abuse of discretion for the district court to conclude otherwise.
We are tempted to reject this contention based on one overarching consideration: Some beneficiaries of the 2020 injunction—who had visa applications pending for more than nine months as of May 21, 2020—still have not had their applications fully adjudicated. The parties stipulated that there were thousands of such beneficiaries in October 2021, and it remains undisputed that at least some of these old applications remain pending.3 If the delay in these adjudications was already unreasonable when the injunction was entered in June 2020, we struggle to see how changed circumstances could make those delays not unreasonable some two years later, when the government moved to terminate the injunction. Of course, intervening changes could affect the reasonableness of additional future delays, as the district court recognized. But post-2020 changes do not retroactively convert past unreasonable delays into reasonable delays. And with the terms of the successor injunction still undetermined, we cannot decide in advance whether it will sufficiently accommodate the increased difficulties faced by the government.
While this problem alone might doom the government’s case, the district court did update its original analysis of the various TRAC factors to account for changed circumstances. We need not consider whether this updating was necessary for the court to deny the motion to terminate the injunction in its
3 The government acknowledged as much at oral argument, Oral Arg. at 2:06–29, and a January 2024 report on the RCIA confirms this point. The RCIA required principal visa applications to be filed by September 30, 2014. RCIA § 1244(c)(3)(C)(iii). Yet the report indicates that some applications are still pending—nearly a decade after the deadline for applying. U.S. Dep’t of State, Report of the Iraqi SIV Program–January 2024. 13 entirety. As we now explain, the court permissibly concluded, based on its updated TRAC analysis, that some continued judicial involvement remains appropriate.
The first two TRAC factors focus on the extent of and reasons for the agency delay. The first factor is that “the time agencies take to make decisions must be governed by a rule of reason.” TRAC, 750 F.2d at 80 (cleaned up). We have described this as the “most important” consideration under TRAC. In re Core Commc’ns, Inc., 531 F.3d 849, 855 (D.C. Cir. 2008). The second factor is that, “where Congress has provided a timetable or other indication of the speed with which it expects the agency to proceed in the enabling statute, that statutory scheme may supply content for this rule of reason.” TRAC, 750 F.2d at 80. The second TRAC consideration thus “gives content to the first.” In re United Mine Workers of Am. Int’l Union, 190 F.3d 545, 549 (D.C. Cir. 1999).
The district court jointly considered and reasonably balanced the first two TRAC factors. It acknowledged that the task of adjudicating visa applications had “undoubtably become more difficult” after the withdrawal from Afghanistan led to a surge in special-immigrant visa applications and closure of the embassy in Kabul. J.A. 995. At the same time, the court noted that the nine-month timeline remained unchanged and that visa applications were “still pending beyond the statutory deadline.” Id. at 996. The court further noted that the COVID-19 pandemic and security risks in Baghdad had been present at all relevant times. Id. at 996–97. For these reasons, the court concluded that the first and second TRAC factors “still favor Plaintiffs, albeit slightly less heavily than before.” Id. at 997. 14 The government contends that the district court gave too much weight to the statutory timeline. But Congress mandated that the Secretaries “shall improve the efficiency” for processing visa applications under the RCIA and AAPA, so that all steps within their control “should be completed not later than 9 months after” a qualifying alien applies for a visa. RCIA § 1242(c)(1); AAPA § 602(b)(4)(A). The government argues that “should” can sometimes be precatory. But here, contextual clues suggest something closer to a command. For one thing, Congress linked the nine-month benchmark that the Secretaries “should” meet to a command that they “shall” improve efficiency for processing visa applications. Also, if the nine- month timeline were solely aspirational, there would have been little reason to carve out “high-risk cases for which satisfaction of national security concerns requires additional time.” RCIA § 1242(c)(2); AAPA § 602(b)(4)(B). The carveout suggests that a nine-month timeline is appropriate unless a case presents distinctively high risks to the national security.
In any event, the district court treated the timeline as only one consideration among many. And TRAC asks only whether Congress has “provided a timetable or other indication of the speed with which it expects the agency to proceed.” 750 F.2d at 80. Even when a statutory timeline is clearly precatory, like one establishing a “sense of Congress” for when the processing of visa applications “should” be completed, we have still treated the deadline as a “ruler against which the agency’s progress must be measured.” Da Costa, 80 F.4th at 344 (cleaned up). Whether we call the nine-month statutory timeline a “rule” or a “ruler,” the implication is the same: It should play an important role in assessing the reasonableness of the government’s pace of adjudication.
