Apache Stronghold v. United StatesApache Stronghold v. United States
OPINION
Appeal from the United States District Court for the District of Arizona Steven Paul Logan, District Judge, Presiding
Argued and Submitted October 22, 2021 San Francisco, California
Filed June 24, 2022
Opinion by Judge Bea; Dissent by Judge Berzon
SUMMARY*
Religious Freedom Restoration Act / Free Exercise Clause
The panel affirmed the district court‘s denial of Apache Stronghold‘s motion for a preliminary injunction seeking to stop a land exchange and prevent any copper mining on Oak Flat, a plot of land in Arizona.
A 2014 act of Congress requires the U.S. Secretary of Agriculture to convey Oak Flat to Resolution Copper, a mining company. In exchange, Resolution Copper will convey to the United States a series of nearby plots of land (the “Land Exchange“). To the Apache American Indians, Oak Flat, known to the Apache as Chi‘chil Biłdagoteel, is sacred ground. Apache Stronghold, a nonprofit organization, sued the government, alleging that the Land Exchange violated the Religious Freedom Restoration Act (“RFRA“), the Free Exercise Clause of the Constitution‘s First Amendment, and a trust obligation imposed on the United States by the 1852 Treaty of Santa Fe between the Apache and the United States.
Concerning Apache Stronghold‘s RFRA claim, the panel began by addressing what constituted a “substantial burden” under RFRA. First, RFRA by its text restored Sherbert v. Verner, 374 U.S. 398 (1963), Wisconsin v. Yoder, 406 U.S. 205 (1972), their “compelling interest” test, and their “substantial burden” inquiry, and defined a “substantial burden” under RFRA as either of the burdens present in those two cases. Second, the Supreme Court has used the phrase “substantial burden” as a Free Exercise Clause term of art that meant only the two burdens within the Sherbert/Yoder framework, and a “substantial burden” under RFRA must hold that same settled meaning. Third, Lyng v. Northwest Indian Cemetery Protective Association, 485 U.S. 439 (1988), and Bowen v. Roy, 476 U.S. 693 (1986), the cases most factually and legally analogous to Navajo Nation v. United States Forest Service, 535 F.3d 1058 (9th Cir. 2008) (en banc), and this case, confirmed that even burdensome government action did not constitute a “substantial burden” (and did not trigger the “compelling interest” test) if that action fell outside the Sherbert/Yoder framework.
The panel next turned to Apache Stronghold‘s main argument that the Land Exchange would hand Oak Flat over to Resolution Copper for its mining plan, thus incidentally making it impossible for Apache Stronghold‘s members to worship on Oak Flat and thereby substantially burdening them. The panel held that this argument could not succeed in light of Navajo Nation. The Land Exchange‘s effect on Apache Stronghold‘s members fell outside of the Sherbert/Yoder framework, and thus outside of RFRA‘s definition of a substantial burden. No government benefits will be lost (as in Sherbert) nor will governmental penalties be imposed (as in Yoder). The Department of Agriculture will simply transfer ownership of a plot of government land to Resolution Copper, and the Land Exchange does not coerce the Apache to abandon their religion by threatening them with a negative outcome. Because Apache Stronghold‘s members have not established that they would suffer a substantial burden under RFRA, Apache Stronghold is not likely to succeed on its RFRA claim. The panel rejected Apache Stronghold‘s and the dissent‘s contentions to the contrary.
Next, the panel addressed Apache Stronghold‘s secondary argument that the Land Exchange did in fact deprive its members of a benefit and subjected its members to a penalty. Namely, the Land Exchange allegedly deprived Apache Stronghold members of the “use and enjoyment of ‘government’ land for religious exercise” and subjected them to penalties for “trespassing on now ‘private’ land.” The panel disagreed. The government does not substantially burden religion every time it ends a governmental benefit that at one time went to religious beneficiaries: there must be an element of coercion. The Land Exchange does not “condition” any government benefits on the Apache violating their religious beliefs. The panel also rejected Apache Stronghold‘s argument that the Land Exchange subjected its members to penalties: liability for trespassing on land that will be private after the Exchange. Apache Stronghold has not shown a sufficiently realistic fear of future criminal trespass liability. Also, Apache Stronghold seeks relief that RFRA cannot provide: RFRA does not authorize Apache Stronghold to enjoin the entire Land Exchange. Similarly, it is not clear that the Apache will be subject to civil trespass liability. But even if Apache Stronghold‘s members were subject to the threat of imminent civil trespass suits, the panel could not enjoin the entire Land Exchange as Apache Stronghold requested.
The panel rejected Apache Stronghold‘s claim that the Land Exchange would violate the Constitution‘s Free Exercise Clause. Apache Stronghold argued that the Land Exchange Provision was neither neutral nor generally applicable and thus was subject to strict scrutiny. The panel held that the Land Exchange was neutral in that its object was not to infringe upon the Apache‘s religious practices. All the evidence suggests that the Land Exchange was meant to facilitate mineral exploration activities – nothing more and nothing less. The panel concluded that the district court properly found that Apache was not likely to succeed on its Free Exercise claim.
Last, the panel considered Apache Stronghold‘s trust claim under the Treaty of Santa Fe. Namely, that the Treaty created an enforceable trust obligation on the U.S. government, and the Land Exchange was inconsistent with the U.S.‘s obligation to pass laws conducive to the prosperity and happiness of the Apache. The panel agreed with the government that on this record, Apache Stronghold has not established that the Treaty of Santa Fe imposes on the United States an enforceable trust obligation. The panel concluded that Apache Stronghold‘s trust claim was unlikely to succeed.
The panel recognized the deep ties the Apache have to Oak Flat, and acknowledged that the Land Exchange may impact the Apache‘s plans to worship at Oak Flat. But RFRA, the Free Exercise Clause, and the 1852 Treaty of Santa Fe do not afford Apache Stronghold the relief that it seeks.
Dissenting, Judge Berzon wrote that the majority applied an overly restrictive test for identifying a “substantial burden” on religious exercise under RFRA. The majority‘s flawed test leads to an absurd result: blocking Apaches’ access to and eventually destroying a sacred site where they have performed religious ceremonies for centuries did not substantially burden their religious exercise. There was no doctrinal basis for limiting the definition of “substantial burden” to the types of burdens imposed in Sherbert and Yoder. The majority‘s proffered practical basis for its constricted definition of “substantial burden” is also flawed. Applying the correct definition of “substantial burden,” Judge Berzon would hold that Apache Stronghold has shown that it is likely to succeed on the merits of its RFRA claim. She would remand for the district court to address the remaining elements of the preliminary injunction test.
COUNSEL
Luke W. Goodrich (argued), Mark L. Rienzi, Diana M. Verm, Joseph C. Davis, Christopher Pagliarella, Daniel D. Benson, And Kayla A. Toney, The Becket Fund for Religious Liberty, Washington, D.C.; Michael V. Nixon, Portland, Oregon; Clifford Levenson, Phoenix, Arizona; for Plaintiff-Appellant.
Joan M. Pepin (argued), Andrew C. Mergen, Tyler M. Alexander, and Katelin Shugart-Schmidt, Attorneys; Jean E. Williams, Acting Assistant Attorney General; Environment and Natural Resources Division, U.S. Department of Justice, Washington, D.C.; for Defendants-Appellees.
Gene C. Schaerr and Joshua J. Prince, Schaerr | Jaffe LP, Washington, D.C.; James C. Phillips, Chapman University, Dale E. Fowler School of Law, Orange, California; for Amici Curiae Jewish Coalition for Religious Liberty, The International Society for Krishna Consciousness, The Sikh Coalition, and Protect The 1st.
Miles E. Coleman, Greenville, South Carolina; Thomas Hydrick and Hunter Windham, Columbia, South Carolina; for Amici Curiae Religious Liberty Law Scholars.
Stephanie Hall Barclay, Associate Professor of Law, Director, Religious Liberty Initiative, Notre Dame Law School, Notre Dame, Indiana; Michalyn Steele, Professor of Law, BYU Law School, Provo, Utah; for Amici Curiae National Congress of American Indians, A Tribal Elder, and Other Federal Indian Law Scholars and Organizations.
William E. Trachman, Mountain States Legal Foundation, Lakewood, Colorado; Timothy Sandefur, Goldwater Institute, Phoenix, Arizona; for Amici Curiae the Towns of Superior, and Hayden, Arizona, and Jamie Ramsey, the Mayor of Kearny, Arizona.
David Debold, Thomas G. Hungar, and Matthew S. Rozen, Gibson, Dunn & Crutcher, LLP, Washington, D.C.; for Amici Curiae American Exploration & Mining Association, Women‘s Mining Coalition, and Arizona Rock Products Association.
OPINION
BEA, Circuit Judge:
A 2014 act of Congress requires the U.S. Secretary of Agriculture to convey Oak Flat, a plot of federal land in Arizona, to a mining company named Resolution Copper. In exchange, Resolution Copper will convey to the United States a series of other nearby plots of land (the “Land Exchange“). Resolution Copper is considering constructing a copper mine under Oak Flat to access one of the world‘s largest undeveloped copper deposits. But to the Apache American Indians, Oak Flat – or as the Apache call it, Chi‘chil Biłdagoteel – is sacred ground. So Apache Stronghold, a non-profit organization formed to preserve and protect American Indian sacred sites, sued the government on the grounds that the Land Exchange violates each of: 1) the Religious Freedom Restoration Act (“RFRA“),
I. Background
A. The At-Issue Land
The Tonto National Forest stretches across nearly 3 million acres (or about 4,500 square miles) across Arizona. See Tonto National Forest, U.S. Dep‘t of Agriculture, https://www.fs.usda.gov/detail/tonto/home/?cid=fsbdev3_018924 (last visited June 15, 2022). Most of the forest is owned by the United States and is managed by the United States Forest Service, a division of the United States Department of Agriculture. See id. Within the Tonto Forest is Oak Flat, a 6.7-square-mile plot of plains, oak groves, and rocky cliffs that sits about 4,000 feet above sea level. Beneath Tonto Forest and extending under part of Oak Flat lies “one of the largest undeveloped copper deposits in the world,” containing an estimated 1,970 billion tons of copper.
