FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

JACOBO JAJATI, No. 22-56015

Plaintiff-Appellant, D.C. No. 3:22-cv-00175- v. RBM-AGS

UNITED STATES CUSTOMS AND

BORDER PROTECTION; TROY A. OPINION

MILLER, in his official capacity as the Acting Commissioner of the United States Customs and Border Protection; DOES, 1 through 10, inclusive,

Defendants-Appellees.

Appeal from the United States District Court for the Southern District of California Ruth Bermudez Montenegro, District Judge, Presiding

Argued and Submitted December 4, 2023 Pasadena, California

Filed May 22, 2024

Before: Carlos T. Bea, Milan D. Smith, Jr., and Lawrence VanDyke, Circuit Judges. 2 JAJATI V. U.S. CUSTOMS & BORDER PROT.

Opinion by Judge Bea; Dissent by Judge VanDyke

SUMMARY *

Administrative Procedure Act

The panel reversed the district court’s order granting the motion of the U.S. Customs and Border Protection (“CBP”) to dismiss for lack of subject matter jurisdiction, and remanded with instructions that the district court consider Jacobo Jajati’s Administrative Procedure Act (“APA”) claim on the merits in the first instance. The Secure Electronic Network for Travelers Rapid Inspection (“SENTRI”) is a “Trusted Traveler Program,” enabled by 8 U.S.C. § 1365b, that allows a member to avoid a full inspection process when crossing the United States- Mexico border. The CBP revoked Jajati’s SENTRI membership, then reinstated it, then revoked it again without explanation. Jajati brought suit seeking an order that CBP’s revocation decision violated the APA and requesting that his SENTRI membership be reinstated. The district court held that CBP’s decisions to revoke SENTRI memberships were not subject to judicial review because administration of SENTRI was committed to agency discretion under 5 U.S.C. § 701(a)(2).

* This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 3

The panel held that § 701(a)(2) does not bar judicial review of Jajati’s APA claims. Although CBP has broad discretion to revoke SENTRI memberships, the APA recognizes that discretion can be abused. The law governing SENTRI provides meaningful standards under which courts can review whether CBP wielded its discretion in a permissible manner. Jajati’s case is, therefore, not one of those rare instances in which the court lacks jurisdiction because there is no law to apply. The panel reversed the district court’s dismissal for lack of subject matter jurisdiction. The panel remanded for the district court to consider whether CBP’s decision to revoke Jajati’s SENTRI membership violated the APA. In making its determination, the district court should consider whether CBP failed to consider the criteria in 8 C.F.R. § 235.7(a)(4)(x) when it determined Jajati was ineligible to participate in SENTRI. The district court should review CBP’s factual findings under the “substantial evidence” standard, and give due deference to the agency’s expertise. Dissenting, Judge VanDyke would hold that SENTRI eligibility is exactly the kind of administrative action committed to agency discretion by § 701(a)(2) of the APA. The criteria included in the SENTRI regulations do not provide a meaningful standard against which to judge the agency’s exercise of its discretion because they are neither exhaustive nor meaningfully defined. In addition, the complicated balancing of the listed criteria when making a risk determination is a matter peculiarly within the agency’s expertise. Judge VanDyke would therefore affirm the district court’s dismissal for lack of subject matter jurisdiction. 4 JAJATI V. U.S. CUSTOMS & BORDER PROT.

COUNSEL

Saman Nasseri (argued), Nasseri Legal, San Diego, California, for Plaintiff-Appellant. Janet A. Cabral (argued), Assistant United States Attorney; Katherine Parker, Assistant United States Attorney, Civil Division Chief; Randy S. Grossman, United States Attorney; Office of the United States Attorney, San Diego, California; for Defendants-Appellees.

OPINION

BEA, Circuit Judge:

We are called upon to decide whether § 701(a)(2) of the Administrative Procedure Act (“APA”), which bars judicial review of agency actions that are “committed to agency discretion by law,” 5 U.S.C. § 701(a)(2), precludes us from reviewing the United States Customs and Border Protection’s (“CBP”) discretionary decisions to revoke individual memberships in the Secure Electronic Network for Travelers Rapid Inspection (“SENTRI”) program. SENTRI is a “Trusted Traveler Program,” enabled by 8 U.S.C. § 1365b. Under § 1365b, the Department of Homeland Security (“DHS”) is authorized to establish programs that allow pre-approved, low-risk travelers to cross the United States border more easily. In 1996, DHS finalized rules that created the “PORTPASS” program of which SENTRI is a part. See 8 C.F.R. § 235.7; 61 Fed. Reg. 53830, 53831–32 (Oct. 16, 1996). One benefit of SENTRI is that a member need not go through the full inspection process when he crosses the United States–Mexico border. A

JAJATI V. U.S. CUSTOMS & BORDER PROT. 5

SENTRI member, in turn, faces a shorter wait time when crossing the border relative to other persons. SENTRI is administered by CBP. In 2018, Plaintiff-Appellant Jacobo Jajati’s estranged ex- wife was arrested for crossing the border with prohibited drugs: methamphetamine. On the same day of her arrest, Jajati received a notice from CBP that his SENTRI membership had been revoked. The notice stated only that Jajati did not meet the guidelines to participate in the SENTRI program. Jajati’s SENTRI membership was later reinstated, and then again revoked. CBP has never explained why it deemed Jajati ineligible, then eligible, and then again ineligible, to participate in SENTRI. In 2022, Jajati brought suit in the United States District Court for the Southern District of California. Jajati claimed that CBP’s decision to revoke his SENTRI membership was arbitrary, capricious, an abuse of discretion, or not in accordance with the law, in violation of the APA, 5 U.S.C. § 706(2)(A). He sought an order declaring that CBP’s revocation decision violated the APA and requested that his membership in SENTRI be reinstated. The district court held CBP’s decisions to revoke SENTRI memberships are not subject to judicial review because the administration of SENTRI is “committed to agency discretion by law.” Id. § 701(a)(2). The district court thus dismissed Jajati’s claim for lack of subject matter jurisdiction. Jajati appeals. We have jurisdiction under 28 U.S.C. § 1291. We hold that § 701(a)(2) does not bar judicial review of Jajati’s APA claims. Although CBP has broad discretion to revoke SENTRI memberships, the APA itself recognizes that discretion can be “abuse[d].” 5 U.S.C. § 706(2)(A). And the 6 JAJATI V. U.S. CUSTOMS & BORDER PROT.

law governing SENTRI provides meaningful standards under which courts can review whether CBP wielded its discretion in a permissible manner. Jajati’s case is therefore not one of those rare instances in which we lack jurisdiction because “there is truly no law to apply.” See Perez Perez v. Wolf, 943 F.3d 853, 861 (9th Cir. 2019) (internal quotation marks omitted). Hence, we reverse the district court’s order which granted CBP’s motion to dismiss for lack of subject matter jurisdiction. We remand with instructions that the district court consider, in the first instance, whether CBP’s decision to revoke Jajati’s SENTRI membership violated the

APA.

I. A. Jajati is a United States citizen who resides in San Diego, California. Prior to the events at issue here, Jajati had been approved to participate in SENTRI. In 2013, Jajati separated from his then-wife, Margarita Rozillio Jajati (“Margarita”). Their divorce became final in May 2015. Jajati and Margarita have two children together, over whom Jajati has full custody. Jajati alleged that neither he, nor his children, have any ongoing personal or business connection with Margarita. Neither Jajati nor his children communicate, jointly own property or businesses, share bank accounts, or engage in any financial transactions with Margarita. Jajati also alleged that he has never been convicted or charged with any criminal misconduct, nor does he have any association with criminal conduct by others. In October 2018, Margarita was arrested for crossing the border with methamphetamine. She posted bail and

JAJATI V. U.S. CUSTOMS & BORDER PROT. 7

absconded. Margarita is subject to an outstanding warrant for her arrest. On the same day Margarita was arrested, Jajati received notice from CBP that his SENTRI membership had been revoked. The notice stated that Jajati did “not meet SENTRI guidelines,” but provided no other explanation for CBP’s decision to revoke Jajati’s SENTRI membership. In February 2019, Jajati had an interview with a CBP agent concerning the revocation of his SENTRI membership. At the interview, CBP did not question Jajati about his ex-wife. In August 2019, Jajati received a notice that his membership in SENTRI had been reinstated. The notice did not relate why his membership had been terminated earlier, nor why it had been reinstated now. In August 2021, however, Jajati received another notice from CBP stating that his membership had again been revoked. The notice stated only that Jajati no longer qualified for the Trusted Traveler Program. The notice did not explain why CBP had determined that Jajati no longer qualified for the program. Jajati contested the revocation of his SENTRI membership with CBP. In September 2021, Jajati sent CBP a letter which requested that the agency send written notice that explained its reasons for revoking Jajati’s SENTRI membership. CBP never provided such an explanation. Around September 30, 2021, Jajati had another interview with a CBP agent regarding the revocation of his SENTRI membership. The agent refused to disclose why Jajati’s SENTRI membership had been revoked. To date, CBP has never provided any reasons for revoking Jajati’s SENTRI membership in 2018, for reinstating it in 2019, or for 8 JAJATI V. U.S. CUSTOMS & BORDER PROT.

revoking it again in 2021. Jajati’s SENTRI membership remains revoked. B. On February 7, 2022, Jajati filed suit in the Southern District of California against CBP and Troy A. Miller, the Acting Commissioner of CBP. Jajati claimed that CBP’s decision to revoke his SENTRI membership and failure to provide an adequate explanation for doing so violated the APA, 5 U.S.C. § 706. Jajati sought a declaration that CBP’s revocation of his SENTRI membership violated the APA. He requested that the court set aside CBP’s revocation decision and order CBP to reinstate Jajati’s SENTRI membership. CBP moved to dismiss Jajati’s complaint for lack of subject matter jurisdiction. CBP argued that decisions to revoke SENTRI memberships are “committed to agency discretion by law,” 5 U.S.C. § 701(a)(2), such that judicial review of Jajati’s claim is unavailable under the APA. The district court granted CBP’s motion and dismissed Jajati’s claim. It concluded that there were no judicially manageable standards to assess how and when CBP should exercise its discretion to revoke an individual’s SENTRI membership. Thus, the court held that it was “precluded from reviewing the agency’s decision.” For the reasons stated below, we reverse and remand.

II.

