KELVIN LEON JONES, BONNIE RAYSOR, et al., Plaintiffs–Appellees, versus GOVERNOR OF FLORIDA, FLORIDA SECRETARY OF STATE, Defendants–Appellants.
No. 19-14551
United States Court of Appeals, Eleventh Circuit
February 19, 2020
D.C. Docket No. 4:19-cv-00300-RH-MJF. [PUBLISH]
(February 19, 2020)
Before ANDERSON and MARCUS, Circuit Judges, and ROTHSTEIN,* District Judge.
On November 6, 2018, Florida voters approved Amendment 4, a state constitutional amendment that automatically restored voting rights to ex-felons who had completed all of the terms of their sentences. Contemporary media reports suggested that as many as 1.4 million felons could be eligible for re-enfranchisement under the law. Accounts differed as to whether this figure made Amendment 4 the single largest act of enfranchisement since the Nineteenth Amendment in 1920, the Voting Rights Act in 1965, or the Twenty-Sixth Amendment in 1971. By any measure, Amendment 4‘s enfranchisement was historic.
Amendment 4 provided that a felon‘s “voting rights shall be restored upon completion of all terms of sentence including parole or probation.” Following its passage, the Florida legislature passed Senate Bill 7066, which implemented the Amendment and interpreted its language to require payment of all fines, fees and restitution imposed as part of the sentence (collectively, “legal financial obligations” or “LFOs“). The Florida Supreme Court later agreed with the legislature‘s interpretation of the Amendment—during the pendency of this appeal, it held that the plain text of Amendment 4 requires payment of all LFOs as a precondition of re-enfranchisement.
Following the passage of SB 7066, the seventeen plaintiffs in this case brought suit, challenging the constitutionality of the LFO requirement. Each plaintiff is a felon who has alleged that he or she would be eligible for re-enfranchisement under Amendment 4 but for non-payment of outstanding LFOs. Each plaintiff has also alleged that he or she is indigent and, therefore, genuinely unable to pay those obligations.
The cases were consolidated in the United States District Court for the Northern District of Florida, which then issued a preliminary injunction requiring the State to allow the named plaintiffs to register and vote if they are able to show that they are genuinely unable to pay their LFOs and would otherwise be eligible to vote under Amendment 4. From this order the State timely appealed to this Court.
Because the LFO requirement punishes those who cannot pay more harshly than those who can—and does so by continuing to deny them access to the ballot box—Supreme Court precedent leads us to apply heightened scrutiny in asking whether the requirement violates the Equal Protection Clause of the Fourteenth Amendment as applied to these plaintiffs. When measured against this standard, we hold that it does and affirm the preliminary injunction entered by the district court.
I. BACKGROUND
Florida has a long history of disenfranchising those who commit serious crimes, a common practice nationwide that dates to the very beginning of the republic. See George Brooks, Comment, Felon Disenfranchisement: Law, History, Policy, and Politics, 32 Fordham Urb. L.J. 851, 852–53 (2005) (“The first disenfranchisement laws in America appeared in the 1600s . . . and were present from the earliest times of the Republic.“); see also Richardson v. Ramirez, 418 U.S. 24, 48 (1974) (noting that, in 1868, 29 states had constitutional provisions authorizing the disenfranchisement of felons); Johnson v. Governor of Fla., 405 F.3d 1214, 1218 (11th Cir. 2005) (en banc) (“Florida‘s policy of criminal disenfranchisement has a long history . . . .“). Indeed, Florida‘s Constitution has authorized the disenfranchisement of felons since before it joined the Union. See
The contemporary voters of Florida, however, are not alone in finding the longstanding policy of categorically depriving felons of voting rights increasingly unpalatable. In the past two decades, nearly half of the states have in some way expanded felons’ access to the franchise.1 While once commonplace, as best as we can tell, only one state maintains a policy of disenfranchising all felons
permanently absent executive clemency,2 although nearly every state continues to disenfranchise felons in some way.3
Regardless of the political trend toward re-enfranchisement, there is nothing unconstitutional about disenfranchising felons—even all felons, even for life. See Richardson, 418 U.S. at 56 (holding that the lifelong disenfranchisement of felons does not violate the Equal Protection Clause). In Richardson, the Supreme Court found in § 2 of the Fourteenth Amendment—which decreases the population of a state that is counted for apportionment purposes when it disenfranchises any male citizens over twenty-one ”except for participation in rebellion, or other crime,”
the United States of the right to vote . . . except as a punishment for such crimes as are now felonies at common law.” Act of June 25, 1868, ch. 50, 15 Stat. 73, 73 (emphasis added).
Prior to the adoption of Amendment 4, all felons in Florida were presumptively disenfranchised for life. See
Thus, for example, in Johnson the plaintiffs alleged that Florida‘s rights-restoration process ran afoul of the Fourteenth Amendment‘s Equal Protection Clause because, though facially neutral, it intentionally discriminated on the basis of race against African-Americans and other voters of color. 405 F.3d at 1217. Sitting en banc, we rejected these claims, determining that the State “enact[ed] the provision [in 1968] without an impermissible motive.” Id. at 1224. Similarly, in Hand v. Scott, we reversed a permanent injunction enjoining the State‘s Executive Clemency Board from enforcing the restoration process, holding that the process
did not have “a discriminatory purpose or effect” with respect to race. 888 F.3d 1206, 1207 (11th Cir. 2018) (emphasis in original).
As a general matter, the Florida Constitution may be amended by a referendum in which sixty percent of the voters agree on the text of an amendment proposed by citizen initiative. See
(a) No person convicted of a felony, or adjudicated in this or any other state to be mentally incompetent, shall be qualified to vote or hold office until restoration of civil rights or removal of disability. Except as provided in subsection (b) of this section, any disqualification from voting arising from a felony conviction shall terminate and voting rights shall be restored upon completion of all terms of sentence including parole or probation.
(b) No person convicted of murder or a felony sexual offense shall be qualified to vote until restoration of civil rights.
In May 2019, the Florida legislature implemented the provisions of Amendment 4 with Senate Bill 7066 (“SB 7066“), which was codified at
SB 7066 further provided that these financial obligations would be deemed completed by payment of the obligation, approval of discharge of the obligation by the person to whom it is owed, completion of community service hours if a court converts the obligation into service hours, or modification of the original sentencing document by a court to remove the obligation.
Finally, SB 7066 established a Restoration of Voting Rights Work Group to study the implementation of Amendment 4.
Pursuant to his statutory authority, see
Amend. 4, The Voting Restoration Amendment, 45 Fla. L. Weekly S10 (2020). Because Florida‘s Supreme Court is the final arbiter of state law, this adjudication conclusively resolves whether SB 7066 properly interpreted the scope of Amendment 4. See Guideone Elite Ins. Co. v. Old Cutler Presbyterian Church, Inc., 420 F.3d 1317, 1326 n.5 (11th Cir. 2005) (“Our objective is to determine issues of state law as we believe the Florida Supreme Court would . . . .“).
II. PROCEDURAL HISTORY
On June 28, 2019, Kelvin Jones sued the Governor and Secretary of State of Florida, as well as the Supervisor of Elections of the county in which he resides, pursuant to
allow him to register to vote, or to bar them from revoking his registration if he qualified to vote without consideration of his outstanding legal financial obligations.
On June 30, 2019, the district court ordered the Jones case consolidated with other cases that had raised similar challenges to Amendment 4‘s LFO requirement. Each of the seventeen named plaintiffs in this consolidated case alleged that he or she would be entitled to vote under Amendment 4 and SB 7066, but for the outstanding LFOs that they are unable to pay. Fifteen of the seventeen named plaintiffs have submitted affidavits attesting to their inability to pay. Two plaintiffs did not, but rather alleged in their complaint that they were unable to pay their outstanding legal financial obligations.
In August 2019, the plaintiffs jointly moved for a preliminary injunction enjoining the enforcement of the portions of the law that required the payment of LFOs as a condition precedent to accessing the ballot. In September, the plaintiffs also moved to certify classes for their Twenty-Fourth Amendment and wealth discrimination equal protection claims. They proposed a broad class for their Twenty-Fourth Amendment poll tax claim, defined as: “All persons otherwise eligible to vote in Florida who are denied the right to vote because they have outstanding LFOs.” As for their wealth discrimination equal protection claim, the plaintiffs proposed to define the subclass as: “All persons otherwise
in Florida who are denied the right to vote solely because they are genuinely unable to pay their outstanding LFOs.” The district court has not yet ruled on class certification.
The district court held an evidentiary hearing on the preliminary injunction on October 7–8, 2019, and, on October 18, 2019, it granted the plaintiffs’ motion for preliminary injunctive relief in part. The court found that the plaintiffs had shown a substantial likelihood of success on the merits of their equal protection claim—that Amendment 4‘s LFO requirement, as applied to those who genuinely could not pay, constituted wealth discrimination in violation of the Fourteenth Amendment. The district court apparently applied strict scrutiny and concluded that the requirement could not pass constitutional muster as applied. Finding the other prerequisites of equitable relief satisfied, the district court issued a preliminary injunction. The district court declined to rule on the plaintiffs’ First Amendment, Due Process Clause or Twenty-Fourth Amendment claims.
