What foreclosure risk and housing loss do to veteran families’ health and stability

Almost every study in this field measures veterans who are already homeless. The families in this record are at an earlier stage: still in the house, behind on the mortgage, facing a sale date. One peer-reviewed study measured that stage directly, and it is the reason this page exists.

DOCUMENTED FACT Bossarte and colleagues, writing in the American Journal of Public Health in 2013, used a probability sample of 1,767 veterans. Their measure of housing instability was not homelessness. It was a single survey question: how often in the past 12 months were you worried or stressed about having enough money to pay your rent or mortgage. S192

The authors state the upstream point themselves, and it is worth reading in their words rather than ours.

More importantly, these results suggest that in addition to realized or chronic homelessness, even financial strain related to housing instability is associated with mental distress and suicidal ideation. That is to say, psychological distress associated with homelessness may be elevated distally at the period preceding actual homelessness.

Bossarte, Blosnich, Piegari, Hill and Kane, American Journal of Public Health, December 2013

Their closing recommendation names the stage precisely: proactive referral may reduce risk associated with "loss or threat of loss of housing." Threat of loss. That is the stage a foreclosure notice creates. S192

PROJECT ANALYSIS The limits of this study have to travel with it, and they are real. It is cross-sectional, so it cannot show which came first. Only 66 of the 1,767 veterans reported the exposure, which is why the confidence interval on the ideation figure runs from 2.6 to 16.1. It is one state, Nebraska, in 2010. The authors themselves call the findings precursory and say point estimates should be read with caution. The honest claim is that veterans under housing-cost strain showed roughly five to six times the odds of distress and ideation in this sample, not that foreclosure causes suicide.

In 2024 RAND published the largest recent study of military and veteran caregivers, funded by the Elizabeth Dole Foundation and drawn from the 2023 RAND Caregiving Survey. It compares caregivers to non-caregivers, which is what makes its numbers usable. S193

DOCUMENTED FACT A separate peer-reviewed study, Delgado and colleagues in PLOS ONE in 2021, surveyed 458 military caregivers and found 23.6 percent reported suicidal ideation since becoming a caregiver. S194

PROJECT ANALYSIS Those two figures look alike and must not be stacked. RAND asks about the past twelve months in a weighted population survey with a comparison group. The PLOS study asks about the whole period since caregiving began, in a self-selected online sample with no comparison group. They are two separate findings about the same population, not two confirmations of one.

On children, the evidence is thinner and this page will not pretend otherwise. RAND reports that around 40 percent of children in military and veteran caregiving households help with caregiving tasks, against about a quarter in civilian caregiving households, and that almost a quarter of caregivers said a caregiving child needed mental health treatment in the past year and did not get it. S193

PROJECT ANALYSIS The child mental-health scores in that report rest on 25 children, and they are reported by the parent rather than assessed by a clinician. That is a small cell and a soft instrument. The household-participation figures rest on a wider base. Treat the direction as real and the precision as weak, and do not put a decimal point on any of it.

RAND also finds that caregivers to those 60 and under report lower relationship quality with their partner, higher family conflict and lower family cohesion than non-caregivers. Housing loss lands on a household that is already carrying this. S193

DOCUMENTED FACT The House Appropriations Committee report accompanying the FY2027 Military Construction and Veterans Affairs bill, H.R. 8469, states as a committee finding that the gap left by ending VASP "has exposed veterans to the risk of foreclosure." The report was filed April 23 2026 and the bill passed the House on May 15 2026. S195

The same report directs VA to report within 90 days of enactment on five things, including "the number of veteran borrowers currently delinquent or at risk of foreclosure" and the timeline for the Partial Claim Program. S195

DOCUMENTED FACT Separately, at page 53, the same committee encourages VA to add to its annual suicide-prevention report metrics including "home loan use and foreclosure assistance and participation in housing and food insecurity programs," in order to show "the correlation between VA benefit usage and the veterans who die by suicide." S195

PROJECT ANALYSIS Congress has asked VA to measure the link between foreclosure and veteran suicide. That request is the clearest official acknowledgement that the link is plausible enough to require data. It is also, as of now, unanswered.

PROJECT ANALYSIS Two things about the status of this language, because they change what it is worth. It is committee report language, not statute. It directs and it does not bind, and it takes effect on enactment. As of this writing the Senate has not produced an FY2027 MilCon-VA bill at all, so the 90-day clock has not started.

