In the

United States Court of Appeals For the Seventh Circuit ____________________ No. 22-2791 CELINA MONTOYA, et al., individually and on behalf of all others similarly situated, Plaintiffs-Appellants,

v.

ROB JEFFREYS, Director of the Illinois Department of Corrections, Defendant-Appellee. ____________________

Appeal from the United States District Court for the Northern District of Illinois, Eastern Division. No. 1:18-cv-01991 — John Robert Blakey, Judge; Gary Feinerman, Judge. ____________________

ARGUED DECEMBER 7, 2023 — DECIDED APRIL 18, 2024 ____________________

Before WOOD, KIRSCH, and JACKSON-AKIWUMI, Circuit Judges. KIRSCH, Circuit Judge. Celina Montoya, Jennifer Tyree, Ronald Molina, and Zachary Blaye brought a class action law- suit against the Illinois Department of Corrections (IDOC) challenging an IDOC policy restricting contact between a 2 No. 22-2791

parent convicted of a sex offense and her minor child while the parent is on mandatory supervised release (MSR). The plaintiffs alleged that this policy violates Fourteenth Amend- ment procedural and substantive due process. The district court entered judgment largely for IDOC. Though we agree with the court that IDOC’s policy does not violate procedural due process, we hold that its ban on phone contact violates substantive due process. On this record, call monitoring is a ready alternative to the phone-contact ban that accommo- dates the plaintiffs’ right to enjoy the companionship of their children at a de minimis cost to IDOC’s penological interests. We therefore affirm in part and reverse in part. I In Illinois, a parent on mandatory supervised release who has been convicted of a sex offense is subject to an IDOC pol- icy restricting her written, phone, and in-person contact with her minor child. Upon her release from prison, she is pre- sumptively banned from contact with her minor child. She may request contact, though she must enroll in sex offender therapy if she does so. She is not limited to seeing an IDOC therapist and may see a therapist in private practice instead. If the parent cannot get an appointment with a therapist within 14 days, an IDOC sex offender therapist will review her file and complete assessments to allow the parent’s con- tainment team to begin making a determination regarding her child-contact request. The containment team, comprised of the parent’s parole agent, a sex offender therapist, and the parole commander for the district, evaluates the risk she poses to her child. The pa- role agent considers the parent’s compliance with MSR, in- cluding compliance with MSR rules. The therapist assesses No. 22-2791 3

the risk of harm to the child and makes a dangerousness de- termination. This assessment holds considerable weight be- cause of the therapist’s expertise in evaluating sex offenders. Each therapist may use her own policies to reach a contact recommendation, including policies related to how long the parent must be enrolled in therapy before the therapist makes a recommendation. The team must give the parent an initial decision on her contact request within 21 days. If the contain- ment team restricts or denies contact, it must provide the par- ent with written reasons for the decision and review of the restrictions or prohibitions every 28 days. A therapist may deny contact on the basis of insufficient therapy to make a recommendation to allow child contact. The parent may appeal the containment team’s decision to IDOC’s Manager of Sex Offender Services, Sarah Brown- Foiles. Brown-Foiles serves as an independent appellate deci- sionmaker. She does not oversee the containment team mem- bers who make the initial contact determinations and does not participate in their decision-making process. Sometimes, Brown-Foiles is not independent, typically when the contain- ment team has solicited her advice about a particular contact request or when she supervises the sex offender therapist on the containment team. But in those instances, a different ap- pellate decisionmaker will hear the appeal: either IDOC’s Chief of Programs, Alyssa Williams, or the Clinical Director at the Big Muddy River Correctional Center, Heather Wright. The named plaintiffs—Celina Montoya, Jennifer Tyree, Ronald Molina, and Zachary Blaye—are parents who are or were subject to IDOC’s policy. Montoya was released from prison on MSR and was prohibited from residing with her youngest child, who was 15 years old at the time of filing, for 4 No. 22-2791

