dissenting in part:
I agree with the majority that the 2012 Biological Opinion (“BiOp”) is not arbitrary and capricious in determining that the Hawaii-based shallow-set fishery expansion would have no appreciable effect on the leatherback sea turtle population, and that the 2012 BiOp adequately considers the impact of global climate change. However, I dissent from the remainder of the majority opinion.
First, the majority errs in rejecting the U.S. Fish and Wildlife Service’s (“FWS”) issuance of a special purpose permit (the “Permit”) under the Migratory Bird Treaty Act (“MBTA”) to the National Marine Fisheries Service (“NMFS”) for the incidental take of migratory birds. The majority determines that issuing the Permit runs afoul of the pertinent regulation’s plain language and the MBTA’s conservation-oriented purpose. That conclusion, however, reflects a misapplication of our deferential standard of review under Auer v. Robbins, 519 U.S. 452, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997), because both the regulation—50 C.F.R. § 21.27—and the MBTA itself accommodate FWS’s view. See Auer, 519 U.S. at 461, 117 S.Ct. 905; Marsh v. J. Alexander’s LLC, 869 F.3d 1108, 1116-17 (9th Cir. 2017). Moreover, the Permit accords with FWS’s past practice, and thereby reflects its considered judgment—another basis for granting deference under Auer. Christopher v. SmithKline, 567 U.S. 142, 132 S.Ct. 2156, 2166, 183 L.Ed.2d 153 (2012).
Second, the majority errs in rejecting the 2012 BiOp’s assessment of the proposed shallow-set fishery expansion’s effects on the endangered loggerhead sea turtle. NMFS’s BiOp concludes that the proposed action would not jeopardize the continued survival and recovery of the loggerhead, as is required to green-light the project under the Endangered Species Act (“ESA”). The majority dismisses the BiOp as arbitrary and capricious because, among other things, it concludes that the scientific evidence does not support NMFS’s no-jeopardy conclusion, and it perceives a conflict with our case law. I disagree. While the record data shows that the loggerhead is in decline, NMFS reasonably concluded that the fishery expansion would not appreciably reduce the likelihood of the loggerhead’s survival and recovery. Nor did NMFS misapply our decision in National Wildlife Federation v. National Marine Fisheries Service, 524 F.3d 917 (9th Cir. 2008) (“NWF”): it considered the incremental impact of the proposed action along with degraded baseline conditions. That is precisely what- NWF requires.
The majority’s contrary. conclusion is a classic example of the judiciary exceeding its authority by substituting an agency’s judgments with its own. This complex case relies on the technical and scientific findings of experts tasked with the responsibility of protecting our Nation’s species-in-peril. It is in this context that our respect for a coordinate branch of government is at its zenith. Indeed,- we are “ ‘at our most deferential’ when reviewing scientific judgments and technical analyses within .the agency’s expertise,” Lands Council v. McNair, 629 F.3d 1070, 1074 (9th Cir. 2010) (quoting Balt. Gas & Elec. Co. v. Natural Res. Def. Council, 462 U.S. 87, 103, 103 S.Ct. 2246, 76 L.Ed.2d 437 (1983)) (adjustment omitted) (“Lands Council II”), and should only reject an agency’s action if it is-plainly arbitrary and capricious, see Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 42-43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983). Yet instead of anchoring its analysis in well-established -principles of agency deference, the majority sets sail on a voyage of discovery, leaving in-its wake our precedent and the doctrinal moorings of Auer v. Robbins. I dissent, respectfully.
I.
A.
Under Auer v. Robbins, we must defer to ah agency’s reasonable interpretation of its own regulation. See Christopher, 132 S.Ct. at 2166. Deference is not warranted, however, “when the agency’s interpretation is plainly erroneous or inconsistent with the regulation,” or when it does not reflect the agency’s “considered judgment.” Id. (internal quotation marks omitted). A lack of “considered judgment” may be evidenced by (i) an “agencyt ] interpretation [that] conflicts with a prior interpretation;” (ii) “when- it appears that the interpretation is nothing more than a convenient litigating position,” or (iii) when the interpretation amounts to a “post hoc rationalization advanced by an agency seeking to defend past agency action against attack.” Id. (internal quotation marks and adjustment omitted).1
At issue is PWS’s issuance of a special purpose permit allowing NMFS to authorize incidental take of migratory birds .that are protected under the MBTA. 50 C.F.R. § 21.27 authorizes FWS to issue permits for the take of migratory birds protected under the MBTA in certain circumstances. In full, the regulation provides that
[p]ermits may be issued for special purpose activities related to migratory birds, their parts, nests, or eggs, which are otherwise outside the scope of the standard form permits of this part. A special purpose permit for migratory bird' related activities not otherwise provided for in this part may be issued to an applicant who submits a' written application containing the general information and certification required by part 13 and makes a sufficient showing of benefit to the migratory bird resource, important research reasons, reasons of human concern for individual birds, or other compelling justification.
50 C.F.R. § 21.27. The majority declines to defer to FWS’s issuance of the Permit because it finds that FWS’s action is plainly contrary to § 21.27 and the MBTA and is therefore ultra vires. Because I conclude that issuing the Permit does not depart from FWS’s past practice, is not inconsistent with § 21.27’s text, and comports with the MBTA’s conservation-oriented purpose, I would defer to FWS’s determination.
1.