Also within the Tonto National Forest are several areas sacred to the Apache American Indians. Oak Flat is one of these areas, as are Devil‘s Canyon (called Ga‘an Bikoh by the Apache), a depression just east of Oak Flat, and Apache Leap (called Dibecho Nadil by the Apache), a steep slope just to Oak Flat‘s west. These three adjacent areas are places where the Apache‘s Ga‘an – beings that the Apache describe as their “creators, [their] saints, [their] saviors, [their] holy spirits” – live and where the Apache can communicate with them. Currently, the federal government owns Oak Flat.1 Devil‘s Canyon is owned partially by Arizona state government trusts2 and partially by the federal
In recent years, Oak Flat has been used for a variety of purposes, both religious and secular. After decades of holding religious rituals on their reservations, the Apache have recently returned to worship in Tonto Forest. In 2014, the Apache held a “Sunrise Dance” on Oak Flat for just the second time in “more than a hundred years.” That 2014 ceremony closely followed another Sunrise Dance held the previous year at Mt. Graham, another sacred site elsewhere in Arizona. Separately, recreational users often camp, hike, or rock-climb throughout Tonto National Forest, including on Oak Flat.
B. The Land Exchange Provision
After nearly a decade of debate, Congress included in the 2014 National Defense Authorization Act a provision (the “Land Exchange Provision“) that requires the Secretary of Agriculture to complete a land swap arrangement with Resolution Copper. See National Defense Authorization Act for Fiscal Year 2015, Pub. L. No. 113-291, § 3003, 128 Stat. 3732-41 (2014) (codified at
Once the Forest Service and Resolution Copper exchange the land specified in the Land Exchange Provision, Resolution Copper expects to take “several years” to conduct a “detailed feasibility study” regarding whether to proceed with a mine on the land it receives. Under Resolution Copper‘s current proposal, it would use a mining technique called “panel caving“; while Resolution Copper would not need to dig a mine on the surface, the land over the mine would eventually subside, “profoundly and permanently alter[ing]” the landscape.
The Land Exchange Provision also requires a series of consultation and mitigation measures. The Secretary of Agriculture must conduct “government-to-government consultation” with all “affected Indian tribes,”
The Secretary of Agriculture must also prepare an environmental impact statement under the National Environmental Policy Act of 1969. See id.
Last, after the Department of Agriculture and Resolution Copper complete the Land Exchange, the Land Exchange Provision prohibits Resolution Copper from mining on Apache Leap and obligates Resolution Copper to surrender all rights to mine on or extract minerals from that land. See id.
C. Administrative and Procedural History
In the years since Congress passed the Land Exchange Provision, the Forest Service has engaged in a consultation process with the public and with American Indian tribes. The Forest Service held eleven public meetings and accepted public comments for 120 days. Over that period, the Forest Service received nearly 30,000 comments. Government officials also met with American Indian tribes on dozens of occasions between 2003 and 2020.
Separately, Resolution Copper has also collaborated with Apache tribe members to conduct a series of surveys that identified 6,906 “salvage locations” in Oak Flat, including 6,871 plant salvage locations, 9 animal salvage locations, and 26 mineral salvage locations. Resolution Copper has committed to removing and relocating the relevant articles from the salvage locations and preserving them at another location. Still, these consultation processes and mitigation measures were not enough to reach a solution that satisfied all parties. This lawsuit stands as evidence of this lack of success.
After these consultations, the Forest Service was scheduled to publish its final environmental impact statement on January 15, 2021. But several days before that scheduled publication date, Apache Stronghold filed this lawsuit, alleging that the Land Exchange violates RFRA, the Free Exercise Clause, and certain trust duties that Apache Stronghold argues were created by the 1852 Treaty of Santa Fe between the U.S. government and the Apache.4 Two days after that, Apache Stronghold filed a motion for a temporary restraining order and for a preliminary injunction to prevent the Land Exchange. The district court denied the temporary restraining order, reasoning that Apache Stronghold “could not show immediate and irreparable injury,” and ordered Apache Stronghold‘s motion for a preliminary injunction to be fully briefed. Apache Stronghold, 519 F. Supp. 3d at 597.
After a full round of briefing on Apache Stronghold‘s motion for a preliminary injunction, the district court held a three-hour hearing, accepted documentary evidence, and heard testimony from witnesses on Apache Stronghold‘s behalf. After considering the evidence and the parties’ arguments, the district court denied Apache Stronghold‘s motion. See id. at 611. As relevant here, the district court found that
Separate from this litigation, the Forest Service had issued its environmental impact statement on-time in January 2021. But in March 2021, soon after Apache Stronghold filed its motion for an injunction pending appeal, the Department of Agriculture ordered the Forest Service to rescind the environmental impact statement. The Department of Agriculture explained that the government needed “additional time” to “understand concerns raised by Tribes and the public” and to “ensure the agency‘s compliance with federal law.” The Forest Service “cannot give a precise length of time for completing the reinitiation of consultation” but estimates that the process will take “several months.”
Returning to this litigation, a Ninth Circuit motions panel heard another full round of briefing, including additional documentary evidence, as to Apache Stronghold‘s motion for an injunction pending appeal. The motions panel eventually denied that motion, concluding that Apache Stronghold had again failed to show that it needed immediate relief to “avoid irreparable harm,” in large part because the Forest Service expected to take “months” to complete its revised environmental review. In dissent, Judge Bumatay disagreed and would have granted Apache Stronghold an injunction pending our resolution of this appeal. Apache Stronghold‘s appeal then reached this panel for a decision on the appeal‘s merits.
Besides this case, there are two other pending cases brought by other plaintiffs who hope to prevent the Land Exchange. Both of these cases were stayed by agreement of the parties after the Forest Service withdrew its original environmental impact statement. See Ariz. Mining Reform Coal. v. U.S. Forest Serv., No. 21-00122 (D. Ariz. Mar. 15, 2021) (order granting, in light of the parties’ joint status report, a stay “pending the Forest Service‘s publication of a future Final Environmental Impact Statement... for the Resolution Copper Project and Land Exchange); San Carlos Apache Tribe v. U.S. Forest Serv., No. 21-00068 (D. Ariz. Mar. 15, 2021) (order granting the parties’ “Joint Motion to Stay Proceedings“).
II. Standard of Review
We review for an abuse of discretion a district court‘s decision to deny a preliminary injunction but review de novo any questions of law underlying that decision. See Env‘t Prot. Info. Ctr. v. Carlson, 968 F.3d 985, 989 (9th Cir. 2020).
III. Discussion
A party seeking a preliminary injunction must show that: 1) it is “likely to succeed on the merits“; 2) it is “likely to suffer irreparable harm in the absence of preliminary relief“; 3) “the balance of equities tips in [its] favor“; and 4) “an injunction is in the public interest.”5 Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008).
In the district court, Apache Stronghold sought a preliminary injunction on the grounds that the Land Exchange violates RFRA, the Free Exercise Clause, and the trust obligations that Apache Stronghold claims were created by the Treaty of Santa
A. Apache Stronghold‘s RFRA Claim
We first address Apache Stronghold‘s claim under the Religious Freedom Restoration Act (again, “RFRA“). Under RFRA, the federal government may not “substantially burden” a person‘s sincere exercise of religion unless that burden is both “in furtherance of a compelling governmental interest” and is “the least restrictive means of furthering that . . . interest.”
Apache Stronghold primarily argues that the Land Exchange – by enabling Resolution Copper to mine on Oak Flat if the company so decides – will render Apache religious exercise on Oak Flat “impossible” and thus substantially burden the religious exercise of Apache Stronghold‘s Apache members.6 Though that argument is where Apache Stronghold focuses its efforts, Apache Stronghold also contends that the Land Exchange substantially burdens its members in another way: by depriving its members of the “government benefit” of their present right to access the government-owned land of Oak Flat and by subjecting its members to the potential penalty of a trespass lawsuit for entering Oak Flat once it becomes the private property of Resolution Copper.
The government, for its part, concedes that Apache Stronghold‘s members seek to exercise sincere religious beliefs by holding ceremonies on Oak Flat, see Apache Stronghold, 519 F. Supp. 3d at 604, and wisely so. The government‘s only response to Apache Stronghold‘s RFRA claim, at least at this stage of the litigation, is to argue the Land Exchange would not “substantially burden” Apache Stronghold under RFRA.
We proceed as follows. First, we summarize the binding Ninth Circuit case law that defines the term “substantially burden” as used in RFRA. Second, we apply that settled understanding of the term to the facts of the Land Exchange and determine whether the Exchange will substantially burden Apache Stronghold under Apache Stronghold‘s primary RFRA argument. And third, we discuss Apache
1. The Definition of a “Substantial Burden”
The parties contest what constitutes a substantial burden under RFRA but fortunately, we do not write on a clean slate. This Court previously addressed this same question in Navajo Nation v. United States Forest Service, 535 F.3d 1058 (9th Cir. 2008) (en banc). In Navajo Nation, our en banc court faced facts that mirror those here. Plaintiffs in Navajo Nation, several American Indian tribes and individuals, sued the U.S. Forest Service to enjoin the Service from allowing a ski resort operating on government land to use recycled wastewater to make artificial snow for skiing. See id. at 1063. Like Oak Flat, the site of the ski resort (a mountain named Humphrey‘s Peak) was “sacred in [the American Indians‘] religion” and was a site for religious ceremonies. Id. Like the Land Exchange, the Forest Service‘s plan in Navajo Nation to permit the ski resort to use recycled wastewater on Humphrey‘s Peak would indisputably “spiritually contaminate” a sacred area and inhibit religious ceremonies. Id. And like Apache Stronghold, the Navajo Nation plaintiffs claimed that the challenged government action would violate RFRA. See id.
Just as the facts in Navajo Nation parallel the facts here, so do the legal issues. On appeal in Navajo Nation was whether Forest Service‘s proposed plan would create a “substantial burden” under RFRA. Id. at 1067.