PROJECT ANALYSIS This page could have been written to sound conclusive. It is not, and the reason is the finding. Across the published literature reviewed here, the evidence divides into two bodies that do not meet.

One body studies veterans who are already homeless. It is large and it is federal. A 2025 study in the American Journal of Preventive Medicine followed 662,682 veterans with housing-instability indicators in VA records against National Death Index mortality, and found that those exiting housing programs had higher odds of suicide, concluding that "exiting programs to address housing instability may be a vulnerable period for Veterans." S196

The other body studies caregiver strain, and it is the RAND and PLOS work above. It documents a population under severe financial pressure with four times the depression and suicidal ideation of non-caregivers.

DOCUMENTED FACT The 319-page RAND caregiver report contains zero occurrences of the words foreclosure, eviction, homeless, or housing instability. Housing appears in it only as a cost line. The Elizabeth Dole Foundation funded the most thorough study of this population that exists, and mortgage loss is not in it. S193

DOCUMENTED FACT There is no published federal figure for veterans who are housing-unstable but not yet homeless. The federal count is a point-in-time homelessness count: 32,882 veterans in January 2024, 13,851 of them unsheltered. There is no corresponding count of veterans behind on a mortgage and still in the house.

PROJECT ANALYSIS So the chain this record documents, servicer conduct to housing loss to family crisis, is not established end to end by the published evidence, and this page will not claim that it is. What is established: the threat of housing loss is associated with sharply elevated distress and suicidal ideation in veterans; caregivers in these households are already at four times the baseline rate; children in them take on adult responsibilities and go without care; and Congress has formally asked VA to produce the foreclosure-to-suicide data that would close the gap. The missing measurement is not evidence of absence. It is a thing nobody has funded anyone to look at, in a population Congress has now twice asked about.

---

Method note. Sources were reviewed 2026-08-22. Four of the academic publisher pages could not be retrieved directly and their citations and abstracts were taken from Europe PMC and Crossref, which carry publisher-deposited records; full texts were not read, so effect sizes from those papers are recorded in the register as unverified rather than published here as numbers. The Bossarte odds ratios above come from the article's own Table 1. Several of the VA-records studies cited share authors and overlapping cohorts and are not independent confirmations of each other. Where a news article and the underlying report disagreed on a figure, the report was used and the discrepancy recorded in the source register.

There is a rule that decides who qualifies for federal housing help, and for years it treated a veteran's disability compensation as ordinary income. A veteran paid at a high rate because of how badly they were hurt could be counted as earning too much to be helped.

DOCUMENTED FACT S213

This is not an inference. The government wrote it down. In August 2024 HUD set the rule aside for its veteran homelessness program, and said why: "the amount of VA service-connected benefits received due to the severity of their disabilities results in the veteran being over the low-income limit."

DOCUMENTED FACT S213

The instruction it replaced it with is plain. A housing authority "must determine the applicant's annual income for purposes of income eligibility by excluding all VA service-connected benefits received by the HUD-VASH applicant."

DOCUMENTED FACT S213

Read the reason again, because it is the finding. The disqualification got more likely as the injury got worse. A veteran hurt badly enough to be paid at the highest rate was the veteran most likely to be told they earned too much for help staying housed.

PROJECT ANALYSIS S213

What the fix does not cover. HUD changed this for one program, the one that houses homeless veterans. It is a waiver for that program, and it changes eligibility only: the same benefits are still counted afterwards when the rent is worked out. Other federal housing programs were not part of it.

DOCUMENTED FACT S213

The other door closed too

When VA ended Veterans Affairs Servicing Purchase (VASP), families were left to find help somewhere else. In Washington, the place most of them would have gone was the Homeowner Assistance Fund, a Treasury program the state ran for homeowners behind on their mortgage.

That program has ended. The Washington State Housing Finance Commission says so on its own site: "After assisting thousands of homeowners over three years, the HAF program has ended."

DOCUMENTED FACT S248

So a family in Washington who lost the federal route when VASP was withdrawn, and then looked for the state one, found it gone as well. Neither closure was announced to the families relying on them.

PROJECT ANALYSIS S248

One limit on this. The Homeowner Assistance Fund was federal money run state by state. This says what happened in Washington. It does not say whether another state's program is still open, and a family elsewhere should check their own.

