Lee, Circuit Judge.

Appeal from the United States District Court for the Eastern District of Wisconsin. No. 2:20-cv-1100 — Lynn Adelman, Judge. Adele D. Nicholas, Attorney, Law Office of Adele D. Nicholas, Chicago, IL, Mark G. Weinberg, Attorney, Law Office of Mark G. Weinberg, Chicago, IL, for Plaintiff-Appellant. Thomas A. Cabush, Dustin Todd Woehl, Attorneys, Kasdorf, Lewis & Swietlik S.C., Milwaukee, WI, for Defendant-Appellee. Before Sykes, Chief Judge, and Rovner and Lee, Circuit Judges. Lee, Circuit Judge.

The Town of Paris is a small, rural community in Kenosha County, Wisconsin. In 2008, Paris enacted its "Sex Offender Residency Restrictions" ordinance, limiting where certain sex offenders—referred to as "designated offenders"—could live within the town. See TOWN OF PARIS. WIS. CODE OF ORDINANCES, §§ 10-19-10-25 (2022) (the "ORDINANCE"). As relevant here, the ordinance prohibits designated offenders from living within 6,500 feet of certain protected locations where children are known to congregate (we will call this the "protected locations restriction"). It also prohibits designated offenders from living within 6,500 feet of any other designated offender (we will call this the "designated offenders restriction"). Id. § 10-21(1)(a)-(b).

Peter Nelson, a former Paris resident and designated offender, was cited for violating the ordinance's designated offenders restriction. He filed suit under 42 U.S.C. § 1983, arguing that the ordinance—both facially and as applied—violates his constitutional right to substantive due process under the Fourteenth Amendment, as well as Article I's prohibition on ex post facto laws. The district court granted summary judgment against Nelson and in favor of Paris on both claims. See Nelson v. Town of Paris, 616 F. Supp. 3d 844 (E.D. Wis. 2022). Nelson appeals.

Applying the analysis espoused in Smith v. Doe, 538 U.S. 84, 123 S.Ct. 1140, 155 L.Ed.2d 164 (2003), we hold that Paris's restriction prohibiting designated offenders from living within 6,500 feet of protected locations does not violate the Constitution's Ex Post Facto Clause because it is not "so punitive either in purpose or effect" as to negate Paris's nonpunitive intent for the restriction. Id. at 92, 123 S.Ct. 1140 (quoting Kansas v. Hendricks, 521 U.S. 346, 361, 117 S.Ct. 2072, 138 L.Ed.2d 501 (1997)). But based on the record before us, we cannot conclude the same about Paris's restriction prohibiting designated offenders from living within 6,500 feet of each other. We therefore remand this issue to the district court for further factual development. As for Nelson's due process claim, because he concedes the ordinance is rationally related to Paris's legitimate interest in protecting children, we affirm the district court's dismissal of that claim.

I. BACKGROUND

A. The Ordinance

Paris enacted the "Sex Offender Residency Restrictions" ordinance in 2008 and amended it in 2018. It applies only to "designated offenders," defined as "any person who is required to register under Section 301.45 and 301.46, Wisconsin Statutes, for any sexual misconduct or violation as a result of being a repeat sexual offender, sexual offender who has used physical violence in committing an offense or who has preyed upon children." ORDINANCE § 10-20(2). The stated intent of the ordinance is not to punish designated offenders, but to "promote, protect and improve the health, safety and welfare" of Paris's citizens "by creating areas around locations where children regularly congregate in concentrated numbers" wherein designated offenders "are prohibited from establishing residency." ORDINANCE § 10-19(3). It is also Paris's stated intent to impose the residency restrictions to "provide protection to children . . . by minimizing immediate access and proximity to children and thereby reducing opportunity and temptation for recidivism." Id. Paris determined that the restrictions would address its "compelling need to protect children where they congregate or play in public places." Id.

