The recent deadline for comments in the FCC’s latest NPRM relating to the Open Internet has me thinking about an article I have spent the past two years looking for time to write. Indeed, I had hoped that Evan Bernick’s recent piece on the constitutionality of agency fact-finding would give me the push I needed to undertake the project this summer. But one of the more difficult aspects of growing up as an academic, it seems, is learning to accept that some articles are a step too-far removed from one’s research agenda – and letting them go. This is one of a couple of posts I hope to make this summer as part of letting go of a couple of idea in a forum where, perhaps, someone else will see and pick them up.
The question that I would like to explore is whether it is problematic for courts to defer to agency interpretations of law that are themselves based on deferred-to factual determinations. Since this is a blog post, I will refer to this phenomenon informally, as double-dawg deference. Were this an article, I may call the phenomenon something more formal-sounding, like multi-level deference or joint fact-law deference. A couple of examples where this has occurred in the context of the FCC are discussed below. I don’t want, however, to dwell on specific examples from the FCC context, however, because I expect that this practice is not unique to the FCC.
Illustrations of Double-Dawg Deference
One example of this comes from the DC Circuit’s 2016 decision upholding the FCC’s 2015 Open Internet Order. In the 2015 Order, the FCC had altered is prior interpretation of the Communications Act, reclassifying broadband Internet access services as Title II Telecommunications Services from their longstanding classification as Title I Information Services (as classification that the FCC had successfully defended before the Supreme Court in Brand-X). One issue before the DC Circuit was whether the Commission’s decision to change its prior interpretation subjected its statutory construction to heightened review. In Fox, the Supreme Court said that such changes do not require heightened review except in certain circumstances, such as when the prior interpretation has engendered significant reliance interests.
In their challenge to the 2015 Order, Petitioners, citing Fox, “argue[d] that the Commission ‘could not rationally abandon its prior policy without accounting for reliance interests that its prior policy engendered.’” The court’s majority opinion found that “The Commission, however, did not fail to ‘account’ for reliance interests. Quite to the contrary, it expressly considered the claims of reliance and found that ‘the regulatory status of broadband Internet access service appears to have, at most, an indirect effect (along with many other factors) on investment.’” Importantly, the Commission’s determination that there were no reliance interests engendered by its prior policy was adopted precisely to address the concerns raised by Fox.
This conclusion – both by the Commission and the court – rankled the sensibilities of many economically-minded commentators. This included Judge Williams, whose “partial dissent finds the Commission’s explanation insufficient and concludes that it failed ‘to make a serious assessment of [Petitioners’] reliance.” This is a polite way of saying that they Commission’s conclusion runs counter to all economic sense so should be subject to substantial scrutiny. The majority, however, was less concerned by the economic (in)coherence of the Commission’s position: “Contrary to our colleague, “[w]e see no reason to second guess these factual determinations, since the court properly defers to policy determinations invoking the [agency’s] expertise in evaluating complex market conditions.”
Because the court deferred to the Commission’s factual determination – a factual determination that was designed to avoid heightened scrutiny of its changed legal interpretation – the court then also deferred to the Commission’s changed legal determination.
A different, but related, situation comes up in the context of the Commission’s authority under Section 706 of the Telecommunications Act, which requires the Commission to ensure timely availability of “advanced telecommunications capabilities” to all Americans and the legal meaning of factual measurement of such capabilities. The Commission has interpreted the term largely to mean broadband Internet service, and has defined broadband Internet in a way that includes only a subset of high-speed Internet services (viz., 25 mbps download speeds and faster, which excludes a significant portion of the market that was technologically capped at 16 mbps or 20 mbps speeds). This definition then fed back into the Commission’s 2015 Open Internet Order as a basis for finding that broadband Internet services were not being timely deployed, justifying affirmative Commission authority under Section 706 to take regulatory action.
This is not an example of “deference” it the traditional judicial review sense, but it is another example where unreviewed (and potentially unreviewable – it is not clear who would have had standing to challenge the Commission’s definition of broadband since that order did not, itself, create or undertake any regulatory action) agency determinations are subsequently used to justify agency actions that are afforded deference (in the case of the Open Internet Order, a finding that the Commission’s use of Section 706 as a basis for affirmative regulation was a permissible construction of the statute).
