Paying for homelessness in hospital beds

Hospital admission data show thousands of inpatient bed days are caused by homelessness. This Is costly, harmful and unsustainable.  

–

Australian hospitals are increasingly managing homelessness as if it were a clinical condition. They are absorbing the consequences of systemic housing failure with no authority to fix it, no control over the upstream drivers, and no sustainable way to discharge people safely. 

New national hospital data show that in 2023-24 more than 1,000 hospital separations recorded homelessness or housing insecurity as the primary diagnosis. These admissions accounted for over 8,000 inpatient bed days and at least $18-19 million in direct hospital costs in a single year. These episodes were not driven by medical illness, but by the absence of somewhere safe to go. 

This is not an anomaly or a coding quirk. It is the health system functioning as a de facto housing provider of last resort. 

Hospitals are shouldering the operational and fiscal burden of housing system failure through avoidable admissions, delayed discharge, and blocked acute beds. Emergency departments have become gateways to nowhere, while discharge planning repeatedly collapses against vacancy rates that simply do not exist. 

Continuing to fund homelessness through hospital beds is not just inefficient; it is an absence of effective housing and health policy. And one Australia can no longer afford. 

Homelessness as a principal diagnosis 

Each year, the Australian Institute of Health and Welfare collects hospital separation data using the International Classification of Diseases. Within this framework, ICD‑10-AM code Z59 identifies problems related to housing and economic circumstances. Crucially, Z59 is only recorded when housing insecurity or homelessness is assessed as the primary reason for the episode of care.  (1) It is important to note that those awaiting aged care placement or in home support are not included in this data set. Z59 is about homelessness, purely and explicitly. 

In 2023-24, there were 1,052 hospital separations nationally where Z59 was recorded as the principal diagnosis. These admissions accounted for 8,152 patient days, with an average length of stay of more than six days. (2) 

Using conservative national benchmarks for public hospital costs, Z59 admissions cost close to $19 million in a single year. (3) This figure captures admitted patient care only. It does not include emergency department presentations, ambulance arrivals, outpatient services, or the costs associated with repeat presentations.

The Z59 data expose a bifurcated system. Some people are discharged on the same day, formally recognised as homeless yet sent back into instability with no recorded resolution. Others remain in hospital for extended periods, occupying acute beds long after they are medically ready, simply because there is nowhere for them to go. 

It is imperative to recognise these are people who often represent the most marginalised and vulnerable of Australia’s population including people sleeping rough, victims of domestic and family violence, and young people exiting state funded care. 

These figures represent the floor, not the ceiling, of homelessness related hospital use. 

Homelessness is rarely coded as the principal diagnosis. More commonly, people experiencing housing insecurity are admitted for infections, injuries, mental health crises, substance related harm, or exacerbations of chronic disease. In these cases, homelessness is a causal factor but not the coded reason for admission. As a result, the vast majority of homelessness driven health system use remains effectively invisible in routine data. (4) 

Emergency Departments: the front door of system failure 

The entrapment of hospitals begins well before admission. 

National emergency department data show that in 2023-24 more than 200,000 presentations were coded under factors influencing health status and contact with health services, a classification that includes housing and social crises. Fewer than 14 per cent of these presentations resulted in admission, meaning the overwhelming majority were managed temporarily in emergency departments and then discharged. (5) 

For people experiencing homelessness, emergency departments are often the only accessible point of care. Primary care is difficult to access without a fixed address. Outpatient appointments are frequently missed due to lack of transport, contactability, or stability. Medication regimes are difficult to maintain without safe storage. None of this is well captured in emergency department datasets. 

What is visible are repeated presentations, prolonged stays, and discharge without resolution of the underlying problem. (6) 

Emergency departments have become the default intake point for housing system breakdown, without the authority, resources, or downstream pathways to resolve it. The result is not continuity of care, but recycling of risk, compounding both harm and cost. 

A data disconnect nobody owns 

One of the most striking features of this problem is the complete absence of integrated data between hospitals and the housing system. 

Hospitals can identify housing insecurity, code it in clinical records, and quantify its impact on bed utilisation. What they cannot see or consistently record is what happens next. There is no routine visibility of referral pathways, response times, interim supports, or housing outcomes. 

