OPINION
This case once again pits the federal government’s efforts to save the bald eagle from extinction against the bird’s profound significance to Native spirituality. Appellant Leonard Fridall Terry Antoine, a member of a Canadian Indian tribe, is spending two years in prison for violating the Bald and Golden Eagle Protection Act (BGEPA), 16 U.S.C. §§ 668-668d. We must decide whether his conviction violates the Religious Freedom Restoration Act (RFRA), 42 U.S.C. §§ 2000bb to 2000bb-4.
1. Antoine is a member of the Cowi-chan Band of the Salish Indian Tribe in British Columbia. He obtained dead eagles in Canada and brought feathers and other eagle parts into the United States, where he swapped them for money and goods. Antoine claims that these exchanges are part of the native custom of “potlatch,” which to him has religious significance. United States authorities charged him with violating BGEPA, which makes it illegal to “knowingly ... take, possess, sell, purchase, barter, offer to sell, purchase or barter, transport, export or import, at any time or in any manner, any bald eagle” or part thereof. 16 U.S.C. § 668(a).
Notwithstanding this prohibition, members of federally recognized Indian tribes can apply for permits to possess and transport eagles or eagle parts for religious purposes. See id. § 668a; 50 C.F.R. § 22.22. Federal wildlife agents who find eagle carcasses send them to a repository in Colorado, which fills applications on a first-come, first-served basis. Because demand significantly exceeds supply, the waiting list is several years long. Antoine is not eligible for a religious use permit at all, however, because his band is not recognized by the United States.
Antoine moved to dismiss the prosecution, arguing that he was exempt from BGEPA under RFRA, which suspends generally applicable federal laws that “substantially burden a person’s exercise of religion” unless the laws are “the least restrictive means of furthering [a] compelling governmental interest.” 42 U.S.C.
*921
§ 2000bb-1(a)-(b);
Guam v. Guerrero,
2. We confronted the intersection of RFRA and BGEPA on one prior occasion:
United States v. Hugs,
Antoine distinguishes
Hugs
on two grounds. He first notes that two years after
Hugs
was decided, the Fish and Wildlife Service proposed removing the bald eagle from the threatened species list because “available data indicate[d] that this species ha[d] recovered.” Proposed Rule To Remove the Bald Eagle in the Lower 48 States from the List of Endangered and Threatened Wildlife, 64 Fed. Reg. 36,454, 36,454 (proposed July 6, 1999). We find the force of this evidence limited. The proposed rule is just that; the Service has not made a final decision to delist. Agencies issue proposed rules in order to educate themselves about their likely effects.
See
5 U.S.C. § 553(c). The Service may well revise its analysis in light of the information it receives. Because the delisting proposal is based on incomplete information, it carries less weight than a final rule.
See Oliver,
The delisting proposal concededly provides some support for Antoine’s argument that the eagle-protection interest is weaker than when
Hugs
was decided. And changed circumstances may, in theory, transform a compelling interest into a less than compelling one, or render a well-tailored statute misproportioned. Nonetheless, the government cannot reasonably be expected to relitigate the issue with
*922
every increase in the eagle population.
Cf. Nixon v. Shrink Mo. Gov’t PAC,
Antoine’s second argument presents a more difficult question. Unlike the defendants in Hugs, Antoine is ineligible for a permit because he is not a member of a recognized tribe. He argues that his exclusion from the permit scheme violates RFRA and so he cannot be prosecuted for obtaining eagles by other means. 4 We do not read Hugs to foreclose this aspect of Antoine’s challenge. The question whether the permit regime in general is valid is distinct from the question whether its restriction to recognized tribe members is valid. The latter question was neither presented nor decided in Hugs; the defendants there were eligible for permits but chose not to pursue them. Hugs defeats arguments that the government must increase the number of eagles available (by allowing people to kill their own, for example), but it doesn’t speak to how that limited supply of eagles is allocated.
