Out of Time? APA Challenges to Old Tax Guidance and the Six-Year Default Limitations Period

We welcome back previous guest blogger Susan C. Morse, who is the Angus G. Wynne Sr. Professor in Civil Jurisprudence and Associate Dean for Academic Affairs at the University of Texas at Austin School of Law.

Is it ever too late to raise an administrative procedure challenge to an old tax regulation?

Consider the pair of cases that has produced a circuit split between the Sixth and the Eleventh Circuits over the adequacy of notice-and-comment for a conservation easement final regulation. (Prior Procedurally Taxing coverage here and here.) The Sixth Circuit held in Oakbrook that the notice-and-comment process was sufficient. In contrast, the Eleventh Circuit concluded in Hewitt that Treasury “violated the Administrative Procedure Act’s requirements” when it promulgated the regulation and that therefore the IRS Commissioner’s application of the regulation was “invalid.” But neither court addressed the question of time. The regulation was promulgated in 1986 – decades before any of the facts arose in either case.

Does time ever limit taxpayers’ ability to raise administrative procedure challenges long after the promulgation of a regulation? Consider 28 U.S.C. § 2401(a), the default limitations period for suits against the federal government. It provides that “every civil action commenced against the United States shall be barred unless the complaint is filed within six years after the right of action first accrues.”

The limitations period analysis turns on when the “right of action” to raise an administrative procedure challenge to a regulation “first accrues.” For instance, in Oakbrook and Hewitt, if this right accrued in 1986, when Treasury promulgated the regulation at issue, then the taxpayers’ claims should have been time-barred. On this theory, the taxpayers were allowed to litigate because the government did not raise 28 U.S.C. § 2401(a) as a defense. (The government can waive the defense, as it’s not jurisdictional.) If the government had raised the six-year limitations period defense, the Oakbrook or Hewitt taxpayer would have had to argue that the right of action first accrued later, when the regulation was applied to the taxpayer’s case, or that an exception to the limitations period should apply.

read more…

Now the government has begun to raise the six-year limitations period defense, first in July 2022, in the Govig case, pending in the federal district court in Arizona. Govig involves Notice 2007-83, which was issued nine years before penalties were first proposed on Govig for the 2016 tax year, relating to the employee welfare benefit arrangement established by the taxpayer in 2015. In Govig, the taxpayer claims that the Notice is invalid because it was issued without notice and comment, and relies on the Sixth Circuit’s decision in Mann Construction. In Mann Construction, though, the government did not raise the six-year limitations period defense.

More than half the Courts of Appeal – the Second, Fourth, Fifth, Sixth, Ninth, Eleventh, D.C., and Federal Circuits – have accepted that for administrative procedure claims, the default six-year limitations period begins to run when the challenged regulation or guidance issues, in other words at the time of final agency action. This limitations period statute exists against the background of sovereign immunity, meaning that it is an exercise of Congressional power to specify on what terms the federal government may be sued. In contrast, for certain other claims, such as claims that the agency exceeded its statutory authority, the limitations period begins to run when the regulation or guidance is applied. This is called the Wind River doctrine after a key 1991 Ninth Circuit case. The Wind River doctrine would say that the limitations period for an administrative procedure challenge to Notice 2007-83 began to run in 2007 and expired in 2013, before any relevant facts arose in the Govig case.

It may seem an awkward reading to suggest that a “right of action first accrues” with the earlier issuance of a Notice, especially when the specific controversy between the taxpayer and the government arises from later enforcement proceedings. And yet that is what the cases hold. As an example, consider Sai Kwan Wong, a 2009 Second Circuit case where the plaintiff sought to challenge a Medicaid rule that treated social security disability income as an amount that offsets Medicaid funding of nursing home care, even if that income was deposited into a special needs trust. The Department of Health and Human Services had promulgated a rule providing this offset treatment in 1980, apparently without using notice and comment. The plaintiff did not have standing until 2006, when his legal guardian began to direct the plaintiff’s disability income to a special needs trust, thus raising the question of whether the offset rule would apply. The Second Circuit held that the six-year limitation period began to run in 1980, when the guidance issued, and not in 2006, when the plaintiff had standing. It then barred the plaintiff’s administrative procedure claim.

The theory that underpins cases like Sai Kwan Wong is articulated in Shiny Rock, a 1990 Ninth Circuit case that preceded Wind River by one year. There, the court explained that any injury “was that incurred by all persons .. in 1964” when the Bureau of Land Management issued a public land order – not in 1979, when Bureau applied the order to deny the plaintiff’s mineral patent application. The Shiny Rock court suggests that the rights that are vindicated by an administrative procedure challenge are the general public rights to participate in the administrative procedure process. A later-accrual approach, added the Shiny Rock court, “would virtually nullify the statute of limitations,” since it would always be possible for the old administrative order to applied later, to a new plaintiff who had later gained standing. Viewed this way, the case law consensus that the limitations period begins to accrue when a regulation is promulgated makes sense.

An alternative reading of 28 U.S.C. § 2401(a) might be that a particular plaintiff’s right of action cannot accrue until the plaintiff has standing. This reading is grounded in a private law understanding of the statutory provision, which envisions the government as party to a contract or tort action that arises from a specific transaction or interaction between the government and a plaintiff. But administrative procedure violations are not like these private law causes of action. They arise not from a specific interaction between government and plaintiff, but rather from the alleged failure of a process that is supposed to serve the general public function of producing better administrative law.

Thus, in Govig, if the District of Arizona follows Ninth Circuit precedent, it should conclude that the administrative procedure challenge to Notice 2007-83 is time-barred – unless, of course, the Govig plaintiffs can persuade the court that an exception to the limitations period applies. There is little in the facts of Govig that would support an equitable tolling or equitable estoppel argument. For instance, the government did not hide information or delay enforcement in order to wait out the limitations period. Instead, the facts of the case did not arise until after the limitations period had expired.

An issue that may arise in Govig relates to intervening case law. This is because the Govig plaintiff arguably relies on CIC Services, a 2021 Supreme Court case that held that some facial or pre-enforcement challenges are permitted in tax, despite the Anti-Injunction Act. (Prior Procedurally Taxing coverage here, here, and here).Historically, intervening case law has restarted the 28 U.S.C. § 2401(a) limitations period when a case has only prospective effect – but not if the case has (as is typical) retroactive effect. Plus, more recent Supreme Court precedent emphasizes that its applications of federal law “must be given full retroactive effect …as to all events, regardless of whether such events predate or postdate the announcement of the rule.” The intervening case law argument seems unlikely to offer the Govig plaintiff an exception to the time limitation of 28 U.S.C. § 2401(a).

The Govig case is one to watch. If the Arizona federal district court follows prevailing case law, it will likely allow the government’s limitations period defense and time bar the plaintiff’s administrative procedure claim. The availability of such time bars would reshape the landscape of administrative procedure in tax by putting APA claims on the clock and replacing the assumption that the government will waive the 28 U.S.C. § 2401(a) limitations period defense.

For further reading, if of interest: I have posted a preliminary draft here with additional analysis of this limitations period issue.

Another Offer Denied, Another Reason for Submitting in Collection Due Process

I keep something of a running tally of cases where my clients would have been unjustly treated if not for the protections of Collection Due Process (CDP) hearings. As a practitioner, I think it helps me in advising and counseling my clients on a course of action: what are the pros and cons of proposing a “collection alternative” prior to doing so in a CDP hearing?

The main “con” is pretty straightforward: if the IRS does something crazy, you’re stuck arguing with the IRS (and not a court) about it. The recent case Richard Dillon et al. v. United States et al. illustrates that point nicely, both in terms of how impossible it is to get judicial review on Offers in Compromise outside of CDP, and how stuck you can be with a completely unreasoned IRS determination.

read more…

The Dillons apparently wanted to submit an Offer but couldn’t get their feet in the door -it was returned as non-processible. Rather than give up on it, they went to federal district court on an Administrative Procedure Act (APA) argument. The Dillon opinion resolves on a jurisdictional issue: whether there is a waiver of sovereign immunity. This, in turn, involves the intersection of the Anti-Injunction Act and the APA. While the APA generally waives sovereign immunity when there is an allegation of agency error and the requested relief is not for money damages, the waiver does not apply “if any other statute that grants consent to suit expressly or impliedly forbids the relief which is sought.” The Dillon court, like other federal courts before, concluded in part that the Anti-injunction Act and the tax exception to the Declaratory Judgment Act directly applied and thus under the APA “forbids the relief” that the Dillons sought.

Because of this jurisdictional issue the underlying substance of the complaint concerning the Offer is not given much attention. However, from the opinion we can glean the following:

The Dillons are married, live in St. Paul and owe about $150,000 in back taxes from 2011 – 2017. They are also “approaching retirement” and have about $180,000 in their retirement account. Every other aspect of their finances goes largely unsaid and apparently did not much matter.

I am not a financial advisor, but that vignette conjures up a married couple that will likely struggle during retirement. Were they to liquidate their retirement accounts to pay their back debts they would have virtually nothing left after taxes. On these facts alone, liquidating their retirement accounts would very likely cause economic hardship in the Dillons most vulnerable years (i.e. during retirement). That’s how I see it, and that is also apparently how the Dillons’ attorney framed the Offer to the IRS.

