Victor B. Skaar v. Robert L. Wilkie

• Court: United States Court of Appeals for Veterans Claims
• Decided: December 6, 2019
• Docket: 17-2574
• Status: Published

Opinion

UNITED STATES COURT OF APPEALS FOR VETERANS CLAIMS

NO. 17-2574

VICTOR B. SKAAR, APPELLANT,

V.

ROBERT L. WILKIE, SECRETARY OF VETERANS AFFAIRS, APPELLEE.

Before BARTLEY, Chief Judge, and PIETSCH, GREENBERG, ALLEN, MEREDITH, TOTH, FALVEY, Judges; and DAVIS and SCHOELEN, Senior Judges.*

ALLEN, Judge, with BARTLEY, Chief Judge, and GREENBERG, TOTH, Judges; and DAVIS, Senior Judge.

SCHOELEN, Senior Judge, concurring in part and dissenting in part.

FALVEY, Judge, with PIETSCH and MEREDITH, Judges, dissenting.

ORDER

United States Air Force veteran Victor B. Skaar was exposed to ionizing radiation while participating in the cleanup of plutonium dust in Palomares, Spain, following a midair aircraft collision. He later developed a blood disorder, leukopenia, which he believes was caused by in- service radiation exposure, even though an Air Force radiation dose estimate found the levels of exposure he suffered far below those required to cause his disability. In an April 14, 2017, decision the Board of Veterans' Appeals (Board) denied him service connection. This appeal followed.

We do not today address the merits of Mr. Skaar's claim. Rather, we consider his motion to certify a class of similarly situated veterans to proceed in an aggregate action. The issue we confront here—class certification in the context of an appeal of an individual Board decision—is one of first impression. For many years, we held this Court categorically lacked the power to certify classes. See Monk v. McDonald, No. 15-1280, 2015 WL 3407451, at *3 (May 27, 2015) (Monk I); Harrison v. Derwinski, 1 Vet.App. 438, 439 (1991) (en banc) (per curiam); Lefkowitz v. Derwinski, 1 Vet.App. 439, 440 (1991) (en banc) (per curiam). This changed when the United States Court of Appeals for the Federal Circuit (Federal Circuit) held we possess, at least in certain contexts, the authority to certify class actions. Monk v. Shulkin, 855 F.3d 1312, 1321-22 (Fed. Cir. 2017) (Monk II). We then held we would, in appropriate cases, certify classes seeking writs of mandamus under the All Writs Act. Monk v. Wilkie (Monk III), 30 Vet.App. 167, 174 (2018); see, e.g., Godsey v. Wilkie, 31 Vet.App. 207, 220-25 (2019); see also Wolfe v. Wilkie, __ Vet.App. __, No. 18-6091, 2019 WL 4254039, at *14-19 (Sept. 9, 2019).

* Judges Davis and Schoelen are Senior Judges acting in recall status. In re: Recall of Reitred Judge, U.S. VET. APP. MISC. ORDERS 16-19, 17-19 (Dec. 4, 2019).

1 This brings us to Mr. Skaar's motion for class certification. We hold (1) the Court may, in appropriate situations, certify classes in the context of an individual appeal of a Board decision; (2) our jurisdiction allows us to include in such classes both persons who have obtained a final Board decision as well as those who have not; and (3) as in the petition context, we will use Federal Rule of Civil Procedure 23 as a guide when deciding whether to grant class certification. Finally, class certification will be reserved for those cases where appellants demonstrate the class device is a superior vehicle for litigating the class claim than a precedential decision. Applying these principles, we grant in part and deny in part the motion for class certification.

2 TABLE OF CONTENTS I. BACKGROUND ........................................................................................................................ 4 II. ANALYSIS ............................................................................................................................... 8 A. Standing ................................................................................................................................. 9 1. Mr. Skaar lacks standing to pursue the § 3.309 claim on behalf of the class. ................... 9 2. Mr. Skaar has standing to pursue the § 3.311 claim on behalf of the class. .................... 10 B. The Power To Certify Class Actions in the Appeal Context ............................................... 13 C. The Utility of Class Actions in the Appeal Context............................................................. 14 D. The Proposed Class Composition ........................................................................................ 15 1. The Present-Future and Future-Future Claimants........................................................... 16 2. The Expired Claimants ..................................................................................................... 22 3. The Past Claimants ........................................................................................................... 24 E. Class Certification Analysis ................................................................................................. 25 1. The proposed class is so numerous that joinder would be impracticable. ....................... 26 2. The proposed class presents a common issue capable of classwide resolution. .............. 27 3. Mr. Skaar's claim is typical of that of the proposed class. ............................................... 28 4. Mr. Skaar will fairly and adequately protect the interests of the class. ........................... 29 5. The requested injunctive relief is appropriate respecting the class as a whole. .............. 29 6. The class action device is a superior method of litigating the class claim. ..................... 30 7. Proposed counsel is adequate. ......................................................................................... 34 8. Generalized notice of class certification is required but opt out rights are not. .............. 35 III. CONCLUSION ..................................................................................................................... 36

3 I. BACKGROUND

In the early morning hours of January 17, 1966, a U.S. Air Force B-52 Superfortress bomber, armed with four thermonuclear weapons, collided with a KC-135 refueling tanker over the small fishing village of Palomares, Spain. See Record (R.) at 28-29, 560, 796-98, 1878-80, 3509, 3557-802. Part of Operation Chrome Dome, a U.S. military plan calling for continuous patrol by nuclear bombers around the airspace of the former Soviet Union, the bomber was supposed to refuel with the tanker for the trip home. R. at 3574-76. The midair collision destroyed both aircraft, and the bomber's atomic payload was scattered across the Spanish countryside. R. at 3605-07. Eventually, one weapon was recovered intact and another fished from the depths of the Mediterranean. R. at 3613-32. Emergency parachutes attached to the other two bombs, however, failed to deploy. R. at 3603-04. Both bombs impacted at high speeds, causing internal, nonnuclear explosives in the devices to detonate. R. at 3606-07. The resulting explosions released a cloud of radioactive plutonium dust over the area, contaminating soil and crops, and spreading radioactive debris for miles. R. at 1878.

Mr. Skaar, along with nearly 1,400 other U.S. military personnel, was sent to the accident site to assist in cleanup and monitoring efforts. While there, to assess possible radioactive exposure, the military personnel gave urine and nasal swab samples. Mr. Skaar was a member of a group of the 26 service members (the High 26) who were determined to be among the most exposed and who were monitored for a period of 18 to 24 months after the accident for signs of radiogenic conditions. R. at 2124-28. The monitoring efforts were discontinued, however, in December 1967 when the Air Force informed Mr. Skaar his "health is in no jeopardy from retention of radioactive materials as a result of participation in the [Palomares cleanup] operation." R. at 2430.

But in 1998, 32 years after the Palomares cleanup, Mr. Skaar was diagnosed with leukopenia, a decrease in white blood cell count. R. 2157. The diagnosing physician opined that exposure to ionizing radiation "[h]istorically does appear to be the positive agent" causing leukopenia, but concluded "we have been unable to prove this." Id. Mr. Skaar then filed a claim with VA, seeking service connection for that condition. R. at 2155. In February 2000, VA denied his claim. See R. at 2090-99. This was so, VA explained, because leukopenia is not a radiogenic disease VA recognizes as resulting from a "radiation-risk activity," and because Mr. Skaar had not presented sound scientific or medical evidence linking the disease to radiation exposure. R. at 2097.

Two separate regulatory paths lead to to service connection for veterans who suffer a disability they believe was caused by exposure to ionizing radiation. Both are at issue here as part of Mr. Skaar's motion for class certification. Under 38 C.F.R. § 3.309(d)(3)(ii), VA recognizes certain nuclear incidents as "radiation-risk activities." Those who participated in a radiation-risk activity listed in § 3.309 and who later developed one or more of the radiogenic diseases enumerated in § 3.309(d)(1) benefit from a presumption of service connection. § 3.309(a). For those who did not participate in a listed radiation-risk activity, § 3.311(a) is available. See Hilkert v. West, 12 Vet.App. 145, 148-49 (1999) (en banc). Under that provision, VA requests exposure data from a veteran's service branch. 38 C.F.R. § 3.311(a)(1)-(2). For those claims that meet certain threshold requirements, the Under Secretary for Benefits then reviews the gathered information

4 and determines whether "sound scientific and medical evidence supports the conclusion [that] it is at least as likely as not" the condition is the result of ionizing radiation exposure. § 3.311(a), (c). The regulation defines "sound scientific evidence" as "observations, findings, or conclusions which are statistically and epidemiologically valid, are statistically significant, are capable of replication, and withstand peer review," and "sound scientific medical evidence" as "observations, findings, or conclusions which are consistent with current medical knowledge and are so reasonable and logical as to serve as the basis of management of a medical condition." § 3.311(c)(3). In making that determination, the Under Secretary for Benefits may request an advisory opinion from the Under Secretary for Health. § 3.311(c)(1). The Under Secretary's final determination is then sent to the agency of original jurisdiction, which considers the opinion as evidence. § 3.311(f). For Palomares veterans, § 3.309's presumption of service connection is unavailable because VA does not recognize the Palomares plutonium dust cleanup as a radiation- risk activity. So instead, veterans such as Mr. Skaar must seek service connection under § 3.311's less favorable provisions. See Ramey v. Gober, 120 F.3d 1239, 1242-43 (Fed. Cir. 1997).

The Air Force provides VA with dose estimates for Palomares veterans. In April 2001, a consulting firm, Labat-Anderson, evaluated the Air Force's dose methodology. See R. at 2682- 2818. This evaluation culminated in a report (the LA Report or the Report) establishing preliminary dose estimates for various subcategories of veterans. R. at 2691. The LA Report stated that the recorded urine dose intakes for Palomares veterans "seemed unreasonably high" compared to "environmental measurements" derived from air sampling some 15 years after the cleanup and "estimates prepared for other plutonium exposure cases – persons residing in the Palomares vicinity and Manhattan Project workers." R. at 2701. These air samples and comparisons "provided a basis for preparing independent estimates of intake and dose using representative scenarios" rather than actual recorded dose intakes. R. at 2691. After comparing those "independent estimates" with the actual recorded dose intakes, the Report "excluded data from the on-site samples and attributed more significance to samples collected at later dates for the High 26 Group." R. at 2795. This exclusion of "unreasonably high" dose estimates forms the basis for Mr. Skaar's allegation that the Air Force's dose estimates do not constitute "sound scientific evidence" as required by law. See Appellant's Apr. 23, 2019, Response (Resp.) at 4. The Report noted its findings "represent preliminary estimates that cannot be considered as definitive" and recommended further study "to develop credible estimates of dose that are compatible with those calculated from environmental data." Id. Despite these reservations, the Air Force adopted the LA Report's dose estimate methodology in full. See R. at 1580-81, 3508-511.

After VA's initial denial in 2000, Mr. Skaar requested that VA reopen his claim in March 2011. R. at 2077. Based on that claim and per § 3.311, the regional office (RO) requested a radiation exposure opinion. R. at 1886. In response, the Air Force estimated in April 2012 that Mr. Skaar's maximum total effective dose during the Palomares cleanup was 4.2 rem with a bone marrow committed dose of 1.18 rem, compared to annual dose limits of 5 and 50 rem, respectively, for occupations typically involving radiation exposure.1 R. at 1888-89. Based on these estimates, the director of the Post 9/11 Environmental Health Program, writing for the Under Secretary for Benefits, advised in May 2012 that "it is unlikely that [Mr. Skaar's] leukopenia . . . can be attributed 1 A rem (roentgen equivalent man) is the unit of measurement for radiation. One unit represents "the dosage of a ionizing radiation that will cause the same biological effect as one roentgen of X-ray or gamma-ray exposure." MERRIAM-WEBSTER DICTIONARY, https://www.merriam-webster.com/dictionary/rem (last visited Oct. 31, 2019).

5 to radiation exposure while in military service." R. at 1877. And, based on this opinion, the RO in June 2012 denied Mr. Skaar's claim. R. at 1869. Mr. Skaar timely disagreed with the RO's denial, but the RO continued to deny the claim in September 2013. R. at 1690-91. He then perfected an appeal to the Board. R. at 1588-89.

In October 2013, a private physician opined that Mr. Skaar's leukopenia "is likely related to exposure to heavy radioactive material in [1966]." R. at 39-40. And 2 months later, the Air Force concluded an evaluation of its radiation dose methodology that revealed "inconsistencies in dose assignment over the past 12 years" since the LA Report. R. at 1580. The Air Force's methodology, derived from the LA Report, "appear[ed] to underestimate doses to some individuals" and, thus, the Air Force intended to "formally standardize [its] response methodology for radiation dose inquiries involving Palomares participants" by establishing dose estimates based on each veteran's specific duties. Id. Finally, the Air Force stated it would reevaluate the individual dose estimates it had already provided for Palomares veterans. R. at 1581. And in June 2014, the Air Force provided VA with Mr. Skaar's revised dose estimate, assigning him a new maximum total effective dose of 17.9 rem and a bone marrow committed dose of 14.2 rem. See R. at 1301, 1274-75.

Meanwhile, the Board in May 2015 found the Air Force's revised dose estimates were new and material evidence warranting reopening of Mr. Skaar's claim. R. at 695-99. The Board then remanded the claim to the RO because the Air Force's "revised assessment [was] significantly higher than the April 2012 assessment" and, thus, "another [dose estimate] opinion [was] warranted." R. at 698. That opinion was provided in August 2016. The RO found that, based on the revised bone marrow committed dose estimate of 14.2 rem, "it is not likely that the Veteran's leukopenia was caused by exposure to ionizing radiation during military service." R. at 131-35. The RO then again denied Mr. Skaar's claim, citing the results of the August 2016 revised dose estimate. R. at 113-14. Nonetheless, in September 2016, a private physician opined that Mr. Skaar's leukopenia was "a result of exposure to ionizing radiation/plutonium." R. at 38.

Mr. Skaar then returned to the Board, which, on April 14, 2017, again denied his claim. See R. at 2-12. The Board concluded VA's May 2012 dose estimate opinion lacked probative value "as it was based on an inaccurate dose estimate." R. at 10. But the Board found the August 2016 dose estimate "highly probative" because it "was based on a review of the entire record," while Mr. Skaar's private medical opinions were not as probative because "none offered any rationale for their statements." R. at 10-11. Mr. Skaar then appealed to this Court, and filed the pending motion for class certification. The Secretary moved to stay proceedings in this matter pending our resolution of Monk III, a request we denied. This matter was assigned to a panel of the Court for decision on the merits but, given the novelty of the issue, the motion for class certification was submitted to the full Court for decision.

Mr. Skaar asks us to certify a class of "all U.S. veterans who were present at the 1966 cleanup of plutonium dust at Palomares, Spain[,] and whose application for service-connected disability compensation based on exposure to ionizing radiation the VA has denied or will deny." Motion (Mot.) for Class Certification at 1. He later clarified the proposed class encompasses (i) "veterans whose claims for service-connected disability benefits related to exposure to ionizing radiation at Palomares the VA has denied at any level, from the RO through the [Board], except for those who have appealed to this Court and received a decision for which the mandate has

6 issued;" (ii) "veterans whose claims the RO or [Board] has denied and for which the deadline for appeal has expired, as well as veterans whose claims are currently pending before a [decision review officer] or the [Board] after an initial RO denial;" and (iii) "Palomares veterans with an appeal currently pending before this Court[.]" Appellant's Apr. 16, 2018, Resp. at 2. The proposed class also includes "veterans with claims that have not yet been filed at the RO," including "those who have not filed a claim for an existing condition, including because they are aware of the VA's history of denial of Palomares veterans' claims or the methodology used to calculate dose exposure" and "those who have only recently developed a radiogenic condition, and those whose claims have been delayed at the RO." Id.

The proposed class raises two claims. The first challenges VA's omission of the Palomares cleanup from the list of radiation-risk activities in 38 C.F.R. § 3.309(d)(3)(ii) (the § 3.309 claim), while the second centers around VA's compliance with § 3.311(c)'s command that when adjudicating Palomares veterans' claims VA rely on dose estimates based on "sound scientific and medical evidence" (the § 3.311 claim). Mr. Skaar's proposed class alleges VA's actions regarding both claims are invalid under the Administrative Procedure Act and violate class members' due process and equal protection rights. The putative class asks us to order VA to (i) recognize the Palomares cleanup as a "radiation-risk activity;" (ii) apply dose estimate methodology that is supported by "sound scientific and medical evidence;" and (iii) re-adjudicate the benefits claims of those class members whose claims have already been denied.

During the Court's review of this matter, it became clear the Board had failed to address several of Mr. Skaar's arguments regarding the § 3.311 claim. See R. at 106-07, 778-83. Thus, we ordered a limited remand to the Agency for it to "provide a supplemental statement of reasons or bases addressing the appellant's expressly raised argument in the first instance." Skaar v. Wilkie, 31 Vet.App. 16, 18 (2019). The Board faithfully complied with our order. In its supplemental statement, the Board stated Mr. Skaar's arguments based on the first, lower 2012 dose estimate "appear moot" as "the Board's April 2017 decision specifically did not rely on [the] May 2012 findings . . . since those findings were based on the April 2012" Air Force dose estimate that had since been found to have inconsistencies. Secretary's Mar. 29, 2019, Resp. at 4.

Regarding the June 2014 revised dose estimate, the Board found that "on its face it is based on sound scientific evidence" because it "was based on then recently re-evaluated internal processes which were initiated to ensure a comprehensive and consistent approach to dose estimates," and because it "considered the Veteran's previously reported intake values based on the application of contemporary models to his bioassay data collected in the 1960's." Id. at 5. As to whether the previous inconsistencies in the Air Force's dose methodology that plagued its earlier April 2012 estimate still plagued the June 2014 revised dose estimate, the Board stated that "just as it is prohibited from exercising its own independent judgment to resolve medical questions, the Board is not in a position to exercise such independent judgment on matters involving scientific expertise." Id. (citing Colvin v. Derwinski, 1 Vet.App. 171, 175 (1991)).

Finally, the Board explained it "is bound by regulations of the Department," and those regulations "provide specific instructions for obtaining dose estimates." Id. at 6 (citing 38 U.S.C. § 7104(c); 38 C.F.R. §§ 19.5, 20.101(a) (2018)). Thus, "[w]ithout an independent dose estimate, and without a rational basis to reject the competent findings of the Air Force," the Board had no

7 evidentiary basis on which to grant service connection. Id. at 5. Armed with a record sufficient for the Court to consider the class certification motion, we turn to that endeavor now.

II. ANALYSIS

First, we confront a threshold issue. We must decide whether Mr. Skaar has the requisite standing to assert the claims on which he seeks to represent a class. We conclude he lacks standing to bring the § 3.309 claim, but has standing to pursue the § 3.311 claim.

We then assess whether we have the power to use the class action device as a matter of law. We conclude we do. We then consider whether, as a normative matter and given our status as an appellate court with the power to issue precedential opinions, we will exercise our discretion to certify class actions in appropriate appeals. We conclude, as we did in the petition context, class actions have a role to play in appeals in appropriate situations.

Returning to the proposed class, we examine the proposed class definition and modify it to exclude those claimants with adverse decisions who chose not to appeal (i.e., their claims have expired). We then address whether we should certify the modified class as to the § 3.311 claim. In this regard, we first make clear, as we did with petitions, see Monk III, 30 Vet.App. at 174, we will use Federal Rule of Civil Procedure 23 as a guide for determining whether class certification is appropriate. We then conclude the modified class satisfies Rule 23(a)'s requirements and is consistent with the functional requirements of Rule 23(b)(2). But we also recognize Rule 23 is only a guide. We are not similarly situated to the Federal district courts, for which Rule 23 was written. Thus, we consider whether our status as an appellant court (both in terms of the use of precedential opinions and the challenges we may face in managing a class action) counsels against certification. We conclude, in the context of this case, our appellate role does not counsel against certification. But we also hold we will presume class actions should not be certified because of our ability to render binding precedential decisions. Claimants seeking class certification can rebut this presumption by showing by a preponderance of the evidence that a class action is "superior to other available methods for fairly and efficiently adjudicating the controversy" before we will exercise our discretion in certifying a class.

Having determined class certification is appropriate, we next consider appointment of class counsel. Following the guidance of Federal Rule of Civil Procedure 23(g), we appoint Michael Wishnie, Esq., of the Jerome N. Frank Legal Service Organization at Yale Law School as class counsel.

Our final consideration concerns whether class members may elect to opt out of this action and what notice, if any, the class should receive of our certification decision. In line with the overwhelming weight of Federal jurisprudence, we hold the nature of this class is such that opt out rights are not required. And, because class members may not opt out, there is no need to provide individualized notice of certification. However, we conclude generalized notice of class certification designed to reach as many class members as possible is appropriate and order the parties to develop a joint plan for effecting such notice.

8 Having summarized our holdings, we now address each point in detail in the balance of this order.

A. Standing

"[S]tanding is a threshold inquiry in all actions," including class actions.2 Allen v. Wright, 468 U.S. 737, 750 (1984). "In an era of frequent litigation, class actions, sweeping injunctions with prospective effect, and continuing jurisdiction to enforce judicial remedies, courts must be more careful to insist on the formal rules of standing, not less so." Ariz. Christian Sch. Tuition Org. v. Winn, 563 U.S. 125, 146 (2011). "Standing is one of the keys necessary to open the door to the federal courthouse." Matte v. Sunshine Mobile Homes, Inc., 280 F. Supp. 805, 826 (W.D. La. 2003). The appellant has the burden of showing standing. See Lujan v. Defs. of Wildlife, 504 U.S. 555, 561 (1992). "That a suit may be a class action . . . adds nothing to the question of standing, for even named plaintiffs who represent a class 'must allege and show that they personally have been injured, not that injury has been suffered by other, unidentified members of the class to which they belong and which they purport to represent.'" Simon v. E. Ky. Welfare Rights Org., 426 U.S. 26, 40 n.20 (1976) (quoting Warth v. Seldin, 422 U.S. 490, 502 (1975)). "[S]tanding cannot be acquired through the back door of a class action." Allee v. Medrano, 416 U.S. 802, 829 (1974) (Burger, C.J., concurring in part and dissenting in part). "If the individual plaintiffs lack standing, the court need never reach the class action issue." Hawecker v. Sorensen, No. 1:10-cv-00085 OWW JLT, 2011 WL 98757, at *2 (E.D. Cal. Jan. 12, 2011).