The statue governing SENTRI is 8 U.S.C. § 1365b, which instructs the DHS Secretary to “establish an international registered traveler program that incorporates available technologies . . . to expedite the screening and processing of international travelers, including United States

JAJATI V. U.S. CUSTOMS & BORDER PROT. 9

Citizens and residents, who enter and exit the United States.” 8 U.S.C. § 1365b(k)(3)(A). The statute directs the Secretary to “ensure that the international registered traveler program includes as many participants as practicable,” by “establishing a reasonable cost of enrollment,” “making program enrollment convenient and easily accessible,” and “providing applicants with clear and consistent eligibility guidelines.” Id. § 1365b(k)(3)(E). The statute also directs the Secretary to “initiate a rulemaking to establish the program, criteria for participation, and the fee for the program.” Id. § 1356(k)(3)(C). Acting pursuant to this authority, DHS promulgated a rule, 8 C.F.R. § 235.7, which establishes the “PORTPASS” program. That program is designed to “provid[e] access to the United States for a group of identified, low-risk, border crossers.” 8 C.F.R. § 235.7(a)(1)(i). SENTRI is one of the programs established under the PORTPASS program. 75 Fed. Reg. 82202, 82202–03 (Dec. 29, 2010). Several provisions of the regulation are relevant here. First, the regulation establishes general eligibility criteria and application requirements for PORTPASS applicants. Applicants must be citizens or lawful permanent residents of the United States, or nonimmigrants determined to be eligible; must agree to furnish all information requested on the application; and must agree to terms set forth by CBP. 8 C.F.R. § 235.7(a)(3). Second, the regulation provides that an “application may be denied in the discretion of the district director having jurisdiction over the [Port-of-Entry (“POE”)] where the applicant requests access.” Id. § 235.7(a)(4)(x). The regulation also enumerates specific criteria that CBP “will . . . consider[]” when it evaluates a SENTRI applicant: 10 JAJATI V. U.S. CUSTOMS & BORDER PROT.

“admissibility to the United States and documentation so evidencing, criminal history and/or evidence of criminality, purpose of travel, employment, residency, prior immigration history, possession of current driver’s license, vehicle insurance and registration, and vehicle inspection.” Id. Notice of a denial must be given to the applicant. Id. Third, the regulation provides that access to a PORTPASS program may be “revoked at the discretion of the district director or the chief patrol agent,” if the participant (1) “violates any condition of the PORTPASS program,” (2) “violated any immigration law or regulation, or a law or regulation of the United States Customs Service or other Federal Inspection Service,” or (3) is “otherwise determined by an immigration officer to be inadmissible to the United States or ineligible to participate in PORTPASS.” Id. § 235.7(b).

III.

We review de novo a district court’s dismissal for lack of subject matter jurisdiction. Perez Perez, 943 F.3d at 860. A. The APA confers a cause of action upon persons “adversely affected or aggrieved by agency action.” Block v. Cmty. Nutrition Inst., 467 U.S. 340, 345 (1984) (quoting 5 U.S.C. § 702). Under the APA, a court shall “hold unlawful and set aside agency action, findings, and conclusions” that the court finds “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). There is a strong “presumption in favor of judicial review of final agency action” under the APA. Perez Perez 943 F.3d at 860; see Bowen v. Mich. Acad. of Fam. Physicians, 476 U.S. 667, 670 (1986) (noting the “strong

JAJATI V. U.S. CUSTOMS & BORDER PROT. 11

presumption that Congress intends judicial review of administrative action”). The presumption of judicial review can be overcome, however, if the challenged agency action is “committed to agency discretion by law.” 5 U.S.C. § 701(a)(2). As the Supreme Court has instructed, this exception to judicial review is read “quite narrowly.” Perez Perez, 943 F.3d at 860 (quoting Weyerhaeuser Co. v. U.S. Fish & Wildlife Serv., 568 U.S. 9, 23 (2018)). Section 701(a)(2) does not preclude judicial review of all discretionary decisions because the APA itself “command[s] that courts set aside agency action that is an abuse of discretion.” Dep’t of Commerce v. New York, 139 S. Ct. 2551, 2568 (2019); see 5 U.S.C. § 706(2)(A). Instead, agency action is “committed to agency discretion by law” only in “those rare instances where statutes are drawn in such broad terms that in a given case there is no law to apply, thereby leaving the court with no meaningful standard against which to judge the agency’s exercise of discretion.” 1 Pinnacle Armor, Inc. v. United States, 648 F.3d 708, 719 (9th Cir. 2011) (cleaned up) (first quoting Webster v. Doe, 486 U.S. 592, 599 (1988); and then quoting Heckler v. Chaney, 470 U.S. 821, 830 (1985)). Accordingly, we have held “an agency’s sole discretionary authority is not inconsistent with judicial

1 Even if a statute grants an agency unfettered discretion, an agency’s decision may “nonetheless be reviewed if regulations or agency practice provide a meaningful standard by which this court may review its exercise of discretion.” ASSE Int’l, Inc. v. Kerry, 803 F.3d 1059, 1069 (9th Cir. 2015) (internal quotation marks omitted) (quoting Spencer Enters., Inc. v. United States, 345 F.3d 683, 688 (9th Cir. 2003)). Hence, whether the meaningful standards derive from a statute or regulation is irrelevant to the question whether § 701(a)(2) bars judicial review. See Trout Unlimited v. Pirazdeh, 1 F.4th 738, 753 (9th Cir. 2021). 12 JAJATI V. U.S. CUSTOMS & BORDER PROT.

review of the agency’s exercise of that discretion.” Perez Perez, 943 F.3d at 863; see Beno v. Shalala, 30 F.3d 1057, 1066 (9th Cir. 1994) (“[T]he mere fact that a statute contains discretionary language does not make agency action unreviewable.”). Indeed, “courts routinely treat discretion- laden standards as providing ‘law to apply’” because “[e]ven if a determination is discretionary, it may still be rooted in a set of requirements or standards” that courts can use to assess whether an agency wielded its discretion in a permissible manner. Perez Perez, 943 F.3d at 862–63. “The fact that an agency has broad discretion in choosing whether to act does not establish that the agency may justify its choice on specious grounds.” Newman v. Apfel, 223 F.3d 937, 943 (9th Cir. 2000). As we have summarized this principle:

[W]here the [agency] has reserved to itself . . . certain decisions as within its “discretion,” or even its “sole discretion,” we will take into account the [agency’s] reservation and expertise and accord it the proper deference. But that does not deprive us of the right to review its actions for an abuse of its discretion or to determine if its actions were otherwise arbitrary and capricious.

ASSE, 803 F.3d at 1071 (citations omitted). This follows because judicial review under the APA concerns not only the particular outcome the agency reaches, but also the process in which the agency engages and the reasoning the agency articulates when it reaches that outcome. See, e.g., Dep’t of Commerce, 139 S. Ct. at 2575–

JAJATI V. U.S. CUSTOMS & BORDER PROT. 13

76 (holding an agency violated the APA when it reached a decision that may have otherwise been permissible, because the agency did not comply with the “reasoned explanation requirement of administrative law”). The fact that an agency reaches an outcome that could conceivably fall within its broad discretion is an issue entirely distinct from whether a court has jurisdiction under the APA to even consider whether the agency “justif[ied] its choice on specious grounds,” see Newman, 223 F.3d at 943, failed to satisfy “the general requirements of reasoned agency decisionmaking,” see Dep’t of Commerce, 139 S. Ct. at 2569, or “fail[ed] to comply with its own regulations,” see ASSE, 803 F.3d at 1069 (quoting Abdelhamid v. Ilchert, 774 F.2d 1447, 1450 (9th Cir. 1985)). See Trout Unlimited, 1 F.4th at 759 (distinguishing “reviewability” from the permissibility of “a particular outcome”). Indeed, the APA “calls for an explanation for agency action” that “can be scrutinized by courts and the interested public.” Dep’t of Commerce, 139 S. Ct. at 2575–76. This is precisely the way in which the APA promotes “political accountability, which itself is the very premise of administrative discretion in all its forms.” Newman, 223 F.3d at 943. It is therefore only in those rare instances in which there is “truly no law to apply” that § 701(a)(2) precludes a court from reviewing whether the agency abused its discretion or acted in an arbitrary and capricious manner. Perez Perez, 943 F.3d at 861 (internal quotation marks omitted). B. Applying these principles here, we conclude that the statute and regulation governing SENTRI provide meaningful standards with which we can review whether 14 JAJATI V. U.S. CUSTOMS & BORDER PROT.

CBP abused its discretion or acted in an arbitrary or capricious manner when it revoked Jajati’s SENTRI membership. See Pinnacle, 648 F.3d at 719; 5 U.S.C. § 706(A)(2). Hence, CBP’s decisions to revoke SENTRI memberships are not absolutely and entirely “committed to agency discretion by law.” See 5 U.S.C. § 701(a)(2). The district court erred when it concluded to the contrary. 1. CBP argues that its decisions to revoke SENTRI memberships are committed to its discretion because 8 C.F.R. § 235.7(b) provides that a membership may be “revoked at the discretion of” the agency if the agency “determine[s]” the person is “ineligible to participate” in SENTRI. According to CBP, there is no standard to review such a discretionary decision because the regulation does not define how an individual is determined to be “ineligible” to participate in SENTRI. We are unpersuaded. As we have explained, “[e]ven if a determination is discretionary, it may still be rooted in a set of requirements or standards” that enable judicial review. Perez Perez, 943 F.3d at 863. Here, the statute and regulation governing SENTRI provide such meaningful standards. First, the regulation establishes the goal of the program: to facilitate border crossing for a group of “low-risk[] border crossers.” 8 C.F.R. § 235.7(a)(1)(i). The enabling statute also requires that CBP “include[] as many participants as practicable by . . . providing applicants with clear and consistent eligibility guidelines.” 8 U.S.C. § 1365b(K)(3)(E). Finally, the regulation “establishes agency duties,” see Perez Perez, 943 F.3d at 863, including

JAJATI V. U.S. CUSTOMS & BORDER PROT. 15

the duty to consider certain criteria when CBP determines a person’s eligibility for SENTRI:

Criteria which will be considered in the decision to approve or deny the application include the following: admissibility to the United States and documentation so evidencing, criminal history and/or evidence of criminality, purpose of travel, employment, residency, prior immigration history, possession of current driver’s license, vehicle insurance and registration, and vehicle inspection.

