COVID-19 Rental Housing and Homelessness Impacts — An Initial Analysis by Tasmanian Council of Social Service Inc

COVID-19: RENTAL HOUSING AND HOMELESSNESS IMPACTS – AN INITIAL ANALYSIS PAWSON, H., MARTIN, C., SISSON, A., THOMPSON, S. (CITY FUTURES RESEARCH CENTRE, UNSW) FITZPATRICK, S. (HERIOT-WATT UNIVERSITY) MARSH, A. (UNIVERSITY OF BRISTOL) A Report for the ACOSS-UNSW Poverty and Inequality Partnership by the City Futures Research Centre at UNSW

ACOSS Partners Acknowledgements The authors gratefully acknowledge the generous contribution of industry, government and NGO representatives through participation in interviews for this research, and provision of requested data. We also thank the UNSW-ACOSS Poverty and Inequality Partnership for supporting the project, as well as the following organisations which are kindly co-funding the research:

B B & A MILLER

FOUNDATION

• Mission Australia • National Shelter (on behalf of Shelter NSW and Shelter WA) • Queensland Shelter Thanks are additionally due to colleagues who kindly commented on an earlier draft of this report: Ryan Harris (Shelter NSW), Adrian Pisarski (National Shelter), Marion Bennett (Mission Australia), Emma Greenhalgh (Queensland Shelter), Michelle Mackenzie (Shelter WA), Carla Treloar (Social Policy Research Centre, UNSW), Ken Gibb (University of Glasgow). We are also indebted to Dr Fatemeh Aminpour, our report designer.

DAVID MORAWETZ’S SOCIAL JUSTICE FUND

Cover image credit: National Institute for Allergy and Infectious Diseases (NIAID). Used under Creative Commons licence (CC BY2.0)

Suggested citation Pawson, H., Martin, C., Sisson, A., Thompson, S., Fitzpatrick, S. and Marsh, A. (2021) 'COVID-19: Rental housing and homelessness impacts – an initial analysis'; ACOSS/UNSW Poverty and Inequality Partnership Report No. 7, Sydney ISSN: 1326 7124 ISBN: 978 0 85871 021 4

HART LINE AND RAETTVISA SUB-FUNDS OF

© Australian Council of Social Service and University of New South Wales This publication is copyright. Apart from fair dealing for the purpose of private study, research, criticism or review, as permitted under the Copyright Act, no part may be reproduced by any process without written permission. Enquiries should be directed to the Publications Officer, Australian Council of Social Service. Copies are available from the address above.

COVID-19: Rental housing and homelessness policy impacts, 2021

Glossary ABS

Contents Australian Bureau of Statistics

ACOSS

Australian Council of Social Service

AHURI

Australian Housing and Urban Research Institute

APRA

Australian Prudential Regulatory Authority

BBB

Building back better

CFRC

City Futures Research Centre

CHP

Community Housing Provider

Coronavirus Supplement Temporary boost to social security payments CPI

Emergency accommodation (temporary accommodation)

JobKeeper Payment

COVID-19 furlough/wage subsidy scheme

JobSeeker Payment

Unemployment payment

LGA

Local government area

National Cabinet

Australian federal decision-making forum including the prime minister, state premiers and territory chief ministers, successor of COAG (Council of Australian Governments)

EXECUTIVE SUMMARY

Overview

Policymaking in the crisis context

Housing market impacts

Pandemic impacts on existing tenancies

Rental housing policy and practice impacts

Homelessness impacts

Conclusion

1. Introduction

1.1 Research purpose and themes

1.2 COVID-19, housing and homelessness

1.3 Australian housing and homelessness governance context

1.4 Research methods

1.5 Report structure

References

2. Crisis policy innovations in response to COVID-19

2.1 COVID-19 national income support programs

2.2 Rental housing regulation innovations in the 2020 crisis

2.3 Homelessness policy responses to the pandemic

References

Consumer Price Index

EA (TA)

PRS

FOREWORD

Private rental sector

COVID-19: Rental housing and homelessness policy impacts, 2021

3. Policy innovation in crisis conditions

6.4 Boarding and lodging

3.1 Introduction

6.5 Social and affordable housing

3.2 Crisis policy theory

6.6 Tenancy management and service provision

102

3.3 Crisis housing and homelessness policymaking 2020: emerging published evidence

6.7 Chapter conclusion

104

3.4 Chapter conclusion

References

105

References

7. Homelessness policy and practice

107

4. Pandemic impacts on the private rental housing market

7.1 Introduction

107

4.1 Introduction

7.2 Policy and practice context for COVID-19 interventions

108

4.2 Private rental housing supply

7.3 Initial pandemic homelessness impacts

111

4.3 Rents

7.4 Homelessness policy action during the pandemic

113

4.4 Vacancy rates, new tenancies & terminations

7.5 Chapter conclusion

124

4.5 Chapter conclusion

References

125

References

8. Housing and homelessness policy change in the United Kingdom: an international comparator

127

5. Pandemic impacts on existing tenancies

8.1 Rental housing policy

127

5.1 Introduction and data sources

8.2 Homelessness policy

131

5.2 Impacts on existing tenancies: triangulating research findings

8.3. Chapter conclusion

133

5.3 Moving out

References

134

5.4 Rent variations

9. Conclusions

135

5.5 Chapter conclusion

Market impacts

135

References

Interpreting crisis policy

135

6. Rental housing policy and practice in the COVID-19 emergency

References

136

6.1 Context: long-term and immediate

6.2 Eviction moratoriums

6.3 Measures regarding rents

COVID-19: Rental housing and homelessness policy impacts, 2021

Figures

Tables

Figure 4.1: Size of private rental sector, Australia, 1994-2018

Figure 4.2: Value of new home loan commitments in Australia, seasonally adjusted, Nov 2010Sep 2020

Figure 4.3: Value of new home loan commitments for investor purchase, seasonally adjusted, Nov 2010-Sep 2020

Figure 4.4: New private dwelling completions, Australia, seasonally 53 adjusted, Sep 2011- Jun 2020 Figure 4.5: New private dwelling approvals, Greater Sydney, 12-month moving average, Nov 2010-Sep 2020

Figure 4.17: Median agreed rents, quarterly & indexed to Mar-19, Melbourne & regional Victoria

Figure 4.18: Median agreed rents 2-BR flats/units, quarterly & indexed to Mar-19, select regions Queensland

Figure 4.19: Median agreed rents for 3-BR houses, quarterly & indexed to Mar-19, select regions Queensland

Figure 4.20: Jobseeker recipients NSW, March & June 2020

Figure 4.21: Jobseeker recipients Victoria, March & June 2020

Figure 4.22: Residential vacancy rates, capital cities, Oct 2019-Oct 2020

Figure 4.23: Net migration, greater capital cities, Oct 2019Oct 2020

Figure 4.6: New private dwelling approvals, Greater Melbourne, 12-month moving average, Nov 2010-Sep 2020

Figure 4.7: New private dwelling approvals, Greater Brisbane, 12-month moving average, Nov 2010-Sep 2020

Figure 4.24: Residential vacancy rates, select regions NSW, Oct 2019-Oct 2020

Figure 4.8: Airbnb listings for entire houses, Sydney, Jan 2020Oct 2020

Figure 4.25: Residential vacancy rates, select regions in Victoria, Oct 2019-Oct 2020

Figure 4.9: Airbnb listings for entire houses, Melbourne, Jan 2020-Oct 2020

Figure 4.26: Residential vacancy rates, select regions in Queensland, Oct 2019-Oct 2020

Figure 4.10: Change in median asking rents, quarter on quarter and year on year to June 2020

Figure 4.27: Proportion of previous quarter bonds refunded, 68 NSW, Jun 2019-Sep 2020

Figure 4.11: Quarterly median asking rent per week, houses, Mar 2019-Sep 2020

Figure 4.28: Quarterly change in total bonds held, NSW, Jun 2019Sep 2020

Figure 4.12: Quarterly median asking rent per week, units, Mar 2019-Sep 2020

Figure 4.29: Quarterly change in total bonds held, Greater Sydney, Jun 2019-Sep 2020

Figure 4.13: Median agreed rents, quarterly & indexed to March-19, Greater Sydney

Figure 4.30: Proportion of previous quarter bonds refunded, 70 Victoria, Jun 2019-Sep 2020

Figure 4.14: Median agreed rents, quarterly & indexed to Mar-19, Greater Melbourne

Figure 4.31: Quarterly change in total bonds held, Victoria, Jun 2019-Sep 2020

Figure 4.15: Median agreed rents, quarterly & indexed to Mar-19, Brisbane LGA

Figure 4.32: New bond lodgements, Greater Melbourne, Mar 2019-Sep 2020

Figure 4.16: Median agreed rents, quarterly & indexed to Mar-19, Sydney & regional NSW

Figure 4.33: Quarterly change in total bonds held, Queensland, Jun 2019-Sep 2020

COVID-19: Rental housing and homelessness policy impacts, 2021

Figure 4.34: New bond lodgements, Brisbane LGA, Mar 2019-Sep 2020

Figure 4.35: Refunded bonds, Brisbane LGA, Jul 2019-Jun 2020

Figure 5.1: Rental survey respondents: breakdown according to whether rent 80 reduction sought and negotiation outcomes Figure 7.1: Overall homelessness numbers; change over time 201116

108

Figure 7.2: ABS Census homelessness operational groups, Australia-wide 2011 and 2016

109

Figure 7.3: Street homelessness trends (from streetcounts) in selected city centres, 2010-20

109

Figure 7.4: Average monthly percentage change in number of persons assisted by specialist homelessness services, 20182020

112

Table A: Key pandemic policy innovations relevant to minimising housing market disruption and homelessness – Australia

Table 5.1: Renter impacts: compendium table

Table 5.2: Renters who asked for variations, were discouraged, and 81 who got variations Table 6.1: Eviction moratorium coverage by jurisdiction

Table 6.2: Emergency measures regarding rents by jurisdiction

Table 7.1: Emergency accommodation placements – flow 15 March-30 September (persons)

120

Table 7.2: Temporary accommodation placements at points in time (persons)

121

Table 7.3: Departures from emergency accommodation – former rough sleepers – 15 Mar30 Sep (persons)

122

Table 7.4: Departures from emergency accommodation – all homeless people – 15 Mar-30 Sep (persons)

123

Foreword This report is the latest published by the Poverty and Inequality Partnership between the Australian Council of Social Service (ACOSS) and UNSW Sydney. It has been researched and written for the partnership by representatives of the City Futures Research Centre at UNSW Sydney, Heriot-Watt University in Scotland and the University of Bristol in the UK. It is also supported by Mission Australia, National Shelter and Queensland Shelter.

We thank the partnership and supporters of this project for their assistance throughout, as well as the ACOSS Board, UNSW Vice Chancellor Ian Jacobs, and UNSW Deputy Vice-Chancellor Equity, Diversity and Inclusion Professor Eileen Baldry.

This report, the first of two planned reports from this particular research project, looks at the impacts of the COVID-19 pandemic – still rampaging throughout the world as this foreword is written – on housing and homelessness in Australia during 2020. Impacts of the pandemic upon housing and homelessness were inevitable, given its effects on the economy as well as public health. This report looks at what housing and homelessness policy changes occurred; how these changes were formulated; and how they were implemented, using both publicly available data and focused research undertaken by the report authors. It also looks at wider policy changes, such as the introduction of JobKeeper and the increase to JobSeeker payments, and what impacts those changes had on housing and homelessness policies. This report is the first report published by the Poverty and Inequality Partnership in 2021. It follows the publication in 2020 of six reports, five of which were part of the core Poverty and Inequality in Australia series, and one of which was a partnership research report on spatial poverty in Australia. The report was written by Hal Pawson, Chris Martin, Alistair Sisson and Sian Thompson from the City Futures Research Centre at UNSW Sydney; Suzanne Fitzpatrick from Heriot-Watt University in Scotland; and Alex Marsh from the University of Bristol in the UK. The Poverty and Inequality Partnership includes researchers from the Social Policy Research Centre, as well as the City Futures Research Centre; the Centre for Primary Health Care and Equity; and the Faculty of Law and Justice at UNSW Sydney. This cross-disciplinary approach allows the partnership to explore the ways in which poverty and inequality are bound to other aspects of disadvantage, such as health and justice, and, of course, housing and homelessness as explored in this particular publication. The Partnership is also made up of the support of non-government organisations from within ACOSS’ membership, along with philanthropists: Anglicare Australia; Australian Red Cross; the Australian Communities Foundation Impact Fund (and two sub-funds – Hart Line and Raettvisa); the BB and A Miller Foundation; the Brotherhood of St Laurence; cohealth; the David Morawetz Social Justice Fund; Good Shepherd Australia New Zealand; Mission Australia; the St Vincent de Paul Society; the Salvation Army; and The Smith Family.

COVID-19: Rental housing and homelessness policy impacts, 2021

Executive Summary Key findings • Australia’s rental housing markets experienced substantial turbulence during 2020, with markedly divergent trends experienced between inner cities and regional areas, between houses and units, and between the eastern state capitals and Perth.

• In the UK, central government’s primary response to pandemic-triggered housing insecurity was, like the Australian Government’s, temporary income support. Legal measures to prevent evictions were patchy, though devolved administrations such as Scotland acted more decisively.

• Median rents fell by nearly 5% in north and west Melbourne between Quarter 1 and Quarter 3 2020, and by as much as 10% in Sydney, but rose by 6% across non-metropolitan Victoria, and by more than 5% in many parts of regional NSW.

• The UK Government was more involved in coordinating and funding efforts to accommodate homeless persons, and the proportion of those placed who were subsequently transitioned to longer-term housing has been double that in Australia.

• There were similar spatial contrasts in rental housing vacancy trends, with Melbourne’s vacancy rate doubling to 4.5% between March and September 2020 while Perth saw its equivalent figure halved to under 1%.

• In contrast with all four comparator countries, Australia’s national government made no coordination or funding input to homelessness EA programs.

• On housing affordability, renters tended to be hit much harder by the COVID-19 pandemic than homeowners. Nationally, renter incomes fell by 5% March-June 2020, while housing costs dropped by only 0.5%; mortgage holders, by contrast, saw a 0.2% decline in incomes alongside a 5% decline in housing costs. • At least a quarter of all private renters lost income during the pandemic, but only a smaller minority got a rent variation from their landlord: between 8-16% of renters, depending on the data source. A similar proportion was refused a variation; more were discouraged from asking and more left their tenancy. • Some 12% of private rental properties were subject to loan payment deferrals authorised by banks. • At least 30% of rent variations merely deferred the rent, rather than reduced it. This implies that tenants with mounting deferred rent debts could number at least 75,000 across Australia in late 2020. • State and territory eviction moratoriums varied significantly, but were generally easily understood and reasonably effective responses to the crisis. Rent variation frameworks were less satisfactory, being reliant on landlordtenant negotiation with little direction from governments. They caused problems for renters during the emergency and going forward. • Income support payments and eviction moratoriums meant the COVID-19 pandemic triggered no immediate increase in homelessness. • Four state governments authorised mass provision of emergency accommodation (EA) for rough sleepers and other homeless people – action that, by September 2020, had benefited at least 40,000 people. 12

• However, these governments and their non-government organisation (NGO) partners were able to facilitate transitions to longer term housing for only a minority of those provided with temporary hotel rooms and similar accommodation. Less than a third (32%) of the 8,000 former rough sleepers who departed EA in the six months to 30 September 2020 had been assisted into longer term tenancies

COVID-19: Rental housing and homelessness policy impacts, 2021

Overview This is the first of two planned reports from an ongoing investigation being undertaken as part of the UNSW-ACOSS Poverty and Inequality Partnership work program, and also supported by Mission Australia, National Shelter and Queensland Shelter. Initiated in response to the COVID-19 pandemic, the research focuses primarily on the domains of rental housing and homelessness. Its main aims are to inform an understanding of: a. What relevant policy shifts or innovations have been prompted by the COVID-19 pandemic b. How these policy innovations have been formulated c. How policy innovations been implemented and with what effect – for both service delivery organisations and service users. In addressing these objectives as they relate to Australia, we draw on the rapidly expanding body of relevant published research and statistical data evidence. This has been complemented by our own primary research and secondary data analysis to provide a fuller, more rounded picture. To place Australian developments in context, some reference is also made to pandemictriggered housing and homelessness policy innovations in the UK and three other comparable countries (Canada, New Zealand and the USA). Research methods included an Australian and international literature review, indepth interviews with government, industry and advocacy sector stakeholders (N=20), a survey of state government emergency accommodation activity, an online survey of private renters (N=312) and a triangulation of findings from other renter surveys and data sources.

Policymaking in the crisis context

Housing market impacts

The COVID-19 emergency unfolded in 2020 as a dual calamity in public health and the economy. It was clear from the start that both aspects would play out in countries’ housing systems. The national economic shutdown rapidly enacted from March prompted widespread fears that resulting mass unemployment would trigger a surge in rental evictions and homelessness, as well as a housing market crash. Faced with this alarming vision, governments rapidly enacted protective measures well above and beyond what most would have previously envisaged.

Partly to further contextualise our investigation of housing and homelessness policy responses to the COVID-19 crisis, the research also analysed rental housing market impacts of the pandemic as these played out across Australia in the period to October 2020. Starkly different trends for different regions and dwelling types are apparent. On the one hand, the initial months of the crisis saw a decline in the demand for inner-metropolitan rental housing, particularly for apartments and units in Sydney, Melbourne, and (to a lesser extent) Brisbane. In these submarkets, median rents fell by 5-10% while vacancy rates rose, and total tenancies contracted by 1-3%.

Table A: Key pandemic policy innovations relevant to minimising housing market disruption and homelessness – Australia Federal Govt

State/ territory govts

Income support – JobKeeper (wage subsidy paid via employers)

Income support – Coronavirus Supplement (temporary boost to designated social security payments)

Access to superannuation savings allowed

Facilitation of banks’ mortgage payment deferral programs*

Rental eviction restrictions (‘moratoriums’)

Rent increase restrictions

Rent relief

Homelessness emergency accommodation programs

* Bank mortgage deferrals were officially facilitated through the Reserve Bank of Australia’s Term Funding Facility (allowing banks to borrow cheaply as their own loans come due) and through the Australian Prudential Regulatory Authority’s relaxation of requirements regarding impaired loans.

While presented as a simple necessity in an extraordinary situation, aspects of these measures can also be analysed and interpreted within the context of the policymaking aphorism ‘never let a good crisis go to waste’. This powerfully indicates a belief that an emergency situation can provide space for innovation and reform beyond the normal bounds of possibility. According to the political science literature, such opportunities may be closely tied to the way crisis policy conundrums are defined and portrayed, with interest groups or ‘advocacy coalitions’ potentially enjoying scope for novel framing of such challenges to advance their goals. Understanding how policy formulation and implementation occurs within a crisis setting may help inform expectations on the prospect that this might provide a basis for ‘building back better’ when the emergency has subsided – or, indeed, for reacting to future crises. It is partly within this context that in-depth research to investigate, document and interpret the housing and homelessness policy innovations of 2020 is important. The key policy innovations that are the focus of this study are items 5-8 in Table A. Similar income support, renter protection and homelessness responses were seen in all of the comparator countries, with varying eligibilities and durations.

COVID-19: Rental housing and homelessness policy impacts, 2021

These developments partly reflect the geographically uneven effects of the pandemic and government responses: lockdowns and working from home prompted changing housing needs and desires, while populations in innermetropolitan areas were disproportionately affected by unemployment and income loss, particularly when the high housing costs of these submarkets are factored in. Importantly, declining median rents do not necessarily point to greater affordability, as ABS household finances suggest that renters’ incomes declined to a greater extent than rents, and survey data indicate a large number of renters in housing stress. In other submarkets, rental housing demand expanded during 2020. By October, vacancy rates in Perth, Adelaide, Hobart, Darwin and Canberra had fallen to 12-month lows of less than 1%, a pattern even more dramatic beyond capital city boundaries. Furthermore, in many non-metropolitan regions of NSW, Victoria and Queensland, and in some outer-metropolitan regions of these states, median rents rose and more renters entered the sector. For instance, in the Newcastle region, median rents rose by 5% and vacancy rates almost halved from March to October. These trends similarly reflect both changing tenant needs and desires and the income and employment effects of COVID-19 and (adequacy of) government support: these submarkets typically offer larger housing and cheaper rents than those in the inner- and middle-rings of the major cities, indicating some geographical movement prompted by aspirations for lower housing costs and/or more dwelling space. These trends raise concerns on the possible deterioration of affordability and availability of private rental housing in regions outside of major capital city cores and provide context for the evaluation of eviction moratoriums and rent variation measures.

Pandemic impacts on existing tenancies Our own online survey of renters, and other sources, indicate that rents in existing tenancies were only modestly impacted by the COVID-19 emergency. Although a loss of income was suffered by at least a quarter of renter households, only a small minority – between 8% and 16%, depending on the data source – actually got a rent variation. In our survey, a similar proportion of renters were refused a variation, and a larger proportion were discouraged from seeking one. A larger proportion moved, than got a variation. For the few who got a variation, the average size appears to have been substantial, but for 37% in our survey,– the variation is a mere deferral. Those amounts are liable to be paid later, on top of a full amount of rent – or will be arrears actionable by the landlord in termination proceedings.

The modest impact of the emergency on rents in existing tenancies is remarkable. Despite the prevalence of income loss, the surrounding framework of eviction moratoriums and official encouragement of negotiations, rents were reduced for a relative few.

Rental housing policy and practice impacts Australia’s emergency measures in rental housing policy were formulated by state and territory governments, in communication with each other and mostly on a common model, but with substantial differences in the details – a familiar pattern from the history of rental regulation. Their eviction moratoriums restricted some termination proceedings against a core COVID-hardship group but there were fewer additional protections for tenants more widely. Tasmania, Western Australia and Victoria had relatively strong moratoriums; the Northern Territory the weakest. Frameworks for rent variations relied on negotiations between tenants and landlords (and landlords’ agents), and mostly eschewed variations determined by state agencies, with only a little influence on negotiations applied through the terms of rent relief schemes (delivered variously through cash payments and land tax rebates). In practice, interviewees found the eviction moratoriums to be a measure readily understood by landlords and tenants that took some pressure out of the private rental sector after the income shocks widely experienced in the early emergency period. The rent variation frameworks, however, were less wellregarded. For some renters, it appears the uncertain timeframes, requirements and outcomes of negotiations were such that they opted instead for the certainty of terminating their tenancy and liabilities; for others, negotiations have produced deferred and accumulating liabilities and, for yet others, there may be uncertainty as to the terms of what they and their landlord agreed. Despite its absence from the design of state and territory measures, the Federal Government was arguably influential on these measures through its increase of income support payments. These probably did more to absorb the sector’s income shock, and let states and territories – and landlords – off the hook for making adjustments that would share income losses through rent variations. One implication of this is that renters who missed out on the income support measures – notably, non-permanent residents – had the benefit only of the moratoriums and rent variation frameworks, which is the lesser part of the whole response. A second implication relates to what happens as the income support measures are withdrawn. With the eviction moratoriums also due to expire in the new year, supported households who do not snap back to their full pre-COVID-19 employment status and income will be at risk, especially if they are carrying deferred rent liabilities from the emergency. We may yet see cause for further extensions, but without the public health impetus of March 2020. We may also see calls, perhaps under the banner of ‘rent relief’, for governments to pay landlords to settle those liabilities. More broadly, the experience of conducting emergency rent relief schemes might, as suggested by a housing policy interviewee, encourage state and territory government to become more involved in the provision of rental housing subsidies on an enduring basis. Similarly, their experiences of greater legal security for tenants, greater regulation of boarding and lodging arrangements, and executive dispute resolution outwith the tribunal, may furnish evidence for further reforms.

COVID-19: Rental housing and homelessness policy impacts, 2021

For the social housing sector, the emergency did not prompt major changes in policy, but there will be lessons to be learnt from the high-handed lockdown of the Flemington public housing towers, as well as the community housing sector’s absence from – and in some cases, dissent from – the early advocacy for eviction moratoriums. Perhaps the most pressing issue for the sector arising from the emergency is the question of its capacity to follow up initiatives in temporary accommodation for homeless persons.

Homelessness impacts The onset of the 2020 COVID-19 pandemic followed on from a period of rising homelessness and housing stress in Australia, especially in the pressured markets of Sydney, Melbourne and Brisbane. However, despite the almost instant loss of over 800,000 jobs due to the March 2020 national lockdown, no measurable rise in new homelessness at once resulted. In Q3 2020 the number of people being provided with homelessness services across Australia remained slightly below that for Q1 2020, immediately prior (in large part) to the pandemic. To varying degrees, policy responses 1-7 in Table A can be credited with having prevented the immediate income shock becoming an immediate homelessness crisis. Also contributing to this outcome will have been the crisis responses of housing cost-burdened tenants themselves, for example through inegotiating rent reductions, doubling up with other renters or returning to the family home. Nevertheless, temporary income and eviction protections are due to be phased out during early 2021 at a time when, as modelled in other research, recovering GDP will remain 4-5% below its 2019 level and where, as a result, unemployment will peak at around 8% on Reserve Bank forecasts – far above its 5.3% level at the start of the crisis. Other researchers 1 have estimated that, compared with the situation at the start of the pandemic, ‘housing affordability stressed’ households will increase from 757,000 to 793,000 (+5%) under a ‘mild recession’ scenario, but to 893,000 (+18%) if a severe recession eventuates. Without new compensatory measures, rising homelessness during 2021 would appear therefore appear highly likely. Soon after the onset of the pandemic four of Australia’s largest states launched emergency accommodation (EA) programs to provide safe temporary housing for existing rough sleepers and homeless people in shelter premises with shared facilities. By September 2020 more than 40,000 people had been assisted in this way, with street homelessness in Adelaide, Melbourne and Sydney – as a result – reduced to residual levels at mid-year. While involving action at unprecedented scale, this built on street homelessness engagement and rehousing efforts that had been already somewhat ramped up in several cities in the immediate pre-2020 period. Pre-existing practice here was not only hugely expanded, but also implemented less restrictively so that it encompassed nonAustralian citizens, and so that service users were booked into hotels for longer periods and with fewer conditions. EA programs were authorised and funded by state governments in NSW, Queensland, South Australia and Victoria. From the perspective of some NGO service providers, much of the impetus here came from the service providers themselves. Moreover, as noted in interviewee testimony, the eroded condition of some state governments when it comes to housing policy capacity, domain 1 Leishman, C., Ong, R., Lester, L. and Liang, W. (2020) Supporting Australia’s housing system: modelling pandemic policy responses, Final Report No. 346, Melbourne: AHURI

knowledge and policy-area-specific senior representation, may have impeded rapid and decisive action. Shared challenges in managing large scale emergency housing and re-housing of homeless people during the pandemic stimulated beneficial collaboration involving government and NGOs, and also across homelessness services sectors in the relevant states. They also provided impetus for a speedy ramping up of productive inter-jurisdictional co-operation and dialogue. All four states made substantial efforts to assist EA service users into longerterm tenancies. This included expansion of private rental subsidy and property headleasing programs, as well as enhanced access to social housing vacancies (albeit that the latter will have displaced other high priority applicants). Nevertheless, while around 8,000 former rough sleepers had departed EA by 30 September 2020, only a third of these (32%) had been assisted into longer term tenancies. England’s parallel EA program has been substantially more successful in transitioning people into move-on accommodation. By late 2020, two thirds of the 30,000 EA placements had been assisted into longer term housing. In Australia, the relatively rapid attrition of the EA population in most of the relevant states and the only modest numbers successfully assisted into longerterm housing exposed some weaknesses of program management – such as inadequacy of support provision and systemic limitations posed by insufficient social housing and inadequate Rent Assistance.

Conclusion Both in Australia and elsewhere 2020 saw emergency policy innovations on rental housing and especially homelessness of scale and significance that few could have imagined possible. At least partly through housing and homelessness emergency measures, Australia succeeded in preventing the spread of disease among the street homeless population and at least forestalling the new homelessness surge that would otherwise have been expected to result from the pandemic-triggered economic downturn. There is much to learn from this experience, about how such crisis policy innovation might be better handled in a future disaster scenario; about how the latter stages of the present crisis may play out for people at particular risk of housing stress and homelessness; and about the dispositions and capacities of the institutions and actors comprising Australia’s housing and homelessness systems. It is only through an in-depth understanding both of the pandemic emergency measures and of their institutional contexts that we can assess the prospects for building back better after the crisis has subsided.

1. Introduction 1.1 Research purpose and themes This is the first of two planned reports from an ongoing investigation by UNSW City Futures Research Centre. The study is being undertaken as part of the UNSW-ACOSS Poverty and Inequality Partnership work program, and also supported by Mission Australia, National Shelter and Queensland Shelter. Initiated in response to the COVID-19 pandemic as the public health crisis hit Australia in March 2020, the research focuses primarily on the domains of rental housing and homelessness. Its main aims are to inform an understanding of: a. What relevant policy shifts or innovations have been prompted by the COVID-19 pandemic b. How these policy innovations have been formulated c. How policy innovations been implemented and with what effect – for both service delivery organisations and service users. The project is pitched at the national level, with fieldwork spanning research participants and organisations located in several of Australia’s eight state/ territory jurisdictions. However, while recognising that there has been substantial geographic variation in the incidence of COVID-19, and in state/ territory-specific policy responses, such variations are not a central concern. Paralleling our Australian fieldwork, colleagues at the UK Collaborative Centre for Housing Evidence (CaCHE) have been undertaking similar research in Britain. To place some of our own findings in context, some reference to findings from these studies is included in Chapter 8. Comparative perspectives on pandemic-triggered housing and homelessness policy innovation in the UK and three other comparable countries (Canada, New Zealand and the USA) are introduced in Chapter 2. The project commenced in July 2020 and runs to April 2021. In this initial report we focus on housing policy and market developments during the first eight months of the COVID-19 pandemic as it affected Australia – that is, MarchOctober 2020.

1.2 COVID-19, housing and homelessness The COVID-19 pandemic unfolded in 2020 as a dual crisis in public health and the economy. Both aspects have played out in countries’ housing systems. As UN Special Rapporteur on Housing (Farha, 2020) put it, housing is the ‘first line of defence against the COVID-19 outbreak’. To shore up this line of defence, the early days of the emergency saw innovations in housing and homelessness policy, and in income support, formulated and implemented at astonishing speed and scale. Importantly, when it comes to Australia, the sudden onset of COVID-19 occurred against a backdrop of gradually intensifying post-millennial housing affordability stress as affecting large parts of the country. In many cities

COVID-19: Rental housing and homelessness policy impacts, 2021

and regions, house prices and rents have continued to trend upwards while homelessness has also outpaced broader population growth (Yates, 2016; Pawson et al. 2018, 2020a). Concurrently, mortgage debt has ballooned and an increasing proportion of banks’ lending books are dedicated to real estate (Conley, 2018; Jordà et al., 2016), with concerning implications for wider economic stability should the market fall (Pawson et al., 2020b). The start of the pandemic saw extraordinary changes to Australia’s social security income support payments, along with temporary wage subsidy systems (as outlined in the Executive Summary and detailed more specifically in Chapter 2). Similarly, and once again in parallel with similar moves overseas, the initial crisis period saw Australia’s state and territory governments rapidly legislating evictions moratoriums and enacting emergency interventions on homelessness. These latter interventions saw many thousands of rough sleepers and shelter-residents temporarily booked in hotels and other safe temporary accommodation. Most of these policy shifts – outlined more fully in Chapter 2 – have been presented as emergency actions, temporary in nature; but their dramatic emergence challenged conventional wisdom about what is politically and economically possible. In housing and homelessness, as in other policy domains, the pandemic has proved to be a ‘focusing event’ (Birkland, 1998) prompting hopes that the experience could open up possibilities for substantial, lasting and overdue policy reform – both in Australia and elsewhere. There is an opportunity for future housing and homelessness policymaking to be informed by an analysis of the recent rapid reform period: both in terms of the policy innovations themselves, and the ways they have been formulated. Moreover, given the pandemic’s global impact and the international parallels on housing and homelessness policy responses, we believe there is substantial scope for fruitful cross-country comparison. With all of the above in mind, and referencing relevant literature from development studies (e.g. Mannakkara et al., 2014), the central question addressed by this research is ‘how can an understanding of COVID-19 shock responses enable us to build back better in housing and homelessness policy?’

1.3 Australian housing and homelessness governance context Policy responsibilities and powers relevant to housing and homelessness are divided in Australia between the two upper tiers of government. Especially within the context of a report which includes an international comparative dimension that encompasses countries with their own specific housing governance frameworks, it is important that these are briefly explained here. In Australia’s federal system of government, the Australian Constitution allocates to the Australian Government (the Commonwealth or Federal Government) powers in numerous areas; these do not expressly include housing, tenancy, homelessness or land-use planning, so these areas are primarily the responsibility of the six states and two territories. Thus, it is the state and territory governments that regulate rental housing markets and providers, as well as owning and managing the bulk of social housing (the state/territory component of this termed ‘public housing’). Moreover, it is state and territory governments that are responsible for ensuring satisfactory housing outcomes for those whose needs are inadequately met by the private market – including people subject to homelessness. 20

COVID-19: Rental housing and homelessness policy impacts, 2021

Local government in Australia is established by state governments and has no statutory responsibilities or powers in these areas, apart from land-use planning. Certain capital city central municipalities (e.g. the City of Sydney) in practice play an active role in co-ordinating and in some instances funding direct service provision within their narrowly-drawn boundaries. Generally, though, Australia’s system is in this respect quite different from countries like the UK where local government is everywhere an important player in rental housing and homelessness service provision. While it has no constitutional responsibility for housing and homelessness, the Australian Government nevertheless plays several significant roles in these areas, through its powers with respect to taxation, grants to states and territories, and social security. The latter is important when it comes to rental housing and homelessness, because – on pre-pandemic figures – about half of renter households receive a government payment as part of their income, and for 27% the payment is their main source of income (ABS, 2019). These payments include: • Age and Disability Support Pension • JobSeeker Payment (formerly NewStart Allowance), Youth Allowance and Parenting Payment • Family Tax Benefit • Rent Assistance – paid as a supplement to other social security payments. Beyond this, reflecting its far more extensive fiscal powers, the Australian Government has a longstanding role when it comes to housing for lower income groups via grants paid to state and territory governments to support their social housing and homelessness activities. Historically, this was mediated through the Commonwealth-State Housing Agreement (CSHA) framework, the predecessor of today’s National Housing and Homelessness Agreement (NHHA). In practice, therefore, Australia’s housing and homelessness policy is formulated and delivered through a complex form of multi-level governance (Dodson et al., 2017). Nevertheless, since it is state and territory authorities that have direct policy responsibility for rental housing and homelessness, it is this tier of government that forms the main focus for this report.

