Learn About the Law
Get help with your legal needs
Treva THOMPSON, individually and behalf of all others similarly situated, Timothy Lanier, individually and behalf of all others similarly situated, Greater Birmingham Ministries, Plaintiffs-Appellants, Darius Gamble, Pamela King, individually and behalf of all others similarly situated, Plaintiff, v. State of ALABAMA, et al., Defendants, Secretary of State for the State of Alabama, Leigh Gwathney, in her official capacity as Chairman of the Board of Pardons and Paroles, James Snipes, III, in his official capacity as Chairman of the Montgomery County Board of Registrars and on behalf of a class of all voter registrars in the State of Alabama, Defendants-Appellees.
Greater Birmingham Ministries (“GBM”), an Alabamian non-profit organization dedicated to aiding low-income individuals, and several Alabamian felons 1 (collectively “Appellants”) appeal the District Court for the Middle District of Alabama's summary judgment denying their Equal Protection Clause, U.S. Const. amend. XIV, § 1, challenge to Amendment 579 of the Alabama state constitution, their Ex Post Facto Clause, U.S. Const. art. I, § 9, cl. 3, challenge to Amendment 579's disenfranchisement provisions, and their National Voting Registration Act of 1993 (“NVRA”), 52 U.S.C. § 20501 et seq., challenge to the format of Alabama's mail voting registration form. Because we hold that (1) Amendment 579 successfully dissipated any taint from the racially discriminatory motives behind the 1901 Alabama constitution; (2) Amendment 579 does not impose punishment for purposes of the Ex Post Facto Clause; and (3) Alabama's mail voting registration form complies with the NVRA, we affirm.
As both the Supreme Court and this Court have previously explained, the 1901 Alabama state constitution was intentionally enacted to discriminate against and disenfranchise black Alabamians. See Underwood v. Hunter (Hunter I), 730 F.2d 614 (11th Cir. 1984) (explaining the history of the 1901 Alabama constitution), aff'd, Hunter v. Underwood (Hunter II), 471 U.S. 222, 105 S. Ct. 1916, 85 L.Ed.2d 222 (1985) (same). The express goal of the 1901 constitutional convention was “to establish white supremacy” in Alabama “within the limits imposed by the Federal Constitution.” Hunter I, 730 F.2d at 619 (quoting John B. Knox, President of the 1901 Convention, I Off. Proceedings of the Const. Convention of the State of Ala., May 21st, 1901, to Sept. 3rd, 1901, at 8 (1940)). To accomplish their goal of disenfranchising black Alabamians, the 1901 drafters resorted to “facially neutral tests that took advantage of differing social conditions. Property tests, literacy tests, residence requirements, the poll tax, and disqualification for conviction of certain crimes all fell into this category.” Id. (internal quotation marks omitted).
While § 182 enumerated a great many crimes resulting in disenfranchisement,2 of relevance to this case is the provision of § 182 disenfranchising individuals convicted of “any ․ crime involving moral turpitude.” In Hunter I, we held that “discriminatory intent was a motivating factor in the adoption of section 182” and that the Alabama registrars could not show that “[t]here was no evidence from which the district court could have found that section 182 would have been adopted had a permissible reason been the sole consideration” under the approach adopted by the Supreme Court in Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 270 & n.21, 97 S. Ct. 555, 566 & n.21, 50 L.Ed.2d 450 (1977). Hunter I, 730 F.2d at 620–21. Accordingly, we struck down the provisions of § 182 “that disfranchise[d] nonprison offenders.” Id. at 621. The Supreme Court unanimously affirmed our decision in Hunter II, specifying that § 182 could not “deny the franchise to persons who commit misdemeanors involving moral turpitude” under the Equal Protection Clause. 471 U.S. at 233, 105 S. Ct. at 1922–23.
However, long before the Hunter decisions, Alabama had already begun the process of repealing and replacing the disenfranchisement provisions of Article VIII of the 1901 Alabama constitution, including § 182. In 1970, Alabama convened a Constitutional Revision Commission to consider potential amendments to the Alabama constitution. As part of that process, Dr. Samuel A. Beatty, a Commission staff member, wrote a report to the Commission members suggesting, inter alia, that § 182 be rewritten to disqualify voters in “general terms” instead of “a long, scattered and redundant list of disqualifying crimes.” Dr. Beatty's proposed language (very similar to what Alabama ultimately adopted in 1996) stated that “[n]o person convicted of a felony involving moral turpitude, or having been adjudicated in this or any other state, territory, or district to be mentally incompetent, shall be qualified to vote until restoration of civil rights or removal of disability.” The Commission agreed with Dr. Beatty's proposed language and submitted it as part of their recommendations to the Alabama legislature in 1973. Nevertheless, the 1973 Alabama legislature did not approve the proposed amendments. The Commission's proposed amendments were reintroduced in 1976, and again were not approved by the legislature.
In 1979, newly elected Alabama Governor Fob James assembled a “working group” to amend the Alabama constitution headed by Michael Waters. Waters testified in a deposition that the working group began with the Commission's 1973 proposal as a starting point. According to Waters, the working group agreed with Dr. Beatty and the 1973 Commission's proposal because “by eliminating that laundry list and keeping it general, you're saying, we don't endorse what the 1901 Constitution was doing.” Further, Waters and his group did not read the “moral turpitude” language from the 1973 draft “as meaning we hereby endorse any segregationist or racial philosophy that was part of the 1901 Constitution.”
Besides Waters's working group, Alabama state legislators also discussed the working group's proposed amendments in the “Joint Interim Committee to Study New Constitution.” There, Representative Tony Harrison praised the proposed amendment to § 182 as “one of the best sections that was proposed” because “it has less language and has chopped out some of the most unnecessary language that was in the Constitution.” Representative Harrison then asked Senator Bob Harris, a member of the working group, about “the legal definition of moral turpitude.” Senator Harris responded
It means doing wrong. I don't know that there is an ironclad definition of moral turpitude. I could probably look in Webster's and there is, Tony. I am not being short about it. You know as well as I do that the Courts have wrestled with this question since we have had Court. And if you go back to the 1901 Constitution, they try to go at it maybe a little bit different. I doubt that you want to go back to that.
Representative Harrison replied, “You know that I don't want to go back to it, Senator.” Following this, the two legislators discussed whether the state constitution should disenfranchise felons at all after they have served their sentence and whether the proposed amendments would allow the state legislature to restore voting rights to felons immediately after completion of their sentence. The discussion then returned to the moral turpitude language:
SENATOR MAC PARSONS: Wouldn't it have been simpler just to have left moral turpitude out? The way I understand it, there is just one or two felonies that don't include moral turpitude. I think stealing whiskey and transporting are about the only two.
SENATOR HARRIS: I think that is all. There may be some other. Somebody around here is bound to tell me. There are a very limited number of felonies that do not involve -- that the Courts have said don't involve moral turpitude.
SENATOR PARSONS: Wouldn't it have just been simpler if you just said “felony,” then?
SENATOR HARRIS: Well, of course, that would disenfranchise some moonshiners, I guess, then. It might not be a bad idea.
Representative Martha Smith then suggested that misdemeanors involving moral turpitude should also result in disenfranchisement. Senator Harris responded:
Well, let me simply say that what we were trying to do is get away from the restraints and restrictions of the 1901 Constitution as far as we could, as safely as we could, in the simplest language that we could, invest in the Legislator as much power as we could consonant with Federal laws and Federal decisions to govern the election process and the qualification of voters.
The joint committee then moved on to discuss other topics.
The joint committee also held public hearings on the proposed amendments in February 1979. At one of these public hearings, Mary Weidler of the Civil Liberties Union of Alabama objected to the felony disenfranchisement provision because “there is no reason to continue to penalize those convicted of felonies once they have served their time.” Weidler also criticized the “moral turpitude” language as “vague and indefinite” and “unwarranted and discriminatory,” arguing that “[i]t was clear from the legislative history of the 1901 Alabama Constitution” that the disenfranchisement section was specifically adopted with the intent to “disenfranchise blacks” and a “continuation of that thinking today is clearly unacceptable.” At another public hearing, Tom Leonard of Appellant GBM argued that felon disenfranchisement punishes those who have served their sentence and is “a constitutionally-imposed disability [that] serves to mark an ex-convict with an additional badge of inferiority.” A 1979 Alabama House bill removed the phrase “moral turpitude” from the proposed amendment, instead proposing that no person convicted “of a felony” shall be qualified to vote. Like the previous proposed amendments in 1973 and 1976, this bill did not gain final passage.
In 1983, the Alabama legislature attempted to replace the 1901 constitution entirely with a new state constitution. This effort, spearheaded by Lieutenant Governor Bill Baxley, State Senator Ryan deGraffenreid, and State Representative Jack Venable, was undertaken by a specially formed constitutional revision committee comprised of members from both chambers of the Alabama legislature. The committee spent seven weeks debating and drafting a new state constitution and, relevantly, adopted the disenfranchisement amendments proposed by the 1973 and 1979 commissions. The Alabama legislature ultimately passed the proposed constitution as Act 83-68. As Senator deGraffenreid explained in a memorandum,
The proposed new constitution completely rewrites the provisions relating to voting and elections in the current constitution. The provisions of the Constitution of 1901 relating to voting and elections were specifically designed to prevent blacks from voting and also prevented women and persons under the age of 21 years from voting. These lengthy and complex provisions of the present Constitution have been held to be unconstitutional under the Constitution of the United States. The new provisions relating to voting and elections are very short and concise and conform to the requirements of the United States Constitution.
Despite the state legislature's approval, however, the proposed 1983 constitution failed because the Alabama Supreme Court held that the Alabama legislature lacked the authority to replace the state constitution. State v. Manley, 441 So. 2d 864 (Ala. 1983). Five months later, on April 10, 1984, this Court decided Hunter I, which the Supreme Court affirmed in 1985 with Hunter II.
In 1995, Representative Venable introduced House Bill 38 to repeal and replace Article VIII of the 1901 constitution. The bill incorporated the language proposed by the 1973 and 1979 commissions and passed in the 1983 constitutional replacement effort, including the provision to disenfranchise persons convicted of felonies involving moral turpitude. House Bill 38 was favorably reported out of committee, received three procedural readings as required by the Alabama Constitution, Ala. Const. art. XVIII, § 284, and then passed the Alabama House unanimously without debate. Representative Seth Hammett, who later served as Speaker of the Alabama House of Representatives from 2001 to 2010, recalled that
The article on voting was brought to the floor by Jack [Representative Venable], and he spoke on the bill. There was limited, really no debate on the bill. The amendment was viewed as non-controversial. We knew Jack was very passionate about revising the constitution, and he had worked hard on these bills. We passed it and there was no controversy.
House Bill 38 then moved to the Alabama Senate, where it was also favorably reported out of committee, received three procedural readings, and approved unanimously without debate or amendment.
The proposed amendment was then placed on the ballot of the June 4, 1996, primary election. Two days before the primary election, Representative Venable was quoted in a newspaper article as saying that the constitutional changes were “strictly housekeeping” to reflect “the voting requirements of the state today, rather than in 1901 when the constitution was written.” The newspaper article also quoted Secretary of State Jim Bennett as saying, “It's an amendment whose time has come” and that “he wished the proposed constitutional amendment could have been on the ballot many years ago. If we had passed the amendment in 1902, we could have avoided all the pain and suffering we went through in the 1950s and 60s.” No public hearings were held on the amendment. The proposed amendment was approved by 76% of the voters, including eight of the ten counties in Alabama with majority black populations, and so became Amendment 579 to the Alabama constitution. Amendment 579, now codified as Ala. Const. art. VIII, § 177, provides that
(a) Every citizen of the United States who has attained the age of eighteen years and has resided in this state and in a county thereof for the time provided by law, if registered as provided by law, shall have the right to vote in the county of his or her residence. The Legislature may prescribe reasonable and nondiscriminatory requirements as prerequisites to registration for voting. The Legislature shall, by statute, prescribe a procedure by which eligible citizens can register to vote.
(b) No person convicted of a felony involving moral turpitude, or who is mentally incompetent, shall be qualified to vote until restoration of civil and political rights or removal of disability.
Before implementing Amendment 579, Alabama was required to receive preclearance by the United States Justice Department or a three-judge panel. See 52 U.S.C. § 10304(a).3 In support of that request, Representative Venable wrote a letter to the Office of the U.S. Attorney General stating that he knew of “no effect” Amendment 579 would have on minority groups “because Alabama is already using the provisions of the proposed Article.”4 Additionally, Representative Venable noted that
The proposed Article has been a part of the last three Constitutional Revision efforts. There was numerous public hearings held during the 1973, 1979 and 1983 efforts, and I recall no opposition to this Article from any group. There were no public hearings when the Article passed the legislature in 1995, and I do not recall any opposition.
The U.S. Attorney General's Office then granted preclearance to Amendment 579 on June 24, 1996.
After Amendment 579 came into effect, Alabama voting registrars had to make determinations about which felonies involved “moral turpitude” because there was no definitive list of morally turpitudinous felonies. See Ala. Op. Atty. Gen. No. 2005-092, 2005 WL 1121853, at *2 (March 18, 2005) (Ala. A.G.) (“[T]his Office cannot provide an exhaustive list of every felony involving moral turpitude.”). In fact, at least some voting registrars refused to register any convicted felon whose civil rights had not been restored after Amendment 579 passed, regardless of whether the felony involved moral turpitude, until the Alabama Attorney General and the Alabama Supreme Court intervened. See Chapman v. Gooden, 974 So. 2d 972, 987–91 (Ala. 2007); Order, Worley v. Gooden, Case No. 1051712 (Ala. Oct. 25, 2006) (“[P]ursuant to Amendment No. 579 the voter registrars cannot deny voter registration to an individual otherwise qualified to vote simply because he or she has been convicted of some felony; denial of voter registration based on a felony conviction is appropriate only if the felony involved moral turpitude.” (emphasis in original)). In an attempt to provide some guidance to voting registrars, in 2007 the Alabama Administrative Office of the Courts (the “AOC”) circulated a non-binding list of felonies that Alabama case law, state statutes, or Attorney General's opinions had identified as involving moral turpitude. However, some Alabama registrars continued to deny voting registration to felons convicted of crimes not on the AOC's list. In 2014, Alabama Secretary of State Jim Bennett also circulated a list of crimes involving moral turpitude based off a Wikipedia entry on federal immigration law. Finally, in 2017, the Alabama legislature passed Ala. Code § 17-3-30.1 to “provide a comprehensive list of acts that constitute moral turpitude for the limited purpose of disqualifying a person from exercising his or her right to vote.” Ala. Code § 17-3-30.1(b)(2)(c).
Prior to 2019, Alabama's mail voting registration form listed as a registration requirement that eligible voters must “[n]ot have been convicted of a disqualifying felony, or if you have been convicted, you must have had your civil rights restored.” In 2018, the federal Election Assistance Commission (the “EAC”) contacted Ed Packard, Alabama's Administrator of Elections, to ask how the passage of Ala. Code § 17-3-30.1 would change Alabama's voter registration form. In response, Packard drafted new language for the registration requirements section of Alabama's voting form. The EAC approved this new language, which took effect in June 2019 and clarifies that “[t]o register in Alabama you must: ․ not have been convicted of a felony involving moral turpitude (or have had our civil and political rights restored). The list of moral turpitude felonies is available on the Secretary of State web site at: sos.alabama.gov/mtfelonies.”
Appellants filed the instant action on September 26, 2016, as a putative class action with fifteen different claims. After extensive discovery, the District Court granted summary judgment in favor of Alabama on all claims on December 3, 2020. Thompson v. Merrill, 505 F. Supp. 3d 1239 (M.D. Ala. 2020). On appeal, Appellants raise only three substantive issues: (1) whether Amendment 579 eliminated the taint of discriminatory intent behind § 182, (2) whether Amendment 579 violates the Ex Post Facto Clause of the U.S. Constitution, and (3) whether Alabama's mail voter registration form violates the NVRA.5
We review grants of summary judgment de novo. Brown v. Nexus Bus. Sols., LLC, 29 F.4th 1315, 1317 (11th Cir. 2022). Summary judgment is proper “if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Id. (quoting Fed. R. Civ. P. 56(a)). On summary judgment review, we view all evidence in “the light most favorable to the nonmoving party” and draw “all justifiable inferences in that party's favor.” Id. (internal quotation marks omitted).
The parties argue about whether the District Court below correctly drew inferences based on incontrovertible facts under the Nunez framework. See Nunez v. Superior Oil Co., 572 F.2d 1119, 1123–24 (5th Cir. 1978). While we have since approved the Nunez framework, we have also explained that our standard of review on appeal is “unaffected by any inferential conclusions reached below.” Fla. Int'l Univ. Bd. of Tr. v. Fla. Nat'l Univ., Inc., 830 F.3d 1242, 1252 (11th Cir. 2016) (quoting Useden v. Acker, 947 F.2d 1563, 1573 n.14 (11th Cir. 1991)). And the parties agree that we should review the District Court's decision in this case de novo. Accordingly, we will proceed under our usual standard of review for appeals from summary judgment.
As the appellants raise three separate issues on appeal, we will address each in turn. In Part A, we discuss the enactment of Amendment 579 and hold that it successfully dissipated any taint from the 1901 convention. In Part B, we explain that Amendment 579's felon disenfranchisement provision does not impose punishment for purposes of the Ex Post Facto Clause. In Part C, we conclude that Alabama's voter form “specifies each eligibility requirement” for voting in compliance with the NVRA.