approximately five months. Tyree was released from prison on MSR, and two of her children were minors at the time of her release. After her release, a state court judge ordered that her minor son live with her, but her parole agent refused to allow him to live with her because of the policy. Molina was released from prison on MSR and was prohibited from all contact with his son, who was 16 years old at the time of filing. Blaye was also released on MSR, and at the time of filing, his son was 13 years old. After first being restricted from having contact with his son under the policy, Blaye was permitted to have phone and in-person contact with him. Montoya, Tyree, Molina, and Blaye brought a class action lawsuit under 42 U.S.C. § 1983 against Rob Jeffreys, in his of- ficial capacity as IDOC’s Director, alleging that IDOC’s policy violates Fourteenth Amendment procedural and substantive due process. After a bench trial, the district court largely up- held the policy, in part because of IDOC’s strong interests in protecting children and rehabilitating sex offenders and testi- mony that there was no assessment or combination of assess- ments that could be an adequate substitute for a sex offender therapist’s post-release evaluation. But it upheld the policy with two exceptions. Uncontested but relevant to the analysis on appeal, one exception held that the policy’s ban on written communications was unconstitutional; the court declared that IDOC must allow a parent to submit a written communication addressed to her child to the containment team for review and decision within seven calendar days. The plaintiffs appealed, challenging the policy’s restrictions on phone and in-person contact. No. 22-2791 5

II Under Article III, a plaintiff must have standing—“a per- sonal stake in the case”—to invoke “the federal judicial power.” TransUnion LLC v. Ramirez, 594 U.S. 413, 423 (2021) (cleaned up). At least one named plaintiff must have standing for a class action to proceed. Kohen v. Pacific Inv. Mgmt. Co., 571 F.3d 672, 676–77 (7th Cir. 2009). It is unclear from the rec- ord whether all the named plaintiffs have standing. But at the time the amended complaint was filed, Molina was still pro- hibited from all contact with his son and thus had standing. See Cnty. of Riverside v. McLaughlin, 500 U.S. 44, 51 (1991) (an- alyzing standing to bring a class action at the time the com- plaint was filed). Because at least one named plaintiff has standing, we may proceed to the merits. In assessing the mer- its, we review the legal conclusions made in a bench trial de novo and the factual findings for clear error. Morisch v. United States, 653 F.3d 522, 528 (7th Cir. 2011). A The plaintiffs first argue that the policy violates proce- dural due process because (1) it fails to provide a sufficiently neutral and impartial decisionmaker, (2) it lacks a pre-depri- vation hearing, and (3) its post-deprivation process is not prompt and fair. To determine whether IDOC’s policy vio- lates procedural due process, we consider the three factors outlined in Mathews v. Eldridge, 424 U.S. 319 (1976): (1) the pri- vate interest affected by the policy, (2) the risk of erroneous deprivation of the interest and the probable value of addi- tional or substitute procedural safeguards, and (3) IDOC’s in- terest, including the administrative or fiscal burdens of the additional safeguards, id. at 334–35. 6 No. 22-2791

1 No party challenges the district court’s conclusion that procedural due process requires a neutral, independent deci- sionmaker in evaluating child-contact requests. Rather, the plaintiffs insist that a parent’s treating therapist is not a suffi- ciently objective and impartial evaluator (because the thera- pist is directly involved in the parent’s treatment) and that there is not always a neutral and impartial appellate deci- sionmaker. In most cases, Brown-Foiles serves as an inde- pendent appellate decisionmaker. When she is not independ- ent, either Williams or Wright will hear the appeal instead. The plaintiffs argue that neither Williams nor Wright is suffi- ciently neutral and impartial, as Williams is Brown-Foiles’s supervisor, and Wright reports to Brown-Foiles. And in their view, the appeals process is also insufficient because it is con- ducted in writing, not through an in-person hearing, and thus a parent does not have a realistic opportunity to present her own evidence and rebut the evidence against her. They assert that IDOC could satisfy procedural due process at a minimal cost by having the Prisoner Review Board (PRB), which al- ready dictates MSR conditions, serve as the initial deci- sionmaker for child-contact requests. The first Mathews factor, the private interest affected by the policy, weighs in the plaintiffs’ favor. Their right to enjoy the companionship of their children “is perhaps the oldest of the fundamental liberty interests recognized by the Supreme Court.” United States v. Lee, 950 F.3d 439, 448 (7th Cir. 2020) (cleaned up). But the second factor, the risk of erroneous deprivation and the probable value of additional safeguards, weighs in IDOC’s favor. If the policy did not provide for a neutral and No. 22-2791 7

