What I Saw Inside Two Residential Facilities — And Why the Better One Still Wasn’t Good Enough
If you’re a parent weighing a placement, a staff member trying to name what’s bothering you about your own workplace, or a survivor looking for language for what happened to you — this one’s for you too.
I’ve worked inside two residential facilities as a paid professional, and volunteered in several more that existed in conditions I still don’t have adequate words for. I’ve been the person doing bathroom care for a nonverbal teenager while his 1:1 sat across the room on her laptop. I’ve been fired over a clean pull-up. I’ve watched a kid flinch at a staff member’s raised hand and watched that flinch go unrecorded, unquestioned, and unrepeated to anyone who might have cared. I am telling you this not because my story is unusual, but because it isn’t, and that’s the actual scandal.
People ask me to rank the facilities I’ve been inside, as if there’s a clean hierarchy from bad to good and the job is just to land your kid or your loved one as high on it as possible. There isn’t. There’s a hierarchy of resourcing — of clean buildings versus rotting ones, of trained staff versus untrained ones, of admin that at least performs accountability versus admin that doesn’t bother. But underneath every tier of that hierarchy sits the same structural rot: a residential facility is a place where the people with the least power are supervised by the people with the most, and almost nothing in the system is built to check what happens when that power goes unwatched.
The best-resourced facility I worked at still ran on fear, just a more polished version of it
I’ll call it Program One, because the actual name doesn’t matter as much as the shape of what happened there. Program One served kids with severe and complex needs — many of whom would call that building home until they aged out into adult services. And by the standards of residential care in this country, it was good. Genuinely good. The caregiving skills staff were expected to bring to the floor were treated as the actual job, not an inconvenient prerequisite to it. Crisis intervention training was CPI-based and consistently reinforced. There was an operating theory of “this is someone’s home” that showed up in small, deliberate ways — decor, routine, the effort put into making a Tuesday feel like a Tuesday and not like a shift change.
None of that made it safe from the thing every institution is vulnerable to: the gap between the stated standard and the enforced one.
Admin at Program One was strict, and the theory behind the strictness was sound — you don’t want loose accountability around vulnerable kids, full stop. But strictness aimed downward and blindness aimed upward is not a system. It’s a hazard with good production values. I watched staff who were skilled, patient, and genuinely attached to their kids get micromanaged over minor infractions, while I watched other staff — the ones with tenure, the ones plugged into the social ecosystem of the building — get a pass on behavior that should have ended their employment. I watched students flinch away from certain staff members’ physical presence. Not once. Repeatedly. Nobody wrote it down. Nobody asked the kid why. A flinch is data. A flinch is a kid’s nervous system telling you something their mouth might not be able to, and in a well-funded, well-trained program, it still went nowhere, because “nowhere” is where uncomfortable data goes when the person generating it has less institutional standing than the person it implicates.
I lost my own job there over a clean pull-up. That’s not an exaggeration for effect — it’s the literal, absurd truth. A coworker had put the pull-up on her student for an outing, against the plan, and after I finished caring for him in the bathroom I didn’t discard it, because it was clean and it wasn’t mine to make that call on. She was on her laptop doing schoolwork when I returned him to her. She reported me. I was gone within the week. I don’t tell you this because I think my termination is the most important thing that happened in that building — it isn’t, not by a long shot — but because it’s a perfect, small-scale model of how these places actually allocate consequence. Petty social conflict between staff can end a career overnight. A kid flinching at an adult’s hand can go unaddressed for months. If you want to know what an institution actually values, don’t read its mission statement. Watch what it’s willing to investigate.
The second facility taught me that “no cameras” is not a coincidence, it’s a policy decision
Program Two served a different population — kids who couldn’t live at home, separated by gender and age or cognitive grouping, trained in TCI rather than CPI. I spent most of my time with the middle-years girls, sometimes with the older group. And the thing that told me everything I needed to know about that building had nothing to do with training model or population. It was the absence of cameras.
Not “insufficient” cameras. Not cameras in the wrong spots. Absent. And I want to be precise about what that absence actually functions as, because it’s easy to hear “no cameras” and think privacy or budget and move on. In a facility housing kids who cannot safely live anywhere else, the absence of a recording mechanism is not a neutral fact about infrastructure. It is a decision about whose word will count when something goes wrong. Without a camera, an allegation is just a kid’s word against a staff member’s, and a kid who has already been told by every adult in his life that his word doesn’t hold much weight is not exactly starting from a position of institutional credibility. Facilities know this. That is precisely why the absence persists.
