REPUBLICAN NATIONAL COMMITTEE v. WETZEL (2025)

ON PETITION FOR REHEARING EN BANC

Treating the petition for rehearing en banc as a petition for panel rehearing (5th CIR. R. 35 I.O.P.), the petition for panel rehearing is DENIED. The petition for rehearing en banc is DENIED because, at the request of one of its members, the court was polled, and a majority did not vote in favor of rehearing (Fed. R. App. P. 35 and 5th CIR. R. 35).

In the en banc poll, five judges voted in favor of rehearing (Judges Stewart, Graves, Higginson, Douglas, and Ramirez), and ten judges voted against rehearing (Chief Judge Elrod, and Judges Jones, Smith, Richman, Haynes, Willett, Ho, Duncan, Engelhardt, and Oldham).

This is not the first time (and it surely won't be the last) that opposing political interests will cross swords over whether and how election deadlines should be enforced. In Foster v. Love, 522 U.S. 67, 118 S.Ct. 464, 139 L.Ed.2d 369 (1997), our court reached the same outcome that we do today—that is, we set aside a state law because it conflicts with deadlines set by Congress. The Supreme Court unanimously affirmed. So we applied the same principles to the state law challenged here.

My distinguished colleague nevertheless suggests that this case may be unusual (and thus remarkable) because “topflight lawyers ․ unaffiliated with the parties” have seen fit to offer their views on this case pro bono. Post, at – (Higginson, J., dissenting from the denial of rehearing en banc). The implication is that the panel decision may be so off the mark that leading members of the profession have felt compelled to speak out.

But there's another explanation: The pro bono activity in this case may just reflect the institutional bias at many of the nation's largest law firms.

Legal scholars have documented how major law firms consistently favor one side in highly charged disputes like this one. See, e.g., Derek T. Muller, Ideological Leanings in Likely Pro Bono Biglaw Amicus Briefs in the United States Supreme Court, Harv. J.L. & Pub. Pol'y Per Curiam (2024); Eugene Scalia, John Adams, Legal Representation, and the “Cancel Culture”, 44 Harv. J.L. & Pub. Pol'y 333, 335–36 (2021).

This evidence has led to the belief that major firms are falling short of “the great traditions of the profession.” Scalia, 44 Harv. J.L. & Pub. Pol'y at 334. The concern is that they have abandoned neutral principles of representation, and instead engage in ideological or political discrimination in the cases that they're willing to take on. See, e.g., id. at 338 (“representing a person with whom we may disagree is a hallowed, essential tradition of the profession”); Lefebure v. D'Aquilla, 15 F.4th 670, 675 n.1 (5th Cir. 2021) (same). Moreover, professional traditions may not be the only thing that firms are failing to uphold. Two years ago, the Justice Department advised our court that much of the institutional bias at major firms may also include workplace policies that discriminate on the basis of race or sex—yet remain fashionable in many legal circles. See Hamilton v. Dallas County, 79 F.4th 494, 509 (5th Cir. 2023) (Ho, J., concurring) (quoting Civil Rights Division asserting that it is unlawful for firms to “only ․ invite women” to certain career events, or to issue “work assignments ․ on the basis of race,” and agreeing that “a lot of law firms” violate Title VII of the Civil Rights Act).

To be sure, evidence of political and other forms of bias doesn't by itself make a particular legal position right or wrong in any given case. But it does mean that we shouldn't be surprised when “topflight” firms consistently jump in on only one side of politically charged disputes, including this one—whether the law actually supports their position or not. Nor should the firms themselves be surprised when others take notice that they are no longer abiding by the principles of the profession, and react accordingly.

It should go without saying, of course, that licensed attorneys have the same right to opine as any other American citizen—whether their views are principled or partisan. Indeed, motivated reasoning is precisely why clients retain counsel.

But then that's exactly how we should treat their work—as motivated lawyering designed to reach a predetermined result. And whatever firms may choose to do, our duty as judges remains the same—to apply the law in a consistent and principled manner, regardless of whose ox is gored. That includes resolving conflicts between state law and the Constitution, no matter what some lawyers might say about it.

I do not recognize the panel decision described by my esteemed dissenting colleagues. With greatest respect, the panel decision did not hold that the States' “common practice” of “count[ing] timely-cast ballots received after Election Day [ ] was preempted by federal law.” Post, at –––– – –––– (Graves, J., dissenting from denial of rehearing en banc). It is absolutely true that many States under different circumstances sometimes accept ballots received after Election Day. See id. at –––– (tallying 28 States). It is also true that federal statutes like the Uniformed and Overseas Citizens Absentee Voting Act (“UOCAVA”) or the Help America Vote Act (“HAVA”) authorize such receipt for narrow classes of voters described by federal law. The question presented to the panel was whether, in the absence of any federal statute authorizing any deviation from the uniform Election Day requirement, States nonetheless have freedom to accept ballots for as long as they would like. The panel held no. But it explicitly recognized that numerous States—including many if not all of the ones invoked as bugaboos by the dissenting opinion—obviously can accept ballots after Election Day under circumstances authorized by federal law. See Republican Nat'l Comm. v. Wetzel, 120 F.4th 200, 211–13 (5th Cir. 2024) (explaining that UOCAVA, HAVA, and other federal laws authorizing States to accept ballots received after Election Day show that Congress knows how to make such exceptions to the general federal deadline); compare post, at –––– (ignoring what the panel actually held and asserting the panel held “that ballot receipt laws in at least twenty-eight states and the District of Columbia are preempted by federal law[ ]”).