To determine the definition of a “substantial burden” under RFRA, Navajo Nation turned to RFRA‘s text. RFRA‘s stated statutory purpose is to “restore the compelling interest test as set forth in [two landmark Free Exercise Clause cases,] Sherbert v. Verner, 374 U.S. 398 (1963) and Wisconsin v. Yoder, 406 U.S. 205 (1972) and to guarantee [that test‘s] application in all cases where free exercise of religion is substantially burdened.”
But Sherbert and Yoder did not only “set forth the compelling interest test.” Navajo Nation, 535 F.3d at 1069. These two cases also “define[d] what kind or level of burden on the exercise of religion is sufficient to invoke” that test – in other words, what burden counts as a “substantial burden.” Id. So, because RFRA expressly “restore[d]” Sherbert and Yoder‘s compelling interest test,
Accordingly, to define a “substantial burden” under RFRA, Navajo Nation looked to the type of burden on religion that was imposed in Sherbert and in Yoder. In Sherbert, the Supreme Court held that denying government benefits on account of religion imposes a substantial burden on religion. See 374 U.S. at 410. South Carolina thus violated the Free Exercise Clause by withholding unemployment benefits from a worker who was fired because she refused to work on her faith‘s day of rest. See id. In Yoder, the Supreme Court held that imposing a government
A second textual clue also supports our holding in Navajo Nation. RFRA explicitly defined numerous terms but not the phrase “substantial burden.” See
How did the Supreme Court define this “substantial burden” term of art? By reference to the Sherbert/Yoder
framework. In Jimmy Swaggart Ministries v. Bd. of Equalization of California, 493 U.S. 378, 391-92 (1990) (quoting Hobbie v. Unemployment Appeals Comm‘n of Florida, 480 U.S. 136, 141 (1987))), for instance, the Supreme Court held that a generally applicable tax “impose[d] no constitutionally significant burden on [the] appellant‘s religious practices or beliefs” because “in no sense has the State ‘conditioned receipt of an important benefit upon conduct proscribed by a religious faith, or . . . denied such a benefit because of conduct mandated by religious belief.‘” Other Free Exercise cases echoed this understanding of when the Free Exercise Clause applies-in other words, this understanding of when the government has created a substantial burden.7 See, e.g., Thomas v. Rev. Bd. of Indiana Emp. Sec. Div., 450 U.S. 707, 717-18 (1981) (“Where the state conditions receipt of an important benefit upon conduct proscribed by a religious faith, or where it denies such a benefit because of conduct mandated by religious belief . . . a burden upon religion exists. While the compulsion may be indirect, the infringement upon free exercise is nonetheless substantial.“). With this background in mind, Navajo Nation‘s conclusion about the meaning of “substantial burden” is even stronger.
burden” in the Free Exercise context-a burden within the Sherbert/Yoder framework-governs that same phrase‘s meaning under RFRA.
Navajo Nation drew further support from Lyng v. Northwest Indian Cemetery Protective Association, 485 U.S. 439 (1988) and Bowen v. Roy, 476 U.S. 693 (1986), two Free Exercise Clause cases that asked and answered essentially the same question that we ask here: what constitutes a “substantial burden” on religion? In Lyng, a group of American Indians challenged a federal plan to build a road over, and permit logging on, land those American Indians held sacred. See id. at 441-42. The challengers claimed that the planned construction would “physically destroy the environmental conditions and the privacy without which the [American Indian] religious practices [could not] be conducted.” Id. at 449. In Bowen, the petitioners challenged a federal statute that required state agencies to use social security numbers to identify welfare benefit recipients; according to the challengers’ American Indian religion, assigning a numerical identifier to their daughter would rob her of “spiritual power.” See id. at 695-96. But the petitioners in neither Lyng nor Bowen had stated a valid Free Exercise claim because in “neither case . . . would the affected individuals be coerced by the Government‘s action into violating their religious beliefs; nor would either governmental action penalize religious activity by denying any person an equal share of the rights, benefits, and privileges enjoyed by other citizens.” Lyng, 485 U.S. at 449; see Bowen, 476 U.S. at 700. This was true, the Supreme Court held, even if the government‘s action in Lyng would “virtually destroy the . . . Indians’ ability to practice their religion.” Id. at 451. Lyng and Bowen thus confirmed that a “substantial burden” in the Free Exercise context consists only of those government actions that fall within the Sherbert/Yoder framework-actions that impose a penalty or deny a benefit-no matter how otherwise burdensome the government might be.8 The same must be true for substantial burdens under RFRA, given that ”Sherbert, Yoder, and federal court rulings prior to Smith“-that is, rulings like Lyng and Bowen-“control [RFRA‘s] ‘substantial burden’ inquiry.” Navajo Nation, 535 F.3d at 1069.
To summarize, Navajo Nation held that a “substantial burden” under RFRA consists only of burdens within the Sherbert/Yoder framework for three reasons. First, RFRA by its text “restored” Sherbert, Yoder, their “compelling interest” test, and their “substantial burden” inquiry, thus defining a “substantial burden”
Applying Navajo Nation‘s “substantial burden” standard to that case‘s facts, we held that under RFRA the Navajo suffered no “substantial burden” on their religion and thus had no RFRA claim against the Forest Service. See id. at 1070. To the Navajo, the Forest Service‘s decision to permit wastewater on Humphrey‘s Peak would “spiritually desecrate a sacred mountain.” Id. But that government decision lay outside the Sherbert/Yoder framework to which RFRA applies. The Forest Service did not “coerce the [Navajo] to act contrary to their religion” by imposing a penalty or denying a governmentally granted benefit when it authorized the ski resort to use wastewater on the peaks. Id. The Service thus imposed no substantial burden under RFRA. See id. This was so, we held, “[e]ven were we to assume . . . that the government action in this case will ‘virtually destroy the . . . [Navajo‘s] ability to practice their religion.‘” Id. at 1072 (quoting Lyng, 485 U.S. at 451). Where there is no substantial burden, there is no ground to apply the “compelling interest” test, and thus no RFRA violation-no matter how dire the practical consequences of a government policy or decision. Any other result would be inconsistent with RFRA‘s text and with the Supreme Court‘s understanding of what constitutes a substantial burden on religious exercise.
While Navajo Nation‘s “substantial burden” holding has firm doctrinal roots, we noted further that our holding there also has a strong practical basis. Were the scope of a substantial burden under RFRA broader than the Sherbert/Yoder framework, “any action the federal government were to take, including action on its own land, would be subject to the personalized oversight of millions of citizens.” Id. at 1063. And in the specific factual context of Navajo Nation-federal land use decisions-“giving one religious sect a veto over the use of public park land would deprive others of the right to use what is, by definition, land that belongs to everyone.” Id. at 1063-64.
2. Apache Stronghold‘s Primary RFRA Argument
With this background in mind, we turn to Apache Stronghold‘s arguments. Apache Stronghold‘s main argument is that the Land Exchange would hand Oak Flat over to Resolution Copper for the latter‘s mining plan, thus incidentally making it “impossible” for Apache Stronghold‘s members to worship on Oak Flat and thereby substantially burdening them. Even assuming that the Land Exchange would in fact make Apache Stronghold‘s members worship “impossible,” this argument cannot succeed in light of Navajo Nation.
The Land Exchange‘s effect on Apache Stronghold‘s members falls outside of the Sherbert/Yoder framework and thus outside of RFRA‘s definition of a substantial burden. Under RFRA, the government imposes a substantial burden on religion in two-and only two-circumstances: when
This is true even if the Land Exchange makes worship on Oak Flat “impossible.” The government makes exercises of religion more difficult all the time. Doing so is not inherently coercive. As one example, the United States has a special visa program for “[m]inisters of [r]eligion.” See Visas for Immigrant Religious Workers, U.S. Dep‘t of State, https://travel.state.gov/content/travel/en/us-visas/immigrate/visa-religious-workers.html (last visited June 15, 2022). When the government denies one of these visas, the government no doubt makes it more difficult for that minister‘s following to exercise their faith. But the visa denial does not coerce those followers by threatening them with a negative outcome (i.e., a penalty or the denial of a governmental benefit) if they continue to worship despite that hardship. So too here: the Land Exchange does not coerce the Apache to abandon their religion by threatening them with a negative outcome. Accordingly, Apache Stronghold‘s members have not established that they would suffer a substantial burden under RFRA. Apache Stronghold is not likely to succeed on its RFRA claim.
Between them, Apache Stronghold and the dissent offer three arguments in response. First, the dissent argues that Navajo Nation misread RFRA and should have held that the definition of a “substantial burden” under RFRA extends beyond the Sherbert/Yoder framework. Second, both the dissent and Apache Stronghold contend that Navajo Nation contains exceptions that permit the panel to find a substantial burden here. And third, the dissent would hold that intervening Supreme Court precedent since Navajo Nation is “clearly irreconcilable” with Navajo Nation, permitting the panel to disregard Navajo Nation in its entirety. Miller v. Gammie, 335 F.3d 889, 893 (9th Cir. 2003) (en banc). None of these responses persuades us.
The dissent first argues that Navajo Nation misread RFRA in concluding that RFRA defines a “substantial burden” as those burdens falling within the Sherbert/Yoder framework. As an initial matter, our en banc decision in Navajo Nation binds this panel-we cannot overrule Navajo Nation even if we do not agree with it. See Robbins v. Carey, 481 F.3d 1143, 1149 n.3 (9th Cir. 2007). But even considering the points that the dissent raises as grounds for overruling Navajo Nation, we find them unconvincing.
At the outset, the dissent contends that RFRA was not “concern[ed]” with defining a “substantial burden” but instead with “ensuring that the compelling interest standard would be applied once a substantial
We do not agree. The two cases that RFRA explicitly cited and “restored“-Sherbert and Yoder-both defined the “compelling interest” test and set out the two burdens that satisfy the predicate “substantial burden” inquiry: a penalty imposed and a governmental benefit denied. Navajo Nation, 535 F.3d at 1069. Moreover, the phrase “substantial burden” was not defined in RFRA‘s text but was a term of art in Free Exercise Clause doctrine that referred to those same two burdens set out in Sherbert and Yoder. See id. at 1074. With this background in mind, the best reading of RFRA‘s text is that RFRA “restore[d]” both Sherbert and Yoder‘s “compelling interest” test and their “substantial burden” inquiry.
