Collaborative Governance for Affordable Housing in Toronto and Melbourne

Policy formation and implementation have largely shifted from a top-down government-led process, to a collaborative governance approach characterised by complex and opaque partnerships, weakly steered by the state.  We use 36 interviews, undertaken in Toronto and Melbourne between 2015 and 2018, to assess procedural accountability in these two cities: the extent to which policy outputs developed through a partnership approach are fair, transparent, rational, and intentional.  We find that both cities fail the basics of procedural accountability, and that there is little shared understanding amongst key partners – local and provincial/state policymakers, non-profit and private sector housing providers, and philanthropic and private sector finance providers – about the definition and missing quantum of affordable housing, let alone a sense of how to move forward.


Introduction
In this article: we ask the question: "How well are collaborative governance processes working to develop policy serving the public good?" We use as case studies the housing regimes of two mid-sized cities of the Global North -Toronto, Canada and Melbourne, Australia -that have moved over the last 50 years from top-down housing policy formulation to largely privatised affordable housing delivery accomplished though diffuse partnerships.
We begin by developing a conceptual framework for the evaluation of procedural accountability in policy formation through the discussion of the literature, drawing on evaluation frameworks derived from public administration and urban planning literatures. We then describe methods applied in this paper and situate the research within a broader international comparative study conducted between 2012 and 2019. We provide an overview of the housing and governance contexts in Melbourne and Toronto before highlighting aspects of procedural accountability in housing policy in both cities. While the institutional contexts of both cities are integral to analysis, our primary research question focuses on whether housing processes and policies are fair, transparent, legitimate and reasoned and whether collaborative governance is achieving its purported aims in these cities. Our main contribution is a framework for assessing procedural accountability in cross-sectoral partnerships and insights into the contextually dependent ways in which process may be co-opted, impoverished or supported. We conclude with a discussion of how this analytic framework might be applied in other contexts.

Collaborative Governance
Under neoliberalism, policy formation and implementation have largely shifted from a top-down government-led process to an increasingly crosssectoral and multi-scalar form of governance, that shapes policy areas such as housing, education, health and transport (Emerson, Nabatchi, & Balogh, 2012;Geddes, 2006). In this context, goods and services once directly delivered, or at least strongly controlled by government, are now the product of complex and opaque partnerships, weakly steered by the state (Fainstein, 2010;Milward & Provan, 2000). This shift is often rationalised as a response to governments' necessity to partner with other sectors to access critical competencies, maximise the return from scarce resources, and provide goods and services it can no longer deliver on its own (Andrews & Entwistle, 2010;Kettl, 2015). These benefits may be achieved through "collaborative governance," defined as "the processes and structures of public policy decision making and management that engage people constructively across the boundaries of public agencies, levels of government, and/or the public, private and civic spheres in order to carry out a public purpose that could not otherwise be accomplished" (Emerson et al., 2012, p. 2). For proponents, collaborative governance provides a mechanism to generate and implement policy and represents "an alternative to confrontation, top-down decision making or paralysis" (Innes & Booher, 2003, p. 33). At its best, it may elicit new knowledge, innovative approaches, a sense of common purpose that outlasts political change and individual and collective empowerment (Emerson et al., 2012;Innes & Booher, 2010).
However, collaborative governance also presents significant challenges. The shift from senior governments exerting 'command and control' for public goods, to a less efficient and poorly controlled set of funding relationships with local governments, private and non-profit providers, has been analysed as a contributory factor to increasing social inequality. In the case of housing, these indirect partnerships in the face of growing 'marketisation' of affordable housing has been described by many researchers as a case study of failure to secure a basic right (see Walks & Clifford (2015), and Kalman-Lamb (2017) for Canada; Mullins, Milligan & Nieboer (2018) and Power & Bergan (2018) for Australia).
Collaborative governance is often viewed as the handmaiden of neoliberalism by detractors who point out that power imbalances, political manoeuvring and differentiated access to resources often renders collaborative governance a tokenistic process that further entrenches existing power dynamics (Parés, Boada, Canal, Hernando, & Martínez, 2017). For Swyngedouw (2005Swyngedouw ( : 1991, collaborative governance raises concerns about power, repre-sentation, accountability, legitimacy, scale of governance and point of intervention, as "governancebeyond-the-state" can reinforce the powerlessness of government and civil society to steer positive social outcomes in public-private partnerships.

