Policies and programmes to end homelessness in Australia: Learning from international practice

Authors


Cameron Parsell, Institute for Social Science Research, The University Queensland, GPN3, St Lucia, Queensland, 4072, Australia. E-mail: c.parsell@uq.edu.au

Abstract

Parsell C, Jones A, Head B. Policies and programmes to end homelessness in Australia: Learning from international practice.

Many welfare states throughout the industrialised world have recently implemented policies to achieve targeted reductions in homelessness. These policy and welfare initiatives differ across national contexts. They are similar, however, in moving away from social programmes that have essentially ‘managed homelessness’ towards interventions that seek to permanently end homelessness. Australia has recently adopted similar homelessness policy objectives. This article examines the manner in which Australian homelessness policy has been converging with international policy directions. More specifically, the article scrutinises Australian social programmes adopted from the UK and USA as a means to achieve strategic goals of reducing homelessness. It argues that although Australian homelessness policy objectives are converging with international policy, Australian programmes modelled on international successes do not have some of the elements shown elsewhere to be crucial for achieving sustainable reductions in homelessness. This may become central to explaining programme outcomes in future years.

Key Practitioner Message:Strategies aimed at permanently ending homelessness represent a significant shift to contemporary professional practice;Homelessness programmes internationally are now characterised by their branding or identification with evidence-based models;It is important to critically scrutinise these models, examining their core elements and the manner in which they are appropriated and incorporated across jurisdictions.

Background

Governments throughout the industrialised world have long implemented policies and social programmes to respond to homelessness. Perennial debates about what constitutes homelessness and how it is counted (Anderson, 2007; Veness, 1993; Williams, 2010) mean that definitive data are not available for making clear comparisons across time and place. Indeed, comparing homelessness across developing and developed states (Tipple & Speak, 2009) or states with different institutional arrangements (Busch-Geertsema & Fitzpatrick, 2008) is problematic. Nevertheless, it is commonly agreed that homelessness in many industrialised states has increased between the 1980s and early 2000s (National Alliance to End Homelessness, 2000; United States Interagency Council on Homelessness, 2010). In response to the growing and observable homelessness problem, welfare systems in the USA, Canada, Europe, the UK and Australia have recently adopted new strategic policy directions with the specific objective of achieving permanent reductions in homelessness (Australian Government, 2008; Benjaminsen, Dyb & O'Sullivan, 2009; Calgary Homeless Foundation, 2011; Cunningham, Lear, Schmitt & Henry, 2006; European Federation of National Organisations working with the Homeless, 2010; Mayor of London, 2009; Randall & Brown, 2002).

This contemporary policy is not homogeneous. As Benjaminsen et al. (2009) pointed out, similar homelessness strategies in different welfare states will have different outcomes even though the broad policy is underpinned by a desire to move away from ‘managing’ to reducing homelessness and maintaining people in sustainable housing. This policy direction is premised on the assumption that, with an adequately resourced welfare system and targeted social programmes, the problem of homelessness can be solved.

In Australia, setting targets to reach measurable reductions in homelessness represents a shift away from crisis-based responses that merely manage homelessness. This new focus is presented as an ambitious and once-in-a-generation opportunity (Australian Government, 2008). Through increased funding into affordable housing, preventive strategies and the adoption of new models of intervention from international contexts, Australia is working towards achieving this ambitious target. Given this context of change in Australia's contemporary homelessness policy direction and the introduction of social programmes based on international models, this article aims to locate Australia's new strategic direction of achieving measurable reductions in homelessness through the lens of policy convergence, drawing lessons from abroad. First, it outlines the conceptual literature about the ways in which policy lessons are learned and adopted across international contexts. Next, it demonstrates how Australia's homelessness policy direction is changing, and how these changes are in line with and influenced by similar policies from the UK and USA. Third, building on the discussion on Australia's national homelessness policy and its convergence with international policy initiatives, the article critically examines the ability of Australia's Street to Home social programmes to achieve reductions in homelessness. Drawing upon empirical research from a larger study (Phillips & Parsell, 2012) with Street to Home programmes in two major Australian cities, Brisbane and Sydney, the latter part of this article considers the extent to which these programmes are implemented in a manner consistent with the international models on which they are based. It is argued that the initial implementation of these international programmes in Australia has not been well-resourced or supported with appropriate institutional arrangements. As a result, they are shown to: (i) diverge from their original model, and (ii) lack crucial elements that contributed towards their claimed successes in their original countries.