Appellants Center for Biological Diversity, et al. (“CBD”) argue that FWS’s Permit should not be accorded Auer deference because, CBD asserts, it does not align with FWS’s past practice.
To determine whether an agency has departed from past practice, the first step is—manifestly—to definé the practice. Christopher, 132 S.Ct. at 2167-68. A practice is a policy or mode of operating that is defined by articulable parameters; simply showing that a current action differs from a prior one in some way does not establish a departure from past practice. Cf. FCC v. Fox Television Stations, Inc., 556 U.S. 502, 538, 129 S.Ct. 1800, 173 L.Ed.2d 738 (2009) (agency departed from past practice by deeming broadcasts of non-literal uses of expletives as actionable only upon repetition); Dillmon v. Nat’l Transp. Safety Bd., 588. F.3d 1085, 1090-91 (D.C. Cir. 2009) (agency departed from past practice of deferring to an' ALJ’s credibility determinations).
CBD argues that by issuing the Permit, FWS has changed: course from its prior position that it lacks authority to grant permits to allow unintentional bird taking—i.e., incidental taking—for an activity that is not directed at migratory birds. The majority-does not base its .decision on this rationale and for good reason:- FWS has long-issued incidental - take permits for all manner of activities whose only relationship to migratory birds is that they affect the birds. For example, since at least 1996, FWS has.authorized incidental take of migratory birds for commercial activities through Endangered Species Act (“ESA”) Habitat Conservation .Plans (“HCPs”).2 A benefit of entering into an HGP is that it comes with an ESA §' 10 incidental take permit. See 16 U.S.C. § 1539(a)(1)(B), (a)(2). That permit “doublets]” as a § 21.27 special purpose permit under the MBTA. See Dep’t of Commerce, Habitat Conservation Plan Assurances (“No Surprises”) Rule, 63 Fed. Reg. 8859, 8862-63 (Feb. 23, 1998). Critically, the take that occurs results from activities that are'unrelated to migratory birds—e.g., natural gas drilling, homebuilding, and myriad other types of land development—except that they result in incidental bird deaths—the very ill that CBD insists infects the Permit at issue here.
FWS has also issued incidental take permits for bald and golden eagles—-which are migratory birds—for activities that, too, are not directed at migratory birds. See 50 C.F.R. §§ 22.11; 22.26. And in 1998, FWS issued a special purpose permit allowing the incidental take of migratory raptors by a wind farm due to collisions and electrocutions. See FWS Region 6, Fed. Fish & Wildlife Permit No. PRT-808690 (1998). In short, CBD’s categorical assertion that “FWS has always understood [§ ] 21.27 does not authorize incidental take as the Permit allows” is plainly wrong.
Identifying one error in CBD’s consistency-with-past-practice argument reveals another. CBD asserts that, “until [FWS] issued to NMFS the permit at issue exempting commercial longline fishing from the MBTA’s take prohibition, the only Special Use Permits FWS had ever issued authorizing incidental take of non-endangered migratory birds were specifically intended to promote migratory bird conservation .... ” If CBD means to say that past permits were always associated with activities that had as their purpose bird conservation, then the preceding paragraph refutes this contention. But if CBD means something more capacious—i.e., that such activities must incorporate bird conservation strategies—then the Permit addresses this concern. NMFS regulates the Hawaii-based shallow-set longline fishery under a program that is expressly geared at reducing seabird bycatch. See 50 C.F.R. § 665.815(a)(1), (2), (4). Indeed, since the program took effect in 2004, incidental take of seabirds by the fishery has plunged nearly 90 percent. Thus, whatever CBD means by activities that “promote migratory bird conservation,” FWS’s issuance of the Permit is consistent with the agency’s historical practice of tying incidental take permits to conservation measures. If there is a past practice to be discerned, that is it.3
2.
CBD insists that FWS’s past statements belie the agency’s assertion that the Permit accords with historical practice. CBD points to a 2009 regulation governing take under the Bald and Golden Eagle Protection Act (“BGEPA”). See Dep’t of the Interior, Eagle Permits; Take Necessary to Protect Interests in Particular Localities, 74 Fed. Reg. 46,836, 46,862 (Sept. 11, 2009). BGEPA allows for the take of bald and golden eagles—which species also fall under the purview of the MBTA—pursu-ant to an MBTA permit. 50 C.F.R. § 22.11; see 50 C.F.R. § 22.26. In response to a public comment, the regulation’s preamble notes that “[n]o permit is currently available to authorize incidental take under the MBTA.” 74 Fed. Reg. at 46,862. CBD seizes on this language as evidence that the Permit is unlawful.
CBD’s argument proves too much. If the cited statement means that FWS does not issue incidental take permits for migratory birds as a categorical rule, then all other instances of such permits would be unlawful. Yet CBD spills pages of ink distinguishing the Permit here from other take permits granted under the aegis of § 21.27, without suggesting that those permits are similarly unlawful. Moreover, under CBD’s interpretation, the cited statement is irreconcilable with FWS’s other pronouncements permitting take for, e.g., migratory birds that are also ESA-listed species. See 2016 HCP Handbook at 16-9. Cf. Boise Cascade Corp. v. EPA, 942 F.2d 1427, 1432 (9th Cir. 1991) (where possible, courts avoid statutory interpretations that result in inconsistencies).
