In Harm’s Way

Agency’s efforts to repeal harm definition for the ESA carries significant legal risks

The Trump Administration has finalized a rule that would eliminate the definition of harm from the regulations implementing the Endangered Species Act (ESA).  Harm is one component of the definition of take in the ESA, and Section 9 of the ESA prohibits take of members of endangered species.  As I noted last spring, when the rule was first proposed, the upshot of this change would be to eliminate protections of ESA-listed species from habitat modification under Section 9 of the ESA.  While other habitat protections for listed species under Section 7 would continue to apply, Section 7 only covers federal government actions, while Section 9 covers actions by private, state, local, and tribal actors.

The final rule, to my mind, appears to take the most legally precarious approach to repealing the definition of harm.  The proposed rule argued that the best interpretation of the term “take” (and thus harm) in the ESA was articulated in the dissent in the Supreme Court case Sweet Home.  The majority in Sweet Home upheld the definition of harm in the regulations, concluding that it was a reasonable interpretation of the term under Chevron.  The dissent argued that take should be limited to intentional acts to harm listed species, such as hunting, and therefore could not include habitat destruction (as under the regulatory definition of harm).  The proposed rule argued that in the wake of the Supreme Court’s overruling of Chevron in Loper Bright, the agency (the U.S. Fish and Wildlife Service) should adopt the dissent’s interpretation of “take” because this is the “best” interpretation – the Court in Loper Bright concluded that Chevron was improper in part because all statutes have a “best” interpretation.  And therefore, the agency concluded that the regulatory definition of harm was illegal.

But there is a catch.  In Loper Bright, the Supreme Court also held that all prior decisions that had upheld agency interpretations of law under Chevron as reasonable were still good precedent under stare decisis.  And that would seem to include the Court’s decision in Sweet Home to uphold the meaning of “harm” in the agency’s regulations.

As I noted in my earlier post, under Loper Bright it appears that the agency could make a choice between keeping the harm definition under the regulations or adopting the dissent’s interpretation (and hoping that a friendly Supreme Court would conclude that the Sweet Home dissent did indeed articulate the “best” meaning of “take under the ESA, and upheld the agency’s change in position).

But that is not what the agency did.  Instead, it concludes that it has a non-discretionary duty to adopt the “best” interpretation of the statute under Loper Bright, and that best interpretation is the dissent in Sweet Home, meaning that the agency must repeal the harm definition from its regulations.  The agency’s claim that it has a mandatory duty to adopt the best interpretation led it to a number of conclusions.

First, the agency argues that it need not consider the reliance interests of parties that have built business models, or otherwise relied, on the harm regulation.  Most importantly for these purposes, there are a lot of businesses that focus on providing mitigation to harm to endangered species habitat – eliminate the Section 9 prohibition on habitat modification for endangered species, and that revenue stream is gone.  The Supreme Court has regularly noted that reliance interests should be considered by agencies when they change regulations.  But if the agency has no choice, then reliance interests have no bearing on the analysis.  The agency also relies on a sentence in Dep’t of Homeland Security v. Regents of Univ. of California (the successful challenge to the first Trump Administration’s repeal of DACA) that reliance interests might weigh less when the agency believes its prior actions were illegal – but that language is really undeveloped.

Second, arguing that it had a duty to change its interpretation of the statute means the agency could short-circuit any policy analysis – all it has to say is we think the law is X, and we have to do it.  That avoids consideration of whether the change would have significant policy implications under the ESA.  It is possible that, without the mandatory duty, the failure to consider those policy implications might be contrary to the arbitrary and capricious standard under the Administrative Procedure Act.  (It is possible that an agency’s purely legal interpretation of a statute is not subject to that standard, but I think that’s an open question.  In particular, in Regents the Supreme Court held that policy considerations are relevant to determine how an agency should respond to a finding of illegality, and that in turn is subject to arbitrary and capricious review.  At the very least, the agency could articulate alternative transition periods for its change in position on legality, and in fact the agency did state it was keeping all existing permits that had been issued under the harm definition.)

Third, arguing that it has a duty to act means that the agency can claim the change is nondiscretionary, and thus exempt from review under the National Environmental Policy Act (NEPA).  (The agency also relies on two categorical exemptions under NEPA, which I noted in my prior analysis probably do not apply.  And the agency claims NEPA does not apply to legal interpretations by agencies, arguing that otherwise NEPA review would be required for agency legal briefs in courts.  That last argument, however, does not address the distinction between briefs (which are arguments to a court) and regulations (which bind the public as a whole) – and indeed, the agency’s argument would allow agencies to simply claim a regulation has a legal interpretation component to avoid any NEPA review of regulations at all.)

So the agency’s approach got it some benefits in terms of its analysis.  But this approach is also high-risk.  For the agency to have a nondiscretionary to adopt the “best” position under Loper Bright, it really seems like it cannot maintain the prior regulatory interpretation as a matter of law.  But the Supreme Court in Loper Bright said that judicial precedent relying on Chevron was still good law.  That seems to mean that the agency could continue to rely on the Sweet Home decision upholding the harm definition.  And thus the agency has no mandatory duty to change its legal interpretation.

And if that is the case, all the arguments above probably fail.  The agency cannot just dismiss reliance interests on the grounds that it has no choice but to act.  And since its current position is not illegal, it’s not even clear it can minimize those reliance interests under Regents.

Generally, courts only uphold agency decisions on the grounds which the agency articulates when it makes its decision – so the agency cannot try to point to an alternative basis for its position.  (There is some discussion in Regents that perhaps if the agency faces a dichotomous choice, the failure to articulate alternative reasonings is irrelevant.  I don’t the agency faced a dichotomous choice based on questions about how to transition to the new regulation, as I discuss above.)

And the agency’s arguments that NEPA does not apply because there is a non-discretionary duty for the agency to comply with do not work either – the agency had a choice, and made one.

Why did the agency take this more legally perilous course?  One possible reason, flagged by J.B. Ruhl at Vanderbilt, is that the agency was surprised that so many businesses that have developed based on ESA mitigation banking would argue that they had significant reliance interests in the harm definition.  The agency realized that it would have to address those reliance interests if it phrased its decision as discretionary, so it chose to claim that it has no discretion but to change its decision.

Also, there is a potentially high reward here for the Trump Administration – the agency can try to push this case to the Supreme Court and get the Court to overrule Sweet Home (and potentially more broadly its language in Loper Bright preserving all past Chevron decisions).  And that would mean that a future administration could not restore the harm definition (whereas if there is a choice, and Sweet Home remains good law, future administrations could always go back to the prior definition if they wished).  I would be very skeptical this would succeed though – traditionally the Court has set a very high bar for overturning interpretations of statutes, both on reliance grounds, and on the grounds that Congress can always change the law if it doesn’t like the Court’s interpretation.  And it would also require the Court to take back statements it made just two years ago in Loper Bright.

Of course, the beauty and curse of a common law system is that courts ultimately can do what they want, especially the Supreme Court.  So any prediction is full of uncertainty…

Chevron deference, Endangered Species Act, harm, Loper Bright, Section 9, Trump Administration