Christina Robinson sat on a bench outside the county library at 7:30 AM, waiting for the doors to open at 9:00. She’d walked four miles from the shelter because the bus didn’t run early enough. The library had computers, and somewhere in her tote bag was the notice about her Medicaid work requirements. She needed to report her work hours by tomorrow or risk losing coverage.
Christina didn’t have work hours in the traditional sense. She worked day labor when she could, when her chronic pain wasn’t too severe, when she could get to the pickup location by 6 AM. Sometimes three days a week. Sometimes none. The work was cash, handed to her at shift end. No paystubs. No verification. No record she existed in any administrative system.
Her phone had been stolen six weeks ago, the third time in eighteen months. She couldn’t afford to replace it. She’d missed calls from the eligibility worker, missed text reminders, missed automated alerts. The notice said to log into the online portal, but without a phone she couldn’t receive the two-factor authentication code. It said to call a helpline, but she had no phone. It said to visit a county office fifteen miles away, requiring two bus transfers she couldn’t afford, closing at 4:30 PM before she could arrive if she took a day labor job that morning.
Christina had been experiencing homelessness for fourteen months, since fleeing her apartment after the violence from her boyfriend escalated beyond what she could endure. She’d couch surfed until hospitality expired. The shelters had rules about how long you could stay, and most didn’t allow mail in residents’ names. She’d worked consistently before homelessness began. But living without stable housing destroys everything. The chronic pain worsened. The diabetes went uncontrolled without regular meals and medication storage. The administrative capacity required to navigate work requirements simply didn’t exist when her entire life focused on surviving each day.
The library opened. Christina tried to log into the portal but couldn’t remember her password, and the reset function required a phone number she no longer had. She had the notice but no way to prove she’d worked at all. No way to document the three days of day labor she’d managed. Tomorrow the deadline would pass. Her coverage would terminate. And she’d lose access to the medication that kept her diabetes from spiraling into crisis.
Demographics and Scope#
People experiencing homelessness face work requirements designed for housed stability while navigating the chaos that housing instability creates.
Approximately 2-3% of expansion adults experience homelessness during any given year, representing 370,000 to 550,000 people subject to work requirements. The January 2024 point-in-time count found 771,480 people experiencing homelessness on a single night, an 18% increase from 2023. Chronic homelessness affects 152,585 people, roughly 20% of the homeless population. Episodic homelessness, cycling between housing and homelessness, affects approximately 45-50%. Another 15-20% have recently lost housing for the first time within the past six months. The split between sheltered and unsheltered varies by region but nationwide approximately 36% sleep outside while 64% access shelters or transitional housing.
Men represent about 60% of the homeless population, women 40%. Veterans number nearly 33,000, representing 4.3% of homeless individuals. Racial disparities are stark: Black Americans constitute 32% of the homeless population despite being only 12% of the general population. Native Americans are similarly overrepresented. Hispanic and Latino individuals represent 30%, a figure that increased 32% from 2023 to 2024. The average age among chronically homeless populations is 50, though this population exhibits accelerated aging where a 50-year-old homeless adult has the health profile of a housed 65-year-old. Family homelessness increased 39% from 2023 to 2024, with approximately 33% of homeless individuals being part of families with children.
The health status of people experiencing homelessness explains why Medicaid access matters urgently and why coverage loss carries severe consequences. Between 45-50% have serious mental illness. Between 50-60% have substance use disorders, with alcohol affecting 38% and drugs affecting 26%. Co-occurring mental illness and substance use disorders affect 40-45%. Chronic physical health conditions affect 75-85%, with over one-third reporting difficulty with activities of daily living. The mortality consequences are devastating: average life expectancy for chronically homeless individuals ranges from 42 to 52 years, compared to 78 for housed populations. Homelessness is lethal.
The health profile reflects both causes and consequences of housing loss. Many became homeless partly because mental illness or substance use disorders disrupted employment and relationships. Once homeless, conditions worsen dramatically. Chronic diseases go unmanaged without consistent access to care, medication storage, or regular meals. Injuries from violence, falls, and exposure accumulate. Dental problems become severe infections. Foot problems from constant walking create chronic pain. The street ages people rapidly: a 50-year-old who has been homeless for five years has the physiological profile of a housed 65-year-old.