A more natural reading of FWS’s statements—and one that comports with FWS’s past practice—is that the agency recognizes that the MBTA lacks a programmatic framework for issuing incidental take permits. To be sure, a comprehensive regulation governing incidental take would be preferable. It could set forth uniform criteria for issuing permits, thereby offering predictability for the regulated and environmental communities.4 But the fact that there exists a better way to authorize incidental take does not mean that it is the only lawful way of doing so. Neither the majority nor CBD provides a persuasive explanation for why § 21.27 does not support case-by-case issuance of permits authorizing incidental take.5
Undeterred, CBD takes aim at yet another non-MBTA regulation. This one— the so-called “No Surprises Rule”—implements the HCP provision of the ESA. See 63 Fed. Reg. at 8862-63. The rule explains that an ESA § 10 incidental take permit, issued in conjunction with an HCP, may “double” as a special purpose permit under the MBTA for ESA-listed species. FWS explains that issuing an ESA § 10 permit in lieu of an MBTA § 21.27 special purpose permit is appropriate because the ESA is more species-protective than the MBTA. Id. For example, HCPs require an “operating conservation program designed to conserve the species and minimize and mitigate the impacts of take of the listed species of migratory birds to the maximum extent practicable.” Id. at 8863. CBD extracts from this statement the conclusion that special purpose permits may not be used to cover incidental take of non-ESA-listed species because such species will not enjoy the superior protections of the ESA.
OBD’s reasoning founders on a logical fallacy. The No Surprises Rule provides that, because an ESA take permit comes with greater protections than-an MBTA permit, a party need not also apply for an MBTA permit: the latter is subsumed un-. der the former. See id, at 8862-63. But that does not mean that ESA-level protections áre necessary to authorize take under the MBTA. Put another way, the No Surprises Rule says nothing about whether it is appropriate to issue a special purpose-permit for incidental take under the MBTA for non-ESA-listed species.6
By -analogy, consider a hypotheticalstate’s labeling requirements for perishable foodstuffs. The default regulation for all perishable- foods requires the use-by date to be no more’ than thirty days from the sell-by date. But certain perishable foods-are on a “highly perishable” list, and are subject to stricter regulations requiring the affixed use-by date to be no more than a week from the sell-by date. Now consider a particular perishable food that is not subject to the stricter regulations because it is not on-the applicable list. Does this mean it is not governed by the laxer default rule? Not at all. Yet that is CBD’s logic here: that because the ESA’s heightened protections apply to some migratory birds, other non-ESA. birds are not subject to the MBTA’s take provision. In fact, nothing about FWS’s incidental take policy toward ESA-listed migratory birds forecloses the agency from issuing incidental take permits for non-ESA-listed migratory birds.
B.
While. FWS’s issuance of the shallow-set fishery incidental take permit reflects its considered judgment and is consistent with its past practice, we may still .be compelled to withhold deference if its interpretation of § 21.27 is “plainly erroneous or inconsistent with the regulation.” Christopher, 132 S.Ct. at 2166 (internal quotation marks omitted).- The majority relies on this rationale in concluding that we should not afford FWS’s action Auer, deference, but its reasoning is based on flawed logic and a misinterpretation of the MBTA. .
1.
The majority claims that the “special purpose activity]” exception to the general ban on permitting take does not apply here because fishing lacks an “articulable special purpose.” What qualifies a purpose as “special”? The majority never quite answers this question, except to obliquely note that “special purpose” must be read “in the context of the regulation’s other requirements .... ” Those requirements are, according to the majority, that the activity authorized by the permit “relate[ ] to migratory birds,” be paired with a “compelling justification,” and have a conservation purpose.7 But the majority never explains what it means to “relate[ ] to migratory birds,” except to posit a proposition in the negative—namely, that not all activities that risk killing migratory birds “relate[ ] to those birds,” Landowners, environmental practitioners, and FWS will be hard-pressed to decipher this delphic explanation. Do some activities that do not have as their purpose the conservation of migratory birds “relate to those birds”? Which ones? And how do we know?
The Auer inquiry is more straightforward. We consider the agency’s interpretation relative to the regulation and the governing statute. Marsh, 869 F.3d at 1116— 17. We must assure ourselves that the agency has fairly construed its own .regulation, while also keeping one eye trained on Congress’ intent. Id. To that end, “‘[we] need not accept an agency’s interpretation of its own regulations if that interpretation is ... inconsistent with the statute under which the regulations were promulgated.’ ” Id. at 1117 (quoting Mines v. Sullivan, 981 F.2d 1068, 1070 (9th Cir. 1992)).
My analysis proceeds as follows: I disaggregate § 21.27 into its relevant textual parts, consider each part against the regulation’s broader structure and context, and then assess FWS’s interpretation against the MBTA.
• ‘‘Permits may be issued for special purpose activities ... which are otherwise outside the scope of the standard form permits of this part.” The regulation does not define “special purpose activit[y].” It is also a regulatory term of art that is not susceptible to interpretation by reference to dictionary definitions. ■
Deploying a wider net, we expand our analysis to the regulation’s structure and context. The latter part of the sentence is instructive. It indicates, that a “special purpose activit[y]” is one that is not covered by an expressly identified permitting scheme. Contrary to CBD’s assertion, nothing in the context of the regulation indicates that to be “special” an activity’s purpose must be directed at migratory birds.8 See Klem, 208 F.3d at 1092.
