Unsigned and Electronically Signed Refund Claims

Last year I blogged on the case of Gregory v. United States, 149 Fed. Cl. 719 (2020), in which the Court of Federal Claims denied a refund claim because the taxpayers did not sign the amended return.  It turns out the Gregory case is part of a larger group of cases involving the same or similar issues.  The tax clinic at Harvard filed an amicus brief in the Federal Circuit regarding one of the cases in the group, Brown, on behalf of the Center for Taxpayer Rights.  The Court of Federal Claims has recently addressed another case in the group and one that has a slightly different fact pattern.

In Mills v. United States, No. 1:20-cv-00417 (Fed. Cl. 2021), the CFC issued another opinion in the long line of cases caused by accountant and lawyer John Anthony Castro, who is doing contingency fee refund suits involving IRC 911 exclusions for numerous people overseas. The first case decided by the Court of Federal Claims was last year and Gregory was among the first wave.


In the earlier cases, the problem was that the Forms 1040X were substantively reviewed by the IRS and denied on their merits.  Only when the cases got to the CFC did the DOJ object that Castro signed the Forms 1040X, though his POA did not authorize him to sign for the taxpayers.  In these earlier cases, the taxpayers, relying on a Supreme Court opinion from the 1940s, Angelus Milling Co. v. Commissioner, 325 U.S. 293 (1945), argued that the IRS waived the signature defect by disallowing the claims on the merits rather than based on the lack of signature. 

In all of the cases decided so far, the Court of Federal Claims has held that the signature requirement is jurisdictional and not subject to waiver.  The judges dismissed all such suits without deciding whether a waiver might have occurred if the requirement were not jurisdictional.  An appeal to the Fed. Cir. has been taken only in one such suit, Brown, where the Center for Taxpayer Rights, as amicus, has argued that the whole requirement to file a refund claim is no longer jurisdictional under recent Supreme Court case law, so the signature requirement is also one that can be waived or forfeited by IRS inaction.

The Center cited, among other things, an opinion of the Fed. Cir. from 2020 named Walby, a case Carl Smith blogged here.  In Walby v. United States, 957 F.3d 1295 (Fed. Cir. 2020), a pro se tax protester case, a panel of the Federal Circuit joined a panel of the Seventh Circuit in Gillepsie v. United States, 670 F. Appx. 393 (7th Cir. 2016), in questioning, in dicta, their Circuits’ precedents holding that the administrative tax refund claim filing requirement at section 7422(a) is a jurisdictional requirement to the bringing of a refund suit, as probably no longer good law under recent Supreme Court case law

Mills differs from the prior cases in that, after the first set of Forms 1040X were signed by Castro, the IRS denied them on the grounds that Castro improperly signed them — i.e., not on the merits.  So, there could be no argument in Mills (unlike in the prior cases) that the IRS waived the signature requirement.  Castro responded by getting the CFC to dismiss without prejudice a prior suit Mills had brought and tried again. 

Mills filed new Forms 1040X on which to bring suit.  Mills was in Afghanistan, however, where he could not print out the Forms 1040X and sign them.  So, he affixed his digital signature to the new Forms 1040X.  This was done before the IRS authorized using digital signatures on Forms 1040X.  Waiting more than 6 months after the second amended returns were filed, Mills brought suit for refund.  The opinion just issued (which had initially been issued 2 weeks ago as a sealed opinion, but is no longer sealed) again holds that the claim filing requirement is jurisdictional — citing prior Fed. Cir. case law, but not mentioning Walby.  The court goes on to hold that Forms 1040X at the time these were filed could not be electronically signed.  Thus, the court dismisses this second Mills suit for lack of jurisdiction. All may not be lost for Mills, however, who can now, if he so chooses, go back and perfect his Forms 1040X for the third time, this time complying with IRS procedures for electronic signatures.

It is Carl’s view that Mills would not make a good test case for asking the Fed. Cir. to overrule its prior precedent on whether claim filing is jurisdictional to a refund suit because there is no possible way for Mills to argue for waiver on these facts (and, indeed, no waiver argument was ever made by Mills in this case).  Although the Court of Federal Claims in Mills may have been wrong to call the requirement jurisdictional, in any appeal, the Fed. Cir. would likely duck the jurisdictional question as not necessary to decide the case — just as it did in Walby.  In Brown, deciding whether or not the filing requirement is jurisdictional cannot be avoided, since the taxpayer has an argument for waiver that has not yet been considered by the CFC.

Fellow blogger, Jack Townsend, wrote to Les and I with the following comment:

In the conclusion, the court says:

“This result here is admittedly harsh. Mr. Mills was working overseas under difficult circumstances and made demonstrable efforts to sign the returns as best he could. The IRS could have accepted his digital initials as an appropriate signature but chose not to do so. If it is to process millions of tax-forms each year, the IRS may insist on strict compliance with its procedural requirements. The plaintiff’s second amended returns did not comply with IRS requirements and were therefore not duly filed. Under these facts, the Court lacks jurisdiction over the plaintiff’s refund suit and must dismiss the complaint.” (emphasis added by Jack)

In the bold-face, the court says IRS had authority/discretion to accept the returns but chose not to do so.  Could that decision be reviewed under the APA arbitrary and capricious standard or some review of discretionary decisions?

Les responded “I would think that an APA claim under 706(2)(A) could have been brought as part of the refund suit. Having failed to raise it the taxpayer is out of luck as the APA would not confer independent jurisdiction.”

Unintentionally, a whole group of taxpayers has been keeping lawyers interested in procedural issues occupied.  It’s possible that the Brown case could create some new precedent in the Federal Circuit that could have an impact on other circuits.  The digital signature problem faced by Mr. Mills is in some ways similar to the mailing problem faced by Guralnik and Organic Cannabis.  In both situations, the rules changed shortly after their failure to follow the old rules.  The old rules were changed because they no longer fit the circumstances but the last taxpayers to incorrectly attempt action under the old rules are paying the price.