Appellants first claim that Amendment 579's felon disenfranchisement provision violates the Equal Protection Clause because the re-enactment process did not adequately dissipate the taint of the discriminatory intent behind § 182 of the 1901 constitution. To start with, “[a] state's decision to permanently disenfranchise convicted felons does not, in itself, constitute an Equal Protection violation.” Johnson v. Gov. of Fla., 405 F.3d 1214, 1217 (11th Cir. 2005) (en banc) (citing Richardson v. Ramirez, 418 U.S. 24, 53–55, 94 S. Ct. 2655, 2670–71, 41 L.Ed.2d 551 (1974)). However, the Equal Protection Clause does prevent states from disenfranchising voters based on race, and a “facially-neutral law violates the Equal Protection Clause if adopted with the intent to discriminate against a racial group.” Id. at 1222 (citing Washington v. Davis, 426 U.S. 229, 239, 96 S. Ct. 2040, 2047, 48 L.Ed.2d 597 (1976)). Determining whether a facially neutral law violates the Equal Protection Clause involves a two-step analysis. First, we “examine whether racial discrimination was a substantial or motivating factor in the state's decision to deny the right to vote to felons.” Id. at 1223. If the plaintiffs succeed in making this showing, “we then ask whether the state can show that the provision would have been enacted in the absence of any racially discriminatory motive.” Id.
However, here Appellants do not contend that Amendment 579 was enacted with discriminatory intent in 1996.6 Instead, Appellants argue that Amendment 579 failed to eliminate the discriminatory intent behind § 182 by re-enacting the “moral turpitude” language of § 182. To determine “whether a subsequent legislative re-enactment can eliminate the taint from a law that was originally enacted with discriminatory intent,” we consider whether the law was re-enacted “through a deliberative process” while paying special attention to whether the re-enactment resulted in any substantive changes. Johnson, 405 F.3d at 1223–24 (citing Cotton v. Fordice, 157 F.3d 388, 391 (5th Cir. 1998)).
In Johnson, this Court sitting en banc considered whether a felon disenfranchisement provision in the 1968 Florida constitution successfully dissipated the assumed racial motivations behind the criminal disenfranchisement provisions of Florida's 1868 constitution.7 Id. at 1220–22. Because Florida had followed its normal deliberative procedures and “narrowed the class of disenfranchised individuals to those convicted of felonies” when enacting its 1968 provision, we found that “Florida's 1968 re-enactment eliminated any taint from the allegedly discriminatory 1868 provision.” Id. at 1223–24.
Likewise, Amendment 579 was also passed through a deliberative process. Alabama considered reforms to § 182 as part of three different constitutional reform efforts in 1973, 1979, and 1983 before Amendment 579 finally passed the Alabama legislature unanimously in 1995 and was ratified by 76% of the Alabama population in 1996. Further, Amendment 579 also resulted in substantive changes to Alabama law; while § 182 disenfranchised all felons, Amendment 579 expanded the franchise by only disenfranchising persons convicted of felonies involving moral turpitude. Accordingly, Alabama has successfully eliminated any taint from the racially discriminatory motives behind § 182 under the test we set forth in Johnson.
Appellants argue that because Amendment 579 was described as “strictly housekeeping,” it lacked the legislative intent necessary to cleanse the discriminatory motivations behind § 182. However, in Johnson we rejected the proposition that a state legislature must demonstrate an intent to remove the discriminatory intent of previous provisions when re-enacting a law. 405 F.3d at 1224–25. The Johnson plaintiffs argued that “Florida must affirmatively prove that racial discrimination was not a substantial or motivating factor behind the disenfranchisement law in 1968” by “demonstrat[ing] that it acknowledged that racial discrimination tainted the 1868 provision, and yet it knowingly reenacted the disenfranchisement provision for non-discriminatory reasons in 1968.” Id. We rejected that argument because “[t]he result would be to reverse the presumption that a State's laws are constitutional, and plunge federal courts into far-reaching expeditions regarding the sins of the past in order to question the laws of today.” Id. at 1225 n.21. The question is whether the re-enactment was done through a deliberative process and without discriminatory intent, not whether the legislature intended the re-enactment to eliminate the earlier provision's discriminatory intent. Appellants point to no evidence on appeal that Amendment 579 was enacted with discriminatory intent. Id. at 1225.
Appellants also argue that the amendment process was not sufficiently deliberate because there was no debate when the amendment passed the Alabama legislature in 1995 and because the prior reform efforts did not, in their view, adequately consider the history and potential impacts of disenfranchising individuals convicted of felonies involving moral turpitude. Appellants urge us to review the “nature of the deliberations” when deciding whether Alabama re-enacted its felony disenfranchisement provision through a deliberative process. However, Appellants’ argument misstates the “deliberative process” requirement for re-enactment. When evaluating Florida's deliberative process in Johnson, we looked only to see whether Florida had passed its 1968 provisions in accordance with its normal deliberative procedures for amending the state constitution:
The provision first was considered by the Suffrage and Elections Committee. The Committee sent its final proposal to the [Constitutional Revision Commission]. The CRC reviewed the changes to the Constitution and sent a draft to the legislature, which approved the new Constitution. Finally, the voters approved the new Constitution. Thus, as in Cotton v. Fordice, Florida's 1968 re-enactment eliminated any taint from the allegedly discriminatory 1868 provision.
405 F.3d at 1224. Cotton v. Fordice, the Fifth Circuit case Johnson relied on, performed a similar analysis:
The [Fifth Circuit] emphasized the deliberative process through which the provision had twice been amended: First, both houses of the legislature had to pass the amendment by a two-thirds vote; then the Mississippi Secretary of State had to publish the full text of the provision at least two weeks before the popular election; finally, a majority of the voters had to approve the full text of the provision. Thus, the Fifth Circuit held that “because Mississippi's procedure resulted both in 1950 and in 1968 in a reenactment of the provision, each amendment superseded the previous provision and removed the discriminatory taint associated with the original version.”
Id. (citing Cotton, 157 F.3d at 391) (internal citations and alterations omitted).
Alabama adopted Amendment 579 through its normal deliberative process for amending the state constitution, which required three readings of the amendment in each chamber of the state legislature, approval by three-fifths of each chamber of the state legislature, publication to the public, and then ratification by the Alabama electorate. Ala. Const. art. XVIII, § 284. Further, Alabama even had to receive preclearance from the federal government before Amendment 579 could come into effect. See 52 U.S.C. § 10304(a). As the deliberate process by which Amendment 579 was enacted was like the deliberate process used in Johnson and Cotton, it suffices under our re-enactment test. Appellants’ proposed standard, by contrast, would “reverse the presumption that a State's laws are constitutional” by requiring the Alabama legislature to show, to this Court's arbitrary satisfaction, that it sufficiently debated the moral turpitude standard. See Johnson, 405 F.3d at 1225 n.21. The Equal Protection Clause only permits federal courts to review state legislation for discriminatory intent or purpose. See Arlington Heights, 429 U.S. at 265, 97 S. Ct. at 563 (“Proof of racially discriminatory intent or purpose is required to show a violation of the Equal Protection Clause.”). Even if we agreed with Appellants that Alabama did not sufficiently deliberate over the moral turpitude standard, the Equal Protection Clause would not permit us to overturn a validly enacted, nondiscriminatory state law. Accordingly, we reject Appellants’ invitation to review the extent the Alabama legislature debated the “moral turpitude” language of Amendment 579.
Finally, Appellants purport to make a separate argument that Alabama's re-enactment of the moral turpitude standard “perpetuated” the “racially discriminatory substance” of § 182. They argue that there is “substantial evidence that the moral turpitude standard itself gave substance to the 1901 framers’ discriminatory intent” and that “[e]liminating redundancies and restating the provision in ‘general terms’ did not change the underlying discriminatory purpose of the moral turpitude standard.” This argument fails for two reasons.
First, the moral turpitude standard is not inherently discriminatory. Both federal and state statutes use the standard in other contexts. See 8 U.S.C. § 1227(a)(2)(A)(i) (permitting deportation of any alien “convicted of a crime involving moral turpitude”); Ala. Code § 5-17-55(c)(1) (providing for removal of members of financial boards “convicted of a felony or any other crime involving moral turpitude”); Ala. Code § 34-8A-16(a)(1) (permitting the revocation of counselor licenses for individuals convicted “of a felony or any offense involving moral turpitude”). While Alabama once used the moral turpitude standard as part of a racially discriminatory disenfranchisement scheme, it is not forever barred from disenfranchising individuals convicted of felonies involving moral turpitude.
Second, this argument is merely a restatement of Appellants’ argument that the Alabama legislature needed to affirmatively intend to eliminate the discriminatory intent behind § 182 when enacting Amendment 579. Under Johnson, it is sufficient that Alabama re-enacted its disenfranchisement provision through a deliberative process and with a substantial change. Even if the Alabama legislature and electorate only enacted Amendment 579 for “strictly housekeeping” purposes like Appellants contend, that is sufficient to eliminate the discriminatory taint from § 182 of the 1901 constitution. Because we hold that Amendment 579 successfully dissipated the racially discriminatory taint from § 182 and because Appellants do not argue the Alabama legislature had discriminatory intent when enacting Amendment 579, we need not reach the second step of the Equal Protection Clause analysis and determine whether Alabama would have enacted Amendment 579 without discriminatory intent.
The Ex Post Facto Clause prohibits “retroactive punishment.” Smith v. Doe, 538 U.S. 84, 92, 123 S. Ct. 1140, 1146, 155 L.Ed.2d 164 (2003). Accordingly, for a state law to violate the Ex Post Facto Clause, it must either have been intended by the state legislature to “impose punishment” or it must be “so punitive either in purpose or effect as to negate the State's intention to deem it civil.” Id. at 92, 123 S. Ct. at 1147 (quotation and alteration omitted). Appellants argue that the lack of a definitive list of felonies involving moral turpitude before 2017 violated the Ex Post Facto Clause by failing to give any Alabamian felon fair warning of whether his or her crime involved moral turpitude. Appellants further argue that retroactive application of Ala. Code § 17-3-30.1(c) to felons convicted before 2017 is “itself an ex post facto violation” because it changed what qualified as a felony involving moral turpitude under Amendment 579. However, Appellants do not argue on appeal that Alabama intended Amendment 579's felon disenfranchisement provision to impose punishment or that felon disenfranchisement is so punitive as to override the intent of the Alabama legislature. Instead, Appellants urge us to affirm the District Court's determination that binding Eleventh Circuit precedent held that disenfranchisement is inherently punishment. Thompson, 505 F. Supp. 3d at 1262–63. We begin by analyzing whether we are bound by our prior panel precedent rule to hold that felon disenfranchisement is punishment for purposes of the Ex Post Facto Clause. As we conclude we are not, we then analyze whether Amendment 579's disenfranchisement provision constituted punishment under the Supreme Court's two-prong test.
“The prior-panel-precedent rule requires subsequent panels of the court to follow the precedent of the first panel to address the relevant issue, ‘unless and until the first panel's holding is overruled by the Court sitting en banc or by the Supreme Court.’ ” Scott v. United States, 890 F.3d 1239, 1257 (11th Cir. 2018) (quoting Smith v. GTE Corp., 236 F.3d 1292, 1300 n.8 (11th Cir. 2001)). Later panels “must faithfully follow the first panel's ruling” even when “convinced the earlier panel is wrong.” Id. (quoting United States v. Steele, 147 F.3d 1316, 1317–18 (11th Cir. 1998) (en banc) (alteration omitted)). Although we are not bound by “mere dictum,” “we must follow the reasoning behind a prior holding if we cannot distinguish the facts or law of the case under consideration.” Id. (citations omitted). Appellants point to three Eleventh Circuit decisions they contend held that disenfranchisement is punishment for purposes of the Ex Post Facto Clause: Johnson v. Gov. of Fla., 405 F.3d at 1218 n.5, 1228, Jones v. Gov. of Fla. (Jones I), 950 F.3d 795 (11th Cir. 2020), and Jones v. Gov. of Fla (Jones II), 975 F.3d 1016 (11th Cir. 2020) (en banc). We examine each case in turn.
In Johnson, this Court considered whether a section of the Voting Rights Act of 1965, 52 U.S.C. § 10301, “applie[d] to Florida's felon disenfranchisement provision.” 405 F.3d at 1227. Congress enacted § 10301 “for the remedial purpose of eliminating racially discriminatory voting practices” and, to achieve that purpose, allowed plaintiffs to challenge state voting regulations based on racially disparate impacts “without proving discriminatory intent.” Id. We held that § 10301 did not apply to felon disenfranchisement provisions under the doctrine of constitutional avoidance after examining the legislative history of § 10301. Id. at 1227–34. As part of our discussion about the history and constitutionality of felon disenfranchisement provisions, we stated that “[f]elon disenfranchisement laws are unlike other voting qualifications. These laws are deeply rooted in this Nation's history and are a punitive device stemming from criminal law.” Id. at 1228 (citing Richardson, 418 U.S. at 48–52, 94 S. Ct. at 2668–70). We also noted that
Indeed, throughout history, criminal disenfranchisement provisions have existed as a punitive device. See Harvard Law Review Association, One Person, No Vote: The Laws of Felon Disenfranchisement, 115 Harv. L. Rev. 1939, 1939–42 (2002). When the Fourteenth Amendment was ratified, twenty-nine of thirty-six states had some form of criminal disenfranchisement law. See Richardson, 418 U.S. at 48, 94 S. Ct. [at 2668]. Today, forty-eight states have some form of criminal disenfranchisement provision. Although Florida's felon disenfranchisement law may be among the most restrictive, Florida hardly stands alone in its long-standing use of these laws.
Id. at 1218 n.5.
Our two references to felon disenfranchisement as historically “a punitive device” in Johnson did not constitute a holding that all felon disenfranchisement provisions are punishment for purposes of the Ex Post Facto Clause. Johnson did not concern the Ex Post Facto Clause or any other constitutional or statutory provision where disenfranchisement being “punishment” is analytically relevant. If it had, we would have performed the two-prong analysis set forth by the Supreme Court for determining whether a law is penal or nonpenal. See Doe, 538 U.S. at 92, 123 S. Ct. at 1147; see also Trop v. Dulles, 356 U.S. 86, 96–97, 78 S. Ct. 590, 2 L.Ed.2d 630 (1958) (plurality opinion) (explaining that a felon disenfranchisement provision can be penal or nonpenal).
Instead, Johnson concerned whether § 10301 of the Voting Rights Act applied to felon disenfranchisement provisions, a separate statutory analysis that turned on felon disenfranchisement's unique constitutional status and the legislative history of the Voting Rights Act. 405 F.3d at 1227–34. Our two off-hand references to felon disenfranchisement as historically a “punitive device” were thus non-binding dicta. See United States v. Gillis, 938 F.3d 1181, 1198 (11th Cir. 2019) (“[D]icta is a statement that neither constitutes the holding of a case, nor arises from a part of the opinion that is necessary to the holding of the case.”) (internal quotation marks omitted).
By contrast, we did outright hold in Jones I that “[d]isenfranchisement is punishment.” 950 F.3d at 819. Jones I involved an appeal from a preliminary injunction requiring Florida to restore the voting rights of felons who had completed their sentence and parole but who could not pay the fines and costs imposed as part of their sentence. See Jones v. DeSantis, 410 F. Supp. 3d 1284 (N.D. Fla. 2019); see also Fla. Const. art. VI, § 4. The Jones I panel held that Florida's requirement that indigent felons pay all fines and fees before being re-enfranchised constituted a wealth classification subject to heightened scrutiny under the Supreme Court's decisions in Griffin v. Illinois, 351 U.S. 12, 76 S. Ct. 585, 100 L.Ed. 891 (1956) and Bearden v. Georgia, 461 U.S. 660, 103 S. Ct. 2064, 76 L.Ed.2d 221 (1983). Jones I, 950 F.3d at 817–25. As part of its holding,8 the Jones I panel concluded that “[d]isenfranchisement is punishment.” Id. at 819. The Jones I panel did not reach this conclusion after performing the Supreme Court's two-prong analysis for determining whether a law is penal or nonpenal. Id; see Doe, 538 U.S. at 92, 123 S. Ct. at 1147. Instead, the Jones I panel relied on the dicta in Johnson, an off-hand reference to disenfranchisement as “not an unusual punishment” in a separate opinion by Justice Scalia,9 a vacated Second Circuit panel opinion,10 and two law review articles 11 for its conclusion that disenfranchisement is punishment. Id. at 819.
However, Jones I is no longer good law. After the Jones I panel affirmed the district court's preliminary injunction, the district court issued a permanent injunction. Jones II, 975 F.3d at 1027–28. Florida appealed and petitioned this Court for initial hearing en banc, which we granted. Id. at 1028. The en banc Court then explicitly overruled Jones I and held that the “Bearden and Griffin lines of precedent are limited to the contexts in which they arose,” i.e., “poverty-based imprisonment.” Id. at 1032–33. Because disenfranchisement is not poverty-based imprisonment, Bearden and Griffin were inapplicable and only “rational basis review applie[d].” Id. at 1033. The en banc Court did note that “even if Bearden applied beyond poverty-based imprisonment,” Florida's re-enfranchisement scheme was distinguishable from Bearden because Florida did not impose “additional punishment” on convicted felons as “Florida automatically disenfranchises all felons upon conviction, and the challenged laws only lift that punishment for felons who have completed all terms of their sentences.” Id. at 1032 (emphasis in original). But as with Johnson, Jones II’s off-hand reference to disenfranchisement as punishment, made without analysis or citation, was neither a holding of the case nor a necessary component of the case's holding. Accordingly, Jones II’s reference to disenfranchisement as punishment was non-binding dicta. Gillis, 938 F.3d at 1198.12 As both Johnson and Jones II’s references to disenfranchisement as punishment are dicta and Jones I has been overruled by this Court sitting en banc, we will proceed to analyze de novo whether Amendment 579's felon disenfranchisement provision constitutes punishment for purposes of the Ex Post Facto Clause.