Circuits have split over the exclusion of nonmember Indians from the permit program. In
Gibson v. Babbitt,
The Tenth Circuit saw things differently in
United States v. Hardman,
We do not believe RFRA requires the government to make the showing the Tenth Circuit demands of it. Although the record contains no data on the number of nonmembers who would seek permits if eligible, the consequences of extending eligibility are predictable from the nature of the repository program. The supply of eagles is fixed because the government distributes every eagle and eagle part that comes into the repository; Hugs’s conclusion that the permit program is the least restrictive means of protecting eagles forecloses any challenge to the government’s refusal to increase supply beyond that. 6 If the government extended eligibility, every permit issued to a nonmember would be one fewer issued to a member. This is the inescapable result of a demand that exceeds a fixed supply.
RFRA requires least restrictive means to avoid substantial burdens on religion. But, in this case, the burden on religion is inescapable; the only question is whom to burden and how much. Both member and nonmember Indians seek to use eagles for religious purposes. The government must decide whether to distribute eagles narrowly and thus burden nonmembers, or distribute them broadly and exacerbate the extreme delays already faced by members. Religion weighs on both sides of the scale. The precise burdens depend on how many nonmember applicants there would be, but not in any illuminating way: Fewer nonmember applicants means shorter additional delays for each member if the restrictions are removed, but also fewer people burdened if they are left in place.
Our cases enforcing RFRA’s least-restrictive-means requirement have involved the pursuit of some secular interest in a manner that burdens religion.
See Mockaitis v. Harcleroad,
The permit program does not discriminate facially on the basis of religion in any way that harms Antoine.
Cf. Davey v. Locke,
Antoine’s prosecution did not violate RFRA, so the district court properly rejected his claim. He raises several other challenges to his conviction and sentence, which we address in a separately filed memorandum disposition.
AFFIRMED.
Notes
. We treat Antoine's foreign citizenship and residency as irrelevant to his RFRA standing because the government has not asked us to consider them. RFRA does apply to any "person's” exercise of religion, 42 U.S.C. § 2000bb-1, which arguably suggests coverage of all individuals subject to the government’s jurisdiction.
Cf. United States v. Verdugo-Urquidez,
. Antoine was accused of dealing in eagle parts, not just using them in traditional Native ceremonies. This fact does not defeat his RFRA claim. Antoine asserts that, to him, potlatch exchange of eagle parts has religious significance. The government counters that potlatch is a cultural practice rather than a religious one. But what matters is its significance to Antoine, not to others. The district court assumed all of Antoine's beliefs were sincere, and we have no basis for disturbing that assumption.
.Hugs never actually says that the defendants were tribe members, only that they were engaged as hunting guides on a reservation.
See
. Because Antoine is ineligible, he need not apply for a permit in order to challenge the permit scheme.
See Desert Outdoor Adver., Inc. v. City of Moreno Valley,
. As in
Hardman,
. Antoine brought his eagle parts into the United States from Canada. He argued below that, by doing so, he actually increased supply and alleviated demand on the repository. Eagles, however, are migratory, 50 C.F.R. § 10.13, so trafficking in Canada necessarily affects the American population. The United States's interest in eagle protection therefore extends across the border.
. The regulations authorize bald eagle possession permits for scientific and exhibition purposes as well as religious purposes. 50 C.F.R. § 22.21. A nonmember Indian might argue that scientific and exhibition uses are not compelling, or that the inclusion of these secular permits renders the program poorly tailored. But Antoine does not raise these arguments; he focuses exclusively on the disparate treatment of religious users.
. The permit program does discriminate facially on the basis of religion within the class of recognized tribe members.
See
50 C.F.R. § 22.22 (authorizing permits only for "members of [recognized] Indian entities” who are "engaged in
religious
activities” (emphasis added)). A recognized tribe member who wanted eagle feathers for nonreligious purposes might have standing to challenge the program on this basis. But Antoine is excluded by the secular classification in the statute — the restriction to recognized tribe members. As to him, the regulation is neutral with respect to religion.
Cf. Rupert v. Dir., U.S. Fish & Wildlife Serv.,
. Antoine argues that the government’s treaty interest is "plagued with inconsistency” because his tribe was the beneficiary of a treaty allowing free passage and commerce between Canada and the United States. The government, however, need not restore either all treaty rights or none. It could rationally decide to restore the treaty rights of only those tribes with which it currently has relations. Its undoubted power to target benefits to recognized tribes implies the more specific power to select only their treaty rights for restoration.