8 C.F.R. § 235.7(a)(4)(x) (emphasis added). The regulation, in turn, lists criteria CBP will consider when it determines whether a person is ineligible for SENTRI. “The word ‘will,’ like the word ‘shall,’ is a mandatory term, unless something about the context in which the word is used indicates otherwise.” Nat. Res. Def. Council, Inc. v. Perry, 940 F.3d 1072, 1078 (9th Cir. 2019) (citation omitted) (quoting Washington v. Harper, 494 U.S. 210, 221 (1990)). Hence, the use of the term “will,” “prescribe[s] what the agency is required (or forbidden) to do,” and “unambiguously imposes a mandatory duty that constrains whatever discretion” CBP “might otherwise have possessed.” See id. at 1078–79. Thus, in determining whether a SENTRI membership can be revoked because a person is “otherwise . . . ineligible to participate,” CBP has constrained its discretion and required itself to consider, at least, the criteria outlined in 8 C.F.R. § 235.7(a)(4)(x). This is consistent with the enabling statute’s requirement that CBP shall “include[] as many 16 JAJATI V. U.S. CUSTOMS & BORDER PROT.

participants as practicable by . . . providing applicants with clear and consistent eligibility guidelines.” 8 U.S.C. § 1365(b)(K)(3)(E). Moreover, the regulation establishes the overarching objective of the program, which guides CBP when it applies the mandatory criteria: to facilitate border crossing for a group of “low-risk[] border crossers.” 2 8 C.F.R. § 235.7(a)(1)(i). We have made clear that such requirements and standards provide courts with “law to apply,” even if the standards are broad and the decision is otherwise discretionary. In Keating v. F.A.A., for example, a passenger pilot argued that the Federal Aviation Administration (“FAA”) violated the APA when it declined to grant him an exemption from a rule that made him ineligible to hold a pilot license after he turned 60-years-old. 610 F.2d 611, 612 (9th Cir. 1979). The sole standard for judging the agency’s discretion was a statute that provided the administrator could grant such an exemption “if he finds that such action would be in the public interest.” Id. We held that the agency’s action was not committed to agency discretion by law under § 701(a)(2) of the APA, because “the ‘public interest’ standard provides law to be applied by the administrator sufficient to permit judicial review.” Id. Similarly, in City of Los Angeles v. United States Department of Commerce, a group of municipalities sued the Department of Commerce after that agency had declined to “adopt statistically adjusted population data” for purposes of

2 Indeed, the revocation notice CBP sent Jajati in August 2021 provided that criminal convictions, pending criminal charges, outstanding warrants, or other circumstances “that indicate to CBP that you have not qualified as low risk” “may make you ineligible for participation” in

SENTRI.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 17

redistricting after the 2000 Census. 307 F.3d 859, 864 (9th Cir. 2002). The relevant statutory language stated only that “the Secretary shall, if he considers it feasible, authorize the use” of such statistical methods. Id. at 869. Again, we held the “Secretary’s discretion to ‘consider’ whether sampling is feasible does not defy meaningful judicial review” under the APA. Id. at 869 n.6. As we explained:

[A]lthough the phrase “if he considers it feasible” confers broad discretion, it has its limits. For example, if the Secretary had considered statistical adjustment but decided against it due to his political disinclinations, his decision would violate § 195. Thus, this is not a situation in which there is “no law to apply.”

Id. (citation omitted); see also Trout Unlimited, 1 F.4th at 756–57 (holding an agency’s failure to comply with its own regulations was reviewable because the agency “chose to constrain its discretion” by stating that the agency “shall” take certain actions in the regulation); Taslimi v. Holder, 590 F.3d 981, 985–86 (9th Cir. 2010) (holding a standard that required an alien to apply for asylum within a “reasonable period” of experiencing changed personal circumstances provided a meaningful standard because “the regulations themselves provide a non-exhaustive list of potential changed circumstances,“ and “refine[d] the standard by which the ‘reasonable period’ may be evaluated by requiring an adjudicator to consider” the alien’s awareness of his changed circumstances); Beno, 30 F.3d at 1066–67 (holding an agency’s decisions to waive federal welfare requirements that “in the judgment of the Secretary [were] likely to assist in promoting the [program’s] objectives,” were reviewable 18 JAJATI V. U.S. CUSTOMS & BORDER PROT.

because the objectives of the statute were “set forth with some specificity” in the statute); Pac. Nw. Generating Co- op v. Bonneville Power Admin., 596 F.3d 1065, 1076–77 (9th Cir. 2010) (holding a statutory requirement to operate in a manner “consistent with sound business principles” provided a meaningful standard). These cases establish that the standards and criteria enumerated in the SENTRI regulation provide law to apply which allows a court to judge whether CBP has abused its broad discretion in this case. The “low-risk” standard articulated in the SENTRI regulation is “at least as specific” as the public interest standard that we held to facilitate judicial review under the APA in Keating. See Bonneville, 596 F.3d at 1076–77 (comparing “sound business principles” to the “public interest” standard in Keating). And like the “if he considers it feasible” standard in City of Los Angeles, the “low-risk” standard articulated 8 C.F.R. § 235.7(a)(1)(i) “has its limits.” See City of Los Angeles, 307 F.3d at 869 n.6. 3 Particularly so given that, unlike in Keating or City of Los Angeles, there are mandatory criteria in the SENTRI regulation, 8 C.F.R. § 235.7(a)(4)(x), which provide “a partial adjudicative standard in and of” themselves by describing the type of information the agency deems relevant when it determines whether an individual is low risk. See Husyev v. Mukasey, 528 F.3d 1172, 1181 (9th Cir. 2008). CBP might abuse its discretion under the APA if, for example, it revoked a SENTRI membership solely

3 The dissent reasons that Keating and City of Los Angeles represent this Circuit’s “disappointing willingness to micromanage the discretionary affairs of administrative agencies.” Diss. Op. at 49. But as a three-judge panel, we are required to adhere to our Circuit authority in the absence of intervening, clearly irreconcilable authority. Miller v. Gammie, 335 F.3d 889, 900 (9th Cir. 2003) (en banc).

JAJATI V. U.S. CUSTOMS & BORDER PROT. 19

because a CBP agent has personal animosity toward the SENTRI member, because such a reason would be untethered from the criteria in the regulation and would have no bearing on whether the individual SENTRI member was a “low-risk[] border crosser.” 8 C.F.R. § 235.7(a)(1)(i). The law governing SENTRI, then, “do[es] not leave [CBP’s] discretion unbounded.” See Dep’t of Commerce, 139 S. Ct. at 2568. We therefore cannot agree with the dissent, which would hold that judicial review is unavailable in every case regarding SENTRI—regardless whether CBP had any valid reasons to revoke a SENTRI membership in a particular case—merely because CBP has broad discretion when it determines whether an individual is ineligible for SENTRI. 4 That is not the law of this Circuit, nor is it consistent with the purpose of judicial review under the APA. See Newman, 223 F.3d at 943.

4 The dissent, for example, posits that “the agency would be well within its rights to rely on the commonsense notion that association with felons is a red flag.” Diss. Op. at 45. Even if that is correct, the dissent “conflate[s] reviewability with a particular outcome.” See Trout Unlimited, 1 F.4th at 759. The mere fact that CBP could have exercised its broad discretion permissibly with respect to Jajati does not establish that we lack jurisdiction to review whether CBP failed to comply with its own regulation or revoked a SENTRI membership on specious grounds. See, e.g., Perez Perez, 943 F.3d at 863; ASSE, 803 F.3d at 1071; Pinnacle Armor, 648 F.3d at 720; Newman, 223 F.3d at 943. The dissent’s hypothetical—that CBP may have revoked Jajati’s membership due to his association with a felon—does not establish that the agency in fact relied on the reason the dissent proffers for the agency (which reasons, to date, CBP has never disclosed to Jajati or any court), nor does it establish that a court could never determine that CBP had relied on specious or arbitrary grounds in any case when it deems a person ineligible for SENTRI. 20 JAJATI V. U.S. CUSTOMS & BORDER PROT.

2. CBP, as well as the dissent, maintain that there is no law to apply because the criteria are undefined and non- exhaustive, and there is nothing with which a court can surmise which factors (or combination of factors) would suffice to support either approval or denial of a SENTRI application. Diss. Op. at 32–33, 43–44. To the contrary, the weighing of non-exhaustive criteria like those in the SENTRI regulation is squarely within the province of the judiciary, even if the regulation involves a “flexible standard that draws considerably on the agency’s expertise and judgment.” See Trout Unlimited, 1 F.4th at 759; accord Gonzalez-Caraveo v. Sessions, 882 F.3d 885, 891–92 (9th Cir. 2018) (holding “six descriptive, though non-exhaustive factors,” such as “the reason administrative closure is sought,” alone provided “a ‘sufficiently meaningful standard’ by which to evaluate” the IJ or BIA’s administrative closure decisions, because, “[a]lthough some of the . . . factors . . . may be issues with which an IJ is more familiar based on his or her experience, [they] are not so unique to the agency that this Court would be unable to evaluate them with the assistance of the parties’ briefing”). For example, in Newman, we held that the Social Security Commissioner’s decision to refrain from increasing a beneficiary’s Supplemental Social Security Income (“SSI”) benefits was not committed to agency discretion by law. 223 F.3d at 942–43. The statute directed the Commissioner to promulgate a rule that enabled him to increase a person’s benefits so long as he used information about the beneficiary’s financial status that was “reliable” and “currently available.” Id. at 939. The Commissioner had discretion to “determine[]” whether specific information offered by a beneficiary was indeed reliable and currently

JAJATI V. U.S. CUSTOMS & BORDER PROT. 21

available. Id. He also had discretion to refrain from increasing a person’s benefits even if such information existed. Id. The Commissioner promulgated a rule that defined the terms “reliable” and “currently available,” but concluded that “no reliable information exists which is currently available,” such that the Commissioner would not increase any benefits under the provision. Id. at 940 (citing 20 C.F.R. § 416.420(c)). A recipient whose benefits were not increased despite a change in financial status challenged the regulation under the APA. Id. at 940–42. The Commissioner claimed his decision was committed to agency discretion by law, because (1) there were no standards by which to define the terms “reliable” and “currently available,” and (2) even if there were, no review was available because “even if [he] determines that such information does exist, he can choose not to use it.” Id. at 943 (alteration in original). We disagreed. Id. We held that the interpretation and application of the terms “reliable” and “currently available” did not defy review because the “basic definition and application of those terms” did not “involve[] a complicated balancing of a number of factors that are so peculiarly within the agency’s expertise that jurisdiction is necessarily defeated.” Id. And as to the agency’s discretion to discount even reliable and currently available information, we similarly rejected the claim that such discretion foreclosed judicial review. Id. Rather, we had jurisdiction under the APA to review the Commissioner’s determination that no reliable or currently available information existed. Id. Here, as in Newman, we can review CBP’s application of the criteria, even if CBP can rely on some other, unenumerated factors when it deems a person ineligible for 22 JAJATI V. U.S. CUSTOMS & BORDER PROT.