Disenfranchisement can be penal or nonpenal. As a plurality of the Supreme Court explained in Trop v. Dulles,
The [Supreme] Court has recognized that any statute decreeing some adversity as a consequence of certain conduct may have both a penal and a nonpenal effect. The controlling nature of such statutes normally depends on the evident purpose of the legislature. The point may be illustrated by the situation of an ordinary felon. A person who commits a bank robbery, for instance, loses his right to liberty and often his right to vote. If, in the exercise of the power to protect banks, both sanctions were imposed for the purpose of punishing bank robbers, the statutes authorizing both disabilities would be penal. But because the purpose of the latter statute is to designate a reasonable ground of eligibility for voting, this law is sustained as a nonpenal exercise of the power to regulate the franchise.
356 U.S. at 96–97, 78 S. Ct. at 596. Accordingly, courts must determine the legislative intent behind the felon disenfranchisement statute or constitutional provision under consideration before holding that it is penal or nonpenal for constitutional purposes. Id.
Alabama urges us to hold that all felon disenfranchisement provisions are always nonpenal based on the decisions of three of our sister circuits. See Johnson v. Bredesen, 624 F.3d 742, 753 (6th Cir. 2010) (“Moreover, in Trop v. Dulles, the Supreme Court expressly stated that felon disenfranchisement laws serve a regulatory, non-penal purpose. Accordingly, as a matter of federal law, disenfranchisement statutes do not violate the Ex Post Facto Clause of the U.S. Constitution.” (citation omitted)); Simmons v. Galvin, 575 F.3d 24, 43 (1st Cir. 2009) (“The Supreme Court has stated that felon disenfranchisement provisions are considered regulatory rather than punitive.” (citing Trop, 356 U.S. at 96–97, 78 S. Ct. at 596)); Green v. Bd. of Elections of City of N.Y., 380 F.2d 445, 450 (2d Cir. 1967) (“Depriving convicted felons of the franchise is not a punishment but rather is a ‘nonpenal exercise of the power to regulate the franchise.’ ” (citing Trop, 356 U.S. at 97, 78 S. Ct. at 596)).
But those three decisions all reached the conclusion that felon disenfranchisement is always nonpenal based on a misreading of Trop. Trop used felon disenfranchisement as the quintessential example of a “consequence” that “may have both a penal and a nonpenal effect.” 356 U.S. at 96, 78 S. Ct. at 596. “The controlling nature” of these provisions “normally depends on the evident purpose of the legislature.” Id. Thus, in Trop’s bank robbery example, if disenfranchisement was “imposed for the purpose of punishing bank robbers,” the disenfranchising provision “would be penal.” Id. But if the purpose of the disenfranchising provision “is to designate a reasonable ground of eligibility for voting,” then the provision is “a nonpenal exercise of the power to regulate the franchise.” Id. at 96–97, 78 S. Ct. at 596. Consequently, we reject Alabama's argument that disenfranchisement is always nonpenal.
The first step in determining whether a statute or constitutional provision imposes retroactive punishment for purposes of the Ex Post Facto Clause is to “ascertain whether the legislature meant the [provision] to establish ‘civil’ proceedings.” Doe, 538 U.S. at 92, 123 S. Ct. at 1146–47 (quoting Kansas v. Hendricks, 521 U.S. 346, 361, 117 S. Ct. 2072, 2082, 138 L.Ed.2d 501 (1997)). And determining “[w]hether a statutory [or constitutional] scheme is civil or criminal ‘is first of all a question of statutory construction.’ ” Id. at 92, 123 S. Ct. at 1147 (quoting Hendricks, 521 U.S. at 361, 117 S. Ct. at 2081). Thus, we begin by considering Amendment 579's “text and its structure” to determine whether the Alabama legislature “indicated either expressly or impliedly a preference for one label or the other.” Id. at 92–93, 123 S. Ct. at 1147 (quoting Hudson v. United States, 522 U.S. 93, 99, 118 S. Ct. 488, 493, 139 L.Ed.2d 450 (1997)).
Alabama's felony disenfranchisement provision (as amended by Amendment 579) is located in Article VIII of the Alabama constitution. Ala. Const. art. VIII, § 177(b). This article, entitled “Suffrage and Elections,” sets forth Alabama's requirements for voter eligibility and empowers the Alabama legislature to pass legislation regulating voter registration and election administration. Ala. Const. art. VIII, § 177. An individual is eligible to vote in Alabama if he or she is (1) a U.S. citizen, (2) “who has attained the age of eighteen years,” (3) “has resided in [Alabama] and in a county thereof for the time provided by law,” (4) has registered to vote “as provided by law,” (5) is not “mentally incompetent,” and (6) has not been “convicted of a felony involving moral turpitude” without having his or her “civil and political rights” restored. Id. While Article VIII is not expressly labeled as “civil,” it is clear from plain text of the article that the Alabama legislature intended Article VIII to set forth a civil scheme for regulating the franchise. There is no indication from the text that Alabama intended to criminally punish non-U.S. citizens, non-Alabama residents, non-registered voters, minors, or the mentally incompetent by denying them the franchise in Article VIII. Nor could Alabama. Further, Article VIII disenfranchises individuals convicted of felonies involving moral turpitude by other sovereigns, see Ala. Code § 17-3-30.1(c)(48), despite Alabama lacking the authority to punish felons convicted by other sovereigns. By placing Alabama's felon disenfranchisement provision amongst its other voter eligibility provisions and in the article concerned with voter eligibility requirements and election administration, Alabama implicitly indicated a preference that its felon disenfranchisement provision be considered civil instead of criminal.
Besides the text and structure of the law, “[o]ther formal attributes of a legislative enactment, such as the manner of its codification or the enforcement procedures it establishes, are probative of the legislature's intent.” Doe, 538 U.S. at 94, 123 S. Ct. at 1148. The felon disenfranchisement provision of Amendment 579 is codified as Article VIII, “Suffrage and Elections,” of the Alabama constitution, which Amendment 579 repealed and replaced. Ala. Const. amend. DLXXIX. Codifying Alabama's felon disenfranchisement provision within an article of the Alabama constitution likewise indicates a civil, regulatory intent by the legislature, as the vast majority of the Alabama constitution—and all provisions of Article VIII—sets forth civil regulations.
Further, while Article VIII does not set forth any enforcement provisions, the voter registration provisions of Article VIII are enforced by Chapter 3, “Voter Registration,” of Title 17, “Elections,” of the Alabama Code. See Ala. Code § 17-3-1 et seq. Chapter 3 sets forth a civil, regulatory scheme whereby registrar boards evaluate registration applications by potential voters to determine their eligibility without any discussion of criminal liability. Id. Of particular note is Ala. Code § 17-3-30.1, the Alabama statute defining felonies involving moral turpitude. Section 17-3-30.1, entitled the “Felony Voter Disqualification Act,” repeatedly describes conviction of a felony involving moral turpitude as “disqualifying,” not as punishment, and states that one of the purposes of the Act is “[t]o ensure that no one is wrongly excluded from the electoral franchise.” See Ala. Code § 17-3-30.1(b),(c),(d),(e). While § 17-3-30.1 passed 21 years after Amendment 579 (unlike other provisions of Chapter 3), Alabama's enforcement mechanisms clearly treat felon disenfranchisement as “a nonpenal exercise of the power to regulate the franchise.” Trop, 356 U.S. at 97, 78 S. Ct. at 596.
Considering the text and structure of Article VIII, its codification in the Alabama constitution, and its enforcement mechanisms, we hold that the Alabama legislature intended Amendment 579's felon disenfranchisement provision to be a nonpenal regulation of the franchise. This conclusion, however, does not end our analysis; “we must further examine whether the statutory scheme is so punitive either in purpose or effect as to negate the State's intention to deem it ‘civil.’ ” Doe, 538 U.S. at 92, 123 S. Ct. at 1147 (quotation, quotation marks, and alterations omitted). “Because we ordinarily defer to the legislature's stated intent, only the clearest proof will suffice to override legislative intent and transform what has been denominated a civil remedy into a criminal penalty.” Id. (internal quotations and citations omitted). This analysis is conducted under the “useful framework” of the seven non-exhaustive and non-dispositive factors set forth in Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168–69, 83 S. Ct. 554, 567–68, 9 L.Ed.2d 644 (1963). Doe, 538 U.S. at 97, 123 S. Ct. at 1149. Those factors are:
 Whether the sanction involves an affirmative disability or restraint,  whether it has historically been regarded as a punishment,  whether it comes into play only on a finding of scienter,  whether its operation will promote the traditional aims of punishment—retribution and deterrence,  whether the behavior to which it applies is already a crime,  whether an alternative purpose to which it may rationally be connected is assignable for it, and  whether it appears excessive in relation to the alternative purpose assigned.
Mendoza-Martinez, 372 U.S. at 168–69, 83 S. Ct. at 567–68 (numbering added and citations omitted). Of these seven factors, the most important are whether the sanction “has been regarded in our history and traditions as a punishment; imposes an affirmative disability or restraint; promotes the traditional aims of punishment; has a rational connection to a nonpunitive purpose; or is excessive with respect to this purpose.” Doe, 538 U.S. at 97, 123 S. Ct. at 1149. We examine each factor in turn.
First, disenfranchisement is not an “ ‘affirmative disability or restraint’ as that term is normally understood.” Hudson, 522 U.S. at 104, 118 S. Ct. at 496. In Hudson, the Supreme Court considered whether occupational disbarment for violations of certain federal banking statutes was punishment for purposes of the Double Jeopardy Clause. Id. at 95–96, 118 S. Ct. at 491. The Court concluded that occupational disbarment did not impose an “affirmative disability or restraint” as disbarment is “certainly nothing approaching the ‘infamous punishment’ of imprisonment.” Id. at 104, 118 S. Ct. at 496 (quoting Flemming v. Nestor, 363 U.S. 603, 617, 80 S. Ct. 1367, 1376, 4 L.Ed.2d 1435 (1960)). While the Hudson Court did not consider this factor further after concluding it did not impose imprisonment, we observe that felon disenfranchisement and occupational disbarment are similar in many ways. See Simmons, 575 F.3d at 44–45 (also comparing occupational disbarment to felon disenfranchisement under this factor). Both remove the civil rights of individuals due to their criminal behavior as part of the State's regulatory power. And Hudson ultimately held that occupational disbarment is nonpunitive. 522 U.S. at 105, 118 S. Ct. at 496.
The second factor, whether felon disenfranchisement has been historically regarded as punishment, is neutral. As the Supreme Court explained in Trop, felon disenfranchisement has “both a penal and a nonpenal effect,” and the “controlling nature” is the “evident purpose of the legislature.” 356 U.S. at 96, 78 S. Ct. at 596. And there is evidence that American courts and legislatures have considered felon disenfranchisement provisions as both penal and nonpenal. Compare Washington v. State, 75 Ala. 582, 585 (1884) (“It is quite common also to deny the right of suffrage, in the various American States, to such as have been convicted of infamous crimes. The manifest purpose is to preserve the purity of the ballot box ․ The presumption is, that one rendered infamous by conviction of felony, or other base offense indicative of great moral turpitude, is unfit to exercise the privilege of suffrage, or to hold office, upon terms of equality with freemen who are clothed by the State with the toga of political citizenship. ․The exclusion must for this reason be adjudged a mere disqualification, imposed for protection, and not for punishment--withholding an honorable privilege, and not denying a personal right or attribute of personal liberty.”), and Anderson v. Baker, 23 Md. 531, 626 (1865) (observing that the felon disenfranchisement provision of the 1864 Maryland constitution “is dissociated from any reference to penalty, and made the consequence of conviction, in the same connection with lunacy or persons non compos,” for purposes of the Ex Post Facto Clause), and Simmons, 575 F.3d at 45 (“[F]elon disenfranchisement has historically not been regarded as punitive in the United States, as the Supreme Court indicated in Trop v. Dulles. Indeed, in holding that felon disenfranchisement has ‘affirmative sanction’ in § 2 of the Fourteenth Amendment of the U.S. Constitution, Richardson, 418 U.S. at 54, 94 S. Ct. [at 2670–71], the Supreme Court noted the historical prevalence of state felon disenfranchisement laws and never characterized even California's broad disqualification of former felons as punitive. Id. at 55, 94 S. Ct. [at 2671].), with Richardson, 418 U.S. at 51–52, 94 S. Ct. at 2669–70 (explaining that Congress readmitted states to the Union after the Civil War on the condition that their state constitutions “never be so amended or changed as to deprive any citizen or class of citizens of the United States of the right to vote who are entitled to vote by the constitution herein recognized, except as a punishment for such crimes as are now felonies at common law” (citations omitted)), and Johnson, 405 F.3d at 1218 n.5, 1228 (noting in dicta that felon disenfranchisement has historically functioned as a “punitive device”).
The third and fifth factors—whether the sanction requires a finding of scienter and whether the behavior to which the sanction applies is already criminal—both weigh in favor of felon disenfranchisement being nonpunitive. There is no scienter requirement for felon disenfranchisement; it is sufficient that the person be convicted of a disqualifying felony. Likewise, felon disenfranchisement only sanctions behavior that is already criminal. That felon disenfranchisement laws are “tied to criminal activity ․ is insufficient to render the [laws] punitive.” United States v. Ursery, 518 U.S. 267, 292, 116 S. Ct. 2135, 2149, 135 L.Ed.2d 549 (1996); see also Simmons, 575 F.3d at 45 (also finding the third and fifth factors weigh in favor of felon disenfranchisement being nonpunitive).
Further, felon disenfranchisement does not “promote the traditional aims of punishment—retribution and deterrence” under the fourth factor. It is very unlikely that an individual considering whether to commit a felony would be willing to risk imprisonment but not disenfranchisement. And even if there were a deterrent effect, “[t]o hold that the mere presence of a deterrent purposes renders such sanctions ‘criminal’ ․ would severely undermine the Government's ability to engage in effective regulation.” Doe, 538 U.S. at 102, 123 S. Ct. at 1152 (quoting Hudson, 522 U.S. at 105, 118 S. Ct. at 496). Moreover, the text of Amendment 579 reveals no retributive intent, and it is clear from the legislative history of the amendment that its “strictly housekeeping” purpose was to update Alabama's voting requirements for the modern day.
Under the sixth factor, Alabama has a clear “alternative purpose” for felon disenfranchisement besides punishment. “The States have long been held to have broad powers to determine the conditions under which the right of suffrage may be exercised, absent of course the discrimination which the Constitution condemns.” Lassiter v. Northampton Cnty. Bd. of Elections, 360 U.S. 45, 50, 79 S. Ct. 985, 989, 3 L.Ed.2d 1072 (1959) (internal citation omitted). “Residence requirements, age, [and] previous criminal record are obvious examples indicating factors which a State may take into consideration in determining the qualifications of voters.” Id. at 51, 79 S. Ct. at 990 (internal citations omitted). And Alabama, which could lawfully disenfranchise all felons permanently, Richardson, 418 U.S. at 56, 94 S. Ct. at 2671, has not exceeded its interest per the seventh factor by choosing only to disenfranchise individuals who commit felonies Alabama considers especially heinous.
Besides the second factor, which is neutral, all seven Mendoza-Martinez factors weigh in favor of finding that Alabama's disenfranchisement provision is not “so punitive either in purpose or effect as to negate the State's intention to deem it ‘civil.’ ” Doe, 538 U.S. at 92, 123 S. Ct. at 1147 (quotation, quotation marks, and alterations omitted)). Even if the second factor weighed in the Appellants’ favor, we could not say that factor alone is “clearest proof” sufficient to “transform what had been denominated a civil remedy into a criminal penalty.” Id. (quoting Hudson, 522 U.S. at 100, 118 S. Ct. at 493). Accordingly, we hold that the disenfranchisement provision of Amendment 579 does not constitute punishment for purposes of the Ex Post Facto Clause.
Finally, Appellants contend that Alabama's mail voter registration form violates the NVRA because the form does not explicitly list all the disqualifying felonies under Alabama law. The NVRA requires the EAC to “develop a mail voter registration application form” in consultation with the states. 52 U.S.C. § 20508(a)(2). This form must “include a statement that ․ specifies each eligibility requirement.” 52 U.S.C. § 20508(b)(2)(A) (emphasis added). The verb “specifies” is not defined by the NVRA. Under Appellants’ view, a state that disqualifies voters for some felonies but not others can only sufficiently specify its eligibility requirements on its mail voting form by listing each disqualifying felony. So, the portion of Alabama's form informing registrants that felonies involving moral turpitude are disqualifying and that registrants can access the list of disqualifying felonies by following a specified link is insufficient.
Appellants, however, propose an absurd, unworkable, and internally inconsistent interpretation of § 20508(b)(2)(A). As the District Court observed, Appellants’ interpretation would require Alabama to list every state, federal, and foreign felony involving moral turpitude to sufficiently specify disqualifying felonies under Alabama law. Thompson, 505 F. Supp. 3d at 1271. After all, Alabama's felon disenfranchisement provision applies to all individuals who have committed felonies involving moral turpitude, whether convicted by Alabama or another sovereign. Ala. Code § 17-3-30.1(c)(48) (disqualifying individuals who have committed “[a]ny crime as defined by the laws of the United States or by the laws of another state, territory, country, or other jurisdiction, which, if committed in this state, would constitute one of the offenses listed in this subsection” as involving moral turpitude). Assuming Alabama could even identify every such felony, the resulting registration form would be of monstrous size. Appellants may as well ask Alabama to attach a copy of each state, federal, and foreign criminal code to its voting form. And any time any state, federal, or foreign government amended their criminal code, Alabama would have to update its list of disqualifying felonies and print anew its prodigious voter registration forms.
Appellants respond that this is a “strawman argument” and that listing the catchall provision of Ala. Code § 17-3-30.1(c)(48) would suffice for non-Alabama felony convictions. But this argument renders Appellants’ position internally inconsistent. As the District Court put it, if “a catchall provision which generally refers to particular crimes is sufficiently specific to satisfy the requirement that the form ‘specify’ the qualification, then a specification of a qualification—disqualifying felony—which generally refers to particular crimes must also be specific enough.” Thompson, 505 F. Supp. 3d at 1271. We agree.