As relevant here, the ordinance imposes two discrete residency restrictions. The protected locations restriction prohibits designated offenders from establishing a residence "within six thousand five hundred (6,500) feet of a Protected Location." Id. § 10-21(1)(a). "Protected Locations" are defined as any school property, day care center, library, park, recreational trail, playground, athletic field used by children, place of worship, swimming pool, specialized school for children (e.g., gymnastics or dance academy), and any other place designated by Paris as a place where children congregate. Id. § 10-20(6). Although these locations are broadly defined, Paris has specifically designated only ten Protected Locations (such as the local school, preschool, and town hall). See Map of Protected Locations, ECF No. 31-3. The ordinance does provide a significant exception: a designated offender will not be held in violation of the ordinance if a new Protected Location opens within 6,500 feet of that person's already established and registered residence. ORDINANCE § 10-21(6)(c).

The ordinance also establishes the designated offenders restriction. This restriction prohibits designated offenders from residing "within a six thousand five hundred (6,500) foot radius of an existing [residence] of another Designated Offender." Id. § 10-21(1)(b).

Maps that depict the scope of the designated offenders restriction, as well as of the collective effect of the two restrictions, were produced in litigation. See Map of Designated Offenders Locations, ECF No. 31-4; Map of All 6,500-Foot Exclusion Zones, ECF No. 31-5. A designated offender who violates the ordinance faces a daily fine of $500. ORDINANCE § 10-24.

B. Peter Nelson

In June 2017, Nelson and his wife moved to the Bristol Motel, a 12-unit motel in Paris providing both nightly rentals and longer-term leases. The motel is not within 6,500 feet of any of Paris's protected locations. But in June 2019, to Nelson's surprise, he received a letter from the town notifying him that he was violating the ordinance because another designated offender lived within 6,500 feet of the motel. Nelson had never known about, met, or spoken to this other offender, who was one of three other designated offenders living in Paris.

Nelson sought an exemption through the appeals process provided by the ordinance, id. § 10-25, but his exemption was denied, and he later received a $500 citation from the Kenosha County Sheriff. Because he and his wife were unable to find another affordable home within Paris that complied with both provisions of the ordinance, they moved out of Paris to nearby Racine, Wisconsin, where they currently reside.

C. The Ordinance's Impact

In addition to the Bristol Motel, Paris has two other multi-unit motels that provide long-term rentals: the 11-unit Paris Motel, and the 21-unit Oasis Motel. Of these three motels, the Bristol Motel and the Paris Motel are unavailable to designated offenders. Although neither is located within 6,500 feet of a protected location, both are within 6,500 feet of another designated offender's residence.

The motels are the only multi-unit residences within Paris. The remaining housing stock in Paris is predominantly single-family homes. Indeed, approximately 75% of Paris is zoned as agricultural parcels, which, absent permission from the Wisconsin Department of Agriculture, may accommodate only one single-family house. The average market value of homes in Paris is $388,000, and the average of a home on an agricultural parcel is $426,600.

At the time Nelson filed his lawsuit, the combined effect of the two residency restrictions put 70.4% of Paris's total land area and 58.5% of its residential units off-limits to designated offenders. But this does not paint a complete picture of the restrictions' impact. As soon as a designated offender establishes a residence in one of the 41.5% of allowable residential units, a new 6,500-foot buffer zone is automatically created around that residence. That means, for example, that only one of the 21 units of the Oasis Motel is available for a designated offender. Once a single designated offender takes up residence in one of the motel's 21 units, the remaining 20 immediately become off-limits. And if and when another designated offender establishes a residence in whatever allowable residences remain, the percentage of allowable residences decreases materially.

II. DISCUSSION

We review the district court's order granting summary judgment to Paris de novo, viewing all facts and making all reasonable inferences in the light most favorable to Nelson. Hope v. Comm'r of Ind. Dep't of Corr., 9 F.4th 513, 523 (7th Cir. 2021).

A. Ex Post Facto Clause

The United States Constitution prohibits any state from passing ex post facto laws—those that "retroactively alter the definition of crimes or increase the punishment for criminal acts." Collins v. Youngblood, 497 U.S. 37, 43, 110 S.Ct. 2715, 111 L.Ed.2d 30 (1990). Thus, a law violates the Constitution if it is both retroactive and penal in nature. Koch v. Village of Hartland, 43 F.4th 747, 748 (7th Cir. 2022).