Evaluating Collaborative Governance: Measuring Procedural Justice and Rationality
We take these problematic aspects of collaborative governance as the starting point of this paper, situating our analysis in the context of neoliberal policy agendas. In particular, we analyse the "democratic deficit" often attributed to collaborative governance, where opaque and complex relationships between partners blur the lines of responsibility and accountability (Steets, 2010: 32). In order to improve outcomes within neoliberal regimes that see themselves as relationship managers rather than as direct providers, a high degree of accountability to the public as well as to key actors is necessary. Page, Stone, Bryson, and Crosby (2015) suggest that crosssectoral partnerships should be judged based on three accountability criteria: democratic , procedural and performance. Democratic accountability refers to the degree to which the collaborative partnership reflects the interests of affected stakeholders. Procedural accountability refers to whether policy outputs are fair, transparent, rational and intentional. Performance accountability refers to the outcomes and whether they achieved public good at a reasonable cost.
In this article we focus on the notion of procedural accountability. We focus on procedural elements in this paper to reflect the most common criticisms levelled against collaborative governance. Procedural justice refers to stakeholder's perceptions of fairness and transparency in collaborative processes. Procedural rationality requires planning decisions typified by evidence (data that underlies decisions) and intentionality (knowing a goal and working towards it). We extend Page et al's. (2015) framework by proposing specific criteria for judging levels of procedural rationality and justice in the context of housing policy.
Based on this, as well as the analytic frameworks of Page et al (2015), we ask the following questions to evaluate procedural justice:

1.
Was engagement based on the expressed desire to co-design policy? 2.
Are there signs of trust, commitment, and mutual understanding being developed? 3.
Is there evidence of new institutional arrangements, leadership, knowledge or resources? 4.
Are there joint actions arising from this deliberation?
While it is too early to judge the outcomes of policies developed within the past three years, it is possible to judge procedural rationality: the extent to which policies reflect evidence-based targets and clear decisions. We use criteria developed through Berke and Godschalk's meta-evaluation (2009) of metropolitan strategies in 12 US states delivered under smart growth legislation. Berke and Godschalk (2009) address internal validity -whether the plans provide sufficient information about goals, possible alternatives, spatialisation, costs, monitoring and evaluation -as well as external validity -coherence with legal, regulatory and fiscal instruments of other levels of government. We thus ask the following questions of both sets of policies: 1.
Are the plan directions and support mechanisms based on data that establishes needs and problems, including clear definitions, spatialisation of the phenomena, and a focus on particularly affected population groups? 2.
How does the identification of issues lead to numeric targets, and a monitoring and evaluation framework? 3.
Is there a selection of options based on a transparent analysis, leading to an estimation of costs and sources of revenue as part of recommendations, as well as mechanisms to overcome any fiscal or legal impediments? 4.
Are there clear leads (governmental, private or non-profit) for initiatives?

Interviews
This article arises from a research program based in Melbourne, Australia, that seeks to improve both the amount and quality of affordable housing, by engaging with a range of key actors in local and state government, private and non-profit housing development, and philanthropic and institutional investment. For five years, the advisory committee has set a research agenda focusing on better policy outputs and scaling up innovative projects, along with capacity building in sectors and amongst researchers (Whitzman, 2017). As part of the research program, we have compared housing policies and partnerships in Melbourne, Australia; Toronto and Vancouver, Canada; and Portland, the United States. These cities were identified as matching pairs through a comparative review of governance, policies and partnerships described on their respective local government websites as well as both academic and media coverage of their affordable housing outcomes. In Vancouver and Portland, productive partnerships were generating modest improvements in affordable housing policy, funding, and completion outcomes despite national government inattention. In Toronto and Melbourne, weak partnerships were failing to generate shared understandings of affordable housing and need. The outcomes in terms of policy, funding and meeting affordable housing needs were weak.
In 2015 and again in 2018, matched sets of 6-8 interviews were undertaken in each of the four cities, with local and state government senior housing policy officers, private developers active in affordable housing provision (or industry associations where these were active in affordable housing policy advocacy), non-profit housing developers (or industry associations representing this group), philanthropic and institutional investors, and an affordable housing partnership involving these sectors (if applicable -see Table 1). The interview participants were selected based on their engagement with affordable housing partnerships, as identified by housing researchers in each city and a desktop review of affordable housing policy documents. The most engaged organisations, and the people leading those organisations, changed more rapidly in Toronto and Melbourne than they did in Vancouver and Portland (Raynor and Whitzman, 2020). A limitation of this approach is that we spoke to housing and planning policy makers rather than treasury officials, who may have had more say in state/provincial policymaking (Jacobs, 2015).
The interview participants were asked questions about their organisation's understanding of affordable housing; what their organisation's motivations were in working on affordable housing, the current need for affordable housing (expressed if possible in number of dwellings, as well as size, location and design concerns); how and with whom they worked in partnerships to create and advocate for affordable housing; what mechanisms (financial, regulatory and design/construction) they considered most important in terms of affordable housing; and what they considered successful in terms of affordable housing partnerships. In 2018, they were asked to focus on changes that had taken place over the past three years.