Standing requires the appellant show (1) an injury-in-fact; (2) traceability; and (3) redressability. See Defs. of Wildlife, 504 U.S. at 560-61; see also Friends of the Earth, Inc. v. Laidlow Envtl. Servs. (TOC), Inc., 528 U.S. 167, 180-81 (2000). An injury-in-fact is one that is "concrete," "particularized," "not abstract," and "actual or imminent." Defs. of Wildlife, 504 U.S. at 560-61. Claimants cannot simply "allege a bare procedural violation, divorced from any concrete harm" to satisfy the injury requirement. Spokeo, Inc. v. Robins, 136 S. Ct. 1540, 1549 (2016). Standing is determined on a claim-by-claim basis. See, e.g., McGuire v. BMW of N. Am., LLC, No. 13-7356 (JLL), 2014 WL 2566132, at *6 (D.N.J. June 6, 2014). In class actions with multiple claims, at least one named representative must have standing with respect to each claim. See Keepseagle v. Veneman, No. Civ.A.9903119EGS1712, 2001 WL 34676944 (D.D.C. Dec. 12, 2001); Prado-Steiman ex rel. Prado v. Bush, 221 F.3d 1266, 1279 (11th Cir. 2000). Without it, the claims must be dismissed. See, e.g., King Cty. v. IKB Deutsche Industriebank AG, Nos. 09 Civ. 8387(SAS), 09 Civ. 8822(SAS) 2010 WL 2010943 (S.D.N.Y. May 18, 2010). Accordingly, we separately analyze Mr. Skaar's standing to challenge both §§ 3.309 and 3.311.

1. Mr. Skaar lacks standing to pursue the § 3.309 claim on behalf of the class.

The § 3.309 claim alleges VA's omission of the Palomares incident from its list of recognized radiation-risk activities under § 3.309 is arbitrary and capricious, violates the Administrative Procedure Act, and is unconstitutional. Section 3.309 establishes a presumption of service connection for veterans who have (i) a listed radiogenic disease (ii) resulting from a

2 This Court has adopted Article III of the Constitution's case-or-controversy requirement. See Mokal v. Derwinski, 1 Vet.App. 12, 13 (1990).

9 recognized radiation-risk activity. So, for Mr. Skaar to show an injury-in-fact he must demonstrate VA's exclusion of Palomares from the regulation's list of radiation-risk activities harmed him in a concrete and particularized way. See Defs. of Wildlife, 504 U.S. at 560-61. But the Board decision before us denied service connection for leukopenia, which is not one of § 3.309's enumerated radiogenic conditions. Thus, if we were to grant the requested relief as to this claim, Mr. Skaar would not benefit from the regulation's presumption. Mr. Skaar attempts to sidestep this by arguing Palomares' recognition as a radiation-risk activity would entitle him to enroll in VA's Ionizing Radiation Registry (IRR). This program provides certain health screening benefits for veterans exposed to ionizing radiation. See VHA Directive 1301 (Apr. 6, 2017).

We hold Mr. Skaar lacks standing to challenge § 3.309 because he would not benefit from the relief requested as his condition, leukopenia, is not a listed radiogenic condition under that regulation. Thus, the inclusion of Palomares as a radiation-risk activity, while it may assist many unnamed class members, would not entitle him to § 3.309's presumption of service connection. Further, any harm Mr. Skaar has suffered from not having access to the IRR is distinct from the alleged harm suffered by veterans with qualifying radiogenic diseases. The unavailability of IRR enrollment also fails to meet the proposed class definition. Mr. Skaar seeks to represent "all U.S. veterans who were present at the 1966 cleanup of plutonium dust at Palomares, Spain[,] and whose application for service-connected disability compensation based on exposure to ionizing radiation the VA has denied or will deny." Mot. for Class Certification at 1. But IRR enrollment, to the extent Mr. Skaar has been denied it and to the extent it represents a "benefit," is not an "application for service-connected disability compensation" and, thus, cannot serve as the basis for Mr. Skaar's standing to represent the proposed class as to the § 3.309 claim.

"It is not enough that the conduct of which the plaintiff complains will injure someone. The complaining party must also show that he is within the class of persons who will be concretely affected. Nor does a plaintiff who has been subject to injurious conduct of one kind possess by virtue of that injury the necessary stake in litigating conduct of another kind, although similar, to which he has not been subject." Blum v. Yaretsky, 457 U.S. 991, 999 (1982) (emphasis in original). But see Gratz v. Bollinger, 539 U.S. 244, 262-63 (2003) (declining to answer whether respondent, who was an undergraduate transfer student, had standing to represent a class that included both undergraduate transfer students and freshmen or whether the issue was more properly analyzed under Rule 23's typicality analysis). Mr. Skaar may very well be correct he has suffered some type of harm from not having access to the IRR, but his proper remedy for that particular injury is to pursue relief from VA, not this Court. Thus, we dismiss Mr. Skaar's challenge to VA's omission of Palomares from § 3.309's list of radiation-risk activities as he lacks standing to bring the claim. See Rosinski v. Shulkin, 29 Vet.App. 183, 190-92 (2018); Prado-Steiman ex rel. Prado, 221 F.3d at 1279.

2. Mr. Skaar has standing to pursue the § 3.311 claim on behalf of the class.

However, we hold Mr. Skaar does have standing to challenge VA's reliance on the Air Force's dose estimate methodology in deciding claims under § 3.311. First, he has suffered an injury-in-fact. Certain qualifying radiogenic conditions not listed in § 3.309 are analyzed under § 3.311, which requires evidence of radiation exposure and dosages for the award of service connection. See 38 C.F.R. § 3.311(a)(1); see also Hilkert, 12 Vet.App. at 145-49. Leukopenia is

10 not listed as a qualifying radiogenic condition. See § 3.311(b)(2). But § 3.311, unlike § 3.309, provides that, for conditions other than those specifically listed by VA as qualifying radiogenic diseases, "VA shall nevertheless consider the claim under the provisions of this section provided the claimant has cited or submitted competent scientific or medical evidence that the claimed condition is a radiogenic condition." § 3.311(b)(4). And the Board favorably found Mr. Skaar's private medical opinions linked his leukopenia to radioactive exposure. See R. at 6. Thus, Mr. Skaar's leukopenia qualifies for the dose estimate procedures of § 3.311.

VA regulations require dose estimates be supported by "sound scientific and medical evidence." 38 C.F.R. § 3.311(c)(1)(i). Mr. Skaar, both individually and on behalf of the class, argues the Air Force's dose methodology, which VA relies on in adjudicating service connection claims by Palomares veterans, fails to meet that standard. Unlike the class claim under § 3.309, in his class claim under § 3.311 Mr. Skaar was subject to the challenged conduct.

For claims under § 3.311, "an assessment will be made as to the size and nature of the radiation dose or doses." § 3.311(a). For claims based on exposure other than from atmospheric nuclear weapons testing or the military occupations of Hiroshima or Nagasaki, VA must request "any available records concerning the veteran's exposure to radiation," such as service medical records and "other records which may contain information pertaining to the veteran's radiation dose in service." § 3.311(a)(2)(iii).

Mr. Skaar filed a service connection claim for leukopenia in March 2011. VA then requested a dose estimate from the Air Force. That estimate stated Mr. Skaar's maximum total effective dose was 4.2 rem. In May 2012, the VA Environmental Health Program found that, because Mr. Skaar's effective dose was less than 5 rem, "it is unlikely that his leukopenia . . . can be attributed to radiation exposure." R. at 1877. VA then denied his claim in June 2012. However, in December 2013, the Air Force increased its assigned dose values for Palomares veterans after determining its previous methods led to inconsistent dose estimates. VA then again denied Mr. Skaar's leukopenia claim in March 2014, choosing not to apply the revised dose methodology to his claim. The Air Force then again revised its assigned dose value for Mr. Skaar to 17.9 rem, a more than quadruple increase from its previous assigned dose value. The Board then reopened Mr. Skaar's leukopenia claim in May 2015 because of the new dose estimate and remanded the claim to the RO, which again denied the claim. Mr. Skaar perfected an appeal to the Board, which then yet again denied service connection. R. at 2-12. The proposed class here challenges VA's reliance on both the Air Force's pre- and post-2013 dose estimate methodologies. See Appellant's Apr. 8, 2019, Resp. at 3.

The parties spill a great deal of ink discussing Mr. Skaar's standing to represent the class challenge. The Secretary argues there is a crucial distinction between the pre-2013 and post-2013 methodologies.3 See Secretary's Apr. 18, 2019, Resp. at 1-3. He contends Mr. Skaar lacks standing to challenge the pre-2013 methodology because that method was derived from air sampling, while his dose estimates came from urine sampling. Id. at 2. Mr. Skaar counters that "[t]he pre-2013 and

3 As stated above, the Air Force adopted the LA Report in 2001. See R. at 1580-81; 3508-511. Thus, and because Mr. Skaar challenges only VA's reliance on dose estimates prepared using the Report's methodology, he does not have standing to challenge denials of claims due to ionizing radiation exposure from the Palomares cleanup that were based on dose estimates pre-dating 2001.

11 post-2013 distinction is meaningless because [he] challenges the VA's reliance on the LA Report as a whole." Appellant's Apr. 23, 2019, Resp. at 4. In his view, "the LA Report's original sin is that it excluded the urine samples with the highest plutonium measurements." Id. Mr. Skaar alleges this exclusion of the highest dose estimates applies equally to both the pre-2013 and post-2013 methodologies.

Whether one considers the question of differences in the pre- and post-2013 methodologies as one of constitutional standing or under Rule 23's typicality analysis is largely one of semantics here, involving significant overlap. Thus, we analyze the pre- and post-2013 distinction in the context of both standing and typicality.

First, Mr. Skaar has standing to challenge the post-2013 methodology because the Air Force's post-2013 methodology excluded the highest measurements recorded. In a December 2013 document, the Air Force stated it was revising Palomares dose estimates by setting the estimated dose intake for the High 26 group as "their established intake estimates," and by using, for all other Palomares veterans, the lowest dose intake from the High 26. R. at 1580-81. But Mr. Skaar contends the established plutonium intakes for the High 26 are artificially deflated by the earlier decision to exclude "unrealistically high" measurements taken on-site. Thus, the Air Force's revised methodology does nothing to correct the exclusion of the urine samples with the highest plutonium measurements as to Mr. Skaar, and he has sufficiently shown an injury-in-fact as to the post-2013 methodology. Appellant's Apr. 23, 2019, Resp. at 4.

Second, debating whether Mr. Skaar has standing to represent those class members solely challenging VA's reliance on pre-2013 Air Force dose estimates is almost certainly an academic exercise. As discussed in the balance of this order, we will certify a modified class of claimants that excludes those whose claims related to ionizing radiation exposure from the Palomares cleanup have been denied by VA or this Court and those whose appeals windows for those denials have expired. Put differently, our decision affects only claimants who will file claims after the date of this order or those whose claims are currently before the Court or pending before VA. That means it's exceedingly unlikely there are any remaining class members who will only have a dose estimate based solely on the pre-2013 methodology.

But, even if there are class members whose claims were denied solely on the basis of the Air Force's pre-2013 methodology, Mr. Skaar has sufficient standing to represent them. He has shown injury-in-fact from the pre-2013 methodology, which was derived from the LA Report. See R. at 1580-81. This methodology was applied to Mr. Skaar in the form of the May 2012 advisory opinion implementing the LA Report's dose estimate methodology, which specifically "excluded data from the on-site [urine] samples and attributed more significance to samples collected at later dates for the High 26 Group," of which Mr. Skaar was a member. R. at 2795. The Secretary argues, however, the exclusion of the urine samples from the pre-2013 methodology is irrelevant here because, in the decision on appeal, the Board expressly discounted the findings of the May 2012 advisory opinion as they were "based on an inaccurate dose estimate." R. at 10. But it is unclear how this makes any difference. It is undisputed that the dose estimate methodology under § 3.311, whether it be from pre- or post-2013, excluded certain urine dose samples. If Mr. Skaar is successful in showing this exclusion is not based on "sound scientific evidence" as required by VA's own regulations, then he will have suffered an injury-in-fact.

12 Mr. Skaar's injury is also "fairly traceable to the challenged conduct of the defendant." Spokeo, 136 S. Ct. at 1547. VA's own regulations require it to use "sound scientific evidence" in adjudicating radiation exposure claims, see 38 C.F.R. § 3.311, and VA is free to request dose estimates from private entities or to establish its own dose estimates procedures. Finally, Mr. Skaar's injury is "likely to be redressed by a favorable judicial decision." Spokeo, 136 S. Ct. at 1547. An order from us holding the Secretary is in noncompliance with § 3.311 and directing him to comply with the law would immediately give Mr. Skaar relief because he could not again be subject to the same allegedly unlawful process. Thus, Mr. Skaar has standing to bring the § 3.311 claim.

Having concluded Mr. Skaar has standing to challenge § 3.311 but not § 3.309, we have occasion to modify Mr. Skaar's proposed class definition to reflect our legal conclusions. See Suchanek v. Sturm Foods, Inc., 764 F.3d 750, 757 (7th Cir. 2014) (courts should modify proposed class definitions that are slightly overbroad rather than deny certification outright); Schorsch v. Hewlett-Packard Co., 417 F.3d 748, 750 (7th Cir. 2005) ("Litigants and judges regularly modify class definitions . . . ."); In re Monumental Life Ins. Co., 365 F.3d 408 (5th Cir. 2004); Robidoux v. Celani, 987 F.2d 931, 937 (2d Cir. 1993) ("A court is not bound by the class definition proposed in the complaint and should not dismiss the action simply because the complaint seeks to define the class too broadly."). But first, we must consider whether, as a matter of law, we have the power to certify class actions in the appeal context at all. We conclude we do.

B. The Power To Certify Class Actions in the Appeal Context

Before the passage of the Veterans' Judicial Review Act (VJRA), Pub. L. 100-687, 102 Stat. 4105 (1988), veterans were free to aggregate challenges to VA regulations in the limited context in which judicial review was available. See, e.g., Johnson v. Robison, 415 U.S. 361 (1974); Wayne State Univ. v. Cleland, 590 F.2d 627 (6th Cir. 1978); Giusti-Bravo v. U.S. Veterans Admin., 853 F. Supp. 34 (D.P.R. 1993); Nehmer v. U.S. Veterans' Admin., 118 F.R.D. 113 (N.D. Cal. 1987); In re "Agent Orange" Prod. Liab. Litig., 506 F. Supp. 762 (E.D.N.Y. 1980). Yet, until recently this Court did not recognize its authority to entertain class actions. See Monk II, 855 F.3d at 1320-21; Harrison, 1 Vet.App. at 439. In Monk II, the Federal Circuit disagreed, reasoning there was "no persuasive indication that Congress intended to remove class action protection for veterans when it enacted the VJRA." 855 F.3d at 1320 (emphasis in original). "Rather, Congress gave the Veterans Court express authority to prescribe rules of practice and procedure for its proceedings." Id. Thus, "[o]n the basis of th[is] express statutory authority . . . , the Veterans Court may prescribe procedures for class actions or other methods of aggregation." Id.

Although Monk II concerned a petition and this is an appeal, nothing in that decision indicates our authority to certify classes is limited to the petition context. Indeed, when describing the bases on which we had the power to certify classes, the Federal Circuit stated: "We hold that the Veterans Court has such authority [to certify and adjudicate class action cases] under the All Writs Act, other statutory authority, and the Veterans Court's inherent powers." Monk II, 855 F.3d at 1318. Although the reference to the All Writs Act arguably could be confined to the context of a petition (although that is not necessarily the case), the other two sources of authority to certify classes are not so limited. Moreover, the Federal Circuit specifically discussed our authority in the

13 context of an appeal. See id. at 1320. To be sure, that court had no occasion to rule on the question of class actions in the appeal context because Monk II concerned a petition. Nevertheless, its analysis is instructive. At a minimum, our inherent authority supports the use of the class action device as does our ability to craft rules of practice and procedure. See 38 U.S.C. § 7264(a). There is no principled distinction between the authority the Federal Circuit recognized for petitions from appeals. Thus, faithfully applying the Federal Circuit's logic in Monk II, we hold we possess the authority to certify class actions in the appeal context.

Having concluded we possess the power to aggregate claims and certify class actions in the appeal context, we now address whether we will exercise that power. We hold that, in appropriate circumstances, we will.

C. The Utility of Class Actions in the Appeal Context

Class actions are "an exception to the usual rule that litigation is conducted by and on behalf of the individual named parties only." Califano v. Yamasaki, 442 U.S. 682, 700-01 (1979). They are "a procedural device intended to advance judicial economy by trying claims together that lend themselves to collective treatment." Blaz v. Belfer, 368 F.3d 501, 504 (5th Cir. 2004). And they have a long history, originating with English "bills of peace," which allowed courts to consolidate numerous persons with the same claim against the same defendant. See Benjamin Kaplan, Continuing Work of the Civil Committee: 1966 Amendments of the Federal Rules of Civil Procedure, 81 HARV. L. REV. 356, 376 (1967). They have been an established part of Federal practice since the original version of Rule 23 was promulgated in 1937 and established three types of class actions plaintiffs could bring. See FED. R. CIV. P. 23(b) advisory committee's note to 1937 adoption. The Rule was revised to its current form in a landmark 1966 amendment laying out the procedural "measures which can be taken to assure the fair conduct of [class] actions." FED. R. CIV. P. 23(b) advisory committee's note to 1966 amendment; see also In re Gen. Motors Corp. Pick-Up Truck Fuel Tank Prods. Liab. Litig., 55 F.3d 768, 785 (3d Cir. 1995).

"Class relief is 'peculiarly appropriate' when the 'issues involved are common to the class as a whole' and when they 'turn on questions of law applicable in the same manner to each member of the class." Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 155 (1982) (quoting Yamasaki, 442 U.S. at 700-01). "[T]he class action device saves the resources of both the courts and the parties by permitting an issue potentially affecting every" class member "to be litigated in an economical fashion under Rule 23." Yamasaki, 442 U.S. at 701.

Class actions can also be an effective force for institutional change. As one court has observed, "[u]nless we can use the class action and devices built on the class action, our judicial system is not going to be able to cope with the challenges of [] mass repetitive wrongdoing." Cimino v. Raymark Indus., Inc., 751 F. Supp. 649, 652 (E.D. Tex. 1990), aff'd in part, vacated in part on other grounds by 151 F.3d 297 (5th Cir. 1998). The Federal Circuit has observed that "[c]lass actions can help [this Court] . . . by promoting efficiency, consistency, and fairness, and improving access to legal and expert assistance by parties with limited resources." Monk II, 855 F.3d at 1320. Further, "[c]lass actions may help [this Court] consistently adjudicate cases by increasing its prospects for precedential opinions," help "prevent the VA from mooting claims

14 scheduled for precedential review," and "could be used to compel correction of systemic error and to ensure that like veterans are treated alike." Id. at 1320-21.

We agree with the Federal Circuit's views on the utility of the class action device. Although that court made its comments in the petition context, the concepts of "efficiency, consistency, and fairness" apply equally to appeals. It is true this Court has the power to issue precedential decisions that, in some measure, mimic the effect of a class action. However, that power does not mean there is no use for the class action device. We conclude although our ability to issue binding precedent is a factor we should consider when deciding whether to certify a class (a matter we return to below), that ability does not counsel in favor of categorically rejecting the use of this procedural device.

Thus, as we have the power to certify class actions and will exercise our discretion to do so in appropriate cases, we now consider whether this matter is appropriate for certification. To do so requires precisely defining the proposed class. See FED. R. CIV. P. 23(c)(1)(B) (class action orders "must define the class and the class claims, issues, or defenses"). To do so we must have "a readily discernible, clear, and precise statement of the parameters defining the class or classes to be certified" that "provid[es] the parties with clarity and assist[s] class members in understanding their rights and making informed opt-out decisions." Marcus v. BMW of N.A., LLC, 687 F.3d 583, 591 (3d Cir. 2012).

D. The Proposed Class Composition

Mr. Skaar asks us to certify a class of "all U.S. veterans who were present at the 1966 cleanup of plutonium dust at Palomares, Spain[,] and whose application for service-connected disability compensation based on exposure to ionizing radiation the VA has denied or will deny." Mot. for Class Certification at 1. Combined with his later clarification of the class definition, the proposed class contains five subgroups.4 They are the following:

x Past Claimants: those Palomares veterans whose claims based on ionizing radiation exposure were denied before reaching the Board but who did not perfect an appeal of that denial; x Expired Claimants: those Palomares veterans whose claims based on ionizing radiation exposure the Board has denied but whose appeal windows to this Court have expired without the filing of a Notice of Appeal; x Present Claimants: those Palomares veterans whose claims based on ionizing radiation exposure the Board has denied and whose appeal windows to this Court have not yet expired or who have already appealed an adverse decision to this Court; x Present-Future Claimants: those Palomares veterans who have filed claims based on ionizing radiation exposure that remain pending before VA at any level and that VA will deny; and

4 We separate the class into subgroups merely for purposes of analyzing our jurisdiction as to each subgroup and do not divide the class into formal subclasses. See FED. R. CIV. P. 23(c)(5) (permitting district courts to divide a class into subclasses).

15 x Future-Future Claimants: those Palomares veterans who have developed a radiogenic condition but have not yet filed claims based on ionizing radiation exposure.

The proposed class composition depends on whether we have jurisdiction over each subgroup. First, we clearly have jurisdiction over the Present Claimants because they possess final Board decisions and either their 120-day windows to appeal those decisions to this Court have not yet expired or these claimants have already appealed within the 120-day time period. See 38 U.S.C. §§ 7252(a), 7266(a). We consider the remaining subgroups in turn.

1. The Present-Future and Future-Future Claimants

The Present-Future and Future-Future Claimants pose a unique jurisdictional issue. Neither subgroup has had final Board decisions dispose of its claims. Indeed, the Future-Future Claimants have not yet even filed disability compensation claims. We must decide whether our jurisdictional statute prohibits the inclusion of class members without a final Board decision as we have "an independent obligation to ensure that [we] do not exceed the scope of [our] jurisdiction." Henderson ex rel. Henderson v. Shinseki, 562 U.S. 428, 434 (2011). Relying on the Supreme Court's holding in Bowen v. City of New York, 476 U.S. 467 (1986), we conclude our jurisdictional statute does not prohibit their inclusion.

i. There is no indication Congress intended veterans to receive fewer procedural protections under the VJRA than they enjoyed before its enactment.

