United States Fish and Wildlife Serv. v. Sierra Club, Inc.
The Supreme Court ruled that two federal wildlife agencies could keep their internal draft assessments private under the Freedom of Information Act, even though those drafts proved to be the last thing written about the EPA's 2013 proposed rule.
The decision reinforces that what matters under the deliberative process privilege is not whether a document was the last one written, but whether agency decision-makers actually adopted it as their final position — documents shelved and never approved remain protected.
“What matters, then, is not whether a document is last in line, but whether it communicates a policy on which the agency has settled.”
The majority's core test for distinguishing a protected deliberative draft from a disclosable final agency decision.
How it got here: Sierra Club sued the agencies in federal district court; the district court ordered disclosure; the Ninth Circuit affirmed; the government asked the Supreme Court to step in and the Court agreed to hear it.
The Case in Depth
What happened
The EPA proposed rules governing industrial cooling water intake structures, which can trap and kill fish and other aquatic wildlife. Because the rules might harm protected species, the EPA was required to consult with two federal wildlife agencies — the U.S. Fish and Wildlife Service and the National Marine Fisheries Service. Staff members at both agencies wrote draft opinions in late 2013 concluding the EPA's proposed rule would likely harm protected species. Their supervisors shelved the drafts, extended negotiations, and eventually cleared a revised 2014 rule as safe for wildlife. Sierra Club, an environmental organization, sought the 2013 draft opinions through public records requests.
The question before the Court
Can federal wildlife agencies withhold internal draft biological opinions under a government-records law, even when those drafts turned out to be the last written assessment of a proposed environmental rule?
The Court's answer
Yes — the deliberative process privilege protects the draft biological opinions from disclosure because, even though they proved to be the last written word on the EPA's 2013 proposed rule, the agencies never treated them as final. The key question is not whether a document was the last one produced, but whether agency decision-makers adopted it as their settled position. Here, supervisors at both agencies never approved the drafts or sent them to the EPA — they shelved the documents and kept negotiating. The Court characterized the disputed documents not even as full draft biological opinions, but as "drafts of draft biological opinions."
The Court also rejected the argument that the drafts had "operative effect" because they prompted the EPA to revise its rule. Under the Court's precedent, "operative effect" means direct legal consequences — which only a final biological opinion can trigger — not the practical ripple effects that any internal communication might cause. An approach tying finality to real-world influence, the Court said, would gut the deliberative process privilege entirely.
Curious how the Court got there? See the step-by-step legal reasoning →
Why it matters
Environmental groups, journalists, and others who request government records will face greater difficulty obtaining draft agency assessments, even when those drafts appear to capture the agencies' real conclusions. As long as agency supervisors never formally approved a document and continued their internal deliberations, the draft stays shielded from public disclosure — even if it shaped how another agency acted.
What changes now
The case is sent back to lower courts. On remand, the district court must determine whether any portions of the withheld draft documents are "segregable" — meaning parts that do not contain privileged information must still be disclosed. The broader ruling is final: draft biological opinions that were never approved by agency decision-makers are protected under FOIA's deliberative process privilege. Future disputes about whether a particular draft was truly unapproved will be decided case by case.
What this does not decide
The Court left open whether a properly completed draft biological opinion that agencies did treat as their settled conclusion might be disclosable. It also did not resolve whether a draft sent to the EPA (rather than shelved) would remain protected. The segregability question — whether non-privileged portions of the documents must still be released — goes back to the district court.
Concurrences and dissents
Dissent — Justice Breyer
Justice Breyer argued that a properly completed Draft Biological Opinion — as distinct from a true draft-of-a-draft — is functionally equivalent to a Final Biological Opinion: both explain the agencies' findings, both present the same alternatives, and both leave the EPA with the same four choices. He contended that because Draft Biological Opinions regularly become public and serve the same practical purposes as final ones, they should not be shielded by the deliberative process privilege. He would have remanded to determine whether the specific documents at issue were genuine Draft Biological Opinions or merely unfinished drafts of them.
How the Court got there
The legal reasoning, step by step
- FOIA Exemption 5 incorporates the deliberative process privilege — a protection that shields documents reflecting the give-and-take of agency policymaking from public disclosure. The privilege is designed to encourage candor among officials, who might otherwise soften their views if every internal memo could become front-page news. But the privilege does not protect documents that embody a final agency decision.
- The core test asks whether an agency treated the document as its final view, concluding the deliberative process and giving the document 'real operative effect.' The key is not whether a document was last in the sequence — sometimes proposals simply die on the vine — but whether it communicates a policy on which the agency has settled.
- The regulations governing biological opinions draw a structural distinction between drafts and finals: a draft must be shared with the action agency for review, and the final may not be issued while that review is ongoing. This framework specifically contemplates that a draft may change before a final opinion issues, confirming that draft opinions are inherently non-final.
- The supervisors at both wildlife agencies never approved the 2013 draft opinions or transmitted them to the EPA. Instead, they concluded that 'more work needed to be done,' extended the consultation period, and ultimately cleared a significantly revised 2014 rule. Because decision-makers never treated the 2013 drafts as their settled view, the Court called them 'drafts of draft biological opinions' — even further from finality than an ordinary draft.
- Sierra Club's argument that the drafts had 'operative effect' — because they led the EPA to revise its proposed rule — was rejected. Under NLRB v. Sears, Roebuck & Co., a 1975 case that set the framework for this privilege, 'operative effect' means direct legal consequences, not practical influence. Only a final biological opinion triggers legal obligations; a draft carries no such force, and any internal agency communication might influence behavior without becoming a final decision.
- The Court acknowledged that agencies cannot abuse the 'draft' label to hide functionally final decisions — if evidence showed an agency disguised a settled conclusion as a draft, the privilege would not apply. But no such concealment occurred here: the consultation continued, the EPA changed its rule significantly, and the opinion that actually concluded the process was the agencies' joint 'no jeopardy' finding about the 2014 rule.
Doctrinal impact
Cases affected by this decision
Reaffirms NLRB v. Sears, Roebuck & Co. (421 U. S. 132)
Reaffirmed as the governing framework for when the deliberative process privilege shields agency documents from FOIA disclosure.
Distinguishes Bennett v. Spear (520 U. S. 154)
'Operative effect' means direct legal consequences only — not practical influence on another agency's behavior.