According to the dissenting opinion, States should be free to accept ballots for as long as they'd like after Election Day. That is, of course, a question for Congress. But even if it was a question for federal judges, do our dissenting colleagues really think that federal law imposes no time limits at all on ballot acceptance? True, statutory deadlines prescribe dates by which the certification of presidential electors must occur. Post, at –––– (Graves, J., dissenting from denial of rehearing en banc). At best, those provide a last-ditch backstop—several weeks after Election Day. At worst, they permit States to engage in gamesmanship, experiment with deadlines, and renew the very ills Congress sought to eliminate: fraud, uncertainty, and delay. See Republican Nat'l Comm., 120 F.4th at 204 (citing Cong. Globe, 28th Cong. 2d Sess. 14–15, 29 (1844)). And nothing whatsoever prevents the States from innovating with ever-later ballot receipt deadlines 2 months, or even 2 years, after Election Day in congressional elections. The dissenting opinion's only response is to say States are unlikely to do that—but such pragmatic assurances only underscore that the dissenting opinion lacks any legal limit.

The opinion the dissenters wanted to rehear en banc and the opinion the panel wrote are very different animals. I concur in the court's decision not to pretend they're the same.

I would grant the petition for rehearing. At a minimum, this case presents a question of exceptional importance: whether federal law prohibits states from counting valid ballots that are timely cast and received by election officials within a time period designated by state law. The substantial, if not overwhelming, weight of authority—including dictionary definitions, federal and state caselaw, and legislative history—counsels against the preemptive interpretation that the panel adopted. Moreover, the opinion conflicts with the tradition that forms the bedrock for our nation's governance—federalism—which vests states with substantial discretion to regulate the intricacies of federal elections. Simply stated, federal law does not mandate that ballots be received by state officials before Election Day's conclusion, and the panel's contrary holding is erroneous.

I.

In 2020, during the COVID-19 pandemic, Mississippi lawmakers worked in bipartisan fashion to craft and pass House Bill 1521. The legislation amended the Mississippi Code to allow the counting of absentee ballots that were (1) mailed by Election Day, and (2) received by the applicable county registrar within five business days of Election Day. Miss. Code. § 23-15-637(1)(a).

Upon H.B. 1521's enactment, Mississippi joined a significant number of states that permit the counting of timely-cast ballots received within a designated period after Election Day. During the 2024 presidential election, eighteen states and the District of Columbia counted valid absentee ballots that were postmarked by Election Day and received by election officials within a pre-established, post-Election Day period. Deadlines for a ballot's postmark and timely receipt by election officials are determined on a state-by-state basis. Mississippi, as discussed above, requires an Election Day postmark and receipt within five business days of Election Day, while Texas instructs that absentee ballots be “placed for delivery by mail or common or contract carrier before election day” and arrive “not later than 5 p.m. on the day after election day.” Miss. Code. § 23-15-637(1)(a); Tex. Elec. Code § 86.007(a)(2). Ten additional states allow for post-Election Day receipt of ballots cast by overseas voters. In all, then, twenty-eight states and the District of Columbia allow the counting of at least some absentee ballots that were mailed by Election Day and received by election officials during a post-Election Day period. See Table 11: Receipt and Postmark Deadlines for Absentee/Mail Ballots, Upon H.B. 1521's enactment, Mississippi Nat'l Conf. of State Legislatures (Feb. 22, 2025), https://www.ncsl.org/elections-and-campaigns/table-11-receipt-and-postmark-deadlines-for-absentee-mail-ballots.

Despite more recent developments, absentee voting is not a pandemic-created phenomenon. The method has roots dating back to colonial America: in 1775, the town of Hollis, New Hampshire allowed soldiers who were away fighting in the Continental Army to vote “as if the men were present themselves.” Hon. Samuel T. Worcester, Town of Hollis, N.H., in the War of the Revolution, in 30 New England Historical & Genealogical Register 288, 293 (1876). During the Civil War, states devised a variety of methods to collect and transport battlefield ballots back to home precincts in time for canvassing. See generally Josiah Henry Benton, Voting in the Field: a Forgotten Chapter of the Civil War (1915). And throughout the twentieth century, numerous states extended the benefits of absentee voting to ordinary civilians; many, as noted above, allowed ballots to arrive after Election Day to broaden voting access and ease burdens on mail carriers and election officials.

In October 2024, a panel of this court concluded that this common practice—for states to count timely-cast ballots received after Election Day—was preempted by federal law. Republican Nat'l Comm. v. Wetzel, 120 F.4th 200 (5th Cir. 2024). The opinion contends that a federal statute establishing a uniform day for federal elections, when paired with the Constitution's Elections Clause, requires all ballots to be both “cast by voters and received by state officials” by the end of Election Day. Id. at 204 (emphases in original). For the reasons discussed below, I am wholly unconvinced that federal law supports the panel's conclusion.

II.

