United States Fish and Wildlife Service v. Sierra Club, Inc., 592 U.S. ___ (2021)

Docket No. 19-547
Granted: March 2, 2020
Argued: November 2, 2020
Decided: March 4, 2021
Justia Summary

The Environmental Protection Agency (EPA) proposed a rule in 2011 regarding cooling water intake structures. Because aquatic wildlife can become trapped in intake structures, the Endangered Species Act required the EPA to consult with the Fish and Wildlife Service and National Marine Fisheries Service before proceeding. Issuance of a “jeopardy” biological opinion would require the EPA either to implement alternatives, to terminate the action, or to seek an exemption. After consulting with the Services, the EPA changed its proposed rule. Staff members at the Services concluded that the 2013 proposed rule was likely to jeopardize certain species and sent drafts of their opinions to the decision-makers within the Services. Those decision-makers neither approved the drafts nor sent them to the EPA but extended the consultation. In 2014, the EPA produced a revised proposed rule that differed significantly from the 2013 version. The Services issued a final “no jeopardy” biological opinion. The EPA issued its final rule.

Primary Holding

The deliberative process privilege protects from FOIA disclosure in-house draft biological opinions that are pre-decisional and deliberative, even if the drafts reflect the agencies’ last views about a proposal.


NOTE: Where it is feasible, a syllabus (headnote) will be released, as is being done in connection with this case, at the time the opinion is issued. The syllabus constitutes no part of the opinion of the Court but has been prepared by the Reporter of Decisions for the convenience of the reader. See United States v. Detroit Timber & Lumber Co., 200 U.S. 321, 337.




certiorari to the united states court of appeals for the ninth circuit

No. 19–547. Argued November 2, 2020—Decided March 4, 2021

NOTICE: This opinion is subject to formal revision before publication in the preliminary print of the United States Reports. Readers are requested to notify the Reporter of Decisions, Supreme Court of the United States, Washington, D. C. 20543, of any typographical or other formal errors, in order that corrections may be made before the preliminary print goes to press.



No. 19–547



on writ of certiorari to the united states court of appeals for the ninth circuit

[March 4, 2021]

Justice Barrett delivered the opinion of the Court.

The Freedom of Information Act (FOIA) requires that federal agencies make records available to the public upon request, unless those records fall within one of nine exemptions. Exemption 5 incorporates the privileges available to Government agencies in civil litigation, such as the deliberative process privilege, attorney-client privilege, and attorney work-product privilege. This case concerns the deliberative process privilege, which protects from disclosure documents generated during an agency’s deliberations about a policy, as opposed to documents that embody or explain a policy that the agency adopts. We must decide whether the privilege protects in-house drafts that proved to be the agencies’ last word about a proposal’s potential threat to endangered species. We hold that it does.



In April 2011, the Environmental Protection Agency (EPA) proposed a rule on the design and operation of “cooling water intake structures,” which withdraw large volumes of water from various sources to cool industrial equipment. EPA’s stated goal was to require industrial facilities to use “the best technology available” for “minimizing adverse environmental impact.” 76 Fed. Reg. 22174 (2011). But it was unclear whether the proposed rule would achieve that goal, at least when it came to aquatic wildlife. The water withdrawn by these structures typically contains fish and other organisms that can become trapped in the intake system and die. If the EPA’s rule did not adequately guard against this risk, it would jeopardize species protected under the Endangered Species Act of 1973, 87Stat. 884, 16 U. S. C. §1531 et seq.