Fortunately, the Treasury Regulations provide for exactly these sorts of Offers -ones where the taxpayer could theoretically full-pay, but a parade of horribles would ensue if they did. They are called “Effective Tax Administration” Offers and some examples are provided at Treas. Reg. § 301.7122-1(c)(3).

Unfortunately, the IRS is often loathe to accept Effective Tax Administration Offers. From both anecdotal evidence and this rather dated article the evidence suggests that getting an Effective Tax Administration Offer accepted is a massively uphill battle.

At its worst, the Dillon case shows just how bad the IRS may be at evaluating these Offers. Apparently, the Offer was “returned” (that is, not processed) without Appeal rights because the IRS determined that it was submitted “solely to hinder or delay” collection. See IRC § 7122(g) and Treas. Reg. § 301.7122-1(f)(5)(ii).

That’s pretty bold, and given the scant information I have, pretty ridiculous. For one, the debts aren’t likely to expire imminently. For two, the stated rationale (“we can collect more than you’re offering”) obviates the entire purpose of ETA Offers, which will always involve an Offer less than “Reasonable Collection Potential” -that’s their whole raison d’etre. Virtually every ETA Offer would be denied (indeed, go unprocessed) if submitting one for less than RCP was seen as intended to hinder or delay collection.

However, I’ve also been informed by those in the know that the IRS tends not to make “solely to hinder or delay” determinations lightly. Sometimes this appears to happen when the taxpayers account was assigned to a Revenue Officer (RO) and the taxpayers file an Offer to essentially take the matter out of the RO’s hands. If the RO has seen some bad taxpayer behavior (transferring property to nominees, etc.) they will reach out to the Offer unit and advise them to make a “solely to hinder or delay” determination.

So perhaps there are good reasons why the Offer unit didn’t let this ETA Offer through the door. I’d sure like to see some more facts…

Alas, no more facts are to be found. Indeed, the Dillons want more than just additional reasoning behind the IRS conclusion: they demand that the Offer be processed via a “writ or order” of the court. That remedy is why they went to court, and ultimately why they are unsuccessful on jurisdictional grounds. Note that if they were in the Tax Court on a CDP determination, however, not only would jurisdiction be clear cut, but the reasoning behind the “solely to hinder or delay” determination would be front and center. And to me, that would provide an important check on what may (or may not) be an irrational decision by the IRS that you’re otherwise stuck with.

Lessons Learned: Advising Clients on Offers

In the past, I mostly considered how “complex” my client’s Offer was in my advice about whether it was worthwhile to wait for a CDP hearing. If it was even remotely complex there was a good reason to wait for CDP.

Dillon emphasizes how low of a bar “complex” is in the Offer context. In my experience, virtually any Offer where the client wasn’t going to pay the full present value of their retirement account (say, because they were entering retirement) or the full equity in their home (say, because they had horrible credit and couldn’t borrow against it) has been enough of a reason to counsel waiting for CDP. I have seen too many times where the IRS review is just to look at the far-right column of the Form 433-A OIC and compare it to the total liability, ignoring any reasons why that is inappropriate that we’ve put forth in a narrative, and leave it at that: “Oh, you’re 64 and have $40,000 in retirement? Then it should be no problem at all to full pay a $30,000 liability.” I assure you this is barely a caricature of how the analysis tends to play out.

But Dillon (and to an extent, the Brown saga) raise other reasons to wait on CDP even if it isn’t a “complex” Offer. Two reasons immediately come to mind.

First, and obviously, in the absence of CDP you are stuck with bad or unjustifiable preliminary decisions made by the IRS as to whether your Offer is “processible.” Without CDP you can’t even get your foot in the door to dispute it, no matter how slam-dunk the underlying Offer may be.

Dillon underscores that problem. I’ve had cases where obvious Offer candidates are “returned without appeal rights” for failing to make quarterly payments where no such payments were required or failing to file a return when they did, in fact, file but the IRS rejected the return for ID verification issues. One was saved by the grace of CDP. The other will have a student assigned to it this semester.

Second, CDP keeps the IRS honest. And frankly, I’ve seen time-and-again that such value cannot be understated. Many of my clients just want to be heard: none of them are doing particularly well financially as virtually all of my clients being under 250% of the federal poverty line, per IRC § 7526. Yes, the IRS has a fair amount of latitude in when to accept an Offer, but if it is going to be rejected my client and I want to know the reasons why. Absent CDP I find that we are never given anything beyond a boilerplate “we’ve considered your circumstances and determined they do not justify accepting your Offer.” When pressed on this (which I really only can do with any value in CDP reviewing the administrative file) I often find that the IRS never really appeared to “consider the circumstances” at all.

A Parting Thought: The Value of Judicial Review on Collection Actions

As an academic who cares about tax administration, I’m not a huge fan of incentivizing people to “wait” on addressing their tax problems. It’s inefficient and costly.

(Note that the Tax Court may have inadvertently (and in my opinion incorrectly) further incentivized waiting until CDP in the context of arguing the underlying merits of the liability under IRC § 6330(c)(2)(B). As covered here, here, and here, if you are proactive in arguing against the underlying tax you may miss out on the chance to get court review later. I definitely don’t trust IRS Appeals enough on deficiency issues with low-income taxpayers (the common “prove the kid lived with you” scenario) to foreclose judicial review.)

I hope that the (much needed) increase in IRS funding (see Les’s post) will ameliorate some of the issues I’ve seen through better training and support. But even if it improves the quality of review on collection issues (a big “if” since it isn’t clear how much of that money would be going to exam, etc.) I believe the need for judicial review of collection actions remains. If you ever have a position contrary to the IRM, you have virtually no chance of success without judicial review.

And sometimes the IRS rules, frankly, are nonsensical or needlessly hurt low-income taxpayers. Vinatieri is the classic example, and Keith has noted other times where it seems that those in positions of power seem to just “make-up” rules. A judiciary check against that power, even if modest, I think, is in order. Most of my low-income clients are collection cases, and collection (the actual taking of their modest property) is a serious concern that can carry serious consequences if you get “the wrong person” at the IRS reviewing the case.

As a parting example, I once submitted an Offer for a homeless client. It was preliminarily rejected by the IRS on their determination that the taxpayer had disposable income because… yep, they weren’t paying for housing. Because the IRM (basically) says “take actual housing expenses,” and since stable housing was just a dream of theirs, they shouldn’t be allowed $1,000 in anticipated monthly housing costs while trying to leave a domestic violence shelter.

I kid you not.

I am convinced this particular case was resolved favorably only because I was able to say to Appeals, as they informed me they were upholding the determination: “do you really want this to play before a Tax Court judge?” I have heard from other colleagues (especially in California) that they’ve run into the homelessness housing expense issue before, with both failure and success. I don’t think the IRS, as an entity, endorses the position that homeless people shouldn’t be allowed the means to pay for housing. But that’s what the frontline workers (and the “Independent Office of Appeals”) end up doing based on their reading of the IRM. Part of me wished my case went to trial so that it could have gained some notoriety, and either led to the IRS changing the IRM on its own or a Tax Court decision that forced its hand.

But for now, I’ll settle for the fact that at least my particular client was given a positive outcome that would not have occurred in the absence of CDP. I’m sure there are other such examples from practitioners nationwide.

APA Provides No Basis To Compel IRS To Provide Access To Appeals

Add Rocky Branch Timberlands v US to the list of interesting cases invoking the APA to challenge IRS actions.  As the case shows, the APA can be a taxpayer’s friend, but it is does not provide the means to get a court to second guess what the court believes are decisions that are solely committed to the agency’s discretion.

The case involves yet another audit of taxpayers claiming conservation easement deductions. The taxpayers, subject to the not quite dead TEFRA partnership audit procedures, initially refused to extend the SOL on assessment. With the three-year period coming close, the IRS issued a Notice of Final Partnership Administrative Adjustment (FPAA). The taxpayers changed their mind about the extension, and sent a signed 872-P extension form to the audit team and sought review before Appeals. IRS declined to sign the 872-P and refused to revoke its FPAA. The taxpayers sued, asking the court to order the IRS (1) to rescind the FPAA and (2) to sign the Form 872-P that would extend the time to assess, all with the hope of getting before Appeals and not having to take the case to Tax Court.

read more...
 

The government filed a motion to dismiss for lack of jurisdiction.

The court quickly found in favor of the government. As to the taxpayer’s request to rescind the FPAA, the court noted that the Anti Inunction Act prevented the court from ordering an injunction that would restrain assessment.  The Supreme Court has carved an exception to the AIA when there is no adequate remedy and the plaintiff shows that the government could not prevail. Here, the opinion notes that the FPAA did in fact trigger a separate remedy: the chance to challenge the denial in Tax Court, one in which the taxpayer through the tax matters partner had in fact done.

As to the demand that IRS sign the extension/Form 872-P, the court looked to the APA, noting that while it waives sovereign immunity there is no basis for a court to gain jurisdiction if the decision is committed solely to the agency’s discretion. In addition, there is no jurisdiction if the agency decision is not final. Both of these exceptions have a fair bit of nuance described in Saltzman and Book, IRS Practice & Procedure in our soon to be revised but still really good discussion of the APA.

But there was not much nuance presented here. As to not being final agency action:

The IRS’s decision not to sign the Form 872-P, and thereby decline to extend the statutory period, was not a final agency action within the meaning of [5 U.S.C] § 704.  Rather, it was an intermediary and procedural step leading up to the issuance of the FPAA and did not alter Plaintiffs’ rights or obligations.  The IRS’s decision not to sign the Form 872-P did not alter the limitations period.