Both parties agree that, under our decision in Koch, the ordinance is retroactive because it "applies to [citizens] convicted for acts committed before the provision's effective date." 43 F.4th at 752 (quoting Weaver v. Graham, 450 U.S. 24, 31, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981)). Accordingly, the only question before us is whether the ordinance is punitive.

In answering this question, we look to the Supreme Court's decision in Smith, which sets forth a two-part inquiry. First, we ask "whether the legislature intended to enact a punitive, rather than civil, law." Hope, 9 F.4th at 530 (citing Smith, 538 U.S. at 92, 123 S.Ct. 1140). If we find that "the intention of the legislature was to impose punishment, that ends the inquiry"—the law is penal. Smith, 538 U.S. at 92, 123 S.Ct. 1140. On the other hand, if we determine that the legislature intended the law to be a civil one, we do not stop there. We proceed to assess whether the law is "so punitive either in purpose or effect as to negate [the legislature's] intention to deem it civil." Id. (cleaned up).

This inquiry requires the consideration of five factors: whether the law (1) "has been regarded in our history and traditions as a punishment"; (2) "imposes an affirmative disability or restraint"; (3) "promotes the traditional aims of punishment"; (4) "has a rational connection to a nonpunitive purpose"; or (5) "is excessive with respect to this purpose." Id. at 97, 123 S.Ct. 1140 (citing Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168-69, 83 S.Ct. 554, 9 L.Ed.2d 644 (1963)). The factors are "neither exhaustive nor dispositive, but are useful guideposts." Id. (cleaned up). Moreover, the Supreme Court has "expressly disapproved" focusing on "the effect that [the law] has on a single individual." Seling v. Young, 531 U.S. 250, 262, 121 S.Ct. 727, 148 L.Ed.2d 734 (2001) (citing Hudson v. United States, 522 U.S. 93, 118 S.Ct. 488, 139 L.Ed.2d 450 (1997)). Put another way, "an 'as-applied' analysis would prove unworkable" because "[s]uch an analysis would never conclusively resolve whether a particular scheme is punitive." Id. at 263, 121 S.Ct. 727. Therefore, while our analysis may be informed by the ordinance's impact on Nelson, we must consider whether the ordinance is punitive on its face.

Nelson rightly concedes that Paris intended its ordinance to be a civil regulatory scheme. See Smith, 538 U.S. at 92, 123 S.Ct. 1140 ("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.") (cleaned up). Here, the stated intent of the ordinance is to protect children, and it expressly disclaims any intent to impose additional punishment on sex offenders. ORDINANCE § 10-19(3). Given the significant deference owed to legislatures, we take Paris at its word. We will only question it if our review of the Smith factors leads us to conclude, by the "clearest proof," that the ordinance is so punitive in purpose or effect as to undermine the veracity of the claim. See Smith, 538 U.S. at 92, 123 S.Ct. 1140.

Before embarking on that analysis, however, we note as a prefatory matter that Paris's ordinance is severable. ORDINANCE § 10-23. Although we cannot ignore their combined impact, the two restrictions do not necessarily rise and fall together. We now turn to the five Smith factors.

1. Historical and Traditional Forms of Punishment

A civil scheme may be punitive when it resembles traditional forms of punishment: those that "h[o]ld the person up before his fellow citizens for face-to-face shaming or expel[ ] him from the community," such as "public shaming, humiliation, and banishment." Smith, 538 U.S. at 98, 123 S.Ct. 1140. The law need not impose burdens identical to a traditionally recognized form of punishment, so long as the burdens sufficiently resemble those typically associated with that punishment. See Does #1-5 v. Snyder, 834 F.3d 696, 703 (6th Cir. 2016) (law was "not identical to any traditional punishments" but met "the general definition of punishment" and had "much in common" with banishment, public shaming, and probation). Here, Nelson argues that the ordinance resembles the traditional punishments of supervised release and banishment.

