In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 22-1007
KARSTEN KOCH,
Plaintiff-Appellant,
v.
VILLAGE OF HARTLAND,
Defendant-Appellee.
____________________
Appeal from the United States District Court for the
Eastern District of Wisconsin.
No. 21-cv-503 — William E. Duffin, Magistrate Judge.
____________________
ARGUED MAY 27, 2022 — DECIDED AUGUST 8, 2022
____________________
Before ST. EVE, KIRSCH, and JACKSON-AKIWUMI, Circuit
Judges.
ST. EVE, Circuit Judge. The Village of Hartland, Wisconsin
(“the Village”) passed an Ordinance (“the Ordinance”) plac-
ing a moratorium against any new sex offenders residing
there, whether on a temporary or permanent basis. Karsten
Koch is a registered sex offender who would like to move into
the Village to be closer to work and family. He sued the
2 No. 22-1007
Village over the Ordinance, asserting that it violated the Ex
Post Facto Clause of Article I, Section 10 of the Constitution.
A law must be both retroactive and penal to transgress the
Ex Post Facto Clause. Ruling for the Village on cross-motions
for summary judgment, the district court concluded that the
retroactivity rule from two Seventh Circuit opinions—United
States v. Leach, 639 F.3d 769 (7th Cir. 2011) and Vasquez v. Foxx,
895 F.3d 515 (7th Cir. 2018)—controlled. Under this prece-
dent, a law is not retroactive, and therefore cannot violate the
Ex Post Facto Clause, if it applies “only to conduct occurring
after its enactment.” Id. at 520. The Ordinance, then, applied
prospectively, and there was no need to determine whether it
was also penal.
While the district court faithfully applied circuit prece-
dent, we no longer believe the Leach-Vasquez rule governing
retroactivity is tenable. We reverse and remand; the Ordi-
nance is retroactive. The district court, on remand, must con-
sider in the first instance whether it is “punitive.”
I.
A.
On September 24, 2018, the Village of Hartland, Wisconsin
enacted Ordinance No. 850-18, which prohibited the estab-
lishment of “Temporary or Permanent Residence” by a “Des-
ignated Offender,” that is, a sex offender, within the Village
“until such time as the saturation level for Designated Offend-
ers in the Village of Hartland reaches a factor of 1.1 or lower
….” A sex offender is a “person who has been convicted of …
a sexually violent offense and/or a crime against children.”
The “saturation level” is
No. 22-1007 3
determined by adding the number of Designated Of-
fenders per square mile in Hartland plus the number
of Designated Offenders per 1,000 population in Hart-
land and dividing the resulting figure by the sum of
the number of Designated Offenders per square mile
in Waukesha County net of Hartland plus the number
of Designated Offenders per 1,000 population in
Waukesha County net of Hartland.
At the time the Ordinance went into effect, Hartland’s satura-
tion level was 6.75.
According to the Ordinance’s “findings and intent” sec-
tion, the Village recently learned that there were thirty-five
sex offenders living within the Village, an allegedly high
number compared to neighboring areas. The ordinance was a
“regulatory measure aimed at protecting the health and
safety of the children of the Village of Hartland from the risk
that a convicted sex offender may re-offend in locations close
to a Designated Offender’s residence ….” The U.S. Supreme
Court has “recognized that the risk of recidivism posed by sex
offenders is high and when convicted sex offenders re-enter
society, they are much more likely than any other type of of-
fender to be rearrested for a new rape or sexual assault.” That
sex offenders “suffer a high rate of recidivism,” the Village
believed, “has a basis in fact,” and they collectively “are a se-
rious threat to public safety,” pose specific dangers to chil-
dren, and “are more likely to use physical violence.” The “po-
tential of psychological trauma to citizens of the Village is real
but difficult to calculate.”
The Village represents that the moratorium allows local
police more time and flexibility in developing its dedicated
community policing program to give officers a chance to
4 No. 22-1007
monitor sex offenders, address experiences, and decrease “re-
cidivism and community conflict.” “As a result of the commu-
nity policing program and the moratorium,” the Village
maintains, “resident designated offenders have not commit-
ted any sex offenses in the Village.” The ordinance also gives
more time to pass “a sex offender residency ordinance that
will satisfy Constitutional requirements.”
B.
Koch is a registered sex offender. Before the Ordinance
was passed, he was convicted of one count of engaging in re-
peated acts of sexual assault on a child and two counts of sec-
ond-degree sexual assault of a child. He served seven years in
prison before being released. His convictions qualify him as a
“Designated Offender” under the Ordinance.
Since his conviction, Koch has worked to get his life on a
positive track. He found employment and now wishes to live
in Hartland to be closer to work and family, as the Village
provides more suitable rental properties than the town where
he currently resides. A property owner was even willing to
rent to Koch, but the Village’s Ordinance prevents any land-
lord from doing so. Instead, Koch must continue to live with
his parents and commute a longer distance to work.
C.
Koch sued the Village, alleging that the Ordinance de-
prived him of a constitutional right under the Ex Post Facto
Clause by criminally punishing his conduct before its enact-
ment. See 42 U.S.C. § 1983. Both parties moved for summary
judgment, and the district court granted the Village’s motion.