Strategies
In addition, as part of a related masters student project, we interviewed a total of 15 housing actors, most of whom were the same actors in the same roles, in 2017, asking slightly different questions about their engagement in the development of affordable housing and metropolitan planning strategies published in 2016/17, and their assessment of the likelihood of the strategies succeeding in providing more and better affordable housing. The specific policies were: 1.
The revision of the Long Term Affordable Housing Strategy (LTAHS) for the Province of Ontario, first developed in 2010; the goal of March 2016 strategy update was that "Every person has an affordable, suitable and adequate home to provide the foundation to secure employment, raise a family and build strong communities" and the implementation relies on "consultation and collaboration with our partners" (Ontario Government, 2016, pp. 3, 39); 2.
The development of Victoria's first affordable housing strategy in March 2017, Homes for Victorians, intended to "to give every Victorian every opportunity to find a home" using "a coordinated approach" (Victorian State Government, 2017a, p. 2); 3.
The revision of the metropolitan planning strategy of Plan Melbourne 2017-2050 in February of 2017, whosegoals include "housing choice in locations close to jobs and Canadian Planning and Policy 2020  The revision of the Places to Grow: Growth Plan for the Greater Golden Horseshoe Region, which encompasses Greater Toronto as well as smaller cities in the region, published in May 2017, with guiding principles including "support a range and mix of housing options, including second units and affordable housing, to serve all sizes, incomes, and ages of household" and mechanisms include the "need for stakeholders to work collaboratively" (Ontario Government, 2017, pp. 6, 12)

Canada and Australia: two neo-liberal welfare regimes
Urban governance is both context and issue dependent, as policy networks do not form in vacuums.
Over the past four decades, affordable housing policy in much of the developed world has followed an increasingly marketized trajectory, transitioning from government-led provision of public housing accompanied by strong land use controls in the mid-20th century, to partnerships based on largely private sector provision of affordable housing (Aalbers, 2015;Austin, Gurran, & Whitehead, 2014). Funding for social (public and non-profit) rental housing provision in both Canada and Australia decreased, even as need for affordable housing increased. Private ownership and rental housing both became increasingly commodified, with the emphasis on increasing the wealth of homeowners, rather than meeting a basic need (Dalton, 2009;Suttor, 2011).
As nations, Australia and Canada are two "'liberal welfare' regimes alike in history, culture, economy, federalism and residual social housing" (Suttor, 2011, p. 255). Australia, influenced by British social policy, invested heavily in public housing aimed at the working class after World War II, much of which was subsequently sold off to tenants between 1955 and 1970. While Canada had a smaller postwar public housing program, it took a more ambi-tious approach in the 1960s and 1970s, supporting non-profit cooperatives, municipal and community housing providers, and private developers to create mixed-income affordable housing (Suttor, 2011).
In both countries, the bulk of direct and indirect federal housing subsidy has been directed at increasing homeownership, with most beneficiaries being middle and high income households (Daley, Coates, & Wiltshire, 2018;Walks & Clifford, 2015). This has occurred through loans to individual homeowners to build or renovate their homes, mortgage insurance, exempting primary residences from capital gains taxes, grants such as First Time Home Buyers' Allowance, and in Australia, allowing losses on secondary residences to be written off against taxable income ("negative gearing"). These mechanisms are often taken for granted in public discourse around housing policy, with political focus placed on the costs of direct subsidy to low income housing, even though the costs of subsidizing homeownership are much greater (Austin et al., 2014;Hulchanski, 2006;Marcuse, 2012).
In Canada, the federal government gradually withdrew funding from supply-side affordable housing programs from the 1980s onwards (Suttor, 2011). The federal government downloaded responsibility for social housing policy and provision to provinces, while maintaining some cost-sharing programs. The province of Ontario subsequently downloaded responsibility to local governments, including the City of Toronto (Suttor, 2011). However, in 2015, a new federal government promised investment in new social housing as well as increased rent assistance and repairs to existing housing as part of an ambitious infrastructure strategy. While still not addressing the previous four decades' social infrastructure deficit (Ruckert & Labonté, 2016), the federal government's change of direction informed Ontario's preparation of a revised affordable housing strategy.
In Australia, there has been an overall erosion of policy and programs supportive of affordable housing at the federal level since the 1980s, with the most precipitous decline in the last five years. The election of a right wing national government in 2013 had as an almost immediate effect the abandonment of a social housing infrastructure funding program as well as a separate tax concession-based program to private developers of affordable rental housing (Austin et al., 2014). A series of nationalstate agreements on provision of affordable housing have emphasised demand-side rent subsidies to households, most of whom rent in the private sector. However, there is an increasing gap between the rent subsidy and market rents in most cities, and the absence of a national build to rent sector has meant that both supply-side and demand side housing policies are failing low income households (Hulse, Reynolds, Stone, & Yates, 2015). As a consequence, both researchers and policy-makers appear trapped in an endless cycle of "busy work" government consultations and inquiries that lead nowhere (Gurran & Phibbs, 2015, p. 711).