1.4 Research methods 1.4.1 Overview As further explained below, the Australian background research and fieldwork has involved: • Literature review • Qualitative fieldwork: in-depth interviews with government and NGO stakeholders, and with homelessness and rental housing service delivery organisations • Quantitative analysis of housing market trends and pandemic policy impacts. 21

Parallel research across a broader range of housing topics is being undertaken in the UK by the Glasgow University-led CaCHE consortium. Beyond this, in the second phase of the research (2021) UNSW and CaCHE colleagues will interview academic or other policy experts in six comparator countries beyond Australia and the UK. The second phase of the project will also include rental housing/homelessness service user interviews.

• NSW (8)

In addition to the original fieldwork being undertaken specifically for this purpose, this report is also informed by other related work that has involved the research team concurrently – including Australian Homelessness Monitor 2020 and the evaluation of the Institute for Global Homelessness rough sleeping reduction program in Sydney.

• Western Australia (1) Respondent targeting prioritised individuals with senior and/or indepth experience of policymaking and/or implementation in relation to rental housing or homelessness. Interviews were undertaken online and recorded with interviewee permission.

1.4.2 Literature review

1.4.4 Quantitative analysis

As reported mainly in Chapter 3, this touched on the academic literature on crisis policymaking, agenda setting and focusing events. It also encompassed recently emerging rapid research findings on housing policy and housing system consequences of the 2020 pandemic. Especially given the recent nature of this crisis – one that was continuing to unfold at the time of writing in late 2020 – reference is also made to media reports of government actions and market developments.

Rental market analysis To generate an overview of evolving rental market conditions through the COVID-19 crisis, we analyse changes in rents for new tenancies, turnover of tenancies, and vacancy rates. We focus on New South Wales, Queensland and Victoria in particular. Key data sources include rental bonds data, released quarterly by state government agencies, as well as the SQM housing market data repository and the Real Estate Institutes of each state.

1.4.3 In-depth interviews

Analysis of state government records on rent reduction negotiations In collaboration with state government Fair Trading or Consumer Affairs departments, this element of the research involves monitoring the outcomes of mediation/conciliation processes where landlords and tenants utilise formal channels to resolve disputed claims for rent reduction.

Fieldwork underlying this research was informed by in-depth interviews with 20 expert government, industry and advocacy stakeholders and service providers as follows: • (State) Government housing policy official (3) • (State) Government consumer affairs official (1) • Homelessness services provider (3) • Community housing provider (1) • NGO/Housing/Homelessness peak (3) • Tenants representative (4) • Tenants advocate (2) • Refugee advocate (2) • Real estate agent (1)2 The geographical base/remit of the 20 respondents was as follows: • National (5)

• Queensland (3) • Victoria (3)

Rent reduction negotiations survey We conducted an online survey to collect data on experience of landlord-tenant negotiations on rent reduction to relieve affordability pressure resulting from tenant loss of income. Survey participants were recruited with the help of tenants unions. Participation was incentivized through entitlement to enter a prize draw for shopping vouchers, according to standard survey practice. In total 312 tenants participated in the survey. Although it was open to renters anywhere in Australia, 70% of respondents were from New South Wales. While this is a relatively small sample whose representativeness is not assured, data collected through its custom-designed format provide a valuable complement to larger more representative surveys from which findings are triangulated in Chapter 5. Data collection on homelessness emergency accommodation programs This aims to help quantify the extraordinary homelessness action undertaken by state governments in response to the pandemic. Focused on the four state governments concerned, it involves collection of basic statistics via a custom-designed pro forma. This was designed to complement the Specialist Homelessness Services (SHS) homelessness statistics routinely published by the Australian Institute of Health and Welfare (AIHW). It relates to both initial TA placements of former rough sleepers and other homeless ‘at risk’ populations into hotels and other safe premises, and government/NGO action in rehousing people out of TA into longer term housing.

2 We sought interviews with representatives of the Real Estate Institutes in two states; both declined, although one referred us to the real estate agent who participated. The agent was active in the REI as a member of its property management committee, but expressly did not participate as a representative of the REI.

COVID-19: Rental housing and homelessness policy impacts, 2021

1.5 Report structure Following this introductory chapter, and expanding on Executive Summary Table A, Chapter 2 presents a factual synopsis of the key income support, rental housing and homelessness policy innovations that emerged in response to the pandemic, both in Australia and – more briefly – in the UK and the three other comparator countries. Next, to frame our primary research on the formulation and implementation of these measures, Chapter 3 probes existing literature on policymaking within a crisis context. It also reviews published evidence on pandemic policy initiatives and impacts (primarily in Australia) as this is emerging from research by others. Next, in Chapter 4, we present an analysis of rental housing market conditions during 2020 to explore the initial demand and supply impacts of the pandemic as these are revealed by indicators such as asking rents, vacancy rates and activity levels. Forming the heart of the report, Chapters 5-7 then detail the impacts of the key policy innovations initially triggered by the COVID-19 pandemic with respect to rental housing and homelessness, and explore their consequences. First, in Chapter 5 we analyse the impacts of the 2020 crisis on existing tenancies – with a particular focus on rent reduction negotiations involving renters who had lost income due to the economic effects of the pandemic. Here, we draw on the work of housing research colleagues elsewhere in Australia, as well as on our own survey of landlord-tenant rent reduction negotiation outcomes (see Section 1.4.4)

Farha, L. (2020) ‘Housing, the front line defence against the COVID-19 outbreak says UN expert’, media release, 18 March; UNOHCHR https://www.ohchr.org/ EN/NewsEvents/Pages/DisplayNews.aspx?NewsID=25727&LangID=E. Jordà, Ò., Schularick, M., Taylor, A. M., Ellison, M., & Pappa, E. (2016) The great mortgaging: housing finance, crises and business cycles, Economic Policy, 31, 85, 107-152 Mannakkara, S. & Wilkinson, S. (2014), “Re-conceptualising “Building Back Better” to improve post-disaster recovery”, International Journal of Managing Projects in Business, vol. 7 no. 3: 327-341 Pawson, H., Parsell, C., Liu, E., Hartley, C., & Thompson, S. (2020a) Australian Homelessness Monitor 2020; Launch Housing https://www.launchhousing.org. au/ending-homelessness/research-hub/australian-homelessness-monitor-2020 Pawson, H., Milligan, V., & Yates, J. (2020b) Housing Policy in Australia: A case for system reform; Palgrave Macmillan Pawson, H., Parsell, C., Saunders, P., Hill, T., & Liu, E. (2018) Australian Homelessness Monitor 2018; Launch Housing https://cms.launchhousing.org.au/ app/uploads/2018/05/LaunchHousing_AHM2018_Report.pdf Yates, J. (2016) Why Does Australia Have an Affordable Housing Problem and What Can Be Done About It?, Australian Economic Review, 49, 3, 328-339

Chapter 6 focuses on the formulation and terms of rental eviction moratoriums, while Chapter 7 investigates the emergence and implementation of emergency accommodation programs for rough sleepers and other homeless people at risk during the pandemic. In both these chapters our key concern is the policymaking process, with our interpretations primarily informed by in-depth interviews as detailed above (see Section 1.4.3). Finally, ahead of our conclusion, Chapter 8 summarises comparable UK pandemic rental housing and homelessness policy developments as identified by CaCHE colleagues. Finally, in Chapter 9, some brief conclusions and next steps are discussed.

References Australian Bureau of Statistics, Household Income and Wealth 2017-18 (Catalogue No 6523DO008, 25 July 2019), Table 8.3. Birkland, T. (1998) Focusing events, mobilization and agenda setting, Journal of Public Policy, 18, 1, 53-74 Cassells, R. and Duncan, A. (2020) JobKeepers and JobSeekers: How many workers will lose and how many will gain? Curtin BankWest Research Centre Research Brief COVID-19 https://bcec.edu.au/publications/job-keepers-andjob-seekers-how-many-workers-will-lose-and-how-many-will-gain/ Conley, T. (2018) The stretched rubber band: banks, houses, debt and vulnerability in Australia, Australian Journal of Political Science, 53, 1, 40-56 Dodson, J., de Silva, A., Dalton, T. and Sinclair, S. (2017) Housing, multi-level governance and economic productivity, Final Report no. 284, Melbourne: AHURI

COVID-19: Rental housing and homelessness policy impacts, 2021

2. Crisis policy innovations in response to COVID-19 Key points: • Pandemic policy responses relevant to rental housing and homelessness in Australia included income support payments, access to superannuation savings and facilitation of mortgage payment deferrals, as well as more specifically focused measures: in particular, restrictions on tenant evictions and extraordinary assistance to street homeless populations. • In different forms and at different scales, 2020 saw approximately parallel actions enacted in all of our comparator countries (Canada, New Zealand, the United Kingdom and the United States). • As in Australia, responsibilities for housing and homelessness being effectively shared across different levels of government under federal arrangements posed substantial co-ordination and implementation challenges in Canada and – especially – the United States. To flesh out the concise listing presented in Executive Summary Table A, this chapter presents a synopsis of the key policy innovations relevant to rental housing and homelessness enacted in Australia in response to the 2020 crisis. The chapter is structured in three main sections: • National income support programs • Rental housing regulation innovations • Homelessness policy responses As well as summarising Australian policy innovations under each of these headings, each section also more briefly outlines approximately equivalent measures implemented in 2020 in a small range of anglophone countries – Canada, New Zealand, the USA and the UK. The second and final report to be published from this research (mid 2021) is intended to incorporate a more substantial international comparative component, encompassing a wider range of countries. Some of Australia’s rental housing and homelessness measures are further elaborated and analysed (e.g. to highlight inter-jurisdictional variations) in Chapters 5-7 of this report.

2.1 COVID-19 national income support programs 2.1.1 Australia

Photo © Andrew Wilson, used under licence from Austockphoto

COVID-19: Rental housing and homelessness policy impacts, 2021

While rental housing and homelessness measures are the main focus of this report it was income protection policy innovation that provided a fundamental underpinning for these former responses. As listed in Executive Summary Table A, the two key initiatives here were the Coronavirus Supplement and JobKeeper – announced by the Australian Government early in the emergency. 27

The Coronavirus Supplement is a supplement to JobSeeker Payment (formerly Newstart Allowance), Australia’s unemployment benefit. It was also applied to Youth Allowance and Parenting Payment. Announced by the Commonwealth Government on 22 March 2020, the Coronavirus Supplement was initially in the amount of $550 per fortnight – effectively doubling the amount received by most JobSeeker recipients. Distinct from parallel policy developments in some other countries (e.g. New Zealand – see below), the Coronavirus Supplement entitlement extended to all recipients of the designated programs – rather than being restricted to those made unemployed due to the public health crisis. The specific justification for the Coronavirus Supplement remained somewhat ambiguous, being at times described by Ministers in terms of the humane need to cushion individuals affected by sudden job loss, and at other times being referenced as if its primary purpose was as a Keynesian economic demand stimulant. Announced by the Prime Minister on 30 March 2020, the JobKeeper Payment was a wholly new wage subsidy, chiefly justified in terms of the need to ensure retained connectivity between pandemic-affected employers and their workers. It is paid to eligible employers and sole traders, rather than through the social security system. JobKeeper was initially a $1,500-per-fortnight payment for each full-time and part-time employee in businesses suffering a reduction in turnover in excess of 30% (50% for businesses with a turnover of over $1 billion). Non-qualifying under the scheme have been casual workers with less than 12 months service to their current employer – a group estimated as numbering 950,000 (Cassells & Duncan, 2020). Moreover, Jobkeeper also excluded most non-permanent residents – an estimated 1.3 million people (ACOSS, 2020), most of whom also lack eligibility for mainstream social security payments. Probably in large measure thanks to the structure and initial rates of JobKeeper and the Coronavirus Supplement, the average income of the 10% of Australian households with the lowest incomes (decile 1) increased in the initial months of the pandemic (Biddle et al., 2020). Similarly, it is estimated that, when housing costs are taken into account, 2.6 million people were experiencing poverty in June 2020; whereas, in the absence of the special income protection measures this figure would have been 5.8 million (Phillips et al., 2020). Leishman et al. (2020) similarly estimate that 1.3 million households would have been living in housing affordability stress without the income support interventions, though they note a rise in housing affordability stress from 758,000 to 861,500 households under JobKeeper and JobSeeker Coronavirus Supplement at their initial rates. Or, to put this another way, emergency income supports successfully mitigated the pandemic-triggered increase in housing affordability stress so that numbers affected initially rose by only 14%, rather than 76%. Both the Coronavirus Supplement and JobKeeper were initially envisaged as operating for six months from April to September 2020 (inclusive). Subsequently, while the programs were retained, payment rates were cut. JobKeeper beneficiaries saw their fortnightly payment reduced to $1,200 for the period September-January, and to $1,000 from January-March 2021. The Coronavirus Supplement was cut to $250 from September-December 2020. Modelling by Phillips et al. (2020) suggested that, combined with the coincidental JobKeeper cut, this change would push more than 700,000 people into poverty. The supplement was further reduced to $150 per fortnight from 1 January-31 March 2021. 28

COVID-19: Rental housing and homelessness policy impacts, 2021

One additional notable associated policy innovation enacted by the Australian Government in March 2020 involved allowing the withdrawal (‘early release’) of up to (a total of) $20,000 in superannuation savings in tax years 2019-20 and 2020-21. Announced the same day as the Coronavirus Supplement, this scheme was intended to ‘assist [people] in dealing with the adverse economic impact of the pandemic’ (ABC News, 2020), and eligibility was to be restricted to people out of work or having otherwise experienced a significant cut in earnings due to the pandemic. By August 2020 some 2.7 million people had made use of this provision, with withdrawals totalling $31 billion (ibid). 2.1.2 Comparator countries Emergency income protection programs of one kind or another were enacted as a pandemic response by all our comparator countries – Canada, New Zealand, the UK and the USA. As in Australia, this has typically featured wagereplacement measures (paid via employers) backed by supplementary social security payments (administered through the income support system). However, measures have differed significantly in their value, breadth of coverage and duration.

Canada Relevant measures enacted by the Canadian Federal Government included: • Canada Emergency Response Benefit (CERB) – a CAN$500 per week taxable payment, payable for 24 weeks to workers losing employment due to the pandemic • Canada Emergency Wage Subsidy (CEWS) for employees of companies with reduced revenues – 75% of wages paid, up to CAN$847 per week – payable until December 2020 (Parliamentary Budget Office of Canada, 2020) Importantly, unlike Australia, these measures were not restricted to Canadian citizens.

New Zealand In New Zealand, emergency income protection programs were implemented as follows: • COVID-19 Income Relief payment, a NZ$490 per week payment (for a maximum of 12 weeks) – eligibility limited to those losing full-time jobs due to the pandemic • Wage subsidy scheme payable to employers losing income due to the pandemic – initially for 3 months, subsequently extended by 10 weeks, and a further 2 weeks to cover those impacted by Auckland’s second lockdown in August. Since it extended only to ‘new unemployed’, the Income Relief payment was characterised as ‘a state-funded pandemic redundancy package’ (Fletcher, 2020). From this perspective it ‘create[d] a massive inequity between those who qualify and those who must rely on standard welfare benefits and welfare eligibility rules’ (ibid). The former received around double the entitlement of the latter.

United Kingdom In the UK, the Furlough Scheme enacted at the start of the pandemic involved a government payment to qualifying employees at 80% of previous earnings, subject to a cap set at just above median earnings, and conditional on employer-retention of workers on their payroll. From October 2020 a replacement program, the Job Support Scheme (payments at 67% of previous earnings) was instituted to assist ‘employees in firms forced to close because of national or local restrictions’ (Handscomb et al., 2020, p. 1).

Regarding the eviction moratoriums, the approach of most states and territories was to implement a set of restrictions on termination for a narrow class of COVID-19 affected tenants (defined differently in each jurisdiction), and a second less onerous set of restrictions of wider application. The exception is Tasmania, where one set of onerous restrictions applied generally; the NT also distinguished between tenancies already in existence at the commencement of its emergency provisions (for which it made limited provision) and new tenancies (which could be subject to more onerous protection).

Alongside its wage replacement schemes, the UK Government’s pandemic response also included boosted social security payment rates and alterations and relaxation of restrictions – at an initial overall cost of £7 billion (Partington, 2020). Some of these measures were announced as explicitly temporary in nature. In others there was less clarity. One important such instance involved the substantial uplift in Housing Benefit (HB) payments, whereby HB Local Housing Allowance rates were restored to the 30th percentile market rent in the local housing market concerned. However, while this was later stated as ‘permanent’ (Hobson et al., 2020), a further subsequent announcement gave notice that the relevant budget would be once more frozen from 2021 (HM Treasury, 2020).

For the COVID-19-affected group, rent arrears terminations were stopped in South Australia, Tasmania, Victoria, Western Australia and the ACT; and have been subject to increased notice requirements and scrutiny by the tribunal in NSW, Queensland and the NT. Also for this group, no-grounds terminations are stopped in Queensland, Tasmania, Victoria and Western Australia; and are subject to increased notice requirements and scrutiny by the tribunal in New South Wales and the NT, and increased scrutiny (only) in South Australia, including consideration of the objective of avoiding homelessness.

United States Finally, in the USA, the March 2020 Coronavirus Aid, Relief and Economic Security (CARES) Act included two major temporary income protection measures, implemented at the national scale: • A $600 weekly unemployment benefit to supplement state payments • The Paycheck Protection Program which provided employers with ‘forgivable loans’ on condition that they retain staff and continue to pay wages. However, the CARES Act jobless benefit rate and the PPP program had expired by August, succeeded only by a substantially lower unemployment insurance payment which ran for only a matter of weeks (Bertolet & Gabobe, 2020). Subsequent pre-Presidential election efforts to institute a new compendium economic relief package were blocked by the US Senate.

2.2 Rental housing regulation innovations in the 2020 crisis 2.2.1 Australia On 29 March 2020, the National Cabinet announced that there would be a six-month moratorium on evictions for commercial and residential tenants in financial hardship because of COVID-19, and encouraged landlords and tenants to negotiate about rent payment obligations. Within days, the National Cabinet drafted a ‘mandatory code of conduct’ for commercial tenancies. Over the month of April, each state and territory government formulated and implemented its own moratorium for residential tenancies, and introduced measures regarding rents. For jurisdictions in which, for decades, residential tenancies laws had allowed ready termination of tenancies and rents set by the market, these were extraordinary measures. The broad outlines of these two sets of measures are presented here, with a more detailed discussion to come in Chapter 6.

COVID-19: Rental housing and homelessness policy impacts, 2021

For tenants generally, most grounds for termination were stopped or subject to greater procedural scrutiny in Tasmania, Western Australia and Victoria; in NSW, some grounds for termination were subject to longer notice periods. No jurisdiction, therefore, implemented a complete moratorium on evictions, and some continued to allow numerous types of termination proceedings – including, in NSW, Queensland and the NT, rent arrears terminations against COVID-affected tenants. Regarding rent measures, states and territories implemented one or more of the following three measures regarding rents: • Prohibitions on rent increases (Tasmania, SA, Victoria and WA; the ACT prohibited rent increases for COVID-affected tenants only) • Rent variation provisions, ranging from informal encouragement (Tasmania, SA, ACT and NT) to formal conciliation processes (NSW, Victoria, Queensland and WA), with Victoria also providing for rent variations determined by the state’s Civil and Administrative Tribunal (Queensland, SA and NT also provide for determined variation in very limited circumstances) • Rent relief programs, in the form of cash payments to landlords and/or land tax rebates where parties had made a rent variation agreement. All jurisdictions originally set a six-month timeframe on their emergency provisions, but all except Queensland later extended them: to 31 January for Tasmania and the ACT; 6 February for SA; 23 March for the NT, 26 March for NSW; 28 March for Victoria and WA. An important backdrop to these measures is the program of housing loan deferrals that commenced late March. From 20 March 2020, Australian banks offered home and business borrowers in hardship deferrals of up to six months, underpinned by the very-low cost finance extended to banks by the Reserve Bank of Australia’s Term Funding Facility (established 19 March), and changes by the Australian Prudential Regulatory Authority to requirements regarding impaired loans (RBA, 2020; APRA, 2020a). Housing loan deferrals peaked at the end of May 2020, when 11% of all housing loans were deferred, with loans 31

to landlords representing more than one-third of all housing loans deferred (APRA, 2020b). 2.2.2 Comparator countries

Canada In Canada, eviction moratoriums were introduced across all provinces early in the pandemic. While these had lapsed by September 2020, some associated restrictions remained in place. In British Columbia, for example, landlords were required to allow tenants until July 2021 to repay accumulated arrears, provided that monthly payments were made in the interim. Up to $500 per month in rent assistance had also been made available to tenants by the provincial government from March-August (Britneff, 2020; Zussman, 2020). At the same time, the Canadian Mortgage and Housing Corporation (CMHC) undertook to provide mortgage forbearance for ‘borrowers that have a CMHC financed loan in good standing’. Landlords benefiting from such assistance were expected to ‘extend support to renters and households so that their housing needs continue to be met’ (CHRA, 2020).

New Zealand A national eviction moratorium and rent freeze was enacted in New Zealand in March 2020. With only very limited exceptions (anti-social behaviour), evictions were banned for three months and rent increases outlawed for six months. The block on evictions even applied to instances where landlords wished to move into the property themselves – because of an intent ‘to ensure as little disruption as possible and keep everyone in stable housing regardless of tenure’ (Wade, 2020). United Kingdom In the UK, residential tenancies in England and Wales were subject to a moratorium on possession actions from March-September 2020. While possession cases were allowed to resume from September, a six months’ notice requirement for evictions (except in cases involving anti-social behaviour or domestic abuse) was also instituted. Similar measures were put in place by the Scottish Government, although with the ban on new possession actions extended in September 2020 to March 2021 (Wilson, 2020). As in most Australian jurisdictions, UK COVID-19 restrictions on the initiation of possession actions, and extended termination notice periods applied to councils and housing associations (i.e. social housing) as well as to private landlords. In tune with this, the representative body for housing associations in England pledged at the start of the pandemic that ‘No one should be evicted [from a not-for-profit housing tenancy] because of the coronavirus’ (National Housing Federation, 2020). Subsequently, however, the Scottish Government’s pledge to extend Scotland’s moratorium to March 2021 triggered some protest at the continued inclusion of social housing within the restrictions (Bookbinder, 2020).

United States Meanwhile, in the USA, the CARES Act also brought in a 3-month national eviction moratorium from March 2020. This covered an estimated 19.3 million of the 43 million renter households (45%) living in dwellings covered by federally assisted programs – including federally-backed mortgages and Housing Choice Vouchers (Congressional Research Service, 2020). 32

COVID-19: Rental housing and homelessness policy impacts, 2021

A further national eviction moratorium was imposed by the Centres for Disease Control (CDC) and Prevention on 4 September for the remainder of 2020 ‘to prevent the further spread of COVID–19’. This was justified in medical terms as follows: ‘Evicted renters must move, which leads to multiple outcomes that increase the risk of COVID–19 spread. Specifically, many evicted renters move into close quarters in shared housing or other congregate settings’ (Centers for Disease Control and Prevention, 2020). However, researchers raised many concerns about the practicability of this measure (American Prospect, 2020). Alongside national, Federal Government-initiated action, 43 states and Washington DC ‘enacted some kind of eviction moratorium’, although ‘many had little to no enforcement mechanism’ (Dougherty, 2020).

2.3 Homelessness policy responses to the pandemic 2.3.1 Australia Homelessness policy responses to the COVID-19 pandemic have concentrated primarily on people sleeping rough, or otherwise on the fringes of Australia’s urban ‘street homeless’ population in early 2020 at the start of the crisis. Of particular concern within this latter group have been those accommodated in buildings with shared bedrooms/dormitories or other facilities. From late March, focusing on these groups, four state governments activated programs to rapidly move as many people as possible into safe temporary accommodation. For the most part this involved large scale hotel bookings. The governments concerned – NSW, Queensland, South Australia and Victoria – authorised substantial extraordinary funding to meet associated costs including: • Hotel charges • Meals • Floating support for hotel-housed residents provided by contracted NGOs. The initial ‘get them safely housed’ impetus had subsided by June 2020 and hotel placements had begun to dwindle. However, the second wave COVID-19 outbreak in Melbourne from July 2020 prompted a further tranche of hotel bookings enabled by renewed funding from the Victorian Government. Beyond the provision of temporary accommodation (for days, weeks or months), governments also initiated extraordinary homelessness policy action in efforts to enable (some of) the hotel homeless to transition into secure housing. For example, a package of measures announced by the Victorian Government in July 2020 was motivated by the aim that ‘2,000 [hotel-housed] Victorians are supported to access stable, long term housing’ (Victorian Government, 2020). Programs of this kind encompassed funding for: • Private rental subsidies enabling individuals to bridge the gap between rental charges and an affordable portion of social security (or other) income • Rent payments to private landlords willing to ‘head lease’ dwellings to notfor-profit housing providers to accommodate former homeless people • Staff costs incurred by contracted NGOs in providing floating support for former homeless people placed in private rental or head-leased accommodation. 33

According to Mason et al. (2020) expenditure commitments associated with these activities had already totalled $192 million by June 2020, although – especially in Victoria, for reasons explained above – the eventual full cost is likely to have been far higher. Parsell et al. (2020) estimate that, by September 2020, $229 million had been committed to extraordinary homelessness expenditure by the five mainland state governments. For a sense of scale, annual state/territory government homelessness services expenditure in 201819 totalled $990 million (Productivity Commission, 2020). 2.3.2 Comparator countries In all four of our comparator countries the onset of COVID-19 in around March 2020 triggered rapid and assertive action to temporarily accommodate homeless people on an extraordinary scale. Consistent with national governance structures, these were more centrally directed and monitored in New Zealand and the UK than in Canada and the USA. In contrast with Australia, however, in all four countries, national governments materially contributed to such programs through designated funding.

Canada A number of Canadian cities and provinces (including Toronto, Montreal, British Columbia and Saskatchewan) moved early in the pandemic to facilitate safe emergency shelter for homeless people. This included booking hotel rooms, acquiring rental buildings and opening additional shelters (Lee & Hamidian, 2020). It is estimated that 10,000 rough sleepers and other homeless people were accommodated under these programs (Pomeroy, 2020). To underpin support to homeless persons, the Federal Government provided $157 million in April 2020 for activities such as temporary accommodation procurement, personal protective equipment for staff, and associated additional salary costs (CHRA, 2020). In September the Federal Government committed a further $237 million ‘to help extend … the emergency measures that have been successful in reducing the risk of potential outbreaks among people experiencing homelessness, as well as provide them the flexibility to deliver affordable housing solutions’ (CMHC, 2020).

New Zealand The New Zealand Government also initiated rapid action to provide temporary housing for rough sleepers and other homeless people at the start of the pandemic. An emergency funding allocation of NZ$108 million underpinned booking of 1,600 hotel and motel rooms through to April 2021, as well as associated support services ‘to enable people to stay housed until [the NZ Government] can secure more long-term housing supply’ (NZ Government, 2020). Thus, by 22 May, the NZ Herald headlined its story: ‘Covid 19 coronavirus: Rough sleeping all but eliminated in New Zealand as pandemic crisis frees up motels and housing’ (Davison, 2020a). However, by late August, it was reported that Auckland CBD street homelessness numbers were once again on the increase (Davison, 2020b). United Kingdom In the UK, the national government’s ‘Everyone In’ announcement of 26 March instructed local authorities in England to move everyone sleeping rough and in communal shelters into a safe place, ideally in self-contained accommodation, over the following two days. Dormitory-style homelessness provision was subsequently taken out of use (Lewer et al., 2020). 34

COVID-19: Rental housing and homelessness policy impacts, 2021

The EI initiative has been described as ‘a bold and unprecedented move that inspired confidence on the part of both local authorities and their voluntary sector partners that they had the Government’s backing to take decisive action’ (Fitzpatrick et al., 2020, p. 3). Particularly commended was the clear Ministerial signal that in crisis circumstances, standard eligibility requirements for housing assistance – status under the homelessness legislation and in relation to citizenship – could be set aside. A subsequent official survey showed that the Everyone In initiative had, by May 2020, provided emergency accommodation to 14,610 ‘people sleeping rough or at risk of sleeping rough’ (Ministry of Housing Communities & Local Government, 2020). As further detailed in Section 8.2.1, by late 2020, government allocated funding specifically earmarked for emergency accommodation and move-on housing had totalled some £296 million (or AUD$500 million).

United States Comparable pandemic-triggered action on homelessness in the USA was mainly initiated and undertaken by local Consortiums of Care (CoCs) involving municipalities and local NGOs. Such action was, in part, enabled through funding allocated under the CARES Act which provided some $12 billion for Housing and Urban Development (HUD) programs, including $9 billion that could be utilised for emergency action on homelessness. Part of this funding was specifically intended to ‘help prevent [a COVID-19] outbreak among … people experiencing homelessness and very low-income households … at risk of homelessness’ (National Low Income Housing Coalition, 2020). Thus, also utilising funding from state and local governments, as well as from philanthropy, CoCs expanded accommodation provision in the initial pandemic phase. Most of this involved hotel/motel rooms – procured in part to facilitate quarantine for homeless people testing positive for COVID-19. In line with CDC guidance, older adults and those with pre-existing medical conditions were typically prioritised for such placements. Additionally, to maintain homelessness shelter capacity while facilitating social distancing, some providers kept winter accommodation open through the summer (Moses, 2020; Rice et al., 2020). More generally, in the absence of any nationally directed program of the kind seen in England, there was no equivalent policy to close congregate accommodation.