This gap is particularly evident in same day Z59 discharges. More than half of people with a single Z59 admission were discharged on the same day, including the majority of those formally coded as homeless. Housing risk is recognised, documented, and then disappears from view. The dataset does not capture whether housing services were contacted, whether support was available, or whether discharge occurred in the absence of any viable housing pathway. 

Housing data tell a disconnected story. National Specialist Homelessness Services data show that fewer than 2 per cent of clients are formally referred from hospitals. (7) On paper, this suggests hospitals play only a minor role in identifying people at risk of homelessness. 

Yet that picture does not align with hospital morbidity, discharge, or emergency data, which consistently show large numbers of patients experiencing housing instability at the point of admission or discharge. (8)

The disconnect deepens when looking at where people were coming from immediately before seeking homelessness assistance. Thousands of clients each year report leaving non‑psychiatric hospitals, psychiatric units, or rehabilitation services before presenting to homelessness services. (9) These individuals clearly passed through health settings, but the systems that touched them do not talk to each other. 

Because hospital and housing datasets are not linked, it is impossible to determine whether homelessness pre‑dated hospitalisation or was created, intensified, or left unresolved at discharge. In data terms, homelessness following hospital care is indistinguishable from homelessness that existed beforehand. 

Long hospital stays appear as inefficiencies. Same day discharges appear efficient. The system cannot see or account for whether housing need was resolved or simply displaced. 

Paying for homelessness the hard way 

Australia spends roughly $107 billion a year on hospitals, with operating costs approaching $300 million per day. (10) Against that backdrop, $18-19 million in coded homelessness admissions may appear marginal. 

But these are among the most avoidable costs in the entire system. 

The Z59 figure captures only a small, visible slice of the problem. It excludes emergency department costs, ambulance transport, repeated short‑stay admissions, mental health presentations, and the cumulative impact of unsafe discharge. Each represents expenditure driven not by medical complexity, but by unmet housing need. 

Evidence consistently shows that supported housing costs less than prolonged hospital care, particularly once repeat admissions and emergency presentations are considered (11, 12) 

From a fiscal perspective, homelessness is already being funded. It is just being funded inefficiently, reactively, and through the most expensive part of government. 

Siloed solutions inside hospitals 

Hospitals have not been idle in the face of rising homelessness. Across Australia, health services have developed local responses: specialist homelessness teams, dedicated discharge coordinators, expanded emergency department social work, GP clinics embedded within hospitals, and outreach partnerships with community organisations. (13, 14) 

These initiatives reflect ingenuity, ethical commitment, and frontline necessity. In many locations, they demonstrably reduce harm. (15)  Taken together, however, they reveal a deeper systemic problem. 

Australia’s hospital response to homelessness is fragmented, localised, and structurally siloed. There is no national framework governing how hospitals should identify, manage, or be supported to address homelessness. Instead, responses depend on local leadership, time limited funding, philanthropic backing, or the determination of individual clinicians. (16, 17) 

Many hospital-based homelessness programs operate as peripheral projects rather than core system functions. They rely on short funding cycles, bespoke local partnerships, and informal referral pathways. This makes them difficult to sustain, hard to replicate, and vulnerable to leadership or budget changes. (18)  

Ironically, their success can entrench the underlying problem. Demonstrating reduced bed days or cost savings may reassure policymakers that homelessness is being managed, even as the structural drivers remain untouched. Good programs reduce immediate harm, but they risk becoming pressure valves that prevent deeper reform. (19)  

What is missing is not evidence of what works, but a system capable of acting on that evidence at scale. 

Policy implications 

Key reform directions are clear: 

• Housing must be recognised as health infrastructure. Targeted investment in rapid access and post discharge housing would free hospital capacity at far lower cost than inpatient care. 

• Discharge into homelessness should be treated as a safety issue, not an administrative outcome. Unsafe discharge should trigger accountability and transparency, not be normalised. 

• Health and housing data must be integrated. Linking datasets would allow governments to track referrals, response times, and outcomes, distinguishing recognition of housing need from its resolution. 

• Incentives must be aligned across systems. Hospitals should not be penalised for housing driven delayed discharge, and housing systems should be accountable for supporting timely and safe discharge. 

• Australia’s international obligations on housing should be reflected in domestic frameworks.  