We have explained before that “[c]ourts should avoid slicing a single word from a sentence, mounting it on a definitional slide, and putting it under a microscope in an attempt to discern the meaning of an entire statutory provision.” Wachovia Bank, N.A. v. United States, 455 F.3d 1261, 1267 (11th Cir. 2006). Appellants’ attempt to do so here would produce a clearly absurd result. See United States ex rel. Hunt v. Cochise Consultancy, Inc., 887 F.3d 1081, 1091 (explaining that courts “should refrain from interpreting a statute in a way that produces a result that is not just unwise but is clearly absurd” (quotation and quotation marks omitted)). Boiled down, § 20508(b)(2)(A) is a notice statute enacted for the convenience of voting registrants. Alabama's mail-in voting form has provided sufficient notice by informing registrants that persons convicted of disqualifying felonies are not eligible to vote and providing an easily accessible link 13 whereby voters convicted of felonies can determine their voter eligibility. Accordingly, Alabama has complied with the requirements of § 20508(b)(2)(A).
For the reasons stated above, we affirm the District Court's grant of summary judgment in favor of Alabama.
Deceiving an elector in preparation of her ballot.1 Altering another person's ballot.2 Failing to count legally cast absentee votes.3 Illegally voting more than once in an election (second violation).4 Willfully and intentionally signing the name of another elector in a poll book.5 Bribery of public servants.6 And perjury.7
Perhaps this recitation sounds like a list of felonies that would disqualify an Alabamian from voting under Amendment 579 to Alabama's constitution—Alabama's felon-disenfranchisement provision. Nope. Those convicted of any of these voting-fraud-related felonies are A-okay, good to go when it comes to voting in Alabama. Alabama exempts them from its felon-disenfranchisement provision, Amendment 579. Under that provision, only other felons—those convicted of felony crimes that Alabama says are crimes of “moral turpitude”—can't vote.
Even worse, in the nearly thirty years since Alabama amended its felon-disenfranchisement provision, Alabama has defined the phrase “moral turpitude” in contradictory or non-uniform ways. At one point, Alabama even allowed each local registrar to interpret the term for herself. In other words, when Alabama precluded those convicted of felony crimes of “moral turpitude” from voting, it may as well have excluded those convicted of “whatever felonies Alabama (or any of its local registrars) at any point in the future might say disqualify a voter,” as Alabama had no definition of the phrase “moral turpitude” in mind.
All of this raises the question: just what was Alabama trying to accomplish with its felon-disenfranchisement provision?
Going back some time to when Alabama adopted its original felon-disenfranchisement provision (which disenfranchised all felons) in 1901, Alabama did so to further white supremacy and suppress Black voting. Hunter v. Underwood, 471 U.S. 222, 229, 105 S.Ct. 1916, 85 L.Ed.2d 222 (1985) (noting that no one “seriously dispute[d] the claim that this zeal for white supremacy ran rampant at the convention”) (“Hunter”).
So maybe when Alabama amended its felon-disenfranchisement provision in 1996, narrowing those prohibited from voting from all felons to only those convicted of felony crimes of moral turpitude, it sought to cleanse the taint of racism from the provision's history? Nope. We can't say that. Alabama has never stated that it amended its felon-disenfranchisement provision to correct the provision's racist origins.
Rather, when Alabama adopted Amendment 579, it described the purpose as “strictly housekeeping.” Doc. 257-17 at 52. But that description is hard to believe both because of (1) the substantive changes the 1996 amendment made to Alabama's felony-disenfranchisement provision and (2) Alabama's ongoing struggle to define what it means by felony crimes of moral turpitude.
Indeed, if Alabama's purpose was “strictly housekeeping,” Alabama needs a new housekeeper. When Alabama amended its felon-disenfranchisement provision to preclude those convicted of felonies of moral turpitude from voting, it left felonies strewn all over without identifying whether they went into or outside the “moral turpitude” closet. In fact, Alabama had no idea what was in its closet and even less of an idea about what it wanted to put there. And the mismatch between Alabama's stated purpose for amending its felon-disenfranchisement provision and the disarray in which Alabama's amendment left its felon-disenfranchisement provision sure makes it seem like Alabama's purpose was not “strictly housekeeping.”
So let's review the facts: (1) Alabama's felon-disenfranchisement provision undisputedly began as a racist mechanism to suppress the Black vote; (2) Alabama's only stated purpose for reenacting an amended form of that provision is inconsistent with what Alabama actually did; and (3) Alabama's construction of its felony-disenfranchisement provision—allowing those convicted of voting-related fraud to continue to vote—is inconsistent the purpose of disenfranchising those convicted of felony crimes of “moral turpitude” (i.e., not letting those involved in fraud-related crimes vote because they damage the collective honor of the community).
But our Equal Protection Clause precedent requires us to ignore all these facts. Rather, we simply ask whether the amended version of the law that was originally enacted for discriminatory reasons went through both chambers of the legislature and was properly effected into law. See Johnson v. Governor of State of Fla., 405 F.3d 1214, 1223–25 (11th Cir. 2005) (en banc). What kind of test is that? None at all, for every law that comes before this Court has experienced bicameralism and presentment. So a law that is a law passes muster under our precedent.
Nor is our deficient test consistent with Supreme Court precedent. When a law is challenged as discriminatory for the first time, the Supreme Court applies a variety of factors (known as the Arlington Heights factors)—including, among others, whether the law has a discriminatory impact, the historical background of a law, and the “substantive departures” the law makes from the reasons stated for its enactment—to assess whether the law violates the Equal Protection Clause.8 So if a federal court concludes under these factors that a law violates the Equal Protection Clause, and that law is later reenacted, why should that law that continues to have a disparate impact get a free pass on the factors and be reviewed instead solely to ensure it went through the proper legislative process? That makes no sense. It should be harder, not easier, for a law to survive an Equal Protection Clause challenge when it has been reenacted after a federal court has found the law to violate the Equal Protection Clause.
Applying the Arlington Heights factors to assess whether a reenacted version of a law that a federal court has previously held to violate the Equal Protection Clause not only makes far more sense, but it is truer to Supreme Court precedent than is our test. And applying the Arlington Heights factors to Amendment 579 and its interpretive statute yields only one conclusion: a material issue of fact exists about whether these laws were adopted for a discriminatory purpose.
Indeed, when, as here, the amended law does nothing to advance its stated purpose, it cannot cleanse the taint of its discriminatory origins. For that reason, if I were not bound by our precedent, I would hold that Alabama's felon-disenfranchisement provision violates the Equal Protection Clause. But since I am bound, I cannot and must instead conclude that, under our case law, the provision does not violate the Equal Protection Clause.
That said, though, Alabama's felon-disenfranchisement statute and its voter registration form do violate the Ex Post Facto Clause and the National Voter Registration Act, respectively. So I would reverse the district court's denial of those challenges.
My discussion proceeds in five parts. In Section I, I recount Alabama's unfortunate history of discrimination against Black voters. Section II tells the story of how Amendment 579 came to be. I address the Equal Protection Clause issue in Section III. In Section IV, I explain why Section 17-3-30.1 violates the Ex Post Facto Clause. And Section V shows that Alabama's voter-registration form does not comply with the National Voter Registration Act's requirements.
I. ALABAMA'S HISTORY OF RACIAL DISCRIMINATION IN VOTING
Black suffrage in the United States was a long time in coming. But after the Civil War, things began looking up for Blacks during Reconstruction. For instance, the United States ratified the Fourteenth Amendment on July 9, 1868. Among other functions, the Fourteenth Amendment began ensuring that Black people “born or naturalized in the United States” enjoyed citizenship and “equal protection of the laws.” U.S. Const. amend. XIV. And it reduced a state's representation in the House of Representatives in proportion to that state's refusal to allow male citizens over 21 to vote (other than for “participation in ․ [as relevant here,] crime”). Id.
Two years later, on February 3, 1870, the United States ratified the Fifteenth Amendment. The Fifteenth Amendment guarantees that “[t]he right of citizens of the United States to vote shall not be denied or abridged ․ by any State on account of race, color, or previous condition of servitude.” U.S. Const. amend. XV.
These rights translated into results at the ballot box in Alabama. For the 42nd, 43rd, and 44th United States Congresses, Alabama elected a Black man to serve as one of its Representatives.9 But Alabama's burgeoning nineteenth-century tradition of Black congressional representatives quickly and unceremoniously ended with the 44th Congress, which closed in 1877. Not coincidentally, 1877 was also when Reconstruction ended with the Compromise of 1877.10 Seminole Tribe of Fla. v. Florida, 517 U.S. 44, 120, 116 S.Ct. 1114, 134 L.Ed.2d 252 (1996) (Souter, J., dissenting).
The Compromise was this: following a disputed presidential election, Democratic southern states would accede to the Republican northern states’ choice—Rutherford B. Hayes—provided that the new President withdraw federal troops from the south. Id. While “President-elect Hayes received assurances that the Democratic governments of the southern states would take upon themselves the responsibility to protect the civil rights of their [B]lack citizens ․ [t]h[o]se assurances ․ probably were ․ disingenuous.” Michael McConnell, The Forgotten Constitutional Moment, 11 Const. Commentary 115, 130–31 (1994). “Once power shifted back to the southern states and away from Congress, the promises of continued respect for the rights of [B]lack Americans quickly proved illusory. This probably came as no surprise to anyone.” Id.
Alabama took this opening and ran with it. See Underwood v. Hunter, 730 F.2d 614, 618 (11th Cir. 1984) (“Underwood”). As we have recognized, “the white citizens of Alabama moved to reassert their once unquestioned political supremacy” over Black Alabamians. Id. In 1901, Alabama adopted a state constitution designed to restrict Black voting. Id. at 619.
“When the Alabama constitutional convention assembled in May 1901, the question was not whether to disfranchise the Negro but rather how to do so constitutionally.” Id. at 619. As the president of the Alabama constitutional convention put it, “[a]nd what is it that we want to do? Why it is within the limits imposed by the Federal Constitution, to establish white supremacy in this State.” John B. Knox, President of the 1901 Convention, Off. Proceedings of the Const. Convention of the State of Ala., May 21st, 1901, to Sept. 3rd, 1901, at 2278 (1940). He continued, explaining that “if we would have white supremacy, we must establish it by law—not by force or fraud.” Id. at 2279. As the Supreme Court later noted, no one “seriously dispute[s] the claim that this zeal for white supremacy ran rampant at the convention.” Hunter, 471 U.S. at 229, 105 S.Ct. 1916.
To disenfranchise Black voters without expressly disenfranchising Black voters, Alabama's so-called “Suffrage” Committee adopted a Swiss-cheese approach, layering various voting restrictions on top of each other. Together, those layers “would subvert the guarantees of the fourteenth and fifteenth amendments without directly provoking a legal challenge.” Underwood, 730 F.2d at 619. The committee settled on a combination of property tests, literacy tests, residency requirements, a poll tax, and, as relevant here, a felon-disenfranchisement provision, resulting in an “exceptionally byzantine suffrage scheme.” Id.
The lynchpin of the plan—or, as newspapers owned by Alabama's governor put it at the time, the “milk in the cocoanut [sic]”—were the county registrars. Doc. 270-3 at 46. The registrars, through the ability to require character affidavits and impose other opaque exclusions, had “a powerful set of instruments to shape the state's voting rolls nearly at-will.” Id. For instance, registrars could impose (then-permissible) tests, like an “understanding test” on particular would-be voters and not others. Id. at 47.
Over the years, “[s]ome of the more blatantly discriminatory selections” were “struck down by the courts.” Hunter, 471 U.S. at 233, 105 S.Ct. 1916. Still, though, Black voter registration in Alabama lagged behind (and still lags behind) non-Black voter registration. For instance, in 1957, only 1% of Blacks in Dallas County, Alabama, were registered to vote. James Blacksher, Voting Rights in Alabama: 1982–2006, 17 S. Cal. Rev. L. & Soc. Just. 249, 252 (2008) (herein-after “Voting Rights”). Every time the Department of Justice eliminated one disenfranchising device, Dallas County implemented a new one. Id. In fact, in the legislative history for the Voting Rights Act, Congress cited Dallas County as an example of “[t]he insufficiency of existing remedies and the need for stronger measures.” H.R. Rep. 89-439 (1965), U.S.C.C.A.N. 2437, 2441 (1965). And Alabama's racial gerrymandering of districts led to the seminal one person, one vote case, Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964). Blacksher, Voting Rights, supra, at 272.
Unfortunately, the Voting Rights Act (“VRA”) didn't end some Alabamians’ determination to restrict Black Alabamians from voting. In the seventeen years between 1965 and 1982, the Department of Justice had to object fifty-nine times to proposed restrictions on voting by Alabama or a political subdivision in Alabama because the Department found those restrictions racially discriminatory. Blacksher, Voting Rights, supra, at 254. And in 1982, Congress amended the VRA in response to (yet another) Alabama voting restriction. Id. at 253 (citing City of Mobile v. Bolden, 446 U.S. 55, 100 S.Ct. 1519, 64 L.Ed.2d 47 (1980)).
The list of cases in which federal courts had to enjoin or strike down Alabama voting restrictions in the 1970s, 80s, and 90s just goes on and on. See, e.g., Pleasant Grove v. United States, 479 U.S. 462, 472, 107 S.Ct. 794, 93 L.Ed.2d 866 (1987) (concluding that Pleasant Grove engaged in purposeful discrimination to minimize future Black voting strength); United States v. Marengo Cnty. Comm'n, 731 F.2d 1546, 1574 (11th Cir. 1984) (“We hold that the record shows a clear violation of the results test adopted by Congress in section 2 of the Voting Rights Act. ․ We note, however, that despite our repeated requests at oral argument, counsel for the defendants did not provide this Court with any sign that would indicate that the political opportunities for Marengo County blacks have improved since 1978.”); Harris v. Siegelman, 695 F. Supp. 517, 525 (M.D. Ala. 1988) (“The plaintiffs also presented compelling evidence that this history of racial inhumanity continues into today, and, more specifically, that, in the tradition established by this state, white poll officials continue to harass and intimidate black voters. Witnesses detailed numerous instances of where white poll officials refused to help illiterate black voters or refused to allow them to vote, where they refused to allow black voters to cast challenged ballots, and where they were simply rude and even intimidating toward black voters.”); Dillard v. Crenshaw Cnty., 640 F. Supp. 1347, 1360 (M.D. Ala. 1986) (“From the late 1800s through the present, [Alabama] has consistently erected barriers to keep black persons from full and equal participation in the social, economic, and political life of the state.” (cleaned up)); Dillard v. Baldwin Cnty. Bd. of Educ., 686 F. Supp. 1459, 1461 (M.D. Ala. 1988) (“[T]he evidence established that the legislature engaged in a century-long pattern and practice of switching between local at-large systems and local single-member district systems as needed to diminish black voting strength.”); Shaw v. Reno, 509 U.S. 630, 640, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993) (“Alabama redefined the boundaries of the city of Tuskegee ‘from a square to an uncouth twenty-eight-sided figure’ in a manner that was alleged to exclude black voters, and only black voters, from the city limits.” (citing Gomillion v. Lightfoot, 364 U.S. 339, 340, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960))).
And this discrimination did not end with the close of the twentieth century. Rather, in 2010, the FBI recorded members of the Alabama State Senate referring to Black Alabamians as “Aborigines” and aiming to quash a referendum that might increase Black voter turnout. Shelby Cnty. v. Holder, 570 U.S. 529, 584, 133 S.Ct. 2612, 186 L.Ed.2d 651 (2013) (Ginsberg, J., dissenting) (citing United States v. McGregor, 824 F. Supp. 2d 1339, 1344–48 (M.D. Ala. 2011)). That didn't happen in 1901 but just a few short years ago.
II. ALABAMA'S HISTORY OF FELON DISENFRANCHISEMENT
With this background, I return to Alabama's felon-disenfranchisement statute in particular. At the 1901 Convention, to choose which crimes—felonies and misdemeanors—would trigger loss of the right to vote, the Suffrage Committee perused the Alabama criminal code for statutes that were “peculiar to the Negro's low economic and social status.” Underwood, 730 F.2d at 619. Though, on paper, the Committee disenfranchised all felons, Alabama did not equally enforce that restriction against Black and white felons, id. at 620. The Committee also selected a long list of disenfranchising non-felony crimes, adding the catchall category of “moral turpitude”:
All idiots and insane persons; those who shall by reason of conviction of crime be disqualified from voting at the time of the ratification of this Constitution; those who shall be convicted of treason, murder, arson, embezzlement, malfeasance in office, larceny, receiving stolen property, obtaining property or money under false pretenses, perjury, subornation of perjury, robbery, assault with intent to rob, burglary, forgery, bribery, assault and battery on the wife, bigamy, living in adultery, sodomy, incest, rape, miscegenation, crime against nature, or any crime punishable by imprisonment in the penitentiary, or of any infamous crime or crime involving moral turpitude; also, any person who shall be convicted as a vagrant or tramp, or of selling or offering to sell his vote or the vote of another, or of buying or offering to buy the vote of another, or of making or offering to make a false return in any election by the people or in any primary election to procure the nomination or election of any person to any office, or of suborning any witness or registrar to secure the registration of any person as an elector.
Ala. Const. art. VIII, § 182 (1901) (emphasis added).
The Committee achieved its aim. Just two short years later, by 1903, the felon-disenfranchisement provision applied to roughly ten times as many Blacks as whites. Underwood, 730 F.2d at 620. The disparity continues until today. Doc. 215-13 ¶ 28.