Is IRS Appeals Using the Taxpayer First Act to Restrict Taxpayer Access?

As the name and its suggests, the Taxpayer First Act (TFA) made numerous changes to the tax code with the general intent of improving taxpayer rights and interaction with the IRS. These changes ranged from the possibly consequential (Sec. 1101’s requirement that the IRS submit a comprehensive customer service strategy to Congress, found here) to the somewhat misguided (Sec. 1203’s “clarification” on Innocent Spouse as noted here), to the outright absurd (Sec. 1406 requiring the IRS to play helpful information on their phone line while placing taxpayers on hold. I’ll take the scratchy “muzak,” thank you very much. Keith anticipated the issue before TFA with his own suggestion here.)

One provision that went largely unheralded, as far as I can tell, was the requirement that most taxpayers be provided their case file prior to meeting with IRS Appeals (codified at IRC § 7803(e)(7)). Few taxpayers (or practitioners) would oppose greater access to case files, so this seems to be a straightforward win for taxpayer rights -most pertinently, the taxpayer “Right to be informed.” IRC § 7803(a)(3)(A).

And yet in my experience IRC § 7803(e)(7) may actually result in less access to information, rather than more. How is this possible? The devil is in the details…

read more…

To be fair, as I will argue, the devil isn’t really in the “details” of the statute. Rather, the devil is in the implementation of the statute. But that lacks alliteration and “the impishness is in the implementation” is not a commonly used phrase. Still, it is more of an “impish” problem than a devilish one. That is to say, the problem is more of an impish gremlin than a glaring devil. It causes problems precisely by going unnoticed. Allow me to explain.

Typically, when IRS Appeals asks me for numerous documents to support my client’s case, I ask Appeals for numerous documents in return. Even prior to the TFA I would ask for the administrative case file. In the past some Appeals Officers (AOs) would be confused about what I was asking for, and some would be on top of things. Post-TFA there is much less initial confusion and clarification in response to these requests, because Appeals has specifically been instructed to provide taxpayer’s case files.


Or not. In my experience, AOs reply quickly to my request for the administrative case file because they interpret it as a “TFA Request” for the case file. Indeed, that is what the subject line has read on the faxed documents I have received from the IRS over the last year.

But who cares what the IRS calls it, so long as it is the case file, right?

Not so fast. A “TFA” case file does not contain all the same information that a non-TFA case file might: importantly, it contains less. Let’s look at the statutory language to see why:

“In any case in which a conference with the Internal Revenue Service Independent Office of Appeals has been scheduled upon request of a specified taxpayer, the Chief of Appeals shall ensure that such taxpayer is provided access to the nonprivileged portions of the case file on record regarding the disputed issues (other than documents provided by the taxpayer to the Internal Revenue Service) not later than 10 days before the date of such conference.”

IRC § 7803(e)(7)(A), emphasis provided.

That emphasized parenthetical makes all the difference. One of the main things I want from the IRS is a copy of correspondence from the taxpayer to the IRS. This is because (1) I want to know exactly what it is the taxpayer has so far argued or provided for consistency purposes, and (2) because of how many IRC § 6662 penalty cases I deal with, and specifically because of the holding of Walquist, (see my take here) it is critical to know if a taxpayer responded to an automated exam, period. (In my view, this is because it takes it from the realm of being “fully automated” if the taxpayer responds to the exam, and supervisory approval of penalties arguably should then ensue.)

Further, where you are working with Appeals on a Collection Due Process case having the full administrative file (including, per the Treasury Regulation, the documents sent by the taxpayer to the IRS -see Treas. Reg. § 301.6330-1(f)(2)(A-F4)) is critical because of record review issues (see posts here and here). But when the IRS responds to your request for the case file with a “TFA case file” you aren’t getting the full picture. This is a gremlin rather than a devil because you (the practitioner) might not immediately (or ever) notice it. In my experience IRS Appeals used to provide communications from the taxpayer to the IRS in a request for the administrative file because… well, it is part of the administrative file. An important part.

Practice Tips

This whole issue came to my attention precisely because I don’t ask for the generic “case file” when I’m working with Appeals. Rather, I ask for very specific things in addition to the full administrative file. I also usually ask for clarification on whether something isn’t included in the documents Appeals sends to me because (1) it doesn’t exist, or (2) the AO just doesn’t think they need to provide it. These specific requests make the gremlin of a perfunctory TFA case file pop out pretty quickly.

As a real-life example, I recently asked for a client’s administrative case file, as well as any call log notes from the IRS pertaining to the taxpayer, and any documents submitted by the taxpayer to the IRS. Because the issue of whether my client had contacted the IRS was critical, I specifically asked Appeals for clarity as to whether the IRS did not have records of any communication, or whether Appeals simply was not planning on providing them. Appeals replied with a fax that said “TFA Case File” and provided a copy of the tax return and some automated exam notices. That was it. I responded by saying “Thank you for the TFA Case File I didn’t ask for. When you get a chance, could you send the documents I actually did ask for?”

(Disclosure, especially for impressionable students: I didn’t actually respond that snarkily. That type of attitude doesn’t help your client and runs afoul of the general proposition that you shouldn’t be a jerk to the IRS. Or anyone, really.)

My take is that this is an issue of training. Appeals Officers are essentially told in their IRM that  “taxpayers have a right to x and y documents under the TFA, so when they ask for their case file you have to give them x and y documents.” This gets internalized on-the-fly as “these are the only documents you ever have to give taxpayers, and those are the only documents you need to give on any request for additional information.”

Post-TFA, is Appeals Right to Limit the Case File?

It is not yet clear to me that IRS Appeals is taking the legal position that the TFA actually intended to constrain taxpayer access by supplanting any previous, broader taxpayer right to see the documents or communications they’ve sent to the IRS. My bet is that some individual Appeals officers may take a harder line on what practitioners are entitled to than others. When people are time-crunched and emotions are running high, I worry that the TFA could lead to more Appeals Officers digging in and saying “look, this is all I have to give you so it is all that I am giving you.”