impartial decisionmaker, there would be a significant risk of erroneous deprivation. However, IDOC’s policy does. Brown-Foiles does not oversee the containment team mem- bers who make the initial contact determinations and is not involved in their decision-making process. See Felce v. Fiedler, 974 F.2d 1484, 1499 (7th Cir. 1992) (“Of course, a deci- sionmaker need not be external to an institution to be inde- pendent.”). Williams may not be sufficiently independent be- cause she supervises Brown-Foiles. See id. at 1499–500 (hold- ing that supervisors of the primary decisionmaker were insuf- ficiently independent because they had “individual interests” in supporting their subordinate’s decision). But we need not reach that issue because the plaintiffs did not dispute Wright’s neutrality below, and on appeal, they simply note that Wright reports to Brown-Foiles, providing no further ex- planation for why Wright is insufficiently independent and citing to no case to support their argument. Plaintiffs thus for- feited their argument on appeal that Wright is insufficiently independent. Love v. Vanihel, 73 F.4th 439, 449 (7th Cir. 2023) (“Arguments inadvertently not raised in the district court are forfeited, and in the civil context, ordinarily unreviewable on appeal, because we review forfeited claims only in excep- tional cases.”) (footnote omitted). Similarly, the probable value of the additional safe- guard—a decisionmaker at the initial decision level who is not directly involved in a parent’s treatment or supervision— is minimal. The policy already provides for a decisionmaker who is not directly involved. And though the plaintiffs argue that the appeals process should be conducted through an in- person hearing rather than in writing, they do not explain why a parent does not have a realistic opportunity to present 8 No. 22-2791

her own evidence and rebut the evidence against her through writing. 1 The second Mathews prong thus favors IDOC. The third Mathews factor—IDOC’s interest, including the administrative or fiscal burdens of the additional safe- guards—may cut in the plaintiffs’ favor. It is unclear from the record what burden IDOC would face if the PRB had to serve as the initial decisionmaker, and IDOC does not address on appeal the potential burden of this additional safeguard. Sim- ilarly, in assessing the third Mathews factor, the district court addressed the burden to IDOC of providing an independent appellate decisionmaker (such as Williams or Wright) but not the burden of providing an independent decisionmaker in the first instance. The burden of this additional safeguard may be minimal. But regardless, the third factor does not outweigh the second, which favors IDOC not only because the risk of erroneous deprivation is low but also because the probable value of the additional safeguard is negligible. IDOC’s policy therefore provides for a sufficiently neutral and impartial de- cisionmaker and does not violate procedural due process on this basis. 2 The plaintiffs argue that the policy’s lack of a pre-depriva- tion hearing is impermissible because fundamental rights

1 In a short footnote in their brief, the plaintiffs mention that many parents are never afforded an opportunity to appeal because they do not receive a written explanation for contact restrictions from their therapist. But the policy requires that parents receive written reasons if contact is restricted or denied. Therefore, a parent who does not receive written rea- sons might be able to bring an as-applied challenge to the policy but not a facial challenge. No. 22-2791 9

cannot be categorically abrogated. They rely on Justice Scalia’s concurrence in Connecticut Department of Public Safety v. Doe, 538 U.S. 1 (2003), which stated that a properly enacted law can eliminate a liberty interest unless the interest is so fundamental that it implicates substantive due process, id. at 8. And they state that the PRB could consider the need for contact restrictions in a pre-deprivation hearing at the same time that it determines other MSR conditions. But an “im- portant government interest, accompanied by a substantial assurance that the deprivation is not baseless or unwarranted, may in limited cases demanding prompt action justify post- poning the opportunity to be heard until after the initial dep- rivation.” Doyle v. Camelot Care Ctrs., Inc., 305 F.3d 603, 618 (7th Cir. 2002) (quotation omitted). Again, the first Mathews factor weighs in the plaintiffs’ fa- vor, as they have a fundamental right to enjoy the compan- ionship of their children. However, their argument that fun- damental rights cannot be categorically abrogated falls short. That the concurrence referred to substantive (not procedural) due process aside, nothing in it suggested that a law can never restrict a fundamental liberty interest without a pre-depriva- tion hearing. Instead, the post-deprivation process must sat- isfy due process. The second factor favors IDOC because the probable value of the additional safeguard of a pre-deprivation hearing is low. To be sure, without a pre-deprivation hearing, the plain- tiffs face a substantial risk of erroneous deprivation of contact. There is no pre-release process to make a contact decision based on the risk of harm to the child, and thus the policy may prevent a parent who poses no harm to her child from having contact with the child until a contact decision is made. But the 10 No. 22-2791

