On Win-Loss Records on Appeal (with War Stories) (5/5/25; 5/6/25)

Note, some of my statistics reported below have been corrected on 5/6/25 11:00am.

I have sprinkled some of my blog entries with war stories from my time with DOJ Tax (1969-1977, with just over 4 years in Appellate and just over 3 years in a refund trial section, called Refund 2, which covered roughly Virginia through Texas. In the Refund 2 trial section, I handled cases in South Carolina, the Northern and Middle Districts of Georgia, Eastern District of North Carolina, and 2 life insurance company cases in Florida and Texas (I forget which districts, but I think Middle District of Florida and Northern District of Texas (because Vester Hughes was opposing counsel) although those cases were not resolved by the time I left DOJ Tax. Today, inspired by the article I quote below, I offer some more war stories through statistics.

The inspiration is a recent article, Stephen K. Cooper, DOJ Tax Chief Touts Winning Court Record On Appeals, 2025 Law360 16-164 (1/16/25) [free link unavailable]. The article covers a talk by Francesca Ugolini, the acting chief of the DOJ Tax Division—there has been no Presidentially appointed and Senate-approved Assistant Attorney General for some time. The article includes the following: 

          The U.S. Department of Justice's Tax Division won an overwhelming majority of appeals in tax cases last year by prioritizing strong legal arguments in disputes that had the potential to significantly affect federal tax administration, the head of the division said Thursday.

          Francesca Ugolini, chief of the DOJ's Tax Division, said in the last fiscal year, the government prevailed in 94% of appeals brought by taxpayers and had an unexpected success rate of 75% for its own appeals.

          "We usually do prevail in over 90% of the taxpayer appeals," Ugolini said at the D.C. Bar Tax Conference, held in Washington, D.C., and online. Regarding the government's appeals, "it's usually over 50%," she said.

          "It's not always as high as 75%, but we have some pretty, pretty good success in the appellate courts," she said.

          Ugolini attributed the higher-than-normal success rate to the division's thorough review process that was used to decide whether to appeal a case the government has lost at the trial level. This includes assessing the strength of the legal arguments, the potential impact on federal tax administration and whether the case presents the best vehicle to address the issue.

          "We don't like to lose on appeal, so we're looking at the strength of the case," she said. "That includes … the strength of our legal argument, what the standard of review is [and] what the precedent is in the circuit.

          "We're more likely to appeal cases that involve legal questions, because those are reviewed de novo on appeal, whereas adverse fact findings are reviewed for clear error, and they're really hard to reverse on appeal."

          The administrative implications of the case are also important to the DOJ in its decision-making process, Ugolini said.

          "We're also looking at what is the broader effect on the federal tax system," she said. "Is this issue important to the IRS administratively? What's the impact on federal tax administration? And then we're also looking at, is this the best case to present the issue?"

I focus on the boldface statements above of an overall Appellate win rate for the fiscal year of 94% for taxpayer appeals and 75% for government appeals. I don’t know the details of those statistics. And, as I say in a footnote in Federal Tax Procedure (Practitioner Edition 2024), p.  882 n. 4044 (as presented the quote is slightly revised for the 2025 Edition):

   n4044 Keep in mind that these are statistics–the worst kind of lies according to the famous saying often attributed to Mark Twain (“There are lies, damn lies and statistics.”). See Wikipedia entry, “Lies, Damned Lies, and Statistics” (last edited 4/24/25 and viewed 5/1/25) (reporting that Twain attributed the saying to Benjamin Disreali, although not found in his extant works). There is, as usual, a more nuanced aphorism: “It is easy to lie with statistics, but easier to lie without them,” attributed to Fred Mosteller, one of the most eminent statisticians of the 20th Century. Wikiquote entry, “Statistics” (last edited 1/13/25 and viewed 5/1/25).

With that caveat about statistics, I offer my statistics in cases I handled while with DOJ Tax Appellate June 1969-March 1974. Readers not interested in my statistics and war stories about them should stop here because the discussion requires nuance and thus lots of words.