As to discretion, the court similarly dismissed the taxpayer’s argument:

The IRS’s decision not to extend the statutory period was also discretionary.  Plaintiffs identify no requirement that the IRS agree to an extension, and the Court is aware of none.  To the contrary, the law provides the statutory period may be extended only upon agreement by the taxpayer and the IRS.  See 26 U.S.C. § 6501(c)(4); Feldman v. Comm’r, 20 F.3d 1128, 1132 (11th Cir. 1994).  This provision clearly provides the IRS discretion—co-equal to a taxpayer’s discretion—as to whether it will extend the statutory period.  It is strange that Plaintiffs would deny the IRS the same discretion is previously exercised in the very same review. 

A somewhat more interesting issue involved the taxpayer’s argument that the court should give some teeth to IRC 7803(e)(4), which as part of the Taxpayer First Act provides that review by Appeals (renamed Independent Office of Appeals) “shall be generally available to all taxpayers.”

As the taxpayer had sought access to Appeals prior to the issuance of the FPAA, the opinion noted that the request was moot in light of its decision to find that the AIA precluded the IRS from rescinding the FPAA. For good measure, the opinion stated that the APA would not help, as the discretion to decline to provide access to Appeals was solely vested in IRS. In finding that this was committed to the agency’s sole discretion, the court accepted the government’s analogy settlement power generally, as it is settled law that an agency’s decision to settle (or not) is solely a matter for agencies to determine.  

The opinion also concluded that the point in time that the taxpayers sought relief in the form of Appeals consideration, prior to the issuance of the FPAA, was merely an interim step in the agency process, and thus not final agency action.

Conclusion

The taxpayers have challenged the IRS’s denial of their deduction in Tax Court, so it is not as if they are without recourse to independent review on the merits. And the taxpayers’ initial refusal to sign the extension contributed to their problems. Add to the mix that there is not a lot of sympathy around those claiming easement deductions.

Yet, access to Appeals is a fundamental part of tax administration and embedded in the Taxpayer Bill of Rights. The language in this opinion, as well as other cases we have blogged about (including the Facebook case I discussed here)have held that access to Appeals is something that IRS can allow, or not allow, for essentially any reason. The TFA provides protections for taxpayers denied access to Appeals following the issuance of a stat notice.  Prior to that time, IRS controls the process, and the APA will likely not help.

The More Things Change The More They Remain The Same

Today’s post is the last in the three-part series from Professor Bryan Camp addressing how to classify tax regulations under the APA. In today’s post, Bryan considers whether current tax administration involves social policies and non revenue raising functions more so today than the past. He concludes by offering observations on two recent cases, Oakbrook Land Holdings v Commissioner and Rogerson v Commissioner. Les

To recap: we are concerned with the question of how must Treasury regulations be promulgated to be in conformity with the APA.  All agencies must conform to the APA.  No one doubts that.  The discussion is about the proper relationship—or fit—of Treasury Regulations to the APA.  Jack Townsend posted his views that the APA does not require most Treasury regulations to be issued through the notice and comment process because they are interpretive regulations and not legislative regulations.  Kristin Hickman posted her views that ALL Treasury Regulations are legislative.  And she says her views are the new orthodoxy. 

But if hers is the new orthodoxy, there are still heretics.  I’m one.  I agree with Jack that most Treasury Regulations are properly classified as interpretive regulations.

read more...
 

In my last post I attempted to show how everyone in the 1940’s and 1950’s believed most Treasury Regulations were interpretive.  Kristin properly responds with a “so what.”  Times change and Kristin pushes the idea that the function of tax administration has transformed from revenue raising to social policy implementation.  She’s not the only one.  I encourage readers to check out these two great articles:  Susannah Camic Tahk, Everything is Tax: Evaluating the Structural Transformation of U.S. Policymaking, 50 Harv. J. Legislative 67 (2013); Linda Sugin, The Great and Mighty Tax Law: How the Roberts Court has Reduced Constitutional Scrutiny of Taxes and Tax Expenditures, 78 Brooklyn L. Rev. 777 (2103).

So today’s post is to take a look at the claim that tax administration has changed sufficiently to change how we classify Treasury Regulations.  Big caveat here: it also may be that other changes in the law or in society now make it appropriate to classify Treasury Regulations as legislative and not interpretive.  I’m not going there today. 

Much of the recent scholarship on how tax administration has changed focuses on various transfer programs—notably the Earned Income Tax Credit (EITC) and the Affordable Care Act (ACA). Some commentators have suggested that the social welfare function served by the EITC should trigger a reformed due process analysis for tax administration.  I really like this article by Megan Newman, Low-Income Tax Gap: The Hybrid Nature of the Earned Income Tax Credit Leads to its Exclusion from Due Process Protection, 64 Tax Lawyer 719 (Spring 2011).  And Les keeps telling me that tax issues for Low Income Taxpayers are really issues about subsidies.   He gives a great presentation of these views in his article Nina Olson: A Champion for Taxpayer-Centered Tax Administration, 18 Pitt. Tax Rev. 117 (2020).  Surely it seems that the IRS is now tasked with jobs would be foreign to those living in the 1940’s and 1950’s.

I agree with much of that.  The basic idea is that much of the tax laws serve non-revenue raising functions.  That necessarily means that what are ostensibly tax regulations may serve those functions as well, including poverty relief, etc.  I get that.  What I am skeptical about are claims that tax administration today involves social policies more than it did at or before enactment of the APA.  Remember, that’s our inquiry: has tax administration changed such that tax regulations today are doing something qualitatively different than they did in the 1940’s and 1950’s when everyone agreed with Professor Davis that “the great bulk of Treasury Regulations under the tax laws clearly are interpretative rules, not legislative rules….”  Davis, Administrative Law Text (1959) at p. 87. 

I have three reasons to doubt claims that tax administration today involves social policies more than it has in the past. 

First, I don’t even know the baseline.  How do you measure the extent to which tax laws are “oriented” towards or away from revenue collection?  In her “Administering The Tax System We Have” article, Kristin made a good stab at it.  She tried to quantify by studying tax regulations over a three year period.  She created a fairly elaborate coding system and concluded that a substantial % of tax regulations were “oriented away” from revenue raising.  One of several difficulties with that project was that she had no historical baseline.  So while she could draw some (debatable) conclusions about the current allocation of administrative efforts between “revenue-raising” and “social policy” implementation, she could draw no historical conclusions. 

Second, it is not clear whether this recent tax scholarship reflects real change or just a change in awareness among tax academics.  Collecting taxes has always been intimately bound up with social policy.  Scholars in other disciplines have known this for forever.  Economics know this.  Check out Joseph Schumpeter’s 1918 classic “The Crisis of the Tax State”, an extended (and rather hyperbolic) examination of the relationship between taxation and social economy.  Historians know this.  See, Isaac William Martin, Ajay K. Mehrotra, and Moica Prasad, The New Fiscal Sociology:  Taxation in Comparative and Historical Perspective (Cambridge University Press 2009), collecting essays from many historians.  No free link to the book, but you can check out my review of it in the Am. Journal of Legal History.  In particular, historians of slavery know this.  See e.g. Robin Einhorn, American Taxation, American Slavery (U. Chicago Press, 2006).  Political Scientists know this. See, e.g. Julian E. Zelizer, Taxing America:  Wilbur Mills, Congress, and the State, 1945-1975 (Cambridge University Press 1998).  Folks, that’s just pulling off the top of my head from stuff I’ve actually read.  I’m sure a little digging will reveal much more.

Third, while I agree that the objects of Congressional solicitude have expanded (EITC, ACA premium credits, clean energy tax credits), I am skeptical that this represents a fundamental shift in the use of the tax laws.  I think it rather represents a shift in which social policies get put in the tax laws.  History gives us many examples of how Congress has used the taxing power for purposes other than revenue raising.  I think a few examples from this history shows the difficulty in arguing that tax administration is now, in any relevant sense, more “oriented away” from revenue raising than it was before the APA. 

We tend to think that because historical actors had simpler technology they also had simpler minds.  They did not.  They just did not have the internet.  Throughout history there have always been really strong non-revenue raising policies embedded in tax provisions, and folks have always been acutely aware of the non-revenue effects of tax laws. 

Let’s take a look.  

Start with Formation of the Republic.  The first use of tax law to further a social policy and not raise revenue is right there in Article 1, Section 2, Clause 3 of the U.S. Constitution.  It’s the great compromise:  in exchange for allowing enslaved persons to be counted for representation purposes, the Constitution also created a tax break: “direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Numbers, which shall be determined by adding to the whole Number of free Persons, including those bound to Service for a Term of Years, and excluding Indians not taxed, three fifths of all other Persons.”  That’s a tax law. And it was about must more than revenue raising!  It was about embedding slavery into the new American republic.

Not everyone bought into the compromise.  Some Southerners opposed it because they feared Congress would use taxation to achieve abolition.  Since enslaved persons were property, abolitionists would use the federal power to impose “a grievous and enormous tax on it, so as to compel owners to emancipate their slaves rather than pay the tax.”  That’s from the formerly sainted Patrick Henry, as quoted in Robin Einhorn, American Taxation, American Slavery (U. Chicago Press, 2006). at 181.  Others disagreed with Henry’s analysis, but the very robustness of the debate demonstrates the awareness in the 1780’s of how tax systems are used for non-revenue purposes.

