SOJOURNER FAMILY PEACE CENTER FAILED TO PROTECT SIX-MONTH-OLD BABY — RESIDENTS HAD TO INTERVENE WHILE AN INFANT’S SAFETY HUNG ON INSTITUTIONAL PARALYSIS

A six-month-old baby cannot call 911. He cannot call Child Protective Services. He cannot photograph a bruise, explain why he is crying, tell a police officer what happened after everyone left the room, file a grievance against the adults responsible for his safety, or walk away from someone frightening him. He cannot read a confidentiality policy or understand why adults who believe he is being hurt are debating whether they are legally allowed to tell anyone outside the building. He has exactly one protection available to him: adults who recognize danger and act. Inside Sojourner Family Peace Center, that protection broke down so badly that residents eventually began discussing whether they were willing to physically intervene themselves—and risk arrest, injury, or losing their own shelter placements—because somebody had to protect a six-month-old baby.

That sentence should make people angry. It should make Sojourner's Board angry. It should make government agencies financing domestic-violence services angry. It should make child-welfare officials, grant administrators, donors, regulators, elected officials and every organization placing its name beside Sojourner's angry enough to demand the records and determine exactly what happened. This was not a disagreement over whether a toddler received the correct bedtime. The conduct documented during my investigation involved an infant approximately six months old, a child so developmentally helpless that every element of his safety depended upon adults recognizing behavior he could neither escape nor report.

I personally witnessed conduct involving this infant that went far beyond a parent appearing tired or frustrated. I witnessed the caregiver screaming profanities at the baby, including telling him to “shut the fuck up.” I witnessed her become angry at a six-month-old because he could not make the kind of decisions about food that she apparently expected him to make. I witnessed rough and punitive handling, including squeezing or pinching during an interaction involving a high chair, forcefully placing him into that chair, and then violently pushing the occupied high chair across the cafeteria before walking away. Other residents separately described seeing bruising and described the infant being placed alone in a bathroom with the door closed. Staff were present for at least portions of the conduct. Residents repeatedly raised concerns.

What residents did not see was equally important. We did not see an obvious child-protection response capable of reassuring us that the baby was being protected. We did not see Child Protective Services arrive. We did not see law enforcement arrive to investigate the conduct we were witnessing. We did not see a medical assessment of the infant. We did not see a structured child-safety intervention that explained to frightened residents that qualified professionals had taken control of the situation. Instead, the concern continued for weeks until residents themselves were discussing what they were going to do the next time they believed that baby was being hurt.

Read that again in the context of where this happened. This was not a crowded apartment building where neighbors wondered whether they should intervene in another family's business. This was Sojourner Family Peace Center, an institution built around recognizing violence, assessing danger, protecting vulnerable people, coordinating with law enforcement and helping human beings who cannot safely remain where violence is occurring. Yet inside that institution, residents who had themselves entered shelter because they needed protection reached the point of discussing whether they would have to become the emergency response for a defenseless infant.

Several residents openly discussed whether they were willing to be arrested for physically removing the baby from the caregiver or fighting the caregiver if she harmed him again. That is not a minor operational failure. Residents in emergency domestic-violence shelter should never have to calculate whether protecting a baby requires them to put their own bodies between a child and another resident while wondering whether police will arrest them afterward. They should never have to decide whether an assault charge, physical injury, shelter discharge or retaliation is an acceptable price for doing what they believe the professionals around them are refusing to do. When traumatized residents begin planning an improvised child-protection response inside a professional domestic-violence institution, leadership has lost control of something fundamental.

Then the situation became even more disturbing. Residents, including me, were told that we could not independently contact CPS or law enforcement about the abuse because doing so would violate shelter confidentiality and could jeopardize our shelter placements. Whatever institutional policy was intended by that instruction, its practical message was unmistakable: speak outside the institution about what you believe is happening to this baby, and you could lose the shelter protecting you from your own danger.

That is an obscene position in which to place a domestic-violence survivor. A woman may have arrived at Sojourner after escaping an offender who used housing, money, dependency, threats and consequences to control her choices. She may have nowhere safe to return to. She may have children of her own. She may be terrified of being located. Then she sees conduct toward an infant that frightens her and believes the child needs outside protection. Her decision is no longer simply whether to make a report. She has to calculate whether making that telephone call could cost her the bed standing between her and the person she fled.

