The Court having been polled at the request of one of its members, and a majority of the judges who are in regular *303 service and not disqualified not having voted in favor ( Fed. R. Ap. P. 35 and 5th Cir. R. 35 ), rehearing en banc is DENIED. In the en banc poll, three judges voted in favor of rehearing (Judges Jones, Smith, and Ho), and twelve judges voted against rehearing (Chief Judge Stewart and Judges Dennis, Clement, Prado, Owen, Elrod, Southwick, Haynes, Graves, Higginson, Costa, and Willett).

Judge Smith, joined by Judges Jones and Ho, dissents from the Court's denial of rehearing en banc, and his dissent is attached.

JERRY E. SMITH, Circuit Judge, joined by EDITH H. JONES and JAMES C. HO, Circuit Judges, dissenting from the denial of rehearing en banc:

The panel opinion mutilates the test for plain-error relief. Every one of the panel's multiple mistakes favors Marroquin. And even under the relaxed standard that the panel accidentally announces, Marroquin falls far short of satisfying the test. Because the court should have vacated this aberrant opinion for en banc rehearing, I respectfully dissent.

To obtain appellate relief from forfeited error, Marroquin must meet the difficult four-prong test. He must show (1) an error (2) that is plain and (3) affects substantial rights. "Fourth and finally, if the above three prongs are satisfied, the court of appeals has the discretion to remedy the error-discretion which ought to be exercised only if the error 'seriously affect[s] the fairness, integrity or public reputation of judicial proceedings.' " Puckett v. United States , 556 U.S. 129 , 135, 129 S.Ct. 1423 , 173 L.Ed.2d 266 (2009) (quoting United States v. Olano , 507 U.S. 725 , 736, 113 S.Ct. 1770 , 123 L.Ed.2d 508 (1993) ) (brackets in Puckett , some internal quotation marks omitted).

I.

Although, to its credit, the panel carefully avoids misquoting Puckett , it misstates the test by changing the words in three different places, in a way that is hugely misleading. I address each of those in turn.

A.

First, the panel omits the crucial word "seriously" from the paragraph in which it applies the Puckett test. The panel says, "That leaves the requirement that Marroquin show the error affected the fairness, integrity, or reputation of the proceeding." United States v. Marroquin , 874 F.3d 851 , 855 (5th Cir. 2017). 1 There are no quotation marks, and there is no citation to Puckett or any other authority; the panel just sets forth an easier test, likely from inadvertence. Without even trying to explain why it believes the sentence affects-much less "seriously affects"-fairness, integrity, or reputation, the panel only gives the conclusional justification that "[w]e choose to correct this error in light of its effect on the sentence combined with the nature of the error." Id.

The Supreme Court would not have used the word "seriously" unless it was "serious" about sending the message that fourth-prong relief is available only in exceptional circumstances. See Edward Goolsby, Comment, Why So Serious? Taking the Word "Seriously" More Seriously in Plain Error Review of Federal Sentencing Appeals , 51 HOUSTON L. REV. 1449 (2014). The panel thus not only misstated *304 but misapplied the fourth prong, and that is error.

B.

This is not the panel's only misstatement of the fourth prong. In describing that prong in the introductory part of the opinion, 874 F.3d at 853 , the panel explains that if the appellant can satisfy the first three prongs, "then we have the discretion to remedy the error if it 'seriously affect[s] the fairness, integrity or public reputation' of the proceeding " (emphasis added) (quoting Puckett , 556 U.S. at 135 , 129 S.Ct. 1423 , but only inside the single quotation marks). The panel uses similar, inaccurate language in the analytical part of the opinion, where it states and applies the fourth prong: "That leaves the requirement that Marroquin show the error affected the fairness, integrity, or reputation of the proceeding ." Id. at 855 (emphasis added).

That also is error. Outside the quotation marks are the words "of the proceeding." That is not what the Supreme Court said. Instead, Puckett refers to "the fairness, integrity or public reputation of judicial proceedings ." Puckett , 556 U.S. at 135 , 129 S.Ct. 1423 (emphasis added). The difference is huge. It is one thing for an observer to think that an erroneous sentence makes that particular proceeding seem "unfair." It is an exponentially higher burden for the appellant to demonstrate that the sentence is so wrong that it seriously casts a pall on judicial proceedings generally.

It is no guess that the Supreme Court meant to give meaning to the phrase "of judicial proceedings" as distinguished from "of the proceeding." At the end of its opinion, the Court helpfully explained what it meant by "judicial proceedings." Puckett dealt with the failure to live up to a plea agreement. Applying the fourth prong, the Court opined that "[i]t is true enough that when the Government reneges on a plea deal, the integrity of the system may be called into question." Puckett , 556 U.S. at 142-43 , 129 S.Ct. 1423 (emphasis added). And further, referring to the sentence at hand, the Court concluded that "receipt of a sentencing reduction ... would have been so ludicrous as itself to compromise the public reputation of judicial proceedings ." Id. at 143 , 129 S.Ct. 1423 (emphasis added).