The district court carefully circumscribed the preliminary injunction. It enjoined the defendants (other than the Governor and the Supervisor of Orange County) from preventing any of the seventeen plaintiffs from registering to vote based solely on an inability to pay outstanding legal financial obligations, where each plaintiff asserts that he or she is genuinely unable to pay. It further enjoined
the same defendants from preventing the plaintiffs from actually voting if indeed they could establish that they are unable to pay.
Moreover, the injunction specifically provided that it “does not prevent the Secretary from notifying the appropriate Supervisor of Elections that a plaintiff has an unpaid financial obligation that will make the plaintiff ineligible to vote unless the plaintiff shows that the plaintiff is genuinely unable to pay the financial obligation.” Nothing in the preliminary injunction precludes the State from requiring additional proof of a plaintiff‘s inability to pay beyond what had been offered in district court. Notably, the court did not order the State to follow any specific procedure or adopt any regime in complying with the injunction. Rather, it suggested that the State‘s existing voter registration procedure could provide an appropriate process. Moreover, the court did not define the term “genuine inability to pay,” again leaving it to the State to make a reasonable, good faith determination of how it could implement that term consistent with the court‘s order. In short, the court‘s order left the State with substantial discretion in choosing how to comply.
The Governor and Secretary timely appealed to us on November 15, 2019. A district court‘s grant of a preliminary injunction is an appealable interlocutory order over which we have jurisdiction.
1248 (11th Cir. 2016). A two-week trial in this case is currently scheduled for April 6, 2020, which will allow for the complete development of a factual record.
On December 12, 2019, this Court asked the parties to address whether the Governor has standing to appeal the district court‘s order, since the order enjoined all defendants “other than the Governor and Supervisor [of Elections] of Orange County.” We carried this issue with the case. We agree with all of the parties that regardless of whether the Governor has standing, a question on which we do not rule, the Secretary of State clearly has
On December 19, 2019, on motion from the Secretary of State, the district court partially stayed the preliminary injunction. It stayed only the portion of its ruling enjoining the defendants from preventing the plaintiffs from actually voting;
the court declined, however, to stay its order enjoining the defendants from preventing the seventeen plaintiffs from registering to vote. The district court provided that its partial stay would expire either when this Court ruled on the propriety of its preliminary injunction or on February 11, 2020, whichever occurs first.7
III. ANALYSIS
A preliminary injunction is an “extraordinary remedy.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24 (2008). A district court may grant preliminary injunctive relief only when a party establishes each of four separate requirements:
(1) it has a substantial likelihood of success on the merits; (2) irreparable injury will be suffered unless the injunction issues; (3) the threatened injury to the movant outweighs whatever damage the proposed injunction may cause the opposing party; and (4) if issued, the injunction would not be adverse to the public interest.
Siegel v. LePore, 234 F.3d 1163, 1176 (11th Cir. 2000) (en banc).
We review a district court‘s grant of preliminary injunctive relief for abuse of discretion. See Democratic Exec. Comm. of Fla. v. Lee, 915 F.3d 1312, 1317 (11th Cir. 2019); see also Carillon Imps., Ltd. v. Frank Pesce Int‘l Grp., Ltd., 112 F.3d 1125, 1126 (11th Cir. 1997) (“The review of a district court‘s decision to grant or deny a preliminary injunction is extremely narrow in scope.“). In so
doing, we review the district court‘s underlying legal conclusions de novo and its findings of fact for clear error. See Democratic Exec. Comm., 915 F.3d at 1317. This deferential standard follows from “[t]he expedited nature of preliminary injunction proceedings,” in which “judgments . . . about the viability of a plaintiff‘s claims and the balancing of equities and the public interest . . . are the district court‘s to make.” BellSouth Telecomms., Inc. v. MCIMetro Access Transmission Servs., LLC, 425 F.3d 964, 968 (11th Cir. 2005) (quoting Cumulus Media, Inc. v. Clear Channel Commc‘ns, Inc., 304 F.3d 1167, 1171 (11th Cir. 2002)).
The plaintiffs claim that Amendment 4‘s requirement that they pay their LFOs as a precondition to voting, as applied to them and to the extent that they are genuinely unable to pay, violates the
correct. Cf. GeorgiaCarry.Org, Inc. v. U.S. Army Corps of Eng‘rs, 788 F.3d 1318, 1329 (11th Cir. 2015) (noting that if the plaintiffs “have not shown a substantial likelihood of success on the merits, we need not consider the remaining factors in the preliminary injunction test“).
Supreme Court precedent leads us to apply some form of heightened scrutiny in our consideration of the constitutionality of the LFO requirement, as applied to these plaintiffs. It is undeniable that the LFO requirement punishes those who cannot pay more harshly than those who can. Denying access to the franchise to those genuinely unable to pay solely on account of wealth does not survive heightened scrutiny. Because we further hold that the district court did not abuse its considerable discretion in balancing the equitable factors for a
preliminary injunction, and because under Florida law the unconstitutional application of the LFO requirement is easily severable from the remainder of Amendment 4, we affirm the district court‘s preliminary injunction.
A. Likelihood of Success on the Merits
The Constitution guarantees that “[n]o State shall . . . deny to any person within
We look through a different prism, however, when the law classifies in certain suspect ways—prototypically, on the basis of race, gender, or national origin—or classifies in a way that burdens fundamental rights. Such classifications are subject to heightened scrutiny, a more exacting form of review. See, e.g., Loving v. Virginia, 388 U.S. 1, 11 (1967) (applying heightened scrutiny to race
classification); Craig v. Boren, 429 U.S. 190, 197–99 (1976) (applying heightened scrutiny to gender classification); Harper v. Va. State Bd. of Elections, 383 U.S. 663, 670 (1966) (applying heightened scrutiny to laws that limit access to the franchise on the basis of wealth).
The first and most critical question we face is deciding what lens we ought to look through in analyzing Florida‘s felon re-enfranchisement scheme. The parties disagree about the appropriate standard of scrutiny that we should employ. The plaintiffs urge a heightened form of scrutiny employing the framework the Supreme Court used in Bearden v. Georgia, 461 U.S. 660 (1983); the defendants argue that rational basis review applies.
The appropriate level of scrutiny is not immediately obvious. This case does not neatly fit the traditional categories that call for heightened scrutiny—those making classifications affecting fundamental rights or suspect classes. Wealth, the Supreme Court has repeatedly held, is not a suspect classification. See San Antonio Indep. Sch. Dist. v. Rodriguez, 411 U.S. 1, 29 (1973); Ortwein v. Schwab, 410 U.S. 656, 660 (1973). And felons may be stripped of their right to vote permanently. See Richardson, 418 U.S. at 56. On the other hand, it is not obvious that rational basis review is appropriate. In M.L.B. v. S.L.J., the Supreme Court instructed us that wealth classifications are subject to heightened scrutiny in two circumstances—where they are used to restrict access to the franchise and in the
administration of criminal justice—both of which are plainly implicated by Amendment 4 and SB 7066. 519 U.S. 102, 124 (1996). Moreover, the Supreme Court has routinely referred to the right to vote as “fundamental.” See, e.g., Reynolds v. Sims, 377 U.S. 533, 561–62 (1964).
The only two courts to face this precise claim—wealth discrimination in automatic felon re-enfranchisement schemes that, as a practical matter, deny indigent felons access to the franchise—concluded that rational basis scrutiny applied because felons do not have a fundamental right to vote and wealth is not a suspect classification. See Johnson v. Bredesen, 624 F.3d 742 (6th Cir. 2010); Madison v. State, 163 P.3d 757 (Wash. 2007).
We
1. Rational Basis Scrutiny
Under rational basis review, a law must be rationally related to a legitimate governmental interest and it “must be upheld against equal protection challenge if there is any reasonably conceivable state of facts that could provide a rational basis for the classification” between persons. F.C.C. v. Beach Commc‘ns, Inc., 508 U.S. 307, 313 (1993). Traditional rational basis review is highly deferential to government action, although it is not “toothless.” Schweiker v. Wilson, 450 U.S. 221, 234 (1981) (quoting Mathews v. Lucas, 427 U.S. 495, 510 (1976)). When we review a statute for rationality, generally we ask whether there is any rational basis for the law, even if the government‘s proffered explanation is irrational, and even if it fails to offer any explanation at all. “[W]e will not overturn such a statute unless the varying treatment of different groups or persons is so unrelated to the achievement of any combination of legitimate purposes that we can only conclude that the legislature‘s actions were irrational.” Vance v. Bradley, 440 U.S. 93, 97 (1979).
The question we ask is whether the legislature could have conceived of a rational basis for the classification it drew. Indeed, “a legislative choice is not subject to courtroom factfinding and may be based on rational speculation unsupported by evidence or empirical data.” Beach Commc‘ns, Inc., 508 U.S. at 315. In upholding a statute that set a mandatory retirement age of 60 for foreign service officers, for instance, the Supreme Court explained that although “empirical proof that health and energy tend to decline somewhat by age 60” would have been powerful evidence for upholding the statute, those challenging a legislative judgment must show not only that the facts do not support the classification, but even more onerously that “the legislative facts on which the classification is apparently based could not reasonably be conceived to be true by the governmental decisionmaker.” Bradley, 440 U.S. at 110–11.