Nonetheless, it is clearly bad policy for Appeals to hold back non-privileged case file information. The mission of Appeals isn’t to play “gotcha!” It is to try to resolve, without litigation, controversies in a fair and impartial manner that will enhance voluntary compliance and public confidence in the Service. It would be a perversion of this mission if Appeals held back information directly relevant to the merits of the case on the grounds that it isn’t covered by the TFA, though specifically requested by the taxpayer.

But beyond cutting against their own policy, it would really just amount to a waste of time since a practitioner can eventually get it through either FOIA, Branerton, or discovery anyway (at least in docketed cases). Appeals should want to resolve cases, not kick the can down the road. I would encourage the IRS to look at the TFA as a floor, not a ceiling, in taxpayer rights and services. Where necessary, Appeals should do more than just provide the documents covered by IRC § 7803(e)(7)(A) if they are serious about their mission and taxpayer rights more broadly. When holding back information they have (and that has been asked for), perhaps Appeals should ask themselves “what do you think of someone that does the bare minimum?”

A Final Plea

I have mixed feelings on the TFA, writ large. I think it was enacted with the right intentions, but I think that it would have seriously benefited from more practitioner input at an earlier stage. Yes, Congress did ask for comments in this instance but they gave a turn-around time of about two weeks (actually less) to get them in. This was covered by Procedurally Taxing at the time (here), and it is hard to imagine that the few comments that did come in were given much weight that late in the game.

Yet we’ve already seen some of the unintended consequences of the legislative language play out with innocent spouse, and I worry that this “right to the case file” may carry similar baggage. Both are issues that I think the practitioner community would have seen from the outset.

So here’s the plea. As previously noted, the TFA required the IRS to come up with a comprehensive customer service strategy. That report came out fairly recently (January 2021). When it has been fully digested, let’s not only have the IRS make some internal changes, but even possibly bring Congress back to the table. Unlike most political issues (and especially “substantive” tax law changes), changes to tax administration garners broad bipartisan support. The TFA saw essentially unanimous support from Congress, signed into law by a president that was impeached (at the time) along party lines. Going back to (arguably) the most important change to tax administration prior to TFA, you have the IRS Restructuring and Reform Act of 1998… which saw essentially unanimous support from Congress, signed into law by a president that was impeached along party lines.

At the end of the day, real respect for taxpayer rights (putting “taxpayers first”) is going to require something of an attitude change with the IRS. To me, part of this means striving to help the taxpayer, rather than looking for ways to make things difficult in the hope of scoring a default “win.” As something of a coda to that point, I’ve recently had another adventure crop up with the TFA “case file” provision. Apparently, some Appeals Officers read the statute as requiring an Appeals conference to be scheduled within 10 days of receiving the TFA file… In other words, receipt of the TFA file starts a ticking clock for the taxpayer to act quickly or miss out on a conference.

This is not at all what the statute says. Granted, Congress used needlessly convoluted language: that the case file must be delivered to the taxpayer “not later than 10 days before” the conference.  But I truly believe that the misreading of the statute (somehow seeing it as constraining the taxpayer) is much easier to do when you come from an adversarial mindset. Let’s hope, as the IRS is pulled more and more into the realm of benefits delivery, that this mindset adapts.

Additional OIC Comments Not Specifically Related to the Mason Case

When Bryan was writing his post, we had an exchange about OICs.  Some of the comments I provided to him I might have provided in posts over the years, but I will state them here in case we have new readers or old readers with memories like mine.  Most of these comments relate to the history of OIC provisions or the IRS administration of the OIC.


The Current OIC Program

Although Congress authorized the IRS to compromise collection cases in the 1860s, the IRS very rarely did so.  In Virginia when I started working in Chief Counsel’s office in 1980, one RO in the state was assigned to work offers.  He would travel the state working the offers and reject every offer after careful consideration.  I think what was happening in Virginia was happening across the US.  Then in 1990, Congress extended the statute of limitations on collection from 6 to 10 years.  I was working in the National Office at the time in the group of attorneys specializing in collection matters.  I watched firsthand the reaction to the change in the statute of limitations, which came as a surprise to the IRS.

The extension from 6 to 10 years was an idea that Congress had to collect more money without having to say they raised taxes.  They could score this as something which would bring in dollars and use that score to reduce taxes elsewhere or spend more money and have it look as though the event was revenue-neutral.  We see the same discussion today as Congress debates whether to give the IRS more money so it can collect billions of additional dollars.  While I think the IRS could use more money, the amount it will collect as a result of receiving more money is tricky to predict.

When Congress changed the statute of limitations on collection in 1990, it did not consult with Treasury or the IRS.  Had they been consulted, Congress would have learned that the IRS collects very little money after the first two years.  The bill got passed at a time when Congress, and therefore the IRS, was very concerned with the Accounts Receivable due to the IRS and was asking lots of questions about how the IRS could reduce the ARDI (I can’t remember all of the acronym.)  The IRS even had an executive whose only job was to reduce accounts receivable. 

The IRS knew that the extended statute of limitations was going to cause the ARDI to balloon because of the uncollectable accounts that were going to stay on the books four years longer.  So, it started casting about looking for fresh ideas to reduce it and one of them was to actually accept offers instead of rejecting them.  It was the wild west for the first few years as the IRS tried to get its policy completely lined up.  In the mid-1990s an excellent attorney, now an IRS executive, Carol Campbell, created the income and expense guidelines as well as the exempt asset guidelines based on IRC 6334.  In 1998, Congress came behind and codified some of the things going on, including requiring the IRS to have income and expense guidelines which it had already created.  So, the current OIC program is less than 30 years old and resulted not from a change in the statute but from a change in administration because Congress gives the IRS almost complete discretion in OICs.