References ABC News. (2020) Paul Keating says people are being allowed early access to their super with no scrutiny whatsoever. Is he correct?; RMIT ABC Fact Check https://www.abc.net.au/news/2020-09-03/-are-people-being-allowed-toaccess-their-super-without-scrutiny/12618002 ACOSS. (2020) Who misses out on access to Australia’s safety net under Covid19? ; ACOSS https://www.acoss.org.au/wp-content/uploads/2020/04/ Who-misses-out-Briefing-ACCESSIBLE-FINAL.pdf American Prospect. (2020) Federal “Eviction Moratorium” Is a Leaky Boat, 2 September https://americanprospect.activehosted. com/index.php?action=social&chash=979d472a84804b9f647bc 185a877a8b5.219&s=bad97c655476f96a390a72c05a742011

Australian Prudential Regulatory Authority (2020a) ‘Letter to authorised deposit-taking institutions: Treatment of loans impacted by COVID-19 – Thursday 9 July 2020’ https://www.apra.gov.au/letter-to-authorised-deposittaking-institutions-treatment-of-loans-impacted-by-covid-19

Dougherty, C. (2020) Millions of Evictions Are a Sharper Threat as Government Support Ends; New York Times 7 August https://www.nytimes. com/2020/08/07/business/economy/housing-economy-eviction-renters.html

Australian Prudential Regulatory Authority (2020b) ‘Temporary loan repayment deferrals due to COVID-19’ https://www.apra.gov.au/temporary-loanrepayment-deferrals-due-to-covid-19-september-2020-accessible-version

Fitzpatrick, S., Watts, B., & Sims, R. (2020) Homelessness Monitor England 2020: COVID-19 Crisis Response Briefing; Crisis https://www.crisis.org.uk/ media/242907/homelessness_monitor_england_2020_covid19_crisis_ response_briefing.pdf

Bertolet, D., & Gabobe, N. (2020) Will the next US Federal aid package rescue renters? Trouble looms for renters unless Congress acts fast to re-up emergency benefits; Sightline Institute 20 July https://www.sightline.org/2020/07/20/ will-the-next-us-federal-aid-package-rescue-renters/

Fletcher, M. (2020) Kindness doesn’t begin at home: Jacinda Ardern’s support for beneficiaries lags well behind Australia’s; The Conversation 3 June https:// theconversation.com/kindness-doesnt-begin-at-home-jacinda-ardernssupport-for-beneficiaries-lags-well-behind-australias-139387

Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2020) COVID-19 and mortgage and rental payments: May 2020 ANU Centre for Social Research and Methods https://csrm.cass.anu.edu.au/sites/default/files/docs/2020/6/COVID-19_and_ housing_FINAL.pdf

Handscomb, K., Pacitti, C., Slaughter, H., & Tomlinson, D. (2020) Back to the furlough: U-turn to retain furlough scheme in closed sectors paves way for fresh lockdowns; Resolution Foundation https://www.resolutionfoundation.org/ publications/back-to-the-furlough/

Bookbinder, D. (2020) Eviction protections hit the wrong target in social housing, and its tenants who’ll pay the cost; Scottish Housing News 18 August https://www.scottishhousingnews.com/article/david-bookbinder-evictionprotections-hit-the-wrong-target-in-social-housing-and-its-tenants-who-ll-paythe-cost

HM Treasury (2020) Spending Review 2020 https://www.gov.uk/government/ publications/spending-review-2020-documents/spending-review2020#economic-and-fiscal-context-1

Britneff, B. (2020) As coronavirus eviction bans end, advocates worry homelessness will rise; Global News 14 August https://globalnews.ca/ news/7268526/coronavirus-renters-homeless-survey/ Cassells, R., & Duncan, A. (2020) JobKeepers and JobSeekers: How many workers will lose and how many will gain? ; Curtin BankWest Research Centre https://bcec.edu.au/publications/job-keepers-and-job-seekers-how-manyworkers-will-lose-and-how-many-will-gain/ Centers for Disease Control and Prevention. (2020) Temporary Halt in Residential Evictions To Prevent the Further Spread of COVID–19. https://www. govinfo.gov/content/pkg/FR-2020-09-04/pdf/2020-19654.pdf CHRA. (2020) COVID-19 response page. https://chra-achru.ca/news/covid-19updates/ CMHC. (2020) ‘Canada to Rapidly Create Affordable Housing and Support the Homeless’, media release, 21 September; https://www.cmhc-schl.gc.ca/ en/media-newsroom/news-releases/2020/canada-rapidly-create-affordablehousing-support-homeless Congressional Research Service. (2020) CARES Act Eviction Moratorium. https://crsreports.congress.gov/product/pdf/IN/IN11320

Hobson, F., Kennedy, S., & Mackley, A. (2020) Coronavirus: Withdrawing crisis social security measures; House of Commons Library http://researchbriefings. files.parliament.uk/documents/CBP-8973/CBP-8973.pdf Lee, M., & Hamidian, A. (2020) Comparing provincial economic responses to COVID-19; Policy Note 23 April https://www.policynote.ca/provincialresponses-covid/ Leishman, C., Ong, R., Lester, L., & Liang, W. (2020) Supporting Australia’s housing system: modelling pandemic policy responses; Australian Housing and Urban Research Institute https://www.ahuri.edu.au/research/final-reports/346 Lewer, D., Braithwaite, I., Bullock, M., Eyre, M. T., White, P. J., Aldridge, R. W., . . . Hayward, A. C. (2020) COVID-19 among people experiencing homelessness in England: a modelling study, The Lancet Respiratory Medicine https://doi. org/10.1016/S2213-2600(20)30396-9 Mason, C., Moran, M., & Earles, A. (2020) Policy coordination and housing outcomes during COVID-19; Australian Housing and Urban Research Institute https://www.ahuri.edu.au/research/final-reports/343 Ministry of Housing Communities & Local Government. (2020) Coronavirus (COVID-19) emergency accommodation survey data: May 2020. Coronavirus (COVID-19) emergency accommodation survey data: May 2020

Davison, I. (2020a) Covid 19 coronavirus: Rough sleeping all but eliminated in New Zealand as pandemic crisis frees up motels and housing; NZ Herald 22 May https://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=12333758

Moses, J. (2020) Responding to COVID-19: Conversations with Homeless System Leaders; National Low Income Housing Coalition https:// endhomelessness.org/wp-content/uploads/2020/08/Promising-Practices-Pubclean-with-logos-.pdf

Davison, I. (2020b) Covid 19 coronavirus: Auckland charities bracing for second wave of homelessness; NZ Herald 27 August https://www.nzherald.co.nz/nz/ news/article.cfm?c_id=1&objectid=12359788

National Housing Federation. (2020) No one should lose their home because of coronavirus. https://www.housing.org.uk/news-and-blogs/news/no-oneshould-lose-their-home-because-of-coronavirus/

COVID-19: Rental housing and homelessness policy impacts, 2021

National Low Income Housing Coalition. (2020) Congressional Leaders Agree to Coronavirus Response Package with Funding for Homelessness and Housing. https://nlihc.org/resource/congressional-leaders-agree-coronavirus-responsepackage-funding-homelessness-and-housing NZ Government. (2020) ‘Keeping vulnerable people housed’, media release, 26 April; https://www.beehive.govt.nz/release/keeping-vulnerable-people-housed Parliamentary Budget Office of Canada. (2020) Costing of Canada’s COVID-19 response plan. https://www.pbo-dpb.gc.ca/en/covid-19 Parsell, C., Clarke, A., & Kuskoff, E. (2020) Understanding responses to homelessness during COVID-19: an examination of Australia, Housing Studies, 1-14 doi:10.1080/02673037.2020.1829564 Partington, R. (2020) Millions in UK ‘could slip through virus wage safety net’; The Guardian 2 April https://www.theguardian.com/business/2020/apr/02/ millions-uk-slip-through-coronavirus-pay-wage-safety-net-self-employedwage-subsidy Phillips, B., Gray, M., & Biddle, N. (2020) COVID-19 JobKeeper and JobSeeker impacts on poverty and housing stress under current and alternative economic and policy scenarios; ANU Centre for Social Research Methods https://csrm. cass.anu.edu.au/research/publications/covid-19-jobkeeper-and-jobseekerimpacts-poverty-and-housing-stress-under Pomeroy, S. (2020) Recovery for All: Proposals to Strengthen the National Housing Strategy and End Homelessness; Canadian Alliance to End Homelessness https://caeh.ca/recovery-for-all-report/ Productivity Commission. (2020) Report on Government Services 2020: Part G Housing and Homelessness – Table 19A.1 https://www.pc.gov.au/research/ ongoing/report-on-government-services/2020/housing-and-homelessness/ homelessness-services/rogs-2020-partg-section19-data-tables.xlsx Reserve Bank of Australia (2020) ‘Term Funding Facility’ https://www.rba.gov. au/mkt-operations/term-funding-facility/overview.html Rice, E., Treglia, D., Culhane, D., Moses, J., & Janosko, J. (2020) Community-level Responses of Homelessness Assistance Programs to COVID-19: Data from May 2020; National Low Income Housing Coalition https://endhomelessness.org/ wp-content/uploads/2020/05/Community-COVID-Survey-5.28.2020-Final.pdf Victorian Government. (2020) ‘Homes For Homeless Victorians During Pandemic And Beyond’, media release, 28 July; https://www.premier.vic.gov.au/ homes-homeless-victorians-during-pandemic-and-beyond Wade, A. (2020) Coronavirus Covid-19 lockdown: Rents frozen and evictions banned under emergency legislation; NZ Herald 25 March https://www. nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=12319849 Wilson, W. (2020) Coronavirus: Support for landlords and tenants; House of Commons Library http://researchbriefings.files.parliament.uk/documents/CBP8867/CBP-8867.pdf Zussman, R. (2020) B.C. eviction ban to end Sept. 1, tenants have until July 2021 to pay unpaid rent; Global News 16 July https://globalnews.ca/news/7185583/ bc-eviction-ban-lifted-sept-1-unpaid-rent-2021/ 38

COVID-19: Rental housing and homelessness policy impacts, 2021

3. Policy innovation in crisis conditions Key points: • Crises such as the COVID-19 pandemic have the potential to serve as ‘focusing events’ – opportunities for ‘thinking outside the box’, the contemplation of policy options previously considered unthinkable • The scope for ‘building back better’ in these circumstances may depend on how associated policy conundrums are defined and portrayed; interest groups or ‘advocacy coalitions’ potentially enjoy scope for novel framing of such challenges to advance their goals • In the homelessness policy realm, the extraordinary mass placement of rough sleepers and others into hotels early in the pandemic followed from the issue being re-framed as a public health concern for the community, rather than a problem largely impacting only on those directly concerned • While it has been suggested that Australian pandemic-triggered housing and homelessness policy innovations exhibited multi-level coordination, our own evidence suggests this claim is highly questionable.

3.1 Introduction This report focuses on the extraordinary housing and homelessness policy initiatives triggered by the COVID-19 pandemic during 2020. Within this, there is a particular interest in the policymaking processes involved. That is, the formulation, elaboration and implementation of these measures amid a public health crisis and its associated economic devastation. Therefore, in framing the domain-specific investigation that forms the heart of our report, this chapter begins in Section 3.2 by reviewing broader political science thinking on the way that crisis conditions and ‘focusing events’ (Birkland, 1998) can have a profound impact on policy innovation and policy trajectories. Then, to contextualise our own empirical research on these developments, Section 3.3 reviews newly emerging evidence on these innovations as rapidly rolled out in Australia in response to the pandemic.

3.2 Crisis policy theory There are several bodies of literature dealing with policy-making during and after a crisis. On the one hand are perspectives relating to the opportunities crises bring for advocacy groups and other non-governmental actors to advance their agenda, largely in the realm of political science, public policy and geography, while on the other hand there is a more practically-focused literature on ‘building back better’. This latter canon is located in the realm of planning and disaster management and relates to how those within government and those on the ground can best reconstruct policy frameworks post-crisis. As such, these two loose groupings consider crisis policy-making from a critical outsider perspective (agenda setting and focusing events) and from an insider perspective (building back better). This section briefly outlines these two areas, following a discussion of the term ‘crises’. 39

3.2.1 Defining crises It is important to note that an event in and of itself is not a crisis; it requires interpretation and intersubjective agreement to give it this status (Foye et al., 2020). Different parties can interpret the event as varying in severity or impact, and indeed crises can be manufactured or ignored to serve political ends. Boin et al. (2009, p. 83) define crises as 'events or developments widely perceived by members of relevant communities to constitute urgent threats to core community values and structures', and argue that actors seek to frame events in ways that further their interests – as unfortunate incidents, or otherwise as critical threats or opportunities. Crises are therefore inherently subjective. From a building back better standpoint, crises (and specifically disasters) are 'unique circumstances, in which everything seems to move faster, information is constrained, money flows more freely, and new opportunities arise' (Kim & Olshansky, 2014, p. 289). They may also be seen as 'a "helpful interruption" to previously unchallenged inadequate policies and practices' (Fernandez & Ahmed, 2019, p. 1) – what Boin et al. (2009) would describe as a ‘critical opportunity’ framing. 3.2.2 Agenda setting The agenda setting literature considers how specific issues come to be prioritised in policy-making. Its origins can be traced to Cobb & Elder’s (1971) model of agenda building, and key current perspectives include the Advocacy Coalition Framework (Sabatier, 1988), the Multiple Streams Approach (Kingdon (1984)) and Punctuated Equilibrium (Baumgartner & Jones, 1993). Zahariadis (2016) argues that, regardless of the perspective taken to understand it, agenda setting entails ‘four 'P's’ leading to prioritisation: Power, Perception (of people), Potency and Proximity (of issue). Actors use power to shape perceptions of the crisis, framing its proximity and potency (Foye et al., 2020). The Advocacy Coalition Framework deals with how ‘people mobilize, maintain and act in advocacy coalitions’ (Weible et al., 2011, p. 349) and the factors influencing policy change under both business-as-usual and crisis conditions. It takes a decades-long view of policy change, and focuses on ‘policy subsystems’, defined as the ‘interaction of actors from different institutions interested in a policy area’ (Sabatier, 1988, p. 131). Together, these actors form ‘advocacy coalitions’ based on shared normative and causal beliefs as well as coordination of action (Sabatier, 1988; Weible et al., 2011). Importantly, advocacy coalitions include those in power and those with an insufficient majority, and members may be policymakers, researchers, activists or other parties. This means different advocacy coalitions are often set against each other. In a crisis, advocacy coalitions will strategically frame events to alter or support existing policy beliefs, engaging 'motivated reasoning' and belief-system defenses’ (Weible et al., 2011, p. 356). This leads to framing contests, where ‘coalitions pursue different 'causal stories' or 'crisis narratives' on the nature, causes, and severity of a crisis, as well as arguments related to responsibility for its occurrence or escalation’ (Nohrstedt, 2008, p. 261). This then leads to differing proposed solutions to the crisis. The Multiple Streams Approach posits that the separate streams of problems, policies and politics shape what comes to the fore, and when these streams combine around an issue (through ‘policy windows’ or the actions of ‘policy entrepreneurs’), it is likely to be prioritised. If policymakers believe they can do little about a particular problem such as housing, they are likely to 40

COVID-19: Rental housing and homelessness policy impacts, 2021

suppress attention to it or minimise the issue (Kingdon, 1984). The Punctuated Equilibrium model assumes policy attention focuses around particular problems and attention shifting is difficult, a ‘dynamic that benefits minor deviations and similar issues over time’ (Zahariadis, 2016, p. 13). Under normal circumstances, neglected issues become more serious and arouse more emotion until a ‘punctuation’ upsets the policy process and encourages a reset or innovation to address the neglected issue. This punctuation may be a macro body stepping in, or it may be an event. 3.2.3 Focusing events Within the agenda-setting literature, Birkland (1998) advanced the concept of ‘focusing events’, which ‘serve as important opportunities for politically disadvantaged groups to champion messages that had been effectively suppressed by dominant groups and advocacy coalitions’ (Birkland, 1998, p. 54). The Global Financial Crisis is an example of a focusing event in the Australian housing policy context, where the resulting Federal government economic stimulus included an extraordinary funding boost for social housing (Milligan & Pinnegar, 2010). However, as in much of the crisis-policy literature (Nohrstedt, 2008), policymakers are assumed to defend the status quo with the impetus for innovative responses coming from outside parties. Perhaps significantly in our particular context, Birkland (1998) points out that while a policy domain may become busier after a focusing event and have heightened public scrutiny, it may not necessarily change in substance. Similarly, Foye et al. (2020) identify a centralisation of power in immediate responses to crises, but note this centralisation may hinder policy innovation and collaboration. The nature of the focusing event and the existing network of advocacy groups within the relevant domain influences whether significant and lasting policy innovation results. If there is no organised advocacy focused on longer-term changes, responses will tend to be 'stop-gap measures and reflexive reactions to the immediate event' (Birkland, 1998, p. 67), focusing on relief efforts and not future mitigation. Focusing events can also harden the positions of all parties in regards to existing policy, increasing conflict and potentially obstructing change (Birkland, 1998). Focusing events can also provide opportunities for those already vested with relative power. Peck (2006) analyses the swift intervention of right wing neoliberal thinktanks in New Orleans post Hurricane Katrina, where 'the policy vacuum sucked in prefabricated conservative diagnoses of the American urban condition—the usual cocktail of race, crime, and dependency' (Peck, 2006, p. 694), greatly influencing the response. More widely, he advances a theory of ‘fast policy’, with policy solutions spreading rapidly between jurisdictions and implying convergence around apparently-neutral ‘best-practice’. However, 'the marketplace for social-policy ideas and models is deeply structured by powerful institutional actors and enduring asymmetries' (Peck, 2011, p. 172), with neoliberal ideas proliferating. The policy transfer literature (Dolowitz and Marsh, 2000) also has clear relevance here. Focusing events may also have little effect on enduring policy, depending on framing (Boin et al., 2009). Nohrstedt (2008) uses the ACF to analyse why the Chernobyl disaster did not lead to meaningful change in Swedish nuclear policy and accelerate nuclear decommissioning, despite conditions seemingly apt for it. In opposition to usual ACF assumptions, the government pursued a transparent, balanced inquiry into their existing policy and were ostensibly 41

open to change, though the inquiry findings did not lead to substantially revised policy. Nohrstedt (2008) argues that the findings did not sufficiently change the problem definition or challenge the government’s core beliefs, and that the inquiry was successful in characterising the government as 'trustworthy and efficient' (p.272), and so effective politically. 3.2.4 Building back better The term ‘building back better’ derives from former US President Clinton’s (2006) ‘lessons learned’ report from the 2004 Indian Ocean earthquake and tsunami, as the UN Secretary-General’s Special Envoy for Tsunami Recovery. The concept has had a lasting impact on disaster risk management practice, and over the past decade and a half academic interest has been growing (Fernandez & Ahmed, 2019), particularly in the domain of crisis management and resilience. Rather than the external advocacy perspective favoured in the agenda setting, focusing event and advocacy coalition literature, research on building back better considers crises from a policymaking and implementation perspective. Mannakkara and Wilkinson (2014, p. 328) observe that recovery efforts 'often focus on quick restoration … which can replicate and worsen existing vulnerabilities', and BBB aims to overcome this. However, in some instances, ‘building back better’ may simply be 'an engaging and attractive slogan', and even where parties are committed to following BBB principles 'success stories are rare' (Fernandez & Ahmed, 2019, p. 3). The definition of ‘better’ is also open to interpretation (Fernandez & Ahmed, 2019); while Kim and Olshansky (2014, p. 289) advise 'building back better, stronger, greener, and more equitably', the phrase allows users to co-opt the concept to advance their particular concerns. 3.2.5 Summary In sum, the perception of crises and responses to them owe a considerable amount to the framing employed by various actors, who may leverage these focusing events to advance their policy beliefs. Crises may also be seen as opportunities to ‘reset’ and fix weaknesses in the previous status quo, as encouraged by the philosophy of ‘building back better’. The necessary speed of responses forces innovation; Foye et al. (2020, p. 6) argue crises demand 'flexibility, improvisation, redundancy, and the occasional breaking of rules'. At the same time, it is important to acknowledge that it is very far from guaranteed that those rare instances of progressive crisis policymaking will bring about lasting systemic change or reform. Indeed, what was arguably the last pre-2020 occasion when an economic emergency prompted a radical departure in Australia’s housing policy – the Rudd Government’s social housing stimulus prompted by the 2008 Global Financial Crisis – pointedly failed to do so. No sooner than the moment of stress had passed, social housing construction reverted right back to the residual level of the previous decade, where it has remained ever since.

3.3 Crisis housing and homelessness policymaking 2020: emerging published evidence The rapid onset of the COVID-19 pandemic produced a sudden shift in attention towards mitigating the health and economic threat, with policy in unrelated areas paused (Weible et al., 2020). As decisions were made, it was unclear how long policies would need to be in effect, and which would be temporary (Weible et al., 2020). As early as 18 March, UN Special Rapporteur on the right 42

COVID-19: Rental housing and homelessness policy impacts, 2021

to housing Farha (2020) identified good practices in several jurisdictions, including ‘moratoriums on evictions due to rental and mortgage arrears; deferrals of mortgage payments for those affected by the virus; extension of winter moratoriums on forced evictions of informal settlements; and increased access to sanitation and emergency shelter spaces for homeless people’. However, she warned of ‘a risk that such measures will enable global financial actors to use the pandemic and the misfortunes of many to dominate housing markets without regard for human rights standards’. Research evidence on Australian housing and homelessness policymaking in the COVID-19 crisis had already begun to emerge in late 2020. This section reviews this early analysis and identifies gaps in understanding and coverage that the remainder of the report seeks to fill. 3.3.1 Evidence on rental housing policy impacts Mason et al. (2020) report that private rental support schemes announced March-June 2020 involved 21 policy initiatives underpinned by spending commitments totalling $1.21b – the notional total cost (including revenue foregone) of tenant subsidies and landlord land tax relief measures. The efficacy of these policies in ensuring market stability and positive outcomes for households is uncertain, however. As Leishman et al. (2020) note, the volatility of health and economic conditions in 2020 made it extremely difficult to adequately predict scenarios. In September, the Reserve Bank released an analysis of the impacts of COVID-19 on the rental market, noting that this market was disproportionately affected due to the characteristics of renters, who tend to be younger and working in industries most affected by crisis-related job losses (Evans et al., 2020). Renters also spend a greater proportion of their incomes on housing costs than owners (Evans et al., 2020), and are therefore likely to be less insulated from any drop in income. Evans et al. (2020, p. 78) state that rental policy measures in combination with income support ‘helped offset the acute fall in rental demand and stabilise the rental market’. Similarly, Mason et al. (2020, p. 9) argue that the Reserve Bank’s action in facilitating liquidity helped in ‘enabl[ing] banks to extend loan repayment holidays to … investors who have lost income due to the crisis (… which can in turn extend that flexibility to tenants)’. According to data from the Australian Prudential Regulatory Authority (APRA), 305,000 investor loans were in such deferral arrangements at the peak in May, implying about 12% of private rental properties were subject to loan deferrals. By comparison, our review of various data sources on rent variation arrangements suggests 8-16% of private renter households got some form of variation. These sources are discussed in Chapter 5. Views are also mixed on the adequacy of coordination on rental housing regulation. Mason et al. (2020, p. 3) claim ‘there was multilevel coordination [i.e. directly involving the Australian Government as well as the states and territories] as evidenced by harmonisation of laws to protect tenants through eviction moratoria and suspension of rental increases’. They also state that ‘Through the National Cabinet the states and territories harmonised regulations to enable eviction moratoria and provide rent relief and adopted a range of measures and initiatives in an attempt to mitigate the twin health and economic impacts of the crisis’ (Mason et al., 2020, p. 56). However, as highlighted by our 43

own research (see Chapters 6 and 7) this claim is highly questionable. Moreover, others argue that national coordination and leadership was often lacking in this policy realm. Leishman et al. (2020) note that initial policy attention focused heavily on income support measures, with little consideration of housing system interventions or their impacts. Oswald et al. (2020) found that measures did not provide certainty for tenants and landlords, with confusion about what support was available. 3.3.2 Evidence on homelessness policy impacts Mason et al. (2020, p. 54) argue the homelessness emergency accommodation programs rolled out in several states in 2020 have been ‘widely seen as a successful public health emergency operation’, and the collaboration between state governments and specialist homelessness services on this program is ‘a stand-out example of coordination’ (p. 55). However as cited by Kirby (2020), Associate Professor Lisa Wood of the University of Western Australia observed ‘Sadly, Australia has no coordinated strategy whatsoever regarding homelessness and COVID nationally … Both in our national and state governments, homelessness crosses a number of government portfolios, and it seems to be falling through the cracks.’ As highlighted by our own research (see Chapter 7) a number of homelessness responses were spearheaded by advocacy groups and later folded into official policy directions, paralleling developments elsewhere in the world where frontline workers developed new ways of working to cope at their discretion (Weible et al., 2020). Berger and Reupert (2020) also criticise the speed of official Australian response on homelessness as relatively slow, reflecting its lower official prioritisation compared with support for the unemployed and businesses affected by the pandemic. Emergency accommodation program impacts are also contested. Mason et al. (2020) claim that, thanks to such action ‘for the first time [in the early months of the pandemic], rough sleeping was briefly eliminated’ (p27). However, our own research suggests that – while it might reflect the situation in central Melbourne during Victoria’s COVID-19 second wave – this assertion is an overstatement for other cities, and for the country as a whole (Pawson et al., 2020). The homelessness response has also emphasized the significance of the way that social problems are framed. Parsell et al. (2020, p. 5) argue that during COVID-19 in Australia the representation of homelessness departed from ‘the prevailing neoliberal/individualising framework that has dominated homelessness policy in recent decades’. They contend that status quo problematisation of homelessness is characterised by ‘sick talk’ – emphasis placed on illness and disability – and ‘sin talk’ – emphasis on misbehaviour and irresponsible choices. Instead, during 2020 homelessness has been re-framed as a threat to public health of the broader population. The perception of homeless people as vectors for community transmission set in motion a COVID-19 crisis response that would normally have been considered impossible, benefiting those with very little power. In common with efforts to rationalise action to combat homelessness on the grounds of resulting public expenditure savings, this ‘public health of the broader population’ framing avoids justifying such action with reference to the resulting benefits enjoyed by the people directly involved – i.e. homeless people themselves. Nevertheless, epidemiological research from the UK (Lewer 44

COVID-19: Rental housing and homelessness policy impacts, 2021

et al., 2020) has estimated the extent to which the provision of safe temporary accommodation in fact helped in preventing virus infection and, therefore, in saving lives within England’s homeless population. The research concluded that, during the COVID-19 first wave, the Everyone In program prevented 266 deaths, as well as 21,092 infections, 1,164 hospital admissions and 338 ICU admissions during this period. By implication – at least within the context of the significant second wave of the virus experienced in Victoria July-October 2020 – housing homeless people in safe temporary accommodation will also have saved lives within the homeless population in Australia.

3.4 Chapter conclusion In 2015, Jacobs (2015, p. 61) pessimistically remarked ‘it seems most unlikely that the direction of Australian government housing policy will change course’ to redress inequality and lack of affordability in the housing system. COVID-19 proved a focusing event for housing policy, as for several other domains, and a number of policy innovations were rapidly rolled out to mitigate the impacts of the pandemic and associated recession, including eviction moratoriums and emergency accommodation programs. These innovations clearly provided much-needed support to millions of people in Australia, with poverty rates and housing affordability stress reducing (Biddle et al., 2020; Leishman et al., 2020). However, it is yet to be seen if responses to COVID-19 might provide the stimulus for the systemic reform needed to address the housing affordability problem. Mason et al. (2020, p. 10) argue ‘the crisis exposed a social housing system deeply under strain, lacking a supply pipeline to meet current (and future) demand’. Orders to stay home revealed the precarity and/or inadequacy of many people’s accommodation, from those who are homeless to those unable to meet rent or mortgage payments on suddenly-restricted income. A significant frame-shift was apparent for responses to homelessness, while support for tenants focused on individualised income support and rental negotiations, continuing a long-standing market-based framing. Further investigation is needed on the implementation and effects of these policies, particularly in how they differentially affect different sectors of society.

References Baumgartner, F. R., & Jones, B. D. (1993) Agendas and instability in American politics. Chicago, IL: University of Chicago Press. Berger, E., & Reupert, A. (2020) The COVID-19 pandemic in Australia: Lessons learnt, Psychological Trauma: Theory, Research, Practice, and Policy, 12, 5, 494496 doi: 10.1037/tra0000722 Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2020) COVID-19 and mortgage and rental payments: May 2020 ANU Centre for Social Research and Methods https://csrm.cass.anu.edu.au/sites/default/files/docs/2020/6/COVID-19_and_ housing_FINAL.pdf Birkland, T. A. (1998) Focusing Events, Mobilization, and Agenda Setting, Journal of Public Policy, 18, 1, 53-74. Retrieved from www.jstor.org/ stable/4007601 Boin, A., ‘t Hart, P., & McConnell, A. (2009) Crisis exploitation: political and policy impacts of framing contests, Journal of European Public Policy, 16, 1, 81106 doi: 10.1080/13501760802453221 45

Clinton, W. J. (2006) Lessons Learned from Tsunami Recovery: Key Propositions for Building Back Better; Office of the UN Secretary-General’s Special Envoy for Tsunami Recovery https://www.preventionweb.net/files/2054_VL108301.pdf

Milligan, V., & Pinnegar, S. (2010) The Comeback of National Housing Policy in Australia: First Reflections, International Journal of Housing Policy, 10, 3, 325344 doi:10.1080/14616718.2010.506747

Cobb, R. W., & Elder, C. D. (1971) The Politics of Agenda-Building: An Alternative Perspective for Modern Democratic Theory, The Journal of Politics, 33, 4, 892915 doi: 10.2307/2128415

Nohrstedt, D. (2008) The Politics of Crisis Policymaking: Chernobyl and Swedish Nuclear Energy Policy, Policy Studies Journal, 36, 2, 257-278 doi:http:// dx.doi.org/10.1111/j.1541-0072.2008.00265.x

Evans, R., Rosewall, T., & Wong, A. (2020) The Rental Market and COVID-19; Reserve Bank of Australia https://www.rba.gov.au/publications/bulletin/2020/ sep/the-rental-market-and-covid-19.html

Oswald, D., Moore, T., & Baker, E. (2020) Post pandemic landlord-renter relationships in Australia; Australian Housing and Urban Research Institute https://www.ahuri.edu.au/research/final-reports/344

Farha, L. (2020) '"Housing, the front line defence against the COVID-19 outbreak," says UN expert’, media release, United Nations Human Rights Office of the High Commissioner https://www.ohchr.org/EN/NewsEvents/Pages/ DisplayNews.aspx?NewsID=25727&LangID=E

Parsell, C., Clarke, A., & Kuskoff, E. (2020) Understanding responses to homelessness during COVID-19: an examination of Australia, Housing Studies, 1-14 doi:10.1080/02673037.2020.1829564

Fernandez, G., & Ahmed, I. (2019) “Build back better” approach to disaster recovery: Research trends since 2006, Progress in Disaster Science, 1, 100003 doi: https://doi.org/10.1016/j.pdisas.2019.100003 Foye, C., Gibb, K., & Marsh, A. (2020) Crises: a working literature review; UK Collaborative Centre for Housing Evidence Jacobs, K. (2015) A reverse form of welfarism: some reflections on Australian housing policy, Australian Journal of Social Issues, 50, 1, 53-68 doi: http:// dx.doi.org/10.1002/j.1839-4655.2015.tb00334.x Kim, K., & Olshansky, R. B. (2014) The Theory and Practice of Building Back Better, Journal of the American Planning Association, 80, 4, 289-292 doi: 10.1080/01944363.2014.988597 Kingdon, J. W. (1984) Agendas, alternatives, and public policies. Boston, MA: Little, Brown. Kirby, T. (2020) Efforts escalate to protect homeless people from COVID-19 in UK, The Lancet Respiratory Medicine, 8, 5, 447-449 doi: 10.1016/S22132600(20)30160-0 Leishman, C., Ong, R., Lester, L., & Liang, W. (2020) Supporting Australia’s housing system: modelling pandemic policy responses; Australian Housing and Urban Research Institute https://www.ahuri.edu.au/research/final-reports/346 Lewer, D., Braithwaite, I., Bullock, M., Eyre, M. T., White, P. J., Aldridge, R. W., . . . Hayward, A. C. (2020) COVID-19 among people experiencing homelessness in England: a modelling study, The Lancet Respiratory Medicine doi: https://doi. org/10.1016/S2213-2600(20)30396-9

Pawson, H., Parsell, C., Liu, E., Hartley, C., & Thompson, S. (2020) Australian Homelessness Monitor 2020; Launch Housing https://www.launchhousing.org. au/ending-homelessness/research-hub/australian-homelessness-monitor-2020 Peck, J. (2006) Liberating the City: Between New York and New Orleans, Urban Geography, 27, 8, 681-713 doi:10.2747/0272-3638.27.8.681 Peck, J. (2011) Global Policy Models, Globalizing Poverty Management: International Convergence or Fast-Policy Integration?, Geography Compass, 5, 4, 165-181 doi:10.1111/j.1749-8198.2011.00417.x Sabatier, P. A. (1988) An advocacy coalition framework of policy change and the role of policy-oriented learning therein, Policy Sciences, 21, 2, 129-168 doi:10.1007/BF00136406 Weible, C. M., Nohrstedt, D., Cairney, P., Carter, D. P., Crow, D. A., Durnová, A. P., . . . Stone, D. (2020) COVID-19 and the policy sciences: initial reactions and perspectives, Policy Sciences, 53, 2, 225-241 doi:10.1007/s11077-020-09381-4 Weible, C. M., Sabatier, P. A., Jenkins-Smith, H. C., Nohrstedt, D., Henry, A. D., & deLeon, P. (2011) A Quarter Century of the Advocacy Coalition Framework: An Introduction to the Special Issue, Policy Studies Journal, 39, 3, 349-360 doi:10.1111/j.1541-0072.2011.00412.x Zahariadis, N. (2016) Setting the agenda on agenda setting: definitions, concepts, and controversies. In Handbook of public policy agenda setting (pp. 1-24). Cheltenham: Edward Elgar Publishing.

Mannakkara, S., & Wilkinson, S. (2014) Re-conceptualising “Building Back Better” to improve post-disaster recovery, International Journal of Managing Projects in Business, 7, 3, 327-341 doi: http://dx.doi.org/10.1108/ IJMPB-10-2013-0054 Mason, C., Moran, M., & Earles, A. (2020) Policy coordination and housing outcomes during COVID-19; Australian Housing and Urban Research Institute https://www.ahuri.edu.au/research/final-reports/343

COVID-19: Rental housing and homelessness policy impacts, 2021

4. Pandemic impacts on the private rental housing market Key points: • Australia’s rental housing markets experienced substantial turbulence during 2020, with markedly divergent trends experienced between inner cities and regional areas, between houses and units, and between the eastern state capitals and Perth • Median rents fell by nearly 5% in north and west Melbourne between Q1 and Q3 2020, and by as much as 10% in Sydney, but rose by 6% across nonmetropolitan Victoria, and by more than 5% in many parts of regional NSW • There were similar spatial contrasts in rental housing vacancy trends, with Melbourne’s vacancy rate doubling to 4.5% between March and September 2020 while Perth saw its equivalent figure halved to under 1%. • On housing affordability, renters tended to be hit much harder by the pandemic than homeowners. Nationally, renter incomes fell by 5% MarchJune 2020, while housing costs dropped by only 0.5%; mortgage holders, by contrast, saw a 0.2% decline in incomes alongside a 5% decline in housing costs.

4.1 Introduction Over the course of the COVID-19 pandemic and recession there have been numerous reports of declining costs and rising vacancy rates in Australian private rental markets. In some submarkets there has been a sudden surplus of properties for rent, particularly as large numbers of tenants terminated their leases in response to unemployment, underemployment, and the need to relocate – intrastate, interstate or internationally – at short notice. In other submarkets, vacancy rates are lower than their pre-pandemic level, as households have relocated to areas where there are fewer restrictions, perhaps lower rental costs, and more housing stock that is better suited to working (and socialising) from home. In short, COVID-19 has had differentiated impacts on private rental housing submarkets nationwide. As this chapter outlines, falling rents and rising vacancy rates are largely a phenomenon of the inner-metropolitan regions in Sydney, Melbourne and (to a lesser extent) Brisbane. Broadly, this is where rents have been highest over the past decade and, as such, where tenancies will have been most difficult to sustain on reduced incomes. These regions are also those in which apartments and units tend to be concentrated. While housing preferences shifted away from such dwellings and toward larger ones during the pandemic, declining demand in inner city submarkets has been also related to the uneven impacts of the recession across different demographics, tenures and geographies.

Photo © Brayden Howie, used under licence from Austockphoto 48

COVID-19: Rental housing and homelessness policy impacts, 2021

Importantly, the limited data available on household income trends during the pandemic suggests that in the initial months of the crisis renters’ incomes tended to fall much more sharply than their housing costs. Official statistics to June 2020 show a 2.3% reduction in renters’ average income nationwide 49

but only a 0.5% reduction in rents (ABS 2020). Mortgage holders, by contrast, experienced a 0.9% increase in average incomes and a 5% decline in mortgage payments (ibid.). An ANU survey of over 3,000 respondents found that, in May 2020, 27% of renters were struggling to pay rent on time, including 44% of respondents aged 18 to 24, compared to 17% of mortgagors (Biddle et al. 2020). Inner-city apartment dwellers, who are disproportionately younger, employed in lower-wage service sectors like retail and hospitality, and non-citizens (particularly international students), are more likely to have experienced sharper income declines due to unemployment, underemployment and exclusion from income support (Berg & Farbenblum 2020; Equity Economics 2020; Randolph & Sisson 2020). Furthermore, given new dwelling pipelines and short-term private rental sectors skew toward the inner cities and higher density areas, there has been a simultaneous increase in supply in these submarkets as new buildings have been completed and as dwellings previously listed on Airbnb and similar platforms are converted to longer-term private rental. As a result of this combination of demand and supply-side factors a result, inner city apartment submarkets have exhibited the steepest falls in median rents (as much as 10%, in inner Sydney) and highest rates of termination (as much as 13%, again for inner Sydney). In the outer rings of Australiaâ€™s three largest cities, and in smaller cities and regional areas, median rents have remained roughly stable while vacancy rates have fallen during the pandemic to date (December 2020). Indeed, in some areas, median rents have risen. As above, there are many factors at play. On the one hand, there appears to have been a greater demand for larger dwellings better suited to spending more time at home. There also appears to have been growth in demand for housing in regional areas, particularly coastal locations and others with natural amenities, which is likely supported by the rise of working from home. On the other hand, many of these outer-metropolitan and regional areas also offer cheaper submarkets than the inner and middle rings of major cities, even after the aforementioned declines in median rents. As such, demand growth in these submarkets may also be a response to income reductions and affordability pressures in more expensive locations.

impacts were more severe and restrictions harsher and longer), these states are the focus of the chapter. Data availability was also a consideration in the decision to focus on these jurisdictions.

