HAYNES, Circuit Judge:

In 2011, Texas ("the State") passed Senate Bill 14 ("SB 14"), which requires individuals to present one of several forms of photo identification in order to vote. See Act of May 16, 2011, 82d Leg., R.S., ch. 123, 2011 Tex. Gen. Laws 619. Plaintiffs filed suit challenging the constitutionality and legality of the law. The district court held that SB 14 was enacted with a racially discriminatory purpose, has a racially discriminatory effect, is a poll tax, and unconstitutionally burdens the right to vote. See Veasey v. Perry, 71 F.Supp.3d 627, 633 (S.D.Tex.2014).

We VACATE and REMAND the Plaintiffs' discriminatory purpose claim for further consideration in light of the discussion below. If on remand the district court finds that SB 14 was passed with a discriminatory purpose, then the law must be invalidated. However, because the finding on remand may be different, we also address other arguments raised by the Plaintiffs. We AFFIRM the district court's finding that SB 14 has a discriminatory effect in violation of Section 2 of the Voting Rights Act and remand for consideration of the proper remedy. We VACATE the district court's holding that SB 14 is a poll tax and RENDER judgment in the State's favor. Because the same relief is available to Plaintiffs under the discriminatory effect finding affirmed herein, under the doctrine of constitutional avoidance, we do not address the merits of whether SB 14 unconstitutionally burdens the right to vote under the First and Fourteenth Amendments. We therefore VACATE this portion of the district court's opinion and DISMISS Plaintiffs' First and Fourteenth Amendment claims.

I. Factual Background and Procedural History

A. Senate Bill 14

Prior to the implementation of SB 14, a Texas voter could cast a ballot in person by presenting a registration certificate — a document mailed to voters upon registration. TEX. ELEC.CODE §§ 13.142, 63.001(b) (West 2010). Voters appearing without the certificate could cast a ballot by signing an affidavit and presenting one of multiple forms of identification ("ID"), including a current or expired driver's license, a photo ID (including employee or student IDs), a utility bill, a bank statement, a paycheck, a government document showing the voter's name and address, or mail addressed to the voter from a government agency. Id. §§ 63.001, 63.0101 (West 2010).

With the implementation of SB 14, Texas began requiring voters to present certain specific forms of identification at the polls. These include: (1) a Texas driver's license or personal identification card issued by the Department of Public Safety ("DPS") that has not been expired for more than 60 days; (2) a U.S. military identification card with a photograph that has not been expired for more than 60 days; (3) a U.S. citizenship certificate with a photo; (4) a U.S. passport that has not been expired for more than 60 days; (5) a license to carry a concealed handgun issued by DPS that has not been expired for more than 60 days; or (6) an Election Identification Certificate ("EIC") issued by DPS that has not been expired for more than 60 days. TEX. ELEC.CODE § 63.0101 (West Supp.2014).1

SB 14 states that DPS "may not collect a fee for an [EIC] or a duplicate [EIC]," TEX. TRANSP. CODE § 521A.001(b) (West 2013), and allows DPS to promulgate rules for obtaining an EIC. Id. § 521A.001(f); § 521.142. To receive an EIC, DPS rules require a registered voter to present either: (A) one form of primary ID, (B) two forms of secondary ID, or (C) one form of secondary ID and two pieces of supporting information. 37 TEX. ADMIN. CODE § 15.182(1). Thus, any application for an EIC requires either one Texas driver's license or personal identification card that has been expired for less than two years, or one of the following documents, accompanied by two forms of supporting identification: (1) an original or certified copy of a birth certificate from the appropriate state agency; (2) an original or certified copy of a United States Department of State Certification of Birth for a U.S. citizen born abroad; (3) U.S. citizenship or naturalization papers without a photo; or (4) an original or certified copy of a court order containing the person's name and date of birth and indicating an official change of name and/or gender. Id. § 15.182(3).2

Before May 27, 2015, a statutory provision distinct from SB 14 imposed a $2 or $3 fee for a certified copy of a birth certificate.3 TEX. HEALTH & SAFETY CODE § 191.0045 (West 2010). As discussed below, after the district court issued its judgment, the Texas Legislature passed Senate Bill 983 during the 2015 legislative session and eliminated this fee.

Persons who have a disability are exempt from SB 14's photo ID requirement once they provide the voter registrar with documentation of their disability from the U.S. Social Security Administration or Department of Veterans Affairs. TEX. ELEC. CODE § 13.002(i) (West Supp.2014). Other persons may vote by provisional ballot without a photo ID if they file affidavits either asserting a religious objection to being photographed or that their SB 14 ID was lost or destroyed as a result of a natural disaster occurring within 45 days of casting a ballot. Id. § 65.054. Additionally, voters who will be 65 or older as of the date of the election may vote early by mail. Id. § 82.003.

If a voter is unable to provide SB 14 ID at the poll, the voter can cast a provisional ballot after executing an affidavit stating that the voter is registered and eligible to vote. Id. § 63.001(a), (g). The vote counts if the voter produces SB 14 ID to the county registrar within six days of the election. Id. § 65.0541.

SB 14 requires county registrars to inform applicants of the new voter ID requirements when issuing voter registration certificates, id. § 15.005, and requires both the Secretary of State and voter registrar of each county with a website to post SB 14's requirements online. Id. § 31.012(a). The requirements must also be placed prominently at polling places. Id. § 62.016. Additionally, the Secretary of State must "conduct a statewide effort to educate voters regarding the identification requirements for voting." Id. § 31.012(b). The district court found that SB 14 allocated a one-time expenditure of $2 million for voter education.4 Veasey, 71 F.Supp.3d at 649.

B. Procedural History

The State began enforcing SB 14 on June 25, 2013.5 The plaintiffs and intervenors (collectively, "Plaintiffs") filed suit against Defendants to enjoin enforcement of SB 14, and their suits were consolidated before one federal district court in the Southern District of Texas. See Veasey, 71 F.Supp.3d at 632. Plaintiffs claim that SB 14's photo identification requirements violate the Fourteenth and Fifteenth Amendments to the United States Constitution and Section 2 of the Voting Rights Act because SB 14 was enacted with a racially discriminatory purpose and has a racially discriminatory effect. Plaintiffs also claim that SB 14's photo ID requirement places a substantial burden on the fundamental right to vote under the First and Fourteenth Amendments, and constitutes a poll tax under the Fourteenth and Twenty-Fourth Amendments. The State defends SB 14 as a constitutional requirement imposed to prevent in-person voter fraud and increase voter confidence and turnout.

The district court conducted a nine-day bench trial at which dozens of expert and lay witnesses testified by deposition or in person. Following that bench trial, the district court issued a lengthy and comprehensive opinion holding:

Veasey, 71 F.Supp.3d at 633. Shortly before in-person early voting was scheduled to begin for the November 2014 elections, the district court "enter[ed] a permanent and final injunction against enforcement of the voter identification provisions [of SB 14], Sections 1 through 15 and 17 through 22," not enjoining sections 16, 23, and 24 in accordance with SB 14's severability clause.6 Id. at 707 & n. 583. Since it struck the State's voter ID law so close to the impending November 2014 election, the district court ordered the State to "return to enforcing the voter identification requirements for in-person voting in effect immediately prior to the enactment and implementation of SB 14." Id. The district court retained jurisdiction to review any remedial legislation and to pre-approve any administrative remedial measures. Id. at 707-08.

In October 2014, the State appealed the district court's final judgment, and this court granted the State's emergency motion for stay pending appeal, grounding its decision primarily in "the importance of maintaining the status quo on the eve of an election." Veasey v. Perry, 769 F.3d 890, 895 (5th Cir.2014). Plaintiffs filed emergency motions before the Supreme Court, seeking to have this court's stay vacated. The Supreme Court denied these motions to vacate the stay of the district court's judgment. See Veasey v. Perry, ___ U.S. ___, 135 S.Ct. 9, 190 L.Ed.2d 283 (2014). Therefore, this court's stay of the district court's injunction remained in place, and SB 14 continues to be enforced.