The dissent also argues that Navajo Nation‘s “substantial burden” definition “lacks a basis in pre-Smith precedent.” Dissent at 64. Not so. The dissent has identified some cases where courts may have suggested that Free Exercise Clause violations could fall outside of the Sherbert/Yoder “substantial burden” framework. But the two cases that RFRA specifically “restore[d]” and cited in its very text were indeed Sherbert and Yoder.
Further, and as noted above, the Supreme Court‘s post-Lyng but pre-Smith Free Exercise doctrine reinforces Navajo Nation‘s understanding of the scope of a “substantial burden.” Pre-Smith, the Free Exercise Clause applied only when the government “placed a substantial burden” on religious exercise. Hernandez, 490 U.S. at 699. And a “substantial burden” referred only to burdens within the Sherbert/Yoder framework. See Lyng, 485 U.S. at 449; Jimmy Swaggart Ministries, 493 U.S. at 391-92.
With the above in mind, we also reject the dissent‘s suggestion that Navajo Nation “constricted” the definition of a “substantial burden” relative to pre-Smith Free Exercise Clause doctrine. Dissent at 67. As just shown, and setting aside the potential outliers that the dissent identified, pre-Smith Free Exercise Clause doctrine already defined a “substantial burden” as only those burdens that fall within the Sherbert/Yoder framework: coercion caused by the government either imposing a penalty or denying a benefit. See Lyng, 485 U.S. at 449; Jimmy Swaggart Ministries, 493 U.S. at 391-92. So, when Navajo Nation recognized that this same framework also defines the scope of a “substantial burden” under RFRA, Navajo Nation did not narrow or constrict the definition of a “substantial burden.” Rather, Navajo Nation stayed faithful to a substantial burden‘s already settled scope.
The dissent also points to the Religious Land Use and Institutionalized Persons Act (“RLUIPA“),
Yet we disagree with the dissent here too: RLUIPA‘s definition of a “substantial burden” casts no doubt on how Navajo Nation defined that term as to RFRA. We have previously interpreted a “substantial burden” under RLUIPA to be defined not by the Sherbert/Yoder framework but by the “plain meaning” of the phrase “substantial burden.” San Jose Christian Coll. v. City of Morgan Hill, 360 F.3d 1024, 1034 (9th Cir. 2004). But unlike RFRA, RLUIPA‘s text does not even mention, much less cite, either Sherbert or Yoder. Compare
The dissent also equates the two contexts covered by RLUIPA-prisons and local land regulation-to situations involving “Native American sacred sites located on government land.” Dissent at 62. In all three contexts, the dissent contends, the government substantially burdens religion by “denying access” to “religious locations and resources.” Id. at 63. But while RLUIPA covers the first two contexts (again, prisons and local land regulation), the third context-the context actually at issue here-falls to RFRA. Compare
Second, Apache Stronghold and the dissent both argue that even under Navajo Nation, the Land Exchange may substantially burden religious exercise. Both reach this conclusion two ways. Neither approach persuades us.
They first seize onto a statement from Navajo Nation that any “burden imposed on the exercise of religion short of that described by Sherbert and Yoder is not a ‘substantial burden’ within the meaning of RFRA,” 535 F.3d at 1070 (emphasis added), and argue that the Land Exchange constitutes a substantial burden because it imposes a “greater burden on religious exercise” than that imposed in Yoder or
Properly understood, Navajo Nation did not set out a quantitative floor for a “substantial burden” such that all “greater” burdens qualify. Rather, Navajo Nation singled out two specific qualitative burdens-denying a benefit or imposing a penalty-that together form the complete universe of “substantial burdens” under RFRA. For evidence, look no farther than the sentence immediately before the “short of” phrase, which reads: “Under RFRA, a ‘substantial burden’ is imposed only when individuals are forced to choose between following the tenets of their religion and receiving a governmental benefit (Sherbert) or coerced to act contrary to their religious beliefs by the threat of civil or criminal sanctions (Yoder).” Navajo Nation, 535 F.3d at 1069-70 (emphasis added). Further proving this point, immediately after the “short of” phrase Navajo Nation applies the test that it announced in the preceding sentences: “[T]here is no ‘substantial burden’ on the Plaintiffs’ exercise of religion in this case. The [challenged government action] does not force the Plaintiffs to choose between following the tenets of their religion and receiving a governmental benefit, as in Sherbert. The [challenged action] also does not coerce the Plaintiffs to act contrary to their religion under the threat of civil or criminal sanctions, as in Yoder.” Id. at 1070. Navajo Nation did not further ask if the Forest Service had imposed a burden greater than that imposed in Sherbert or Yoder, reinforcing that such a step is not necessary. Other passages in Navajo Nation similarly belie the dissent and Apache Stronghold‘s reading of the case.9 Accurately read, Navajo Nation recognized that the government imposes a substantial burden under RFRA only when the government denies the delivery of a benefit (as in Sherbert) or imposes a penalty (as in Yoder). The “short of” language did not change the character or type of government action that is required to constitute a “substantial burden” under RFRA.
Apache Stronghold and the dissent contend also that both Navajo Nation and Lyng are limited to cases where the government action would interfere with “subjective spiritual experience,” not cases where the government “objectively and severely interfere[s] with a plaintiff‘s access to religious locations or resources.” Dissent at 72 (quoting Navajo Nation, 535 F.3d at 1063). (Apache Stronghold‘s
formulation of the same idea is that Navajo Nation and Lyng do not apply to cases involving a “physical impact” on land.) Because Resolution Copper‘s mining plan would have such an “objective” or “physical” impact here, they argue that Navajo Nation and Lyng do not control. True enough, in dicta, Navajo Nation pointed out that the challenged government action would not make any “places of worship . . . inaccessible” or physically affect any “religious ceremonies.” 535 F.3d at 1063. Similarly, dicta in Lyng states that “[n]o sites
Neither Navajo Nation nor Lyng turned on whether the challenged government action “objectively” interfered with religious exercise or “physically” affected sacred land. The rule that Navajo Nation drew from RFRA‘s text and from ”Sherbert, Yoder, and federal court rulings prior to Smith” was clear: “Under RFRA, a ‘substantial burden’ is imposed only when individuals are forced to choose between following the tenets of their religion and receiving a governmental benefit (Sherbert) or coerced to act contrary to their religious beliefs by the threat of civil or criminal sanctions (Yoder).” Id. at 1069-70 (emphasis added). This rule contains no exception for when the government neither imposes a penalty nor denies a benefit but “objectively” or “physically” interferes with religious exercise.
A close examination of the claimed burden on religion in Lyng further refutes the dissent and Apache Stronghold‘s argument. It was true that “[n]o sites where specific rituals take place were to be disturbed.” Lyng, 485 U.S. at 454. But those opposed to the government action argued that “the proposed road w[ould] ‘physically destroy the environmental conditions and the privacy without which the [American Indian] religious practices [could not] be conducted.‘” Id. at 449. And even so-despite this “objective,” “physical” impact that could “virtually destroy” the American Indians’ “ability to practice their religion,” the Supreme Court found no substantial burden there.10 See id. In sum, we cannot differentiate between physical and intangible damage to religious sites as Apache Stronghold asks because the Sherbert/Yoder framework turns on the nature of government action, not on the severity of the government‘s encroachment on a religious site. See Lyng, 485 U.S. at 451 (noting that the substantial burden inquiry “cannot depend on measuring the effects of a governmental action on religious exercise“); Navajo Nation, 535 F.3d at 1070 n.12 (“[I]n Yoder, it was not the effect . . . on the children‘s subjective religious sensibilities that constituted the undue burden on the free exercise of religion. Rather, the undue burden was the penalty of criminal sanctions on the parents.“); accord Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 723-24 (2014) (noting that courts have “have no business addressing” whether the RFRA substantial burden analysis changes if a religious adherent would only be forced
to participate in a religiously prohibited act in an “attenuated” way).
If any doubts about Navajo Nation‘s meaning survive the arguments above, the many Ninth Circuit cases that have applied Navajo Nation put those doubts to
As the dissent notes, none of these post-Navajo Nation cases addressed the precise facts at issue here. Dissent at 72 n.4. None need have. RFRA defined a “substantial burden” according to the Sherbert/Yoder framework. See Navajo Nation, 535 F.3d at 1069-70. This is an across-the-board definition that applies in all cases under the statute, not a “restricted railroad ticket, good for th[at] day and train only.” Smith v. Allwright, 321 U.S. 649, 669 (1944) (Roberts, J., dissenting). And dispositive here, this definition contains no exceptions for burdens on religion thought to be quantitatively “greater” than the burdens in Sherbert and Yoder or for burdens that neither impose a penalty nor deny a benefit but “objectively” or “physically” interfere with religious exercise in an incidental way.