• Special purpose permits must be “related to migratory birds ....” The term “relate” has several dictionary definitions (an inauspicious start for the majority), including, as is pertinent here: “[t]o refer to,” “[t]o have reference to,” “[t]o have some connection with; to stand in relation to,” or “[t]o connect, to link; to establish a relation between.” Oxford English Dictionary (3d ed. 2009) (goo.gl/grzBqC) (last accessed Dec. 8, 2017). Whether the first two definitions could flex to embrace an activity whose purpose is not directed at migratory birds is debatable. But we need not parse'those definitions because the last two plainly do: an activity like commercial fishing indisputably has “some connection with” migratory birds.
•An applicant for a special purpose permit must “make[] a sufficient showing of benefit to the migratory bird resource, important research reasons, reasons of human concern for individual birds, or other compelling justification.” FWS invoked the “other compelling justification” category as the regulatory hook for issuing the Permit. FWS discerned a “compelling justification” in its determination that the Permit would “provide a[n economic] net benefit to the Nation” and would “serve[ ] as a benchmark internationally for employing effective seabird mitigation techniques and serves as an example of responsible conservation practices by a fishery.”
The majority concludes that FWS’s rationale is inadequate, observing that FWS fails to “read the ‘special purpose’ provision in the context of the regulation’s other requirements that, taken together, fail to turn § 21.27 into a general incidental take exception.”9 The problem for CBD and the majority, however, is that nothing in § 21.27 suggests—let alone requires—that all special purpose activities must have as their objective migratory bird conservation to satisfy the “compelling justification” standard. In fact, § 21.27’s text reveals just the opposite. The first eligible category is for activities that provide a “benefit to the migratory bird resource.” Thus, one type of permit is for an activity that is directed at bird conservation. But , another listed category—“important research reasons”—includes not even a gloss of conservation intent. Nor does anything in § 21.27 indicate that a characteristic of the first stand-alone category—“benefit to the migratory bird resource”—modifies all those that follow. Rather, the most natural reading is that special purpose permits are appropriate for activities that are either directed at bird conservation or at other activities that may or may not have a conservation purpose—e.g., scientific research.
Lest there be any doubt, the immediately following subsection makes clear that permits may be issued for non-conservation-related purposes. Section 21.27(a) describes the criteria for issuing a special purpose permit. See 50 C.F.R. § 21.27(a). It explains that such a permit “is required before any person may sell, purchase, or barter captive-bred, migratory game birds ....” Id. Nothing in this subsection suggests that selling, purchasing, or bartering birds serves the purpose of conserving those birds. Nor do those terms have an inherent conservation-oriented meaning— quite the opposite.10
In sum, the catch-all category “other compelling justification” is not limited to activities whose purpose is conserving migratory birds. And the majority provides no other limiting condition, except to warn against transforming § 21.27 into a “general incidental take exception.” But no party argues that § 21.27 grants FWS a roving license to permit incidental take whenever it chooses. The question is, instead: where the agency’s interpretation is not irreconcilable with the regulation’s text and reflects the agency’s “considered judgment” (i.e., it is consistent with past practice), who gets to decide, the courts or the agency? Auer provides the answer: we defer to the agency in which Congress has vested regulatory authority to craft policy based on its expert judgment. See Christopher, 132 S.Ct. at 2166-67 (internal quotation marks omitted). Accordingly, I conclude that FWS’s interpretation of “other compelling justification” as including economic benefits and the benefit of teaching other nations good conservation techniques is not “plainly erroneous or inconsistent with the regulation.” Id. at 2166 (internal quotation marks omitted).
2.
The Permit also comports with the MBTA’s conservation purpose. The majority is correct that in passing the MBTA Congress sought to promote migratory bird conservation.11 But the statute also expressly provides for non-conservation-related take of migratory birds. As is relevant here, the MBTA allows FWS to consider economic factors in determining whether to permit, among other things, the taking, killing, possessing, or sale of migratory birds or their parts. 16 U.S.C. § 704(a). Stated in full, § 704(a) provides that:
Subject to the provisions and in order to carry out the purposes of the [migratory bird treaty] conventions ... the [FWS] is authorized and directed, from time to time, having due regard to the zones of temperature and to the distribution, abundance, economic value, breeding habits, and times and lines of migratory flight of such birds, to determine when, to what extent, if at all, and by what means, it is compatible with the terms of the conventions to allow hunting, taking, capture, killing, possession, sale, purchase shipment, transportation, carriage, or export of any such bird, or any part, nest, or egg thereof, and to adopt suitable regulations permitting and governing the same ....
Id. (emphasis added).
But how—the reader may ask—can we reconcile the statute’s conservation-oriented focus with its provisions allowing for the killing of migratory birds? One way is to interpret § 704(a) as permitting bird deaths—by way of hunting, incidental take, or other means—to the extent that doing so does not threaten the overall conservation of migratory birds. Indeed, we would not be the first court to adopt this interpretation. See Humane Soc’y v. Watt, 551 F.Supp. 1310, 1319 (D.D.C. 1982), aff'd, 713 F.2d 865 (D.C. Cir. 1983) (“It does not necessarily follow from the MBTA’s evident purposes of conservation that the statute creates a presumption against hunting .... ”).