C. Senate Bill 983

On May 27, 2015, after oral argument was heard on this appeal, Senate Bill 983 ("SB 983") was signed into law, eliminating the fee "for searching or providing a record, including a certified copy of a birth record, if the applicant [for the record] states that the applicant is requesting the record for the purpose of obtaining an election identification certificate...." Act of May 25, 2015, 84th Leg., R.S., ch. 130, 2015 Tex. Sess. Laws Serv. Ch. 130 (West) (to be codified as an amendment to TEX. HEALTH & SAFETY CODE § 191.0046(e)) (hereinafter "SB 983"). SB 983 became effective immediately. Id. §§ 2-3 (to be codified as Note to TEX. HEALTH & SAFETY CODE § 191.0046); see also S.J. of Tex., 84th Leg., R.S. 1449-50 (2015) (reporting unanimous passage out of the Texas Senate); H.J. of Tex., 84th Leg., R.S., 4478-79 (2015) (reporting passage by 142 to 0, with one member absent, in the Texas House). SB 983 provides that "a local registrar or county clerk who issues a birth record" required for an EIC that would otherwise be entitled to collect a fee for that record "is entitled to payment of the amount from the [D]epartment [of State Health Services]." Act of May 25, 2015, 84th Leg., R.S., ch. 130 (to be codified as an amendment to TEX. HEALTH & SAFETY CODE § 191.0046(f)). SB 983 did not appropriate funds to spread public awareness about the free birth records.

The parties filed Federal Rule of Appellate Procedure 28(j) letters noting SB 983's passage.7 The State emphasizes that SB 983 would prevent voters from being charged $2 to $3 for birth certificates necessary to obtain EICs, would eliminate fees to search for those records, and that "[t]he State will reimburse local governments any amount they would have retained had a fee been charged." Therefore, the State argues that the Legislature "does not harbor some invidious institutional purpose" and that SB 983 "eliminates the core factual premise of plaintiffs' already-unavailing claims that SB14 imposes an [unconstitutional] burden [under the First and Fourteenth Amendments], violates VRA § 2, and constitutes a poll tax." Id. Plaintiffs also filed Rule 28(j) letters, asserting that SB 983 does not affect the district court's discriminatory purpose or effect analyses or its unconstitutional burden analysis. Plaintiffs highlight that the Legislature passed SB 983 only after oral argument was held in this case and that the Legislature ignored many more comprehensive bills that were submitted during this legislative session.

II. Standing

Article III standing cannot be waived or assumed, Rohm & Hass Tex., Inc. v. Ortiz Bros. Insulation, Inc., 32 F.3d 205, 207 (5th Cir.1994), and we review questions of standing de novo. See Nat'l Rifle Ass'n of Am., Inc. v. Bureau of Alcohol, Tobacco, Firearms, & Explosives, 700 F.3d 185, 190 (5th Cir.2012). As most of the private, political, and organizational plaintiffs have standing, we have jurisdiction to consider the claims raised on appeal. Nat'l Rifle Ass'n of Am., Inc. v. McCraw (McCraw), 719 F.3d 338, 344 n. 3 (5th Cir.2013) ("Only one of the petitioners needs to have standing to permit us to consider the petition for review." (quoting Massachusetts v. EPA, 549 U.S. 497, 518, 127 S.Ct. 1438, 167 L.Ed.2d 248 (2007))). However, a court should not permit a party that it knows lacks standing to participate in the case. See id.

In its brief, the Texas League of Young Voters Education Fund ("Texas League") states that it has "ceased operations." "A claim becomes moot when `the parties lack a legally cognizable interest in the outcome.'" Id. at 344 (quoting Powell v. McCormack, 395 U.S. 486, 496, 89 S.Ct. 1944, 23 L.Ed.2d 491 (1969)). Thus, the mootness doctrine "ensures that the litigant's interest in the outcome continues to exist throughout the life of the lawsuit ... including the pendency of the appeal." McCorvey v. Hill, 385 F.3d 846, 848 (5th Cir.2004) (citation and internal quotation marks omitted). Because the Texas League no longer suffers the injury allegedly imposed by SB 14, we conclude that its claims are moot. See McCraw, 719 F.3d at 344. As other Plaintiffs have standing, we nonetheless have jurisdiction over the appeal. Id. at 344 n. 3.

III. Discussion

A. Discriminatory Purpose

The State appeals the district court's judgment that SB 14 was passed with a discriminatory purpose in violation of the Fourteenth and Fifteenth Amendments and Section 2 of the Voting Rights Act. We review this determination for clear error; as the district court did, we apply the framework articulated in Village of Arlington Heights v. Metropolitan Housing Development Corp., 429 U.S. 252, 265-68, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977), which remains the proper analytical framework for these kinds of cases. See Price v. Austin Indep. Sch. Dist., 945 F.2d 1307, 1312 (5th Cir.1991). "If the district court's findings are plausible in light of the record viewed in its entirety, we must accept them, even though we might have weighed the evidence differently if we had been sitting as a trier of fact." Id. (citation and internal quotation marks omitted). However, if the district court committed an error of law in making its fact findings in this case, we may set aside those fact findings and remand the case for further consideration. See Pullman-Standard v. Swint, 456 U.S. 273, 291-92, 102 S.Ct. 1781, 72 L.Ed.2d 66 (1982). In the words of the Supreme Court, when the district court's "findings are infirm because of an erroneous view of the law, a remand is the proper course unless the record permits only one resolution of the factual issue." Id. Although the district court properly cited the Arlington Heights framework, we conclude that some "findings are infirm," necessitating a remand on this point.

"Proof of racially discriminatory intent or purpose is required to show a violation of the Equal Protection Clause." Arlington Heights, 429 U.S. at 265, 97 S.Ct. 555. However, "[r]acial discrimination need only be one purpose, and not even a primary purpose, of an official action for a violation to occur." United States v. Brown, 561 F.3d 420, 433 (5th Cir.2009) (citation and internal quotation marks omitted). Arlington Heights enumerated a multi-factor analysis for evaluating whether a facially neutral law was passed with a discriminatory purpose, and courts must perform a "sensitive inquiry into such circumstantial and direct evidence of intent as may be available." See 429 U.S. at 266, 97 S.Ct. 555. The appropriate inquiry is not whether legislators were aware of SB 14's racially discriminatory effect, but whether the law was passed because of that disparate impact. See Pers. Adm'r of Mass. v. Feeney, 442 U.S. 256, 278-79, 99 S.Ct. 2282, 60 L.Ed.2d 870 (1979).8 Importantly, although discriminatory effect is a relevant consideration, knowledge of a potential impact is not the same as intending such an impact. See id.; see also Ashcroft v. Iqbal, 556 U.S. 662, 676, 129 S.Ct. 1937, 173 L.Ed.2d 868 (2009) ("Under extant precedent purposeful discrimination requires more than `intent as volition or intent as awareness of consequences.'" (quoting Feeney, 442 U.S. at 279, 99 S.Ct. 2282)); Arlington Heights, 429 U.S. at 266, 97 S.Ct. 555 (noting that "[t]he impact of the official action ... may provide an important starting point" under a discriminatory purpose analysis (emphasis added)).

The Court articulated the following non-exhaustive list of factors to guide courts in this inquiry: (1) "[t]he historical background of the decision ... particularly if it reveals a series of official actions taken for invidious purposes," (2) "[t]he specific sequence of events leading up to the challenged decision," (3) "[d]epartures from normal procedural sequence," (4) "substantive departures ... particularly if the factors usually considered important by the decisionmaker strongly favor a decision contrary to the one reached," and (5) "[t]he legislative or administrative history ... especially where there are contemporary statements by members of the decision making body, minutes of its meetings, or reports." Arlington Heights, 429 U.S. at 267-68, 97 S.Ct. 555. "Once racial discrimination is shown to have been a `substantial' or `motivating' factor behind enactment of the law, the burden shifts to the law's defenders to demonstrate that the law would have been enacted without this factor." Hunter v. Underwood, 471 U.S. 222, 228, 105 S.Ct. 1916, 85 L.Ed.2d 222 (1985). If the law's defenders are unable to carry this burden, the law is invalidated. See id. at 231, 105 S.Ct. 1916.9

The State's stated purpose in passing SB 14 centered on protection of the sanctity of voting, avoiding voter fraud, and promoting public confidence in the voting process. No one questions the legitimacy of these concerns as motives; the disagreement centers on whether there were impermissible motives as well. We recognize that evaluating motive, particularly the motive of dozens of people, is a difficult enterprise. We recognize the charged nature of accusations of racism, particularly against a legislative body, but we also recognize the sad truth that racism continues to exist in our modern American society despite years of laws designed to eradicate it.