a) Supervised Release

The Supreme Court acknowledged in Smith that a comparison between sex offender restriction laws and supervised release "has some force." 538 U.S. at 101, 123 S.Ct. 1140. Nevertheless, we have rejected such an argument twice, and we see no reason to veer from our precedents here. See Hope, 9 F.4th at 531-32; see also Vasquez v. Foxx, 895 F.3d 515, 521 (7th Cir. 2018), overruled on other grounds by Koch, 43 F.4th at 756. As in those cases, Paris's restrictions control only where designated offenders may live; they do not control any other aspects of their lives, such as where they may work or congregate, or with whom they can interact. See, e.g., Vasquez, 895 F.3d at 521. Moreover, as we noted in Hope, a key characteristic of supervised release is the supervisor's ability to seek revocation of the release based on the original offense. 9 F.4th at 531-32. By contrast, a designated offender's violation of the ordinance does not subject him to revocation, but instead to a separate $500 fine untethered to his original offense. Accordingly, in keeping with our prior holdings in Hope and Vasquez, we hold that neither the protected locations restriction nor the designated offender restriction resembles the traditional punishment of supervised release.

b) Banishment

A civil scheme resembles the traditional punishment of banishment when it forces individuals to leave their communities or to have difficulty finding a new one. See Smith, 538 U.S. at 98, 123 S.Ct. 1140; Vasquez, 895 F.3d at 521. In Vasquez, we concluded that Illinois's sex offender residency statute, which proscribed child sex offenders from knowingly residing within 500 feet of a "day care home" or "group day care home," did not amount to banishment because it "merely ke[pt] child sex offenders from living in very close proximity to places where children are likely to congregate; it d[id] not force them to leave their communities." 895 F.3d at 518, 521. We followed that same line of reasoning in Hope, concluding that the difference between Indiana's 1,000-foot restriction at issue there and the 500-foot restriction in Vasquez was "not constitutionally significant" because it did not render Indiana's statute "any more similar to banishment." Hope, 9 F.4th at 531.

Relying on these cases, the district court determined that Paris's restrictions do not amount to banishment because "[e]ven with the restrictions in place . . . more than 40% of the Town's housing stock [is] available," making it at most "difficult" for sex offenders to find housing in Paris. Nelson, 616 F. Supp. 3d at 851. Moreover, the district court noted, Paris's ordinance is less restrictive than those at issue in Vasquez and Hope in some ways, because it provides that individuals who have established residency need not move if a new protected location opens within 6,500 feet of their residence. Id. at 852.

Nelson argues that Vasquez and Hope should not control here. As he sees it, unlike in those cases, the collective effect of the two restrictions eliminates the only affordable, long-term rental units in Paris (i.e., the three motels). This, he asserts, effectively banishes designated offenders who are likely unable to afford a single-family home within the town. Nelson also contends that the ordinance amounts to banishment because it makes it impossible for more than a few designated offenders to live in Paris, given that no offender can live within 6,500 feet (i.e., 1.23 miles) of another offender (or a protected location) in a town that is only roughly six-by-six miles.

We are not persuaded that the protected locations restriction, on its own, rises to the level of banishment. We acknowledge that a buffer of 6,500 feet is six and a half times larger than the one we allowed in Hope. But the inquiry is not limited to the mere size of the restriction, but how the restriction affects an offender's ability to live in the town. Here, the protected locations restriction, taken alone, impacts less than 30% of the total housing stock of Paris and leaves the three motels available to designated offenders. Although the radius of the residency restriction may border on being excessive and is much larger than any we have addressed, it does not have the effect of banishing designated offenders from the town. This is particularly so given the exception in the ordinance that allows designated offenders to remain in their residences if a new protected location opens nearby.