Federal law sets “the day for the election” as the first Tuesday that occurs after the first Monday in an even-numbered year. 2 U.S.C. § 7 (House of Representatives); 3 U.S.C. § 21 (presidential and vice-presidential electors); see also 2 U.S.C. § 1 (establishing the same date for the Senate). The preemption analysis thus turns on whether “the day for the election” also mandates that all ballots be in the custody of state officials at the conclusion of the exact same day. Our judicial role is limited: we cannot “rewrite the statute that Congress has enacted,” Dodd v. United States, 545 U.S. 353, 359, 125 S.Ct. 2478, 162 L.Ed.2d 343 (2005), or impose our “individual policy preferences,” Loper Bright Enters. v. Raimondo, 603 U.S. 369, 403, 144 S.Ct. 2244, 219 L.Ed.2d 832 (2024). Instead, we must “interpret the words consistent with their ‘ordinary meaning ․ at the time Congress enacted the statute.’ ” Wis. Cent. Ltd. v. United States, 585 U.S. 274, 277, 138 S.Ct. 2067, 201 L.Ed.2d 490 (2018) (alteration in original) (quoting Perrin v. United States, 444 U.S. 37, 42, 100 S.Ct. 311, 62 L.Ed.2d 199 (1979)).

A.

Dictionaries are a useful starting point, as their “definitions inform the plain meaning of a statute.” United States v. Radley, 632 F.3d 177, 182–83 (5th Cir. 2011) (citing United States v. Ferguson, 369 F.3d 847, 851 (5th Cir. 2004) (per curiam)). A review of dictionaries published around the time of the statute's enactment reveals that an “election” referred to the day that voters made choices for public officers—nothing more. See, e.g., Election, An American Dictionary of the English Language (1st ed. 1828) (“The day of a public choice of the officers.”); Election, Black's Law Dictionary (1st ed. 1891) (“The selection of one man from among several candidates ․”).

While the panel opinion acknowledges these definitions, it declares that they are not suitable because they “make no mention of deadlines or ballot receipt.” 120 F.4th at 206 n.5. Respectfully, this sounds like an answer in search of a question. The date for holding an election is distinct from a deadline for receiving ballots cast in that election. C.f. Bush v. Gore, 531 U.S. 98, 116, 121 S.Ct. 525, 148 L.Ed.2d 388 (2000) (Rehnquist, C.J., concurring) (“After the election has taken place, the canvassing boards receive returns from precincts ․” (emphasis added)). “[T]here is no compelling reason not to read Elections Clause legislation simply to mean what it says.” Arizona v. Inter Tribal Council of Ariz., Inc., 570 U.S. 1, 15, 133 S.Ct. 2247, 186 L.Ed.2d 239 (2013). And if, at the time of the statute's enactment, the ordinary meaning of “election” carried no mandate as to when ballots were to be received, our inquiry should end.

Instead, the panel resorted to its own flawed reading of Foster v. Love, 522 U.S. 67, 118 S.Ct. 464, 139 L.Ed.2d 369 (1997). In Foster, the Supreme Court invalidated a Louisiana statute that allowed for its senatorial elections to be held, and potentially completed, a month before Election Day. Id. The panel culled a trio of factors from the decision that purportedly “guide” the contours of an election: “(1) official action, (2) finality, and (3) consummation.” 120 F.4th at 207.

But the decision to employ Foster as a vehicle for analysis is questionable. The Foster court expressly renounced any intention of “paring the term ‘election’ in [2 U.S.C.] § 7 down to the definitional bone.” 522 U.S. at 72, 118 S.Ct. 464. Yet the panel employs the opinion for that very purpose, declaring that three factors, selectively chosen from the opinion's text, encapsulate an election's defining features. What is more, Foster conceded that its holding did not “turn on any nicety in isolating precisely what acts must cause to be done on federal election day (and not before it) to satisfy the statute.” Id. But again, the panel uses the decision to support that end—cutting it into various pieces to determine whether ballot receipt must occur by the conclusion of Election Day. In my view, it is ill-advised to employ a decision that characterizes its approach as “narrow” and simply “enough to resolve” the dispute at hand, for the very purpose that the opinion disclaims. Id. at 71–72, 118 S.Ct. 464.

B.

Even assuming that the panel's trio of factors is proper, problems remain with its application to ballot receipt deadlines. The opinion defines the first prong, “official action,” as “the combined actions of voters and officials meant to make a final selection of an officeholder.” 120 F.4th at 207 (quoting Foster, 522 U.S. at 71, 118 S.Ct. 464). There are two logical ways to contextualize this factor within voting processes. The first is that the “combined action” consists of election officials providing the means of suffrage—i.e., setting up in-person voting stations, offering a ballot drop box, or providing an absentee ballot—and eligible voters casting ballots. Under this reading, processing actions, including ballot receipt, voter validation, and tabulation, are administrative duties that occur after the election and finalize its result.¹ The second, alternative reading would circumscribe all of the above-mentioned activities within a single election. As the panel opinion acknowledges, this second reading is disfavored, as it would require election officials to complete counting and certify an election by the end of Election Day—a practically impossible task. Id. at 208 (“Of course, it can take additional time to tabulate the election results.”).

Curiously, the panel advances a third reading: that among all of the processing duties that election officials perform after voters have cast ballots, only ballot receipt must occur by the end of Election Day. It bases this splicing on the notion that a ballot cannot be cast unless and until it is received by election officials. Id. at 207. That is error. Consulting a dictionary confirms that a “cast” object is not necessarily received by its intended recipient. Broadly, the term means “to throw,” and in the electoral context, it means “to deposit (a voting paper or ticket).” Cast, Oxford English Dictionary (online ed., 2025). Neither definition expressly conditions the act of casting on ultimate receipt of the item being cast.