Against this backdrop, we respect and appreciate the district court's efforts to address this difficult inquiry. We now examine the evidence upon which the district court relied and find some of it "infirm." In seeking to discern the Legislature's intent under the Arlington Heights framework, the district court relied extensively on Texas's history of enacting racially discriminatory voting measures. See Veasey, 71 F.Supp.3d at 633-39. It noted, for instance, Texas's use of all-white primaries from 1895-1944, literacy tests and secret ballots from 1905-1970, and poll-taxes from 1902-1966. Id. at 634. All of the most pernicious discriminatory measures predate 1965. See Shelby Cnty. v. Holder, ___ U.S. ___, 133 S.Ct. 2612, 2628, 186 L.Ed.2d 651 (2013) (noting that "history did not end in 1965"). In McCleskey v. Kemp, the Supreme Court held that "unless historical evidence is reasonably contemporaneous with the challenged decision, it has little probative value." 481 U.S. 279, 298 n. 20, 107 S.Ct. 1756, 95 L.Ed.2d 262 (1987) (resolving that laws in force during and just after the Civil War were not probative of the legislature's intent in 1972). More recently, Shelby County also counseled against reliance on non-contemporary evidence of discrimination in the voting rights context. ___ U.S. at ___-___, ___, 133 S.Ct. at 2618-19, 2631 (voiding Section 4 of the Voting Rights Act because "the conditions that originally justified these measures no longer characterize voting in the covered jurisdictions"). In light of these cases, the relevant "historical" evidence is relatively recent history, not long-past history.10 We recognize that history provides context and that historical discrimination (for example, in education) can have effects for many years. But, given the case law we describe above and the specific issue in this case, we conclude that the district court's heavy reliance on long-ago history was error.

We also recognize that not all "history" was "long ago" and that there were some more contemporary examples of discrimination identified by the Plaintiffs in the district court. However, even the relatively contemporary examples of discrimination identified by the district court are very limited in their probative value in connection with discerning the Texas Legislature's intent. In a state with 254 counties, we do not find the reprehensible actions of county officials in one county (Waller County) to make voting more difficult for minorities to be probative of the intent of legislators in the Texas Legislature, which consists of representatives and senators from across a geographically vast, highly populous, and very diverse state. See Miss. State Chapter, Operation Push, Inc. v. Mabus (Operation Push), 932 F.2d 400, 409-10 (5th Cir.1991) (stating that "evidence of disparate registration rates or similar registration rates in individual counties could not provide dispositive support" for the claim that plaintiffs could not participate in the political process at the state level (emphasis added)).

The only relatively contemporary evidence regarding statewide discrimination comes from a trio of redistricting cases that go in three directions, thus forming a thin basis for drawing any useful conclusions here. The first, Bush v. Vera, 517 U.S. 952, 116 S.Ct. 1941, 135 L.Ed.2d 248 (1996), found discrimination in redistricting to create more minority representation. The second found voter dilution affecting Hispanics in the redrawing of one congressional district. See League of Latin Am. Citizens v. Perry, 548 U.S. 399, 439-40, 126 S.Ct. 2594, 165 L.Ed.2d 609 (2006). Although citing discussions of the historic discrimination against Hispanics in Texas, the Court did not base its decision on a conclusion that the legislature intentionally discriminated based upon ethnicity. Id. Instead, it looked at history as a context for the disenfranchisement of voters who had grown disaffected with the Hispanic Congressman the legislature sought to protect by its redrawing of the district. Id. at 440, 126 S.Ct. 2594. The Court did not find any voter dilution as to African-Americans in the drawing of a different district. Id. at 444, 126 S.Ct. 2594. The third case, Texas v. United States, 887 F.Supp.2d 133 (D.D.C.2012), vacated and remanded on other grounds, ___ U.S. ___, 133 S.Ct. 2885, 186 L.Ed.2d 930 (2013), was a preclearance case where the burden of proof was different and which was vacated in light of Shelby County and remains unresolved as of this date. Thus, these cases do not support a finding of "relatively recent" discrimination.

The district court's heavy reliance on post-enactment speculation by opponents of SB 14 was also misplaced. Discerning the intent of a decisionmaking body is difficult and problematic. Hunter, 471 U.S. at 228, 105 S.Ct. 1916. To aid in this task, courts may evaluate "contemporary statements by members of the decisionmaking body, minutes of its meetings, or reports. In some extraordinary instances the members might be called to the stand at trial to testify concerning the purpose of the official action ...." Arlington Heights, 429 U.S. at 268, 97 S.Ct. 555. Where the court is asked to identify the intent of an entire state legislature, as opposed to a smaller body, the charge becomes proportionately more challenging. Hunter, 471 U.S. at 228, 105 S.Ct. 1916. As United States v. O'Brien explained:

391 U.S. 367, 383-84, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968).

To ascertain the Texas Legislature's purpose in passing SB 14, the district court relied to a large extent on speculation by the bill's opponents about proponents' motives (rather than evidence of their statements and actions). For instance, it credited the following: Representative Hernandez-Luna's simple assertion that two city council seats in Pasadena, Texas were made into at-large seats "in order to dilute the Hispanic vote and representation"; Representative Veasey's testimony that his appointment as vice-chair for the Select Committee on Voter Identification and Voter Fraud was only for appearances; repeated testimony that the 2011 session was imbued with anti-immigrant sentiment;11 testimony by the bill's opponents that they believed the law was passed with a discriminatory purpose; and testimony by Senator Uresti that he knew SB 14 was intended to impact minority voters.

"The Supreme Court has ... repeatedly cautioned — in the analogous context of statutory construction — against placing too much emphasis on the contemporaneous views of a bill's opponents."12 Butts v. City of New York, 779 F.2d 141, 147 (2d Cir.1985) (citing, inter alia, Ernst & Ernst v. Hochfelder, 425 U.S. 185, 204 n. 24, 96 S.Ct. 1375, 47 L.Ed.2d 668 (1976)). We too have held that such statements are entitled to "little weight." Mercantile Tex. Corp. v. Bd. of Governors of Fed. Reserve Sys., 638 F.2d 1255, 1263 (5th Cir. Unit A Feb.1981). The Second Circuit considered such speculation in Butts and held that "the speculations and accusations of ... [a] few opponents simply do not support an inference of the kind of racial animus discussed in, for example, Arlington Heights." 779 F.2d at 147. The Tenth Circuit has likewise concluded that "discriminatory intent cannot be ascertained by eliciting opinion testimony from witnesses, often out of context and accumulating those responses as substantive evidence of the motive of the [enactment]." Dowell by Dowell v. Bd. of Educ. of Okla. City Pub. Schs., Indep. Dist. No. 89, 890 F.2d 1483, 1503 (10th Cir.1989) rev'd sub nom. on other grounds, Bd. of Educ. of Okla. City Pub. Sch., Indep. Sch. Dist. No. 89 v. Dowell, 498 U.S. 237, 111 S.Ct. 630, 112 L.Ed.2d 715 (1991). We agree with our sister circuits. Conjecture by the opponents of SB 14 as to the motivations of those legislators supporting the law is not reliable evidence.13

Moreover, the district court appeared to place inappropriate reliance upon the type of postenactment testimony which courts routinely disregard as unreliable. See Barber v. Thomas, 560 U.S. 474, 485-86, 130 S.Ct. 2499, 177 L.Ed.2d 1 (2010) ("And whatever interpretive force one attaches to legislative history, the Court normally gives little weight to statements, such as those of the individual legislators, made after the bill in question has become law."); see also Edwards v. Aguillard, 482 U.S. 578, 596 n. 19, 107 S.Ct. 2573, 96 L.Ed.2d 510 (1987) ("The Court has previously found the postenactment elucidation of the meaning of a statute to be of little relevance in determining the intent of the legislature contemporaneous to the passage of the statute."). While probative in theory, even those (after-the-fact) stray statements made by a few individual legislators voting for SB 14 may not be the best indicia of the Texas Legislature's intent.14 See Operation Push, 932 F.2d at 408 (finding "isolated and ambiguous statements made by ... legislators" were not compelling evidence of that law's discriminatory purpose); Jones v. Lubbock, 727 F.2d 364, 371 n. 3 (5th Cir.1984) (refusing to "judge intent from the statements [made by] a single member" of the legislative body).