The Ex Post Facto Clause proscribes “retroactive punish-
ment.” For a law to violate this protection, it must be
No. 22-1007 5
retroactive and punitive. The district court only considered
the retroactivity prong of the two-part test because two Sev-
enth Circuit opinions—Leach and Vasquez—dictated the out-
come. Under our precedent, a law creating only “new, pro-
spective legal obligations” is not retroactive. Therefore, the
Ordinance operates only prospectively because it “limits a
Designated Offender’s housing options based on [] prior his-
tory.” “In other words, the Ordinance only applies to Koch’s
current desire to move to Hartland.” The district court could
not “accept Koch’s invitation to reject Leach and Vasquez and
follow the reasoning employed by other circuits when consid-
ering Ex Post Facto Clause challenges.” And because the law
was not retroactive, the district court did not need to consider
whether it punished the targeted offenders.
Koch filed a timely appeal. We review a grant of summary
judgment de novo, drawing all reasonable inferences “in the
light most favorable to the nonmoving party on each motion.”
Birch|Rea Partners, Inc. v. Regent Bank, 27 F.4th 1245, 1249 (7th
Cir. 2022) (quoting Lalowski v. City of Des Plaines, 789 F.3d 784,
787 (7th Cir. 2015)).
II.
The Constitution provides that “[n]o State shall … pass
any … ex post facto Law,” U.S. Const. art. I, § 10, cl. 1, defined
as an act that “retroactively alter[s] the definition of crimes or
increase[s] the punishment for criminal” deeds, Cal. Dep’t of
Corr. v. Morales, 514 U.S. 499, 504 (1995) (citing Calder v. Bull,
3 U.S. (3 Dall.) 386, 391–92 (1798)). See also U.S. Const. art. I,
§ 9, cl. 3 (“No … ex post facto Law shall be passed.”). “Statutes
that transgress the Ex Post Facto Clause [] share two charac-
teristics: They are ‘both retroactive and penal.’” Hope v. Comm’r
6 No. 22-1007
of Ind. Dep’t of Corr., 9 F.4th 513, 530 (7th Cir. 2021) (en banc)
(quoting Vasquez, 895 F.3d at 520).
A.
We have held that a regulatory scheme applying “only to
conduct occurring after the law’s enactment” is merely pro-
spective and thus cannot violate the Ex Post Facto Clause.
Vasquez, 895 F.3d at 520. Koch concedes that under current
precedent the Ordinance is not retroactive because it targets
only future conduct, that is, taking up residency in the Village.
Nonetheless, he urges us to overturn this rule, which conflicts
with the history and values of the Ex Post Facto Clause, Su-
preme Court precedent, and the consensus among other cir-
cuits and state courts. We take this opportunity to overrule it.1
1.
The Ex Post Facto Clause safeguards the legal principle
that there can be no punishment without law, nulla poena sine
lege. This maxim has a long history. The Digest of Justinian, a
sixth-century codification of Roman law, declared, “The pen-
alty for a past wrong is never increased ex post facto.” Robert
G. Natelson, Statutory Retroactivity: The Founders’ View, 39
Idaho L. Rev. 489, 500 (2003). William Blackstone, in his influ-
ential Commentaries on the Laws of England, recounted with
horror how the emperor Caligula posted laws that could not
be seen by the public and then prosecuted his subjects. 1 Wil-
liam Blackstone, Commentaries *46; see generally Evan C.
Zoldan, The Civil Ex Post Facto Clause, 2015 Wis. L. Rev. 727,
1 Because this opinion overrules circuit precedent, we circulated it to all
active members of the court under Circuit Rule 40(e). A majority of judges
did not wish to rehear the case en banc.
No. 22-1007 7
737. An ex post facto law, he described, happens when, “after
an action is committed,” a rogue legislator “for the first time
declares it to have been a crime.” 1 William Blackstone, Com-
mentaries *46. “It is impossible that the party could foresee that
an action, innocent when it was done, should be afterwards
converted to guilt by a subsequent law; he had therefore no
cause to abstain from it; and all punishment for not abstaining
must of consequence be cruel and unjust.” Id.
Despite this venerable history, Parliament still enacted
acts that resembled ex post facto laws. David F. Forte & Mat-
thew Spalding, The Heritage Guide to the Constitution 203.
Against this backdrop, the Framers included the Ex Post Facto
Clauses in the Constitution. See U.S. Const. art. I, §§ 9–10.
These provisions garnered prominent support from early
leaders and jurists. James Madison opined that “ex-post-facto
laws … are contrary to the first principles of the social com-
pact, and to every principle of sound legislation.” The Federal-
ist No. 44 (James Madison). The constitutional protections
against such laws then, Alexander Hamilton posited, provide
essential “securities to liberty and republicanism.” The Feder-
alist No. 84 (Alexander Hamilton). Justice Samuel Chase ob-
served, “The prohibition against their making any ex post
facto laws was introduced for greater caution, and very prob-
ably arose from the knowledge, that the Parliament of Great
Britain claimed and exercised a power to pass such laws …
inflicting … punishment.” Calder, 3 U.S. at 389. Examples in-
clude “declaring acts to be treason, which were not treason,
when committed, at other times, [] violat[ing] the rules of ev-
idence (to supply a deficiency of legal proof) … inflict[ing]
punishments, where the party was not, by law, liable to any
punishment; and in other cases, [inflicting] greater punish-
ment, than the law annexed to the offence.” Id.