The spreadsheet from which I derived the statistics is in Excel format. I will post it to Google Docs in Google spreadsheet format later and post a link here [to come]. I think the Google Docs Spreadsheet will permit the sorting and filtering features carried from Excel. Here are some statistics for cases that I briefed and, in most cases argued:

Caveat 5/5/25: I had some major busts in the numbers originally reported but have corrected thenmon 5/6/25 11:00am]

• Total cases: 54, with 44 Government wins, 9 Government losses, and 1 Inconclusive. 
• Appeals from the Tax Court: 21, with 19 Government wins and 2 Government losses. 
• Taxpayer appeals from the Tax Court: 21, with 19 Government wins and 2 Government losses 
• Government appeals from the Tax Court: 2 with 1 Government win and 1 Government loss
• Appeals from Article III District Courts: 33, 25 Government wins, 7 Government losses, and 1 inconclusive (I have the opening brief, but no reply brief, and no reported decision, which leads me to believe that the Government dropped the appeal because the SG did not authorize it)
• Taxpayer appeals from the district courts: 21, with 21 Government wins
• Government appeals from the district courts: 11, with 4 Government wins and 7 Government losses

Those are the bare numbers, a form of statistics. Those bare numbers indicate that my statistics were far worse for Government appeals than the statistics in the article above for the current fiscal year. However, I do have backstories or side stories about the losses that may help explain what appear to be dismal statistics.

Let me start with my general observation that DOJ Tax Appellate assigned new unproven lawyers the easier cases unlikely to lose. More difficult cases, often Government appeals because one judge or jury  or more than one judge in the case of reviewed Tax Court opinions, have already rejected the Government position. My win rate on Government appeals is not good on the surface.

I start with a chart of the Courts of Appeals (numerical, all Circuits except 1st Cir. and D.C. Circuit; the 5th Circuit had not yet split off the 11th Circuit) and Supreme Briefs in Opposition (reported as Government wins because petitions for certiorari were not granted in any of the cases). The Mo. Ct. App. is the state court Missouri Court of Appeals which I discuss below in Vittert Constr.:

The wins in my Government appeals were:

• Economy Finance Company v. United States, 501 F.2d 466 (7th Cir. 1974)
• Superior Life Insurance Company v. United States, 462 F.2d 945 (4th Cir. 1972)
• Liberty Loan Corp. v. United States, 498 F. 2d 225 (8th Cir. 1974)
• Mid-America Industries, Inc. v. United States, 477 F.2d 1029 (8th Cir. 1973)

The losses in Government appeals were:

• American Realty Trust v. United States, 498 F.2d 1194 (4th Cir. 1974)
• Fletcher v. United States, 436 F.2d 413 (7th Cir. 1971)
• Gant v. United States, 441 F.2d 1130 (5th Cir. 1971)
• Grinstead v. United States, 447 F.2d 937 (7th Cir. 1971)
• Jacob v. United States, 493 F.2d 1294 (3d Cir. 1974)
• United States v. Malnik, 489 F.2d 682 (5th Cir. 1974), cert. den 419 U.S. 826 (1974)
• Vittert Constr. & Inv. Co. v. Wall Covering Contractors, Inc. , 473 S.W.2d 799 (Mo. Ct. App. 1971)

 SOME BACK OR SIDE STORIES

• AMERICAN REALTY TRUST V. UNITED STATES, 498 F.2D 1194 (4TH CIR. 1974), here.