The designated offenders restriction, on the other hand, looks much more akin to banishment. On its face, this restriction—whether taken alone or in combination with the protected locations restriction—creates a possible future in which no new offenders will be permitted to live within Paris at all, because any available residences will be within 6,500 feet of another designated offender (or a protected location). It is true that the restriction will not oust designated offenders from their current homes in Paris, but as the Supreme Court explained, banishment traditionally means that a person "could neither return to their original community nor . . . be admitted easily into a new one." Smith, 538 U.S. at 98, 123 S.Ct. 1140 (emphasis added) (citation omitted); accord Hope, 9 F.4th at 531; Vasquez, 895 F.3d at 521. The designated offenders restriction creates a ceiling on the number of designated offenders that may be able to reside in Paris and, therefore, has a direct impact on the ability of designated offenders to move into Paris. As for this restriction, the first factor favors Nelson.

2. Affirmative Disabilities or Restraints

Next, we consider whether the restrictions subject designated offenders to an "affirmative disability or restraint." This factor focuses on how the effects of the law are felt by those subjected to it. Smith, 538 U.S. at 99-100, 123 S.Ct. 1140. "If the disability or restraint is minor and indirect, its effects are unlikely to be punitive." Id. at 100, 123 S.Ct. 1140. The paradigmatic example of an affirmative disability or restraint is imprisonment—a physical restraint. Id.

In the context of residency restrictions, some courts have found that the inability or substantial difficulty in selecting or changing residences due to the restriction may be considered a direct and substantial restraint. See Doe v. Miami-Dade County, 846 F.3d 1180, 1185 (11th Cir. 2017); see also Snyder, 834 F.3d at 703 (restrictions on where offenders may live, work, and loiter, as well as requiring in-person registration, "are direct restraints on personal conduct"); Hoffman, 249 F. Supp. 3d at 958 (restrictions limiting designated offenders to ten percent of Village's land area, most of which was non-residential, were "severe restraints"). Recognizing that "[t]he boundaries of this factor are undefined," however, we have declared that "very few burdens are significant enough to tip the scale" of this factor in favor of the challenger. Hope, 9 F.4th at 532.

Neither of the restrictions at issue impose any burdens significant enough to "tip the scale" in favor of Nelson. The only restraint imposed by the restrictions—whether taken together or in isolation—relates to where a designated offender may live. The ordinance does not, for example, force offenders to leave their homes if a new protected location opens within 6,500 feet of their established residence. See Vasquez, 895 F.3d at 518, 522 (finding no affirmative disability or restraint where plaintiffs were forced to move after a new child day care opened within 500 feet of their homes). Indeed, it does not force offenders to leave their homes at all, let alone impose an explicit ban on new offenders moving into Paris. See Hoffman, 249 F. Supp. 3d at 958 (permanently banning offenders who were neither residents of the town at the time of their most recent offense nor at the time the ordinance was passed). Nor does it restrict where designated offenders can work or require offenders to appear in person to register. See Snyder, 834 F.3d at 703. And it does not otherwise "restrain activities sex offenders may pursue but leaves them free to change jobs or residences." Smith, 538 U.S. at 100, 123 S.Ct. 1140. This factor favors Paris.

3. Promoting Traditional Aims of Punishment

The next factor does not sway us one way or the other. As we—and many other courts—have noted, determining whether sex offender residency restrictions promote the traditional aims of punishment provides little value to the over-all Smith inquiry. This is because all such regulations inevitably overlap with the traditional aims of punishment, such as deterrence, retribution, or incapacitation. See Smith, 538 U.S. at 102, 123 S.Ct. 1140 ("Any number of governmental programs might deter crime without imposing punishment."); Hope, 9 F.4th at 533 (noting we have been "unpersuaded that . . . residency restrictions . . . further[ ] traditional punitive aims" because they have the "obvious aim" of protecting children); Snyder, 834 F.3d at 704 (the law's "very goal is incapacitation insofar as it seeks to keep sex offenders away from opportunities to reoffend"); Hoffman, 249 F. Supp. 3d at 958 (this factor "is of limited importance because punishment goals often overlap with legitimate civil regulatory goals"). Accordingly, this factor typically favors the municipality where such "punitive aims" flow naturally from the civil scheme's non-punitive goals.