Tuberculosis, hepatitis, and HIV rates run higher in homeless populations due to crowded shelter conditions, shared needles, and barriers to treatment. Skin infections become chronic. Respiratory conditions worsen without shelter from weather. The cascade of health problems makes employment increasingly difficult, which makes housing increasingly unattainable, which worsens health further. The spiral operates in one direction without intervention.
Why Work Requirement Systems Fail Homeless Populations#
People experiencing homelessness cannot meet work requirements designed for housed populations because every system assumption proves false.
The address assumption collapses immediately. Work requirement notices require mailing addresses. Shelters often prohibit using their address for official mail. Post office boxes require fees homeless individuals cannot afford and identification many have lost. General delivery exists but requires awareness of the option and proximity to post offices. Notices mailed to last known addresses never reach recipients who moved months ago. The verification deadline arrives, the person never receives notice, coverage terminates. The system interprets non-response as noncompliance when the actual problem was communication impossibility.
The communication infrastructure failure compounds address problems. Work requirement systems assume phone access for reminders, two-factor authentication, and eligibility worker contact. Homeless individuals lose phones to theft, damage, or inability to maintain service. A stolen phone means losing all saved numbers, missing all calls, and inability to access portals requiring SMS codes. Library computers provide limited access but require remembering passwords set months ago. The eligibility worker calls, gets no answer, sends texts reaching no one, and documents failed contact that looks like disengagement when the actual problem is infrastructure absence.
The documentation deadlock creates impossible verification requirements. Day labor employers pay cash at shift end with no paystubs. Gig economy work requires smartphones homeless people don’t maintain. Under-the-table employment provides no verification trail. Community service at meal programs or shelters generates no official documentation. The homeless person who works three days has nothing to submit as proof. The system demands documentation their employment pattern cannot produce, then terminates coverage for failure to verify work that actually occurred.
The survival bandwidth collapse explains cognitive capacity limitations. Housed people underestimate the cognitive load homelessness imposes. Where will I sleep tonight. Where can I charge my phone. Where will meals come from. How do I stay safe. These questions consume mental capacity constantly. Adding work requirement deadlines, portal navigation, document maintenance, and appointment schedules exceeds available bandwidth. The housed person juggles administrative tasks alongside stable housing. The homeless person juggles them alongside survival and loses.
The episodic capacity mismatch reflects how homelessness affects work capacity inconsistently. Someone works steadily for three weeks, then chronic health conditions flare and work becomes impossible for two weeks, then capacity returns. Monthly requirements of 80 hours don’t accommodate this pattern. Systems designed for consistent monthly activity fail people whose capacity fluctuates with health status, shelter availability, weather exposure, and crisis events. The person who works 100 hours in good months and zero in crisis months averages adequate activity across time but fails monthly verification repeatedly.
The systems fail not because people experiencing homelessness refuse to engage but because engagement requires infrastructure homelessness destroys. Stable address. Reliable phone. Document storage. Cognitive bandwidth beyond survival. Consistent capacity across monthly cycles. Housed populations possess these prerequisites automatically. Homeless populations lack them entirely.
Exemption Frameworks That Acknowledge Reality#
Accommodating homeless populations requires fundamentally different verification approaches that work within street-world constraints rather than demanding housed-world capacities.
Automatic exemption triggered by homelessness status represents the most effective approach. States with integrated HMIS systems can identify Medicaid members enrolled in homeless services and process exemptions without requiring any member action. Someone entering a shelter generates HMIS enrollment. That enrollment triggers automatic Medicaid exemption through data matching. The member never navigates administrative processes because the processes happen automatically based on service engagement.
Trusted intermediary verification authorizes shelter case managers, street outreach workers, and Continuum of Care staff to submit exemptions and work verification on members’ behalf. The outreach worker who finds someone under a bridge documents homelessness and initiates exemption during the encounter. The shelter case manager submits monthly work hours for residents engaged in shelter work programs. Navigation happens through existing service relationships rather than requiring separate administrative engagement.
Simplified attestation allowing member self-report with random audit balances fraud concerns against documentation impossibility. The member states they worked day labor on specific days for approximately eight hours each. That attestation suffices for verification. States conduct random audits contacting employers or observing work sites for pattern analysis rather than demanding universal documentation homeless employment cannot produce.
Alternative contact methods acknowledge phone and address limitations. States establish shelter-based voicemail systems where automated calls route to kiosks that shelter staff monitor. Text messages route to library terminals where staff alert members about deadlines. Email communications include plain language and direct action links rather than requiring portal login with credentials members cannot maintain. Extended response windows of 30 days rather than 10 accommodate communication barriers.