Province of Ontario/State of Victoria: Second-tier governments with primary housing responsibility
Toronto and Melbourne are subject to metropolitan planning and housing policy primarily created and implemented by provincial (in Canada) and state (in Australia) governments. Although their tax mechanisms and revenue streams are much more constrained than is the case for federal governments, state/ provincial governments have several tools at their disposal. Taxes can be waived for affordable housing or used to encourage housing closer to jobs and infrastructure. Second-tier governments can donate or lease government land to social housing providers, and guarantee construction and maintenance loans. State/ provincial governments can implement landlord-tenant legislation to control rents, limit conversion of rental units to owned condominiums, and make evictions more difficult. They can also allow local governments to enact mechanisms like inclusionary zoning (requiring a proportion of new development to be available at specific price points), support planning mechanisms such as secondary dwellings on lots, and better coordinate the regulation of housing densities in relation to employment centres, public transport and social infrastructure. Both Ontario, Toronto's provincial government, and Victoria, Melbourne's state government, have been conservative in most of these approaches. In particular, there has been reluctance to allow local governments to use mechanisms such as inclusionary zoning, although both Ontario and Victoria have recently paved the way for local govern-ments to use this mechanism (Davison et al., 2012;Zon & Dragicevic, 2014).
As metropolitan regions, both Toronto and Melbourne are typified by extremely unaffordable housing markets, weak planning powers, a history of political and policy instability and a small and poorlyresourced social housing sector. Toronto has the highest proportion of households in core housing need in Canada, 20.7% of all households, with "core housing need" defined as those who are paying over 30% of their income on housing and/or are living in overcrowded or inadequately maintained dwellings (CMHC, 2018). Australia uses a slightly different measurement, housing stress, which measures the proportion of the lower two household income quintiles who are paying over 30% of their income on housing costs, but does not include overcrowding or state of repair. Melbourne has the highest rate of Australian homeowners in mortgage stress, 21.1%, with 35.6% of renter households in stress (Allen, 2019). This affordable housing deficit represents over 400,000 dwellings in each city.
Toronto and Melbourne both had metropolitan government structures abolished by their respective provincial/state governments of Ontario andVictoria (in 1997 and1988, respectively). Both cities have experienced a high degree of policy shift towards neoliberalism in both local and state/provincial politics over the past two decades (Raynor and Whitzman, 2020).

Toronto and Melbourne: Two Weak Cities
The 6.6 million people in Toronto's metropolitan area, the most populous in Canada, are governed by one large local city government and four regional governments. Ontario had a left-of-centre government from 2003 to 2018, with a change in that year to the right. However, relative political stability and an avowedly progressive provincial government was not accompanied by interest in affordable housing policy, and Toronto has experienced increasing disinvestment in its public housing stock, along with rapidly increasing unaffordability of private ownership and rental options (Walks & Clifford, 2015).
Melbourne is the second largest city in Australia.
With 4.9 million people and the fastest growth rate in the country, it is projected to become the nation's most populous metropolis by 2030 (Australian Bureau of Statistics, 2018). After the abolition of metropolitan governance, further reforms have left 32 local governments within the metropolitan area, all with very weak taxation and development control powers (Gleeson, Dodson, & Spiller, 2012). As is the case for Toronto, public housing construction in the 1950s through to the 1970s left a legacy of units poorly supported by federal-state agreements, accompanied by rapidly increasing housing and rental unaffordability (Austin et al., 2014, Suttor, 2011). An exacerbating factor in Victoria is what Sandercock (1976, p. 56) identified many decades ago as political instability leading to planning powerlessness: the five decades since the 1970s have seen five metropolitan strategies, all of which were intended to cover 30-40 years of infrastructure planning, none of which outlasted a change in state government (Whitzman & Ryan, 2014).
Without adequate taxation sources beyond property tax and developer contributions, local governments' contributions to affordable housing are limited. In Toronto, the local government can manage public housing assets, limit conversions of affordable rental housing to ownership, apply limited rent control, and manage property owners' objections to new affordable housing developments. Local governments in Melbourne cannot access even these limited affordable housing mechanisms. Some local (and senior) governments have provided land towards affordable housing development, and some Melbourne local governments have created housing policies that try to use these limited powers to encourage well-located affordable housing to meet local population needs.

Analyzing Procedural Justice in Toronto and Melbourne: Is Collaborative Governance really happening?
In both Toronto and Melbourne, weak and inconsistent steering has caused key housing actors to fluctuate between optimism and despair. What has been consistent, however, is a profound reluctance of state/provincial governments to genuinely engage in co-design of policy, despite lip service to this ideal.

