In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 17‐3256
DONALD LACY, on behalf of himself and all
others similarly situated,
Petitioner‐Appellee,
v.
KEITH BUTTS, Warden,
Respondent‐Appellant.
____________________
Appeal from the United States District Court for the
Southern District of Indiana, Indianapolis Division.
No. 13‐CV‐00811 — Richard L. Young, Judge.
____________________
ARGUED OCTOBER 26, 2018 — DECIDED APRIL 25, 2019
____________________
Before WOOD, Chief Judge, and SYKES and SCUDDER, Circuit
Judges.
WOOD, Chief Judge. When the state wants to encourage sus‐
pects, defendants, or incarcerated offenders to admit guilt, it
has many tools at its disposal. Before or during trial, prosecu‐
tors may hold out the prospect of a plea bargain. Judges may
reward defendants with a sentence reduction for accepting re‐
sponsibility. Prison rehabilitation programs may offer
2 No. 17‐3256
benefits and incentives by conditioning visitation rights, work
opportunities, housing in a lower‐security unit, and other
privileges on an offender’s willingness to admit responsibility
for the crime of conviction. McKune v. Lile, 536 U.S. 24, 40
(2002).
But the Fifth Amendment draws one sharp line in the
sand: no person “shall be compelled in any criminal case to be
a witness against himself.” U.S. CONST. amend. V. (emphasis
added). This case requires us to decide whether Indiana’s Sex
Offender Management and Monitoring (INSOMM) program
crosses that line with its system of revoking good time credits
and denying the opportunity to earn such credits for
convicted sex offenders who refuse to confess their crimes. In
an action brought by a class led by Donald Lacy, an inmate
subject to INSOMM, the district court ruled that Indiana’s
system as currently operated impermissibly compels self‐
incrimination and must be revised. We affirm.
I
Indiana requires all inmates convicted of a sex offense to
complete the INSOMM program before release. The program
is divided into three phases. During Phase 1, which focuses
on “Consent and Assessment,” participants must fill out a Sex
Offender Questionnaire that asks them to identify which ille‐
gal sexual acts (for instance, rape, child molestation, or pros‐
titution) they committed and how often. Based on their of‐
fense history, participants are then sorted into one of three
risk groups for Phase 2, “Risk Based Sex Offender Treatment,”
which consists of group therapy sessions led by counselors.
Inmates meet only with others assessed at the same risk level;
this becomes their “core group.” The higher‐risk groups must
complete more hours of therapy. In preparation for the
No. 17‐3256 3
therapy sessions, participants must fill out workbooks that re‐
quire them to describe in detail all past acts of sexual violence
and abuse, regardless of whether they were ever charged for
those offenses. Finally, after release participants must com‐
plete Phase 3, “Community Management and Monitoring,”
which is not at issue in this case.
As INSOMM stresses to its participants throughout the
program, they enjoy neither immunity nor confidentiality for
any of the disclosures they make at any stage. Moreover, par‐
ticipation is an all‐or‐nothing affair: inmates may not opt out
of any part of the program, and they are required to respond
fully to all questions asked. A counselor who suspects that a
participant has been either deceptive or less than forthcoming
has the discretion to order polygraph testing. Such an order
triggers a requirement for the participant to fill out a detailed
Polygraph Sex History Questionnaire. A participant is ex‐
cused from admitting responsibility for an offense only if the
polygraph examination indicates no deception and the coun‐
selors agree that the participant is being truthful.
Failure to participate satisfactorily in INSOMM is costly.
It is treated as a Class A or Major Conduct disciplinary viola‐
tion. That is the same class assigned to rioting, escape, rape,
or assault on prison staff. The particular Class A violation at
issue in our case—Code 116, “Refusing a Mandatory Pro‐
gram”—applies to a much broader range of conduct than a
flat‐out refusal to participate. Inmates who refuse a polygraph
examination, deny parts of their offenses, give answers that
are deemed to be incomplete or dishonest, or otherwise fail to
“adhere to treatment expectations” also qualify for a Code 116
violation. The INSOMM counselors make the final decision
whether an inmate’s conduct merits disciplinary action.
4 No. 17‐3256
For a first Code 116 offense, inmates are penalized (among
other ways) by placement in a credit class that denies them
the opportunity to accrue good‐time credits. These are credits
to which they would otherwise be entitled by statute. IND.