 American Realty Trust was a precursor to Frank Lyon Co. v, United States, 435 U.S. 561 (1977), here, involving the tax characterization of a sale-leaseback (where the putative seller-lessee was Harry Helmsley, the husband of the notorious tax cheat, Leona Helmsley, and ART was the putative buyer-lessor). In Frank Lyon, the documents described a sale-leaseback transaction as transferring ownership to the nominal buyer-lessor; on that basis, the nominal buyer-lessor claimed the depreciation. The IRS disallowed the depreciation on the basis that the nominal buyer-lessor was in substance a secured lender rather than owner entitled to depreciation. The Supreme Court in Frank Lyon held that the nominal buyer-lessor was the owner entitled to depreciation. That was the issue in American Realty Trust which preceded Frank Lyon. At trial, the Government attorney insisted that the ownership issue was a fact issue that the jury (rather than the judge) must try. The jury rendered verdict for the taxpayer. In the Government’s appellate brief, the Government argued that the tax characterization was a legal issue based on proper interpretation of the documents to determine who, as between nominal buyer-lessor and nominal seller -lessee, had the attributes of ownership and thus entitled to depreciation. That is a pure legal issue that should be reviewed de novo on appeal. (The logical consequence of this argument is that the Government's trial attorney should not have insisted on a jury.) However, the appellate panel treated the issue as a fact issue (p. 1197) because the Government attorney at oral argument (my reviewer) “suggested that the proper characterization of the 'substance' of the transaction was essentially a factual matter.” That is not the case we presented in the briefs (Opening and Reply). That meant that the Court of Appeals reviewed the jury verdict under the much stricter reasonable man standard finding support in the record requiring rather minimum support in the record. (On minimum support for jury verdicts, see discussion of Grinstead below.) That "suggestion" at oral argument lost the case for the Government.

One not-so-fun fact from American Realty Trust is that, although the Court treated the issue as factual where the jury verdict can be supported by any evidence in the record, the panel decided the case without the benefit of a trial transcript that could have been informative on the support or lack of support for the jury verdict. In a pretrial conference, the trial judge said he thought that the issue was a legal issue for which there should be no jury. The trial attorney, a feisty attorney who fancied herself invincible before juries, insisted to the trial judge that she should get a jury trial on the ownership issue and even told the judge if he denied a jury trial, she would get him reversed on appeal.(The is called chutzpah.) So, the judge let her have a jury trial. She lost to the jury. Thinking that the Solicitor General would not authorize an appeal because the jury verdict was the end of the matter, she did not order a transcript in the time mandated for an appeal. I recommended appeal on the legal issue, and the Solicitor General approved the appeal. When I found  out the trial attorney had not ordered the transcript, I directed her to order the transcript, which she did belatedly under the Rules. The taxpayer moved to dismiss the appeal for untimely ordering the transcript. We opposed, requesting the court to await the transcript. The court carried action on the motion to the merits decision after oral argument. The merits decision was then made before the transcript was produced—meaning that all fact matters in dispute were resolved against the Government because the Government had not produced a transcript that would permit the appellate panel to assess all of the evidence the jury considered.

Another not-so-fun fact is that, after American Realty Trust was decided, Frank Lyon was decided in the trial court involving the same issue of ownership for depreciation purposes. The Government lost at the trial level in a trial to the judge rather than a jury. That loss required that the case go through the Appellate Section for a recommendation as to whether to appeal to the Eighth Circuit. I found out who was assigned to write the DOJ Tax Appellate recommendation. My only interest in Frank Lyon was that I was involved in American Realty Trust, and was not happy with the appellate decision. I went to that Appellate attorney to discuss with him. He had not yet concluded his consideration or written the recommendation. He told me that he was inclined not to recommend appeal. I “lobbied” him hard to recommend appeal. He later recommended appeal (I can’t say that I was a major influence for the recommendation but maybe I was some influence), the SG authorized the appeal, and the Eighth Circuit held for the Government in a masterful opinion assessing the burdens and benefits of ownership. Frank Lyon Co. v. United States, 536 F. 2d 746 (8th Cir. 1976), here. I highly recommend that those interested read that opinion and for fun and games compare it to the Supreme Court opinion in Frank Lyon. As I describe the Supreme Court's Frank Lyon opinion in Federal Tax Procedure (Practitioner Edition), p. 882 n. 4044:

Frank Lyon is the poster child for the saying that tax cases are too important to turn loose on the Supreme Court. See Bernard Wolfman, The Supreme Court in the Lyon's Den: A Failure of Judicial Process, 66 Cornell L. Rev. 1075, 1098 (1981). It has thus been noted by a thoughtful observer that that “few [tax] shelters are shoddier than those approved by the Court in Lyon and Brown.” Charles I. Kingson, How Tax Thinks, 27 Suffolk U. L. Rev. 1031, 1034-35 (2004). Brown is the earlier decision in Commissioner v. Clay B. Brown, 380 U.S. 563 (1965).

