Before WILLETT, Ho, and DUNCAN, Circuit Judges. STUART KYLE DUNCAN, Circuit Judge: In response to the coronavirus pandemic, Texas Governor Greg Abbott has issued various proclamations about the upcoming November election. Among other things, these measures have expanded the options for Texans to vote in-person early or to vote by absentee ballot. Take, for instance, early in-person voting. Normally, early voting would have started October 19. Now it will start October 13, six days earlier. Or take absentee (“mail-in”) ballots. Normally, if a voter wanted to hand-deliver her mail-in ballot, she would have had only one day to do it—Election Day. Now, under the Governor’s expanded policy, she can deliver the ballot anytime until Election Day. That effectively gives voters forty extra days to hand-deliver a marked mail-in ballot to an early voting clerk. And the voter still has the traditional option she has always had for casting a mail-in ballot: mailing it. To make the situation clear, this chart compares what we will call “pre-COVID” early-voting and absentee-ballot rules and “COVID” rules: Pre-COVIDCOVID Early voting starts October 19.Early voting starts October 13 (six extra days). Absentee ballots may be mailed.Absentee ballots may be mailed. Absentee ballots may be hand- delivered only on Election Day.Absentee ballots may be hand- delivered before and up to Election Day (forty extra days). The controversy we now face involves the rules for hand-delivering absentee ballots. As noted, the Governor’s prior proclamation expanded the timeframe for doing that by forty days. But it happened that a few large Texas counties wanted to set up multiple delivery locations for these ballots. The Governor disagreed with this policy which, in his view, threatened election uniformity and security. Consequently, on October 1, the Governor issued a new proclamation. This proclamation, refining the previous one, specified that mail-in ballots could be delivered only to one designated location per county. But it left in place the previous forty-day expansion for delivering mail-in ballots and the always-available option of the U.S. mail. A coalition of Plaintiffs sued, claiming the October 1 proclamation violated their right to vote by restricting absentee voting options. The district court agreed and enjoined the October 1 proclamation. The Texas Secretary of State appealed and now seeks an emergency stay. Among other things, the Secretary argues that the district court fundamentally misunderstood the context of the October 1 proclamation. She explains that the proclamation is part of the forty-day expansion of Texans’ opportunities to hand-deliver absentee ballots beyond what state election rules normally permit. The proclamation refines that expanded voting period by specifying where ballots are to be delivered. But it leaves the enlarged period in place, and also does nothing to prevent Texans from mailing in their absentee ballots, as they have done in the past in election after election. Properly understood, the Secretary tells us, the October 1 proclamation is part of an expansion of absentee voting in Texas, not a restriction of it. We agree with the Secretary and grant the stay. I. A. Texas law allows eligible voters to vote early, either by mailing in a ballot or by personally appearing at an early voting place. TEX. ELEC. CODE §§ 81.001(a), 82.001–82.005. In response to the coronavirus pandemic, on July 27, 2020, Governor Abbott issued a proclamation (the “July 27 Proclamation”) expanding early voting opportunities for the upcoming November 3 election beyond those provided in the Texas Election Code. Allegedly issued pursuant to authority granted the Governor by the Texas Disaster Act of 1975, see TEX. GOV’T CODE §§ 418.001–418.261, this measure was one of a series of pandemic-driven changes to Texas election effected taken since the Governor’s March 13 coronavirus disaster declaration. See In re Steven Hotze, et al., — S.W.3d —, 2020 WL 5919726, at *1–2 (Tex. Oct. 7, 2020) (explaining “[t]he Governor has repeatedly asserted his authority under the [Texas Disaster] Act to modify election procedures beginning shortly after his March 13 disaster proclamation,” and listing measures including the July 27 Proclamation).[1] Among other things, the July 27 Proclamation authorized early in- person voting to begin on October 13, instead of October 19, thus allowing six extra days to vote in person.[2] The proclamation also expanded opportunities for delivering marked mail ballots in person to an early voting clerk’s office. Previously, voters wishing to hand-deliver their mail ballots could do so “only while the polls are open on election day.” TEX. ELEC. CODE § 86.006(a-1).[3] The July 27 Proclamation, however, suspended this requirement by “allow[ing] a voter to deliver a marked mail ballot to the early voting clerk’s office prior to and including on election day” (emphasis added). This had the effect of extending the amount of time for a voter to hand- deliver a mail ballot: as the Texas Director of Elections explained in this case, “ because counties began sending mail ballots on or before September 19, 2020, the proclamation increased the opportunities for voters to hand-deliver their marked mail ballots from only one day—election day—to over forty days.” Additionally, of course, voters retained the ability to send in their mail ballots the traditional way—through the mail. See id. § 86.006(a)(1), (2). Following the July 27 Proclamation, at least four Texas counties— Harris, Travis, Fort Bend, and Galveston—announced their intention to have multiple mail