OICs for Low-Income Taxpayers

One of the few restrictions placed upon the IRS as the IRS codified additional OIC provisions in 1998 was the requirement that it not reject OICs simply because the taxpayer did not offer some minimum amount.  This restriction is located in IRC 7122(c)(3).  You can give credit for that code section primarily to Nina Olson and partially to me.  Nina had a Tax Court case with my office back when she directed the Community Tax Law Project in Richmond, Virginia.  The IRS had determined that the taxpayer owed a lot of money.  The case had a messy factual background that was going to require a lengthy and difficult trial. 

If the IRS won, it was unlikely to collect anything from the taxpayer, whose business had ended and who had spent time in prison.  I suggested that, instead of a trial, she concede the liability and we compromise the debt.  Nina liked the idea, but we fought over the compromise because I wanted a minimum amount to make the effort worthwhile.  Subsequent to the case but not long thereafter, Nina was asked to testify before Congress.  In her testimony to Congress leading up to the 1998 changes, she convinced Congress that requiring a minimum amount to compromise the debt of a low-income taxpayer was wrong.  Her testimony resulted in the passage of IRC 7122(c)(3), for which I claim partial credit since I was the person at the IRS who “inspired” her testimony.

OIC Stats

For those who can peek behind the paywall, David Van Den Berg recently wrote an article for Law 360 building on the current National Taxpayer Advocate’s comments regarding OICs in the mid-year report.  In her report, she provides stats on the declining number of OICs over the past decade. According to the NTA, fiscal year 2020 marked the seventh consecutive year of decline in OIC receipts, and total OIC receipts for FY 2020 were the lowest they had been since 2008.  She states that TAS is looking for ways to increase the number of successful OICs.  Unemployment is a big driver of receipts.  Perhaps the unemployment situation caused by the pandemic will cause many more OICs. 

In his article, David mentions that the IRS is considering investing in robotics and exploring digitization of the OIC forms as well as creating an ability to submit the OIC online.  I mentioned to him when he contacted me about the article that I would like to see the IRS articulate its goals for the OIC program as part of deciding whether it has too many or too few OICs.  It started the modern program over 30 years ago as a reaction to a surprise change in the statute.  It started the program to reduce accounts receivable rather than to necessarily benefit collection or benefit taxpayers as a whole.  With a tightly crafted goal for the program, it would be easier to determine if it was meeting the goal for accepting offers rather than just saying a goal exists to increase the number of offers.

Submitting an Offer in Compromise Through Collection Due Process

The case of Mason v. Commissioner, T.C.M. 2021-64 shows at least one benefit of submitting an offer in compromise (OIC) through a request for a collection due process (CDP) hearing.  As part of his lessons from the Tax Court series, Bryan Camp has written an excellent post both on the case and the history of offers.  I will try to provide some insights not explicitly covered in Bryan’s post, but if you have time to read just one, I suggest reading his post.

In my clinic, we try to file offers in CDP if possible because we get the chance to go to Tax Court, where the possibility of a reversal exists, such as occurred here.  Except in CDP cases, taxpayers cannot go to court to contest the denial of an OIC.  Bryan prefers that the door to court remains shut because he thinks the process should be inquisitorial and not adversarial.  The current system works primarily in that way, with only a small percentage of cases having the opportunity to go to court and a still smaller group actually going to court with an even still smaller group getting a favorable outcome by going to court.  His point, which is a valid one, is that the cost of going to court is too high compared to the benefits it brings to the system.  My thought is that it benefits the system to occasionally have someone outside the IRS look at how the system is working and set some parameters for the IRS.


In Mason, the taxpayer had submitted an OIC prior to requesting a CDP hearing.  The IRS returned rather than rejected the OIC request.  The IRS returns OICs that either do not meet its processability standards or which relate to cases in which the taxpayer fails to complete some action requested by the OIC specialist.  The IRS returns OICs at the outset because the taxpayer has unfiled returns, has an ongoing bankruptcy case, or because of a host of other reasons stated in section of the Internal Revenue Manual.  My clinic knows the reasons for the return of offers at the outset and almost always has no problem with those.  In every case, after the OIC unit determines that an offer meets the processability requirements and does not quickly return it, an OIC examiner is assigned who then works the case and eventually, several months later, reaches out to the clinic to tell us what additional information it needs.  In almost 15 years of submitting offers for clinic clients, I may have seen one or two offers get accepted without the request for additional information, but in 99% of the cases, the OIC examiner calls after reviewing the file and wants more.  When the OIC examiner calls, the taxpayer and the representative usually have a short window within which to obtain the additional information.  The OIC examiner will state a date in the conversation by which the material must arrive on their desk and state “if X does not arrive by Y date, your offer will be returned and you will not have appeal rights.”  We try very hard to get the newly requested material to the OIC examiner by the requested date and only fail if the lengthy period of darkness waiting for the review of the OIC to conclude has caused us to lose touch with the client.

In the Mason case, a rare reason for returning the offer occurred.  The Masons owed over $150,000.  Owing that much usually buys you the chance to have a revenue officer (RO) work your case.  As someone who primarily represents low-income taxpayers who do not owe enough to buy the services of an RO, I am envious.  I would much rather work with an RO than an Automated Call Site (ACS).  Of course, some ROs are difficult to work with and some ACS responders are good to work with, but by and large I would prefer to work with an RO because most ROs will be responsive and, if it is possible to frame an RO’s work this way, reasonable.  Having an RO means you have someone who can exercise much more judgment and therefore work toward reasonable solutions that ACS would find difficult.

The Masons, however, experienced the downside of having an RO work their case.  The RO assigned to their case appears diligent, competent, and committed to reaching the appropriate result.  For them, this proved a bad combination.  They had valuable assets with which they did not want to part in order to satisfy their large tax obligation.  Shortly after their conversation with the RO, who let them know that they needed to work to pay the liability, they filed an OIC offering about $5,000.  The RO did something I cannot remember seeing before.  The RO wrote to the OIC unit and said that the taxpayers submitted the OIC for purposes of delay and not in a good faith attempt to satisfy their liability.