CODE § 35‐50‐6‐3 (accrual of credit time for people convicted
up to June 30, 2014); § 35‐50‐6‐3.1 (same, for people convicted
after June 30, 2014); § 35‐50‐6‐5 (deprivation of credit time). If
an inmate persists in whatever conduct gave rise to the disci‐
plinary action—such as by refusing to admit a particular
crime or answer a particular question—he is regarded as com‐
mitting a continuing Code 116 violation punishable by revo‐
cation of 180 days of already‐acquired good‐time credits for
every 60 days during which the noncompliance continues.
While the violation is ongoing, the inmate is barred from
earning new good time credits. Some class members have lost
all of their accrued good‐time credits as a result of this system.
Lacy brought this case on May 16, 2013, under 28 U.S.C.
§ 2254, on behalf of a class of Indiana inmates who have been
punished with a loss of good‐time credits and a demotion in
credit class for failure to abide by the requirements of IN‐
SOMM. On September 20, 2015, the district court certified the
class. (For ease of exposition, we refer to Lacy’s arguments, on
the understanding that he presents them on behalf of the en‐
tire class.) Lacy argues that the disclosures required by IN‐
SOMM and the penalties imposed for non‐participation,
taken together, amount to a violation of his Fifth Amendment
right to be free from compelled self‐incrimination. The district
court agreed and ordered the issuance of the writ of habeas
corpus to restore the class’s lost good‐time credits. It also va‐
cated all disciplinary actions and sanctions for failure to par‐
ticipate in INSOMM.
No. 17‐3256 5
On appeal, Indiana argues both that the INSOMM pro‐
gram does not carry a sufficiently serious risk of incrimination
to trigger the protections of the Fifth Amendment, and that
even if it does, the revocation of credit time and the demotion
in credit class do not add up to unconstitutional compulsion.
Because the district court resolved the case on summary judg‐
ment, we consider its decision de novo. Thomas v. Clements, 789
F.3d 760 (7th Cir. 2015).
II
To succeed in his Fifth Amendment challenge to IN‐
SOMM, Lacy must demonstrate that the disclosures to which
he objects are (1) testimonial, (2) incriminating, and (3) com‐
pelled. Hiibel v. Sixth Judicial Dist. Court of Nevada, Humboldt
Cnty., 542 U.S. 177, 189 (2004). Because the parties do not dis‐
pute that the disclosures are testimonial, we need decide only
the latter two issues: whether the program calls for incrimi‐
nating information; and if so, whether participation is “com‐
pelled” as that term is used in the Supreme Court’s Fifth
Amendment jurisprudence.
A
Lacy’s first task is to show the risk of self‐incrimination
that flows from an inmate’s participation in INSOMM is “real
and appreciable,” rather than “imaginary and insubstantial.”
Marchetti v. United States, 390 U.S. 39, 48 (1968). He can do so
even if the required disclosures would not “in themselves
support a conviction,” so long as they would “furnish a link
in the chain of evidence” that could lead to a criminal prose‐
cution. Hoffman v. United States, 341 U.S. 479, 486 (1951).
A closer look at INSOMM shows that Lacy has met that
burden. The extensive information that INSOMM requires of
6 No. 17‐3256
its participants—with the express warning that neither im‐
munity nor confidentiality will be available—exposes Lacy
and his fellow class members to the risk of future criminal in‐
vestigation and prosecution. Although Indiana contends that
“the INSOMM program permits participants to describe their
actions in the most general terms,” a few examples illustrate
why this benign description is not supported by the record.
First, in the disclosure assignments that make up part of
the core‐group workbooks, the program directs the partici‐
pant to be “detailed and specific” about each victim he has
harmed, using a separate sheet for each one. In the Medium
Risk Core Group Workbook, the questions that the partici‐
pant must answer for each past act of sexual violence or abuse
include the participant’s age and the victim’s age at the time
of the offense; the first name of the victim; the participant’s
relationship to the victim; “what parts of [the participant’s]
body” touched “what parts of [the victim’s] body”; “how
many times [the participant] offended the victim and over
what period of time”; “where and when” the abuse occurred;
how the victim was selected and groomed; and the types of
force used to ensure the victim’s compliance. The High Risk
Core Group Workbook includes these questions and more: it
also asks the participant to describe “in detail the set‐up of the
sexual abuse of each victim”; patterns among his victims, such
as their age, appearance, and race; “when and how” the abuse
started with each victim; and “in what ways [the partici‐
pant’s] … sexual offending behavior change[d] over time.”