ballot delivery locations in their counties for the November election. In response to this development, Governor Abbott issued a second proclamation on October 1 (the “October 1 Proclamation”) amending the previous one. This new proclamation first reiterated that early in-person voting would begin on October 13. It then clarified that the prior suspension of section 86.006(a-1), concerning hand-delivery of mail ballots, applied only under certain conditions. First, a voter must deliver the ballot “at a single early voting clerk’s office location that is publicly designated by the early voting clerk for the return of marked mail ballots under Section 86.006(a-1) and this suspension.” Second, the early voting clerk must “allow[] poll watchers the opportunity to observe” the ballot delivery. Finally, the proclamation specified that “[a]ny marked mail ballot delivered in person to the early voting clerk’s office prior to October 2, 2020, shall remain subject to the July 27, 2020 proclamation.” B. On October 2, 2020, three individuals and several organizations (collectively, “Plaintiffs”)[4] challenged the October 1 Proclamation by filing two separate actions in federal district court against Governor Abbott, Texas Secretary of State Ruth Hughs, and four local election officials. They requested a preliminary injunction against the October 1 Proclamation on the grounds that it (1) places an undue burden on their right to vote under the First and Fourteenth Amendments and (2) violates the Equal Protection Clause of the Fourteenth Amendment. Consolidating the two cases for purposes of the preliminary injunction motion, on October 9, 2020, the district court granted the motion in part, enjoining Secretary Hughs and the local officials from enforcing the Proclamation’s restriction on hand- delivering mail ballots to a single designated early voting clerk’s office. Initially, the court ruled that various threshold issues did not prevent it from deciding Plaintiffs’ claims. First, the court found that both the individual and organizational plaintiffs had standing. Second, the court rejected Secretary Hughs’ Eleventh Amendment argument, concluding she had sufficient connection to enforcing the proclamation for purposes of Ex parte Young, 209 U.S. 123 (1908). The court did, however, dismiss Governor Abbott on Eleventh Amendment grounds based on our decision in In re Abbott, 954 F.3d 772 (5th Cir. 2020). Additionally, the court declined to abstain under the Pullman doctrine, despite the fact that the October 1 Proclamation is currently being challenged in state litigation. See R.R. Comm’n of Tex. v. Pullman Co., 312 U.S. 496, 501 (1941); Moore v. Hosemann, 591 F.3d 741, 745 (5th Cir. 2009). Finally, the court declined to stay its hand under the so-called Purcell principle that a federal court should avoid altering state election rules close to an election. See Purcell v. Gonzalez, 549 U.S. 1, 6 (2006); see also Republican Nat’l Comm. v. Democratic Nat’l Comm., — U.S. —, 140 S. Ct. 1205, 1207 (2020) (per curiam). The court reasoned that it was the October 1 Proclamation, and not its injunction, that caused voter confusion and that therefore its “injunction supports the Purcell principle.” On the merits, the district court evaluated the Plaintiffs’ voting claims under the Anderson-Burdick balancing framework. See Burdick v. Takushi, 504 U.S. 428, 434 (1992); Anderson v. Celebrezze, 460 U.S. 780, 789 (1983). It found that the proclamation’s burden on Plaintiffs’ voting rights was “somewhere between ‘slight’ and ‘severe,” because it forced absentee voters to “choose between risking exposure to coronavirus to deliver their ballots in-person or disenfranchisement if the [United States Postal Service] is unable to deliver their ballots on time.” The court found this burden outweighed the State’s professed interests in ballot security, election integrity, and voter safety. Consequently, the court found Plaintiffs were likely to succeed on their voting claims. The court also found likelihood of success on Plaintiffs’ equal protection claims because the proclamation disproportionately burdened absentee voters in larger counties by subjecting them to “increased distance, increased wait time, and increased potential for exposure to the coronavirus,” without any countervailing justification. On October 9, Secretary Hughs timely appealed and, the next day, sought an emergency stay and a temporary administrative stay. On October 10, we granted a temporary stay and requested a response to her emergency stay motion, which was timely filed earlier today on October 12. II. In deciding whether to grant a stay pending appeal, “[w]e evaluate ‘(1) whether the stay applicant has made a strong showing that he is likely to succeed on the merits; (2) whether the applicant will be irreparably injured absent a stay; (3) whether issuance of the stay will substantially injure the other parties interested in the proceeding; and (4) where the public interest lies.’” Texas Democratic Party v. Abbott, 961 F.3d 389, 397 (5th Cir. 2020) (TDP I) (quoting Nken v. Holder, 556 U.S. 418, 426 (2009)). The first two factors carry the most weight. Barber v. Bryant, 833 F.3d 510, 511 (5th Cir. 2016). The party seeking the stay bears the burden of showing its need. Clinton v. Jones, 520 U.S. 681, 708 (1997). III. We first consider whether Secretary Hughs has made a strong showing she will likely succeed on the merits. We conclude she has done so on at least one ground: that the district court erred in analyzing the Plaintiffs’ voting-rights and equal protection claims. We therefore need not address, and so express no opinion about, the Secretary’s arguments concerning standing, Ex parte Young, Purcell, or Pullman abstention.