Monday, July 27, 2015

The Triumph of Chevron Step Zero?

By Michael Dorf

Tomorrow I will be speaking at the Practicing Law Institute's annual Supreme Court Review session. This is the seventeenth such session and, if memory serves, I have been a panelist since just about the beginning. It's always a fun-filled day. UC-Irvine Dean Erwin Chemerinsky and Touro Emeritus Professor Marty Schwartz organize the festivities as we range over the SCOTUS Term just completed. PLI is a non-profit provider of continuing legal education but it has a lot of expenses (including travel expenses but no pay for us speakers) so it charges for attendance or remote access. That's my way of saying that if you want to attend or watch, you'll need to pay. (Some scholarships are available.)

In addition to commenting on cases presented by other speakers, each speaker is assigned the responsibility for presenting a portfolio of cases. My portfolio includes Texas Dep’t of Housing v. Inclusive Communities Project, Inc., which holds that disparate-impact claims are cognizable under the Fair Housing Act (FHA). Because the case was decided at the end of the Term, when my attention was mostly focused on other cases, I only recently had occasion to read it carefully in preparation for tomorrow's panel. My presentation will focus chiefly on the big-picture disagreement between the majority and dissent over the availability of disparate impact claims under the FHA and more broadly, but here I want to note something that cuts across the Court's statutory jurisprudence: the possibly declining influence of Chevron deference.

In 1984, in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc. the Court held that where a statute contains unclear language, courts ought to defer to a reasonable agency interpretation of that language. Chevron was not the first case to say this but it became the standard citation for the proposition that courts defer to administrative agencies. And Chevron deference was grounded in principles of both expertise and accountability. As between agencies--staffed by experts and accountable to the People through the president--and courts--staffed by generalists with life tenure--it should be assumed that Congress preferred agencies to fill in statutory gaps. Over the years, Chevron was refined into a two-step test: (1) Is the statutory language unclear? and (2) If so, is the agency interpretation reasonable?

Yet as Cass Sunstein argued in an important paper ten years ago, borrowing a term coined by Tom Merrill and Kristin Hickman, there has emerged an important threshold question: Chevron step zero. In Chevron step zero the court asks whether the Chevron framework applies at all or whether instead the lack of clarity in a statute should be treated as calling for judicial rather than agency clarification. Sunstein argued that the emergence of Chevron step zero was a regrettable occurrence, as it marked a decline of deference to agencies. I'm not sure I agree with that normative assessment but my aim in this post is simply expository rather than normative. I want to suggest that two cases in the past Term vindicate the view that Chevron step zero is real and increasingly important.

The SCOTUS itself has never expressly used the term "Chevron step zero" but a few federal appeals court decisions have. More important than the terminology, however, is the substance of what the Supreme Court cases do, and there the evidence is quite clear.

In King v. Burwell, for example, CJ Roberts declines to decide the permissibility of subsidies on federal exchanges based on Chevron, even though his opinion makes clear that the Internal Revenue Service's interpretation of the Affordable Care Act would survive Chevron's two-step process. He offers a classic Chevron-step-zero account of why the issue goes to the Court rather than an agency:
Whether those credits are available on Federal Exchanges is thus a question of deep “economic and political significance” that is central to this statutory scheme; had Congress wished to assign that question to an agency, it surely would have done so expressly. [Citation] It is especially unlikely that Congress would have delegated this decision to the IRS, which has no expertise in crafting health insurance policy of this sort.
Perhaps even more striking than that passage is the Court's resolution of Texas Dep't of Housing at the Chevron-step-zero stage without even pausing to consider whether to apply the Chevron framework. Justice Kennedy's majority opinion simply assumes that the question presented is whether the FHA permits disparate impact claims rather than first considering the possibility that the issue should be whether the Housing and Urban Development (HUD) regulation interpreting the FHA to encompass disparate impact liability satisfies Chevron. The Court decides the case at the Chevron-step-zero stage without even discussing the Chevron-step-zero issue. It's a kind of Chevron-step-negative-one ruling. (Justice Alito, in dissent, rejects the HUD reg at Chevron step one, arguing that the statute's language clearly encompasses only intentional discrimination.)