Moreover, the burden of proof is placed squarely on the party challenging the lawfulness of the legislative classification. Statutes that do not draw lines based on suspect classes and do not burden fundamental rights “come[] to us bearing a strong presumption of validity, and those attacking the rationality of the legislative classification have the burden ‘to negative every conceivable basis which might support it.‘” Beach Commc‘ns, Inc., 508 U.S. at 314–15 (quoting Lehnhausen v. Lake Shore Auto Parts Co., 410 U.S. 356, 364 (1973)) (citation omitted).
Proceeding to the merits, we note that the parties disagree—not only about the choice between rational basis review and heightened scrutiny—but also about the proper focus of rational basis review. The
i. As applied to these plaintiffs, the LFO requirement is irrational
If the question on rational basis review were simply whether the LFO requirement is rational as applied to those unable to pay, we think it is clearly not. The continued disenfranchisement of felons who are genuinely unable to pay LFOs and who have made a good faith effort to do so, does not further any legitimate state interest that we can discern.
The irrationality of the LFO requirement as applied is apparent when we consider the State‘s interests. The longstanding policy of felon disenfranchisement has been justified on two grounds, suggesting what those interests could be: (1) punishment for those who have breached the social contract by committing crimes, and (2) shielding the ballot box from those who have manifested antagonism to society‘s laws. See Brooks, Felon Disenfranchisement, 32 Fordham Urb. L.J. at 852–54; Note, The Disenfranchisement of Ex-Felons: Citizenship, Criminality, and ‘the Purity of the Ballot Box,’ 102 Harv. L. Rev. 1300, 1301–03 (1989); see also Johnson, 405 F.3d at 1218 n.5 (describing felon disenfranchisement as a punitive device); Green v. Bd. of Elections of City of N.Y., 380 F.2d 445, 451 (2d Cir. 1967) (“A man who breaks the laws he has authorized his agent to make for his own governance could fairly have been thought to have abandoned the right to participate in further administering the compact. On a less theoretical plane, it can scarcely be deemed unreasonable for a state to decide that perpetrators of serious crimes shall not take part in electing the legislators who make the laws, the executives who enforce these, the prosecutors who must try them for further violations, or the judges who are to consider their cases.“).
It appears that Florida‘s disenfranchisement of felons was adopted with similar motivation. See Act of June 25, 1868, ch. 70, 15 Stat. 73, 73 (authorizing Florida‘s re-entry to the Union while prohibiting any change to its constitution that “deprive[d] any citizen or class of citizens of the United States of the right to vote . . . except as punishment for such crimes as [were then] felonies at common law“) (emphasis added).
Florida‘s interest in disenfranchisement as punishment is multifaceted. Perhaps most obviously, the State has a strong interest in promoting the payment of restitution and other financial obligations, for which ongoing punishment of disenfranchisement may serve as an incentive. Indeed, as to some class of felons, the State surely has a legitimate interest in making victims whole by encouraging restitutionary payments and, more generally, Florida has a legitimate interest in ensuring compliance with the lawful sentencing orders of its courts. Similarly, as a general matter, the State has an obvious interest in deterring crime and in employing punitive measures designed to raise the costs associated with criminal conduct. Moreover, the State argues that it has an additional interest in punishment beyond any practical effects it may have—it argues that the State has a lawful interest in punishing criminals simply because they have done a bad thing and deserve to be punished. See Glossip v. Gross, 135 S. Ct. 2726, 2769 (2015) (Breyer,
First, the State asserts an interest in facilitating the collection of outstanding fines and restitution and broadly encouraging the repayment of felons’ debts to society. We agree that the State has an “interest in the collection of revenues produced by [the] payment of fines.” Williams v. Illinois, 399 U.S. 235, 238 (1970). With regard to those felons who have the ability to pay, it is rational for the State to withhold benefits until they do so. However, as for these seventeen plaintiffs, who are indigent and genuinely unable to pay despite good faith efforts, collection is obviously futile and further punishment makes collection no more likely. If a felon is truly indigent and unable to pay his LFOs, Florida‘s requirement “obviously does not serve [revenue collection]; the defendant cannot pay because he is indigent.” Tate v. Short, 401 U.S. 395, 399 (1971); cf. Bearden, 461 U.S. at 670–71 (“Revoking the probation of someone who through no fault of his own is unable to make restitution will not make restitution suddenly forthcoming. Indeed, such a policy may have the perverse effect of inducing the probationer to use illegal means to acquire funds to pay in order to avoid revocation.“).
The problem with the incentive-collections theory is that it relies on the notion that the destitute would only, with the prospect of being able to vote, begin to scratch and claw for every penny, ignoring the far more powerful incentives that already exist for them—like putting food on the table, a roof over their heads, and clothes on their backs. The simple truth is that a collection-based rationale for those who genuinely cannot pay, and who offer no immediate prospects of being able to do so, erects a barrier “without delivering any money at all.” Zablocki v. Redhail, 434 U.S. 374, 389 (1978) (striking down a statute that required an individual to show he had satisfied court-ordered child support before being able to marry). If withholding marriage from those who cannot pay will not incentivize them to do so, surely withholding access to the franchise will not get the State any further.
Although the State has an interest in avoiding asset concealment among felons, we think that interest is inapplicable here.
As for the State‘s interest in deterrence, enforcing continued disenfranchisement against these plaintiffs does not further that interest. To the extent that losing the right to vote is a punishment that could give a would-be criminal pause (unlikely, in the best of circumstances), these plaintiffs have already lost that right. And because we are analyzing the rationality of the LFO requirement as applied to these plaintiffs, any interest the State might have in its general deterrent effect with respect to society as a whole is not implicated.
Next, we address the State‘s interest in punishment for its own sake, perhaps the most challenging interest at issue because it is stated at the highest order of abstraction. Nevertheless, under any plausible theory of retribution, punishment must at least bear some sense of proportionality to the culpability of the conduct punished to be rational. And this correlation must only be between the culpability of the conduct and the punishment:
Retribution, which has as its core logic the crude proportionality of “an eye for an eye,” has been regarded as a constitutionally valid basis for punishment only when the punishment is consistent with an individualized consideration of the defendant‘s culpability, and when the administration of criminal justice works to channel society‘s instinct for retribution.
Tison v. Arizona, 481 U.S. 137, 180–81 (1987) (Brennan, J., dissenting) (quotations and citations omitted); Enmund v. Florida, 458 U.S. 782, 800 (1982) (“As for retribution as a justification for [imposing a particular punishment], we think this very much depends on the degree of [the defendant‘s] culpability . . . .“).
Here, these plaintiffs are punished more harshly than those who committed precisely the same crime—by having their right to vote taken from them likely for their entire lives. And this punishment is linked not to their culpability, but rather to the exogenous fact of their wealth. Indeed, the wealthy identical felon, with identical culpability, has his punishment cease. But the felon with no reasoned prospect of being able to pay has his punishment continue solely due to the impossibility of meeting the State‘s requirement, despite any bona fide efforts to do so. Whatever interest the State may have in punishment, this interest is surely limited to a punishment that is applied in proportion to culpability.
Punishment aside, the State could conceivably have an interest in safeguarding the ballot box from a singularly unqualified group of potential voters. While such an interest has been tightly circumscribed by the Supreme Court, see, e.g., Harper, 383 U.S. at 665–66, it can be legitimate, see, e.g., Lassiter v. Northampton Cty. Bd. of Elections, 360 U.S. 45, 52 (1959) (“[A] State might conclude that only those who are literate should exercise the franchise.“); see also Green, 380 F.2d at 451.
This interest, however, is not plausibly furthered by the distinction Amendment 4 draws between those who have paid their financial obligations and those who have not. This is because the classification at issue is not between people who have disregarded the laws of society and those who have not, nor among groups of felons who have committed crimes that demonstrate that they are more hostile to democracy and the rule of law than are others. Rather, the classification is simply drawn between those who have paid their financial obligations and those who have not. This distinction is unrelated entirely to a felon‘s prudent exercise of the franchise and cannot be said to reasonably further that purpose. See Harper, 383 U.S. at 666 (“Voter qualifications have no relation to wealth nor to paying or not paying this or any other tax.“).
The State additionally asserts that the LFO requirement satisfies rational basis review because it is justified by the administrative costs associated with conducting individualized determinations of ability to pay. The State is correct that ease of administration and reduction of costs may be a legitimate state interest. See Weinberger v. Salfi, 422 U.S. 749, 784 (1975). But with regard to the seventeen plaintiffs at issue in this appeal, we readily reject this asserted justification. The district court‘s preliminary injunction leaves the State ample discretion in designing the procedure to be used to provide an opportunity for these plaintiffs (and others) to demonstrate their indigency and inability to pay for reasons beyond their control. The State has had, and still has before upcoming elections, ample time to provide such opportunity for these seventeen plaintiffs.
Moreover, reducing administrative costs is not a stand-alone interest in this case—Amendment 4 was not passed by a large majority of Florida‘s voters as a cost-saving measure. Rather, the administrative burdens associated with the process of re-enfranchisement are more properly a factor the State may legitimately consider in pursuing its other interests. Bradley, 440 U.S. at 109 (noting that rational basis review accepts over- or under-inclusive classifications “because it is in turn rationally related to the secondary objective of legislative convenience“) (emphasis added).
