Thesis – Social Housing Supply and Homelessness Outcomes in Melbourne

Abstract

This thesis examines the relationship between social housing supply and homelessness outcomes in Greater Melbourne between 2015 and 2024. An explanatory sequential mixed methods design was employed: a quantitative phase analysed published administrative indicators drawn from the Australian Institute of Health and Welfare, the Australian Bureau of Statistics and the Victorian Housing Register, followed by a qualitative phase of eight semi-structured stakeholder interviews that explained the observed patterns. The indicator analysis shows that average annual social housing completions almost quadrupled after the Victorian Big Housing Build commenced, yet the housing register grew by 62 per cent and the estimated homelessness rate rose from 41.2 to 46.1 per 10,000 residents. Stakeholders attributed this divergence to construction and allocation lags, the narrowing of lettings to highest-need applicants, and the dependence of tenancy sustainment on support services. The thesis concludes that supply expansion is necessary but not sufficient, and recommends mandatory inclusionary zoning, expanded community housing provider capacity and protected support funding.

Introduction

Melbourne is experiencing the most sustained housing affordability pressure in its post-war history. Advertised rents across the metropolitan area outpaced wage growth in each year from 2021 to 2024, vacancy rates remained below 2 per cent, and on Census night 2021 more than 30,000 Victorians were estimated to be homeless, most of them within Greater Melbourne (ABS 2023). Specialist homelessness services in Victoria assist more clients per capita than in any other state while still turning away a substantial share of those who seek help (AIHW 2024). At the same time, social housing has fallen to approximately 3 per cent of Victorian dwellings, the lowest share of any Australian state, a position housing scholars characterise as slow structural failure rather than cyclical downturn (Pawson, Milligan & Yates 2020; Productivity Commission 2024).

The Victorian Government’s $5.3 billion Big Housing Build, announced in November 2020, is the largest single social housing investment in the state’s history and promises more than 12,000 new dwellings. Whether investment on this scale improves homelessness outcomes, and over what time horizon, is an empirical question of national significance for the Housing Australia Future Fund and successor programs. This thesis addresses the question for Greater Melbourne across the decade 2015-2024, guided by three research questions:

1. RQ1: How has social housing supply in Greater Melbourne changed relative to assessed need between 2015 and 2024?
2. RQ2: What relationship is observable between supply indicators and homelessness outcomes over this period?
3. RQ3: How do sector stakeholders explain the mechanisms linking supply, support services and homelessness outcomes?

The extract below summarises the literature, describes the mixed methods design, presents the quantitative and qualitative findings, and discusses their implications for Victorian policy.

Literature Review

Housing First and its Australian adaptations

Housing First reverses the treatment-first staircase by providing immediate permanent housing with voluntary wraparound support, and its effectiveness for chronically homeless adults is among the most replicated findings in homelessness research (Padgett, Henwood & Tsemberis 2016). Australian programs influenced by the model, including Melbourne’s Journey to Social Inclusion, report strong tenancy retention. Yet Clarke, Parsell and Vorsina (2020) show that Australian implementation is frequently conditional in practice, because scarce social housing forces services to ration access, contradicting the model’s premise of housing as an unconditional starting point. Parsell (2018) likewise argues that responses to homelessness reflect what governments are prepared to provide rather than what evidence supports. The Australian Housing First literature therefore leads back to the supply question.

Supply-side and demand-side explanations

A second debate concerns whether homelessness is best addressed through housing supply or demand-side transfers such as Commonwealth Rent Assistance. Longitudinal Australian research finds that housing and labour market conditions, especially the availability of affordable rental stock, outweigh individual risk factors in explaining entries into homelessness (Johnson et al. 2019; Wood et al. 2015). Hulse et al. (2019) further demonstrate that even when aggregate rental supply grows, the stock affordable and available to the lowest income quintile contracts as new supply filters to higher rent bands. This supports supply-side intervention targeted at the bottom of the market, while cautioning that headline construction figures can mask shortages where they matter most.

The decline of Australian social housing stock

Nationally, social housing has declined from more than 6 per cent of dwellings in the mid-1990s to around 4 per cent, with Victoria persistently ranked last among the states (Pawson, Milligan & Yates 2020; Productivity Commission 2024). Lawson et al. (2018) estimate a national requirement exceeding 700,000 additional social dwellings by 2036 and argue that social housing should be financed as economic infrastructure through capital grants and low-cost bond finance. What remains under-examined is the metropolitan-scale relationship between supply pipelines and homelessness outcomes, and how practitioners understand the mechanisms involved. This thesis addresses that gap for Melbourne.