The Permit is consistent with this accommodation of competing statutory directives: it allows for the take of migratory birds when paired with measures designed to minimize such take. Neither CBD nor the majority contends that, if such measures are followed, the MBTA’s broad goal of conserving migratory birds is threatened,
3.
The majority has one lure left in its tackle box, but I decline to take the bait. The majority suggests that because the MBTA generally prohibits take, a presumption attaches against reading § 21.27 as authorizing incidental take. The majority reasons that “although § 21.27 is intended to allow the FWS to authorize activities not otherwise permitted by the regulations, it is still a narrow exception to the MBTA’s general prohibition on killing migratory birds.”
While it is true that the MBTA generally prohibits taking migratory birds, the majority’s observation is a red herring because the statute and regulations provide for numerous exceptions to the general rule.12 The pertinent question turns on the scope of the exception to the. prohibition, not the existence of the general prohibition in the first place. As discussed, § 21.27 is ambiguous and accommodates FWS’s view that the Permit supports a “special purpose activit[y]” that is anchored in .a “eom-pelling.justification.”
[[Image here]]
Because issuing the Permit follows FWS’s past practice, is not plainly erroneous or inconsistent with § 21.27, and comports with the MBTA’s conservation-oriented purpose, I would hold it to be a lawful exercise- of FWS’s authority.
II.
The majority also errs in rejecting NMFS’s loggerhead turtle BiOp as arbitrary and capricious. The majority’s analysis rests on a misapprehension of both binding case law and the administrative record in this case.'
‘ Section 7(a)(2) of the ESA requires all federal agencies to ensure that any discretionary “action” they authorize, fund, or implement does not “jeopardize the continued existence” of an ESA-listed species. 16 U.S.C. § 1536(a)(2); 50 C.F.R. § 402.03.13
To “jeopardize” means “to engage in an action that reasonably would be expected, directly or indirectly, to reduce appreciably the likelihood of both the survival and recovery of a listed species in the wild by reducing the reproduction, numbers, or distribution of that species.” 50 C.F.R. § 402.02. Put another way, “[t]o ‘jeopardize’—the action ESA prohibits—means to ‘expose to loss or injury’ or,to ‘imperil.’” NWF, 524 F.3d at 930, As we have previously explained,
[ejither of these [terms] implies causation, and thus some new risk of harm. Likewise, the suffix “-ize” in “jeopardize” indicates some active change of status: an agency may not “cause a species to be or to become” in a state of jeopardy or “subject a species to” jeopardy
[[Image here]]
[A]n agency may not take action that will tip a species from a state of precarious survival into a state of likely extinction. Likewise, even where baseline conditions already jeopardize a species, an agency may not take action that deepens the jeopardy by causing additional harm.'
Id. (emphasis added).
Our discussion of “j'eopardy” ih NWF must be read in the context of the regulatory standard. To “deepen[ ] the jeopardy” of a species is to “reduce appreciably” a species’ chance at continued survival and recovery. See 50 C.F.R. § 402.02. It cannot—as CBD and the majority suggest— simply mean exacerbating a species’ already “imperiled” existence, no matter how de minimis the impact. An “endangered species” like the loggerhead is, by definition, a “species which is in danger of extinction throughout all of a significant portion of its range.” 16 U.S.C. § 1632(6) (defining “endangered”). If the ESA prohibited any action that worsened—no matter how marginally—a species’ current plight, then it is difficult to conceive of an action that could survive § 7 consultation. That is not the standard: the question is not whether the agency action will negatively affect the species, but whether in doing so it will appreciably reduce its likelihood of survival and recovery. NWF, 524 F.3d at 930 (the operative inquiry is whether the action will “cmse[] some new jeopardy”—i.e., whether it will “tip a species from a state of precarious survival into a state of likely extinction” (emphasis added)).
In NWF, we rejected a BiOp that excluded certain discretionary agency actions from the jeopardy analysis, and which also failed to consider degraded baseline conditions. Id. at 933. The BiOp assessed the effects of dam operations on the Chinook salmon, an ESA-listed species* Id. at 926-26. We faulted. NMFS for departing from its past practice and taking a novel approach in evaluating dam operation impacts. First, NMFS labeled several operations as nondiscretionary, thereby “excluding them from the requisite ESA jeopardy analysis.” Id. at 928-29. Second, NMFS considered only the marginal impact of certain discretionary dam operations in its jeopardy analysis. Id. at 929-30. As concerns the second error, NMFS considered only whether those actions were “ ‘appreciably5 worse than baseline conditions.” Id. at 930. Only if they were would NMFS then conduct a jeopardy analysis. Id.-
We held that-NMFS’s methodology collided with the plain text of the regulations. Section 402.02 explains that an agency action “jeopardizes” a species if it “reduce[s] appreciably the likelihood of’ the species’ “survival and recovery,” when considering the action’s direct, indirect, and cumulative impacts measured against the environmental baseline. 50 C.F.R. §§ 402:02; 402.14(g)(4). NMFS executed a different procedure. Instead of weighing the proposed action in the context of the species’ continued existence, it assessed the action against then-current baseline conditions. See NWF, 524 F.3d at 930.
