Key findings Public housing transfers and affordable housing industry development Since the mid-1990s, but especially since around 2007, property transfers by Australian state and territory housing authorities have added significantly to the housing stock and capacity of community housing providers (CHPs). Latterly, there has been increasing interest among housing policy-makers in the prospect of further transfers, at a larger scale, as reflected by the 2009 Housing Ministers Conference target of community housing achieving a 35 per cent share of the social housing sector. This interest has been particularly motivated by the increasingly financially stressed and physically run-down condition of public housing, and by the perceived benefits of 'contestability' arising from a multi-provider social housing system. Crucial here are the revenue advantages afforded CHPs under current subsidy settings-in particular, CHP tenant eligibility for Commonwealth Rent Assistance (CRA) paid through the social security system, thus enabling CHPs to charge higher rents without reducing tenant net incomes. Until recently, however, public housing transfer programs have been relatively small in scale and experimental in nature (Pawson, Milligan et al. 2013). Public housing transfers post-2012 Since 2012, three Australian states with little prior experience of transfers have commenced ground-breaking new transfer programs. Tasmania's Better Housing Futures (BHF), South Australia's Better Places, Stronger Communities (BPSC) and Queensland's Logan Renewal Initiative (LRI) are considered here as case studies on contemporary public housing transfer policy and practice. BHF involved four parcels of properties (about 500-1,100 dwellings each, with some vacant land for development), representing, in total, 35 per cent of Tasmania's public housing stock; these were transferred to four CHPs, three based interstate. BPSC involved transfer of two parcels (about 500-600 properties each) to two SA-based CHPs. These two programs proceeded through to management handovers in 2014 and 2015, and the respective state governments are, at this writing, progressing further transfer initiatives with more ambitious objectives. By contrast, LRI, which would have been the largest and most far-reaching transfer program yet undertaken in Australia (about 5,000 properties to a partnership of two interstate CHPs), has recently been terminated, along with other planned Queensland transfers, after protracted preparations, political controversy and a change of government. Transfer objectives and models Relative to the objectives and models of transfers identified in our 2013 research (Pawson, Milligan et al. 2013), the case study transfers examined here consolidated, extended and innovated in various ways. All three programs embodied the objective of capturing CRA as the most important motivating factor. In South Australia (SA) and Tasmania, this enhanced revenue was directed to increased spending to address maintenance backlogs. More ambitiously, the Queensland Government aspired to leverage funding for large scale estate renewal and housing construction, which would have involved investment reportedly totalling $800 million. Building the capacity of the not-for-profit housing industry was an important secondary objective of each of the programs. Recognition of local industry capacity limitations was reflected in the initially modest objectives for renewal and growth adopted by the SA and Tasmanian governments. The case study transfers also consolidated the model of transferring property management rights by lease or agency agreement, rather than freehold title. However, they also extended the model by transferring for longer periods (terms of 10 years for Tasmania, and 20 years for SA and Queensland (as planned)). Competitive selection processes, without a role for tenants, also consolidated previous practice-although the openness to interstate providers represented a new development. Transfer processes, terms and tenant implications In all the case study states, the transfer selection, contracting and transition processes built capacity in government and housing providers, but were costly-most of all, of course, in Queensland, where the Logan transfer was aborted despite an extensive tendering process and a subsequent lengthy period of preparation for handover. In the two states where transfers progressed to completion, significant process issues were encountered-particularly in connection with Centrelink payments and the transfer of tenant credits and liabilities. The transfer contracts contain notable provisions relating to: • Government termination of contracts-raising questions of security. • Backlog maintenance liabilities-subject to spending caps that mitigate risk for CHPs. • CHP organisational management and tenancy management in accordance with social housing policies-raising questions as to the proper place for this level of regulation. Large-scale transfers raise questions around the employment of public housing staff; however, BHF and BPSC largely avoided the issue through prior recruitment freezes and internal redeployments within the public service-approaches that could not be replicated in a larger- scale transaction or program. LRI would have required the successor CHP to employ ex-public housing staff-but with the project's cancellation, the associated organisational challenges and opportunities were not seen through to implementation. None of the case study transfers sought to build the capacity or agency of tenants in the transfer process, but CHPs' post-transfer engagement with tenants and service improvement initiatives appear to have been well received. Transfer finances, accounting and CHP asset bases Financial modelling indicates that, through CRA-enhanced rent revenues, transfers of public housing to CHPs may be a viable way of achieving maintenance backlog reduction and, at the same time generating a modest revenue surplus to underpin other designated CHP activities. Employing social landlord income and expenditure assumptions derived from transfer tendering practice, this modelling focused on a number of '30-year business plan' scenarios for a notional 1,000 dwelling public housing transfer. These scenarios were compared with a 'continuing public housing management' base case. Allowing for the elimination of a maintenance backlog averaging $15,000 per dwelling, our transfer 'base case' generated an operational surplus over the business plan period sufficient to leverage construction of 113 new homes. Of these, 13 would be for market sale, 29 to replace obsolete transferred public housing, and 71 as additional affordable housing units. Alternatively, in the most favourable scenario-where strong government action facilitates access to cost-effective private finance (through a financial intermediary) and free land (through planning interventions)-it was estimated that leveraging could yield as many as 557 new homes. Of these, 77 would be for market sale, as well as 143 units to replace worn out public housing and 337 additional affordable dwellings. However, any transfer contract commitment for a recipient CHP to carry out larger scale catch-up repairs and/or to undertake non-landlord activities (e.g. place making, housing advice and support) would quickly erode and eliminate this development capacity. And, even in the most favourable circumstances imaginable, the social housing financial regime would (over 30 years) enable the successor landlord to replace only a very small proportion of the ageing transferred portfolio. The above objectives may be achieved where CHPs are granted a mere leasehold, as opposed to freehold, interest in transferred properties. This is because lenders consider 'long-lease' acquisitions of former public housing as potentially sufficient to underpin cash flow-based lending to reputable providers. For accounting purposes, 'long lease' is now being interpreted as including contracts of as little as 10-year duration. As confirmed through recent practice, proper accounting treatment of such transactions involves the asset concerned being recorded as a 'disposal' on the public accounts. It must be emphasized that significant questions linger as to what, under current policy settings, transfers may be reasonably expected to deliver. In particular, there is no validated information about the true scale of dwelling condition impairment in transferred property portfolios, nor on the time needed for 'catch-up repair' programs to eliminate such problems.