King and Texas Dep't of Housing are somewhat unusual Chevron-step-zero decisions in that both cases uphold the underlying agency interpretation of the statute. Given that the Court thinks in each case that the agency got it right, it follows a fortiori that the Court would have thought Chevron was satisfied if the Chevron framework applied. As Adam White noted in a SCOTUSblog post  after King v. Burwell, Chevron step zero first emerged in cases in which the Court struck down administrative interpretations of statutes. It was a doctrine embraced by conservatives seeking to rein in the administrative state. Indeed, that's part of the reason why Sunstein--a fanboy of the administrative state if ever there was one--fretted over Chevron step zero.

Despite its conservative origins and what may yet prove to be its conservative bias, Chevron step zero has been most vigorously opposed on the Supreme Court by Justice Scalia. His lone dissent in the 2001 case of United States v. Mead Corp. provides a particularly vigorous defense of a presumption in favor of Chevron's applicability.  However, liberals who share Sunstein's concern about the long-term implications of Chevron step zero should not take too much comfort from Justice Scalia's opposition. As cases like King and Texas Dep't of Housing show, he is quite willing to reject agency interpretations that do not fit his normative views at Chevron step one, finding language that others deem unclear to be dispositive. Thus, to take an example I blogged about last week, even though Justice Scalia has argued for Chevron deference to the Equal Employment Opportunity Commission (EEOC), don't count on him voting to uphold the EEOC conclusion that sexual orientation discrimination violates Title VII's prohibition on sex discrimination.


Steve Finnell said...


Is praying to God a form of worship? Yes, without a doubt. Is praying to men or women who are dead or alive worship? Yes, without a doubt.

Who should men worship? Luke 4:7-8 "Therefore, if You will worship before me, all will be Yours." 8 And Jesus answered and said to him, "Get behind Me, Satan! For it is written, 'You shall worship the Lord your God, and Him only you shall serve."(NKJV)


Worship Defined: 1. (Ecclesiastical terms) to show profound religious devotion and respect to; adore or venerate (God or any person or thing considered divine) 2. to be devoted to and full of admiration for. 3. to have or express feelings of profound adoration......7. (Ecclesiastical) the formal expression of religious adoration; rites, prayers, etc. [REF. The Free Dictionary.]

Praying to God is worship and men are to only worship God.

A Costa Rican woman has told how she recovered from a brain aneurysm after praying to Blessed Pope John Paul ll ---the second miracle attributed to the pontiff, who died in 2005. [REF: catholicherald.]

Praying to any man dead or alive is worship and to worship anyone but God is sin.

Through the selfless "yes" of the Virgin Mary, Jesus Christ, our Savior, was brought into the world. It is appropriate, therefore , that we offer prayers of petition and praise to the Mother of God.


Praying to anyone is worship and to worship anyone but God is sin.



Posted by Steve Finnell at 3:52 PM No comments:
Email This
Share to Twitter
Share to Facebook
Share to Pinterest
Links to this post

Joe said...

A very important inside baseball issue.

Paul Scott said...

Spell caster spam again!

Justin said...

Because I agree with Justice Roberts, I am going to make a few points - mostly related to the real case we're all discussing here, City of Arlington v. FCC.

1) Ah, the perils of writing in the middle of a timeline. Consider this, from the majority opinion:

But Mead denied Chevron deference to action, by an agency
with rulemaking authority, that was not rulemaking.
What the dissent needs, and fails to produce, is a single
case in which a general conferral of rulemaking or adjudicative
authority has been held insufficient to support
Chevron deference for an exercise of that authority within
the agency’s substantive field. There is no such case, and
what the dissent proposes is a massive revision of our
Chevron jurisprudence.

Well, now we have a case. THANKS, SCALIA.

2) The majority continues:

Rather, the dissent
proposes that even when general rulemaking authority is
clear, every agency rule must be subjected to a de novo
judicial determination of whether the particular issue was
committed to agency discretion.

So true. But where does Scalia, a committed originalist, get a contrary argument? Not from Article I, which says unequivocally, "All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives." Thomas's argument-that the regulatory state itself is unconstitutional-might be right from an originalist standpoint, although I fear that dooms originalism more than the regulatory state. So let's look at Chevron's justification for itself:

If Congress has explicitly left a gap for the agency to fill, there is an express delegation 844*844 of authority to the agency to elucidate a specific provision of the statute by regulation. Such legislative regulations are given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute.[12] Sometimes the legislative delegation to an agency on a particular question is implicit rather than explicit. In such a case, a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency.