In recent years, homelessness programmes and policy initiatives have been branded as or identified with ‘proven models’. These include: Common Ground, the Foyer Model, the Rough Sleepers Initiative (RSI), Housing First and Street to Home. However, these programmes have yet to be fully evaluated in relation to actual service delivery arrangements to determine whether their implementation is consistent across contexts, to discern which aspects of the original model are most successful and to consider their effectiveness in reducing homelessness.

Policy transfer and policy convergence

The transfer of homelessness programmes and policies across national boundaries reflects the growing prevalence and significance of policy transfer in a globalised world (Banks, Disney, Duncan & Reenen, 2005). Its importance in contemporary public policy analysis (James & Lodge, 2003) is evidenced by the emerging empirical and theoretical work examining the nature and processes of policy transfer across national borders (Dolowitz & Marsh, 2000; Gilardi, 2010; Knill, 2005; Pawson & Hulse, 2011; Rose, 1991; Schmitt, 2011). In one of the seminal papers in this area, Dolowitz and Marsh (2000) viewed policy transfer as an iterative process ‘involving the adoption of policies across a number of different nations, and subsequent adaptations within individual nations’ (p. 6). As an iterative process, policy development, transfer and diffusion are a constant process of implementation, evaluation and improvement, and the likelihood of policy convergence across contexts is high given the benefits of information technology.

According to Gilardi (2010), policy diffusion explains ‘why and how policy choices in one country are influenced by prior decisions in other countries’ (p. 651). According to Rose (1991), when policy makers are confronted with problems, they can draw ‘lessons’ from other areas (countries, states and regions) to assist them in responding to similar problems in their own area. In this way, Rose (1991) explained that, ‘if the lesson is positive, a policy that works is transferred, with suitable adaptations. If it is negative, observers learn what not to do from watching the mistakes of others’ (p. 4).

Schmitt's (2011) recent study of the telecommunications sector in OECD countries suggests that governments only superficially learn from each other's policies. Furthermore, the transference usually takes place between countries that are near-neighbours and already have similar policy and economic institutions. Scholars engaged in the study of policy diffusion have sought to explain why the transfer of policy ideas has taken place; for instance, the ideological assumptions and motivations of leaders to identify carefully what is transferred, including the transfer process and to evaluate the specific outcomes of each policy transfer process (Dolowitz & Marsh, 2000; Gilardi, 2010; Knill, 2005).

In the context of significant scholarly interest in this burgeoning area, James and Lodge (2003) made the important observation that ‘it is hard to think of any form of rational policy making that does not, in some way, involve using knowledge about policies in another time or place to draw positive or negative lessons’ (p. 181). The essence of their argument is that policy transfer is a step in conventional rational or evidenced-based models of policy development that involve reviewing problems and searching for answers from available information. Indeed, they argue that ‘even rational policy makers’ preferences for the status quo in their own jurisdiction could be seen as implicitly involving negative lessons about alternatives in other countries or in other times' (James & Lodge, 2003, p. 181).

According to Knill (2005), policy transfer and policy convergence are different processes. Policy transfer is essentially concerned with processes wherein ideas about policy are moved from one context to another, while policy convergence is the end result of the transfer process when policy similarities emerge in different contexts over a period of time. ‘Successful’ policy transfer inevitably leads to policy convergence. The relationship between policy transfer, implementation and convergence is of central concern to the discussion that follows.

Dolowitz and Marsh (2000) identified several policy transfer states: (i) ‘successful transfer’ where policy convergence is achieved; (ii) ‘uninformed transfer’ where the borrower lacked understanding of the policy or context from which it is borrowed; (iii) ‘incomplete transfer’ where central features or dimensions of the policy are not transferred; and (iv) ‘inappropriate transfer’ where the proposed change was premised on a poor understanding of the social, political, ideological or economic differences between the host and borrowing contexts. Generally, however, policy transfer is not an ‘all or nothing’ process but represents a continuum from learning about a policy to adapting it to fit new contexts to implementing it and evaluating it, or deciding that it is not applicable at all.