"Courts created by statute," like ours, "can have no jurisdiction but such as the statute confers." Christianson v. Indus. Operating Corp., 486 U.S. 800, 818 (1988). Subject-matter jurisdiction "can never be waived or forfeited." Gonzalez v. Thaler, 565 U.S. 134, 141 (2012). "A statute affecting federal jurisdiction must be construed both with precision and with fidelity to the terms by which Congress has expressed its wishes." Kucana v. Holder, 558 U.S. 233, 252 (2010). Guided by the Federal Circuit, we hold that, pursuant to our statutory authority under 38 U.S.C. §§ 7252 and 7261, we have the authority to certify class actions that include veterans who have not yet received a final Board decision and those who have not yet filed a claim. See Monk II, 855 F.3d at 1318.

We have only one source of jurisdiction: 38 U.S.C. § 7252. See Henderson, 562 U.S. at 434. It gives us "exclusive jurisdiction to review [Board] decisions," allowing us to "affirm, modify, or reverse" Board decisions and "remand the matter, as appropriate." 38 U.S.C. § 7252(a). Essentially, a final Board decision operates as the jurisdictional "trigger" that gives us the authority to hear a particular appeal. See Ledford v. West, 136 F.3d 776, 779 (Fed. Cir. 1998) (our Court's "jurisdiction is premised on and defined by the Board's decision concerning the matter being appealed"); Wick v. Brown (In re Wick), 40 F.3d 367, 373 (Fed. Cir. 1994) (a Board decision is a "statutory prerequisite for [this Court's] jurisdiction"). 38 U.S.C. § 7261 then lays out our scope of review in cases in which we already possess jurisdiction under section 7252 and "does not provide an independent basis for jurisdiction." Wick, 40 F.3d at 371; see also Dixon v. McDonald, 815 F.3d 799, 803 (Fed. Cir. 2016). Instead, this provision delineates what types of relief we may provide. See 38 U.S.C. §§ 7252(b) ("The extent of [this Court's judicial] review shall be limited to the scope provided in section 7261 . . . ."), 7261(a)(1)-(4) (laying out the various actions this Court can take

16 when deciding appeals). Both statutes play important, but differing roles. First, for jurisdiction to be proper in a given matter, it must lie under section 7252. Then, once jurisdiction is proper, section 7261 informs us what, if any, actions we may take.

In Harrison, we decided we lacked the authority to hear class actions because, among other reasons, section 7252 limited our jurisdiction to review of Board decisions. 1 Vet.App. at 439. But in Monk II, the Federal Circuit addressed that, stating Harrison "reflect[ed] a concern that the Veterans Court would exceed its jurisdiction if, for example, it certified a class that included veterans that had not yet received a Board decision or had not yet filed a notice appealing a Board decision." 855 F.3d at 1320. The Federal Circuit "disagree[d] that [our] authority is so limited," explaining that 38 U.S.C. § 7264(a), which authorizes us to create the procedures necessary to exercise our jurisdiction, allows us to "prescribe procedures for class actions or other methods of aggregation." Monk II, 855 F.3d at 1320. The Federal Circuit also noted that "[b]efore the VJRA, veterans seeking to enforce veterans benefit statutes were able to file class actions in some circumstances." Id. at 1319. In essence, the Federal Circuit's holding was supported by the notion that veterans should be afforded more procedural protections after the VJRA's enactment, not less.

Thus, absent any express indication from either Congress or the Federal Circuit that veterans in the context of an appeal should be afforded less procedural protections than were available to them before the VJRA's enactment, rather than more, we will not place such a restriction on this most favored class of citizens and their ability to pursue their disability benefits claims in the manner and fashion of their choosing. See Henderson, 562 U.S. at 441 ("We have long applied 'the canon that provisions for benefits to members of the Armed Services are to be construed in the beneficiaries' favor.'" (quoting King v. St. Vincent's Hosp., 502 U.S. 215, 220-21 n.9 (1991))).

The dissent asserts that section 7252(a) "contains the nonwaivable, jurisdictional elements that a veteran must have both filed a claim and received a Board decision." Post at 49, 50. The dissent goes on to reason that "[t]he majority's focus on determining whether to waive the requirement of a Board decision is at best premature because it did not explain why it determined that our jurisdictional statute has waivable components." Id. But, the dissent misreads our decision. We do not today hold that the requirement of a final Board decision is waivable. Rather, we hold that because Mr. Skaar, as class representative, has obtained a final Board decision pursuant to section 7252, the jurisdictional door has been opened, and we may use our other authorities, as explained in Monk II, to aggregate Mr. Skaar's claims with those of the remaining class members.

Our reasoning can be analogized to a magistrate judge's exercise of jurisdiction over a class action. 28 U.S.C. § 636 is jurisdictional in nature, and, in sum, provides that a magistrate judge can exercise jurisdiction over proceedings in civil matters with the consent of the parties. Roell v. Withrow, 538 U.S. 580, 585-86 (2003). Yet, even though section 636 is jurisdictional in nature, a magistrate can enter judgment in a class action without each class member giving consent. Koby v. ARS Nat'l Servs., Inc., 846 F.3d 1071, 1078-79 (9th Cir. 2017); Day v. Persels & Assocs., LLC, 729 F.3d 1309, 1324-25 (11th Cir. 2013); Dewey v. Volkswagen Aktiengesellschaft, 681 F.3d 170, 180-81 (3d Cir. 2012); Williams v. Gen. Elec. Capital Auto Lease, Inc., 159 F.3d 266, 268-69 (7th Cir. 1998). Thus, the jurisdictional mandates of section 636(c) are satisfied when only the named plaintiff in a class action has consented to proceed before a magistrate.

17 The courts to have considered the issue of consent in a class action have not "waived" the jurisdictional requirement of consent. Rather, they have held that the jurisdictional requirement is satisfied for all class members through the named plaintiff providing consent. Williams, 159 F.3d at 269 ("[T]he named representative . . . is the 'party' to the lawsuit who acts on behalf of the entire class, including with regard to the decision to proceed before a magistrate judge. This is an inherent part of representational litigation."). We find that Mr. Skaar's satisfaction of our jurisdictional requirement of a final Board decision, see 38 U.S.C. § 7252(a), is sufficient to vest this Court with subject matter jurisdiction, much in the same way a named plaintiff's consent to proceed before a magistrate is sufficient to grant the magistrate jurisdiction to enter final judgment as to all class members.

ii. We may certify classes that include claimants without final Board decisions.

The Secretary argues we lack jurisdiction to certify a class of veterans that includes those without a final Board decision "[b]ecause a Board decision is a jurisdictional prerequisite to review in this Court[.]" Secretary's Resp. to Mot. for Class Certification at 5. Thus, in his view, we could never certify a class of veterans without first ensuring there is a final Board decision as to each veteran in the class. In support, he relies on three Social Security cases: Weinberger v. Salfi, 422 U.S. 749 (1975); Yamasaki, 442 U.S. at 682; and City of New York, 476 U.S. at 467. We examine each in turn.

In Salfi, the District Court certified a class of claimants challenging a Social Security regulation that required a marriage to have existed at least 9 months before the death of a wage earner for a surviving spouse to receive benefits. The District Court held jurisdiction was proper under 28 U.S.C. § 1331 (the general Federal question jurisdictional statute), certified the class, and held the regulation unconstitutional. On direct appeal, the Supreme Court reversed, finding jurisdiction lay under 42 U.S.C. § 405 instead. That statute requires a final decision after a hearing by the Secretary of Health and Human Services before claimants can appeal adverse Social Security decisions to a district court. The Court concluded the District Court erred by certifying a class that included claimants who had not yet filed an application for benefits because "the [class] complaint was deficient in that it contain[ed] no allegations that [claimants] ha[d] even filed an application with the Secretary, much less that he has rendered any decision, final or otherwise." But, the Court went on to also hold that the District Court did not err in certifying a class of claimants who had filed a benefits application but had not yet been afforded a hearing—a nonjurisdictional requirement of § 405(g). The Court reasoned that the exhaustion requirement was not necessary when the issue was one that would be futile to bring before an agency. When read in isolation, Salfi is clearly disadvantageous to the proposed class members who do not have final Board decisions. However, as we will see, the lack of a final agency decision for each of a proposed class's members was not a concern for the Court 11 years later in City of New York.

Although the Secretary argues Yamasaki weighs against our having jurisdiction over the proposed class, we find it inapposite. There, the Supreme Court was confronted with a nationwide class of Social Security claimants whom the Government had overpaid. The Government sought to recoup those overpayments by withholding the respondents' future benefits. The respondents requested reconsideration or waiver of the recoupment. Two district courts then certified a

18 nationwide class of claimants and granted injunctive relief requiring the Agency to provide every class member with a pre-recoupment oral hearing. On appeal, the Court of Appeals for the Ninth Circuit affirmed. The Supreme Court needed to determine, among other things, whether section 405(g) "permits a federal district court to certify a nationwide class and grant injunctive relief." The Court concluded it did, reasoning Congress would have explicitly proscribed class actions in the Social Security context if it had intended to do so. Yamasaki is relevant here only to the extent the Court discusses the relief granted, not the lower courts' jurisdiction. The Court held "[w]ith respect to that relief, the classes certified were plainly too broad" as both classes "included persons who had not filed requests for reconsideration or waiver in the past and would not do so in the future." But that discussion was not key to the Court's holding, as it explained: "The Secretary's objection to the class definition is well taken, but it provides no basis for altering the relief actually granted in this case." 442 U.S. at 682. Thus, Yamasaki sheds no light on the question before us.

City of New York, however, bears a striking similarity to the matter before us. There, the Supreme Court considered a class of claimants challenging an internal policy of the Social Security Administration that operated to deny otherwise deserving claimants benefits to which they were entitled. "The gravamen of respondents' complaint was that petitioners had adopted an unlawful, unpublished policy under which countless deserving claimants were denied benefits." The District Court found the Government's internal policy invalid and certified a class that included both (i) claimants who had not appealed Social Security's decision within the required 60-day timeframe, thus requiring equitable tolling, and (ii) claimants who had not received a final agency decision. The Court of Appeals for the Second Circuit affirmed. 476 U.S. at 467.

The Supreme Court grappled with two issues in City of New York. The first, which we discuss elsewhere in this order, concerned whether the District Court erred by equitably tolling the statute of limitations for class members who had not timely appealed the Government's decision. The second issue, however, concerned whether the District Court lacked jurisdiction to certify a class that included claimants who had not received a final agency decision, as required by section 405(g). In Salfi, the Court called this requirement "central to the requisite grant of subject- matter jurisdiction" and, thus, claimants without a final decision could not be certified as part of a class. 422 U.S. at 764. But this time, in City of New York, the Court concluded section 405(g) was not a bar to class certification, even for claimants who had not received a final decision. This was so because (i) the class claims were "collateral to the claims for benefits that class members had presented administratively;" (ii) "the claimants . . . would be irreparably injured were the exhaustion requirement now enforced against them;" and (iii) "[t]he purposes of exhaustion would not be served by requiring these class members to exhaust administrative remedies." The Court further explained the class

stand[s] on a different footing from one arguing merely that an agency incorrectly applied its regulation. Rather, the District Court found a systemwide, unrevealed policy that was inconsistent in critically important ways with established regulations. Nor did this policy depend on the particular facts of the case before it; rather, the policy was illegal precisely because it ignored those facts. . . . Under these unique circumstances, there was nothing to be gained from permitting the compilation of a detailed factual record, or from agency expertise.

19 In addition, the relief afforded by the District Court is fully consistent with the policies underlying exhaustion. The court did not order that class members be paid benefits. Nor does its decision in any way interfere with the agency's role as the ultimate determiner of eligibility under the relevant statutes and regulations. Indeed, by ordering simply that the claims be reopened at the administrative level, the District Court showed proper respect for the administrative process. It did no more than the agency would have been called upon to do had it, instead of the District Court, been alerted to the charge that an undisclosed procedure was illegal and had improperly resolved innumerable claims.

476 U.S. at 485.

The Court also found its decision in Mathews v. Eldridge dispositive. There, the Court held "cases may arise where a claimant's interest in having a particular issue resolved promptly is so great that deference to the agency's judgment is inappropriate." 424 U.S. 319, 330 (1976). The Court in City of New York explained that "[t]wo factors influenced the Court's judgment that Eldridge was a case in which deference to the [A]gency's determination of finality was not necessary. First, the constitutional challenge brought there was 'entirely collateral to [a] substantive claim of entitlement.' Second, the claim rested 'on the proposition that full relief cannot be obtained' [as a result of the district court's decision]." 476 U.S. at 483 (citation omitted) (quoting Eldridge, 424 U.S. at 330-31). The City of New York Court was "especially sensitive to this kind of harm where the Government seeks to require claimants to exhaust administrative remedies merely to enable them to receive the procedure they should have been afforded in the first place." Id. at 484. The purposes of exhaustion include (i) permitting evidentiary development; (ii) allowing the agency to bring its expertise to bear on an issue before judicial review; and (iii) giving due respect to the agency's established procedures. City of New York, 476 U.S. at 486.

City of New York tells us an administrative exhaustion-of-remedies requirement can be waived where (i) the challenged conduct is collateral to a claim for benefits; (ii) enforcing the exhaustion requirement would irreparably harm the claimant; and (iii) the purposes of exhaustion would not be served by its enforcement. Turning to the instant appeal, we hold we have jurisdiction to certify a class action that includes members who do not have a final Board decision provided (i) the challenged conduct is collateral to the class representative's administratively exhausted claim for benefits—i.e., the class representative has obtained a final Board decision; (ii) enforcing the exhaustion requirement would irreparably harm the class; and (iii) the purposes of exhaustion would not be served by its enforcement.

Applying this test here, we have jurisdiction over the proposed class and will not require exhaustion of administrative remedies by each and every class member. First, jurisdiction over Mr. Skaar's appeal is proper under section 7252(a), for he has exhausted his administrative remedies, and the challenged conduct is collateral to both his and the unnamed class members' benefits claims because granting the requested relief would not be an "order that class members be paid benefits." City of New York, 476 U.S. at 486. "[A] claim is collateral when the 'plaintiffs' claims are essentially to the policy itself, not its application to them, nor to the ultimate substantive determination of their benefits.'" Stengel v. Callahan, 983 F. Supp. 1154, 1159 (N.D. Ill. 1997)

20 (quoting Johnson v. Sullivan, 922 F.2d 346, 353 (7th Cir. 1990)). Second, the alleged harm here, if shown to be true, is precisely the type of "harm where the Government seeks to require claimants to exhaust administrative remedies merely to enable them to receive the procedure they should have been afforded in the first place" the Supreme Court was concerned with in City of New York. 476 U.S. at 484. And, finally, the purposes of exhaustion would not be served by enforcement of section 7252(a)'s exhaustion requirement on the unnamed class members. The parties have compiled and agreed on a detailed factual record containing the Board's findings and conclusions. VA, through the Board, has brought its agency expertise to bear by providing a supplemental statement of reasons or bases addressing Mr. Skaar's challenge to § 3.311. See generally Secretary's Mar. 29, 2019, Resp. And, if the requested relief is granted, our order would not "in any way interfere with the [A]gency's role as the ultimate determiner of eligibility under the relevant statutes and regulations." See City of New York, 476 U.S. at 486. Thus, we waive the exhaustion requirement for the Present-Future and Future-Future Claimants, permitting them to be included in the proposed class.

Our reading of this caselaw is consistent with class action adjudication in the veterans' benefits context before the VJRA's enactment. For example, the lack of final Board decisions was not an impediment to pre-VJRA class certification in Nehmer. There, a district court certified a class of veterans challenging VA's implementation of 38 U.S.C. § 354, the Dioxin and Radiation Exposure Compensation Act, even though "[n]one of the named plaintiffs presented the claims raised in this lawsuit to the VA, either during their individual claim adjudications or in a petition for rulemaking[.]" The court reasoned the class members did not need to exhaust administrative remedies because (i) although VA may have had expertise in creating its procedures, "it does not possess particular expertise in determining what procedures adhere to the statutory mandate of the Dioxin Act and the Administrative Procedure Act;" (ii) a full record would be available through discovery; (iii) "the Court's hearing of the plaintiff's claims will not engender disrespect for the [A]gency's procedures;" (iv) the likelihood of the plaintiff's success by exhausting their administrative remedies was "low" because "the VA itself has adopted a system-wide policy; any errors committed in adopting the policy were made by the VA itself, not an individual fact-finder;" (v) "the class attack on the VA's procedural irregularities is distinct from any individual's attack on their denial of benefits;" and (vi) requiring exhaustion of remedies would place a "substantial burden" on the class members. Nehmer, 118 F.R.D. at 113. Nehmer, which predated the VJRA, thus fits with our holding today and, again, there is "no persuasive indication that Congress intended to remove class action protection for veterans when it enacted the VJRA." Monk II, 855 F.3d at 1320 (emphasis in original).

iii. This Court is the appropriate forum to hear challenges that are collateral to a benefits claim.

The remaining class claim here is collateral to Mr. Skaar's claim for benefits. Veterans cannot preemptively bring such collateral claims to VA seeking only to invalidate a specific procedure or practice. Instead, their only avenue would be to proceed to exhaust their administrative remedies by asking the Board to provide relief it is powerless to give. See 38 U.S.C. § 7104(c) (Board decisions are "bound by the regulations of the Department"). Congress cannot have intended such a result. Requiring every class member to have a final Board decision when the Board is powerless to provide the relief sought does not comport with the principle that, when interpreting statutory finality requirements, "[t]he prevailing rule of construction is that crucial

21 collateral claims should not be lost and that irreparable harm should be avoided." Mental Health Ass'n of Minn. v. Heckler, 720 F.2d 965, 969 (8th Cir. 1983). If veterans cannot aggregate actions to collaterally challenge alleged systemic wrongdoing before us, where should they seek such review? It is not enough to say Palomares veterans instead should have petitioned for rulemaking when the regulations at issue were drafted. See 38 U.S.C. § 553(e). If the class claim is proven, veterans could not have known and should not be required to have known their benefits claims would be subject to a legally invalid process. Thus, this Court is the appropriate forum to hear their collateral challenges to benefits claims.

Having concluded the Present, Present-Future, and Future-Future Claimants are members of the proposed class, we next consider the Expired Claimants.

2. The Expired Claimants

The Expired Claimants require a different analysis because they received final Board decisions but did not appeal them to this Court. Mr. Skaar asks us to exercise our discretion and waive section 7266(a)'s 120-day Notice of Appeal filing requirement, allowing their expired benefits claims to be revived before us, aggregated as part of the proposed class, and then, if the class prevails on the merits, returned to the Agency for readjudication. See Appellant's Mar. 21, 2018, Resp. at 3-4; see also Bove v. Shinseki, 25 Vet.App. 136, 140 (2011) (per curiam order), overruled on other grounds by Dixon v. McDonald, 815 F.3d 799 (Fed. Cir. 2016). We decline to do so.

As the Supreme Court explained in Henderson, section 7266(a)'s 120-day appeal window for obtaining review before this Court "does not have jurisdictional attributes" but nonetheless was "an important procedural rule," leaving it to us to determine whether and when waiver applied. 562 U.S. at 441. In Bove, we explained waiver is warranted "when circumstances precluded a timely filing despite the exercise of due diligence." 25 Vet.App. at 140. Those circumstances include (1) mental illness that renders one incapable of handling one's own affairs or other extraordinary circumstances beyond one's control; (2) reliance on incorrect statements by VA officials; or (3) misfilings at the regional offices or the Board. See, e.g., Brandenburg v. Principi, 371 F.3d 1362, 1364 (Fed. Cir. 2004) (misfiling); Barrett v. Principi, 363 F.3d 1316, 1321 (Fed. Cir. 2004) (mental illness); Bailey v. West, 160 F.3d 1360, 1365-68 (Fed. Cir. 1998) (en banc) (incorrect statement by VA official); McCreary v. Nicholson, 19 Vet.App. 324 (2005) (extraordinary circumstances). But this is not an exhaustive list because there are no bright line rules in the equitable tolling context. As the Federal Circuit recently reminded us, "the extraordinary circumstances element [of equitable tolling] necessarily requires a case-by-case analysis and not a categorical determination." James v. White, 917 F.3d 1368, 1373 (Fed. Cir. 2019).5

The Supreme Court dealt with a similar issue in City of New York. Recall there the Court upheld certification of a class of Social Security claimants that included those who had not

5 Given the case-by-case analysis equitable tolling requires and the prohibiting of the use of categorical rules under James, it is difficult to see how equitable tolling matters could be resolved through aggregate action. We leave for another day whether such a class would be appropriate, but the uncertainty on that question is an additional reason to exclude the Expired Claimants from the class here.

22 appealed adverse benefits determinations within the relevant appeal window. 476 U.S. at 486. The Court concluded equitable tolling was warranted. Id. at 482. This was so, the Court reasoned, because equitable tolling "served the purpose of the [Social Security] Act where . . . 'the Government's secretive conduct prevents plaintiffs from knowing of a violation of rights.'" Id. at 481 (quoting City of New York v. Heckler, 742 F.2d 729, 738 (1984)). But see Pittson Coal Grp. v. Sebben, 488 U.S. 104, 123 (1988) (finding equitable tolling was not warranted where "[t]he agency action was not taken pursuant to a secret, internal policy, but under a regulation that was published for all to see"). To the Court, the Government's conduct in City of New York represented one of the "cases [that] may arise where the equities in favor of tolling . . . are 'so great that deference to the agency's judgment [of finality] is inappropriate.'" 476 U.S. at 480 (quoting Eldridge, 424 U.S. at 330). Mr. Skaar essentially asks us to equate VA's adjudication of Palomares veterans' claims with the secretive conduct the Supreme Court found so reprehensible in City of New York, to extend Bove to such situations, and to allow equitable tolling here. We will not.