Note, either way, that the grounds for deference is based on delegation. You cannot defer on the question of the very delegation that allows the deference. That would be the tail wagging the dog.

Justin said...

3) Now, I disagree with the policy concerns of the dissent. Roberts's throwaways regarding "the danger posed by the growing power of the administrative state" are just that. And, in most cases, the question of whether deference was intentional can be gleamed from the structure of the statute. Using Scalia's own example, the use of a term of art such as "common carrier" without providing a definition is an implicit delegation. So is the use of the term "unreasonable."

In contrast, the idea that Congress intended to delegate the IRS the issue of tax credits in federal exchanges is itself silly. So is the idea that Congress intended to delegate the question of whether disparate impact discrimination is barred by the statute. This is not merely because they are "big, important, questions" but instead because they are questions regarding the structure of the law in which it appears plain that Congress intended to answer that question for itself rather than delegate.

4) But this shows quite an inconsistency between Roberts' view and textualism, or at least the type of narrowish textualism that Scalia adopts which prevents consideration of the structure and purpose of statutes. It is only by looking at the structure and purpose of the statute that one could determine that Congress did not delegate the question of tax credits in King. It is only by looking at the structure of the FHA that allows us to determine precisely whether this is an implied delegation of authority. And even in more run of the mill cases, one would be hard-pressed to determine, e.g., if "because of" provided a precise meaning or an implied ambiguity without looking at structure, since sometimes the term is ambiguous, and other times it is not. Roberts' view is inconsistent with a narrow interpretation of textualism.

5) And, to go off point a little, this is the most fundamental part of Roberts' opinion in King. In King, Roberts found an "ambiguity" and resolved the ambiguity using structure and purpose. I wouldn't think of this as finding an ambiguity in the first place - an ambiguity would only exist if the answer was indeterminate even after looking at structure and purpose. Nevertheless, Roberts' express invocation of structure and purpose as part of textualism will hopefully change the way we look at statutes in general. If this change was done in order to overrule Arlington, so much the better.

Greg said...

Justin - The problem I see with your argument is that it works for well-drafted laws, but it risks giving too much power to the courts in the case of poorly drafted laws. If every law were well-drafted then there would be no need for Chevron Step 1 (or at least the answer would be obvious as it would be finding obvious deference, not ambiguity), and we could jump right to Chevron Step 2.

King is perhaps the exception that proves the rule, and I'm not sure that Texas Dep't of Housing is nearly so clear that the right thing was to remove deference.

King was an excellent example of a poorly drafted law where Chevron deference would normally be warranted based on one part appearing to tell the agency to do one thing and a different part assuming that they would do almost the exact opposite. Even though this isn't delegation, the agency tasked with implementing the law is often in a better position to figure out what the heck Congress meant based on how the various parts of the law are going to interact (and thus the best implementation) than the courts are.

King is ultimately the exception because the stakes are INCREDIBLY high, and thus there is NO reasonable way to believe that Congress had ANY intention of leaving this decision up to the agency. Further, there is no way to believe that, regardless of their intent, Congress would be okay with this decision changing with a changing administration. In this exceptional case, Roberts was right that deference to the agency is not the right solution, and a binding response from the courts is absolutely required.

I'm not so sure about Texas Dep't of Housing because it is quite possible that Congress was doing a broad grant of authority that was rightfully subject to Chevron deference so long as the interpretation is reasonable. Here the decision should be squarely within Chevron Step 2 (insofar as the steps can be reasonably separated from each other, which sometimes they can not.)

As applied by Roberts, this is not Chevron Step 0, but Chevron Step 1.5. Instead of asking up front if it was a deference, he was asking in the case of an already identified ambiguity (Step 1) was the decision so important that, despite passing Chevron Step 1, the courts must make a determination because the question is so significant that the will of Congress must be to establish an answer permanently and not subject to the whims of the administration, and that having a permanent answer is more important than having the correct one.

Now, I'm assuming here for the sake of discussion that Prof. Dorf's description of Texas Dep't of Housing as Chevron Step 0 is correct. I'm not sure I read the opinion that way, but I'm chosing to defer to him on this point.