4.2 Private rental housing supply In the decade prior to March 2020, when COVID-19 restrictions came into force throughout Australia, the size of the private rental sector grew to over one quarter of all households. In 2009-10, Figure 4.1 shows, households renting from a private landlord constituted approximately 23.7% of all households; in 2017-18, they constituted 27.1%. While the volume of investor finance issuance (i.e. the volume of finance for private landlords) declined over the three years preceding the pandemic, the prior trend, illustrated in Figure 4.2, was one in which investor purchasers were receiving a growing share of new home loan commitments, besides a contraction during 2015-16. In COVID times, owner-occupier finance has rebounded much more sharply than investor finance, which only recovered to pre-pandemic levels as of September 2020. This suggests that the relative growth of the private rental sector may be temporarily stalled, though no up-to-date national data is available at the time of writing. Figure 4.1: Size of private rental sector, Australia, 1994-2018

One important implication of these trends is that private renting may be becoming more expensive for low-income private renters in regional and outer-metropolitan areas. Such a development could be expected to intensify as extraordinary Commonwealth Government income supports are scaled back and ended in coming months. Observed 2020 trends also suggest that eviction moratoriums may have been more effective in some areas than others, depending on how easily the property might attract another tenant at the same price. These two points are elaborated in the conclusion to the chapter. The remainder of the chapter is organised as follows. The first section briefly outlines supply trends within the private rental sector in the lead up to and during the COVID-19 pandemic and recession, including trends in dwelling approvals and completions, investor finance issuance, and Airbnb listings. The second section details trends in median rents, highlighting differences among different regions and housing types. Similarly, the third section describes trends in new tenancies and lease terminations, as well as residential vacancy rates, across different regions and housing types. The chapter concludes with a brief reiteration of the uneven impacts of COVID-19 on private rental markets. Given that the most dramatic changes and starkest comparisons are found within NSW, Victoria and Queensland (unsurprisingly, given this is where pandemic 50

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: ABS Cat 4130.0

Figure 4.2: Value of new home loan commitments in Australia, seasonally adjusted, Nov 2010- Sep 2020

Investors tend to be disproportionately represented among purchasers of new apartments and units (CoreLogic 2016). As Figures 4.4 through 4.7 illustrate, the mid-2010s saw a boom in dwelling construction of this type, corresponding to the waves of investor finance. Furthermore, as a comparison between the below figures suggests, this growth was primarily within Australiaâ€™s three largest cities. In Sydney, Melbourne and Brisbane, apartments and units became the dominant source of new dwellings over 2013-14 to 2015-16. Indeed, in Sydney it has remained the majority. Sydney is the only state capital for which up-to-date dwelling completions data is available, and these data reveal that over 167,000 multi-unit dwellings were completed from 2015-16 to 2019-20, or 65% of all new dwellings (DPIE 2020). These were concentrated in the inner ring and middle rings of the city, as were new apartment and unit approvals for Melbourne and Brisbane (ABS Cat 8752.0). Since 2018, the pipeline of new dwellings has contracted, and particularly since March 2020. Figure 4.4 suggests that it may be reorienting towards detached homes; however, given the natural month-to-month fluctuation in apartment and unit approvals, and notwithstanding the months immediately following the implementation of COVID-19 restrictions, it is too soon to determine how lasting this may be.

Source: ABS Cat 5601.0

Two waves of investor finance are clear from the data for NSW, Queensland and Victoria, where the vast majority of investment has flowed. Figure 4.3 illustrates that these waves of loan commitments, either side of 2015-16, corresponded with substantial growth in the supply of privately rented dwellings. From the beginning of 2010-11 to the end of 2013-14, approximately 262,000 additional households entered the sector, or over 70% of net new households (ABS Cat 4130.0). In 2017-18, another 245,000 households entered the private rental sector, or nearly 80% of net new households (ibid.)

Figure 4.4: New private dwelling completions, Australia, seasonally adjusted, Sep 2011- Jun 2020

Figure 4.3: Value of new home loan commitments for investor purchase, seasonally adjusted, Nov 2010-Sep 2020

Source: ABS Cat 8752.0

Source: ABS Cat 5601.0

COVID-19: Rental housing and homelessness policy impacts, 2021

Figure 4.5: New private dwelling approvals, Greater Sydney, 12-month moving average, Nov 2010-Sep 2020

Source: ABS Cat 8731.0

Figure 4.6: New private dwelling approvals, Greater Melbourne, 12-month moving average, Nov 2010-Sep 2020

Figure 4.7: New private dwelling approvals, Greater Brisbane, 12-month moving average, Nov 2010-Sep 2020

Source: ABS Cat 8731.0

In recent years, a significant portion of investor-owned housing stock – both new build and established – has been leased on the short-term rental market. In particular locations it has accounted for more than 10% of rental dwellings (Alizadeh et al. 2018). However, over the course of 2020 the number of listings has declined dramatically, as travel restrictions have limited international and, until recently, domestic travel. The latest data from Inside Airbnb indicate that listings for entire homes have contracted by 17% in Sydney and 22% in Melbourne since COVID-19 restrictions were introduced. This equates to 4,317 dwellings in Sydney and 3,661 dwellings in Melbourne, most of which have presumably been either available for long term lease or purchase. The decline is most pronounced in the inner rings of Sydney and Melbourne, as Figures 4.8 and 4.9 clearly show: 72% of removed listings in Sydney and 80% of the removed listings in Melbourne were located in the inner ring. While many of these properties may have been sold, re-occupied by their owners, or left unoccupied, these data suggest a substantial increase in the supply of longer term private rental properties in inner-Sydney and Melbourne. A recent report by Buckle et al. (2020) similarly notes the perceptible effect of the reduced number of Airbnb listings on vacancy rates and median rents in Sydney, particularly in high-demand LGAs including the City of Sydney, North Sydney and Northern Beaches. Furthermore, the authors note that in Hobart, where Airbnb has penetrated housing markets to an even greater extent, a decline of approximately 17% of listings was responsible for between 46% and 65% of the decline in median rents across the city’s four LGAs.

Source: ABS Cat 8731.0

COVID-19: Rental housing and homelessness policy impacts, 2021

Figure 4.8: Airbnb listings for entire houses, Sydney, Jan 2020-Oct 2020

rentals have been converted to long-term. During the pandemic, there has been a marked decline in demand within these submarkets and a corresponding increase in demand for larger outer suburban or regional rental homes.

4.3 Rents In most submarkets, COVID-19 caused a sudden halt to or decline in median rents. Analysis from the Reserve Bank of Australia describes this as â€œthe first quarterly fall in rents in the history of the Consumer Price Index (CPI)â€™ (Evans et al. 2020: 76) and the fastest price adjustment in several decades. As this section details, declines were steeper and more enduring for apartments and units, particularly in the inner areas of Melbourne and Sydney. The following analysis of median asking (advertised) rents and agreed rents points to a significant fall in demand within these inner city, higher density submarkets and a concomitant (though not equal) rise in demand within lower density, outer metropolitan and regional submarkets.

Source: data adapted from InsideAirbnb

Figure 4.9: Airbnb listings for entire houses, Melbourne, Jan 2020-Oct 2020

Figure 4.10 compares the decline in median advertised rents from both the first quarter of 2020 and the June quarter of 2019, for both houses and units in each of the state capitals. It illustrates the immediate impact of the pandemic on unit rents in Sydney, Melbourne and Hobart, though only in Sydney and Melbourne did asking rents fall to lower than the median price for the same time the previous year. By contrast, the effect on median asking rents for houses was much smaller, barring Hobart. Indeed, in Adelaide, Perth and Canberra, median asking rents remained higher than they had been during the June quarter of 2019. Furthermore, the impact of COVID-19 on house rents seems to have been shorter lived. Figure 4.11 shows that in every state capital bar Hobart the median asking rent for houses increased during the September quarter 2020, to at least its prepandemic level and in several cities to a greater level. This is also true of asking rents for units in all cities besides Sydney, Melbourne and Hobart; as Figure 4.12 summarises, apartments in Brisbane, Adelaide and Perth were advertised for higher rates in the September quarter than in the March quarter of 2020. Figure 4.10: Change in median asking rents, quarter on quarter and year on year to June 2020

Source: data adapted from InsideAirbnb

This section has briefly outlined the pre-pandemic growth of the private rental sector in both its long- and short-term forms. It has particularly highlighted the growth of apartments and units as submarkets which are disproportionately driven by investor purchasers. Finally, it has noted the concentration of this growth in Sydney, Melbourne and Brisbane, and in particular the inner rings of these cities. Pre-pandemic, there was a large stock of privately rented flats and units within the inner-regions of Australiaâ€™s three largest cities, and this stock has likely increased as new dwellings have been completed and as short-term 56

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: Domain Rental Reports

Figure 4.11: Quarterly median asking rent per week, houses, Mar 2019-Sep 2020

bedroom units and 3.4% for two-bedroom units from the March quarter to the September quarter. In Sydney, median rents declined by 8.2% and 5.6% respectively. In Melbourne, median rents declined by 8.1% and 8.7%. In contrast, median rents for three- and four-bedroom houses have declined to a lesser extent in Brisbane (1.1% and 0%) and Melbourne (2.4% and 2.2%) and have risen in Sydney (by 2.0% and 1.6%). Figure 4.13: Median agreed rents, quarterly & indexed to March-19, Greater Sydney

Source: Domain Rental Reports

Figure 4.12: Quarterly median asking rent per week, units, Mar 2019-Sep 2020

Source: NSW DCJ Rent and Sales reports

Figure 4.14: Median agreed rents, quarterly & indexed to Mar-19, Greater Melbourne

Source: Domain Rental Reports

The diverging trends of houses and units are further demonstrated in data relating to agreed rents for new tenancies, depicted below in Figures 4.13 to 4.15. As opposed to asking or advertised rents, agreed rents represent the rental price at which a property is contracted to a tenant. Median agreed rents for one and two-bedroom apartments and units have declined in Sydney, Melbourne and Brisbane. In Brisbane, median rents declined by 5.4% for one58

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: VIC DHHS rental reports

Figure 4.15: Median agreed rents, quarterly & indexed to Mar-19, Brisbane LGA

Across Queensland and outside of the state capital, median rents for apartments were less significantly impacted, falling slightly in parts of the Gold Coast and other larger centres but rising in others, including the Sunshine Coast (+2.6%), Mackay (+3.4%) and Rockhampton (+4.3%). Median house rents also rose or returned to pre-pandemic levels for most regions outside of innerBrisbane and Gold Coast, most notably in Gladstone (+4.0% for 3-BR houses), Rockhampton (+3.2%), Caboolture (+2.9%) and Hervey Bay (+2.9%). Steeper rises were generally recorded for four-bedroom houses. Figure 4.16: Median agreed rents, quarterly & indexed to Mar-19, Sydney & regional NSW

Source: QLD RTA Median rents quarterly data

Across all dwelling types, the drop in median rents has been most significant in inner-metropolitan regions. As Figure 4.16 shows, agreed rents declined most markedly in inner-Sydney, falling more than 10% for all dwellings from the March quarter to September quarter, while median rents state-wide declined 2.1%. Middle-ring median rents also fell (-4.9%), while outer-ring median rents returned to their pre-pandemic levels after an initial decline. Figure 4.16 also illustrates that rents remained the same in the Wollongong region and rose in the Newcastle region (+5.0%) and the rest of NSW (+2.9%). Increases of more than 5% were recorded in Queanbeyan, Orange, Ballina and Lismore, among other, smaller, LGAs. A similar trend is hinted at in the data for median agreed rents in Melbourne and Victoria. As Figure 4.17 displays, there was a steep decline of 4.8% in the North and West Metro region, and a lesser decline and quicker rebound in the Eastern and Southern Metro regions. Outside of Melbourne, median rents rose by 6.3%, most notably in the Latrobe (+8.0%), Wellington (+7.3%), Moira (+7.4%) and Glenelg (+15.4%) LGAs. Finally, falling median rents in Queensland were largely confined to apartment stock in inner-Brisbane and Gold Coast. These data are summarised in Figures 18 and 19, which present quarterly median agreed rents for two-bedroom flats and units (Figure 4.18) and for three-bedroom houses (Figure 4.19). (The data are presented in this way due to the absence of data for median rents across all dwelling types.) Between the March and September quarters, inner Brisbane regions registered declines between 2.2% and 10.8% for one bedroom units and between 5.0% and 13.8% for two bedroom units (though in South East Inner Brisbane there was a slight increase in the former and no change in the latter). Median rents for four-bedroom houses also declined in some inner metropolitan regions over this period. Median rents for three-bedroom houses largely rebounded after an initial three-month decline, besides in the City Inner region where they fell by 3.5%. 60

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: NSW DCJ Rent and Sales reports

Figure 4.17: Median agreed rents, quarterly & indexed to Mar-19, Melbourne & regional Victoria

Source: VIC DHHS rental reports

Figure 4.18: Median agreed rents 2-BR flats/units, quarterly & indexed to Mar-19, select regions Queensland

Source: QLD RTA Median rents quarterly data

Figure 4.19: Median agreed rents for 3-BR houses, quarterly & indexed to Mar-19, select regions Queensland

particularly within non-metropolitan regions and in outer-metropolitan regions. Simply put, demand for inner-city apartments has dropped and demand for larger houses, outside of the capital cities or within their outer rings, has risen. It is safe to assume that as many households transitioned to and remained working from home, and as social lives also withdrew further into the private sphere, larger housing became more desirable and close proximity to inner-city employment and amenities became less so. Larger housing may also have been more desirable for the formation of new sharehouses in response to housing stress, as research by Raynor & Panza (2020) suggests. At the same time, the supply of inner-metropolitan rental housing has increased due to new dwelling completions and the decline of the short-term rental sector. However, falling rents in inner-metropolitan apartment submarkets are not simply due to renters desiring different dwelling types due to COVID-19 restrictions; it is also because inner-metropolitan regions are some of the most severely impacted by unemployment, income loss and restrictions on international migration. Per Figure 4.20 below, the number of residents receiving the JobSeeker payment increased during the June quarter by 190% in inner-Sydney and 147% in the middle ring, while for the outer ring the increase was 97%, Wollongong 70%, and the rest of NSW 54%. Similar, albeit somewhat less stark, variation occurred in Victoria, as illustrated in Figure 4.21. During the same period, Commonwealth Rent Assistance recipients grew by 77% in the inner ring, double the rate of increase for the middle ring and more than double the rest of NSW (DSS 2020).

Figure 4.20: Jobseeker recipients NSW, March & June 2020

Source: QLD RTA Median rents quarterly data

In sum, median asking rents and agreed rents have fallen most significantly for apartments and units, particularly within the inner-metropolitan regions of Sydney, Melbourne and (to a lesser extent) Brisbane. Conversely, the pandemicâ€™s impact on median rents for houses has been far less dramatic, 62

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: DSS Payment Demographic Data

Figure 4.21: Jobseeker recipients Victoria, March & June 2020

Source: DSS Payment Demographic Data

Furthermore, while the total number of new recipients was larger in outer and middle Sydney, JobSeeker (and JobKeeper) payments and Commonwealth Rent Assistance are less likely to be sufficient to sustain rental costs in innermetropolitan regions, even at their supplemented rate, without a rent variation being granted. Yet, as the next chapter outlines, many households were unable to negotiate an adequate rent reduction. Additionally, the inner-metropolitan regions are home to larger numbers of non-citizen residents, particularly international students, who are ineligible for income support and who in large numbers left the country or did not return for on-campus study (Berg & Farbenblum 2020; Morris et al. 2020; Unions NSW 2020).

Among the state and territory capitals, Sydney and Melbourne stand apart as having higher residential vacancy rates than prior to COVID-19 restrictions. (Note that these data relate to all dwellings, regardless of tenure.) As Figure 4.22 illustrates, vacancy rates in all other cities are below 1% â€“ bar Brisbane, where a sharp increase in vacancies was followed by a steady decline over May to October. These differences roughly correlate with the severity of pandemic impacts and restrictions, net migration trends for the capital cities (shown in Figure 4.23), and aforementioned supply-side differences, including larger pipelines of new dwellings in the three larger cities and greater impacts of the short-term rental sector. The rise in vacancy rates in Sydney is most pronounced in the CBD (from 3.6% in October 2019 to 11% in October 2020), Eastern Suburbs (2.6% to 5.0%), Lower North Shore (3.8% to 4.9%) and Inner West (3.0% to 4.4%). The rise in vacancy rates in Melbourne was most pronounced in the City (2.6% to 6.4%) and Inner East (2.6% to 6.2%). Conversely, vacancy rates rose less sharply in the middle ring of Sydney and fell in most outer suburbs, including the South West (2.6% to 1.1%), Sutherland Shire (2.5% to 1.0%), Hills District (4.8% to 3.3%) and Western Sydney (3.3% to 2.8%). In Melbourne, moderate declines were more widespread, though vacancy rates fell in the South East (1.9% to 1.3%) and Mornington Peninsula (1.0% to 0.7%). Vacancy rates fell in almost all non-metropolitan regions of NSW, Victoria and Queensland. Trends for the largest of these regions are summarised in Figures 4.24 to 4.26 Thus, vacancy rate trends further confirm that relative demand for housing outside of the inner and middle rings of the major cities has increased over the course of COVID-19 to date.

Figure 4.22: Residential vacancy rates, capital cities, Oct 2019-Oct 2020

As the next section outlines, higher density, inner-metropolitan submarkets experienced the highest rate of tenancy terminations, as renters sought both cheaper and more suitable housing elsewhere or exited the private rental sector altogether. One consequence has been increased demand in other submarkets, represented by both higher rents and lower vacancy rates. The latter are discussed in the next section. In short, it has become cheaper and less competitive to rent an apartment in the inner city. However, it remains more expensive than renting a house in the outer suburbs or regions, leading to the growing competitiveness and cost of renting in these, previously less coveted, submarkets.

4.4 Vacancy rates, new tenancies & terminations Another way of measuring the impact of COVID-19 on private rental demand is by analysing trends in vacancy rates, bond lodgements (indicating new tenancies) and bond refunds (indicating tenancy terminations). While these data are less consistently available from state to state, they confirm that there has been a significant fall in demand for apartments and units, particularly in inner-metropolitan regions, and relative rise in demand for houses in outer urban areas and regions. 64

COVID-19: Rental housing and homelessness policy impacts, 2021

Source: SQM Research

Figure 4.23: Net migration, greater capital cities, Oct 2019-Oct 2020

Source: ABS Regional internal migration estimates, provisional June 2020

Figure 4.24: Residential vacancy rates, select regions NSW, Oct 2019-Oct 2020

Source: SQM Research

COVID-19: Rental housing and homelessness policy impacts, 2021

Figure 4.25: Residential vacancy rates, select regions in Victoria, Oct 2019-Oct 2020

Source: SQM Research

Figure 4.26: Residential vacancy rates, select regions in Queensland, Oct 2019-Oct 2020

Source: SQM Research

While the above figures describe vacancy rates for all dwellings, rather than for the private rental sector specifically, they are supported by data relating to new and terminated tenancies in NSW, Victoria and Queensland. The most comprehensive data are available for NSW. Figures 4.27 and 4.28 point to the termination of 75,769 tenancies during the June quarter, a quarter-onquarter increase of 8% and year-on-year increase of 17%. This was primarily due to inner-Sydney rental submarkets, where more than 13% of tenancies were terminated. Indeed, outside of the inner and middle metropolitan regions, there was little to no change in the rate of tenancy terminations, as the diverging trends in Figure 4.27 indicate. In both the middle and outer rings of Sydney, and in most other regions in NSW, the total number of tenancies increased from the March quarter to September quarter â€“ most significantly in the LGAs of Byron (+7.7%), Tamworth (+4.8%) Shellharbour (+3.9%) and Ballina (+3.9%), among steeper increases in small LGAs â€“ whereas in inner-Sydney the total number of tenancies contracted by 1.7%.

Figure 4.28: Quarterly change in total bonds held, NSW, Jun 2019-Sep 2020

Figure 4.27: Proportion of previous quarter bonds refunded, NSW, Jun 2019-Sep 2020

Source: NSW DCJ Rent and Sales reports

The aforementioned trend towards greater turnover in higher density submarkets is further indicated in Figure 4.29 which highlights that flat and unit rental dwellings were disproportionately represented among terminated tenancies. The total number of such tenancies has returned to roughly prepandemic levels, after a sharp decline in the total number of such tenancies during the March quarter. Figure 4.29: Quarterly change in total bonds held, Greater Sydney, Jun 2019-Sep 2020

Source: NSW DCJ Rent and Sales reports

COVID-19: Rental housing and homelessness policy impacts, 2021

In Victoria, as Figure 4.30 indicates, the rate of terminations was not as high as in NSW, perhaps due to policy differences discussed in Chapter 6. Across the state, 631,675 tenancies were terminated in the June quarter, fewer than the March quarter. Nevertheless, there was a notable decline in total bonds held across all metropolitan regions (between 1.0 and 1.6%), whereas total bonds held in the rest of the state rose by 2.7%. The most significant growth in tenancies was recorded in the LGAs of Surf Coast (+11.1%), Baw Baw (+4.3%) and Benalla (+4.6%), among some other, smaller LGAs with steeper rises. These data are summarised in Figure 4.31. While no data is available for total bonds held for different dwelling types, new bond lodgements similarly suggest a movement away from apartments. As Figure 4.32 illustrates, new lodgements fell by 38.43% for one-bedroom apartments and 30.4% for two-bedroom apartments, while for three and four-bedroom houses they fell by 25.4% and 21.5% respectively.

Figure 4.31: Quarterly change in total bonds held, Victoria, Jun 2019-Sep 2020

Figure 4.30: Proportion of previous quarter bonds refunded, Victoria, Jun 2019-Sep 2020

Source: VIC DHHS rental reports

Figure 4.32: New bond lodgements, Greater Melbourne, Mar 2019-Sep 2020

Source: VIC DHHS rental reports

COVID-19: Rental housing and homelessness policy impacts, 2021

Data availability for Queensland is also somewhat more limited than NSW. Nevertheless, it too points to a movement away from inner-Brisbane and towards outer-Brisbane and other cities and regions beyond the state capital. These shifts are summarised in Figures 4.34 and 4.35. Between the March quarter and September quarter 2020, the total number of tenancies in innerBrisbane contracted by 0.4% while rising by 1.9% in outer-Brisbane, 2.6% in Gold Coast and 2.8% in the rest of Queensland. Among more populous regions, Noosa (+4.8%), Townsville (+4.3%) and Caboolture (+3.4%) recorded the most significant growth in bonds held. There was no sharp increase in the rate of tenancy terminations throughout the state, with fewer bonds refunded during the June quarter 2020 than the same period in 2019. There was an increase in terminations in Brisbane, where the number of refunded bonds declined quarter-on-quarter but increased 5.4% year-on-year to June quarter 2020. As Figures 35 and 36 indicate, the increase was predominantly within apartment submarkets; however, this is to some extent due to seasonal variation, and new bond lodgements in the September quarter indicate that these submarkets have recovered to a greater extent than in Sydney and Melbourne.

Figure 4.34: New bond lodgements, Brisbane LGA, Mar 2019-Sep 2020

Figure 4.33: Quarterly change in total bonds held, Queensland, Jun 2019-Sep 2020 Source: QLD RTA Median rents quarterly data

Figure 4.35: Refunded bonds, Brisbane LGA, Jul 2019-Jun 2020

Source: QLD RTA Median rents quarterly data

COVID-19: Rental housing and homelessness policy impacts, 2021

In sum, and while the data is somewhat patchy, trends in residential vacancy rates and bond lodgements and refunds similarly point to steeper falls in demand within inner-city, higher density submarkets. At the same time, the data suggest that demand has stabilised, or indeed risen, within some submarkets beyond the major cities, or beyond their inner and middle suburbs. As the previous section discussed, these trends are likely caused by both changing preferences and the uneven geography of COVID-19 impacts, as well as differences in supply trends both prior to and during the pandemic. As the next chapter outlines, some renters terminated their tenancies to take advantage of cheaper rents within some submarkets, or changes to their working and living needs, but many others terminated because they could no longer afford them, with or without a rent variation. Some renters from both categories have moved into submarkets that they otherwise mightn’t have, such as outer-suburbs, smaller cities and regional areas. Their actions have, in part, contributed to the falling vacancy rates discussed in this section and the rising median rents discussed in the previous section. Thus, as the conclusion to this chapter outlines, COVID-19 is proving a double-edged sword in the private rental sector.

4.5 Chapter conclusion The impacts of the COVID-19 pandemic, along with associated restrictions and recession, have varied significantly across geographies and among different types of dwellings. In sum, demand in the inner areas of Sydney, Melbourne and (to a lesser extent) Brisbane has fallen the most dramatically, particularly for flats and units. This, in combination with increases to supply due to new dwelling pipelines and falling numbers of short-term rentals, has led to lesser competition for rental housing and declining median rents. Yet while rents may be lower in these submarkets, they are not necessarily more affordable, given widespread unemployment and income losses. Higher rates of termination and vacancy within these submarkets do not simply indicate shifting lifestyle preferences; they also represent a response to housing stress due to lost income and the absence or inadequacy of government support. Chapter 5 explores this in greater detail, particularly in relation to eviction moratoriums and rent variation regimes. On the other hand, rental costs and competitiveness elsewhere – in the other states, in other regions of the eastern seaboard, and in the outer-metropolitan regions of its three largest cities – have largely either remained at pre-pandemic levels or increased. Rising median rents and falling vacancy rates point to increased demand for housing outside of the inner rings of Sydney, Melbourne and Brisbane, for larger homes in particular. Again, this is a response to housing stress as well as to lifestyle preferences or housing needs, given that rents are often cheaper in these submarkets than in the inner cities and, as Chapter 5 explains, few tenants received a rent reduction. As such, housing may be becoming less affordable for many existing renters in such locations – a trend that will be exacerbated when eviction moratoriums cease. Geographical variations in price and demand also have implications for landlord and property manager responses to tenants’ requests for rent variations under various eviction moratoriums. In submarkets where demand has grown and median rents risen, landlords and agents may be less inclined to offer a rent reduction and more inclined to permit debt accrual or encourage termination; it may be that this approach best meets their financial interests. In submarkets 74

COVID-19: Rental housing and homelessness policy impacts, 2021

where demand has fallen and median rents declined, they be more likely to offer a (larger) rent reduction. In other words, the negotiating position of tenants is substantively affected by the submarket in which they live; contrary to some accounts, few (if any) have become “renters’ markets”. While no comprehensive data are available on this subject, further insight into the varied design and performance of eviction moratoriums is presented in Chapter 5.

References ABS (2020) Household financial resources https://www.abs.gov.au/statistics/ economy/finance/household-financial-resources/june-2020#about-thisrelease Alizadeh, T., Farid, R., & Sarkar, S. (2018). Towards Understanding the SocioEconomic Patterns of Sharing Economy in Australia: An Investigation of Airbnb Listings in Sydney and Melbourne Metropolitan Regions. Urban Policy and Research, 36(4), 445–463 Berg, Laurie and Farbenblum, Bassina, As if We Weren’t Humans: The Abandonment of Temporary Migrants in Australia during COVID-19 https://ssrn. com/abstract=3709527 Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2020). COVID-19 and mortgage and rental payments: May 2020. Australian National University Buckle, C., Gurran, N., Phibbs, P., Harris, P., Lea, T. & Shrivastava, R. (2020) Marginal Housing During COVID-19. AHURI Final Report No. 348 CoreLogic (2016) Profile of the Australian residential property investor https:// www.corelogic.com.au/resources/pdf/reports/CoreLogic%20Investor%20 Report_June%202016.pdf DPIE (2020) Metropolitan Housing Monitors https://www.planning.nsw.gov.au/ Research-and-Demography/Metropolitan-Housing-Monitors DSS (2020) DSS Demographics June 2020 https://data.gov.au/data/dataset/ dss-payment-demographic-data Equity Economics (2020) A wave of disadvantage across NSW: Impact of the COVID-19 recession; New South Wales Council of Social Services (NCOSS) Evans, R., Rosewall, T. and Wong, A. (2020) The Rental Market and COVID-19; RBA Bulletin pp75-84; Sydney: Reserve Bank of Australia Morris, A., Hastings, C., Wilson, S., Michell, E., Ramia, G. & Overgaard, C. (2020). The experience of international students before and during COVID-19: Housing, work, study, and wellbeing. UTS Randolph, B. & Sisson, A. (2020) Who lives in higher density housing? https:// cityfutures.be.unsw.edu.au/cityviz/who-lives-higher-density-housing/ Raynor, K. & Panza, L. (2020). The impact of COVID-19 on Victorian share households. University of Melbourne; Hallmark Research Initiative for Affordable Housing Unions NSW (2020). No worker left behind: survey results

5. Pandemic impacts on existing tenancies Key points: • At least a quarter of all private renters lost income during the pandemic, but only a smaller minority got a rent variation from their landlord: between 8-16% of renters, depending on the data source. • A similar proportion was refused a variation; more were discouraged from asking and more left their tenancy. • By comparison, at the May peak in loan deferrals, some 12% of private rental properties were subject to deferrals authorised by banks. • At least 30% of rent variations merely deferred the rent, rather than reduced it. This implies that tenants with mounting deferred rent debts could number at least 75,000 across Australia in late 2020.

5.1 Introduction and data sources This chapter examines the quantitative evidence about impacts of the COVID-19 pandemic and economic crisis on existing tenancies. These impacts are the result of the interplay of wider market conditions (reviewed in the previous chapter) and emergency policy measures intended to protect households from default and eviction. Those measures have been discussed in outline in Chapter 2, and are further elaborated in Chapter 6; here the focus is on quantitative evidence of impact. The analysis relies on the following sources: • The City Futures COVID-19 renter survey. We conducted this survey especially for the present research, with an online questionnaire in the field from mid-August to the end of October. Persons aged 18 years and older who were living in rental housing in Australia at 29 March 2020 (the date the National Cabinet announced the eviction moratorium) were eligible to participate; in total 312 persons took part, all but one were private tenants. • State government agency data from New South Wales, Victoria, Queensland and Western Australia. State government agencies in these four jurisdictions provided data about rent variations especially for this research. The Victorian data were a sample of de-identified records of rent variation agreements (n=200) registered with Consumer Affairs Victoria in April and May 20203. The New South Wales data were de-identified records of all rental variations (n=47) conciliated by NSW Fair Trading to 29 June 2020. The Queensland data were a sample of de-identified records of rent variation agreements (n=195) conciliated by the Residential Tenancies Authority to the end of July 2020. The Western Australia data were a sample of de-identified records of rent variations (n=203) conciliated by the WA Mandatory Conciliation Service June-September 2020. 3

Victoria was the only jurisdiction to implement such a registration scheme.

Photo by Dimi Katsavaris on Unsplash 76

COVID-19: Rental housing and homelessness policy impacts, 2021

• The Australian National University (ANU) COVID-19 and Housing Stress survey (Biddle, et al. 2020). The ANU Survey is a nationally representative survey of 3,249 respondents conducted in May 2020. • Better Renting’s ‘Rent Due: renting and stress during COVID-19’ survey (Dignam, 2020). This was conducted by the rental housing advocacy group Better Renting in May and June 2020. The online questionnaire was answered by 967 respondents, with a large proportion from eastern states and the ACT. • AHURI’s Renting in the Time of COVID-19 survey (Baker, et al. 2020). The AHURI survey is a nationally representative survey of 13,289 renter households (PRS and social housing) conducted in July. • The ABS Household Impacts of COVID-19 surveys of May and October 2020. • The Reserve Bank of Australia bulletin ‘The Rental Market and COVID-19’ (Evans, et al. 2020). The bulletin cites data sourced from the rent payments platform, MRI, which reportedly handles about one fifth of PRS tenancies in Australia.

5.2 Impacts on existing tenancies: triangulating research findings A selection of key data points from this group of sources is presented in Table 5.1. This triangulation shows correspondence between the results of our survey and the other much larger surveys in the group, indicating that the results generally are plausible. They also show some points of variation, and areas of impact not captured or not reported in the other sources, which will be discussed.

5.3 Moving out As we saw in Chapter 3, one early response to the emergency, particularly in New South Wales, was an unusually large movement out of existing tenancies – all the more remarkable since this coincided with the national lock-down and general strategy of ‘staying home’ as officially advocated across Australia in the March-June 2020 period. Only a little of this movement is captured in the surveys published by other researchers in the field. In the ANU survey, 3% of respondent renters had requested ‘early termination’ of their tenancies, and 1% said they had been ‘evicted’ (Biddle et al, 2020: 12). The report of the AHURI survey does not enumerate tenants who moved for all reasons, but does report that 6% ‘had to move out’ as an effect of financial hardship, and 5% received ‘an eviction notice’, with 3% actually evicted (Baker et al. 2020: 11-12). In our survey, however, almost one in five respondents (19%) had moved since the start of the emergency period (i.e. within 5-7 months of the onset of the pandemic in Australia). 29% of movers did so to reduce their housing costs, while 48% moved for another reason of their own; 23% moved because the landlord requested or required it. Group households and single parent households were more likely than other households to move.