Thus, if the question on rational basis review were simply whether the LFO requirement was rational as applied to the truly indigent—those genuinely unable to meet their financial obligations to pay fees and fines, and make restitution to the victims of their crimes—we would have little difficulty condemning it as irrational. Quite simply, Florida‘s continued disenfranchisement of these seventeen plaintiffs is not rationally related to any legitimate governmental interest.
ii. As applied to the whole class of felons, on this record, the plaintiffs have failed to meet their burden to demonstrate the law is irrational
But the State argues, and not without some force, that this is the wrong way to conduct rational basis review. It says we must look at the rationality of the Amendment in a more general sense, rather than asking whether it is irrational when applied to this particular class of plaintiffs, or as applied to any other discrete group. See, e.g., Califano v. Jobst, 434 U.S. 47, 55 (1977) (“The broad legislative classification must be judged by reference to characteristics typical
The Supreme Court has on occasion, however, considered the rationality of a statute as applied to particular plaintiffs without opining on its rationality more generally. See, e.g., City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 435 (1985). Indeed, in Cleburne, the Supreme Court specifically rejected the lower court‘s application of heightened scrutiny and repeatedly described its analysis as rational basis review as applied to the plaintiffs. Id. at 442, 446 (“[W]e conclude . . . that the Court of Appeals erred in [subjecting a classification based on mental retardation to] a more exacting standard of judicial review. . . . [L]egislation that distinguishes between the mentally retarded and others must be rationally related to a legitimate governmental purpose.“).
The Supreme Court has never overturned Cleburne or disavowed its logic. However, the case has come to be seen as an exception to ordinary rational basis review that applies a more demanding standard where the measure at issue has no purpose “other than a ‘bare . . . desire to harm a politically unpopular group.‘” Trump v. Hawaii, 138 S. Ct. 2392, 2420 (2018) (quoting Dep‘t of Agric. v. Moreno, 413 U.S. 528, 534 (1973)); see also Lawrence v. Texas, 539 U.S. 558, 580 (2003) (O‘Connor, J., concurring) (describing Cleburne as applying “a more searching form of rational basis review“).
If rational basis review, then, generally is designed to ask only if the codification has some conceivable relation to a legitimate interest of the state, we would readily say that the LFO requirement as applied to the whole class of felons is rational. The analysis becomes more difficult, however, when the requirement is irrational as applied to a class of felons genuinely unable to pay if this class of the impecunious actually resembles the mine-run felon who has otherwise completed the terms of his sentence. Put another way, if the LFO requirement is irrational as applied to those felons genuinely unable to pay, and those felons are in fact the mine-run of felons affected by this legislation, then the requirements may be irrational as applied to the class as a whole.
The State appears to almost concede this point, arguing in its brief that “[a]bsent any evidence that felons unable to pay their outstanding legal financial obligations vastly outnumber those able to pay,” we cannot conclude that the requirement is irrational. And while we understand that a classification subject to rational basis review “must be upheld against equal protection challenge if there is any reasonably conceivable state of facts that could provide a rational basis for the classification,” Beach Commc‘ns, Inc., 508 U.S. at 313, if a substantial enough proportion of felons cannot pay their LFOs, it may be irrational to demand that they do.
Jobst and Murgia illustrate the point. In Jobst, the Supreme Court upheld a Social Security Act scheme that continued benefits for disabled children who married someone receiving social security and cut off benefits for those who married someone not receiving social security. 434 U.S. at 54–55. The Court rested its conclusion on the common-sense observation
In Murgia, the Court upheld Massachusetts‘s mandatory retirement age of 50 for police officers. 427 U.S. at 310–11. As in Jobst, the Court found that the generalization on which the statute was based broadly corresponded to reality. See id. at 315 (“Since physical ability generally declines with age, mandatory retirement at 50 serves to remove from police service those whose fitness for uniformed work presumptively has diminished with age.“). If the state‘s age classification had not been broadly accurate—if Massachusetts only let people over 50 be police officers, for example—it is hard to imagine that the statute would have been upheld. So too here—if the plaintiffs in this case, as to whom the requirement at issue is clearly irrational, are the rule rather than the exception, we would have serious doubt about the requirement‘s rationality.
The district court has made no factual findings on the ability to pay of the mine-run felon or the reasonableness of the scheme‘s assumption that some non-trivial number of felons are able to pay their legal financial obligations, and it is not our place to do so. See, e.g., United States v. Mock, 523 F.3d 1299, 1304 (11th Cir. 2008) (vacating and remanding for further fact-finding where “the district court‘s failure to make specific findings of law and fact precludes meaningful appellate review“).
Nevertheless, what evidence there is in this preliminary injunction record, taken in concert, casts a shadow on the State‘s theory that the impecunious plaintiffs are the exception rather than the rule. Three pieces of evidence offered during the hearing suggest instead that a substantial number of felons—maybe a majority, maybe even a great majority—are like these plaintiffs who claim to be genuinely unable to pay their LFOs and would be precluded from voting under Florida‘s scheme on account of wealth.
First, plaintiffs’ expert, Professor Daniel Smith, compiled data from 58 of the 67 Florida counties and identified 542,207 individuals with felony convictions who had completed their terms of incarceration, parole, or probation. He found that more than 436,000, or 80.5%, of these felons had outstanding LFOs. Moreover, he found that 59% of these individuals had at least $500 of obligations outstanding, and 37.5% had at least $1,000 outstanding. Because Professor Smith was unable to obtain data from some of Florida‘s most populous counties, his analysis was arguably a conservative one. While, notably, he did not conduct any analysis of whether these felons would be able to pay their LFOs, his report suggests that an overwhelming majority of felons have substantial unpaid financial obligations. Dr. Smith‘s findings are not surprising. We know from elsewhere
In the second place, the Florida Court Clerks and Comptrollers’ 2018 Annual Assessments and Collections Report—also a part of the preliminary injunction record—lists a collection rate for court-related fines, fees, penalties, charges, and costs of just 20.55% for felonies in the criminal circuit courts of Florida. That is, of the more than $250 million in outstanding court-related fines and fees, only approximately $50 million was collected in the 2017–2018 fiscal year. The report describes 85.79% of collections as at risk for nonpayment because of (1) incarceration (55%),
In the absence of any fact-finding by the district court, and on this limited record, we cannot say that the plaintiffs have carried their burden of establishing that a substantial proportion of felons (let alone a substantial majority of them) are indigent and, therefore, that the plaintiffs represent the mine-run felon. The facts presented are only suggestive. What they do tell us, however, is that the State‘s implicit claim that a non-negligible number of felons are capable of paying their LFOs may be wrong.
Accordingly, we have some pause in concluding that the LFO requirement as a general matter is grounded in a rational basis. Further development of the factual record at trial may shed more light on these issues. Thus, it appears to us plausible that rather than representing an idiosyncratic, impoverished exception, the plaintiffs in this case, if they are genuinely unable to pay, resemble the mine-run felon otherwise eligible to vote under Amendment 4. We know from the record that the Florida legislature has acknowledged as much, that the Florida courts have an extremely pessimistic outlook about their collection prospects, and that the overwhelming majority of felons have more than $500 in outstanding LFOs.
If that turns out to be the case, then under Jobst the focus of the rationality evaluation would be on indigent felons just like these plaintiffs because their characteristics are those “typical of the affected class[].” 434 U.S. at 55. And we have already determined that enforcement of Florida‘s scheme against such indigent persons is clearly irrational.
It is the plaintiffs’ burden to demonstrate that the LFO requirements are irrational
Although we do not affirm the district court‘s preliminary injunction under a rational basis review, this does not end our analysis. For the reasons we detail below, Supreme Court precedent leads us to apply heightened scrutiny to this case and when measured against that standard, the requirement that the genuinely indigent pay back all of their legal financial obligations as a condition precedent for access to the ballot violates the Equal Protection Clause of the Fourteenth Amendment.
2. The LFO Requirement, As Applied to These Plaintiffs, Cannot Be Sustained Under Heightened Scrutiny
Where a law draws suspect classifications or burdens fundamental rights, we look beyond the benign glance of rational basis review and demand a tighter fit between the classification and the government‘s interests. At first blush, this case does not neatly fit into either category demanding heightened scrutiny. As we have said, generally, wealth is not a suspect classification. See Ortwein, 410 U.S. 656 (1973).
at 660. And states may—even permanently—disenfranchise a felon on the basis of his conviction. See Richardson, 418 U.S. at 56.
But the Supreme Court has told us that wealth classifications require more searching review in at least two discrete areas: the administration of criminal justice and access to the franchise. M.L.B., 519 U.S. at 123 (“[O]ur cases solidly establish two exceptions to that general rule [of rational basis for wealth classifications]. The basic right to participate in political processes as voters and candidates cannot be limited to those who can pay for a license. Nor may access to judicial processes in cases criminal or ‘quasi criminal in nature’ turn on ability to pay.” (citations omitted)). Because Florida‘s re-enfranchisement scheme directly implicates wealth discrimination both in the administration of criminal justice and in access to the franchise, we are obliged to apply some form of heightened scrutiny. Florida has implemented a wealth classification that punishes those genuinely unable to pay fees, fines, and restitution more harshly than those able to pay—that is, it punishes more harshly solely on account of wealth—and it does so by withholding access to the franchise. The observation that Florida may strip the right to vote from all felons forever does not dictate that rational basis review is proper in this case. To the contrary, settled Supreme Court precedent instructs us to employ heightened scrutiny where the State has chosen to “open the door” to alleviate punishment for some, but mandates that punishment continue for others, solely on account of wealth.
i. Florida‘s re-enfranchisement scheme unconstitutionally punishes a class of felons based only on their wealth
We trace the principle of equal criminal justice for the “poor and rich, weak and powerful alike” to Griffin v. Illinois, 351 U.S. 12, 16 (1956). In Griffin, the state of Illinois required criminal defendants to purchase a
The Court‘s conclusion flowed from two constitutional doctrines: “Both equal protection and due process emphasize the central aim of our entire judicial system—all people charged with crime must, so far as the law is concerned, ‘stand on an equality before the bar of justice in every American court.‘” Id. at 15 (quoting Chambers v. Florida, 309 U.S. 227, 241 (1940); citing Yick Wo v. Hopkins, 118 U.S. 356, 359 (1886)). But whether sounding in equal protection or due process, Griffin‘s equality principle is straightforward: the state may not treat criminal defendants more harshly on account of their poverty. See United States v. Plate, 839 F.3d 950, 955–56 (11th Cir. 2016) (applying the Griffin–Bearden line of cases and concluding that “[i]t is apparent that [the defendant] was treated more harshly in her sentence than she would have been if she (or her family and friends) had access to more money, and that is unconstitutional“).