When an agency plans to undertake action that might “adversely affect” a protected species, the agency must consult with the U. S. Fish and Wildlife Service (FWS) and National Marine Fisheries Service (NMFS) (together, “Services”) before proceeding. See 16 U. S. C. §1536(a)(2); 50 CFR §§402.01–402.17 (2019).[1] The goal of the consultation is to assist the Services in preparing an official “biological opinion” on whether the agency’s proposal will jeopardize the continued existence of threatened or endangered species. §402.14(g)(4). These opinions are known as “ ‘jeopardy’ ” or “ ‘no jeopardy’ ” biological opinions. §402.14(h)(1)(iv), as amended, 84 Fed. Reg. 45017 (2019). If the Services conclude that the action will cause “jeopardy,” they must propose “reasonable and prudent alternatives” to the action that would avoid harming the threatened species. 16 U. S. C §1536(b)(3)(A); 50 CFR §402.14(h)(2). And if a “jeopardy” biological opinion is issued, the agency must either implement the reasonable and prudent alternatives, terminate the action altogether, or seek an exemption from the Endangered Species Committee. 16 U. S. C. §§1536(b)(4), (g), 1538(a).

The EPA began informally consulting with the Services about its proposed regulations on cooling water intake structures in 2012, see 50 CFR §402.13, and it requested a formal consultation in 2013, see §402.14. Throughout this period, the Services and the EPA conducted meetings, held conference calls, and exchanged emails and draft documents on the proposed rule and its potential effect on endangered species.

As a result of the consultation, the EPA made changes to its proposed rule, and the Services received the revised version in November 2013. Soon after, the Services tentatively agreed to provide the EPA with draft biological opinions by December 6, 2013, and final opinions by December 20, 2013. See §402.14(g)(5) (requiring the Services to provide a “draft biological opinion” to action agency upon request).

Staff members at NMFS completed a draft biological opinion on December 6, and staff members at FWS completed a draft on December 9. Both drafts concluded that the proposed rule was likely to jeopardize certain species and identified possible reasonable and prudent alternatives that the EPA could pursue. Staff members sent the drafts to the relevant decisionmakers within each Service and prepared to circulate them to the EPA.

But decisionmakers at the Services neither approved the drafts nor sent them to the EPA. Instead, concluding that “more work needed to be done,” the decisionmakers decided to continue discussions with the EPA. App. 37, 58–59. The EPA was still engaged in an internal debate about key elements of the rule, and the Services wanted a better grasp of what the EPA proposed to do. So the Services shelved the draft opinions and agreed with the EPA to extend the period of consultation.

Over the next several months, the Services and the EPA continued to discuss the rule, and in March 2014, the EPA sent the Services a proposed rule that differed significantly from the 2013 version. Satisfied that the revised rule was unlikely to harm any protected species, the Services issued a joint final “no jeopardy” biological opinion, thereby terminating the formal consultation. See 50 CFR §402.14(m)(1), as amended, 84 Fed. Reg. 45016. The EPA issued its final rule that same day.


Sierra Club, an environmental organization, later submitted FOIA requests for records related to the Services’ consultations with the EPA. The Services turned over thousands of documents, but they invoked the deliberative process privilege for others—including the draft biological opinions analyzing the EPA’s 2013 proposed rule. The deliberative process privilege shields documents that reflect an agency’s preliminary thinking about a problem, as opposed to its final decision about it. The Services asserted that as drafts, the withheld documents were necessarily nonfinal and therefore protected.

Sierra Club sued the Services in the Northern District of California, alleging that the withheld documents were subject to disclosure under FOIA. The District Court agreed with Sierra Club, and the Ninth Circuit affirmed in part. 925 F.3d 1000 (2019). As relevant here, it held that the draft biological opinions were not privileged because even though they were labeled as drafts, they represented the Services’ final opinion that the EPA’s 2013 proposed rule was likely to have an adverse effect on certain endangered species.[2] Judge Wallace dissented in part on the ground that the drafts were part of the ongoing consultation process rather than summaries of the Services’ final views.

We granted certiorari. 589 U. S. ___ (2020).



FOIA mandates the disclosure of documents held by a federal agency unless the documents fall within one of nine enumerated exemptions. See 5 U. S. C. §552(b). The fifth of those exemptions protects “inter-agency or intra-agency memorandums or letters that would not be available by law to a party other than an agency in litigation with the agency.” §552(b)(5). As the text indicates—albeit in a less-than-straightforward way—this exemption incorporates the privileges available to Government agencies in civil litigation. That list includes the deliberative process privilege, attorney-client privilege, and attorney work-product privilege. See Department of Interior v. Klamath Water Users Protective Assn., 532 U.S. 1, 8 (2001).