By way of example, consider a hypothetical scenario in which a residential subdivision is planned for an area inhabited by the endangered arroyo toad. See Rancho Viejo, LLC v. Norton, 334 F.3d 1158, 1160 (D.C. Cir. 2003) (Roberts,' J., dissenting from denial of rehearing en banc). The development requires a federal permit, thereby triggering ESA § 7 consultation. Sierra Club v. Bureau of Land Mgmt., 786 F.3d 1219, 1224 (9th Cir. 2015) (consultation required where a private project is “funded, authorized, or constructed by any federal agency”). The toad is already threatened by the combined effects of climate change and habitat fragmentation.14 Existing developments have substantially reduced the toad’s habitat, and it teeters on the precipice between survival and extinction. The proposed development would reduce the toad’s habitat by an additional 10 percent, which, in the agency’s estimation, does hot amount to an “appreciable” negative impact when compared to the habitat destruction that has already taken place. Thus, under the methodology rejected by this court in NWF, the agency would not have engaged in a jeopardy, analysis.
The pertinent question under NWF, however, is whether the proposed development would have an appreciable impact on the toad’s survival and recovery. Comparing only the marginal impact against already degraded baseline conditions conceals this inquiry. Only by considering the impact of the proposed development “ ‘within the context of other existing human activities that impact the listed species’ ”—i.e., in the context of climate change effects and an already diminished natural habitat—can the agency determine whether the proposed action will consign the toad to a fate of oblivion. See NWF, 524 F.3d at 930. Similarly, the flaw NWF identified in that case was NMFS’s failure to account for the “existing human activity” of dam operations, which impacted the salmon’s survival. See id. at 930-31. The court held that NMFS should have considered the proposed agency action—continued dam operations—together with degraded baseline conditions, instead of against those conditions. See id. at 931.
Turning to the matter before us, NMFS undertook the analysis required by NWF. NMFS considered, among other things, the (i) the current status of the loggerhead sea turtle, (ii) the direct effects of the proposed action on the loggerhead based on climate-based and classical modeling, (iii) the impact of climate change and other cumulative effects, and (iv) whether the proposed action would result in an appreciable reduction in the likelihood of the loggerhead’s survival and recovery. The majority arrives at a contrary conclusion by fixating on the BiOp’s statement that the incremental harm of the proposed action is “the death of a single adult, female loggerhead per year,” which is an “‘extremely small ... level of take from the action.’ ” The majority insists that NMFS ran afoul of NWF by comparing the marginal impact of the fishery “to the much greater harm resulting from factors beyond the fishery.” But NMFS’s consideration of the marginal impact of the fishery did not drive its jeopardy analysis á la NWF. Instead, NMFS considered the “adverse effect on the overlying population ... when considered together with all impacts considered in the Status of the Species, Baseline and Cumulative Effects sectims, including other federally authorized fisheries and foreign fisheries.” NMFS explained that,
[djespite the projected population decline over one generation, we expect the overall population to remain large enough to maintain genetic heterogeneity, broad demographic representation, and successful reproduction. The proposed action will have a small effect on the overall size of the population, and we do not expect it to affect the loggerheads’ ability meet their lifecycle requirements and to retain the potential for recovery.
Thus, unlike in NWF, where NMFS failed to consider direct, indirect, and cumulative effects, here, NMFS incorporated the marginal impact of the fishery in assessing whether the action—combined with baseline conditions—would “tip [the loggerhead] from a state of precarious survival into a state of likely extinction.” See id. at 930. It concluded it would not, and we owe that determination deference.15 See Lands Council II, 629 F.3d at 1074 (“Review under the arbitrary and capricious standard is narrow and we do not substitute our judgment for that of the agency.”) (internal quotation marks omitted)).
The majority also criticizes NMFS for relying on “the conservative nature of its calculations to support the difference between its conclusion and the climate-based model’s results.” As a first matter, the majority does not explain where the model results diverge from NMFS’s finding of no-jeopardy. Nor could it plausibly do so: an analysis of the record data in the BiOp supports NMFS’s conclusion. The climate-based model showed that, in 99.5 percent of the tests, the loggerhead would fall below the quasi-extinction threshold (“QET”) in 25 years without the proposed action. NMFS similarly found that “[w]hen the same model is run with the proposed action, the mortality of 1 adult female, the results are similar with 99.5% to 100% of the runs falling below the QET.”16 Indeed, the model showed that while the proposed action would have a “detectable influence on the loggerhead population, there is no significant difference in the risk of extinction between the default, climate-based trends and the forecast considering the direct effects of the proposed action.” In other words, the risk of extinction is virtually the same whether or not the shallow-set fishery is expanded. Accordingly, NMFS reasonably concluded that the proposed action would not “reduce appreciably the likelihood” of the loggerheads’ “survival and recovery.” See 50 C.F.R. § 402.02.
At any rate, the majority is simply wrong that NMFS relied on its conservative estimates to arrive at its no-jeopardy conclusion. In fact, NMFS relied on (i) the results of the climate change model showing no statistically significant difference in the risk of extinction to the loggerhead with or without the proposed agency action; and (ii) a “qualitative analysis” reflecting that the loss of one additional female loggerhead per year would still allow the loggerhead population to “remain large enough to maintain genetic heterogeneity, broad demographic representation, and successful reproduction.”17
Accordingly, because NMFS’s path “may reasonably be discerned” and “a reasonable basis exists for its decision,” I would affirm. NMFS’s loggerhead BiOp. Pac. Coast Fed’n of Fishermen’s Ass’ns v. Blank, 693 F.3d 1084, 1091 (9th Cir. 2012) (internal quotation marks and citation omitted); Bowman Transp., Inc. v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) (“[W]e will uphold a decision of-less than ideal clarity if the agency’s path may reasonably be discerned.”).