Move to the early republic.  Understand that before 1862 the federal government’s revenue came almost entirely from tariffsHistorical Statistics of the United States, 1789-1945, Table Series P 89-98 (“Federal Government Finances”), pp. 295-298.  Ok.  Do YOU want to argue that tariff legislation did not have a significant non-revenue function?  I thought not.  But if you really think that tariffs were even mostly about revenue raising, I recommend you read this commentary from one of the leading tariff lawyers of his day:  William McKinley, The Tariff in the Days of Henry Clay and Since: An Exhaustive Review of Our Tariff Legislation from 1812 to 1896 (Henry Clay Publishing Co., 1896).  He’ll set you straight.  And the title is not misleading.  It’s an exhaustive read.  While we might today ignore tariffs because they account for so little of the federal government’s revenue, remember again that in the mid-1800’s, tariffs were pretty much the entire funding mechanism. 

And yet it was during this very time—mid 1800’s, when tax administration was chiefly focused on tariff administration—that courts created the doctrines regarding both issuance and authority of tax guidance that lawyers in the 1940’s believed to be in harmony with the commands of the APA.  That’s in my History of Tax Regulations article.

Move to the Civil War.  Few would argue that the Revenue Act of 1862 was anything other than a revenue raiser.  Yet even there Congress wrestled with social policy, in the form of what  tax breaks to give various classes of taxpayers.  That’s what we call “tax expenditures” today.  For example, shortly after Congress enacted the very first income duty in 1862, folks pointed out that the failure to tax owner-occupied housing constituted a tax subsidy that discriminated against renters.  There was back and forth on this issue for several years.  Congress even created a special Commission to study the problem.  But Congress refused to tax the imputed income of self-owned property, for policy reasons.  In fact, Congress explicitly excluded the rental value of self-owned homes from gross income but, to equalize, permitted a deduction for house-renters of their house-rents. Revenue Act of 1864, 13 Stat. 223, 281 (§117).  When the income tax was reinstated in 1913, Congress dropped the house-rent deduction (but kept the imputed income exclusion), over the objection of then Senator (and future Justice) Sutherland, who protested this discrimination against house renters. See Seidman’s Legislative History at 992-993.

Move to 1913.  Let’s talk charities.  When it revived the income tax in 1913, Congress had social policy reasons for exempting certain organizations if they were organized “exclusively” for various purposes, including religious, charitable, scientific, or educational purposes, so long as “no part of the net income of which inures to the benefit of any private stockholder or individual.” 38 Stat. 114, 172. 

Whatever you believe the social policy for tax exempt organizations to be—and Professor Atkinson gives a thorough review of many of them in his article Theories of the Federal Income Tax Exemption for Charities:  Thesis, Antithesis, and Syntheses, 27 Stetson L. Rev. 395 (1997)—the policy is in considerable tension with the revenue function of tax, even if the JCT (for reasons I do not understand) does not include this exclusion in its yearly list of tax expenditures.

The BIR was left to grapple with drawing the social policy lines in the 1910’s and 1920’s, long before the APA.  For example, would income derived from unrelated business activities be subject to tax if it was used entirely for an organization’s exempt purpose?  The BIR said yes; the Supreme Court said no. Trinidad v. Sagrada Orden de Predicadores de la Prvincia del Santisimo Rosario de Filipinas, 263 U.S. 578 (1924). What about a company formed to provide employment to members of a certain labor union, all of whose the profits went to the labor union?  Was its income exempt?  The BIR said no.  And here’s a funny thing:  that position was embedded in sub-Treasury guidance and never challenged.  Legislative rule?  Interpretive rule?  The original decision was in Office Decision (O.D.) 523, 2 C.B. 211 (1920).  The position was re-affirmed by Rev. Rul. 69-386, 1969-2 C. B, 123, and up until the TE/GE “scandal” was found in IRM 7.25.5.2.4. 05-21-2014)(“Nonqualifying Activities”). I don’t know where it is now. 

But What About the EITC?  It is commonly believed that Congress created the EITC in 1975.  See, e.g. Dorothy Brown, The Tax Treatment Of Children: Separate But Unequal, 54 Emory L. J. 757, 766 (2004).  Not quite.  It is more accurate to say that Congress revived the EITC in 1975.  The EITC first appeared in §206 of the Revenue Act of 1924 in the form of a tax credit equal to 25% of earned income. The credit continued until 1932 when it was demoted to a deduction.  This smaller relief eventually died in 1944, succumbing to the revenue demands of WWII, the expansion of the class tax to a mass tax, and the creation of the standard deduction.  

The social policy behind the 1924 EITC was similar to that of the 1975 EITC and was equally in tension with revenue-raising.  Actually, it appears more in tension with revenue raising.  Both policies were intended to “orient” the law towards promoting social justice and “orient away” from revenue raising.  But guess what, folks.  Concepts of social justice change.  The current EITC is viewed as being in lieu of welfare payments to the poor, the idea being that this netting mechanism is more efficient than having one hand of the government paying out benefits while the other hand collects taxes.  The 1924 EITC had a similar social policy but directed at a different class of taxpayers: wealthy wage earners who had the same income as those whose income was due to returns from capital.  And why did Congress view that as social justice?  Why because those wage-earners needed help to save for their retirement in a way that those who were able to generate income from capital did not need.  And since those earning income from captial had gotten a HUGE tax subsidy starting in 1921, wage earning accounts, lawyers, and doctors believed they were getting shafted.  Equalizing those classes of taxpayers involved a potentially larger hit to revenue than shifting welfare payments to the poor into the tax laws.  The legislative history contains back-and-forth debates in Congress over the extent to which revenue needs outweighed the social policy, and vice versa.  You can find the grody details in Bryan Camp, Franklin Roosevelt and the Forgotten History of the Earned Income Tax Credit, 20 Green Bag 2d. 337 (2017).

My Bottom Line: Certainly tax administration has changed since enactment of the APA.  But the changes that are perhaps most relevant to administrative law are procedural changes, not a reorientation of the tax system or a re-balancing of revenue and non-revenue functions.  That is, while I am skeptical that the tax system created by Congress has changed in ways that create a different relationship with the APA, I am concerned that the tax system administered by the Service has changed, and dramatically.  As I spell out in excruciating detail elsewhere, tax administration is now driven by computer and computer coding. Bryan T. Camp, Theory and Practice in Tax Administration, 29 Va. Tax Rev. 227 (2009). This debate about “legislative” and “interpretive” regulations is less important, I think, than recognizing that agency rules encoded in computers have a “force of law” in a much more immediate and real-world way than do the issuance of tax regulations.  Just ask anyone whose electronic filing is rejected!  But I don’t hear Kristin or anyone demanding that every computer coding change go through notice and comment process!  I think these changes create a need to rethink conceptions of due process, but leave that discussion to a different time and place.

Concluding Thoughts: The Weird Results If All Regs Are Legislative

Kristin’s views are definitely trending and becoming the new orthodoxy. In fact, you are likely committing malpractice if you are not mounting a procedural attack on any and all administrative guidance adverse to your client!  

But some weirdness is starting to appear.  Go read Judge Guy’s concurrence in Oakbrook.  Then go read Judge Toro’s recent opinion in Rogerson v. Commissioner, T.C. Memo 2022-49 (May 12, 2022).  Or just keep reading my scribbles here.

The issue in Oakbrook was whether the taxpayer was entitled to a massive deduction for a conservation easement.  The statute requires that a conservation purpose must be guaranteed in perpetuity. §170(h)(5)(A).  Treasury has issued a regulation saying that to meet the perpetuity requirement easement agreements must provide for a particular distribution of proceeds if unexpected events require the later sale of the property subject to the easement. Treas. Reg. 1.170A-14(g)(6)(ii).  The IRS did not think the Oakbrook easement met the statutory perpetuity requirement because it did not meet the regulatory requirement.

Channeling Kristin’s views of tax regulations, the taxpayer argued that the regulation was invalid because Treasury had not properly promulgated it.  The government said “no, we issued the regulation properly.” Two of the judges in Oakbrook agreed with the government and found that Oakbrook did not comply with this validly issued regulation, splitting with the 11th Circuit.

Judge Guy, however, took a different approach.  He agreed with the taxpayer that the Treasury Regulation was improperly issued and so was invalid.  In fact, he recites the Myth of Mayo by using it to justify, without explanation, his sub-silentio conclusion that the proceeds regulation was a legislative regulation.  Here’s what he says, using the money quote from Mayo:

“The Department of the Treasury must play by the same rules as other federal agencies. The Supreme Court made that clear when it refused to carve out an approach to administrative review good for tax law only and expressly recognized the importance of maintaining a uniform approach to judicial review of administrative action.”  (internal quotes omitted)

I hope you see the myth there.  The battle in Oakbrook was not a battle about what deference to give the regulation; it was about whether the regulation was validly issued.  A void regulation gets no deference because there’s just no there there, just empty space where a regulation might have been.  So who cares what Mayo says about deference?  Kristin is correct that courts do not appear to care whether Treasury regulations are or are not subject to notice and comment requirements.  Everyone in this case assumed the regulation was legislative.  