8 No. 22-1007
The clauses serve at least two purposes. See Stogner v. Cal-
ifornia, 539 U.S. 607, 612 (2003); see also Wayne A. Logan, The
Ex Post Facto Clause and the Jurisprudence of Punishment, 35 Am.
Crim. L. Rev. 1261, 1276 (1998). First, they prevent govern-
ments from promulgating “manifestly unjust and oppres-
sive” laws. Calder, 3 U.S. at 391. A “Constitution that permits
… legislatures to pick and choose when to act retroactively[]
risks both ‘arbitrary and potentially vindictive legislation,’
and erosion of the separation of powers.” Stogner, 539 U.S. at
611 (quoting Weaver v. Graham, 450 U.S. 24, 29 (1981)); see also
Landgraf v. USI Film Prods., 511 U.S. 244, 266 (1994) (“[A legis-
lature’s] responsivity to political pressures poses a risk that it
may be tempted to use retroactive legislation as a means of
retribution against unpopular groups or individuals.”). Sec-
ond, they ensure that a defendant is given “fair warning” to
allow him to conform his actions before embarking upon a
course of conduct. Carmell v. Texas, 529 U.S. 513, 533 (2000). At
bottom, the government must always “play by its own rules.”
Id.
2.
The Supreme Court has recognized the values embodied
by the Ex Post Facto Clauses and adhered to a broader under-
standing of retroactivity than used by our court.2 The clearest
formulation of the retroactivity inquiry—and the one we
2 See, e.g., Stogner, 539 U.S. at 612; Smith v. Doe, 538 U.S. 84, 92–95 (2003);
Carmell, 529 U.S. at 533; Morales, 514 U.S. at 504–05; Collins v. Youngblood,
497 U.S. 37, 41–42 (1990); Weaver, 450 U.S. at 28; Dobbert v. Florida, 432 U.S.
282, 298 (1977); Lindsey v. Washington, 301 U.S. 397, 401 (1937); Rooney v.
North Dakota, 196 U.S. 319, 324–25 (1905); In re Medley, 134 U.S. 160, 171
(1890); Gut v. Minnesota, 76 U.S. 35, 37 (1869); Cummings v. Missouri, 71 U.S.
(4 Wall.) 277, 325–26 (1866).
No. 22-1007 9
adopt today—comes from Weaver v. Graham, 450 U.S. 24.
There, Florida charged the prisoner with a crime on January
31, 1976, and he was convicted six months later. Id. at 26. At
the time, state law provided a system of “gain-time credits”
for every prisoner who maintained good behavior. Id. The
state then passed a new law that went into effect January 1,
1979, making it more difficult to receive these credits; it ap-
plied to all prisoners, even those sentenced before its effective
date. Id. at 27. In defending the law against an Ex Post Facto
challenge, the state argued that, given its effective date, it was
not retrospective. Id. at 28. The Supreme Court soundly re-
jected that position—“it is the effect, not the form, of the law
that determines whether it is ex post facto.” Id. When analyzing
whether a law is retroactive, “[t]he critical question is whether
the law changes the legal consequences of acts completed be-
fore its effective date.” Id. A law that “applies to [citizens] con-
victed for acts committed before the provision’s effective
date” is retroactive. Id.
More recently, Vartelas v. Holder, 566 U.S. 257 (2012), reaf-
firmed this understanding of retroactivity. In 1996, Congress
overhauled the immigration system by enacting the Illegal
Immigration Reform and Immigrant Responsibility Act
(“IIRIRA”). Id. at 260. The law abolished the old two types of
proceedings, deportation hearings and exclusion hearings,
and replaced them with a simple “removal proceeding.” Id. at
261–62. Under the old system, “lawful permanent residents
who had committed a crime of moral turpitude could … re-
turn from brief trips abroad without applying for admission
to the United States.” Id. at 263. The question before the court
was whether IIRIRA applied to lawful permanent residents
who committed crimes of moral turpitude before the Act’s ef-
fective date. Id. at 265. The dissenting justices did not believe
10 No. 22-1007
the law applied retroactively. Id. at 276–77 (Scalia, J., dissent-
ing) (“Although the class of aliens affected by
§ 1101(a)(13)(C)(v) is defined with respect to past crimes, the
regulated activity is reentry into the United States.”). The ma-
jority disagreed. “The presumption against retroactive legis-
lation … ‘embodies a legal doctrine centuries older than our
Republic.’” Id. at 266 (quoting Landgraf, 511 U.S. at 263). The
case presented a “firm” example of why the principle exists.