CONCLUSION
FWS acted within its authority when it issued a special purpose permit to NMFS under the MBTA. Its decision aligns with past practice, is not “plainly erroneous or inconsistent with [50 C.F.R. § 21.27],” and comports with the MBTA’s conservation-oriented' purpose. The majority errs in holding otherwise. Similarly, NMFS’s no-jeopardy finding -for the loggerhead sea turtle is rationally related to the evidence in the record, satisfies its statutory obligation to consider direct, indirect, and cumulative impacts, and is faithful to our decision in NWF. Because we should uphold the MBTA Permit and the loggerhead BiOp, I must respectfully dissent.
. Auer's continued vitality is a matter of considerable debate. Justice Antonin Scalia, the progenitor of the doctrine named after the 1997 case, Auer v. Robbins, which he authored, called for its abolition eighteen years later in Perez v. Mortgage Bankers Association, - U.S. -, 135 S.Ct. 1199, 1213, 191 L.Ed.2d 186 (2015) (Scalia, J„ concurring). He appears to have shared this view with at least two other justices, Justices Samuel Alito and Clarence Thomas. See id. at 1210 (Alito, J., concurring); id. at 1213 (Thomas, J., concurring). See also John C. Eastman, The President's Pen and the Bureaucrat's Fiefdom, 40 Harv. J.L. & Pub. Pol'y 639, 641 (2017). Also, Justice Neil Gorsuch has openly criticized Chevron deference, see Gutierrez-Brizuela v. Lynch, 834 F.3d 1142, 1156 (10th Cir. 2016) (Gorsuch, J., concurring) (citing Marbury v. Madison, 5 U.S. (1 Cranch) 137, 2 L.Ed. 60 (1803))—a less controversial deference doctrine because it provides for a check-and-balance between two branches of government (Congress and the Executive), whereas Auer involves the Executive’s interpretations of its own actions. At any rate, my conclusion that the Permit is a lawful exercise of FWS’s authority does not rely on the continued validity of Auer. Applying traditional tools of statutory construction,, the Permit is lawful agency action because it is consistent with (i) the regulatory text of § 21.27, (ii) § 21.27’s greater context, and (iii) the purposes of both § 21.27 and the MBTA itself.
. See Fish and Wildlife Service and National Marine Fisheries Service,. Habitat Conservation Planning and Incidental Take Permit Processing Handbook App’x 5 (Nov. 4,. 1996) ("1996 HCP Handbook”); see also Fish and Wildlife Service and National Marine Fisheries Service, Habitat Conservation Planning and Incidental Take Permit Processing Handbook 16-9 (Dec. 21, 2016) ('-'2016 HCP Handbook”) ("FWS routinely issues consolidated ESA and [MBTA] permits for ESA-listed-bird speciés.”).
. To be sure, what I articulate as FWS’s past practice does not precisely align with FWS's own description of its policy for issuing special purpose permits, which broadly encompasses "incidental take of migratory birds” pursuant to agency "activities.” Courts are not permitted to make sense of an agency action by supplying a rationale not offered by the agency itself. Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 285-86, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974) (citing SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1760, 91 L.Ed. 1995 (1947)). But my description of the agency’s past practice does not supply a rationale for an otherwise arbitrary and capricious agency action. My observation that FWS’s issuance of the Permit is consistent with FWS’s historical policy simply demonstrates that CBD has not met its burden of showing that FWS has departed from past practice.
. FWS is in the process of drafting a regulation that would do just that, though it appears the process has stalled. See Dep’t of the Interi- or, Migratory Bird Permits; Programmatic Environmental Impact Statement, Notice of Intent, 80 Fed. Reg. 30,032 (May 26, 2015).
. CBD also references statements from a 1996 version of FWS’s Habitat Conservation Handbook. The Handbook describes the process governing HCPs under the ESA. Because the Handbook is, at most, a guidance document, it lacks the force and effect of law. See Perez v. Mortg. Bankers Ass'n, — U.S.—, 135 S.Ct. 1199, 1203-04, 191 L.Ed.2d 186 (2015); see generally 1996 HCP Handbook. And to the extent it is probative of FWS’s ‘‘past practice,” it is of little value because the current Handbook is internally Contradictory. One chapter states that ‘‘[n]on ESA-listed, migratory birds can be covered or otherwise addressed in the HCP and incidental take permit.” 2016 Handbook at 3-28. But another chapter states that "if an MBTA protected species is not ESA-listed, the FWS does not have a way to authorize incidental take." 2016 Handbook at 7-7. An internal contradiction is archetypal evidence of a lack of "considered [agency] judgment,” and so the Handbook's description of FWS's MBTA permitting authority is neither persuasive nor deserving of deference. See Christopher, 132 S.Ct. at 2166.
. CBD offers no reason why the rationale for issuing ESA'§ 10 permits in lieu "of an MBTA § 21.27 permit—that the ESA affords species greater protections—is not equally applicable to standalone § 21,27 permits for non-ESA-listed species, FWS, in its discretion, may require a § 21.27 permittee to implement the same types of conservation measures that are codified under the ESA. FWS effectively did just that with the shallow-set fishery here. Because the fishery incorporates conservation measures that have dramatically reduced seabird bycatch, FWS’s issuance of the Permit is consistent with its rationale for covering migratory birds under ESA § 10.