The use of a term in common parlance can also aid statutory interpretation. Here, it confirms the dictionary definition: while casting an object connotes an expectation that the object is received by its intended recipient, receipt is neither necessitated nor guaranteed. Consider this example: if a law clerk tells me that he spent the weekend casting fishing lines, I would expect that the lines were “received”—that is, his hooks landed in the water. “But,” the law clerk continues, “it was incredibly windy, and my lines were caught in a nearby tree.” However poor my clerk's casting abilities are, his statements are correct from a vocabulary-based perspective.

A similar principle applies when analyzing the “cast” term in the context of an election. While the word “deposit” suggests that a ballot must be placed in the custody of someone else, that does not compel the conclusion that the panel insists on: that a ballot must be “received by state officials” to be considered cast. 120 F.4th at 204 (emphasis omitted). Instead, lawmakers in each state exercise their legislative prerogative, and establish to whom a ballot must be deposited by the end of Election Day for it to be eligible for counting once timely received by election officials. In Mississippi, for example, voters may place ballots in the custody of the United States Postal Service or other “common carrier, such as United Parcel Service or FedEx Corporation.” Miss. Code § 23-15-637(1)(a). And while an absentee voter would likely expect that their cast ballot is timely received by the appropriate election official, that may not always be the result.

Indeed, the Supreme Court has recently recognized that the casting and receipt of a ballot are distinct actions. In staying a Wisconsin district court's ruling that required election officials to count ballots that were cast after a primary election day, the Court noted that “[e]xtending the date by which ballots may be cast by voters—not just received by the municipal clerks but cast by voters—for an additional six days after the scheduled election day fundamentally alters the nature of the election.” Republican Nat'l Comm. v. Democratic Nat'l Comm., 589 U.S. 423, 424, 140 S.Ct. 1205, 206 L.Ed.2d 452 (2020) (per curiam) (emphasis added). The Court's express differentiation between the casting and receipt of a ballot further highlights the panel's error in insisting that a ballot can only be cast if it is also received—and derivatively, that receipt is necessary for any “official action” requirement to be satisfied.

The panel's second factor, “finality,” fares no better. The opinion relies heavily on two conclusions: (1) a voter's individual selections are different from the polity's aggregate election of a candidate, and (2) an election cannot be considered “final” until all eligible ballots are in the custody of state officials. 120 F.4th at 207. The first conclusion is uncontroversial, particularly considering—as the opinion concedes—that “count[ing] (or recount[ing]) ballots after Election Day” is acceptable. Id. at 208. But the second conclusion—that an election cannot be “final” unless “the proverbial ballot box is closed,” fails to withstand scrutiny. Id. at 207.

To start, the statement is not supported by any caselaw. The only federal case cited in this section merely addresses whether the Constitution authorizes the federal government to regulate political party primaries (the answer is no). Newberry v. United States, 256 U.S. 232, 41 S.Ct. 469, 65 L.Ed. 913 (1921). And even if Newberry was topical, the cited portion—that the “general significance” of an election is the “final choice of an officer by the duly qualified electors”—says nothing about when that “final choice” must occur, or whether the receipt of ballots has anything to do with that “final choice.” Id. at 250, 41 S.Ct. 469.

Ironically, the other cited case, Maddox v. Board of State Canvassers, 116 Mont. 217, 149 P.2d 112 (1944), supports a conclusion different from the panel's: that state statutes govern when each voter has made the “final choice” that Newberry identifies. The panel summarized the Montana Supreme Court's holding as follows: “the Electors Clause preempted a state law that allowed receipt of ballots after Election Day.” 120 F.4th at 208. But that description omits a critical feature: at the time Maddox was decided, Montana's laws required that ballots be “delivered to the election officials and deposited in the ballot box before the closing of the polls on election day.” 149 P.2d at 115 (emphases added) (citations omitted). The following paragraph confirms the significance of this important distinction: “federal and state laws must be read together, and since the state law provides for ballots deposited with the election officials, that act must be completed on the day designated by state and federal laws.” Id. (emphases added).

Mississippi's current laws are materially different: to be counted, an absentee ballot must be deposited, with a confirming postmark, to a mail carrier “on or before the date of the election”; and be received by election officials “no more than five (5) business days after the election.” Miss. Code. § 23-15-637(1)(a). And while the panel opinion identifies a pair of Mississippi election regulations that, within the tabulation process, classify a ballot as “final” only after election officials validate and count that ballot, 120 F.4th at 207–08 (citing 1 Miss. Code R. 17-2.1, 17-2.3(a)), that regulatory label does not in practice ² —and cannot, as a matter of principle—invalidate the statute's directive allowing for post-Election Day ballot receipt. After all, a Mississippi agency “may not make rules and regulations which conflict with, or are contrary to, the provisions of a statute, particularly the statute it is administering or which created it.” Miss. Pub. Serv. Comm'n v. Miss. Power & Light Co., 593 So. 2d 997, 1000 (Miss. 1991) (citation omitted).