• FLETCHER V. UNITED STATES, 436 F.2D 413 (7TH CIR. 1971), here.

I have no side or backstory on Fletcher. The Government lost on the merits. A solid loss.

• GANT V. UNITED STATES, 441 F.2D 1130 (5TH CIR. 1971), here.

Gant involved the mitigation provisions of the Code. These provisions are notoriously complex unless and until one spends the time to understand their beauty and fairness. I discuss the case in this blog entry: Giants in Tax Law: Roger John Traynor (Federal Tax Procedure Blog 10/26/19; 8/26/23), here. Traynor was one of the authors of the mitigation provisions in the 1930s and also one of the authors of the leading law review articles (2 in a Yale Law Journal Series in 1939) on the mitigation provisions. On the appeal, perhaps because I was not persuasive in briefs or argument, the panel did not get past the complexity so as to understand the beauty and fairness of the mitigation provisions. The Government lost. Hence, since there was a clear conflict (in my mind) with another appellate court opinion, I recommended the SG seek cert. The SG at the time was Erwin Griswold, former tax professor and Dean of the Harvard Law School. The “Dean” as we called him, rejected my recommendation, noting in handwriting on my recommendation that (and this is a paraphrase but pretty close to the actual quote) “We can't take a mitigation case to the Supreme Court, for they will never understand it.”

• GRINSTEAD V. UNITED STATES, 447 F.2D 937 (7TH CIR. 1971), here.

In Grinstead, the Government appealed a jury verdict in a widow bonus case (was the payment by the corporation to the widow of the employee a gift or compensation in recognition of past services). I have previously written on that case. Justice Thomas and Tax -- The Plot Sickens (Federal Tax Procedure Blog 10/29/23; 10/7/24), here. The probability of overturning a jury verdict on a fact issue is virtually nil, unless there is some error in the jury instructions (a legal error). We argued that there was such an error; the court of appeals said there was no such error even though the court of appeals, like the trial judge, said that there would likely have been a different result if the trial were to the judge rather than the jury. Which all goes to show that it is hard to overcome a jury verdict on appeal.

A fun fact from Grinstead. Readers might look at the linked blog above for a fun discussion as to why the trial judge did not give the Government’s requested instructions straight out of the controlling Supreme Court authority—Duberstein, known to every student of the tax law. The reason for the trial judge’s refusal is a classic hallucination—he had it on good authority that a Justice’s clerk had written the opinion and that no Supreme Court Justice (including the nominal author) agreed with the opinion in Duberstein. One of the judges on the appellate panel chuckled at that explanation because the trial judge was a wild card well known to the Seventh Circuit. But it did not help.

• JACOB V. UNITED STATES, 493 F.2D 1294 (3D CIR. 1974), here.

On Jacob, I have no back or side story, except that my reviewer, Bill Friedlander, a great person and lawyer, argued the case. He and I made a deal that he could take oral argument in Jacob, a Government appeal that we lost, and that I could take oral argument in Liberty Loan Corp. v. United States, 498 F. 2d 225 (8th Cir. 1974), a Government appeal that we won. That was a good trade off, particularly since word got back indirectly to the Chief of DOJ Tax Appellate that panel of judges thought I had done an outstanding oral argument where they were inclined at the start of oral argument to reject the Government appeal.

One back or side story on Liberty Loan is that the trial decision that the Government appealed was by Judge William Webster who, by the time the appeal was heard, was on the Eighth Circuit serving alongside the panel on Liberty Loan, a fact that opposing counsel touted in his oral argument; Judge Webster later served as Director of the FBI and wrote the famous report on IRS Criminal Investigation Division. Opposing counsel mentioned more than once that reversal would require the panel to decide the case differently than the fellow  and esteemed Circuit Judge. That ploy did not help.

• UNITED STATES V. MALNIK, 489 F.2D 682 (5TH CIR. 1974), CERT. DEN 419 U.S. 826 (1974), here.