Including the Expired Claimants in the class offends the very notion of finality. Each of them received Board decisions and could have challenged VA's treatment of Palomares veterans just like Mr. Skaar, yet each chose not to. Mr. Skaar has presented no reason for us to depart from Bove's principle that the 120-day Notice of Appeal window to this Court will only be waived "when circumstances precluded a timely filing despite the exercise of due diligence." 25 Vet.App. at 140. Indeed, he has never alleged the Expired Claimants were precluded from timely filing appeals to this Court for any reason other than VA's historical practice in adjudicating claims from Palomares veterans. But before a claimant succeeds in changing the law, VA will always (presumably) adjudicate claims in accord with its own interpretation of that law and our legal pronouncements. Thus, there is no principled way to distinguish the Expired Claimants here and any other claimants who have been denied benefits, failed to appeal to this Court, and later discovered their benefits denial was based on an incorrect reading of the law. The proper course for such claimants is to file supplemental claims based on new and relevant evidence with VA, see 38 C.F.R. § 3.2501, not to attempt to skirt finality and existing precedent merely because of the novel procedural nature of this case.

The unfair substantive legal advantage the Expired Claimants would enjoy if we permitted them to join the class is illustrated by a recent Court decision, Ray v. Wilkie, 31 Vet.App. 58 (2019). There, a panel of the Court held VA's historical practice of refusing to define a key regulatory phrase in 38 C.F.R. § 4.16(b) frustrated judicial review, warranting remand in cases where the phrase is undefined. Id. at 73-74. The Ray decision surely benefited the named appellant. And it also benefited any claims involving that regulation currently pending before the Court or VA. But it certainly provided no retrospective relief for claimants who had been denied benefits previously but whose appeal windows had expired.

Or consider this matter. Had Mr. Skaar filed the instant appeal, not sought class certification, and succeeded on the merits, his appeal would be decided through precedential decision. That decision would bind Mr. Skaar and any and all claimants with claims currently pending before VA and the Court (the Present, Present-Future Claimaints) as well as any claimants with claims filed in the future (Future-Future Claimants). But there would be no authority to support that precedential decision reviving expired claims, as Mr. Skaar asks us to do here.

23 At first glance, our exclusion of the Expired Claimants may seem unduly harsh. But claimants in the veterans benefits system do not face the same consequences of finality as litigants in traditional civil litigation. Instead, under 38 U.S.C. § 5108(a) and 38 C.F.R. § 20.1105(a), if the class succeeds on the merits, then the Expired Claimants can file supplemental claims based on new and relevant evidence. The Expired Claimants may not enjoy the same effective date protections as the other subgroups within the class, but they would still have an avenue to service connection available to them.

In sum, that this is a class action does not and should not change this analysis as the class action device is a procedural rule that, if we are to employ it, should not yield substantive legal benefits. We will not now excuse the Expired Claimants' lack of diligence in pursuing their claims, depart from precedent, and grant retrospective relief merely because this is a class action. Thus, we decline to equitably toll the Expired Claimants' claims and modify the proposed class to exclude them. See FED. R. CIV. P. 23(c)(5); Suchanek, 764 F.3d at 757; Schorsch, 417 F.3d at 750; Robidoux, 987 F.2d at 937.

3. The Past Claimants

The Past Claimants were denied by VA but never reached the Board because they did not perfect an administrative appeal. For our purposes, they are akin to the Expired Claimants in that they have no final Board decisions. But unlike the Expired Claimants, that is not because they failed to appeal their denials to this Court. Instead, these claimants were denied by some part of VA other than the Board. Thus, if they are to be included in the class, they require equitable tolling of their appellate review windows before VA. See Jaquay v. Principi, 304 F.3d 1276, 1286 (Fed. Cir. 2002), overruled on other grounds by Henderson v. Shinseki, 589 F.3d 1201 (Fed. Cir. 2009); Hunt v. Nicholson, 20 Vet.App. 519, 522 (2006) ("[T]he same principles that guided the Federal Circuit in allowing equitable tolling of the deadline for filing appeals to this Court apply with equal force to tolling the deadline for filing Substantive Appeals."). For the same reasons we decline to equitably toll the appeal windows for the Expired Claimants, we decline to do so for the Past Claimants as well and modify the proposed class to exclude them. There is simply no principled distinction between the proposed class here and any other individual challenge to VA action that warrants excusing the Past Claimants' lack of diligence in preserving their claims.

Considering Mr. Skaar lacks standing to bring the § 3.309 claim but possesses standing to pursue the § 3.311 claim and considering our exclusion of the Expired and Past Claimants from class certification, we must modify the proposed class definition. See FED. R. CIV. P. 23(c)(5); see also Suchanek, 764 F.3d at 757. Thus, we modify the proposed class definition as follows: All U.S. veterans who were present at the 1966 cleanup of plutonium dust at Palomares, Spain, and whose application for service-connected disability compensation based on exposure to ionizing radiation VA has denied or will deny by relying, at least in part, on the findings of dose estimates requested under 38 C.F.R. § 3.311, except those whose claims have been denied and relevant appeal windows of those denials have expired, or those whose claims have been denied solely based on dose estimates obtained before 2001. With this modified definition in mind, we now turn to the class certification analysis.

24 E. Class Certification Analysis

At this time, our Court has no rule of procedure governing class actions. Indeed, as far as we are aware, we are the only appellate court in the Nation with the authority to aggregate actions in the first instance. But while we are unique in that regard, we are not starting with a blank slate. As alluded to before, the Federal Rules of Civil Procedure provide for class actions in Rule 23. As we did in the petition context, see Monk III, 30 Vet.App. at 174, we adopt Rule 23 as a guide for class certification in the appeal context. Also, as with petitions, see id., we have at least some limited factfinding ability in the context of determining whether a class should be certified.

"Rule 23 does not set forth a mere pleading standard. A party seeking class certification must affirmatively demonstrate his [or her] compliance with the Rule[.]" Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350 (2011). A party seeking class certification must demonstrate by a preponderance of the evidence the four requirements of Rule 23(a), and at least one of the requirements of Rule 23(b).6 See N.J. Carpenters Health Fund v. Rali Series 2006-Q01 Tr., 477 F. App'x 809, 812 (2d Cir. 2012); see also Wal-Mart Stores, Inc., 564 U.S. 338 at 351 ("A party seeking class certification . . . must be prepared to prove that there are in fact sufficiently numerous parties, common questions of law or fact, etc.") (emphasis in original).

Rule 23(a) requires (1) the class be "so numerous that joinder of all members is impracticable;" (2) there be common questions of law or fact; (3) the claims or defenses of the named representative be typical of the class; and (4) the class representatives "fairly and adequately protect the interests of the class." FED. R. CIV. P. 23(a). These requirements "effectively 'limit[] the class claims to those fairly encompassed by the named plaintiff's claims.'" Falcon, 457 U.S. at 156 (quoting Gen. Tel. Co. of Sw. v. EEOC, 446 U.S. 318, 330 (1980)). Rule 23(b)(2), the relevant subsection here, states class actions are appropriate when "the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that final injunctive or corresponding declaratory relief is appropriate respecting the class as a whole." Taken together, the Rule 23 analysis tells us "whether the named plaintiff's claim and the class are so interrelated that the interests of the class members will be fairly and adequately protected in their absence," while protecting defendants' rights. Falcon, 457 U.S. at 157.

We must conduct "a rigorous analysis" of the proposed class, Falcon, 457 U.S. at 160-61, that may "entail some overlap with the merits of the plaintiff's underlying claim" as the "class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff's cause of action," Comcast Corp. v. Behrend, 569 U.S. 27, 33 (2013).

6 Although not explicitly listed under Rule 23, many courts have required that class membership be "ascertainable." See, e.g., Ward v. EZCorp, Inc., 679 F. App'x. 987 (11th Cir. 2017); Leyse v. Lifetime Entm't Servs., LLC, 679 F. App'x 44, 47 (2d Cir. 2017); Sandusky Wellness Ctr., LLC v. Medtox Sci., Inc., 821 F.3d 992, 996 (8th Cir. 2016); Mullins v. Direct Dig., LLC, 795 F.3d 654, 659 (7th Cir. 2015); see also McKeage v. TMBC, LLC, 847 F.3d 992, 998 (8th Cir. 2017) ("A class may be ascertainable when its members may be identified by reference to objective criteria."); Rikos v. Procter & Gamble Co., 799 F.3d 497, 525 (6th Cir. 2015). We need not weigh in on this debate here because it is clear ascertainability is not required for Rule 23(b)(2) classes such as the one at issue here. See Shook v. El Paso City, 386 F.3d 963, 972 (10th Cir. 2004) ("while the lack of identifiability is a factor that may defeat Rule 23(b)(3) class certification, such is not the case with respect to class certification under Rule 23(b)(2)"); Yaffe v. Powers, 454 F.2d 1362, 1366 (1st Cir. 1972); Shelton v. Bledsoe, 775 F.3d 554, 561 (3d Cir. 2015); Cole v. City of Memphis, 839 F.3d 530, 541-42 (6th Cir. 2016).

25 But, crucially, "[i]n determining the propriety of a class action, the question is not whether the plaintiff or plaintiffs have stated a cause of action or will prevail on the merits, but rather whether the requirements of Rule 23 are met." Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 177-78 (1974). "Rule 23 grants courts no license to engage in free-ranging merits inquiries at the certification stage." Amgen, Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 466 (2013). Instead, "[m]erits questions may be considered to the extent—and only to the extent—that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied." Id. As we explain below, the proposed class meets the requirements for class certification for the remaining class claim.

1. The proposed class is so numerous that joinder would be impracticable.

To warrant certification under the Federal Rules of Civil Procedure, the proposed class must be "so numerous that joinder of all members is impracticable." FED. R. CIV. P. 23(a)(1). This requirement is a bit of a square peg in a round hole at this Court. In Federal district court, parties have numerous devices they may use to "join" additional parties. See, e.g., FED. R. CIV. P. 19 (mandating joinder of certain parties), 20 (allowing joinder of certain other parties), 22 (interpleader), 24 (intervention). The rules thus make the class action a more exceptional device with stringent requirements because there are alternative means for parties to join others in a proceeding that do not require the binding of absent parties. We have no comparable joinder devices.7 Thus, asking if joinder in an appeal is "impracticable" does not make the same sense here as doing so in a district court. If anything, given the difficulty in terms of "joinder" before our Court, the numerosity standard would likely be met on a lesser showing than in a district court. In any event, it is met here under any standard.

Numerosity need not be proven exactly. See, e.g., Hinman v. M&M Rental Ctr., Inc., 545 F. Supp. 2d 802, 806 (N.D. Ill. 2008). "[C]ourts generally find that the numerosity factor is satisfied if the class comprises 40 or more members and will find that it has not been satisfied when the class comprises 21 or fewer." Celano v. Marriott Int'l, Inc., 242 F.R.D. 544, 549 (N.D. Cal. 2007); see Lightfoot v. District of Columbia, 246 F.R.D. 326, 335 (D.D.C. 2007) ("Courts in this District have generally found that the numerosity requirement is satisfied and that joinder is impracticable where a proposed class has at least forty members."). But "[t]here is no minimum number of members needed for a suit to proceed as a class action." Marcus v. BMW of N. Amer., LLC, 687 F.3d 583, 595 (3d Cir. 2012). "[I]t is permissible for a plaintiff to make reasonable inferences drawn from available facts" and "an 'information monopoly [by the party opposing the class] will not stand in the way of persons seeking relief.'" Violette v. P.A. Days, Inc., 214 F.R.D. 207, 213 (S.D. Ohio 2003) (quoting Jackson v. Foley, 156 F.R.D. 538, 542 (E.D.N.Y. 1994)). Additionally, the numerosity requirement is relaxed for classes seeking injunctive relief. Sueoka v. United States, 101 F. App'x. 649, 653 (9th Cir. 2004) ("Because plaintiffs seek injunctive and declaratory relief, the numerosity requirement is relaxed and plaintiffs may rely on the reasonable inference arising from plaintiffs' other evidence that the number of unknown and future members . . . is sufficient to make joinder impracticable."). And although "[n]umerosity is more than a numbers game," Howard's Rexall Stores, Inc. v. Aetna U.S. Healthcare, Inc., No. CIV. oo-CV- 7 Indeed, our rules do not even expressly allow for joinder, much less describe how parties are to seek it. Thus, in so far as the numerosity requirement asks whether "joinder of all members is impracticable," it would appear to always be answered in the affirmative in proposed class actions before us until we craft such a rule.

26 31B, 2001 WL 501055, at *6 (D. Me. May 8, 2001), "[w]hen class size reaches substantial portions, . . . the impracticability requirement is usually satisfied by numbers alone." In re Am. Med. Sys., Inc., 75 F.3d 1069, 1079 (6th Cir. 1996).

In response to a Court order requesting more information, the Secretary stated that, per the Department of Defense, 1,388 U.S. military personnel participated in the Palomares nuclear cleanup. See Secretary's Dec. 13, 2018, Resp. The order also asked him to provide information relating to certain categories of veterans in the proposed class. But instead, the Secretary explained VA's "internal databases are not equipped to furnish the Court with the number of veterans falling within the" class's various subcategories. Id. In reply, Mr. Skaar questioned the Secretary's compliance with our order and noted the record reflects there are "at least nineteen veterans who had filed claims for Palomares-related disabilities with the VA, 'including three appeals for reassessment for a total of 22 claims.'" Appellant's Jan. 4, 2019, Resp. at 3 (quoting R. at 1580). Given the overall number of veterans present at Palomares, the relaxed numerosity standard for classes seeking injunctive relief, see Sueoka, 101 F. App'x. at 653, and Mr. Skaar's additional information concerning the claims made, we may reasonably infer the proposed class contains potentially up to 1,388 veterans and at least 22, a number sufficient to satisfy the numerosity requirement. See, e.g., Lightfoot, 246 F.R.D. at 335. Thus, we hold the class satisfies the numerosity requirement.

2. The proposed class presents a common issue capable of classwide resolution.

The second Rule 23 requirement for class certification, commonality, "requires the plaintiff to demonstrate that the class members have suffered the same injury. This does not mean merely that they have all suffered a violation of the same provision of law." Wal-Mart, 564 U.S. at 350. Rather, "[t]heir claims must depend upon a common contention." Id. "That common contention, moreover, must be of such a nature that it is capable of classwide resolution – which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke." Id. "[F]or purposes of Rule 23(a)(2) [e]ven a single [common] question will do." Id. "What matters to class certification . . . [is] the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation." Id. "The critical point is 'the need for conduct common to members of the class.'" Suchanek, 764 F.3d at 756 (quoting In re IKO Roofing Shingle Prods. Liab. Litig., 757 F.3d 599, 602 (7th Cir. 2014)) (emphasis in original). "Where the same conduct or practice by the same defendant gives rise to the same kind of claims from all class members, there is a common question." Suchanek, 764 F.3d 750, 756 (7th Cir. 2014); see In re Nat'l Football League Players Concussion Injury Litig., 821 F.3d 410 (3d Cir. 2016).

The Secretary concedes the proposed class would satisfy the commonality requirement if the class is limited "to include only those veterans whose applications [for service-connected disabilities] were denied based on § 3.311[.]" See Secretary's Feb, 20, 2018, Resp. at 17. Considering our dismissal of the class challenge to § 3.309 and corresponding modification of the class definition, this is an effective concession of commonality as to the class challenge under § 3.311 as only "those veterans whose applications were denied based on § 3.311" would qualify as class members. Further, we agree commonality is met for this claim. The class members' claims "depend upon a common contention"—that VA's dose estimate procedures do not rely on "sound

27 scientific and medical evidence" in contravention to § 3.311(c)(1)(i)—that "is capable of classwide resolution"—in the form of an order enjoining the Secretary from denying claims under § 3.311 until VA's procedures comply with the regulation. Wal-Mart, 564 U.S. at 350.

3. Mr. Skaar's claim is typical of that of the proposed class.

Class certification also requires that "the claims or defenses of the representative parties are typical of the claims or defenses of the class." FED. R. CIV. P. 23(a)(3). This inquiry focuses on whether "in pursuing his own claims, the named plaintiff will also advance the interests of the class members." In re Am. Med. Sys., 75 F.3d 1069, 1082 (6th Cir. 1996). Or, put differently, "as goes the claim of the named plaintiff, so go the claims of the class." Sprague v. Gen. Motors Corp., 133 F.3d 388, 399 (6th Cir. 1998). Although distinct, the typicality requirement overlaps with certain other requirements of Rule 23(a). In particular, "[t]he commonality and typicality requirements . . . tend to merge." Falcon, 457 U.S. at 157 n.13.

Courts will deny class certification "when the variation in claims" between a class representative and absent class members "strikes at the heart of the respective causes of actions." Deiter v. Microsoft Corp., 436 F.3d 461, 466-67 (4th Cir. 2006). The class representative's claims need not be identical, but must "share the same essential characteristics as the claims of the class at large." Haggart v. United States, 89 Fed. Cl. 523, 534 (2009); Arreola v. Godinez, 546 F.3d 788, 798 (7th Cir. 2008). "The test of typicality 'is whether other members have the same or similar injury, whether the action is based on conduct which is not unique to the named plaintiffs, and whether other class members have been injured by the same conduct.'" Wolin v. Jaguar Land Rover N.A., LLC, 617 F.3d 1168, 1175 (9th Cir. 2010) (quoting Hanon v. Dataproducts Corp., 976 F.2d 497, 508 (9th Cir. 1992)). "[T] he typicality prong of Rule 23(a) sets a relatively low threshold." Karvaly v. eBay, Inc., 245 F.R.D. 71, 82 (E.D.N.Y. 2007); see, e.g., Stirman v. Exxon Corp., 280 F.3d 554, 562 (5th Cir. 2002); Lightbourn v. Cnty. of El Paso, Tex., 118 F.3d 421, 426 (5th Cir. 1997). Typicality is also easier to satisfy where classes seek injunctive relief. See Baby Neal ex. Rel. Kanter v. Casey, 43 F.3d 48 (3d Cir. 1994).

The Secretary argues Mr. Skaar's claim is not typical enough to permit him to serve as class representative because the reason for any denials of Palomares veterans' claims related to ionizing radiation exposure may not turn on the results of dose estimates requested under § 3.311. See Secretary's Feb. 20, 2018, Resp. at 17-19; Secretary's July 27, 2018, Resp. at 8-11. Much like any concerns regarding commonality and standing, this concern is alleviated by our restructuring of the class. As explained above, because we are dismissing the class challenge to § 3.309 for lack of standing, the only issue before us concerns those claims that have either been denied or will be denied under § 3.311.

And as discussed above regarding Mr. Skaar's standing to represent the class, the Secretary's argument that Mr. Skaar lacks standing to represent class members whose claims had been denied under the Air Force's pre-2013 methodology also presents potential typicality concerns. But, as we explained, the pre- and post-2013 distinction is largely theoretical. Put simply, Mr. Skaar shares the same injury from VA's reliance on Air Force's dose estimates as any conceivable claimant falling within the modified class. Thus, his claim "share[s] the same essential

28 characteristics as the claims of the class at large," and his claim is typical enough to permit him to serve as class representative. Haggart, 89 Fed. Cl. at 534.

4. Mr. Skaar will fairly and adequately protect the interests of the class.

The final Rule 23(a) inquiry asks whether "the representative parties will fairly and adequately protect the interests of the class." FED. R. CIV. P. 23(a)(4). "A decision with respect to the class is conclusive only if the absent members were adequately represented by the named litigants and class counsel." In re Bridgestone/Firestone, Inc., Tires Prods. Liab. Litig., 333 F.3d 763, 768 (7th Cir. 2003), abrogated on other grounds by Smith v. Bayer Corp., 564 U.S. 299 (2011). 8 "Adequacy is twofold: the proposed class representative must have an interest in vigorously pursuing the claims of the class, and must have no interests antagonistic to the interests of other class members." In re Literary Works in Elec. Databases Copyright Litig., 654 F.3d 242, 249 (2d Cir. 2011). Thus, "[t]he adequacy inquiry under Rule 23(a)(4) serves to uncover conflicts of interest between named parties and the class they seek to represent." Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 626 (1997). Class representatives serve as fiduciaries for certified classes. See London v. Wal-Mart Stores, Inc., 340 F.3d 1246, 1254 (11th Cir. 2003).

To be adequate, class representatives must possess the claim asserted on behalf of the class, have interests otherwise aligned with and not antagonistic to those of the class, and be able to advocate vigorously and competently for the interests of the class. See Kirkpatrick v. J.C. Bardford & Co., 827 F.2d 718, 727 (11th Cir. 1987). For much of the same reasons typicality and commonality are present here, we hold Mr. Skaar is adequate to serve as class representative. He possesses the same claim as the unnamed class members, his interest in VA complying with § 3.311(c)(1)(i) is aligned with the class, and there is no indication he is unable to vigorously and competently advocate for the interests of the class. Id. Moreover, we see no conflict of interest that would prevent Mr. Skaar from advancing the interests of the class.

5. The requested injunctive relief is appropriate respecting the class as a whole.

Federal Rule of Civil Procedure 23(b)(2) permits aggregation when all Rule 23(a)'s perquisites have been met, and "the party opposing the class has acted or refused to act on grounds that apply generally to the class, so that injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole." The Supreme Court has instructed that "[t]he key to the (b)(2) class is 'the indivisible nature of the injunctive or declaratory remedy warranted – the notion that the conduct is such that it can be enjoined or declared unlawful only as to all of the class members or as to none of them." Wal-Mart, 564 U.S. at 360 (quoting Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. REV. 97, 132 (2009)). Rule 23(b)(2) requires that "a single injunction or declaratory judgment . . . provide relief to each member of the class." Id. Thus, if there are class members who would not benefit from a class- wide injunction (or declaration), certification under Rule 23(b)(2) would not be appropriate. See Jennings v. Rodriguez, 138 S. Ct. 830, 852 (2018) (commenting in action concerning claims by detained aliens that, because some members of the class may not be entitled to the requested relief, certification under Rule 23(b)(2) might be inappropriate).

8 We consider the adequacy of class counsel below.

29 We hold the proposed class meets Rule 23(b)(2)'s requirements for certification. The class seeks a single class-wide injunction ordering VA to comply with the provisions of § 3.311. And with the dismissal of the class challenge to § 3.309 and the restriction of the class to those claimants who have been or will be subject to § 3.311, there is no question that, if the class succeeds on the merits, "injunctive relief or corresponding declaratory relief"—in the form of an order from this Court to the Secretary that he comply with the provisions of § 3.311—"is appropriate respecting the class as a whole." FED. R. CIV. P. 23(b)(2).