Looking beyond the timeline, the government argues that the district court failed to account for a host of changed facts 15 involving either (i) improvements in how it processes special- immigrant visa applications or (ii) increased difficulties in doing so. Without question, the intervening circumstances cited by the government are significant. For example, the withdrawal of the United States from Afghanistan in 2021 caused the number of monthly visa applications under the AAPA to increase by 443 percent. Likewise, closure of the U.S. embassy in Kabul and travel restrictions imposed by the Taliban have made it substantially more difficult to conduct necessary interviews to process these applications. And the government provides at least some evidence of improved efficiencies between 2019 and 2022. These developments plainly would bear on whether a court could find unreasonable delay as of today or in the future. They also would bear on the contours of any future injunction, as would considerations regarding continuing fallout from the pandemic and continuing security concerns regarding the U.S. embassy in Baghdad.
But again, the only question before us is whether the district court was legally compelled to entirely terminate an adjudication plan for individuals with visa applications pending for more than nine months as of May 21, 2020. The district court found that these delays had already become unreasonable by then, and the government has not challenged that finding. Whatever improvements it might have made in processing other applications, beneficiaries of the 2020 adjudication plan had applications pending for over three-and- a-half years when the district court declined to terminate the plan. Likewise, increased difficulties in processing visa applications may make future delays more understandable, and they may thus be relevant to the terms of any modified plan. But they cannot retroactively make reasonable the already- unreasonable delays for the beneficiaries of the 2020 plan. 16 In sum, the district court permissibly concluded that the first two TRAC factors do not warrant entirely terminating the adjudication plan.
The third and fifth TRAC factors focus on the interests affected by agency delay. The third factor says “delays that might be reasonable in the sphere of economic regulation are less tolerable when human health and welfare are at stake.” TRAC, 750 F.2d at 80. The fifth factor addresses “the nature and extent of the interests prejudiced by the delay.” Id.
The district court assessed these factors jointly. In its 2020 opinion, the court concluded that they support the plaintiffs because many class members face significant risks of harm for having assisted the United States. J.A. 14–15. Then, in its 2022 opinion, the court concluded that these factors weigh “more heavily for Plaintiffs” than before, because continuing delays have put all class members at continued and increasing risk. Id. at 997–98.
The government points to competing interests—including health and safety ones—implicated by the RCIA and AAPA. Specifically, it invokes concerns about foreign policy, national security, terrorism, and COVID. The parties dispute whether these concerns bear on the third and fifth TRAC factors at all. According to the plaintiffs, these factors address only burdens to the party seeking agency action, not burdens to the government or systemwide risks from proceeding too quickly. We need not address this question because the government’s concerns plainly bear on the overall reasonableness of its conduct and the appropriateness of continuing judicial oversight. So the question of which, if any, of the open-ended, non-exhaustive TRAC factors they fall under is largely 17 academic. See TRAC, 750 F.2d at 80 (factors provide only “the hexagonal contours of a standard,” which is “hardly ironclad”).
The district court declined to address the government’s arguments under the third and fifth TRAC factors, but it did reasonably address the government’s underlying concerns. In its 2020 opinion, the court noted that the existence of foreign- policy and national-security concerns was apparent when Congress established the special-immigrant visa programs for Iraqi and Afghan allies and then set the nine-month timeline for processing applications. J.A. 16 n.12. Likewise, in its 2022 opinion, the court noted that difficulties in processing applications from COVID and terrorist attacks in Iraq had not substantially worsened since its 2020 order. Id. at 996–97. Moreover, although the situation in Afghanistan had obviously become much worse, that not only affected processing times, but also increased the risk of harm to Afghans who had supported the United States. Id. at 998. And in any event, the passage of two additional years only heightened the interests of the original beneficiaries of the 2020 adjudication plan. Id. The district court reasonably explained why the injunction should not be fully eliminated.4
4 The government contends more broadly that courts should leave the administration of visa programs to Congress and the Executive Branch. But it does not contend that the doctrine of consular non-reviewability, which bars judicial review of decisions to grant or deny visas to aliens outside the United States, applies to the claims at issue in this case. Cf. Saavedra Bruno v. Albright, 197 F.3d 1153, 1156–64 (D.C. Cir. 1999). Nor does it contend that Kleindienst v. Mandel, 408 U.S. 753 (1972), which permits only limited constitutional scrutiny of certain visa decisions, requires modifying the TRAC standards for addressing unreasonable-delay claims brought under the APA. Absent such categorical limits on our review, we are left to consider whether the district court 18 C
The fourth TRAC factor addresses “the effect of expediting delayed action on agency activities of a higher or competing priority.” 750 F.2d at 80. In its 2020 opinion, the district court concluded that the government had not identified any “specific agency activities” that would be harmed if it had to process RCIA and AAPA visa applications more quickly. J.A. 16. The court then adhered to this conclusion in its 2022 opinion. Id. at 999.