additional safeguard of a pre-deprivation hearing has low probable value. The district court did not clearly err in finding that a pre-release evaluation of dangerousness is less accurate than a post-release evaluation, in part because a pre-release evaluation does not consider the parent’s adjustment to com- munity life and adaptation to stressors outside prison. Given this low probable value, the second Mathews factor weighs in IDOC’s favor. The third Mathews factor favors IDOC, too. IDOC has a strong interest in making an accurate contact determination. The PRB could consider the need for contact restrictions when it determines other MSR conditions. But the district court did not clearly err in crediting testimony that there is no assess- ment or combination of assessments that could be an ade- quate substitute for a sex offender therapist’s post-release evaluation, even though the PRB could hear testimony from a therapist before it conducts a pre-release evaluation. A ther- apist’s post-release evaluation may still be more accurate than the PRB’s pre-deprivation evaluation because she can deter- mine the time at which she can accurately make a dangerous- ness determination. The policy’s failure to provide for a pre- deprivation hearing does not violate procedural due process. 3 The plaintiffs also argue that the post-deprivation process is not prompt and fair, particularly because it lacks criteria and deadlines for contact determinations. They warn that a therapist could deny contact to incentivize a parent to comply with other aspects of her therapy or for no particular reason. We address this procedural due process argument to the ex- tent it is distinct from their substantive due process argument concerning the lack of criteria and deadlines for contact No. 22-2791 11

determinations. As with the plaintiffs’ other procedural due process arguments, the first Mathews factor favors them. But the second and third factors weigh against them. There is a substantial risk of erroneous deprivation, as a parent who poses no danger to her child may be denied a final contact determination while her therapist gathers enough infor- mation to make a contact recommendation to the containment team. But the plaintiffs mischaracterize IDOC’s policy by stat- ing that a therapist could decline to make a recommendation to incentivize a parent to comply with other aspects of her therapy or for no reason whatsoever. IDOC’s policy does not allow contact to be denied for reasons unrelated to safety. And though Dr. Jerry Blain, a sex offender therapist who works with IDOC, suggested in testimony that a therapist could withhold contact approval to incentivize a person on MSR to comply with other aspects of her therapy, this view is an aberrant misunderstanding of the policy and one IDOC has rejected. We note that to the extent such misunderstand- ings persist, a person on MSR may still be able to bring an individual lawsuit claiming that IDOC has applied the policy against her in a way that violates the Fourteenth Amendment. For the same reason, the plaintiffs’ additional safeguard of objective standards and criteria has low probable value; IDOC’s policy already restricts the bases upon which a con- tact determination may be made. We expect that IDOC will ensure that containment team members understand the per- missible bases for a contact recommendation or decision. The plaintiffs’ additional safeguard of deadlines also has low probable value because the post-deprivation hearings yield an initial decision within 21 days of the parent’s contact re- quest and review every 28 days of any restrictions or prohibi- tions. And in any event, the third Mathews factor outweighs 12 No. 22-2791

