Highly Recommended Article on Chevron and History of Deference (7/1/20; 7/7/20)

Readers of this blog will already know that I have been interested in the interface of administrative law and tax, and as subsets, interface of the Administrative Procedure Act (“APA”) and tax and the interface of Chevron deference and tax.  I have written an article posted on SSRN.  The Report of the Death of the Interpretive Regulation Is an Exaggeration (SSRN 6/20/19, as revised 1/25/20), here.  In that article, I argue against the notion, apparently current among scholars, that the Treasury interpretive regulation is no more.  I discuss what I think are the relevant facets of administrative law, the APA and Chevron.

One of the principal arguments I make is that Chevron did not create deference ex nihilo, but simply refined and regularized deference that already existed.  Justice Scalia said of Chevron that it was “a stable background rule against which Congress can legislate.”  City of Arlington v. FCC, 569 U.S. 290, 296 (2013).  In the article, I go through some of the history of deference prior to Chevron that shows that the Supreme Court often explicitly deferred to reasonable agency interpretations, noting some prominent tax examples of such deference.

For readers interested in the subject, I recommend a new article that dives deeper into the history of deference. Craig Green, Chevron Debates and the Constitutional Transformation of Administrative Law, 88 Geo. Wash. L. Rev. 654 (2020), here.  The summary of the article at the beginning is:

             Chevron v. NRDC is under attack. Chevron deference to agencies’ statutory interpretation is a pillar of modern government that judges and bureaucrats have used almost every day for thirty years. Until recently, most observers dismissed efforts to overrule Chevron as impossible or absurd, yet one of Justice Anthony Kennedy’s last acts on the Supreme Court suggested that Chevron deference might violate the separation of powers. 

            Constitutional threats to Chevron are surprisingly recent and grave. In 2015, Justice Clarence Thomas was the first judge in history to write that Chevron is unconstitutional. Anti-Chevron critiques by Justices Neil Gorsuch and Brett Kavanaugh were featured elements of their Supreme Court nominations. Justice Samuel Alito joined an opinion in 2019 that condemned all administrative deference. And even though Chief Justice John Roberts’s concerns have been more nuanced, his ambivalence may be decisive. A landmark ruling seems imminent—one way or the other—and now is the time to analyze relevant arguments and consequences. 

            This Article examines the history and merit of Chevron’s constitutional critiques. Reagan-era conservatives like Antonin Scalia used to celebrate Chevron as compatible with the separation of powers, and the Supreme Court viewed administrative deference as a perfectly ordinary practice for almost two hundred years. That historical evidence supports normative arguments that Chevron is consistent with basic structures of constitutional law. Overturning Chevron would be the most radical decision in modern history about constitutional structure, upsetting hundreds of precedents, thousands of statutory provisions, and countless agency decisions. Such a ruling would transform constitutional law itself, as judges apply newly aggressive theories to destroy established tools of democratic self-governance.

JAT Comments:

1.  Again, I think the article is excellent at developing the history of deference.  One key helpful feature is the Appendix at the end (beginning on p. 734, here) that presents a table of key historical deference cases back to 1827.  The author concludes (p. 660, here):  “Even though pre-Chevron cases were sometimes inconsistent and certain details changed over time, the concept of administrative deference has been routinely accepted as permissible since the earliest decades of administrative government.”

2.  A good quote (pp. 693-694, here):

             In the end, this Article’s complex history of administrative deference supports a very simple conclusion. Various forms of administrative deference—which only somewhat and sometimes resembled Chevron—have existed for an immensely long time, and until recently, courts have not defined the limits of administrative deference [*694] by citing principles of constitutional law. Modern critics have claimed that deference to agencies offends constitutional traditions and values, but it is important to know that most lawyers and judges throughout American history consistently failed to notice.