Malnik was a summons enforcement case where, based on his attorney’s representation that Malnik would assert the Fifth Amendment to all questions, the IRS agreed that Malnik (a reputed mobster) did not have to show up and listen to questions. I did not write the DOJ Tax Appellate appeal recommendation, but was assigned to brief and argue the case. It is black letter law that a taxpayer must show up for compulsory process (either subpoena or IRS summons) and invoke the 5th Amendment on a question-by-question basis. The validity of the 5th Amendment claim can only be assessed by considering the question asked. In this case, the IRS agent did not require the taxpayer to show up and therefore asked no questions to put the 5th Amendment claim in context for the appeal. Yet the SG authorized appeal to require the taxpayer to show up, listen to the questions, and properly assert the 5th Amendment claim. I was assigned the task of briefing and arguing the case. I had to argue that the taxpayer’s (through his lawyer) 5th Amendment claim to answer any and all questions was an improper blanket assertion of the 5th Amendment privilege because the IRS could have asked (but did not ask) some unknown questions for which the taxpayer would have no proper 5th Amendment claim. When I showed up at the beginning time for the slate of oral arguments, either Judge Wisdom or Judge Ingraham told me that the case would be reached in just over an hour, but that I might use the time to go to the library to see if I could pull a rabbit out of the hat on the need to ask the questions which the IRS agent waived. Of course, I knew there was no rabbit there. Needless to say, the Government lost that case.

Fun fact from Malnik. Although the taxpayer won the case, the taxpayer petitioned for certiorari. I wrote the DOJ Tax Appellate draft of the brief in opposition, which was finalized by the SG’s assistant and filed. The brief in opposition argued primarily that the taxpayer won below and thus there was nothing for the Supreme Court to review. The reason the taxpayer filed the petition was delay. The taxpayer knew the IRS could issue another summons and this time not waive his appearance. But, they also knew that, at least until the Supreme Court denied cert, the IRS was unlikely to do that. I have no idea whether the IRS ever issued another summons with respect to the tax years involved.

VITTERT CONSTR. & INV. CO. V. WALL COVERING CONTRACTORS, INC. , 473 S.W.2D 799 (MO. CT. APP. 1971), here.

In Vittert Constr. I did not write the appeal recommendation but was assigned to brief and argue the case. In briefing, I found a key discrepancy between the official Missouri court reports and the West Southwest Reporter system. The appeal had been authorized by the SG based on the Southwest Reporter system. However, based on the official reporter, the Missouri court reports, the Government should lose the case, and the appeal should never have been authorized. I tried to get the SG’s assistant on the case to permit the Tax Division to confess error in the case. He refused but told me that I could tell the court at oral argument how to hold against the Government. I did that;  hence, the nominal “loss” was really a Government win because DOJ Tax had no interest in winning cases they clearly should lose. (“Clearly” is somewhat of a fuzz word, but I stand by it for now, because the Government should not have won the case on appeal.)

One fun fact from that oral argument is that, at the conclusion, while I was packing up at the counsel table after opposing counsel had left, one of the judge’s clerks came up to invite me to have lunch with the judges. I was startled by that; I have never had anything like that happen in an appellate or trial level case; but how can one refuse that type of invitation? So I went. At lunch, nothing was said about the case. The judges were interested in my accent and where I was from. The five-judge panel had talked among themselves during oral argument with each trying to identify the region for my accent. The Chief Judge was the only one who got it—he said I was from upstate South Carolina or North Carolina. (I am from Greenwood, SC.) The other judges were way off the mark. Otherwise the meal was pleasant (I paid for my own meal) with no discussion of the case, but had there been any discussion I would likely have just referred to my argument.

So, the foregoing explains my win loss record.

BRIEF COMMENTS ON THE WINS FOR GOVERNMENT APPEALS.

Of the 4 win cases, the consensus among the superiors after briefing was that we would lose all except Mid-America Industries (where oral argument was by Tom Stapleton, the reviewer). I should caveat that two of the cases—Economy Finance and Superior Life Ins. involved qualification for life insurance tax status that was pretty esoteric at the time. We won the cases on appeal. However, the Government subsequently lost on the issue in the Supreme Court, a case I did not handle. United States v. Consumer Life Ins. Co., 430 US 725 (1977).