That is not meaningful choice. That is institutional leverage over conscience. And Wisconsin law is considerably more complicated than the slogan residents say was effectively handed to them. Wisconsin's child-abuse reporting law, Wis. Stat. § 48.981, creates mandatory reporting obligations for specified professionals under defined circumstances. The critical threshold for a covered mandatory reporter is not courtroom proof that abuse occurred. The statute addresses professionals who have reasonable cause to suspect that a child seen in the course of their professional duties has been abused or neglected, or reason to believe that abuse or neglect has been threatened and will occur. Wisconsin's courts have treated that reporting requirement as an actual legal obligation, not a suggestion that disappears whenever an institution becomes uncomfortable with disclosure. The Wisconsin Court of Appeals in State v. Ziehr emphasized the statutory command that a mandatory reporter “shall immediately” inform the appropriate authorities and rejected the proposition that a covered reporter is automatically relieved of the obligation merely because somebody else may have independently reported the incident.

That does not mean every person employed by Sojourner is automatically a mandatory reporter. Wisconsin law does not create a blanket professional category called “domestic-violence shelter employee.” The actual employee's profession, credentials, duties and circumstances matter. A social worker, healthcare professional, counselor, child-care professional or another person falling within a statutory category can present a different legal question from a shelter employee who does not. That is precisely why an institution housing children cannot operate on organizational folklore about whether “Sojourner employees” are or are not reporters. The law attaches obligations to people and professional roles under specified circumstances; the institution's responsibility is to know which of its people carry which obligations.

Wisconsin law also does something critically important that destroys the simplistic idea that a person who is not a mandatory reporter therefore has no legal avenue to report. Section 48.981 separately provides a permissive-reporting pathway for people outside the mandatory categories under the statutory conditions. The legal architecture therefore contains two entirely different questions: Who must report, and who may report? Those questions cannot be collapsed into “nobody is allowed to report.”

Federal domestic-violence confidentiality law does not rescue that categorical proposition. It makes the need for competent institutional policy even clearer. The Family Violence Prevention and Services Act contains strong confidentiality protections for victims because disclosing the identity or location of someone fleeing domestic violence can have catastrophic consequences. But Congress did not turn confidentiality into a gag order protecting suspected child abuse from outside scrutiny. The current federal statutory text expressly says its nondisclosure provisions do not prohibit an FVPSA grantee or subgrantee from reporting abuse or neglect where reporting is mandated or expressly permitted by the state or tribe involved.

That language matters enormously here. Wisconsin distinguishes mandatory reporting from permissive reporting. FVPSA expressly preserves reporting that state law mandates or expressly permits. The federal framework therefore does not answer a Wisconsin child-safety question with the word “silence.” It requires an institution receiving covered funding to understand what Wisconsin law permits and requires and then protect survivor confidentiality while complying with that legal architecture.

Confidentiality is not the enemy here. In a domestic-violence shelter, confidentiality can literally keep people alive. An offender learning where a survivor is sleeping, where her children attend school, what telephone number she is using, where she plans to move or when she intends to appear in court can create immediate danger. Sojourner should protect that information aggressively. But protecting a survivor's confidential information and activating a legally authorized child-protection process are not mutually exclusive concepts. The answer to sensitive information is competent disclosure control: know what can be disclosed, know what must be disclosed, know who receives it, disclose no more than the law requires or permits, and continue protecting everything else.

A six-month-old baby should not absorb the consequences because an institution failed to master that distinction.

The situation becomes harder to comprehend when the institution's mission is considered. Domestic-violence organizations exist because society has already rejected the idea that violence occurring within a family should remain hidden merely because it occurs inside a private relationship. Child-protection laws exist for the same fundamental reason. A child does not cease to require protection because the adult responsible for that child is herself a survivor of domestic violence. Survivorship and harmful parenting behavior can coexist. Trauma can explain behavior without making dangerous behavior safe. A parent can desperately need protection from an offender while her infant simultaneously needs protection from her conduct.

That is not an attack on survivors. It is what taking survivors seriously looks like when more than one vulnerable human being exists in the same family.

A sophisticated domestic-violence organization should be exceptionally capable of understanding that complexity. A parent entering shelter may be traumatized, exhausted, sleep deprived, economically destabilized, frightened and overwhelmed while caring for a baby whose crying, feeding difficulties and constant dependence increase stress. That predictable reality is precisely why family shelter should contain strong parenting support, infant-development education, respite, childcare, behavioral intervention and pathways for escalating concerns before somebody gets hurt. The answer should never be to ignore dangerous behavior because the parent is a survivor. The answer should also not be to wait until the situation becomes intolerable and then simply throw the mother and baby out.