The transfer process is thus not linear or apolitical and is often messy and incomplete (Lendvai & Stubbs, 2007; Needham, 2010). As Dolowitz (2000) recognised, the process is influenced by the ideological and political position of the borrowing parties in terms of where they look for policies and what policies they accept. Similarly, the role of epistemic communities and international consultants can shape the policy transfer process. Hulme's (2006) work demonstrates this complex and fluid process by taking into account the multi-levelled nature of the stakeholders involved, such as politicians, bureaucrats, academics, advisors and practitioners. Lesson learning and policy transfer are not simply an objective process of adopting the ‘best evidence’. Dobbin, Simmons and Garrett (2007) noted how stakeholders often ‘learn’ lessons that cohere with their beliefs, and, sometimes, they are based on fads and popular global norms rather than empirical evidence.

Rose (1991) developed a typology to capture the complexity of the transfer process, believing that, ‘in the real world, we would never expect a programme to transfer from one government to another without history, culture and institutions being taken into account’ (p. 21). Thus, varying degrees of copying, emulating, hybridising, synthesising and inspiring take place as lessons are learned in the transfer process. Dolowitz and Marsh (2000) reduced Rose's typology to four degrees of learning: copying, emulating, combining and inspiring.

This literature is helpful for providing a framework to understand the process of policy transfer and the factors affecting ‘incomplete transfers’ in relation to contemporary Australian homelessness policy. The discussion now turns to the influences of homelessness policy from the UK and USA.

Recent developments in Australian homelessness policy

A policy platform of the Labor Government elected to office in Australia in 2007 was the national crisis of housing affordability and the problem of homelessness. Incoming Prime Minister Kevin Rudd described homelessness as a national disgrace, lamenting the presence of more than 100,000 homeless people on any given night in Australia (Chamberlain & MacKenzie, 2008). This was seen as unacceptable, not only in the context of Australia's economic prosperity, but also in relation to the country's egalitarian culture. Rudd (2008) suggested that, unlike many countries where homelessness was tolerated or accepted, Australians ‘are not like that’. These comments by the new Prime Minister gave a heightened sense of urgency to the need to address the homelessness issue as a policy priority.

One of the first tasks of the Rudd Labor government was to develop a policy vision and plan for responding to the urgent problem of homelessness. After a broad public consultation process (in response to a Green Paper), the government released a White Paper in December 2008 (Australian Government, 2008) outlining Australia's homelessness policy agenda for the next 12 years. Central to this policy was the explicit objective for services ‘to end homelessness for their clients permanently’ (Australian Government, 2008, p. 46). Two targets were set that subsequently shaped homelessness initiatives and programmes: (i) ‘to halve overall homelessness by 2020’, and (ii) ‘to offer supported accommodation to all rough sleepers who need it by 2020’ (Australian Government, 2008, p. viii).

These objectives represented a fundamental change in Australian homelessness policy. In the face of the homelessness crisis, which Prime Minster Rudd recognised, prior policies and programmes funding short- to medium-term homeless accommodation and street-based outreach programmes focusing on harm minimisation were deemed inadequate (Bullen, 2010; Parsell, 2011). The prior crisis response since the inception of the Supported Accommodation Assistance Program in the mid-1980s, was not in any way intended, designed or resourced to achieve a permanent reduction in homelessness but merely managed the problem (Chesterman, 1988; Erebus Consulting Partners, 2004).

The Australian Government (2008) described its new homelessness policy direction as ‘a once in a generation opportunity to drastically reduce homelessness’ (p. iii). Following the White Paper, national and state governments committed to a 55 per cent increase in funding for homelessness strategies (Australian Government, 2008). The National Partnership Agreement on Homelessness provided for an investment of AU$800 million over 5 years as part of a broader AU$6.2 billion National Affordable Housing Agreement (Council of Australian Governments, 2010).

The new homelessness strategy converged with similar initiatives in other industrialised democracies. Indeed, when establishing and implementing the new policy direction, Australian politicians and policy documents explicitly cited successes achieved in the UK and USA in permanently reducing homelessness (Australian Government, 2008; Borger, 2010).