Id. at 267. “Neither [the] sentence, nor the immigration law in
effect when [the immigrant] was convicted and sentenced,
blocked him from occasional visits” abroad. Id. Applying the
provision of IIRIRA “would thus attach ‘a new disability’ to
conduct over and done well before the provision’s enact-
ment.” 3 Id. The majority reemphasized the Weaver principle,
“The essential inquiry … is ‘whether the new provision at-
taches new legal consequences to events completed before its
3 In suggesting a different test, the concurrence places undue weight on
statements from the dissenting opinion in Vartelas, which the Court never
adopted. The majority opinion, however, is entirely consistent with
Weaver: a law that attaches a “new disability,” Vartelas, 566 U.S. at 267,
changes the “legal consequences of acts completed before its effective
date,” Weaver, 450 U.S. at 31. If any additional confirmation is needed, it
then expressly incorporated Weaver’s language. Vartelas, 566 U.S. at 273
(quoting Landgraf, 511 U.S. at 269–70). Proclamations about unrelated
laws, including footnote seven, were dicta. Only a single provision of
IIRIRA was before the Court—opining about different statutes was not
germane to the question presented. Vartelas, 566 U.S. at 265; see also id. at
270–72 (suggesting—but not adopting—different theories for RICO, Fer-
nandez-Vargas, § 922(g), and recidivism sentencing enhancements). To the
extent the concurrence is worried about laws not before us, we “observe
the wise limitations on our function and [] confine ourselves to deciding
only what is necessary to the disposition of the immediate case.”
Whitehouse v. Illinois Cent. R. Co., 349 U.S. 366, 372–73 (1955).
No. 22-1007 11
enactment.’ That is just what occurred here.” Id. at 273 (quot-
ing Landgraf, 511 U.S. at 269–70).
The Court has also implicitly acknowledged, in Smith v.
Doe, 538 U.S. 84 (2003), that sex-offender laws applying to
people with convictions before the effective date are retroac-
tive. Smith involved Alaska’s Sex Offender Registration Act,
which had two components: a registration requirement and a
notification system. Id. at 89–90. A sex offender had to register
in the state promptly and verify the information on a regular
basis. Id. Alaska chose to publish much of this information on
the internet. Id. at 91. Two sex offenders challenged the law,
and the Ninth Circuit agreed that the system violated the Ex
Post Facto Clause. The Supreme Court reversed because the
law did not punish pre-Act offenders, as the scheme had nei-
ther the purpose nor the effect of punishing past conduct. Id.
at 92–95. But the Court never considered the possibility that
the law was only prospective, noting even the components of
the system were “retroactive.” Id. at 90.
Other circuits employ a similar retroactivity analysis and
recognize that sex-offender laws applying to conduct before
their enactment date are retroactive. See, e.g., Doe v. Pataki, 120
F.3d 1263, 1265, 1273 (2d Cir. 1997) (acknowledging implicitly
that the law was retroactive and assessing then “whether a
retroactively imposed burden constitutes ‘punishment’
within the meaning of the Ex Post Facto Clause”); E.B. v. Ver-
niero, 119 F.3d 1077, 1092 (3d Cir. 1997) (examining a sex of-
fender law and asking whether the state “has inflicted ‘pun-
ishment’”); United States v. Young, 585 F.3d 199, 203 (5th Cir.
2009) (per curiam) (“Indeed, for a penal law to be considered
ex post facto, it ‘must apply to events occurring before its en-
actment.’ Although SORNA does relate to old conduct that
12 No. 22-1007
was criminal when done, the question is whether SORNA
punishes this old conduct.”); Does 1-5 v. Snyder, 834 F.3d 696,
698 (6th Cir. 2016) (“It is undisputed on appeal that SORA’s
2006 and 2011 amendments apply to them retroactively. That
law has had a significant impact on each of them that reaches
far beyond the stigma of simply being identified as a sex of-
fender on a public registry.”); Doe v. Miller, 405 F.3d 700, 718–
19 (8th Cir. 2005) (“A final, and narrower, challenge advanced
by the Does is that § 692A.2A is an unconstitutional ex post
facto law because it imposes retroactive punishment on those
who committed a sex offense prior to July 1, 2002.”); Russell v.
Gregoire, 124 F.3d 1079, 1083 (9th Cir. 1997) (“[T]he focus of
the ex post facto inquiry is not on whether a legislative change
produces some sort of ‘disadvantage,’ … but on whether any
such change alters the definition of criminal conduct or in-
creases the penalty by which a crime is punishable.” (quoting
Morales, 514 U.S. 499, 506 n.3); Shaw v. Patton, 823 F.3d 556, 560
(10th Cir. 2016) (“In 1998, when Mr. Shaw was convicted, Ok-
lahoma did not have any residency or loitering restrictions for
sex offenders. … Thus, Mr. Shaw is subject to restrictions on
reporting, residency, and loitering only because Oklahoma
changed its laws years after Mr. Shaw’s criminal conduct. By
definition, these restrictions are being retroactively applied to
Mr. Shaw. The resulting issue is whether these restrictions
constitute punishment.”). 4
4 While two circuits have cited footnote seven from Vartelas, those state-
ments were cursory, and later cases in both respective circuits turned back
to Weaver when analyzing retroactivity under the Ex Post Facto Clauses.