Had the OIC unit accepted the OIC for processing, it would have taken several months before the OIC came to an end.  It would have ended with a rejection (assuming the Masons stuck to the very low dollar offer) and would have provided them the opportunity to go to Appeals, which could have taken several more months.  At the end of this lengthy process which almost certainly would have resulted in a rejection of the OIC, the case would have gone back into the active collection inventory but might have been assigned to a less diligent RO, if one was assigned at all.  By going the OIC route, the Masons would have suspended the statute of limitations on collection but would have also had access to their assets for the period that the OIC was under consideration and might have benefited on the back end.

To keep this from happening, the diligent RO wrote to try to convince the OIC unit to return the offer at the outset without processing it.  The OIC unit agreed and returned the offer as submitted for delay based upon IRC 7122(g).  This kept the momentum of the collection of the liability going; however, the next step for the RO with taxpayers who would not voluntarily liquidate their assets and pay the tax was to levy.  Before the RO could levy, a CDP notice needed to occur.  While the Masons did not diligently pursue payment, they did diligently pursue their rights and they quite properly filed a timely CDP request.  In making the request they stated a desire for an OIC.

After the normal delays present in a CDP case, an employee of Appeals worked their CDP request.  The Appeals employee looked at their OIC, which mirrored the OIC returned to them earlier, and determined that the OIC unit properly returned the OIC.  Because the OIC unit properly returned the OIC, the Appeals employee determined that they should not receive an OIC through the CDP process without looking at the merits of the OIC but only at the merits of the earlier decision by the OIC unit to return the OIC.

As mentioned before, only a small number of OICs are subject to judicial review and only a tiny fraction of those involve the issue of returned OICs.  Judge Holmes looks at the prior case law involving OICs and gives a good background on those cases.  He determines that the Appeals employee working a CDP case must consider the merits of the offer and cannot simply determine that the prior decision to return the offer was correct.  So, he remands the offer to allow Appeals a second chance.  Unless Ms. Mason, who now pursues the case alone after Mr. Mason’s death, substantially modifies the offer or unless her circumstances have materially changed over the period of this process, it seems likely that Appeals will make a determination that the IRS should not accept the OIC as a collection alternative to levy.  If it makes that decision and the Tax Court does not overturn it, then the case may go back to the diligent RO, or perhaps that RO has now retired, moved, been promoted, or has too large an inventory. 

Jurisdiction of Bankruptcy Courts to Hear Innocent Spouse Cases

The case of In re Bowman, No. 20-11512 (E.D. La. 2021) denies debtor’s motion for summary judgement that Ms. Bowman deserves innocent spouse relief.  On its own, the court reviews the issue of its jurisdiction to hear an innocent spouse issue as part of her chapter 13 bankruptcy case and decides that it has jurisdiction to make such a decision.  The parties did not raise the jurisdiction issue, which is not surprising from the perspective of the plaintiff but may signal a shift in the government’s position since it had previously opposed the jurisdiction of courts, other than the Tax Court, to hear innocent spouse cases.


The court addresses the issue of its jurisdiction at the outset of the opinion.  It first cites 28 U.S.C. § 1334 and the Order of Reference from the district court before stating that this is a core proceeding.  This part of the opinion addresses the basic issue of bankruptcy courts’ jurisdiction in all issues, stemming from the litigation in the Marathon Oil case from 40 years ago (challenging the basic authority of bankruptcy courts under the then-newly-created bankruptcy code).

Moving past the bankruptcy court’s basic basis for jurisdiction, the court hones in on its ability to hear an innocent spouse case.  It first states:

Although it is true that “Section 6015(f) does not allow a bankruptcy court to exercise initial subject matter jurisdiction over an innocent spouse defense because only the Secretary [of the IRS] receives the equitable power to grant innocent spouse relief under that Section,” here, it is undisputed that the Debtor sought such relief from the Secretary in July 2019 and the Secretary denied the request.  

This aspect of jurisdiction would apply to any court hearing an innocent spouse case.  In essence, the statute requires a taxpayer claiming this relief to exhaust their administrative remedies before seeking to have a court determine relief.

Next, the court turns to its specific ability to hear an innocent spouse case and cites heavily from an earlier case from Texas:

Section 6015(e)(1) states that, in a case where an individual requests equitable relief under Section 6015(f), “[i]n addition to any other remedy by law, the individual may petition the Tax Court to determine the appropriate relief available to the individual under this section . . . .” 26 U.S.C. § 6015(e)(1)(A). It is unambiguous that a Tax Court—and not just the Secretary—may grant relief to an individual. Moreover, the remedy available in the Tax Court is “[i]n addition to any other remedy provided by law.” 26 U.S.C. § 6015(e)(1)(A).

11 U.S.C. § 505 is another “remedy provided by law.” Section 505(a)(1) specifically provides bankruptcy courts with remedial power over tax liabilities and penalties . . . . This statutory language provides a bankruptcy court with the power to determine the legality of taxes and tax penalties.

Pendergraft v. United States Dep’t of the Treasury IRS (In re Pendergraft), 119 A.F.T.R.2d (RIA) 2017-1229 (Bankr. S.D. Tex. Mar. 22, 2017)

Because it determines that the tax liability directly impacts the administration of the bankruptcy case and because the IRS has filed a proof of claim seeking to have Ms. Bowman pay the liability for which she seeks relief, the court finds that it has jurisdiction while also noting that the IRS has not objected to its jurisdiction.

The opinion is important for being only the second court to deal with the issue of whether a bankruptcy court has jurisdiction to decide § 6015 relief.  The court says that it does have such jurisdiction because 6015(e)(1)(A) (giving the Tax Court jurisdiction) is only “in addition to any other remedy provided by law” and that the bankruptcy court is another such remedy.  The court cites the Pendergraft case, which is the only other opinion from a bankruptcy court on this matter.  The court conveniently doesn’t mention all the district court opinions holding that 6015 relief jurisdiction does not exist in collection suits or (in one opinion) in refund suits, but resides only in the Tax Court.