SENTRI. See id. Similarly, as in Newman—where we held judicial review was available even though the agency had discretion to “determine” whether specific information offered by a beneficiary was reliable and currently available—the mere fact that CBP has discretion to “determine” which information satisfies the various criteria does not establish that we lack jurisdiction to consider whether CBP applied those criteria on specious grounds. See id. Indeed, even more so than the “reliable” and “currently available” criteria in Newman, the interpretation and application of the eligibility criteria in 8 C.F.R. § 235.7(a)(4)(x), such as employment status and criminal history, are not so “peculiarly within the agency’s expertise that jurisdiction is necessarily defeated.” 5 See Newman, 223

5 Federal courts frequently evaluate such criteria in other contexts. For example, district courts may evaluate the nature of a criminal defendant’s criminal history when computing his offense level under the Sentencing Guidelines. See, e.g., U.S.S.G. § 4A1.3; United States v. Henderson, 993 F.2d 187, 189 (9th Cir. 1993) (explaining that the Sentencing Guidelines “permit[] a sentencing court to consider upward departure when a defendant’s criminal history category does not adequately reflect the seriousness of the defendant's past criminal conduct”). So too with respect to an individual’s employment status when an agency denies his eligibility for unemployment benefits. See, e.g., N.L.R.B. v. Adrian Belt Co., 578 F.2d 1304, 1308–09 (9th Cir. 1978) (reviewing for substantial evidence an agency’s finding that a claimant had retained her employment status, despite a leave of absence, when the agency denied the individual’s claim for unemployment benefits). The dissent reasons that the criteria in Henderson and Adrian Belt were better defined, see Diss. Op. at 40–42, but the point is that these factors are not so unfamiliar to the judiciary so as to evade our review. Indeed, the primary example of a discretionary decision that involved “a complicated balancing of a number of factors which are peculiarly within [the agency’s] expertise” came in Heckler, which involved a challenge to the FDA’s decision not to enforce its drug safety regulations. Heckler, 470 U.S. at 831–33. Nothing in our decision encroaches on CBP’s ability to choose whether

JAJATI V. U.S. CUSTOMS & BORDER PROT. 23

F.3d at 943. Although the criteria are non-exhaustive and undefined, the criteria are still “descriptive,” see Gonzalez- Caraveo, 882 F.3d at 892, and “refine the [low-risk] standard,” see Taslimi, 590 F.3d at 986, by elucidating the type of information that is relevant to determine which individuals are “low-risk” so as to qualify for SENTRI.6 Accordingly, we can review whether CBP abused its discretion, or acted in an arbitrary and capricious manner, when it weighed those criteria and determined that an individual was ineligible to participate in SENTRI because he was not a low-risk border crosser. Moreover, because the regulation imposes mandatory duties on CBP to consider the criteria, we can review whether the agency failed to consider those criteria and, in doing so, “fail[ed] to comply with its own regulations.” See ASSE, 803 F.3d at 1069 (quoting Abdelhamid, 774 F.2d at 1450). For example, in Perez Perez, the plaintiff challenged a United States Citizenship and Immigration Services (“USCIS”) decision to deny him a U Visa. Perez Perez, 943

to initiate revocation proceedings against an individual SENTRI member. See Gonzalez-Caraveo, 882 F.3d at 892. Once the agency chooses to act, however, its subsequent actions are governed by the standards in the SENTRI regulation, which are refined by the criteria elucidated in the regulation. See id. at 893; Newman, 223 F.3d at 943. 6 For this reason, the dissent’s assertion that we cannot review CBP’s decision because it is “tasked with applying the regulatory standards to the underlying question of eligibility” is unconvincing. Diss. Op. at 43. The same was true in Gonzalez-Caraveo, in which we held judicial review was available because the agency was tasked with applying “six descriptive, though non-exhaustive factors” when it determined the underlying question “whether administrative closure [was] appropriate in a given case.” 882 F.3d at 891–92. 24 JAJATI V. U.S. CUSTOMS & BORDER PROT.

F.3d at 856. An alien is eligible for a U Visa if he (1) has suffered physical or mental abuse as a result of being a victim of a qualifying crime, (2) possesses information about the crime, and (3) has been helpful in investigating the crime. Id. at 862–63 (quoting 8 U.S.C. § 1101(a)(15)(U)(i)). The U Visa statute establishes application procedures and requires the agency to “consider any credible evidence relevant to the petition.” Id. at 863 (quoting 8 U.S.C. § 1184(p)(3)–(4)). The statute, however, provides that the agency “determines” whether an alien has met the eligibility requirements. Id. The regulation, moreover, gives USCIS “sole discretion” to weigh evidence and provides that USCIS has “sole jurisdiction” over U Visa petitions. Id. The plaintiff argued that USCIS violated the APA in part because it failed to consider all credible evidence and made an erroneous factual finding when it denied him a U Visa. Id. at 864–65. USCIS claimed its decision was committed to agency discretion by law. Id. at 860. We held in Perez Perez that § 701(a)(2) of the APA did not bar review of the plaintiff’s claim. Id. at 863. We recognized that a U Visa “determination is discretionary,” but reasoned that discretion was “rooted in a set of requirements or standards” that enabled judicial review. Id. at 863. For example, the statute’s mandate that USCIS “consider any credible evidence relevant to the petition provides a meaningful standard by which to review” whether the agency failed to consider such evidence. Id. at 864 (citation omitted). And we held that we could still review a claim that the agency made an erroneous factual finding, even though the regulation gave the agency “sole discretion” to weigh evidence. Id. at 865 (citing ASSE, 803 F.3d at 1071).

JAJATI V. U.S. CUSTOMS & BORDER PROT. 25

Here too, the SENTRI regulations establish mandatory criteria that CBP must consider when it evaluates whether a person is “ineligible” for SENTRI. 8 C.F.R. §§ 235.7(a)(4)(x), 235.7(b). Just as in Perez Perez, where we held that we could review a claim that agency had “fail[ed] to consider all credible evidence” as required by the statute—even though “the determination of what is relevant evidence and the weight to be given to that evidence” was within the “sole discretion” of the agency—so too here may we review whether CBP failed to consider the required the criteria or abused its discretion when it evaluated those criteria. See Perez Perez, 943 F.3d at 864. Indeed, as we have explained, the objective of the program—to identify “low- risk” border crossers—offers a guidepost for courts to evaluate whether CBP applied the mandatory criteria on permissible grounds. See 8 C.F.R. § 235.7(a)(1)(i); Perez Perez, 943 F.3d at 863–64; Gonzalez-Caraveo, 882 F.3d at 892. Hence, although the criteria in the SENTRI regulation are non-exhaustive and undefined, CBP’s discretion to revoke a SENTRI membership is “rooted in a set of requirements or standards” by which courts can review whether CBP abused its discretion or acted in an arbitrary or capricious manner. See Perez Perez, 943 F.3d at 863; Newman, 223 F.3d at 942–43; Gonzalez-Caraveo, 882 F.3d at 892. 3. In arguing to the contrary, the dissent directs us to cases involving statutes and regulations that provided no criteria to constrain an agency’s discretion. But those cases illustrate just how rarely agency action is unreviewable under § 701(a)(2). 26 JAJATI V. U.S. CUSTOMS & BORDER PROT.

The dissent relies on Ekimian v. INS, 303 F.3d 1153, 1156–59 (9th Cir. 2002), where a group of aliens sought review of a BIA decision to refuse to reopen removal proceedings sua sponte. Diss. Op. at 47. The statute provided that the BIA “may at any time reopen or reconsider on its own motion any case in which it has rendered a decision.” Id. at 1156 (emphasis omitted). The regulation provided that the BIA could reopen sua sponte “in unique situations where it would serve the interest of justice.” Id. at 1157 (emphasis omitted). The only other guidance came from an agency adjudication in which the BIA explained that the regulation allowed it to reopen in “exceptional situations.” Id. at 1158. We held the BIA’s refusal to reopen proceedings was unreviewable because there was not “even a discretionary standard” to apply to determine whether exceptional circumstances existed. 7 Id. at 1157. Specifically, the relevant statute and regulation did not state when the BIA could sua sponte reopen proceedings, and the BIA had never elucidated “what constitute[s] ‘exceptional situations.’” Id. at 1158. We explained that the phrase, “‘exceptional situations,’ without more,” did not provide a meaningful standard of review. Id. (emphasis added). Only because there were no criteria—at all—by which we could measure the BIA’s exercise of discretion did we lack jurisdiction in Ekimian. Id.; see also Idrees v. Barr, 923 F.3d 539, 542–43 (9th Cir. 2019) (holding the BIA’s decision to refuse to certify a claim of ineffective assistance of counsel was

7 We later clarified that we may review BIA decisions denying sua sponte reopening “for the limited purpose of reviewing the reasoning behind the decisions for legal or constitutional error” because, in that instance, there is “law to apply.” Bonilla v. Lynch, 840 F.3d 575, 588 (9th Cir. 2016). Ekimian, then, is limited to the BIA’s determination as to whether “there were truly exceptional circumstances.” Id.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 27

unreviewable because the BIA would certify such claims only in “exceptional circumstances” and “no other regulation or statute provide[d] guidance” to determine “which circumstances are considered to be exceptional”); 8 Diaz-Covarrubias v. Mukasey, 551 F.3d 1114, 1117–19 (9th Cir. 2009) (holding judicial review was unavailable with respect to the BIA’s decision to deny a request for administrative closure where the only standard applied to the decision was “administrative convenience”). In contrast, when the term “extraordinary circumstances” is elucidated with criteria in a regulation, we have held that we can review a discretionary determination under that capacious standard. See, e.g., Husyev, 528 F.3d at 1180–81 (holding a statute that allowed an alien to file a late asylum application under “extraordinary circumstances” was not rendered standardless because the “regulations set out a non-exhaustive list of six potentially qualifying ‘extraordinary circumstances’” that provided “a partial adjudicative standard in and of itself”). Here, in contrast to Ekimian and Idrees, in which the guidelines “exceptional situations” and “exceptional circumstances”—without more—had no objective reference point, the SENTRI regulation offers several mandatory, objective criteria and standards with which courts are quite

8 The dissent reasons that in Idrees, there were criteria to guide the BIA’s decision: namely, that the BIA could reopen “if it determines that the parties have already been given a fair opportunity . . . regarding the case, including the opportunity to request oral argument and to submit a brief.” Diss. Op. at 47–48 (quoting Idrees, 923 F.3d at 542–43). But the “criteria” the dissent cites had no bearing as to which circumstances were “exceptional.” Here, by contrast, the SENTRI regulation describes factors that speak directly to whether an individual is low risk. Cf. Gonzalez-Caraveo, 882 F.3d at 892. 28 JAJATI V. U.S. CUSTOMS & BORDER PROT.

familiar. Those criteria and standards provide “a partial adjudicative standard in and of [themselves],” and are therefore sufficient to make CBP’s discretionary decision subject to judicial review. See Husyev, 528 F.3d at 1181; Gonzalez-Caraveo, 882 F.3d at 892. In sum, our precedent forecloses the dissent’s application of § 701(a)(2) to this case. The SENTRI regulation, 8 C.F.R. § 235.7, provides sufficient “standards” and “requirements” which enable courts to review whether CBP, in fact, considered the criteria as required by the regulation, see Perez Perez, 943 F.3d at 863. The regulation’s objective— to identify “low-risk” border crossers—also provide a standard by which we can review whether CBP abused its discretion, or acted in an arbitrary and capricious manner, when it interpreted and applied those criteria with respect to an individual SENTRI member and deemed that person “ineligible” to participate in SENTRI. See Gonzalez- Caraveo, 882 F.3d at 892; Newman, 223 F.3d at 943. Hence, the “strong presumption” in favor of judicial review under the APA has not been rebutted. 9 See Bowen, 476 U.S. at 670.

9 The dissent raises the policy concern that our decision will “force the agency to divulge its reasons for revoking memberships . . . some of which might pertain to sensitive matters of national security.” Diss. Op. at 55–56. But such “a weak connection to foreign policy is not enough to commit an agency action to the agency’s discretion.” ASSE, 803 F.3d at 1069. Were Congress so concerned with keeping secret the agency’s reasons for determining eligibility for SENTRI, it would not have required the agency to provide applicants with “clear and consistent eligibility guidelines.” 8 U.S.C. § 1365b(k)(3)(E). Moreover, Congress retains the right to “preclude judicial review” by statute. 5 U.S.C. § 701(a)(1).