But was it?  Judge Guy’s concurrence undercuts that assumption.  After finding the regulation void he then went on to say that the taxpayer loses.  He still thought that the easement agreement did not satisfy the perpetuity requirement.  Why?  Well, gosh, he interprets the statute to reach that conclusion!  He applies traditional tools of statutory interpretation to decide what the word “perpetuity” means and then finds that Oakbrook’s easement agreement did not satisfy his judicial interpretation of what “perpetuity” requires. 

Hmmm.  So let me get this straight: when Judge Guy does it we call it “interpretation.”  But a Treasury Regulation doing the same thing is not an interpretive regulation?  It’s a legislative regulation?  That makes no sense to me. 

Rogerson presents a similar scenario.  There, the taxpayer had a bunch of losses from a yacht leasing activity in 2014, 2015, and 2016.  They were passive losses and, as readers know, that meant he could only take them against gains from passive activities.  Thus says §469(a).  It just so happened that he reported a bunch of gains from ownership in three S corporations in those years, all of which he reported as passive activities.  The trouble was, for the nine years prior to 2014, he had reported his involvement in the predecessor corporation as active. 

The issue in the case was whether Mr. Rogerson materially participated in the S corps in the three years at issue.  The statute is notoriously unhelpful here, saying only that a taxpayer’s participation in an activity is active when the taxpayer “materially participates” in the activity.  The statute then says a “taxpayer shall be treated as materially participating in an activity only if the taxpayer is involved in the operations of the activity on a basis which is—(A) regular, (B) continuous, and (C) substantial.” §469(h)(1).  Ok.  Sure.  But what do those words mean?  How are we to interpret those words?  Well, there are regulations for that. 

The applicable regulations say that a taxpayer can satisfy the statutory requirement if they meet any one of seven tests laid out in the regulation.  One of the tests is that the taxpayer has materially participated in the activity for five of the ten years before the year in question.  Temp. Treas. Reg. § 1.469-5T(a)(5). 

Yep.  It’s one of those Temp. regs from long, long ago that was grandfathered when Congress revised §7805 to limit Temp regs to three years.  This one is from 1988.  To complicate matters, Treasury issued associated regulations after that and then modified them with the result, says Judge Toro, that “today, the five of ten test appears in a temporary regulation, while the rule explaining how the five of ten test should be applied appears in a final regulation.”  Op. at 18.

Flourishing the exceptionalism myth, the taxpayer’s attorney attacked the Temp. Reg., trying to avoid having his material participation in prior years count against him in the years at issue.  If the taxpayer could get that nasty regulation voided, then it would appear the taxpayer would be home free. 

Not so fast.  It is true that Judge Toro thought the regulation was valid and applied it.  Bummer for Mr. Rogerson.  But then Judge Toro adds an analysis much like Judge Guy’s concurrence in Oakbrook; he looks at whether Mr. Rogerson’s activities were material participation within the meaning of the statute, disregarding the regulations. 

“To summarize, Mr. Rogerson would not prevail even if he were correct about the procedural validity of the five of ten test, because we find that he was regularly, continuously, and substantially involved in the operations of RAEG during 2014, 2015, and 2016 within the meaning of section 469(h). Accordingly, we need not decide whether the five of ten test is procedurally valid and turn instead to Mr. Rogerson’s final argument.”  Op. at 27.

Hmmm.  So let me get this straight: if you maintain (as Kristin does) that the Treasury Regulation creating the five in ten rule was a “legislative” action and not an “interpretive” action, then is not Judge Toro also engaging in a “legislative” action?  Just like Judge Guy?  Contrariwise, if you say Judge Toro was engaging in mere “interpretive” actions, then how is a Treasury Regulation that does the same thing any different?  How is it not an interpretive action as well? 

There are several answers to that. Kristin’s answer, of course, is the Myth of Mayo: Mayo transformed ALL Treasury regulations into legislative rules because they now carry the mythic “force of law.”  Treasury would agree that this regulation was a legislative one, but for a different reason: the regulation was not issued under the general authority in §7805 but was instead issued under specific authority given Treasury to “specify what constitutes…material participation…for purposes of this section.” § 469(l)(1). 

To me neither are satisfactory answers.   Both are form-over-substance reductions.  Kristin’s answer that “force of law” makes a rule legislative is not only reductionist but also circular.  It obliterates the APA distinction, despite her protestations (she writes “The fact that Treasury regulations do not qualify as interpretative rules does not mean that no agency pronouncements qualify as interpretative rules; plenty of agency pronouncements, including by the IRS, fall under the interpretative rule category for APA purposes. Just not Treasury regulations.”).  Treasury’s answer used to make some sense but has become disconnected from substance over time as Congress randomly authorizes regulations in specific statutes when the general authority would suffice.  In Rogerson, for example, Treasury did not need specific authority to interpret the statutory test for material participation.  Ya got three statutory words: regular, substantial, continuous.  What those words mean is an interpretive task, whether done by a single judge or by the Treasury Department.

I think there’s a better way to distinguish “interpretive” from “legislative” rules.  It involves looking at each agency, and evaluating what the agency action is attempting to do in relation to its organic statute.  It’s an agency-by-agency determination, not some unified field theory of administrative law.  The APA is not a hammer.  But I promised Les I’d keep this under 4,000 words and I’m already over that, so I must leave those thoughts for another day. 

The APA Is Not A Hammer

Professor Bryan Camp follows up from his post yesterday, as he explores the history of the APA and tax regulations to support his view that all tax regulations are not legislative rules under the APA. For Professor Hickman’s post, see It’s Time To Let Go: Treasury Regulations Are Not Interpretative Rules. While this issue may seem a bit academic, it is important, as litigants increasingly challenge the procedural validity of tax guidance in cases like Oakbrook Land Holdings v Comm’r and Hewitt v Comm’r. Les

Kristin Hickman loves the APA.  To channel Jed Rakoff, it’s her Stradivarius, her Colt 45, her Louisville Slugger, her Cuisinart, and her True Love.  It’s her Hammer, her righteous Mjölnir

And when you have a hammer, everything looks like a nail.  Including ALL Treasury regulations.  This is a follow-up post from yesterday to explain why I disagree with Kristin’s contention that ALL Treasury regulations are “legislative” for APA purposes. 

read more...

To recap: we are concerned with the question of how must Treasury regulations be promulgated to be in conformity with the APA.  To answer that question, Kristin starts her analysis with the APA text.  All agencies must conform to the APA.  It’s a hammer.  Kristin has spent her academic career looking for a unified theory of administrative law and she views the APA as the enforcement mechanism to whack all the governmental agencies that pop up their unruly heads.  Agencies that do not conform to a strict reading of the APA must be claiming to be “exceptional” from the law.  That’s the Myth of Tax Exceptionalism I discussed yesterday.

To achieve this trans-agency uniformity, Kristen looks at the words in the APA and gives them a strict, but abstract, meaning:  “legislative” means “force of law.”  She then applies that meaning to various agencies to whack them into conformity.  When she applies it to the Treasury Department she concludes that ALL Treasury regulations have the “force of law” and are, therefore, legislative for issuance purposes under the APA.  For that proposition she looks to the Supreme Court’s decision in Mayo as imbuing all Treasury regulations with that magic “force of law.” 

I certainly agree there are certain uniform principles of law that apply to all agencies.  The biggest one is “do what Congress tells you to do.”  I just disagree that the APA is the right place to start.  I believe one should start with the agency’s organic statute and the case law about that agency.  I start there because no lawyer actually practices something called “administrative law.”  Lawyers practice environmental law, or securities law, or SSA disability law….or tax law.  While Kristin and I might study “administrative law” in the abstract, that’s not how it works in the real world. 

I think the history of the APA shows that it was not intended to be a hammer.  It was not enacted to override or intrude on specific laws applicable to specific agencies.  I think the history of the APA supports reading and applying it not so much as a hammer as a safety net, providing a set of legal principles that agencies should follow.  But there are many ways to obey those principles.  Thus, it’s entirely possible that the APA applies differently to different agencies, depending on the agency’s organic statute.  You need to look at the history. That is particularly true for the Treasury Department. 

Kristin seems to say APA history does not matter, for two reasons.  First, she appears to believe that the APA wiped out all prior agency rules and practices.  It hammered out all that came before.  Second, as to tax administration, she appears to believe that the nature and function of tax administration has dramatically changed since the APA’s enactment.  The 1940’s and 1950’s are no longer relevant.  Specifically, she believes that the tax regulations are now more “oriented away” from revenue raising and “oriented towards” using tax laws to serve non-revenue social policies. She writes “Although the tax system has always served multiple goals, recent decades have seen a dramatic escalation in tax programs and provisions serving purposes other than traditional revenue raising.” “Administering the Tax System We Have,” 63 Duke L.J. 1717, 1728 (2014). 

I disagree with Kristin on both counts.   It’s part of what I call the “Myth of Change”  that I mentioned yesterday.  First, I think it is critical to understand that Treasury was promulgating regulations long, long before the APA was enacted.  The APA was enacted on top of an existing tax guidance structure.  Second,I think tax administration has always been an exercise in balancing revenue raising needs with social policies.  To be sure, the particular social policies that Congress wants to affect through taxation have changed over time, but not the use of the tax laws to do more than raise revenue.   I have not seen convincing evidence that Congress is using the tax laws now more than ever for social policy as opposed to revenue raising.  That’s the Myth of Change.