The panel further justifies its line-drawing of finality by claiming that “the result is fixed when all of the ballots are received and the proverbial ballot box is closed.” 120 F.4th at 207. But this notion of a “fixed” result is belied by the panel's artificial separation of the receipt of a ballot from other administrative functions. As the opinion acknowledges, Mississippi election officials are required to perform a multi-faceted verification process to ensure ballot integrity. That process undoubtedly strikes some number of received ballots from being counted, and thus undermines the panel's conclusion that receipt ensures a fixed result.³

Lastly, in insisting that federal law requires that ballots be in the custody of an elections official, and not a designated intermediary, by Election Day's conclusion, the panel claims that a sender's purported ability “to recall mail” means that “voters can change their votes after Election Day.” Id. at 208. Even without considering the practical issues ⁴ with this reasoning, the logic is fatally flawed. Any piece of mail that is successfully recalled will receive a new postmark when mailed again, and if that new postmark bears a post-Election Day date, Mississippi election officials must reject it. Miss. Code § 23-15-647 (“All such absentee ballots returned to the registrar after the cutoff time shall be safely kept unopened” and are ultimately “destroyed.”). And if a voter fails to resubmit a recalled ballot, the result remains the same: the election official has neither received nor tabulated any ballot from that voter. In no scenario is a voter abusing delivery services to meaningfully “change their vote after Election Day.” 120 F.4th at 208.

The final prong of the panel's triad, “consummation,” purportedly evaluates the theoretical point at which an election is complete. The opinion cites to two cases, Voting Integrity Project, Inc. v. Bomer, 199 F.3d 773 (5th Cir. 2000) and Foster, 522 U.S. 67, 118 S.Ct. 464, 139 L.Ed.2d 369, for the proposition that “so long as the State continued to receive ballots, the election was ongoing and had not been consummated.” 120 F.4th at 208.

But that framing confuses a derivative act for the predicate requirement. Both Bomer and Foster emphasize that the consummation of an election cannot occur prior to Election Day because doing so would infringe upon the electorate's ability to vote on the day that Congress designated for voting. Bomer, 199 F.3d at 775–76 (affirming the constitutionality of a Texas early voting law because polls remained “open on federal election day and most voters cast their ballots that day”); Foster, 522 U.S. at 74, 118 S.Ct. 464 (noting the Louisiana Attorney General's concession that the state's voting “system certainly allows for the election of a candidate in October, as opposed to actually electing on Federal Election Day”). If the panel was correct that consummation was strictly linked to ballot receipt, a state could comply with federal election laws by only receiving absentee ballots on Election Day, while also barring in-person voting itself. That notion, of course, is soundly rejected by Foster and Bomer.

III.

Though not dispositive, “[h]istorical practice [is] particularly pertinent when it comes to [interpreting] the Elections and Electors Clauses.” Moore v. Harper, 600 U.S. 1, 32, 143 S.Ct. 2065, 216 L.Ed.2d 729 (2023). Given that 2 U.S.C. § 7 was enacted in the 1870s, absentee ballot practices employed during the Civil War are particularly illuminative of Congress's understanding. And the historical record demonstrates that many states accommodated wartime voting by counting timely-cast ballots that were received after Election Day.

The panel opinion casts Civil War absentee voting methods as a binary: either local “election officials brought ballot boxes to the battle-field, where soldiers cast their ballots,” or soldiers prepared ballots in the field and gave “them to a proxy for deposit in the ballot box of the soldier's home precinct.” 120 F.4th at 209–10. Both methods, according to the panel, underscore that a soldier only “voted when the vote was received by election officials”—either by placing the ballot in a box brought to the battlefield by election officials, or when a proxy-transported vote was deposited at a soldier's county of residence. Id. at 210.

But this summation oversimplifies the historical record in two material respects. For one, at least three states allowed Civil War soldiers to vote on Election Day, with administrative oversight performed not by local election officials, but by military personnel.⁵ After voting concluded, ballots were transported back to home precincts for counting, and necessarily arrived after Election Day, given inferior transportation methods that existed in the 1860s. These three states counted these ballots upon receipt, even though they were neither deposited in a box “brought ․ to the battle-field” by election officials, nor “deposit[ed] in the ballot box of the soldier's home precinct” before Election Day. Id. at 209–10.

For example, until at least 1862, Pennsylvania's General Election Law allowed soldiers to “exercise the right of suffrage at such place as may be appointed by the commanding officer of the troop or company.” Benton, supra, at 189. Each soldier's selections were treated “as fully as if [the soldiers] were present at the usual place of election,” and “their votes were counted,” even though those ballots necessarily arrived after Election Day. Id. at 190. In a similar vein, Nevada required that “[o]n election day the vote was to be taken under the direction of the commanding officers,” with words such as “Constitution yes” or “Constitution no” used to signify each soldier's vote. Id. at 171. Once the ballots arrived in the Silver State—undoubtedly after a days-long trek across the Great Plains—they were duplicated for verification purposes, assessed for eligibility, and then counted. Id. at 171–72.

Rhode Island's laws were even more explicit: “on the day of such elections,” a solider was allowed to “deliver a written or printed ballot with the names of the persons voted for ․ to the officer commanding the regiment or company to which he belongs.” Id. at 186–87. Once all soldiers had voted, the commanding officer was tasked with returning each ballot “to the Secretary of State within the time prescribed by law for counting votes in such elections.” Id. at 187 (emphasis added). These three states, in turn, represent at least a third approach intended to facilitate Civil War suffrage: soldiers completed their ballots on Election Day in the field, without the presence of local election officials, and then those ballots were transported by proxy—necessarily arriving after Election Day—to home precincts for counting.