This case concerns the deliberative process privilege, which is a form of executive privilege. To protect agencies from being “forced to operate in a fishbowl,” EPA v. Mink, 410 U.S. 73, 87 (1973) (internal quotation marks omitted), the deliberative process privilege shields from disclosure “documents reflecting advisory opinions, recommendations and deliberations comprising part of a process by which governmental decisions and policies are formulated,” NLRB v. Sears, Roebuck & Co., 421 U.S. 132, 150 (1975) (internal quotation marks omitted). The privilege is rooted in “the obvious realization that officials will not communicate candidly among themselves if each remark is a potential item of discovery and front page news.” Klamath, 532 U. S., at 8–9. To encourage candor, which improves agency decisionmaking, the privilege blunts the chilling effect that accompanies the prospect of disclosure.

This rationale does not apply, of course, to documents that embody a final decision, because once a decision has been made, the deliberations are done. The privilege therefore distinguishes between predecisional, deliberative documents, which are exempt from disclosure, and documents reflecting a final agency decision and the reasons supporting it, which are not. See Renegotiation Bd. v. Grumman Aircraft Engineering Corp., 421 U.S. 168, 186 (1975). Documents are “predecisional” if they were generated before the agency’s final decision on the matter, and they are “deliberative” if they were prepared to help the agency formulate its position. See Sears, 421 U. S., at 150–152; Grumman, 421 U. S., at 184–186, 190. There is considerable overlap between these two prongs because a document cannot be deliberative unless it is predecisional.

It is not always self-evident whether a document represents an agency’s final decision, but one thing is clear: A document is not final solely because nothing else follows it. Sometimes a proposal dies on the vine. National Security Archive v. CIA, 752 F.3d 460, 463 (CADC 2014) (Kavanaugh, J.). That happens in deliberations—some ideas are discarded or simply languish. Yet documents discussing such dead-end ideas can hardly be described as reflecting the agency’s chosen course. See Sears, 421 U. S., at 150–151. What matters, then, is not whether a document is last in line, but whether it communicates a policy on which the agency has settled.

To decide whether a document communicates the agency’s settled position, courts must consider whether the agency treats the document as its final view on the matter. See id., at 161. When it does so, the deliberative “process by which governmental decisions and policies are formulated” will have concluded, and the document will have “real operative effect.” Id., at 150, 160 (internal quotation marks omitted). In other words, once cited as the agency’s final view, the document reflects “the ‘consummation’ of the agency’s decisionmaking process” and not a “merely tentative” position. See Bennett v. Spear, 520 U.S. 154, 177–178 (1997) (discussing finality in context of obtaining judicial review of agency action). By contrast, a document that leaves agency decisionmakers “free to change their minds” does not reflect the agency’s final decision. Grumman, 421 U. S., at 189–190, and n. 26.


The deliberative process privilege protects the draft biological opinions at issue here because they reflect a preliminary view—not a final decision—about the likely effect of the EPA’s proposed rule on endangered species.[3]

We start with the obvious point that the Services identified these documents as “drafts.” A draft is, by definition, a preliminary version of a piece of writing subject to feedback and change. That is not to say that the label “draft” is determinative. As we have explained before, a court must evaluate the documents “in the context of the administrative process which generated them.” Sears, 421 U. S., at 138. Here, though, the administrative context confirms that the drafts are what they sound like: opinions that were subject to change.