1. Tax Regulations Came Before The APA

The APA, 60 Stat. 237 was enacted June 11, 1946.  It resulted from the Attorney General Office’s monumental study of federal agencies, published in a famous 1941 Final Report.  That Report is still highly influential on how courts apply the APA. see Joanna Grisinger,  Law in Action: The Attorney General’s Committee on Administrative Procedure, 20 J. of Policy History 379 (2008) (reviewing the influence of the Final Report on the APA).  The Final Report, in turn, grew out of a detailed study of then-existing agencies, a study contained in 27 Monographs written by staff, each running hundreds of pages.  Each one is a book.  Monograph 22 focused on the tax administration, back at a time when the IRS was called the Bureau of Internal Revenue (BIR).

Our understanding how the APA applies to tax regulations should thus start in the 1940’s because unlike chicken and eggs, we actually know what came first: tax regulations!  And then, yes indeedy, we need to see whether tax administration or general principles of administrative law have changed so much as to require a change in that relationship. 

What lessons do we learn from this history? 

(a) The APA was not intended to be a hammer. 

The AG’s Committee “had initially hoped to be able to suggest uniform rules for agency practice” similar to the Walter-Logan bill that Congress had passed and President Roosevelt had vetoed. Final Report at 22 (emphasis supplied).  That’s what Kristin wants.  In light of the information produced in the 27 monographs, however, the Final Report backed away considerably from that aspiration and instead prescribed a general framework for balancing the goals of agency efficiency and autonomy with the goals of agency transparency and protection of individuals from arbitrary agency actions.  See generally, Roni A. Elias, The Legislative History of the Administrative Procedure Act, 27 Fordham Envtl. L. Rev. 207 (2008)(nice short student note). 

That is why the resulting APA was widely understood as standing for the proposition that “procedural uniformity was not well suited to the administrative process.” Grisinger, supra, at 402.  That is, the APA provided generalized standards for controlling administrative actions rather than detailed and strict prescriptions.  

As enacted, the APA incorporated broad conceptual principles of administrative law.  Fundamentally was the binary notion of what agencies did.  Agency action was either an “adjudication” or a “rulemaking.”  Roughly speaking, rulemaking was forward-looking, the process of creating some kind of general statement that would apply to a broad set of situations in the future.  In contrast, adjudication was backwards-looking, the process of applying the law to an existing set of facts. 

As to rulemaking, the APA created large conceptual baskets for types of rules, with associated procedures for their issuance.  Again, as Jack Townsend properly noted and as I discussed yesterday, the APA says nothing about what weight various types of rules should carry with the courts.  The default issuance procedure was notice-and-comment.  Some regulations had to go through a more formal process, but only when Congress specified by using the magic language “on the record after a hearing.”  United States v. Florida East Coast Ry., 410 U.S. 224 (1973).  And other regulations could be issued with less formal process if they were either “interpretative rules, general statements of policy, or rules of agency organization, procedure, or practice” or if the issuing agency had “good cause” to find that “notice and public procedure thereon are impracticable, unnecessary, or contrary to the public interest.”  5 U.S.C. §553(b).  These foundational concepts were roomy enough to accommodate a wide variety of guidance modalities. 

Let’s see what lesson we can find about tax guidance.

(b) Treasury Regulations were not believed to require notice and comment for issuance. 

The contemporary view in the 1940’s was that existing issuance procedures were consistent with this new “constitution” of administrative law.  The concerns expressed in the Final Report related mostly to the new agencies created by the New Deal, agencies that “have been devised by Congress under the pressure of events for the exercise of new powers in new fields.” Final Report at 213 (emphasis supplied).  Thus, nothing in either Monograph 22 nor the Committee’s Final Report suggests there were concerns with Treasury regulations.  Far from it.  Monograph 22 acknowledged that the “considerable history” behind the BIR made it a very different subject from other agencies “which are working in areas only recently occupied by the Federal Government.” Monograph 22 at 146.

The writers of Monograph 22 discuss in some detail the issuance of a variety of tax guidance documents.  They were fine with the then current process.  Public hearings?  That “would probably be of small practical value, since the problems to be studied are of a highly technical or “legal” character…” Id. at p 147.  Moreover, “time…is often a problem.”  Id. at 146.  Well, Duh!  The writers suggested notice and comment would be appropriate “when time allowed.”  Id.  Thus, the vision was that Treasury Regulations would normally be issued without notice and comment unless time allowed or other circumstances required.  When the writers did express concerns about agency guidance, it was with sub-Treasury guidance being too prolific and de-centralized. Id. at 150-156.  Anyone reading Monograph 22 today will find it very familiar: what was true then is largely still true today. 

Given this history, it is not surprising that the common view was that the APA did not require Treasury to issue most regulations through notice-and-comment process.   That, perforce, meant the rules should be classified as interpretive.  The great administrative law scholar Kenneth Culp Davis reflected this general understanding.  Writing in 1959 for law students, he explained that “the great bulk of Treasury Regulations under the tax laws clearly are interpretative rules, not legislative rules, despite the provision of §7805….  Without the grant of power by §7805, the power of the Secretary or his delegate would be the same…” Administrative Law Text (Foundation Press Hornbook Series) (1959) at 87.   

Again, to reiterate Jack’s point: the APA is notoriously silent on the extent to which a court must follow agency rules when deciding a dispute.  While §706 instructs the reviewing court to “decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action,” that directive is silent on the effect of agency guidance on the court’s task.  The general understanding in 1946—expressed in the influential Attorney General’s Manual as well as elsewhere—was that the APA neither added to nor subtracted from the law of judicial review as it had developed to that point, but instead was only a restatement of existing law.

Into this silence post-APA tax cases continued the pre-APA approach of evaluating tax guidance depending on the level of authority behind its issuance.  Under that view, pre-APA case law had long of distinguished between Treasury and sub-Treasury guidance.  The post-APA case law continued that distinction, with no discussion of how the Treasury regulations were issued.  For example, in 1948 the Supreme Court considered the validity of a regulation that required taxpayers using the installment sale rules to compute deductions consistently with income.  In its opinion, the Court said nothing about whether the recently enacted APA now required changes in how regulations were issued.  Instead, it just went ahead and explained the proper level of deference to be given, proposing this standard: “Treasury regulations must be sustained unless unreasonable and plainly inconsistent with the revenue statutes and…should not be overruled except for weighty reasons.” Commissioner v. South Texas Lumber Co. 333 U.S. 496, 501 (1948).  Hmmm.  Sounds a lot like Chevron!  And Mayo.  And just as in those opinions, the Court’s South Texas Lumber opinion makes no linkage between how the Treasury regulations were issued and the level of deference accorded them.  ‘Cause that was a different issue.

Lower courts as well continued to distinguish Treasury guidance from sub-Treasury guidance, just as they did in pre-APA.  Post APA decisions repeatedly follow the pre-APA approach that sub-Treasury guidance “not promulgated by Secretary” did not have the same weight with courts as Treasury regulations in interpreting tax statutes.  See, e.g. Biddle v. Commissioner, 302 U.S. 573 (1938), 383  (1938); Helvering v. New York Trust Co., 292 U.S. 455, 468 (1934).  But again, these cases do not turn on how the guidance was issued.  They turned on who was issuing it: Treasury or IRS. 

The proper application of the APA to tax administration must therefore start with the recognition that, at the time of its enactment, tax administration was generally considered to be obedient to the APA’s restatement of basic administrative principles.  The APA’s statutory language did not require any changes to how tax guidance was issued or how it was evaluated by courts when adjudicating tax disputes. 

2.  Tax Administration Is Same As It Ever Was

Kristin properly pushes back that the analysis cannot end there.  The next step in the analysis, is to see what post-APA events might have changed this relationship between the APA and tax administration.  Hmmm.  Well….certainly the language of the APA has not changed. The concepts are still broadly defined.  They still do not imagine that only documents titled “regulations” are agency rules subject to the APA, nor do they imagine that every agency document titled “regulation” is a rule that must either be issued in the same way as all other agency rules or that ought to receive the same kind of judicial deference as all other similarly-titled guidance.  These are roomy concepts.  So the inquiry is whether either tax administration or judicial interpretation of the APA, or both, have changed in a way that creates a different relationship between the APA and tax administration. 

We’ll save that inquiry for a different post, next week. 

It’s Time To Get Real: Treasury Regulations Can Certainly Be Interpretive Rules

Professor Bryan Camp, the George Mahon Professor of Law at Texas Tech School of Law, offers his perspective on the recent posts from Jack Townsend and Kristin Hickman that have addressed the controversial and difficult issue concerning the proper classification of tax regulations under the APA. For Jack’s original post, see here. Kristin’s response. Jack’s reply (on his blog), see here. Les

It was interesting to read the recent back-and-forth between Jack Townsend and Kristin Hickman.  Jack takes the position that most Treasury Regulations are properly categorized as interpretive rules under the APA for purposes of evaluating whether they are validly promulgated.  Kristin disagrees, telling Jack “it’s time to let go.” She says that anyone who fails to recognize that ALL Treasury Regulations are legislative rules under the APA is clinging to an outdated concept of tax exceptionalism and has simply failed to see that ship has sailed.