International homelessness policy

UK example

The UK represents one of the first successful examples of setting and meeting targets of reducing homelessness (Randall & Brown, 2002). UK homelessness policy from the early 1990s sought to eradicate rough sleeping. The 1990 Single Homelessness Initiative – later and more commonly referred to as the RSI (see Anderson, 1993) – was implemented against a backdrop of public concern about people sleeping rough on London's streets. In the early stages of the RSI, Anderson (1993) claimed that the programme was neither able to reduce rough sleeping substantially nor to address the structural causes of homelessness. Anderson's critiques notwithstanding, the RSI was shown to have reduced rough sleeping in London in the early 1990s (Randall & Brown, 1999). The precise extent to which rough sleeping was reduced was unknown, and, by 1994, the reductions had reached a plateau, but the evaluated success of the RSI in London meant that the initiative was then implemented more widely across England. Likewise, Scotland adopted a RSI based on the English model (Anderson, 2007), and, in 1999, the Scottish Executive set a target of ‘eliminating the need to sleep rough in Scotland by the end of 2003’ (Fitzpatrick, Pleace & Bevan, 2005, p. 7). While shortages of suitable housing were a barrier to facilitating exits from homelessness in some areas, access to housing primarily occurred through the large social housing sector and the RSI objective was deemed to be ‘largely fulfilled’ (Fitzpatrick et al., 2005, p. 103).

The Blair/Brown Labour Government (1997–2010) positioned homelessness as an extreme form of social exclusion and extended the prior Conservative government policy of setting elimination targets. The 1999 Rough Sleepers Unit (RSU) was established to oversee a reduction in the number of people sleeping rough in England by two-thirds by April 2002 (Randall & Brown, 2002). Building on the former RSI, the RSU consisted of street outreach workers adopting a persuasive approach to help people sleeping rough move into accommodation. The RSU achieved the two-thirds reduction (Randall & Brown, 2002), and, in 2008, the national policy No One Left Out set a target of completing the task of ending rough sleeping in England by 2012 (Communities and Local Government, 2009).

At the local level, contemporary homelessness policy in the UK has sought to build on national targets of reducing homelessness, with a commitment to permanently end rough sleeping in London (Mayor of London, 2009). In 2009, the London Delivery Board decreed that: ‘By the end of 2012 no one will live on the streets of London, and no individual arriving on the streets will sleep out for a second night’ (The Pavement, 2010). While targeted reductions in rough sleeping have been critiqued on the basis of not addressing structural causes (Anderson, 1993) and pursuing social cohesion ends over those of social justice (Fitzpatrick & Jones, 2005), they have unambiguously reduced – although not eliminated – the incidence of rough sleeping (Jones & Pleace, 2010; Wilson, 2011), and thus the public display of homelessness. Moreover, UK lobby groups, peak bodies and homelessness service providers have highlighted an overarching objective of permanently ending homelessness (Crisis, n.d.; Homeless Link, n.d.; St Mungo's, 2011).

USA example

Efforts to end homelessness have been embraced at the local, State and Federal levels in the USA. Nationally, the establishment of strategies to achieve an end to homelessness in the USA can be traced to the National Alliances to End Homelessness' (2000) A Plan, Not a Dream: How to End Homelessness in Ten Years. Set against a welfare context that prioritised shelters, transitional housing and soup kitchens, the plan to end homelessness was described as a new vision for the USA (Cunningham et al., 2006). The plan to end homelessness placed emphasis on prevention and active engagement among all stakeholders across the government and not-for-profit sectors, as well as the development of detailed data to assist planning at the local level. It was premised on the assumption that prior policy focused on ‘managing’ homelessness was inadequate and a firm belief that ending homelessness was within the nation's grasp (National Alliance to End Homelessness, 2000). By 2006, commentators were able to report that ‘220 communities have undertaken efforts to end homelessness and 90 communities have completed plans to end homelessness’ (Cunningham et al., 2006, p. 3). Four years later, Howard (2010) reported that these initiatives had spread throughout the USA, with 1,000 mayors and counties developing plans to end homelessness.