Toronto: Cycling from despair to cautious hope and then back to despair
In Toronto in 2015, six of seven interview subjects agreed that provincial and federal government had lost the ability to effectively steer affordable housing outcomes. The provincial government policy officer concluded his interview by saying: "We don't know how to expertly guide this ship". But many interview participants felt they had innovative projects they wanted to scale up, including a private developer working on a large-scale public housing redevelopment, and a non-profit developer creating limited-equity homeownership for low income households. A new provincially funded affordable housing partnership, the GTA Housing Lab, was dedicated to bringing government, the private sector, non-profit providers, and researchers together to generate solutions to the seemingly insoluble affordable housing crisis. Interview subjects expressed cautious hope for better policy at the local and provincial scales.
Two years later, five of six Toronto interview participants reported having extensively engaged with the provincial government in the development of affordable housing policy, and indeed, there were many signs of engagement in co-design. For instance, the City of Toronto representative argued that they had effectively advocated to enable inclusionary zoning, although overall the LTAHS was not much of a "heavyweight policy document". A long-time affordable housing advocate described structured invitational workshops involving key stakeholders like private developers and landlords, investors such as banks, and social housing providers and advocacy groups. These acted as "sparring grounds" where contentious issues such as inclusionary zoning could be debated. The private developer industry association representative agreed that they had been "heavily consulted… very involved in the development of the strategy", through attending workshops, responding to drafts and oneon-one meetings. The social investor specifically mentioned how the 30 or so stakeholders in the Housing Action Lab had developed proposals that they felt were included in the housing strategy around recapitalising aging affordable private rental stock.
The participants were describing trust, co-learning, and co-design of the new strategy.
Unfortunately, the 2018 interviews, which occurred the month after the change to a right-wing provincial government, suggested a return to cynicism about good policy outputs, and despair about outcomes of the process. The social investor who was so excited about the GTA Housing Lab co-design of new policies and programs now said: "we were ineffective as a group in moving the issue forward". The same representative of a non-profit housing industry association who said in 2017 that "we finally have staff in provincial government offices who understand housing" now spoke of the "huge governance vacuum". A spike in engagement-related enthusiasm did not survive political shift, suggesting that co-design was not accompanied by new institutional arrangements that would allow these gains to be politically resilient.

Melbourne: From optimism to pessimism and back to limited optimism
Melbourne participants described a different trajectory: from optimism, to pessimism, to limited optimism again. In 2015, a change of state government had recently occurred, and the new left-of-centre government had promised a higher priority for affordable housing provision. An institutional investor described co-research with a local government on community land trusts, a private developer talked about the potential for creating investment opportunities for private market-driven low-cost rentals, and the general tone of all seven interviews was that substantive change was imminent, and that partnerships were working well to generate new knowledge and mutual understanding.
By 2017, there was a sharp division between the two state government informants and five of six other interview participants. The Victorian planning and housing department actors spoke of mutual "support and collaboration" across "the best state government interdepartmental committee I've ever been on", leading to "coordinated delivery of the strategy". One advantage of having representatives from all departments was that the emphasis shifted from affordable homeownership to encompassing the entire spectrum, including social housing and homelessness prevention. In other words, the horizontal integration across state government departments was considered a success, in terms of developing trust, commitment and mutual understanding and potentially the capacity to effectively steer affordable housing delivery.
The advocacy organisation for low income renters felt they had effectively used media campaigns and "one on one conversations with public services and with [the ministers and their political staff]" to engage in co-design. However, five other affordable housing actors were disappointed with their engagement. One of two non-profit housing informants put it most succinctly: "Two years of waiting and nothing [except] being told there were rivers of gold [coming]." A partnership made up of private and non-profit affordable housing providers met several times with state government staff, but according to the other non-profit informant, "it seemed very clear that state government bureaucrats couldn't say anything at all, which was frustrating". A local government housing officer felt that while private "development and community housing" sectors were being engaged, "there was no active consultation with local government". Yet the representative of the private development industry association said they had not been consulted. The philanthropic investor felt that despite committing millions of dollars into affordable housing construction, they had not been invited to the table.
In 2018, the division remained between state government self-congratulation and the feeling expressed by other actors that there was no accountability from the state government. The same local government officer stated: "the glass is half full… [the state government] have kind of slipped back into old habits of 'we'll tell you when we want to tell you what we think we need to tell you'", while a private developer active in affordable housing advocacy said the state government was "tinkering around the edges". However, some actors who had recently become directly engaged in new state government programs were more optimistic. One nonprofit developer said that relationships with state government "are strong. We've gotten more con-tracts". The philanthropist who felt excluded from the policy table in 2017 now pointed to "quite extensive" relationships with state government, including the possibility of providing matching government funding to philanthropic donations. Table 2. summarises our analysis of democratic accountability. In Toronto, co-design of the revised provincial LTAHS released in 2017 did not lead to new institutional arrangements or joint actions before a change of government in 2018 cast the future of that policy into doubt. In Melbourne, new trust, commitment and mutual understanding within state government was not matched by engagement in codesign with other actors: local government, private and non-profit providers, or investors. However, the roll-out of the policy may lead to new institutional arrangements.