We also have concerns about undue reliance on the procedural departures enumerated in the district court's opinion as evidence of intentional discrimination. See Veasey, 71 F.Supp.3d at 645-59. While we do not reweigh evidence for the district court, we have noted that "objection[s] to typical aspects of the legislative process in developing legislation," such as increasing the number of votes a law requires for passage, may not be sufficient to demonstrate intent. Cf. Operation Push, 932 F.2d at 408-09 & n. 6. The rejection of purportedly ameliorative amendments does not itself constitute a procedural departure; rather, the court must evaluate whether opponents of the legislation were deprived of process. See Allstate Ins. Co. v. Abbott, 495 F.3d 151, 161 (5th Cir.2007) (holding that the Texas Legislature did not deviate from procedural norms sufficient to demonstrate discriminatory intent where the Legislature held well-attended committee hearings, those opposed to the legislation were allowed to testify, and legislators met with private parties harboring concerns about the proposed law). Finally, we observe that context also matters; the procedural maneuvers employed by the Texas Legislature occurred, as the district court notes, only after repeated attempts to pass voter identification bills were blocked through countervailing procedural maneuvers. See Veasey, 71 F.Supp.3d at 645-46. Given this context, the district court must carefully scrutinize whether the tactics employed by the Texas Legislature are indeed evidence of purposeful discrimination.15

While the district court's comprehensive opinion included some evidence supporting its finding of discriminatory purpose, given the degree of attention paid to the evidence discussed above, we cannot gauge whether the district court would have reached the same conclusion after correct application of the legal standard weighing the remaining evidence against the contrary evidence. This is particularly true in light of the extensive discovery of legislators' private materials that yielded no discriminatory evidence.16 We are mindful that it is not our role to reweigh the evidence for the district court. See Pullman-Standard, 456 U.S. at 291-92, 102 S.Ct. 1781 ("When an appellate court discerns that a district court has failed to make a finding because of an erroneous view of the law ... there should be remand for further proceedings to permit the trial court to make the missing findings." (emphasis added)); N. Miss. Commc'ns, Inc. v. Jones, 951 F.2d 652, 656-57 & n. 21 (5th Cir.1992) (citing Pullman-Standard, 456 U.S. at 291, 102 S.Ct. 1781) (remanding a case, for the fourth time, for factual findings under the proper standard). Thus, instead of ourselves evaluating any remaining evidence and drawing a conclusion as to discriminatory purpose, we conclude that the proper procedure is to vacate this portion of the district court's judgment (and its accompanying remedies) and remand to the district court for a reexamination of the probative evidence underlying Plaintiffs' discriminatory purpose claims weighed against the contrary evidence, in accord with the standards elucidated above.

B. Discriminatory Effect

If the district court again finds discriminatory purpose on remand, then it would not need to address effect. However, because the result could be different on remand and because the district court addressed, and the parties fully briefed, discriminatory effect, we now turn to consideration of it. Plaintiffs allege that SB 14 has a discriminatory effect in violation of Section 2 of the Voting Rights Act, which proscribes any "voting qualification or prerequisite to voting or standard, practice, or procedure ... which results in a denial or abridgement of the right of any citizen ... to vote on account of race or color." 52 U.S.C. § 10301(a). Unlike discrimination claims brought pursuant to the Fourteenth Amendment, Congress has clarified that violations of Section 2(a) can "be proved by showing discriminatory effect alone." Thornburg v. Gingles, 478 U.S. 30, 35, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986); see also 52 U.S.C. § 10301(b).

To satisfy this "results test," Plaintiffs must show not only that the challenged law imposes a burden on minorities, but that "a certain electoral law, practice, or structure interacts with social and historical conditions to cause an inequality in the opportunities enjoyed by black and white voters to elect their preferred representatives." Gingles, 478 U.S. at 47, 106 S.Ct. 2752 (emphasis added).

We now adopt the two-part framework employed by the Fourth and Sixth Circuits to evaluate Section 2 "results" claims. It has two elements:

League of Women Voters of N.C. v. North Carolina, 769 F.3d 224, 240 (4th Cir.2014) (citations and internal quotation marks omitted), cert. denied, ___ U.S. ___, 135 S.Ct. 1735, 191 L.Ed.2d 702 (2015); see also Ohio State Conf. of NAACP v. Husted, 768 F.3d 524, 554 (6th Cir.2014), vacated on other grounds by No. 14-3877, 2014 U.S.App. LEXIS 24472, at *2 (6th Cir. Oct. 1, 2014) (applying the two-part framework above); cf. Frank v. Walker, 768 F.3d 744, 754-55 (7th Cir.2014), cert. denied, ___ U.S. ___, 135 S.Ct. 1551, 191 L.Ed.2d 638 (2015).17

While courts regularly utilize statistical analyses to discern whether a law has a discriminatory impact, see e.g., Operation Push, 932 F.2d at 410-11, the Supreme Court has also endorsed factors ("the Senate Factors") enunciated by Congress to apprehend whether such an impact exists and whether it is a product of current or historical conditions of discrimination. Gingles, 478 U.S. at 44-45, 106 S.Ct. 2752. These factors include:

Id. at 36-37, 106 S.Ct. 2752 (quoting S.Rep. No. 97-417, at 28-29 (1982), reprinted in 1982 U.S.C.C.A.N. 177, 206-07). Two additional considerations are:

Id.

These factors are not exclusive, and "`there is no requirement that any particular number of factors be proved, or that a majority of them point one way or the other.'" Id. at 45, 106 S.Ct. 2752 (quoting S. Rep. at 29). While the State argues that these factors are inapposite in the "vote denial" context, we disagree.18 See Operation Push, 932 F.2d at 405-06 (affirming the district court's application of the Senate Factors in a vote denial case).

Guided by these two frameworks, we evaluate the district court's discriminatory effect finding for clear error. See id. at 410. Of course, we review legal questions de novo. Gingles, 478 U.S. at 79, 106 S.Ct. 2752.

1. Disparate Impact

The district court found that 608,470 registered voters, or 4.5% of all registered voters in Texas, lack SB 14 ID. Veasey, 71 F.Supp.3d at 659. Of those, 534,512 voters did not qualify for a disability exemption from SB 14's requirements. Id. The latter figure, which was derived by comparing the Texas Election Management System with databases containing evidence of who possesses SB 14 ID, is known as the "No-Match List."19 Id.

Plaintiffs' experts then relied on four distinct methods of analysis to determine the races of those on the No-Match List.20 Id. at 659-61. Those included: (1) ecological regression analysis, (2) a homogenous block group analysis, (3) comparing the No-Match List to the Spanish Surname Voter Registration list, and (4) reliance upon data provided by Catalist LLC, a company that compiles election data. Id. at 661. The ecological regression analysis performed by Dr. Stephen Ansolabehere, an expert in American electoral politics and statistical methods in political science, which compared the No-Match List with census data, revealed that Hispanic registered voters and Black registered voters were respectively 195% and 305% more likely than their Anglo peers to lack SB 14 ID. Id. According to Dr. Ansolabehere, this disparity is "statistically significant and highly unlikely to have arisen by chance." The block group analysis yielded similar results, and other experts arrived at similar conclusions. Id. These statistical analyses of the No-Match List were corroborated by a survey of over 2,300 eligible Texas voters, which concluded that Blacks were 1.78 times more likely than Whites, and Latinos 2.42 times more likely, to lack SB 14 ID. Id. at 662-63. Even the study performed by the State's expert, which the district court found suffered from "severe methodological oversights," found that 4% of eligible White voters lacked SB 14 ID, compared to 5.3% of eligible Black voters and 6.9% of eligible Hispanic voters. Id. at 663 & n. 239. The district court thus credited the testimony and analyses of Plaintiffs' three experts, each of which found that SB 14 disparately impacts African-American and Hispanic registered voters in Texas.21 Id. at 663.

The district court likewise concluded that SB 14 disproportionately impacted the poor. Id. at 664-65. It credited expert testimony that 21.4% of eligible voters earning less than $20,000 per year lack SB 14 ID, compared to only 2.6% of voters earning between $100,000 and $150,000 per year. Id. at 664. Those earning less than $20,000 annually were also more likely to lack the underlying documents to get an EIC. Id. Dr. Jane Henrici, an anthropologist and professorial lecturer at George Washington University, explained that:

Id.