JAJATI V. U.S. CUSTOMS & BORDER PROT. 29

C. Having established the availability of judicial review, we turn to the standard the district court should apply on remand. For this standard, we look to § 706 of the APA, under which we assess whether the revocation of an individual’s SENTRI membership was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,” or in violation of a statutory, procedural, or constitutional requirement. 5 U.S.C. § 706(2). In making this determination, the district court should consider whether CBP failed to consider the criteria in 8 C.F.R. § 235.7(a)(4)(x) when it determined Jajati was ineligible to participate in SENTRI. See Perez Perez, 943 F.3d at 864. The district court should also assess whether CBP abused its discretion, or acted in an arbitrary and capricious manner, when it interpreted and applied those criteria to Jajati in light of the overarching objectives of the program: to facilitate border crossing for low-risk border crossers, 8 C.F.R. § 235.7(a)(1)(i), and to ensure the program “includes as many participants as practicable by . . . providing applicants with clear and consistent eligibility guidelines,” 8 U.S.C. § 1365b(k)(3)(E). See Newman, 223 F.3d at 943. Moreover, the district court should review CBP’s factual findings under the “substantial evidence” standard. See ASSE, 803 F.3d at 1072 (“‘[A]s a practical matter, the arbitrary and capricious standard incorporates the substantial evidence test,’ and we use that test for review of agency factfinding in informal proceedings as well.”) (quoting Ursack Inc. v. Sierra Interagency Black Bear Grp., 639 F.3d 949, 958 n.4 (9th Cir. 2011)); see also McLean v. Morgan, 2020 WL 5094683, at *6–7 (D. Kans. Aug. 28, 2020) (remanding to the agency to reconsider revocation of a 30 JAJATI V. U.S. CUSTOMS & BORDER PROT.

Global Entry membership because substantial evidence did not support a finding that the plaintiff had multiple convictions). Substantial evidence review “is generally confined to a review of the administrative record.” See Perez Perez, 943 F.3d at 865; see also 5 U.S.C. § 706 (“[T]he court shall review the whole record or those parts of it cited by a party . . . .”). Finally, in evaluating Jajati’s APA claim, the district court should “give due deference to the agency’s expertise.” Perez Perez, 943 F.3d at 865; ASSE, 803 F.3d at 1071 (“[W]e will take into account the [agency’s] reservation and expertise and accord it the proper deference.”). With this said, the agency may not “justify its choice on specious grounds.” Newman, 223 F.3d at 943.

IV.

We hold that § 701(a)(2) of the APA does not bar judicial review of Jajati’s claim. Hence, we reverse the district court’s order which granted CBP’s motion to dismiss for lack of subject matter jurisdiction. We remand with instructions that the district court consider Jajati’s APA claim on the merits in the first instance.

REVERSED AND REMANDED.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 31

VANDYKE, Circuit Judge, dissenting:

Not every enabling statute or regulation provides a “meaningful standard against which to judge the agency’s exercise of discretion.” Spencer Enterprises, Inc v. United States, 345 F.3d 683, 688 (9th Cir. 2003) (quoting Heckler v. Chaney, 470 U.S. 821, 830 (1985)). Courts should not assume, as the majority does today, that the mere existence of criteria an agency must consider before acting definitively answers the question of whether there is “law to apply” on review, id., because applying such criteria often requires “complicated balancing” that is “peculiarly within the agency’s expertise.” Newman v. Apfel, 223 F.3d 937, 943 (9th Cir. 2000). Even where an agency is required by law to consider certain criteria before acting, those standards might very well be all bark and no bite, leaving the decision “committed to agency discretion by law.” 5 U.S.C. § 701(a)(2). This case demonstrates the truth of that proposition. SENTRI’s enabling regulations require United States Customs and Border Protection to consider nine listed criteria, see 8 C.F.R. § 235.7(a)(4)(x), when determining whether a SENTRI applicant is sufficiently “low risk” to qualify for the privilege of expedited screening at the United States–Mexico border, id. § 235.7(a)(1)(i). 1 The majority concludes that because these criteria are “mandatory” and “objective,” they are “sufficient to make CBP’s discretionary decision subject to judicial review.” I disagree both with the majority’s characterization of the criteria and with its conclusion that they provide a

1 Like TSA precheck, which travelers are perhaps more familiar with, SENTRI membership is a privilege, not a right. 32 JAJATI V. U.S. CUSTOMS & BORDER PROT.

meaningful standard against which to judge the agency’s revocation decisions. No matter whether the criteria are “mandatory” (they are, but they aren’t exhaustive) or “objective” (they aren’t), they ultimately provide “no law to apply” on review for several reasons. First, the criteria are neither meaningfully defined nor self-evident. Second, the regulations do not indicate the “right” or “wrong” way to apply each factor when assessing an applicant’s risk. And third, there is no guidance as to how to weigh the factors against one another or what combination of factors might suffice to establish eligibility. Because the criteria do not constrain the agency’s discretion, they do not meaningfully facilitate judicial review. How can a reviewing court grade the agency’s papers when the agency’s enabling regulations don’t provide an answer key? It can’t. What will inevitably happen here is what always happens when courts step in to “judge” the executive branch’s highly discretionary decisions: those decisions end up being mostly governed by the judges’—not the executive branch officials’—discretion. Even if the regulation’s criteria did provide justiciable guidance, the majority faces another problem. Because there is no indication the listed criteria are exhaustive, CBP remains free to consider an infinite number of other, unenumerated factors when making its eligibility determinations. So even if a court were to consider how the agency applied each of the listed criteria to Jajati’s case, it would be no closer to discovering whether the CBP abused its discretion by revoking his SENTRI membership. Given the nebulous, incomplete nature of the listed criteria, I cannot help but conclude that SENTRI eligibility is exactly the kind of discretionary policy decision insulated

JAJATI V. U.S. CUSTOMS & BORDER PROT. 33

from judicial review by § 701(a)(2) of the APA. The majority’s contrary conclusion further “devalues the separation of powers” and invites judicial henpecking of the agency’s countless, discretionary, and national security- laden decisions about who gets a fast pass across our hopelessly busy Southern border—a topic about which courts can claim no particular expertise. Perez Perez v. Wolf, 943 F.3d 853, 868 (9th Cir. 2019) (Callahan, J., dissenting). Because SENTRI eligibility determinations are quite sensibly “committed to agency discretion by law,” 5 U.S.C. § 701(a)(2), I respectfully dissent. I. The majority begins its analysis by defending a proposition that neither the district court nor any of the parties seem to contest: the “mere fact that a statute contains discretionary language does not make agency action unreviewable.” Beno v. Shalala, 30 F.3d 1057, 1066 (9th Cir. 1994). I agree with that uncontroversial and well-settled reading of the APA. As the Supreme Court has recognized, it cannot be the case that every discretionary act is “committed to agency discretion by law” because the APA itself contemplates judicial review of agency action for “abuse of discretion.” See Heckler, 470 U.S. at 829 (quoting 5 U.S.C. § 706(2)(A). Instead, as the majority correctly notes, the proper test is whether the relevant statutes and regulations “are drawn in such broad terms that in a given case there is no law to apply, thereby leaving the court with no meaningful standard against which to judge the agency’s exercise of discretion.” Pinnacle Armor, Inc. v. United States, 648 F.3d 708, 719 (9th Cir. 2011). As this court is prone to do, however, the majority errs by so overstating the stringency of the standard as to render 34 JAJATI V. U.S. CUSTOMS & BORDER PROT.

“th[e] exception without import or content.” Trout Unlimited v. Pirzadeh, 1 F.4th 738, 761 (9th Cir. 2021) (Bress, J., dissenting). For one thing, while I agree that the existence of discretionary language is not itself dispositive, such language “certainly places additional weight on that side of the scale.” City & County of San Francisco v. U.S. Dep’t of Transp., 796 F.3d 993, 1003 (9th Cir. 2015). Indeed, the SENTRI regulations are “peppered with the classic language of discretion.” Id. at 1002. 2 And while it is certainly true that § 701(a)(2) is read “quite narrowly” and that courts characterize its application as “rare,” Perez Perez, 943 F.3d at 860, “rare” does not mean “impossible.” Unsurprisingly, both the Supreme Court and this court have applied the § 701(a)(2) exception with some regularity to insulate discretionary agency action from judicial review. See Pirzadeh, 1 F.4th at 761 (Bress, J., dissenting) (collecting a long string cite of cases). For further support, the majority also invokes Department of Commerce v. New York, 139 S. Ct. 2551 (2019), in which the Supreme Court explained that “the reasoned explanation requirement of administrative law” “is meant to ensure that agencies offer genuine justification for important decisions” that “can be scrutinized by courts and

2 See 8 C.F.R. § 235.7(a)(4)(x) (“An application may be denied in the discretion of the district director having jurisdiction over the POE where the applicant requests access.”); id. (“There is no appeal from the denial.”); id. § 235.7(b) (“A PORTPASS program participant … who is otherwise determined by an immigration officer to be … ineligible to participate in PORTPASS, may have the PORTPASS access revoked the discretion of the district director or the chief patrol agent.”); id. § 235.7(c) (“Nothing in this section is intended to create any right … enforceable in law or equity by a party against the [agency].”) (emphasis added throughout).

JAJATI V. U.S. CUSTOMS & BORDER PROT. 35

the interested public.” Id. at 2575–76. But applying that principle here to conclude that the SENTRI regulations are amenable to judicial review begs the question. In Department of Commerce, the Supreme Court invoked “the reasoned explanation requirement of administrative law” only after it had rejected the notion that the challenged agency action—“[t]he taking of the census”—was “committed to agency discretion.” Id. at 2568. The majority’s suggestion that every agency decision must be conducive to “scrutin[y] by courts and the interested public” is fundamentally inconsistent with the APA’s recognition that some actions are “committed to agency discretion by law.” 5 U.S.C. § 706(2)(A). 3 But the majority’s most serious errors stem from its heavy reliance on 8 C.F.R. § 235.7(a)(4)(x), which provides that an applicant’s “[1] admissibility to the United States and

3 The majority’s policy concern that SENTRI memberships might be revoked solely because “a CBP agent has personal animosity toward the SENTRI member” is similarly circular. It cannot be the case that we have jurisdiction to review every potential abuse of discretion because the potential for unreviewed abuses of discretion arises every time a decision is committed to agency discretion by law. The majority’s jurisdictional rule essentially boils down to “bad things might happen if courts can’t intervene.” But that approach is inconsistent with our role as courts of limited jurisdiction. If the majority’s approach to 5 U.S.C. § 706(2)(A) wins out, then nothing will ever be committed to agency discretion by law, and we will always have jurisdiction. Courts “do not, or should not, sally forth each day looking for wrongs to right.” United States v. Sineneng-Smith, 590 U.S. 371, 376 (2020). Nor are we “a small group of fortunately situated people with a roving commission to second- guess Congress … concerning what is best for the country.” William H. Rehnquist, The Notion of a Living Constitution, 54 TEX. L. REV. 693, 698 (1976). If, as here, a decision is committed to agency discretion by law, we ought to leave it at that, even if agency decisionmakers aren’t perfect. We certainly aren’t. 36 JAJATI V. U.S. CUSTOMS & BORDER PROT.