So I guess I’m blind, too.  I also know, moreover, that ships have a habit of returning to port.  The law has never been static and just because the winds may favor one direction now does not mean they will not change.  I think Jack is mostly right in what he says.  I think Kristin’s assertion that ALL Treasury regulations are “legislative” for APA purposes is incorrect, both as a descriptive and normative matter.  Kristin is a prolific and entrepreneurial academic activist with an impressive list of articles to her credit.  More than I will ever write, that’s for sure.  But that does not make her right.  And it certainly does not give her the right to condescend.

Les and the PT gang have kindly agreed to let me offer you my thoughts on the matter in two posts, one today and one tomorrow.  In my view, Kristin’s conclusion is based on three myths: the myth of Mayo, the myth of Tax Exceptionalism, and the Myth of Change. 

read more...

First, I will start where Jack started.  He started his post with the observation that the APA contains guidance on how agencies are supposed to promulgate of agency regulations (called “Rules” in the APA).  But the APA does not similarly give guidance as to what weight an agency rule is supposed to carry when a court seeks to resolve a dispute relating to that rule.  To Jack, the question of deference to an agency rule is different that the question of what procedures an agency must follow to issue the rule.  That’s incontrovertible. 

And Kristin does not contradict Jack.  ‘Cause she can’t.  She instead just ignores the distinction.  ‘Cause she can.  She seems to believe that the distinction between interpretive and legislative rules for APA promulgation purposes was eliminated by the Supreme Court’s opinion in Mayo Foundation for Medical Educ. and Research v. U.S., 562 U.S. 441 (2010).  Her reasoning is that any regulation that carries the “force of law” is legislative and in Mayo, she says in her blog post, “the Supreme Court laid the jurisprudential groundwork for treating all Treasury regulations as legally binding for APA purposes.”

Her reasoning is based on myths.  The first one you can call the Myth of Mayo. 

1. The Myth Of Mayo

In Mayo, the Supremes decided that Treasury Regulations should be given the same weight (i.e. level of deference) as would be given any other agency’s regulations under the Supreme Court’s guidance in Chevron.   Let’s call that “Chevron deference.”  [That’s also a myth, as beautifully demonstrated in Ann Graham, “Searching for Chevron in Muddy Watters: The Roberts Court and Judicial Review of Agency Regulations” 60 Ad. Law Rev. 229 (2008).  But that’s a subject for a different post.]

Mayo says nothing about the distinction between interpretive and legislative rules.  It was a decision about deference.  There was no issue in Mayo over how the regulation at issue had been promulgated.  Again, the issue in Mayo was on the proper level of deference, not on the proper mode of promulgation.  As to deference, the Mayo court sure did say that ALL Treasury regulations get the same treatment as other agency regulations: Chevron deference.  Again, it’s beyond this short post to go into all the reasons for why that is so, but it basically comes down to what is the right way to balance the duties of a court with the duties of administrative agencies.

The myth of Mayo is to say that it stands for the proposition that all Treasury regulations are legislative for promulgation purposes.  The Court said nothing of the sort!  Kristin, however, conflates the two issues that Jack presents as separate: (1) how an agency rule must be promulgated, and (2) what deference should a court give an agency rule. 

Why does she do that? Well, because for the past 15 years or so Kristin has been beating on a drum that Treasury violates the APA when it issues ANY regulation without following the “appropriate” notice and comment procedure.  I use the word “appropriate” in quotes because what is “appropriate” under the APA is a moving target.  Thus, for example, we have the current disagreement on the Conservation Easement proceeds regulation between the 11th Cir. in Hewitt (proceeds regs did not follow proper APA procedure) and the 6th Circuit in Oakbrook (regs were properly promulgated).  I’ll come back to the Oakbrook/Hewitt disagreement in tomorrow’s post.

Kristin started beating on this drum when she looked at a set of 232 Treasury regulation projects published in the Federal Register in 2003, 2004, and 2005.  She concluded that Treasury had not followed what she considered the appropriate APA promulgation procedure in 40% of those projects.  Those were almost all Temp Regs.  In that article she was not entirely sure that this violated the APA because she was not entirely sure what was a “legislative rule” for APA promulgation purposes.  Kristin studied the matter, wrote a couple more law review articles, and adopted the idea that if a regulation has “the force of law” it must be categorized as a legislative regulation under the APA, and therefore must follow the “appropriate” notice and comment procedure.  And Kristin believes all Treasury Regulations have “force of law” because of Mayo.  Once the Supreme Court said that courts should defer to Treasury Regulations on the same basis as other agency regulations, then that made them “force of law” in Kristin’s view.  Again, she writes: “the Supreme Court laid the jurisprudential groundwork for treating all Treasury regulations as legally binding for APA purposes.”

So the first myth that Kristin invokes to support her claim that all Treasury Regulations are properly categorized as legislative regulations is the myth of Mayo.  You only get to her conclusion if you follow her logic.  But her logic ignores the distinction Jack presents, the distinction between the rules governing issuance of agency rules (which is in the APA) and the rules governing what weight courts ought to give agency regulations (which is judge-made).  And her logic creates a second myth, the myth of Tax Exceptionalism. 

2. The Myth of Tax Exceptionalism

Once Kristin concludes that all Treasury regulations have the magic “force of law” she concludes that makes them legislative regulations and so they must be issued, per the APA, thorough the au currant judicial interpretation of APA notice and comment requirements. As a good textualist, Kristin believes all agencies are strictly bound by the APA unless and until Congress specifically exempts them.  So one must find a specific statutory exception to avoid the APA.  She finds no text in the Tax Code that excludes tax regulations from the APA.  Simple!  Case closed! 

And that’s how we get the myth of tax exceptionalism. That myth says that anyone who disagrees with her analysis must be saying that tax administration is not bound by the APA, must be trying to—gasp!—avoid the law!  

No one, ever, has said that the APA does not apply to Treasury regulations.  No one is trying to avoid the law.  It’s an intellectually empty claim.  It’s what in debate we used to call a straw man fallacy.  Remember the difference that Jack explains (and that Kristin simply ignores) between questions of how must an agency issue regulations and how must a court give deference to agency regulations.  In neither branch of the analysis has there ever been a claim that Treasury regulations are somehow exempted from the APA.  Let’s look at each separately.

First, as to on what deference to give regulations, Jack points out correctly that the APA is simply silent on the deference issue.  So as to that there is nothing to be exceptional from!  What courts used to say is that they would give Treasury regulations different authoritative effect than they gave other agency regulations.  Courts did that because the S.Ct. told them to.  See National Muffler.  However, as Kristin rightly points out, the winds of law have changed.  The S.Ct. changed its mind in Mayo.  But, again, that opinion did not turn on whether the regulation there was interpretive or legislative.  The Court was there dealing with the question of what deference to give a 2004 amendment to a 1951 regulation and not whether either had been validly promulgated.  Promulgation was not at issue.  There is nary a word in Mayo about whether the regulations at issue in that case were legislative or interpretive.  It simply did not matter.  And the money quote from Mayo says nothing about promulgation!  It’s all about deference.  Here’s the standard money quote: “we are not inclined to carve out an approach to administrative review good for tax law only. To the contrary, we have expressly recognized the importance of maintaining a uniform approach to judicial review of administrative action.” 562 U.S. at 55. (cleaned up, emphasis supplied).  It’s all about “review” and not about “promulgation.”  Oh, and funny note: Kristin keeps chastising the Treasury Department for claiming “tax exceptionalism.”  But in Mayo it was the government that was asking the Court to apply Chevron. 

Second, as to what is the proper process for promulgation of Treasury Regulations, no one, ever, has said Treasury Regulations are not subject to the APA.  The disagreement is how the APA applies, and how should it apply, to Treasury Regulations.  Kristin’s answer is simplicity itself: she links the deference issue to the promulgation issue.  Read her sentence again: “the Supreme Court [in Mayo] laid the jurisprudential groundwork for treating all Treasury regulations as legally binding for APA purposes.”

Her rationale is that because Mayo blessed ALL Treasury regulations with what Kristin calls “force of law,” then by doing so the Supreme Court, unwittingly, transformed ALL Treasury Regulations into legislative regulations, obliterating without comment over 60 years of understanding.

How can that be?  Well, this takes us to the third myth: the myth of change. 

3. The Myth That Tax Administration Has Changed.

No one, ever, has argued that all Treasury regulations are interpretive for APA purposes.  Jack points out, correctly, that when the APA was first enacted, in the 1940’s, everyone did believe that most Treasury regulations fell into the interpretive category of APA.  I go in to tedious detail on why that is so in “A History Of Tax Regulations Prior to the Administrative Procedure Act,” 63 Duke L.J. 1673 (2014). 

Importantly, Kristin does not (and cannot) deny Jack’s basic point.  Her response is basically “that was then, this is now.”  Her longer explanation is in “Administering the Tax System We Have,” 63 Duke L.J. 1717 (2014).

Her point is well taken, but can easily be pushed too far.  It is certainly possible that even though tax regulations were considered mostly as interpretive regulations under the APA, circumstances now require them to be moved into the legislative regulation box.  Or perhaps we should have a different criteria for typing tax regulations than those who wrote the APA came up with in the 1940’s.

In tomorrow’s post I’ll examine Kristin’s claims on why that should now be different and try to give some refinement to her stark claim that ALL Treasury Regulations are legislative rules under the APA.   I think Kristin’s claim that tax administration has changed has some definite truth to it, but is considerably overstated. 

