Section 7805(a) - Legislative or Interpretive and Problem of Retroactivity (4/2/20; 4/24/20)

In my article, The Report of the Death of the Interpretive Regulation Is an Exaggeration (SSRN last revised 2/28/20, here), I argued that interpretive regulations for agencies generally and for the IRS under § 7805(a) remain viable despite some claims to the contrary.  Today, I summarize a key facet of my argument in the article – that § 7805(a) authorizes only interpretive regulations (rather than legislative regulations as some scholars argue) and that the limitations on retroactivity in § 7805(b) are properly viewed as limitations on the general rule of retroactivity for interpretive regulations rather than grants of authority to promulgate retroactive legislative regulations.  Since I provide copious citations in the article, I will not lard up this blog post with citations except as necessary.  (Variations on this theme have been addressed in prior blogs, particularly, Treasury Regulations and the APA Categories of Legislative and Interpretive Regulations (Federal Tax Procedure Blog 1/12/19; 1/19/19), here, and Legislative Rules And Chevron Deference An Oxymoron? (Federal Tax Procedure Blog 1/31/20; 2/10/20), here, but I recommend that readers sort through this blog before going to those other offerings.)

Section 7805, here, is a familiar Code Section.  It is titled:  Rules and regulations.  It’s first subsection is:

(a) Authorization
Except where such authority is expressly given by this title to any person other than an officer or employee of the Treasury Department, the Secretary shall prescribe all needful rules and regulations for the enforcement of this title, including all rules and regulations as may be necessary by reason of any alteration of law in relation to internal revenue.

Note that the authorization is for “rules and regulations.”  I focus here just on regulations (although I will sometimes use the term rules, which is an APA category that may include both regulations and subregulatory guidance (such as in the case of the IRS, Revenue Rulings).

The antecedents of the current § 7805(a) trace back to the dawn of the modern income tax (stated by scholars as 1917, 1918 or 1921, with some antecedents ever further back in general administrative law). Section 7805(a) is often referred to as a “general authority” provision in contrast to specific authority in a Code section to promulgate regulations to either define (interpret) a term in the statutory text (interpretive regulations) or to prescribe the law (a legislative regulation).  Most Treasury regulations are general authority regulations issued under § 7805(a).  (There is some fuzziness here because some would argue that a specific authority statutory direction to define a statutory term makes the regulations legislative in character rather than interpretive, even though all the statute does is direct the IRS to interpret the statutory term; but let’s not get hung up on that because that would mean that the specific authority regulations are all legislative rather than interpretive and would not affect the dividing line between § 7805(a) general authority regulations and specific authority regulations.)

The question I address here is whether regulations issued under general authority statutes such as § 7805(a) (or even under inherent authority derived from Congress’ assigning administrative authority to the agency) are interpretive or legislative in character.  I focus here on § 7805(a) for Treasury Regulations, but the issue is presented for other agency general authority regulations.

In the current context, § 7805 had a pretty clear meaning for most of its statutory life.  The meaning was that § 7805(a) authorized the Treasury (IRS) to prescribe guidance in the form of “rules and regulations” which did not create new law but interpreted ambiguous text in existing statutes.  The understanding was that § 7805(a) authorized interpretive regulations and did not authorize legislative regulations.

I think most judges and scholars were comfortable with understanding at least until Chevron and more likely until later cases interpreting Chevron.  Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984), and then principally United States v. Mead Corp., 533 U.S. 218 (2001).  Then some new ways of viewing § 7805(a) crept into the academic discourse.  Section 7805(a), so the new notion goes, is a grant of authority to the IRS to create legislative regulations (legislative rules must be regulations).  A key feature of legislative regulations is that they create the law just as if they were statutes and are not interpretations of the law (statutes).  Because legislative regulations are the law (and not interpretations of the law), they are said to have the force of law.  Thus, morphing that traditional modeling of the legislative regulations to deference concepts under Chevron and its progeny, the new notion is that § 7805(a) regulations, within the scope of the authority granted, are the law.  Why?  Because, once you say that § 7805 regulations have the force of law, you say that they are the law just like statutes and not interpretations of the law (statutes).

So what?