The last district court case to render an opinion on this issue was Hockin v. United States, 400 F. Supp. 3d 1085 (D. Or. 2019).  We blogged on the Hockin case here.  In Hockin, the district court rejected earlier district court opinions and found it had jurisdiction to hear an innocent spouse case.  The case never went to trial because the parties settled.  The issue highlighted differences in the Tax Division of the Department of Justice where the trial sections argued that the district courts lacked jurisdiction while the appellate section simultaneously argued that they had jurisdiction and used that argument as a basis for dismissing innocent spouse cases filed late in the Tax Court as having missed a jurisdictional deadline.  The Tax Clinic at Harvard filed an amicus brief in Hockin pointing out the dissonance in the positions taken within DOJ, and the court noted the conflicting positions.

Perhaps the failure to raise jurisdiction as an issue in Bowman means that DOJ has abandoned the issue that only the Tax Court has jurisdiction to hear innocent spouse cases, or perhaps a split now exists within the trial sections at DOJ.  Another possibility is that DOJ distinguishes between district court and bankruptcy court cases raising this issue.  In its motion to dismiss in the Hockin case, DOJ stated:

The language of Section 6015(e)(3) explicitly strips the Tax Court of jurisdiction once a refund suit is filed in district court, which avoids parallel proceedings. But another court explicitly rejected Boynton. In re Pendergraft, 16-33506, 2017 WL 1091935, at *3 (S.D. Tex. B.R. Mar. 22, 2017). That court held that it could consider an innocent spouse defense as part of a bankruptcy court’s powers to determine the amount or legality of a tax under 11 U.S.C. § 505. The court was unconcerned with the possibility of inconsistent judgments, finding that jurisdiction cannot be “based on a hypothetical possibility that concurrent proceedings could produce inconsistent results. That issue, if it ever exists, should be left to Congress.” Id.

Pendergraft is an outlier decision, and it ignores Boynton’s most convincing point: if Congress intended to provide two equally accessible lanes for a taxpayer to seek review of an innocent spouse determination, why does Section 6015(e)(3) treat the process as a one-way street? The Tax Court is clearly divested of jurisdiction when a refund suit is filed in district court, yet the statute is silent on the reverse scenario. Section 6015 sets out a clear, detailed process for funneling review of innocent spouse determinations to the Tax Court. That statute provides no such scheme for the district courts.

DOJ did not try to distinguish Pendergraft because bankruptcy is different.  In Pendergraft, the DOJ argued that the availability of a Tax Court 6015 action precluded 6015 relief under BC 505. The Pendergraft opinion provides a lengthy response disagreeing with the DOJ and its citations — but one that does, in part, rely on the purpose of BC 505.  Section 505 grants jurisdiction to bankruptcy courts to resolve tax merits issues.  The Pendergraft court says that BC 505 is a remedy encompassed by the “in addition to any other remedy provided by law” clause in 6015(e)(1)(A).

Going past the jurisdictional issue, the court in Bowman declined petitioner’s invitation to grant her relief based on summary judgment.  Here is her motion and here is the DOJ response.  She sought relief under 6015(f) but did not submit an affidavit or much other information related to the factors that the IRS has established as required for relief in Rev. Proc. 2013-34.  The court found that insufficient evidence was presented to allow it to grant relief at this stage.  Of course, she can still succeed if she puts on adequate evidence at trial.  At least, based on the court’s finding of jurisdiction, she will have that opportunity.

IRS Wins Lien Priority Fight with Bank

In Citizens Bank, N.A. v. Nash, No. 2:20-cv-00351 (E.D. Pa. 2021) a lien priority fight occurred between the IRS and the bank holding the taxpayer’s mortgage.  In many ways the bank’s problem reminded me of problems that routinely plague the IRS in lien priority fights.  The bank erroneously recorded a release and that causes it to lose the lien priority fight. 


Mr. and Mrs. Nash borrowed money to buy property in Warrington, PA.  The bank recorded a mortgage to secure the loan in 2006.  After buying the home, the Nash’s ran up a fair amount of federal tax debt causing four notices of federal tax lien to be recorded against them.  On March 19, 2019 the bank executed a satisfaction of mortgage and had it recorded.  Not too long thereafter the bank realized its mistake and brought an action for erroneous satisfaction. 

Because they had filed liens, the IRS and the state of PA were named as defendants.  The IRS removed the case to federal district court which it has the right to do and which it will do in almost every case in which it is named.  In their answer the Nashes conceded the mortgage had not been paid in full and consented to the relief sought by the bank.  The IRS meanwhile moved for judgment on the pleadings. 

The bank asks the court to strike the erroneous recording of the release nunc pro tunc and declare it void ab initio to restore it to its place before the filing of the erroneous release.  The court cited state precedent which had held “a satisfaction “entered by accident or inadvertence . . . may be set aside and the mortgage reinstated, except as the rights of third persons may prevent.”  Because the Nashes admitted the recordation was a mistake, the court set aside the release and reinstated the mortgage.  The court, however, refused to declare that the release was void ab initio. 

The court then addressed the priorities between the lienholders.  The court noted the state law which returned the bank to its former position with the proviso for the rights of third parties.  It then briefly discussed federal lien law citing to the seminal cases of Aquilino v. United States, 363 U.S. 509, 514 (1960) which holds that federal law governs priority after state law establishes property rights and then United States v. New Britain, 347 U.S. 81, 85-86 (1954)) and United States v. McDermott, 507 U.S. 447, 449 (1993) which hold that the lien arising first will take priority.

Here, the federal tax liens were filed between 2012 and 2016.  The IRS argued that although the bank’s 2006 mortgage had priority over the federal tax liens prior to its release, the release of the bank’s mortgage made its lien interest inchoate and only the decision to reinstate the mortgage rendered the mortgage lien choate again.  Since the reinstatement occurred in 2021 after the filing of the notices of federal tax lien, the IRS argued that its lien had priority over the mortgage at this point.  The court agreed.  As a result, the mistake in releasing the mortgage causes the bank to lose priority. 