PROJECT ANALYSIS S248

If you are in Washington and behind on your mortgage now. The state Homeownership Hotline is 877-894-4663. It is free, it is confidential, and a housing counsellor will talk through the options that are left. The Commission points there itself now that the fund has closed.

DOCUMENTED FACT S248

What this page does not yet say, and why

Families in this record have said they were turned down for mortgage help from the Homeowner Assistance Fund for this same reason, that their VA disability compensation was counted as income. That is not published here yet, because it is not yet sourced to a document.

The reason it is harder to pin down than it looks: the Homeowner Assistance Fund does not use one national income rule. The Treasury guidance leaves the method to each state program under a ceiling of roughly 150 percent of area median income, so whether disability compensation counted depends on which state a family applied in. A national sentence would be wrong even where the individual experience is real.

What would settle it: the written denial itself, or a state program's own published income definition. Until one of those is in hand, this section states the rule HUD wrote down and stops there.

UNVERIFIED · HELD

In November 2022, fourteen unhoused veterans with severe disabilities sued the Secretary of Veterans Affairs over housing at the West Los Angeles VA campus. They were represented by Public Counsel and sued under the Rehabilitation Act. After a bench trial, Judge David O. Carter ordered the Department to build permanent and temporary supportive housing on the campus.

DOCUMENTED FACT S244

One part of that ruling is the reason it appears on this page. The court found the Department's housing policy discriminated against disabled veterans by allowing income limits that counted disability payments as income, which kept the veterans in the greatest need from qualifying.

DOCUMENTED FACT S244

That is the same mechanism this page describes: a payment made because a veteran is disabled, counted as though it were earnings, with the result that being more disabled makes a family less eligible for help.

What this does not do, and the distinction matters. This is not a foreclosure case and it does not decide anything about mortgage assistance. It concerns housing on one VA campus. It does not settle the Homeowner Assistance Fund question held open above, which still needs a written denial or a state program's own income definition. What it establishes is narrower and still worth having: a federal court has looked at this way of counting and called it discrimination.

PROJECT ANALYSIS S244

The Department's own appeal history in that case has not been read for this page. Anyone citing it in a filing should read the opinion first.

READ THE OPINION BEFORE FILING S244

There is a version of this argument that says helping families costs money and letting them fall does not. The Department's own research says the opposite.

PROJECT ANALYSIS

VA's research arm published a study in January 2026 covering 229,096 veterans across 693,383 patient-trials. Veterans enrolled in the Department's supportive housing services had better health outcomes, lower hospital costs, and an estimated 14 percent lower risk of dying.

DOCUMENTED FACT S218

That is not an advocacy claim. It is the Department measuring its own program and finding that housing people keeps them alive.

PROJECT ANALYSIS S218

The other direction is measured too. Research for the Council on Criminal Justice found that veterans with a history of homelessness were 7.8 times more likely to attempt suicide than veterans who had never experienced it.

DOCUMENTED FACT S219

That figure has to be read for exactly what it measures, and this page will not stretch it. It is about homelessness. It says nothing about foreclosure, and foreclosure is not homelessness. A family that loses a house and moves in with relatives has not been counted in it. What it establishes is the severity of the far end of the road, not the distance a particular family has travelled down it.

PROJECT ANALYSIS S219

Put the two together and the shape of the spending is the finding. There is a budget for crisis lines, for emergency mental health care, for housing people who have already lost their homes. There is a documented saving from stopping the loss earlier. The money is being spent at the end of the sequence rather than the start of it.

PROJECT ANALYSIS S218 S219

Which is what makes the rules described above expensive rather than careful. A family turned down because it recovered, or because a disability payment counted as income, does not stop needing help. It arrives later, further along, and costs more.

PROJECT ANALYSIS

Why these two sources and not a hearing quote

This argument is made elsewhere in this project's own earlier material using testimony from a hearing. That version is not used here. Hearing testimony transcribed from video without verification sits in a lower reliability tier in this record, and an argument this load-bearing should not rest on the weakest available source when better ones exist.

Both sources above are published research: one from the Department's own research service, one from a criminal justice commission. Both were read in full before being cited. Neither is about the home loan program. They are about what happens to a household after the housing is gone, which is the part of the cost that never appears in a foreclosure ledger.

PROJECT ANALYSIS

The approach behind the research described above has a name. Housing First means putting someone into stable housing before requiring anything else of them, on the reasoning that almost nothing else can be treated while a person has nowhere to live. It has been the framework behind federal veteran homelessness work for years.

