« Fair Housing and the Federal Executive | Main | “An Antidemocratic and Largely Foreign Conspiracy” »

Thursday, June 25, 2015

Chevron After King v. Burwell

As Richard already observed today, in King v. Burwell, the Court upheld the government's interpretation of the Affordable Care Act to allow people to get subsidies on healthcare exchanges created by the federal government. Chris Walker has a nice post at JREG discussing the case.    

I agree with Chris that the way the Court reached today's outcome could have broader consequences for administrative law.  I offer some very, tentative off-the-cuff thoughts about the opinion below the fold.

As I mentioned in an earlier post, there were a few different ways that the Court could have sided with the government. First, the Court could have found that the statute favored the government's position by using a somewhat novel application of the "unconstitutional avoidance" doctrine.  The Court could have used that doctrine to reason that Obamacare could not be read to punish the citizens of states that failed to set up health exchanges without raising serious federalism questions under the Constitution.  The Court sometimes uses avoidance canons like this before it decides to defer to an agency (at, what administrative law people call "Chevron Step 1," discussed below). See, e.g., Edward J. DeBartolo Corp. v. Fla. Gulf Coast Bldg. & Constr. Trades Council, 485 U.S. 568, 574-75 (1988); Solid Waste Agency v. U.S. Army Corps of Eng'rs, 531 U.S. 159 (2001).  

Second, the court could have found that the statute  favored the government under cases like PennhurstGregory, and Gonzales, which require Congress to speak unambiguously before imposing new onerous conditions on the states.  The Court could have reasoned that Congress could not have done so in such an awkward way, buried in an obscure subsection of the tax code.  Some commentators, like Mila Sohoni and  Nick Bagley, observed that this doctrine would not raise the same kinds of constitutional doubts about the how Congress can interact with states in the future as the avoidance doctrine described above. The Court would simply be saying that Congress didn't write the law with sufficient clarity.

Third, the Court could have found that the provision favored the government because, even though the text was somewhat awkward and unclear, when read as a whole, the structure and purpose of the law unambiguously favored the government.  This was the government's primary argument.  

Fourth, the Court could have found that the statute was ambiguous, but then deferred to the government's interpretation of the statute under Chevron.  One thoughtful commenter argued that this approach would respect the government's position without raising any new questions, under any doctrine, about how Congress interacts with state government.  

All of those tacks would have been consistent with the way the court is supposed to apply Chevron--determining (1) whether Congress has spoken to the issue using traditional canons of construction, and if not, (2) deferring to the agency's permissible reading of a statute.  But Court didn't do any of those things.  The Court instead said the ACA was ambiguous.  But rather than defer to the government's interpretation of an ambiguous statute under Chevron, it held that courts, not agencies, have the primary role in interpreting statutes that raise questions of “deep economic and political significance.”

When analyzing an agency’s interpretation of a statute, we often apply the two-step framework announced in Chevron. Under that framework, we ask whether the statute is ambiguous and, if so, whether the agency’s interpretation is reasonable. Id., at 842–843. This approach “is premised on the theory that a statute’s ambiguity constitutes an implicit delegation from Congress to the agency to fill in the statutory gaps.”FDA v. Brown & Williamson Tobacco Corp., 529 U. S. 120, 159 (2000). “In extraordinary cases, however, there may be reason to hesitate before concluding that Congress has intended such an implicit delegation.”Ibid.

This is one of those cases. 

This "major questions doctrine"—the presumption that Congress “does not . . . hide elephants in mouseholes”—has been around for a while. See Whitman v. American Trucking (2001), FDA v. Brown & Williamson (2000).  But it does seem unusual for the Court to use it here for two reasons.  

First, the Court relied on this proposition by citing FDA v. Brown & Williamson.  But that case didn't acknowledge or claim that it was side-stepping Chevron.  Rather, Brown & Williamson said that whether or not the statute gave the FDA authority to regulate tobacco had to be considered against Congress' long history of regulating tobacco in other ways.  And when it was considered in that light, Congress clearly could not have given the FDA that power under Chevron's first step. Brown & Williamson ("[W]e find that Congress has directly spoken to the issue here and precluded the FDA’s jurisdiction to regulate tobacco products.")