Over the years, Alabama considered revisiting the felon-disenfranchisement provision from time to time, though it never made any changes until 1995. But during the late 1960s and early 1970s, Governor Albert Brewer convened a commission to investigate the need to amend the constitution and to provide a final report to the legislature. That commission then formed a subcommission on Suffrage and Elections. And the subcommission sent a draft report to the full commission, which suggested moving from a list of enumerated crimes to one that proposed “general terms,” whereby “[a]s statutory offenses grow or change, their inclusion or exclusion becomes a matter of constitutional interpretation or constitutional amendment.” The subcommission on Suffrage and Elections also proposed “eliminating a long, scattered[,] and redundant list of disqualifying crimes.”
The new draft proposal denied the vote (1) to only those convicted of a felony involving moral turpitude and (2) the incompetent. Though the commission's minutes do not reflect any debate on the proposed changes, the proposed changes were approved by a vote of 11-2. The commission's proposed constitution—with the new language on felon disenfranchisement—was introduced into the Alabama Legislature with no debate or changes to the language. Ultimately, the Alabama Legislature did not pass the proposed language.
In the late 1970s, Alabama tried again to amend its constitution. Governor Fob James convened an informal working group to amend the constitution, stressing that the “subjects in which [he] was most interested” were “popular democracy ․ and low property taxes.” The working group briefly discussed which felonies revealed “moral turpitude.” Along these lines, the working group heard testimony from the ACLU that the phrase “moral turpitude” was “vague and indefinite” and appeared “unwarranted and discriminatory.” In the end, the working group suggested eliminating the list of enumerated felonies, thereby simplifying the provision. Again, the proposed new constitution was never passed.
After a false start in the 1980s, Alabama returned to the felon-disenfranchisement question in 1995. State Representative Jack Venable introduced a bill that would amend the felon-disenfranchisement provision to prohibit only those convicted of felonies of moral turpitude (and the mentally incompetent) from voting. No debate on the amendment occurred in committee. And no hearings on the amendment occurred in committee, either. Rather, the day after Representative Venable introduced the amendment, the Standing Committee on Constitution and Elections simply returned it to the House with a favorable report. Then the full House passed the amendment without any debate or discussion. At no point did the legislature explain the meaning of “moral turpitude” in the amendment.
Once the amendment went to the Senate, it quickly moved through committee to the floor and passed. After it passed both houses of the Alabama Legislature, the amendment went to a vote before the Alabama voters. In an editorial to the public, Representative Venable urged the public to pass the amendment and described it as “strictly housekeeping.” The Alabama public approved the change.
The bill had one final step before it became a law. In 1996, because Alabama had been part of perpetrating “an insidious and pervasive evil ․ through unremitting and ingenious defiance of the Constitution” (meaning Jim Crow laws and segregation), it was required to obtain “preclearance” from the Department of Justice before it could implement a change to state election law. Shelby Cnty., 570 U.S. at 535, 537, 133 S.Ct. 2612. While Alabama obtained the necessary approvals, it falsely told the Department of Justice that the bill didn't make a change because Alabama was already using the provisions. Maj. Op. at –––– – ––––. But the bill did make a change—it removed a laundry list of felonies and left in its place the ambiguous “moral turpitude” language, giving county registrars free rein to reject voting registrations for any crime they saw fit.
And thus began Alabama's decades-long struggle to explain what it meant by felonies of “moral turpitude.” As the history of Alabama's varied interpretations of the phrase from Amendment 579 show—since 1996, Alabama officials (including local registrars in the initial years) have construed crimes of “moral turpitude” to include anywhere from fifteen crimes to all felonies—Alabama's selection of qualifying crimes is most accurately described as arbitrary.
As I've noted, upon adoption of Amendment 579, Alabama registrars were inconsistent in deciding which felonies were those of moral turpitude. Indeed, as the Majority Opinion acknowledges, some registrars refused to register any felons at all. Maj. Op. at –––– – ––––. As a result, the moral-turpitude provision could be construed so that the same crime could be disqualifying in one county but not in another.
Various Alabama governmental actors stepped up to fix the problem. First, in 2005, the Alabama Board of Pardons and Paroles asked the Attorney General which felonies involved moral turpitude. Ala Op. Atty. Gen. No. 2005-092 (Ala. A.G.), 2005 WL 1121853. Conceding the vague nature of the phrase, the Attorney General explained that he could not provide “an exhaustive list” of felonies involving moral turpitude but that he could generate a list of crimes that Alabama courts had determined to be crimes of moral turpitude. Id. at *2. The Attorney General's list had about fifteen crimes. Id.
Second, between 2007 and 2008, the Alabama governor and the Alabama Administrative Office of the Courts (“AOC”) produced dueling lists. Doc. No. 269-12 at 2. As background, the United States sued Alabama to comply with the Help America Vote Act. Id. at 11 (citing United States v. Alabama, 06-cv-392-WKW (M.D. Ala.)). The federal judge handling the case appointed the Alabama governor as a special master. Id. The governor “believed that [his] obligations as [s]pecial [m]aster included making a determination of what felonies involved moral turpitude.” Id. According to the AOC, the governor's office had an attorney and a law-student intern compile a list that ultimately took the position that about 480 felonies (out of the then-existing 575 felonies under Alabama law) involved moral turpitude. Id. at 11–12. The governor asked the AOC to review the list. Id. at 12.
The AOC responded that the governor's list was too expansive and instead identified only seventy or so felonies that it said involved moral turpitude. Id. In the AOC's view, the governor's list was too broad because it included more than those crimes that the Alabama courts had said involved moral turpitude. Id. “In light of the preeminent importance of the right to vote in our country and also the relatively recent history of voting rights strife, including voter and voter registrant intimidation, in [Alabama],” the AOC wrote, it would “not be a party” to denying citizens the right to vote. Id.
The Alabama Secretary of State insisted that the shorter AOC list, not the governor's longer list, be supplied to county registrars. Id. But the next fall, the AOC received disturbing reports: county registrars were refusing to register people to vote because those people had been convicted of crimes on the governor's list. Id. In September 2008, the AOC learned that registrars were being sent lists entitled “Weekly Felon Check” that listed crimes that were definitively not crimes of moral turpitude. Id. at 13. For instance, one registrar refused to allow someone to register to vote because that person had been convicted of “having more beer in their car than the law allowed.” Id. Beer, for Pete's sake!
The arbitrariness didn't end there. When the AOC followed up with the governor's office, it learned that the governor's list had been used to disqualify voters. Id. at 15. The AOC sought to correct the problem so that only the AOC list would be used to disqualify voters. But the governor's office resisted, replying that “the big list” was “in the system” and taking it out would be “very expensive.” Id. Ultimately, the governor's office did not agree to change the report until the “end of October” 2008. Id. But that meant that Alabama enforced the wrong list and arbitrarily prohibited the registration of voters with nonqualifying felonies until just a week before the 2008 general election. Id.
Unsurprisingly, the dueling lists caused confusion. At the time, two registrars (out of three) had to agree to register someone to vote. So the process (a best two-out-of-three system that involved no standardized list) led to unpredictable enforcement.
And somehow, the situation got worse still.
In 2014, Secretary of State John Merrill's office tasked a college-student-campaign-staffer-turned-employee 11 with compiling an official list of which crimes involved moral turpitude for inclusion in a registrar's handbook. The student didn't know much about moral turpitude, so he asked Wikipedia. Yes, Wikipedia. There, the student pulled the federal statutory list of crimes used to disqualify applicants from obtaining visas or citizenship and copied it. The student didn't check the list against Alabama law and didn't do anything to learn more about moral turpitude. Then this arbitrarily assembled list was sent to state registrars.
Alabama took its next stab at what it meant by felonies of “moral turpitude” in late 2015. At that time, Secretary Merrill formed a “Voter Disenfranchisement and Restoration of Voting Rights Exploratory Committee.” (“The Exploratory Committee”). The Committee had a subcommittee, chaired by Edward Packard—Alabama's Administrator of Elections since 1994—on moral turpitude. That subcommittee recommended a list of fifteen disqualifying felonies to satisfy the moral-turpitude provision. The list of crimes of moral turpitude didn't include crimes like bribery, public corruption, and perjury. Packard (who drafted the list) admitted that “there wasn't any particular reason for excluding” those crimes. In other words, he conceded the arbitrariness of his subcommittee's suggested list.
The subcommittee presented its findings to the Exploratory Committee. The Exploratory Committee considered the list and added one other crime: human trafficking. Had the Exploratory Committee's list been adopted, Black Alabamians would not have been disenfranchised at a higher rate than their representation in the Alabamians-with-felony-convictions population.
Despite the Exploratory Committee's recommendation, Secretary Merrill “redirected the discussion” to a longer list (of about forty crimes) patterned after a 2015 Alabama bill, H.B. 344. Packard—who was at the meeting and who had drafted Secretary Merrill's longer list—didn't know why Secretary Merrill had ignored the Committee's recommendation. He didn't remember being surprised that no one objected—even though the Committee had approved a shorter list the previous hour—and said that any impact the Committee had was in adding crimes—not removing them. Packard recalled that the Committee had gone through the longer list and “asked if people thought [particular crimes] should be removed from the list.”
Along these lines, Packard didn't remember any objection to including crimes like voter fraud or public corruption. In fact, Packard said, there were no set criteria by which crimes were included: none were discussed and everyone used their own personal criteria. He didn't “recall there being that much discussion” about the list all.
The longer list was sent to the legislature, introduced as H.B. 282, and ultimately become Alabama Code Section 17-3-30.1, which lists felonies that automatically disenfranchise those convicted of them as felony crimes of moral turpitude. Ala. Code § 17-3-30.1. Section 17-3-30.1 has a discriminatory impact: of those Alabamians convicted of a felony, Black Alabamians are more likely to be disenfranchised than white Alabamians by a rate of 35% of whites to 40% of Blacks.
This sad history brings me to our case. The Plaintiff-Appellants are a group of Black Alabamians with convictions that currently qualify as felony crimes of “moral turpitude” under Section 17-3-30.1. They are therefore unable to exercise their constitutionally protected right to vote under Amendment 579 and Section 17-3-30.1, and they challenge those provisions. The district court dismissed or granted summary judgment on all their claims. On appeal, they press just three: (1) that Alabama's 1996 reenactment of its felon-disenfranchisement provision in its constitution is still tainted by the 1901 racist enactment and therefore violates the Equal Protection clause; (2) that Alabama's 1996 reenactment and subsequent statutory clarification of which felons are disenfranchised constitute “punishment” and thus violate the Ex Post Facto Clause; and (3) that Alabama's voter-registration form doesn't comply with the National Voter Registration Act's requirements.
I must concur with the Majority Opinion as to the first claim because our precedent requires it, even though it seems clear to me that Alabama has not done enough to cleanse the taint of Amendment 579's discriminatory origins. As to the Ex Post Facto Clause and VRA claims, though, I respectfully dissent.
III. PROPERLY ANALYZED, ALABAMA'S REENACTMENT VIOLATES THE EQUAL PROTECTION CLAUSE
On a blank slate, I would remand this case to the district court for a trial on the Appellants’ Equal Protection Clause claim because Amendment 579 didn't erase the racist taint of Alabama's felon-disenfranchisement provision. Unfortunately, I don't write on a blank slate, and our precedent requires me to vote to affirm.
But in my view, our precedent is wrong: it allows states to simply reenact racially discriminatory laws without debate or discussion, entirely undercutting the purpose of the Equal Protection Clause litigation that found the initial provision to have been enacted for a discriminatory reason in the first place. We should revisit our precedent. In Part III.A, I explain why our precedent requires me to vote to affirm and why our precedent is wrong. In Part III.B, I sketch out what the proper test should look like and apply it to this case.
A. Our precedent sets the bar too low to remove discriminatory taint from an earlier law.
Our precedent sets a meaningless “bar” to reenact discriminatory laws and is out of step with the Supreme Court's jurisprudence. Indeed, the standard to survive an Equal Protection Clause challenge in this Circuit is lower when a state reenacts a law that a court has already found to be enacted with discriminatory intent than when a state enacts a law without a predecessor that was found to have been enacted with discriminatory intent. In other words, it's easier for a state to reenact a law with racist origins than it is for a state to enact a law without racist origins.
Johnson v. Governor of State of Florida is our seminal precedent on reenactments of laws that were originally enacted with discriminatory intent. 405 F.3d at 1223. There, we approved Florida's reenactment of its discriminatory disenfranchisement provision because the bill went through what we said was a deliberative process: the law went through a committee, then the full legislature, and then was approved by the voters. Id. at 1224. We said that was enough. Id. And we rejected the notion that the Florida legislature had to explicitly discuss (and discount) the racist motivation underlying the earlier law. Id. We also did not consider requiring the state to provide a legitimate, non-discriminatory reason for reenacting the law that had previously been found to be discriminatory.
Alabama's reenactment sufficiently parallels Florida's process. Like Florida's version, Alabama's Amendment 579 went through a committee, then the full legislature, and then was approved by the voters.
To be sure, Alabama's process was less deliberative than Florida's. Representative Venable introduced the amendment into committee, where it spent a single day before being reported out favorably. No hearings were held. The next week, Representative Venable offered it for a second reading, and it was brought to the House floor, where in Representative Venable's words, “limited, really no debate on the bill” occurred. But I can't say that Alabama's process is distinguishable in any meaningful way from the process we approved in Johnson. After all, as Judge Barkett pointed out in her dissent, as we used the term, “the adjective ‘deliberative’ describes the procedural aspects of the decision, [but] it need not include any substantive component at all.” Johnson, 405 F.3d at 1246 (Barkett, J., dissenting). So I am bound to vote to affirm.
Our precedent is wrong, though. Our precedent requires only bicameralism and presentment to “cleanse” a law of discriminatory taint. I see two problems with our minimal “test”: it is inconsistent with the Supreme Court's jurisprudence in Arlington Heights, and it is so low a bar as to be meaningless.
As to Arlington Heights, the Supreme Court has explained that we should consider a variety of indicators to determine whether the government enacted a law for discriminatory reasons (I discuss these factors later), in violation of the Equal Protection Clause. Arlington Heights, 429 U.S. at 267–68, 97 S.Ct. 555. But in Johnson, we skipped Arlington Heights’s multifactor test altogether in favor of a single question: did the law go through bicameralism and presentment? 405 F.3d at 1224. We didn't even mention the factors the Supreme Court enumerated in Arlington Heights. Id.
It cannot be that the bar is lower in a case like this—with a past finding of discrimination—than in an ordinary case. That is, if Alabama had passed this amendment with a disparate impact in the first instance, we would perform an Arlington Heights analysis to determine whether Alabama had discriminatory intent. But now that Alabama is reenacting an amended version of a provision the Supreme Court determined was originally adopted for discriminatory reasons, somehow, Alabama doesn't have to pass an Arlington Heights analysis? That can't be right.
Second, even taken on its own terms, the Johnson bicameralism-and-presentment requirement is meaningless. How could that requirement ever be unsatisfied in the real world? As a federal court, we will likely never be asked to rule on a law or constitutional amendment that didn't go through that very process. In fact, a challenge before that process is completed might be unripe. But our standard—which doesn't require the legislature to state a nondiscriminatory basis for the new law that is actually furthered by the new law—lays the bar reenactments must clear flat on the floor. Any law can shuffle right over it so long as legislators are not obtuse enough to state out loud any discriminatory intentions. Such a requirement is meaningless. Cf. Dep't of Comm. v. New York, ––– U.S. ––––, 139 S. Ct. 2551, 2575–76, 204 L.Ed.2d 978 (“[W]e cannot ignore the disconnect between the decision made and the explanation given. Our review is deferential, but we are ‘not required to exhibit a naiveté from which ordinary citizens are free.’ United States v. Stanchich, 550 F.2d 1294, 1300 (2d Cir. 1977) (Friendly, J.). The reasoned explanation requirement of administrative law, after all, is meant to ensure that agencies offer genuine justifications for important decisions, reasons that can be scrutinized by courts and the interested public. Accepting contrived reasons would defeat the purpose of the enterprise. If judicial review is to be more than an empty ritual, it must demand something better than the explanation offered for the action taken in this case.”).
Think about how reenacting court-enjoined discriminatory legislation unfolds under our precedent. At Time 1, a legislature votes with a discriminatory purpose to prevent minority citizens from voting. Plaintiffs challenge the law or regulation and prove that the legislature acted with discriminatory intent, no small feat. At Time 2 (which, under our precedent, could be next year), a legislature reenacts the exact same statute, or tweaks the original statute a little bit, and the law will sail through a court challenge, so long as no legislator is foolish enough to say the quiet part out loud. And after all, what legislator in modern times is? Cf. Rollins v. TechSouth, Inc., 833 F.2d 1525, 1528 (11th Cir. 1987) (“It is rare that direct evidence of discrimination exists.”); see also Price Waterhouse v. Hopkins, 490 U.S. 228, 271, 109 S.Ct. 1775, 104 L.Ed.2d 268 (1989) (O'Connor, J., concurring), superseded in part by The Civil Rights Act of 1991, Title I, § 107(a), 105 Stat. 1075 (codified at 42 U.S.C. § 2000e-2(m)) (“[T]he entire purpose of the McDonnell Douglas prima facie case is to compensate for the fact that direct evidence of intentional discrimination is hard to come by.”).
Our precedent effectively clothes laws with a presumption of good faith—even if the law is a reenacted version of a law that has been established to have been enacted originally for a discriminatory purpose. See Johnson, 405 F.3d 1214. But when the impetus for the original version of a law was racial discrimination and the law is reenacted after an express finding of taint, the Supreme Court has suggested that the presumption that the government acted in good faith should not persist: “In these cases, we do not confront a situation like the one in Hunter [where the law was never struck down and therefore never reenacted]. Nor is this a case in which a law originally enacted with discriminatory intent is later reenacted by a different legislature.” See Abbott v. Perez, ––– U.S. ––––, 138 S. Ct. 2305, 2325, 201 L.Ed.2d 714 (2018) (reversing lower courts for not applying presumption of good faith).