6. The class action device is a superior method of litigating the class claim.

Having concluded Rule 23(a) and Rule 23(b)(2) are satisfied, we could stop our certification analysis were we sitting as a district court. However, we are not. We have used Rule 23 as a "guide" for class certification. But we are not bound by it. See Int'l Union, UAW, Local 283 v. Scofield, 382 U.S. 205, 217 n.10 (1965) (the "Federal Rules of Civil Procedure . . . apply only in the federal district courts"); FED. R. CIV. P. 1 ("These rules govern the procedure in the United States district courts."). As we mentioned earlier in our discussion, to our knowledge, we are the only appellate body in the Nation with the authority to aggregate actions in the first instance. Our appellate nature and national jurisdiction make us stand apart from the ordinary course of aggregate litigation in Federal district courts, which are empowered to find facts and conduct discovery while we are not, absent some limited circumstances. See 38 U.S.C. § 7261(c) ("In no event shall findings of fact made by the Secretary or the Board of Veterans' Appeals be subject to trial de novo by the Court."); § 7252(b) ("Review in the Court shall be on the record of proceedings before the Secretary and the Board."); but see Monk III, 30 Vet.App. at 171 (holding this Court "has authority to conduct limited factfinding to determine whether class certification is warranted"); Bove, 25 Vet.App. at 143 ("[T]his Court . . . may seek facts outside the record before the Board and independently weigh the facts to determine if equitable tolling is appropriate."); Erspamer v. Derwinski, 1 Vet.App. 3, 10 (1990) (Court may consider facts not before the Board when considering the merits of a petition for extraordinary relief). Moreover, we are different than district courts because we can issue precedential decisions that bind those not before the Court. In other words, unlike district courts, our decisions can have something like the effect of a class action judgment without receiving class treatment.

As we explain below, class actions before us will serve as a special procedural device for certain types of claims that lend themselves to aggregate adjudication. This is because class actions "conserve judicial resources by allowing courts to treat common claims together, obviating the need for repeated adjudications of the same issues." Cochran v. Volvo Grp. N.A., LLC, No. 1:11- CV-927, 2013 WL 1729103, at *1 (M.D.N.C. Apr. 22, 2013). They also relieve absent class members from having to bring and litigate complex claims individually. "[A]n absent class-action plaintiff is not required to do anything. He [or she] may sit back and allow the litigation to run its course, content in knowing that there are safeguards provided for his [or her] protection." Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 810 (1985). Especially in an adjudicatory system involving large numbers of unrepresented claimants, class actions may allow claimants, such as Mr. Skaar, who have the resources, knowledge, and desire to challenge VA conduct and regulations to step forward and represent similarly situated claimants and, through notice of certification, educate

30 other class members about the existence of a legal claim against the VA. See Watkins v. Simmons & Clark, Inc., 618 F.2d 398, 404 (6th Cir. 1980).

But our unique nature requires considerations beyond those applicable to district courts under Rule 23. Just as there are reasons in favor of exercising our discretion to certify a class in a particular matter, so, too, are there reasons counseling against certification. In Harrison, we declined to adopt class action procedures because (i) we believed we lacked the power to adopt such procedures; (ii) the potential difficulties in managing class actions in the first instance at the appellate level; and (iii) the availability of precedential decision-making as a superior form of litigation. 1 Vet.App. at 439. As we stated in Monk III, the Federal Circuit has expressly overruled Harrison's first factor, lack of authority. 30 Vet.App. at 171 n.5. In Monk III, we declined to decide whether the remaining two Harrison factors were appropriate considerations for class certification. Id. We now explain that the remaining two Harrison factors—manageability and the availability of precedential decisions—stem from the unique nature of this Court and are relevant considerations in the class certification analysis before this Court, even if they are not categorical reasons to decline to certify class actions.

While we recognize for traditional Rule 23(b)(2) class actions, "superiority [is] self- evident," Wal-Mart, 564 U.S. at 363, our national jurisdiction makes the inquiry different here. Requiring claimants to justify the use of the class action device considering the available alternatives, such as single-party precedential decisions, consolidation, petitions for rulemaking, and the ability to issue writs of mandamus, is necessary to justify the expenditure of judicial time and energy required to adjudicate class actions as an appellate court in the first instance and assume the risk of prejudicing the rights of absent veterans. See Pipefitters Local 636 Ins. Fund v. Blue Cross Blue Shield, 654 F.3d 618, 630-31 (6th Cir. 2011). Thus, considering our appellate nature and limited factfinding abilities and guided by Rule 23, class actions before this Court are the exception, not the rule. In other words, we will presume classes should not be certified because our ability to render binding precedential decisions ordinarily will be adequate. Claimants seeking class certification can rebut this presumption by showing by a preponderance of the evidence that a class action is "superior to other available methods for fairly and efficiently adjudicating the controversy" before we will exercise our discretion in certifying a class. FED. R. CIV. P. 23(b)(3). This is a "fact-specific analysis" that "will vary depending on the circumstances of any given case." Madison v. Chalmette Ref., L.L.C., 637 F.3d 551, 555 (5th Cir. 2011).

Rule 23(b)(3) lists several factors for determining the superiority of a class action. This is at least a useful starting point. Of these, only 23(b)(3)(D) is relevant here.9 That factor addresses "the likely difficulties in managing a class action," a highly relevant concern given our previously

9 Subsection (A) looks at "the class members' interests in individually controlling the prosecution or defense of separate actions. FED. R. CIV. P. 23(b)(3)(A). But absent claimants are already bound by our precedential decisions, see 38 U.S.C. § 7269, and thus we need not require this factor. Subsection (B) considers "the extent and nature of any litigation concerning the controversy already begun by or against class members." FED. R. CIV. P. 23(b)(3)(B). Our national jurisdiction addresses this factor. See 38 U.S.C. § 7269. Duplicative legal issues can already be brought in this Court and we have adequate means to address them. See U.S. VET. APP. R. 5(a)(3) (allowing us to stay matters pending before the Court "in the interest of judicial efficiency"). Finally, subsection (C) is not relevant here as we are the appropriate forum for claimants to challenge VA's denial of benefits. See FED. R. CIV. P. 23(b)(3)(C) (listing "the desirability or undesirability of concentrating the litigation of claims in the particular forum" as a 23(b)(3) factor); see also 38 U.S.C. §§ 7252, 7261.

31 discussed limitations. FED. R. CIV. P. 23(b)(3)(D). Manageability "encompasses the whole range of practical problems that may render the class action format inappropriate for a particular suit." Eisen, 417 U.S. at 164. Courts have declined to certify classes because of manageability concerns where individual class members brought claims in different states under different state laws, see Riordan v. Smith Barney, 113 F.R.D. 60, 66 (N.D. Ill. 1986); communication with some class members would be unduly difficult, see Mateo v. The M/S Kiso, 805 F. Supp. 761, 774 (N.D. Cal. 1991); individual damages calculations would be too complex, see Abrams v. Interco, Inc., 719 F.2d 23, 31 (2d Cir. 1983); the class required too many individualized determinations, see Danvers Motor Co., Inc. v. Ford Motor Co., 543 F.3d 141, 149 (3d Cir. 2008); and the sheer size of the class made effecting notice and providing opt out rights unmanageable, see Gaffney v. United States, 834 F. Supp. 1, 6 (D.D.C. 1993). Importantly, the "focus is not on the convenience or burden of a class action suit per se, but the relative advantages of a class action suit over whatever other forms of litigation might be realistically available" to claimants. Klay v. v. Humana, Inc., 382 F.3d 1241, 1269 (11th Cir. 2004); see also Johnston v. HBO Film Mgmt., 265 F.3d 178, 194 (3d Cir. 2001) (class action must represent the best available method for fair and efficient adjudication to warrant certification). But again, we only use Rule 23 as a guide. It is imperfectly crafted for our appellate setting and Rule 23(b)(3)(D)'s baseline is only the starting point of our analysis. In the balance of this section, we provide a non-exhaustive set of factors we will consider when deciding if a claimant has rebutted the presumption against aggregate action.

After canvassing federal class action jurisprudence and considering our unique appellate nature, we hold that, when considering whether the presumption against aggregate action has been rebutted, the Court will consider, as appropriate, whether (i) the challenge is collateral to a claim for benefits; (ii) litigation of the challenge involves compiling a complex factual record; (iii) the appellate record is sufficiently developed to permit judicial review of the challenged conduct; and (iv) the putative class has alleged sufficient facts suggesting a need for remedial enforcement. No one of these factors is more or less important than the others, rather the Court must engage in a case-by-case balancing to determine whether class certification is appropriate.

The first factor, whether the challenge is collateral to a claim for benefits, focuses on whether "the 'plaintiffs' claims are essentially to the policy itself, not its application to them, nor to the ultimate substantive determination of their benefits.'" Stengel, 983 F. Supp. at 1159 (quoting Johnson, 922 F.2d at 346). Such claims are "not essentially a claim for benefits" because they do "not merely challeng[e] the merits of the" agency's ultimate benefits determination. Id. In appeals involving clear regulatory or constitutional attacks on VA's application of a regulation such as this one, determining whether a matter is collateral will likely involve a simpler analysis than those instances where the regulatory or constitutional challenge is necessarily intertwined with VA's merits determination. Thus, the proper focus is whether the class challenge "is bound up with the merits so closely that our decision would constitute 'interference with agency process.'" Johnson, 922 F.2d at 353 (quoting Salfi, 422 U.S. at 765).

The second factor, whether litigation of the challenge involves compiling a complex factual record, is meant to reserve the class device for challenges that will likely require extensive record development at the Agency beyond the class representative's individual benefits claim. Without such factual development, many claimants could find it extraordinarly difficult to litigate such challenges as they would lack the ability to obtain the information necessary to substantiate the

32 class claims. Additionally, class certification centralizes litigation in a single appellate record, obviating the need for unnamed class members to collect evidence or request information from VA and for VA to adjudicate duplicative information requests.

The third factor requires considering whether the record is sufficiently complete for adjudication. This reflects the fact that "the focal point for judicial review [of agency conduct] should be the administrative record already in existence, not some new record made initially in the reviewing court." Camp v. Pitts, 411 U.S. 138, 142 (1973). Further, the putative class representatives have control over this factor as ordinarily the completeness of the record is strongly influenced by claimants expressly raising arguments before the Board and entering relevant evidence into the record. As stated above, we do, just as in the petition context, have some limited factfinding ability when deciding motions for class certifications in the appeal context. See Monk III, 30 Vet.App. at 174. But factfinding is "typically unnecessary to judicial review of agency decisionmaking." Fla. Power & Light Co. v. Lorion, 470 U.S. 729, 744 (1985). This is doubly so for our court, which, as discussed, has unique limitations on its factfinding ability above and beyond those of a federal district court. See 38 U.S.C. §§ 7261(c), 7252(b). But see Monk III, 30 Vet.App. at 171; Bove, 25 Vet.App. at 143; Erspamer, 1 Vet.App. at 10. Thus, the extent to which a proposed class will require additional factfinding is an important consideration in determining whether the presumption against aggregate action is rebutted.

The final factor deals with enforcement. When this Court issues a favorable precedential decision, it certainly binds VA in all pending and future claims. See 38 U.S.C. § 502. But claimants not party to that decision who may be subject to errors affecting their rights, whether due to VA's non-compliance with our decision at a later date or otherwise, do not have any right to prompt remedial enforcement. Their only recourse is bringing the allegedly invalid agency action before us by fully exhausting agency review before filing a notice of appeal. And in some cases, this will be an ordinary feature of litigation. But where the facts suggest a need for prompt remedial enforcement, claimants may instead seek class certification. This is a fact-specific analysis that will vary based on the unique facts of each individual appeal. So, for example, one need not find that the Agency is likely to disobey—we find such willful noncompliance unlikely in all but the most extreme case. Instead, a special need for remedial enforcement might be the result of the class members' age or some similar factor suggesting the need for especially timely relief.

Applying these factors here, class certification is the superior method for litigating the remaining class claim. The class claim is collateral to Mr. Skaar's claim for benefits because it challenges VA's adherence to a generally applicable regulation and is not "bound up with the merits [of Mr. Skaar's claim for disability benefits] so closely that our decision would constitute 'interference with agency process,'" Johnson, 922 F.2d at 353 (quoting Salfi, 422 U.S. at 765), as a favorable decision on the merits would not be an "order that class members be paid benefits" nor would it "in any way interfere with the agency's role as the ultimate determiner of eligibility" for benefits. City of New York, 476 U.S. at 485. In fact, a merits decision in the class's favor would do "no more than the agency would have been called upon to do had it, instead of [us], been alerted to the" alleged deficiencies in the Air Force's dose estimate methodologies. Id. Thus, this factor weighs in favor of certification.

33 So, too, does the second. The record in this case is complex and voluminous, containing numerous documents related to technical and scientific matters, e.g., R. at 2635-50, 2682-3501, and decades old records, e.g., R. at 3558-4148. Centralizing the class challenge in one litigation strikes us as a far better use of our limited judicial resources and avoids the specter of both unnamed class members and VA engaging in duplicative record development.10

The third factor also weighs in favor of certification. Mr. Skaar and the proposed class have submitted scientific evidence challenging the validity of the Air Force's dose estimates. See R. at 2635-50. We are also equipped with the Board's supplemental statement addressing Mr. Skaar's challenge to VA's adherence to § 3.311. See generally Secretary's Mar. 29, 2019, Resp. We require no additional information to decide the class challenge on the merits. Importantly, if the class sought not only to challenge VA's compliance with § 3.311 but also proffered an alternative dose methodology, we would likely require significant amounts of additional information such that class certification could prove impractical. However, here, the record is complete.

Finally, the class has alleged sufficient facts suggesting a need for timely remedial enforcement, and thus the final factor also weighs in favor of certification. The Palomares nuclear cleanup occurred on January 17, 1966, nearly 54 years ago. The advanced age of the class members, especially considering they all must suffer from a radiogenic disability to qualify, suggests a need for the availability of prompt remedial enforcement. VA already considers claimants' ages when determining whether to expedite appeals. See 38 U.S.C. § 7107. Thus, we think it an apt consideration in the class certification context as well. Additionally, the requested relief is identical across the class—a Court order to VA that it comply with § 3.311. It is more efficient and prudent to administer the requested class relief here collectively through an orderly and consistent process amenable to judicial supervision, rather than through piecemeal litigation.

All four factors weigh in favor of certification. Thus, we hold class certification is a superior method of litigating the remaining class claim.

7. Proposed counsel is adequate.

Having now concluded a class action is appropriate in this appeal as to the § 3.311 claim, we turn to the appointment of class counsel who is adequate to protect the interests of absent class members. Although Rule 23(a)(4) historically included an analysis of the adequacy of class counsel, that inquiry is now codified in 23(g). See Sheinberg v. Sorensen, 606 F.3d 130, 132-35 (3d Cir. 2010). Despite the rule change, the analysis is largely the same. See Kalish v. Karp & Kalamotousakis, LLP, 246 F.R.D. 461, 463 (S.D.N.Y. 2007). The Rule provides a set of factors courts must consider when judging class counsel's adequacy: (i) the work already done investigating and developing the claims; (ii) counsel's class action and substantive legal experience; (iii) counsel's relevant legal knowledge; and (iv) counsel's willingness to litigate the claim. FED. R. CIV. P. 23(g)(1)(A)(i)-(iv). Courts are not limited to these factors and "may consider any other matter pertinent to counsel's ability to fairly and adequately represent the interests of the

10 As an example of the type of duplicative recordmaking we hope to discourage, Mr. Skaar indicated that several other putative class members with claims at the Board would "shortly submit in their own cases the same records" he has already submitted to the Court. Appellant's June 20, 2018, Resp. at 14, n.4. Such duplicative recordmaking cannot be in the interest of systemic efficiency.

34 class." FED. R. CIV. P. 23(g)(1)(B). We adopt these Rule 23(g) factors as guides for our assessment of the adequacy of class counsel.

Proposed class counsel in this action is Michael Wishnie, Esq., of the Veterans Legal Services Clinic of Yale Law School's Jerome N. Franks Legal Services Organization. He is adequate. Counsel has done extensive work developing the claims at issue in this matter, demonstrated both "relevant legal knowledge" of and experience in both class action litigation and veterans law through prior aggregate actions before us, see, e.g., Monk III, 30 Vet.App. at 174, and shown a willingness to commit the necessary resources to lead this action through counsel's extensive work on this matter. Thus, and because there are no "other matter[s] pertinent to counsel's ability to fairly and adequately represent the interests of the class," counsel is adequate and will be appointed to represent the class. See FED. R. CIV. P. 23(g)(1)(B).

8. Generalized notice of class certification is required but opt out rights are not.

We have two final matters to consider, although they are related. We must first determine whether to afford class members the opportunity to opt out of the class we have certified. Next, we must determine what type of notice, if any, to provide to the class about this certification. The issues are related because if opt out rights are available, ensuring actual notice of the pendency of the class action takes on greater importance.

Classes certified under Rule 23(b)(2) generally do not require opt-out rights for absent class members. See Stoetzner v. U.S. Steel Corp., 897 F.2d 115, 119 (3d Cir. 1990). This is so because the indivisible nature of injunctive relief means it applies to every member of the class no matter what. See In re Allstate Ins. Co., 400 F.3d 505, 506 (7th Cir. 2005) (commenting that "[t]he thinking behind this distinction [concerning opt out rights] is that declaratory and injunctive relief will usually have the same effect on all members of the class as individual suits would"). This same indivisible nature of the injunctive relief requested here combined with this Court's national jurisdiction counsel against allowing opt-out opportunities for members of the class we have certified. See 38 U.S.C. § 7269.

Federal Rule 23(c) states "[f]or any class certified under Rule 23(b)(1) or (2), the court may direct appropriate notice to the class" while for those certified under (b)(3) "the court must direct to class members the best notice practicable under the circumstances." (emphasis added). Because we have determined the class members do not have the right to opt out of the class we have certified, notice at this stage of the proceedings is less critical than if class members could remove themselves from the class. Nonetheless, we believe it is the best practice to take reasonable steps to inform class members of the pendency of this action. Such notice need not be individualized for each member of the class but, rather, may be a generalized notice. As directed at the conclusion of this order, the parties are to jointly submit a proposed class notice and plan for effecting notice, both of which we must approve. If the parties are unable to agree, they should submit separate sections and include them in the joint submission.

35 III. CONCLUSION

We are, as we have observed before, "in uncharted waters." Monk v. Shulkin, No. 15-1280, 2018 WL 507445, at *2 (Jan. 23, 2018). We recently recognized our authority to aggregate actions in the petition context, see Monk II, 30 Vet.App. at 170-71, and we will now do so in the appeal context as well. Our decision today heralds the beginning of an era in which we will entertain, but by no means always certify, class actions in the first instance, making us the only Federal appellate court in the Nation to do so. Grappling with the complexities of the law of aggregate action while also maintaining fidelity to the VJRA and congressional intent to benefit those who have served the Nation has been—and no doubt will continue to be—a challenge we must face. But if class action procedures can lead to more consistent, efficient, and effective adjudication, then our Nation's veterans deserve no less.

Upon consideration of the foregoing, it is

ORDERED that the motion for class certification is GRANTED IN PART and DENIED IN PART. It is further

ORDERED that the proposed class definition is modified as explained herein and the following class is certified in this matter: All U.S. veterans who were present at the 1966 cleanup of plutonium dust at Palomares, Spain, and whose application for service-connected disability compensation based on exposure to ionizing radiation VA has denied or will deny by relying, at least in part, on the findings of dose estimates requested under 38 C.F.R. § 3.311, except those whose claims have been denied and relevant appeal windows of those denials have expired, or those whose claims have been denied solely based on dose estimates obtained before 2001. It is further

ORDERED that Michael J. Wishnie, Esq., is appointed as class counsel. It is further

ORDERED that, within 30 days, the parties jointly submit a proposed class notice and plan for effecting notice. If the parties are unable to agree, they are to submit separate sections and include them in the joint submission. It is further

ORDERED that this matter is returned to the original panel appointed to this appeal for management of the class action and a decision on the merits.

DATED: December 6, 2019

SCHOELEN, Senior Judge, concurring in part and dissenting in part:

I agree with my colleagues in the majority generally as to the usefulness of the class action mechanism in the context of appeals before this Court. I particularly agree that class certification could be a useful device for dealing with broad, ancillary issues such as the potentially flawed dose estimate methodology challenged in the case before us. That issue exists outside the boundaries of traditional veterans law litigation, and having a system in place to address a discrete legal issue divorced from class members' underlying benefits claims will increase judicial efficiency and agency adjudication rates. Nonetheless, I respectfully disagree with the majority's ill-explained

36 finding that our jurisdictional statute permits us to include Future-Future Claimants as class members. I also disagree with their unwillingness to include Past Claimants and Expired Claimants in the class. In my view, the majority's interpretation and application of Bowen v. City of New York, 476 U.S. 467 (1986), is flawed, and their flawed view systematically precludes vulnerable veterans from receiving full and fair hearings. Additionally, I am very concerned about reconciling our role as an appellate court that can issue precedential decisions with the necessity and superiority of class actions. To that end, I propose additional factors for the balancing test analyzing whether class actions are superior to precedential decisions.

I. THE FUTURE-FUTURE CLAIMANTS SHOULD BE EXCLUDED FROM THE CLASS

The majority states that City of New York "bears a striking similarity to the matter before us." Majority at 19. I strongly agree, and find our jurisdictional statute, 38 U.S.C. § 7252, to be properly analogous to the Social Security jurisdictional statute, 42 U.S.C. § 405(g), at issue in City of New York, which is why I find the majority's inclusion of the Future-Future Claimants in the class troubling.

At the outset, I agree with my dissenting colleagues insofar as they find that section 7252 includes the nonwaivable, jurisdictional requirement that a veteran's claim be presented preliminarily to VA, just as the Supreme Court in Mathews v. Eldridge held that presentment was a nonwaivable, jurisdictional requirement for Social Security claimants to obtain judicial review under section 405(g). Dissent at 46-48; 424 U.S. 319, 328 (1976) ("The waivable element is the requirement that the administrative remedies prescribed by the Secretary be exhausted. The nonwaivable element is the requirement that a claim for benefits shall have been presented to the Secretary."). This is so because, intuitively, there can be no decision under either statute absent a claim.

