Isn’t Textual Ambiguity a “Yes” or “No” Question?
In Kisor v. Wilkie, the Supreme Court purportedly upheld the Auer doctrine of judicial respect for an agency’s interpretation of its own regulation. At the same time, however, the court “reinforced and expanded on” the doctrine’s limits. To this end, Justice Kagan’s opinion did two things. First, Kisor imported a “Step One”: henceforth, lower courts should give deference only if they identify a textual ambiguity after employing “all the standard tools of interpretation.” Second, even if the regulation is “genuinely” ambiguous, “not all reasonable agency constructions are entitled to deference.” To warrant binding judicial respect, a reasoned agency interpretation also must survive a series of “markers” or conditions.
To be sure, I expected that Kisor would dramatically change the Auer doctrine (unlike some others). That said, I thought that Kagan’s “markers” would prove to be the most consequential limitation—particularly the requirement for interpretive consistency. I reasoned that if a regulatory provision is of any importance, and if it has existed for longer than one presidential administration, then there are vanishingly small odds that the agency’s interpretation has remained consistent. Thus aware, I assumed that Kisor’s call for consistency would provide the greatest check on Auer deference.
Over the last few months, this blog has monitored how the Ninth Circuit is receiving Kisor. To date, the circuit has adopted a forceful take on Kisor, just as I’d expected. But the circuit is doing so in a manner that is far different from what I’d foreseen. Instead of Kagan’s “markers,” the Ninth Circuit has focused on Kisor’s requirement for robust textual scrutiny at “Step One.”
Previously, I reported on I.R.S. v. Amazon. In that case, a Ninth Circuit panel conceded that the agency had set forth a reasoned interpretation of an ambiguous regulatory text. Prior to Kisor, this combination of factors (an ambiguous text & a reasoned interpretation) undoutbedtly would have triggered the Auer doctrine. And yet the Amazon panel denied deference. After consulting the rule’s purpose and history, the court decided that the regulation was merely ambiguous when “viewed in isolation,” but not “genuinely ambiguous” when considered in context. In this manner the court found sufficient clarity to “leave little room for the Commissioner’s proffered meaning.”
On September 11th, the Ninth Circuit did the same thing in Secretary of Labor v. Seward Ship’s Drydock. At issue was an occupational health rule promulgated by the Labor Department. In applying the Kisor framework, the panel investigated whether the text was “truly ambiguous.” After considering the rule’s “text, structure, history, and purpose,” the court concluded that the provision was “sufficiently clear” and therefore not “genuinely ambiguous.”
Frankly, I’m torn over how Kisor is developing in the Ninth Circuit. On the one hand, I was among the Auer doctrine’s critics, so I appreciate its diminishment. On the other hand, there’s something troubling about the Ninth Circuit’s judicial methodology. It seems to me that the presence of a textual ambiguity is a binary question that affords only a “yes” or “no” answer. By treating textual ambiguity as a range—from merely ambiguous to “truly” ambiguous—I fear the court is inviting confusion. In particular, I fear that agencies won’t be able to discern when the Ninth Circuit issues a definitive regulatory interpretation.
Here’s why it matters. If, in any given controversy over a regulaton’s meaning, the court finds that a provision is unambiguous, then agencies have no room for interstitial policymaking. But if the court finds that the language is ambiguous, then agencies would have the discretion to introduce a new interpretation in the future. By obfuscating its ambiguity analysis, the Ninth Circuit also clouds these policy options for agencies.
To avoid this sort of uncertainty, I’d prefer that the Ninth Circuit reconceptualized at least part of its search for “genuine” ambiguity within the Kisor framework. To the extent that a regulation’s “purpose and history” inform judicial review, these factors tend to speak more loudly about a regulatory interpretation’s reasonableness than they do about whether the underlying rule is ambiguous. Put differently, an investigation of a regulation’s “purpose and history” usually fits best in Step Two rather than Step One.
Administrative‐like Procedures in Article III Regulatory Regimes
When it comes to congressional grants of policymaking power across branches of government, administrative law focuses on legislative delegations to regulatory agencies. Yet lawmakers also provide courts the initiative to effectuate regulatory policy. In this context, the “delegation” takes the form of a statutory cause of action to vindicate federally‐created rights. Depending on the regime, these Article III regulatory programs establish original jurisdiction in federal courts to hear suits brought by the government, private citizens, or both. Such regulatory regimes include the Americans with Disabilities Act, the Fair Credit Reporting Act, the Stored Communications Act, and many more.
Each month, as I review cases for these posts, I’ve become increasingly fascinated by the diverse administrative‐like procedures associated with these judicial regimes.