For good reason. A presumption of good faith when a past finding of discriminatory animus exists isn't consistent with other areas of Equal Protection Clause jurisprudence. Take employment-discrimination cases, for instance. We have explained that discrimination claims under the Equal Protection Clause (and 42 U.S.C. § 1981) and discrimination claims under Title VII “have the same requirements of proof and use the same analytical framework.” Lewis v. City of Union City, 918 F.3d 1213, 1220 n.5 (11th Cir. 2019) (en banc) (citation and quotation marks omitted).
In those types of cases, in the usual situation—when no direct evidence of discrimination exists—we use the McDonnell Douglas burden-shifting framework to evaluate a claim. See id. at 1220–21. Under that framework, the plaintiff must carry the initial burden of setting out a prima facie case of discrimination “by showing (1) that she belongs to a protected class, (2) that she was subjected to an adverse employment action, (3) that she was qualified to perform the job in question, and (4) that her government employer treated ‘similarly situated’ employees outside her class more favorably.” Id. Once the plaintiff establishes these elements, the burden shifts to government defendant “to articulate a legitimate, nondiscriminatory reason for its actions.” Id. at 1221. If the government satisfies that burden, the plaintiff then must establish that the government's stated reason was “a pretext for unlawful discrimination, an obligation that merges with the plaintiff's ultimate burden of persuading the factfinder that she has been the victim of intentional discrimination.” Id. (cleaned up).
And that framework makes sense. It accounts for reality—in particular, both the near-impossibility of proving discriminatory purpose through direct evidence in modern times and the truth that the government doesn't act without a purpose—so if its purpose is not discriminatory, the government should be able to articulate that purpose and show how that purpose is advanced by its action.
The employment-discrimination Equal Protection Clause analysis differs from the legislative Equal Protection Clause analysis in one key way: there is no presumption that the governmental employer acted in good faith. And that's true even though there's no past finding of intentional discrimination by the government (employer) against the plaintiff.
In the discriminatory-taint-legislation context, though, the situation is just the opposite: we've found discriminatory intent about the current law's immediate predecessor. But under our precedent, we still give the current law a presumption of good faith. That is illogical.
Our precedent's sole taint-cleansing “requirement” (going through the “deliberative process”) is wholly insufficient for ensuring the government is not continuing to violate the Equal Protection Clause. The requirement provides no assurance that the reenacted law that was previously found to have been adopted for discriminatory reasons was reenacted for nondiscriminatory reasons.
B. A Better Way
In my view, if we are really interested in ensuring the enactment of an amended version of an acknowledged racist and discriminatory law has been cleansed of its taint, we should be faithfully applying the Arlington Heights factors to the amended version of the reenacted law. This type of inquiry considers, among other things, whether the amended law has a disparate impact and if so, whether the new law advances the stated purpose for the law so as to outweigh the harm that the disparate impact of what began as a tainted law imposes. Only an inquiry like this, which peers under the hood to evaluate the legitimacy and sincerity of a state's reason for reenactment, addresses the discriminatory-taint problem. When, as here, the amended law has a disparate impact and does nothing to further (and in fact contradicts) its stated purpose, it cannot cleanse the taint of its discriminatory origins.
i. How to Remove Discriminatory Taint
I begin by reviewing the framework under which we evaluate Equal Protection Clause challenges to legislation that has not previously been found to be discriminatory. I then compare that situation to the one here, where we must consider whether an amended version of a reenacted statute that was previously found to have been enacted for a discriminatory reason has been cleansed of its taint.
The Supreme Court has explained that the Equal Protection Clause prohibits only laws that have both (1) a racially disproportionate impact and (2) a racially discriminatory purpose. Hunter, 471 U.S. at 228, 105 S.Ct. 1916 (citing Washington v. Davis, 426 U.S. 229, 239, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976)). Once plaintiffs have shown a disparate impact, we consider direct and indirect evidence of the legislature's motivation or purpose in enacting the statute. Arlington Heights, 429 U.S. at 267, 97 S.Ct. 555; Burton v. Belle Glade, 178 F.3d 1175, 1189 (11th Cir. 1999) (“We evaluate all available direct and circumstantial evidence of intent in determining whether a discriminatory purpose was a motivating factor in a particular decision.”).
In a traditional Equal Protection analysis, we consider several factors: “(1) the impact of the challenged law; (2) the historical background; (3) the specific sequence of events leading up to its passage; (4) procedural and substantive departures; [ ] (5) the contemporary statements and actions of key legislators;” (6) “the foreseeability of the disparate impact; (7) knowledge of that impact, and (8) the availability of less discriminatory alternatives.” Greater Birmingham Ministries v. Sec'y of State of Ala., 992 F.3d 1299, 1322 (11th Cir. 2021) (citing Arlington Heights, 429 U.S. at 267–68, 97 S.Ct. 555). We've explained that these factors are not exhaustive. Jean v. Nelson, 711 F.2d 1455, 1485 (11th Cir. 1983). To determine whether a law violates the Equal Protection Clause in cases where no court has ruled that a predecessor provision of the challenged provision violated the Equal Protection Clause, we apply these Arlington Heights factors (the Supreme Court has not yet decided a reenactment case). See Hunter, 471 U.S. at 233, 105 S.Ct. 1916 (“As such, [the provision] violates equal protection under Arlington Heights.”).
Although Arlington Heights governs our Equal Protection Clause analysis of laws with a discriminatory effect, our precedent does not appear to consider the Arlington Heights factors at all in a challenge to a reenacted law that was previously found to have been enacted for a discriminatory reason. Rather, our sole reenactment case assumed that the original statute had been passed with discriminatory animus. Johnson, 405 F.3d at 1223. Yet without mentioning the Arlington Heights factors and based solely on the reenacted law's bicameralism and presentment, we concluded that the readoption had no indicia of discriminatory animus. Id.
In my view, this makes no sense. It should be harder, not easier, for a law that has been reenacted after a discriminatory finding to survive an Equal Protection Clause challenge than it is for a law facing an Equal Protection Clause challenge for the first time to satisfy such an inquiry. So the Arlington Heights factors should govern reenactment challenges to at least the same extent that they control the analysis in Equal Protection challenges to laws that have not previously been found to be discriminatory.
The Arlington Heights factors allow for consideration of bicameralism and presentment, as Johnson considers. More specifically, the fourth Arlington Heights factor requires consideration of “procedural and substantive departures.” Greater Birmingham Ministries, 992 F.3d 1299 at 1322. A failure to engage in bicameralism and presentment could indicate a “procedural ․ departure[ ].” Id.
But that factor does not end the Arlington Heights inquiry. Rather, we consider seven other factors, as well as the other half of the fourth factor—“substantive departures.” The term “substantive departures” refers to substantive irregularities—for instance, a law that does the opposite of its claimed purpose. The category of “substantive departures” accounts for the fact that the state has either failed to give a reason for reenacting its law that a federal court has previously found to have been enacted for discriminatory reasons or failed to provide a reason that matches the effect of the reenacted law. See Arlington Heights, 429 U.S. at 564–65, 97 S.Ct. 555 (“Substantive departures too may be relevant, particularly if the factors usually considered important by the decisionmaker strongly favor a decision contrary to the one reached.”); cf. Dep't of Comm., 139 S. Ct. at 2576 (“If judicial review is to be more than an empty ritual, it must demand something better than the explanation offered for the action taken in this case.”).
Either way, a substantive departure is a telltale sign that the state lacks a nondiscriminatory reason for reenactment. After all, as I've noted, the government doesn't act without a purpose. And when what the legislature says it wants to do and what it actually does don't match, we can find evidence of animus. Cf. Williams v. Valdosta, 689 F.2d 964, 975 (11th Cir. 1982) (reversing summary judgment where employee was denied promotion for not passing an exam but adherence to the promotional policy was “inconsistent and arbitrary at best”). As Justice Thomas has recognized, “if a policy remains in force, without adequate justification and despite tainted roots and segregative effect, it appears clear—clear enough to presume conclusively—that the State has failed to disprove discriminatory intent.” United States v. Fordice, 505 U.S. 717, 747, 112 S.Ct. 2727, 120 L.Ed.2d 575 (1992) (Thomas, J., concurring). For this reason, when an earlier finding of discriminatory intent exists, in a reenactment challenge, the most telling aspect of the Arlington Heights factors is any “substantive departure.”
To see why, look at employment actions. Except for discrimination, a true reason for taking adverse action against an employee favorably resolves an employment-discrimination action for the employer. So why offer a false reason—unless, of course, the true reason is discrimination, which would impose liability on the employer. The same is true in the legislation context. Just as a false reason for an employment action against a member of a protected group can suggest discrimination under the Equal Protection Clause, a false reason for the enactment of a law that has a discriminatory impact can indicate pretext for discrimination. And that is especially the case when the law is the reenactment of a law that a federal court has previously expressly found to have been enacted for a discriminatory purpose.
To purge that taint, to assure citizens and courts that a legislature is not merely “taint laundering,” the reenacting legislature must identify its policy goal and justify any disparate impact, see W. Kerrel Murray, Discriminatory Taint, 135 Harv. L. Rev. 1190, 1244 (2022), as part of its showing that no impermissible substantive departures have occurred in the legislative process. Otherwise, “taint is a time-wasting paper tiger,” id. at 1264, because a state can reenact the same law that's already been struck down as discriminatory—as long as the state doesn't say it intends to discriminate. See Palmer v. Thompson, 403 U.S. 217, 225, 91 S.Ct. 1940, 29 L.Ed.2d 438 (1971) (“[T]here is an element of futility in a judicial attempt to invalidate a law because ․ [i]f the law is struck down ․ it would presumably be valid as soon as the legislature or relevant governing body repassed it for different reasons.”).12
ii. Alabama's reenactment doesn't remove the discriminatory taint that the Supreme Court found in Hunter.
With that in mind, I apply the Arlington Heights factors here. And in the light most favorable to the Plaintiffs-Appellants (as the nonmoving party), those factors reveal a genuine issue of fact about whether Alabama adopted Amendment 579 and its definitions of “moral turpitude” with discriminatory animus.
First, the Plaintiffs-Appellants have submitted evidence that the reenactment has a racially discriminatory effect. Of Alabamians convicted of felonies, 40% of Blacks and 35% of whites are disenfranchised.13 Comparatively, that means that 14% more of the Black felon population than the white felon population is disenfranchised.14
So I proceed to the Arlington Heights factors. Of those, all but two weigh against Alabama.
I begin with the second part of the fourth factor—“procedural and substantive departures”—because, as I've explained, the inquiry into substantive departures is the most revealing (and therefore the most important) in a case when a state reenacts a law that a federal court has previously established was enacted originally for discriminatory reasons.
Here, the “substantive departures” factor weighs heavily against Alabama for four reasons.
One, Alabama's only stated reason for adopting Amendment 579 is inconsistent with what Amendment 579 actually did. That makes Alabama's stated reason suspect.
Indeed, Alabama's only stated purpose for adopting Amendment 579—“strictly housekeeping”—was, at best, an inaccurate description of what the amendment actually did to Alabama's felon-disenfranchisement law. When we think of housekeeping, we think of cleaning things up—in the context of legislation, clarifying the law.
Amendment 579 did the exact opposite. Before the state adopted the amendment, Alabama's constitution disenfranchised all convicted felons. Though it did so for an impermissible and discriminatory reason, the provision was clear on its face: convicted felons knew the law said they could not vote. But after adoption of Amendment 579, no one—not even the State of Alabama—knew which convicted felons were prohibited from voting. The law was in such a state of disarray that different local registrars enforced the amendment to apply to different crimes. An amendment that throws so much uncertainty into a once-clear law cannot accurately be described as one adopted for “strictly housekeeping” reasons.
Nor does Amendment 579—at least as Alabama has construed it—seem to match with the usual reasons given for disenfranchising those convicted of felonies of moral turpitude. Criminals historically were disenfranchised because they did “damage to the ‘collective honor’ of the polis.” Aristotle, for instance, “believed that ‘criminals who break laws cannot govern themselves and hence are not fit to govern others.’ ” And Rousseau asserted that “every offender who attacks the social right becomes through his crimes a rebel and a traitor to his homeland; he ceases to be one of its members by violating its laws, and he even wages war against it.”
As for disenfranchising those convicted of felonies of moral turpitude in particular, “moral turpitude,” insofar as it has generally been defined, involves “fraud” or “fraudulent conduct,” see Jordan v. De George, 341 U.S. 223, 227–28, 71 S.Ct. 703, 95 L.Ed. 886 (1951).15 Indeed, the Supreme Court has said that, “[w]ithout exception, federal and state courts have held that a crime in which fraud is an ingredient involves moral turpitude.” Id. at 227, 71 S.Ct. 703.
But that doesn't seem to be what Alabama thinks a crime of “moral turpitude” is. Consider what's not on Alabama's list. Alabama's version of “moral turpitude” doesn't disenfranchise those who've committed voting-fraud-related felonies like deceiving an elector in preparation of her ballot,16 altering another person's ballot,17 failing to count legally cast absentee votes,18 illegally voting more than once in an election (second violation),19 willfully and intentionally signing the name of another elector in a poll book,20 bribery of public servants,21 and perjury.22 So Alabama doesn't disenfranchise those who've been convicted of election-related or government-corruption-related fraud.
This underinclusiveness “raises serious doubts about whether the government is in fact pursuing the interest” traditionally invoked for felon-disenfranchisement provisions that pertain to crimes of moral turpitude. Cf. Brown v. Ent. Merchants Ass'n, 564 U.S. 786, 802, 131 S.Ct. 2729, 180 L.Ed.2d 708 (2011); see also Fitzpatrick v. City of Atlanta, 2 F.3d 1112, 1125 (11th Cir. 1993) (“[T]his underinclusiveness shows that the City's claim that it was motivated by safety concerns is in fact pretextual.”). In other words, Alabama says that it is concerned about disenfranchising people who do collective damage to the polis's honor (including election integrity) through what have historically been viewed as fraud-related crimes. But then Alabama exempts from its felon-disenfranchisement provision those convicted of election-related fraud—which goes to the heart of the collective honor of our polis.
Let's just say it: that's really odd. And the incongruity between the traditional purpose for disenfranchising those convicted of crimes of moral turpitude and those Alabama has actually disenfranchised—or more accurately, has not disenfranchised—should make a court question whether the traditional purpose of such a provision is in fact Alabama's purpose here.
Two, Amendment 579 employs vague language—“moral turpitude”—that was selected in Amendment 579's predecessor for a discriminatory purpose and was enforced in a discriminatory way. Not coincidentally with that history, it has long been recognized that “the loose terminology of moral turpitude hampers uniformity.” Note, Crimes Involving Moral Turpitude, 43 Harv. L. Rev. 117, 121 (1930). As Justice Robert Jackson pointed out seven decades ago, the phrase was “not one which has settled significance from being words of art in the profession. If we go to the dictionaries, the last resort of the baffled judge, we learn little except that the expression is redundant, for turpitude alone means moral wickedness or depravity and moral turpitude seems to mean little more than morally immoral.” Jordan, 341 U.S. at 235, 71 S.Ct. 703 (Jackson, J., dissenting).
Using a vague term like “moral turpitude” gives local registrars, who were the ultimate arbiters of what qualified as felonies of moral turpitude in the years immediately after Amendment 579 was adopted, carte blanche to discriminate. Indeed, the government in Jordan conceded that “moral turpitude” was “a term that [was] not clearly defined,” and that “[t]he various definitions of moral turpitude provide no exact test by which we can classify the specific offenses here involved.” Id. In Justice Jackson's view, “moral turpitude” was an “undefined and undefinable standard.” Id. In other words, we know that Alabama in 1901 used the phrase “moral turpitude” for racially discriminatory reasons. Hunter, 471 U.S. at 233, 105 S.Ct. 1916. So its use of that same phrase again should trigger closer inquiry.
Alabama is of course free to use the same phrase again. But its unexplained choice to do so—especially when Alabama itself had not even settled on a definition for the term—counsels against finding the earlier impermissible intent was cleansed.
Three, and relatedly, Alabama arbitrarily enforced the amendment in the immediate aftermath of its ratification, leaving it entirely to local registrars to construe the meaning of the term “moral turpitude.” As Clay Helms, the Assistant Director of Elections and Supervisor of Voter Registration, acknowledged in his deposition, the moral-turpitude provision meant different things in different counties because local registrars made individual decisions without binding guidance from the state. As a result, the moral-turpitude provision could be interpreted so that the exact same crime could be disqualifying in one county but not another. That arbitrary enforcement of Amendment 579 undermines any conceivable legitimacy behind Alabama's reason or the traditional reason for disenfranchising those convicted of felonies of moral turpitude.
Four, Alabama's whipsaw change in definitions of “moral turpitude” further belies the notion that Alabama adopted Amendment 579 for a legitimate purpose. As the history of the provision's interpretation shows, Alabama went from leaving it up to the local registrars to construe the phrase to categorizing just 15 crimes as involving moral turpitude to including over 480—with stops along the way. Compare the Exploratory Committee's list, Doc. 215-13 ¶ 39, with the Governor's list, Doc. 269-12 at 11–12. The huge variance in Alabama's definition of felonies of “moral turpitude” over the years is perhaps unsurprising given that Alabama never had a meaning of the phrase in mind when it adopted Amendment 579 and instead treated the task of determining qualifying convictions as a complete afterthought, assigning everyone from interns to committees to tell Alabama whom it had disenfranchised. But the point is this: Alabama's process of determining what qualifies as a felony of moral turpitude was completely arbitrary. That sheer arbitrariness casts doubt on any conclusion that Alabama adopted Amendment 579 for a legitimate purpose.