Apache Stronghold (but not the dissent) also points to a scattered set of cases that apply a definition of “substantial burden” in a manner broader than the Sherbert/Yoder framework.12 But for a variety of reasons,
affect our interpretation of Navajo Nation. As an initial matter, even were courts from other circuits to take approaches different than ours in Navajo Nation, Navajo Nation binds this panel and this Circuit.13 But turning to the substance of the in-circuit cases that Apache Stronghold cites, they either interpret RFRA but predate Navajo Nation14 or interpret not RFRA but RLUIPA instead.15 To the extent our pre-Navajo Nation RFRA cases defined a “substantial burden” differently than did Navajo Nation, our later en banc decision in Navajo Nation controls. See Robbins, 481 F.3d at 1149 n.3. And the RLUIPA cases are similarly unpersuasive. As we have explained, we have interpreted RFRA and RLUIPA to apply different substantial burden standards. Compare Navajo Nation, 535 F.3d at 1069-70 (“Under RFRA, a ‘substantial burden’ is imposed only when individuals are forced to choose between following the tenets of their religion and receiving a governmental benefit (Sherbert) or coerced to act contrary to their religious beliefs by the threat of civil or criminal sanctions (Yoder).“), with San Jose Christian Coll., 360 F.3d at 1035 (holding that under RLUIPA, the government imposes a “substantial burden” on religion when it “imposes a ‘significantly great’ restriction or onus” on religious exercise). Apache Stronghold‘s RLUIPA cases thus give us no guidance for how to interpret the phrase “substantial burden” under RFRA.16
Last, the dissent argues that Navajo Nation is “clearly irreconcilable” with recent Supreme Court precedent, allowing the panel to ignore Navajo Nation entirely. Dissent at 74 (quoting Miller, 335 F.3d at 900). Miller does permit Ninth Circuit panels to treat as “effectively overruled” any Ninth Circuit cases that are “clearly irreconcilable” with “intervening Supreme Court authority.” 335 F.3d at 900. But the “‘clearly irreconcilable’ requirement ‘is a high standard.‘” Fed. Trade Comm‘n v. Consumer Def., LLC, 926 F.3d 1208, 1213 (9th Cir. 2019) (quoting Rodriguez v. AT & T Mobility Servs. LLC, 728 F.3d 975, 979 (9th Cir. 2013)). If, as a panel, “we can apply our precedent consistently with that of the higher authority, we must do so.” Consumer Def., 926 F.3d at 1213.
In our view, Navajo Nation is fully reconcilable with the Supreme Court‘s recent cases. The dissent highlights Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682 (2014), Holt v. Hobbs, 574 U.S. 352 (2015), and Ramirez v. Collier, 142 S. Ct. 1264 (2022). To this list we add Tanzin v. Tanvir, 141 S. Ct. 486 (2020), a case that Apache Stronghold cites, and Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012 (2017). When we compare these cases to Navajo Nation, we do not see any clear irreconcilability.
Turning first to Hobby Lobby, that case does not contradict Navajo Nation‘s “substantial burden” holding. Hobby Lobby held that closely held corporations can maintain a RFRA claim but it provided no comprehensive definition of “substantial burden.” See 573 U.S. at 719. In fact, Hobby Lobby framed a substantial burden in precisely the way Navajo Nation did: Hobby Lobby suffered a substantial burden because it would have had to “pay an enormous sum of money” to the government-a government penalty-“if [it] insist[ed] on providing insurance coverage in accordance with their religious beliefs.” Id. at 726.
As the dissent rightly notes, Hobby Lobby made clear that RFRA claims need not perfectly track pre-Smith Free Exercise doctrine in every single way. RFRA plaintiffs are not limited to those who “fell within a category of plaintiffs [who] had brought a free-exercise claim that [the Supreme] Court entertained in the years before Smith” because RFRA did not “merely restore[ the Supreme] Court‘s pre-Smith decisions in ossified form.” Id. at 715-16.
But Navajo Nation did not assume otherwise. Rather, Navajo Nation observed that RFRA, by its own terms, “restore[d]” pre-Smith Free Exercise doctrine in a single, limited way: it incorporated Sherbert and Yoder‘s “compelling interest test” and predicate “substantial burden” inquiry.
Next is Holt. There, the Supreme Court stated that RLUIPA “allows prisoners ‘to seek religious accommodations pursuant to the same standard as set forth in RFRA.‘” 574 U.S. at 358 (quoting Gonzales v. O Centro Espírita Beneficente União do Vegetal, 546 U.S. 418, 436 (2006)). From this connection, the dissent argues that RFRA, like RLUIPA, recognizes a “substantial burden” “when the government denies access to religious locations or resources.” Dissent at 64. But we do not read Holt‘s dicta to support the dissent‘s position. This quotation from Holt is best read as applying to the “compelling interest” test-that is, the stage of the RFRA (and RLUIPA) analysis at which individuals “seek religious accommodations” and have those accommodations assessed against the government‘s justification-not as applying to the predicate “substantial burden” stage. The dissent seems to recognize this nuance as well, observing that “RLUIPA sets forth the ‘same standard’ for evaluating governmental justifications for imposing substantial burdens on religion as RFRA-strict scrutiny.” Dissent at 68-69.
Further, the actual “substantial burden” standard that Holt applied matches the Sherbert/Yoder framework almost perfectly. Holt challenged a prison grooming policy that required him to “shave his beard and thus to ‘engage in conduct that seriously violates his religious beliefs.‘” Holt, 574 U.S. at 361 (quoting Hobby Lobby, 573 U.S. at 720). If Holt violated that policy, he would “face serious disciplinary action” and the Supreme Court reasoned that “[b]ecause the grooming policy puts [Holt] to this
individuals are “coerced to act contrary to their religious beliefs by the threat of civil or criminal sanctions.” Navajo Nation, 535 F.3d at 1070. The government action in Holt—requiring a prisoner to violate his religious beliefs or “face serious disciplinary action,” id. at 361—falls squarely within that framework. So here, too, “we can apply our precedent consistently with that of the higher authority.” Consumer Def., 926 F.3d at 1213.
For similar reasons, we dismiss the dissent‘s appeal to Ramirez. First, Ramirez was a
Finally, Apache Stronghold points to Tanzin v. Tanvir, 141 S. Ct. 486 (2020), in which the Supreme Court held that
We also reject the idea that Tanzin implied any substantial burden holding through its choice of lower-court cases to cite. Tanzin included a “See, e.g.,” citation to DeMarco v. Davis, 914 F.3d 383 (5th Cir. 2019), a Free Exercise Clause case involving a prison officials’ destruction of a prisoner‘s personal property—his legal and religious books.18 See id. at 389-90. From that citation, Apache Stronghold divines the principle that the government can violate
We also add an overarching consideration that further supports our conclusion that Navajo Nation and the Supreme Court‘s decisions cited by the dissent can be reconciled. We must read Hobby Lobby, Holt, Ramirez, and Tanzin in conjunction with the Supreme Court‘s other precedents. And the Supreme Court reaffirmed as recently as 2017 that a “substantial burden” on religion is still defined by the Sherbert/Yoder framework recognized in Navajo Nation. In Trinity Lutheran Church of Columbia, Inc. v. Comer, the Supreme Court quoted Lyng‘s “substantial burden” rule: even actions that “would interfere significantly with private persons’ ability to pursue spiritual fulfillment according to their own religious beliefs” pose “no free exercise violation ... [if] the affected individuals were not being ‘coerced by the Government‘s action into violating their religious beliefs.‘” 137 S. Ct. 2012, 2020 (2017) (quoting Lyng, 485 U.S. at 449). That reasoning matches ours here perfectly. So when the dissent‘s cases and Trinity Lutheran are taken together, as they must be, they cast no doubt on the scope of the Sherbert/Yoder framework or on Navajo Nation‘s “substantial burden” holding.19 Given that we decline to apply our past precedents only when more recent Supreme Court decisions are “clearly irreconcilable” with those precedents, Miller, 335 F.3d at 893, we must apply Navajo Nation here and we do so without hesitation.
We thus conclude that under Navajo Nation, the Land Exchange does not substantially burden Apache Stronghold within the meaning of
As we observed in Navajo Nation, were the definition of “substantial burden” under
The dissent is surely right that some government action swept into
Apache Stronghold‘s broader definition of “substantial burden” would also create another, deeper problem: It would force judges to make decisions for which we are fundamentally unsuited. The dissenters in Navajo Nation were correct on one important point: “[R]eligious exercise invariably, and centrally, involves a ‘subjective spiritual experience.‘” 535 F.3d at 1096 (Fletcher, J., dissenting); see also id. at 1070 n.12 (majority opinion) (agreeing with the dissent on this point).
Who are we to say whether government action has an “objective” impact on religious observance or merely “diminishes [a worshipper‘s] subjective spiritual fulfillment“? Id. Questions like this raise issues on which judges must not pass. As we are often reminded, it is outside the “judicial ken to question the centrality of particular beliefs or practices to a faith.” Hernandez, 490 U.S. at 699. The straightforward Sherbert/Yoder framework avoids these problems.
Of course, the U.S. government may propose future projects that, like the Land Exchange here, would impose no substantial burden but still have an incidental impact on religious observance or fulfillment. And someone must decide whether the government should ultimately pursue each such project. But
3. Apache Stronghold‘s Secondary RFRA Argument
Apache Stronghold‘s secondary argument is that the Land Exchange does in fact deprive its members of a benefit and subject its members to a penalty. Apache Stronghold contends that the Exchange deprives its members of “the use and enjoyment of ‘government’ land for religious exercise” and subjects them to penalties for “trespassing on now ‘private’ land.” We disagree.
Turning first to Apache Stronghold‘s argument that the Land Exchange denies its members a benefit, that argument has a problem. The government does not substantially burden religion every time it ends a “governmental benefit” that at one time went to religious beneficiaries. There must be an element of coercion: the government must “condition” the benefit upon conduct that would violate sincerely held religious beliefs. Navajo Nation, 535 F.3d at 1067. Consider this example. Suppose that for many years, the Forest Service has paid Apache Stronghold‘s members to host educational sessions to teach local children about the Apache‘s history and culture, including the Apache‘s religious traditions. But this year, the Forest Service says to Apache Stronghold: “our budget‘s been cut—we can‘t renew your contract for more sessions next year.” Apache Stronghold‘s members have just been deprived of a benefit—payment for the educational sessions that they previously held—but they have not been coerced to abandon their religious beliefs. We need not apply strict scrutiny to every contract cancellation or revision.
Under this rubric, the Land Exchange thus presents no “substantial burden.” The Exchange does not “condition” any government benefits on the Apache violating their religious beliefs. Like the cancelled educational sessions in the hypothetical above, the Land Exchange does not force Apache Stronghold‘s members to choose between following their religion and losing a benefit (the “use and enjoyment” of Oak Flat). The Land Exchange just incidentally keeps everybody—Apache Stronghold‘s members included—from using Oak Flat: No conditioning of a benefit; no coercion. Were the rule otherwise, the federal government would substantially burden religion any time it cancels a contract with a religious entity or repeals a program that subsidized both parochial and secular private schools.