See, e.g., McGill v. Shinn, 16 F.4th 666, 700–01 (9th Cir. 2021); Rhines v.
Young, 899 F.3d 482, 495 (8th Cir. 2018); Nevarez v. Barnes, 749 F.3d 1124,
1128 (9th Cir. 2014).
No. 22-1007 13
The Tenth Circuit’s analysis in Shaw v. Patton, 823 F.3d 556,
offers one illustrative example. After Oklahoma passed a law
limiting sex offenders from living within certain feet of
schools, playgrounds, parks, and childcare centers, the plain-
tiff claimed that the law violated the Ex Post Facto Clause. Id.
at 559–60. The state argued that because these obligations did
not apply until an offender entered Oklahoma, the law was
not retroactive. Id. at 560. That argument, though, was quickly
dismissed. Id. While an offender might not be subject to a law
until he moves into the state, “the date of his move does not
affect whether the statute is being enforced retroactively.” Id.
“A statute is enforced retroactively,” the Tenth Circuit rea-
soned, “if it governs conduct that preceded the statute’s en-
actment.” Id.
States, when interpreting a federal or state ex post facto
clause, have reached the same conclusion. See, e.g., Gonzalez v.
State, 980 N.E.2d 312, 316 (Ind. 2013) (“This provision prohib-
its, in relevant part, the passage of any law ‘which imposes a
punishment for an act which was not punishable at the time
it was committed; or imposes additional punishment to that
then prescribed.’” (quoting Weaver, 450 U.S. at 28); Common-
wealth v. Baker, 295 S.W.3d 437, 442 (Ky. 2009) (“As a threshold
question, for a law to be considered ex post facto, ‘it must be
retrospective, that is, it must apply to events occurring before
its enactment, and it must disadvantage the offender affected
by it.’”); State v. Letalien, 985 A.2d 4, 13–15 (Me. 2009) (summa-
rizing retroactivity law and moving to whether the law pun-
ishes sex offenders, as the state made that principal argu-
ment); State v. Walls, 558 S.E.2d 524, 525–26 (S.C. 2002) (“First,
the law must be retroactive so as to apply to events occurring
before its enactment. … The Act meets the first prong of de-
termining whether it falls within ex post facto prohibitions. The
14 No. 22-1007
Act is retroactive because it applies to events occurring before
its enactment. In particular, it applies to appellant whose of-
fense was committed in 1973, prior to the enactment of the
Act.”); Kitze v. Commonwealth, 475 S.E.2d 830, 832 (Va. 1996)
(acknowledging implicitly that the law was retroactive and
holding that “the sex offender registration requirement is not
penal and that the General Assembly ‘intended to facilitate
law enforcement and protection of children.’” (citation omit-
ted)).
3.
Our caselaw has departed from this history and judicial
consensus. Our departure began in United States v. Leach. 639
F.3d 769. There, Donald Leach, a convicted sex offender, chal-
lenged his conviction under the federal Sex Offender Regis-
tration and Notification Act (“SORNA”), which requires all
sex offenders to register in the jurisdiction where they work,
reside, and attend school, as contravening the Ex Post Facto
Clause. Id. at 770–71. The registration requirements of
SORNA, we concluded, however, were not retrospective be-
cause the law “merely create[d] new, prospective legal obli-
gations based on the person’s prior history.” Id. at 773. Thus,
Leach’s argument failed for that reason alone. Id. We also re-
marked that Smith v. Doe proved indistinguishable, as both
involved similar registration requirements that did not vio-
late the Constitution. Id. In reaching this outcome, we joined
every other circuit to have considered SORNA’s residency re-
quirements. Id.
Vasquez v. Foxx, 895 F.3d 515, then adopted the retroactiv-
ity analysis employed in Leach. At issue in the case was Illi-
nois’s residency restriction for child sex offenders, which pro-
hibited them from residing within 500 feet of facilities with
No. 22-1007 15
children, such as daycares. Id. at 518. Joshua Vasquez raised
several challenges, including one based on the Ex Post Facto
Clause. Id. at 520. We concluded that the law was neither ret-
roactive nor punitive. See id. at 520–22. A regulatory scheme
that “applies only to conduct occurring after the law’s enact-
ment” is prospective only. 5 Id. at 520. Like in Leach, the Illinois
residency statute’s “requirements and any criminal penalty
apply only to conduct occurring after its enactment—i.e.,
knowingly maintaining a residence within 500 feet of a child
day-care home or group day-care home.” Id. Additionally, the
statute was not “‘so punitive either in purpose or effect as to
negate’ the legislature’s nonpunitive intent.” Id. at 521 (quot-
ing Smith, 538 U.S. at 92).