In sum, on the substantive-departure inquiry, the mismatch between Alabama's stated purpose in adopting Amendment 579 and the actual effect of that Amendment (not to mention the sheer arbitrariness of how Alabama has construed Amendment 579) strongly weighs in favor of finding discriminatory intent.
Almost all the other Arlington Heights factors also weigh in favor of finding discriminatory intent.
As to the first factor—“the impact of the challenged law”—as I've already noted, Alabama's felon-disenfranchisement provision has a racially disparate effect because under it, of Alabamians convicted of felonies, 40% of Blacks are disenfranchised and 35% of whites are disenfranchised.23 So this factor weighs against Alabama.
On the second factor—“the historical background”—that also figures heavily in favor of a finding of intent here. Alabama has a long, long history of racial discrimination in general—especially in the voting context. See supra at –––– – ––––. Of course, the past is the past. But here, some of the discriminatory past is not so long ago. After all, it was less than fifteen years ago when Alabama state senators were caught on video referring to Blacks as “Aborigines” and strategizing to reduce the Black-voter turnout. Shelby Cnty., 570 U.S. at 584, 133 S.Ct. 2612 (Ginsberg, J., dissenting). So we have direct evidence of discriminatory intent by Alabama legislators not just in the distant past but even after Amendment 579 was adopted. We cannot ignore that under the Arlington Heights framework. And this factor also weighs heavily against Alabama.
I combine my consideration of the third factor—“the specific sequence of events leading up to [the law's] passage”—the “procedural ․ departures” aspect of the fourth factor, and the fifth factor—“the contemporary statements and actions of key legislators.” As I've mentioned, after its introduction on the floor, the amendment underwent no discussion in committee. Yet the day after introduction, the Standing Committee on Constitution and Elections returned it to the House with a favorable report. The full House passed the amendment without any debate or discussion. The amendment moved to the Senate, and it quickly moved through committee to the floor and passed. After it passed both houses of the Alabama Legislature, the amendment survived a vote by the Alabama voters. To be sure, the fact that the bill went through bicameralism and presentment (and was approved by the voters) reveals no “procedural ․ departures” and puts some weight on the scale in Alabama's favor.
But then again, Alabama adopted the amendment (complete with the “moral turpitude” language) with no discussion—so we can't evaluate “the contemporary statements and actions of key legislators.” And the law went through bicameralism and presentment without discussion after the unsuccessful lead-ups to the 1996 adoption, where Alabama heard evidence from the ACLU that the amendment's use of the phrase “moral turpitude” was “vague and indefinite” and appeared “unwarranted and discriminatory.” Given this objection, Alabama's historic use of the phrase “moral turpitude” to disenfranchise Blacks through Amendment 579's predecessor, and Alabama's lack of definition for “moral turpitude” at the time of Amendment 579's adoption, that the amendment endured utterly no discussion on its way to adoption is certainly curious. And in that respect, it seems to me to outweigh any weight for Alabama that bicameralism and presentment otherwise would have accorded.
As to the sixth factor—“the foreseeability of the disparate impact”—that also weighs against Alabama. For starters, Alabama had historically used the phrase “moral turpitude” to disenfranchise Black voters with Amendment 579's predecessor. Plus, the ACLU warned Alabama previously that the phrase was “discriminatory” and “vague.” Even with that historic knowledge, Alabama still chose to use the phrase “moral turpitude” without defining it—and then it assigned its local registrars to construe it. It's hard to imagine how it would have been unreasonable for Alabama to have foreseen the likelihood of the Amendment's discriminatory effect.
On the seventh factor—“knowledge of [the discriminatory] impact”—the record does not include direct evidence that legislators who voted in favor of what became Amendment 579 would have a discriminatory impact. So I do not weigh that against Alabama. Still, it's hard to understand how Alabama would not have known, given its history of using the phrase “moral turpitude” for the very purpose of disenfranchising Blacks, the ACLU's warning, and Alabama's failure to attempt to define the phrase “moral turpitude” in the Amendment. Indeed, against that background, use of the phrase “moral turpitude” almost seems designed to inflict a discriminatory impact.
On the eighth factor—the availability of less discriminatory alternatives—the Plaintiffs-Appellants do not point to any. So I weigh this factor in Alabama's favor.
In short, only one factor weighs against a finding of discriminatory intent. But several—and some heavily—support the notion that Amendment 579 flunks its Equal Protection Clause challenge. If our precedent were aligned with the Supreme Court's instructions—and common sense—we would remand for a jury to decide whether these indicators prove that the legislator acted with discriminatory intent. So if I were not confined by our precedent, I would conclude that Amendment 579 did not cleanse the discriminatory taint of Alabama's 1901 felon-disenfranchisement provision.
IV. THE FELON DISENFRANCHISMENT STATUTE VIOLATES THE EX POST FACTO CLAUSE.
I would also conclude that Alabama's felon-disenfranchisement law is an unconstitutional ex post facto punishment.24
The Ex Post Facto Clause prevents governments from imposing new criminal sanctions on conduct that has already occurred. Weaver v. Graham, 450 U.S. 24, 30, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981) (citing U.S. Const. Art. I, § 9, cl. 3). That prohibition ensures that people have “fair notice” of the consequences for breaking the law. Id.
The Ex Post Facto Clause prohibits only punishments. Id., So we begin our Ex Post Facto Clause analysis by evaluating whether a law is punitive, which turns on the legislative intent. Smith v. Doe, 538 U.S. 84, 92, 123 S.Ct. 1140, 155 L.Ed.2d 164 (2003). If the legislature expressed its intention for the statute to be punitive, then our inquiry ends. Id. But even if the legislature didn't so intend, a law may still be punitive. Id. If the statutory scheme is “so punitive either in purpose or effect,” then we deem the scheme to be criminal. Id. We use the Kennedy v. Mendoza-Martinez factors to assess whether a statutory scheme is sufficiently punitive to implicate the Ex Post Facto Clause. Id. at 97, 123 S. Ct. 1140, 1146 (citing 372 U.S. 144, 83 S.Ct. 554, 9 L.Ed.2d 644 (1963)). Here, the amendment and the statute contain no express indication that they are intended as punishment. So we must turn to the Mendoza-Martinez factors.
Before I begin, I note two assumptions my analysis makes because they don't matter to the outcome, and the Majority Opinion has opined on them. First, I assume that even though we've previously said that felon disenfranchisement is always punitive,25 those statements were dicta and therefore nonbinding. Maj. Op. at –––– – ––––. Second, I assume that Trop v. Dulles, 356 U.S. 86, 78 S.Ct. 590, 2 L.Ed.2d 630 (1958), stands for only the proposition that disenfranchisement may (or may not) be punitive. Maj. Op. at –––– – ––––. From that assumption, the Majority Opinion's conclusion that our sister circuits have misread Trop follows.
At the first step of the Ex Post Facto analysis, I also agree with the Majority Opinion that the text of Amendment 579 isn't clear on whether Alabama intended felon disenfranchisement to be civil or criminal. Maj. Op. at –––– – ––––. But that's not the end of the analysis because the Supreme Court has instructed lower courts to determine if—even though the legislature didn't affirmatively intend for a regulation to be punitive—the regulations are so severe that the law is punitive in effect. Smith, 538 U.S. at 91–92, 123 S.Ct. 1140.
For this analysis, we must consider the Mendoza-Martinez factors. These factors include the following: (1) whether the sanction is an “affirmative disability or restraint as that term is normally understood,” Hudson v. United States, 522 U.S. 93, 104, 118 S.Ct. 488, 139 L.Ed.2d 450 (1997); (2) whether the sanction was historically considered civil or criminal, Mendoza-Martinez, 372 U.S. at 168, 83 S.Ct. 554; (3) whether the sanction-triggering conduct requires scienter, id., (4) whether the law serves the traditional aims of punishment: (deterrence and retribution), id.; (5) whether the behavior to which it applies is already a crime; id.; (6) “whether the alternative purpose to which it may rationally be connected is assignable for it,” id. at 168–69, 83 S. Ct. 554; and (7) whether the law appears excessive in relation to the alternative purpose assigned, id. at 169, 83 S. Ct. 554.
I must vigorously dissent from the Majority Opinion's analysis of the Mendoza-Martinez factors because the Majority Opinion misconstrues and misapplies those factors. Maj. Op. at –––– – ––––.
The first Mendoza-Martinez factor asks whether the sanction is an “affirmative disability or restraint as that term is normally understood.” Hudson v. United States, 522 U.S. 93, 104, 118 S.Ct. 488, 139 L.Ed.2d 450 (1997). The Majority Opinion concludes that losing the ability to vote is more like occupational disbarment than imprisonment. Maj. Op. at ––––.
That conclusion defies logic. As the Supreme Court has explained, the “right to vote freely for the candidate of one's choice is of the essence of a democratic society, and any restrictions on that right strike at the heart of representative government.” Reynolds, 377 U.S. at 555, 84 S.Ct. 1362. Indeed, voting is a fundamental right. Id. And Amendment 579 and Alabama Code § 17-3-30.1 prevent citizens from affirmatively exercising their fundamental rights. On the other hand, though certainly important, “the right to practice a particular profession is not a fundamental one.” Locke v. Shore, 634 F.3d 1185, 1196 (11th Cir. 2011) (citation omitted). Because voting is a fundamental right and practicing a particular profession is not, losing the right to vote more closely resembles imprisonment than losing the right to practice a particular profession. See Simmons v. Galvin, 575 F.3d 24, 65 (1st Cir. 2009) (Torruella, J., dissenting) (“Disenfranchisement, though neither physical nor permanent, deprives U.S. citizens of a fundamental right, and as such, is undoubtedly an affirmative disability.”).
Also under the first Mendoza-Martinez factor (whether the sanction is an “affirmative disability or restraint”), we must consider the arbiter of the sanction: an agency or a jury. In Hudson, the Supreme Court relied on the fact that the business license could be revoked by an agency as evidence that occupational disbarment was civil. 522 U.S. at 103, 118 S.Ct. 488. But that consideration cuts the other way here: only a jury can find an Alabama voter guilty of a felony of moral turpitude. So I would conclude that losing the right to vote constitutes an affirmative disability or restraint.
As to the second Mendoza-Martinez factor—whether the sanction was historically considered civil or criminal, Mendoza-Martinez, 372 U.S. at 168, 83 S.Ct. 554—the Majority Opinion, faced with the overwhelming weight of history, cites to some cases from the 1800s and declares a draw. Maj. Op. at –––– – ––––. But a Maryland Supreme Court case from 1865 and an Alabama Supreme Court case from 1884 aren't probative of how history has viewed felon disenfranchisement. Instead, we should look at history in full.
Let's start pre-Founding. Ancient Greece and Rome had “infamy” laws that revoked political rights as an additional punishment for committing crimes. See Alexander Keyssar, The Right to Vote: the Contested History of Democracy in the United States 62–63 (2000); Mirjan R. Damaska, Adverse Legal Consequences of Conviction and their Removal: A Comparative Study, 59 J. Crim. L., Criminology & Police Sci. 347, 351 (1968) (explaining that the Roman Republic also employed infamy as a penalty for those convicted of crimes involving moral turpitude). Medieval continental Europe assessed “civil death” for committing crimes, and Medieval England had “attainder” laws. Id. In nineteenth-century Europe, it wasn't uncommon for penal statutes to disenfranchise felons automatically. Damaska, Adverse Legal Consequences of Conviction and their Removal, supra, at 352–53 (citing statutes from Belgium, Italy, Luxembourg, Monaco, and Spain).
Turning to this country, we have historically viewed disenfranchisement as a punitive device. In readmitting Confederate States to the Union, Congress required that states amend their constitutions to ensure universal (male) suffrage except as a punishment for felonies. See Ramirez, 418 U.S. at 48–51, 94 S.Ct. 2655. The upshot of this is that Congress recognized that felon disenfranchisement was a punitive sanction.
Indeed, the Second Circuit has described the “nearly universal use of felon disenfranchisement as a punitive device.” Muntaqim v. Coombe, 366 F.3d 102, 123 (2d Cir. 2004) (vacated en banc on other grounds). And the American Law Institute's Model Penal Code described prisoner disenfranchisement as “an integral part of the criminal law.” Model Penal Code § 306.3 (Proposed Official Draft 1962). We've even said so ourselves. In our en banc Johnson opinion, we recognized that “throughout history, criminal disenfranchisement provisions have existed as a punitive device.” 405 F.3d at 1218 n.5. When we don't cherry-pick history, it shows that felon disenfranchisement is a punitive sanction.
Even if this factor were neutral—which it isn't—the Majority Opinion's analysis reveals that this factor deserves little weight here. After all, the Majority Opinion relies on two cases from the 1860s and 1880s, when voting wasn't considered a fundamental right. Richard Briffault, The Contested Right to Vote, 100 Mich. L. Rev. 1506, 1512 (2001) (“Moreover, suffrage was a state matter, not an aspect of national citizenship. The federal Constitution largely sidestepped the question of the vote ․ Voting was in no sense a federal constitutional right.”). So even if those two cases represented the legal consensus at the time—and they don't—whether felon disenfranchisement was viewed as punitive in 1860 (when voting wasn't yet understood to be a fundamental right) cannot inform whether felon disenfranchisement is punitive today.
Turning to the third Mendoza-Martinez factor, Alabama's felon-disenfranchisement law unquestionably requires scienter. Maj. Op. at ––––. The Majority Opinion claims (without analysis or citation) that the provision doesn't have a scienter requirement. Maj. Op. at ––––. But that's just not correct. The statute disenfranchises only those who commit crimes of moral turpitude. See Ala. Code. § 17-3-30.1. How can one commit a crime of moral turpitude without a mental state? The Majority Opinion doesn't list any strict-liability-moral-turpitude crimes (nor can I think of any). Nor does the Alabama Legislature's list include any strict-liability crimes. Id. (including murder, manslaughter, kidnapping, rape, sodomy, sexual torture, terrorism, human trafficking).26
On the fourth Mendoza-Martinez factor, the felon-disenfranchisement provision serves the traditional aims of punishment: retribution and deterrence. Felon disenfranchisement is retributive because the restriction, by “deny[ing] the civic and human dignity of persons who have been convicted of doing wrong,” is “emblematic of the denunciatory function of criminal law.” Simmons, 575 F.3d at 66 (citing Bell v. Wolfish, 441 U.S. 520, 593–93 n.26, 99 S.Ct. 1861, 60 L.Ed.2d 447 (1979) (Stevens, J., dissenting)).
Losing the right to vote is also a deterrent to criminal behavior. The Majority Opinion claims (again without citation or analysis) that “[i]t is very unlikely that an individual considering whether to commit a felony would be willing to risk imprisonment but not disenfranchisement.” Maj. Op. at ––––. There are many problems with this claim. The first problem is that there is no record evidence for that assertion. And we are, after all, talking about a fundamental right. See Simmons, 575 F.3d at 65 (Torruella, J., dissenting) (“Of course, the threat of being deprived of a fundamental right will, to a certain extent, always operate to deter a rational person from engaging in unlawful conduct.”). So we should hesitate to base a decision on the Majority Opinion's suppositions in this respect.
And, even if we accepted the Majority Opinion's hypothesis, the assertion also proves too much. Try this: “[i]t is very unlikely that an individual considering whether to commit a felony would be willing to risk” nineteen years’ imprisonment but not twenty years. Under the Majority Opinion's logic, would the twentieth year of imprisonment be non-punitive because it didn't increase the deterrent effect? Of course not. Neither is disenfranchisement.
The Majority Opinion also doesn't explain the positive case for its conclusion. In other words, the Majority Opinion doesn't explain why, if felon disenfranchisement isn't retributive or a deterrent, felon disenfranchisement is instead compensatory or regulatory. How, exactly, does not letting people who have served their sentence and paid their legal financial obligations vote “compensate” their victims?
Finally, there is another reason why felon-disenfranchisement statutes are punitive and not regulatory: states can't screen for “good character” in voters as a regulatory matter. After Reynolds, states can't require “good character” for voting because voting is a fundamental right. Pamela S. Karlan, Convictions and Doubts: Retribution, Representation, and the Debate over Felon Disenfranchisement, 56 Stan. L. Rev. 1147, 1155 (2004). Given that, “it would be perverse to rely on criminal convictions as evidence that individuals lack qualities that voters are not required to have.” Id. In sum, this fourth factor—whether disenfranchisement serves the traditional aims of punishment, Mendoza-Martinez, 372 U.S. at 168, 83 S.Ct. 554, also supports a finding of punitiveness.
As to the fifth Mendoza-Martinez factor, the Majority Opinion admits that “felon disenfranchisement only sanctions behavior that is already criminal.” Maj. Op. at ––––. The Majority Opinion seems to feel that this fact favors the conclusion that a restriction is civil. Maj. Op. at –––– (citing United States v. Ursery, 518 U.S. 267, 292, 116 S.Ct. 2135, 135 L.Ed.2d 549 (1996)). It does not. The Supreme Court has explained that the fact that sanctions are linked to criminalized behavior “suggests” that the sanctions are, in fact, “criminal in nature.” United States v. One Assortment of 89 Firearms, 465 U.S. 354, 365, 104 S.Ct. 1099, 79 L.Ed.2d 361 (1984). In other words, the Majority Opinion's admission that felon disenfranchisement is linked directly to criminal behavior (such as commission of a felony) supports finding that disenfranchisement is a punitive sanction.
To be sure, the Supreme Court has given this factor little weight because “Congress may impose both a criminal and a civil sanction in respect to the same act or omission.” Helvering v. Mitchell, 303 U.S. 391, 399, 58 S.Ct. 630, 82 L.Ed. 917 (1938); 89 Firearms, 465 U.S. at 365, 104 S.Ct. 1099 (“But that indication is not as strong as it might seem at first blush.”); Ursery, 518 U.S. at 292, 116 S.Ct. 2135 (“[T]hough both statutes are tied to criminal activity, as was the case in 89 Firearms, this fact is insufficient to render the statutes punitive.”). In other words, a link between a crime and a sanction is not sufficient (without more) to show a statute is punitive. But we have more here—five of the six other Mendoza-Martinez factors favor a finding of punitiveness.