• Stronger rights-based recognition would support whole of government accountability across health, housing, and social policy. 

Conclusion: Hospitals trapped between systems 

As housing shortages deepen and social systems fragment, hospitals have become the point where structural failure concentrates and where it is paid for. Sometimes through rapid discharge into risk. Sometimes through prolonged admission as a substitute for housing. Both pathways are costly, harmful, and unsustainable.  

Homelessness is no longer merely intersecting with healthcare. It is shaping hospital flow, length of stay, discharge safety, and public expenditure. 

Until health and housing systems are aligned through shared accountability, integrated data, and upstream investment, hospitals will remain trapped, treating the consequences of homelessness while funding its persistence. 

The question is no longer whether hospitals should respond to homelessness. They already do. The real policy choice is whether governments will continue using hospitals as an expensive and ineffective substitute for housing policy. 

Image by Tony McDonough for Homeless HealthCare, used with permission.

Endnotes 

1. Australian Institute of Health and Welfare 2025a, Separation statistics by principal diagnosis (ICD‑10‑AM 12th edition), Australia, 2023–24 [Data set], https://www.aihw.gov.au/reports/hospitals/principal-diagnosis-data-cubes 

2. Ibid. 

3. Independent Health and Aged Care Pricing Authority 2025, National Hospital Cost Data Collection: Public sector report 2022–23, https://www.ihacpa.gov.au/publications/nhcdc-public-sector-report-2022-23  

4. Australian Institute of Health and Welfare 2025b, Health of people experiencing homelessness, https://www.aihw.gov.au/reports/australias-health/health-of-people-experiencing-homelessness  

5. Australian Institute of Health and Welfare 2024, Emergency department presentations: Principal diagnosis by ICD‑10‑AM chapter (National Non‑admitted Patient Emergency Department Care Database,. https://www.aihw.gov.au/hospitals/topics/emergency-departments/presentations  

6. Currie J, Stafford A, Hutton J & Wood L 2023, Optimising access to healthcare for patients experiencing homelessness in hospital emergency departments, International Journal of Environmental Research and Public Health, 20(3), 2424, https://doi.org/10.3390/ijerph20032424  

7. Australian Institute of Health and Welfare 2025c, Specialist homelessness services annual report 2024–25: Data tables, https://www.aihw.gov.au/reports/homelessness-services/specialist-homelessness-services-annual-report/data  

8. Australian Institute of Health and Welfare 2025a, op cit. 

9. Australian Institute of Health and Welfare 2025c, op cit, 

 10. Australian Institute of Health and Welfare 2024,  Hospital resources 2022-23, https://www.aihw.gov.au/reports-data/myhospitals/resources  

11. Australian Housing and Urban Research Institute 2013a. The cost of homelessness and the net benefit of homelessness programs: A national study (AHURI Final Report No. 205), https://www.ahuri.edu.au/research/final-reports/205   

12. Flatau P, Zaretzky K, Brady M, Haigh Y & Martin R 2017, The cost effectiveness of Housing First support (AHURI Final Report No. 271), https://www.ahuri.edu.au/research/final-reports/271  

13. Australian Institute of Health and Welfare 2025b, op cit. 

14. Australian Medical Association 2020, Social determinants of health: AMA position statement, https://www.ama.com.au/sites/default/files/2020-12/Social%20Determinants%20of%20Health%202020%20-%20AMA%20Position%20Statement_1.pdf 

15. Australian Housing and Urban Research Institute 2022, Housing, homelessness and mental health: Towards systems change, https://www.ahuri.edu.au/sites/default/files/documents/2022-02/Housing-homelessness-and-mental-health-towards-systems-change.pdf 

16. Ibid. 

17. Productivity Commission 2026, Housing. Report on Government Services, https://www.pc.gov.au/ongoing/report-on-government-services/housing-homelessness/housing 

18. Australian Housing and Urban Research Institute 2013b, The cost effectiveness of homelessness programs (AHURI Research & Policy Bulletin No. 104), https://www.ahuri.edu.au/sites/default/files/migration/documents/AHURI_RAP_Issue_104_The_cost_effectiveness_of_homelessness_programs.pdf 

19. Ibid. 

This article first appeared in Parity magazine’s April 2026 edition. Read more about Parity here.