The majority glosses over this requirement and instead summarily concludes that we have jurisdiction over the Future-Future Claimants. It is unclear to me whether the majority finds that we have jurisdiction over nonpresenting Palomares veterans because we have jurisdiction over Mr. Skaar or because they should be treated in like manner to the Present-Future Claimants under the administrative exhaustion analysis. If it is the former, the Social Security cases we rely upon throughout this opinion counsel that the jurisdictional requirement that someone file a claim is an individual requirement that cannot be waived; if it is the latter, the majority improperly conflates the concepts of presentment and exhaustion. Nothing in our caselaw or the analogous Social Security cases leads me to believe that either of these theories is a faithful interpretation of our jurisdictional statute. To the contrary, section 7252 is, on its face, sufficiently comparable to section 405(g) and this Court should find that presentment is a jurisdictional requirement. Simply put, it cannot possibly be true that our jurisdictional statute is waivable in its entirety for potential class members who have never filed a claim.

Further, I find no Social Security caselaw that allows a District Court to assert jurisdiction over nonpresenting individuals pursuant to section 405(g). In fact, when nonpresenting individuals have been consolidated with other Social Security class members, courts have invoked creative mechanisms such as mandamus jurisdiction under 28 U.S.C. § 1361. See Clark v. Astrue,

37 274 F.R.D. 462, 467 (S.D.N.Y. 2011) ("[I]ndividuals failing to present their claims can still be part of the class because the Court may exercise mandamus jurisdiction over their claims pursuant to 28 U.S.C. § 1361."); see also City of New York v. Heckler, 742 F.2d 729, 739 & n.7 (2d Cir. 1984); Ellis v. Blum, 643 F.2d 68, 77-82 & n.10 (2d Cir. 1981). Our closest analogue is the All Writs Act, which does not provide an independent source of jurisdiction, but rather allows us to protect our future jurisdiction. See Syngenta Crop Prot., Inc. v. Henson, 537 U.S. 28, 33 (2002) (affirming that the All Writs Act does not confer jurisdiction on the federal courts); see also Clinton v. Goldsmith, 526 U.S. 529, 534-35 (1999) (noting that the express terms of the All Writs Act confine a court "to issuing process 'in aid of' its existing statutory jurisdiction; the Act does not enlarge that jurisdiction"). Because other federal courts have found the need to invoke an independent source of jurisdiction for nonpresenting class members, and because we have no other statutory grant of jurisdiction outside section 7252, it follows that our Future-Future Claimants cannot be consolidated as part of the class.

Despite the fact that I believe the Future-Future Claimants should not be part of the class, it is worth noting that this group of veterans is unlikely to be harmed by exclusion. In some ways, the exclusion of the Future-Future Claimants presents a legal fiction unique to this Article I appellate court – the precedential effect of our decision will bind them regardless of their nonpresenting status, and as soon as they file, they will be subject to whatever rule VA has been judicially mandated to follow. Although the Future-Future Claimants are necessarily implicated in this litigation, our authority to issue precedential decisions means they will not suffer any injustice during these proceedings, and our jurisdictional statute should not be skirted to establish a false equivalent with the Present-Future Claimants.

II. THE PAST AND EXPIRED CLAIMANTS SHOULD BE INCLUDED IN THE CERTIFIED CLASS

I also take exception with the majority's exclusion of the Past and Expired Claimants from the class. City of New York addressed the same legal issues we now face in deciding class composition – exhaustion of administrative remedies and equitable tolling – but, here, the majority has only adopted the Supreme Court's holding insofar as it pertains to the exhaustion of remedies issue. I do not believe the majority's application of that case is uniform or consistent.

In City of New York, the Supreme Court, in affirming the rulings of both the District Court and the Court of Appeals, notes that the District Court included claimants in the class who had not exhausted their administrative remedies. City of New York, 476 U.S. at 475-76 (citing Eldridge, 424 U.S. at 319). The Supreme Court then recounts the District Court's analysis as to why the class properly included those who had not complied with the 60-day statute of limitations:

The [District] [C]ourt noted that the 60-day requirement is not jurisdictional . . . [and] found that "the same reasons which justify implying waiver of the exhaustion requirement are stronger for the sixty[-]day requirement because the statute of limitations is not, as is the exhaustion requirement, 'central to the requisite grant of subject-matter jurisdiction.'"

Id. at 476 (emphasis added) (citations omitted).

38 Effectively, the majority properly applies City of New York's analysis as to the jurisdictional question (at least insofar as it pertains to the Present-Future Claimants), but chooses to impose a higher burden on the claimants in the nonjurisdictional portion of the case. This should not be so. Here, as in City of New York, the same rationales for waiver of the administrative exhaustion requirement are applicable to, and indeed stronger for, the equitable tolling issue. Succinctly stated, this Court should not waive the jurisdictional requirements for one class of veterans and then exclude other classes of veterans who present no jurisdictional impediments.11

Moreover, it is unclear to me whether the majority purports to adopt City of New York's equitable tolling framework and chooses to find that the nonsecretive nature of VA's dose estimate methodology distinguishes the matter, or whether they do not believe that framework applies at all to the Past and Expired Claimants simply because the specter of equitable tolling "offends the very notion of finality." Majority at 23. Regardless, I respectfully find their interpretation far too narrow.

A. Proper Application of Equitable Tolling Framework

This Court should endorse a wholesale import of City of New York's framework. That means that, when analyzing whether equitable tolling is warranted for Past and Expired Claimants in a class context, two questions are presented: (1) "[W]hether equitable tolling is consistent with Congress' intent," and (2) "whether tolling is appropriate on these facts." City of New York, 476 U.S. at 480.

The first question should be answered now and applied to all future class certification analyses: Yes, equitable tolling in the context of the Expired Claimants and Past Claimants is consistent with congressional intent. Just like 42 U.S.C. § 405(g) at issue in City of New York, Congress designed the applicable veterans benefits statutes to be "unusually protective" of claimants. Id.; see Henderson ex rel. Henderson v. Shinseki, 562 U.S. 428, 437 (2011) ("The Social Security disability benefits program, like the veterans benefits program, is 'unusually protective' of claimants.") (quoting Heckler v. Day, 467 U.S. 104, 106-07 (1984)). As the U.S. Court of Appeals for the Federal Circuit has stated, "Congress' intent in crafting the veterans benefits system is to award 'entitlements to a special class of citizens, those who risked harm to serve and defend their country. This entire scheme is imbued with special beneficence from a grateful sovereign.'" Barrett v. Nicholson, 466 F.3d 1038, 1044 (Fed. Cir. 2006) (quoting Bailey v. West, 160 F.3d 1360, 1370 (Fed. Cir. 1998) (en banc) (Michel, J., concurring)); see also Jaquay v. Principi, 304 F.3d 1276, 1286 (Fed. Cir. 2002) (en banc); Hensley v. West, 212 F.3d 1255, 1262 (Fed. Cir. 2000). That "special beneficence" is noted time and again in caselaw, and "in the context of veterans' benefits where the system of awarding compensation is so uniquely pro-claimant, the importance of systemic fairness and the appearance of fairness carries great weight." Hodge v. West, 155 F.3d 1356, 1363 (Fed. Cir. 1998).

11 I note that, although I agree with the dissent's point regarding the nonwaivability of section 7252's presentment requirement, I diverge from their thinking as to exhaustion. I agree with the majority's finding that our jurisdictional statute is sufficiently analogous to section 405(g) to warrant the same exhaustion analysis conducted in City of New York.

39 Keeping in mind this rationale as to why equitable tolling is appropriate in veterans law cases generally, we must assess whether tolling is appropriate on the facts of this case. That must be done by comparing this case to City of New York and determining whether the conduct at issue here warrants our tolling of the filing deadline.

The majority tersely states that they will not equate VA's adjudication of Palomares veterans' claims with the secretive conduct at issue in City of New York, then asserts that "there is no principled way to distinguish the Expired Claimants here and any other claimants who have been denied benefits, failed to appeal to this Court, and later discovered that their benefits denial was based on an incorrect reading of the law." Majority at 23. In context, this means that the majority has (1) implicitly held that "secretive conduct" must be at issue to trigger equitable tolling, and (2) placed this case on equal footing with conventional challenges to denials of veterans' disability compensation claims.

Other courts have not applied City of New York so strictly. For instance, the U.S. Court of Appeals for the Eighth Circuit analyzed whether secretive conduct is "an absolute prerequisite" for equitable tolling to be appropriate and held that "although a secret, internal policy is probably not a prerequisite to equitable tolling, some type of misconduct on the part of the agency or gross, but good-faith, error on the part of the claimant should justify this extraordinary remedy." Medellin v. Shalala, 23 F.3d 199, 204 (8th Cir. 1994), rehearing denied (June 2, 1994). Similarly, the Southern District of Ohio has previously held that equitable tolling was appropriate for a class of plaintiffs challenging the former practice of the Secretary of Health and Human Services in calculating the amount of supplemental security income (SSI) benefits. Though the policy at issue was not secret or clandestine, the District Court found equitable tolling was warranted because the calculation of SSI benefits was not made pursuant to an established regulation and claimants "might well be unaware of the specific factors taken into account by the Secretary." Gould v. Sullivan, 131 F.R.D. 108, 112 (S.D. Ohio 1989). Additionally, when certifying a class of claimants, the Southern District of New York in Hill v. Sullivan stated that it did "not believe it necessary to determine whether . . . behavior amounts to a 'clandestine policy' to 'prevent[ ] plaintiffs from knowing of a violation of [their] rights.'" 125 F.R.D. 86, 95 (S.D.N.Y. 1989) (citations omitted). Rather, the court agreed with the plaintiffs that the Secretary's failure to publish challenged rulings "had the same practical effect on claimants as the defendant's secretive conduct in [City of New York]." Id. (citations omitted).12

I do not attempt here to explicitly import another court's test or draw a bright line that can be applied in future cases. Rather, when taken together, these cases demonstrate that equitable tolling can be appropriate in instances where the conduct complained of falls short of "secretive," and I believe that, on the facts of this specific case, tolling is warranted. See Toomer v. McDonald, 783 F.3d 1229, 1239 (Fed. Cir. 2015) (citing Holland v. Florida, 560 U.S. 631, 649 (2010)) (stating that equitable tolling is a matter assessed by the Court on a case-by-case basis with an

12 Additionally, although not arising in the equitable tolling context, the District Court in Nehmer v. U.S. Veterans' Admin. did not require secretive conduct by VA to include the "Expired Claimants" – i.e., the pre-1985 claimants – in the class. 118 F.R.D. 113 (N.D. Cal. 1987). Nevertheless, they were allowed to participate in the class because they shared a threat of "future harm" with other class members. Id. at 117. This harkens to the analysis by the majority that surely Congress did not expect veterans to have fewer rights after the Veterans' Judicial Review Act than they did before its enactment.

40 acknowledgment of the "need for flexibility" and "for avoiding mechanical rules"). The U.S. Air Force originally worked with consultants who developed a methodology for deriving dose estimates for Palomares veterans, which was detailed in the LA Report; the inputs for this methodology included vast amounts of scientific data not easily understood by laypersons, including dosimetry readings, bioassay data, environmental testing, and multiple complex computer models; over 12 years after the LA Report was published, the Air Force – not VA – determined that inconsistencies existed in dose estimates; thereafter, the Air Force began using a revised methodology when providing VA with dose estimates for Palomares veterans; and the revised methodology also contained highly complex measurements and datasets (which may or may not be flawed). There is no doubt in my mind that this development-and-assignment exercise, conducted outside VA's purview and essentially devoid of oversight, prevented veterans from continuing administrative appeals and pursuing benefits they may have been entitled to, and thus is sufficient under City of New York's framework that the equities in this case favor tolling.

B. The Majority's Other Contentions

Further, the majority should not equate a flawed dose estimate methodology with a misapplication of law. City of New York itself states that claimants who were subject to the systemwide, unrevealed policy "stand on a different footing from one arguing merely that an agency incorrectly applied its regulation."13 476 U.S. at 485. The dose estimates produced by that methodology function as scientific facts ancillary to administrative proceedings, not as a legal interpretation subject to future revision. And the development of this methodology behind a veil at the Department of Defense (DoD) "prevented [the claimants] from realizing that they had valid grounds for seeking administrative review." McDonald v. Sec'y of Health & Human Servs., 834 F.2d 1085, 1090 (1st Cir. 1987). The flawed dose estimates did not function like a new legal interpretation that was disadvantageous to veterans, but rather provided a flawed factual basis that prevented claimants from even accessing the veterans benefits system.

Additionally, the majority says there is no principled way to distinguish the Past and Expired Claimants from any other claimants who have been denied benefits, failed to appeal to this Court, and later discover their benefits denial was based on an incorrect reading of the law. Majority at 23. But I would assert that the same rationales for inclusion of the Present-Future Claimants apply with equal – if not greater – force to the Past and Expired Claimants. The majority views it a "substantive advantage" that veterans' claims will be relitigated maintaining their effective dates, but to frame this advantage as more substantive than the inclusion of those claimants over whom we do not typically have jurisdiction is incorrect. Equitable tolling is a procedural tool the Court can use just like waiver of administrative exhaustion. The fact that veterans can file supplemental claims under 38 U.S.C. § 5108(a) and 38 C.F.R. § 20.1105(a) is of no consequence. Moreover, they may very well lose their original effective date, and thus it is not a similar remedy. Veterans who are effectively barred from an entire administrative system via a

13 The Supreme Court made this statement when discussing claimants who had not exhausted their administrative remedies as opposed to those who argued equitable tolling was warranted. Nevertheless, the phrase is easily extended to the claimants seeking equitable tolling, as its purpose is merely to distinguish the policy challenge from an illegal application of a regulation. In other words, regardless of which group within the proposed class we are discussing, a claimant's challenge to the underlying obscured policy differs from a claimant's challenge to a regulation.

41 factual error developed by an agency we have no direct authority over would not be "substantively advantaged" in any way by including them in the class; instead, they would only be given what they were improperly denied initially under the law.

Further, for the sake of argument, even if I agreed with the majority's premise that utilizing the class device here renders substantive benefits for the Past and Expired Claimants, it is unclear to me why that precludes this Court from including them in the class. City of New York clearly endorsed certification of just such a group of Social Security claimants. Those claimants arguably were privy to the same types of "substantive benefits" that our Past and Expired Claimants would be, but were still included in the class. I believe it error to first invoke a categorical rule that class certification should never be used for a substantive advantage, then label inclusion in the class a substantive advantage, all while overlooking that City of New York did the very thing the majority prohibits.

At the end of the day, Article III caselaw is not controlling, but this Court has chosen of its own volition to import the narrowest interpretation possible of City of New York to justify certifying an unjustly narrow class. 14 Our failure to equitably toll in this case does not show reverence for existing interpretations of law or respect for the administrative process, but rather provides tacit endorsement of DoD-developed policies and facts to be used later by VA, no matter the consequences within VA's regulatory scheme.15 It is a statement that a group of vulnerable veterans should not have full and fair hearings because they were not legally savvy enough to challenge a complicated and convoluted dose reconstruction methodology developed by consultants at an agency wholly separate from VA. As the U.S. Court of Appeals for the Second Circuit stated in City of New York v. Heckler, "[a]ll of the class members who permitted their administrative or judicial remedies to expire were entitled to believe that their Government's determination of ineligibility was the considered judgment of an agency faithfully executing the laws of the United States." 742 F.2d at 738. The Past and Expired Claimants should be allowed their (legitimate) day in court, just like the Present-Future Claimants over whom we would not traditionally have jurisdiction.

III. SUPERIORITY TEST

Another significant issue involves the determination of when we will grant class certification versus when we will issue a precedential decision – a question unique to this appellate court engaging in an activity typically committed to District Courts. Because we possess the authority to issue precedential decisions that bind all future VA decisions, class actions would likely be more appropriate in rare and unique circumstances. When assessing whether the class action device is superior to a precedential decision, I agree with the majority that a balancing test is appropriate; however, it must be a sufficiently robust test. To that effort, I would add two factors

14 See Henderson, 562 U.S. at 437-38 ("[N]one of the precedents cited by the parties controls our decision here. All of those cases involved review by Article III courts. This case, by contrast, involves review by an Article I tribunal as part of a unique administrative scheme."). 15 That is not to say that I necessarily agree with Mr. Skaar as to the merits underlying this case. But I believe the majority to be saying that no matter how far removed from the veterans benefits process or the agency which oversees it, and no matter how scientifically dense or ill-conceived the policy, we lack the power as an institution to equitably toll veterans' cases if the alleged misconduct is not clandestine.

42 to their analysis. The first additional factor addresses whether litigation of the challenge involves complex technical or scientific matters. The second addresses whether the alleged conduct is "systemic" – that is, whether a significant number of VA claims involve this issue.

A. Technical or Scientific Complexity

This first additional factor is meant to reserve the class device for challenges that will likely require sophisticated knowledge beyond the normal level of savvy needed by claimants or their attorneys to litigate veterans' individual benefits claims. Many claimants could find it extraordinarily difficult to litigate challenges involving technical data or complex scientific concepts, as they would lack the ability to obtain or understand the information necessary to substantiate their claims. Class certification centralizes litigation, obviating the need for unnamed class members to independently construct theories based on data not readily available or understandable.

This factor is related to, but separate from, the majority's second prong, which contemplates whether "litigation of the challenge involves compiling a complex factual record." One of these considers whether the underlying concepts that will be contemplated in merits litigation are complicated to a litigant and one considers whether development before the agency is extensive and onerous (essentially making it complicated for the Court). Future cases can and should contemplate both factors when asking whether class certification is superior.

Here, the additional factor is clearly met. Understanding how DoD constructed dose estimates for Palomares veterans, and understanding whether or how those dose estimates were miscalculated, is a highly complex exercise that requires skills far beyond those of individual litigants. This lends extra weight to the majority's findings as to superiority.

B. Systemic Complaint

The second factor I propose adding – whether the issue in the appeal is a systemic complaint – is a distinct inquiry from the numerosity prong of the class certification test set out under Rule 23(b), where the concerns are more related to whether the class is so numerous as to make individual adjudication of claims at the Court impractical. The systemic-complaint factor looks at the question from VA's perspective – are there so many claims at VA involving this issue that this decision will have a significant effect on the agency and will the agency likely benefit from a single-stroke class action decision rather than one-by-one appeals?

I would find that this factor weighs against a class action and favors a precedential decision. Although 1,600 veterans is a significant number of parties affected (and sufficient to satisfy Rule 23's numerosity requirement), it is not sufficient to be deemed a systemic complaint when VA handles over a million claims per year.16 Nevertheless, as the superiority test is a balancing test, failing one factor does not foreclose class certification. When taken as a whole, I concur with the majority that class certification is superior in this case to a precedential decision.

16 See VA, CONGRESSIONAL SUBMISSION, FY 2020, VOL. III: BENEFITS AND BURIAL PROGRAMS AND DEPARTMENTAL ADMINISTRATION 146 (2019), https://www.va.gov/budget/docs/summary/fy2020VAbudgetvolume IIIBenefitsBurialProgramsAndDeptmentalAdministration.pdf.

43 FALVEY, Judge, with whom PIETSCH and MEREDITH, Judges, join, dissenting:

The majority boasts that "we are the only appellate court in the Nation with the authority to aggregate actions in the first instance." Ante at 25. There are sound reasons why no other appellate court has undertaken this innovation. Given the limited nature of our jurisdiction and scope of review, we question the efficacy of the majority's action, and, considering our ability to issue precedential decisions that direct VA practices nationwide, we also question its necessity. We believe that the majority has created a class that exceeds our jurisdiction and offers a comparable outcome to members of that class that a precedential decision could provide without the manageability and preclusion problems inherent in class litigation. Because we disagree with the substance of the majority's order, the rationale underlying it, and the way the majority has developed this case, we respectfully dissent.

I. ANALYSIS

Although there is much in the majority's order with which we disagree, we will focus here on those matters related to our jurisdiction to conduct class actions in the appellate context and the utility of doing so even if we have such jurisdiction, and how that applies to Mr. Skaar's proposed class.

A. The Power to Certify Class Actions in the Appeal Context

1. Our authority to certify a class is derived from our procedural statutes.

Under our jurisdictional statute—38 U.S.C. § 7252—the Court's review is limited to Board decisions and the record of proceedings before the Secretary and the Board. We agree that, if a proposed class satisfies the jurisdictional requirements of section 7252, then the Court has the authority to certify that class. Under such circumstances, if the Court chooses to exercise that authority, it may certainly utilize procedural statutes, such as 38 U.S.C. § 7264(a), to aggregate a class. In our view, our jurisdictional statute restricts classes that we may certify in the appeal context under our procedural statutes to those containing only class members who have obtained a final Board decision. And, our review of those members' cases is limited to the record of proceedings.

The majority goes much further. It finds the authority to conduct class actions in an esoteric "inherent authority." Citing the United States Court of Appeals for the Federal Circuit (Federal Circuit) in Monk v. Shulkin (Monk II), 855 F.3d 1312 (Fed. Cir. 2017), the majority contends that our "inherent authority" supports our use of class actions. Ante at 13-14. The majority fails to explain the source and scope of the term "inherent authority." More importantly, it does not explain how "inherent authority" expands our jurisdiction beyond that provided by our jurisdictional statute, aside from a conclusory statement that it does. It is equally unclear why vague references to "inherent authority" are necessary to justify class actions where section 7264(a)—which provides that proceedings of the Court "shall be conducted in accordance with such rules of practice and procedure as the Court prescribes"—allows for such aggregation, so long as the jurisdictional requirements under section 7252 are first met.

44 The Federal Circuit in Monk II cited the All Writs Act (AWA) as the basis for this Court's authority to certify class actions in the petition context. The majority itself questions whether the AWA grants us authority to certify classes in the appeal context. It does not. Neither the AWA nor Monk II can stand as the legal basis for aggregating appeals because, unlike the wide authority the AWA gives us to protect our prospective jurisdiction, our authority to review appeals has been tightly circumscribed by Congress.

Thus, we would find that, although the Court has authority to certify a class in the appeal context when jurisdictional requirements are satisfied, such authority is derived from the procedural discretion granted to us by Congress, not the AWA or any purported "inherent authority."