the probable value of the safeguards. As with the lack of pre- deprivation hearings, IDOC has a strong interest in accurate contact determinations and thus in information about a par- ent’s post-release adjustment to community life and adapta- tion to stressors outside of prison. Due process does not re- quire a prompter post-deprivation hearing or additional cri- teria in IDOC’s policy. B The plaintiffs next assert that the policy violates substan- tive due process because (1) it lacks criteria and deadlines for contact determinations, (2) the contact ban amounts to an un- constitutional presumption of dangerousness, and (3) the phone-contact ban is irrational. To determine whether IDOC’s policy violates substantive due process, we apply the stand- ard from Turner v. Safley, 482 U.S. 78 (1987), which asks whether the policy “is reasonably related to legitimate peno- logical interests,” id. at 89. Turner identified four factors rele- vant to the reasonableness determination: (1) whether there is a “valid, rational connection” between the regulation and the governmental interest; (2) whether alternative means of exer- cising the right at issue remain available to inmates; (3) the impact accommodation of the right will have on guards, other inmates, and the allocation of prison resources; and (4) the ab- sence of ready alternatives. Id. at 89–90 (quotation omitted). Regarding the absence of ready alternatives, “if an inmate claimant can point to an alternative that fully accommodates the prisoner’s rights at de minimis cost to valid penological in- terests, a court may consider that as evidence that the regula- tion does not satisfy the reasonable relationship standard.” Id. at 91. IDOC does not have to prove the validity of its policy; No. 22-2791 13

rather, the burden is on the plaintiffs to disprove it. Overton v. Bazzetta, 539 U.S. 126, 132 (2003). The Court explained in Turner that this deferential stand- ard is necessary because prison administrators, not courts, are tasked with making difficult decisions related to institutional operations. 482 U.S. at 89 (citation omitted). It sought to avoid subjecting “every administrative judgment … to the possibil- ity that some court somewhere would conclude that it had a less restrictive way of solving the problem at hand.” Id. And in Felce, we applied Turner to the MSR context. Felce, who was on MSR, sued under § 1983, alleging that a condition of his MSR violated due process. Felce, 974 F.2d at 1485–87. In deter- mining that Turner’s reasonableness standard applied, we discussed Washington v. Harper, 494 U.S. 210 (1990), which stated that the standard “applies to all circumstances in which the needs of prison administration implicate constitutional rights,” id. at 224. We concluded, “We have no reason to doubt that this basic analysis [in Washington] is just as applicable to parole as to prison situations.” Felce, 974 F.2d at 1494. The plaintiffs argue that the district court erred in apply- ing Turner’s standard because its logic is inapplicable to the community supervision context, which does not present the same institutional security needs as do jails and prisons. They also argue that the standard should not apply where an MSR condition interferes with a fundamental right. But Felce squarely held that Turner’s standard applies to the MSR con- text, and Turner and Felce made no distinction between the different constitutional rights that administrative needs could implicate. And though the plaintiffs point to United States v. Baker, 755 F.3d 515 (7th Cir. 2014), and United States v. Good- win, 717 F.3d 511 (7th Cir. 2013), in which we vacated 14 No. 22-2791

conditions of supervised release related to sex offenders’ con- tact with their children, those cases involved special, discre- tionary conditions imposed by district courts, Baker, 755 F.3d at 523; Goodwin, 717 F.3d at 522, whereas Felce involved an MSR condition drawn up by Felce’s parole agent and the agent’s supervisor, Felce, 974 F.2d at 1487. Turner concerned deference to administrative, not judicial, decisions. Thus, Felce controls, and the district court did not err in applying Turner’s standard to evaluate whether IDOC’s policy violates substan- tive due process. 1 As with their procedural due process argument, the plain- tiffs maintain that the policy lacks any criterion or deadline for a therapist’s final contact recommendation and that a ther- apist can interfere with a parent’s right to enjoy the compan- ionship of her child indefinitely, without making any deter- mination of dangerousness. For instance, they argue that a therapist could continually deny contact approval on the ba- sis of insufficient therapy sessions to make an assessment. Looking to the first Turner factor, there is a valid, rational connection between the particular criteria and deadlines in the policy and IDOC’s interests in protecting children and re- habilitating sex offenders on MSR. It is rational to restrict par- ent-child contact before the parent’s stability and reintegra- tion into the community are assessed as part of a dangerous- ness determination. Further, it is rational to pair a 21-day deadline for the initial contact determination with a 28-day deadline for each subsequent review. These deadlines ensure that therapists have sufficient sessions with the parent to make an accurate recommendation. This factor thus weighs in IDOC’s favor. No. 22-2791 15