That is precisely the case here. Paris's ordinance pursues its legitimate interests in protecting children through residency restrictions that seek to deter recidivism. This factor favors Paris as to both restrictions, although we afford it little weight for the reasons noted.

4. Rational Connection to Nonpunitive Purpose

"Whether the law has a 'rational connection to a nonpunitive purpose' is 'a most significant factor in our determination that the statute's effects' are not punitive." Hope, 9 F.4th at 533 (quoting Smith, 538 U.S. at 102, 123 S.Ct. 1140). Our inquiry begins with identifying a nonpunitive purpose and then determining whether the ordinance's requirements are rationally connected to that purpose. Id. We will not fault the legislature for lacking a "close or perfect fit" with its nonpunitive aims, so long as any imprecision in the ordinance "does not suggest that the Act's nonpunitive purpose is a 'sham or mere pretext.' " Smith, 538 U.S. at 103, 123 S.Ct. 1140 (quoting Kansas, 521 U.S. at 371, 117 S.Ct. 2072 (Kennedy, J., concurring)).

The goal of Paris's ordinance is to protect children from harm. Specifically, its stated intent is to "promote, protect and improve the health, safety and welfare of the citizens of the Town by creating areas around locations where children regularly congregate in concentrated numbers wherein certain sex offenders and sex predators are prohibited from establishing residency." ORDINANCE § 10-19(3). It also seeks "to provide protection to children in the Town by minimizing immediate access and proximity to children and thereby reducing opportunity and temptation for recidivism," and to "protect children where they congregate or play in public places." Id.

Protecting children is a legitimate nonpunitive purpose, and Nelson concedes for the purposes of appeal that the restrictions at issue have a rational connection to that purpose. Given that Nelson provides no basis for us to question the stated rationale of the ordinance, we find this factor to favor Paris as to both restrictions.

5. Excessive With Respect to Nonpunitive Purpose

Instead of challenging the ordinance's rational connection to its purpose of protecting children, Nelson places his stock in the final factor, arguing that the restrictions are excessive with respect to that purpose. At this step, which is related to the fourth factor, see Vasquez, 895 F.3d at 522, we do not ask "whether the legislature has made the best choice possible to address the problem it seeks to remedy." Smith, 538 U.S. at 105, 123 S.Ct. 1140. Rather, "[t]he question is whether the regulatory means chosen are reasonable in light of the nonpunitive objective." Id. (emphasis added).

We begin with the protected locations restriction. Nelson acknowledges that courts have regularly upheld these types of restrictions for various distances. E.g., Vasquez, 895 F.3d at 522 (upholding a 500-foot restriction); Hope, 9 F.4th at 534 (upholding a 1,000-foot restriction); Doe v. Miller, 405 F.3d 700, 722-23 (8th Cir. 2005) (upholding a 2,000-foot restriction). But Nelson argues that there is no precedent for upholding a 6,500-foot restriction and that the town has failed to show that such a large radius is necessary or that it would be effective in promoting its aims. In response, Paris correctly notes that it need not provide empirical evidence of efficacy to support its legislative decisions. See F.C.C. v. Beach Commc'ns, Inc., 508 U.S. 307, 315, 113 S.Ct. 2096, 124 L.Ed.2d 211 (1993) ("[A] legislative choice is not subject to courtroom fact-finding and may be based on rational speculation unsupported by evidence or empirical data."). Instead, Paris gives three reasons to justify the scope of its protected locations restriction: the rural nature of the town, the frequent travel of children within the town without adult supervision, and the lack of police presence in the town.

We see no need to parse Paris's rationales with a fine-toothed comb. Rather, we are persuaded by the fact that Paris's 6,500-foot protected locations restriction, taken alone, impacts less than 30% of the available housing stock in Paris. That leaves over 70% of the town's residential units—including its affordable long-term rental motels—available to designated offenders. While Nelson is correct that (as far as we can tell) no court has ever upheld, let alone had an opportunity to evaluate, a 6,500-foot residency restriction, he has failed to show by the "clearest proof" that such a buffer around the ten protected locations in Paris is constitutionally excessive. Indeed, we suspect that if this were the only restriction at issue, we would not be here today. Thus, we find the excessiveness factor to favor Paris as to this restriction.