In the more than 60 years since it was announced, ”Griffin‘s principle of ‘equal justice,’ . . . has been applied in numerous other contexts.” Bearden, 461 U.S. at 664 (citing, among others, Douglas v. California, 372 U.S. 353 (1963) (indigent entitled to counsel on first direct appeal); Roberts v. LaVallee, 389 U.S. 40 (1967) (indigent entitled to free transcript of preliminary hearing for use at trial); Mayer v. Chicago, 404 U.S. 189 (1971) (indigent cannot be denied an adequate record to appeal a conviction under a fine-only statute)).
The Supreme Court has also determined that a state may not extend punishment on account of inability to pay fines or fees. See Bearden, 461 U.S. at 672–73 (holding that a state may not revoke probation—thereby extending a prison term—based on the failure to pay a fine the defendant is unable, through no fault of his own, to pay); Tate, 401 U.S. at 399 (holding that a state cannot imprison under a fine-only statute on the basis that an indigent defendant cannot pay a fine); Williams, 399 U.S. at 240–41 (holding that a period of imprisonment cannot be extended beyond the statutory maximum on the basis that an indigent cannot pay a fine).
In Bearden, the Court confronted “whether the
We think the Griffin–Bearden principle straightforwardly applies here too, where the State has chosen to continue to punish those felons who are genuinely unable to pay fees, fines, and restitution on account of their indigency, while re-enfranchising all other similarly situated felons who can afford to pay. This is so because continued disenfranchisement is indisputably punitive in nature, and because felons who are unable to pay are subject to continued punishment solely because of their inability to pay. Just like in Bearden and in Griffin, the fact that the State originally was entitled to withhold access to the franchise from felons is immaterial; rather, heightened scrutiny is triggered when the State alleviates punishment for some, but mandates that it continue for others, based solely on account of wealth.
Disenfranchisement is punishment. We have said so clearly: “[f]elon disenfranchisement laws are unlike other voting qualifications” in that they “are deeply rooted in this Nation‘s history and are a punitive device stemming from criminal law.” Johnson, 405 F.3d at 1228 (emphasis added); see also Harmelin v. Michigan, 501 U.S. 957, 983 (1991) (Scalia, J., writing separately) (discussing disenfranchisement in the context of the
What‘s more, disenfranchisement is a continuing form of punishment. It is true that two similarly situated felons—two felons who had committed the same crime—were subject to the same punishment (disenfranchisement) on the front end of the sentencing procedure. But much like a prisoner is punished throughout a carceral term, not just when he is initially sentenced to incarceration, so too disenfranchised felons continue to be punished as
Because the State has opted to automatically re-enfranchise felons who complete their terms of sentence and pay all of their LFOs, similarly situated felons who have otherwise completed their sentences except for the payment of LFOs they are unable to pay are treated differently on account of their inability to pay. The felon with money in the bank will be re-enfranchised. But the felon who can‘t will continue to be barred. Merely because the State could strip the rights of both felons does not mean it can continue punishment for some and not others. In Griffin, after all, none of the convicts at issue had a right to a state appeal. But once the State opened the door to appeal, and as soon as it did, it subjected itself to a more heightened level of scrutiny by including wealth as the differentiating factor. So too in Bearden: the state was entitled to demand that every felon remain incarcerated. Felons have no constitutional right to probation. But by opening the door and offering probation to some, while continuing the punishment of incarceration for others, again based solely on wealth, heightened scrutiny was triggered. The same principle applies here: Florida is permitted to disenfranchise all felons. But as soon as the law demanded that punishment continue for some and not others based on wealth, we became obliged to examine this disparity with a sharper focus, because differential punishment on account of wealth strikes at the heart of Griffin‘s equality principle.
ii. Heightened scrutiny is further compelled because the punishment the plaintiffs suffer implicates access to the franchise
Moreover, and even more significantly, where the punishment itself takes the form of denying access to the franchise—the second arena where Griffin‘s equality principle applies—we are all the more convinced that some form of heightened scrutiny is required.
The Supreme Court expressly and repeatedly extended Griffin‘s equality principle beyond the realm of criminal justice; it has been applied to state action that burdens important constitutional interests, such as fundamental associational and political participation interests. See M.L.B., 519 U.S. at 111 (”Griffin‘s principle has not been confined to cases in which imprisonment is at stake.“); see also Zablocki, 434 U.S. at 389, 402 (striking down a statute that required an individual to show he had satisfied court-ordered child support before being able to marry); id. at 402 (Powell, J., concurring) (citing Griffin to distinguish from the traditional wealth-discrimination rational basis analysis); Lubin v. Panish, 415 U.S. 709, 719–21 (1974) (Douglas, J., concurring) (invoking Griffin to strike down a law requiring a filing fee to be placed on a state ballot); Cruz v. Hauck, 404 U.S. 59, 63–64 (1971) (invoking Griffin and declaring “[Griffin‘s] equal protection concept is not limited to criminal prosecutions” (citations removed)); Boddie v. Connecticut, 401 U.S. 371, 382–84 (1971) (striking down state procedures imposing filing fees on an indigent couple who sought a divorce, deciding on due process grounds but stating “the rationale of Griffin covers this case,” while Justice
Voting is squarely among the interests that fall within Griffin‘s grasp. More than a century ago, the Supreme Court held that voting “is regarded as a fundamental political right, because [it is] preservative of all rights.” Yick Wo, 118 U.S. at 370. More recently, the Court has affirmed the idea that “the right of suffrage is a fundamental matter in a free and democratic society,” and “any alleged infringement of the right of citizens to vote must be carefully and meticulously scrutinized.” Reynolds, 377 U.S. at 561–62. Indeed, in Harper v. Virginia State Board of Elections, the Court expressly held, citing to Griffin, that “a State violates the Equal Protection Clause of the
The State argues, however, that Harper is distinguishable because it involved the fundamental right to vote, whereas felons have no fundamental right to vote because access to the ballot box can be constitutionally stripped from them. We are aware that other courts have applied rational basis scrutiny to felon re-enfranchisement schemes requiring the payment of LFOs as a precondition to voting on the ground that felons no longer enjoy the fundamental right to vote. See, Bredesen, 624 F.3d at 746; Harvey v. Brewer, 605 F.3d 1067, 1079 (9th Cir. 2010); Madison, 163 P.3d at 766–67. Only Bredesen and Madison, however, involved an as-applied challenge to the LFO requirement by those who could not pay. Indeed, Justice O‘Connor, writing for the Ninth Circuit in Harvey, expressly noted that no plaintiff in the case alleged that he was indigent. 605 F.3d at 1079. Her discussion left open the constitutional question arising from a re- enfranchisement scheme that continues to disenfranchise felons solely because they were indigent and truly unable to pay their fines. Id. at 1080.
For the proposition that felons can no longer claim a fundamental right to vote, other courts have ultimately relied on Richardson. See Owens v. Barnes, 711 F.2d 25, 27 (3d Cir. 1983) (“Plaintiff‘s argument fails because the right of convicted felons to vote is not ‘fundamental.’ That was precisely the argument rejected in Richardson.“); see also Bredesen, 624 F.3d at 746, 750 (citing Richardson and Sixth Circuit precedent including Owens); Harvey, 605 F.3d at 1079 (“But [plaintiffs] cannot complain about their loss of a fundamental right to vote because felon disenfranchisement is explicitly permitted under the terms of Richardson . . . .“); Madison, 163 P.3d at 768 (“[W]e conclude that Richardson dictates that we hold that the right to vote is not fundamental for convicted felons.“).
But Richardson does not control the result in this case. All it holds, and no more, is that states have the green light to exclude felons from the franchise. 418 U.S. at 54. This broad statement of law is undoubtedly true, and indeed, no party disputes that Florida can withhold access to the franchise for all felons. But Richardson did not address what may happen when a state chooses to adopt the automatic re-enfranchisement of felons. And it surely did not address the critical factual circumstances permeating this case, where the State‘s re-enfranchisement scheme automatically restores the right to vote to some felons who stand in materially the same shoes as these plaintiffs, but continues to disenfranchise these seventeen plaintiffs solely on account of their indigency and inability to pay for reasons wholly beyond their control. Instead, the Court made it clear that the abridgement of a felon‘s right to vote is still subject to constitutional limitations; states do not have carte blanche to deny access to the franchise to some felons and not others, based on race or religion, for example. See id. at 56 (remanding to the Supreme Court of California for consideration of equal protection arguments); see also Harvey, 605 F.3d at 1079 (“[A] state could not choose to re-enfranchise voters of only one particular race . . . .” (citing Hunter v. Underwood, 471 U.S. 222, 233 (1985))).