Next is Apache Stronghold‘s argument that the Land Exchange subjects its members to penalties: liability for trespassing on land that will be private after the Exchange. We also reject this argument.
Turning first to criminal trespass liability, when a religious plaintiff has a “sufficiently realistic fear” that the government will punish him for exercising his religious beliefs, he can sue the government under
But Apache Stronghold‘s argument faces two problems. For one, Apache
And even had Apache Stronghold shown a “sufficiently realistic fear” of criminal prosecution, it seeks relief that
Next, when we consider potential civil trespass suits brought by Resolution Copper, we again see two problems with Apache Stronghold‘s argument.20 The first problem is factual. At this early stage in the litigation, it is not clear whether the Apache will in fact be subject to civil trespass liability. Even after the Land Exchange, Resolution Copper “will ensure ongoing public access to the Oak Flat Campground, recreational trails and climbing,” and will “accommodate requests to periodically close the campground to the public for traditional and ceremonial purposes.” Resolution Copper also committed to “permit harvesting of the Emory oak groves by individuals, or commercially through an authorization.” And the Apache need not rely on Resolution Copper‘s goodwill alone. The Land Exchange Provision itself obligates Resolution Copper to “provide access to the surface of the Oak Flat Campground to members of the public, including Indian tribes.”
B. Apache Stronghold‘s Free Exercise Clause Claim
We next address Apache Stronghold‘s claim that the Land Exchange would violate the Constitution‘s Free Exercise Clause. See
First, the Land Exchange is “neutral” in that its “object” is not to infringe upon the Apache‘s religious practices. Id. at 533. The Land Exchange Provision never mentions religion, and when it comes closest to doing so, the Provision shows solicitude towards religion, not intent to infringe. See
Apache Stronghold disagrees, arguing that the Land Exchange “targets religious conduct for distinctive treatment.” As evidence, it posits that Congress must have known the adverse impact that the Land Exchange would have on the Apache. But even assuming that 535 distinct Congresspersons could have a single collective “knowledge” or “purpose,” Congress‘s knowledge is not enough to prove its purpose.22
The Land Exchange is also generally applicable: it does not selectively “impose burdens only on conduct motivated by religious belief.” Lukumi, 508 U.S. at 543. Rather, the Land Exchange will also burden all manner of secular activities on the areas to be transferred to Resolution Copper. After the Land Exchange, parts of the Tonto National Forest will “no longer [be] accessible to hikers, rock climbing enthusiasts, cyclists, equestrians, campers, hunters, and other recreational users.”
Apache Stronghold responds that the Land Exchange is not generally applicable because it is “designed to apply to only one piece of land,” but this argument misconstrues the legal standard. We do not ask if the law was “designed to apply to only one piece of land.” Indeed, the statute challenged in Smith—and upheld there as neutral and generally applicable—was designed to apply to only one type of conduct: the “knowing or intentional possession of a ‘controlled substance.‘” 494 U.S. at 874 (quoting
C. Apache Stronghold‘s Trust Claim
We last consider Apache Stronghold‘s trust claim. As relevant background, the Apache and the U.S. government signed the Treaty of Santa Fe in 1852. In that treaty, the U.S. promised to “designate, settle, and adjust [the Apache‘s] territorial boundaries, and pass and execute in their territory such laws as may be
The government responds that this trust claim fails for three reasons: 1) Apache Stronghold cannot bring a trust claim under the Treaty of Santa Fe because it is a non-profit group, not the Apache tribe that signed the treaty; 2) the Treaty of Santa Fe does not create an “enforceable trust duty“; and 3) the Land Exchange Provision abrogated the Treaty of Santa Fe by statute. We need address only the second reason, as it is dispositive here.23
We agree with the government that on this record, Apache Stronghold has not established that the Treaty of Santa Fe imposes on the United States an enforceable trust obligation. As a general matter, the U.S. government shoulders a trust obligation with respect to an American Indian tribe when the U.S. government “takes on or has control or supervision over tribal monies or properties.” United States v. Mitchell, 463 U.S. 206, 225 (1983) (quoting Navajo Tribe of Indians v. United States, 224 Ct. Cl. 171, 183 (Ct. Cl. 1980)). But here, the government does not control or supervise tribal properties at Oak Flat. Oak Flat belongs to the government, a fact that Apache Stronghold does not presently contest. Apache Stronghold argues that title over Oak Flat is irrelevant, as it seeks not title but “usufructuary rights to use land for traditional purposes.” But the Treaty‘s language explicitly tied any obligations that it created to the Apache‘s title to land. The government promised to “designate, settle, and adjust [the Apache‘s] territorial boundaries, and pass and execute in their territory such laws,” “their” referring to the Apache treaty signatories. Even assuming that Oak Flat was once Apache land according to historical maps, Apache Stronghold has not pointed to any evidence indicating that the government designated any boundaries of the Apache‘s territory after the 1852 Treaty, let alone boundaries that encompass Oak Flat. Because Apache Stronghold points to no evidence establishing that the U.S. government “designate[d] . . . territory” on which the government has any obligation to “pass and execute”
This conclusion accords with how both we and other courts have interpreted identical treaty language in other cases. The Treaty with the Utah, just like the Treaty of Santa Fe, required the United States to “designate, settle, and adjust [the American Indians‘] territorial boundaries, and pass and execute such laws, in their territory, as the [United States] may deem conducive to the happiness and prosperity of said [American] Indians.” Treaty with the Utah, Dec. 30, 1849, art. VII, 9 Stat. 984. But that language only “reserves for a future date the final delineation of boundaries.” Uintah Ute Indians of Utah v. United States, 28 Fed. Cl. 768, 788, 789 (Fed. Cl. 1993). The Treaty with the Utah “contains no obligations with respect to property” and created neither “a trust relationship [n]or trust protection,” at least not until “the Government established boundaries” that delineated American Indian land upon which the United States could have some obligations. Id. We agreed in Robinson v. Jewell when we held that the Treaty with the Utah did not “create[] any enforceable property rights.” 790 F.3d 910, 916 (9th Cir. 2015); see also id. at 917.
So too here. Apache Stronghold has not adduced evidence which establishes that the U.S. government implemented the Treaty of Santa Fe by designating any land or recognizing any title vested in the Apache. And without title vested in the Apache, there can be no trust relationship arising from the Treaty of Santa Fe and no trust obligations relating to “usufructuary rights.” Apache Stronghold‘s trust claim is thus unlikely to succeed.
* * *
We are a “cosmopolitan nation made up of people of almost every conceivable religious preference.” Braunfeld v. Brown, 366 U.S. 599, 606 (1961). This pluralism is a source of strength, but it places demands on us all. In some cases, the many must accommodate the needs of the few—we accept that the government must sometimes “expend additional funds to accommodate citizens’ religious beliefs.” Hobby Lobby, 573 U.S. at 730. But in other cases, our need to “maintain an organized society that guarantees religious freedom to a great variety of faiths requires that some religious practices yield to the common good.” Lee, 455 U.S. at 259. This give-and-take suits perfectly neither the religious nor the secular. The “diversity of beliefs in our pluralistic society” demands as much. Bowen, 476 U.S. at 712 (plurality opinion). Here, for the reasons given above, this case is the second of those two types.
As we reach this conclusion, we do not rejoice. Rather, we recognize the deep ties that the Apache have to Oak Flat and to the nearby Apache Leap and Devil‘s Canyon. And we acknowledge that the Land Exchange may impact the Apache‘s plans to worship on Oak Flat. But
The district court‘s denial of Apache Stronghold‘s motion for a preliminary injunction is AFFIRMED.
BERZON, Circuit Judge, dissenting:
The majority applies an overly restrictive test for identifying a “substantial burden”
First, the doctrinal argument rests on the notion that
The majority ignores the reality that pre-Smith federal cases applied a broader definition of “substantial burden,” particularly in the prisoner context. Those cases recognized that when a plaintiff depends on the government for access to religious resources, the government‘s withholding of those resources can constitute a substantial burden on religious exercise. By making religious practice impossible, instead of merely discouraging or penalizing it, such a burden can be greater than those imposed in Sherbert and Yoder.
The majority derives its definition of “substantial burden” from Navajo Nation v. U.S. Forest Service, 535 F.3d 1058 (9th Cir. 2008) (en banc). Because that case held that
Second, the “practical basis” for the majority‘s definition stems from the concern that “giving one religious sect a veto over the use of public park land would deprive others of the right to use what is, by definition, land that belongs to everyone.” Majority Op. 24-25 (quoting Navajo Nation, 535 F.3d at 1063-64). But redefining “substantial burden” to exclude great burdens on religious exercise because accommodating a religious practice could interfere with other uses of federal land is a disingenuous means of reconciling those competing claims. Instead of denying the burden exists, the appropriate way to address the conflicting interests is at the justification stage. If accommodating the religious practice would cause other societal harms, then the government may well be able to show that applying the burden is the “least restrictive means of furthering [a] compelling governmental interest.”
I.
I begin with the majority‘s principal doctrinal argument—that
imposed by laws neutral toward religion“; and that “the compelling interest test as set forth in prior Federal court rulings is a workable test for striking sensible balances between religious liberty and competing prior governmental interests.”
The majority relies on Navajo Nation for the conclusion that ”Sherbert and Yoder must also control [RFRA‘s] ‘substantial burden’ inquiry.” Majority Op. 19 (alteration in original) (quoting Navajo Nation, 535 F.3d at 1069). As explained in more detail below, I do not read Navajo Nation as so instructing. And the idea that RFRA—a statute intended to restore religious freedom—silently limited the concept of “substantial burden” to the two types of burdens found in Sherbert and Yoder requires an inferential leap justified neither by logic nor by the pre-Smith federal case law.