If counselors find any of the responses to those questions
to be inaccurate or incomplete—a determination they have
unfettered discretion to make—they can refer the participant
for a polygraph examination. Such a referral starts with the
No. 17‐3256 7
obligation to fill out yet another, equally intrusive, question‐
naire. The polygraph questionnaire asks highly specific ques‐
tions, such as: “How many children have you physically
forced into sexual activities? Describe what you did.”; “How
many times have you had sexual contact with someone who
is handicapped? Describe.”; and “How many times have you
made child pornography (taken pictures, video‐tapes, films,
etc.) of nude children or children engaged in sex acts? De‐
scribe.” The participant must also “write the number of
times” that he has masturbated in each of 31 different public
places, indicate the most recent time he exposed himself to
someone, and describe times when he had sexual contact with
someone who was drunk or under the influence of drugs. For
each victim, the participant must indicate whether he en‐
gaged in any one of 27 specific actions and how many times,
including whether he “put [his] penis inside [the victim’s]
vagina (even a little bit),” “threatened [the] victim with
harm,” or “masturbated in front of [the] victim.”
Truthful and complete answers to questions such as
these—and there are many, many more—are highly pertinent
to crimes beyond those of conviction. The odds that some par‐
ticipants would be investigated and successfully prosecuted
for past uncharged crimes are high. The required disclosures
could provide enough detail to kickstart a brand‐new investi‐
gation. (By the time the state knows the victim’s first name,
age, and relationship to the offender, as well as when and
where the abuse took place, it often will require little outside
investigation to complete the story.) INSOMM disclosures
might advance or substantiate an ongoing investigation or
one that was halted for lack of evidence. In current or poten‐
tial prosecutions, disclosures from INSOMM could confirm
identity or modus operandi. FED. R. EVID. 404(b)(2). And when
8 No. 17‐3256
the questions require an admission of responsibility for the
crime of conviction rather than for an uncharged crime, a par‐
ticipant who pleaded not guilty and testified in his own de‐
fense could be exposed to perjury charges. In short, when In‐
diana requires inmates to describe all of their past sexual of‐
fenses in minute detail, the risks of self‐incrimination are far
from “remote” or “speculative.” Zicarelli v. N.J. State Commʹn
of Investigation, 406 U.S. 472, 478 (1972). They are self‐evident.
Indiana’s arguments to the contrary are unpersuasive. It
places great weight on the fact that “[p]articipants are not re‐
quired to identify their victims or specific dates of the sexual
activity” and that they “are specifically advised not to provide
identifying information.” But that general statement collides
with the specific command to provide highly detailed ques‐
tions about each incident of past sexual abuse, omitting only
the date and the victim’s last name (or even full name). The
required details easily raise a significant risk of incrimination.
The state has not explained how it is possible to answer these
more specific questions—such as the age of the victim and the
victim’s relationship to the participant—while taking ad‐
vantage of the general admonishment to avoid “identifying
information.” Ignoring the actual questions, the state insists
that the disclosures are at such a high level of generality that
they “could not support a link in the chain sufficient to sup‐
port an investigation or even admission at trial.”
Saying so does not make it so. This ipse dixit does not ex‐
plain why granular descriptions of the circumstances sur‐
rounding specific sex crimes and patterns of criminal sexual
behavior would prove useless to investigators or prosecutors.
The Fifth Amendment “protects against any disclosures
which the witness reasonably believes could be used in a
No. 17‐3256 9
criminal prosecution or could lead to other evidence that might
be so used.” Kastigar v. United States, 406 U.S. 441, 445 (1972)
(emphasis added). The questions posed to an INSOMM par‐
ticipant would yield answers that any competent sex‐crimes
investigator or prosecutor would love to have.
Indiana also stresses that “no prosecutions have been ini‐
tiated as a result of the INSOMM program.” But this fact—
assuming it is true—is not determinative. It is not the case that
“a witness’ constitutional privilege against self‐incrimination
depends upon a judge’s prediction of the likelihood of prose‐
cution. Rather … it is only where there is but a fanciful possi‐
bility of prosecution that a claim of fifth amendment privilege
is not well taken.” In re Folding Carton Antitrust Litigation, 609
F.2d 867, 871 (7th Cir. 1979) (emphasis added).