Note that even such luminaries as Justice Scalia, a Chevron fan, failed to notice any constitutional infirmities in deference.  And, as to whether the pre-Chevron deference resembled Chevron, as I note in my article, pre-Chevron deference did not have the Chevron two-step and was not always consistently applied, but at least in prominent tax cases, that deference was expressly articulated as deference to a reasonable agency interpretation.  One of the key cases I cite is United States v. Correll, 389 U.S. 299, 307 (1967), where the Court said:  “The role of the judiciary in cases of this sort begins and ends with assuring that the Commissioner's regulations fall within his authority to implement the congressional mandate in some reasonable manner.”  In Correll, Treasury interpreted the statutory “away from home” requirement for business travel deductions to require “sleep or rest,” an interpretation certainly not compelled by the statutory text but simply a reasoned choice by the agency charged with administering the statutory “away from home” requirement so that it meant something.  At least in theory, there were other reasonable choices to interpret that statutory text; the agency chose one and the Court sustained the agency’s choice of reasonable interpretation.  (As an aside, Correll and Chevron itself should dispell the notion that just because the agency interpretation is not compelled by the text of the statute does not mean that it is not an interpretation tested as such rather than an agency "legislative" activity, meaning that it is a legislative regulation and tested as such.)

3.  The author also makes this key point (pp. 707, here, and 710, here; footnotes omitted):

The constitutional parallel between statutory interpretation by judges and similar activity by agencies has consequences for many federal statutes. Whenever it is unconstitutional for agencies to “legislate” by providing substantive content for vague federal statutes, it is also unconstitutional for courts to provide substantive content for vague federal statutes. Because only Congress can write federal laws, neither agencies nor courts can undertake legislation-in-fact while claiming that they are only doing “interpretation.” Conversely, whenever it is permissible for federal courts to interpret vague statutes without unconstitutionally “legislating,” it is equally reasonable for agencies to do the same. 

* * * *  

Chevron’s nondelegation critics cannot distinguish circumstances where courts interpret ambiguous statutes from other contexts where agencies do the same thing On the facts of Chevron, if it was an unconstitutional delegation for Congress to let the EPA decide what “stationary source” means, it should be equally unconstitutional for Congress to let courts make that decision. Either it is legislation to prescribe a specific meaning for “stationary source,” or it is not. As a matter of nondelegation doctrine, constitutional limits should apply regardless of whether the arguably legislating entity is an agency or a court. Contrary to Chevron’s critics, the charmingly oversimplified argument that only Congress “writes the laws” cannot stop agencies and courts from resolving statutory ambiguities. Chevron’s critics have never explained how much “nonlegislative” interpretation they think agencies and courts can exercise, but that line-drawing problem is one reason that Scalia—unlike Thomas and Gorsuch—rejected nondelegation cases like A.L.A. Schechter and Panama Refining as unredeemed outcasts.

Added 7/3/20 11:45 am:  I made similar points in my article linked above.  The following is the first point it surfaces in the article (at page 6, here, footnote omitted):

Another concept that threads through the issues that readers should keep in mind is the constitutional roles of Congress, the Executive and the Courts.  Congress legislates.  The Executive, generally through agencies, administers many of the laws enacted by Congress.  In the process of administration, the Executive interprets the laws.  Courts resolve disputes and, in the process, interpret the laws enacted by Congress.  Although Congress may delegate to the Executive some carefully circumscribed legislative functions via authority for legislative rules making the law, Congress cannot and does not delegate legislative functions to Courts.  Accordingly, when Courts interpret ambiguous statutory text and choose among reasonable interpretations, Courts are not legislating, a power Courts do not have.  When Congress has assigned to Executive agencies the power to interpret statutes in their administration of a regulatory scheme, that power is not a legislative power to make the rules, it is just a power to interpret the rules enacted by Congress as statutory text.  And, to close that loop, when Courts interpret the statute by deferring to an agency interpretation, the Courts are still interpreting and applying the statute, the constitutional role of the Courts.  (The statements I just made embed arguments reflecting my view that are presented in detail in this article, but I think this is the usual conceptualization of the constitutional roles of the branches of our Government.)

  4  Another excerpt (p. 714, here; footnotes omitted):

[S]ome critics have cheerfully speculated that, after Chevron is overthrown, courts might return to some kind of pre-Chevron regime where case-by-case deference depends on a particular agency’s expertise, the legal dispute’s technical nature, and the decision’s procedural thoroughness. For example, Gorsuch wrote that “[w]e managed to live with the administrative state before Chevron. We could do it again.” As a matter of constitutional law, however, the opposite is true. For purposes of Marbury and Article III, Chevron’s categorical deference is identical to pre-Chevron deference that varied in different circumstances. Administrative deference either violates Marbury by misallocating adjudicative authority, or it does not. That is why every living judge until recently reached the latter conclusion.