Well, let’s see what that means for the rest of § 7805.  But first a quick foray into administrative law and the APA in particular.  Legislative regulations, like statutes, must generally be prospective only.  As with statutes, there can conceivably be exceptions to nonretroactivity for legislative regulations, but those exceptions must be authorized by Congress in a statute, and indeed exceptions to the APA’s requirements for prospectivity must be expressly stated in statutory text.  5 U.S.C. § 559 ("Subsequent statute may not be held to supersede or modify [the APA] * * * * except to the extent that it does so expressly.").

Now focus on the rest of § 7805 but let’s frame the focus on the two competing views of § 7805:

• § 7805(a) authorizes only interpretive rules; when those rules are stated in regulations the regulations are interpretive regulations.
• § 7805(a) authorizes interpretations, but when those interpretations are stated in regulations the resulting “rules” are legislative and not interpretive.

If legislative, then a § 7805(a) regulation generally cannot be retroactive without express statutory approval permitting retroactivity as an exception to the APA general command for prospectivity.  But, perusal of § 7805 shows that there is no express exception to the APA’s requirements for prospectivity of legislative rules.  There is a key provision for retroactivity but it does not expressly state that it is an exception to the APA general requirement for prospectivity.

If interpretive (rather than legislative), then a § 7805 regulation can indeed be retroactive back to the effective date of the statute that the regulation interprets.  This was the traditional understanding of § 7805(a).

So, which is it?

Here is how I address it in the article (slightly modified and footnote omitted):

• If § 7805(a) is authority only for interpretive regulations, then the balance of § 7805 makes sense and works.  As interpretive authority, the § 7805(a) regulations can be issued without Notice and Comment (thus permitting Temporary Regulations) and can be retroactive subject to such limited retroactivity or prospectivity as determined by Treasury or imposed by Congress in the balance of § 7805.  Specifically, § 7805(b) limits retroactivity otherwise allowed. But, the § 7805(b) limits do not apply to regulations interpretations of statutes enacted prior to July 30, 1996, thus leaving in place retroactivity for § 7805(a) regulations for such pre-July 30, 1996 statutes if § 7805(a) regulations are interpretive.  This is consistent with the Traditional Understanding that § 7805(a) regulations, as interpretive regulations, could be applied retroactively.

• If, however, § 7805(a) is authority for legislative regulations, then the balance of § 7805 wobbles.  Section 7805(b) must then either be viewed as a nullity because violating the Prospectivity requirement for legislative regulations or must be viewed as an unauthorized unstated exception to APA required prospectivity for legislative regulations.  And, § 7805(e)’s blessing of temporary regulations which have not undergone Notice and Comment must be viewed as an unstated exception to the APA requirement of Notice and Comment’s prospectivity requirement for legislative regulations.  There is no authority supporting the notion that Congress had this legislative regulation paradigm in mind, for those § 7805 subsections when enacted or revised before there was even a credible claim that § 7805(a) authorized only legislative regulations in the APA sense.  Indeed, both key consequences of treating § 7805(a) as authority for legislative regulations require that § 7805(b) and § 7805(e) are unauthorized unstated exceptions to the APA.  The APA, 5 U.S.C. § 559, prohibits any modification to the APA’s requirements “except that it does so expressly.”  I have not seen any persuasive claim that, by viewing § 7805(a) as a grant of legislative authority, this APA requirement for express modification was met for § 7805(b) or (e).

I ask readers to consider this: What do you think the Republican controlled Congress thought it was doing with the 1996 amendment to § 7805(b) and addition of § 7805(e)?  Do you think those Republicans intended to limit the IRS’s authority to promulgate retroactive interpretive regulations under § 7805(a), an intention consistent with the Republicans political agenda.  Or do you think those Republicans intended to grant the IRS retroactive authority for § 7805(a) regulations which, if they were legislative regulations, could not be retroactive without that express authority, an intention inconsistent with the Republicans' political agenda?

That really is the end of the major topic of this blog, but let me posit an example that may add some nuance.

1. On 1/15/01, Congress adds IRC section (the statute), effective immediately.  The statute imposes an additional tax liability on certain conduct.  The text of the statute is ambiguous – i.e., has a zone of reasonable ambiguity.