Moreover, even if felons do not have the right to vote, nothing in Harper‘s analysis turned on the assumption that those who would be unable to pay the fee personally had a fundamental right to vote. Indeed, the Harper Court expressly declined to rule on whether anyone has a fundamental right to vote in state elections, a proposition for which there would have been no authority. See 383 U.S. at 665 (“[T]he right to vote in state elections is nowhere expressly mentioned [in the Constitution]. It is argued that the right to vote in state elections is implicit . . . . We do not stop to canvass the relation between voting and political expression. For it is enough to say that once the franchise is granted to the electorate, lines may not be drawn which are inconsistent with the Equal Protection Clause of the
Further, Harper distinguished Lassiter v. Northampton County Board of Elections—an earlier case that had held that denying access to the franchise to those who failed literacy tests did not violate the Equal Protection Clause—not by holding that the poor have a fundamental right to vote while the illiterate do not, but rather because it concluded that wealth has no relation to the prudent exercise of the franchise, while literacy does. See Harper, 383 U.S. at 665–66; see also Lassiter, 360 U.S. at 51. In the end, Harper makes clear that it is the fundamental nature of the right to vote in general, not its particular possession by the poor, that dictated the application of heightened scrutiny. See id. at 670 (“[W]ealth or fee paying has . . . no relation to voting qualifications; the right to vote is too precious, too fundamental to be so burdened or conditioned.“) (emphasis added).
That Harper‘s application of heightened scrutiny to wealth discrimination in the context of access to the franchise was based on the importance of the right in general, rather than the possession of the right by particular individuals, is not surprising. Indeed, the Griffin–Bearden line of cases all illustrate this point. For example, each of Williams, Tate, and Bearden involved the “right to remain free.” Williams, 399 U.S. at 263 (Harlan, J., concurring). But in all three cases, the plaintiffs, who had committed crimes, could lawfully have had their liberty taken away—even permanently. To the same extent that felons are not entitled to vote, the plaintiffs in Williams, Tate, and Bearden were no longer entitled to their liberty.
Nevertheless, because the interest in liberty is so important, the Court held that the state could not rely on the plaintiffs’ wealth in deciding whether to deprive them of liberty. See Bearden, 461 U.S. at 672–73 (striking down revocation of probation that “would deprive the probationer of his conditional freedom simply because, through no fault of his own, he cannot pay the fine“); Tate, 401 U.S. at 398 (“[T]he same constitutional defect condemned in Williams also inheres in jailing an indigent for failing to make immediate payment of any fine, whether or not the fine is accompanied by a jail term and whether or not the jail term of the indigent extends beyond the maximum term that may be imposed on a person willing and able to pay a fine.“); Williams, 399 U.S. at 271 (“[O]nce the State has defined the outer limits of incarceration necessary to satisfy its penological interests and policies, it may not then subject a certain class of convicted defendants to a period of imprisonment beyond the statutory maximum solely by reason of their indigency.“). Williams, Tate, and Bearden establish, as Harper suggests, that the state‘s ability to deprive someone of a profoundly important interest does not change the nature of the right, nor whether it is deserving of heightened scrutiny when access to it is made to depend on wealth. Moreover, as we‘ve discussed, the holding in Griffin did not turn on whether there is a fundamental right to an appeal—it could not have, because such a right does not exist. See Griffin, 351 U.S. at 18.
The long and short of it is that once a state provides an avenue to ending the punishment of disenfranchisement—as the voters of Florida plainly did—it must do so consonant with the principles of equal protection and it may not erect a wealth barrier absent a justification sufficient to overcome heightened scrutiny.
Our conclusion that heightened scrutiny applies here is not contrary to the general rule that rational basis scrutiny applies when analyzing felon disenfranchisement or re-enfranchisement schemes. Most classifications in this context will be subject only to rational basis review. Thus, by way of example, a state need only offer a rational basis for distinguishing in re-enfranchisement between those felons currently incarcerated and those who have completed prison terms, or among felons who have committed different kinds of crimes, or even between federal and state felons. See, e.g., Hayden v. Paterson, 594 F.3d 150, 171 (2d Cir. 2010) (upholding under rational basis review a New York constitutional provision that disenfranchised only incarcerated felons and those on parole); Owens, 711 F.2d at 28 (upholding under rational basis review a Pennsylvania law that disenfranchised only incarcerated felons).
For this reason, our decision to apply heightened scrutiny is not in conflict with the precedent of this Circuit in Shepherd v. Trevino, 575 F.2d 1110 (5th Cir. 1978).11 In Shepherd, the former Fifth Circuit examined a Texas re-enfranchisement mechanism and held that rational basis scrutiny applied. There, the Texas Constitution disenfranchised all persons convicted of a felony, subject to exceptions made by the legislature. Id. at 1111. A Texas statute created a mechanism whereby persons convicted of a felony in Texas courts could, after the satisfactory fulfillment of and
explained, Griffin and Bearden mandate that we apply some form of heightened scrutiny when punishment continues for one class and not another simply based on wealth, while Harper applies this same principle to voting.
Nor does Katzenbach v. Morgan compel that we apply rational basis review, as the State suggests. 384 U.S. 641 (1966). In Katzenbach, the Court held that the prohibition on literacy tests as a prerequisite to voting contained in
Moreover, Katzenbach did not involve disenfranchisement as a criminal punishment grounded on indigency, and thus did not involve the context of this case in which the Harper line of cases, and the Griffin–Williams–Tate–Bearden line of cases counsel in favor of applying heightened scrutiny. In sum, Katzenbach tells us nothing about the level of scrutiny applicable to our review of the State‘s LFO requirement.
On the other side of the coin, we also disagree with the district court that this Court‘s en banc decision in Johnson controls the resolution of this case. In Johnson, the plaintiffs challenged the constitutionality of Florida‘s then-existing voting rights restoration scheme, which consisted solely of discretionary executive clemency, primarily on the grounds that it was racially discriminatory. 405 F.3d at 1217. The plaintiffs also brought a wealth discrimination claim. As to the wealth discrimination claim, they argued that the clemency policy‘s distinction between those who had paid their LFOs and those who had not—the former being entitled to have their applications decided on the papers and the latter having to appear for a hearing—violated the Equal Protection Clause. Id. at 1216 n.1.
In a footnote disposing of the wealth discrimination claim, our en banc Court
Quite simply, two strands of Supreme Court law—those embodied in its Griffin–Bearden line of cases outlawing different levels of punishment for similarly situated defendants, solely on account of wealth, and those found in the Harper line of cases underscoring the importance of access to the ballot—run together in this case. These weighty interests are directly implicated and they yield the conclusion that we must examine the Amendment‘s impact on the seventeen named plaintiffs through the lens of heightened scrutiny.
iii. Applying heightened scrutiny, the LFO requirement is unconstitutional as applied to those who genuinely cannot pay
Although Griffin gave us the broad principle of equal treatment in criminal justice without regard to wealth, its progeny has elaborated on the application of heightened scrutiny to laws said to run afoul of Griffin‘s command. In Bearden, the Court concluded that revoking probation based on failure to pay a fine that a defendant is unable to pay amounted to unconstitutional wealth discrimination. 461 U.S. at 672–73. The disparity in treatment implicated both equal protection and due process principles, but, the Court said, “[w]hether analyzed in terms of equal protection or due process, the issue cannot be resolved by resort to easy slogans or pigeonhole analysis.” Id. at 666.
The form heightened scrutiny took in Bearden was comprised of four considerations: (1) “the nature of the individual interest affected“; (2) “the extent to which it is affected“; (3) “the rationality of the connection between legislative means and purpose“; and (4) “the existence of alternative means for effectuating the purpose.” Id. at 666–67 (quoting Williams, 399 U.S. at 260 (Harlan, J., concurring) (analyzing a wealth discrimination claim under due process rather than equal protection but reaching the same result as the majority)).12 Taking these one
by one, we believe they support a determination that the LFO requirement is unconstitutional as applied to felons who genuinely cannot pay.
First, as we have explained, voting is undoubtedly a weighty interest. See, e.g., Harper, 383 U.S. at 667-68; Reynolds, 377 U.S. at 561-62; Yick Wo, 118 U.S. at 370.
Second, the interest is profoundly affected; with the LFO requirement in place, felons sustain the “deprivation of a meaningful opportunity to enjoy th[e] benefit” of the ability to vote. Rodriguez, 411 U.S. at 20. The defendants argue, however, that felons have three alternative avenues to regain their access to the ballot: (1) by terminating their LFOs “[u]pon the payee‘s approval,” as SB 7066 allows, see
It is exceedingly unlikely that the first avenue is practically available for most felons. For starters, many felons have had their debts assigned to private collection agencies, which would understandably be loath to voluntarily forgive these debts. And as far as restitution is concerned, it is similarly unlikely that the victims of crime would agree to forgo financial payouts—however unlikely their receipt may be—to allow the people who have victimized them to be able to vote.