 5.  And from the conclusion (p. 732-733, here]:

From the very start, Chevron has been a judicial interpretation of statutory defaults and administrative practice, much like similar doctrines from the 1940s and diverse precedents before that. The success of political conservatives in the 1980s brought Chevron into the world, [*733] and a new generation of political conservatives may very well take it out. Whatever one thinks of reforming Chevron through ordinary mechanisms of judicial pragmatism or political choice, eliminating administrative deference as a matter of constitutional law would contradict established precedents, long traditions, and basic governmental stability. From that perspective, reliance on newfangled visions of constitutional “theory” or “structure” would not only transform the operation of administrative government, it would also change the fundamental nature of constitutional law itself.

6.  Added 7/7/20 11:00am:  I loop back to point #3 above about the parallel between judicial interpretation of statutes and administrative interpretation of statutes.  Both are exercises in the interpretive authority rather than legislative authority.  A related point was made  In Barr v. AAPC, __ U..S. ___, ___ S.Ct. ___ (2020), here, the Supreme Court invalidated the government debt exception to the general prohibition on robocalls to cell phones in the Telephone Consumer Protection Act of 1991.  That exception had been generally thought applicable, so there were an indeterminate number of government debt exception robocalls from enactment through yesterday’s opinion that might thus be liable for their past conduct, hence the notion of “retroactive” liability.  But, as I and many others have noted, the liability is not retroactive because all the Court did was to announce what the law has been all along (i.e., the general prohibition excluding the exception the Court recognized as invalid).

Will Baude, here, a prominent scholar and thoughtful commentator, offers a pungent comment on the exercise of judicial interpretation (Footnote 12 and retroactive liability (The Volokh Conspiracy 7/6/20), here) on Barr v. AAPC.  Baude quotes the relevant part of the plurality opinion (page 22 n. 12):

As the Government acknowledges, although our decision means the end of the government-debt exception, no one should be penalized or held liable for making robocalls to collect government debt after the effective date of the 2015 government-debt exception and before the entry of final judgment by the District Court on remand in this case, or such date that the lower courts determine is appropriate. See Reply Brief 24. On the other side of the ledger, our decision today does not negate the liability of parties who made robocalls covered by the robocall restriction.

Baude then comments (emphasis supplied):

This conclusion is perfectly logical, but its legal basis is a bit mysterious to me. (Page 24 of the government's reply brief, the only authority cited, says: "It is doubtful that a person who made automated calls to collect government-backed debts before the exception was held invalid could be said to have violated the TCPA. But because no question of retroactive liability is presented here, those issues may be reserved for a future case.") 

So what gives? Judicial decisions are inherently retroactive, because they are supposed to declare what the law is, not change it into something else. (The Court wrestled with special cases of prospective judicial decision making a few decades ago, but then abandoned it as a mess.) Lower courts have recognized a "good faith" defense when people relied on a statute or precedent that was later invalidated, but the Supreme Court has not decided that issue.

So it seems to me that those who joined this part of the plurality opinion (Justice Kavanaugh, the Chief Justice, and Justice Alito) must either want to experiment again with prospective-only judicial decisionmaking, or they must believe that the Court should recognize a new good faith defense after all.

The key point I want to make is that implicit in this comment and the footnote that it discusses is the notion that interpretation is not a legislative exercise of power--a power courts do not have under any concept of the separation of powers.  (Indeed, Congress cannot confer legislative powers on courts, but can delegate some carefully circumscribed legislative powers on agencies, such as the consolidated return delegation in § 1502.)  Just as courts when interpreting are not exercising legislative powers, so agencies when interpreting are not exercising legislative powers. 

One of the examples I use in my articles is the interpretation of a statutory term "white" to include "off-white."  If a court did that, the court would be interpreting; if an agency did that, the agency would be interpreting.  Chevron would test the interpretation to determine if it is reasonable within the scope of the statute's textual ambiguity (a broad test that really encompasses both Chevron Step One and Chevron Step Two).  The agency  interpretation might also be tested for procedural regularity under the 706/State Farm test, but that goes to procedural regularity rather than the reasonableness of the interpretation.  But the  key point is that it is an exercise in  interpreting, not in legislating.