The government’s response stresses In re Barr Laboratories, Inc., 930 F.2d 72 (D.C. Cir. 1991). There, one drug manufacturer sought an injunction compelling the government to adjudicate more quickly its pending applications to sell generic drugs. Id. at 73. We denied relief because an order putting one applicant “at the head of the queue simply moves all others back one space and produces no net gain.” Id. at 75. Other precedents likewise stress concerns about “line-jumping.” Da Costa, 80 F.4th at 339; see also Mashpee, 336 F.3d at 1100–02. But here, there is no such concern within the RCIA and AAPA programs. The class includes all individuals with visa applications pending for more than nine months, and the 2020 plan established guidelines for disposing of all applications pending the longest—those pending for more than nine months as of May 21, 2020. So the plan did not lead to anyone jumping the line to move ahead of others with longer-pending applications.
Of course, we must also consider the possibility of harm to other programs. The government highlights its efforts to resettle Iraqi and Afghan nationals in third-party countries, to
permissibly balanced the various competing interests in declining to terminate the 2020 adjudication plan. 19 alleviate the humanitarian crisis in Ukraine, and to reduce the unprecedented visa backlog caused by the pandemic. And because agencies have only finite resources, any order to step up activity in one program may harm others. Yet this concern can go only so far—agency resources always are limited, and we have never suggested that the possibility of their shifting is always enough to make any TRAC relief inappropriate. Sometimes the right answer is to grant relief with a light touch—for example, by requiring periodic progress reports and reserving the possibility of “additional appropriate relief” in the future. In re United Mine Workers, 190 F.3d at 556.
Given these considerations, the district court reasonably discounted the fourth TRAC factor. The government ticks off its multiple competing priorities, many of which arose after the district court’s 2019 and 2020 orders. But it has not explained with precision the personnel or resource tradeoffs that it faces, or why those constraints justify a complete end to this case. Given this lack of specifics, the district court reasonably concluded that the other TRAC factors still outweigh the government’s generalized appeals to resource constraints.
Furthermore, the district court has accounted for resource constraints. For example, its 2019 decision did not order any immediate government action on pending visa applications, but instead required the development of a plan that the parties submitted jointly. Moreover, the plan did not threaten contempt if the government failed to meet its timelines, but only required the government to explain the reasons for its failure and potentially meet and confer with the plaintiffs. This flexible, measured approach substantially mitigates the government’s concerns about satisfying its various obligations. Finally, as noted above, the district court recognized that intervening circumstances had made the government’s task “undoubtably … more difficult,” J.A. 995, thus warranting a 20 new plan that may include even greater accommodations. And if the government concludes that such further accommodations do not go far enough, it may of course appeal the entry of any successor injunction.
D
The sixth TRAC factor notes that “the court need not find any impropriety lurking behind agency lassitude in order to hold that agency action is unreasonably delayed.” 750 F.2d at 80 (cleaned up). In its 2019 order, the district court found it unnecessary to address this factor. J.A. 9. In its 2022 order, the court assumed that there was no impropriety yet continued to discount this factor. Id. at 999.
The government argues that its good faith affirmatively tips the scales in its favor. But in TRAC itself, we said that bad faith was unnecessary to support judicial intervention. 750 F.2d at 80. And in Da Costa, we described the sixth TRAC factor as “neutral” when there was no plausible evidence of bad faith. 80 F.4th at 345–46. Moreover, even if the government were correct, it strains credulity to suppose that placing some weight on its side of the scale, to account for good faith, would have materially affected the district court’s overall assessment of unreasonableness.
IV
The district court reasonably refused to order a complete termination of the adjudication plan. And no dispute about the details of the original plan—or of any future successor plan— is currently before this Court.
Affirmed.