Methodology

The study employed an explanatory sequential mixed methods design, in which a quantitative phase is completed first and a subsequent qualitative phase explains the observed patterns (Creswell & Plano Clark 2018). Figure 1 illustrates the sequence.

Phase 1 compiled descriptive indicators for Greater Melbourne for 2015-2024 from published administrative sources: the AIHW Specialist Homelessness Services annual collection (AIHW 2024), ABS Census homelessness estimates (ABS 2023), Victorian Housing Register quarterly releases and the Report on Government Services (Productivity Commission 2024). Three headline indicators were constructed: applications on the Victorian Housing Register, average annual social housing completions, and the estimated homelessness rate per 10,000 residents. Values were aggregated into two-year bands to smooth definitional and reporting changes across sources.

Phase 2 comprised eight semi-structured interviews of 45-60 minutes conducted between March and May 2024 with a purposive sample: three community housing provider executives, two state government housing officials, two specialist homelessness service managers and one metropolitan council strategic planner. Transcripts were coded using reflexive thematic analysis (Braun & Clarke 2022), and the university’s Human Research Ethics Committee approved the protocol, with participants identified by role only. Integration occurred through a joint display comparing indicator trends with stakeholder explanations. The design is explanatory rather than causal: the indicator series cannot isolate supply effects from concurrent shocks such as pandemic-era rental market disruption.

Findings

Indicator trends, 2015-2024

Table 1 presents the indicator series. Three patterns stand out. First, supply effort increased materially: average annual completions rose from 980 in 2015-16 to 3,640 in 2023-24 as Big Housing Build projects reached practical completion, an almost fourfold increase. Second, assessed need grew faster: register applications rose from 35,600 to 57,800, an increase of 62 per cent (22,200 / 35,600 = 0.62). Third, homelessness outcomes did not improve: the estimated rate rose from 41.2 to 46.1 per 10,000, dipping only in 2021-22 (43.9) when emergency accommodation programs such as Homelessness to a Home temporarily housed rough sleepers in hotels (VAGO 2024). Even at the 2023-24 delivery peak, annual completions equalled just 6.3 per cent of the register (3,640 / 57,800 = 0.063), so a household joining the register in 2024 faces a multi-year wait unless granted priority status.

Table 1: Social housing supply and homelessness indicators, Greater Melbourne, 2015-2024 (two-year bands)

Note: Rounded study estimates compiled from AIHW (2024), ABS (2023), Productivity Commission (2024) and Victorian Housing Register quarterly releases; completions are averaged within each band.

The community housing provider (CHP) share of social stock rose from 19 to 26 per cent over the decade, reflecting Big Housing Build contracting arrangements and management transfers, a compositional shift whose implications for sector capacity are discussed below.

Stakeholder explanations

Table 2 summarises the five themes generated from the interviews.

Table 2: Stakeholder interview themes (n = 8)

All eight participants described a structural lag between funding and lettings. A state official observed: “What is completing now was funded three budgets ago; nothing approved today reaches the register before 2027.” Seven described allocation narrowing: because lettings are rationed to priority categories such as family violence and chronic homelessness, new stock initially changes who is housed rather than how many wait, leaving the register “effectively two queues” in the words of one homelessness service manager. Six argued that tenancy sustainment depends on support: “A key without support is a tenancy at risk within six months” (CHP executive). The same number identified financing constraints capping provider growth, and five viewed the planning system as friction, with affordable housing contributions negotiated “deal by deal” and yielding uncertain dwelling numbers.

Discussion

The qualitative phase reframes the apparent paradox in Table 1. Housing supply operates on a pipeline measured in years: land assembly, planning approval and construction each add 12-18 months, and allocation rules direct new stock to the most acute need first, so headline register length is the last indicator to move. The divergence between rising completions and worsening outcomes is therefore consistent with, not contrary to, an effective supply program observed mid-pipeline. This supports the argument of Lawson et al. (2018) that social housing requires continuous infrastructure-style investment rather than episodic programs whose benefits arrive after the political cycle that funded them.

Second, gross completions overstate net additions. A share of Big Housing Build delivery replaces demolished estate stock, and the Victorian Auditor-General’s Office has noted the program’s reliance on completions concentrated late in the decade (VAGO 2024). Combined with population growth and the contraction of affordable private rental documented by Hulse et al. (2019), net effective supply at the lowest rent band grew far more slowly than the completion figures imply. On these dynamics, measurable stabilisation of the register and the homelessness rate is plausibly a late-2020s outcome, a hypothesis testable as post-2024 data are published.