Table 2. Procedural Justice in Affordable Housing Collaborations-Melbourne and Toronto (according to interview participants) Analyzing Procedural Rationality in Toronto and Melbourne: Are partnerships generating policy that promotes the public good?
Procedural rationality focuses on the manner in which a decision is made, and the policy outputs, rather than the outcomes of that policy (Munda, 2006). The quality of the policies released in 2016/17 does not bode well for affordable housing outcomes in either Toronto or Melbourne.

Ontario: Weak Policies and Inadequate Supports
Neither Ontario's Places to Grow regional strategy update nor the Long Term Affordable Housing Strategy update includes any research into affordable housing need. There is $2.5 million pledged towards "innovation, evidence and capacity building" (Ontario Government, 2016, p. 4), but no use of the excellent data on need, including spatialization and discussion of specific population groups produced by the Neighbourhood Change Project, the Ontario Non-Profit Housing Association (ONPHA), the Wellesley Institute, or the Social Planning Council. For instance, ONPHA has been collecting data on social housing wait lists since 2003; its 2016 report found over 175,000 people registered on these wait lists, with averaging four years before they are offered non-profit housing (ONPHA, 2016, p. 5). In terms of specific population groups, the planning strategy does mention the importance of good housing design to accommodate aging communities, but it does not discuss this within an affordability context.
In Places to Grow, the Ontario Government (2017, p. 22) places the onus for developing and monitoring housing targets firmly in local government hands, stating that: "municipalities must establish targets for affordable housing as defined in document and mechanisms to support those targets". However, without aggregate targets, this is an abrogation of provincial government steering, and has resulted in substantially different housing outcomes  Government, 2016, p. 5) proposes several regulatory reforms that do not have upfront costs for provincial government, and which support the intensification and well-located affordable housing aims of Places to Grow. These include enabling legislation so that local governments can enact inclusionary zoning and as-of-right accessory dwelling units. There is additional funding provided from the provincial government budget, although these are mostly in the form of demand-side rental subsidy in markets with low vacancy rates (ONPHA, 2016) rather than new affordable dwelling units. Examples are: $17 million over three years on a portable housing benefit targeted towards victims of domestic violence that would eventually assist "up to 3000 survivors"; $45 million added to the $294 million already in homelessness prevention initiatives; $100 million into supportive housing "to assist 4000 households in 3 years"; and "supporting construction of 1500 new units eventually assisting 6000 individuals and families" (Ontario Government, 2016, p. 2). While these options are costed, there is limited information on how they will be monitored and evaluated.
Finally, there is a very limited sense of how the provincial government will steer this partnership model and who is taking the lead. As discussed above, Places to Grow requires municipalities to set targets, but without a strong set of numbers on overall spatialised need, targets become program-based (how they will spend money, why they did not accomplish more) rather than needs-based (how they will meet their local population's current and projected needs). The Ontario strategies take pride in their consultative process, with "30 stakeholder meetings and 113 formal written submissions" underpinning the LTAHS update (Ontario Government, 2016, p. 10). There is generic wording about "consultation and collaboration with our partners" (Ontario Government, 2017, p. 39), but certainly no ongoing sense of how collaborative governance with effective steering might take place. In terms of needs assessment, targets, costed options and a partnership model, the Ontario policy documents fail the test of procedural rationality.