From this, there can be no doubt that the fourth prong requires the appellant to show "serious" damage not only to the fairness of the result in his own case but-much more broadly-to the fairness, integrity, or public reputation of judicial proceedings generally. That is a more than a gentle reminder of how steep the fourth prong really is and of how, by subtle changes to its articulation, the panel has "seriously" (pun intended) reduced the burden for this appellant. 2

*305 C.

So far, I have shown that the panel misstated the Puckett test in two respects: by omitting the crucial word "seriously" and by altering the phrase "of judicial proceedings." There is yet a third alteration: The panel changed "public reputation" to "reputation." The panel begins its penultimate paragraph by stating, "That leaves the requirement that Marroquin show the error affected the fairness, integrity, or reputation of the proceeding." Marroquin , 874 F.3d at 855 (emphasis added).

It is significant that the Supreme Court referred to "public reputation" and that the panel omitted "public." 3 That underscores the fact that, as explained above, the Court was referring to judicial proceedings generally and not to the specific proceeding at hand. It is difficult to see how a forfeited-error sentence in one particular case, such as this one, would sully the public reputation of the system except in the most notorious case. The panel's phrase "reputation of the proceeding" hardly makes sense here.

II.

The foregoing discussion shows that, for whatever reason, the panel stumbles in its articulation of the fourth prong by misrepresenting it in at least three crucial respects. That needs to be examined by the en banc court so that we carefully adhere to what the Supreme Court requires. But even accepting the majority's more-than-questionable description of the four prongs, the panel errs in finding all those prongs satisfied so as to afford plain-error relief to this appellant.

A.

Marroquin fails the second prong, because the error is not "plain." "[T]he legal error must be clear or obvious, rather than subject to reasonable dispute." Puckett , 556 U.S. at 135 , 129 S.Ct. 1423 (citing Olano , 507 U.S. at 734 , 113 S.Ct. 1770 ). "Under plain error review, errors that are apparent only after 'traversing a somewhat tortuous path,' through a 'careful parsing of all the relevant authorities, including the sentencing guidelines and applicable decisions[,]' are not 'clear or obvious.' " United States v. Singleton , 707 Fed.Appx. 298 , 299 (5th Cir. 2017) (per curiam) (quoting United States v. Rodriguez-Parra , 581 F.3d 227 , 231 (5th Cir. 2009) ). 4

The purported error here is anything but obvious. The question is the interpretation *306 of a North Carolina statute. The Marroquin panel admits that the Fifth Circuit addressed that very statute in United States v. Rodriguez-Prieto , 491 Fed.Appx. 514 , 515 (5th Cir. 2012) (per curiam). The panel concedes that Rodriguez-Prieto contains "some language ... that suggests it would not be error to separately assess criminal history points for each of the consolidated [North Carolina] offenses." Marroquin , 874 F.3d at 854 .

The Marroquin panel yields instead to a Fourth Circuit decision, United States v. Davis , 720 F.3d 215 , 219 (4th Cir. 2013), which decides the North Carolina statutory question favorably to Marroquin. It is fine, with benefit of hindsight, to decide that, as between Davis and Rodriguez-Prieto , the former has the better view of the law, but it is not fine for this panel to do so in the context of plain error. In light of the "reasonable dispute," including a Fifth Circuit decision going the other way, the error is not "clear or obvious," and plain-error review here should end at the second prong. 5

B.

The third prong requires the appellant to show that a plain error affected his substantial rights. Instead of requiring Marroquin to shoulder the burden of establishing that his substantial rights were affected, the panel flips the burden. The panel closes its discussion of the third prong with the statement that "the government has not shown this to be a case in which prejudice did not result from an error that affected the Guidelines range." Marroquin , 874 F.3d at 855 . Laying the burden on the government, and requiring it to prove a negative, is blatantly contrary to proper plain-error methodology and only compounds the mistakes made by this panel in a published opinion that confuses circuit law.

C.

Marroquin also fails at the fourth prong. The panel's explanation is both scant and conclusory: "This error ... based on a misinterpretation of the state criminal laws ... would create doubt about the integrity of the process." Marroquin , 874 F.3d at 855 . The panel gives no reason for that conclusion, because there is none. Moreover, the panel ignores the government's compelling explanation of why no one would reasonably think the sentence was unfair or lacked integrity. That includes Marroquin's numerous imprisonments, with at least four convictions for controlled substances and two illegal reentries. At bottom, the sentence of 25 months is only four months above the top of the allegedly correct guideline range. No one viewing the record as a whole should see Marroquin as deserving of the extraordinary gift of plain-error reversal.

III.

In Puckett , the Supreme Court warned of "a reflexive inclination by appellate courts to reverse because of unpreserved error." Puckett , 556 U.S. at 134 , 129 S.Ct. 1423 (citation omitted). That unfortunate tendency is evident here in the numerous *307 mistakes that all favor the appealing defendant. The panel gravely undermines the incentive for criminal defendants to raise meritorious issues at sentencing instead of finding ways to succeed on forfeited issues on appeal. 6

Our plain-error jurisprudence continues to struggle. See United States v. Suarez , 879 F.3d 626 , 642-46 (5th Cir. 2018) (Smith, J., dissenting). Granting en banc rehearing would vacate the mistaken panel opinion, which would be an expeditious way of correcting the myriad errors identified above. 7

I respectfully dissent from the denial of rehearing en banc.