The second Turner factor weighs slightly against the plain- tiffs. The policy’s presumptive ban on contact applies equally to phone and in-person contact, but the district court declared the ban on written contact unconstitutional. Therefore, a par- ent may exercise her right to enjoy the companionship of her child via written communications. But while written commu- nication is an alternative means of parent-child contact, we are not oblivious to the fact that written communication with a child is far different than phone and in-person communica- tion, which are interactive. Similarly, the third factor weighs in IDOC’s favor. Impos- ing a formal deadline for a final determination may not sig- nificantly affect the allocation of its resources. But, as dis- cussed above, IDOC’s policy already contains criteria and a deadline for the initial contact determination, and the plain- tiffs do not identify what specific criteria should be added to the policy or how the addition of that criteria would affect the allocation of prison resources. The fourth factor favors IDOC, too. Imposing a formal deadline for final contact determinations would pose more than a de minimis cost to IDOC’s interest in protecting chil- dren, as a therapist may have to make a final contact recom- mendation without the information needed to make an accu- rate assessment of the parent’s dangerousness. And again, be- cause the plaintiffs do not specify what criteria they would add to the policy, it is unclear what cost the additional criteria would pose to IDOC’s penological interests. The policy is therefore reasonably related to IDOC’s interests; it does not lack criteria or deadlines such that it violates substantive due process. 16 No. 22-2791

2 The plaintiffs also argue that the policy contains a 35-day irrebuttable presumption against phone and in-person con- tact that amounts to an unconstitutional presumption of dan- gerousness. They note that there is an established methodol- ogy for assessing risk—an evaluation by a licensed sex of- fender evaluator—and that IDOC is already legally mandated to conduct a pre-release sex offender evaluation on every per- son released from IDOC who has been convicted of a sex of- fense, see 730 ILCS 5/3-6-2(j). IDOC has noted that contact may be approved in fewer than 35 days, so not every parent is subject to a 35-day period in which she awaits an initial con- tact determination. On appeal, IDOC seemingly concedes that the policy makes a presumption of dangerousness but stresses that the presumption is not irrebuttable. With respect to the first Turner factor, there is a rational connection between the presumptive ban and IDOC’s inter- ests in protecting children and preventing sexual harm. Though not every parent convicted of a sex offense will pose a danger to her child, there is at least a rational connection between the policy of disallowing contact until a dangerous- ness determination is made and IDOC’s interest in protecting children. The second factor also slightly favors IDOC, as there is no longer a presumptive ban on written contact, but written contact is substantially different than phone or in-person con- tact. The third factor favors the plaintiffs, as IDOC is already legally mandated to conduct a pre-release sex offender evalu- ation on each parent subject to the policy. But the fourth factor favors IDOC and outweighs the others. The alternative of a pre-release evaluation would pose more than a de minimis No. 22-2791 17

cost to IDOC’s goals of protecting children and rehabilitating sex offenders. As previously mentioned, a post-release evalu- ation is more accurate than a pre-release evaluation because it considers a parent’s response to life and stressors outside prison. The presumptive ban is reasonably related to IDOC’s legitimate penological interests and does not violate substan- tive due process. 3 Finally, the plaintiffs challenge the policy’s ban on phone contact as irrational for four reasons: (1) the parents regularly communicated with their children via phone while they were in prison, (2) phone contact does not present the same risk of sexual abuse as unsupervised in-person contact, (3) a custo- dial parent or IDOC agent could easily monitor phone calls, and (4) the abrupt disruption of phone contact can only have negative consequences for the children. In support of the ban, IDOC argues that it has a strong interest in assessing a par- ent’s stability and reintegration into the community before al- lowing parent-child contact and that, as the district court found, predatory behaviors are possible through phone con- tact. It thus warns that allowing phone contact between a par- ent and her child immediately upon the parent’s release from prison would require it to monitor the calls in real time and to potentially intervene to protect the child, which would pose a substantial administrative burden. Both the plaintiffs and IDOC have significant interests at stake. But phone contact is materially different than physical, in-person contact. Those speaking on an audio call cannot see each other, which is a substantial, meaningful difference. There is no exchange of facial expressions or body language, and there are no forms of physical affection typical of a 18 No. 22-2791

