In Granbury, homelessness doesn't look like a shopping cart. It looks like a sofa.

Why nobody sees her

A homeless woman almost never looks like the picture in your head.

She goes to great lengths not to. She showers at a friend's before work so that nobody at work can tell. She keeps her clothes in the boot of the car in a laundry basket, because a laundry basket looks like errands and a bin bag looks like what it is. She parks somewhere different every night. She stays polite and useful and easy to have around, because the sofa she is on belongs to someone whose patience is a resource she is spending down, and she can feel it running out.

And when the sofa ends, very often the only offer on the table comes from a man. Which is how a woman ends up back in a house she escaped from — not because she forgot what he did, but because it was February and she had a nine-year-old in the back seat.

Women's homelessness is not hidden by accident. It is hidden on purpose, by the woman, because being visibly homeless is more dangerous for her than for almost anyone else.

Our board member Karen Kalinka spent seven years building and running a 24-bed women's recovery programme inside a downtown Edmonton shelter — a programme that stayed full, serving roughly fifty women a year — while also managing the day drop-in and the night shelter. She has looked at this population for a very long time. Here is what she says about Granbury:

"Granbury's homelessness is invisible. In a wealthy community, it does not look like shopping carts or street corners. It looks like a woman sleeping in her car. A woman couch-surfing from house to house. A woman in a tent by the river. A woman staying in a dangerous relationship because she has no other option."
— Karen Kalinka, Mercy Manor board

The count cannot see her — and now it doesn't even try

This is not a feeling. It is a matter of federal regulation, and you can check every line of it.

1. The sofa is excluded by rule

Once a year, communities carry out a Point-in-Time count. It enumerates people who are "literally homeless" — in a shelter, in transitional housing, or in a place not meant for human habitation. Doubling up — staying temporarily with friends or family — is not in that definition. (24 CFR § 578.3) She isn't overlooked. She is out of scope.

2. And the sofa is where nearly everyone is

Schools use a broader definition that does include doubling up. In Texas, in the 2024–25 school year, 74.31% of students identified as homeless were doubled up — against 8.55% in shelter and 7.27% unsheltered. (Texas Education Agency, 2025)

Three out of four. The category the federal count leaves out is not a rounding error — it is nearly the whole thing.

3. The government's own auditors say the count is low

The Government Accountability Office found the Point-in-Time count "likely underestimates the number of people experiencing homelessness," and noted that none of the researchers and participants it consulted thought the count overestimates. (GAO-20-433, 2020)

GAO is equally clear that the size of the undercount is unknown. So we will not put a number on it, and you should distrust anyone who does.

4. Researchers have found for decades that this is where women go first

An international evidence review published by the Scottish Government names women, and women who are parents, first among the groups consistently found in hidden homelessness. It reports that women "tend to exhaust their informal options of accommodation, like staying with friends and family, before presenting to homeless services," and that this "leads to an undercount in methods like street counts and counts that use homeless shelter's administrative data." When the definition widens to include concealed homelessness, women appear in higher numbers. (Scottish Government, 2023)

The same review records that women sleeping rough conceal themselves — and conceal their gender — as a survival strategy.

5. Where the count still looked, the pattern held

Nationally, in the 2024 federal count, 72% of the homeless women counted were sheltered, against 60% of the men — women, far more than men, found a roof of some kind rather than the open street. (HUD 2024 Point-in-Time, national; percentages our arithmetic) In Texas the same gap held: 34.2% of counted women were unsheltered, against 49.7% of men. (HUD 2024, Texas)

Being scrupulous: this is consistent with women avoiding the street, but it does not prove an undercount — it is equally consistent with shelters prioritising women and children. We offer it as the pattern the argument predicts, not as proof of it.

6. And as of the latest count, she is not reported at all

What this looks like in Hood County

Hood County sits in the Texas Balance of State Continuum of Care — the enormous rural catch-all that covers most of the state outside the big cities. In the January 2025 count it recorded 9,234 people homeless, 5,351 of them unsheltered — about 58%. (HUD, TX-607, 2025)

But here is the part that matters locally. The agency that runs that count says this about how it works:

"Because of the TX BoS CoC's large geography, we depend on communities to volunteer to perform a PIT Count… We can't provide data on communities who don't perform a count."
— Texas Homeless Network

Hood County does not appear among the network's published Point-in-Time reports by area for 2022, 2023, 2024 or 2025.

So when somebody tells you there isn't really a homelessness problem in Granbury, understand precisely what is being said. It is not that we looked and found nobody. It is that the count depends on volunteers turning out to walk a county, and nobody published a result for this one. Absence of evidence has been quietly promoted to evidence of absence.

The schools, meanwhile, use the definition that can see a sofa — and they see far more. Granbury ISD reports roughly 220 students who qualify as homeless under the federal McKinney-Vento definition: children living on the street, in an abandoned building, in a car, couch-surfing from house to house, or in transitional housing. (Granbury ISD) The state's single-day PEIMS snapshot, which captures only a slice of the year, logged 147. (TEA PEIMS, 2025–26)

Two hundred and twenty children. Every one of them attached to an adult who is also, by that same definition, without a stable home. Mostly mothers. In a county with no published street count of its own.

The one door that exists is full

Hood County does have an emergency shelter, and it is a good one: Mission Granbury's Ada Carey Shelter for Women and Children, serving the county since 1998. We will not pretend otherwise in order to make our own case look better.

The problem is not that the door doesn't exist. The problem is what happens when it's full. The Texas Council on Family Violence surveyed survivors in 2025 and found:

Texas Council on Family Violence, State of the State (survivor research, 2025)

Read those three numbers in sequence and you have the whole argument. Half the women who ask for a safe bed in Texas are told there isn't one. A third of them go back to him. A quarter of them sleep in the car.

A full shelter is not a neutral outcome. It is a decision, made by arithmetic, to send a measurable number of women back to the man who hurt them.

What she needs first

That last figure deserves a moment. The story we are told is that homeless people with addictions refuse help. What the research found is a queue of people asking for treatment and not getting any. The shortage is not of willingness. It is of beds.

Where Mercy Manor comes in

Mercy Manor is being built to be the thing Hood County does not have: not another emergency bed for a night, but a home for a year.

The distinction matters more than it sounds. An emergency shelter solves tonight, and it must — that is its job, and Mission Granbury does it. But a woman who is handed back to the street after thirty days, with the same phone and the same debts and the same man's number in it, has not been rescued. She has been paused.

The one-year structure comes directly from what Karen built in Edmonton. A year is long enough to stabilise, to heal, to learn boundaries and budgeting, to practise new habits, and to rebuild an identity — long enough to become a woman for whom the old number is no longer the obvious call.

Mercy Manor's full admissions criteria — who can come, cost, children, detox, medication, and how intake works — are laid out plainly on Who We Can Help.

If we're not the right door

We would rather tell you the truth than take you.

Not sure if Mercy Manor can take you? We have written it down plainly — including whether you can bring your children, what happens about detox, and the two histories we cannot accept: who we can help, and who we can't.

Go deeper

• Understanding homelessness among women
• Shelter vs. transitional housing vs. residential recovery — which one is she actually looking for
• The state of the need in North Texas — the whole picture, sourced
• The cycle of abuse: why she stays
• How to help a woman leave
• Local help in Granbury and Hood County

Common questions