[5] A. As to the Plaintiffs’ voting-rights claims, the district court applied Anderson-Burdick balancing. Under this framework, a court “must weigh the character and magnitude of the asserted injury” to voting rights “against the precise interests put forward by the State as justifications for the burden imposed by its rule.” Voting for Am., Inc. v. Steen, 732 F.3d 382, 387–88 (5th Cir. 2013) (quoting Burdick, 504 U.S. at 434; Anderson, 460 U.S. at 789) (cleaned up). A “severe burden” on voting can be justified only by state rules “narrowly drawn to advance a state interest of compelling importance.” Id. at 388 (quoting Burdick, 504 U.S. at 434). “Lesser burdens, however, trigger less exacting review, and a State’s ‘important regulatory interest’ will usually be enough to justify ‘reasonable, nondiscriminatory restrictions.’” Id. (quoting Timmons v. Twin Cities Area New Party, 520 U.S. 351, 358 (1997)). The district court classified the burden on Plaintiffs’ right to vote as “somewhere between ‘slight’ and ‘severe’” because due to the order, “absentee voters must choose between risking exposure to coronavirus to deliver their ballots in-person or disenfranchisement if the USPS is unable to deliver their ballots on time.” This burden, the court reasoned, outweighed the State’s asserted interests in ballot security, uniformity, and lessening voter confusion. The court asserted the Governor’s proclamation was “the true source of confusion and disparate treatment among voters.” Further, it found the State had not introduced evidence of voter fraud or shown that the security measures at additional drop-off locations were subpar. Therefore, the State, “by merely asserting an interest in promoting ballot security,” could not “establish that the interest outweighs a significant burden on voters.” Finally, the court added that the Governor’s authority to issue the orders was rooted in his emergency powers, which enable him to act to protect public health and safety, but the “justifications for the October 1 Order’s limitation on ballot return centers bear no relationship to protecting public health and safety.” Assuming Anderson-Burdick applies,[6] for at least two reasons we conclude the Secretary will likely show the district court erred in applying it. First, the district court vastly overstated the “character and magnitude” of the burden allegedly placed on voting rights by the October 1 Proclamation. Steen, 732 F.3d at 387 (quoting Burdick, 504 U.S. at 434). Indeed, one strains to see how it burdens voting at all. After all, the proclamation is part of the Governor’s expansion of opportunities to cast an absentee ballot in Texas well beyond the stricter confines of the Election Code. Previously, as we have explained, mail ballots could be hand-delivered to the early voting clerk only on Election Day. See TEX. ELEC. CODE § 86.006(a-1). The Governor’s July 27 Proclamation effectively extended that hand-delivery option by forty days, and the impact of the October 1 Proclamation can be measured only against that baseline. To be sure, the proclamation requires a single designated drop-off location per county during the expanded forty-day period. But that represents merely a partial refinement of the bounds of a still-existing expansion of absentee voting opportunities. In a related context, we have recently explained that to “abridg[e]” the right to vote means to “ place a barrier or prerequisite to voting, or otherwise make it more difficult to vote.” Texas Democratic Party v. Abbott, — F.3d —, 2020 WL 5422917, at *15 (5th Cir. Sept. 10, 2020) (TDP II). By contrast, “a law that makes it easier for others to vote does not abridge any person’s right to vote.” Id. The July 27 and October 1 Proclamations— which must be read together to make sense—are beyond any doubt measures that “make[] it easier” for eligible Texans to vote absentee. Id. How this expansion of voting opportunities burdens anyone’s right to vote is a mystery.[7] But even if we focused myopically on the voting options restricted by the October 1 Proclamation, as the district court did, we would still find no more than a de minimis burden on the right to vote. The district court emphasized that some absentee voters would have to travel farther to drop off mail ballots at a centralized location and that, as a result, would confront a higher risk of exposure to coronavirus. The court also warned that voters unwilling or unable to do so would “risk[] . . . disenfranchisement if the USPS is unable to deliver their ballots on time.” This drastic picture painted by the district court fails to account for the numerous ways Texans can vote early or absentee in the November 3 election. As we have recounted, under the Governor’s expansion of voting opportunities, Texans can (1) vote early in-person for an expanded period starting on October 13 (as opposed to the previous early-voting period starting on October 19); (2) hand-deliver a marked mail ballot during a forty-day period starting on September 19 (as opposed to the previous one day—Election Day—on which this was permitted); or (3) drop an absentee ballot in the mail. In light of those options, the October 1 Proclamation’s partial refinement of one avenue for absentee voting does not amount to a “severe restriction” on the right to vote. Burdick, 504 U.S. at 434; see also A. Philip Randolph Inst. of Ohio v. LaRose, — F. App’x —, 2020 WL 6013117 (6th Cir. Oct. 9, 2020) (unpublished) (concluding a similar drop-box restriction