Second,  I wonder how the Court's latest use of the "major questions" doctrine is supposed to interact with two other court cases: Mead and City of Arlington.  Under Mead, the Supreme Court said it is supposed to consider two things before determining whether or not to apply Chevron: (1) whether Congress delegated power to an agency to interpret a statute and (2) whether the agency exercised its authority to do so. But, as the Solicitor General observed at oral argument, Section 36B(g) of the ACA expressly gives the IRS the specific authority to make any decisions necessary to implement Section 36B (the provision at issue). And the IRS did so through a formal, deliberative process.  The Court never grapples with Mead, but perhaps the court is suggesting that when Congress delegated the IRS interpretive power, it didn't mean to delegate authority to resolve this kind of question.  

That seems to square with Justice Kennedy's concerns at oral argument: that the IRS could not have authority to make this interpretation because the decision involved "billions of dollars" of subsidies.  But I'm not sure how that fits with City of Arlington.  Under, City of ArlingtonChevron is supposed to apply to all decisions--jurisdictional and non-jurisdictional, big and small. The purported reason for this was because it's so difficult and subjective for judges to determine what counts as a big and small question before deciding to defer to an agency.  After all, agencies like the EPA, the FCC, and the FTC often are delegated with power to make billion dollar decisions.  How specific does Congress have to be when it gives agencies authority to regulate?

If anything, the Court's discussion of Chevron, mirrors the kind of rigorous judicial review that appears in Justice Robert's dissent in City of Arlington-- searching provision-by-provision to determine “whether [that] delegation covers the ‘specific provision’ and ‘particular question’ before the court": 

By the same logic, even when Congress provides interpretive authority to a single agency, a court must decide if the ambiguity the agency has purported to interpret with the force of law is one to which the congressional delegation extends. A general delegation to the agency to administer the statute will often suffice to satisfy the court that Congress has delegated interpretive authority over the ambiguity at issue. But if Congress has exempted particular provisions from that authority, that exemption must be respected, and the determination whether Congress has done so is for the courts alone.

(emphasis added).

More broadly, I wonder whether this case reflects growing unease by the Chief Justice--and possibly other members of the court--with opinions that defer to agency's interpretations of statutes and rules. Or perhaps, this decision just reflects a lack of unanimity on the court about what Chevron really means.  (If it's the latter, that's equally surprising given that no member of the court in King v. Burwell questioned the decision to bypass Chevron.)                    

Posted by Adam Zimmerman on June 25, 2015 at 06:00 PM | Permalink


Adam, I really enjoyed your post and agree that King's "deep questions" exception, read for all it may be worth, is hard to reconcile with some recent cases. I've added a word or two (see http://prawfsblawg.blogs.com/prawfsblawg/2015/06/chevrons-magical-disappearing-act.html) based upon yesterday's fair housing decision.

Asher, I think you're right to wonder what lower courts should make of the "oscillating" Chevron doctrine. Perhaps we shouldn't be surprised to see a "Chevron shuffle" in the lower courts, to refer back to one of my earlier posts.

Posted by: Seth Davis | Jun 26, 2015 2:50:04 PM


Thanks again for your thoughtful comments. I think you make an interesting point about Skidmore (and, in case my hastily-written post wasn't clear, I agree with you that the case could have been decided at Chevron Step One). Your post reminds me a lot of Cass Sunstein's Chevron Step Zero, 92 Va. L. Rev. 187 (2006)--both in the way you frame why the Court could have easily resolved the case at Chevron Step One, as well as the justice's oscillating views about Chevron itself. As to the latter, Sunstein argued that the unsettled doctrine in this area, to some extent, reflected whether or not one considered Chevron to be "evolutionary" (a summary of the way courts had always reviewed different agencies' legal interpretations using previously established principles, as Justice Breyer would have it), or "revolutionary" (a decision that fundamentally altered the relationship between how agencies and reviewing courts interpret law, as Scalia would have it). I agree that the Court appears to vacillate between these ideas almost as much, if not more, than they do over how to interpret a statute. That uncertainty not only has implications for the lower courts, but whether or not agencies themselves can confidently create stable regulations.