COVID-19: Rental housing and homelessness policy impacts, 2021

5.4 Rent variations 5.4.1 The incidence of variations

Table 5.1: Renter impacts: compendium table CFRC

ANU

AHURI

Better Renting

RBA/MRI

Renters who moved (all reasons)

19%

Renters who were evicted/ moved at request of landlord

Renters who lost income

41%

Renters who asked for a variation

38%

Renters who were discouraged from asking

24%

Of renters who asked, those who got a variation

43%

73%

60%

42%

Renters who got a variation

16%

12%

10%

13%

15%

Renters who got a reduction

10%

Renters who got a deferral

Renters who got a combined reduction and deferral

17%

ABS May

ABS Oct

NSW Govt

Vic Govt

Qld Govt

WA Govt

As for the proportion of renters who actually got a variation, the results across the group of surveys are closer, mostly between 8% (ABS October) and 16% of respondents (our survey).

3% 24%

63%

16%

32%

In our survey, as in the ANU, AHURI and Better Renting surveys, we find only a minority of renters sought a rent variation, though the proportion varies across surveys: 38% of our respondents had asked for a variation, similar to the Better Renting survey result (32%), but significantly higher than the 16-17% recorded in the AHURI and ANU surveys. It may be that our respondents (and Better Renting’s) tend to be more active as regards tenancy issues (e.g. following tenancy-related media and social media where our survey was advertised); another factor might be our survey running later in the emergency period.

In our survey, 38% of respondents had asked for a rent variation, and of those asking about 43% got a variation – for 16%, the variation they asked for, and for 27% an acceptable counter-offer made by the landlord or agent. Half of those requesting variations had been met with refusal or no response, while 12% received a counter-offer they could not accept.

Of renters who got a variation, those who got a reduction

57%

84%

74%

64%

50%

60%

75%

47%

12%

Notably, among those who did not ask for a variation, almost a quarter (24%) said this reflected fears that it could trigger negative consequences, and 15% said they felt it would be pointless, while about half said they had no need for a reduction. This discouraged group suggests the ‘success’ rate of requests is apt to be overstated. In total, almost half of our respondents were either discouraged from asking for variation, refused a variation, or received an unacceptable counter-offer. Only 16% got an acceptable variation. Without quantifying it, the AHURI survey also reports that discouragement, fear of negative consequences (a black mark on their tenancy record) and embarrassment were prominent themes in comments from respondents who had not asked for a variation (Baker et al. 2020: 11).

Of renters who got a variation, those who got a deferral

37%

16%

22%

36%

50%

36%

13%

42%

57%

Figure 5.1: Rental survey respondents: breakdown according to whether rent reduction sought and negotiation outcomes

Of renters who got a variation, those who got a combined reduction and deferral

Of renters who got a reduction, mean reduction weekly (%) Of renters who got a deferral, mean deferral weekly (%)

13%

$105

$153

$146

$116

$86

(-14%)

(-22%)

(-29%)

(-24%)

(-20%)

$216

$201

$139

$173

$108

(-31%)

(-24%)

(-30%)

(-34%)

(-25%)

Source: Authors survey

About two-fifths (41%) of our survey respondents had lost income during the emergency period. Just over half (55%) of those who lost income asked for a rent variation (and they comprised the majority (67%) of all respondents seeking variations) â€“ see Figure 5.1. 44% of those who lost income and sought a variation received a variation, and 42% were refused. Of those who did not ask for a variation, more than half were discouraged by fear of negative consequences (38%) or the feeling that asking was pointless (20%). These results, and results for selected other categories of respondent, are presented in Table 5.2 below. Renters who received the Coronavirus Supplement or the JobKeeper payment were more likely to ask for variations, and to get them. Single parents, persons in group households and persons born overseas were more likely to ask; group households and overseas-born persons were also less likely to get a variation. Young persons (18-24 year-olds) were also less likely to get a variation, and more likely to be discouraged from asking. Table 5.2: Renters who asked for variations, were discouraged, and who got variations Proportion of respondents

Asked for variation

Discouraged from asking

Of those who asked, got a variation

38%

24%

43%

41%

55%

26%

44%

Renters who 28% received Coronavirus Supplement

42%

21%

47%

Renters who received JobKeeper

19%

55%

16%

58%

Single parents

50%

30%

45%

Group households

20%

55%

18%

33%

18-24 year-olds

38%

33%

Born overseas

23%

50%

14%

34%

Women

62%

38%

25%

47%

All respondents Renters who lost income

We also found only 30% of respondents who said they were in rent arrears had got a variation, and 18% who were keeping up with rent but feeling hardship in other ways had got a variation. In the ANU survey, younger renters (18-24) were more likely to request variations (21% did so); but as in our survey were less likely to be successful (53%). By income, renters in the second quintile were more likely than other income quintiles to request variations (24% did so), and somewhat more likely to be successful (81%); the middle quintile was both least likely to ask (10%) and least successful (53%). (Interestingly, 14% of renters in the top quintile asked for variations, and scored a 100% success rate.) (Biddle, et al. 2020: 14). The AHURI survey does not report on the characteristics of renters seeking and receiving rent variations, except whether they have dealt directly with their landlord or 81

with an agent: perhaps significantly, those dealing with agents were more likely to make a request, and less likely get a variation. As noted in Chapter 3, another perspective on the small proportions of renters getting rent variations in all the surveys is offered by APRA’s loan deferral data, which show almost 305,000 rental investor loans were deferred at the May peak (APRA 2020). This suggests about 12% of PRS properties were associated with a deferred loan at that time. Furthermore, calculations by Leishman et al. (2020) on the prevalence of housing stress in the emergency suggest just 4% of landlord households were themselves in housing stress. We also asked whether respondents had sought help from a government conciliation service. Of those who sought a rent variation, the large majority (88%) had done so themselves, and not accessed conciliation. 41% of respondents said they did not know that such were available, while 22% felt there was no point seeking assistance, and 15% feared negative consequences if they applied. Among those who had lost income, the rates of not knowing of the government services (43%) and fear of negative consequences (17%) were higher. 5.4.2 The terms of variations: reduction/deferral, and quantum For the minority of renters securing rent variations in 2020, most of the sources indicate that rent reductions were the more common type of variation. The less common arrangement was deferral, such that a renter is not required to pay some part or the whole of the rent that would otherwise be payable now, but is liable to pay it later. In our survey, 57% of respondents who secured variations got a reduction, and 37% a deferral (and 6% made an arrangement that involved a combination of the two). This rate of deferral arrangements is similar to that found in the Better Renting survey (36%), and lower than the RBA/MRI estimate (50%), but significantly higher than the rates in the ANU and AHURI surveys (16% and 22% respectively)4. Possible explanations for our higher rate include our survey recruiting more ‘active’ respondents who wanted to ventilate this type of arrangement; on the other hand, these respondents might know more accurately the nature of their variation, whereas the representative samples in the other surveys might tend to confuse reductions and deferrals – an issue to which we return in the next chapter. There may also be a geographic factor, and differences between jurisdictions. NSW government data show its conciliation process produced deferrals at the same rate reported in our (NSW heavy) survey. In Victoria, the sample of registered variation agreements shows 74% were for reductions, and 14% for deferrals (and the remainder could not be categorised on the information recorded). In Queensland, 47% of conciliated variations were reductions, and 42% deferrals; while in Western Australian, merely 12% of conciliated variations were reductions, and – strikingly – 57% were deferrals. Finally, there may also be a time factor: it may be that reductions became less common, and deferrals more common, as the emergency period went on. Few sources give information on the quantum of variations. In our survey, the average size of the variation was substantial: $105 per week (14%) for reductions, and $216 (31%) per week for deferrals. Substantial deferrals are, of course, a mixed blessing – they mean more money to pay later. The government data show similarly substantial average reductions (from 20-29% across the

jurisdictions) and average deferrals (24-34%). The other surveys in the group do not indicate the quantum of variations, although the Better Renting survey reports one-third of reductions were of a ‘trivial’ amount. Focusing specifically on the rent deferral scenario, if the typical size of deferral is indeed $216 per week, that would imply that if continued for 6-9 months the accumulated debt would total some $5,600-$8,400 – a very substantial amount for a low income renter to repay, should their landlord allow the arrangement to continue over such a period. Despite the substantial average size of rent variations, our survey found that 63% of those who had secured them were still in financial difficulty, with 19% stating they remained in rent arrears, and 44% keeping up with (varied) rent payments but feeling the difficulty in other ways. Just over one-fifth (22%) of respondents who had got a variation but still experienced hardship subsequently moved. Crucially, the various surveys show that those granted financial relief on the basis of deferred payment rather than actual reduced rent, equated to between 2% and 7% of all tenants surveyed. While this is rather a large range, we should arguably discount the ANU survey because of being undertaken early in the crisis (May 2020) and, for the AHURI figure, allow for the fact that the 2% deferral rate reported from that research relates to a survey that included social as well as private rental housing. On this basis, it would seem reasonable to conclude that the proportion of all private tenants subject to a rent deferral agreement by the second half of 2020 lay in the range 3-7%. Bearing in mind the sector’s estimated total size – 2.5 million tenancies in 2017-18 according to the ABS Survey of Income and Housing – this implies that deferred rent tenancies in the second half of 2020 could number between 75,000 and 175,000. Even at the lower end, this is a numerically large cohort of people who must be considered especially vulnerable to eviction and possibly homelessness when eviction moratoriums are lifted.

5.5 Chapter conclusion Our survey and other sources indicate that rents in existing tenancies were only modestly impacted by the COVID-19 emergency. Although a loss of income was suffered by a significant minority of renter households, only a small minority of renters – between 8% and 16% – actually got a rent variation. In our survey, a similar proportion of renters were refused a variation, and a larger proportion were discouraged from seeking one. A larger proportion moved, than got a variation. For the few who got a variation, the average size appears to have been substantial, but for a significant proportion – 36% in our survey, variously lower and higher in other sources – the variation is a mere deferral. Those amounts are liable to be paid later, on top of a full amount of rent – or will be arrears actionable by the landlord in termination proceedings. The quantitative modesty, even weakness, of the impact of the emergency on existing tenancy rents is remarkable, considering the surrounding framework of eviction moratoriums and official encouragement for parties to negotiate. These frameworks are considered in detail in the next chapter.

4 The ANU report uses the term ‘freeze on payments’, and it is apparent from the context that it means a deferral of payments. For consistency with other sources, we use ‘deferral’.

COVID-19: Rental housing and homelessness policy impacts, 2021

References ABS (2020) Cat 4940.0 Household impacts of COVID-19 survey; https://www. abs.gov.au/statistics/people/people-and-communities/household-impactscovid-19-survey/latest-release Australian Prudential Regulatory Authority (2020) ‘Temporary loan repayment deferrals due to COVID-19’ https://www.apra.gov.au/temporary-loanrepayment-deferrals-due-to-covid-19-september-2020-accessible-version Baker, E., Bentley, R., Beer, A. and Daniel, L. (2020) Renting in the time of COVID-19: understanding the impacts, Final Report No. 340, Melbourne: AHURI https://www.ahuri.edu.au/research/final-reports/340 Biddle, N., Edwards, B., Gray, M. and Sollis, K. (2020) COVID-19 and housing stress: COVID-19 and mortgage and rental payments: May 2020. ANU Centre for Social Research and Methods https://csrm.cass.anu.edu.au/research/ publications/covid-19-and-mortgage-and-rental-payments-may-2020 Dignam, J. (2020) Rent Due: renting and stress during COVID-19. Canberra: Better Renting https://www.betterrenting.org.au/coronavirus_survey_report Evans, R., Rosewall, T. and Wong, A. (2020) The Rental Market and COVID-19; RBA Bulletin pp75-84; Sydney: Reserve Bank of Australia https://www.rba.gov. au/publications/bulletin/2020/sep/the-rental-market-and-covid-19.html Leishman, C., Ong, R., Lester, L. and Liang, W. (2020) Supporting Australia’s housing system: modelling pandemic policy responses, Final Report No. 345, Melbourne: AHURI

6. Rental housing policy and practice in the COVID-19 emergency Key findings • State and territory eviction moratoriums varied significantly. No jurisdiction implemented a complete moratorium: Tasmania, Victoria and Western Australia implemented relatively strong restrictions; other jurisdictions were considerably less complete. • According to interviewees, the moratoriums were generally easily understood and reasonably effective responses to the crisis. • States and territories also implemented different emergency rent measures: some prohibited rent increases; some implemented frameworks for rent variations; and most offered rent relief through cash payments and/or land tax rebates. • The rent variation frameworks have relied on landlord-tenant negotiation with little direction from governments, and interviewees suggested this caused problems during the emergency and going forward. • In the social housing sector, evictions and rent variations were less prominent issues than managing relations with tenants in lockdown. Actions ranged from participation in the drastic, high-handed ‘hard lockdown’ of the Flemington public housing towers, to new forms of light-touch welfare checks. This chapter examines the pandemic-response emergency policy changes to rental housing regulation enacted by Australia’s state/territory governments in 2020. The largest part of the discussion concerns the key policy innovations of the eviction moratoriums and the rent measures, which were primarily directed to the private rental sector. The chapter also discusses specific policy issues in the social housing sector, and pandemic impacts on management and service delivery in both sectors, including property and tenancy management, tenants advice service provision, and access to dispute resolution. We examine here in more detail the pandemic-triggered rental housing regulation innovations outlined more broadly in Chapter 2, with a view to explaining some of the impacts on markets and individual tenancies indicated in Chapters 4 and 5. The examination also includes discussion of the emergency measures, their formulation and their implementation, informed by perspectives offered by sector stakeholders and service providers interviewed for this research. The chapter is structured as follows. Following a brief review of the policy context for the research, we examine the eviction moratoriums, comparing the different measures implemented in each state and territory, and the perspectives of stakeholders. Next we similarly examine the rent measures: variously restrictions on increases, frameworks for rent variations, and rent relief schemes. Stakeholders were more critical of these measures than the eviction moratoriums, and we consider their criticisms in detail. Then we briefly note the extension of regulation to boarding and lodging arrangements (and flag this for further investigation in the second report of this project),

COVID-19: Rental housing and homelessness policy impacts, 2021

before considering some aspects of the emergency response in relation to social housing. Finally, we briefly note and flag for further investigation some pandemic-driven changes in practices in tenancy management, tenants advice service provision, and tribunal procedures.

6.1 Context: long-term and immediate The long-term context is the development of rental housing regulation as an area of state/territory responsibility, in which all jurisdictions have adopted a broadly common legislative model of • mild consumer protection – i.e. standard forms agreement, prescribed terms and notice periods, accessible dispute resolution through a tribunal5; • market rents; and • ready but orderly termination of tenancies through a tribunal. There are, however, many differences between jurisdictions in the details of their legislation and, as jurisdictions have reviewed and amended their legislation over the years, they have done so independently: with only one exception6, there has been no concerted effort to harmonise laws nationally. Nor is there any recent history of residential tenancy regulation being part of a response to a public health emergency, or an economic emergency: the nearest precedent is the regime of rent and eviction controls implemented nationally at the commencement of the Second World War and continued patchily in some jurisdictions in the post-war period. The immediate context for the emergency measures can be quickly sketched. Among our interviewees, the tenant advocates observed the first effects of the pandemic in early February 2020, when they started taking enquiries from international students unable to commence their courses and tenancies7. By the third week of March, public health emergencies had been declared by all Australian governments, imposing distancing requirements and bans on gatherings; banks were offering loan deferrals, with the support of financial regulators; and tenant organisations and other housing advocates had begun calling for a national eviction moratoriums in line with measures already being adopted in some international jurisdictions (Maalsen et al. 2020). In the fourth week of March, non-essential businesses were ordered closed, the Australian Government announced the Coronavirus Supplement to the Jobseeker payment and opened access to superannuation funds. The Tasmanian and New South Wales parliaments passed eviction moratoriums and other emergency legislation. And on 29 March the National Cabinet announced moratoriums on evictions of commercial and residential tenants, and encouraged landlords and tenants to negotiate about rent obligations, with the state and territories to give the announcement effect. 5 In all states and territories this is the jurisdiction’s Civil and Administrative Tribunal, except Western Australia and Tasmania, where the Magistrates Court hears tenancy matters and operates in a less formal mode than usual. 6 The exception is the more or less nationally consistent regulation of residential tenancy databases, the result of an intergovernmental process commencing 2003, and finally concluding in 2018 when the last jurisdiction (the NT) enacted the provisions. 7 One advocate reflected: ‘I remember when we got the first enquiry and I thought oh wow, I never thought COVID would impact on my work!’ [Tenant advocate 1].

COVID-19: Rental housing and homelessness policy impacts, 2021

Most of the states’ and territories’ emergency measures were implemented by the third week of April, with some rent relief measures coming later. Those postannouncement/pre-implementation weeks were tumultuous – our real estate agent interviewee recalled ‘it turned into the wild wild west for a couple of weeks’ (real estate agent) – with reports of agents directing tenants to access their superannuation to keep paying rent (Karp, 2020) (at least until they were warned off by ASIC) and 16,000 renters pledging a rent strike (Bagshaw and Duke, 2020). This period also saw the announcement by the federal government of the new JobKeeper payment (4 April 2020). This measure, coming on top of previous income support measures, substantially relieved cash flow pressures on households; it probably also relieved pressure on policymakers grappling with questions of the eviction moratoriums and rents. As noted in Chapter 2, all jurisdictions originally set a six-month timeframe on their emergency provisions, but all except Queensland have extended them: to 31 January for Tasmania and the ACT; 6 February for SA; 23 March for the NT, 26 March for NSW; 28 March for Victoria and WA.

6.2 Eviction moratoriums The eviction moratoriums are the stand-out rental housing policy innovation of the COVID-19 emergency. However, nowhere in Australia was there a complete stop on evictions, and the degree of restriction has differed significantly between jurisdictions. As indicated in Chapter 2, the broad approach of all jurisdictions except Tasmania was to restrict evictions for a narrow group of COVID-affected tenants, with most also imposing restrictions of wider application. Each jurisdiction defined the COVID-affected group differently. Table 6.1 summarises jurisdictions’ measures in relation to both the COVID-affected and wider groups.

Table 6.1: eviction moratorium coverage by jurisdiction NSW

Vic

Qld

Tas

ACT

COVID group – rent arrears COVID group – no-grounds COVID group – other grounds Wider group – rent arears Wider group – no-grounds Wider group – other grounds Generally not allowed Allowed subject to scrutiny or limitations Allowed subject to increased notice periods

Tasmania did not distinguish between groups and implemented the most complete moratorium: it stopped landlords from seeking termination on all grounds except nuisance, illegal use of premises and a new COVID-specific hardship ground; although from June it also allowed termination on the grounds that the premises were destined for sale, renovation or for housing the landlord’s own family. After Tasmania, Western Australia implemented the next most complete moratorium, stopping rent arrears terminations for renters in COVID-related hardship, and no-grounds terminations for all renters, as well as instituting a compulsory mediation process prior to termination proceedings on most other grounds. Victoria similarly made all termination proceedings subject to a new dispute resolution scheme and, for matters proceeding to the tribunal, higher qualifications were required for a termination order: in particular, no termination order was allowable for a tenant in rent arrears because of a COVID-related reason and where payment would cause severe hardship. The remaining jurisdictions’ moratoriums were considerably less complete. South Australia stopped rent arrears terminations for COVID-hardship cases; otherwise, the usual grounds for termination remained available, subject to a direction to the tribunal that in determining most types of termination proceedings it consider ‘the need to avoid homelessness’ during the emergency. The ACT stopped rent arrears terminations for COVID-impacted households, without anything more. Queensland stopped rent arrears terminations and nogrounds terminations for tenants in excessive hardship because of COVID-19 (defined more stringently than other jurisdictions), but also added new grounds for termination of tenancies with fixed terms still in force where the landlord is preparing the property for sale or occupying the property themselves. For Queensland tenants not in COVID-related excessive hardship, the usual provisions of termination remained available, as well as the new grounds – so their security was actually diminished by the ‘moratorium’. In New South Wales, termination proceedings for rent arrears, other grounds, and without grounds remained available, but subject in most cases to longer notice periods and, in cases of COVID-impacted tenants in rent arrears, a requirement that the landlord participate in conciliation regarding the rent. The Northern Territory uniquely distinguished between tenancies already existing at announcement of its emergency measures, and those commencing after. Regarding already existing tenancies in COVID-related hardship, rent arrears termination notice timeframes were merely extended and the tribunal allowed to suspend possession orders if satisfied the tenant could pay 30% of their income in rent and eventually repay all arrears. Regarding new tenancies, parties were required to make good faith efforts to resolve the breach or negotiate a new rent, and the tribunal was given a freer hand to alleviate hardship – but overall, the NT’s measures amount to the weakest ‘moratorium’ in Australia. In interviews, stakeholders gave insights into the making of the moratoriums and related emergency measures in housing policy. After the 29 March announcement, the federal government was ‘completely out of it’ (Consumer Affairs officer), as indeed it was with housing and homeless policy in the pandemic generally:

COVID-19: Rental housing and homelessness policy impacts, 2021

The Commonwealth is absent, just vacant. That’s from the ministerial level down. They don’t contribute. [Government housing policy official 1] Stakeholders from government and the NGO sectors both indicated that they were initially expecting a more nationally co-ordinated approach, with certain states to lead on a model approach:

We understood a couple of states were going to go away, work out the framework and we’d work together, consistently from that. It didn’t take long to realise that that wasn’t necessarily happening…. So there were lots of phone calls and emails between people at my level to try to work out what’s happening and keep it consistent. But there were other processes in play as well. [Government consumer affairs official] Government interviewees said it was each jurisdiction’s Consumer Affairs minister and, in some states, its treasurer, who most shaped the form and extent of its moratorium. They agreed with NGO interviewees that there was little consultation with non-government stakeholders, although it is evident from the public record in Queensland that a public relations blitz by that state’s Real Estate Institute caught the attention of the government and led to dilution of its proposed emergency measures. A tenant organisation from another jurisdiction with a relatively weak moratorium also noted REI lobbying, but suggested that the government’s mild approach reflected its own priorities and dispositions, rather than external pressure:

Even though there was a principle of staying safe at home – like, if people were forced out they’d be in the community and creating a risk to public health – it was only if you had been a good working person who had lost income that you got any protection. It was very clearly built into the framework: you deserve to be protected if you are a worthy worker who had suddenly lost work, as opposed to someone who had already been struggling and couldn’t afford the rent, for whatever reason, and don’t elicit sympathy in the same way. Or if the landlord had some other plans [for the property] going on. So the idea that ‘we all need to be safe at home’ was undermined by the limited protections in place [Tenants representative 2] In practice, the tenant advocate interviewees described the moratoriums as ‘decent’, and reasonably easy to communicate to parties in disputes that were covered: ‘The eviction rules – like, ‘you cannot just kick them out’ – worked quite well – you could send a copy and [the landlord or agent] would understand that.’ (Tenants advocate 2). The real estate agent interviewee considered that the moratorium had been ‘fairly positive’, reflecting that it ‘got people a bit of breathing space when it really was manic’, and had helped things ‘settle down’ as households accessed Jobseeker or JobKeeper payments and negotiated arrangements about rent.

I think initially we needed it to come in, and while originally it was the wild west it put a stop to knee jerk reactions – like, there were landlords who would have put tenants out. But I think it’s served its purpose, and it cannot stay forever. [Real estate agent] Outside the mainstream of the rental market, however, a refugee advocate suggested that the moratorium had not really reached the informal housing arrangements in which many of their clients live:

We had people who were evicted from their housing. Often, our clients aren’t on a lease; they’ve renting a room they’ve found on a website, or a friend of a friend…. I know there was a pause on evictions, but it didn’t really translate to our clients. I mean, a lot of them are still housed, but it is like it always is: they scrape by, they do favours. [Refugee advocate 1] The Consumer Affairs interviewee’s perspective was that ‘mostly, people just got used to it, and the sky didn’t fall in’, and that for the purposes of future consideration of reforms to improve tenants’ security, ‘it’ll be evidence that it’s not the end of the world if you don’t have no-grounds termination.’

6.3 Measures regarding rents Aside from the moratoriums, states and territories also implemented a variety of measures regarding rents. As outlined in Chapter 2, jurisdictions generally adopted one or more of three measures: • Prohibitions on rent increases – for the core COVID-affected group, or more widely • Frameworks for rent variations – including statutory conciliation processes, and in some cases provision for a variation determined by a decision-maker • Rent relief schemes – being either cash payments from the state government, or land tax rebates Table 6.2 summarises the mix of measures implemented by states and territories (none were implemented by the federal government). We discuss the details of each in turn below, and then discuss the perspectives of interviewees on the implementation of the measures and practices around rent payments and variations in the crisis. Table 6.2: emergency measures regarding rents by jurisdiction NSW

Vic

Qld

No increases – COVID group

Tas

ACT

Generally speaking, the approach to rents has been an extension of the early advice given by the Prime Minister on the announcement of the eviction moratorium: that tenants should negotiate with landlords and agents and make their own individual arrangements. Governments have largely eschewed mandating variations – whether for tenants as a group, or even in individual cases – and have not even provided much guidance to negotiating parties, although the terms of some of the rent relief programs do provide some. Compared with the eviction moratoriums, the rent variation measures drew stronger criticism from non-government interviewees; their comments, along with the data presented in Chapter 4, point to problems that may persist longer than the emergency periods. 6.3.1. Prohibitions on rent increases Four jurisdictions – Victoria, Western Australia, South Australia and Tasmania – straightforwardly prohibited landlords from increasing rents for existing tenancies for the duration of their emergency period; the ACT prohibited rent increases for COVID-impacted tenancies only. The Consumer Affairs interviewee is from a jurisdiction that prohibited rent increases generally, and explained the factors in the decision:

So we thought, okay do we do an across-the-board cap on rent increases – not more than 5%, or a bit more than CPI but not too much. Treasury really baulked at that because they worried about that being the message: that all rents should increase by 5%. So we thought, do we limit rent increases where people are in COVID hardship – say on JobKeeper – and we thought, that creates massive uncertainty. How would a landlord know if someone is in COVID hardship?... One of [the Real Estate Institute’s] proposals was to prohibit rent increases just for people in COVID financial hardship, and we said, well, way to put a target on someone’s back. They become further marginalised in the rental market, if they need to find another tenancy for whatever reason. [Government consumer affairs official The interviewee was clear that the prohibition was a temporary measure specific to the COVID-19 emergency, and was not regarded as a testbed for other proposals for regulating rent increases, such as caps in rent pressure zones. 6.3.2. Frameworks for rent variations

No increases – wider group

While governments in all jurisdictions informally encouraged negotiations about rents, in four – New South Wales, Victoria, Queensland and Western Australia – statutory conciliation processes were established to facilitate variations. (These schemes are the source of the data analysed in Chapter 4.)

Rent variations - conciliated variations Rent variations - determined variations Rent relief - cash payment Rent relief - land tax rebate Implemented Implemented but limited

The New South Wales scheme co-opts an existing complaints conciliation scheme operated by NSW Fair Trading. It cannot compel landlords to participate, but in any subsequent rent arrears termination proceedings the tribunal will consider whether they participate in good faith or refused ‘a reasonable offer about rent’. Nor can the scheme determine a rent variation; where a variation cannot be negotiated, the only course for the tenant is seek termination of the tenancy. In Western Australia, the new Mandatory Conciliation Service is empowered to require both parties to participate in conciliation, but not to determine a variation without the agreement of the parties.

Implemented but very limited

COVID-19: Rental housing and homelessness policy impacts, 2021

In Queensland, the Residential Tenancies Authority was tasked with formalising rent variations through mandatory conciliation where the partiers were not able to agree. The legislation limited mandatory conciliation to where rent was unpaid and the landlord had issued a show cause notice. The Queensland Civil and Administrative Tribunal was also empowered, after a failed conciliation, to determine the disputeabout unpaid rent, not the rent going forward.

By contrast, in Western Australia the qualifying rent variation could be a reduction or a deferral – so there was no encouragement of reductions, which may be a factor in the large proportion of deferrals recorded in that state’s conciliation data. Also, the income eligibility requirement was different: the rent before the reduction must have been more than 30% of the tenant’s (reduced) income, so there was no discouragement as to the size of the variation.

Like Western Australia, Victoria established a new statutory pre-tribunal conciliation scheme but, unlike the former, Victoria empowered the new scheme (and the tribunal) to make orders to reduce the rent or set out a payment plan. This made Victoria the only jurisdiction to afford tenants a clear right to make an application regarding rental liabilities that could be determined by a decision-maker reducing the rent in arrears and going forward.

The function of the payment differed between jurisdictions. Victoria and Western Australia stipulated that the payment was a credit to the rent account – a benefit to the tenant (Victoria went as far as expressly stating that it was not compensation for the landlord for reducing the rent). So, in Victoria the tenant received both a rent reduction and some amount of their reduced rent paid for them; in Western Australia, the variation might have merely deferred the rent but some amount of rent was paid for them. By contrast, the South Australian and Tasmanian schemes allowed the payment to be treated as compensation to the landlord for the rent reduction.

South Australia and the Northern Territory, it should be said, also made very limited provision for determined variations: specifically, in termination proceedings by landlords. Tasmania, with the most complete moratorium, made no formal provisions regarding rent variations at all. The ACT countenanced the possibility of rent variation agreements, and stipulated that such agreements would reduce the rent, not merely defer it – a small but significant stipulation implemented in no other jurisdiction8. Aside from the ACT stipulation, the terms of rent variations – quantum, duration, whether reduced or deferred – were in other jurisdictions left open. It is possible, however, that state and territory governments have exerted some influence on the terms of variations through their rent relief schemes. 6.3.3. Rent relief schemes Most states and territories implemented rent relief schemes in relation to tenants in hardship – although which party received the benefit of the ‘relief’ is not always clear and in practice differs between jurisdictions and the form of relief. The schemes came in two forms: cash payments and land tax rebates. Victoria, Western Australia, South Australia and Tasmania implemented cash payment programs that looked similar at first glance but differed significantly in the details9. All required that the landlord and tenant make a rent variation agreement, and then the scheme paid an amount (subject to certain caps) directly to the landlord. Other requirements for the payment, the method for calculating the amount, and whether it functions as a benefit to the tenant or landlord, differ between jurisdictions. Victoria, South Australia and Tasmania stipulated that the rent variation was a reduction, not a mere deferral. They also required that tenant was still paying more than 30% of their household income in rent after the reduction. This would tend to encourage reductions, but not so much that the rent is reduced below this threshold. 8 Queensland had proposed to stipulate rent reductions, not deferrals, but backed down after the REIQ’s campaign. 9 At this writing, Western Australia has just announced changes to its scheme, to take effect in January 2021. The discussion here does not include the changed scheme. Queensland also implemented an emergency cash payment program early in the crisis, before the implementation of the Jobkeeper payment and Queensland’s own eviction moratorium and related measures. The scheme made payments where rent negotiations were unsuccessful, and closed 29 April – so it has not influenced rent variation agreements as other jurisdictions’ schemes have potentially done.

COVID-19: Rental housing and homelessness policy impacts, 2021

Land tax rebates were introduced in all jurisdictions except Western Australia, Tasmania and the Northern Territory (the latter does not have a land tax regime). All these schemes required a rent reduction (not a mere deferral), and as a credit on the landlord’s tax bill function as compensation to landlord (rather than a benefit to the tenant). In the ACT – where, notably, almost all rental properties are subject to land tax – landlords who reduce rents for COVID-impacted household by not less than 25% could access the rebate, which was equivalent to half the total amount of the reduction, capped at $2,600. In other jurisdictions, the rebate matched the rent reduction, up to certain caps. In these jurisdictions, however, land tax is charged on only small minorities of rental properties (because their owners’ total holdings sum to less than the minimum land value thresholds): from ATO data, we estimate just 17% of rental properties in New South Wales; 37% in Victoria; 6% in Queensland and 28% in South Australia (ATO, 2020: Table 23). This means the land tax rebate, even though apparently encouraging of reductions, is a weak lever on variations in the large majority of cases. 6.3.4. Rent variations in practice Interviewees gave insights into the practice of rent negotiations, and the functionality of the frameworks and schemes implemented by governments.

Entering negotiations, discouragement and uncertainty As in our survey and the AHURI survey (Baker, et al. 2020: 11), the issue of discouragement from even broaching negotiations was highlighted by several interviewees. ‘We know people were scared’, said tenant organisation representative 4. Refugee advocates observed that, to their knowledge, none of their clients had engaged in negotiations about rent, although many had lost work and they had been excluded from the federal government’s income support measures. The Consumer Affairs interviewee reflected on a case that eventually came up for conciliation, after the tenant had avoided engaging because of embarrassment: This happened the other week – one of my conciliators was almost in tears because she had this bloke, a tenant, who was unemployed for the first time in 30 years, and… he said ‘I’ve lost my job, I lost it four months ago’. And the property manager said, ‘well why didn’t you tell me’ and he said, ‘you don’t understand, I’m not telling anybody.’ It’s heartbreaking. [Government consumer affairs official] 93

The uncertainty of negotiations within the framework – particularly where no provision was made for variations to be determined by a third party, as in all jurisdictions except Victoria – was itself discouraging, according to tenant representatives and advocates.

On rents, the complete [official] failure to give directions on affordable rents, or mandate landlords to give affordable rents, caused massive uncertainty. Not empowering the tribunal to vary rents, or order that people don’t have to pay back rent arrears, have caused massive problems.… Many people just left their tenancies because they had no confidence that they could enter into agreements, or that they’d have protection from eviction once the moratorium ended. [Tenant advocate 2] No-one at any stage in the process can say ‘this is what the rent should be – taking into account all the circumstances.’ No-one can take this decision…. It’s a real problem. And it was because [the Department] and the Tribunal went ‘we don’t want that. We don’t have the training and the expertise to make that decision.’ Which is bizarre. [Tenant representative 2] Both tenant organisation representatives and the real estate agent interviewee highlighted the lack of direction from governments as a practical challenge.