documentation so evidencing, [2] criminal history and/or evidence of criminality, [3] purpose of travel, [4] employment, [5] residency, [6] prior immigration history, [7] possession of current driver’s license, [8] vehicle insurance and registration, and [9] vehicle inspection” “will be considered in the decision to approve or deny the application” (emphasis added). The majority makes a series of related errors regarding these criteria, all of which undermine its contention that the SENTRI regulations provide “law to apply” on review. First, citing Perez Perez, the majority repeatedly implies that a discretionary determination is per se reviewable whenever it is “rooted in a set of requirements or standards.” 943 F.3d at 863. But Perez Perez does not stand for that far- reaching proposition, and our caselaw elsewhere gives no indication that the mere presence of statutory or regulatory criteria, standing alone, is somehow a dispositive answer to the question of reviewability. Instead, as Perez Perez itself makes clear, the question remains whether the requirements or standards, if they exist, meaningfully constrain the agency’s discretion to act, thus “furnish[ing] meaningful standards” on review. Id. at 862. The connection between meaningful constraint and meaningful review is entirely intuitive. To effectively review agency action for abuse of discretion, a court must be able to avail itself of some discernible standard that imposes a limitation on the agency’s ability to act. No discernible standard, no real constraint. No real constraint, no basis to declare agency actions unlawful. In other words, if the regulatory standards imposed on an agency do not provide the court with any insight into what an agency should have done (or should have done differently), then those standards are essentially a paper tiger. They offer a reviewing court

JAJATI V. U.S. CUSTOMS & BORDER PROT. 37

nothing useful in its endeavor to decide whether the agency acted arbitrarily, unlawfully, or in a manner that abused its discretion. Under such circumstances, who are we to intervene, and even assuming we do, what legal basis do we have to conclude the agency acted unlawfully? For these reasons, the mere fact that certain criteria are listed, without more, is insufficient. 4 Those criteria must meaningfully constrain the agency’s discretion to provide “law to apply” on review.

II.

Perhaps sensing this truth, the majority also contends that the factors listed in 8 C.F.R. § 235.7(a)(4)(x) “offer[] several mandatory, objective criteria” that the agency must consider when making an ultimate determination as to whether an applicant is sufficiently “low-risk” to qualify for SENTRI membership. See 8 C.F.R. § 235.7(a)(1)(i). In the majority’s view, the combination of these criteria with the regulatory objective to facilitate “low-risk” crossings provides a meaningful standard that will facilitate judicial review. I disagree both with the majority’s characterization of the criteria as “mandatory” and “objective,” and with its

4 For anyone doubting the notion that an agency may be required to consider a statutory or regulatory standard that nevertheless fails to meaningfully constrain its discretion, consider a hypothetical statute providing that “when acquiring new buildings, the General Services Administration shall consider the building’s architectural style.” Would such a statute authorize judges to police the federal government’s well- known, much-maligned penchant for brutalist architecture that is intentionally, soul-crushingly ugly? See Theodore Dalrymple, The Architect as Totalitarian, City Journal (Nov. 19, 2009), https://www.city-journal.org/article/the-architect-as-totalitarian. As much as I might wish otherwise: of course not. 38 JAJATI V. U.S. CUSTOMS & BORDER PROT.

conclusion that these criteria provide a meaningful standard enabling judicial review. A. First, consider the majority’s characterization of the criteria as “objective.” That description adequately describes some of the listed factors, including, for example, “admissibility to the United States,” “possession of [a] current driver’s license,” “vehicle insurance and registration,” and “vehicle inspection.” These checkbox criteria require little to no exercise of discretion or analysis from the agency. An applicant’s inadmissibility to the United States obviously means that he should not receive an expedited path across the border. Likewise, an individual who lacks a valid license, insurance, registration, and inspection shouldn’t be driving a vehicle at all, let alone driving one across the border. These few checkbox criteria may provide a brightline and arguably justiciable basis to deny someone membership in the SENTRI program. But the fact that just these baseline criteria are satisfied cannot provide a justiciable basis to insist that they must be given such membership. In short, these checkbox criteria will always matter little in any case like this where someone like Jajati is complaining about his denial in the program. If they aren’t met, then obviously the agency didn’t abuse its discretion in denying membership. If they are met, that likewise means pretty much nothing—those basic criteria are met by the overwhelming majority of all the people who drive across the southern border every day, most of whom are not in the SENTRI program. All the other listed criteria that are not effectively checkboxes cannot be characterized as “objective” in any

JAJATI V. U.S. CUSTOMS & BORDER PROT. 39

meaningful sense, including an applicant’s “purpose of travel,” “employment,” “residency,” and “criminal history.” These criteria are not “objective” because the proper way to apply them to the overarching assessment of risk is not readily apparent, nor does Jajati “point to any statutory, regulatory, or caselaw definition” that might assist a reviewing court in making sense of them. Ekimian v. I.N.S., 303 F.3d 1153, 1159 (9th Cir. 2002). 5 How, for example, should the agency assess the purpose of an applicant’s travel? Are beachgoers bound for Baja less risky than truckers transporting goods from Tijuana? If so, why? What regulatory justification could the court possibly provide for that conclusion? And what about employment? As far as I can see, nothing in the SENTRI regulations constrains the agency from deciding that a given profession is inherently riskier than another, and thus the mere fact that “employment” is a consideration does not meaningfully facilitate judicial review. Similar examples could be drawn up for “residency” and “criminal history” ad nauseum. The point is that the listed criteria are not “objective” because they are not meaningfully defined, and how the agency should go about applying them is thus entirely unclear. The majority takes a different view, asserting that assessing “criteria like those in the SENTRI regulation is squarely within the province of the judiciary.” But saying so

5 Even the majority is forced to repeatedly acknowledge that the criteria are “undefined.” 40 JAJATI V. U.S. CUSTOMS & BORDER PROT.

does not make it so, and the cases it cites undermine that contention instead of supporting it. The majority first cites United States v. Henderson for the proposition that courts regularly consider the “seriousness of the defendant’s past criminal conduct” when “consider[ing] upward departure.” 993 F.2d 187, 189 (9th Cir. 1993). While that’s true, there are at least three flaws with the suggested analogy. First, it “is a longstanding tradition in American law” that sentencing courts “exercise a wide discretion in the sources and types of evidence used to craft appropriate sentences.” Concepcion v. United States, 597 U.S. 481, 486 (2022) (emphasis added, citations and internal quotation marks omitted). If anything, the “wide discretion” afforded courts in assessing criminal history when deciding a sentence underscores the discretion afforded the agency here. Second, it is not all apparent that a court deciding an appropriate sentence, which presumably accounts not only for a defendant’s risk of recidivism, but also for his culpability and society’s retributive interests, is performing a function that is even remotely analogous to the agency’s assessment of a SENTRI applicant’s risk. Third, and most importantly, even if the analogy were sound, the sentencing guidelines offer courts considerably more guidance regarding how to assess a defendant’s criminal history than 8 C.F.R. § 235.7(a)(4)(x)’s passing reference to “criminal history” does. See U.S.S.G. § 4A1.3(a)(2)(A)–(E). Indeed, the guidelines list specific examples—including “foreign and tribal convictions,” “prior similar misconduct established by a civil adjudication,” and “conduct not resulting in a criminal conviction”—justifying upward departure. See id. Unsurprisingly, Henderson relies on comparison to these examples—not its vague appraisal of the severity of the

JAJATI V. U.S. CUSTOMS & BORDER PROT. 41

appellant’s “criminal history” more generally—to conclude that an upward departure from the guidelines was unjustified. See 993 F.2d at 189. 8 C.F.R. § 235.7(a)(4)(x) includes no framework to the sentencing guidelines that would afford a reviewing court any explanation of what the SENTRI guidelines mean by “criminal history,” “employment,” or “purpose of travel.” The majority’s second case, NLRB v. Adrian Belt Co., 578 F.2d 1304 (9th Cir. 1978), is no more compelling than the first. In Adrian Belt, the court affirmed the Labor Board’s conclusion that an employee who had taken a leave of absence was still properly classified as an employee of the defendant company at the time of a union election despite the leave. Id. at 1307–08. The majority cites this case as evidence that courts are well acquainted with the task of analyzing an employee’s employment status. While that may very well be true, it is ultimately irrelevant. This court’s ability to determine whether people are employed says nothing about its ability to assess how their employment might affect the risks they pose when crossing the southern border, which is obviously a very different question. 6

6 As should be sufficiently clear from the analysis above, my point is not just that, in the majority’s words, “the criteria in Henderson and Adrian Belt were better defined.” While that is true, my point is that those cases employ the “criminal history” and “employment” standards in much more direct—and more traditionally judicial—ways. This court’s ability to assess criminal history when crafting a criminal sentence or to assess the mere fact of employment status when deciding a labor dispute says nothing about its ability to assess how those factors affect the risk a traveler poses at the border. It is this latter task that courts face in a case like this one. For that reason, neither Henderson nor Adrian Belt get the majority any closer to proving that courts have competency to deal with the criteria listed in 8 C.F.R. § 235.7(a)(4)(x) in a way that would render 42 JAJATI V. U.S. CUSTOMS & BORDER PROT.

Many of the majority’s other attempts to cite caselaw in its defense fall victim to the same fundamental error. See Gonazlez-Caraveo v. Sessions, 882 F.3d 885, 891–92 (9th Cir. 2018) (holding that the “administrative closure” standard was amenable to judicial review because the BIA had enumerated “six descriptive, though non-exhaustive factors,” but only because those factors were “not so unique to the agency that this Court would be unable to evaluate them”). Two particularly good examples are Husyev v. Mukasey, 528 F.3d 1172 (9th Cir. 2008), and Taslimi v. Holder, 590 F.3d 981 (9th Cir. 2010), both of which concern “extraordinary circumstances relating to the delay in filing [an asylum] application,” see 8 U.S.C. § 1158(a)(2)(D). In Husyev and Taslimi, the court concluded that the statutory “extraordinary circumstances” standard was amenable to judicial review notwithstanding its breadth for two reasons. First, the accompanying “regulations set out a non- exhaustive list of six potentially qualifying ‘extraordinary circumstances.’” Husyev, 528 F.3d at 1181 (citing 8 C.F.R. § 1208.4(a)(5)(i)–(vi)). Second, an adjudicator was required to consider specific facts, including “an applicant’s delayed awareness of changed circumstances,” when deciding whether extraordinary circumstances existed. Taslimi, 590 F.3d at 986 (citing 8 C.F.R. § 208.4(a)(4)(ii)). The majority contends that the regulatory criteria in 8 C.F.R. § 235.7(a)(4)(x) inform CBP’s risk-based eligibility assessment in much the same way as the regulations expounding on “extraordinary circumstances,” providing “a partial adjudicative standard in and of itself.” Husyev, 590

them not “peculiarly within the agency’s expertise.” Newman, 223 F.3d at 943.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 43