Indeed, in its reply brief the state all but concedes that the
necessary level of risk is present. It admits there that the pos‐
sibility of prosecution is anything but “fanciful.” It does so in
the course of explaining why INSOMM does not provide im‐
munity for its participants. Offenders, the state insists, must
feel the risk of prosecution for the program to achieve its ther‐
apeutic goals. “Immunity is not an option. If prisoners know
they will never be prosecuted for past offenses, they may be
led to believe that society does not consider their crimes to be
serious.” No one says that the state is obliged to grant immun‐
ity, of course; Lile holds otherwise. 536 U.S. at 27. But the
state’s recognition that the disclosures run enough risk of self‐
incrimination that a prisoner might want immunity is telling.
It seems to us that the state is trying to have it both ways: on
the one hand, it wants us to believe that the rehabilitative pur‐
pose of INSOMM depends on the disclosures’ carrying a
meaningful risk of prosecution; but on the other hand, it
10 No. 17‐3256
maintains that this risk is too remote to implicate the Fifth
Amendment. Our own independent assessment of the pro‐
gram convinces us that the disclosures do give rise to a real
and appreciable risk of prosecution.
B
That brings us to the central question on appeal: whether
INSOMM, through its denial of the opportunity to earn good‐
time credits and its revocation of credits already earned, as a
means of inducing participants to furnish information, com‐
pels self‐incrimination in contravention of the Fifth Amend‐
ment.
We acknowledge that the line between permissible pres‐
sure and impermissible compulsion can be difficult to draw.
That may explain why in Lile, the Supreme Court’s latest pro‐
nouncement on the subject, the justices failed to produce a
majority opinion. 536 U.S. 24. Nonetheless, many cases are not
close to the line, and ours is one of them. All five justices join‐
ing the judgment in Lile singled out the extension of a term of
incarceration as a penalty more severe than the one before
them. We look for guidance to Justice O’Connor’s plurality
opinion, which we find reflects the narrowest grounds ac‐
cepted by the concurring justices and thus is controlling un‐
der the rule of Marks v. United States, 430 U.S. 188, 193 (1977).
Justice O’Connor summarized the relevant restrictions im‐
posed on inmate Lile as follows:
These consequences involve a reduction in
incentive level, and a corresponding transfer from a
medium‐security to a maximum‐security part of the
prison. In practical terms, these changes involve
restrictions on the personal property respondent can
No. 17‐3256 11
keep in his cell, a reduction in his visitation privileges,
a reduction in the amount of money he can spend in
the canteen, and a reduction in the wage he can earn
through prison employment.
536 U.S. at 50–51.
Justice O’Connor’s concurrence specifically identified a
longer term of incarceration as an impermissible penalty, at
least if it were imposed automatically. 536 U.S. at 52 (“The
penalties potentially faced in these cases—longer incarcera‐
tion and execution—are far greater than those we have al‐
ready held to constitute unconstitutional compulsion in the
penalty cases. Indeed, the imposition of such outcomes as a
penalty for refusing to incriminate oneself would surely im‐
plicate a ‘liberty interest.’”). Her rationale leads to the conclu‐
sion that Lacy faced unconstitutional compulsion. Cf. United
States v. Antelope, 395 F.3d 1133, 1138 (9th Cir. 2005) (“Justice
O’Connor made clear in her McKune [v. Lile] concurrence that
she would not have found a penalty of ‘longer incarceration’
such as that here to be permissible.”). See also Glover v. United
States, 531 U.S. 198 (2001) (error leading to longer sentence is
prejudicial for Sixth Amendment purposes, even if the exten‐
sion is not a long one).
The plurality opinion in Lile is not to the contrary. The four
justices in the plurality went out of their way to distinguish
the penalties they found permissible from the sort of punish‐
ment at issue in this case. They noted that Lile’s “decision not
to participate … did not extend his term of incarceration. Nor
did his decision affect his eligibility for good‐time credits or
parole.” Lile, 536 U.S. at 38 (plurality opinion; emphasis
added). Like Justice O’Connor, the plurality suggested that
when the penalty in question was as serious as a longer term
12 No. 17‐3256
of incarceration, the automatic imposition of such a penalty
would amount to unconstitutional compulsion. “States may
award good‐time credits and early parole for inmates who ac‐
cept responsibility because silence in these circumstances
does not automatically mean the parole board, which consid‐
ers other factors as well, will deny them parole.” Id., citing
Baxter v. Palmigiano, 425 U.S. 308, 317–18 (1976). For this pur‐
pose, the distinction between a discretionary award of less
time and the automatic deprivation of an earned sentence re‐
duction is critical.