One of the things that we and the REI kept pushing for was more guidance around rent reductions. Landlords and certainly tenants hadn’t been in this space before: what should I be pushing for as a fair and reasonable outcome? [Tenant representative 1] Like, we were basically told just work it out for yourselves, with no guidelines and no power. That makes it very difficult…. It has fallen to property managers to resolve issues without a lot of training or resources to make it happen, which has put a lot of pressure on individual staff – who don’t earn a lot of money, and don’t have a lot of training to do that. And we are talking about working with people facing homelessness and no-one wants to push people out onto the street. [Real estate agent] Ultimately, said the real estate agent, ‘how we’ve managed it is just putting it to the landlord and making it their decision.’

At the end of the day we have to act on the owner’s instructions, and sometimes we’re being instructed to do things we don’t necessarily agree with. So the stress on agents shouldn’t be underestimated, and it has been a bit overlooked at this stage. [Real estate agent] A significant aspect of the uncertainty concerned the timeframes implicit in an unguided process of informal negotiation and then, where available, conciliation. Tenant representative 1 reflected that ‘policymakers had a different timeframe as to what would be reasonable for something to play out without an arrangement in place’:

By that I mean: if you’re a tenant and you lose your job, and you have to pay rent again in two weeks, those two weeks are potentially $1500, depending on your household. So two weeks, four weeks – that’s a long time and a lot of money…. So a landlord might be thinking, ‘oh, well, let me think about this, let me see the evidence and let me work out what my finances are’; their timeframes are going to be a few weeks, and they think it is reasonable to reply in one week or two. In ordinary times we 94

COVID-19: Rental housing and homelessness policy impacts, 2021

might think that’s reasonable too, but from a tenant’s point of view, it is not reasonable, because they are accumulating debt at such a high rate. [Tenant representative 1] The tenant advocates also observed specific problems in the negotiation process around information requirements and privacy concerns. It was implicit in the eviction moratoriums’ distinction of core affected group that rent variations were for individual hardship cases only, and not a change in the market level of rents, so information about tenants’ income and finances became a key issue in negotiations.

There was a lot of uncertainty about what could be required and what tenants could say no to. Really intrusive requests for info by real estate agents. I think the REI may have sent around a template for rent reduction requests, because they were very similar, and they had some incredibly invasive stuff, and focused on people’s savings, not income. The government wasn’t controlling that, and that was a significant problem. [Tenant advocate 2] The real estate agent, who worked in the same jurisdiction as tenant advocate 2, confirmed:

The REI has put together an application form, that asks for information – what their income was, what it has been reduced by – and we pass that on to the owners and ask what they can afford to reduce the rent by, and then it’s basically negotiation. [Real estate agent] In this way rent negotiations were subject to gatekeeping, with agents administratively admitting for consideration hardship cases but excluding tenants seeking reductions to reflect changing market conditions rather than their own circumstances.

Dispositions in negotiations Interviewees gave their impressions as to the wide range of approaches taken by parties to negotiations. The tenant advocates and representatives acknowledged that the nature of their work meant they tended to see the difficult and unsatisfactory cases, and ‘there were definitely examples of landlords proactively offering rent reductions, or responding really well’ (tenant advocate 2), but they highlighted that the negotiation framework did not prevent intransigence. I had no luck in budging agents on rent. They’d just say, ‘the landlord’s in financial hardship, we’re not in a position to do it.’ And you’d say ‘prove it’ and they’d say ‘the landlord’s in financial hardship, and is not in a position to do it’. There’s no stick in that case. They’re thinking: we’re going to live with this person here, for the moment we can’t kick them out but they’ll just accrue a debt and we’ll force them to pay it back later on. [Tenant advocate 2] The government put a lot of faith in landlords understanding that it was in their interest to agree to a reduction, and the market would regulate it. Whereas we saw a lot of people whose landlords and agents just refused to negotiate, who said ‘we will not enter into negotiations, we understand we cannot evict you for three months, but then we will try to evict you and get the full rent by money order when the three months is up.’ [Tenant advocate 1] 95

Tenant representative 2, who worked in a jurisdiction with a falling rental market, suggested that the particular circumstances of the emergency gave landlords a reason to delay or refuse to negotiate:

One of the assumptions of the eviction moratorium is that a landlord will always evict if a tenant isn’t paying the rent, which in a regular context is true. But in this context, it is very rational for a landlord not to evict a tenant. Because if they are seeing a whole lot of properties go vacant around them, even if they are not sure their tenant can ever pay off the debt, they have a paper debt, a potential for payment. If they kick them out, they have the definite of a vacant property. So, it becomes more rational to keep the tenant, don’t kick them out but don’t bother negotiating them, and hope that they get a new job and pay off the debt at some point in the future. [Tenant representative 2] This interviewee added ‘the role of agents in this is a bit unexamined’. Whereas our real estate agent interviewee said they acted on landlords’ instructions, tenant representative 2 suggested that for some agents, ‘their interpretation of the landlord’s best interest is not actually the landlord’s best interest’, and may refuse or delay variations so that landlords ‘still owe the agent the commission at the higher rate’ – even if it is on a paper debt (tenant representative 2). The real estate agent was more positive about the disposition of most parties in negotiations: ‘people really do love being given the opportunity to give without costing them anything – you know, a feel-good factor – and people are good by nature.’

I think the majority of people, if they say no to something like that, it’s because they just physically can’t for whatever reason, not because they don’t want to. And if they can help someone else, they will do it if they can. The compassion and the understanding. And even with tenants as well: a landlord might come back and say ‘we can give you a $50 a week rent reduction’ and the tenant will go ‘no, that’s too much, how about a $30 a week rent reduction’. People really are, when push comes to shove, not as cynical as you see in the media. Where people don’t help each other, a lot of the time it’s because in their personal life it’s just not possible, not because they don’t want to. [Real estate agent] The qualification placed by the interviewee on this positive appraisal – ‘if they can’ – is significant. This was an assessment by landlords of what was financially possible and for some, this did not admit suffering a loss.

What it comes back to is where the landlord is financially, with their loans and their banks…. If they are in a tight financial position themselves and they have to repay money themselves and they don’t know where they’ll get it from, then ‘I can’t waive it but I can defer it’ may be all they can financially do. But by and large, when people can do it, they are waiving it. [Real estate agent] The terms of variations The terms of variation agreements – in particular, deferrals – were a major concern for tenant advocates and representatives. Tenant advocate 1 saw deferrals as the unfair exploitation of tenants’ vulnerability: Yeah, there were a lot of tenants agreeing to deferrals because they were terrified of being evicted. Without pushing for more. [Tenant advocate 1] 96

COVID-19: Rental housing and homelessness policy impacts, 2021

The Consumer Affairs interviewee saw tenants in conciliation entering into deferrals of their own volition, but was also concerned that they were a potential problem going forward:

A lot of the agreements are like: the tenant will pay off the arrears at $50 or $100 per fortnight, but we’re trying not to lock in too many end dates – it will just continue until the arrears are repaid, so if there is slippage along the way they can be a bit flexible. It’s something I’ve been really mindful of, and I keep bringing the conciliators back to it, it does concern me that at face value there is a lot of deferral… but the conciliators keep assuring me that they do raise it with tenants in private session. For tenant organisation representative 4, the different types of arrangements meant there was a high risk of confusion in individual cases, and confusion across the rental sector.

There’s a bunch where there’s no agreement. There’s a bunch where it’s deferrals – either cajoled or they’ve Pollyanna-ised that they are going to get their job back. And there are some where they’ve got reductions…. To be honest, I don’t think anyone knows what’s going on. Like, I’ll get a Facebook message from someone, they’ve filled in the form asking for a reduced rent, provided details of their income, and later they get a message to pay their rent – like, they thought they’d done everything to get a rent reduction. All this shit is going to hit the fan further down the line. [Tenant representative 4] In New South Wales and Queensland in particular, each state’s Real Estate Institute had publicly pressed for deferrals in preference to reductions: for example, the President of REINSW addressed tenants on social media advising ‘I need to be clear that any rent reduction that is agreed is not waived. It is just deferred and you still have an obligation to make those payments.’10 Tenant representative 2 characterised this as ‘leaning on the negotiation process at a macro level’:

They were trying to affect the parameters of negotiation from above, and they did that with the stuff about deferral rather than waiver. [Tenant representative 2] Rent variations and rent relief – for whom? Interviewees also reflected on rent variations in the wider contexts of rent relief schemes and income support. The real estate agent’s perspective was that tenants, rather than landlords, had (universally) benefited from the Jobseeker Coronavirus Supplement and/or JobKeeper payment, which detracted from the case for using rent variations to share the costs of lost income: The REI’s position was concern for transferring the hardship on one group of people directly onto another group without the benefits that tenants sometimes do have: JobKeeper, JobSeeker with the additional benefit. [Real estate agent]. By the same token, the agent was apprehensive about the impending reduction and withdrawal of these supports: 10 @REINSWnews (Twitter, 16 April 2020, 1.05 pm AEST) <https://twitter.com/REINSWnews/ status/1250621411871211520>.

When the financial support from the government is rolled back, we could see a second wave of people in financial distress and rents may again have to be looked at. I’m assuming [laughs] the government has a plan for that, and they’ll tell us when it is announced, but it doesn’t take Einstein to see that when you stop paying people what’s been keeping them afloat you’ll go back to having issues with affordability. [Real estate agent] The Consumer Affairs representative observed that the Coronavirus Supplement, in particular, ‘has definitely benefited lower income people who have always struggled’ – and they made the point that this was to the advantage of landlords too:

We have an advisory committee and [the Real Estate Institute] have been saying rent arrears and utilities are the lowest they’ve ever been, particularly for the Jobseeker people. They’ve been able to keep up. [Government consumer affairs official] Tenant representative 2, on the other hand, considered that ‘the support given to landlords far outweighed the support given to tenants’:

Even things like the Victorian and Queensland $2000 payments. As a tenant, you had to impoverish yourself before you were eligible. You had to have less than $5000 savings. Landlords had no equivalent requirement to lose their wealth before getting – weak as it was – the mortgage holiday, or land tax exemptions or any of that. All of that came in no matter how wealthy you were, and in fact the land tax thing is geared to wealthy people more than others! So there was an inequality in that. That’s not so much an unintended outcome, as coming from an unexamined ideology, and they’re setting up some people to come out of COVID much better than others. (Tenant representative 2)

6.4 Boarding and lodging The application of the eviction moratoriums to boarding and lodging arrangements is another aspect of the regulatory response to the COVID-19 emergency that deserves attention – more attention than it will be given here, though we intend to remedy that in the final report of this project. In most jurisdictions, boarders and lodgers are excluded from the mainstream provisions of residential tenancies and subject to their own legislative regime that affords even fewer prescribed rights and less security to resident. In some jurisdictions, landlords can terminate boarding and lodging arrangements without going through the tribunal. In Western Australia, boarding and lodging arrangements are excluded from the Residential Tenancies Act and do not have their own legislation either. When they introduced their eviction moratoriums, all jurisdictions except the ACT applied the measures to boarding and lodging arrangements. This was a significant extension of regulation and scrutiny to a sector that has historically maintained that it needs a free hand to operate. The Consumer Affairs interviewee said their department was monitoring the experience with a view to the sector’s post-emergency regulation.

COVID-19: Rental housing and homelessness policy impacts, 2021

6.5 Social and affordable housing 6.5.1 Policy innovation and social landlords Although the eviction moratoriums were primarily addressed to potential problems in the private rental sector, public housing and community housing were also covered, except in New South Wales, where social housing was expressly excluded from the emergency measures. Prior to the 29 March announcement, the community housing sector was absent from the calls for a moratorium, and certain community housing provider representatives publicly challenged the proposed broad moratorium. According to one CEO cited in a local newspaper:

Tenants whose circumstances haven't changed may be confused by the broad messaging [that eviction is outlawed], stop paying their rent in the short term and accrue debt they will struggle to pay down the track. (Latifi 2020) Considering the role community housing providers play in Australian housing advocacy organisations and campaigns, tenant organisation representatives were disturbed by what they perceived as having been revealed by the pandemic:

The difference in approach between public and community really came out… community housing leaders were out there writing op-eds about how if their tenants got a moratorium, they’d stop paying their rents, they’d bankrupt community housing and the whole thing would collapse… the community housing sector really fought against the eviction moratorium. And it showed that they don’t have good relations with their tenants, some of them … Partly this is because the community housing sector are like small-holding landlords, and they’re operating on credit: they don’t have the finances that a state government does to weather a big storm. And like private landlords they punch down instead of up. [Tenant representative 2] The CEO views cited above chime with concerns voiced by a NFP housing peak body in Scotland. As noted in Chapter 2, this involved a contention that social landlords were mistakenly included within Scotland’s 2020-21 eviction moratorium, and a claim that this had encouraged wilful rent non-payment by housing association tenants believing themselves immune from legal action (Bookbinder 2020). At the same time, the extent to which the rather similar community housing CEO perspective mentioned above was representative of the Australian NFP housing sector more broadly is uncertain. Relevant here is that – as part of a wider statement on housing management responses to the pandemic (and irrespective of the different moratoria terms applicable across the country) – the Community Housing Industry Association committed its members to ‘[avoid] evicting anyone for rent arrears resulting from the coronavirus outbreak’ (CHIA 2020). It is also fair to acknowledge that this is a relatively narrow form of words which does not proscribe evictions in any or all circumstances. While formally excluding social housing from the moratorium, the NSW social housing minister did approve a marked softening of the standard approach to managing rent arrears in public housing. This involved toleration for an unusually low rate of arrears repayment as a condition for holding off eviction action.

Determining rent-assessable income At the start of the pandemic public housing authorities and community housing organisations committed to not assessing, for the purposes of rent calculations, any additional tenant income being received under Commonwealth Government special measures (CHIA 2020). This refers, in the main, to those in receipt of the Coronavirus Supplement which was made available during 2020 to recipients of JobSeeker Payment, Parenting Payment or Youth Allowance (see Chapter 2). The significance of this landlord commitment comes from the fact that in Australia most social housing tenants pay rents subject to a rebate that reduces the amount paid to a percentage of the household’s income – usually 25%. Special purpose payments are typically excluded from assessable income; the treatment of the emergency payments was consistent with this approach, although excluding payments of this size – the amount of the Supplement was initially almost equal to the standard single person JobSeeker rate – was unprecedented. As a result, recipient tenants retained 100% of the Supplement rather than 75%, and had the benefit of a disposable income boost of 130% - rather than approximately 100%, had the additional payment been treated as assessable. 6.5.2 Social housing management impacts and responses

Community housing experiences, sector-wide As implicit in the previous section, the economic shutdown at the start of the pandemic raised some alarm in the social housing world. Given their status as non-government entities subject to commercial risk, many community housing providers were initially concerned that reduced rental income could imperil financial solvency. In practice, at least as far as their mainstream social housing business has been concerned, it seems that any hit to social landlords’ rental income has been typically minimal. More in-depth research has reported that CHPs generally experienced only a small rental income reduction during the early phase of the pandemic – typically less than 3% (NHFIC 2020). Indeed, a CHP stakeholder interviewed in our own research reported that:

We’re actually at historically low levels of arrears – we’ve never had arrears as low as we have at the moment. And that’s a combination of factors: people want to stay in their houses and they want to pay their rent; but also the pro-active phone calls, because people aren’t in the field, and the level of contact has really driven the arrears down as well. [Community housing provider] This alludes to new working practices adopted by social landlords during the COVID-19 pandemic. Just as in many other fields, there was a substantial switch to home-working and online – rather than face to face – communication. Alongside this, was a re-focusing of tenancy management and support activity towards ‘vulnerable tenants’:

In the hard lockdown, we weren’t encouraging anybody to go out and do any non-essential property inspections, like the annual visit. We replaced that with a [tenant] wellbeing check – we already had a roster for elderly residents and we expanded it out to everyone at risk – elderly, mental health, disability support. We’ve been maintaining that. [Community housing provider]

100 COVID-19: Rental housing and homelessness policy impacts, 2021

Whether or not it reflects universal practice among CHPs, this is in tune with the pandemic-practice commitment made by the Community Housing Industry Association on behalf of member organisations to:

[Provide] support for older people, Indigenous Australians, and vulnerable tenants – for example, by phoning every tenant to assess need in the event of self-isolation, and using staff and volunteers to help by funding or linking with other local charities and coordinating volunteers in circumstances where people are having difficulty accessing essential food or medication [CHIA 2020 p4]. More in-depth research has also reported that additional services typically provided by CHPs at some level during this period has included ‘temporary rental discounts for tenants who had lost work, welfare calls, referrals for support and practical support, and direct delivery of practical support such as food hampers’ (Stuart 2020 p3). The more intensive engagement with tenants is perceived as having yielded positive benefits:

We were actually in a lot more contact with our residents, and had been able to pick up on those who had support needs that hadn’t been picked up before, or mental health issues that they’d had in the past and that started to flare up because of the pandemic … That phone contact, the wellbeing check, was really important to keep the connection and make sure they knew there was someone they could contact if something was going wrong or they weren’t feeling well. [Community housing provider]. Affordable rental housing impacts Affordable rental housing, in this context, refers to housing targeted at low-tomoderate income households such as low-income workers. Qualifying tenants are usually charged rent set on a ‘discount to market’ basis – often calculated as 75-80% of the rent for an equivalent property on the open market. Given their economic profile, it would be expected that tenants of affordable rental housing would be vulnerable to possible loss of income due to the recession to a much greater extent than most social renters. That small minority of community housing provider portfolios containing substantial affordable rental representation is potentially more vulnerable to rental income losses with implications for financial viability. Given its small scale, however, our own research was unable to investigate this directly. Nevertheless, neither of the two more focused studies that have researched pandemic impacts on community housing (NHFIC 2020; Stuart 2020) has suggested that serious problems of this kind have yet occurred. 6.5.3. High-rise lock downs The most drastic response to COVID-19 in the sphere of social housing management was the ‘hard lockdown’ of public housing tenants in nine towers in the Melbourne suburbs of Flemington and North Melbourne. This occurred on 4 July, at the beginning of Victoria’s pandemic second wave, following an initial outbreak of 23 cases across 12 households among the estates (ABC 2020). The rationale provided by the Victorian Government was that the towers and residents were uniquely vulnerable due to the density of the estates, their many shared facilities, and their demographic and socio-economic characteristics. Furthermore, Government submissions to the Victorian Ombudsman’s report into the matter suggested that the estates were seen to be rife with crime and non-compliance (Victorian Ombudsman 2020). A detention order was 101

issued to come into effect immediately following the Premier’s announcement. While the directive applied for 14 days, the Government’s intention was to limit them to five days. The restrictions were relaxed after five days for all bar one tower, whose residents were deemed close contacts of positive cases and thus required to self-isolate for the full two weeks. The restrictions were immediately enforced by a large police presence of some 500 officers, suggesting that tenants were expected to be non-compliant with the health advice (Melbourne Activist Legal Support, 2020). For some residents and commentators, this reflected a prejudiced view towards public housing tenants and the over-policing of African residents (e.g. Kelly et al. 2020; Murray-Atfield 2020; Button & Szego 2020). It is also true that, given the decision not to forewarn tenants so as to allow them adequate time to prepare, some degree of non-compliance could be reasonably expected. The decision to enforce an immediate and severe lockdown was also followed by a suite of co-ordination problems, including the allocation of culturally inappropriate supplies, the obstruction of community-led relief efforts, confrontations between police and residents and legal observers, the erection of fencing around the estate, and generally poor communication with tenants (Melbourne Activist Legal Support, 2020).

inspections; other forms of property inspection and access were still allowed subject to restrictions. These restrictions and rules were implemented in state and territory public health orders, and some jurisdictions made further provisions in their emergency legislation restricting rights of entry to rented premises. One of our tenant organisation representatives highlighted how stressful continued access had been for some tenants; the real estate agent interviewee said it was stressful for agents too. The agent also noted that in response to restrictions on showing prospective tenants through properties, they had changed their application process so that interested tenants had to apply before viewing the properties, with the agent showing the property only to applicants assessed suitable for an offer. The agent was concerned about the implications for access to housing:

I do worry that that is going to be difficult for people who are not ideal on paperwork. We’re not meeting them beforehand, we’re not talking to them, and for tenancies that don’t look amazing on paper, they’re going to have more trouble getting a go. [Real estate agent] However, the agent did not expect this would be a permanent change, as it was also more laborious for agents.

The Victorian Ombudsman found that ‘people found themselves without food, medication and other essential supports. Information was confused, incomprehensible, or simply lacking. On the ground few seemed to know who was in charge. No access to fresh air and outdoor exercise was provided for over a week’ (Victorian Ombudsman 2020: 4). For former refugee residents with past experience of state violence and oppression overseas, the heavyhanded response was particularly distressing (ibid.). The Ombudsman concluded that the lockdowns appeared to breach the law and violate the human rights of the estates’ residents.

6.6.2. Tenants advice

The episode in Flemington and North Melbourne portrays a decision-making process that failed to take adequate account of the lived realities of residents, their rights and needs, and their capacity to be willing partners in a public health response. It is well known that social housing cohorts are increasingly ‘high needs’ (e.g. Pawson et al. 2020), yet the preparations that might be required for, for example, a person with a disability, chronic illness or addiction appear to have been ignored or dismissed. At the same time, many residents expressed a desire for a consultative and collaborative approach that is at odds with the enforcement approach of the Victorian Government and Police (Carrasco et al. 2020; Taha 2020).

The tenant advocate interviewees thought the shift to working from home had gone ‘fairly smoothly’ and they had demonstrated that they could effectively provide phone information and advice from home. However, tenant advocate 1 was concerned that ‘there’s a certain section of tenants who’ve we’ve just not been able to connect with under COVID’:

6.6 Tenancy management and service provision Aside from housing-specific policy changes, general measures such as social distancing, restrictions on gatherings and business closures had implications for tenancy management and service provision in the rental housing sector. In this final section of the chapter, we briefly review the issues identified by interviewees. Rather than exhaustively documenting changes in practice, the review is selective, highlighting issues where enduring change may happen, or where further examination is required. 6.6.1. Tenancy management practices Early in the emergency, the National Cabinet announced restrictions on gatherings and social distancing rules that specifically prohibited open house 102 COVID-19: Rental housing and homelessness policy impacts, 2021

Most states and territories have specialist NGO tenant advice services whose primary mode of work is information and advice provided by phone, although many also offer face-to-face drop-in sessions, representation in the tribunal – including duty advocacy – and community education. At the start of the emergency period the services experienced an ‘avalanche’ of calls from tenants for assistance (tenant advocate 1); at the same time, they also had to close their offices and shift to working from home.

There are some clients you just need to see face-to-face, or they can drop-in – when they happen to be out of bed, or happen to be in the area, or have the bus money and can drop in and see you – and that’s just not possible under COVID. [Tenant advocate 1] Even more of a concern was the impact on their tribunal work. 6.6.3. Tribunals The COVID-19 emergency produced two major issues for the states and territories’ tribunals: first, the tribunals stopped their usual mode of in-person hearings and shifted to phone hearings and, secondly, the eviction moratorium and rent variation frameworks in some jurisdictions have expanded or created new conciliation processes outwith the tribunals, which could endure after the emergency. The tenant advocates saw the shift to phone hearings as a significant change with potentially positive and negative impacts. One positive was that phone hearings made it possible for advocate to offer representation services to more tenants: ‘we’ve found we can take on representation more, because previously 103

we couldn’t take on multiple hearings in one day’ (tenant advocate 1). However, the advocates said that the shift to phone hearings had meant that they were not doing duty advocacy as previously, and so worried that fewer tenants – particularly those unaware that assistance was available – were actually being represented. They also felt less sure in phone hearings whether the hearing would produce procedural orders (directions) for the determination of the matter at a subsequent hearing, or orders to determine the matter immediately. They also noted that a missed call or mistaken number could mean someone missing their hearing and having the matter determined in the absence:

My case was where the tribunal called the tenant 20 minutes after the hearing started, and asked if he was a different person, he said no. They’d got the numbers mixed up on the files, and when they did call it was like ‘oh I was waiting for the call’ and ‘we just terminated your tenancy’. [Tenant advocate 1] Our real estate agent interviewee had had no experience of the tribunal during the emergency period; nor did we collect data from or interview tribunal officers. We will investigate the tribunals’ experience more for the second report of this research. Regarding the new extra-tribunal conciliation processes, the consumer affairs interviewee said their government was monitoring its effectiveness with a view to keeping it permanently, to divert matters from the tribunal. A tenant organisation representative in another jurisdiction understood that their government was taking a similar stance. Our real estate agent interviewee ‘wholeheartedly agreed’ with measures to divert matters from the tribunal as a general policy. For tenant advocates and representatives, though, the shift from open tribunals, already less formal and independent than courts, is more contentious. These issues will also be investigated further in our second report.

6.7 Chapter conclusion Australia’s emergency measures in rental housing policy were formulated by state and territory governments, in communication with each other and mostly on a common model, but with substantial differences in the details – a familiar pattern from the history of rental regulation. Their eviction moratoriums restricted some – but not all – termination proceedings against a core COVID-hardship group, with rather less – or no – additional protection for tenants more widely. Tasmania, Western Australia and Victoria had relatively strong moratoriums; the Northern Territory the weakest. Frameworks for rent variations relied on negotiations between individual parties, mostly eschewing a determinative role for state agencies, and applying only a little influence through the terms of rent relief schemes (delivered variously through cash payments and land tax rebates). In practice, interviewees found the eviction moratoriums to be a readily understood measure that took some pressure out of the private rental sector after the income shocks widely experienced in the early emergency period. The rent variation frameworks, however, were regarded less well. For some renters, it appears the uncertain timeframes, requirements and outcomes of negotiations were such that they opted instead for the certainty of terminating their tenancy and liabilities; for others, negotiations have produced deferred and accumulating liabilities and, for yet others, there may be uncertainty as to the terms of what they and their landlord agreed. 104 COVID-19: Rental housing and homelessness policy impacts, 2021

Although it was evidently absent from the design of the states’ and territories’ measures, the Australian Government was arguably influential on them, through its income support measures. These probably did more to absorb the sector’s income shock, and let states and territories – and landlords – off the hook for making adjustments that would share income losses through rent variations. One implication of this is that renters who missed out on the income support measures – notably, non-permanent residents – had the benefit only of the moratoriums and rent variation frameworks, which is the lesser part of the whole response. A second implication relates to what happens as the income support measures are withdrawn. With the eviction moratoriums also due to expire in the new year, supported households who do not snap back to their full pre-COVID-19 employment status and income will be at risk, especially if they are carrying deferred rent liabilities from the emergency. We may yet see cause for further extensions – but without the public health impetus of March 2020. We may also see calls, perhaps under the banner of ‘rent relief’, for governments to pay landlords to settle those liabilities. More broadly, the experience of conducting emergency rent relief schemes might, as suggested by one government housing policy interviewee, encourage state and territory government to become more involved in the provision of rental housing subsidies on an enduring basis. Similarly, their experiences of greater legal security for tenants, greater regulation of boarding and lodging arrangements, and executive dispute resolution outwith the tribunal, may furnish evidence for further reforms. For the social housing sector, the emergency did not prompt major changes in policy, but there will be lessons to be learnt from the high-handed lockdown of the Flemington public housing towers, as well as the community housing sector’s absence from – and in some cases, dissent from – the early advocacy for eviction moratoriums. Perhaps the most pressing issue for the social housing sector arising from the emergency is the question of its capacity to follow up initiatives in temporary accommodation for homeless persons – which is the focus of the next chapter.

References ABC (2020) Coronavirus forces lockdown of Melbourne public housing towers in 3051 and 3031 postcodes. This is what that means for the residents; 4 July https://www.abc.net.au/news/2020-07-04/coronavirus-victoria-melbournepublic-housing-estates-lockdown/12423042 Australian Taxation Office (2020) Taxation Statistics 2017-18 https://www. ato.gov.au/About-ATO/Research-and-statistics/In-detail/Taxation-statistics/ Taxation-statistics-2017-18/ Bookbinder, D. (2020) Eviction protections hit the wrong target in social housing, and its tenants who’ll pay the cost; Scottish Housing News, 18 August https://www.scottishhousingnews.com/article/david-bookbinder-evictionprotections-hit-the-wrong-target-in-social-housing-and-its-tenants-who-llpay-the-cost Button, J. & Szego, J. (2020) ‘'We'll see you after dark': how Melbourne police targeted African men in high-rises’, The Guardian, 11 July, https://www. theguardian.com/australia-news/2020/jul/11/well-see-you-after-dark-howmelbourne-police-target-african-men-in-high-rises 105

Carrasco, S., Faleh, M. and Dangol, N. (2020) Our lives matter – Melbourne public housing residents talk about why COVID-19 hits them hard; The Conversation 24 July https://theconversation.com/our-lives-matter-melbournepublic-housing-residents-talk-about-why-covid-19-hits-them-hard-142901 Community Housing Industry Association (2020) COVID-19: Our Response and Current Priorities https://www.communityhousing.com.au/wp-content/ uploads/2020/04/CHIA-statement-Covid-19.pdf?x12261 Duke, J. and Bagshaw, E. (2020) ‘Real estate agents 'abused' as coronavirus rent-strike grows’, SMH 7 April 2020 https://www.smh.com.au/politics/federal/ real-estate-agents-abused-as-coronavirus-rent-strike-grows-20200407p54hwo.html Kelly, D., Shaw, K. & Porter, L. (2020) Melbourne tower lockdowns unfairly target already vulnerable public housing residents. The Conversation, 6 July https:// theconversation.com/melbourne-tower-lockdowns-unfairly-target-alreadyvulnerable-public-housing-residents-142041 Latifi, A. (2020) Housing Trust CEO: rental payments vital to keep tenants out of debt; Illawarra Mercury https://www.illawarramercury.com.au/ story/6709335/housing-trust-ceo-rental-payments-vital-to-keep-tenants-outof-debt/ Melbourne Activist Legal Support (2020) Interim report on public housing detention directions in Flemington and North Melbourne https:// melbactivistlegal.org.au/2020/07/16/interim-report-on-public-housingdetention-directions-in-flemington-and-north-melbourne/ Murray-Atfield, Y. (2020) Melbourne public housing tower resident says community is being 'treated like criminals' amid coronavirus lockdown, ABC News, 5 July https://www.abc.net.au/news/2020-07-05/coronavirus-lockdownmelbourne-public-housing-residents-speak/12423170 NHFIC (2020) Australia’s affordable and social housing sector: A resilient response to COVID-19 https://www.nhfic.gov.au/media/1414/australias-socialand-affordable-housing-sector-28-october-2020.pdf Pawson, H., Milligan, V. & Yates, J. (2020) Housing Policy in Australia: A case for system reform; Singapore: Palgrave Macmillan Stuart, C. (2020) Responding to the COVID-19 Crisis: What we learnt from talking to community housing organisations; Community Housing Industry Association https://www.communityhousing.com.au/wp-content/ uploads/2020/10/Covid-19-ImpactsProject-webFinal.pdf?x12261 Taha, A. (2020) ‘It was a shock to see police surround our tower blocks. But I'm happy Victoria is helping us at last’. The Guardian, 5 July https://www. theguardian.com/commentisfree/2020/jul/05/it-was-a-shock-to-see-policesurround-our-tower-blocks-but-im-happy-victoria-is-helping-us-at-last Victorian Ombudsman (2020) Investigation into the detention and treatment of public housing residents arising from a COVID-19 ‘hard lockdown’.

7. Homelessness policy and practice Key points: • Income support payments and eviction moratoriums meant the COVID-19 pandemic triggered no immediate increase in homelessness. • Four state governments authorised mass provision of emergency accommodation (EA) for rough sleepers and other homeless people – action that, by September 2020, had encompassed at least 40,000 people. • These governments and their NGO partners were able to facilitate transitions to longer term housing for only a minority of those provided with temporary hotel rooms and similar accommodation. Less than a third (32%) of the 8,000 former rough sleepers who departed EA in the six months to 30 September 2020 had been assisted into longer term tenancies • The rapid attrition of the hotel-housed population and the modest numbers successfully assisted into longer-term housing exposed some weaknesses of program management, as well as the more fundamental systemic flaws of the housing system, especially the limitations posed by inadequate availability of social housing.

7.1 Introduction With the onset of the COVID-19 crisis in 2020, homelessness abruptly shot up the agenda as a policy concern for Australian governments. As the pandemic began to rage across the globe, there was a sudden realisation that a country with at least 8,000 people sleeping rough on any given night and many thousands more in overcrowded or insanitary buildings was thus at higher risk from a public health perspective. With the sudden emergence of such concerns, these forms of homelessness gained a new profile as national vulnerabilities demanding urgent and extraordinary mitigating action. This chapter explores the pandemic’s impacts on homelessness, the emergency policy responses that quickly materialised to counter associated public health and other risks, and the ways that these actions were shaped and implemented. First, to set these events in context, Section 7.2 briefly reviews homelessness trends in Australia over the past decade. Likewise, we present an overview of the significant pre-pandemic homelessness policy and practice developments which, as hindsight now makes plain, provided something of a platform for the emergency action of 2020. This section draws on our recently published companion report Australian Homelessness Monitor 2020 (Pawson et al. 2020). Next, in Section 673, and building on our housing market analysis in Chapter 4, we analyse emerging statistical data on initial COVID-19 impacts on homelessness causation. The primary source here is the monthly statistical data series drawing on homelessness service user records and launched this year by AIHW. Section 7.4, the heart of the chapter, first unpacks the emergency homelessness policy initiatives announced by state governments in the months from March 2020. Next, drawing on key stakeholder interviews, we explore how these initiatives were

106 COVID-19: Rental housing and homelessness policy impacts, 2021

107

formulated and rolled out – which parties were involved and in what ways. We then attempt to calibrate the scale of these programs in terms of temporary and permanent housing placements for formerly homeless people. Before concluding, the chapter then proceeds to draw some comparisons between the somewhat parallel emergency homelessness actions seen in Australia and England in 2020.