Consider the regulatory process that generates a draft biological opinion. The governing regulation distinguishes between draft and final biological opinions by separating the steps at which each is produced. If the Services prepare a biological opinion, they must “make available” to the action agency—in this case, the EPA—a “draft” of that opinion and generally may not issue the final opinion “while the draft is under review’’ by the action agency. 50 CFR §402.14(g)(5). This provision thus specifically contemplates further review by the agency after receipt of the draft, and with it, the possibility of changes to the biological opinion after the Services send the agency the draft.[4]

Consistent with this understanding, the agreement between the Services and the EPA allowed for the possibility of postcirculation changes. The Services were scheduled to provide the EPA with draft copies of the biological opinions on December 6 and final versions by December 20. If the drafts were to be final and immune from change, there would have been little reason to include a two-week period between the Services’ circulation of the drafts and their submission of the final product. The logical inference is that the Services expected the EPA to provide comments that they might incorporate into the final opinion.

Sierra Club contends, though, that while these documents may have been called “drafts,” they were actually intended to give the EPA a sneak peek at a conclusion that the Services had already reached and were unwilling to change. And Sierra Club says that the EPA responded accordingly: Once the EPA knew that a jeopardy opinion was coming, it revised its proposed rule. Sierra Club insists that the draft opinions thus had an “operative effect” on the EPA and must be treated as final under our precedent. See Sears, 421 U. S., at 160.

Sierra Club misunderstands our precedent. While we have identified a decision’s “real operative effect” as an indication of its finality, that reference is to the legal, not practical, consequences that flow from an agency’s action. Ibid. (noting that the relevant memorandum has “real operative effect” because it “permits litigation before the Board”); id., at 159, n. 25 (comparing the “operative effect” of the memorandum to that of a district court order). In this regulatory scheme, a final biological opinion leads to “direct and appreciable legal consequences” because it alters “the legal regime to which the action agency is subject, authorizing it” to take action affecting an endangered species “if (but only if ) it complies with the prescribed conditions.” Bennett, 520 U. S., at 178. That is not true of a draft biological opinion.

To be sure, a draft biological opinion might carry a practical consequence if it prompts the action agency to change its proposed rule. For example, the agency might adopt an alternative approach that avoids jeopardizing an endangered species. But many documents short of a draft biological opinion could prompt an agency to alter its rule. An agency might make changes in response to the Services’ views—or, for that matter, the views of the agency’s own officials—at any stage of the consultation process. And even Sierra Club does not contend that any email or memorandum that has the effect of changing an agency’s course constitutes a final administrative decision. That approach would gut the deliberative process privilege.

Sierra Club’s proposed effects-based test is therefore not the right one. To determine whether the privilege applies, we must evaluate not whether the drafts provoked a response from the EPA but whether the Services treated them as final.

They did not. The drafts were prepared by lower-level staff and sent to the Services’ decisionmakers for approval. Sierra Club characterizes the drafts as polished documents lacking only an autopen signature. But the determinative fact is not their level of polish—it is that the decisionmakers at the Services neither approved the drafts nor sent them to the EPA. Instead, the decisionmakers concluded that “more work needed to be done” and extended the time for consultation with the EPA. These documents, then, are best described not as draft biological opinions but as drafts of draft biological opinions. Sierra Club’s argument thus fails on its own terms: Even assuming that a draft biological opinion would have expressed the Services’ settled conclusion, a draft of a draft is a far cry from an “agency decision already made.” Grumman, 421 U. S., at 184.

It is true, as Sierra Club emphasizes, that the staff recommendations proved to be the last word within the Services about the 2013 version of the EPA’s proposed rule. But that does not change our analysis. The recommendations were not last because they were final; they were last because they died on the vine. See Sears, 421 U. S., at 151, n. 18 (“[C]ourts should be wary of interfering” with drafts that “do not ripen into agency decisions”). Further consultation with the Services prompted the EPA to alter key features of its 2013 proposal, so there was never a need for the Services to render a definitive judgment about it. The opinion that came to fruition was the Services’ joint “no jeopardy” opinion about the 2014 version of the EPA’s proposed rule. The staff recommendations were thus part of a deliberative process that worked as it should have: The Services and the EPA consulted about how the rule would affect aquatic wildlife until the EPA settled on an approach that would not jeopardize any protected species.