Sixth, I concede that Alabama could have a legitimate purpose for disenfranchising felons. States, even after the Fourteenth Amendment, can restrict felons from the franchise. See U.S. Const., amend XIV, § 2 (recognizing that States may exclude from the franchise those who participated in rebellion or other crime without losing representation in Congress). But see Ramirez, 418 U.S. at 72–73, 94 S.Ct. 2655 (Marshall, J., dissenting) (arguing that there is little legislative history or explanation of the meaning of “any crime” in the Fourteenth Amendment).
But seventh, that purpose is “excessive in relation to the alternate purpose assigned.” Mendoza-Martinez, 372 U.S. at 169, 83 S.Ct. 554. Alabama says that it wants to exclude those who have engaged in crimes of moral turpitude. Fair enough. But if that's true, why doesn't Alabama want to exclude those convicted of bribery of public servants 27 ; perjury 28 ; deceiving an elector in preparation of her ballot 29 ; altering another person's ballot 30 ; failing to count legally cast absentee votes 31 ; illegally voting more than once in an election (second violation)32 ; and willfully and intentionally signing the name of another elector in a poll book?33 Of course, it isn't my place to tell Alabama which felons to exclude from the franchise (or who not to exclude). But it is my job to point out when Alabama is drawing lines at odds with its stated purpose—lines that suggest its purpose might be something else. Cf. Weinberger v. Wiesenfeld, 420 U.S. 636, 648, 95 S.Ct. 1225, 43 L.Ed.2d 514 (1975) (“But the mere recitation of a benign, compensatory purpose is not an automatic shield which protects against any inquiry into the actual purposes underlying a statutory scheme.”). The mismatch between Alabama's stated aim and its method suggests that the restriction may be punitive.
To recap, then, six of the seven factors support finding that Alabama's felon-disenfranchisement statute is punitive.
This conclusion—that the felon-disenfranchisement statute is punitive—means that the Ex Post Facto Clause prohibits application of the statute to the Plaintiffs-Appellants. That is so because the statute took effect after the Plaintiffs-Appellants committed their crimes, and the Plaintiffs-Appellants committed their crimes after Alabama adopted Amendment 579, which made Alabama's disenfranchisement provision applicable to only a then-undefined group of crimes—felonies of “moral turpitude.” Simply put, because the Plaintiffs-Appellants didn't know that they would lose their right to vote when they committed their crimes, it is unconstitutional to retroactively strip them of that right. Lynce v. Mathis, 519 U.S. 433, 449, 117 S.Ct. 891, 137 L.Ed.2d 63 (1997). I therefore disagree with the Majority Opinion's conclusion to the contrary.
V. ALABAMA'S VOTER REGISTRATION FORM VIOLATES THE NATIONAL VOTER REGISTRATION ACT
Finally, I would also reverse as to the National Voter Registration Act (“NVRA”) claim. As I see it, the Majority Opinion ignores the text of the statute and the Plaintiffs-Appellants’ actual arguments. Maj. Op. at –––– – ––––.
The NVRA requires that mail voter-registration forms “shall” include a statement that “specifies each eligibility requirement (including citizenship).” 52 U.S.C. § 20508(b)(2)(A) (emphasis added). To “specify” something means “to name or state explicitly or in detail.” Specify, Merriam-Webster Online Dictionary (https://www.merriamwebster.com/dictionary/specify (last accessed Apr. 20, 2023). Alabama's voter-eligibility requirements mandate that the voter not be “convicted of a felony involving moral turpitude” unless his or her civil and political rights were restored. Ala. Const., Art. VIII, § 177(b). Alabama's voter form then, must “specify” what it means to be convicted of a felony involving moral turpitude.
But Alabama's form does not “name or state explicitly in detail” what it means to have been convicted of a felony involving moral turpitude. Rather, Alabama's form says only that a prospective voter can register by swearing—under penalty of perjury—that he or she is “not barred by reason of a disqualifying felony conviction. (The list of disqualifying felonies is available on the Secretary of State's web site[.] [URL omitted]).” A form that identifies a requirement but doesn't give any information about that requirement can't be said to state it “explicitly in detail.” For instance, the form doesn't even mention the requirement that the felony be one of moral turpitude. Nor does it have a list of the felony convictions that are disqualifying. The form provides only a website link. The form is problematic both because not everyone has access to the Internet but also because a website link can't possibly be said to “state explicitly in detail” the disqualifying felonies. And it doesn't take much imagination to conclude that the form's overly general description of who can't vote discourages some who are eligible to vote from voting because, unsure of their eligibility based on the form's instructions, they do not want to take the chance of violating the law.
The Majority Opinion concludes that the website URL is sufficiently specific. But it never explains why a URL link is sufficiently specific and never grapples with the text of the statute. In fact, the Majority Opinion claims that courts should avoid picking out individual words to discern an entire statute's meaning. Maj. Op. at ––––. While I agree with that approach, it doesn't give us license to ignore the text entirely. See Van Buren v. United States, ––– U.S. ––––, 141 S. Ct. 1648, 1654, 210 L.Ed.2d 26 (2021) (“[W]e start where we always do: with the text of the statute.”). Here, the Majority never analyzes the only verb in the section of the statute—the only thing that Alabama must do. How can we know if Alabama has performed its required duty without so much as mentioning what that duty—as codified in the verb “specify”—is? Nor am I construing the meaning of an entire statute based on one word: I'm ensuring that the one word—enacted by Congress and signed by the President—is given its proper meaning. See id. (discussing the meaning of “so” in the phrase “entitled so to obtain.”).
Finally, the Majority Opinion sets up and knocks down a strawman version of the Plaintiffs-Appellants’ argument. Contrary to the Majority Opinion's (mis)-characterization, the Plaintiffs-Appellants aren't asking Alabama to list “every state, federal, and foreign felony involving moral turpitude.” Maj. Op. at –––– – –––– (emphasis added). They just want the list of crimes that Alabama says qualify—the forty-seven enumerated crimes and the catchall provision.
And contrary to the Majority Opinion's suggestion, the Plaintiffs-Appellants’ position isn't internally inconsistent. Maj. Op. at –––– – ––––. The Plaintiffs-Appellants just want more specificity than a generic description and a website URL and don't want (and have never asked for) a constantly shifting list of federal and foreign crimes of moral turpitude that are irrelevant to whether they can vote. Asking for a middle position isn't inconsistent—it's moderate.
Ultimately, I agree with the Majority Opinion that Alabama's reenactment of its felon-disenfranchisement provision satisfies our precedent's required procedure to remove discriminatory taint from an earlier statute. But in my view, our precedent imposes a meaningless test to ascertain whether a reenacted version of a law that a federal court has previously found to have been enacted for a discriminatory reason violates the Equal Protection Clause. Because our precedent requires us to apply that meaningless test, though, I am constrained to concur in affirming the district court's judgment.
But I dissent from the Majority Opinion's holding that Alabama's felon-disenfranchisement statute is civil and therefore isn't subject to the Ex Post Facto Clause. In fact, the statute is punitive and therefore can be applied only prospectively. I also dissent from the Majority Opinion's holding on the National Voter Registration Act claim because Alabama's voter-registration form doesn't “specify” which individuals convicted of felonies cannot vote and therefore violates federal law.
1. This case was initially filed as a putative class action with ten named defendants. The District Court denied class certification, and on appeal only two individual plaintiffs remain: (1) Treva Thompson, a black woman convicted of theft of property in the first degree, and (2) Timothy Lanier, a black man convicted of attempted murder and two counts of burglary in the first degree. Ala. Code § 17-3-30.1(c) specifies that each of these felonies involve moral turpitude.
2. Section 182 disenfranchised the following individuals:All idiots and insane persons; those who shall by reason of conviction of crime be disqualified from voting at the time of the ratification of this Constitution; those who shall be convicted of treason, murder, arson, embezzlement, malfeasance in office, larceny, receiving stolen property, obtaining property or money under false pretenses, perjury, subornation of perjury, robbery, assault with intent to rob, burglary, forgery, bribery, assault and battery on the wife, bigamy, living in adultery, sodomy, incest, rape, miscegenation, crime against nature, or any crime punishable by imprisonment in the penitentiary, or of any infamous crime or crime involving moral turpitude; also, any person who shall be convicted as a vagrant or tramp, or of selling or offering to sell his vote or the vote of another, or of buying or offering to buy the vote of another, or of making or offering to make a false return in any election by the people or in any primary election to procure the nomination or election of any person to any office, or of suborning any witness or registrar to secure the registration of any person as an elector.Ala. Const. art. VIII, § 182.
3. Section 10304(a) of the Voting Rights Act of 1965 requires certain covered states and counties, which in 1996 included Alabama, to receive preclearance from a three-judge panel before implementing any changes to the state or county's voting requirements unless the Attorney General declined to object to the change within sixty days. In 2013, the Supreme Court struck down the coverage formula used in § 10304(a), 52 U.S.C. § 10303(b), in Shelby County, Ala. v. Holder, 570 U.S. 529, 133 S. Ct. 2612, 186 L.Ed.2d 651 (2013), rendering § 10304(a)’s preclearance requirement inoperative.
4. This statement was not entirely correct. Section 182 prohibited persons convicted of all felonies from voting, not just felonies involving moral turpitude. Accordingly, Amendment 579 expanded Alabama's franchise to those convicted of felonies not involving moral turpitude.
5. Appellants also contend that the District Court applied the wrong legal standard for summary judgment, which we address infra in Part II.
6. Appellants did make this argument in the District Court. Thompson, 505 F. Supp. 3d at 1259–61. The Court found that Appellants had presented insufficient evidence of discriminatory intent to survive summary judgment under the Arlington Heights factors. Id. Appellants have forfeited this issue on appeal by not raising it in their opening brief, so we decline to address it. United States v. Campbell, 26 F.4th 860, 873 (11th Cir. 2022) (en banc).
7. As Johnson was an appeal from summary judgment, we assumed without deciding that the 1868 Florida constitution's felon disenfranchisement provision was racially motivated. Johnson, 405 F.3d at 1223.
8. Because the Jones I panel held that Bearden and Griffin applied whenever a state “alleviate[s] punishment for some, but mandates that punishment continue for others, solely on account of wealth,” its determination that disenfranchisement constituted punishment was essential for its holding. See Jones I, 950 F.3d at 817–820.
9. Harmelin v. Michigan, 501 U.S. 957, 983, 111 S. Ct. 2680, 2695, 115 L.Ed.2d 836 (1991) (Scalia, J., separate opinion).
10. Muntaqim v. Coombe, 366 F.3d 102, 123 (2d Cir. 2004), vacated en banc, 449 F.3d 371 (2d Cir. 2006).
11. Pamela A. Wilkins, The Mark of Cain: Disenfranchised Felons and the Constitutional No Man's Land, 56 Syracuse L. Rev. 85, 133–34 (2005); Note, One Person, No Vote: The Laws of Felon Disenfranchisement, 115 Harv. L. Rev. 1939, 1939–42 (2002).
12. Jones II also briefly referred to disenfranchisement as punishment in a later section of the opinion analyzing whether Florida's re-enfranchisement scheme was an unconstitutional poll tax. 975 F.3d at 1039 (“Some punishments, like disenfranchisement, are imposed on all felons alike regardless of the severity of their crimes.”). This reference, again without analysis or citation, is also dicta because it was not a holding of Jones II nor necessary to the en banc Court's holdings. Gillis, 938 F.3d at 1198.
13. Appellants argue in a footnote that “[i]ncluding a link to the State's website does not satisfy the NVRA” as “[r]egistrants at a motor vehicle or other government agency may not have access to the Internet, defeating the purpose of promoting on-site registration.” But § 20508(b)(2)(A) governs state mail voting registration forms, not on-site voting registration forms. Registrants may thus access Alabama's website for the list of disqualifying felonies at their leisure.
1. Ala. Code § 17-17-19.
2. Ala. Codev § 17-17-24(a).
3. Ala. Code § 17-17-27.
4. Ala. Code § 17-17-36.
5. Ala. Code § 17-17-15.
6. Ala. Code § 13A-10-61.
7. Ala. Code § 13A-10-101.
8. See Vill. of Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252, 97 S.Ct. 555, 50 L.Ed.2d 450 (1977).
9. From 1871 through 1873, in the 42nd Congress, Congressman Benjamin Sterling Turner represented Alabama. https://history.house.gov/Exhibitions-and-Publications/BAIC/Historical-Data/Black-American-Representatives-and-Senators-by-State-and-Territory/ (last visited Apr. 20, 2023). Congressman James Thomas Rapier represented the state from 1873 through 1875, in the 43rd Congress. Id. And Congressman Jeremiah Haralson had the honor from 1875 through 1877 in the 44th Congress. Id. Alabama didn't elect another Black man or woman to Congress for almost 120 years. Id.
10. The Compromise of 1877 resolved the disputed 1876 Presidential election. Nathan Colvin & Edward Foley, The Twelfth Amendment: A Constitutional Ticking Time Bomb, 64 U. Miami L. Rev. 475, 502 (2010). In the 1876 Presidential election, Rutherford B. Hayes, the Republican nominee, ran against Samuel Tilden, the Democratic nominee. Id. At the time, 185 votes would clinch an Electoral College majority. But on election night, Tilden had 184 votes and Hayes had 165. Chris Land & David Schultz, On the Unenforceability of the Electoral Count Act, 13 Rutgers J. Law & Pub. Pol'y 340, 350–51 (2016). Twenty votes remained outstanding because, in three states (including Florida, which shows that history does indeed repeat itself), both political parties’ electors said they were the winners. Id. at 351. Congress created an electoral commission to resolve the dispute. Id. Eventually, the commission awarded all twenty disputed electoral votes to Hayes, so Hayes won by one vote (185 to 184). Id. But Southern Democrats would not acknowledge Hayes as the President unless Republicans met certain demands, including giving Democrats in southern states the ability to legislate about Black citizens without northern interference. McConnell, The Forgotten Constitutional Moment, infra at 130–31.
11. The record indicates that he was a college student at the time, given his testimony that he graduated college in 2015.
12. Cf. Ne. Fla. Chapter of Associated Gen. Contractors v. Jacksonville, 508 U.S. 656, 662, 113 S.Ct. 2297, 124 L.Ed.2d 586 (1993) (refusing to deem a case moot if defendant repealed a challenged statute and replaced “it with one that differ[ed] only in some insignificant respect”).
13. Because the history of Alabama's adoption and application of the constitutional amendment shows that the amendment's reference to felony crimes of “moral turpitude” is meaningless in the absence of the statute, I use the statutory definition interchangeably with the amendment's reference to felonies of “moral turpitude.”
14. 40 divided by 35 is 1.14. See Hunter, 471 U.S. at 227, 105 S.Ct. 1916 (noting that the Eleventh Circuit there found that Amendment 579's predecessor made it “at least 1.7 times as likely [that Blacks would] suffer disenfranchisement” as whites).
15. The term “moral turpitude” was first used in an 1891 immigration statute accompanied without comment or debate. Brian C. Harms, Redefining “Crimes of Moral Turpitude”: A Proposal to Congress, 15 Geo. Imm. L.J. 259, 262 (2001) (“The lack of legislative history makes it impossible to know whether Congress considered the term to be a new criterion, or ․ merely a synthesis of previously recognized distinctions.”) (cleaned up).
16. Ala. Code § 17-17-19.
17. Ala. Code§ 17-17-24(a).
18. Ala. Code § 17-17-27.
19. Ala. Code § 17-17-36.
20. Ala. Code § 17-17-15.
21. Ala. Code § 13A-10-61.
22. Ala. Code § 13A-10-101.
23. Consideration of a law's disparate impact occurs twice in an Equal Protection Clause analysis—once on its own as a requirement and again as part of the factor-based inquiry into intent. Of course, the analysis is the same here.
24. The Plaintiffs-Appellants also challenge the 1996 amendment—not just the 2017 statute—under the Ex Post Facto Clause. As far as I can tell, only one plaintiff was convicted of a crime before 1996: a woman who was convicted of first-degree murder in 1995. But those with murder convictions were disqualified from voting both before and after 1996, so there is no ex post facto problem with her disqualification. Compare Ala. Const. § 182 (1901), with Ala. Const. § 177(b) (1996).
25. Johnson, 405 F.3d at 1218 n.5, 1228; Jones v. Gov. of Fla. (“Jones I”), 950 F.3d 795 (11th Cir. 2020); and Jones v. Gov. of Fla (“Jones II”), 975 F.3d 1016 (11th Cir. 2020) (en banc).
26. To be fair, the statute includes “possession” crimes, like possessing a bacteriological or biological weapon. Id. § 17-3-30.1 (citing id. § 13A-10-193). I'm unaware of any authority as to whether those crimes require scienter. But I presume that they do because the general rule in Alabama is that “all criminal statutes must contain the element of scienter.” McCrary v. State, 429 So. 2d 1121, 1124 (Ala. Crim. Ct. App. 1982). Given that general statement of law, the Majority Opinion's claim that the listed crimes don't require scienter cannot be right.
27. Ala. Code § 13A-10-61.
28. Ala. Code § 13A-10-101.
29. Ala. Code § 17-17-19.
30. Ala. Code § 17-17-24(a).
31. Ala. Code § 17-17-27).
32. Ala. Code § 17-17-36.
33. Ala. Code § 17-17-15.
Tjoflat, Circuit Judge:
Was this helpful?
Get help with your legal needs
Search our directory by legal issue
Enter information in one or both fields (Required)