Justin said...

1) I think both you and Dorf are wrong regarding either "step 0" or "step 1.5." Rather, "step 1" as normally defined is whether "Congress has directly spoken to the precise question at issue." But this is precisely because sometimes "the legislative delegation to an agency on a particular question is implicit." What Roberts is doing is not adding a step, etiher prior to Step One or after Step One. Rather, he's being faithful to the actual basis for Step One - the explicit or implicit deference - and admitting that not every time there is an ambiguity was such delegation implied. This is no less true of poorly drafted statutes as well drafted statutes. Poor writing is poor writing, not an intent of greater or lesser delegation.

2) I think your interpretation of Roberts's opinion of "is so significant" misstates Roberts's views. Rather, Roberts looked at the structure of the various statutes in question and concluded that there was significant and sufficient evidence that Congress intended to answer the question posed in the Bill itself.

3) I think you overstate the problem with Roberts's view generally, While I agree it can be used to erode Chevron substantially, instead of faithfully apply it, in the usual case the structure of the Act will imply delegation. As Chevron says, the *usual* result of an ambiguity is the finding of deference. All Roberts (appropriately) says, is that just because that is true, the existence of an ambiguity is not itself irrebuttable proof of an implied delegation.

4) For the reasons cogently explained by the majority opinion in the FHA case, I don't think any delegation was intended regarding whether FHA applied to disparate impact. Rather, focusing on applying the standards of Title VII, Congress intended it to apply. Had the Reagan Administration issued regs saying that disparate impact does not violate the FHA, it would be countering Congress's specific intent, and would not be acting in accordance with their delegated power.

Justin said...

5) " Even though this isn't delegation, the agency tasked with implementing the law is often in a better position to figure out what the heck Congress meant based on how the various parts of the law are going to interact (and thus the best implementation) than the courts are."

While this appears to be a post-hoc justification of Chevron by at least some academics (Strauss), it is not a justification for Chevron that, as far as I am aware, has ever been accepted by the Court. Why would the agency be in any better place than the Court in determining Congressional intent? That's not delegation, that an usurpation of the judicial function.

Justin said...

One last point - I think you and Dorf are really arguing for Vermuele and Stephenson's argument that there's only one Chevron step anyway. But I think it is telling that the word "delegate" does not appear in any of its forms anywhere in that essay.

- J

Anonymous said...

I don't see King in the same light as you do Professor. Assuming that the normative grounds for Chevron are correct--expertise and accountability--then Roberts' point about the IRS buttresses those claims rather than undercuts them. If the point is to defer to agencies because they have special expertise it logically follows that if an agency doesn't have that expertise one should not defer to them. So Roberts' opinion doesn't undermine Chevron; it is wholly consistent with Chevron's normative outlook.

Joseph said...

Justin & James Longfellow, while I agree with your outlook, I think criticism of Prof. Dorf's exposition is misplaced. Justice Roberts wrote that King was an exceptional case in which Chevron was not applicable. I was bothered by Roberts' framing because I think that it should not be exceptional to first establish whether there is delegation. However, I do think it is fair (and correct) to say that the Court has, at best, traditionally elided that consideration. In asserting that initial consideration, Chevron, as traditionally applied, is weakened. U.S. v. Mead shows the contrasting views well. Even after King, it is not clear that considering whether there has been delegation is strictly called for, but instead might be invoked at the Court's discretion, or elided yet.

Anonymous said...

@Joseph Simmons.

I'm simply not convinced that the distinction is meaningful. It seems to me that in the case at hand the dispute is whether a lack of expertise is best categorized as a Chevron "step zero" or a Chevron "step one" failure. Roberts frames it as a Chevron step zero failure but that position seems to me more rhetorical than substantial. Either way the IRS is not due deference.

There is too much trying to read meaning into the halos of penumbras for my taste.

Justin said...

I wasn't intending to criticize Professor Dorf except in two minor ways - one, I think it is more appropriate labeled as part of Step 1 than as "step 0" (this is more a criticism of Sunstein, anyway), and two - I suppose I implicitly criticized the post for not discussing City of Arlington.

The rest was a defense of Roberts- I don't take Dorf to be taking a specific position in this post on Roberts' view, other than noting that it might have a politically conservative bias.