Furthermore, the court found that the poor are less likely to avail themselves of services that require ID, such as obtaining credit and other financial services. Id. They are also less likely to own vehicles and are therefore more likely to rely on public transportation. Id. at 665, 672-73. As a result, the poor are less likely to have a driver's license and face greater obstacles in obtaining photo identification. Id. Even obtaining an EIC poses an obstacle — the district court credited evidence that hundreds of thousands of voters face round-trip travel times of 90 minutes or more to the nearest location issuing EICs.22 Id. at 672. Of eligible voters without access to a vehicle, a large percentage faced trips of three hours or more to obtain an EIC. Id.

Although the State does not dispute the underlying factual findings, it raises several purported legal errors in the district court's decision. We conclude that the district court did not reversibly err in determining that SB 14 violates Section 2 by disparately impacting minority voters.

Foremost, the State disputes the propriety of using statistical analyses to determine the racial composition of the No-Match List. Citing Bartlett v. Strickland, 556 U.S. 1, 17-18, 129 S.Ct. 1231, 173 L.Ed.2d 173 (2009), the State argues that the Supreme Court foreclosed using statistical analysis to determine the racial composition of a group of voters. That is a mischaracterization. Strickland cautions against adopting standards that require judges to make complicated, race-based predictions in redistricting cases, a concern that is not implicated here. Id. It is well within the district court's purview to assess whether minorities are disproportionately affected by a change in the law, based on statistical analyses. See e.g., Operation Push, 932 F.2d at 410-11. Using accepted statistical methodologies to estimate the racial composition of Texas voters does not require the type of race-based predictions that the Court referenced in Strickland.23 Instead, this case is more akin to Operation Push, in which this court approved using surveys and "independent statistical tests" to project the impact on minorities of newly enacted voter registration procedures. Id.

The State also relies on Strickland to argue that the canon of constitutional avoidance militates against requiring the State to ensure that voters of various races possess voter ID in equal measure. See 556 U.S. at 18, 129 S.Ct. 1231. The district court's discriminatory effect finding, if affirmed, would do no such thing; nor does Section 2 mandate the sort of remedy to which the State objects. Section 2 merely prohibits the State from imposing burdens on minority voters that would disproportionately diminish their ability to participate in the political process.24 Cf. Texas Dep't of Hous. & Cmty. Affairs v. Inclusive Communities Project, Inc. ("Inclusive Communities"), ___ U.S. ___, 135 S.Ct. 2507, 2524, 192 L.Ed.2d 514 (2015) ("Remedial orders in disparate-impact cases should concentrate on the elimination of the offending practice .... If additional measures are adopted, courts should strive to design them to eliminate racial disparities through race-neutral means. Remedial orders that impose racial targets or quotas might raise more difficult constitutional questions." (citation omitted)).

Next, the State argues that the analyses relied upon by the district court are unreliable because one source of data — the State's voter registration database — does not list the race or ethnicity of voters. The State contends that Plaintiffs' expert should have relied instead on data provided by the Department of Public Safety ("DPS"). The district court rightly rejected this argument. The DPS database did not allow registrants to identify themselves as "Hispanic" until May 2010. As the Texas Director of Elections conceded, the number of Hispanic registered voters is "exponentially higher" than the DPS records would suggest. We cannot fault the district court for refusing to rely on inaccurate data, particularly in light of the State's failure to maintain accurate data.

Finally, the State suggests that conveying the disparity in ID possession in comparative percentages is misleading. See Frank, 768 F.3d at 755 n. 3 (stating that purveying data as a comparative percentage is a "misuse" that "produces a number of little relevance to the problem"). Instead, the State believes a less deceptive method is to state that 2% of Anglo, 5.9% of Hispanic, and 8.1% of African-American registered voters lack SB 14 ID. Even assuming the State is correct, conveying the disparities in the way the State suggests does not change the analysis. The district court did not err in concluding that SB 14 disproportionately impacts Hispanic and African-American voters.25

2. The Senate Factors

We next consider the district court's finding that SB 14 "produces a discriminatory result that is actionable because [it] ... interact[s] with social and historical conditions in Texas to cause an inequality in the electoral opportunities enjoyed by African-Americans and Hispanic voters." Veasey, 71 F.Supp.3d at 698. The district court found Senate Factors 1, 2, 5, 6, 7, 8, and 9 probative. Id. at 697.

(a) Senate Factor 1: History of Official Discrimination

As part of this "searching practical evaluation of the past and present reality," Gingles, 478 U.S. at 45, 106 S.Ct. 2752 (citation and internal quotation marks omitted), the district court again found that Texas's history of discrimination in voting acted in concert with SB 14 to limit minorities' ability to participate in the political process. We repeat Shelby County's admonishment that "history did not end in 1965," 133 S.Ct. at 2628, and emphasize that contemporary examples of discrimination are more probative than historical examples. Even discounting this factor and the district court's analysis of it, however, we conclude that the other factors support its finding that SB 14 has a discriminatory effect.

(b) Senate Factor 2: Racially Polarized Voting

The district court relied primarily on the testimony of Dr. Barry Burden, a political science professor, and Mr. George Korbel, an expert on voting rights, in concluding that racially polarized voting exists throughout Texas. The court stated that "[r]acially polarized voting exists when the race or ethnicity of a voter correlates with the voter's candidate preference." Veasey, 71 F.Supp.3d at 637 (citing Gingles, 478 U.S. at 53 n. 21, 106 S.Ct. 2752). For support, the district court noted that the gap between Anglo and Latino Republican support is between 30 and 40 percentage points, the Supreme Court has previously acknowledged the existence of racially polarized voting in Texas, and that in other litigation, Texas has conceded that racially polarized voting exists in 252 of its 254 counties. The State did not contest these findings before the district court.

For the first time in its reply brief, the State argues that the district court erred by examining whether race and voting patterns exhibited a correlated, rather than causal, link. We generally do not consider arguments raised for the first time in a reply brief. See Baris v. Sulpicio Lines, 932 F.2d 1540, 1546 n. 9 (5th Cir.1991).

(c) Senate Factor 5: Effects of Past Discrimination

Next, the district court appraised "[t]he extent to which minority group members bear the effects of past discrimination in areas such as education, employment, and health, which hinder their ability to participate effectively in the political process." Gingles, 478 U.S. at 45, 106 S.Ct. 2752. The disparity in education, employment, and health outcomes between Anglos, African-Americans, and Hispanics is manifest by fact that the 29% of African-Americans and 33% of Hispanics in Texas live below the poverty line compared to 12% of Anglos. Veasey, 71 F.Supp.3d at 665. The unemployment rate for Anglos is also significantly lower. At trial, the court found that 6.1% of Anglos were unemployed compared to 8.5% of Hispanics and 12.8% of African-Americans. Id. at 666. Furthermore, 91.7% of Anglo 25-year-olds in Texas have graduated from high school, compared to 85.4% of African-Americans, and only 58.6% of Hispanics. Id. Anglos are also significantly more likely to have completed college — 33.7% of Anglos hold a bachelor's degree, compared to 19.2% of African-Americans and 11.4% of Hispanics. Id. Finally, the district court credited testimony that African-Americans and Hispanics are more likely than Anglos to report being in poor health, and to lack health insurance. Id. at 666-67.

According to the district court, "[t]hese socioeconomic disparities have hindered the ability of African-Americans and Hispanics to effectively participate in the political process. Dr. Ansolabehere testified that these minorities register and turn[ ]out for elections at rates that lag far behind Anglo voters."26 Id. at 697. This is significant because the inquiry in Section 2 cases is whether the vestiges of discrimination act in concert with the challenged law to impede minority participation in the political process. See League of United Latin American Citizens, Council No. 4434 v. Clements (LULAC), 999 F.2d 831, 866-67 (5th Cir.1993) (en banc). The district court concluded in the affirmative, and the State does not contest these underlying factual findings on appeal.

The district court credited expert testimony that tied these disparate educational, economic, and health outcomes to Texas's history of discrimination. According to Dr. Vernon Burton, a professor with an expertise in race relations, past state-sponsored employment discrimination and Texas's maintenance of a "separate but equal" education system both contributed to the unequal outcomes that presently exist. Veasey, 71 F.Supp.3d at 636. Although Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954), mandated desegregated schools in 1954, Dr. Burton testified that Texas maintained segregated schools until roughly 1970. Veasey, 71 F.Supp.3d at 634. The district court found that the disparity in educational outcomes is also due, in part, to unequal administration of discipline. For instance, African-American students are three times more likely than Anglos to be removed from school for an otherwise comparable infraction, and African-Americans are 31% more likely to face school disciplinary procedures. Id. at 666. According to Dr. Burton, students that face serious disciplinary action are less likely to graduate from high school. Id. Again, the State does not dispute the underlying data or methodologies, and as such we cannot conclude that the district court clearly erred.