As made clear, the Leach-Vasquez rule for analyzing retro-
activity can no longer stand. This is not the first time we have
expressed doubt over our approach. In Hope v. Commissioner
of Indiana Department of Correction, the en banc majority noted
“[t]here is tension in the caselaw regarding the requirement
of the retroactivity prong.” 9 F.4th at 530. But because the
plaintiffs did not ask us to overrule our prior decisions and
the Indiana SORA was not punitive, we declined to revisit our
decisions at the time. Id. The concurrence, however, observed
that “[o]ur case law on the retroactivity prong need[ed] a
course correction.” Id. at 535 (Scudder, J., concurring). “What
Leach and Vasquez failed to account for [was] that the registra-
tion obligations did not apply at the time the sex offenders
5It is impossible to square the concurrence’s conclusion that the Ordi-
nance is retroactive with our prior holding in Vasquez. Both concerned a
prohibition against living in certain areas based on sex-offender status.
The only difference here is the Ordinance lacks tailoring—but a few hun-
dred feet means nothing for purposes of retroactivity.
16 No. 22-1007
committed the offenses triggering registration—meaning that
the sex offender registration laws imposed obligations be-
yond those prescribed at the time of the offense.” Id. We
agree—and resolve the tension in our caselaw—as “[t]he issue
is sure to surface in future cases.” Id.
We overturn our previous rule governing the retroactivity
inquiry of the Ex Post Facto Clause, announced in Leach and
followed in Vasquez, that a law which “targets only [] conduct
undertaken … after its enactment” is not retroactive. 6 Leach,
639 F.3d at 773; see also Vasquez, 895 F.3d at 520 (holding that a
law applying “only to conduct occurring after the law’s enact-
ment” is not retroactive). Instead, “[t]he critical question is
whether the law changes the legal consequences of acts com-
pleted before its effective date.” 7 Weaver, 450 U.S. at 31; see also
Vartelas, 566 U.S. at 273 (“The essential inquiry … is ‘whether
the new provision attaches new legal consequences to events
6 While the retroactivity analysis from Leach and Vasquez must be dis-
carded, we do not disturb the opinions’ alternative holdings. Neither case
involved a violation of the Ex Post Facto Clauses: SORNA’s registration
requirements mirrored the one at issue in Smith v. Doe, the Supreme
Court’s recent decision, see Leach, 639 F.3d at 773, and the Illinois statute
prohibiting sex offenders from residing too close to daycares did not
amount to punishment, see Vasquez, 895 F.3d at 521. We leave undisturbed
those determinations even as we depart from the reasoning elsewhere.
7 The test proposed by the concurrence would render the Ex Post Facto
Clause’s protections a nullity. Any legislature could argue persuasively
that a law is addressing “postenactment dangers,” even the most restric-
tive punishments. For example, a law adding ten years to certain sex-of-
fender sentences could be justified as prevention against the future danger
of additional crimes.
No. 22-1007 17
completed before its enactment.’” (quoting Landgraf, 511 U.S.
at 269–70)).
***
“There is no question that the obligations imposed by”
many sex-offender laws “apply retroactively.” Hope, 9 F.4th at
535 (Scudder, J., concurring); see also Does 1-5, 834 F.3d at 698;
Shaw, 823 F.3d 560. The Village’s Ordinance is one example. It
attaches new legal consequences to pre-Act conduct. See
Weaver, 450 U.S. at 31. Specifically, those convicted of qualify-
ing sex offenses now face additional burdens that did not exist
at the time of their offenses; they cannot establish residence in
the Village of Hartland. 8
B.
The retroactivity prong, though, is only half the analysis:
to violate the Ex Post Facto Clause, a law must also be puni-
tive. Hope, 9 F.4th at 530. In determining whether a law is pe-
nal, courts employ the two-part “intent-effects test.” Id. The
first inquiry is “whether the legislature intended to enact a
punitive, rather than a civil, law.” Id. If not, the second inquiry
is whether the law is “so punitive” in “effect as to negate [the
8 The Village advances the startling argument that because the Ordinance
is “temporary,” it cannot be retroactive. Assuming with skepticism that a
four-year ban is still temporary, we fail to see how an allegedly shorter
deprivation of a constitutional right has any bearing on our analysis. A
suspect taken unreasonably into custody for up to four years finds little
solace in the temporary nature of his unlawful seizure—so, too, a prisoner
subjected to a “brief” four-year ex post facto punishment. The importance
of constitutional rights depends, in part, on their permanence, even in
times of great social and political upheaval. This imaginative position
from the Village borders on the frivolous and can be quickly dismissed.
18 No. 22-1007
State’s] intention to deem it ‘civil.’” Smith, 538 U.S. at 92. Five
factors inform the effects analysis: 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 connec-
tion to a nonpunitive purpose; or [5] is excessive with respect
to this purpose.” Hope, 9 F.4th at 530 (quoting Smith, 538 U.S.
at 97).
Relying on Hope and Vasquez, the Village asks us to con-
duct this fact-intensive inquiry ourselves. But the Ordinance
drastically differs from the targeted laws at issue in our prior
cases. See Hope, 9 F.4th at 520 (a sex-offender residency ban
within 1,000 feet of a school, daycare, youth program, or pub-
lic park); Vasquez, 895 F.3d at 522 (a 500-foot buffer zone
around daycares). Given that the district court never consid-
ered the punitive prong, we remand to allow it to do so in the
first instance. See N.J. by Jacob v. Sonnabend, 37 F.4th 412, 426–
27 (7th Cir. 2022).