The second avenue is similarly illusory. As the plaintiffs point out, community service conversion is unavailable for those felons whose debts have been converted into civil liens and for those with federal convictions. Moreover, the conversion rate—determined by the district court to be at “low hourly rates“—could tee up a long process for rights-restoration, during which otherwise-eligible felons may miss many opportunities to vote.
The third option could be more promising but is largely illusory as it stands now. Although the Executive Clemency Board just recently revised its procedural requirements to allow a felon to make his case for rights-restoration without paying restitution, we have no evidence of how effective, speedy, or accessible this process will be. What we do know is that the Board‘s five-year and seven-year waiting periods before any felon may even seek rights-restoration remain intact, see Fla. R. Exec. Clemency 9, 10 (2020), https://www.fcor.state.fl.us/docs/clemency/clemency_rules.pdf, and the district court found that Florida‘s executive clemency process has historically moved at glacial speed.
All three avenues suffer from a common and basic infirmity—they are entirely discretionary in nature. No one owed money by a felon—whether a collection company or a victim—is required to forgive the debt. No court is required to convert LFOs to community-service hours. And no member of the Clemency Board is required to grant rights restoration. We can hardly call pursuing a discretionary grant that at best will take years to receive, if received at all, a suitable alternative. After all, those felons who are able to pay enjoy near immediate, automatic re-enfranchisement as of right. None of the alternatives is a viable stand-in for the automatic re-enfranchisement enjoyed by felons whose circumstances are materially similar to those of these plaintiffs except for their wealth. Accordingly, we have little difficulty in concluding that the individual‘s weighty interests are profoundly affected.
As for the third Bearden consideration—“the rationality of the connection between legislative means and purpose“—we have already expressed our strong reservations regarding the State‘s interest in continuing the disenfranchisement of these plaintiffs even under rational basis review. Under heightened scrutiny, we can readily say that this law fails the test when applied to felons who are genuinely unable to
The State also has no real collection interest as applied to these felons. We agree, as we have already observed, that the State generally has an “interest in the collection of revenues produced by payment of fines.” Williams, 399 U.S. at 238. But as to these plaintiffs, who despite their best efforts are genuinely unable to pay, collection is futile. The State cannot draw blood from a stone. Florida‘s requirement “obviously does not serve [revenue collection]; the defendant cannot pay because he is indigent.” Tate, 401 U.S. at 399; see also Bearden, 461 U.S. at 670-71. If a felon is truly unable to pay, it makes no sense to assert that he will be incentivized to pay his LFOs with money that he does not have. The Supreme Court concluded that this kind of incentive was powerless even in the face of barring the ability to marry; surely the State cannot argue that the incentive is more powerful here. See Zablocki, 434 U.S. at 389. To whatever degree the State has a collection interest in the class as a whole, it surely does not as applied to felons who cannot pay.
Nor does the State have a readily identifiable interest in protection of the ballot box, because, as the Supreme Court has made clear, “[v]oter qualifications have no relation to wealth nor to paying or not paying this or any other tax.” Harper, 383 U.S. at 666. Two hypothetical felons here have committed the same crime, with identical culpability, yet one is deemed more qualified to vote based on ability to pay. Harper demands that we reject such a classification.
Finally, the State maintains an abstract interest in punishment for punishment‘s sake. The question here is not whether states may punish felons with disenfranchisement—they can. See Richardson, 418 U.S. at 56. The question, as we see it, is whether a state may continue to punish one felon but not another, after opening the door to relief. Punishment must be assigned in some proportion to culpability or it is merely vindictive. Here, the plaintiffs are not punished in proportion to their culpability but to their wealth—equally guilty but wealthier felons are offered access to the ballot while these plaintiffs continue to be disenfranchised, perhaps forever. Thus, the third Bearden consideration falls in the plaintiffs’ favor.
Turning finally to the fourth consideration—“the existence of alternative means for effectuating the purpose“—we believe that the State has more than adequate ways to promote its interest in debt collection. As the Supreme Court explained in Zablocki, “the State already has numerous other means for exacting compliance with [financial] obligations, means that are at least as effective as the instant statute‘s and yet do not impinge upon” the right to vote. 434 U.S. at 389. Under Florida law, the State can refer debts to collection agencies, garnish wages, and conduct civil forfeiture. See
In short, applying Bearden‘s heightened scrutiny, we have little difficulty in concluding that the LFO requirement is likely unconstitutional as applied to these seventeen plaintiffs.
In a final attempt to defeat this conclusion, the State argues that we cannot hold that the LFO requirement violates the
Moreover, the Supreme Court has never required proof of discriminatory intent in a wealth discrimination case, see Mayer, 404 U.S. 189; Tate, 401 U.S. 395; Roberts, 389 U.S. 40; Douglas, 372 U.S. 353; Griffin, 351 U.S. 12, including in cases that post-date Davis, see Bearden, 461 U.S. 660; Zablocki, 434 U.S. 374; see also M.L.B., 519 U.S. at 127 (“[U]nder respondents’ reading of Washington v. Davis, our overruling of the Griffin line of cases would be two decades overdue.“). The race discrimination cases cited by the State are inapposite. The plaintiff‘s failure to allege or establish discriminatory intent was not a barrier to the Court‘s holding wealth discrimination unconstitutional in Bearden, and it is not here.
B. Irreparable Injury
Having concluded that these plaintiffs are substantially likely to succeed on the merits, we are also required to consider whether they will suffer irreparable injury. Since the preliminary injunction factors other than the likelihood of success on the merits turn on equitable considerations and factual findings, we owe substantial deference to the district court‘s conclusions. See BellSouth Telecomms., Inc., 425 F.3d at 968.
The district court did not abuse its discretion in finding that the plaintiffs will suffer an irreparable injury if they are precluded by the enforcement of Amendment 4 from voting in an election in which they were constitutionally entitled to vote. An injury is “irreparable if it cannot be undone through monetary remedies.” Scott v. Roberts, 612 F.3d 1279, 1295 (11th Cir. 2010) (quotation omitted). Casting a vote has no monetary value. It is nothing other than the opportunity to participate in the collective decisionmaking of a democratic society and to add one‘s own perspective to that of his or her fellow citizens. Each vote provides a unique opportunity to do that. No compensation a court can offer could undo that loss. The denial of the opportunity to cast a vote that a person may otherwise be entitled to cast—even once—is an irreparable harm. Indeed, several of our sister circuits have similarly concluded that missing the opportunity to vote in an election is an irreparable harm for the purposes of a preliminary injunction. See, e.g., League of Women Voters of N.C. v. North Carolina, 769 F.3d 224, 247 (4th Cir. 2014) (“[O]nce the election occurs, there can be no do-over and no redress. The injury to these voters is real and completely irreparable if nothing is done to enjoin this law.“); Obama for Am. v. Husted, 697 F.3d 423, 436 (6th Cir. 2012) (“A restriction on the fundamental right to vote . . . constitutes irreparable injury.” (citing Williams v. Salerno, 792 F.2d 323, 326 (2d Cir. 1986))). The district court did not abuse its discretion in agreeing.
C. Balance of Hardship
Likewise, the district court did not abuse its discretion in finding that the injury to the plaintiffs outweighs the damage to the State. The State argues that it has four discrete interests that weigh against the issuance of a preliminary injunction: (1) a general concern in enforcing its statutes and the integrity of the electoral process; (2) if the injunction were expanded, it “could change the outcome of the upcoming elections“; (3) the difficulty and cost of administering compliance with the court‘s order; and (4) the expense associated with “making individual determinations about the socioeconomic status of hundreds of thousands of individuals,” which would all be for naught if the State ends up prevailing. We think these interests, while significant, are unavailing as compared to the plaintiffs’ interest in their opportunity to exercise the core democratic right of voting. In the first place, although we have no doubt that the State has a weighty interest in upholding its statutes and ensuring the integrity of the electoral process, these interests are stated at too high an order of abstraction to be persuasive. The State‘s broad interest in enforcing its statutes, standing alone, would be applicable any time a statute‘s constitutionality is challenged and a preliminary injunction issued against its enforcement. The State‘s argument in this case, if it carried the day, would prove too much—hardly any preliminary injunction could ever issue. C.f. Lebron v. Sec‘y, Fla. Dep‘t of Children & Families, 710 F.3d 1202, 1217 (11th Cir. 2013) (affirming a preliminary injunction enjoining the State from enforcing one of its laws on constitutional grounds).
Second, the State‘s interest in avoiding altered election results—which it posits would happen if the injunction were expanded—is entirely speculative. For one, the preliminary injunction itself requires the State to allow only seventeen plaintiffs to vote. As close as Florida elections may be, the votes of these seventeen are vanishingly unlikely to be outcome dispositive. There is similarly no evidence in the record at this stage regarding how many felons would in fact vote nor that the population of ex-felons is a cohesive political unit. In any event, as the district court observed, the State‘s interest in preventing votes by ineligible voters is no greater than its interest in allowing votes by eligible voters. The plaintiffs’ concrete, personal, and measurable injury in the denial of access to the franchise, if in error, far outweighs the State‘s concerns—which could only ever be speculative—that the outcome of a particular election was determined by the influence of felon voters ultimately found to be ineligible.