However, Cunningham et al. (2006) noted that, while setting a positive precedent, a significant number of local plans to end homelessness did not set clear numerical indicators, establish timelines or identify funding sources. Conscious of these limitations, they highlighted the importance of taking ‘plans off the shelf and moving from planning to action’ (Cunningham et al., 2006, p. 5). New York City's Common Ground Street to Home intervention worked towards achieving targeted reductions in homelessness at a neighbourhood level, and reported significant successes in permanently reducing chronic homelessness (Common Ground, n.d.).

The Federal Government played an increasing role in taking action to end homelessness by endorsing local initiatives and plans. In 2001, the Department of Housing and Urban Development supported goals to end chronic homelessness within 10 years, and, in 2003, President Bush announced the Administration's commitment to do the same (Cunningham et al., 2006). The Homeless Emergency Assistance and Rapid Transition to Housing Act of 2009 established the US Interagency Council on Homelessness with a mandate to develop a national strategic plan to end homelessness (Maine Democratic Party, 2010). In June 2010, a report entitled Opening Doors: Federal Strategic Plan to Prevent and End Homelessness was submitted to President Obama and Congress. It was described as the first comprehensive strategy to end homelessness, and, for the first time, national goals and measurable outcomes were set (United States Interagency Council on Homelessness, 2010). Central to the Opening Doors policies were the objectives of ending veteran and chronic homelessness by 2015, and ending homelessness among children, families and youth by 2020 (United States Interagency Council on Homelessness, 2010). While homelessness persists in the USA, arguably at a growing extent due to housing foreclosures in the economic downturn (The United States Conference of Mayors, 2008), the contemporary homelessness policy direction represented a significant shift from the emergency and crisis responses that characterised the 1987 Stuart B. McKinney Homelessness Assistance Act (Lee, Tyler & Wright, 2010).

Australian parallels

Australia's contemporary policy objectives of setting measurable targets to end homelessness reflect the initiatives, plans and policies implemented in the UK and USA over the last 20 years. As already noted, Prime Minister Rudd (2007–2010) elevated homelessness to the status of an important social policy problem (Australian Government, 2008; Rudd, 2008), and set ambitious targets to reduce homelessness within a 12-year period. These targets, together with recognition of the inadequacies of existing policies and programmes, created the need to devise new strategies and set the scene for learning from the experience of others. In this context, Australian policy makers looked to the perceived successes of initiatives in the UK and USA as providing models for programmes to be implemented within Australia (Australian Government, 2008; Borger, 2010). For example, two Australian State Governments, South Australia and Tasmania, commissioned Rosanne Haggerty from the New York Common Ground initiative to visit Australia to advise on ways to reduce homelessness based on the New York experience (Haggerty, 2006, 2008).

At the macro level, the goal of reducing homelessness had been transferred to Australia from the UK and USA. Australian policy makers were explicit in drawing upon lessons learned from the UK and USA when establishing national policy directions. The degree of policy convergence is unclear, however, since the outcomes have yet to be measured. That is to say, it cannot be assumed that Australia's adoption of policy objectives from the UK and USA to achieve targeted reductions in homelessness was coupled with the measures and resources those countries, or smaller locations within them, used to execute their policies into practical programmes. In order to examine the extent to which Australian national homelessness policy was resourced appropriately – consistent with the international policies and programmes on which it was based – it is important to examine the practical implementation of the policies. By adopting international homelessness goals and programmes, is Australia on track to reduce homelessness, or are crucial elements that ‘made the policy or institutional structure a success in the originating country not being transferred’ (Dolowitz & Marsh, 2000, p. 17)? The discussion later considers the extent to which recent Australian approaches to achieve targeted and measurable reductions in homelessness have been consistent with the models on which they were based.