Rhode Island's approach also highlights a second distinction that further undercuts the opinion's skewed recollection of Civil War suffrage. Many states attempted to facilitate voting among soldiers by allowing battlefield ballots to be counted if they arrived before a post-Election Day deadline. A quartet of Confederate states illustrates this principle: Alabama waited until at least November 26 to count soldiers' ballots; North Carolina directed that votes be counted “twenty days after they were cast in the field”; Georgia required that soldiers' votes be counted “within fifteen days after the election, that is, after the day on which they were cast”; and Florida tabulated such ballots on “the twentieth day after the election.” Id. at 317–18. Among Union states, Maryland required its canvassing officer, the Governor, to wait “fifteen days after the election ․ to allow the returns of the soldiers' vote to be made.” Id. at 318. And other northern states generally understood that “a sufficient period would elapse between the day of the election, which was the day on which the soldiers were to vote on the field, and the counting of the votes of the State by the officers who were to count them, to enable the votes to reach them.” Id. The post-Election Day receipt and counting of these Civil War ballots are consistent with absentee ballot procedures currently present in twenty-eight states and the District of Columbia: ballots that are timely cast by eligible voters are counted if received within an established time period after Election Day.

The varying approaches that states adopted to facilitate suffrage during the Civil War separately encapsulates a “clearly established” principle: “the policy of Congress for so great a part of our constitutional life has been, and now is, to leave the conduct of the election of its members to state laws.” United States v. Gradwell, 243 U.S. 476, 485, 37 S.Ct. 407, 61 L.Ed. 857 (1917). As Madison remarked at the Constitutional Convention, the Elections Clause left “the regulation of [federal elections], in the first place, to the state government, as being best acquainted with the situation of the people.” 3 Records of the Federal Convention of 1787, at 312 (M. Farrand ed., 1911). States accordingly hold a “constitutional duty to craft the rules governing federal elections,” Moore, 600 U.S. at 29, 143 S.Ct. 2065, and are tasked with crafting regulations for a variety of ancillary tasks, including the “counting of votes.” Smiley v. Holm, 285 U.S. 355, 366, 52 S.Ct. 397, 76 L.Ed. 795 (1932).

Congress, of course, has the power to preempt a state's election regulations through the Elections Clause. U.S. Const. art. I, § IV. But when it chooses to, “it has done so by positive and clear statutes.” Gradwell, 243 U.S. at 485, 37 S.Ct. 407 (emphases added). We should refrain from “giv[ing] an unduly broad interpretation to ambiguous” language and risk “overextending a federal statute's pre-emptive reach.” Inter Tribal, 570 U.S. at 21, 133 S.Ct. 2247 (Kennedy, J. concurring in part and concurring in judgment). And we should be especially wary of adopting an ambitious—if not strained—reading of a statute that is also inconsistent with historical practice at the time of the law's enactment.

IV.

“Legislative history” can also help “ascertain the intent of the legislative authority” that enacted a statute. Rainbow Gun Club, Inc. v. Denbury Onshore, L.L.C., 760 F.3d 405, 410 (5th Cir. 2014) (quoting In re Hammers, 988 F.2d 32, 34 (5th Cir. 1993)). Fortunately, the Supreme Court has already spoken as to Congress's intent when it crafted the statute that set a uniform day for federal elections. 2 U.S.C. § 7 was enacted to address two concerns: “the distortion of the voting process threatened when the results of an early federal election in one State can influence later voting,” and “the burden on citizens forced to turn out on two different election days to make final selections of federal officers.” Foster, 522 U.S. at 73, 118 S.Ct. 464. Construing the statute to compel ballot receipt by the end of Election Day does not redress (or even implicate) either concern. As one of our sister circuits has noted, “there is no reason to think that simply because Congress established a federal election day it displaced all State regulation of the times for holding federal elections.” Millsaps v. Thompson, 259 F.3d 535, 549 (6th Cir. 2001).

Subsequent legislative developments are also useful in determining whether a statute was intended to achieve a preemptive effect on state laws. In particular, “the case for federal pre-emption is particularly weak where Congress has indicated its awareness of the operation of state law in a field of federal interest, and has nonetheless decided to ‘stand by both concepts and to tolerate whatever tension there [is] between them.’ ” Bonito Boats, Inc. v. Thunder Craft Boats, Inc., 489 U.S. 141, 166–67, 109 S.Ct. 971, 103 L.Ed.2d 118 (1989) (quoting Silkwood v. Kerr–McGee Corp., 464 U.S. 238, 256, 104 S.Ct. 615, 78 L.Ed.2d 443 (1984)). And candidly, a legislative history review illustrates that Congress has not been merely “aware[ ]” or “tolera[nt]” of post-Election Day receipt deadlines. Id. Instead, the body has accommodated each state's distinct ballot receipt laws while crafting federal legislation to improve the ability for overseas citizens to vote.