2. On 6/1/01, Taxpayer, T1, enters a transaction knowing the statute has more than one reasonable interpretation, but one of those reasonable interpretations avoids the tax for the particular transaction.  T1 does not report any tax on the transaction or make disclosure on the return (filed 4/15/02).

3.  On 1/15/02, Treasury issues a temporary regulation and a proposed final regulation adopting a reasonable interpretation of the statute that imposes the tax on T’s transaction.  The temporary regulation is effective immediately; the proposed regulation states an effective date applying to conduct from the date of the temporary regulation (1/15/02).

4. On 2/1/02, another Taxpayer, T2, enters a transaction exactly like T1's and when subsequently filing his tax return (say 4/15/03) does not report the tax and includes a disclosure that he is taking a position contrary to the temporary regulation on the basis that the temporary regulation is not a reasonable interpretation of the statute and thus flunks Chevron Step Two.

5.  On 1/15/03, Treasury finalizes the regulation, after notice and comment and reasoned decisionmaking, on 7/1/04, with the effective dates as note in paragraph 2.

6. On 9/1/03, the IRS starts an audit of T1’s tax year 01 focused only on T1's transaction.

7.  On 10/1/03, the IRS (through same agent) starts an audit of T2's tax year 02 focused only on T2's tax year 02 transaction.

8. On 1/15/04, the IRS issues notices of deficiency taxing the transaction both transactions.  The T1 notice of deficiency and underlying documents base the deficiency on an interpretation of the statute but do not refer to the temporary or proposed regulations because of the effective dates in § 7805(b).  The T2 notice of deficiency and underlying documents base the deficiency on the interpretation by specific reliance on the temporary and proposed regulations.

9. On 3/1/04, T1 and T2 file petitions with the Tax Court.

10.  On 1/15/04, the parties in both cases, after fully stipulating facts and submitting the case for decision, the parties file their respective briefs.  In the T1 case, the Commissioner’s brief urges that the proper interpretation of the statute taxes the transaction.  The Commissioner's brief does mention the temporary and final regulations but does have a footnote to the regulations.  The Commissioner does not argue Chevron deference for the interpretation.  In the T2 brief, the Commissioner relies on the interpretations in the temporary regulations and the now final regulations, urging that (i) the Commissioner’s interpretation is the best interpretation of the statute and (ii), in any event, is a reasonable interpretation entitled to Chevron deference.

What does the Tax Court judge do.  I think the analysis without nuance for now is:

1. In T1's case, the judge can apply the judge's best interpretation of the statute, because Chevron deference does not apply.

2. In T2's case, the judge must apply the reasonable interpretation in the temporary and now final regulations because Chevron does apply.

But, if this example (or some variation of it) were to happen in the real world with both cases decided at the same time, would a judge impose a different result for T2 than he does for T1?  Certainly, the theory permits the judge to impose different results in those (I submit rare) instances where the judge believes the IRS interpretation is reasonable but the judge thinks another interpretation is better.  But how much will a Chevron-entitled regulation interpretation affect the judge’s determination to T1's transaction where Chevron does not apply?  Can the judge consider the Chevron-entitled interpretation after 1/15/02 in some type of Skidmore analysis to determine whether it is persuasive as to the most reasonable interpretation?  See Skidmore v. Swift & Co., 323 U.S. 134 (1944). (This is even assuming that Skidmore actually means anything in terms of deference.)  I suspect that there may be some dynamic that the Chevron-entitled regulations applicable only to subsequent years will influence the judge’s determination of the best interpretation to apply in T1's case.  Will a judge having both cases to decide at the same time really find that an interpretation favoring T1 is most persuasive and then apply a different Chevron entitled interpretation taxing T2?  Since, under the facts I posit, the judge must tax T2 because of Chevron deference, will he really let T1 off the tax by insisting that the judge's interpretation is better than the IRS's reasonable interpretation?  And will the appellate court sustain such a holding?  All of this is just to say that a judge might make the determination of what is the most reasonable interpretation by working from the interpretation that all other taxpayers will be subject to in the future.