My colleagues who focus on telecommunications issues surely can find issues with my characterizations of both of these examples. I elide over a paragraph and a half of discussion from the DC Circuit Open Internet opinion, for instance, in which the majority does discuss the Commission’s determination that there were no reliance interests – even though the court expressly says that it defers to the agency’s determination, one could argue that the court is engaging in at least a somewhat probing review of the agency’s determination. And in the example of Section 706, the issue is not really of courts deferring to a factual determination but that determination going unchallenged (even if there were no way for such a challenge to be brought outside the context of the subsequent application of those facts). If this were an article-length exploration of these issues I would get into that discussion further. But this is not such an exploration – rather, it is my hope that these examples demonstrate at least a hypothetical concern.
Is Double-Dawg Deference Uncouth?
My general intuition is that double-dawg deference presents a situation somewhat analogous to that of double for-cause protection. As the Court framed the issue in PCAOB: “May the President be restricted in his ability to remove a principal officer, who is in turn restricted in his ability to remove an inferior officer, even though that inferior officer determines the policy and enforces the laws of the United States?” We could reframe this in the language of Article III and deference: “May the courts be restricted in their ability to interpret ambiguous statutes, where permissible constructions of those statutes are in turn constrained by interpretation of ambiguous facts, even though construction of those ambiguous facts determines the permissibility of a given statutory interpretation?” And just as the Court held “that such multilevel protection from removal is contrary to Article II’s vesting of the executive power in the President,” such multilevel protection of interpretation seems contrary to Article III’s vesting of judicial power in the courts.
This line of thinking is premised on an understanding of deference as a compromise between two competing interests – much like our acceptance of for-cause–protected agency heads is premised on a compromise between the need for Officers to be accountable to the President balanced against the need to depend, in some circumstances, for Officers “to maintain an attitude of independence against the [President’s] will.” On the one hand, we have the Judiciary’s constitutional role as arbiter of what the law is. On the other hand, we have the traditional bases for deference: (implied) Congressional intent that agencies resolve ambiguities, that such resolution is based on delegated legislative (not judicial) power, that agencies are both more politically accountable and more expert than courts so are in a better position to exercise such authority, and all of this wrapped up in a great big ball of separation of powers concerns.
Chevron tells us that, in order to ensure our interests in those latter concerns, the Judiciary needs to forego in some circumstances its role as arbiter. By analogy to the Article II context, PCAOB suggests that the Judiciary by cautious not to forego this role to too great an extent.
Another way of thinking about this concern is to analogize double-dawg deference to Auer deference – some of the concerns that many have about Auer deference map reasonably well on double-dawg deference. Just as deferring to agency interpretations of their own regulations creates perverse incentives for agencies to write their regulations ambiguously, deferring to agency interpretations of law based on agency interpretations of fact creates perverse incentives for agencies to interpret facts in ways to support their preferred legal interpretations.
On the other hand, Auer is the law of the land. Agencies interpretations of their own regulations are afforded substantial deference. And this is not PCAOB. The justifications for affording certain Officers independence from the Executive are different from the justifications for affording expert agencies in interpreting legislative ambiguities. And Congress has substantial latitude in deciding how agencies approach fact-finding. For instance, in the example of Section 706 discussed above, Congress decided both to entrust the FCC to report on the state of deployment of advanced telecommunications capabilities and to take action based upon those reports. It could have just as easily tasked another agency – NTIA, for instance – to generate those reports and a separate agency – the FCC – to act upon those reports. The same separation of powers principles that lead the courts to defer to implied Congressional intent that agencies interpret statutory ambiguities may suggest that courts defer to agencies’ exercise of implied delegation of non-reviewable legislative fact-finding.
In the end, I don’t know where I come out on the soundness of double-dawg deference – be that soundness defined in Consititutional, utilitarian, efficiency, or some other normative terms. But I think, at least, that the idea presents an interesting puzzle – one that I believe is important in my own area of substantive law and that I expect comes up in other contexts. Sadly, with that, I think I must let this idea float into the ether. Perhaps, however, it will prompt discussion by others down the line.
Gus Hurwitz is an Assistant Professor of Law and Co-Director of the Space, Cyber and Telecom Law Program at the University of Nebraska College of Law.