III.
For these reasons, we reverse and remand the judgment of
the district court.
No. 22-1007 19
KIRSCH, Circuit Judge, concurring in the judgment. I agree
with the result. In reaching that result, though, the majority
relies on Weaver v. Graham, 450 U.S. 24 (1981), rather than ap-
plying the retroactivity rule as set forth by the Supreme Court
in Vartelas v. Holder, 566 U.S. 257 (2012), which limits the
breadth of Weaver. The majority concludes that the limiting
principle announced in Vartelas is mere dicta to be ignored.
That conclusion will render retroactive “countless laws that
… impose ‘new disabilities’ related to ‘past events.’” See id. at
281 (Scalia, J., dissenting). Shrugging off the precedential ef-
fect of its holding, the majority does not attempt to explain
why the sex offender statutes in Leach and Vasquez are retro-
active but laws like 18 U.S.C. § 922(g)(1) or (g)(4) are not, dis-
counting the latter laws because they are “not before us.” Su-
pra at 10 n.3. Vartelas provides the answer: These laws, and
countless others like them, are not retroactive, because they
“address dangers that arise postenactment.” 566 U.S. at 271
n.7. Because I would recognize the Weaver retroactivity rule’s
obsolescence, leave be Leach and Vasquez, and apply Vartelas’s
rule to find Hartland’s ordinance retroactive, I concur only in
the judgment.
The majority opines, “[t]he clearest formulation of the ret-
roactivity inquiry,” supra at 8–9, comes from Weaver: “The
critical question is whether the law changes the legal conse-
quences of acts completed before its effective date.” Weaver,
450 U.S. at 28. But that broad rule is not the most recent law
on retroactivity. In Vartelas, the Supreme Court limited
Weaver, holding that certain laws otherwise satisfying
Weaver’s boundaries are not retroactive when they target
postenactment dangers. Vartelas, 566 U.S. at 271 n.7 (holding
that the IIRIRA statute at issue operated retroactively because
it did not target postenactment dangers).
20 No. 22-1007
The majority correctly notes that Vartelas did not involve
an Ex Post Facto challenge like Koch’s case does. But a law is
retrospective or prospective. Retroactivity cannot mean one
thing for Ex Post Facto cases and another for general pre-
sumption-against-retroactivity cases. Two circuits have ap-
plied Vartelas—without even mentioning Weaver—as the law
of retroactivity in Ex Post Facto Clause cases. See United States
v. Elk Shoulder, 738 F.3d 948, 958 (9th Cir. 2013); Bremer v. John-
son, 834 F.3d 925, 932 (8th Cir. 2016). 1
The first question in determining whether a statute’s ap-
plication is prospective or retrospective is: What is the refer-
ence point—the “moment in time to which the statute’s effec-
tive date is either subsequent or antecedent”? Vartelas, 566
U.S. at 277 (Scalia, J., dissenting). In his Vartelas dissent, Justice
Scalia considered this moment to be when the regulated party
does what the statute forbids or fails to do what it requires. Id.
If that moment is postenactment, then the statute is not retro-
active to that conduct. If preenactment, then retroactive.
I am not, contrary to the majority’s charge, suggesting we
apply Justice Scalia’s dissent. Rather, I cite to it because the
Vartelas majority clarified its rule in response to Justice Scalia,
rejecting his proposed reference point. Instead, the majority
held that courts must choose between two possible reference
points—preenactment misconduct or postenactment dan-
gers—by identifying which is the target of the legislature’s
regulation. Vartelas, 566 U.S. at 269–70. No longer is the
1 The majority recognizes that both circuits have been inconsistent on their
applications of Weaver and Vartelas. The later decisions from both circuits
cited by the majority did not even cite, let alone address, their respective
earlier opinions applying Vartelas.
No. 22-1007 21
“critical question … whether the law changes the legal conse-
quences of acts completed before its effective date.” Weaver,
450 U.S. at 31. Rather, the critical question is: What is the law’s
expressed “reason for the new disability imposed on” regu-
lated individuals? Vartelas, 566 U.S. at 269. If the reason is to
target “present … wrongful activity,” i.e., postenactment dan-
gers, then that activity is the reference point and the law—
enacted antecedent to that conduct—is nonretroactive. Id. at
269–70. If the reason is to target “past misconduct,” i.e., a pre-
vious conviction, then the conviction is the reference point
and the law—enacted subsequent to the conviction—is retro-
active. Id. In other words, if a law’s expressed reason for im-
posing new obligations is, for example, a previous sex offense,
then that law is explicitly retroactive. See, e.g., Smith v. Doe,
538 U.S. 84, 90 (2003) (finding Alaska SORA retroactive where
law’s subsection expressly making provisions retroactive did
not address hazardous conduct occurring postenactment but
did address preenactment sex offender status). Vartelas’s
holding is not dicta; that is guidance on retroactivity from a
Supreme Court majority opinion (and not just in a footnote,
either). Rather than simply “reemphasiz[ing]” the Weaver
principle, supra at 10, Vartelas modified what qualifies as ret-
rospective and left us lower courts “the unenviable task of
identifying new-disabilities-not-designed-to-guard-against-
future-danger-and-also-lacking-a-prospective-thrust.” Var-
telas, 566 U.S. at 282 (Scalia, J., dissenting); see id. at 271 n.7
(majority opinion) (“[M]entally unstable persons purchasing
guns[]” was a danger cropping up in the future, so 18 U.S.C.