on absentee ballots “surely does not impose a ‘severe restriction[] on the right to vote’”) (quoting Mays v. LaRose, 951 F.3d 775, 779 (6th Cir. 2020)). We are especially unpersuaded by the district court’s view that voters must have multiple drop-off locations in order to “avoid the delays involved with mailing their ballots through the U.S. Postal Service.” The USPS recommends voters request absentee ballots at least fifteen days before election day to ensure timely arrival. Texas law, however, allows voters to request ballots up to eleven days before election day. Therefore, the district court concluded, a voter may legally request an absentee ballot that is not guaranteed to arrive on time. The court thus concluded that absentee voters face an insoluble choice between “risk[ing] infection with a deadly disease to return their ballots in person or disenfranchisement if the USPS is unable to deliver their ballots in time.” This kind of speculation about late-arriving ballots comes nowhere close to rendering Texas’s absentee ballot system constitutionally inadequate. Neither Plaintiffs nor the district court have cited any authority suggesting that a State must afford every voter multiple infallible ways to vote. As we explained in TDP I, mail-in ballot rules that merely make casting a ballot more inconvenient for some voters are not constitutionally suspect. 961 F.3d at 405. The principle holds true even if “circumstances beyond the state’s control, such as the presence of the [coronavirus,]” or, here, possible postal delays, make voting difficult. Id.; see also McDonald, 394 U.S. at 810 & n.8 (explaining that a State is not required to extend absentee voting privileges to all classes of citizens, even those for whom “voting may be extremely difficult, if not practically impossible,” such as persons caring for sick relatives or businessmen called away on business). We cannot conclude that speculating about postal delays for hypothetical absentee voters somehow renders Texas’s absentee ballot system constitutionally flawed. Second, the district court undervalued the state interests furthered by the October 1 Proclamation.[8] Even assuming the proclamation poses any burden on voting rights, that burden is minimal and would “trigger less exacting review,” meaning that “a State’s important regulatory interests will usually be enough to justify reasonable, nondiscriminatory restrictions.” Steen, 732 F.3d at 388 (internal quotations and citation omitted). The district court found that Texas’s “vague interests” in ballot security, election uniformity, and avoiding voter confusion were insufficiently substantiated to outweigh the proclamation’s “significant burden on voters.” The Secretary is likely to show on appeal that the district court erred. States have critically important interests in the orderly administration of elections and in vigilantly reducing opportunities for voting fraud. See Crawford v. Marion Cty. Election Bd., 553 U.S. 181, 195–96 (2008) (“There is no question about the legitimacy or importance of the State’s interest in counting only the votes of eligible voters” and “in orderly administration and accurate recordkeeping[.]“). Indeed, both the Supreme Court and our court have recognized that “mail-in voting” is “far more vulnerable to fraud” than other forms of voting. Veasey v. Abbott, 830 F.3d 216, 263 (5th Cir. 2016) (en banc) (observing that, in contrast to in-person voting, record evidence showed “mail-in voting . . . is far more vulnerable to fraud, particularly among the elderly”); id. (criticizing challenged law because it “does nothing to address the far more prevalent issue of fraudulent absentee ballots”); see also Crawford, 553 U.S. at 195–96 & n.12 (discussing examples “in recent years” of “fraudulent voting . . . perpetrated using absentee ballots and not in-person fraud . . . demonstrat[ing] that not only is the risk of voter fraud real but that it could affect the outcome of a close election”); TDP I, 961 F.3d at 414 (Ho, J., concurring) (observing that “courts have repeatedly found that mail-in ballots are particularly susceptible to fraud”) (and collecting authorities). In sum, “[w]hile the most effective method of preventing election fraud may well be debatable, the propriety of doing so is perfectly clear.” Crawford, 553 U.S. at 196. It is therefore evident that Texas has an “important regulatory interest” in policing how its citizens’ votes are collected and counted. Steen, 732 F.3d at 388. This interest is acute when it comes to mail-in ballots. Crawford, 553 U.S. at 195–96 & n.12; see also Veasey, 830 F.3d at 239 (concluding “mail-in ballot fraud is a significant threat”). It is likely, then, that the Secretary will prevail on appeal in showing that the October 1 Proclamation was justified by the State’s important interests in election integrity. As explained, that proclamation is part-and-parcel of a sizable expansion of absentee voting options occasioned by the Governor’s pandemic-related orders. See In re Steven Hotze, 2020 WL 5919726, at *1–2. Opportunities for absentee voters to hand-deliver ballots ballooned from a pre-COVID one day (Election Day itself) to an in-COVID forty days. The evidence showed this expansion of absentee voting provoked an increase in drop-off locations in certain counties. While this reaction is understandable, also understandable is the Governor’s goal of centralizing delivery locations, and deploying poll watchers there, in order to maximize ballot security. The Secretary is thus likely to show on appeal that the October 1 Proclamation was a “reasonable, nondiscriminatory” measure