Added 4/6/20 2:00 pm:

On the last point as to the retroactive effect of an interpretation, see Smiley v. Citibank (South Dakota), 517 U.S. 735, 744 n. 3 (1996) ("Where, however, a court is addressing transactions that occurred at a time when there was no clear agency guidance, it would be absurd to ignore the agency's current authoritative pronouncement of what the statute means.")  Smiley is not directly in point, but sufficiently in point that, at least in theory it supports a court considering the subsequent regulation in applying an interpretation for T's transaction in year 01 (whether it is called Skidmore deference or just plain common sense).

Smiley does raise a key issue in my frame of reference and my difference with those who claim that regulations interpreting ambiguous statutory text are legislative rather than interpretive.  In Smiley, the issue was whether the statutory term "interest" included bank credit card late fees.  The Comptroller of Currency promulgated a notice and comment regulation defining interest to include late fees.  The Court held that that interpretation was subject to Chevron deference because the statutory term "interest" was ambiguous and the interpretation was reasonable.  In my view the regulation incorporating that interpretation was an interpretive regulation.  The Smiley Court certainly did not give any hint otherwise, and that was the paradigm it operated under.  (In this regard, the  notion now asserted that notice and comment interpretations are legislative did not morph into such claims until after Smiley.)

As I have noted in several blog entries, some, probably most, now characterize an interpretation in a regulation adopted with notice and comment and subject to Chevron deference (as in Smiley) are legislative regulations.  Professor Kristin E. Hickman is, I think,  the most full throated advocate of that position.  I address that claim in my article linked above.  As to her position on Smiley, I recommend Kristin E. Hickman & Richard J. Pierce, Jr., Administrative Law Treatise § 4.7 Retroactive Rulemaking (Walters Kluwer 6th ed. 2019).  The retroactivity issues the authors discuss in that section are real issues only if notice and comment regulations that do no more than reasonably interpret ambiguous statutory text and thus attract Chevron deference are legislative regulations rather than interpretive regulations.  If, however, regulations that do not more than reasonably interpret statutory text and thus attract Chevron deference are interpretive regulations rather than legislative regulations, the regime for testing retroactivity is not legislative regulation retroactivity but interpretive regulation retroactivity.  Generally, interpretative regulations with reasonable interpretations inherent in the statutory text since enactment of the ambiguous may be retroactive to the effective date of the statute because, as many have noted (and the Smiley quote indicates), there is no unfairness in applying that interpretation because where no reasonable fixed and reasonable expectations to the contrary existed.  Where, however, there is something that might have given rise to a some fixed and reasonable different interpretation (e.g., as intervening judicial decision in Brand X or some prior agency interpretation) in the interim, then there are tools within the interpretive regulation paradigm (such as prospective application for the new interpretation) for dealing with any settled expectations that may have arisen in the meantime.

Added 4/24/20 1:52pm: 

On the last point as to retroactive effect of an interpretation, in Maui v. Hawaii Wildlife Fund, ___ U.S. ___, ___ S.Ct. ___  (4/23/20), here, Justice Breyer in the majority opinion said (Slip Op. 12):

Neither the Solicitor General nor any party has asked us to give what the Court has referred to as Chevron deference to EPA’s interpretation of the statute. See Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U. S. 837, 844 (1984). Even so, we often pay particular  attention to an agency’s views in light of the agency’s expertise in a given area, its knowledge gained through practical experience, and its familiarity with the interpretive demands of administrative need. See United States v. Mead Corp., 533 U. S. 218, 234–235 (2001); Skidmore v. Swift & Co., 323 U. S. 134, 139–140 (1944). But here, as we have explained, to follow EPA’s reading would open a loophole allowing easy evasion of the statutory provision’s basic purposes. Such an interpretation is neither persuasive nor reasonable.

I have bold-faced the language I focus on here.

Thus, if in the example, the transaction or conduct in issue in the litigation occurred before the maximum retroactivity permitted by § 7805(b), presumably Chevron would not apply but, for the reasons stated by Justice Breyer, the court may consider with "particular attention" the agency's views and, perhaps in theory, that might be enough to convince the court that the agency interpretation  is the best interpretation (perhaps a form of Skidmore deference even without whatever retroactivity means in § 7805(b)).  That is not deference in the Chevron sense, but it is something more than leaving the court without guidance as to the interpretation to apply to resolve the litigation at hand.