Sierra Club warns that ruling against it here would permit the Services to stamp every document “draft,” thereby protecting even final agency decisions and creating “ ‘secret [agency] law.’ ” Id., at 153. It is true that a draft document will typically be predecisional because, as we said earlier, calling something a draft communicates that it is not yet final. But determining whether an agency’s position is final for purposes of the deliberative process privilege is a functional rather than formal inquiry. If the evidence establishes that an agency has hidden a functionally final decision in draft form, the deliberative process privilege will not apply. The Services, however, did not engage in such a charade here.

*  *  *

The deliberative process privilege protects the draft biological opinions from disclosure because they are both predecisional and deliberative. We reverse the contrary judgment of the Court of Appeals and remand the case for further proceedings consistent with this opinion.[5]

It is so ordered.

1  The FWS and NMFS administer the statute on behalf of the Secretaries of Interior and Commerce, respectively. See 50 CFR §§17.11, 222.101(a), 402.01(b).
2  The Ninth Circuit also concluded that several other draft documents, including certain documents meant to accompany the draft biological opinions and a March 2014 draft of reasonable and prudent alternatives, were not privileged.
3  Like the parties, we focus on the draft biological opinions. But the logic applied to these drafts also applies to the other draft documents.
4 Sierra Club contends that the regulations treat a “jeopardy” finding as final, even though the opinion triggers a discussion of reasonable and prudent alternatives. See 50 CFR §402.14(g)(5) (requiring the Services to make a draft biological opinion available “for the purpose of analyzing the reasonable and prudent alternatives”). As explained below, a critical question is whether the Services treat a draft opinion as final. See infra, at 9. So we do not foreclose the possibility that a draft biological opinion is final because, for example, the Services have made clear that they would not incorporate into that opinion responses made by the action agency, as to reasonable and prudent alternatives or other matters. See infra, at 9–10. We need not resolve that issue because, as we explain below, the Services’ opinions in this case did not count even as drafts under the regulation—they were merely drafts of draft biological opinions.
5  We agree with the parties that the District Court must determine on remand whether any parts of the documents at issue are segregable. See 5 U. S. C. §552(b) (Agencies must disclose “[a]ny reasonably segregable portion” of a document containing some exempt information); §552(a)(4)(B); NLRB v. Sears Roebuck & Co., 421 U.S. 132, 161, n. 27 (1975).


No. 19–547



on writ of certiorari to the united states court of appeals for the ninth circuit

[March 4, 2021]

Justice Breyer, with whom Justice Sotomayor joins, dissenting.

Because the word “draft” may here prove misleading, it should help the reader understand my argument if he or she keeps in mind three different but related kinds of documents: “Final Biological Opinions,” “Draft Biological Opinions,” and “Drafts of Draft Biological Opinions.” A Final Biological Opinion, as its name suggests, embodies a final agency decision, for example, a decision by the Services that a proposed Environmental Protection Agency (EPA) action will jeopardize an endangered species. We all agree, I believe, that a Final Biological Opinion is not deliberative and that Exemption 5 of the Freedom of Information Act (FOIA) does not protect it from disclosure. I also agree with the Court about the third kind of document, a Draft of a Draft Biological Opinion. That kind of document normally is not final. It normally is deliberative. And Exemption 5 normally protects it from disclosure.

But what about the second kind of document, a Draft Biological Opinion? Does it normally set forth a “final” Services view, or is it normally a “deliberative” document? I agree with the Court that whether a document is “final” or “deliberative” primarily depends upon its “function[ ]” within an agency’s decision-making process. Ante, at 10; see also NLRB v. Sears, Roebuck & Co., 421 U.S. 132, 138 (1975) (“[T]he function of the documents” and “the context of the administrative process which generated them” is “[c]rucial” to understanding whether the deliberative process privilege applies). I believe that, in the context before us, the Services’ Draft Biological Opinions reflect “final” decisions regarding the “jeopardy” the EPA’s then-proposed actions would have caused. Hence, they would normally fall outside, not within, Exemption 5.