justified by Texas’s important interests in election integrity. Steen, 732 F.3d at 388. The Secretary is also likely to show that the district court erred in scrutinizing whether the proclamation furthered those interests. Cf. id. (requiring “less exacting review” for laws placing “[l]esser burdens” on voting rights). For example, the district court demanded evidence of “actual examples of voter fraud” justifying the centralization of mail ballot delivery locations. Such evidence has never been required to justify a state’s prophylactic measures to decrease occasions for vote fraud or to increase the uniformity and predictability of election administration. For instance, in Crawford, the Supreme Court upheld Indiana’s voter identification law, despite the fact that “[t]he record contain[ed] no evidence of [in-person voter] fraud actually occurring in Indiana at any time in its history.” 553 U.S. at 181. “Here, as in Crawford, Texas need not show specific local evidence of fraud in order to justify preventive measures.” Steen, 732 F.3d at 394; see also Munro v. Socialist Workers Party, 479 U.S. 189, 194–95 (1986) (“We have never required a State to make a particularized showing of the existence of voter confusion, ballot overcrowding, or the presence of frivolous candidacies prior to the imposition of reasonable [voting regulations].”). Furthermore, the Secretary articulated other interests beyond fraud, such as uniformity across counties. The district court too quickly discounted the state’s interest in promoting uniformity in how mail ballots are delivered. In doing so, the court appears to have overlooked evidence showing the unusual nature of the developing absentee ballot process. As the Director of Elections stated, following the July 27 Proclamation, only four of Texas’s 254 counties announced their intention “to utilize more than one location for in-person delivery of mail ballots,” and there were concerns about whether the additional locations complied with Texas election law. In sum, Secretary Hughs has shown she is likely to prevail on the merits of the Plaintiffs’ voting-rights claims. B. We turn to Plaintiffs’ equal protection claims.[9] Because the right to vote is fundamental, Burdick, 504 U.S. at 433, “once the franchise is granted to the electorate, lines may not be drawn which are inconsistent with the Equal Protection Clause of the Fourteenth Amendment.” Harper v. Virginia State Bd. of Elec., 383 U.S. 663, 665 (1966); Gray v. Sanders, 372 U.S. 368, 379 (1963) (“[A]ll who participate in the election are to have an equal vote— whatever their race, whatever their sex, whatever their occupation, whatever their income, and wherever their home may be in [the] geographical unit. This is required by the Equal Protection Clause.”). The district court held the Plaintiffs were likely to show that the October 1 Proclamation violated these principles by imposing disproportionate burdens on voters based on where they live. Plaintiffs argued that the elimination of additional drop-off locations would force voters in large and populous counties to travel farther, wait longer, and risk increased exposure to the coronavirus. Meanwhile, voters in smaller and less populous counties would not face the same difficulties. Because the district court concluded that the proclamation resulted in disparate treatment of voters based on county of residence, it applied the Anderson-Burdick framework. It reasoned that absent evidence that drop-off locations have posed or will pose a threat of voter fraud, Texas’s proffered interest in election integrity was not “sufficiently weighty” to justify the differential burdens on voters. The Secretary is likely to show this analysis was mistaken. As with the voting-rights claim, the district court misconstrued the nature of the alleged burden imposed by the October 1 Proclamation. It is true that the Equal Protection Clause guarantees that “every voter is equal to every other voter in his State,” regardless of race, sex, occupation, wealth, or residence. Gray, 372 U.S. at 380. But the district court accepted Plaintiffs’ contention that the proclamation “dol[es] out electoral opportunity based on county lines.” That is incorrect. The proclamation establishes a uniform rule for the entire State: each county may designate one early voting clerk’s office at which voters may drop off mail ballots during the forty days leading up to the election. That voters who live further away from a drop-off location may find it inconvenient to take advantage of this particular, additional method to cast their ballots does not “limit[] electoral opportunity,” as the district court thought. As we have explained, the October 1 Proclamation was part of an expansion of absentee voting opportunities beyond what the Texas Election Code provided. The fact that this expansion is not as broad as Plaintiffs would wish does not mean that it has illegally limited their voting rights. Moreover, the cases relied on by the district court are easily distinguishable. The court cited Moore v. Ogilvie, 394 U.S. 814 (1969), and Gray v. Sanders, 372 U.S. 368, to support its conclusion that the October 1 Proclamation necessarily treats voters differently on the basis of county residence. But both Moore and Gray confronted state election laws that effectively gave more weight to the votes cast by voters in certain counties. The Illinois statute in Moore required independent candidates for the offices of electors to obtain a set number of voters’ signatures from each of at least fifty counties. The Court invalidated the law because it gave voters in