. See Klem v. County of Santa Clara, 208 F.3d 1085, 1092 (9th Cir. 2000) (‘‘the question ... is whether the Secretary’s • interpretation is justified when considered together with the text of [the regulation], taken in context”); cf. FDA v. Brown & Williamson Tobacco Corp,, 529 U.S. 120, 133, 120 S.Ct, 1291, 146 L.Ed.2d 121 (2000) (noting the "fundamental canon . of statutory construction that the words of a statute must be read in their context and with a view to their place in the overall statutory scheme” (internal quotation marks omitted)).
. CBD asserts that an "ongoing fishing business ... has no ‘special purpose' beyond catching fish.” But this observation,only begs the question; what is a "special purpose”? CBD offers no explanation, except to march out a parade of horribles, warning that if the Permit is allowed to stand then the court will have ushered in a brave new world in which “every activity that happens to somehow harm birds” will qualify for an incidental take permit.
. The majority correctly adheres to the doctrine that "all the words used in a list should be read together and given related meaning when construing a statute or regulation.” Aguayo v. U.S. Bank, 653 F.3d 912, 927 (9th Cir. 2011).
. To be sure, the quoted phrase applies only to captive-bred birds. But the point is that the regulation expressly contemplates issuing special purpose permits for something other than conserving migratory birds.
. See Humane Soc'y of U.S. v. Watt, 551 F.Supp. 1310, 1319 (D.D.C. 1982), aff'd, 713 F.2d 865 (D.C. Cir. 1983) (‘“The United States ... [and] Great Britain ..., being desirous of saving from indiscriminate slaughter and insuring the preservation of such migratory birds as are either useful to men or are harmless, have resolved to adopt some uniform system of protection which shall effectively accomplish such objects ,...’”) (quoting 39 Stat. 1702 (Convention on the Protection of Migratory Birds) incorporated by reference into the MBTA at 16 U.S.C. § 703(a)).
. See 16 U.S.C. § 703(a) ("except as permitted by regulations ... it shall be unlawful ... to-... take .,.. any migratory bird." (emphasis added)); 50 C.F.R, §§ 21.13 (taking certain mallard ducks); 21.15 (incidental take for military readiness activities); 21.23 (taking fqr scientific research); 21.24 (taking for taxidermy); 21.25 ("dispos[ing]” of migratory waterfowl); 21.26 (killing Canada geese); 21.27 ("special purpose activities” not covered by other permits); 21.29 (taking for raptors).
. “Section 7 . , applies] to all actions in which there is discretionary Federal involvement or control.” 50 C.F.R. § 402.03. -
. See U.S. Fish and Wildlife Service, Arroyo Toad 5-Year Review: Summary and Evaluation 10, 16 (Aug.-2009).
. NMFS included in its analysis an assessment of “spillover” effects—i.e., the impact of the expanded domestic shallow-set fishery on foreign fisheries. NMFS found that without the expansion, foreign fisheries would move in and occupy the area. And because the implicated foreign nations generally have weaker environmental laws than does the United States, NMFS concluded "with reasonable certainty, that [under the agency action] there will be a reduction of [loggerhead and leatherback sea turtle] mortalities as a result of the spillover effect.” NMFS estimated the reduction to be “11 fewer interactions in the central and north Pacific ... or four fewer [loggerhead and leatherback sea turtle] mortalities.”
This data amply supports NMFS’s no-jeopardy conclusion. However, NMFS did not incorporate its findings into the jeopardy analysis because it concluded that “data on foreign fisheries is likely incomplete or inaccurate.” Thus, while the "spillover” effects data is compelling, I—like the agency—do not rely on it in assessing the reasonableness of NMFS’s ultimate determination,
. The additional loss of one adult female per annum from the proposed action results in a projected reduction in the overall population of 4 to 11 percent, due to a loss of that single turtle's "reproductive potential” over the course of generations. But, contrary to the majority’s assessment, NMFS did not credit this numerical loss because it had low confidence in the data. NMFS noted that the estimated loss does "not account for the high mortality rate expected of these hatchlings from other sources, including climate-based threats.” In other words, the reduction due to a loss of reproductive potential is significantly overstated.
. NMFS’s use of conservative data inputs is relevant not because it is the sole basis for its no-jeopardy conclusion (as discussed, it isn’t), but because it reflects the reasonableness of its findings. For example, NMFS considered the lost "reproductive potential” of all "unborn hatchlings," even though hatchlings have a "high mortality rate.” It also assumed that the shallow-set fishery would immediately operate at 5,500 sets every year, even though the increase is likely to be gradual over time. And its climate model did not incorporate the results of anticipated indirect effects—namely, beneficial "spillover” effects—of the domestic fishery!s displacement of international fisheries.
As discussed, NMFS’s no-jeopardy conclusion is not unreasonable even without considering the conservative nature of its inputs. Recognizing that those inputs are more con-1 servative than actual conditions warrant therefore only weakens the majority’s erroneous conclusion that NMFS's action is arbitrary and capricious. See George v. Bay Area Rapid Transit, 577 F.3d 1005, 1011 (9th Cir. 2009) ("The party challenging an agency’s action as arbitrary and capricious bears the burden of proof ,...”).