Post-housing grace periods recognize that employment capacity develops gradually after housing stabilizes. The member who exits homelessness into housing receives automatic six-month exemption extension while establishing routines. During months five and six, care managers contact housing providers to coordinate support. At month seven, graduated requirements activate at 40 hours monthly rather than 80, increasing to full requirements at month twelve.
Medical exemptions during homelessness should presumptively apply based on housing status alone given health condition prevalence. States can establish policy that homelessness constitutes medical exemption category without requiring separate health documentation. Research shows 75-85% of homeless individuals have chronic health conditions, 45-50% have serious mental illness, and 50-60% have substance use disorders. Presuming health barriers exist makes statistical and practical sense.
Coverage suspension rather than termination keeps the person enrolled but pauses benefits, restorable without full reapplication when documentation arrives. For homeless populations, termination creates months-long gaps suspension would avoid.
The Accountability Question#
Automatic exemptions generate legitimate objections deserving engagement rather than dismissal.
The perverse incentive concern asks whether exemptions might discourage housing. This dramatically underestimates how destructive homelessness is. Average life expectancy for chronically homeless people ranges from 42 to 52 years compared to 78 for housed populations. Violence risk is constant. Weather exposure kills. Nobody chooses chronic homelessness to avoid 80 hours of monthly work requirements.
The structure argument suggests work requirements provide beneficial structure. This has merit but timing matters. Imposing requirements during active homelessness creates barriers rather than structure. The person who can’t receive mail, can’t keep a phone, can’t store documents, and can’t maintain focus beyond immediate survival cannot benefit from work requirement structure. Better to provide structure supporting housing stability first, then graduated work expectations after housing stabilizes. Housing provides the foundation. Work requirements can provide structure once foundation exists. Demanding structure before foundation exists guarantees collapse.
The Housing First research supports this sequencing. Programs providing housing without preconditions then layering in support show better outcomes than programs requiring compliance before housing. Applying this insight to work requirements suggests coverage should be unconditional during active homelessness, with requirements activating only after housing stabilizes.
The fraud concern notes simplified verification invites gaming. The policy question is whether fraud prevention catching 2-3% of fraudulent claims justifies requirements that 30-40% of honest homeless people cannot complete. Focus fraud detection on pattern analysis identifying outliers rather than universal documentation requirements failing most honest claimants.
The fiscal calculation changes when emergency care enters. Someone losing coverage doesn’t stop having health needs. They present to emergency departments, receive care hospitals can’t refuse under EMTALA, and generate uncompensated care costs. Emergency treatment for diabetic ketoacidosis costs tens of thousands. The person cycling between coverage loss and emergency care generates higher costs than the person maintaining coverage through accommodated exemptions. The question isn’t whether we pay for homeless healthcare. The question is whether we pay for managed care or emergency care.
The Housing Policy Intersection#
Work requirement implementation cannot be divorced from housing policy context. The President’s FY2026 budget proposed 43-51% HUD funding reductions, converting vouchers to state block grants with significantly reduced funding. Emergency Housing Vouchers faced elimination. Two-year time limits would apply to rental assistance.
Congress rejected these proposals in July 2025 reconciliation. The final One Big Beautiful Bill Act preserved existing HUD program structures, increased Low-Income Housing Tax Credit allocations, and maintained voucher funding levels. The rejected proposals did not become law.
But funding uncertainty triggered practical effects that continue. Housing authorities reduced voucher payment standards anticipating cuts, making vouchers insufficient in many markets. New voucher issuance slowed. Affordable housing developers paused projects lacking certainty about long-term commitments.
These effects matter for work requirements because housing stability is prerequisite for employment capacity. The person exiting homelessness into voucher-supported housing develops routines and builds employment capacity. Voucher payment standards not covering market rents force impossible choices. Slowed issuance extends homelessness duration. Stalled development reduces supply.
The policy interaction creates problematic scenarios. Homeless individuals subject to work requirements but unable to access housing due to voucher scarcity face verification requirements while still homeless. Individuals exiting homelessness into inadequate vouchers may lack stable foundation for employment. States implementing work requirements in December 2026 must design systems accommodating potentially larger homeless populations if housing policy shifts.
Prudent state planning acknowledges uncertainty through multiple mechanisms. Automatic exemption for homelessness should not include time limits. Post-housing grace periods should extend to twelve months. Care coordination should include housing navigation as core service. Work requirements cannot substitute for housing as foundation for stability.