Sherbert and Yoder both addressed situations occurring in private life—that is, life outside an institutional setting such as a prison. In private life, “government inhibitions on voluntary religious practice are the exception rather than the norm.” Stephanie Hall Barclay & Michalyn Steele, Rethinking Protections for Indigenous Sacred Sites, 134 Harv. L. Rev. 1294, 1301 (2021). Two common tools the government uses to influence behavior “in contexts in which voluntary choice is the baseline” are so-called “carrots and sticks.” Id. at 1326. The government offers carrots, or government benefits, to induce desired behavior, and uses sticks, or penalties, to deter undesired behavior. As Sherbert and Yoder recognized, the government substantially burdens religious exercise when it denies carrots, or threatens sticks, based on a person‘s religious activity. Or, as the majority puts it: “the government imposes a substantial burden on religion... ‘when individuals are forced to choose between
But some Americans seek to practice their religion in contexts in which voluntary choice is not the baseline. In these contexts, the government controls access to religious locations and resources. See Barclay & Steele, supra, at 1301. Three main examples of these contexts are prisons, Native American sacred sites located on government land, and zoning.
Prisoners “are unable freely to attend to their religious needs and are therefore dependent on the government‘s permission and accommodation for exercise of their religion.” Cutter v. Wilkinson, 544 U.S. 709, 720–21 (2005). Prisons may allow or prevent access to resources such as prison chapels or religious texts. Many traditional Native American religious sites are located on federal land. The government controls access to and other aspects of these sites, leaving Native Americans “at the mercy of government permission to access sacred sites.” Barclay & Steele, supra, at 1301. And through zoning decisions, local governments can limit religious groups’ ability to “build, buy, or rent” “a place of worship... adequate to their needs and consistent with their theological requirements,” which is “at the very core of the free exercise of religion.” Int‘l Church of Foursquare Gospel v. City of San Leandro, 673 F.3d 1059, 1069 (9th Cir. 2011) (quoting Vietnamese Buddhism Study Temple in Am. v. City of Garden Grove, 460 F. Supp. 2d 1165, 1171 (C.D. Cal. 2006)). In these three contexts, the government may exercise its sovereign power more directly than by using carrots and sticks. By simply preventing access to religious locations and resources, the government may directly prevent religious exercise.
It would be an exceedingly odd statute that recognized and provided remedies for government-created substantial burdens on religious exercise only when the government uses carrots and sticks to influence people‘s behavior indirectly but not when it directly prevents access to religious resources. Yet the majority reaches just that illogical interpretation of RFRA in this case, without acknowledging its incoherence.
Of course, Congress can enact illogical laws if it chooses. But there is no basis for concluding that RFRA is such a statute, and several reasons for concluding it is not.
First, as discussed, the majority relies primarily on RFRA‘s invocation of Sherbert and Yoder in reinstating the compelling interest test. RFRA also refers generally to “Federal court rulings” “prior” to Smith.
Second, the Supreme Court has held repeatedly that courts should apply the “same standard” in deciding cases under RFRA and the Religious Land Use and Institutionalized Persons Act (“RLUIPA“),
Third, recent Supreme Court case law makes evident that pre-Smith cases should not be read to cabin RFRA‘s reach. As the Supreme Court has explained, there is “no reason to believe” that RFRA “was meant to be limited to situations that fall squarely within the holdings of pre-Smith cases.” Burwell v. Hobby Lobby Stores, Inc., 573 U.S. 682, 706 n.18 (2014). If this court held otherwise in Navajo Nation, 535 F.3d 1058—which I do not believe it did—then Navajo Nation is in irreconcilable conflict with subsequent Supreme Court case law and is no longer binding. See Miller, 335 F.3d at 900.
A.
A review of the pre-Smith Free Exercise cases ignored by the majority demonstrates that the majority‘s constrained definition of “substantial burden” lacks a basis in pre-Smith precedent. In Cruz v. Beto, 405 U.S. 319 (1972) (per curiam), for example, a Buddhist prisoner in Texas alleged that the prison denied him access to the prison chapel and prohibited him from corresponding with his religious adviser. The Supreme Court reversed the dismissal of the complaint, noting that if the allegations were “assumed to be true,” “Texas has violated [the Free Exercise Clause of] the First and Fourteenth Amendments.” Id. at 322. Later, in O‘Lone v. Estate of Shabazz, 482 U.S. 342 (1987), Muslim prisoners assigned to an outside work detail were prevented from attending “Jumu‘ah, a weekly Muslim congregational service.” Id. at 345. The Supreme Court held that the policy requiring the prisoners to remain outside did not violate the Free Exercise Clause, but not because there was no burden on the prisoners’ religious exercise. Assuming a burden, the Court went on to evaluate the question whether the burden was justified by “legitimate penological objectives” and found that it was. Id. at 352–53. In both these cases, the claim was not that the plaintiffs were “forced to choose between following the tenets of their religion and receiving a governmental benefit” or “coerced to act contrary to their religious beliefs by the threat of civil or criminal sanctions,” Majority Op. 20 (internal quotation marks omitted), but that they were directly denied access to religious resources.
Similarly, in McElyea v. Babbitt, 833 F.2d 196 (9th Cir. 1987), we reversed a grant of summary judgment in favor of an Arizona prison because the plaintiff had raised triable issues of fact regarding his claims that “(1) there were no weekly Jewish services conducted at the prison; (2) he was unreasonably denied permission to attend a special service on the High Holy Days; (3) he was unable to obtain a kosher diet; and (4) there were no Jewish religious writings available at the prison.” Id. at 197. Our reversal for further factual development recognized that, if true, the allegations not only raised equal protection concerns but also showed a burden on religious exercise that the government must justify. Id. at 199. We explained, for example, that “the defendants cannot erect a barrier to an inmate‘s access to religious reading material absent a security or penological interest.” Id.2
In short, federal cases prior to Smith accepted that governments substantially burden religious exercise—and so must justify their actions—when they control access to religious resources and deny plaintiffs access to those resources. The notion that pre-Smith cases recognized a substantial burden only when the government denied a benefit or threatened a penalty is revisionist history not supported by the case law.
B.
Nor is there any reason to believe that Congress, in enacting RFRA, narrowed the definition of “substantial burden” from what it had been in the pre-Smith Free Exercise cases. Congress enacted RFRA as a reaction to Smith, “which held that neutral, generally applicable laws that incidentally burden the exercise of religion usually do not violate the Free Exercise Clause of the First Amendment.” Holt, 574 U.S. at 356–57. “Following... Smith, Congress enacted RFRA in order to provide greater protection for religious exercise than is available under the First Amendment.” Id. at 357 (emphasis added). The majority‘s implicit suggestion that in so doing, Congress silently constricted the definition of “substantial burden” is exceedingly difficult to credit in light of the overall thrust of RFRA.
If there were any question whether Congress intended for RFRA‘s definition of “substantial burden” to be broad enough to encompass governmental denial of access to religious resources, it is laid to rest by Congress‘s passage of RLUIPA seven years later. By then, City of Boerne v. Flores, 521 U.S. 507 (1997), had “invalidated RFRA as applied to States and their subdivisions, holding that the Act exceeded Congress’ remedial powers under the Fourteenth Amendment.” Cutter, 544 U.S. at 715. “Congress responded to City of Boerne by enacting RLUIPA, which applies to the States and their subdivisions and invokes congressional authority under the Spending and Commerce Clauses.” Holt, 574 U.S. at 357.
Section
Section 3 of RLUIPA governs religious exercise by institutionalized persons, such as prisoners. “Section 3 mirrors RFRA and provides that ‘[n]o government shall impose a substantial burden on the religious exercise of a person residing in or confined to an institution... even if the burden results from a rule of general applicability, unless the government demonstrates that imposition of the burden on that person—(1) is in furtherance of a compelling governmental interest; and (2) is the least restrictive means of furthering that compelling governmental interest.‘” Holt, 574 U.S. at 357–58 (alterations in original) (quoting
Given that Congress enacted RLUIPA to restore part of RFRA‘s original reach, that RLUIPA uses the same “substantial burden” language as RFRA, and that RLUIPA sets forth the “same standard” for evaluating governmental justifications for imposing substantial burdens on religion as RFRA—strict scrutiny—there is no reason to believe that “substantial burden” means something different under RFRA and RLUIPA. Cases decided under RLUIPA, in both the prison and zoning contexts, confirm that the definition of “substantial burden” includes the denial of access to religious locations and resources.
For example, in Greene v. Solano County Jail, 513 F.3d 982 (9th Cir. 2008), a county jail denied the plaintiff, a maximum-security prisoner, the opportunity to attend group worship services. We had “little difficulty in concluding that an outright ban on a particular religious exercise is a substantial burden on that religious exercise.” Id. at 988. Similarly, in Yellowbear v. Lampert, 741 F.3d 48 (10th Cir. 2014) (Gorsuch, J.), the Tenth Circuit held that it did not “take much work to see that” a prison substantially burdened the plaintiff‘s religious exercise by “flatly prohibiting” him from using the prison‘s sweat lodge. Id. at 56. And in Haight v. Thompson, 763 F.3d 554 (6th Cir. 2014), the Sixth Circuit held that prison officials substantially burdened plaintiffs’ religious exercise by denying them permission to buy ceremonial foods for an annual event. Id. at 565.
Most recently, the Supreme Court stayed the execution of a prisoner who requested that “his long-time pastor be allowed to pray with him and lay hands on him while he is being executed.” Ramirez, 142 S. Ct. at 1272. The Court held that Ramirez was entitled to a preliminary injunction because, among other things, he was “likely to succeed in showing that Texas‘s” refusal to permit audible prayer or religious touch “substantially burdens his exercise of religion.” Id. at 1278.3
In the zoning context, we have held that a county “imposed a substantial burden”
As demonstrated by this case law in the prison and zoning contexts, when the government controls access to religious locations and resources, it substantially burdens religious exercise by directly—rather than indirectly through the use of carrots and sticks—denying access to those locations or resources, objectively interfering with the plaintiff‘s religious practice.
C.