F.3d at 1181. But again, the suggested analogy does not stand up to scrutiny. As explained above, in Husyev and Taslimi, the court considered a statutory phrase— “extraordinary circumstances”—against the backdrop of an extensive regulatory explanation of what that phrase meant. Thus, even though every conceivable “extraordinary circumstance” was not listed in the regulation, the court still had the benefit of fairly detailed guidance about the kind of circumstances that might fairly be considered extraordinary. Contrast that with the regulations at issue here. While 8 C.F.R. § 235.7(a)(4)(x) lists some factors the agency must account for when assessing the risk posed by an applicant, it nowhere explains what affect those factors—particularly “purpose of travel,” “employment,” “residency,” and “criminal history”—are supposed to have on the outcome. And unlike the task presented in cases like Husyev, Taslimi, and Adrian Belt, which is simply to determine whether a statutory or regulatory standard is met, the agency here is tasked with applying the regulatory standards to the underlying question of eligibility. Because the regulations fail to provide CBP with any meaningful guidance about how to do so, the relevant standards are nowhere near as “objective”—or helpful—as the majority suggests. B. Next, consider the majority’s reliance on the “mandatory” nature of the criteria listed in 8 C.F.R. § 235.7(a)(4)(x). In the majority’s view, the fact that the agency has stated these criteria “will be considered in the decision” means that it has self-imposed “a mandatory duty that constrains whatever discretion” it “might otherwise have possessed.” Nat’l Res. Def. Council, Inc. v. Perry, 940 F.3d 1072, 1078–79 (9th Cir. 2019). But that idea cannot be 44 JAJATI V. U.S. CUSTOMS & BORDER PROT.

squared with the fact that the listed criteria are non- exhaustive. The regulation’s command that certain criteria “will be considered” requires only what its plain terms suggest: the agency must consider the listed criteria when making an eligibility determination. As the majority seems to recognize, that command does not limit the agency’s consideration to only those criteria on the list. Nor does that command dictate how much weight to give each “mandatory” factor. Thus, notwithstanding 8 C.F.R. § 235.7(a)(4)(x), the agency remains free to consider the entire universe of factors that are in its judgment relevant to assessing the risk posed by an applicant, and weigh each factor it considers however the agency wishes. This case provides a ready example of other, inherently reasonable criteria that the agency might in its discretion choose to consider. Though Jajati asserts he has no criminal history of his own, he suspects that the agency originally revoked his SENTRI membership because his estranged ex- wife has a criminal history of smuggling drugs across the United States–Mexico border. Jajati assigns error to the agency on that basis, but I fail to see how such reasoning is necessarily erroneous or what source of law we would invoke on review to say so. 7

7 The majority casts this reasoning as “conflat[ing] reviewability with a particular outcome.” Not only is that not true, but it also misses the point of the hypothetical. I do not speculate as to CBP’s motives to pass on the underlying merits of the dispute or to assert with any certainty why the agency decided to revoke Jajati’s SENTRI membership. Indeed, for the reasons explained in this opinion, I believe this court has no jurisdiction to do so. Instead, the hypothetical demonstrates that CBP might plausibly rely on an unlisted factor as the driving force for its decision—and how neither the enabling statute nor the SENTRI

JAJATI V. U.S. CUSTOMS & BORDER PROT. 45

Instead, it seems to me that the agency would be well within its rights to rely on the commonsense notion that association with felons is a red flag. 8 The relevance of that risk is only heightened where the felony involved—drug smuggling across an international border—is so closely related to the privilege sought: expedited screening across an international border. Nothing in the regulations constrains the agency’s ability to reason in this manner, and I cannot imagine how this court could ever conclude such reasoning was arbitrary and capricious without imposing its own ivory tower notions of fairness upon the agency’s efforts to administer the SENTRI program. The regulations relied on by the majority certainly do not provide us with any basis to do so. The majority concludes that the mere existence of the mandatory criteria means that at the very least, a reviewing

regulations provide this court with any tools with which to review its decision to do so. 8 This court regularly engages in a similar kind of reasoning when it upholds conditions of supervised release preventing convicted criminals from associating with known felons. E.g., United States v. King, 608 F.3d 1122, 1128 (9th Cir. 2010). Indeed, we have so thoroughly acknowledged the severity of that risk as to suggest that conditions of supervised release can sometimes prohibit association with a so-called “life partner.” See United States v. Napulou, 593 F.3d 1041, 1047 (9th Cir. 2010) (“A condition of supervised release that prohibits association with convicted felons without the permission of a probation officer is a standard condition …. When, however, such a condition goes beyond the standard prohibition on contact with convicted felons, and singles out a person with whom the individual on supervised release has an intimate relationship, the sentencing court must undertake an individualized review of that person and the relationship at issue, and must provide a justification for the imposition of such an intrusive prohibitory condition.”) (internal citations omitted). 46 JAJATI V. U.S. CUSTOMS & BORDER PROT.

court can test the agency’s fidelity to the self-imposed mandatory criteria. But where, as here, the agency remains free to consider any potentially relevant factor and assign whatever weight it wants to any factor, and the regulation provides no instruction about which combination of factors is potentially dispositive, the agency is well within its discretion to (as it apparently has done here) downplay the importance of the listed criteria and rely predominantly on unlisted criteria when making its risk-based eligibility determination. Because the agency’s decision can rightfully rest in significant part on unlisted criteria, I do not see the utility in reviewing “whether CBP failed to consider the required criteria.” Nor do I see how it would differ meaningfully from reviewing whether the agency “abused its discretion when it evaluated those criteria.” All of this is especially true given the agency is under no obligation to share its reasoning with individuals whose SENTRI memberships are revoked. See 8 C.F.R. § 235.7(a)(4)(x) (requiring only notice of, not the reasons for, revocation). At bottom, once it has been established that 8 C.F.R. § 235.7(a)(4)(x) does not impose any practical restriction on the agency’s essentially unlimited ability to consider whatever factors it deems relevant to eligibility, and weigh those factors however it wants, it is hard to understand how the stipulation that some listed criteria “will be considered” imposes any meaningful constraint on the agency’s discretion. Thus, while I agree with the majority that the criteria listed in 8 C.F.R. § 235.7(a)(4)(x) are mandatory, I do not see how that controls the question of whether there is a “meaningful standard against which to judge the agency’s exercise of discretion.” Spencer Enterprises 345 F.3d at 688.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 47

C. For the reasons explained above, I am deeply skeptical of the majority’s conclusions that the SENTRI regulations provide “mandatory, objective criteria” that meaningfully facilitate review. The listed criteria are not “objective” because they are not defined, and even though they are “mandatory,” they are not exhaustive. Taken together, these facts mean the regulations fail to provide any objective reference point for the type of applicant that is sufficiently “low-risk” to be eligible for SENTRI membership. This court has held time and again that where an enabling statute or regulation fails to meaningfully define the key standard against which agency action should be judged, the decision involved is “committed to agency discretion by law.” 5 U.S.C. § 701(a)(2). Consider, for example, Ekimian v. I.N.S., in which the court again faced the question of whether an “exceptional circumstances” standard provided a meaningful basis for review. 303 F.3d at 1158–59. Unlike the plaintiffs in Husyev and Taslimi, cases which also involved “exceptional circumstances,” the plaintiffs in Ekimian “[could] not point to any statutory, regulatory, or caselaw definition of ‘exceptional circumstances’ applicable to the [agency’s] … power under” the relevant regulations. Id. at 1159. Thus, the court concluded it could “not discover a sufficient meaningful standard against which to judge the [agency’s] decision.” Id. Relatedly, in Idrees v. Barr, the court considered whether the BIA’s decisions regarding whether to certify an ineffective assistance of counsel claim were unreviewable. 923 F.3d 539, 542 (2019). The relevant regulation provided that the BIA “in its discretion may review any such case by certification … if it determines that the parties have already 48 JAJATI V. U.S. CUSTOMS & BORDER PROT.

been given a fair opportunity … regarding the case, including the opportunity to request oral argument and to submit a brief.” Id. The BIA later clarified that it would only certify such claims under “exceptional circumstances.” Id. at 542–43. Despite regulatory criteria that are at least as specific as the criteria the majority relies upon here—“fair opportunity,” “opportunity to request oral argument,” “opportunity … to submit a brief,” and “exceptional circumstances”—the Idrees court concluded that “[t]he regulation contain[ed] no standard for how the agency should exercise its discretion.” Id. at 542. And because (1) “no other regulation or statute provide[d] guidance on th[e] issue,” and (2) the BIA “ha[d] not elaborated on which circumstances are considered to be exceptional,” the issue was “committed to agency discretion.” Id. at 542–43. The court has come to similar conclusions outside the context of immigration. See, e.g., Hyatt v. Office of Mgmt. & Budget, 908 F.3d 1165, 1174 (9th Cir. 2018) (concluding that an agency’s discretionary decision to “take appropriate remedial action … only if necessary” “is beyond judicial review” because “[t]here is no express standard … to guide the OMB in determining whether any particular remedy is either ‘appropriate’ or ‘necessary’”). The majority criticizes the analogy to Idrees, reasoning that “the ‘criteria’ the dissent cites had no bearing as to which circumstances were ‘exceptional.’” But that’s exactly my point. Just like the SENTRI regulations, the regulations at issue in Idrees contained language that superficially appeared to guide the agency’s decision-making. But also just like here, the standards involved were ultimately too unconstrained to be practically meaningful. Though both here and in Idrees each agency has somewhat narrowed the

JAJATI V. U.S. CUSTOMS & BORDER PROT. 49

inquiry by committing itself to a certain standard— “exceptional circumstances” in Idrees and “low-risk” here— neither adequately defined the standard such that its discretion was actually constrained, and neither explained what combination of criteria might allow an applicant to meet the burden imposed by the standard. Therefore, as in Idrees, the majority here should have concluded that the SENTRI regulations committed the question of eligibility to CBP as a matter of law. 9 The majority counters these analogies to Ekimian, Idrees, and Hyatt by making its own analogies to prior panel precedent, including this court’s decisions in Keating v. F.A.A., 610 F.2d 611 (9th Cir. 1979), and City of Los Angeles v. United States Department of Commerce, 307 F.3d 859 (9th Cir. 2002). Presumably, the majority selected those cases because they represent the far outer limit of what this court has considered amenable to judicial review in the past. In Keating and City of Los Angeles, the court held that the terms “public interest” and “feasible,” without more, were sufficiently specific to enable judicial review. See 610 F.2d at 612; see also 307 F.3d at 869. In my view, these cases are uniquely good examples of this court’s disappointing willingness to micromanage the discretionary affairs of administrative agencies. From these far-reaching displays of judicial aggrandizement, the majority reasons that the SENTRI statute’s use of the term “low-risk” must be amenable to judicial review because it is more specific than the “public

9 The majority’s attempts to distinguish Ekimian and Idrees by characterizing the criteria in the regulation here as “mandatory” and “objective.” For the reasons already explained, that characterization is unconvincing. 50 JAJATI V. U.S. CUSTOMS & BORDER PROT.