Depending on the value of the house, the action the IRS takes to enforce its lien and the remaining balance on the mortgage, the bank may or may not lose actual dollars from the loss of its priority status.  The IRS does not foreclose on many homes.  If it does not take action against this home, and assuming the Nashes do not otherwise pay the tax liability, it’s possible that their tax liability could fall off of the books due to the statute of limitations. 

In addition to the bank losing, it’s also possible that the Nashes are losers here if the mortgage is a recourse mortgage.  Should the IRS get paid out of the equity in the house, the bank could obtain a personal judgement against the Nashes.  It is much more likely to do so than the IRS would have been had the IRS remained in the second position.  While it’s easy to think of the bank as the loser here, the Nashes might be the real losers.  You see this type of loss sometimes in bankruptcy cases where the IRS fails to properly file a claim but has a nondischargeable debt.  In those cases it might have been paid out of estate assets but instead the estate assets go to pay creditors who might have been discharged.

The lien issue that causes the bank to lose here regularly causes the IRS to lose.  If the IRS fails to refile its lien as the time for refiling expires, the IRS loses its priority and other creditors move up in priority in the same fashion that the IRS has done here.  The case shows the importance of preserving a lien once it exists.  The court does not discuss how the bank came to make the erroneous release, but I expect that a thorough scrub of its procedures has resulted because of this case.  

Boyle Strikes Again: Incarcerated Individual Subject to Delinquency Penalties Even Though Attorney Embezzled Funds and Failed To File His Tax Returns

We have often discussed the reach of the 1985 Supreme Court case United States v. Boyle. Section 6651(a)(1) and (2) impose delinquency penalties for failing to file a tax return or pay a tax unless the taxpayer can establish that the failure was due to reasonable cause and not willful neglect.  Boyle essentially stands for the position that taxpayers have a nondelegable duty to be aware of tax deadlines. An agent’s incompetence or willful misconduct will not excuse the taxpayer from delinquency penalties.

Lindsay v US is the latest case to apply the principle.


Lindsay was incarcerated and executed a POA to Bertelson, an attorney, to manage his affairs, including filing his tax returns.  The attorney assured Lindsay he was doing so for the years 2012-15; instead he failed to file the returns and for good measure embezzled hundreds of thousands of dollars. The actions resulted in Lindsay receiving $705,414.61 in actual damages and $1 million in punitive damages.

Lindsay eventually filed his tax returns and paid over $425,000 in delinquency penalties. He filed a claim for refund; IRS rejected and he filed a suit in district court. The district court, contrary to the magistrate’s recommendation, granted the government’s motion to dismiss, citing Boyle as precluding a claim for relief. Following a timely appeal, the Fifth Circuit affirmed. In so doing it applied Boyle to Lindsay’s somewhat sympathetic circumstances:

Lindsay claims that he exercised ordinary business care and diligence by giving Bertelson his power of attorney and by directing Bertelson to file his income tax returns and to pay his taxes. Lindsay routinely asked Bertelson whether he was handling Lindsay’s tax obligations, and Bertelson said that he was. In Lindsay’s view, he has a reasonable cause for late filings and delayed payments because he used ordinary business care and prudence but was nevertheless unable to file his returns and pay his income taxes due to circumstances beyond his control, i.e., Bertelson’s malfeasance.

Lindsay’s position was rejected in BoyleBoyle established that taxpayers have a non-delegable duty to promptly file and pay their taxes. 469 U.S. at 249–50. Unlike cases where taxpayers seek and detrimentally rely on tax advice from experts, “one does not have to be a tax expert to know that tax returns have fixed filing dates and that taxes must be paid when they are due.” Id. at 251. Lindsay’s argument fails.

The opinion disposed of a couple of other of Lindsay’s arguments. He also raised the IRS’s own definition of unavoidable absence as excusing delinquency. Citing George v Comm’r, a 2019 TC Memo opinion that collects cases on the point, the Fifth Circuit panel emphasized that the mere fact of incarceration itself does not mean there was reasonable cause to miss deadlines.

Lindsay’s final argument was that  Boyle does not control in cases where a taxpayer is not “physically and mentally capable of knowing, remembering, and complying with a filing deadline.” The opinion stated that even if Boyle created an incapability exception he could have done more, “much like he conducted business and employed a CPA while incarcerated.” 


Lindsay, like many other taxpayers, is out of luck when it comes to trying to recover delinquency penalties that can be directly linked to an agent’s misconduct or incompetence. He did have some recourse, however, as he was awarded significant compensatory and punitive damages from his embezzling attorney. 

The Current State of Taxpayer Service (or Lack Thereof) at the IRS

In my written testimony for a recent hearing before the Ways and Means Subcommittees on Select Revenue Measures and Oversight about the tax gap, I discussed some of the current state of taxpayer service at the Internal Revenue Service (IRS) and the many causes for refunds being delayed in the processing of tax returns.  In a normal filing season, refunds may be stopped because of suspicion of identity theft, or omitted or understated wage income or overstated tax withholding.  They may be stopped for “math error” processing for any number of reasons, including incorrect social security numbers.  As the National Taxpayer Advocate, I regularly focused (here and here) on the high “false positive rates” of these programs.  That is, the IRS froze many more returns that ultimately turned out to be legitimate refund requests than were fraudulent.  For example, for the 2020 filing season, the National Taxpayer Advocate reported that IRS refund fraud filters froze 3.2 million individual income tax returns on suspicion of refund fraud.  Of those 3.2 million refund suspended returns, 66 percent – almost two-thirds — were false positives, meaning they were legitimate refund requests.  (See footnote 19 in NTA report.)  For a quarter of the frozen returns, it took the IRS longer than 56 days to release them to normal processing. 

This year is a far from normal filing season.  The IRS is grappling with reconciling various new provisions including the Rebate Recovery Credit, exclusion of unemployment benefits from income, and the “look-back” provision for the Earned Income Tax Credit.  These provisions have resulted in many more returns being suspended and requiring some form of manual review before processing, posting, and refund issuance can resume.