Street to Home: a contemporary Australian homelessness programme

As one initiative to achieve the objectives of reducing homelessness and offering supported accommodation to all rough sleepers who need it by 2020, Australia introduced Street to Home programmes. The National Partnership Agreement on Homelessness, the main agreement between the national and state and territory governments to implement the national goal of reducing homelessness, identified as a core output the establishment of ‘Street to Home initiatives for chronic homeless people (rough sleepers)’ (Council of Australian Governments, 2009, p. 5). State and territory governments across Australia were required under this Partnership Agreement to introduce Street to Home models as a means to achieve sustainable housing and accommodation outcomes to end homelessness (Council of Australian Governments, 2009; New South Wales Government, 2009a). In Australia, Street to Home was promoted as an innovative model distinct from previous crisis approaches seen as underdeveloped and not mandated or resourced to end homelessness (Australian Government, 2008).

New York City-based supportive housing provider Common Ground identified itself as the architect of the Street to Home model (Common Ground, n.d.). Developed in 2003, Common Ground set out four key activities that constituted the essence of the Street to Home model: (i) establishing a registry of people sleeping on the streets; (ii) prioritising housing to those identified as most vulnerable; (iii) actively assisting people to access housing immediately; and (iv) arranging for the provision of personalised services, especially to assist with tenancy sustainment (Common Ground, n.d.). The Common Ground Street to Home approach was modelled on the earlier British Rough Sleeping Initiative.

In Australia, Street to Home developed as a model comprising three key features: (i) street outreach, (ii) permanent housing, and (iii) ongoing housing support. Street outreach involved the provision of services in situ to people outside of the premises or offices of the service provider. Its intention was to target individual need and tailor services accordingly so as to end homelessness (Queensland Government, 2008). In this respect, it was referred to as assertive outreach (Phillips, Parsell, Seage & Memmott, 2011). The Street to Home model was also underpinned by the Housing First approach (Borger, 2010; Queensland Government, 2008), whereby service users were provided with immediate access to permanent housing and not required to transition through homeless accommodation. Finally, the Australian Street to Home model emphasised the availability of an ongoing and integrated suite of support services to help individuals permanently exit homelessness (New South Wales Government, 2009b; Queensland Government, 2008).

Australian policy makers adopted the Street to Home model based on its perceived successes in the UK and USA and its national implementation was bolstered by the reporting of positive outcomes achieved in South Australia (Australian Government, 2008; Haggerty, 2006). It is a quintessential example of ‘lesson-drawing’ from international contexts. The New South Wales Government's (2009a) version was called Way2Home to ‘correlate with the evidence base for the Street to Home approach’ (p. 7).

The three dimensions of the Street to Home model – street outreach, permanent housing and ongoing support services – were seen as evidence based given the UK's success with persistent and assertive outreach (Randall & Brown, 2002) and the demonstrated effectiveness of Housing First in the USA (Stefancic & Tsemberis, 2007; Tsai, Mares & Rosenheck, 2010; Tsemberis, 1999; Tsemberis, Gulcur & Nakae, 2004). The factors leading to the success of the USA model included immediate access to permanent housing and social support and health services provided by Assertive Community Treatment (ACT) teams.

In the early stages of its implementation of Street to Home, Australia diverged from some of the fundamental tenets of the model as enacted in the UK and USA. The remainder of this article draws on a case study research for a larger project (Phillips & Parsell, 2012) which examined the first 12 months of Brisbane's Street to Home and Sydney's Way2Home interventions.

Sydney and Brisbane case studies

Sydney is Australia's largest city and Brisbane is its third largest city. The Street to Home projects were comprehensive street outreach initiatives located in inner-city areas. The fieldwork for the research involved in-depth interviews with 16 key stakeholders employed in the two interventions in funding, policy and practice capacities and documentary analysis of policy material pertinent to the two programmes. Interview participants were purposefully recruited on the basis of their position in or knowledge of the programmes. The documents were subject to a thematic analysis to identify the features, aims and capacities of the programmes.

The programmes were established in April 2010 with funding through the National Partnership Agreement on Homelessness. While they differ in terms of structure, practice dimensions and funding arrangements, they target people sleeping rough through street outreach and aim to achieve sustainable housing solutions. Street outreach workers patrolled particular public places searching for people living on ‘the streets’ and offered a range of health and social support services to people in situ. These workers are skilled in engaging with people sleeping rough and identifying their needs through using the Vulnerability Index Tool, developed in the USA to target the most vulnerable. Although not new to Australia (Parsell, 2011), the street outreach provided by the programmes under study is directed purposively towards assisting people in exiting from homelessness, rather than just managing homelessness with harm minimisation responses.

