The finding
The program treats a death more harshly than an improvement
When a veteran's condition improves and the caregiver is discharged, they get 60 days of written warning and then 90 days of benefits. One hundred and fifty days to find work, find coverage, and land.
When the veteran dies, they get 90 days, starting the day it happens. No warning is possible. No corrective action exists. And it is the one exit where the person is bereaved, unemployed, and losing their health insurance in the same moment.
The system already knows how to give someone a runway. It gives the longest one to the best circumstance and a shorter one to the worst.
Signature
Every way out, and what each one gives you
Advance notice in light blue. Continued benefits in navy. Total runway is the whole bar.
Detail
The exits
What the 90 days lands on
Report difficulty paying bills
60%
post-9/11 caregivers
Have no rainy-day funds
50%+
over half
Have no health insurance
44%
already, before any discharge
Income under 130% of poverty
36%
while still receiving the stipend
Source: RAND, 2023 Caregiving Survey, published 2024. These are the conditions on day 91.
Build on what exists
The Program of Comprehensive Assistance for Family Caregivers
Created by the Caregivers and Veterans Omnibus Health Services Act of 2010 and expanded to all service eras by the MISSION Act of 2018. It requires a service-connected disability rated 70% or more and a need for in-person personal care for at least six continuous months.
It is a real program and a good one. This framework does not replace it. Everything below already exists and should be the first thing a caregiver is told about, because a great many of them do not know it is there.
Also worth knowing
- There is no work restriction. A caregiver may hold outside employment while receiving the stipend, provided the veteran's care needs are still met. Many believe otherwise and give up more income than they had to.
- The general program has no application. Peer mentoring, skills training, coaching and the support line are open without qualifying for the comprehensive program.
- Legacy participants are protected through September 30, 2028, with no stipend reduction on reassessment and certain reassessments suspended.
What is actually given up
A career does not pause. It ends and starts over lower.
Post-9/11 military caregivers reported missing an average of 3.5 days of work a month, against about one day for civilian caregivers. RAND put lost productivity among post-9/11 caregivers at $5.9 billion in 2011 dollars. One quarter of them are the veteran's parents, which usually means leaving a career already in progress.
The stipend replaces some income. It does not replace a career, and the difference compounds for the rest of the caregiver's life through lost promotions, lost seniority, lost retirement contributions, and lost years of earnings history.
The gap the framework has to name
- The stipend is tax-free, which is generous now and costly later. Because it is not wages, it appears not to generate Social Security credits or retirement contributions — so years spent caregiving may register as years of zero earnings in a benefit calculation decades later. Likely — this needs primary-source confirmation and it is the single most important open item here.
- Unemployment insurance almost certainly does not apply. A caregiver who loses the stipend has not lost a job in the sense state unemployment systems recognise. Open
- The legal and financial planning benefit is scoped to the veteran. It covers the veteran's needs. It does not cover the caregiver's own retirement, credentials, or re-entry plan.
- Nothing in the program is designed to keep the caregiver employable. No credential maintenance, no continuing education, no re-entry support.
What HOLDFAST adds
- Credential maintenance as a benefit. Licence renewals, continuing education units, and certifications kept current while caregiving, so a nurse who steps out for nine years is still a nurse at the end of it.
- Earnings-record protection. The legislative ask: caregiving years should not count as zero-earnings years.
- A re-entry track that starts before the end, not after it — the same logic as beginning veteran transition before separation rather than after.
- Respite treated as maintenance, not indulgence. Thirty days a year exists; 31% say it would be difficult to actually take a break. A benefit nobody can use is not a benefit.
The ask
Ninety days becomes six months — and longer where the service was longer
90 → 180+
This is a change to a VA regulation, not a statute. 38 CFR Part 71 sets the extended-benefit periods, and VA set them administratively. That makes this a rulemaking ask rather than a bill, which is a materially easier fight than the § 5313 amendment.
Six months is the floor, not the whole answer. Ten years out of the workforce and thirty years out are not the same problem, and a flat number treats them identically. The schedule below scales the transition to the service that earned it.
The schedule
Thirty days per year served
A severance-style formula. Floor of six months so nobody gets less than the base ask. Ceiling of twenty-four months so the cost stays bounded and predictable, which is what makes it passable.
At the long end this stops being a re-employment benefit. Someone thirty years in is not going back to a career — they are arriving at retirement age with thirty years of no earnings on their record. For that group the transition payment matters far less than the earnings-record fix, and the framework should say so rather than pretend a longer cheque solves it.
Why this one is winnable
- The precedent is inside the same rule. VA already gives 150 days of total runway for an assessment-based discharge. Nobody has to invent a new principle — the ask is that a death be treated no worse than an improvement.
- The population is small, identifiable, and self-limiting. It is the number of PCAFC veterans who die in a year. There is no growth risk and no incentive effect: nobody stays in caregiving longer, or leaves it sooner, because of a survivor run-on.
- It is cheap. Ninety additional days of stipend and CHAMPVA for that group is a rounding error against a program budget in the billions.
- It has no natural opponent. There is no constituency for cutting off a bereaved caregiver at 90 days. The number is 90 because someone picked a round number, not because anyone defended it.
- Bereavement support already exists and proves the principle. VA already recognises that this moment needs specific handling — REACH VA bereavement counselling is offered to exactly this group. The gap is that grief support was built and income support was not.
Model regulatory change
The smallest version
In 38 CFR Part 71, amend the provision governing the period of extended benefits following discharge due to the death of the eligible veteran by striking "90 days" and inserting: "a period equal to 30 days for each year the individual served as the designated Primary Family Caregiver, but not fewer than 180 days and not more than 730 days", and provide that CHAMPVA coverage for the Primary Family Caregiver continues coextensively with that period.
Two supporting changes worth carrying in the same package: extend the same 180-day period where the caregiver is discharged because the veteran was institutionalised and does not return home, and provide that the caregiver's own death triggers a transition payment to their household rather than nothing at all.
The regulatory citation for the specific paragraph is a counsel item — the durations are confirmed from VA's own published fact sheet and 38 CFR Part 71, but the exact subsection should be pinned before anything is filed.
Who carries it
A caregiver who lost their income 90 days after burying the person they cared for is the most credible witness this change could have. As with the § 5313 fix, the ask should be narrow and named: one number in one rule, and a person who can say what happened on day 91.
Veterans Crisis Line Dial 988, then press 1 Free and confidential, 24 hours. Caregivers and family members can use it too — you do not have to be the veteran.
Honest about the evidence
The rules are solid. The outcome data barely exists.
Every duration on the first panel comes from VA's own published fact sheet and the governing regulation. The burden and hardship figures come from RAND's 2023 survey. Both are firm.
What is missing is any real evidence about what happens to military caregivers after the veteran dies. Income, housing, health coverage, time to re-employment — these are not well documented, and the framework should not pretend otherwise. That absence is itself an argument: nobody has measured what happens on day 91 because nobody has looked.
Building that record should be part of the ask.
Full evidence base