NEPA Hard Look Requirement: Prohibiting Arbitrary Agency Departures from Model‐Based Risk Assessments

Introduction

In WildEarth Guardians v. U.S. Forest Service, 10th Cir. No. 24-1187 (May 9, 2025), the Tenth Circuit addressed a challenge under the Administrative Procedure Act (APA) and the National Environmental Policy Act (NEPA) to the Forest Service’s decision authorizing new domestic sheep grazing (“Wishbone Allotments”) in the Rio Grande National Forest. Petitioners WildEarth Guardians and Western Watersheds Project contended the Forest Service acted arbitrarily and capriciously when it overrode the results of a science‐based “risk of contact” model (RCM)—which predicted a high disease transmission risk to vulnerable Rocky Mountain bighorn sheep—and instead relied on untested “local factors” (topographical barriers, shortened season, best management practices) to downgrade the risk to “moderate.” The Court agreed that the agency failed to take the requisite “hard look” at environmental impacts, reversed the district court’s denial of relief, and remanded to determine appropriate remedy.

Summary of the Judgment

• The RCM had consistently predicted a high risk of disease transmission (one event every 4–8 years) if domestic sheep grazed near bighorn herds, a frequency far above the 32-year threshold for “high” risk.
• In prior decisions (2013, 2015), the Forest Service had vacated grazing allotments when the RCM rated them “high,” citing insufficient evidence that “project design criteria” could reliably prevent contact.
• In 2017, however, the agency authorized the Wishbone Allotments despite the RCM’s “high” rating, invoking unquantified “local factors”—river and highway barriers, reduced grazing window, fragmented pastures, and assumed herder effectiveness—to downgrade risk to “moderate.”
• Petitioners objected administratively and then sued in district court. The Tenth Circuit held the Forest Service’s departure from its own model was arbitrary and capricious, as it lacked scientific data or literature to justify overriding the RCM results. The agency also failed to consider cumulative impacts on neighboring bighorn herds. The Court reversed and remanded for remedy under the APA.

Analysis

1. Precedents Cited

NEPA “Hard Look” Standard: Robertson v. Methow Valley Citizens Council, 490 U.S. 332 (1989) – agencies must rigorously examine environmental consequences.
APA Arbitrary and Capricious Review: 5 U.S.C. § 706(2)(A); Motor Vehicle Mfrs. Ass’n v. State Farm, 463 U.S. 29 (1983) – agency must consider relevant factors and articulate rational connection between record and decision.
Unexplained Change in Agency Position: F.C.C. v. Fox Television Stations, 556 U.S. 502 (2009); Cure Land, LLC v. USDA, 833 F.3d 1223 (10th Cir. 2016) – an agency must supply a reasoned explanation when departing from earlier findings.
Remedy for APA Violations: Allied-Signal, Inc. v. NRC, 988 F.2d 146 (D.C. Cir. 1993) – vacatur/remand balancing test.
Climate and Cumulative Impact Analyses: Diné Citizens Against Ruining Our Env’t v. Haaland, 59 F.4th 1016 (10th Cir. 2023); Center for Biological Diversity v. DOI, 72 F.4th 1166 (10th Cir. 2023) – scope of NEPA’s requirement to use best available science and assess cumulative effects.

2. Legal Reasoning

• Reliance on the RCM: The RCM was recognized by the Forest Service as the “best available science” to predict disease transmission from domestic sheep to bighorn herds. It generated a “high” risk rating (disease interval ≤32 years) for the Wishbone Allotments.
• Arbitrary Departures: The agency replaced the RCM output with untested “local factors”—roads, rivers, reduced grazing duration, fragmented pastures, and project design criteria—without citing data or literature that these factors could reliably reduce contact frequency.
• Inconsistent Treatment: Shortly before, the agency had vacated overlapping allotments precisely because the RCM rated them “high,” underscoring that project design criteria alone were insufficient. Having approved the same permit configuration under “local factors,” the agency offered no reasoned explanation for this switch.
• Cumulative Effects Omitted: Although NEPA requires analysis of reasonably foreseeable impacts on neighboring bighorn populations, the agency confined its risk assessment to the three closest herds and ignored data on herd interactions beyond those zones.
• Failure to Take a Hard Look: By supplanting a validated risk model with logical assumptions untethered to empirical evidence, and by omitting cumulative‐impact analysis, the Forest Service failed NEPA’s hard look requirement and acted arbitrarily and capriciously under the APA.

3. Impact on Future Cases

• This decision reinforces that agencies must adhere to their own predictive models unless they can point to solid data or peer‐reviewed literature supporting any departure.
• It underscores NEPA’s mandate to analyze cumulative effects on all affected populations, not only those nearest the proposed action.
• The ruling warns agencies against “cherry‐picking” untested management measures to gloss over adverse model predictions, lest their actions be set aside as arbitrary and capricious.
• Decisionmakers will need to document empirical support whenever they propose to override or adjust a model’s output—and will have to reconcile any new policy with prior inconsistent decisions.

Complex Concepts Simplified

NEPA “Hard Look”: The legal requirement that federal agencies thoroughly evaluate environmental impacts before acting. It is reviewed under the APA’s “arbitrary and capricious” standard.
Administrative Procedure Act (APA): Federal statute governing agency decisionmaking, authorizing courts to set aside actions that are arbitrary, capricious, or inconsistent with law.
Risk of Contact Model (RCM): A statistical tool used by the Forest Service to predict how often domestic sheep grazing near bighorn populations will result in interspecies contact—and thereby disease transmission.
Core Herd Home Range (CHHR): The principal grazing and breeding area of a bighorn herd, derived from GPS collar data and habitat mapping.
Project Design Criteria: Best management practices (e.g., herder staffing, stray‐sheep removal, hazing) intended to keep domestic and wild sheep apart.
Environmental Assessment (EA) vs. Environmental Impact Statement (EIS): Under NEPA, an EA is a concise study to determine whether an EIS (a more exhaustive review) is required. A Finding of No Significant Impact (FONSI) follows if the EA shows no major harm.
Arbitrary and Capricious: A legal standard meaning an agency’s decision is set aside if it fails to examine relevant evidence, offers explanations contrary to the record, or implements a reasoning so implausible it cannot reflect agency expertise.

Conclusion

WildEarth Guardians v. U.S. Forest Service crystallizes essential principles of NEPA and APA review. Federal agencies must not override validated scientific models with untested assumptions unless they disclose—and rigorously document—the empirical basis for doing so. They must also analyze cumulative effects across all impacted ecosystems. The Tenth Circuit’s ruling vacates the Forest Service’s permit and remands for an appropriate remedy, affirming that reasoned, transparent decisionmaking is indispensable when environmental stakes—such as the survival of vulnerable bighorn sheep—hang in the balance.