I watched complaints get quietly reclassified as allegations — a word that sounds procedurally neutral but functions, in practice, as a soft demotion. An allegation, unlike a complaint, doesn’t require the same urgency of response. It can sit. It can wait for corroboration that, without cameras, will never fully materialize. I witnessed abuse directly. I was told, by people who’d been there far longer than I had, that what I’d seen was recent history, not distant history — that there was a long back catalog of things considerably worse than what I’d personally caught. And the machinery around all of it didn’t grind to a halt or throw up red flags. It just kept running, the way water keeps running around a rock in a stream, because the water was never built to stop for the rock.
The difference between the two facilities is real, and it is also not the point
A well-resourced facility with excellent training and a genuine philosophy of care can still fail a child in ways that never make it into an incident report, because the failure isn’t a broken rule — it’s a rule nobody wrote down in the first place.
I want to sit with that distinction for a second, because it matters for how you evaluate a placement. Program One and Program Two are not the same story. One had trained staff, decent facilities, and a leadership team that at least understood the theory of what good care looked like. The other had none of that, plus a structural void where oversight should have lived. If you are a parent choosing between two placements and one of them looks like Program One and the other looks like Program Two, choose Program One. That is not a controversial statement, and I’m not going to pretend otherwise for the sake of a tidier moral.
But here is the part that should actually keep you up at night: the type of failure at Program One — the flinching that never got documented, the petty social power dynamics that decided who kept their job, the blind spots leadership never checked because checking them would have implicated people they trusted — is not a resourcing problem. You cannot buy your way out of it with better funding or better training, because it isn’t a training gap. It’s a structural feature of any institution where the people being cared for have less power than the people being paid to notice when care goes wrong. Program Two’s failures were louder and uglier. Program One’s failures were quieter and better dressed. Neither one was accountable in the way a family, watching from the outside and hoping for the best, would assume it was.
What a flinch actually means, and why nobody in the building is incentivized to ask
If you’ve spent any time around trauma-informed care, you already know that a flinch isn’t a behavior problem. It’s a nervous system doing exactly what it evolved to do — bracing against a stimulus it has learned, through direct experience, to associate with harm. A kid who flinches at a specific staff member’s raised hand is not being dramatic, and he is not confused about who that staff member is. He has data the rest of the building doesn’t have, gathered through a channel — his own body — that the reporting structure isn’t built to receive. Incident reports want words, timestamps, corroborating witnesses. A flinch offers none of that, and so a flinch, however loud it is to the kid experiencing it, registers as silence to everyone else.
This is where “why didn’t anyone do anything” runs into the actual mechanics of institutional life. Doing something about a colleague’s behavior means initiating conflict with someone you’ll be sharing a building with for the next five years. It means risking your own standing in a social ecosystem you still have to survive in tomorrow, and the day after. Admin that is strict about staff conduct on paper but incurious about staff conduct in practice creates an environment where the actual risk calculation for a bystander isn’t should I report this — it’s what happens to me if I do, versus what happens to the kid if I don’t, and which of those costs do I have to live with personally. That calculation, repeated across enough staff and enough kids and enough years, is how a facility with genuinely good bones ends up with the exact same blind spot as a facility with none.
If you’re choosing a placement, you are not choosing safety — you are choosing a set of risks to manage
I’m not going to end this with a tidy list of red flags to watch for, because you’ve read enough of those, and this piece was never meant to be that one. What I want you to take with you instead is this: there is no version of residential placement, however well-funded, however well-trained, that hands you safety as a finished product. What you are actually choosing, every time, is which category of risk you are willing to monitor for, and how hard you are willing to fight to be believed when you see something. Program One made that fight easier because the staff were skilled and the philosophy was sound. It did not make that fight unnecessary. Nothing does. Not yet. Maybe not ever, until the systems themselves are built to reward the person who documents the flinch instead of the person who lets it pass.
That’s the actual work. Not finding the perfect building. Building the muscle, in yourself and in every family and case manager and staff member reading this, to keep asking the uncomfortable question even when the building around you has stopped expecting it.