Three of our sister circuits share this understanding of Lile.
See Antelope, 395 F.3d at 1133 n.1; Ainsworth v. Stanley, 317
F.3d 1, 4 (1st Cir. 2002); Searcy v. Simmons, 299 F.3d 1220, 1225
(10th Cir. 2002); see also United States v. Von Behren, 822 F.3d
1139, 1147–48 (10th Cir. 2016). (There are also two un‐
published decisions that discuss Lile more briefly; they do not
reach any contrary conclusions about the effect of the O’Con‐
nor opinion. See Edwards v. Goord, 362 F. App’x 195 (2d Cir.
2010); Thorpe v. Grillo, 80 F. App’x 215 (3d Cir. 2003).)
More broadly, the Supreme Court has emphasized that
punishments implicating the length of confinement are differ‐
ent in kind from those that affect the conditions of confine‐
ment. Here, Indiana considers no other factors as it evaluates
whether to revoke an inmate’s good time credits or to deny
him the opportunity to earn such credits. The decision to de‐
cline participation in INSOMM is not merely a trigger for a
later stage in which the state takes a more holistic view of an
inmate’s progress toward rehabilitation. Instead, a prisoner’s
choice to invoke his privilege against self‐incrimination is the
direct cause of his loss of credits—credits that otherwise
would be statutorily guaranteed (assuming no independent
No. 17‐3256 13
reason to revoke them). IND. CODE §§ 35‐50‐6‐3, 35‐50–6‐5.
Neither the plurality nor the concurring opinions in Lile sup‐
port such a system. Nor, for that matter, did the dissenters. As
the Ninth Circuit noted in Antelope, 395 F.3d at 1138, the dis‐
senters found even the existing penalties to be unconstitution‐
ally coercive; the more severe sanction of longer incarceration
would have been at least as problematic.
Indiana stresses that its credit‐time statute explicitly con‐
templates the revocation of credit time for an offender who
“refuses to participate in a sex offender treatment program”
during incarceration. IND. CODE § 35‐50‐6‐5(a)(6). But that
does not answer the question before us, which is whether it is
permissible to strip a prisoner of a constitutional right as a
means of forcing participation. We have no quarrel with the
proposition that the state may use other penalties or induce‐
ments.
Indeed, we can assume that there is still some room for
parole revocation, depending on the circumstances. But that
measure must be used with care. Other courts of appeals that
have evaluated these programs since Lile have found that rev‐
ocation of parole as a penalty for non‐participation is permis‐
sible in situations where it does not automatically follow. See
Ainsworth, 317 F.3d at 4–5 (noting that the denial of parole
“was not entirely automatic”); Thorpe, 80 F. App’x 215 (finding
no unconstitutional compulsion where an inmate’s refusal to
participate in a treatment program did not “automatically de‐
prive him of consideration for parole”). In contrast, the Ninth
Circuit held that a district court’s decision explicitly to condi‐
tion a prisoner’s supervised release on his successful comple‐
tion of a treatment program similar to INSOMM violated the
inmate’s right to be free from compelled self‐incrimination.
14 No. 17‐3256
Antelope, 395 F.3d at 1139. Just as in Antelope, Lacy’s refusal to
incriminate himself has rendered him categorically ineligible
for the shorter term of confinement he otherwise would have
received. The Fifth Amendment does not allow Indiana to im‐
pose such a penalty.
Finally, we stress that this case relates to the length of time
Lacy will be incarcerated, not to the conditions of his confine‐
ment. It thus falls on the habeas‐corpus side of the line drawn
by Preiser v. Rodriguez, 411 U.S. 475 (1973), not the section 1983
side. In fact, Lacy originally brought his suit as a section 1983
action, but we held that a petition under section 2254 was the
only appropriate vehicle for his claim. Lacy v. Indiana, 564 F.
App’x 844 (7th Cir. 2014) (unpublished order). His case is thus
distinguishable from Lile, insofar as the sanctions that the Su‐
preme Court endorsed in Lile were all conditions of confine‐
ment, not measures that would affect duration.
III
We AFFIRM the judgment of the district court.