Figure 7.2: ABS Census homelessness operational groups, Australia-wide 2011 and 2016

7.2 Policy and practice context for COVID-19 interventions 7.2.1 Homelessness levels and trends 2006-2020 Over the past 10-15 years Australia has seen homelessness rising significantly ahead of population growth. On the most widely accepted measure of homelessness – the point-in-time statistic generated by the five-yearly ABS census – the affected population grew by 14% in the period 2011-16 (see Figure 6.1), up 30% since 2006. An alternate measure, homelessness service users annually assisted, saw numbers rising by 22% over this period – see Figure 6.1. Subsequently, as shown by published AIHW statistics, the annual volume of assisted service users has remained fairly static at the national scale, rising only marginally from 288,000 in 2016-17 to 290,000 in 2018-19. A new set of ABS homelessness statistics (from census 2021) is expected to be published in late 2022. Importantly, as shown in Figure 7.1, recent homelessness trends have varied considerably across Australia, with a tendency for rising numbers particularly apparent in NSW and Victoria – substantially reflecting the pressured housing markets of their capital cities, Sydney and Melbourne. As made plain in Figure 7.2, the relatively broad ABS definition encompasses population cohorts whose insecure or otherwise grossly inadequate housing situation is officially defined as equating to homelessness. However, while street homelessness or ‘rough sleeping’ accounted for only 8,200 people on Census night 2016 (see Figure 7.2), this represented a 20% increase on the figure five years earlier. Moreover, point-intime calibration understates the true scale of the problem in the sense that, in any given month or year, rough sleeping affects much larger numbers. For example, some 42,000 service users logged by specialist homelessness services in 2018-19 had slept rough in the month preceding presentation (Pawson et al. 2020).

Source: ABS Census

Over the past decade a few city councils have undertaken periodic and systematic street counts covering limited city centre localities. As shown in Figure 7.3, these have indicated fairly diverse trends during the 2010s – in some cases (e.g. Sydney) reflecting episodic policy intervention to stem the numbers.

Figure 7.3: Street homelessness trends (from streetcounts) in selected city centres, 2010-20

Figure 7.1: Overall homelessness numbers; change over time 2011-16

Sources: ABS Census; AIHW Specialist Homelessness Services statistics

108 COVID-19: Rental housing and homelessness policy impacts, 2021

Source: Australian Homelessness Monitor 2020 – statistics assembled from local authority streetcounts (with interpolations for missing years).

109

7.2.2 Policy and practice developments in the immediate pre-pandemic era The upsurge in policymaker concerns on homelessness triggered by COVID-19 in fact followed on from a period which had already seen a marked, if less prominently reported, rise in government attention to the issue across Australia. Albeit entirely confined to state/territory authorities (rather than the Federal Government), this new concern began to emerge in a number of jurisdictions around 2016. This development, seemingly triggered partly by rising street homelessness in capital city CBDs, was marked by a spate of strategies, plans and targets published by several jurisdictions in recent years. Cases in point have included: • Victoria’s Homelessness and Rough Sleeping Action Plan’ (Government of Victoria 2017) • ‘NSW Homelessness Strategy 2018-2023’ (NSW Government 2018) • South Australia’s ‘Our Housing Future 2020-2030’ (Government of South Australia 2019) • ‘All Paths Lead to a Home’ (Government of Western Australia 2019). Perhaps most tangibly of all, the NSW Government committed in 2019 to a state-wide 50% reduction in rough sleeping by 2025. Meanwhile, the South Australian Government pledged to end street homelessness in Adelaide in ‘functional zero’ terms. These commitments and concepts are further discussed in Australian Homelessness Monitor 2020. Although it is often difficult to gauge the exact significance of the dollar sums associated with such plans, they have been generally accompanied by spending pledges of an appreciable scale – e.g. as in strategies published by NSW in 2018 and Western Australia in 2019. Consistent with other indications of expanding activity, state/territory government (real) expenditure on homelessness services has been rising at 7% p.a. in recent years (Productivity Commission 2020). Recently published state/territory homelessness strategies have tended towards a particular focus on reducing rough sleeping, and a largely common set of measures aimed at achieving this objective. Consistent with this orientation, stepped-up activity to tackle street homelessness has been recently seen in all of Australia’s major cities. As explained in more detail by Australian Homelessness Monitor 2020, common strategy components have included commitments on: • Expanded assertive outreach – i.e. where rough sleepers are engaged ‘on site’ with the aim of enabling and supporting a transition from street homelessness to sustainable housing (Phillips and Parsell 2012). • Boosted private rental subsidy programs – time-limited payments to low income households to make private renting affordable • Head-leasing of private rental properties – enabling community housing providers to house former homeless people in dwellings leased from private landlords, e.g. on two-year contracts.

110

COVID-19: Rental housing and homelessness policy impacts, 2021

Enhanced levels of expenditure and activity by state/territory governments have also involved stepped-up interaction and collaboration with NGOs, the main providers of homelessness services. Having now framed the COVID-19 pandemic homelessness story, we move on to consider exactly how that story has unfolded in the initial months of the crisis.

7.3 Initial pandemic homelessness impacts In Australia, as internationally, the deep economic recession brought on by the pandemic triggered a sharp rise in unemployment and loss of earned income for large numbers of people. As shown in Chapter 4, significant housing market impacts quickly emerged, particularly in the private rental sector. The major economic and housing system dislocation that has unfolded in Australia since March 2020 would be expected to exacerbate the nation’s existing homelessness problem. However, pandemic effects on the overall scale of homelessness will have been masked in the short term by the Commonwealth Government’s income protection measures under the JobKeeper and JobSeeker Coronavirus Supplement programs (see Chapter 2), as well as by the rental eviction moratoriums enacted across Australia (see Chapter 6). The initial aggregate impact of the temporary JobKeeper and JobSeeker Coronavirus Supplement programs was to substantially increase pre-pandemic incomes of eligible low-income households (Biddle et al. 2020). For large numbers in this income bracket, rental stress levels – and consequent risk of homelessness – will have temporarily fallen as a result. This effect will, nevertheless, have begun to dissipate after the two programs began to be wound back from September 2020. At the same time, the large cohort of non-permanent residents in Australia – notably international students, skilled migrants and asylum seekers on temporary visas, most New Zealand citizens living in Australia – were excluded from the Commonwealth Government’s COVID-19 income protection programs11. This will have undoubtedly pushed a very substantial number into extreme poverty – and, at least for some, vulnerability to homelessness. The precise scale of this issue nevertheless remains extremely difficult to quantify. The only indicator that facilitates the tracking of homelessness change over time in inter-censal periods is the Australian Institute of Health and Welfare ‘Specialist Homelessness Services’ (SHS) statistical series. This focuses on people being assisted by homelessness services agencies across Australia. The standard headline output from the system is the annual number of people assisted by SHS agencies (e.g. AIHW 2019). During 2020, however, the AIHW has initiated monthly publication of a limited range of SHS statistics that enable some analysis of changes in the scale and nature of homelessness as the pandemic and recession have been unfolding. 11 This refers, in particular, to the estimated 1.1 million non-permanent citizens resident in the country at the start of 2020. Especially since many have relied on low paid employment in hard hit sectors such as hospitality and tourism, these people will have been especially vulnerable to the recession and resulting mass unemployment. Not surprisingly, they are reported to be grossly over-represented among the greatly enlarged numbers of service users logged by foodbanks since the start of the pandemic. Thus, as recently reported to the Senate Coronavirus Committee, food bank users have jumped by 79% to 1.4 million since the start of the public health emergency (Wright & Duke 2020).

111

The headline message from the SHS statistics for Q2 2020 and Q3 2020 is that the COVID-19 public health and economic crisis generated no substantial immediate homelessness impact12. Given the factors explained above this is entirely explicable. More specifically, as shown in Figure 7.4, the six months to Q3 2020 saw a slight reduction in total persons assisted – with numbers falling by an average of 0.31% per month during this period. This contrasts with the year to Q1 2020 when the number of persons assisted remained almost static, and the year to Q1 2019 when the comparable figure had been increasing by an average of 0.37% per month. Bearing in mind widespread concerns that pandemic lockdowns would exacerbate family and domestic violence, it is notable that the period AprilSeptember 2020 saw a slight fall in the number of people being assisted for whom this was a factor implicated in their housing insecurity – see Figure 7.4 Whereas the number of assisted persons in this category had grown by a monthly average of 0.34% in the year to Q1 2029, and by 0.15% in the year to Q1 2020, it marginally declined during mid-2020. By contrast, the number of people being assisted, and who were affected by some form of mental ill health, rose by a monthly average of 0.73% during the pandemic. Importantly, however, this is consistent with an established upward trend in the representation of people experiencing mental ill health among the cohort of SHS assisted service users that saw a 38% increase in their numbers during the four years to 2018-19 (Pawson et al. 2020). Figure 7.4: Average monthly percentage change in number of persons assisted by specialist homelessness services, 2018-2020

At the time of writing it is understood that phasing down of the JobKeeper and JobSeeker Coronavirus Supplement income protection programs will continue in the early part of 2021. Moreover, during and after that phasing down it is anticipated that joblessness will remain at levels well above prepandemic norms and that unemployed persons’ incomes will fall13. With eviction moratoriums also set to expire during the first half of 2021 (see Chapter 5) this will inevitably lead to a resurgence of rental – and mortgage – stress which is highly likely to flow through into increased homelessness at some level. The extent of such an increase is of course highly uncertain since it depends on the changing public health situation during 2021, as well as the timing and vitality of post-pandemic economic recovery. However, on the basis of unemployment rising to 10%, it was recently projected that homelessness in NSW could rise as a result by 21%14.

7.4 Homelessness policy action during the pandemic 7.4.1 Homelessness policy innovation Significant homelessness policy innovation was prompted by the first and second waves of the COVID-19 pandemic in 2020. As summarised in Chapter 2, this primarily involved funded programs to enable: • Rapid placement of rough sleepers and other ‘at risk’ homeless populations into safe temporary accommodation during periods of virus community transmission • Temporarily accommodated former homeless people to transition into longer term tenancies. In the main, the chosen form of ‘safe temporary accommodation’ has been hotel rooms. Especially during the initial national lockdown (March-July), and with the cessation of international tourism, this has been in extremely plentiful supply:

‘You could, through a few phone calls, secure a huge amount of rooms… Over the course of a weekend we pre-booked hundreds of hotel rooms… The policy was to avoid fully booking out any individual place – the target was 40% so you can flex up in case something goes wrong’. [Government housing policy official 2] The one significant exception to the reliance on hotels was in Queensland where the state government took a 9-month lease on a large recently-completed student housing block in inner Brisbane (Boucher 2020). The general preference for hotel bookings (rather than, for example, serviced apartments or student housing) may have reflected a government wish to retain maximum flexibility – maximising the ability to scale up or down accommodation use at short notice. The staff-supervised entry and building oversight that is normal in hotels may also have been seen as desirable. Source: AIHW Specialist Homelessness Services monthly statistics

12 It is worth noting that, since SHS agencies provide what are formally ‘universal services’, resulting service user statistics might be expected to include non-permanent residents affected by severe housing problems – despite their ineligibility for mainstream social security payments and other public services (including social housing tenancies). Whether non-permanent residents at risk of homelessness would be, in fact, likely to seek such assistance is probably more questionable.

112

COVID-19: Rental housing and homelessness policy impacts, 2021

The COVID-19 homelessness emergency accommodation programs seen in Queensland and other states were unprecedented in scale, but not in principle. Especially under the stepped-up attention to street homelessness seen in several jurisdictions in recent years (see Section 7.2.2), assertive outreach 13 In its November 2020 Statement on Monetary Policy, the Reserve Bank of Australia projected that unemployment would peak at nearly 8% and would remain at or above 6.5% until at least June 2022 (RBA 2020). For reference, the unemployment rate in December 2019 was 5.2%. 14 This estimate, by Equity Economics (2020), applied research findings by Guy Johnson and colleagues (Johnson et al. 2019) based on the Journeys Home dataset.

113

and temporary housing placements en route to longer-term tenancies had become somewhat more common prior to 2020. As implemented at scale in the pandemic context, however, temporary accommodation placements involved significant policy relaxations, particularly in terms of placement eligibility and duration restrictions. A critical adjustment to standard practice on eligibility for temporary housing concerned citizenship status. Significantly, standard rules disallowing nonAustralian citizens from such assistance were dropped or disregarded in the COVID-19 temporary rehousing programs:

‘In terms of our rough sleeper service, our main focus at the start of [the pandemic] was to get people off the street into EA [emergency accommodation] very rapidly. And then go back and retro-fit assessments – income checks and all of that – once people were housed’ [Homelessness services provider 3] Thus, in common with the protection afforded to tenants under evictions moratoriums, non-citizens were able to benefit from the offer of emergency safe accommodation. This contrasts with the ineligibility of temporary visa holders, international students and others from Commonwealth Government income protection programs (see Chapter 1). Nevertheless, their non-qualifying status for social security payments and social housing tenancies meant that non-citizens were highly unlikely to benefit from assisted transition to a longterm tenancy that has been the outcome for some of the EA cohort. Concerns about the especially vulnerable position of non-citizens in the pandemic and recession however cut no ice with the Federal Government:

Ministers have been telling us that their message to people on temporary visas is: ‘your options are to go home [and/or] to access your superannuation’’. [Ministers are continuing to] ‘take that hard position that ‘you said you could look after yourself when you came, and that is our expectation’’ [NGO/housing/homelessness peak body 3 ] State government staff (and NGOs acting on their behalf) were also allowed to depart from ‘business as usual’ policy when it came to the duration and conditionality of hotel bookings:

‘To get five nights of temporary accommodation you normally need to show you’re making an effort to get other accommodation [whereas in pandemic conditions] rough sleepers were being booked into hotels for 28 day blocks – with potential for repeat bookings’ [ Government housing policy official 2] Normally [as a rough sleeper] you might be offered two nights of temporary accommodation and then you have to come in and do a full assessment and then you may be offered a bit more. At one stage at the height of the pandemic we would even give 28 days of temporary accommodation’ [Government housing policy official 3] In one state, however, pandemic-era practice for homeless people other than former rough sleepers involved an initial hotel booking restricted to only five nights, with a case by case assessment for subsequent extension. As perceived by a peak body interviewee, this had ‘caused a huge level of anxiety’ and – as a result – significant levels of hotel abandonment. The policy was viewed 114

COVID-19: Rental housing and homelessness policy impacts, 2021

negatively by contrast with the COVID-19 program in another state where people booked into hotels were reportedly assured of a placement for up to three months with a subsequent ‘[long term] housing outcome’. 7.4.2 The policymaking process

A state government focus When we consider the policymaking process that underlay the pandemic-triggered initiatives for rough sleepers and other homeless people described above, the focus rests almost exclusively on state and territory administrations rather than the Commonwealth Government. As explained in Chapter 1, it is states and territory authorities that are directly responsible for housing and homelessness services in Australia. Although they receive some Commonwealth financial assistance to underwrite homelessness management costs, the vast bulk of such expenditures are funded from state/territory governments’ own resources15. As demonstrated in the stance adopted by the 2007-2010 Rudd Government (Milligan and Pinnegar 2010), the Commonwealth may choose to become more active in driving efforts to prevent or relieve homelessness. However, the 2020 public health crisis evoked no such response. It has been suggested that the EA initiatives implemented in a number of states from March 2020 were in some way nationally co-ordinated through the National Cabinet (Mason et al. 2020 p27), but this claim is unsupported and we have detected no evidence for it.

Policy authorisation and the role of NGOs The key decisions to initiate EA programs were taken in late March 2020 by state governments in NSW, Queensland, South Australia, Victoria. Crucially, these decisions were part of a broader set of responses to public health concerns as the pandemic quickly took hold. As explained by one interviewee, the formal trigger for such responses was the state government declaration of a Public Health Order. It is relevant at this point to acknowledge that differing levels of concern in different jurisdictions meant that the arrival of COVID-19 in Australia did not spark large scale EA programs in all states and territories. Most significantly, the West Australian Government drew back from such action after the closure of its domestic and international borders had been judged successful in insulating the WA population from the pandemic. Much to the disappointment of homelessness advocates, the Government’s pilot program was therefore terminated rather than expanded. Crucially underpinning the EA policy announcements in Australia’s other mainland states were commitments to new, previously unbudgeted, expenditures. As this process functioned in one state, programs calling for funding additional to a Department’s allocated budget, required Expenditure Review Committee (ERC) endorsement. Especially in the case of a minister responsible for homelessness but lacking cabinet rank securing such funds was considered by officials as a potentially challenging objective:

‘Our minister must have fought very hard at that [ERC] table’ [State government official 3] 15 Thus, while the National Housing and Homelessness Agreement mandated $125million in Commonwealth Government funding for homelessness services in 2019-20 (Australian Government 2019), state and territory government expenditure on such services was $990 million in 2018-19 (Productivity Commission 2020).

115

However, as revealed by our interviews, the EA funding announcements cannot be, by any means, represented as an outcome of a simple top-down process. At least from the perspective of NGO homelessness service providers in more than one state, this particular policymaking episode involved governments being pressured into commitments as much as being initiators of action. An element of this involved urgent lobbying activity by peak bodies and other homelessness advocates, including formal submissions advocating extraordinary action. In this instance, however, pressure exerted by NGOs went much further. At least in two of the four states, from the perspective of senior non-government interviewees, it was very substantially ‘a push by the sector’ – pre-emptive action – that had evoked Government decisions to authorise the ‘vastly increased funding’ that enabled EA programs to proceed:

We had already put people into EA and run out of money, and said [to government] “so what are you going to do? Do you want us to send them all back to the streets?” So there was a push-pull in that policy creation. Luckily we work for an organisation that was prepared to go beyond our funded means… and then to seek reimbursement from government. [This was] to my CEO’s absolute credit. I wouldn’t have had the gumption for that brinkmanship…’ [Homelessness services provider 3] I don’t think [the EA policymaking impetus] came from the [social housing] minister… I think it was driven from [the homelessness sector]. And from Health – I know their preference was that people [at risk] were in some form of accommodation [Community housing provider]. To some extent these perspectives probably reflect the frenetic atmosphere prevailing as the pandemic first wave rapidly took hold – a context in which orderly policymaking was inherently a tall order:

‘These are crazy times and of course this wasn’t beautiful planning’ [NGO Homelessness services provider 3]. A more specific cited instance of reactive rather than pro-active government policymaking concerned the reportedly belated official acceptance of a need to de-concentrate homelessness shelter accommodation with shared facilities. As viewed from the NGO service provider perspective in one state, this came about only as a result of strong lobbying by accommodation providers. Evidently, for responsible government officials the process of securing EA policy authorisation will have necessitated overcoming significant bureaucratic hurdles and internal political contestation. NGO frustration at what they may have viewed as official hesitation and indecision should perhaps be interpreted in this light. Equally, this is a scenario arguably exacerbated by limitations on homelessness policy responsiveness that result from the erosion of housing expertise and identity within government. In most Australian governments recent decades have seen a loss of housing identity as associated policy and operational functions have been increasingly subsumed within broader human services (and/or other) departments (Pawson et al. 2020). One important consequence is that senior public servants responsible for housing functions are nowadays typically tasked with multiple other responsibilities (e.g. for other aspects of human services). This limits the scope for them to identify with or advocate for housing.

116

COVID-19: Rental housing and homelessness policy impacts, 2021

Collaborative policymaking in the implementation phase Tensions that had quickly emerged in the lead-up to EA announcements seem to have largely dissipated immediately thereafter, giving way to a period of extraordinarily intense and – from the accounts of many involved – often highly productive collaboration. Emergency consultative structures were quickly set up to facilitate interaction between government officials, NGOs and local councils. Collaborative arrangements of this kind were considered by many (although not all) to have been an unusual exemplar of effective partnership working: There was a bit of scramble at the start in some jurisdictions. Then we went into quite a cooperative relationship and there were working groups set up, for example in inner city Sydney … peak groups and other service providers have worked quite collaboratively together with the government. A little bit too late, but you know it's finally happened and that's been good’ [Homelessness services provider 1]. From a public servant perspective, these interactions were also seen by some as unprecedented – not only in interaction intensity, but also in terms of the depth of issues covered:

I don’t recall ever being in forums where we’ve shared so much with the non-government sector [Government housing policy official 2]. As perceived by some, the need for extraordinary pandemic-response action saw many organisations contributing goodwill in large measure, and also enhanced a sense of fraternity within the homelessness services sector itself:

The pandemic brought out the best in us; people did set aside their conflicting interests and rivalries to make it work…[As an example], although CHPs are not funded for outreach, [some of them] have done outreach… CHPs have gone above and beyond the call of duty to do certain things that are not their core business [Government housing policy official 3] One of the delightful outcomes is there are key players in the homelessness services system that work together in ways we probably didn’t before. Some of that competitive nature of our work has had to dissipate – ‘who’s doing mission better’ has had to take a side seat… [Homelessness services provider 3]. Inter-governmental collaboration As noted above, our evidence informs scepticism as to claims that any of the extraordinary homelessness policymaking seen during the pandemic has been in any significant way instigated, shaped or co-ordinated by the Commonwealth Government. For some interviewees, the Commonwealth’s reluctance to become involved in – or contribute to – emergency action on homelessness was striking. Closely mirroring a statement by a senior state government official cited in Section 6.2, another interviewee commented on pandemic policymaking as follows: The Commonwealth are totally absent. Not just politicians, but unfortunately you know the officials as well [NGO/housing/homelessness peak 1] At the same time, it might have been expected that intensity of local pressures exerted by COVID-19 could have accentuated the traditionally rather autonomous operating style of state/territory governments, leading to a 117

problematically fragmented response. At least when it comes to homelessness and housing matters, however, the pandemic in fact appears to have beneficially stimulated inter-jurisdictional collaboration. Such contact was reportedly stepped up through the pre-existing Housing and Homelessness Senior Officers Network. Convened by Victoria, and involving all jurisdictions, HHSON began meeting weekly rather than the previous business as usual frequency which was ‘two or three times a year’. This provided a forum to discuss emerging pandemic challenges including each jurisdiction’s experiences of interacting with hotel quarantine service personnel, homelessness services and community housing providers:

‘When Victoria had its second wave Victoria shared their public housing tower response with everybody… I shared our TA policy changes with everyone… We also invited our colleagues from other departments into that forum to discuss [policy developments such as] our evictions moratorium’ [Government housing policy official 2]. Such interaction also calls into play another motivating factor for policy innovation; the inspirational knowledge that another state government has authorised an exceptional program or measure:

‘We try to shame each other into going further… If another jurisdiction is doing something really clever, it gives us imprimatur to [push for something similar]’ [Government housing policy official 3]. Nevertheless, according to one stakeholder interviewee, the Commonwealth Government was not only reluctant to become involved in policy co-ordination activity, but actively sought to discourage inter-state/territory government interaction to this end. Citing a Commonwealth guidance document issued in late 202016 a senior state government official reported that:

‘[Commonwealth]/ National Cabinet have even killed off housing official meetings’. Policy critique Before concluding this section on pandemic policymaking we briefly recount some of the evaluative comments advanced by interviewees. As already noted above, one frustration shared by many NGOs was the sense that – from their perspective – state governments appeared initially both unprepared and slow to act. Some non-government interviewees also questioned the value-formoney implications of the near-universal official preference for hotel rooms at the start of the pandemic. Instead, as some argued, a more economical approach, also more potentially compliant with housing first principles, would have seen private rental property head-leasing programs initiated at scale right at the start of the pandemic. Beyond this, critical comment tended to focus on the limitations of efforts to transition hotel-housed service users into longer term housing. Interviewees believed that relatively rapid erosion of the hotel-housed cohort had, in part, reflected insufficient support and excessive delays in identifying mainstream affordable move-on tenancy opportunities: 16 The guidance advocates that ’officials groups should be kept to a minimum, to reduce the number of layers of bureaucracy and streamline the processes surrounding how meetings function … Secretariats: Meetings should not have dedicated secretariats. Secretariat functions … should be undertaken by the relevant Commonwealth Department’ (Dept of Prime Minister and Cabinet 2020).

118

COVID-19: Rental housing and homelessness policy impacts, 2021

You lost a lot of people, who couldn’t sustain themselves in TA after a while, because it’s not a long-term option and people can’t sustain it, and there’s the support service issues’ [Community housing provider]. As widely noted, a major limiting factor in moving people out of hotels was the inadequacy of social housing provision, one of the underlying issues believed responsible for the scale of homelessness preceding the pandemic:

The failures of the housing system are laid bare in a crisis. And we haven’t created a systemic response to it now, in the middle of it – it’s still focused on TA and getting them out and not the chronic systemic issues … the pandemic really highlighted the problems in the [name of state] housing system and brought them to the fore. And we’re playing at the edges’ [Community housing provider]. In a similar vein, while acknowledging the potential utility of temporarily expanding the social housing stock through head-leasing as move-on accommodation, it was reflected that:

‘The pandemic created the policy space for additional funding… as you know it’s a big struggle for anyone involved in the social housing system to get any money out of any government…But so far, there’s been no announcement of any17 funding … None of the extraordinary investment so far has done anything to add to the capacity of the system – expansion of permanent housing affordable to low-income households’ [Government Housing policy official 3] 7.4.3 Emergency accommodation program activity and impacts

EA placements Although numbers have been cited in the media, there are no official statistics that capture the extent of state government emergency accommodation activity for homeless people prompted COVID-19. In this section we draw on unpublished statistics provided to the research team by the four state governments concerned: NSW, Queensland, South Australia and Victoria18,19 We also refer to more detailed explanatory information kindly provided by the Queensland Government. Two of the other three state governments helpfully took up our offered opportunity to comment on an initial draft of the following analysis. The new data collected here20 confirm that the programs implemented in these states were, indeed, remarkable in scale. From the point at which the pandemic struck Australia in mid-March to the end of September, emergency accommodation was provided to over 40,000 people – see Table 7.1. In the 17 Note that this statement preceded the extraordinary $5.3 billion social housing stimulus program announced by the Victorian Government in November 2020 and the more modest $400 million committed to additional (or accelerated) social housing construction by the NSW Government in its state budget in the same month. 18 It is understood that any such activity in other jurisdictions was on a small scale only. 19 It must be emphasized that the figures included in this section have been kindly provided by state government colleagues, drawing on record keeping systems that are not necessarily ideal for the purpose, and applying statistical definitions that are unfamiliar. Generating the figures has therefore necessarily required the exercise of judgement and, partly with this in mind, it is acknowledged that the numbers provided by each government may not be wholly comparable. Thus, the figures cited in this section must be treated as indicative rather than definitive. 20 NSW did not submit a return in this survey. However, some statistics were provided in the course of an interview with a NSW Government official. With respect to EA departure (move-on housing placements) statistics, NSW data (relating only to the period to 14 June 2020) is drawn from the State Government’s response to a similar survey undertaken for Australian Homelessness Monitor 2020 research (Pawson et al. 2020).

119

three states for which a finer breakdown is available, almost half of these (48%) were classed as former rough sleepers. The remainder will have included people living in homelessness shelter (and similar) accommodation considered unsafe in pandemic conditions – e.g. due to the need to share facilities. Notably, the number of rough sleepers assisted in Queensland, South Australia and Victoria (7,718) was well over double the point-in-time number of rough sleepers in these entire states as recorded by the 2016 Census (3,246) – see ‘State and territory of usual residence, all persons’ tables (ABS 2018). This further emphasizes that (as argued in Section 7.2.1) citation of the Census estimate for this group (8,200 nationally) substantially understates the scale of the issue in terms of those experiencing (or at risk of) rough sleeping over any significant time period. This, in turn, only highlights the reality that there is a substantial body of people who cycle in and out of actual rough sleeping, far greater in number than the snapshot total on any given night. Table 7.1: Emergency accommodation placements – flow 15 March-30 September (persons) Former rough sleepers

NSW

Other homeless

24,000

Total

24,000

Qld

3,276

1,648

4,924

513

Vic

3,929

6,882

10,811

Total

40,248

Source: authors’ survey; interview with NSW Government official. Notes: 1. NSW TA placements statistic relates to the period 1 April-8 October; 2. South Australia 'placed in EA' includes nine people allocated to hotels but who never checked in

Nationally, the peak number of emergency accommodation placements at a point in time probably occurred in April 2020. By 30 June the number will have already fallen back. Other than in Victoria, as shown in Table 7.2, point in time placements further declined or (in the case of South Australia) entirely ended over the following three months. The key point to make here is that by 30 September, only a very small proportion of the 40,000 people provided with emergency accommodation over the previous six months remained in such premises – a fact no less true in Victoria.

Table 7.2: Temporary accommodation placements at points in time (persons) 30 Jun

30 Sep

Change – 30 Jun-30 Sep

NSW

2,000

900

-1,100

Qld

3,155

1,114

-2,041

154

-154

Vic

1,072

1,962

890

Total

6,381

3,976

-2,405

Source: authors’ survey; interview with NSW Government official. Notes: NSW point in time EA placement figure for 'June 2020' represents April 2020; NSW point in time EA placement figure for '30 Sept' represents 8 October

Rehousing out of EA All four states have made substantial efforts and commitments to assist EA service users into longer-term tenancies, as far as possible – whether in social or private rental housing. For example, as part of a $150 million21 program announced on 28 July 2020, the Victorian Government pledged expansion of headleased private rental stock by 1,100 properties so that ‘2,000 [hotelhoused] Victorians are supported to access stable, long term housing’ (Victorian Government 2020). While the duration of the funding (i.e. the length of leases to be procured) was unspecified in the Victorian announcement, we can note that under the NSW Government’s ‘Together Home’ headleasing program to facilitate rough sleeper rehousing, private rental dwellings were to be acquired for two year terms. Across the four states, collectively, nearly 8,000 former rough sleepers had departed emergency accommodation by 30 September – see Table 7.3. Just under a third (32%) had been assisted into longer term tenancies. In Queensland, this was by far the most common outcome, whereas it was much less true in NSW and Victoria. For NSW it should be borne in mind that these figures reflect activity only up until 14 June (see footnote). Moreover, a further 108 former rough sleepers had an active application on the NSW social housing register at that time.

21 Also including provision for associated tenant support, as well as second wave temporary accommodation charges.

120 COVID-19: Rental housing and homelessness policy impacts, 2021

121

Table 7.3: Departures from emergency accommodation – former rough sleepers – 15 Mar-30 Sep (persons) Rehoused in social housing

NSW

Assisted into private tenancy

Placed in congregate accom

166

Self discharge/ other

1,392

Total departing EA in period

1,558

Qld

1,167

1,000

no data

126

2,293

186

271

513

Vic

Total

2,545

3,466

3,490

5,309

7,854

Source: authors’ survey; NSW data previously provided to the authors and already published in Australian Homelessness Monitor 2020. Notes: 1. ‘Rehoused in social housing’ includes tenancies in headleased properties; 2. ‘Assisted into private tenancy’ involves those assisted by bond loans, rental subsidies etc.; 3. ‘Placed in congregate accom’ involves persons placed in homeless shelters, hostels or similar; 4. Light type figures for ‘total departing EA in period’ indicate our estimate, calculated by subtracting those remaining in EA on 30 Sept from those placed in EA during the period. 5. Light type figures for ‘self-discharge/other’ indicate a residual number calculated by subtracting 'rehoused in social housing' plus 'assisted into private tenancy' from total departing EA during the period; 6. NSW departures from EA statistics relate to the period 1 April-14 June. However, on 17 Jan 2021 it was reported that, NSW had rehoused 794 former rough sleepers from EA into ‘stable housing’ during 2020. This represented ‘more than the [previous] three years combined [750] (https://twitter.com/9NewsSyd/status/1350345269905338369)

of rough sleeping. Notably, by late 2020, rough sleeper numbers in Adelaide and Sydney had risen markedly from their mid-year lows which – at that time – reflected initial EA mass placement programs. According to monthly monitoring statistics published by the Don Dunstan Foundation, rough sleeping in central Adelaide fell from 150 in March 2020 to 83 in June. By September, however, this had once more risen to 129. In Sydney, a base pre-pandemic number is provided by the City of Sydney’s latest street count which recorded 334 rough sleepers in February 2020. According to routine monitoring by the peak body Homelessness NSW, by April the hotel rehousing program had reduced this to 90. By August, however, Homelessness NSW estimated that the total had drifted back up to 170. Table 7.4: Departures from emergency accommodation – all homeless people – 15 Mar-30 Sep (persons) Rehoused in social housing

NSW

Focusing on all homeless people provided with emergency accommodation and departing such premises in the period to 30 September (see Table 7.4) it appears that around a fifth had been successfully assisted into longer-term tenancies. A proportion of others will have been returned to, or placed in, congregate accommodation (homeless shelters, hostels and the like). Some of these people are likely to be assisted into longer-term tenancies at a later date – e.g. reflecting their progression on a social housing register. A proportion of the estimated 18,573 people who departed EA without being assisted directly into longer term tenancies - ‘other EA departees’ (see Table 7.4) - are likely to have been people moving into (or back to) specialist homelessness service accommodation or other transitional housing . As regards South Australia, the state government advises that some may have been Indigenous people returning to remote communities. At the same time, for a significant proportion of other EA departees, this will likely have involved a return to homelessness – in some instances a resumption 122 COVID-19: Rental housing and homelessness policy impacts, 2021

Placed in congregate accom

991

Self discharge/ other

9,187

Total departing EA in period

10,178

Qld

1,774

1,632

no data

404

3,810

186

271

513

Vic

101

Total

In Victoria, with the COVID-19 second wave still ongoing at the end of September, it may be that active efforts to rehouse former rough sleepers out of hotels and into longer term tenancies had yet to get into full swing. The state’s new headleasing program had been only recently announced, and will have been yet to generate lettable vacancies. It should also be emphasized that the number of people enumerated under ‘placed in congregate accom’ and ‘self-discharge/other’ for Victoria is a residual and could therefore include some people rehoused in social or private rental housing but uncounted as such due to missing data.