some counties “greater voting strength” than others, an idea “hostile to the one man, one vote basis of our representative government.” Moore, 394 U.S. at 819. Similarly, Gray examined Georgia’s county unit system of counting votes, under which the candidate who won each county was considered to have “carried the county” and received votes corresponding to that county’s number of representatives. As a result of widely varying populations per county, one “unit vote” in one county represented less than 1,000 residents, while a unit vote in another county represented over 90,000 residents. Such a system “weights the rural vote more heavily than the urban vote and weights some small rural counties heavier than other larger rural counties.” Gray, 372 U.S. at 379. The effects of the October 1 Proclamation are nothing like the effects of the laws in Moore and Gray. As we have explained, supra III(A), the burden imposed by the proclamation is at most de minimis. More to the point, it applies a uniform rule to every Texas county and does not weight the votes of those in some counties more heavily than others. Consequently, Secretary Hughs is likely to show that the October 1 Proclamation does not impermissibly classify voters based on county of residence. Moreover, a “state’s important regulatory interests are generally sufficient to justify reasonable, nondiscriminatory” voting regulations. Anderson, 460 U.S. at 788. As we have explained, supra III(A), the Secretary has articulated important state interests in ensuring election uniformity and integrity that the October 1 Proclamation furthers. IV. Having concluded the Secretary will likely succeed on the merits, we address the remaining Nken factors: “whether [the Secretary] will be irreparably injured absent a stay,” “whether issuance of the stay will substantially injure” other interested parties, and “where the public interest lies.” Nken, 556 U.S. at 426. The Secretary has shown irreparable harm absent a stay. When a district court’s injunction prevents a State from effectuating its own election procedures, put in place by elected officials, it suffers irreparable harm. See TDP I, 961 F.3d at 411 (holding an injunction that effectively required “Texas to institute [an absentee ballot] policy against its will presents significant, irreparable harm”). The remaining two factors are also met. Issuing a stay will not substantially injure Plaintiffs, who retain numerous avenues for casting their absentee ballots under the expanded voting opportunities afforded by the Governor’s proclamations. What we said recently in TDP I applies equally here: “Given the great likelihood that the state officials will ultimately succeed on the merits, combined with the undeniable, irreparable harm that the injunction would inflict on them—factors that we consider ‘the most critical,’—we hold that the balance of harms weighs in favor of the state officials.” 961 F.3d at 412 (citation omitted). Finally, we conclude that public interest favors the Secretary. When a State is the party appealing an injunction, “its interest and harm merge with that of the public.” Veasey v. Abbott, 870 F.3d 387, 391 (5th Cir. 2017) (per curiam). Because the Secretary has met her burden, we exercise our discretion to grant a stay pending appeal. *** Leaving the Governor’s October 1 Proclamation in place still gives Texas absentee voters many ways to cast their ballots in the November 3 election. These methods for remote voting outstrip what Texas law previously permitted in a pre-COVID world. The October 1 Proclamation abridges no one’s right to vote. The Secretary’s emergency motion for stay pending appeal is GRANTED. JAMES C. HO, Circuit Judge, concurring: I concur fully in Judge Duncan’s typically thoughtful opinion. But I also do so grudgingly. I firmly agree that the federal district court usurped the authority that our Constitution vests in state legislatures to set the rules governing federal elections. But so did the Governor of Texas—as Judge Duncan also cautions. See supra at n.7. The district court was wrong to rewrite Texas law. But the distinguished judge who did so was simply following in the Governor’s footsteps. It is surely just as offensive to the Constitution to rewrite Texas election law by executive fiat as it is to do so by judicial fiat. Yet that is what occurred here. Respected legislators and public leaders called on the Governor to call a special session so that legislators in both parties could consider and debate amendments to the state’s election rules to accommodate voter concerns arising out of the pandemic. But the Governor rejected those calls, and instead issued a series of executive proclamations purporting to unilaterally “suspend” various Texas election laws. Those actions have generated significant controversy. Members of the Texas Supreme Court described the Governor’s actions as “a clear abuse of discretion of a public official,” In re Hotze, S.W.3d , (Tex. Oct. 7, 2020) (Devine, J., dissenting) (emphasis in original) (quotations omitted), that “raise[s] important questions about the constitutionality of government action during the coronavirus crisis,” id. at (Blacklock, J., concurring). Only the district court’s rewriting of Texas law is before us today, however. And that leads us to an unfortunate irony: by setting aside only the district court’s rewriting of Texas law, we must restore the Governor’s rewriting of Texas law. It recalls the adage that sometimes it’s only the guy who throws the second punch that gets caught. THE DICTIONARY OF MODERN PROVERBS 209 (2012). I grudgingly concur. I. Under the Constitution, it is the state legislature—not the governor or federal judges—that is authorized to establish the rules that govern the election of each state’s Presidential electors, U.S. Senators, and U.S. Representatives. See U.S. CONST. art. I, § 4, cl. 1 (“The Times, Places and Manner of holding Elections for Senators and Representatives, shall be prescribed in each State by the Legislature thereof . . . .”); U.S. CONST. art. II, § 1, cl. 2 (“Each State shall appoint, in such Manner as the Legislature thereof may direct, a Number of Electors [for President and Vice President].”).