(d) Factor 6: Racial Appeals in Political Campaigns

While the existence of racial appeals in political campaigns is a factor that may be indicative of a law's disparate impact, see Gingles, 478 U.S. at 40, 106 S.Ct. 2752, it is not highly probative here (and racial appeals seem to have been used by minorities and non-minorities). The district court found that such appeals still exist in Texas and cited anecdotal evidence to support its finding. See Veasey, 71 F.Supp.3d at 638-39. While we do not overturn the underlying factual finding, it is not clear how such anecdotal evidence of racial campaign appeals combines with SB 14 to deny or abridge the right to vote.

(e) Senate Factor 7 and Factor 8: Minority Public Officials and Responsiveness to Minority Needs

The extent to which minority candidates are elected to public office also contextualizes the degree to which vestiges of discrimination continue to reduce minority participation in the political process. See Gingles, 478 U.S. at 45, 106 S.Ct. 2752. The district court found that African-Americans comprise 13.3% of the population in Texas, but only 1.7% of all Texas elected officials are African-American. Veasey, 71 F.Supp.3d at 638. Similarly, Hispanics comprise 30.3% of the population but hold only 7.1% of all elected positions. Id. Within the Texas Legislature, however, both groups fare better — African-Americans hold 11.1% of seats in the Legislature while Hispanics hold 21.1% of seats. Id. Again, the State does not contest these findings. Id.

The district court also found that Texas's history of discrimination, coupled with SB 14's effect on minorities in Texas, demonstrated a lack of responsiveness to minority needs by elected officials. See Gingles, 478 U.S. at 45, 106 S.Ct. 2752. It noted that ameliorative amendments that attempted to lessen SB 14's impact on minority communities were repeatedly rejected, without explanation. See Veasey, 71 F.Supp.3d at 650-51, 658, 669, 698, 702. While this does not prove improper intent on the part of those legislators, it nonetheless supports a conclusion of lack of responsiveness.27

(f) Factor 9: Tenuousness of Policies Underlying the Law

Finally, the district court concluded that the policies underlying SB 14's passage were tenuous. While increasing voter turnout and safeguarding voter confidence are legitimate state interests, see Crawford v. Marion Cnty. Election Bd., 553 U.S. 181, 191, 128 S.Ct. 1610, 170 L.Ed.2d 574 (2008), the district court found that "the stated policies behind SB 14 are only tenuously related to its provisions," Veasey, 71 F.Supp.3d at 698. While in-person voting fraud is rare and mail-in fraud is comparatively much more common, SB 14's voter ID restrictions would only combat the former. Id. at 639-41, 653.

The district court likewise found that concerns about undocumented immigrants and non-citizens voting were misplaced. It credited testimony that undocumented immigrants are unlikely to vote as they try to avoid contact with government agents for fear of being deported. Id. at 654. At least one Representative voting for SB 14 conceded that he had no evidence to substantiate his fear of undocumented immigrants voting. Id. Additionally, the district court found that SB 14 would not prevent non-citizens from voting, since non-citizens can legally obtain a Texas driver's license or concealed handgun license, two forms of SB 14 ID. Id.

The district court also found "no credible evidence" to support assertions that voter turnout was low due to a lack of confidence in elections, that SB 14 would increase public confidence in elections, or that increased confidence would boost voter turnout. Id. at 655. Two State Senators and the Director of the Elections Division at the Texas Secretary of State's office all were unaware of anyone abstaining from voting out of concern for voter fraud, and the Director testified that implementing the provisional ballot process might undermine voter confidence. Id. The district court also credited testimony that SB 14 would decrease voter turnout. Id. at 655-56. According to a well-established formula employed by political scientists to assess individuals' likelihood of voting in an election, increasing the cost of voting decreases voter turnout — particularly among low-income individuals, as they are most cost sensitive. Id. at 656. Further, the district court dismissed the argument that increased turnout during the 2008 presidential election was demonstrative of increased voter confidence in two states that had recently passed voter ID laws. Id. at 655. Instead, it found that the increased turnout, nationwide, was due to President Obama's candidacy. Id. Finally, the court also found that public opinion polls — which found high levels of support for photo ID requirements — were not demonstrative that SB 14 itself would promote voter confidence. Id. at 656. The district court discounted the polls because they did not evaluate whether voters supported SB 14 when weighed against its attendant effect on minority voters. Id.

We note that, due to timing, a full election featuring dozens of statewide offices including Governor, federal offices including United States Senator, and numerous local offices was conducted in November 2014 while SB 14 was in effect. During oral argument, we inquired whether it would be appropriate to consider evidence of effect from this election. Both sides declined any such suggestion. Thus, there is no need to remand for consideration of any such evidence.

(g) Discriminatory Effect Conclusion

Given its findings regarding SB 14's disparate impact and the Senate Factors, the district court held that SB 14 acted in concert with current and historical conditions of discrimination to diminish African-Americans' and Hispanics' ability to participate in the political process. Id. at 695, 698. Contrary to the State's assertion, we conclude that the district court performed the "intensely local appraisal" required by Gingles. 478 U.S. at 78-79, 106 S.Ct. 2752. It clearly delineated each step of its analysis, finding that:

Veasey, 71 F.Supp.3d at 664.

The district court thoroughly evaluated the "totality of the circumstances," each finding was well-supported, and the State has failed to contest many of the underlying factual findings. Furthermore, the district court's analysis comports with the Supreme Court's recent instruction that "a disparate-impact claim that relies on a statistical disparity must fail if the plaintiff cannot point to a defendant's policy or policies causing that disparity." Inclusive Communities, 135 S.Ct. at 2523. The district court here acknowledged this principle and tethered its holding to two findings. First, the court found a stark, racial disparity between those who possess or have access to SB 14 ID and those who do not. Second, it applied the Senate Factors to assess how SB 14 worked in concert with Texas's legacy of state-sponsored discrimination to bring about this disproportionate result.

As such, we conclude that the district court did not clearly err in determining that SB 14 has a discriminatory effect on minorities' voting rights in violation of Section 2 of the Voting Rights Act. As discussed below, we remand for a consideration of the appropriate remedy in light of this finding in the event that the discriminatory purpose finding is different.

C. First and Fourteenth Amendment Burden on Right to Vote

Plaintiffs argue that SB 14 also unconstitutionally burdens their right to vote, as forbidden by the First and Fourteenth Amendments. We decline to decide this question, under the "well established principle governing the prudent exercise of this [c]ourt's jurisdiction that normally th[is c]ourt will not decide a constitutional question if there is some other ground upon which to dispose of the case." Escambia Cnty. v. McMillan, 466 U.S. 48, 51, 104 S.Ct. 1577, 80 L.Ed.2d 36 (1984). Since we affirm the district court's determination that SB 14 has a discriminatory effect under Section 2 of the Voting Rights Act, Plaintiffs will be entitled to the same relief they could access if they prevailed on these First and Fourteenth Amendment claims. Cf. Nw. Austin Mun. Util. Dist. No. One v. Holder, 557 U.S. 193, 205, 129 S.Ct. 2504, 174 L.Ed.2d 140 (2009); see also Ketchum v. Byrne, 740 F.2d 1398, 1409-10 (7th Cir.1984) ("There appears to be no difference in the practical result or in the available remedy regardless of how the resulting discrimination is characterized. We therefore shall not explicitly decide the issue of a fourteenth amendment violation...."), cert. denied sub nom. City Council of the City of Chi. v. Ketchum, 471 U.S. 1135, 105 S.Ct. 2673, 86 L.Ed.2d 692 (1985). Put another way, the rights and remedies are intertwined and, therefore, we need not decide the constitutional issue. See Crawford, 553 U.S. at 203, 128 S.Ct. 1610 (indicating that, under the facts of that case, the petitioners did not show that the proper remedy for "an unjustified burden on some voters ... would be to invalidate the entire statute," but not foreclosing this possibility under other circumstances); see also Frank v. Walker, 17 F.Supp.3d 837, 863, 879 (E.D.Wis.2014) (noting that Crawford did not prevent the district court from invalidating a photo ID requirement based on a Fourteenth Amendment claim and invalidating the entire requirement even when there existed a valid Section 2 discriminatory effect claim), rev'd, 768 F.3d 744 (7th Cir.2014) (reversing on the merits, and, in dicta, casting doubt on the remedial decision of the district court, but not foreclosing the option of invalidation of an entire statute based on a Fourteenth Amendment claim), cert. denied, ___ U.S. ___, 135 S.Ct. 1551, 191 L.Ed.2d 638 (2015); Boustani v. Blackwell, 460 F.Supp.2d 822, 827 (N.D.Ohio 2006) (in the absence of a Section 2 claim, holding that amended sections of an Ohio law requiring presentation of a certificate of naturalization unconstitutionally burdened the right to vote and permanently enjoining the statutory sections imposing this requirement); Cotham v. Garza, 905 F.Supp. 389, 400-01 (S.D.Tex. 1995) (permanently enjoining a Texas law that banned the possession of written communications while marking a ballot as an unconstitutional burden on the plaintiffs' right to vote); Pilcher v. Rains, 683 F.Supp. 1130, 1130, 1135-36 (W.D.Tex. 1988) (in the absence of a Section 2 claim, permanently enjoining a Texas statute that required signatures on unrecognized political party petitions to be accompanied by the signer's voter registration number because this unconstitutionally burdened the right to vote).