Stakeholder Roles in Supporting Homeless Populations#
Multiple institutions must coordinate to make work requirements navigable.
Managed care organizations must adapt for populations lacking standard infrastructure. Specialized homeless care management abandons assumptions about phone contact, mailed assessments, and scheduled appointments. Care managers coordinate with street medicine teams, communicate through shelter case managers, and conduct assessments during emergency visits. Proactive exemption identification through claims analysis flags likely homelessness: frequent emergency use, Healthcare for the Homeless visits, multiple address changes, unstable phone numbers. Monthly HMIS data sharing provides automatic exemption triggers without member action.
Panel sizes designed for housed populations fail when care managers must coordinate through intermediaries. A care manager serving 300 housed members might effectively serve only 50 homeless members. MCOs should establish formal relationships with Continuum of Care organizations, authorizing outreach workers to submit exemptions on members’ behalf. Shelter staff receive portal access for direct submission. Street outreach teams use mobile apps interfacing with MCO systems.
Alternative contact infrastructure must support irregular communication patterns. MCOs establish shelter-based voicemail boxes for automated calls. Extended response windows of 30 days acknowledge communication limitations. When HMIS data indicates a member has entered housing, MCOs initiate transition monitoring with five-month communications about upcoming activation.
Providers serving homeless populations need understanding that patients face work requirement deadlines that may disappear from awareness between visits. Healthcare for the Homeless clinics serve as navigation hubs where clinical encounters address health and coverage simultaneously. Street medicine teams can document homelessness and initiate exemptions during outreach. Providers should understand no-shows reflect street life chaos, and documentation they provide may be the only verification patients can produce.
Emergency departments represent critical intervention points because homeless populations use emergency care at rates far exceeding housed populations. Emergency department social workers already conduct discharge planning. Adding work requirement screening catches people who will never connect with primary care. Medical respite programs providing recuperative care after discharge create extended windows for navigation and connection.
Community-based organizations already serving homeless populations become essential partners. Continuum of Care providers using HMIS can add exemption coordination to workflows. Day labor centers tracking workers and hours support simplified verification without requiring documentation workers cannot preserve. Faith communities operating meal programs see the same individuals repeatedly, creating trust government systems cannot replicate.
Street outreach workers represent the most effective navigation resource because they go where homeless people are. The outreach worker finding someone sleeping under a bridge can explain requirements, document homelessness, and initiate exemptions during the same encounter. Navigation happens where homeless people live rather than expecting them to navigate into systems designed for housed populations.
Employers hiring homeless individuals bear responsibility for verification accommodation. Day labor employers should issue simplified work confirmation letters upon request. Some can establish direct verification relationships with MCOs where employers provide monthly lists of workers and hours, MCOs cross-reference with member lists, and verification completes automatically.
Back to Christina#
Christina eventually gave up at the library computer. She couldn’t access the portal, couldn’t remember passwords, couldn’t verify work she’d actually done because cash day labor produces no documentation. Her coverage terminated thirty days later. Six weeks after that, she was hospitalized for diabetic ketoacidosis because without insurance she’d stopped taking insulin she couldn’t afford. The hospital bill was $47,000. She was discharged back to the shelter, weaker than before, with the same verification notice waiting.
Her experience is typical. Verification failure, coverage termination, hospitalization, discharge back to homelessness. This pattern will repeat for hundreds of thousands of homeless expansion adults if states implement work requirements without homeless-appropriate infrastructure.
The question is not whether Christina tried hard enough. She walked four miles to a library to try to comply. The question is whether infrastructure designed for housed stability can function for people experiencing homelessness. Standard verification assumes stable addresses, continuous phone access, document storage, and cognitive bandwidth for deadlines while managing survival. For homeless populations, every assumption is wrong.
The policy question is not whether homeless people can work. Many do work, episodically and informally. The question is whether systems will accommodate how homeless people actually work and live, or whether they will demand housed-world verification from street-world populations and interpret inevitable failure as individual noncompliance.
States face a fundamental choice. Design verification systems that work for homeless populations through automatic exemptions, trusted intermediaries, simplified attestation, and graduated requirements. Or apply standard processes and accept that thousands will lose coverage, generate emergency costs, experience health deterioration, and potentially die from conditions that Medicaid would have managed. The choice determines not whether states pay for homeless healthcare, but whether they pay through managed coverage or emergency crisis response.