§ 922(g)(4) wasn’t retroactive, even though the law changed
the legal consequences of acts completed before its effective
date); id. at 271 (18 U.S.C. § 922(g)(1) was not retroactive be-
cause prohibiting felons from possessing firearms “target[s] a
22 No. 22-1007
present danger, i.e., the danger posed by felons who bear
arms”).
The majority worries Vartelas’s rule might “render the Ex
Post Facto Clause’s protections a nullity” because legislatures
would begin tailoring their laws to address postenactment
dangers. Supra at 16 n.7. But what’s wrong with legislatures’
being more explicit about whether their laws are retrospective
or prospective? Courts should encourage, not discourage,
such legislative clarity. The majority does not explain why in-
centivizing legislative precision in tailoring is so problematic.
And to the extent that Vartelas’s framework is a problem, it is
a problem for the Court that wrote Vartelas, not this court.
Still, the rule announced in Vartelas makes clear that Hart-
land’s Ordinance is retroactive. Vartelas, like this case, in-
volved a law that restricted future movement based on past
misconduct: The IIRIRA statute (8 U.S.C. § 1101(a)(13)(C)(v))
effectively precluded lawful foreign travel by lawful perma-
nent residents convicted of certain crimes by making them in-
eligible for reentry at the conclusion of their foreign travel. In
finding that the statute did not address a danger that arose
postenactment, the Court explained, “[t]he act of flying to
Greece … does not render a lawful permanent resident like
Vartelas hazardous.” Vartelas, 566 U.S. at 271 n.7. “[T]he rea-
son for the ‘new disability’ imposed on [Vartelas] was not his
lawful foreign travel[]” but his pre-IIRIRA conviction. Id. at
269. “That past misconduct, in other words, not present
travel, [wa]s the wrongful activity Congress targeted.” Id. at
269–70. The same logic applies in full force to Hartland and
Koch. Koch is just as dangerous in Hartland as he is in other
towns. The act of moving to Hartland does not render Koch
more hazardous than he was wherever he lived previously.
No. 22-1007 23
Hartland cannot say Koch’s lawful travel is being targeted. In
fact, never does the Ordinance explain what postenactment
danger it addresses. The one and only reason the Ordinance’s
plain text offers for its new residence regulations is the previ-
ous convictions of designated sex offenders. The reference
point is Koch’s conviction. So the Ordinance does not address
a postenactment danger but attaches a new disability (inabil-
ity to move to Hartland postenactment) in respect to past mis-
conduct (sex offense conviction preenactment). See Vartelas,
566 U.S. at 261. Therefore, I agree with the result in this case.
Considering Vartelas makes clear the retroactivity of the
Ordinance, there is no reason to overrule Leach and Vasquez.
Both were correctly decided under Vartelas’s postenactment
dangers approach. The law at issue in Leach expressly ad-
dressed a future danger—unregistered sex offenders moving
to a new jurisdiction. The registration requirements directly
addressed the danger at issue—public ignorance of poten-
tially dangerous sex offenders in the community. Cf. Elk
Shoulder, 738 F.3d at 958 (finding nonretroactive the registra-
tion and notification provisions of the pre-SORNA Wetterling
Act under Vartelas because “they addressed the danger that
the public would not be aware of potentially dangerous sex
offenders living, working, or attending school in its area.”).
And the Illinois statute in Vasquez did the same thing in re-
quiring that sex offenders could not live within 500 feet of day
cares. The new day care requirement explicitly addressed a
postenactment peril recognized by the state—children in day
cares in physical proximity to sex offenders. Neither Leach nor
Vasquez featured a law that imposed new obligations because
of previous convictions. On the contrary, the laws in both
cases plainly expressed that the reasons for the new obliga-
tions were to address present activity. Both cases survive
24 No. 22-1007
under Vartelas because the laws in both cases were antecedent
to the reference points—the postenactment dangers ad-
dressed.
But there’s no doubt that Hartland’s ordinance addresses
no postenactment danger. It is not “impossible to square” this
with the original Vasquez holding. See supra at 15 n.5. Vartelas
instructs us to look at the purpose of the law. The Illinois stat-
ute in Vasquez was tailored to address the specific problem of
sex offenders’ close physical proximity to daycares. The Hart-
land ordinance never even attempts to specify what the resi-
dence ban targets besides sex offender status.
Applying Vartelas, I agree the ordinance is retroactive and
the case should return to the district court for a determination
on punitiveness.