In 1986, for example, Congress passed the Uniformed and Overseas Citizens Absentee Voting Act (“UOCAVA”), a bill that requires states to provide eligible overseas voters with absentee ballots. If an eligible voter fails to timely receive a ballot from the state they are registered to vote in, UOCAVA directs the state to accept a substitute ballot, known as the Federal Write-In Absentee Ballot. The key feature for this discussion relates to ballot receipt: the statute requires that states accept and count a Federal Write-In Absentee Ballot unless a state-issued ballot is received from the same voter “by the deadline for receipt of the State absentee ballot under State law.” 52 U.S.C. § 20303(b)(3). Instead of setting Election Day as a backstop receipt deadline for overriding ballots, Congress chose to expressly incorporate each state's independent ballot receipt deadline within the statute.

More recently, Congress amended UOCAVA's provisions by passing the Military and Overseas Voter Empowerment Act (“MOVE”). The 2009 law requires federal and state designees to facilitate and streamline the delivery of overseas absentee ballots. Id. §§ 20302(a)(10), 20304(b)(1). Once again, Congress was intentional in selecting the target date by which ballots were to be returned to home states: no later than whatever “date by which an absentee ballot must be received in order to be counted in the election.” Id. § 20304(b)(1).

UOCAVA and MOVE are strong examples of a “long history of congressional tolerance, despite the federal election day statute, of absentee” voting procedures and post-Election Day receipt deadlines. Voting Integrity Project, Inc. v. Keisling, 259 F.3d 1169, 1175 (9th Cir. 2001). Indeed, when presented with multiple opportunities to disable, or at least condemn, this “universal, longstanding practice,” Congress chose instead to incorporate post-Election Day ballot receipt deadlines into its statutes. It is this court's stated policy that we not “read the federal election day statutes in a manner that would prohibit such a universal, longstanding practice of which Congress was obviously well aware.” Bomer, 199 F.3d at 776. Yet the panel opinion reneges on that principle, and unabashedly dismisses a panoply of state laws establishing ballot receipt deadlines.

V.

Two other portions of the panel opinion deserve scrutiny. First, the opinion suggests that any contrary conclusion could allow a state to extend its ballot receipt deadline by “100 days.” 120 F.4th at 215. Limiting principles—both statutory and practical—already prevent such an absurd outcome. In presidential and vice-presential election years, states must certify election results by December 31. 3 U.S.C. § 5(a)(1) (requiring certification “6 days before the time fixed for the meeting of the electors”); 3 U.S.C. § 15(a) (setting January 6 as the meeting date). And states that delay tabulation for House and Senate races risk disenfranchising constituents and losing representation in a new Congress. These reasons are perhaps why, to my knowledge, no state—in the near quarter-millennial history of our nation—has risked the undemocratic gambit that the panel opinion postulates.⁶

Second, the opinion intimates that the only constitutionally-acceptable form of absentee voting is “the practice ․ that arose during the Civil War,” and criticizes attempts to expand the benefits of absentee voting to common citizens as “late-in-time outliers.” 120 F.4th at 211 (quoting N.Y. State Rifle & Pistol Ass'n, Inc. v. Bruen, 597 U.S. 1, 70, 142 S.Ct. 2111, 213 L.Ed.2d 387 (2022)). But the Supreme Court has not required that the Bruen-Rahimi historical analysis be applied to election litigation—and even if such an exacting standard applied, the panel's intimation is “as mistaken as applying the protections [of the Second Amendment] only to muskets and sabers.” United States v. Rahimi, 602 U.S. 680, 692, 144 S.Ct. 1889, 219 L.Ed.2d 351 (2024).

Moreover, our role as judges is not to determine whether these so-called “late-in-time outliers”—referring to broadened access to absentee voting—are efficacious or desirable. 120 F.4th at 211. It is axiomatic that “federal courts are supposed to leave policy-preference choices to Congress—not invoke them to rewrite statutes.” Hoyle v. City of Hernando, No. 23-60451, 2024 WL 4039746, at *5 (5th Cir. Sept. 4, 2024) (Oldham, J., concurring). And “when judges test their individual notions ․ against an American tradition that is deep and broad and continuing, it is not the tradition that is on trial, but the judges.” Schad v. Arizona, 501 U.S. 624, 650, 111 S.Ct. 2491, 115 L.Ed.2d 555 (1991) (Scalia, J., concurring in part and concurring in judgment).

Within the context of election regulations, the “American tradition” that Justice Scalia spoke to is clear: “the policy of Congress for so great a part of our constitutional life has been, and now is, to leave the conduct of the election of its members to state laws.” Id.; Gradwell, 243 U.S. at 485, 37 S.Ct. 407. That tradition delegates to the states the responsibility of “provid[ing] a complete code for federal elections”—including the “counting of votes.” Smiley, 285 U.S. at 366, 52 S.Ct. 397. And this responsibility—which “involves lawmaking in its essential features and most important aspect,” should be left to our state legislators, absent clear and express federal language to the contrary. Id.

* * *

The panel opinion holds that ballot receipt laws in at least twenty-eight states and the District of Columbia are preempted by federal laws that merely designate the day on which an election must occur. The strained statutory interpretation that the opinion employs runs counter to all the traditional tools of statutory interpretation—plain meaning, dictionary definitions, common parlance, historical practice, congressional intent, and congressional history—that we deploy on a daily basis. We should have reheard Wetzel en banc, and I respectfully dissent from this court's failure to do so.