What happened before the eventual removal is not entirely an unanswered question, because I was there when the removal occurred. Before that moment, residents had already witnessed conduct toward the six-month-old infant that caused serious concern. Residents had raised concerns. Staff had been present for at least portions of the conduct. Yet the event that finally caused Sojourner leadership to order the caregiver to pack her belongings and leave was not another incident involving the baby. It was not the rough handling residents had witnessed. It was not the screaming and profanity directed at a six-month-old. It was not the forceful high-chair incident. It was not the concerns residents had communicated about the child's safety. The immediate removal occurred after the caregiver gave attitude to a Sojourner supervisor named Rho. Once that confrontation occurred, she was told to pack her belongings and leave.

That sequence makes the institutional response substantially more difficult to explain. Sojourner had apparently tolerated a growing child-safety problem until the caregiver's behavior was directed at institutional authority. The baby could not demand respect from a supervisor. The baby could not complain about how he was being treated. The baby could not challenge staff, violate a behavioral expectation, or create an interpersonal conflict with management. Residents could raise concerns on his behalf, and they did. Yet the consequence that finally removed the caregiver from the shelter came when the conflict reached a supervisor personally. If the chronology is accurately reflected in Sojourner's records, oversight agencies should examine why conduct frightening residents about the physical safety of an infant apparently failed to produce the immediate institutional consequence that disrespect toward a supervisor produced.

The departure did nothing to reassure the residents watching what happened next. Residents watched the caregiver leave with the six-month-old and witnessed her handling the baby while placing him into the vehicle in a manner residents described as literally throwing him into the car seat before driving away. The people who had already been discussing whether they would physically intervene were now watching the infant disappear from the one environment in which other adults could still see what was happening to him. Residents continued discussing whether somebody should call CPS or law enforcement because removal from Sojourner had not answered the question that mattered to them: Was the baby safe?

That is where Sojourner's response becomes extraordinarily difficult to reconcile with child protection. Removing a resident for her conduct toward a supervisor can protect staff and institutional order. It does not protect an infant from the caregiver leaving with him. If Sojourner believed the previous conduct did not create sufficient concern to activate a child-protection response, its records should establish what assessment produced that conclusion. If the institution believed the conduct was sufficiently concerning that the child's safety remained in question, sending the caregiver and infant out the door without a documented protective handoff would not resolve that concern. It would simply move the child beyond the view of the residents who had been watching him.

Those questions become especially important because the investigative material identifies problems with childcare and child-development resources inside the shelter. In a family domestic-violence program, childcare is not merely something that allows a parent to attend an appointment. Under the conditions in which survivors arrive, respite and competent child-development support can become part of violence prevention itself. A caregiver reaching a point of severe dysregulation may need safe separation, support and intervention before frustration becomes physical. If childcare or partner programming deteriorated because of scheduling problems, organizational conflict, funding decisions or another administrative breakdown, the institution needs to explain what replaced it.

A six-month-old cannot wait for two organizations to resolve a scheduling dispute.

Eventually, the caregiver and infant were removed from Sojourner after residents intervened. That did not answer the child-safety question. It made the question more urgent. Residents remained distressed about what they had witnessed and did not know what happened to the baby afterward. The investigative material includes a resident account describing forceful handling of the infant while the caregiver was leaving. Yet the same understanding about confidentiality remained hanging over the residents who were already afraid that contacting outside authorities could jeopardize their own shelter placements.

Removing a baby from the building is not the same thing as protecting the baby.

If Sojourner determined that the caregiver presented no serious child-safety concern, then the institution should be able to explain why the conduct ultimately warranted immediate removal. If the conduct was serious enough to require immediate removal, then Sojourner should be able to explain what child-protection mechanism accompanied sending the caregiver and six-month-old infant into an environment with less institutional observation, not more. Where did they go? Was CPS contacted? Was law enforcement contacted? Was another shelter or service provider notified through a legally authorized mechanism? Was a medical assessment performed? Did anybody confirm that the child reached a safe environment?

But “we don't know” is not reassuring when the person nobody can account for is six months old.