interest” or “feasible” standards we have previously considered. Moreover, it implies that we are bound to hold as much because Keating and City of Los Angeles are prior panel precedent. But while Miller v. Gammie, 893 F.3d 889 (9th Cir. 2003) (en banc) binds us to the holdings that 42 U.S.C. § 1421(c) (Keating) and 13 U.S.C. § 195 (City of Los Angeles) are not “committed to agency discretion,” it does not doom us to repeat the same wrong reasoning relied on in those cases or to extend that reasoning by analogy to every other vaguely worded statute or regulation on the books. Instead, we must decide whether a law contains “meaningful standards” of review as we always do under these circumstances, by reviewing the specific statutory or regulatory language at issue in its context and determining whether it provides “law to apply.” See Perez Perez, 943 F.3d at 861–62 (collecting cases). In addition to misstating the inquiry, the majority’s theory of how prior panel precedent should be applied to this case also proves too much. If every statutory or regulatory standard that is at least as or more specific than “public interest” or “feasible” is automatically susceptible to judicial review, then the majority would be forced to conclude that this court’s decisions in Ekimian, Idrees, and Hyatt were wrongly decided. After all, it is difficult to see how the standards at issue in those cases—“exceptional circumstances” and “if necessary”—are any less specific than “public interest” or “feasible.” To be sure, the majority resists this conclusion. In its view, Ekimian’s and Idrees’s use of the term “exceptional circumstances” provides “no criteria” to apply. Huh? While I actually agree with that characterization for the reasons explained above, it is extremely difficult for me to see how the majority gets there given its prior endorsement of

JAJATI V. U.S. CUSTOMS & BORDER PROT. 51

Keating and City of Los Angeles. Discerning how the terms “public interest” and “feasible” provide law to apply while the term “exceptional circumstances” does not is beyond my capacity for judicial hair-splitting. In my view, what is in the public interest, what is feasible, and what circumstances are exceptional are all equally broad, discretionary inquiries, and all three seem to fit well within that “certain categor[y] of administrative decisions that courts traditionally have regarded as ‘committed to agency discretion.’” Dep’t of Commerce, 139 S. Ct. at 2568 (quoting Weyerhaeuser Co. v. U.S. Fish & Wildlife Serv., 139 S. Ct. 361, 370 (2018)). Rather than admitting our court has overstepped in the past, the majority amplifies these past errors by unnecessarily extending them. And worse, instead of admitting our caselaw is hopelessly irreconcilable, it pretends that this tangled web of precedents is somehow coherent and self- evidencing. Count me as unconvinced. D. Finally, even if one could accurately characterize the criteria listed in 8 C.F.R. § 235.7(a) as “objective” and “mandatory,” the majority’s position fails to account for this court’s prior recognition that “agency enforcement decisions are generally not suitable for judicial review” when they “involve[] a complicated balancing of a number of factors which are peculiarly within its expertise.” City and County of San Francisco, 796 F.3d at 1001–02. The kinds of factors that are usually considered to be “peculiarly within an agency’s expertise” are “the prioritization of agency resources, likelihood of success in fulfilling the agency’s statutory mandate, and compatibility with ‘the agency’s overall policies.’” Newman, 223 F.3d at 943 (quoting Chaney, 470 U.S. at 831) (cleaned up). 52 JAJATI V. U.S. CUSTOMS & BORDER PROT.

Each of these factors favors a finding of unreviewable agency discretion here. In 2023, CBP reported that over 235 million travelers crossed a land border into the United States. 10 The record reflects that wait times at the San Ysidro border crossing—the busiest land port of entry in the Western Hemisphere 11—have historically been as high as five or six hours. As anyone who has ever stood in an airport security line knows, not everyone can qualify for expedited screening. (If everyone’s special, nobody is.) Given the great mass of people seeking to legally and illegally cross our southern border and the national security consequences of letting the wrong people slip by, I cannot think of a decision more naturally suited to the exercise of the agency’s discretion than how it will go about prioritizing its limited resources to maximize security along our fraught southern border. The decision to expedite the screening of some travelers and pay closer attention to others falls squarely within the contours of that discretionary endeavor. But setting aside the traditional reasons such decisions have been left to the agency’s discretion, even the criteria that are explicitly enumerated in 8 C.F.R. § 235.7(a) are the kind of criteria that are “peculiarly within the agency’s expertise.” Newman, 223 F.3d at 943 (cleaned up). As explained above, federal judges have little to no expertise in assessing how a SENTRI applicant’s “purpose of travel,”

10 United States Customs and Border Protection, Traveler and Conveyance Statistics (last updated Mar. 22, 2024), available at https://www.cbp.gov/newsroom/stats/ travel. 11 United States General Services Administration, San Ysidro Land Port of Entry (last updated Dec. 8, 2023), https://www.gsa.gov/about-us/gsa- regions/region-9-pacific-rim/land-ports-of-entry/san-ysidro-land-port- of-entry.

JAJATI V. U.S. CUSTOMS & BORDER PROT. 53

“employment,” “criminal history,” or “residency” affect the risk they pose when crossing the southern border. 12 Border patrol agents, by contrast, are the group of federal officers most directly tasked with enforcing this country’s immigration laws. If anyone has expertise about how a border crosser’s personal characteristics might affect the risk they pose, it is them. For this reason, it should come as no surprise that the SENTRI program’s enabling statute vests the Department of Homeland Security—and not this court— with the task of “establish[ing] an international registered traveler program” and deciding the “criteria for participation.” 8 U.S.C. § 1365b(k)(3)(A). Today’s decision interferes with that sensible delegation of discretion and will only complicate the already-difficult task CBP faces at the United States–Mexico border. E. Having established the myriad reasons why SENTRI’s enabling regulations do not provide a “meaningful standard

12 The majority relies on Newman v. Apfel to support its conclusion to the contrary. But the regulatory criteria considered in Newman— whether relevant information was “currently available” and “reliable”— are readily distinguishable from the disputed criteria here. 223 F.3d at 939. Unlike the multi-pronged risk analysis that 8 C.F.R. § 235.7(a)(4)(x) requires of the agency, courts consider the “current availability” and “reliability” of information all the time. District courts must do so any time they are tasked as a factfinder. And this court reviews such determinations any time it considers whether substantial evidence supports a factfinder’s conclusion. Indeed, one could convincingly argue that deciding the “current availability” and “reliability” of facts are among the key judicial functions. How we could conclude that “the eligibility criteria in 8 C.F.R. § 235.7(a)(4)(x)” are “even more” susceptible to judicial review than the reliability and current availability of information is absolutely mystifying. 54 JAJATI V. U.S. CUSTOMS & BORDER PROT.

against which to judge the agency’s exercise of discretion,” Spencer Enterprises, 345 F.3d at 688, I now turn to comment briefly on the majority’s treatment of this court’s decision in Perez Perez v. Wolf, 943 F.3d 853 (9th Cir. 2019). Like the other cases relied on by the majority, Perez Perez is ultimately distinguishable from this case. But it was also wrongly decided, and so the majority doubly errs in unnecessarily extending it. In Perez Perez, our court considered whether § 701(a)(2) of the APA commits the question of eligibility for a “U visa” to the discretion of the United States Citizenship and Immigration Service (“USCIS”). Id. at 856. “[A] petitioner is eligible for a U visa if the petitioner (1) has suffered ‘substantial physical or mental abuse’ as a result of having been a victim of qualifying criminal activity; (2) ‘possesses information’ about qualifying criminal activity; and (3) ‘has been helpful, is being helpful, or is likely to be helpful’ to an authority ‘investigating or prosecuting’ qualifying criminal activity.” Id. at 863 (quoting 8 U.S.C. § 1101(a)(15)(U)). In evaluating these eligibility criteria, the agency has a statutory “duty to ‘consider any credible evidence relevant to the petition.’” Id. (quoting 8 U.S.C. § 1184(p)(4). Ultimately, our court in Perez Perez found that these statutory eligibility criteria provided meaningful standards of review. Id. Like in Husyev and Taslimi, the task the agency faced in Perez Perez is meaningfully different than the task CBP faces in assessing the risk posed by Jajati’s SENTRI status. In Perez Perez, the three statutory criteria—past abuse, possession of relevant information, and helpfulness— directly inform the question of eligibility for a U visa. How, then, would that assessment work out in practice? If USCIS finds that a petitioner (1) has suffered abuse, (2) possesses

JAJATI V. U.S. CUSTOMS & BORDER PROT. 55

information, and (3) he has been helpful to an investigation, then it can conclude with certainty that the petitioner is statutorily eligible for the visa. By contrast, the disputed regulations in this case provide CBP with no such certainty when determining an applicant’s eligibility for SENTRI status. While 8 C.F.R. § 235.7(a)(1)(i) makes it sufficiently clear that the touchstone for eligibility is risk, the criteria in § 237.1(a)(4)(x) do not speak to the risk assessment with anywhere near the degree of clarity as in Perez Perez because, as has now been exhaustively explained, CBP retains considerably more discretion when determining SENTRI eligibility than USCIS does when assessing U visa eligibility. Thus, even taking its conclusions as true, Perez Perez is readily distinguishable. Even if Perez Perez were indistinguishable, however, I believe it was wrongly decided, and for the reasons laid out in Judge Callahan’s thoughtful dissent, I would vote to take an appropriate case en banc to correct its errors. See id. at 873–75 (Callahan, J., dissenting). And unfortunately, Perez Perez is not even close to the worst offender. See, e.g., Keating, 610 F.2d at 612; see also City of Los Angeles, 307 F.3d at 869 n.6. Results like this one are just the latest in a long line of judicial intrusions into matters prudently committed to the agency’s discretion.

III.

The majority’s decision today not only misinterprets the APA, but it also threatens to make serious practical impositions on CBP’s ability to administer the SENTRI program moving forward. Defending its decisions on appeal will inevitably force the agency to divulge its reasons for revoking memberships to reviewing courts, some of which 56 JAJATI V. U.S. CUSTOMS & BORDER PROT.

might pertain to sensitive matters of national security. See Webster v. Doe, 486 U.S. 592, 600 (1988) (citing considerations of national security as a basis to foreclose judicial review of an ex-CIA employee’s wrongful termination claim); see also Bennett v. Chertoff, 425 F.3d 999, 1001 (D.C. Cir. 2005) (“Because the authority to issue a security clearance is a discretionary function of the Executive Branch and involves the complex area of foreign relations and national security, employment actions based on denial of security clearance are not subject to judicial review.”). Worse, the agency will be forced to defend its discretionary decision-making process even though SENTRI’s enabling regulations nowhere oblige the agency to explain the reasons for a revocation to a former member. Forcing the agency to defend against actions like Jajati’s will funnel more of CBP’s time and resources toward litigation, which will very likely result in the agency granting fewer SENTRI memberships overall. Or maybe the agency will just err on the side of granting such memberships to avoid judicial meddling. Either way, it is hard to see how this will strike the right balance between our nation’s security and expediting lawful border crossings. Because the majority improperly aggrandizes judicial power and “plunges us [further] into the standardless review” of inherently discretionary agency decision-making, Perez Perez, 943 F.3d at 868 (Callahan, J., dissenting), I respectfully dissent.