Accordingly, we need not and do not decide whether SB 14 violates the First and Fourteenth Amendments by placing an unconstitutional burden on the right to vote. See Merced v. Kasson, 577 F.3d 578, 586-87 (5th Cir.2009); Jordan v. City of Greenwood, 711 F.2d 667, 668-70 (5th Cir. 1983) ("If there is one doctrine more deeply rooted than any other in the process of constitutional adjudication, it is that we ought not to pass on questions of constitutionality... unless such adjudication is unavoidable." (quoting Spector Motor Serv., Inc. v. McLaughlin, 323 U.S. 101, 105, 65 S.Ct. 152, 89 L.Ed. 101 (1944))). We therefore VACATE the district court's determination on this issue and DISMISS Plaintiffs' First and Fourteenth Amendment claims.

D. Poll Tax28

The Veasey Plaintiffs29 originally alleged that SB 14 imposed a poll tax under the Fourteenth and Twenty-Fourth Amendments. After the passage of SB 983, the Veasey Plaintiffs filed a Rule 28(j) Letter with this court, stating that "SB14, as amended by SB983, is no longer a poll tax." The Veasey Plaintiffs nevertheless contend that "the poll tax issue is still alive" because it operated as a poll tax for nearly two years, preventing Plaintiffs and others from voting, and because it will take a "long time" for Texas voters to "learn about and acquire free birth certificates." Additionally, even without the $2 to $3 fee, the Veasey Plaintiffs argue that the process of obtaining a free birth certificate and a free EIC constitutes the kind of "burdensome alternative process" that was struck down in Harman v. Forssenius, 380 U.S. 528, 531-32, 541-42, 85 S.Ct. 1177, 14 L.Ed.2d 50 (1965).

To the extent that the Veasey Plaintiffs have not abandoned or conceded this claim,30 we conclude that SB 14, as amended by SB 983, does not impose a poll tax. Although SB 983 was passed when this case was already on appeal, we do not need to remand this issue to the district court for two reasons: (1) we conclude that even before SB 983, SB 14 did not create a facial poll tax; and (2) the issue of SB 983's impact on the poll tax issue is a pure question of law (at least as far as this facial challenge) that does not necessitate any reweighing of evidence or consideration of new evidence.

The Veasey Plaintiffs previously facially challenged SB 14 with respect to Texas voters born out of state (who are unaffected by SB 983's passage). Those voters could face fees in their state of birth to obtain documentation required for an EIC. We conclude that SB 14 does not facially impose a poll tax on those voters. Rather, SB 14 requires all Texas voters to present valid identification at the polls, exercising the State's "legitimate interest in assessing the eligibility and qualifications of voters." Gonzalez v. Arizona, 677 F.3d 383, 408-10 (9th Cir.2012) (en banc), aff'd sub nom. on other grounds, Arizona v. Inter Tribal Council of Ariz., Inc., ___ U.S. ___, 133 S.Ct. 2247, 186 L.Ed.2d 239 (2013); see also Harper v. Va. Bd. of Elections, 383 U.S. 663, 668, 86 S.Ct. 1079, 16 L.Ed.2d 169 (1966) ("But we must remember that the interest of the State, when it comes to voting, is limited to the power to fix qualifications."). The indirect cost on voters born out of state does not constitute a poll tax.31 Cf. Harman, 380 U.S. at 541, 85 S.Ct. 1177 ("Thus, in order to demonstrate the invalidity of [the challenged law], it need only be shown that it imposes a material requirement solely upon those who refuse to surrender their constitutional right to vote in federal elections without paying a poll tax." (emphasis added)).

Likewise, SB 14 did not impose a poll tax on voters before the passage of SB 983. It did not "impose[ ] a material requirement solely on those who refuse[d]" to pay a poll tax, as proscribed by the Twenty-Fourth Amendment. Harman, 380 U.S. at 541-42, 85 S.Ct. 1177. Rather, it drew from the State's power to set voter qualifications by requiring all voters to present a valid form of photo identification at the polls. See Gonzalez, 677 F.3d at 408. Under the Fourteenth Amendment as the Supreme Court interpreted it in Harper, the Court has observed that a state invidiously discriminates when it imposes a cost to vote with a justification that is "irrelevant to the voter's qualifications." Crawford, 553 U.S. at 189, 128 S.Ct. 1610. Although the questions presented to the Supreme Court in Crawford did not include whether Indiana's voter ID law imposed a poll tax, the Court observed that a statute would be invalid under Harper's Fourteenth-Amendment poll tax analysis "if the State required voters to pay a tax or a fee to obtain a new photo identification." 553 U.S. at 198, 128 S.Ct. 1610 (emphasis added). The Court implied that requiring voters to obtain photo identification and charging a fee for the required underlying documentation may not qualify as a poll tax, and we hold that SB 14's similar requirements did not operate as a poll tax. See id. at 198 & n. 17, 128 S.Ct. 1610; see also Gonzalez, 677 F.3d at 407-10.

As amended by SB 983, Texas law no longer imposes any direct fee for any of the documentation required to obtain a qualifying voter ID. In both of the seminal cases addressing what constitutes a poll tax, a state attempted to tax voters a specific amount for the privilege of voting. See, e.g., Harper, 383 U.S. 663, 86 S.Ct. 1079; Harman, 380 U.S. 528, 85 S.Ct. 1177. SB 983 has removed any specific amount the State would have required of those voters who lacked both SB 14 ID and the underlying documentation to obtain it. What remain are the requirements that such voters travel to the local registrar or county clerk's office, gather and present certain forms of documentation to receive the certified record, travel to the DPS office with that record, and present the certified record, along with two forms of supporting identification, to receive an EIC. See 37 TEX. ADMIN. CODE § 15.182(3)-(4). The Veasey Plaintiffs appear to argue in their Rule 28(j) Letter that these obligations make SB 14 unconstitutional under Harman because they "requir[e] voters to follow a burdensome alternative process to avoid paying a ... poll tax." This is somewhat in tension with the Veasey Plaintiffs' initial briefing, which claimed SB 14 was a poll tax based on the fee involved and conceded that "incidental burdens on voters are not taxes," including "[i]ncidental costs such as paying for gas to drive to the polls."

Nevertheless, to the extent the Veasey Plaintiffs now attempt to analogize SB 14 and SB 983 to the scheme in Harman, we reject that analogy. In Harman, the state of Virginia forced those who would vote in federal elections to choose between paying a poll tax and meeting a registration requirement before each election year. 380 U.S. at 531-32, 85 S.Ct. 1177. The Virginia constitution mandated that federal voters file a certificate of residence within a specific date range, beginning on October 1 of the year before the federal election at issue and ending on a date six months before the date of the federal election. Id. at 532, 85 S.Ct. 1177. On a notarized, witnessed certificate, the federal voter had to submit a current address and attest to: (1) being a resident of Virginia, at the time of submission and since the date of voter registration, and (2) an intent not to move from the city or county of residence before the next general election. Id. Those voters who chose to pay federal and state poll taxes were only required to file the certificate of residence one time; those who did not pay the federal poll tax had to file a new certificate of residence in the designated time frame before each election year. Id.