Again, neither Jack nor I assert that all Treasury regulations are interpretive.  Some are.  Some are legislative.  Kristin, in contrast, asserts that ALL Treasury regulations are legislative (although she leaves open—for now—the question of whether sub-regulatory guidance issued by the IRS are APA legislative rules).  What troubles me about that claim, aside from my inherent distrust of categorical claims, is its reliance on a hugely abstract and amorphous concept called “force of law” to classify claims.  It’s a very top-down approach, which is understandable as Kristin is an academician and is used to taking the 30,000 foot view.  But that approach is not particularly helpful for agency attorneys and those they advise on how to produce timely and effective guidance.  More on that tomorrow.   

It’s Time To Let Go:  Treasury Regulations Are Not Interpretative Rules

Today’s post is from Kristin Hickman, the McKnight Presidential Professor in Law at the University of Minnesota Law School. Kristin is a leading authority in the fields of administrative law, tax law, and statutory interpretation, and her writing has greatly influenced courts and other academics. Les

It’s tough to know what to do when an esteemed colleague cannot be persuaded from and continues to perpetuate an idea that is not only legally untenable but that arguably has negative real-world consequences. I have nothing but respect and warm regard for Jack Townsend. A smart and experienced tax practitioner and teacher, Jack was kind to me and engaged seriously with my ideas about tax and administrative law when many others did not. For many years now, I have enjoyed discussing and debating the intersection of tax administration and administrative law requirements, doctrines, and norms with Jack. And when it comes to ideas, I believe in a big tent approach, acknowledging that law is often uncertain and seeing merit even in arguments with which I strongly disagree. In this same vein, I am happy to let other people have their say, and I don’t feel that I have to reply to every thought that someone publishes but that I think is misguided.

read more...
 

At some point, however, one does have to recognize that ideas can have consequences, and that refusing to let go of and continuing to promote a legal argument may encourage real world actors to behave badly. Such is the case, I think, with Jack’s claim that a (potentially large) subset of Treasury regulations are interpretative rules under the Administrative Procedure Act (APA), and thus that they are exempt from notice-and-comment rulemaking procedures (even if Treasury and the IRS purport to follow those procedures in adopting such regulations).

For the last 15 years, I have explained in articles, essays, and amicus briefs (e.g., here and here) why Treasury regulations simply do not qualify as interpretative rules for APA purposes under any contemporary standard, so I do not want to belabor the point here too much.  To summarize briefly, black-letter administrative law holds, and has held for decades now, that legislative rules are legally binding and interpretative rules are not.  No one seriously questions that all Treasury regulations are legally binding, whether issued under an express grant of rulemaking power in a particular Internal Revenue Code provision or the general rulemaking grant of 26 U.S.C. § 7805(a). A different understanding did exist at the time the APA was adopted, but that different understanding was predicated on (1) now-rejected perceptions that the nondelegation doctrine of Article I, §1 of the Constitution did not allow § 7805(b)-type regulations to be legally binding (for more on that, see, e.g., here, esp. pp. 1104-1113), plus (2) very different approaches to the tax legislative process and in how Treasury exercised its general rulemaking authority in the 1940s relative to today. The casual use of the term “interpretive regulations” by this or that justice or judge in describing legally-binding regulations that interpret statutes does not alter the standards the courts employ in distinguishing legislative rules from interpretative ones and does not signal that the justice or judge is questioning the categorization of the regulations for APA purposes.

Even if you think that some Treasury regulations should be considered interpretative rules—whether on originalist grounds, for pragmatic reasons, or otherwise—at this point, it is clear that the courts will not agree with you. Whatever case law may have existed previously, the Supreme Court laid the jurisprudential groundwork for treating all Treasury regulations as legally binding for APA purposes in its 2011 decision in the Mayo Foundation case. Since then, no federal judge has concluded that any Treasury regulation is an interpretative rule, while several have held the opposite. The Tax Court, hardly staffed by APA maximalists, unanimously concluded in 2015 in the Altera case that Treasury regulations are legislative rules, using reasoning that was not specific to the regulations at bar and would apply equally to every other Treasury regulation. The Tax Court’s Altera decision was reversed by the Ninth Circuit, but only on other grounds. The IRS and the Department of Justice have mostly if not entirely stopped arguing in court that Treasury regulations are exempt from notice-and-comment requirements as interpretative rules, I presume because they know that the argument is not colorable, will not succeed, and will undermine the credibility of their other claims. Currently, the active debate in the courts is over whether Internal Revenue Bulletin guidance like IRS Notices or Revenue Procedures are legislative or interpretative rules, not whether Treasury regulations are.  The fact that Treasury regulations do not qualify as interpretative rules does not mean that no agency pronouncements qualify as interpretative rules; plenty of agency pronouncements, including by the IRS, fall under the interpretative rule category for APA purposes. Just not Treasury regulations.

Regardless of the above, one might wonder, what’s the harm in continuing to assert in academic articles and blog posts that some Treasury regulations should be considered interpretative rules? After all, academic works advance all kinds of interesting legal and normative arguments that have no chance of succeeding in real world courts. Academic writing is great for developing and exploring crazy new ideas, or resurrect or maintain old ones. 

In this instance, however, when highly-regarded tax lawyers like Jack refuse to concede that all Treasury regulations are legislative rules and parse legal texts to gin up uncertainty regarding that characterization, they embolden Treasury and IRS regulation drafters to do the same. Even as IRS and Department of Justice litigators decline to argue in court that Treasury regulations are interpretative rules, the IRS in the Internal Revenue Manual has doubled down post-Mayo on the assertion that most of its regulations are exempt from APA procedural requirements as interpretative rules. In my experience, many (though by no means all) Treasury and IRS regulation drafters still do not take APA procedural requirements as seriously as they should, no doubt inspired at least in part not just by the Internal Revenue Manual but also by highly-regarded tax lawyers giving them cover. Treasury and IRS intransigence on the applicability of APA procedural requirements to Treasury regulations undermines taxpayer confidence in tax system fairness and legitimacy, which in turn encourages taxpayer disrespect and discourages tax compliance. Why shouldn’t taxpayers rely on untenable legal arguments and stretch the boundaries of statutory interpretation to justify their tax return reporting positions when the IRS continues to embrace an untenable legal claim regarding its own legal obligations in the Internal Revenue Manual?

Tax system integrity depends in no small part on smart and experienced tax lawyers, both inside and outside the government, knowing when to stop pushing the legal envelope. It is time for the tax bar to let go of and move forward from the idea that Treasury regulations are interpretative rules.

More On The Confusion Surrounding The Difference Between Legislative And Interpretive Rules

Today’s post is from Jack Townsend. Jack is a practitioner whose blogs on tax procedure inspired us to start Procedurally Taxing back in 2013. In addition to working with me as the principal contributing author for the chapter in Saltzman and Book on criminal tax penalties, Jack writes widely on issues of tax procedure. An area he has written extensively about is the intersection of administrative law and tax procedure. In today’s post, Jack comments on my post the other day that discussed the difference between interpretive and legislative rules. Les

Les’ post Update on CIC Services And More On The Legislative vs Interpretive Rule Difference links to my article The Report of the Death of the Interpretive Regulation Is an Exaggeration, which considers whether interpretive regulations generally and Treasury interpretive regulations specifically are a viable APA category.

PT readers might want to know that I provide a Key Point Summary of the article, where I distill over 100 pages into 7 pages with few citations and no footnotes:   A Key Point Summary of The Report of the Death of the Interpretive Regulation Is an Exaggeration

I think the overall point of confusion is blending APA concepts into the deference discussion.  Interpretive and legislative are APA concepts (importing those concepts from pre-APA law).  They are not relevant to the deference discussion, except that (i) legislative regulations are the law and not subject to deference reasonableness of interpretation testing, and (ii) interpretive regulations are interpretations rather than the law and are subject to deference reasonableness of interpretation testing.  Over time, the Chevron deference commotion has morphed the concepts, somehow claiming that Chevron deference applies only to legislative regulations (or rules that create law rather than interpret law).  That erroneous claim is harmless error in the deference context because courts and scholars then apply the reasonableness of the interpretation test.  In other words, they get the right result but get there at the intermediate step of mislabeling interpretive regulations as legislative.  The erroneous claim, however, can do great mischief when imported into the APA distinction between legislative and interpretive regulations.

I offer one more point that may not be clearly stated in my offerings.  Simply because a notice and comment regulation is interpretive and thus does not REQUIRE notice and comment, once the agency promulgates the interpretation in a notice and comment regulation (as Treasury has historically done for many agency interpretations), then the regulation can be tested for procedural regularity (arbitrary and capricious test) under the APA.  That is not because the interpretation is legislative, but because the APA requires procedural regularity for agency action.  For example, if the IRS said in the Reasoned Decisionmaking explanation in a notice and comment regulation that a tax applies only to individuals with blue eyes when there is no such requirement in the statute, the regulation might flunk the reasonableness of the interpretation deference test AND the procedural regularity (arbitrary and capricious) test.  But the tests are not the same (contrary to the claims of some courts and scholars who misread Judulang‘s footnote). See Judulang v. Holder, 565 U.S. 42, 52 n.7 (2011) (As long time tax observers know, seemingly throw-away statements in Supreme Court footnotes can create great mischief, although the Judulang footnote just requires careful reading to know that claims of equivalence in the two tests are false.)