[10] But apparently that is not how federal elections will be administered in Texas this year. If officials were following Texas law, “[t]he period for early voting by personal appearance” would “begin[] on the 17th day before election day”— or Monday, October 19, 2020. TEX. ELEC. CODE § 85.001(a). See also TEX. ELEC. CODE § 85.001(c) (“If the date prescribed . . . for beginning the period is a Saturday, Sunday, or legal state holiday, the early voting period begins on the next regular business day.”). But not this year. On July 27, 2020, the Governor of Texas proclaimed that, due to the COVID-19 pandemic, he will “suspend Section 85.001(a)” so that “early voting by personal appearance shall begin on Tuesday, October 13, 2020″—six days earlier than the date provided by the Texas Legislature. Furthermore, Texas law ordinarily provides that, if qualified individuals elect to receive their ballots by mail but ultimately choose to deliver their marked ballots in person, they may do so “only while the polls are open on election day.” TEX. ELEC. CODE § 86.006(a-1). But once again, not this year. In that same July 27 proclamation, the Governor of Texas announced that he would also “suspend Section 86.006(a-1)” and “allow a voter to deliver a marked mail ballot in person . . . prior to . . . election day”—again in direct conflict with the framework set forth by the Texas Legislature. And on October 1, 2020, the Governor amended his July 27 proclamation to make clear that he would suspend section 86.006(a-1) unless a county designates more than one location for qualified voters to deliver marked mail ballots, or offers a location that is not monitored by poll watchers—again without support in the Texas Election Code. It did not have to be this way. The Texas Constitution imposes strict limits on the number of days the Legislature can meet in regular session to consider legislation—once every two years for 140 days. See TEX. CONST. art. 3, §§ 5, 24. But it also empowers the Governor to call the Legislature back for a special session to focus on any topic of his choosing. See TEX. CONST. art. 3, § 40. So the Governor did not have to act unilaterally to amend Texas election law in the wake of the pandemic. He could have called a special session. Indeed, a number of respected legislators and public leaders urged him to do just that—to quote one particularly emphatic plea, “if ever a special session was justified, now is the time.”[11] But instead, the Governor concluded that a special session was unnecessary because the Texas Disaster Act of 1975 gives him the authority to legislate all by himself. That act, however, only gives the governor limited power to “suspend the provisions of any regulatory statute prescribing the procedures for conduct of state business . . . if strict compliance with [such requirements] would in any way prevent, hinder, or delay necessary action in coping with a disaster.” TEX. GOV’T CODE § 418.016(a) (emphases added). Moreover, the Act, like any other Texas law, must be construed in light of the Texas Constitution. That includes the constitutional provision that “[n]o power of suspending laws in this State shall be exercised except by the Legislature.” TEX. CONST. art. I, § 28. See also In re Hotze, ___S.W.3d ___, ___(Tex. July 17, 2020) (Devine, J., concurring) (“I find it difficult to square [the Texas Disaster Act of 1975], and the orders made under it, with the Texas Constitution.”). It also includes the Governor’s constitutional authority to call special sessions of the Legislature. TEX. CONST. art. 3, § 40. It is difficult to see how it is “necessary . . . in coping with a disaster” for the governor to suspend provisions of the Texas Election Code over three months before the November election. “[W]hen a crisis stops being temporary, and as days and weeks turn to months and years, the slack in the leash eventually runs out.” Capitol Hill Baptist Church v. Bowser, No. 20-cv- 02710 (TNM), slip op. at 15 (D.D.C. Oct. 9, 2020). And that is especially so considering that the Constitution expressly forbids anyone other than the Legislature from suspending Texas laws—and considering that members of the Legislature are not only willing and able but demanding to convene a special session to consider legislation in response to the pandemic.