The designated offenders restriction, however, is a different story. Unlike the protected locations restriction, there is no precedent for this type of residency restriction, and the record offers no explanation why prohibiting designated offenders from residing within 6,500 feet of one another would safeguard Paris's children, particularly given the safeguards already in place by the protected locations restriction.

In enacting the ordinance, Paris did determine that sex offenders were "extremely likely" to recidivate in "locations close to their residences." ORDINANCE § 10-19(2). We defer to these legislative findings. See Frank v. Walker, 768 F.3d 744, 750 (7th Cir. 2014); see also United States v. Kebodeaux, 570 U.S. 387, 395-96, 133 S.Ct. 2496, 186 L.Ed.2d 540 (2013); Turner Broad. Sys., Inc. v. F.C.C., 520 U.S. 180, 195, 117 S.Ct. 1174, 137 L.Ed.2d 369 (1997). But, even if these findings were true and even if, as Paris suggests, residential proximity might allow greater interaction between designated offenders which, in turn, might lead to greater recidivism (a debated proposition), it is entirely unclear from this record what this restriction adds to the security the protected locations restriction already provides to the town's children. See Turner Broad. Sys., Inc. v. F.C.C., 512 U.S. 622, 664, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994) ("That the Government's asserted interests are important in the abstract does not mean, however, that the [regulation at issue] will in fact advance those interests.").

Accordingly, after drawing all reasonable factual inferences in Nelson's favor, we conclude that disputed facts exist as to whether the designated offenders restriction is reasonably related to Paris's legitimate interest in protecting its children.

In summary, we conclude that Nelson has failed to establish by the "clearest proof" that Paris's protected locations restriction violates the Constitution's prohibition on ex post facto laws. Nelson has established, however, that the designated offenders restriction resembles the traditional form of banishment because it effectively establishes a ceiling beyond which no designated offender could ever reside in Paris. And the factual record, as it currently stands, leaves genuine disputes of fact as to the reasonableness of this restriction when considered against Paris's stated goal of protecting its children. We therefore remand this case to the district court for an evidentiary hearing and further consideration of the designated offenders restriction. See Turner Broad., 512 U.S. at 668, 114 S.Ct. 2445 (remanding for factual development where the "paucity of evidence" precluded determination as to the constitutionality of the challenged statute).

B. Substantive Due Process

The remaining issue is easily dispatched. Nelson argues that the ordinance violates his substantive right to due process protected by the Fourteenth Amendment. He acknowledges that his claim is subject to rational basis review because he does not allege that the ordinance infringes on any fundamental right. Nevertheless, he urges us to apply some level of heightened scrutiny, arguing that sex offenders are a despised minority.

We rejected this exact argument in Vasquez. 895 F.3d at 524-25. There, when asked to apply heightened scrutiny to the residency restrictions at issue, we declined to do so, explaining that because "[t]he residency statute [wa]s facially neutral and advance[d] a compelling government interest: protecting children from recidivism by child sex offenders," heightened scrutiny did not apply. Id. at 525. We find the same to be true here. There is no facial animus toward sex offenders in the ordinance, the purpose of which is to protect children. Therefore, only rational-basis review is appropriate. And given that Nelson concedes that the ordinance is rationally related to Paris's interest in protecting children, his due process claim necessarily fails.

III. CONCLUSION

For the foregoing reasons, we AFFIRM the district court's grant of summary judgment in favor of Paris on Nelson's Fourteenth Amendment claim and his Ex Post Facto Clause claim to the extent it applies to the protected locations restriction, ORDINANCE § 10-21(1)(a). But we hold that the district court erred in granting summary judgment in favor of Paris on Nelson's Ex Post Facto Clause claim as it applies to the designated offenders restriction, ORDINANCE § 10-21(1)(b). We therefore VACATE in part and REMAND to the district court for proceedings consistent with this opinion.