Victoria: Inadequate Policies and Weak Supports
The Victorian policies also fail the criteria set by this analytic framework. Plan Melbourne provides an overall need for 1.6 million new dwellings that need to be constructed over the 33 years covered by the strategy (Victorian State Government, 2017b, p. 43), but Homes for Victorians does not provide data as to affordable housing need. The omission of a needs calculation is particularly egregious since an affordable housing gap assessment was commissioned by the state government in 2016, but only released two months after the launch of the affordable housing strategy. This report estimated the need for at least 1,700 social housing units per year for the next 20 years to keep the proportion of social housing at its current level (3.5% of total housing stock), with double that annual amount to meet the needs of low income people in severe housing stress, and more of course to meet all households in housing stress (Victorian State Government, 2017c).
While Plan Melbourne speaks only of generic "housing diversity' assisting both younger and older people accessing appropriate housing, Homes for Victorians mentions specific programs to meet the needs of affected population groups, such as women and children escaping intimate partner violence as well as young people facing homelessness (Victorian State Government, 2017a, p. 39). It also discusses the role of public housing in meeting the needs of people with disability or mental illness (Victorian State Government, 2017a, p. 7). However, there is little detail as to how these needs will be met, nor are there targets for new or renovated homes.
Plan Melbourne mentions "strategies to be put in place" for targets at various price points within six newly established sub-regions of metropolitan Melbourne at an unspecified later date (Victorian State Government, 2017b, p. 46). But by March 2019, two years after the plan was released, there were no signs that the basis for this needs assessment and target development were being supported by the state government. The Homes for Victorians strategy pledged to renew 2,500 public housing units plus create 2,200 new social (public and non-profit) housing units within five years (Victorian State Government, 2017a, p. 31). Given that the estimated need for total new dwellings in both documents is a little less than 50,000 a year, that means that the state government is committing to a target of 0.8% of that new build quantum as affordable housing.
The main funding mechanisms announced in Homes for Victorians are: (1) $1 billion in direct investment; (2) $100 million in direct loans; and (3) $1 billion in loan guarantees to build capacity in social housing providers (Victorian State Government, 2017a, p. 4). It is unclear, however, how much of that investment money would be going directly to new social housing provision. For instance, $50 million was pledged towards a shared equity scheme to assist first time home buyers to copurchase 400 homes, and $187 million was pledged to the Public Housing Renewal Program, which will subsidize renovation of 1,100 public housing units in locations close to central city locations by selling off portions of the land to private developers. The state government has promised a social housing uplift of "at least" 10% on these redeveloped estates, or 110 units (Victorian State Government, 2017a, p. 56), but research on previous public housing renewal in the state suggests that four bedroom units may be replaced by one bedroom units, for a net loss in affordable beds (Shaw, 2013). Further major tax breaks for private homeownership were offered in the form of removal of home purchase tax (called 'stamp duty') for new dwellings priced at less than $600,0000, and there was an increase in First Time Home Owners' grants.
Plan Melbourne will "explore" a trial of voluntary inclusionary zoning agreements in major developments, in return for expedited approvals and other benefits such as increased height and/or density. As part of the very tentative introduction of this developer contribution scheme, it developed a standardized definition of affordable housing, and included "provision of affordable housing" as an aim in the Planning Act (Victorian State Government, 2017b). Anecdotal evidence based on 2018 discussions with local government planners suggest that the voluntary agreements are expected to yield 3-5% of total housing affordable to low-and moderate-income households; again, a drop in the bucket of housing need.
Discussions of collaborative partnerships with clear leads are as vague in Victorian policies as they are in those from Ontario. Plan Melbourne limits its discussion of the Commonwealth government to a landholding party to consult with, as part of the development of several "national employment centres" (Victorian State Government, 2017b, pp. 27-29).
Similarly, it speaks of working with "developers, local governments, and stakeholders to create sustainable, liveable and attractive places" (Victorian State Government, 2017b, p. 50), but not of higher level implementation partnerships that would engage these actors in problem-solving around clearly stated goals. Homes for Victorians, a product of a human services instead of a land use planning bureaucracy, is a much more overtly political document, criticising the federal government for taxation policies that have contributed to housing speculation as well as policy uncertainty about funding for affordable housing construction and ongoing housing subsidies (Victorian State Government, 2017a, pp. 3-5). Homes for Victorians is also stronger on partnerships criterion, discussing the importance of the federal government "coming to the table" with a substantive plan (Victorian State Government, 2017a, p. 3) and specifically advocating for the two levels of government working on a 'bond aggregator' model to stimulate private investment in the hitherto neglected investment rental sector (Victorian State Government, 2017a, p. 24). It concludes with the importance of working "in partnership with the Commonwealth Government and local councils, as well as the community social housing sector, and the development and construction industries" to achieve affordable housing outcomes (Victorian State Government, 2017a, p. 43). However, the strategies do not set out how this partnership will occur on an ongoing collaborative basis, despite key housing actors being the main delivery instruments for the affordable housing outcomes.

Comparing Ontario and Victoria
The comparative policy analysis is summarized in Table 3. The Ontario government's major new policy option is allowing local governments to implement inclusionary housing mechanisms. There is no real funding for new construction, with an emphasis instead on increased rent subsidy and maintaining an aging social housing stock. However, the federal government's affordable housing plan, which includes new supports for social housing, was released several months after this strategy (Government of Canada, 2017). The Victorian government, working with a less generous federal regime, is focused on developing its own funding stream for new construction as well as maintenance of existing stock. However, neither pair of policies provide an evidence base in relation to need. With very weak support from the metropolitan/regional plan in both cases, ways to scale up intervention to meet needs in a deeply unaffordable housing system, building on a collaborative governance model that uses skills and knowledge from other key housing actors, is absent from both plans.