This record reveals that Plaintiffs and those who lack both SB 14 ID and underlying documentation face more difficulty than many Texas voters in obtaining SB 14 ID. Plaintiffs and others similarly situated often struggle to gather the required documentation, make travel arrangements and obtain time off from work to travel to the county clerk or local registrar, and then to the DPS, all to receive an EIC. These greater difficulties receive consideration in the Section 2 discriminatory effect analysis, but Supreme Court jurisprudence has not equated these difficulties, standing alone, to a poll tax. See, e.g., Harper, 383 U.S. at 666, 86 S.Ct. 1079. In Harman, the Court specifically noted:

380 U.S. at 538, 85 S.Ct. 1177; see also Crawford, 553 U.S. at 198-99, 128 S.Ct. 1610 (contrasting the unconstitutionality of a requirement that voters "pay a tax or a fee to obtain a new photo identification" with a requirement that voters without ID "travel to the circuit court clerk's office within 10 days [of the election] to execute the required affidavit").

The State does not offer Texas voters a choice between paying a fee and undergoing an onerous procedural process. Cf. Harman, 380 U.S. at 540-41, 85 S.Ct. 1177. All voters must make a trip to the DPS, local registrar, county clerk, or other government agency at some point to receive qualifying photo identification. Undoubtedly, those who own vehicles, have flexible work schedules, and already possess the required documentation can more easily meet these procedural requirements than some of the Plaintiffs and others who lack these resources. Again, that consideration alone does not make the photo identification requirement a poll tax. See Crawford, 553 U.S. at 198-99, 128 S.Ct. 1610; Harman, 380 U.S. at 538, 85 S.Ct. 1177. Additionally, whether the qualifying identification is a driver's license, passport, or EIC, voters need not undergo this process every election year during a specific time frame six months prior to the election, as was the case in Harman. Instead, the record indicates that an EIC remains valid for six years and must only be obtained sometime before an election.

In light of the recently-enacted SB 983, SB 14 does not impose an unconstitutional poll tax under the Fourteenth or Twenty-Fourth Amendments; nor did it impose a poll tax before SB 983's enactment. Accordingly, we VACATE the district court's judgment for the Veasey Plaintiffs on their poll tax claim and RENDER judgment in the State's favor.

E. Remedy

After finding that SB 14 was enacted with a racially discriminatory purpose, the district court fully enjoined SB 14's implementation, with the exception of several sections of the law that do not relate to photo identification. See Veasey, 71 F.Supp.3d at 707 & n. 583. That remedy is potentially broader than the one to which Plaintiffs would be entitled if, on remand, the district court only found that SB 14 has a discriminatory effect in violation of Section 2 of the Voting Rights Act. Compare Crawford, 553 U.S. at 200, 203, 128 S.Ct. 1610 (noting, in the Section 2 context, that "petitioners have not demonstrated that the proper remedy — even assuming an unjustified burden on some voters — would be to invalidate the entire statute"), with City of Richmond v. United States, 422 U.S. 358, 378, 95 S.Ct. 2296, 45 L.Ed.2d 245 (1975) (holding, in the discriminatory purpose context, that "[a]n official action ... taken for the purpose of discriminating ... on account of [ ] race has no legitimacy at all...."), and Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457, 465-66, 471, 487, 102 S.Ct. 3187, 73 L.Ed.2d 896 (1982) (affirming the permanent injunction of a statewide initiative because its provisions were "effectively drawn for racial purposes" in violation of the Fourteenth Amendment).32

We remand this case for further consideration of the discriminatory purpose finding, vacate the poll tax finding, and uphold at this point only the district court's discriminatory effect finding. Because of the uncertainty of findings on remand, we address the question of remedy assuming only a finding of discriminatory effect. We consider it prudent to provide guidance regarding what would constitute a properly-tailored remedy to address the discriminatory effects of the law.33

"When devising a remedy to a [Section] 2 violation, the district court's `first and foremost obligation ... is to correct the Section 2 violation.'" Brown, 561 F.3d at 435 (quoting Bone Shirt v. Hazeltine, 461 F.3d 1011, 1022 (8th Cir. 2006)). Yet, any remedy must be "sufficiently tailored to the circumstances giving rise to the [Section] 2 violation," id., and to the extent possible, courts should respect a legislature's policy objectives when crafting a remedy, see Perry v. Perez, ___ U.S. ___, 132 S.Ct. 934, 940-44, 181 L.Ed.2d 900 (2012) (per curiam). See also Inclusive Communities, 135 S.Ct. at 2524 ("Remedial orders in disparate-impact cases should concentrate on the elimination of the offending practice that arbitrar[ily]... operate[s] invidiously to discriminate on the basis of rac[e]." (citation and internal quotation marks omitted)). In the context of redistricting,34 the Supreme Court instructed that a legislature's policy objectives may be discerned from the challenged legislation, and those policy choices should be respected as much as possible, even when some aspect of the underlying law is unenforceable. Perry, 132 S.Ct. at 941.

When a statute contains a severability clause, courts must take special care to attempt to honor a legislature's policy choice to leave the statute intact. See Ayotte v. Planned Parenthood of N. New England, 546 U.S. 320, 331, 126 S.Ct. 961, 163 L.Ed.2d 812 (2006) (holding that lower courts should have invalidated only the unconstitutional applications of a statute, rather than the entire statute, given its severability clause). In this case, SB 14's severability clause makes clear that the Legislature intended the photo identification system to be left intact for all valid applications.35 Also clearly underlying SB 14 is the concern that a voter present proper identification that is not easily counterfeited or used by another.

Accordingly, if on remand the district court finds that SB 14 has only violated Section 2 through its discriminatory effects, it should refer to the policies underlying SB 14 in fashioning a remedy. Clearly, the Legislature wished to reduce the risk of in-person voter fraud by strengthening the forms of identification presented for voting. Simply reverting to the system in place before SB 14's passage would not fully respect these policy choices — it would allow voters to cast ballots after presenting less secure forms of identification like utility bills, bank statements, or paychecks. See TEX. ELEC.CODE § 63.001(b) (West 2010). One possibility would be to reinstate voter registration cards as documents that qualify as acceptable identification under the Texas Election Code.36 The court could also decree that, upon execution of an affidavit that a person does not have an acceptable form of photo identification, that person must be allowed to vote with their voter registration card. Cf. TEX. ELEC.CODE §§ 63.008, 63.0101 (West 2010) (allowing a person to present alternate forms of identification upon submitting an affidavit certifying they did not have their voter registration card in their possession). Such a remedy would respect the Legislature's choice to do away with more problematic forms of identification, while also eliminating SB 14's invalid applications.37 See Ayotte, 546 U.S. at 331, 126 S.Ct. 961 ("So long as they are faithful to legislative intent, then, in this case the lower courts can issue a declaratory judgment and an injunction prohibiting the statute's unconstitutional application."). However, we recognize that the district court must assess this potential solution in light of other solutions posited by the parties, including other forms of photo identification. We urge the parties to work cooperatively with the district court to provide a prompt resolution of this matter to avoid election eve uncertainties and emergencies.

IV. Conclusion

For the reasons stated above, we VACATE the district court's judgment that SB 14 was passed with a racially discriminatory purpose and REMAND for further consideration of Plaintiffs' discriminatory purpose claims, using the proper legal standards and evidence. We VACATE the district court's holding that SB 14 is a poll tax under the Fourteenth and Twenty-Fourth Amendments and RENDER judgment for the State on this issue. We need not and do not address whether SB 14 unconstitutionally burdens the right to vote under the First and Fourteenth Amendments; therefore, we VACATE the district court's judgment on that issue and DISMISS those claims. We AFFIRM the district court's finding that SB 14 violates Section 2 of the Voting Rights Act through its discriminatory effects and REMAND for consideration of the appropriate remedy.

Finally, on remand, the district court should: (1) give further consideration to its discriminatory purpose findings as specified herein; and (2) if the district court does not find that SB 14 was imposed with a discriminatory purpose, consider what remedy it should grant due to SB 14's discriminatory effect in violation of Section 2 of the Voting Rights Act, taking account of any impact of SB 983 and this opinion. We leave it to the district court in the first instance to decide whether any additional evidence may be proffered on the matters remanded.