The National Taxpayer Advocate has reported (Figure 2, Filing Season Report) that for Filing Season 2021, through May 21, 2021, the IRS received 148 million individual income tax returns and processed 135.7 million such returns.  The remaining 12.3 million were stuck in various stages of review.  This number under review is almost twice the number of returns unprocessed at the end of the 2020 filing season (6.3 million were suspended as of July 15, 2020).  Elsewhere,  the National Taxpayer Advocate has reported that as of May 21, 2021, the IRS had over 35 million individual and business tax returns either suspended or “in process” but requiring manual processing before the return can move along and refunds issued (Figure 3, Filing Season Report).

Included in these “frozen” returns are 9.8 million individual tax returns sitting in the Error Resolution System (ERS), up 544 percent from the 2020 filing season, and 2.1 million individual income tax returns still suspended for Identity verification, up 91 percent from the 2020 filing season.  Trying to get through to the IRS to resolve these issues is nearly impossible.  During the 2021 filing season, the IRS received 167 million calls on all its lines, up 294 percent from the year before; only 15.67 million of those calls reached a live assistor.  On the 1040 phone line, which is the main phone number for individual income tax assistance, the IRS received 85 million calls, up 978 percent from the 2020 filing season, with only 3 percent reaching a live assistor.  (Figure 5, Filing Season Report.)

If a taxpayer’s refund return is selected for identity verification (on suspicion of identity theft), the taxpayer is required to verify their identity through an online tool, or by telephone, or at an appointment at the Taxpayer Assistance Center (TAC).  Taxpayers have reported being unable to verify online, unable to get through to the phone verification system, unable to reach the TAC appointment line, and if they are able to get a TAC appointment, it is 9 weeks later.  The phone identity verification line received over 6 million calls this filing season, with a 19 percent level of service, meaning 4 out of 5 calls could not get through to verify their identity and get their refund released.

What can be done to fix this?  First, technology, artificial intelligence, and data science can play an important role.  Many of the IRS’ fraud detection and questionable refund filters are rule-based.  That is, a fixed rule is broken, then a return is selected and must be manually reviewed; often the taxpayer must supply additional information.  From year to year, the IRS does not do a good job of learning from the cases where its filters incorrectly identified a return.  The IRS needs to work with data scientists and artificial intelligence experts to design a fraud/error detection system that is not rule-based but rather learns from the returns that actually turned out to be fraudulent, as well as those that were frozen and ultimately determined to be legitimate.  In short, IRS systems need a continuous feedback loop so it minimizes the false positive rate and improves on its initial selection of returns.  To date, the IRS has refused to set goals for reducing the false positive rate on its fraud detection system.  The data cited above show the urgent need for the IRS to set these goals and act on them.  Congress should require it to do so.

Second, the IRS can use programming to minimize taxpayer errors.  In the 2009 filing season, in which taxpayers were reporting the Economic Stimulus Payments (ESPs) they received in 2008, the IRS had a system by which taxpayers (and their preparers) could look up the amount of ESP they received.  The IRS did not replicate this system for the 2021 filing season.  Thus, according to the National Taxpayer Advocate, 5 million returns were suspended in Submission Processing to reconcile Economic Impact Payments with Rebate Recovery Credits.  For the 2022 filing season, Congress should require the IRS to create a similar look-up system for the Advanced Child Tax Credit; otherwise we will have the same return backlog in 2022 as we have today.

The IRS also could program, as part of the submission processing pipeline, the ability to systemically look back to the 2019 modified adjusted gross income, where a taxpayer claimed the “look back” rule for EITC eligibility.  Instead, millions of returns were suspended for manual review because the IRS did not program this.  While I realize this may not have been possible for the 2021 filing season, given the late enactment of the look back provision, if Congress makes the EITC look back rule permanent, it should require the IRS to systemically check for eligibility.  This approach not only reduces the number of returns that must be manually verified but will also identify returns on which the look back was not but should have been claimed.

On top of all this, the Taxpayer Advocate Service (TAS) has at the top of its homepage on its website a statement as follows: “Refund Delayed? Our ability to help may be limited.”  TAS is supposed to be the safety net for taxpayers experiencing significant hardship; yet TAS is unavailable for most taxpayers experiencing refund delays this year.  This is unacceptable, and a violation of the right to a fair and just tax system.  I do not know why TAS has decided it is unable to help these taxpayers; what I do know is that taxpayers are losing faith in TAS.

Third and most importantly, the IRS taxpayer service functions need a significant and steady funding increase.  Congress should scrutinize the IRS projections for the number of calls it receives.  When refunds are frozen, online services are just not satisfactory; taxpayers want to talk to a live human being.  Congress should require the IRS to tell it how many assistors will be required to answer 85 percent of the calls coming in.  Congress should require the IRS to project how many returns it expects to suspend for identity theft and questionable refund issues, and what level of staffing it needs in ERS and the identity theft/refund fraud units to resolve those issues within 14 days.  Congress should require the National Taxpayer Advocate to project the number of cases they would receive if they were willing to assist taxpayers with refund issues.  With all of this information, Congress can then appropriate the funds necessary to get staffing levels up in all these functions so U.S. taxpayers get the assistance and service they deserve.

These are resolvable problems.  The Commissioner regularly points out in oral testimony that Congress only appropriated funding for a 60 percent level of service in FY 2021.  What the Commissioner omits saying is that the President’s budget for that year only requested funding for that level of service.  (See page 85 of FY 2021 President’s Budget Request for Treasury here.)  Thus, Congress funded 100 percent of what the President/IRS requested.  So, to start transforming taxpayer service, the IRS must be transparent about the abysmal quality of service it is currently providing taxpayers and provide Congress with a budget request that reflects the level of service taxpayers need to comply with the tax laws.  No more of this 60 percent nonsense!  Then, as Michelle Singletary wrote in her recent column, the IRS will be able to pick up the damn phone.