I was tempted to say in my original post that Chief Justice Robert's decision to revive the "major questions" doctrine was, like his last opinion in Obamacare, an attempt to play the "long game." Just as he used the last Obamacare case to revive a conservative conception of federalism, he seems to be using this one to restrain the administrative state -- empowering courts, more than they have acknowledged before, to question whether Congress has delegated authority to an agency to regulate. Bruce Huber, today, for example, argues that King's use of Chevron may give the Court more ammunition to strike down the EPA's Clean Power Plan, which also has "deep economic and political significance." (http://www.yalejreg.com/blog/what-king-burwell-means-for-environmental-law). But given the way the justices continue to go back and forth over the meaning of Chevron--from cases like MCI and Brown and Williamson, to Mead and Barnhart, to City of Arlington, and now, to King v. Burwell--I wonder whether it's even possible to play any kind of long game in this area of law.

Posted by: Adam Zimmerman | Jun 26, 2015 1:10:37 PM

Well done. I appreciate the work and insight you put into this.

Posted by: Tortmaster | Jun 26, 2015 2:38:20 AM

Well, this turned out to be an even worse version of the major questions exception decision than I feared. Before I comment on major questions, I have to ask: what's happened to Skidmore? Like most people, I think Chevron's about deferring to agencies' policy choices on issues Congress delegated to agencies, while pre-Chevron deference cases, like Skidmore, were mostly about deferring to agencies' sometimes superior insights into what Congress meant with respect to questions Congress didn't delegate. So when the Court finds, as here, that Congress didn't intend to delegate a question to an agency, that's precisely when Skidmore's supposed to kick in. Skidmore deference is (in)famously variable, because the reasons for epistemic deference are contingent, and maybe this is a case where Skidmore calls for very little deference, so little Skidmore isn't worth talking about. But I doubt that. The Court certainly appears to defer to various expert claims, either made by the government itself or sympathetic economists, about the consequences of invalidating the IRS rule, and arguing that those consequences tell us something about what Congress likely intended. And the Court, not being comprised of economists, isn't very well-positioned to predict those consequences by itself. So why not say that under Skidmore the Court gives some weight to the agency's economic predictions, which inform the government's arguments about what Congress intended? Maybe it's just an oversight - after rejecting the FCC's arguments for Chevron deference in City of Arlington, Roberts didn't consider Skidmore there either - but I do worry that lower courts will read this opinion and think deference is now Chevron or nothing.

As for major questions, the Court's intuition that Congress didn't intend to delegate this question to an agency - that rather, Congress must have had some intent, however awkwardly expressed, on this issue - is very understandable. Indeed, it's probably right. But there's a way to test intuitions of that kind, and that's Step One, which all the major questions cases the Court cites actually performed before holding that Congress hadn't delegated those questions to the agency. If Congress really had an intention on some issue, it will come out at Step One, after a court applies all the traditional tools of interpretation to the statute at hand (which it's free to apply somewhat more vigorously, as in B&W, when it doubts that Congress intended to delegate). But if a court can't find a clear meaning at Step One, then it should acknowledge that, intuitions notwithstanding, Congress may have delegated (knowingly or otherwise) to the agency after all.

Instead, this opinion says that when a court spots a really important question (whatever that might mean), it should assume that Congress must have had some intention on the matter that it wanted courts to discern, no matter what the process of actually interpreting the statute reveals. Surely Congress delegates *some* major questions to agencies. Of course, the Court says this statute's ambiguous, so maybe Step One's not the answer here, but by ambiguous, the Court only means semantically ambiguous; the opinion doesn't suggest that its author had a whiff of doubt about what Congress intended. The opinion could very easily be rewritten as a Step One decision in the mold of Brown and Williamson, one which would have held that, given context and structure and consequences and the anomalies petitioners' reading would create, Congress's meaning is clear. One final observation: so long as the Court's voting and lineups and even opinion assignments in Chevron cases is distorted by the substantive content of agency rules before the Court, Chevron "doctrine" will always unpredictably and rapidly oscillate, perhaps more so even than standing doctrine tends to oscillate depending on the Court's interest in and views on underlying merits issues in standing cases. And given all that oscillation, maybe lower courts should treat new pronouncements about Chevron in the same non-binding way that they treat decisions about grammar canons, or give greater weight to Chevron cases on regulations that the Court doesn't care much about, and/or Chevron cases that the Court grants to decide questions about Chevron itself.

Posted by: Asher | Jun 26, 2015 2:22:11 AM

The comments to this entry are closed.