parent-child relationship. Also, phone calls can be more easily monitored than in-person contact and can be abruptly ended. The risks that IDOC identifies, while certainly not insignifi- cant, are less compelling when only phone contact is in- volved. We are mindful that IDOC has an interest in protect- ing children from their own parents if it has “some definite and articulable evidence giving rise to a reasonable suspicion that [another] child has been abused or is in imminent danger of abuse.” Brokaw v. Mercer Cnty., 235 F.3d 1000, 1019 (7th Cir. 2000). But IDOC has tools at its disposal to minimize or elim- inate the risk of abuse, and we must also consider the parents’ fundamental liberty interest in the enjoyment of the compan- ionship of their children. With that, we move to the applica- tion of the Turner factors. As to the first Turner factor, there is a rational connection between the phone ban and IDOC’s interests in protecting children and rehabilitating sex offenders on MSR. It is rational to restrict contact before the parent’s dangerousness is accu- rately assessed, and therefore this factor weighs in IDOC’s fa- vor. But we note that the rational connection is weaker when only phone contact, not both phone and in-person contact, is restricted. The second factor weighs slightly against the plain- tiffs because though they can contact their children through written communications, written communication is signifi- cantly different than interactive phone contact. Our analysis instead turns on the third and fourth Turner factors, which weigh heavily in the plaintiffs’ favor. The plaintiffs argue that a custodial parent or an IDOC agent could easily monitor a phone call between a person on MSR and her child. IDOC responds that the district court found that call monitoring would pose a substantial administrative No. 22-2791 19

burden. But there is no evidence in the record to suggest that call monitoring by a custodial parent or an IDOC agent while a parent awaits a final contact determination would create such a burden, and the district court did not explain its find- ing. See Pyles v. Nwaobasi, 829 F.3d 860, 868 (7th Cir. 2016) (“Where the evidence underlying a fact … supports only spec- ulation about the existence or nonexistence of the contested point, it is clear error to conclude that the point has been es- tablished.”). In fact, the record indicates that custodial parents are already involved in call monitoring. Steve DeYoung, an IDOC parole commander, testified during the bench trial that if a parent is approved for phone contact, the therapist on the containment team may request that the child’s custodial par- ent monitor the phone calls. And an IDOC agent could also monitor a call: a person on MSR convicted of a sex offense must meet with her parole agent at least once a month and thus could schedule a monitored call during one of those re- quired meetings. The district court additionally found that no evidence sug- gests that IDOC could “recreate at de minimis cost the existing infrastructure used to record or monitor prison phone calls,” but that is not the proper inquiry for a person on MSR. The record gives no indication that IDOC must monitor phone calls made outside prison in the same way it monitors calls inside prison to address its safety concerns. See Heyer v. U.S. Bureau of Prisons, 984 F.3d 347, 364 (4th Cir. 2021) (finding no evidence in the record to support the district court’s conclu- sion that a particular phone call security measure was neces- sary to address security concerns). For instance, Brown-Foiles testified during the bench trial that any potential risk from a phone call would be ameliorated by monitoring the call and did not testify that the monitoring had to occur in the same 20 No. 22-2791

way it does in prison to be effective. And because call moni- toring would minimize the risk of abuse, monitoring would not pose more than a de minimis cost to IDOC’s penological interests in protecting children and rehabilitating sex offend- ers. “Here, as in Turner, there may well be ‘obvious, easy alter- natives to the [phone-contact ban] that accommodate the right [to enjoy the companionship of one’s child] while imposing a de minimis burden’” on IDOC’s child safety and rehabilitation objectives. Riker v. Lemmon, 798 F.3d 546, 558 (7th Cir. 2015) (quoting Turner, 482 U.S. at 98). “Absent significantly more evidence explaining the importance of banning [phone con- tact], Turner does not allow us to accept at face value the De- partment’s unsubstantiated contentions.” Id. IDOC could modify its policy to bring it into compliance with substantive due process. But on this record, the phone-contact ban is not reasonably related to legitimate penological interests. We therefore conclude that IDOC’s policy is unconstitutional in- sofar as it presumptively prohibits parent-child phone contact while a parent awaits a final contact determination and re- verse as to the phone-contact ban. AFFIRMED IN PART AND REVERSED IN PART