PROJECT ANALYSIS

In January 2026 VA's own research service reported that veterans in its supportive housing program had about a 14 percent lower risk of dying.

DOCUMENTED FACT S218

On June 1, 2026, HUD announced it was moving its main homelessness program away from that approach, redirecting $4.04 billion for the year. Secretary Scott Turner's words in the announcement: the approach "promised to end homelessness. Instead, billions of taxpayer dollars were spent while homelessness increased to record levels."

DOCUMENTED FACT S220

And the reasoning, quoted in full rather than summarised, because it is a real argument and deserves to be read as one: "The 'housing first' experiment failed Americans by warehousing the vulnerable without results", and "Housing alone will not solve a crisis driven by addiction and mental illness."

DOCUMENTED FACT S220

That second sentence is a factual claim, and it has been tested on exactly the people it describes. A 2025 study using VA's own records followed nearly 300,000 unhoused veterans with mental illness for three years, comparing the Department's permanent-housing program against its program that requires treatment first.

DOCUMENTED FACT S221

Its finding: "Housing First reduces three-year mortality by 4.6 percentage points relative to a no-program counterfactual. In contrast, Treatment First has no long-term effects on health." The author reports being able to "statistically rule out that Treatment First is as effective as Housing First at reducing mortality." Housing First was also the cheaper of the two, because people used fewer long inpatient stays.

DOCUMENTED FACT S221

The population in that study is not a general one. It is unhoused veterans with mental illness, in VA's own programs. That is the exact group named in the reason given for ending the approach, and on that group the evidence runs the other way.

PROJECT ANALYSIS S220 S221

One thing this page will not do with that study. The other claim in the announcement, that spending rose while homelessness rose, is a question about national totals over time. This research does not measure that, and it is not used here as though it did. A record that stretches its best evidence onto a question that evidence cannot answer has stopped being a record.

PROJECT ANALYSIS

What this record can say is narrower and harder to answer. One federal department measured the approach on its own veterans and published a mortality finding in January. Another federal department ended its funding for the approach in June. Five months apart, and the second announcement does not mention veterans at all.

PROJECT ANALYSIS S218 S220

For a veteran family losing a home, that is the ground the safety net is standing on while the mortgage side of it is also being decided by rules written for someone else.

PROJECT ANALYSIS

One thing this page is not yet stating, and why

There is a report that a federal judge vacated HUD's 2026 funding notice on August 7, 2026, on the grounds that the change required public notice and comment and did not get it. That is not published as fact here, because the court's order has not been read. It is known to this project from a single secondary news report.

A court ruling is exactly the kind of claim that should rest on the order itself, not on a summary of it. What would settle it: the docket entry and the written opinion. Until one of those is in hand, this sits here rather than on the page.

UNVERIFIED · HELD

This section describes Washington's Homeowner Assistance Fund eligibility rule as published in 2024; it is not a description of a currently open Washington program. The source summarized the program as available until funds ran out or June 30, 2024. Under that rule, two tests were measured at different moments: the hardship had to have already happened, while the income test looked at the household's current income at application.

DOCUMENTED FACT S215

The published Washington rule shows both tenses in the same breath. The hardship test looks back: a homeowner "must have suffered a financial hardship" after January 21, 2020. The income test looks at the present: household income "must be 100% or less of the area median income."

DOCUMENTED FACT S215

Now put that next to what a payment pause is for. Forbearance buys a household time. Time to find work, to retrain, to get a second income into the house. That is the point of it. That is what a family is supposed to do with the months it is given.

PROJECT ANALYSIS

A family that uses the time well raises its current income. Raising current income is what fails the income test. The debt from the original hardship is still there, unchanged, still owed. But the household is now measured on the recovery rather than on the harm.

PROJECT ANALYSIS S215

So the two rules point in opposite directions. One says the harm must be real and in the past. The other says you must still be in the condition the harm put you in. A household that stayed down qualifies. A household that climbed does not, and still owes every missed payment from the months it was down.

PROJECT ANALYSIS

Set this beside the income rule described above, and the two compound. A veteran's disability compensation counts toward the current-income test. So does any income the household managed to add while the payments were paused. Neither of those is money that clears the arrears, and either can carry a household over the line.

PROJECT ANALYSIS S217 S215

This project has a sentence it has used since the beginning, and this is the mechanism underneath it. The event that harmed you is the gate that excludes you.