Five features of the Draft Biological Opinion lead me to this conclusion. First, literally speaking, a Draft Biological Opinion is a “final” document with respect to its content. That is in fact the key difference between a Draft Biological Opinion and a Draft of a Draft Biological Opinion. If further deliberation about the draft’s content is likely, the document is not a Draft Biological Opinion. It is a Draft of a Draft. I recognize that in principle a Service might change its mind about the content of even the most final of Draft Biological Opinions. It might then prepare a new Draft Biological Opinion. But, in principle, a Service could also change its mind about a Final Biological Opinion, withdrawing a Final Biological Opinion already issued and substituting a new one in its place. See, e.g., Oregon Natural Resources Council v. Allen, 476 F.3d 1031, 1032 (CA9 2007) (Fish and Wildlife Service “voluntarily reinitiated consultation . . . [and] withdrew its favorable Biological Opinion”); see also 50 CFR §402.16 (2019) (requiring the Services to reinitiate consultation in specified circumstances, including when “new information reveals effects of the action that may affect listed species or critical habitat in a manner or to an extent not previously considered”). The mere possibility of a future change does not alter the finality, or the final effect, of the original document.

Second, a Final Biological Opinion and a Draft Biological Opinion finding jeopardy serve the same functions within the formal administrative process. Both explain the Services’ findings. Both set forth “reasonable and prudent” modifications or alternatives. 16 U. S. C. §1536(b)(3)(A). And both have substantially the same effect on the EPA (the action agency in this case). In response to both documents, the EPA has essentially four options: It can drop the proposed action; it can accept the proposed modifications; it can take the proposed action and potentially expose itself to considerable penalties, or it can seek a Cabinet-level exemption. See National Assn. of Home Builders v. Defenders of Wildlife, 551 U.S. 644, 652 (2007); see also §§1536(b)(4), (g); 50 CFR §402.15; U. S. Fish and Wildlife Service, Consultations | Frequently Asked Questions (June 10, 2020), https://www.fws.gov/endangered/ what-we-do/faq.html. The EPA could also, of course, propose a new course of action embodying modifications not proposed by the Services—possibly generating a new consultation—but that is just as true of a Final Biological Opinion. See ibid. (noting that an agency can always choose to alter its proposed action).

A Draft Biological Opinion differs from a Final Biological Opinion in only one way that matters. The Services must make the Draft Biological Opinion available to the EPA before it issues a Final Biological Opinion. 50 CFR §402.14(g)(5). It then continues its consultation with the EPA but not with an eye toward changing the Services’ environmental analysis or conclusions. Rather, the negotiations are designed to find less damaging alternatives to the original EPA-proposed action. See ibid. (allowing the action agency to request the Draft Biological Opinion “for the purpose of analyzing the reasonable and prudent alternatives”). If the agencies find suitable alternatives, the EPA will then publicly adopt those alternatives, and the process will culminate in a Final Biological Opinion finding no jeopardy.

The function of a Draft Biological Opinion finding jeopardy then is much the same as that of a Final Biological Opinion finding jeopardy. Transmitting the Draft Biological Opinion to the EPA simply allows the EPA to make its choice before a Final Biological Opinion issues. See ibid.

Third, agency practice shows that the Draft Biological Opinion, not the Final Biological Opinion, is the document that informs the EPA of the Services’ conclusions about jeopardy and alternatives and triggers within the EPA the process of deciding what to do about those conclusions. Amici tell us without contradiction that “out of 6,829 formal consultations” between 2008 and 2015, the Fish and Wildlife Service “issued a [Final Biological Opinion finding] jeopardy” “only twice.” Brief for Center for Biological Diversity et al. as Amici Curiae 22–23. If a Final Biological Opinion is discoverable under FOIA, as all seem to agree it is, why would a Draft Biological Opinion, embodying the same Service conclusions (and leaving the EPA with the same four choices), not be?