2. But our procedural statutes do not create jurisdiction.

The majority, relying on Monk II, conflates the procedural statutes, which provide us with the methods to manage cases over which we have jurisdiction, with the statute authorizing our jurisdiction. The Federal Circuit in Monk II noted that Harrison v. Derwinski, 1 Vet.App. 438 (1991) (en banc) (per curiam order), in which the Court found that it lacked power to adopt a class action rule because, inter alia, section 7252 limited our review to Board decisions, reflected a concern that we would "exceed [our] jurisdiction" if we certified a class that included veterans without a Board decision. Monk II, 855 F.3d at 1320. The Federal Circuit then stated that it disagreed that our "authority is so limited," indicating that Congress expressly gave us "the authority to 'compel action of the Secretary unlawfully withheld or unreasonably delayed.'" Id. (quoting 38 U.S.C. § 7261(a)(2)).

The authority to compel action of the Secretary, coupled with our power under the AWA, allows us to aggregate cases in the petition context. It does not help us determine how to handle direct appeals. Anything the Federal Circuit said about direct appeals is dicta. That tribunal has yet to discuss our authority to conduct class actions on direct appeal when that issue was directly presented, properly briefed, and accompanied by an appropriate record. We, therefore, have no precedential guidance concerning that matter and do not rely on any unnecessary statements the Federal Circuit may have made.

After noting that the Federal Circuit disagreed with the Court's finding in Harrison, the majority references section 7264(a). A procedural statute, which authorizes us to create mechanisms necessary to exercise our jurisdiction (i.e., we may utilize such tools once we have jurisdiction), cannot be used to overcome the jurisdictional barrier that the Court identified in Harrison. See Henderson v. Shinseki, 562 U.S. 428, 434 (2013); In re Wick, 40 F.3d 367, 373 (Fed. Cir. 1994) ("If Congress had intended the court's jurisdiction to be broader than that conferred by § 7252, Congress would have expressed that intention legislatively.").

45 3. Based on section 7252(a) and Supreme Court precedent, we are prohibited from waiving any administrative exhaustion requirements and assuming jurisdiction over class members who have not filed a claim and do not have a Board decision.

The majority acknowledges the Secretary's argument that the Court lacks jurisdiction to include veterans without a Board decision in the certified class because such a decision is a jurisdictional prerequisite for Court review. Ante at 18; see 38 U.S.C. § 7252(a). Of those without a Board decision, the majority indicates that such veterans fall into one of two subgroups within the proposed class: (1) Present-Future claimants—those who have filed claims that remain pending before VA; and (2) Future-Future claimants—those who have not yet filed claims. Ante at 15-16. The majority then states that it waives the exhaustion requirement—which, presumably, is that each class member have a Board decision—for these claimants and finds that the Court has jurisdiction over them. Id. at 20-21.

The Supreme Court's Social Security Administration (SSA) cases the Secretary and the majority reference to support their positions regarding jurisdiction are not directly on point as to our judicial review statutes. Although these cases provide helpful guidance as to how jurisdictional requirements should be analyzed, this precedent does not undermine the jurisdictional requirements of section 7252(a) or show that those requirements are waivable by the Court.

a. A statute may contain nonwaivable jurisdictional requirements and waivable administrative exhaustion requirements.

In Mathews v. Eldridge, the Supreme Court explained that its decision in Weinberger v. Salfi, 422 U.S. 749 (1975), identified three conditions17 that must be satisfied to obtain judicial review under 42 U.S.C. § 405(g). 424 U.S. at 328. Of these, the requirement that there be "a final decision of the Secretary made after a hearing" was central to the requisite grant of subject-matter jurisdiction. Id. (citing Salfi, 422 U.S. at 764). The Supreme Court stated that, implicit in Salfi, was the principle that

this condition consists of two elements, only one of which is purely "jurisdictional" in the sense that it cannot be "waived" by the Secretary in a particular case. The waivable element is the requirement that the administrative remedies prescribed by the Secretary be exhausted. The nonwaivable element is the requirement that a claim for benefits shall have been presented to the Secretary. Absent such a claim there can be no "decision" of any type. And some decision by the Secretary is clearly required by the statute.

Id. Recently, the Supreme Court in Smith v. Berryhill reiterated the Eldridge holding that section 405(g) contains both a nonwaivable jurisdictional requirement and a waivable requirement regarding the exhaustion of administrative requirements. 139 S. Ct. 1765, 1773 (2019).

17 The Supreme Court noted that two of these conditions—that civil action be commenced within 60 days after the mailing of notice of such decision and that the action be filed in an appropriate district court—specified a statute of limitations and appropriate venue, and are waivable by the parties. Mathews v. Eldridge, 424 U.S. 319, 328 n.9 (1976).

46 b. There is a difference between a requirement being waivable and determining whether to waive that requirement.

Although the majority notes the axiom that "[s]ubject-matter jurisdiction 'can never be waived or forfeited,'" ante at 16 (quoting Gonzalez v. Thaler, 565 U.S. 134, 141 (2012)), it then proceeds to do just that. It does so by applying a test for determining whether to waive a statutory requirement without first ascertaining whether the statutory requirements in question are in fact waivable.

Lest there be any residual doubt after 30 years of caselaw, section 7252(a) is jurisdictional. Indeed, it's hard to imagine that the English language could produce a more clearly jurisdictional provision. See Fort Bend Cty. v. Davis, 139 S. Ct. 1843, 1849 (2019) (courts should deem a requirement jurisdictional when Congress clearly states that it is). The statute is labeled "[j]urisdiction" and the phrase in question says that this Court "shall have exclusive jurisdiction to review decisions of the Board." The majority here investigates whether it may expand the Court's traditional view of its authority by reaching back into the agency's adjudicatory process and laying hold of claims that have not yet been subject to a Board decision. As we will explain, its actions contravene the intentions of Congress.

In Bowen v. City of New York, 476 U.S. 467 (1986), and Eldridge, the Supreme Court found that the waivable element of section 405(g) was the requirement that the administrative remedies prescribed by the Secretary be exhausted. The Supreme Court then utilized the test referenced by the majority (whether the challenged conduct is collateral to a claim for benefits; exhaustion would cause irreparable harm; and the purpose of exhaustion is not served by its enforcement) when assessing whether deference to the agency's determination of finality was necessary. City of New York, 476 U.S. at 483 (noting that, "[o]rdinarily, the Secretary has discretion to decide when to waive the exhaustion requirement," but that in certain cases deference to the agency's judgment is inappropriate), 484 ("The ultimate decision of whether to waive exhaustion . . . should also be guided by the policies underlying the exhaustion requirement."); Eldridge, 424 U.S. at 328, 330.

In other words, the Supreme Court first determined whether the statutory element was waivable and only then assessed whether those steps created by the Secretary to reach a final decision warranted any deference, a process that the majority did not follow. As discussed below, no portion of section 7252(a) is waivable.

c. Section 7252(a) contains the nonwaivable requirement that a class member must have filed a claim with VA.

Once again, under section 7252(a), our Court "shall have exclusive jurisdiction to review decisions of the Board of Veterans' Appeals," and, by way of comparison, under section 405(g), an individual may obtain review by a court of "any final decision of the Commissioner of Social Security made after a hearing." 18 Section 7252(a) includes the nonwaivable, jurisdictional

18 At the time of Eldridge, the title of the agency head was Secretary of Health, Education, and Welfare and thus this portion of section 405(g) read "any final decision of the Secretary made after a hearing." 424 U.S. at 327. Currently, the title is Commissioner. Aside from this title change, the language of section 405(g) has remained the same.

47 requirement identified in Eldridge—that a claim for benefits shall have been presented to the agency—given that, under both statutes, there could be no decision absent a claim. 424 U.S. at 328 (noting that a decision was "clearly required by the statute"); see Berryhill, 139 S. Ct. at 1773. Therefore, if a veteran has not filed a claim with VA, our Court would not have jurisdiction over that individual. Since this requirement is jurisdictional, we cannot waive it. Thus, as discussed further below, the notion that the majority's so-called "Future-Future" claimants can be part of a class over which we have jurisdiction does not make it past the starting line.

d. Section 7252(a) does not contain the waivable requirement that administrative remedies prescribed by the Secretary be exhausted.

Our jurisdictional statute contains nothing like the waivable element identified in Eldridge—that the administrative remedies prescribed by the Secretary be exhausted. Section 405(g) allows for judicial review of "any final decision" of the Secretary/Commissioner, whereas section 7252(a) requires a decision of the Board. Congress specifically identified the type of VA decision that a claimant must obtain before jurisdiction in this Court is established, while section 405(g) does not specify which component of SSA must have provided the decision.

The Supreme Court relied on the fact that section 405(g) did not identify a particular component of SSA from which a decision need be issued when determining that the exhaustion of administrative remedies could be waived. For context, the SSA administrative review process generally requires that, if an SSA claimant disagrees with the state agency's initial denial of benefits, the claimant may seek (1) reconsideration by the original state agency; (2) if reconsideration is adverse, a hearing by an administrative law judge (ALJ); and (3) if the ALJ's decision is adverse, review by the Appeals Council. See City of New York, 476 U.S. at 471-72. In Salfi, the Supreme Court found that, because the Secretary in that case did not raise an exhaustion of administrative remedies argument, the reconsideration determination was "final." 422 U.S. at 767; see id. at 766 (stating that the term "final decision" was left undefined by the Act and its meaning left to the Secretary to flesh out by regulation).

In Eldridge, the claimant, rather than request reconsideration of the state agency's determination, commenced judicial action challenging the constitutional validity of SSA's administrative procedures and the Supreme Court held that the denial of the claimant's request for continued benefits was a final decision for the purpose of section 405(g) jurisdiction over his constitutional claim. 424 U.S. at 324-25, 332 (noting that Salfi required only that there be a "final decision" with respect to the claim for entitlement to benefits and that denying Mr. Eldridge's substantive claim would not answer his constitutional challenge).

In contrast, section 7252(a) requires a Board decision, rather than any VA decision. The statute, therefore, precludes the Court from using the waivable element identified in Eldridge to find that an agency decision other than a Board decision meets the requirements for section 7252(a) jurisdiction.

Moreover, the Supreme Court in Salfi and Eldridge focused on the fact that the Secretary/Commissioner was responsible for establishing the steps in SSA's administrative process, given that the waivable element was the requirement that the administrative remedies

48 prescribed by the Secretary be exhausted. See City of New York, 476 U.S. at 471-72 (noting that reconsideration of the state agency determination and review by the Appeals Council were prescribed by regulations, not statutes). Those factors are fully inapposite here.

Although the Secretary may establish administrative procedures through regulations, our jurisdictional statute inherently includes the administrative step of appealing an adverse regional office (RO) decision to the Board because the statute itself requires a Board decision. See Am. Legion v. Nicholson, 21 Vet.App. 1, 4-5 (2007) (citing Senate Bill 11); see also 134 Cong. Rec. S9184 (daily ed. July 11, 1988) (Senator Cranston, in outlining the procedure for judicial review under the Veterans Judicial Review Act (VJRA), stated that such review "would be available only after a veteran's claim has been turned down by a VA regional office and, on appeal, by the Board"). Because the administrative steps the majority here is seeking to waive are prescribed by Congress in a statement of jurisdiction, rather than the Secretary (who must also comply with the statute), Eldridge cannot be used as a tool to make a requirement that is plainly jurisdictional and unwaivable into something that is not.

We note also that section 405(g) contains the language "after a hearing." But, in waiving the administrative remedies requirement, the Supreme Court in its SSA cases focused on the fact that this statute did not specify the type of decision required before judicial review, rather than whether the hearing component in the statute could be waived. In Salfi, however, the Supreme Court briefly addressed this requirement and it found that a hearing would be futile once the Secretary determined that the only issue to be resolved was a matter of constitutional law beyond his competence to decide and that the Secretary may award benefits without a hearing. Salfi, 422 U.S. at 767. Our jurisdictional statute does not require a hearing before judicial review. Moreover, although the SSA Secretary may make a benefits determination without a hearing, VA cannot make a benefits determination without issuing a decision. Further, according to the Supreme Court's guidance in Eldridge, 424 U.S. at 328, "some decision . . . is clearly required" by our statute and, as noted, specifies it must be a Board decision.

e. We conclude that section 7252(a) includes no waivable elements.

As the Court and Federal Circuit have assumed for 30 years, section 7252(a) contains the nonwaivable, jurisdictional elements that a veteran must have both filed a claim and received a Board decision. Under the Supreme Court's framework, the Court and the Secretary are unable to waive any requirement of our jurisdictional statute. The majority's focus on determining whether to waive the requirement of a Board decision is at best premature because it did not explain why it determined that our jurisdictional statute has waivable components.

Further, the test that the majority utilizes to determine waivability was used by the Supreme Court to assess whether deference should be given to the administrative steps prescribed by the Secretary to reach a final decision. Because administrative remedies inherent in section 7252(a) are prescribed by Congress rather than the Secretary, the test that the majority cites does not apply.19 Because the requirement of a Board decision under section 7252(a) cannot be waived, we

19 The majority's analogy of Mr. Skaar's case to magistrate judges exercising jurisdiction over proceedings in civil matters with the consent of parties, ante at 17-18, is unpersuasive. First, our analysis for finding that our

49 do not have jurisdiction over individuals who have yet to obtain one. We will now address the two subgroups contained within this category.

f. We do not have jurisdiction over the Future-Future claimants.

As we explained above, the Court cannot take jurisdiction over the majority's so-called Future-Future claimants—i.e., those veterans who have not yet filed a claim. In Salfi, Eldridge, City of New York, and Berryhill, the Supreme Court noted that the requirement that a claim for benefits shall have been presented to the agency was a nonwaivable, jurisdictional statutory element. Accordingly, in Salfi, the Supreme Court found that, as to the unnamed plaintiffs, "the complaint is deficient in that it contains no allegations that [those class members] have even filed an application with the Secretary, much less that he has rendered any decision . . . . The class thus cannot satisfy the requirements for jurisdiction under 42 U.S.C. § 405(g)." Salfi, 422 U.S. at 764; see Califano v. Yamasaki, 442 U.S. 682, 704 (1979) (stating that the certified classes were too broad, but indicating that, at least in this instance, the relief offered by the injunction would not be afforded to individuals until they filed a written waiver request to the Secretary—i.e., met the statutory jurisdictional prerequisites).

Our statute contains the nonwaivable, jurisdictional requirement that a claimant have filed a claim with VA. The majority's conclusion that we have jurisdiction over individuals who have not filed a claim cannot be correct.

g. We also do not have jurisdiction over the Present-Future claimants.

As to the Present-Future claimants—those veterans who have filed claims that remain pending before VA at any level—we would also find that the Court does not have jurisdiction over these individuals. As stated, no element of section 7252(a) is waivable, given that Congress prescribed the administrative remedy necessary to obtain judicial review in our Court and specified that a veteran must have a Board decision before we assume jurisdiction. Therefore, we disagree that the majority has the authority to waive this requirement.

The Present-Future claimant subgroup can be further subdivided: (1) veterans who have filed a claim that remains pending before the RO (i.e., veterans who do not have a VA decision at all); and (2) veterans who have a claim pending before the Board (i.e., veterans who have appealed an RO decision, but have not obtained a Board decision).

The first group is in the same boat as the Future-Future claimants. See Eldridge, 424 U.S. at 328 ("[S]ome decision by the Secretary is clearly required by the statute."). Regarding the

jurisdictional statute contains no waivable requirements is based on Supreme Court precedent regarding SSA benefits, where at least two of those cases pertained to class actions. See generally City of New York, 476 U.S. at 467; Salfi, 422 U.S. at 749. Those Supreme Court cases discussing disability benefits are more analogous to our VA disability benefits cases and provide more guidance than do circuit court cases pertaining to magistrate judges. Second, as we will discuss, there are significant distinctions between trial courts—i.e., where magistrate judges practice—and our Court—i.e., an appellate body, where class certifications generally are not initiated. Third, even though all members of a class need not consent to proceed before a magistrate if the named plaintiff has done so, other jurisdictional requirements must still be met. See 28 U.S.C. § 636(c)(1) (a "magistrate judge . . . may conduct any or all proceedings . . . when specially designated to exercise such jurisdiction by the district court or courts he serves").

50 second group, once more, Congress, not the Secretary, prescribed the administrative steps necessary to obtain review in our Court and insisted that claimants obtain a Board decision before appealing here. The cases discussed by the majority are inapposite, and we have neither jurisdiction over that group nor authority to accrue more power than Congress explicitly intended.

4. Under section 7252(b), we are prevented from reviewing class members' records that were not first reviewed by VA as well as the evidence Mr. Skaar submitted following the Court's limited remand.

Under section 7252(a), we would find that we do not have jurisdiction over a large portion of Mr. Skaar's proposed class because they do not have a Board decision.20 But our jurisdictional statute contains another section, which provides that our review is limited to the record before the Board or the Secretary. 38 U.S.C. § 7252(b). This statutory requirement raises issues not only for the other class members, but for Mr. Skaar as well.

a. We do not have jurisdiction to review other class members' records.

In the appellant's June 20, 2018, response to the Court's May 21, 2018, order, Mr. Skaar explained that three other veterans intended to submit the exhibits he had attached to his merits brief to a decision review officer (DRO) and the Board. The Secretary had moved to strike those documents because they were not in Mr. Skaar's record before the Board. Mr. Skaar asserted that, "[a]s a result, should this Court certify the proposed class, so much of the Secretary's motion to strike as addresses Mr. Skaar's exhibits would likely become moot, because the contested exhibits would indisputably be before the Secretary in the individual records of other class members." Appellant's June 2018 Response (Resp.) at 14, n.4. Yet, the Court could not review these documents, or any other such evidence, and make determinations based on them where the Secretary or the Board had not first reviewed those veterans' records and made findings, in a decision, as to that evidence. See 38 U.S.C. §§ 7252, 7261(c); see also Owens v. Brown, 7 Vet.App. 429, 433 (1995) (holding that the Board is responsible for assessing the credibility and weight of evidence).

b. We do not have jurisdiction to review Mr. Skaar's supplemental record.

Mr. Skaar and the majority faced a significant impediment in reaching class certification. Mr. Skaar's arguments could not result in class certification unless he and the majority found a way to force many hundreds of pages of documents that he did not present to the Board before us. They were not part of his record of proceedings, cannot plausibly be said to have been constructively before the Board, and are not of the kind subject to judicial notice. Mr. Skaar is the only named veteran. We are not aware of any potential class member that has obtained a final decision after submitting the documents in question to the Board.

The Secretary asked us to strike those documents. In a typical case, we certainly would have granted the motion. In this instance, however, members of the majority issued an order on February 1, 2019, which we will refer to as the limited remand order. For reasons we need not

20 And some do not even have a claim filed with VA that would lead to such a decision.

51 repeat here, that order was not in accordance with law. See Skaar v. Wilkie, 31 Vet.App. 16, 22 (2019) (Pietsch, J., dissenting). We cannot condone the Court's decision to use a record created by judicial artifice to certify a class. See Camp v. Pitts, 411 U.S. 138, 142 (1973) (per curiam) ("[T]he focal point for judicial review [of agency conduct] should be the administrative record already in existence, not some new record made initially in the reviewing court.").

c. Mr. Skaar should not have been permitted to submit additional evidence to the Board and we do not have jurisdiction to review those documents.

The Court should not have permitted Mr. Skaar to submit additional evidence after the limited remand. The Court, in Kutscherousky v. West, explained that providing an appellant with 90 days to submit additional evidence and argument to the Board after a Court remand was "consistent with the shift of the claim upon remand by the Court from the Court's adversarial process back to the nonadversarial, ex parte adjudication process carried out on behalf of the Secretary." 12 Vet.App. 369, 372 (1999) (per curiam order); see Williams v. Wilkie, ___Vet.App. ___, No. 16-3988, 2019 WL 4365058, *6 (Sept. 13, 2019). This case never left the adversarial process. The Court explicitly stated in its limited remand order that it retained jurisdiction over the matter. Unlike in Kutscherousky, where the Court stated, as justification for allowing the submission of additional evidence and argument, that the "nonadversarial process should begin anew with a full de novo adjudication," 12 Vet.App. at 372, the majority in Mr. Skaar's case indicated that "what we require from the Board is not a new decision," Skaar, 31 Vet.App. at 19. Rather, the Court required only a supplemental statement of reasons or bases from the Board addressing in the first instance a challenge to the dose methodology that Mr. Skaar made prior to the April 2017 Board decision.

As the word "supplemental" reveals, the April 2017 Board decision remains the jurisdiction-conferring decision on appeal. Mr. Skaar's submissions plainly run afoul of our caselaw stating that we may not review documents postdating the Board decision on appeal. See Obert v. Brown, 5 Vet.App. 30, 32 (1993) ("This Court is a Court of review that may consider only evidence that was in the record and before the Board in its adjudication."); Rogozinski v. Derwinski, 1 Vet.App. 19, 20 (1990). The majority ignored the matter in its decision. It should have explained why what it has done here is not artificial record-building that assisted Mr. Skaar in overcoming the obvious deficiencies in his class certification motion.

Second, in the limited remand order, the majority, by stating that Mr. Skaar could submit additional materials "including the evidence submitted to this Court," 31 Vet.App. at 19, highlighted a method for circumventing procedures that the Court itself and Congress had put in place—i.e., it offered Mr. Skaar and other veterans a way to defeat motions to strike and possibly obtain review of documents that would not otherwise be afforded. See id. at 31 (Pietsch, J., dissenting). It is difficult to read that passage and not conclude that the Court has put a thumb on the scales in this case.

Third and most importantly, we do not have jurisdiction to review the documents Mr. Skaar submitted to the Board following the limited remand order. As we noted in our dissent, id. at 31, the Federal Circuit, in Kyhn v. Shinseki, held that the Court's review of affidavits requested by the Court and generated after the Board decision on appeal "was in contravention of the jurisdictional

52 requirement that '[r]eview . . . shall be on the record of proceedings before the Secretary and the Board,'" 716 F.3d 572, 576-77 (Fed. Cir. 2013) (quoting 38 U.S.C. § 7252(b)). The documents Mr. Skaar submitted following the limited remand, which discuss dose methodology, are evidentiary in nature and were not in the record prior to the Board decision on appeal. See id. (the affidavits were evidentiary in nature and could not be considered by the Court in the first instance).