There is another reason the reporting confusion demands immediate scrutiny. Wisconsin law does not require a mandatory reporter to become a detective before making a report. The system is designed so that the appropriate child-protection or law-enforcement authority can investigate. A frontline professional does not have to conduct an interrogation, diagnose bruising, reconstruct every interaction, establish criminal intent or prove a case beyond a reasonable doubt before the reporting mechanism becomes relevant. The statutory threshold for a covered reporter turns on reasonable cause to suspect abuse or neglect under the circumstances described by the law. The Wisconsin court system's own confidentiality guidance describes § 48.981 as requiring covered professionals to report where they have reasonable cause to suspect abuse or neglect or reason to believe threatened abuse or neglect will occur.

That design exists for an obvious reason: waiting for certainty can be deadly.

Nobody knows which frightening interaction with an infant will remain a frightening interaction and which will become the event that produces catastrophic injury. Nobody watching an increasingly dysregulated caregiver can know with certainty whether the next hour will end uneventfully or in an emergency department. Child-protection systems therefore exist to move qualifying suspicions into the hands of people legally empowered to investigate rather than demanding that every observer wait until the evidence becomes undeniable.

Because undeniable can arrive too late.

Imagine the institutional record if the worst possible outcome had occurred after residents raised concerns. Investigators would reconstruct who saw what. They would determine which staff were present. They would interview residents who complained. They would obtain policies, training materials, case notes, surveillance records, internal communications and shift assignments. They would identify which employees were legally mandatory reporters and what those employees understood their obligations to be. They would ask whether residents were discouraged from calling CPS or law enforcement. They would ask who created the confidentiality interpretation. They would ask whether supervisors delayed action. They would ask what interventions were available before discharge. They would ask where the child went afterward.

And if the child had suffered catastrophic injury or died, nobody would care that the institutional confusion was well intentioned.

That is not a prediction that this particular infant would have died. The evidence does not establish that, and pretending otherwise would be irresponsible. It is the reason child-safety systems are designed around prevention rather than autopsy. Infants cannot compensate for adult hesitation. They cannot absorb one more violent movement while adults schedule a meeting about policy. They cannot explain that yesterday's bruise matters. They cannot decide that the environment has become unsafe and leave.

An infant experiences whatever adults allow to happen while adults are deciding what to do.

That is why the most disturbing fact in this entire episode may not ultimately be any single act by the caregiver. It may be what the residents believed they had to become because the institution around them appeared incapable of producing a protective response they trusted. These were domestic-violence survivors living inside emergency shelter. They were not child-protection investigators, police officers or an emergency-response team. Yet some were discussing whether they would physically take the baby, fight another resident and accept arrest if that was what protecting him required.

Sojourner should not dismiss that as residents becoming overly involved. It should ask why residents believed nobody else was going to act.

The federal government should be asking the same question. Wisconsin child-welfare authorities should be asking it. The City of Milwaukee should be asking it. Organizations funding Sojourner should be asking it. So should every government entity whose money, contracts or partnerships support an institution responsible for housing vulnerable adults and children. This is exactly the kind of incident that oversight exists to examine before the next child becomes the case everyone claims nobody could have foreseen.

The review should begin with the actual records. Sojourner should preserve and produce, to the extent legally authorized for appropriate investigators and oversight bodies, the incident documentation, relevant case-management records, shift assignments, supervisory communications, internal messages, child-safety complaints, reporting-policy materials, mandatory-reporter training, confidentiality training, documentation of resident complaints, relevant surveillance evidence, records concerning the discharge decision, records showing what services or interventions were offered before removal, and records establishing what protective handoff occurred afterward. Oversight should identify every employee who personally witnessed relevant conduct and determine that person's professional role and reporting status at the time.

The institution should also be required to answer what residents were actually told. Not what a polished policy says today. Not what counsel later determines employees were supposed to say. Not what leadership wishes residents had understood. What were survivors living inside the building actually told about calling CPS or law enforcement when they believed a child was being abused? Who said it? Was shelter removal threatened? What written rule supposedly authorized that consequence? Was that rule reviewed against Wis. Stat. § 48.981 and the federal confidentiality provisions applicable to the program? How many previous child-safety incidents were handled under the same understanding?

A technically correct policy buried in a binder is worthless if the operational policy communicated in the hallway is wrong.

That is why oversight should test employees rather than simply request a training certificate. Ask workers from different shifts and different professional roles the same question: You personally observe conduct tonight that gives you reasonable cause to suspect a six-month-old infant is being physically abused. What do you do? If one says call CPS, another says call a supervisor, another says Sojourner cannot report, another says only Aurora can report, another says confidentiality prohibits disclosure, and another does not know, the child-safety system has already failed its most basic test.