I join in full Judge Graves's comprehensive dissenting opinion from our denial of the petition for full court rehearing. To add further credit for my assessment that this case deserves rehearing, I write to acknowledge Attorney Adam Unikowsky's critique of our court's decision. See Adam Unikowsky, The soul of “election day”, Adam's Legal Newsletter (Nov. 3, 2024), https://adamunikowsky.substack.com/p/the-soul-of-election-day.

We benefit from lawyer insight and criticism. Though we receive amicus curiae briefs less frequently than the Supreme Court, they provide primary opportunity for non-party lawyers to give insight, albeit with stringent requirements. It is rarer that topflight lawyers,¹ like Unikowsky, have time to offer scholarly critique of a case neither he, nor Bernstein, was retained to handle.

Chief Justice Roberts recently reminded that “public engagement with the work of the courts results in a better-informed polity and a more robust democracy.” Chief Justice John G. Roberts, Jr., 2024 Year End Report on the Federal Judiciary 4 (2024), https://www.supremecourt.gov/publicinfo/year-end/2024year-endreport.pdf. It is for this reason that the judiciary depends on lawyers, not just as party advocates, but also for all forms of engagement with courts. “[I]nformed criticism” of court opinions from lawyers unaffiliated with the parties is in that vital tradition. Id. at 5 (quoting Chief Justice William H. Rehnquist, 2004 Year End Report on the Federal Judiciary 5 (2004), https://www.supremecourt.gov/publicinfo/year-end/2004yearendreport.pdf).

I regret that I have not made Attorney Unikowky's critique convincingly to my colleagues. Nevertheless, remembering the Supreme Court's acknowledgements when scholars are instructive,² this case provides opportunity for me to acknowledge lawyers like Unikowsky and Bernstein, whose vigilant criticism of our court's work product should help the growth of the law.

FOOTNOTES

FOOTNOTE.  

1.   The opinion rejects this reading by providing hypotheticals that it concedes are “obviously absurd”: “[w]hat if a State changes its law to allow voters to mark their ballots and place them in a drawer? Or what if a State allowed a voter to mark a ballot and then post a picture on social media?” 120 F.4th at 207. To the extent that the hypotheticals can be construed as serious objections, both examples result in ballots that are cast (submitted in accordance with state law) but are neither received nor tabulated by an election official.

2.   The panel's reliance on the two Mississippi regulations conflates the selections made by a voter (while casting an individual ballot) with the processing and tabulation actions that election officials undertake to finalize an election. And, as the opinion concedes, ballot processing and counting can occur after Election Day. 120 F.4th at 208.

3.   At a minimum, then, the panel's conception of a “fixed” result is no better than an alternative reading: that in jurisdictions with post-Election Day receipt deadlines, the results are “fixed” by Election Day because that deadline establishes a definitive range, with a one-way ratchet, of eligible ballots.

4.   While recalling a mailed ballot is theoretically possible, practical considerations render the occurrence incredibly unlikely. For one, mail recall is not automatically available for every ballot mailed through the postal service. A voter must instead have the foresight and means to have a parcel tracking number attached to their ballot envelope (akin to a shipper adding a tracking service onto a package), and the means to pay an additional fee if their package is successfully recalled. Domestic Mail Manual, § 507.5.1.3 (“Package Intercept is not available for ․ [m]ailpieces that do not contain a tracking barcode.”); § 507.5.2 (“Customers must pay a nonrefundable per-price fee once the USPS successfully intercepts the package.”). And for two, it is nearly impossible to meaningfully alter a vote once it has been marked. What is a voter supposed to do with their prior selections—apply whiteout? Cross out their prior marks? Make superseding marks with a different pen, or perhaps permanent marker? The possibilities may be endless, but all would likely be futile.

5.   While some may suggest that these military officers acted as ex officio election officials, military designees were readily distinguishable from civilian election officials. As the merits-stage amicus brief submitted by the United States explains, most states required civilian election officers “to swear oaths to fairly conduct the election or uphold state oaths.” In-field military officers lacked such deputization; they were simply designated as chaperones through their military rank.

6.   In the 2024 election, for example, the latest state ballot receipt deadlines were ten days (Alaska and Maryland) and fourteen days (Illinois) after Election Day. Table 11: Receipt and Postmark Deadlines for Absentee/Mail Ballots, Nat'l Conf. of State Legislatures (Feb. 22, 2025), https://www.ncsl.org/elections-and-campaigns/table-11-receipt-and-postmark-deadlines-for-absentee-mail-ballots.

1.   Unikowsky, in turn, credits Attorney Richard Bernstein's analysis for the Society for the Rule of Law. Richard Bernstein, The Fifth Circuit Was Wrong–Counting Timely-Cast Remote Votes That Are Received After Election Day is as Old as the Founding, Soc'y for the Rule of L. (Nov. 1, 2024), https://societyfortheruleoflaw.org/fifth-circuit-wrong/.

2.   See, e.g., Feist Pub'ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 347, 111 S.Ct. 1282, 113 L.Ed.2d 358 (1991) (O'Connor, J.); Harrington v. Purdue Pharma L. P., 603 U.S. 204, 229, 144 S.Ct. 2071, 219 L.Ed.2d 721 (2024) (Kavanaugh, J., dissenting).

Per Curiam:* FN* Judges Southwick and Wilson did not participate in the consideration of rehearing en banc.