PROJECT ANALYSIS

What this section is careful about

The two tenses are quoted from the published rule. What follows from them is this project's reading, and it is labelled that way on every line. No program document says "we intend to exclude families who recover." The exclusion is a consequence of how the two tests are timed, not a stated aim, and this page does not claim otherwise.

Nor does this section say how any particular household's income was worked out. Families in this record describe exactly this sequence: a pause used to get back on their feet, and then a refusal measured against the income they had rebuilt. Those are their accounts. The rule quoted above is the documented part.

UNVERIFIED · HELD

The federal rule that decides what counts as a household's income for housing help is written backwards from what most people expect. It does not list the income that counts. It says everything counts except the things on one short list of exceptions.

DOCUMENTED FACT S217

The rule's own words: annual income means "all amounts, not specifically excluded" from the exceptions list.

DOCUMENTED FACT S217

Two kinds of veterans' benefit are on the exceptions list, and neither is the ordinary monthly one. The list leaves out disability benefits that arrive "in a lump sum amount or in prospective monthly amounts", which is back pay, and it leaves out payments for veterans who need regular aid and attendance.

DOCUMENTED FACT S217

The regular monthly compensation a disabled veteran lives on is not on that list at all. So it counts. Not by anyone's decision in a particular case, and not by a mistake at a call centre. It counts because the rule is built to count everything it does not name, and it does not name this.

PROJECT ANALYSIS S217

HUD confirmed this from the other direction, by having to undo it. When it set the rule aside for its veteran homelessness program, as set out above, it explained that the compensation was pushing badly injured veterans over the limit. You do not waive a rule that was not applying.

PROJECT ANALYSIS S213 S217

This is a federal definition, not a state one. It is the definition that programs measuring against local median income are built on. That is why a family's experience of this is not particular to one state, and why the fix HUD made for one program did not reach the others.

PROJECT ANALYSIS S217

What this does not settle

This section states what the federal definition does. It does not say which definition any particular state program adopted. States running mortgage relief were allowed to set their own method, and the Washington program never published which one it used, as set out above.

So the honest position is this. The federal rule counts the compensation. A state program measuring against local median income is very likely using that rule, because it is the definition those limits are written for. Very likely is not the same as documented, and this record does not publish the two as if they were.

UNVERIFIED · HELD

The federal mortgage relief that homeowners could apply for during the pandemic was not one program. Congress funded it and the Treasury set the outer limits, and then each state wrote its own rules inside those limits. That design decision is why two families in the same situation, in two different states, could get two different answers.

DOCUMENTED FACT S214

Treasury's ceiling was about 150 percent of the local median income. Washington set its limit at 100 percent. A household earning more than the middle income for its own county was over the line.

DOCUMENTED FACT S215

Now the part that is harder to write about, because there is nothing to quote. The program's own published questions and answers never say how income is worked out. It says eligibility rests on "a variety of factors", assessed "on a case-by-case basis", and tells the reader to call a hotline.

DOCUMENTED FACT S214

So the limit was public and the rule was not. A homeowner could look up the income figure for their county. They could not read the definition that decided whether they were above it or below it. A rule you cannot read is a rule you cannot check, and it is a rule you cannot appeal on its own terms.

PROJECT ANALYSIS S214

Read this next to what HUD wrote down about disability compensation, set out above. HUD said in the Federal Register that counting service-connected compensation pushes badly injured veterans over an income limit, and it removed that effect from one program it runs. A state program with a limit set at two thirds of what Treasury allowed, and no published rule about what counts, is where that same arithmetic would land hardest and be least visible.

PROJECT ANALYSIS S213 S215

What families report, and what this page will not state as fact

Families in this record report being told that every source of household income was counted, disability compensation included, because that is the standard the state applies across its programs. That is what people were told. It is not published anywhere this project has been able to find, and so it is recorded here as reported rather than as a documented rule.

What was searched on August 24, 2026, and what it returned: the program FAQ, read in full, no income definition. The commission's live program page, which returned an error and then a page that will not load without scripting. The successor site the closed program now points visitors to, which carries no income definition either. The Treasury guidance document itself, which this project does not fetch because the site's own rules disallow it.

What would settle it: a written program guideline listing counted income, or an application form showing the income questions. Until one of those exists in this record, the sentence stays in this box and off the page.

UNVERIFIED · HELD