The answer cannot depend upon which employee happens to be standing closest to the baby.

Sojourner also needs to explain whether residents can make good-faith reports without being threatened with homelessness. Wisconsin's statutory structure includes protections surrounding good-faith participation in child-abuse reporting and investigation, while federal FVPSA law expressly preserves state-mandated and state-permitted abuse reporting for covered grantees and subgrantees. Whatever confidentiality obligations properly bind shelter residents, the proposition that a survivor can be removed from emergency shelter simply for lawfully alerting child-protection authorities to suspected serious abuse demands immediate legal scrutiny.

The issue is not solved by choosing child protection over survivor confidentiality. That would be another failure. Domestic-violence confidentiality exists because information can become a weapon. A competent system protects the survivor's confidential location, limits unnecessary disclosure, safeguards unrelated residents, discloses only information legally required or authorized, and simultaneously ensures that a child does not disappear behind confidentiality language that the law itself does not make absolute.

That requires competence, not slogans.

“Confidential” is not the same word as “untouchable.” “Not a mandatory reporter” is not the same thing as “legally forbidden to report.” “Tell your supervisor” is not necessarily the same thing as satisfying an independent statutory reporting obligation. “The family was discharged” is not the same thing as “the child was protected.” And “the baby did not die” is sure as hell not evidence that the system worked.

The federal child-protection framework itself reflects the principle that safety and confidentiality are supposed to coexist. Federal child-abuse law requires state systems to include reporting mechanisms, immediate steps to protect abused or neglected children, good-faith reporting protections, confidentiality safeguards and cooperation among child-protection and law-enforcement authorities. In other words, the architecture is not built around pretending privacy and protection are mutually exclusive. It is built around doing both competently.

Sojourner's Board should therefore treat this episode as a sentinel event whether or not an outside investigation ultimately substantiates every resident concern. A near miss is valuable precisely because it exposes the same structural weakness that can produce catastrophe next time. The question is not whether leadership can say the mother and infant eventually left. The question is how residents reached the point of contemplating violence and arrest to protect a six-month-old while simultaneously believing that calling the people legally empowered to investigate child abuse could cost them their own shelter.

There should be no comfortable institutional answer to that question.

There should be records. There should be names. There should be timestamps. There should be policies. There should be training materials. There should be a reconstruction of who knew what and when. There should be a legal determination of who was required to report, who was permitted to report and what confidentiality provisions actually applied. There should be an explanation of what happened before discharge and what happened after it. There should be an accounting of whether the baby received medical assessment, child-protection intervention or a safe handoff. There should be an examination of whether childcare and parenting-support failures allowed a preventable situation to escalate.

And if those records do not exist, that is an answer too.

A domestic-violence institution does not get to build its public identity around recognizing violence and then become confused when violence may be occurring inside its own walls. It does not get to teach survivors that silence is dangerous in abusive relationships while creating an environment in which residents believe speaking to outside authorities about a defenseless baby could make them homeless. It does not get to invoke trauma-informed care as a reason to avoid confronting dangerous behavior. And it does not get to call removing a mother and infant from the building a child-safety solution unless somebody can explain how the child became safer when the doors closed behind them.

The mother in this story mattered. Her trauma mattered. Her safety mattered. Her need for support mattered. If she was overwhelmed, she deserved intervention before crisis. If she lacked parenting knowledge, she deserved education. If she needed respite, Sojourner should have had a pathway to provide or locate it. If she was experiencing behavioral-health problems, those deserved attention. Being a survivor should have increased the sophistication of the institutional response.

It should never have decreased the value of her baby's safety.

The residents mattered too. They entered shelter because they needed protection, not because they volunteered to become an underground child-protection team. They should never have been forced to choose between the roof protecting them and the infant they believed needed help. They should never have had to discuss whether they were willing to fight another resident because the institutional response left them believing physical intervention was the only option remaining.

But the person with the least power in this entire story was six months old.

He could not tell Sojourner that its policy made no sense. He could not challenge anyone's interpretation of federal confidentiality law. He could not ask whether the employee standing nearby happened to hold a professional credential covered by Wisconsin's reporting statute. He could not call CPS when the adults were afraid to. He could not demand that the Board investigate. He could not ask a government funder why an institution receiving money to protect vulnerable families appeared unable to produce a clear protective response for him.

He could only cry and depend upon adults to understand what that dependence required.

The public should now demand the answer he could not.

Who was protecting the baby?