Third, stakeholders positioned support services as the moderator that converts dwellings into durable exits from homelessness, echoing Housing First evidence that housing is the platform and support the mechanism of stability (Padgett, Henwood & Tsemberis 2016). Where support is rationed, allocating high-need households to tenancies without adequate support risks tenancy failure and a return to homelessness, reproducing the conditionality identified by Clarke, Parsell and Vorsina (2020). The findings thus endorse the structural account of homelessness (Johnson et al. 2019; Wood et al. 2015) while specifying an interaction: supply determines how many households can exit homelessness, and support determines whether those exits endure. Because the study rests on descriptive aggregate indicators and a small purposive sample, these mechanisms stand as explanatory hypotheses rather than causal estimates.

Policy Implications

Mandatory inclusionary zoning

Victoria currently relies on voluntary affordable housing agreements under section 173 of the Planning and Environment Act 1987, which participants described as slow and contested. A mandatory, metropolitan-wide inclusionary zoning schedule requiring fixed contributions of 5-10 per cent of dwellings on rezoned land would convert planning uplift into a continuous supply pipeline that is less exposed to budget cycles than grant-funded construction alone.

Community housing provider capacity

With CHPs now managing roughly a quarter of Melbourne’s social stock, future growth depends on their balance sheets and workforce. Pairing capital grants with low-cost finance through Housing Australia’s bond aggregator, and transferring title where regulatory safeguards permit, would allow providers to leverage assets at scale (Lawson et al. 2018; Pawson, Milligan & Yates 2020).

Protected support funding

Under the National Agreement on Social Housing and Homelessness that commenced in July 2024, support funding should be indexed and allocated alongside Big Housing Build lettings so that high-need allocations arrive with attached support packages. Governments should also publish net additions annually, so that progress is measured against need rather than gross construction counts.

Conclusion

This thesis asked whether the largest social housing investment in Victorian history has yet shifted homelessness outcomes in Melbourne. The indicator analysis shows that supply effort nearly quadrupled while the register grew 62 per cent and the homelessness rate continued to rise (RQ1, RQ2), and the interviews explain the divergence through pipeline lag, allocation narrowing and the dependence of tenancy sustainment on support services (RQ3). The central conclusion is that supply expansion is necessary but not sufficient: without a continuous pipeline, planning mechanisms that capture uplift, financially capable community providers and protected support funding, completions will keep arriving behind need. Future research should link dwelling-level administrative data to tenancy outcomes for Big Housing Build allocations, which would allow the lagged effects hypothesised here to be tested directly.

References

Australian Bureau of Statistics 2023, Estimating Homelessness: Census, 2021, ABS, Canberra.

Australian Institute of Health and Welfare 2024, Specialist Homelessness Services Annual Report 2023-24, AIHW, Canberra.

Braun, V & Clarke, V 2022, Thematic Analysis: A Practical Guide, Sage, London.

Clarke, A, Parsell, C & Vorsina, M 2020, ‘The role of housing policy in perpetuating conditional forms of homelessness support in the era of Housing First: evidence from Australia’, Housing Studies, vol. 35, no. 5, pp. 954-975.

Creswell, JW & Plano Clark, VL 2018, Designing and Conducting Mixed Methods Research, 3rd edn, Sage, Thousand Oaks.

Hulse, K, Reynolds, M, Nygaard, C, Parkinson, S & Yates, J 2019, The Supply of Affordable Private Rental Housing in Australian Cities: Short-term and Longer-term Changes, AHURI Final Report no. 323, Australian Housing and Urban Research Institute, Melbourne.

Johnson, G, Scutella, R, Tseng, YP & Wood, G 2019, ‘How do housing and labour markets affect individual homelessness?’, Housing Studies, vol. 34, no. 7, pp. 1089-1116.

Lawson, J, Pawson, H, Troy, L, van den Nouwelant, R & Hamilton, C 2018, Social Housing as Infrastructure: An Investment Pathway, AHURI Final Report no. 306, Australian Housing and Urban Research Institute, Melbourne.

Padgett, DK, Henwood, BF & Tsemberis, SJ 2016, Housing First: Ending Homelessness, Transforming Systems, and Changing Lives, Oxford University Press, New York.

Parsell, C 2018, The Homeless Person in Contemporary Society, Routledge, Abingdon.

Pawson, H, Milligan, V & Yates, J 2020, Housing Policy in Australia: A Case for System Reform, Palgrave Macmillan, Singapore.

Productivity Commission 2024, Report on Government Services 2024: Housing and Homelessness, Productivity Commission, Canberra.

Victorian Auditor-General’s Office 2024, Delivering the Big Housing Build, VAGO, Melbourne.

Wood, G, Batterham, D, Cigdem, M & Mallett, S 2015, The Structural Drivers of Homelessness in Australia 2001-11, AHURI Final Report no. 238, Australian Housing and Urban Research Institute, Melbourne.