Third, the State‘s remaining two interests relate to the administrative burden that it will shoulder if the preliminary injunction is upheld or expanded. We acknowledge that compliance with the constitutional principle of this case may entail an administrative burden. We also recognize, however, that the question of how substantial a burden the district court‘s injunction will be is largely one of Florida‘s own making. The injunction does not require the State to employ any particular procedures, nor even to adopt a particular definition of “genuine inability to pay.” To comply with the legal principle behind the injunction, the State need make only a good faith effort to ensure that no felon
We too leave the procedural implementation of the preliminary injunction to the State itself, but note that the State has a range of options available that do not appear unduly burdensome in light of the significance of the plaintiffs’ interest. Indeed, Florida‘s voter registration statute already allows for individualized determinations of eligibility if a Supervisor of Elections or the Department of State concludes that a voter is ineligible. See
By leaving to the State substantial discretion in how to comply with the preliminary injunction, the district court‘s order respects the State‘s sovereignty. See Democratic Exec. Comm., 915 F.3d at 1331 (noting that where a preliminary injunction, “instead of throwing out the plausibly legal with the constitutionally problematic . . . narrowly tailored its relief to home in on the one limited aspect . . . [that] unduly burdened . . . [the] right to vote,” “rather than undermining Florida‘s sovereignty, . . . actually respected it“); see also Ga. Muslim Voter Project v. Kemp, 918 F.3d 1262, 1276 (11th Cir. 2019) (mem.) (Jill Pryor, J., concurring) (affirming an injunction that “borrowed heavily from the processes already in place“).
In sum, we think that the harms the State may suffer as a result of the preliminary injunction are demonstrably outweighed by those imposed on the plaintiffs. See Fish v. Kobach, 840 F.3d 710, 755 (10th Cir. 2016) (noting that there is no contest between denial of access to the ballot and a state‘s administrative burden). At any rate, on this record we do not believe we could fairly say that the district court abused its discretion in so concluding.
D. Public Interest
Finally, the district court did not abuse its discretion in concluding that the public interest favors the preliminary injunction. The “cautious protection of the Plaintiffs’ franchise-related rights is without question in the public interest.” Charles H. Wesley Educ. Found., Inc. v. Cox, 408 F.3d 1349, 1355 (11th Cir. 2005). We have also held that the knowledge that otherwise-eligible voters were not counted “would be harmful to the public‘s perception of the election‘s legitimacy,” and that “the public interest is served when constitutional rights are protected.” Democratic Exec. Comm., 915 F.3d at 1327.
E. Severability
Finally, the State says that to the extent the LFO requirement as applied to the plaintiffs violates the Equal Protection Clause, this application cannot be severed from Amendment 4 as a whole.13 Therefore, rather than affirming the partial injunction granted by the district court, the State argues that we must enjoin Amendment 4 as a whole.
Plainly, the severability of state constitutional provisions is “a matter of state law.” Leavitt v. Jane L., 518 U.S. 137, 139 (1996). “Florida law clearly favors (where possible) severance of the invalid portions of a law from the valid ones.” Solantic, LLC v. City of Neptune Beach, 410 F.3d 1250, 1269 n.16 (11th Cir. 2005) (quoting Coral Springs Street Sys., Inc. v. City of Sunrise, 371 F.3d 1320, 1347 (11th Cir. 2004)). Indeed, Florida‘s severance doctrine is “designed to show great deference to the legislative prerogative to enact laws,” Schmitt v. State, 590 So. 2d 404, 415 (Fla. 1991), by “recognizing the obligation of the judiciary to uphold the constitutionality of legislative enactments where it is possible to strike only the unconstitutional portions,” Ray v. Mortham, 742 So. 2d 1276, 1280 (Fla. 1999) (citing State v. Calhoun County, 170 So. 883, 886 (Fla. 1936)). Florida law thus adopts a strong presumption of severability, and squarely places the burden on the party challenging severability. See Ray, 742 So. 2d at 1281 (noting that, because the purpose of the severability doctrine is “to preserve the constitutionality of enactments where it is possible to do so,” the “burden is properly placed on the challenging party“).
Under Florida law, “the remainder of the act [may] stand” where “a part of a statute [has been] declared unconstitutional” so long as four requirements are met:
(1) the unconstitutional provisions can be separated from the remaining valid provisions, (2) the legislative purpose expressed in the valid provisions can be accomplished independently of those which are void, (3) the good and the bad features are not so inseparable in substance that it can be said that the Legislature would have passed the one without the other and, (4) an act complete in itself remains after the invalid provisions are stricken.
Smith v. Dep‘t of Ins., 507 So. 2d 1080, 1089-90 (Fla. 1987) (quoting Cramp v. Bd. of Pub. Instruction of Orange Cty., 137 So. 2d 828, 830 (Fla. 1962)). This test also applies to state constitutional provisions. See Ray, 742 So. 2d at 1281. And contrary
The first and fourth requirements are not challenged by the State. See Women‘s Emergency Network v. Bush, 323 F.3d 937, 949 (11th Cir. 2003) (noting that the first and fourth elements are presumed to be satisfied “in almost any case“). The only provision enjoined by the district court, as applied to the plaintiffs, is the application of the phrase “all terms of sentence” to require payment of legal financial obligations to those genuinely unable to pay them. This application can obviously be excised, leaving Amendment 4 as a complete act. Indeed, we know that this is true because before the Florida Supreme Court issued an advisory opinion, and as noted by the district court, there was a plausible textual argument that this application did not fall within the Amendment‘s language at all.
The State focuses its challenge on the second and third factors. The relevant inquiry is “whether the overall legislative intent is still accomplished without the invalid provisions.” State v. Catalano, 104 So. 3d 1069, 1080-81 (Fla. 2012) (citing Martinez v. Scanlan, 582 So. 2d 1167, 1173 (Fla. 1991)). It is the State‘s burden to show that Amendment 4 would not have been adopted absent the unconstitutional application of the LFO requirement to those who cannot pay. See Ray, 742 So. 2d at 1281. And Florida law is clear that a party that does no more than “cast doubt on whether the amendment would have passed” without its unconstitutional application has not carried its burden. Id.
On this preliminary record, that is all the State has done. It simply argues that the purpose of Amendment 4 was not merely to re-enfranchise felons but to re-enfranchise felons who have “paid their debt to society,” and that the latter qualification was a critical prerequisite to the Amendment‘s passing. This strikes us as a plausible characterization of the evidence in the record as far as it goes, but again, the State‘s assertion in this regard is speculative. It is altogether unclear whether the people of Florida would have voted differently if they knew that the Amendment they adopted could not be constitutionally applied to those felons who were genuinely unable to pay despite their good faith efforts to do so. Nevertheless, the notion that the district court‘s narrow preliminary injunction entirely vitiates the conditioning of access to the franchise on completion of “all terms of sentence” is unpersuasive. The district court‘s injunction still requires felons to complete their carceral sentences and parole or probation before becoming eligible to vote. Moreover, to the extent a felon can pay LFOs, he or she must.14 In short, it is still the State‘s burden to persuade us that enough Florida voters would have voted differently had they known that Amendment 4 could not be used to exclude these plaintiffs who had otherwise completed their sentences but were genuinely unable to pay their LFOs from voting. Absent some concrete evidence or even a persuasive argument to that effect, we conclude the unconstitutional application is severable.15
IV. CONCLUSION
We affirm the district court‘s preliminary injunction enjoining the defendants
AFFIRMED.16
Notes
The district court, in dicta, discussed the plaintiffs’ Twenty-Fourth Amendment claim at some length. In particular, it observed that while none of the financial obligations at issue are formal poll taxes, and fines and restitution plainly do not qualify as “other taxes,” it was debatable whether routine fees imposed during the criminal process would qualify. Nevertheless, because the district court expressly declined to rule on the merits of the Twenty-Fourth Amendment claim, we do not address it either. “[W]hen an appeal is taken from the grant or denial of a preliminary injunction, the reviewing court will go no further into the merits than is necessary to decide the interlocutory appeal.” Callaway v. Block, 763 F.2d 1283, 1287 n.6 (11th Cir. 1985) (citations omitted). Moreover, since consideration of the merits of the claim is at a minimum a mixed question of law and fact, turning in part on undeveloped facts about the function of these fees in the criminal justice system, it would be inappropriate for us to rule on it in the first instance. See Garrett v. Higgenbotham, 800 F.2d 1537, 1539 (11th Cir. 1986) (noting that a “mixed question of fact and law” is “one we are poorly situated to decide in the first instance” and remanding).
Similarly, we decline to rule on the plaintiffs’ procedural due process claims. The plaintiffs alleged that Florida‘s LFO requirement violated the Due Process Clause of the Fourteenth Amendment because the State, which regularly converts outstanding LFOs to civil liens and relegates them to collection agencies, often does not know whether an individual has completely paid off his or her LFOs or how much remains outstanding. However, the district court explicitly declined to rule on this claim too, and, therefore, it is not properly before us. See Callaway, 763 F.2d at 1287 n.6. Moreover, the due process claim turns on factual questions about how Florida‘s LFO collection scheme operates in practice, and in the absence of any factual findings by the district court, we will not attempt to find such facts on this preliminary record. See GeorgiaCarry.Org, Inc. v. U.S. Army Corps of Eng‘rs, 788 F.3d 1318, 1328–29 (11th Cir. 2015) (“This slim preliminary injunction record does not provide nearly enough information to enable a court to fairly engage in a thorough constitutional analysis.“).