[12] But now that the Governor has paved the way for rewriting Texas election law based on personal policy disagreements over how elections should be run during the pandemic, it should surprise no one that a federal district court has seen fit to jump in as well, in response to the “executive- caused voter confusion” resulting from “Governor Abbott’s unilateral decision to reverse his July 27 Order.” Tex. League of United Latin American Citizens v. Abbott, No. 1:20-cv-01006-RP, at 33 (W.D. Tex. Oct. 9, 2020) (emphasis in original). On October 9, a federal district court entered a preliminary injunction that effectively set aside a portion of the Governor’s October 1 proclamation in favor of his July 27 proclamation. Under the preliminary injunction, state officials are enjoined from forbidding counties to establish more than a single location where qualified voters can deliver marked mail ballots. In response, the Secretary of State seeks a stay of the preliminary injunction pending appeal. Tellingly, however, nowhere in her stay papers does the Secretary suggest that the preliminary injunction conflicts with Articles I and II of the U.S. Constitution—perhaps because she recognizes that the Governor’s proclamations suffer from the same defect. That said, no one is asking us to set aside the Governor’s July 27 proclamation. To the contrary, the plaintiffs want that proclamation enforced as is—while the State of Texas wants the July 27 proclamation enforced, but only as amended by the Governor’s October 1 proclamation. So we must take the July 27 proclamation as a given. The only question before us is whether the district court was correct to set aside a portion of the Governor’s October 1 proclamation. I agree with my colleagues that the district court was wrong to do so, and that a stay should therefore be granted. II. None of this is to say, of course, that there are not valid policy reasons to support the conflicting judgments reached by the Governor and the federal district court. The ongoing global pandemic has already roiled the lives and livelihoods of millions of Texans. It is understandable that citizens have strong views on the myriad ways that election rules and procedures might be reformed to maximize voter access in these difficult and challenging times. After all, “[t]o lose the ability to vote in an upcoming election due to fear of the pandemic would be beyond heartbreaking for citizens who are already hurting, for it is a right they will never be able to recover.” Tex. Democratic Party v. Abbott, 961 F.3d. 389, 413 (5th Cir. 2020) (Ho, J., concurring) (quotations omitted). So the Governor may well believe sincerely that expanding the early voting period furthers the goal of maximizing voter access, and that limiting where mail ballots may be delivered in person will help maximize ballot integrity. On the other hand, the plaintiffs counter that the Governor’s approach to mail ballots gets it backward—and that in fact there is a greater risk of fraud when ballots are returned by mail than when they are delivered in-person. To quote the plaintiffs: “the unrebutted evidence demonstrates that allowing voters to return their absentee ballots at the annexes is ‘more secure than returning [ballots] by mail.’” Numerous courts agree. As a distinguished panel of the Seventh Circuit once observed: “Voting fraud is a serious problem in U.S. elections generally . . . and it is facilitated by absentee voting [A]bsentee voting is to voting in person as a take-home exam is to a proctored one.” Griffin v. Roupas, 385 F.3d 1128, 1130–31 (7th Cir. 2004) (collecting authorities). See also Tex. Democratic Party, 961 F.3d. at 414 (Ho, J., concurring) (“[C]ourts have repeatedly found that mail-in ballots are particularly susceptible to fraud.”) (collecting cases).[13] But if changes to Texas election rules are warranted in response to the pandemic, they must be made consistent with the Constitution. And under our Constitution, it is for the Texas Legislature through the legislative process—and not for the Governor or the judiciary by executive or judicial fiat—to determine how best to maximize voter access as well as ballot security. See, e.g., Griffin, 385 F.3d at 1131 (noting that “the striking of the balance between discouraging fraud and other abuses and encouraging turnout is quintessentially a legislative judgment”). What’s more, there may be special cause for concern when unilateral changes to election laws are made by a single elected official. As the Chief Justice once wrote, “those who govern should be the last people to help decide who should govern.” McCutcheon v. FEC, 572 U.S. 185, 192 (2014) (plurality opinion). He made that observation in the context of a case involving campaign finance regulation. But the same principle readily applies to any area of election law. Indeed, skepticism about politicians regulating politics is “deeply engrained in our nation’s DNA.” Stringer v. Whitley, 942 F.3d 715, 725 (5th Cir. 2019) (Ho, J., concurring). “As Americans, we have never trusted the fox to guard the henhouse.” Id. So the Governor’s actions in this case should trouble you regardless of whether you agree or disagree with any of his actions as a policy matter. For there is a more fundamental principle at stake: If a governor can unilaterally suspend early voting laws to reach policy outcomes that you prefer, it stands to reason that a governor can also unilaterally suspend other election laws to achieve policies that you oppose. Want to expand voting by mail? Too bad—the governor can suspend mail-in ballots all by himself, for the same reason restaurants have replaced paper menus with online ones in response to consumer concerns about the pandemic. Want to restrict voting by mail? Sorry—the governor can expand mail-in voting on his own, because some people fear going to the polls during the pandemic. But that of course is not how our Constitution works. The Constitution vests control over federal election laws in state legislatures, and for good reason—that’s where we expect the voice of the people to ring most loudly and effectively. Moreover, change by other means doesn’t just undermine respect for legal process. It threatens to undermine the very legitimacy of the election results—the last thing we need in these divisive and uncertain times.[14] I concur.