Fourth, permitting discovery of Draft Biological Opinions under FOIA is unlikely to chill frank discussion within a Service because the Services’ staff are already aware that these Drafts may well be made public. And for good reason. When a private party prompts the agency action under review, say, by seeking an EPA permit, regulations require the Service to make the Draft Biological Opinion available to the private applicant, removing the Draft Biological Opinion from Exemption 5’s protection. See 50 CFR §402.14(g)(5) (requiring disclosure of Draft Biological Opinions to private applicants if requested); see also Department of Interior v. Klamath Water Users Protective Assn., 532 U.S. 1, 4–5 (2001) (documents exchanged between agency and third party not covered by Exemption 5 because they are no longer inter-agency or intra-agency communications). To hold that Draft Biological Opinions are discoverable when a private party seeks an EPA permit but not when, as here, the EPA seeks to write a generally applicable rule that governs private party conduct seems highly anomalous.

Even where there is no private applicant, the evaluating agencies have a long history of disclosing Draft Biological Opinions to the public. See, e.g., App. 93–98, 102–104 (discussing timing of Draft Biological Opinion disclosure); see also id., at 93–98 (discussing roll-out plan and public talking points for Draft Biological Opinion); Supp. Record in No. 17–16560 (CA9), pp. 164–199 (citing additional examples of public disclosure of Draft Biological Opinions); L. Schiffer, National Oceanic and Atmospheric Admin., Guidelines for Compiling an Agency Administrative Record 10 (Dec. 21, 2012), https://www.gc.noaa.gov/documents/2012/ AR_Guidelines_122112-Final.pdf (“Final draft documents with independent legal significance, such as final draft environmental impact statements, . . . will not be flagged for potential listing on the agency’s Privilege Log” (emphasis in original)). The EPA too may well release a Service’s Draft Biological Opinion. See App. 96 (“EPA [Office of Water] has a track record of putting these drafts on their docket which then show up on regulations.gov”); see also id., at 95 (“I agree that it is likely that EPA will put this draft on their docket”).

Fifth, legal consequences flow from the Services’ completion of a Draft Biological Opinion. The Services’ regulations state that “[i]f requested, the Service shall make available to the Federal agency [i.e., the EPA] the draft biological opinion for the purpose of analyzing the reasonable and prudent alternatives.” 50 CFR §402.14(g)(5). Once the Draft Biological Opinion is under review at the EPA, the Services may not issue a Final Biological Opinion prior to the specified deadline. Ibid. Moreover, as explained, Draft Biological Opinions, like Final Biological Opinions, limit the EPA’s set of available options. Cf. Bennett v. Spear, 520 U.S. 154, 178 (1997) (holding that a Final Biological Opinion has “legal consequences,” even though the action agency is not legally obligated to accept the opinion’s recommendations or conclusions, because the opinion “alter[s] the legal regime to which the action agency is subject”). Why, then, would these same consequences (together with the other factors mentioned above) not also place Draft Biological Opinions outside Exemption 5’s protection?

In sum, the likely finality of a Draft Biological Opinion, its similarity to a Final Biological Opinion, the similar purposes it serves, the agency’s actual practice, the anomaly that would otherwise exist depending upon the presence or absence of a private party, and the presence of at least some regulation-based legal constraints—convince me that a Draft Biological Opinion would not normally enjoy a deliberative privilege from FOIA disclosure.

The question remains whether the particular documents at issue here are Draft Biological Opinions or Drafts of Draft Biological Opinions. As the majority points out, there are reasons to believe some of them may be the latter. See ante, at 7, n. 4, 9. The National Marine Fisheries Service’s documents contain highlighting and editing marks reflective of a work-in-progress. But the Fish and Wildlife Service documents do not, and the record indicates they may have been complete but for a final signature. See App. 105. Given the fact-intensive nature of this question, I would remand to allow the Court of Appeals to determine just how much work was left to be done. If the court determines that the documents are merely Drafts of Draft Biological Opinions, I agree with the majority that a segregability analysis would be appropriate.

For these reasons, I dissent.