Assisted into private tenancy

4,777

8,657

8,849

18,573

23,350

Sources and notes as Table 7.3

Perhaps the most striking figures reported in Tables 7.3 and 7.4 relate to Queensland. They are highly notable, firstly, because the number of people departing EA via ‘self-discharge’ was very low. At least in part, this could possibly reflect the fact that an element of Queensland’s EA capacity was student housing being utilised on leases running for months rather than (as in other states) wholly involving hotel accommodation booked on a nightly or similar basis. This enabled some of those in Queensland to be accommodated in EA on a less insecure footing. It is also possible that more intensive support provision could have enabled formerly homeless residents to sustain their tenure more effectively. Indeed, the Queensland Government believes that the EA retention rate ‘is likely the result of having provided highly intensive support to temporary accommodation placements’. NGO service providers were commissioned to provide ‘round the clock support’ to EA residents. The statistics for Queensland in Tables 7.3 and 7.4 are notable, secondly, in that the number assisted into longer term tenancies in social and private housing was large in both proportionate and absolute terms. For a sense of scale here, we can refer to official statistics showing that the 2018-19 combined total number of lettings to new tenant households in public housing,community housing and Indigenous housing in Queensland was 6,922 (4,368 + 2,315+239) (Productivity Commission 2020). Bearing in mind that the period covered in our analysis is only 6 months, we might expect the appropriate comparator figure 123

(all mainstream social housing lettings in Queensland) to be around 3,461. The Queensland Government advises that the 1,774 people rehoused out of EA into social housing over the period (see Table 7.4) involved 1,016 households. This therefore equated to 29% of ‘steady state’ total social lettings. When bearing in mind that most of the households being transitioned out of EA will have been single people qualifying for 1-bedroom units (a property size cohort likely to account for only a minority of all lettable social housing vacancies) the logged rehousing numbers appear all the more remarkable. As explained by the Queensland Government, a number of strategies contributed to this achievement. These included: • Highly intensive support provided to EA residents in transitioning into longer-term sustainable housing • Rapid expansion of (quasi) social housing stock through headleasing acquisitions • Prioritisation of access to newly built social housing coming onstream during 2020. Queensland’s performance notwithstanding, the broader point here is that – collectively – state governments found it easier to make mass EA placements than to support those concerned for long enough to assist them into secure tenancies. Once the governments had decided to mandate extra spending for the purpose, the pandemic-triggered superfluity of hotel and similar accommodation made the first of these relatively simple to achieve. Helping people into longer term housing posed a far greater challenge – largely because of the insufficient supply of social housing and the inadequacy of Rent Assistance in making private rental housing affordable. Beyond these, the significant proportion of non-Australian citizens within the EA cohort will have constituted a third limiting factor – given this group’s ineligibility for both social housing and maintstream social security payments22.

7.5 Chapter conclusion Despite the extensive economic disruption triggered by the COVID-19 public health emergency, there was no immediate resulting surge in homelessness as reflected by service demand. At least three factors are likely to have been in play. Firstly, the Commonwealth Government’s emergency income protection programs which particularly benefited lower income households (Biddle et al. 2020b). Secondly, the rental eviction moratoriums implemented across the country to protect tenants at risk of falling into arrears due to job loss. Finally, the ability of housing cost-burdened tenants to adjust to their situation by doubling up with other renters. Soon after the onset of the pandemic four of Australia’s largest states launched emergency programs to provide safe temporary accommodation for existing rough sleepers and homeless people in shelter premises with shared facilities. While involving action at unprecedented scale, this built on rough sleeper engagement and rehousing efforts that had been already somewhat ramped up in a number of cities in the immediate pre-2020 period. Pre-existing practice here was not only hugely expanded, but also implemented less restrictively so 22 The Queensland Government advises that it was, nevertheless, able to assist transition to longer term accommodation by brokering private housing outcomes for this cohort.

124 COVID-19: Rental housing and homelessness policy impacts, 2021

that it encompassed non-Australian citizens, and so that service users were booked into hotels for longer periods and with fewer conditions. Emergency accommodation programs were authorised and funded by state governments in NSW, Queensland, South Australia and Victoria. From the perspective of some NGO service providers, much of the impetus here came from the service providers themselves. Moreover, the eroded condition of some state governments when it comes to housing policy capacity, domain knowledge and policy-area-specific senior representation may have impeded rapid and decisive action. The relatively rapid attrition of the hotel-housed population and the only modest numbers successfully assisted into longer-term housing exposed some weaknesses of program management – such as inadequacy of support provision – and similarly the more fundamental systemic flaws of the housing system – in terms of the limitations posed by grossly inadequate availability of social housing. From the start of the pandemic in March 2020 a parallel program to provide emergency accommodation for rough sleepers and others was rolled out in England. As in Australia, this action was prioritised as means of reducing COVID-19 health risks to homeless people themselves, but also to protect population health across the broader community. At least initially, in both countries normally applied assistance eligibility restrictions concerning citizenship were waived. Also in common was the fortuitous opportunity for government and NGO players to platform off several preceding years of stepped-up efforts to tackle street homelessness in the late 2010s. At the same time, participating stakeholders in both countries highlighted scope for more effective program management and use of funds. However, England’s program has been significantly different from Australia’s in certain key respects. Importantly, unlike in Australia, it was directed, coordinated and to a large extent funded by national government. But perhaps the most striking difference is the extent to which people provided with emergency temporary accommodation have been subsequently assisted into transitional or long term housing. While Australian state governments have made substantial efforts to facilitate such moves, those benefiting from such assistance have formed only a small minority of those departing emergency hotel placements. In England, by contrast this appears to be the majority outcome. The main explanation is likely to be relative scale of social housing provision in the two countries – the drastically greater challenge posed to governments and NGOs in Australia, where public and community housing accounts for little more than 4% of all stock – only a quarter of the equivalent English figure.

References Australian Government (2019) National Housing and Homelessness Agreement; Department of Social Services; Canberra: Australian Government https:// www.dss.gov.au/housing-support-programs-services-homelessness/nationalhousing-and-homelessness-agreement Australian Institute of Health and Welfare (2019) Specialist Homelessness Services annual report 2018–19. Cat. no. HOU 318. Canberra: AIHW https://www. aihw.gov.au/reports/homelessness-services/shs-annual-report-18-19 125

Biddle, N., Edwards, B., Gray, M. & Sollis, K. (2020a) Tracking outcomes during the COVID-19 pandemic (May 2020) – Job and income losses halted and confidence rising; ANU Centre for Social Research and Methods; https://csrm. cass.anu.edu.au/research/publications/tracking-outcomes-during-covid-19pandemic-may-2020-job-and-income-losses Biddle, N., Edwards, B., Gray, M., & Sollis, K. (2020) COVID-19 and mortgage and rental payments: May 2020 ANU Centre for Social Research and Methods https://csrm.cass.anu.edu.au/sites/default/files/docs/2020/6/COVID-19_and_ housing_FINAL.pdf Boucher, D. (2020) Student Apartment Block Turns into Crisis Housing; The Urban Developer https://theurbandeveloper.com/articles/student-apartmentsturn-crisis-housing Dept of Prime Minister and Cabinet (2020) Guidance for Intergovernmental meetings; https://tinyurl.com/y6bmlto7 Equity Economics (2020). Supporting Economic Recovery in NSW; Sydney: NCOSS https://www.ncoss.org.au/sites/default/files/public/policy/Equity%20 Economics%20-%20Supporting%20Economic%20Recovery%20in%20NSW_ Final_220620.pdf Mason, C., Moran, M. and Earles, A. (2020) Policy coordination and housing outcomes during COVID-19, Final Report no. 343; Melbourne: AHURI Milligan, V. and Pinnegar, S. (2010) The comeback of national housing policy in Australia: first reflections. International Journal of Housing Policy, 10 (3), pp.325-344 Pawson, H., Milligan, V. & Yates, J. (2020) Housing Policy in Australia: A case for system reform; Singapore: Palgrave Macmillan Pawson, H., Parsell, C., Liu, E., Hartley, C. and Thompson, S. (2020) Australian Homelessness Monitor 2020; Melbourne: Launch Housing https://cityfutures. be.unsw.edu.au/documents/619/Australian_Homelessness_Monitor_2020.pdf Pawson, H., Parsell, C., Liu, E., Hartley, C. and Thompson, S. (2020) Australian Homelessness Monitor 2020; Melbourne: Launch Housing https://cityfutures. be.unsw.edu.au/documents/619/Australian_Homelessness_Monitor_2020.pdf Productivity Commission (2020) Report on Government Services 2020: Part G – Housing and Homelessness; Tables 18A5, 18A6 and 18A7; https://www.pc.gov. au/research/ongoing/report-on-government-services/2020/housing-andhomelessness/homelessness-services Reserve Bank of Australia (2020) Statement on Monetary Policy, November; Sydney: RBA https://www.rba.gov.au/publications/smp/2020/nov/ Victorian Government (2020a) Homes For Homeless Victorians During Pandemic And Beyond; Media Release 28 July https://www.premier.vic.gov.au/ homes-homeless-victorians-during-pandemic-and-beyond Wright, S. & Duke, J. (2020) Record 1.4 million people relying on food charity as recession bites; Melbourne Age, 1 July https://www.theage.com.au/politics/ federal/record-1-4-million-people-relying-on-food-charity-as-recession-bites20200701-p55821.html

126 COVID-19: Rental housing and homelessness policy impacts, 2021

8. Housing and homelessness policy change in the United Kingdom: an international comparator Key points: • In the UK, the central government’s primary response to the risk of evictions was, like the Australian Government’s, income support. Legal measures to prevent evictions were patchy, though devolved administrations such as Scotland acted more decisively • The UK Government was more involved in coordinating and funding efforts to accommodate homeless persons than national government in Australia • Although directly comparable statistics are unavailable, it appears that more of those placed in EA in the UK have been transitioned into longer-term housing than in Australia. This chapter provides an international point of comparison for the policy change and implementation examined in the previous two chapters, by reviewing the main features of analogous policy change in the United Kingdom. As in the Australian chapters, the review encompasses both the substance of the UK’s emergency measures, and the processes by which they were made.

8.1 Rental housing policy 8.1.1 Emergency policy measures The short-term policy response to the pandemic in the UK was somewhat of a patchwork. Emergency measures directed at assistance with housing costs initially emanated from HM Treasury because social security is a reserved power that sits with the UK government in Westminster. In contrast, housing policy is a devolved responsibility – that is, under the direct control of the Scottish Parliament and equivalent bodies in Wales and Northern Ireland. In recent years the UK private rented sector has become a key site through which the implications of devolution are being explored. The devolved administrations have, for example, introduced national systems of landlord registration and/or licensing, a move the Westminster Government has so far rejected as regards possible application in England. More relevant to the current discussion, the tenancy regulation framework in Scotland was changed significantly by the Private Housing (Tenancies) (Scotland) Act 2016. Once Wales acquired the power to shape its own housing policy it also planned to diverge from the framework of tenancy regulation it has shared with England for a quarter of a century, although those changes are yet to be enacted. Although differing in detail between the countries, much recent policy change in the UK – including in England – has tended towards more active policy and strengthened regulatory frameworks. England sits as something of an outlier in the European context in having such a lightly regulated private rented sector.

127

The focus of this broader policy change has been on tenant rights and security of tenure. Scotland legislated in 2016 to extend tenancy lengths and require that eviction should be on the basis of a limited series of specific grounds. In contrast, the default tenancy length in the English private rented sector is six months and so-called ‘no fault’ (or ‘no grounds’) evictions are common. The passage of associated legislative reform was disrupted by the arrival of the pandemic. This was the context in which the crisis broke. The first response to the pandemic came from the Treasury. As part of a broader package of measures, the Chancellor announced on 20 March that nearly a billion pounds would be used to “increase the generosity of housing benefit and Universal Credit, so that the Local Housing Allowance will cover at least 30% of market rents in your area” (Gov.uk, 2020). This move reversed some of the erosion of the value of these benefits produced by a decade of restraints on annual uprating, rationalised using an austerity discourse. Concurrently, the Financial Conduct Authority published guidance to lenders advising them to operate payment holidays of up to three months for mortgagors – including private landlords. This would cushion the blow from any disruption to revenue resulting from tenant income losses. These financial measures were shortly followed by a wide-ranging UK Coronavirus Act 2020 which included extending the notice period for tenancy termination for social and private tenants: for most private tenants in England and Wales this meant an increase in notice from two to three months. Shortly after these initial moves the Scottish Government passed the Coronavirus (Scotland) Act which provided tenants with enhanced security of tenure on a temporary basis, with effect from 7 April. The Act temporarily redefined certain mandatory grounds for eviction as discretionary for a fixed period of time and extended certain notice periods up to September 2020. The initial suite of emergency measures was completed by announcements in relation to evictions. In Scotland the Housing and Property Chamber First-Tier Tribunal announced that from 19th March all hearings and case management discussions would be postponed until 28th May 2020. No new eviction orders would be granted until that date. A week later, the Master of the Rolls similarly announced that all ongoing housing proceedings in England and Wales should be suspended for ninety days. As in Australia, underlying the policy response was the view that where possible the threat of eviction for COVID-induced arrears should be resolved through negotiation between landlord and tenant and through landlord forbearance. This is evident in the statement from Scottish Constitution Secretary Michael Russell, introducing the Coronavirus (Scotland) Bill:

While all tenants experiencing issues with rent arrears should firstly explain their circumstances to their landlords, this new emergency legislation will provide an important backstop to prevent evictions and relieve the financial pressure people may be facing. We are also encouraging all landlords to be as flexible as possible during this unprecedented time and would urge them to also seek assistance, if necessary by speaking to their lenders about mortgage breaks. (gov.scot, 2020) In England early Government thinking about the crisis response placed considerable emphasis upon extending the Pre-Action Protocol (PAP) from 128 COVID-19: Rental housing and homelessness policy impacts, 2021

the social rented sector to the private rented sector. The thinking was that this move would oblige private landlords to engage with their tenants to understand their financial position and wherever possible agree a rent repayment plan and avoid eviction. If a landlord failed to follow this procedure correctly then the court could adjourn the claim or strike it out completely. This approach, if the Government considered it the full extent of its plans to protect renters, attracted considerable criticism. It was felt to be an inadequate response to the emerging crisis not only because it rests almost entirely on forbearance on the part of the private landlord but also because courts would not be well-placed to use the PAP to prevent private rental evictions in a context where rent arrears are a mandatory ground for possession. Several of the emergency measures put in place at the start of the crisis were subsequently extended incrementally. In June the suspension of eviction proceedings in England was extended for further two months; in August it was, at the last minute, extended again for a further month. Court proceedings resumed in September but with new requirements placed upon claimants until March 2021. In Scotland and Northern Ireland relevant restrictions were extended until March 2021. Meanwhile both Wales (July) and England (August) temporarily increased end of tenancy notice periods from three to six months. Given that social security is a reserved power, the Scottish and Welsh Governments innovated to create mechanisms to provide other forms of financial assistance. In April the Scottish Government implemented an interestfree loan system to help eligible landlords, while in September it announced a £10 million Tenant Hardship Loan Fund. In October the Welsh Government launched their equivalent initiative - Tenancy Saver Loans. In the November 2020 Spending Review the Chancellor indicated that the increases to Local Housing Allowance will not extend past March 2021. There is less clarity on whether the Universal Credit uplift will be sustained. Current plans suggest that the incidence of tenants experiencing affordability problems could increase sharply from Q2 2021. 8.1.2 Policymaking process While we have limited first-hand information on policymaking inside government in the UK administrations, discussions with stakeholders within the policy community indicate some key commonalities and areas of divergence. Five brief points are worth noting here (see Foye and Marsh, 2021, for further discussion). First, the response to the pandemic is characteristic of broader housing policy stances. The Scottish Government has a longer-term agenda of rebalancing power in the sector and enhancing tenants’ rights. This translated into more decisive moves to protect tenants from the negative housing consequences from the pandemic, working within the constraints of devolved powers while at the same time seeking to apply pressure to the Westminster Govt on reserved matters. The same can be said of the tenor of policy in Wales. In contrast, the UK Government, acting in relation to England on housing matters, showed much greater inclination towards seeking to minimise the extent and duration of interventions and a desire to get back to the market-dominated status quo ante as soon as possible. Second, the initial moves to up-rate benefits and to suspend eviction proceedings were viewed very positively by landlord organisations, tenant organisations and charities. There was, rather unusually, a broad coalition 129

of support for action from across the sector. Several stakeholders saw the pandemic as having triggered policy changes they had been seeking for some time. In England the crisis was seen by some as signalling the need to accelerate the passage of the Renters Reform Bill: it would be a surer and longer-lasting mechanism to deal with problems of tenant insecurity. However, this was a move the Government very explicitly resisted in the short-term, while maintaining that more fundamental reform of the rental sector was still on the medium-term legislative agenda. Most stakeholders viewed capacity constraints in the face of dealing with the pandemic as the explanation for lack of progress in this area, rather than the delay necessarily indicating a lack of Government enthusiasm for the reforms. Third, the policymaking process in the devolved administrations appeared to be more open and inclusive than in England. This characteristic has long been recognised as a function of the size of each polity, but stakeholders in England – while gaining access to Ministers and civil servants to make their case – did not feel that policymaking took the form of a dialogue. In contrast, the Scottish Government made an early move in response to the crisis of setting up crosssectoral resilience groups, including one focused on the private rented sector. Fourth, there was a general sense that policy was being made largely in the dark. That is, there was a limited amount of high-quality data arriving in a timely manner to inform policy. This allowed stakeholders to feel that, drawing on their own data to make the case for their preferred course of action, they were able to make an impression on politicians. It also meant that there was scope for stakeholders to frame narratives on the basis of the carefully chosen example or anecdotal case and for it to gain traction because there was limited scope to triangulate against other data. Crucial data gaps included variables such as the number of tenancies where the landlord and tenant were able to agree a strategy to deal with rental shortfalls and preserve the tenancy; the number of landlords viewing greater government regulation as the trigger to exit the sector; and the number of evictions that were continuing illegally despite the suspension of proceedings. Fifth, motivations behind policy – even where the policy continues in the same form – can be both more complex than they first appear and can evolve significantly over time. Whereas many stakeholders viewed the initial crisis response in England as having a public health focus, few thought that this was still the case in July or August when measures were extended. In the latter case the rationale was either considered to be primarily political or to be an exercise in last minute can-kicking while solutions were sought. The latter was itself seen as characteristic of the Government’s broader policymaking approach. Even where it was generally agreed that the initial intervention was driven by public health concerns the reasoning was perhaps less obvious than it might initially appear. While one might frame the public health issue as being about the damage to tenants from losing their homes and potentially finding themselves homeless, it appears that a significant factor in the decision was the inability to keep the courts running in a COVID-safe manner. That is, it is possible that the motivation was not so much about trying to reduce the inflow of eviction cases because that is inherently desirable, but rather reducing the inflow because there was no way of safely processing the volume of cases to maintain the corresponding outflow. The constraints of a legal system already under considerable strain appear a stronger driver than might have been initially appreciated. 130 COVID-19: Rental housing and homelessness policy impacts, 2021

8.2 Homelessness policy 8.2.1 Emergency policy measures In the UK, the pandemic prompted a remarkably speedy and successful nation-wide emergency response to the most extreme forms of homelessness (Fitzpatrick et al, 2020a). In sharp contrast to the Australian experience, central government played a pivotal role in terms of both leading and funding this unprecedented effort. Importantly, of course, such a comparison needs to factor in that Australia is a federal system whereas the UK remains (largely) a unitary state, at least as far as England is concerned. As noted in Chapter 1, housing and homelessness are the constitutional responsibility of Australia’s state and territory administrations; not, as in the UK, national government. At the same time, Canberra’s long-established subsidisation of state/territory housing activity23 is in part a recognition of the Commonwealth’s far superior fiscal powers – a reality of equal significance in the UK. The Australian Government may choose to utilise these powers to a greater extent in crisis circumstances – just as was seen in the Rudd Government’s 2008 housing response to the GFC (Milligan and Pinnegar 2010). In 2020, however, no such commitment was forthcoming. A letter was sent by the relevant UK Government Minister to all English local authorities on 26 March 2020 instructing them to move everyone sleeping rough, or living in communal shelters, into safe, ideally self-contained, accommodation over the following two days. As in Australia, this precipitated the widespread utilisation of commercial hotels, especially in larger urban centres, but also suitably configured hostels and supported accommodation. Use too was made of holiday lets, university accommodation and social tenancies. Crucially, the letter from Government that triggered all of this activity signalled that, in the exceptional circumstances of the pandemic, standard eligibility requirements for statutory housing assistance – ‘priority need’ status under the homelessness legislation and citizenship qualifications – could be set aside. This ‘Everyone In’ initiative was preceded by a £3.2 million UK Government funding allocation to local authorities for rough sleeper support, alongside £4.6 billion general funds to help councils cope with the overall financial pressures of the pandemic. In May 2020 the Government brought forward £161 million out of an (increased) £433 million four-year budget to provide 6,000 new supported housing units for ex-rough sleepers. In June 2020 a further £105 million was released to local authorities for interim accommodation shortly followed by £25 million part of which was targeted at areas with the largest numbers of rough sleepers. Faith and community groups received £2 million to make night shelters safer for use in winter 2020-21. All of this emergency expenditure (homelessnessspecific components totalling approximately AUD$500 million) was in addition to pre-existing centrally-funded homelessness programmes, including the third year of a Rough Sleepers Initiative, amounting to £112 million in 2020-21. The devolved administrations in Scotland and in Wales took similarly swift, decisive and comprehensive steps to provide self-contained emergency accommodation (EA) for people sleeping rough, or in unsafe forms of temporary accommodation, with levels of per capita emergency funding that were, in the case of Wales at least, more generous than those in England (Boobis & Albanese, 2020). In Northern Ireland, where rough sleeping levels 23 Under Commonwealth-State Housing Agreements, as from 1945; renamed as from 2009 – currently titled National Housing and Homelessness Agreement

131

are low, there was no specific targeted program, but a memorandum of understanding between government agencies committed to ensuring that rough sleepers were provided with EA. The UK Government estimated that over 90% of rough sleepers known to English councils at the beginning of the COVID-19 crisis were offered selfcontained EA. By September 2020, over 10,000 people continued to be accommodated under these Everyone In arrangements, while almost 19,000 people were reported to have been moved into settled accommodation or ‘a rough sleeping pathway outside of temporary accommodation’ (Ministry of Housing, Communities and Local Government 2020). According to the UK’s National Audit Office, a total of 30,000 people had been subject to EA placements by late 2020. Bearing this in mind, it would appear that those transitioned to move-on housing equated to almost two thirds (63%) of the total accommodated (NAO 2021). It has also been separately reported that, according to the UK Government ‘two-thirds of the 29,000 rough sleepers who were housed under the scheme have been moved into “settled accommodation” – defined as a tenancy of at least six months either in the private sector or with a housing association or council’ (Bulman 2021). The people helped under Everyone In, and the equivalent programmes in Scotland and Wales, included a great many ‘sofa surfers’ ejected by friends and family who were no longer able or willing to accommodate during the pandemic, as well beyond those sleeping rough and in communal shelters on the eve of the crisis. The combined efforts of the UK Government, local authorities and NGOs were successful in keeping COVID-19 infection rates low in the homeless population in England at least (Lewer et al, 2020), and there have been very few COVID-related fatalities amongst homeless people (Office of National Statistics, 2020). 8.2.2 Policymaking and implementation process Even pre-COVID-19, rough sleeping was already a high political priority across England, Scotland and Wales, with a specific central government target to ‘end the blight of rough sleeping’ by 2024 in England. Strengthened statutory duties to prevent homelessness have been enacted relatively recently in both Wales and England, with similar measures currently under discussion in Scotland. A commitment to more interventionist homelessness policies by the Theresa May-led UK Conservative Government, signalled a retreat from the ‘localist’ stance favoured by the 2010-2015 Conservative-led Coalition Government (Fitzpatrick et al, 2020b), was felt by some stakeholders to have laid the foundation of a more effective response to the COVID-19 crisis than would otherwise have been the case:

…we're building on a really strong foundation here, because for a couple of years, [Government] have been funding expert advisors, rough sleeping advisors, and advisors in the homelessness advice and support team, the team responsible for the implementation of the Homelessness Reduction Act. Across the country, there's these really strong relationships between… advisors and the local authorities… that was in place prior to all this happening…so that [all] really helped to make sure that the [COVID-19 emergency accommodation] hotels were set up quickly. [Statutory sector stakeholder]

132 COVID-19: Rental housing and homelessness policy impacts, 2021

While NGOs, local authorities and other stakeholders were united in praising the swiftness and clarity of the early communications and actions taken by central government when the pandemic first hit, many were concerned at what was seen as subsequent ‘mixed messages’ and erosion of commitment as the crisis unfolded. A Ministerial letter on 28 May reminded local authorities that they could only lawfully accommodate people ineligible for benefits (due to citizenship status) following an individual assessment of risks to life. Some councils interpreted this communication as encouragement to take a tougher line on accommodating non-UK nationals:

…although there was that very decisive initial response from [Government] around the ‘Everyone In’ message, it almost feels now that there's been a step back from national government and very much like, 'Over to your local areas to now do it'… local authorities left to work out what to do with people who ordinarily they wouldn't be accommodating, so largely people with no local connection, people with no recourse to public funds. [Voluntary sector stakeholder] The large volume of emergency funding ‘sloshing around’ the homelessness system in England during mid-2020 was widely acknowledged. However, the proliferation of highly specified, short-term funding pots, focused overwhelmingly on rough sleeping, and heavily weighted towards capital over revenue expenditure, was severely criticised by stakeholders. The point was also made that these funds barely begin to compensate for the massive reduction in mainstream revenue funding for homelessness services over the past decade (Thunder & Rose, 2019). As in Australia, hopes were expressed that the enhanced partnership working that emerged in many parts of the country during the COVID-19 crisis could be retained in the longer-term. However, in the UK such comments tended to focus on the strengthening of relationships between the homelessness sector and mainstream health services, rather than amongst homelessness services themselves, where collaborative approaches are often already well established:

…there's a sense in the sector that, yes, although everyone is really exhausted, frontline-wise, they're also really proud of what's been done. There's a huge amount of positive energy around… in terms of working with the NHS, better partnership working locally, some of the flexibilities in the system… particularly…around the more entrenched group… people are just being a bit more creative and willing to take risks. If… some of that can be maintained, that would be great [Voluntary sector stakeholder]

8.3. Chapter conclusion Comparing the UK’s COVID-19 emergency responses in housing and homelessness policy to Australia’s, we see some notable common themes, but also some significant differences. In both countries, the central government responded to the income shock and risk of eviction primarily through incomesupport measures, while also supporting eviction moratoriums – although the latter were the subject of less decisive action. The UK Government’s restrictions on evictions took the form of extended notice periods and suspensions of proceedings, extended at the last minute as the emergency continued. The UK’s devolved administrations, particularly in Scotland, took more decisive legislative action, and like the Australian states and territories, these administrations have innovated new rent relief schemes. 133

In Australia and the UK, street homelessness was radically reduced through exceptionally assertive emergency accommodation programs, building on the pre-pandemic renewal in both countries of political interest in this rising problem. In the UK, this interest had registered with the central government, and it was actively involved in coordinating and funding local governments and NGOs in the COVID-19 emergency response; by contrast, in Australia the revival of homelessness policy has been at the state and territory level, and the Australian Government was similarly absent from the COVID-19 emergency response.

References Bulman, M. (2021) Hundreds of homeless people pushed back on to streets of London during first lockdown ‘due to lack of support’; The Independent, 12 January https://www.independent.co.uk/news/uk/home-news/homelessrough-sleepers-hotels-london-b1785437.html Fitzpatrick, S., Pawson, H. & Watts, B. (2020b) The limits of localism: a decade of disaster on homelessness in England; Policy and Politics; Online first, DOI: https://doi.org/10.1332/030557320X15857338944387 Fitzpatrick, S., Watts, B., & Sims, R. (2020a) Homelessness Monitor England 2020: COVID-19 Crisis Response Briefing; Crisis https://www.crisis.org.uk/ media/242907/homelessness_monitor_england_2020_covid19_crisis_ response_briefing.pdf Foye, C. and Marsh, A. (2021) Policy towards private renting during the pandemic: interim report, UK Collaborative Centre for Housing Evidence Lewer, D., Braithwaite, I., Bullock, M., Eyre, M. T., White, P. J., Aldridge, R. W., . . . Hayward, A. C. (2020) COVID-19 among people experiencing homelessness in England: a modelling study, The Lancet Respiratory Medicine doi: https://doi. org/10.1016/S2213-2600(20)30396-9 MHCLG (2020) Coronavirus (COVID-19) emergency accommodation survey data: September 2020; Ministry of Housing, Communities and Local Government https://www.gov.uk/government/publications/coronavirus-covid19-emergency-accommodation-survey-data-september-2020 Milligan, V. & Pinnegar, S. (2010) The comeback of national housing policy in Australia: First reflections; International Journal of Housing Policy; Vol 10(3) pp325-344 National Audit Office (2021) Investigation into the housing of rough sleepers during the Coronavirus pandemic, Report on Ministry of Housing, Communities & Local Government ONS (2020) Coronavirus and deaths of homeless people, England and Wales: deaths registered up to 26 June 2020; Office for National Statistics

https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/deaths/articles/coronavirusanddeathsofhomelesspeopleenglandandwalesdeathsregisteredupto26june2020/2020-07-10 Thunder, J. & Rose, C.B. (2019) Local Authority Spending on Homelessness: Understanding Recent Trends and their Impact. St Mungo’s and Homeless Link http://wpieconomics.com/site/wp-content/uploads/2019/04/Local-authorityspending-on-homelessness-FULL-FINAL.pdf 134 COVID-19: Rental housing and homelessness policy impacts, 2021

9. Conclusions Both in Australia and elsewhere 2020 has seen emergency policy innovations on rental housing and especially homelessness of scale and significance that few could have imagined possible. Australia’s emergency measures have played some part in preventing the spread of disease among the street homeless population and forestalling, or even preventing, the new homelessness surge that would otherwise have been expected to result from the pandemictriggered economic downturn. There is much to learn from this experience, about how such crisis policy innovation might be better handled in a future disaster scenario; about how the latter stages of the present crisis may play out for people at particular risk of housing stress and homelessness; and about the dispositions and capacities of the institutions and actors comprising Australia’s housing and homelessness systems. It is only through an in-depth understanding both of the pandemic emergency measures and of their institutional contexts that we can assess the prospects for building back better after the crisis has subsided. This concern is intended to form a central theme in the second report to be published out of this research later in 2021. In this report we have sought to shed light on the nature of relevant policy shifts, how these came into being, how they were implemented and with what effect. We have drawn on the rapidly expanding body of published research and statistical data evidence relevant to these issues. This has been complemented by our own primary research and secondary data analysis to provide a fuller, more rounded picture.

Market impacts The rental housing market impacts of the pandemic and recession are a crucial contextual backdrop for the crisis policy innovations that are the main focus of this research. As reported in Chapter 2, such effects have been notably diverse and, indeed, somewhat divergent. In the outer rings of Sydney, Melbourne and Brisbane, as well as in smaller cities and many regional areas, median rents for new tenancies either remained stable or rose during 2020. Falling markets were, on the other hand, characteristic of inner suburbs in the nation’s three largest cities. However, since many such areas also housed residents particularly vulnerable to loss of employment and income, the cost of rental housing did not necessarily become ‘more affordable’ as a result. Thus, changing rates of tenancy turnover, rents and vacancy rates seen during 2020 are not simply a reflection of changing lifestyle preferences.

Interpreting crisis policy formulation and implementation How can we interpret 2020’s rental housing and homelessness policy innovations in relation to theories of crisis policymaking? Deploying the term coined by Birkland (1998), COVID-19 could prove to have been a focusing event in this area, as for other policy domains. As Birkland thought possible, the 2020 public health crisis can be portrayed as having provided the space for advocacy in favour of normally disregarded disadvantaged groups, a punctuation point in the previous equilibrium situation (Zahariadis 2016).

135

In the area of rental housing regulation, two long-term features of the Australian model – ready termination of tenancies, and rent setting by markets rather than consideration of affordability – suddenly became problematic and open, at least temporarily, to modification. Although no jurisdiction effected a complete eviction moratorium – and some were very far from complete – there is a sense that policymakers and sector stakeholders have now experienced what’s possible when the law makes terminations more scrutinised and qualified, and tenancies more secure. There is, on the other hand, less clarity as to the possibilities around rent liabilities. For many sector stakeholders and other actors involved in pandemic response, the payment of rent from tenants to landlords was an ineluctable force in the economy (and was an argument for income support), and variations to liabilities were almost everywhere left to individual negotiation. Rent regulation, it appears, is a lost art in Australia. Particularly within the context of rough sleeper emergency accommodation programs, stakeholder interview evidence seems consistent with the notion that, as argued by Weible et al. (2011), advocacy coalitions can exert influence in such situations. Arguably, the coalitions that took shape in this case spanned citizens groups and NGOs, but also – at least in some instances – policy domain specialist colleagues within government. Backed by the former, the latter were empowered to mount a successful pitch for extraordinary funding within government. In part, this reflected the re-framing of homelessness as a public health – rather than a ‘personal bad choices’ – issue. Our research suggests that the 2020 emergency measures seen in rental housing and homelessness complements were far from simple top-down policymaking. Their formulation and enactment reflected considerable interest group lobbying and contestation. Governments were subject to both externally applied pressure and internal jockeying and debate. Policy refinement and implementation also involved inter-jurisdictional interaction ramped up to unusual levels. At the same time, along with other emerging evidence on pandemic housing and homelessness policy responses, our evidence points to a fundamental absence of national coordination and leadership. By comparison with all four comparator countries, including the two federal states within this group, the Australian Government’s role in the design and implementation of housing and homelessness policy responses (as distinct from cash payments) has been extremely ‘hands off’ in nature. Whether this becomes institutionalised in the new inter-governmental structures as part of the post-COAG national governance architecture under the National Cabinet remains to be seen.

References Birkland, T. A. (1998) Focusing Events, Mobilization, and Agenda Setting, Journal of Public Policy, 18, 1, 53-74. Retrieved from www.jstor.org/ stable/4007601 Weible, C. M., Sabatier, P. A., Jenkins-Smith, H. C., Nohrstedt, D., Henry, A. D., & deLeon, P. (2011) A Quarter Century of the Advocacy Coalition Framework: An Introduction to the Special Issue, Policy Studies Journal, 39, 3, 349-360 doi: 10.1111/j.1541-0072.2011.00412.x Zahariadis, N. (2016) Setting the agenda on agenda setting: definitions, concepts, and controversies. In Handbook of public policy agenda setting (pp. 1-24). Cheltenham: Edward Elgar Publishing. Photo © Gillian Vann, used under licence from Austockphoto 136 COVID-19: Rental housing and homelessness policy impacts, 2021

137