Consider, for example, the distinct character of the government’s friend‐of‐the‐court brief, on which the Ninth Circuit leaned heavily last month in Rudel v. Hawai’I Management Alliance Association. Typically, when an executive branch agency undergoes judicial review, the agency is represented by the Justice Department in court. In Rudel, however, the Labor Department’s Solicitor General directly submitted an amicus brief to inform the court about an Article III regime (Employee Retirement Income Security Act). Labor’s amicus, moreover, was in support of neither party. To my eyes, Labor’s brief in Rudel—and amicus briefs as a general matter—draws an obvious parallel to the role of public comments in the § 553 rulemaking process.
There are other ready comparisons between Article III regimes and aspects of “normal” administrative procedure. In a prior post about a Federal Railroad Safety Act case, I compared jury instructions to “non‐legislative” rules issued by regulatory agencies.
Another obvious parallel is that both courts and agencies engage in interstitial policymaking via statutory interpretation. Of course, when a court respects an agency’s reasonable statutory interpretation of a textual ambiguity, we call it “Chevron deference.” When a court does the same thing within an Article III regime, we call it an enclave of “federal common law.”
Ninth Circuit Demonstrates Karl Llewellyn’s Famous Argument about Dueling Statutory Canons
Almost 60 years ago, Prof. Karl Llewellyn published a celebrated law review arguing that for each canon of statutory construction, there is an equal canon pointing in the opposite direction. Llewellyn’s article came to mind as I encountered the Ninth Circuit’s recent canonical inconsistency.
In Pitt River Tribe v. BLM, which the Ninth Circuit issued on September 19th, a three‐judge panel addressed a controversy over the duration of leases pursuant to the Geothermal Steam Act. Under the statute, the Bureau of Land Management is responsible for issuing leases for the development of geothermal power on public lands. The statute, moreover, allows the agency to manage multiple (individual) leases through a “unit plan,” in order to engender efficient production.
Durations for individual leases and “unit plans” are set forth in discrete subsections of the statute. Pitt River Tribe involved the interaction of these subsections. In a nutshell, the agency sought to treat the two subsections as identical and thereby achieve a forty‐year extension of 26 leases within the Glass Mountain Unit Plan in California.
The court rejected the BLM’s argument. In large part, the court based its reading on the “well‐established canon of statutory interpretation that the use of different words or terms within a statute demonstrates that Congress intended to convey a different meaning for those words.” After observing that only one of the subsections referred to “unit plans,” the panel concluded that “when Congress includes particular language in one section of a statute but omits it in another section of the same Act, it is generally presumed that Congress acts intentionally and purposely in the disparate inclusion or exclusion.”
Taken on its own, the Ninth Circuit’s order in Pitt River Tribe isn’t necessarily objectionable. But it becomes positively problematic on comparison to Gonzalez Romo v. Barr, an immigration controversy on which the Ninth Circuit had delivered an opinion only a month prior.
Ms. Gonzalez Romo faced deportation under 8 U.S.C. § 1182(a)(2), which states that any “alien convicted of … a crime involving moral turpitude … or an attempt or conspiracy to commit such a crime … is [removable].” Previously, the Board of Immigration Appeals claimed that the phrase “attempt or conspiracy to commit such a crime” serves to exclude solicitation crimes from being CIMTs. The agency based its reasoning on the contrast between §§ 1182(a)(2) and 1227(a)(2). For deportation purposes, the two statutory provisions are identical, except that § 1227(a)(2) does not include the phrase “attempt or conspiracy to commit such a crime.” Based on this textual discrepancy, the Board interpreted these provisions differently. Specifically, the Board previously indicated that crimes of solicitation qualify as deportable offenses under § 1227, but not under § 1182.
Ms. Gonzalez Romo was convicted of soliciting marijuana. Before the Board, therefore, she pointed to the agency’s prior interpretation of § 1182(a)(2), which, again, had excluded solicitation crimes. During Ms. Gonzalez Romo’s removal proceedings, however, the Board claimed that its prior interpretation was mere dicta, and the agency reversed course in a precedential order. The Board’s new interpretation is that § 1227(a)(2) and § 1182(a)(2) actually carry identical meanings for deportation, notwithstanding their blatant textual discrepancy. Although the panel conceded that “redundancy and surplusage are not always desirable,” it nonetheless concluded that “it is not unusual to see redundancy occur in clarifying enactments, which do run the risk of introducing other problems while focusing on the task of making sure that a particular problem does not arise.”
The upshot is that these two cases were decided about a month apart, and they reflect an inconsistent, even contradictory, use of statutory canons. Both cases pertained to near‐identical statutory provisions that differed by a single phrase. In one case, the Ninth Circuit relied on the canon of construction that every word and clause should be given effect. And in the other, the panel employed the principle that words and clauses may be rejected as surplusage when they were inadvertently inserted or repugnant to the rest of the statute. If the canons had flipped, so would’ve the outcomes.
I’d say the contrast between these two cases would make Prof. Llewellyn smile, but wishy‐washy justice is no laughing matter.