One fun fact or, rather, an embarrassing fact about Consumer Life which I did not handle because the case came up from the Court of Claims where the DOJ Tax Court of Claims Section handled trials and appeals, I knew about the progress of the case which I followed because of my interest in Economy Finance and Superior Life. I happened to be at a lunch with John Jones, a prominent lawyer in Washington who had formerly served in the Tax Division, and who represented Consumer Finance in the case. I told him that he would lose Consumer Life in the Court of Claims (as it was then called); I assumed he already knew we had won the issue in Economy Finance and Superior Life. Of course, that was my arrogance even if he had lost, but he won Consumer Life.

COMMENTS ON TWO LOSSES IN TAXPAYER APPEALS FROM THE TAX COURT

I lost only two taxpayer appeals from the Tax Court. Both deserve mention 

In Ramos v. Commissioner, 429 F.2d 487 (5th Cir. 1970), here, the Tax Court held that a Texas wife whose husband had left for parts unknown was taxed on one-half her absent husband’s income under Texas community property law, even though she never benefitted from any of his income nor in the tax year even knew where he was. Her appeal was handled by a free local legal society, probably the first tax case they ever handled. She lost in the Tax Court based on strict application of Texas’ community property rules. When I got that taxpayer’s brief and dug into the case, I realized that only about $100 or some other small amount was in issue and it was a real stinker. Accordingly, I offered to my boss whether there was some way I could pay the tax and we could all move on. I’m not sure about how sincere I was in making that offer, but I was trying to convince him that we needed to find some way that we could avoid the appeal that was technically correct but morally (in my view) wrong. He said no. I wrote my brief, it was reviewed, and we filed the brief. Shortly thereafter, the Fifth Circuit summarily without oral argument, rendered a Government “loss.” Within two or three days of the loss, the SG’s tax assistant, Matt Zinn (a wonderful guy and very smart lawyer), came into my office and said that I should not even bother to write a certiorari recommendation in the case, for the decision had already been made not to seek cert. I may have had to write some pro forma cert recommendation against cert (I just don’t remember), because SOP was to write one in every case recommending whether to seek cert or not.

But the fun fact from the Ramos loss was that, after discussing the case with colleagues (one of whom handled the Scudder joint return joint liability case), I prepared draft legislation for innocent spouse relief that the AAG Tax forwarded to the IRS. Thereafter, DOJ Tax struggled mightily with the IRS (including an oral threat that DOJ Tax would forward a recommendation to Congress), with the IRS finally relenting by coming forth with a recommendation to Congress that resulted in enactment of the original innocent spouse provisions. The IRS recommendation was not as broad in relief as my recommendation, but the Acting AAG at the time told me the aphorism that half a loaf is better than no loaf.

In Morris Est. v. Commissioner, 454 F.2d 208 (4th Cir. 1972), here, another case where the technicalities but not the equities were with the Government. The Tax Court in a lop-sided reviewed opinion (I believe 12-4) interpreted the tax law to permit the benefit the taxpayer claimed. At the start of my opening appellate argument, Chief Judge Sobeloff asked what the breakdown in the Tax Court was on this issue of interpretation. When opposing counsel and I said we thought it was 12-4. Judge Sobeloff said something like a rhetorical question (paraphrased): "Well isn't that about it." I then knew we were going to lose. In this regard, although by statute, courts of appeals consider appeals of legal issues (as in Morris Est.) de novo, as a practical matter there is some tendency to defer to Tax Court interpretations of law. There is a long history around the scope of review of Tax Court legal interpretation going back to the 1940s; I won't go into that now, but many have observed some tendency of appellate panels to defer somewhat to Tax Court legal interpretations, a tendency that would certainly be exacerbated with lop-sided reviewed opinions.

CONCLUSION

 Perhaps the quote from the innocent spouse legislation is also the metaphor for my won-loss record—half a loaf is better than no loaf.