Navajo Nation is not to the contrary. There, we held that “[a]ny burden imposed on the exercise of religion short of that described by Sherbert and Yoder is not a ‘substantial burden’ within the meaning of RFRA.” 535 F.3d at 1070 (emphasis added). By excluding burdens “short of” those described in Sherbert and Yoder, we left room for a more severe burden to qualify as substantial.
As discussed, the government‘s denial of access to religious resources may result in a greater burden on religious exercise—potentially preventing religious practice altogether—than when it influences religious exercise indirectly by withholding benefits or threatening penalties. See Cutter, 544 U.S. at 720–21 (explaining that the “degree of control” the government exercises in institutional contexts is “severely disabling to private religious exercise“); Yellowbear, 741 F.3d at 56 (holding that when a “prison refuses any access” to a sweat lodge, the restriction does not present “a situation where the claimant is left with some degree of choice in the matter and we have to inquire into the degree of the government‘s coercive influence on that choice,” but instead “easily” qualifies as a substantial burden); Haight, 763 F.3d at 565 (“The greater restriction (barring access to the practice) includes the lesser one (substantially burdening the practice).“).
Navajo Nation‘s failure to recognize a substantial burden under the facts of that
In short, in Navajo Nation, the government did not deny access to or destroy a religious site, as the en banc court emphasized. So the case did not involve a situation in which the government objectively and severely interfered with a plaintiff‘s access to religious locations or resources.4
Nor does Bowen v. Roy, 476 U.S. 693 (1986), Lyng v. Northwest Indian Cemetery Protective Association, 485 U.S. 439 (1988), or Trinity Lutheran Church of Columbia, Inc. v. Comer, 137 S. Ct. 2012 (2017), support the majority‘s constricted understanding of the term “substantial burden” in RFRA. In Bowen, the plaintiff objected to the government‘s use of a Social Security number in conducting its “internal affairs.” 476 U.S. at 699. Bowen thus did not address a context in which the government controlled the plaintiff‘s access to religious resources. In Lyng, as in Navajo Nation, the government did control access to several religious sites, but the government action at issue—a proposed road in a national forest—did not deny access to or directly damage the sites. “No sites where specific rituals take place were to be disturbed” by the road, and the government sited the road so as to minimize “audible intrusions” and “visual impact” on the religious sites. 485 U.S. at 454. Last, Trinity Lutheran, in discussing the Court‘s Free Exercise Clause jurisprudence, simply noted that the Court had not found government coercion in Lyng. 137 S. Ct. at 2020. That summation is accurate and does not imply the Court would reach the same result in a case in which the government controlled access to religious resources and entirely denied a plaintiff access to those resources.
In sum, there is no doctrinal basis for limiting the definition of “substantial burden” to the types of burdens imposed in Sherbert and Yoder. To the contrary, the case law supports defining “substantial burden” to include, at a minimum, situations in which the government controls access to religious resources and entirely denies access to or destroys those resources, objectively interfering with the plaintiff‘s religious practice.
As discussed, the Supreme Court has repeatedly instructed courts to apply the “same standard” in cases under RFRA and RLUIPA. Holt, 574 U.S. at 358 (quoting Gonzales, 546 U.S. at 436). Recent Supreme Court cases under RLUIPA and RFRA are irreconcilable with Navajo Nation if that case is read, as the majority reads it, to limit “substantial burden” to denied benefits and threatened penalties. In Ramirez, a case under RLUIPA, the Court‘s holding rested on an understanding of “substantial burden” that includes the denial of access to religious resources where the government controls access to those resources. 142 S. Ct. at 1278. And Hobby Lobby emphasized that Congress enacted RFRA “to provide very broad protection for religious liberty” that goes “far beyond what [the Supreme] Court has held is constitutionally required.” 573 U.S. at 693, 706. The Court rejected as “absurd” the notion that “RFRA merely restored [the Supreme] Court‘s pre-Smith decisions in ossified form.” Id. at 715. If Navajo Nation held that RFRA‘s definition of “substantial burden” is limited to the types of burdens described in Sherbert and Yoder, that holding cannot be squared with Holt, Ramirez, and Hobby Lobby, read together. See Miller, 335 F.3d at 900.5
II.
The majority‘s proffered “practical basis” for its constricted definition of “substantial burden” fares no better than its faulty doctrinal analysis. Majority Op. 24-25. Practicality, the majority maintains, requires limiting the concept of “substantial burden” to exclude burdens arising from the government‘s control over access to Native American sacred sites on federal land because “giving one religious sect a veto over the use of public park land would deprive others of the right to use what is, by definition, land that belongs to everyone.” Id. (quoting Navajo Nation, 535 F.3d at 1063–64).
True, recognizing Native Americans’ right of access to traditional religious sites on federal land may sometimes constrain competing uses of the land. But this “practical basis” for the majority‘s definition of “substantial burden” is flawed in two ways. First, there is no justification for resolving competing claims on the uses of federal land by refusing to recognize the Native American claim at the “substantial burden” stage of the analysis. Second, recognizing a substantial burden on religious exercise does not result in an automatic “veto” over other uses of the land. I address these errors in turn.
First, burdens on Native Americans who practice land-based religions and who depend on the federal government for access
Moreover, it is disingenuous to resolve the concern about competing claims on federal land by slipping it into the substantial burden analysis. The majority‘s concern, revealed by its discussion of the “practical basis” for its holding, has nothing to do with whether the Apaches’ religious exercise is substantially burdened and everything to do with how we address competing demands for resources—in this case, federal land that hosts both a traditional sacred site and a copper deposit. By pretending that the question is whether there is a “substantial burden” on the Apaches’ religious exercise, and not whether the government has shown a compelling interest in putting the site to a different use, we avoid a transparent inquiry into the considerations that should determine the allocation of resources for which there are competing demands, one of which is religion-based.
That brings me to the majority‘s second error, its assertion that acknowledging a substantial burden when Native Americans are denied access to sacred sites would give Native Americans an automatic “veto” over competing uses of federal land. Majority Op. 24-25. It would not. Instead, it would lead us to the second step of the analysis, the compelling interest test.
Unlike the substantial burden inquiry, the compelling interest test provides a transparent tool for airing and resolving conflicts between the interests of religious adherents and those of others in society. It gives the government an opportunity to provide a rationale for its action and demonstrate the lack of viable alternatives. It allows the court to engage in an open discussion about balancing competing interests. And it does not result in an automatic loss for the government. “Strict scrutiny is not ‘strict in theory, but fatal in fact.‘” Grutter v. Bollinger, 539 U.S. 306, 326 (2003) (quoting Adarand Constructors, Inc. v. Peña, 515 U.S. 200, 237 (1995)). According to one empirical analysis, federal courts applying strict scrutiny in religious liberty cases between 1990 and 2003 upheld the challenged laws nearly 60 percent of the time. Adam Winkler, Fatal in Theory and Strict in Fact: An Empirical Analysis of Strict Scrutiny in the Federal Courts, 59 Vand. L. Rev. 793, 796–97 (2006).
The majority has demonstrated neither a doctrinal nor a practical basis for its narrow definition of “substantial burden” under RFRA. The case law and history of RFRA instead support recognizing a substantial burden when the government controls access to religious resources and completely denies access to or destroy those resources, objectively interfering with the plaintiff‘s religious exercise. After finding a substantial burden, courts still must apply RFRA‘s compelling interest standard, which permits a transparent inquiry into the strength of the government‘s proffered justification for its action.
III.
Applying the proper definition of the term, there is no doubt that the complete destruction of Oak Flat would be a “substantial burden” on the Apaches’ religious exercise. As the district court found, the “evidence... shows that the Apache peoples have been using Oak Flat as a sacred religious ceremonial ground for centuries.” Apache Stronghold v. United States, 519 F. Supp. 3d 591, 603 (D. Ariz. 2021). And the Oak Flat location is not fungible with
The purpose of the Land Transfer Act, and Resolution Copper‘s planned use of the land, is to extract copper ore from below Oak Flat, using a technique called “block caving” or “panel caving.” Once the ore is removed, the land above the deposit will collapse, creating a “subsidence zone” about 1.8 miles in diameter and about 1,000 feet deep, destroying Oak Flat. According to the government‘s environmental impact statement, “the impacts on archaeological sites, tribal sacred sites, cultural landscapes, and plant and mineral resources caused by construction of the mine would be immediate, permanent, and large in scale.” As the district court found, “the land... will be all but destroyed to install a large underground mine, and Oak Flat will no longer be accessible as a place of worship.” Apache Stronghold, 519 F. Supp. 3d at 606. By preventing the Apache people from using Oak Flat as a site for religious ceremonies as they have for centuries, the Land Transfer Act will “have a devastating effect on the Apache people‘s religious practices.” Id. at 607. “The Western Apaches’ exercise of religion at Oak Flat will not be burdened—it will be obliterated.” Order Denying Emergency Mot. for Injunction Pending Appeal at 9, Apache Stronghold v. United States, No. 21-15295 (9th Cir. Mar. 5, 2021) (Bumatay, J., dissenting), ECF No. 26.
As the government controls access to Oak Flat and the result of the Land Transfer Act will be to make the site inaccessible and eventually destroy it, objectively preventing Apaches from holding religious ceremonies there, I would hold Apache Stronghold is likely to succeed in showing a substantial burden on its members’ religious exercise.6
Once a court finds a substantial burden, “the burden of persuasion shifts to the government to prove that the challenged government action is in furtherance of a ‘compelling governmental interest’ and is implemented by ‘the least restrictive means.‘” Navajo Nation, 535 F.3d at 1068. The government has not attempted to satisfy the compelling interest test here or in the district court, instead limiting its arguments to the substantial burden issue.
Because the government bears the burden of persuasion on the compelling interest test and has not carried it, Apache Stronghold is likely to succeed on the merits of its RFRA claim. See Gonzales, 546 U.S. at 429 (confirming the government bears the burden of satisfying RFRA‘s compelling interest test at the
I therefore respectfully dissent.
Notes
I do not stand principally on this point, however. I am reluctant to lend credence to the notion that a trespass conviction is a substantial burden on religion but complete destruction of an irreplaceable religious location is not.