The Brisbane street outreach programme was found to be hampered by limited available housing options, which USA research had shown to be pivotal (Burt et al., 2004; Kryda & Compton, 2009). Unlike Common Ground and Housing First programmes in the USA, which has housing stock leased from the private rental sector and funded in part by the Federal Government Section 8 housing voucher, the Brisbane programme neither owned nor leased housing from private providers but relied upon public and community housing, and people living rough identified through the Street to Home programme in Brisbane were assessed for social housing on the same basis as other housing applicants. The programme thus did not have priority or guaranteed access to social housing. Nevertheless, outreach workers achieved remarkable success in sourcing housing for its service users through lobbying, networking and building positive relationships with housing providers. This success was due to the ingenuity and practices of the workers on the ground.

The Sydney program similarly did not have access to an adequate supply of housing stock and the majority of service users either continued to sleep rough or were assisted with temporary homeless accommodation. Neither programme had access to permanent housing, the first aspect of the USA Street to Home model.

The so-called wrap around and multidisciplinary support services which the policy boasted (Queensland Government, 2008) were not available to Brisbane's Street to Home service users and there was little in the way of individually tailored health services to cater for the ‘complex needs’. It was difficult to assess the support provided by Sydney's Way2Home programme as most of the 40 service users had only recently been housed. The multidisciplinary health team, comprising drug and alcohol, physical and mental health specialists – similar to the ACT teams from the USA – had the professional capacity to provide the support services needed, but it was too early to determine its effectiveness.

Discussion

Rose's (1991) warnings about policy transfer proved ominous in relation to the Australian experience described earlier. While Australia's homelessness policy converges with USA and UK policy, its programmes have not been adequately resourced and will likely not lead to reductions of homelessness on the scale reported in the regions of the UK and parts of the USA. The emphasis on evidenced-based homelessness policy notwithstanding (Australian Government, 2008; Borger, 2010), without resources for assertive outreach, support services and housing stock, these programmes cannot work. The two brief case studies suggest that difficulties in implementation went beyond taking account of, and responding to, local contexts and individual needs. This has significant implications for the capacities of the programmes to achieve their targets over time. The research assessing these programmes is not in a position to measure housing, health and homelessness outcomes. However, the paucity of housing available, together with the limited capacity to provide follow-up multidisciplinary support to those people housed, suggests that exiting people from homelessness will be difficult, and assisting those who receive housing to sustain their tenancies will be similarly challenging. Compared with transitional models of housing, it is the immediate provision of housing and the availability of client-directed services that are the hallmarks of success for programmes modelled on Street to Home principles. The design and resource limitations that undermine the extent to which Street to Home programmes in Australia can emulate the international models on which they are based are also echoed in the housing supply problems noted by Pawson and Hulse (2011) in relation to the transfer of choice-based letting policy to Australia.

Conclusion

Australia embarked upon an ambitious policy direction aiming to achieve targeted reductions in homelessness but failed to resource it appropriately. Attracted by the successes in the UK and USA, Australian policy makers attempted a piecemeal transfer making it unlikely that the goal of permanently ending homelessness will be achieved. This article has provided a brief discussion of two Street to Home programmes, assessed the extent to which these programmes reflected the international models on which they were based and examined their likely capacity to achieve their targets. It has argued that, while constituting a significant investment, the programmes have been implemented in a manner that does not include the essential features important for achieving sustainable reductions in homelessness. It might be seen as a model of ‘inappropriate’ or ‘incomplete’ transfer should the government decide to resource programmes and implement the overseas models in their entirety.

In the lead up to the 2010 Australian federal election, the parliamentary opposition leader Tony Abbott said the targets were unrealistic and unachievable (Perusco, 2010). Other policy commentators have made similar remarks (Johns, 2010). When policy objectives to achieve social goods are not resourced in a manner to make the targets realistically achievable, the resulting weak outcomes arguably lend support to criticism that these policies were inherently mistaken. The article drew on two Street to Home case studies to argue that it is not sufficient to emulate programmes without appropriate resourcing and ongoing research.

Ancillary