Further, the Board did not make factual findings in the first instance about much of the later-submitted evidence. See Board Mar. 26, 2019, Supplemental Statement at 2-5 (discussing evidence it had previously considered in the April 2017 Board decision, such as the April 2012 and December 2013 Air Force Memorandums and the June 2014 Air Force revised radiation dose estimate). To the extent that it made such findings, the Board addressed only one of the later- submitted documents—a December 2017 publication—and noted that it was published after the April 2017 Board decision and that the author's disagreement with the methodology used by the Air Force "does not necessarily render the June 2014 opinion 'unsound.'" Id. at 4-5. Rather than faithfully undertake the factfinding the limited remand intended, the Board correctly noted that it is limited to reviewing the evidence available at the time it renders its decision. Id. at 5. Ultimately, the Board determined that in April 2017 it had no evidentiary basis to reject the dose estimate offered by the Air Force. Id.

These correct findings mean that the limited remand order did not solve the record problems that the Court faces in this case. The Court is not permitted to review evidence submitted to the Board following the February 1, 2019, limited remand or, even if it were, to make findings of fact as to most of that evidence because the Board has not done so in the first instance. See Kyhn, 716 F.3d at 576-77. The answer remains the same as the one we proffered in our dissent to the limited remand order. The class motion should be denied and this case remanded. Then, the Board, with full jurisdiction, may consider any evidence that Mr. Skaar wishes to submit, and Mr. Skaar, should the Board deny his claim again, will be better positioned to support a class motion.

As we have noted before, "[b]ecause the [Notice of Appeal (NOA)] triggering our jurisdiction relates only to the April 2017 Board decision, the date of the Board's decision governs what materials are considered part of the record of proceedings under section 7252(b)," Skaar, 31 Vet.App. at 30 (Pietsch, J., dissenting) (citing U.S. VET. APP. R. 10(a)(1) (providing that the record before the agency consists of all evidence before the Board "on the date the Board issued the decision from which the appeal was taken" (emphasis added))). The majority, in neither the limited remand nor this order certifying the class, cites any authority indicating that a "supplement" to the Board decision on appeal is legally sufficient for it to deem the date of the supplement to be the decision date and to then augment the record accordingly. See Secretary's Apr. 23, 2019, Resp. at 1 n.1 (arguing that the Board's supplemental statement is not a decision because it does not grant or deny relief as required by 38 U.S.C. § 7104(d)(2)).

Therefore, Mr. Skaar has not met the jurisdictional requirement under section 7252(b) such that he may adequately represent other class members in challenging the dose methodology, where (1) that challenge is based on documents not previously reviewed by the Board, and (2) we are not permitted to review or make findings as to most, if any, of the evidence submitted following the limited remand.

53 5. In addition to our jurisdictional restrictions, our procedural statutes limit our scope of review.

Section 7261(c) provides that "[i]n no event shall findings of fact made by the Secretary or the Board . . . be subject to trial de novo by the Court." 38 U.S.C. § 7261(c). The majority acknowledges this, stating that "[o]ur appellate nature and national jurisdiction make us stand apart from the ordinary course of aggregate litigation in federal district courts, which are empowered to find facts and conduct discovery while we are not." Ante at 30 (citing 38 U.S.C. §§ 7252(b), 7261(c)). Our procedural limitations make it near impossible to develop a motion for class certification as well as adjudicate the merits of the appeal without dubious mechanisms like the limited remand order.

The majority cites to three cases apparently to demonstrate that we are perhaps not so unlike district courts. Ante at 30, 33. The majority first cites Erspamer v. Derwinski, 1 Vet.App. 3, 10 (1990), noting that the Court may consider facts not before the Board when addressing the merits of a petition. But, in considering whether to grant a petition, the Court necessarily requires information not included in the record before the Board, such as evidence of actions taken by VA to process a veteran's claim where delay is alleged. See Cheney v. U.S. Dist. Court, 542 U.S. 367, 380-81 (2004) (an appellate court must determine whether mandamus is appropriate under the circumstances); Cox v. West, 149 F.3d 1360, 1363 (Fed. Cir. 1998) (this Court's jurisdiction is "irrelevant to the question of the [C]ourt's power under the AWA," which provides authority for the Court to grant petitions). That evidence is not used, as the evidence collected here is intended to be used, to address the merits of the underlying claim. It is used only for the purpose of determining whether our prospective jurisdiction has been blocked.

We are restricted by law (but see Wolfe v. Wilkie, __ Vet.App. __, No. 18-6091, 2019 WL 4254039, at *23-24 (Sept. 9, 2019) (granting the petition and invalidating a regulation despite the availability of agency remedies because obtaining a final agency determination would be "'a useless act'")) from using the facts we gather in the petition context for any purposes other than ensuring that our potential jurisdiction is protected. See Lamb v. Principi, 284 F.3d 1378, 1384 (Fed. Cir. 2002) ("'[E]xtraordinary writs cannot be used as substitutes for appeals, even though hardship may result from delay and perhaps unnecessary trial.'" (quoting Bankers Life & Cas. Co. v. Holland, 346 U.S. 379, 382 (1953))). Thus, the Court's ability to review evidence not before the Board in addressing a petition does not provide support for that same type of review of appeals, where section 7252(b) firmly restricts us from conducting discovery. For these same reasons, the majority's citation to Monk v. Wilkie (Monk III), 30 Vet.App. 167, 171 (2018) (en banc order), for the proposition that the Court has authority to conduct limited factfinding to determine whether class certification in the petition context is warranted, ante at 30, is unpersuasive in the context of adjudicating class action appeals on the merits.

Finally, the majority cites Bove v. Shinseki, stating that the Court "'may seek facts outside the record before the Board and independently weigh the facts to determine if equitable tolling is appropriate,'" ante at 30 (quoting 25 Vet.App. 136, 143 (2011) (per curiam order)). As with a petition, however, determining whether to equitably toll a late filing necessarily requires information not in the record before the Board, because the Court must assess whether events that happened after the Board issued the decision on appeal were extraordinary and prevented the claimant from timely filing the document in question despite due diligence. See Toomer v.

54 McDonald, 783 F.3d 1229, 1238 (Fed. Cir. 2015). The merits of the underlying claim are not considered at that stage and section 7252(b) does not apply. If equitable tolling is granted, the merits decision that the Court ultimately issues will be based on the record before the Board at the time of the decision on appeal, as required by law, and not on evidence gathered to determine whether equitable tolling is warranted. If it postdates the Board decision, that evidence will not appear in the record of proceedings.

Thus, the process for determining whether to equitably toll a filing deadline is not analogous to reviewing class action appeals on the merits. The distinction between our Court and district courts remains, as does the issue of how our jurisdictional and procedural statutes impact our ability to adjudicate aggregated appeals.

6. There are distinctions between an appellate court and trial courts.

The Federal Circuit in Monk II and the majority here determined that veterans should be afforded more, not less, procedural protections after the VJRA's enactment and thus, because veterans were previously allowed to aggregate appeals, they should be able to do so now. See Monk II, 855 F.3d at 1319-20; ante at 13, 16-17. The Federal Circuit found "no persuasive indication that Congress intended to remove class action protection for veterans when it enacted the VJRA." Monk II, 855 F.3d at 1320, n.4 (referencing a Congressional Budget Office cost estimate from 1988 that discussed potential litigation challenges, stating that, according to SSA, most challenges to regulations are class actions). Congress, however, created an appellate tribunal with distinct features that separate it from district courts and even other appellate courts. We must account for those differences. See Henderson, 562 U.S. at 441 ("[T]he review opportunities available to veterans before the VJRA was enacted are of little help in interpreting [a statute within the VJRA].").

The VJRA provided a new framework for veterans to pursue their disability benefits and with it a new procedure to ensure that this Court's findings applied to many veterans—i.e., a precedential decision. See 38 U.S.C. §§ 7254, 7267; see also Frankel v. Derwinski, 1 Vet.App. 23, 25-26 (1990). A precedential decision ensures that judicial determinations are broadly and consistently applied across VA and afford similar, if not greater, protections for veterans than did the rare instances of class actions in district courts prior to the VJRA. To the extent that our jurisdictional requirements inhibit our ability to certify a class in the appeal context, we assume that Congress was aware of any such limitations when it enacted the VJRA. If Congress wishes to expand our class action authority in the appeal context, then it should legislate the change to our jurisdictional statute. It is not for us to enhance our own authority by rewriting statutes to suit our preferences.

Three of the five cases cited by the majority (and the Federal Circuit in Monk II) to demonstrate that veterans were previously able to aggregate cases are district court cases. Ante at 13 (citing Nehmer v. U.S. Veterans' Admin., 118 F.R.D. 113 (N.D. Cal. 1987); Giusti-Bravo v. U.S. Veterans Admin., 853 F. Supp. 34 (D.P.R. 1993); In re Agent Orange Prod. Liab. Litig., 506 F. Supp. 762 (E.D.N.Y. 1980)). Nehmer demonstrates a key distinction between an appellate court and a trial court. There, the U.S. District Court for the Northern District of California reasoned that class members did not need to exhaust administrative remedies because, inter alia, a full record

55 would be available through discovery. Nehmer, 118 F.R.D. at 122. All agree that our Court is precluded by statute from relying on discovery to complete the record.

The other two cases the majority and the Federal Circuit reference are from appellate courts. In both, the trial courts determined whether the classes should be certified prior to the cases being appealed. See Johnson v. Robison, 415 U.S. 361, 364 n.3 (1974) (noting that class action was commenced in the U.S. District Court for the District of Massachusetts and that the district court defined the class); Wayne State Univ. v. Cleland, 590 F.2d 627, 628, n.1 (6th Cir. 1978) (indicating that the district court certified the class and remanding in part for the district court to decide a matter in the first instance). The appellate courts reviewed the propriety of decisions regarding class actions but did not, as the majority is trying to do, make certification decisions in the first instance.

The primary tension in this case is that we are an appellate court doing what appellate courts normally should not do. Trial courts are equipped to certify classes and adjudicate aggregated cases because they are not statutorily prohibited from supplementing the record through discovery and making factual findings in the first instance. Our inability to conduct those basic functions vital to certifying a class means that, unless Congress restructures our authority, adjudicating class certification cases that come to us through an appeal will likely present the jurisdictional hurdles that we have seen in this case.

B. The Utility of Class Actions in the Appeal Context

In Harrison, our Court declined to establish class action procedures, in part, because they would be "highly unmanageable" and because class actions are "unnecessary," given the binding effect of the Court's precedential decisions in pending and future cases. 1 Vet.App. at 438-39. Although the Federal Circuit disagreed with our finding in Harrison that we lack authority to certify a class, it did not disturb our determination that class actions are unnecessary and highly unmanageable. See Monk II, 855 F.3d at 1320. That conclusion was correct when Harrison issued and remains so today.

1. Class actions are unnecessary because we can issue precedential decisions, which may be used to attain institutional change and efficiency.

The majority, referencing Monk II, states that class actions will stop VA from preventing judicial review of meritorious arguments by mooting the cases in which they arise. Ante at 14-15. The Federal Circuit authority on which the majority relies applies to petitions, not appeals. The Federal Circuit noted that in Young v. Shinseki, 25 Vet.App. 201, 215 (2012) (en banc per curiam order) (Lance, J., dissenting), the dissent explained that VA's delay in adjudicating appeals evades review at times because VA usually acts promptly to resolve petitions. Monk II, 855 F.3d at 1320- 21. The Federal Circuit noted that, after we order VA to respond to a petition, "the 'great majority of the time' the VA 'responds by correcting the problem within the short time allotted for a response, and the petition is dismissed as moot.'" Id. (quoting Young, 25 Vet.App. at 215 (Lance, J., dissenting)).

56 The Secretary cannot "moot" an appeal in the same manner that he can "moot" a petition. The Secretary may offer to settle an appeal, but that offer must be accepted by an appellant. A motivated appellant who has decided to place his or her own outcome second to the greater cause of veterans rights—in other words, an appellant like Mr. Skaar—can always decline even the most generous settlement offer if a greater victory remains to be won. The Secretary also may concede error before a judge, but the Court is free to ignore or accept his concession and find additional errors. Unlike with petitions, the Secretary cannot unilaterally stop an appeal from proceeding to judicial review or control the outcome once it reaches a judge or panel.

Second, the majority states that class actions "can . . . be an effective force for institutional change" and may be used to correct systemic error and ensure that veterans are treated alike. Ante at 14-15. Leaving aside for a moment the problem of judicial overreach inherent in that declaration, a precedential decision may be used to achieve the same objective. See Harrison, 1 Vet.App. at 438 (finding that class actions were unnecessary due to the binding effect of precedential decisions). If we had an adequate record, a panel might have, months ago, found that the dose methodology VA used in Mr. Skaar's case was flawed and counter to 38 C.F.R. § 3.311. Its decision, a nationwide precedent, would have fixed any such systemic dose estimate problem and VA would have been required to apply the Court's holding consistently to all veterans' cases.

The majority responds that claimants not party to a panel decision and potentially subject to errors affecting their rights, whether due to VA's non-compliance with our decision or otherwise, "do not have any right to prompt remedial enforcement." Ante at 33. The assumption that VA will not comply with our precedential decisions, like the assumption that it will moot every potential embarrassment, is needlessly cynical and suggests that we are acting at least in part with punitive intent. Moreover, all so-called Future-Future claimants' claims will be governed by the precedent, Present-Future claimants can point out the new precedent to VA and ask for it to be considered, claimants on appeal at the Court can ask for a remand based on the new precedent, and claimants who have already received a final decision may seek to reopen or file a supplemental claim.

Furthermore, if we found against an appellant in a precedential decision, other claimants impacted by that decision will have a full and fair opportunity to attempt to distinguish their cases. Bound class members will presumably have no such leverage. Given the difficulty in conveying the meaning of a class litigation, they may be surprised by the fact that their individual cases are subsumed and decided through arguments made by another.

Third, the majority and the Federal Circuit in Monk II tout class actions as an efficient method for correcting VA error. This case is not good support for that position, as we are now, more than 800 days after Mr. Skaar filed his NOA, issuing our third substantive en banc order and have not begun to address the merits.

We see no indication that class certification appeals are going to move as quickly as the average panel decision, particularly where the class appeals would require the additional step of certifying the class. We also are not moved by the novelty of this case. Had the Court waited to develop rules for aggregate litigation rather than issue a string of contested ad hoc decisions, it might have significantly reduced the time and resources it has expended on this case.

57 Finally, the procedural history of Mr. Skaar's case demonstrates that aggregated appeals at our Court may not be as efficient as expected. As we noted above, given our inability to conduct discovery, limited remand decisions or other suspect mechanisms may routinely be necessary to grant future class motions. That can only lead to delay.

2. Class actions are more unmanageable for our appellate Court than they are for trial courts.

The majority states that the Harrison manageability factor stems from the unique nature of the Court and, although it is not a categorical reason to decline class certification, it is a relevant consideration. Ante at 31. The majority indicates that class actions will only be allowed if the appellant demonstrates the superiority of the class action to a precedential decision. It then sets forth a several factor balancing test, cut from whole cloth, for making this determination. Id. at 32. One factor is whether the record is sufficiently complete for adjudication, including whether additional factfinding is needed. Id. at 33 (acknowledging Supreme Court precedent that the record may not be created by a reviewing court and that we have unique limitations on factfinding). That factor is no factor at all if limited remands are to become the norm in class cases. We also believe that there are additional related considerations.

First, when assessing whether the named appellant meets the section 7252(b) requirement—such that we have jurisdiction not only over his or her record but also over class members who themselves do not meet our jurisdictional prerequisites—the Court should not rely on circuitous methods (e.g., limited remands) to find this requirement satisfied, which presumably would become unmanageable over time. Rather, the record of the named appellant should be itself complete before appeal to this Court. By using the limited remand here, the Court has provided a poor and probably misleading example of how these cases should be handled in the future. Its actions do not square with its Harrison analysis.

Relatedly, class actions are more unmanageable for our Court because, for class members who do not meet section 7252 jurisdictional requirements and there is no record of proceedings, we cannot make necessary factual findings in the first instance. Therefore, although some potential class members here purportedly submitted the same scientific evidence to VA that was the subject of the Secretary's motion to strike, we are not persuaded by Mr. Skaar's argument that any problem in reviewing this evidence was resolved. In other words, records not reviewed by VA cannot be used to supplement the named appellant's incomplete record. We do not have jurisdictional authority to review those records even if they contained more favorable evidence than that found in the named appellant's record. The evidence must come before us in the form of a record of proceedings from a properly appealed Board decision.

Second, we once more reiterate that this Court does not have the same discovery and factfinding abilities as trial courts. See Nehmer, 118 F.R.D. at 122 (highlighting a key distinction between those courts and our appellate body when the Northern District of California determined that class members did not need to exhaust administrative remedies because, inter alia, a full record would be available through discovery). Further, the Court also does not have the same ability as a trial court to hear from an expert about complex scientific matters.

58 As indicated throughout our dissent, we would find that the third factor of the majority's balancing test, the completeness of the record, heavily weighs against certifying the class in Mr. Skaar's case, particularly where the Court is not permitted to review the evidence submitted to the Board following the February 1, 2019, limited remand or, even if it were, to make findings of fact as to most of that evidence. See Kyhn, 716 F.3d at 576-77.

C. Class Certification in Mr. Skaar's Case

For the most part, we will not address the majority's class certification analysis. As to numerosity, however, based on our view that we do not have jurisdiction over those veterans without a final Board decision, we would find that Mr. Skaar's proposed class does not satisfy this factor.

The Secretary stated that he knew of only six Palomares veterans who had received a Board decision (adverse or not) from 2001 to the present addressing any claim dealing with claimed ionizing radiation exposure concerning the Palomares cleanup. See Secretary's Dec. 13, 2018, Resp. at 3. Mr. Skaar responded that the record reflects that there are "at least [19] veterans who had filed claims for Palomares-related disabilities with the VA, 'including [3] appeals for reassessment for a total of 22 claims,'" Appellant's Jan. 4, 2019, Resp. at 3 (quoting R. at 1580), but he does not indicate how many of those claims resulted in a Board decision. Even if the six Board decisions referenced by the Secretary pertain to 38 C.F.R. § 3.311 and applied the post- 2013 methodology (given that Mr. Skaar does not have standing to challenge 38 C.F.R. § 3.309 or the pre-2013 methodology)21 and that those decisions were adverse,22 six or seven potential class members is not sufficient to fulfill the numerosity requirement.

Although such an adverse finding on numerosity would be dispositive when assessing whether to certify a class, we will also briefly address the adequacy of the class representative. Mr. Skaar cannot adequately protect the interests of the class because we do not have jurisdiction to review the evidence he submitted to the Board following its April 2017 decision (i.e., the documents that form the basis of his challenge to the VA methodology) or to make any determinations regarding that evidence.

21 Although we recognize that it is unlikely that there are individuals who have a dose estimate based solely on the pre-2013 methodology, we would find that Mr. Skaar lacks standing to challenge the pre-2013 methodology and only has standing to challenge the post-December 2013 methodology. He suffered no injury-in-fact in his current appeal based on the pre-2013 methodology. The Board expressly discounted the findings of the May 2012 advisory opinion, which were based on the pre-2013 methodology. R. at 10. Although there is some overlap between standing and typicality, the majority appears to have conflated these issues when explaining how Mr. Skaar has established standing. The majority states that, if he is successful in showing that the exclusion of urine samples was not based on sound scientific evidence, he will have suffered an injury-in-fact. Ante at 12-13. But that only indicates that he may satisfy the typicality requirement—his issue regarding the urine sample exclusion is typical of class members with dose estimates based on both pre- and post-2013 methodologies. However, this does not show that Mr. Skaar in his current appeal was harmed by the pre-2013 methodology. 22 Although this is unlikely because the Secretary also stated that he knew of three Palomares veterans who had received an adverse Board decision from 2001 to the present addressing any claim dealing with claimed ionizing radiation exposure concerning the Palomares cleanup. See Secretary's Dec. 13, 2018, Resp. at 3.

59 Finally, related to our concern that class actions, if unfavorable to the class, may preclude members from raising different arguments as to the dose methodology, we question whether Mr. Skaar has presented the best argument to challenge this methodology. See McDowell v. Brown, 5 Vet.App. 401, 408 (1993) (noting that "courts will more carefully scrutinize the adequacy of representation afforded to absent [class] members [who are not afforded notice and opt-out protections] . . . before determining that they are bound, by res judicata, by the final judgment or settlement in the prior class action."). Mr. Skaar's argument focuses on the 2001 Labat-Anderson (LA) Report. Although the April 2012 Air Force dose estimate relied, in part, on the LA Report, R. at 1888 (citing seven other references), it is unclear whether the Air Force's post-December 2013 methodology relied on that report. We note that, at least on its face, the June 2014 Air Force memorandum regarding revised radiation dose information does not appear to rely on the LA Report because it does not mention it and instead states that the new dose estimates were based on International Committee on Radiological Protection (ICRP) reports. R. at 1301-02. It may be that Mr. Skaar has presented the best challenge to the VA methodology, but we believe that the majority should "more carefully scrutinize" this matter where preclusion is an issue. See McDowell, 5 Vet.App. at 408.

II. CONCLUSION

This case highlights some of the jurisdictional and practical challenges inherent in entertaining class actions in an appeal context, given the statutory framework that governs our review of Board decisions and the record before the Board or Secretary. See Ledford v. West, 136 F.3d 776, 779 (Fed. Cir. 1998) ("[T]he court's jurisdiction is premised on and defined by the Board's decision concerning the matter being appealed."). The majority has created a mechanism that exceeds our jurisdiction and offers no more benefits than a precedential decision, but with significant manageability and preclusion problems. Although we are sympathetic to the veterans who served in Palomares and who may have suffered injuries as a result, and we applaud Mr. Skaar's efforts to remedy this matter for all veterans, a class action in the appeal context is no answer. A simple precedential decision on this issue, when properly before the Court, would more efficiently provide them and Mr. Skaar with the answers they deserve.

Finally, we are concerned with the manner that this case has been handled. The Court has seized more power than Congress allotted to it with unsound legal innovations.

For these reasons, we respectfully dissent.