Conclusion: Improving Affordable Housing Processes and Policies
The Failure of Collaborative Governance in Two Neoliberal Settings As this comparative case study demonstrates, a stated commitment to collaborative governance is not leading in any meaningful way to procedural justice in relation to housing policy in Toronto or Melbourne. There is little genuine co-design between provincial/state governments and local governments, non-profit or private housing providers, or potential providers of affordable housing finance. There is little in the way of consistent trust, commitment or mutual understanding; leadership, knowledge generation or new resources; or joint actions arising from new arrangements. Nor are these failed partnerships creating procedural rationality: policies that provide clear definitions, reliable estimates of needs, emphasis on equity-seeking groups, transparent costing or clear needs. Both processes and policies are failing miserably in terms of meeting huge and growing unmet affordable housing needs. Swyngedouw (2005) posits six potential explanations for a lack of accountability in "governancebeyond-the-state": 1.
incapacity or unwillingness of potential stakeholders 2.
lack of accountability and transparency in decision making 4.
absence of legitimacy in relation to definitions of problems and solutions 5.
reliance on the wrong scale of government 6.
separation of policy development and implementation. This analysis and comparison of processes and outputs in relation to affordable housing in two cities facing affordable housing crises focuses on two of these explanations. Lack of accountability and transparency in decision-making is a procedural justice issue in both cities. Similarly, an absence of housing need analysis and funded policy options are indicative of an absence of legitimacy in problem definition and solution generation.

A Framework to Evaluate Policy Processes and Outputs-and Relate these to Outcomes
This article does not focus on affordable housing outcomes. However, repeated interviews over a three-year period led to some insights as to how weaknesses in processes and policies can harm outcomes. There were some undoubted wins in Ontario's new strategies, such as allowing local governments the right to mandate inclusionary zoning. However, without the mechanisms to secure new institutional arrangements or leadership, inadequate policy outputs appear to be unsustainable after a change in government. While interview respondents felt at the time that accountability was manifested through the provincial government listening to concerns of various actors, the procedural rationality necessary to effectively steer housing outcomes were missing from both the revised regional planning and housing strategies. Without an expression of the huge need for affordable housing, meaningless targets were set. Without adequately costed options, there was limited opportunity to scale up production of new housing, or even to provide sufficient subsidy to support low income households in limited private rental options. Without either topdown direct provision of housing or adequate steer-ing of strong shared institutions, the optimism of key Toronto housing actors appeared to collapse as soon as there was a change in provincial government.
Melbourne, in contrast, demonstrates that an opaque process led to increased cynicism in relation to a collaborative governance approach. The state government appears to have acted in a more internally integrated manner than in the past in developing an affordable housing strategy. However, expanding this steering capacity into effective partnerships with private sector finance and development, let alone with a struggling community hous-ing sector, local governments, and social investors, is still a challenge. Collective ownership of solutions was not supported through procedural justice. The state government did not provide enough steering in its metropolitan planning and affordable housing policies. There was no real discussion of the urgent need for scaled up affordable housing, no spatialized targets, and only a vague sense of partnerships going forward. There is some possibility of implementation partnerships forming through some of the new options, including funding for new affordable housing and inclusionary zoning. However, the sense of collective efficacy in tackling the huge problem of affordable housing

Clear Leads
Emphasis on direction and enabling legislation to local governments (federal government plans released afterwards), little mention of ongoing co-development of programs with private or non-profit housing providers Limited direction and enabling legislation to local government and metropolitan sub-regions (Plan Melbourne); downloading public housing to social housing providers and working on increased investment with federal government (Homes for Victorians)

Table 3. Procedural Rationality-Affordable Housing Strategies in Toronto and Melbourne Compared
was not supported through collaborative policy codesign.
The purpose of this article was to develop a framework to evaluate policy processes and outputs and relate these to failures to achieve the right to housing. We present the framework from this research, as summarized in Figure 1, as a useful mechanism for practitioners and academics interested in assessing cross-sectoral partnerships across policy areas. We have argued that procedural accountability requires both procedural justice and procedural rationality and have proposed metrics for assessing the degree to which procedural justice and rationality are evident in a partnership. Procedural justice, with its focus on co-design, trust, new institutional arrangements and joint actions, forms the basis of ideal versions of collaborative governance. It can be assessed using interviews, focus groups and participant observation and it can be encouraged through actions that build mutual understanding and joint actions based on shared goals. This paper provides a series of questions and objectives that could be used to shape more procedurally just policy processes.
In addition, procedural rationality encourages an emphasis on the internal consistency and legitimacy of policy documents. The integral factors identified in policy documents -ie useable data, shared definitions, targets, monitoring mechanisms, cost estimations and delegation of responsibility -constitute elements of good planning more broadly. These elements add accountability and can aid in coalition building (Raynor and Whitzman, 2020). They can be easily assessed through content analysis of policy documents and can serve as a 'shopping list' of desired elements for practitioners developing plans and strategies. This conceptual framework brings together theoretical and practical insights from urban planning and public administration literatures to propose an actionable framework for assessing and developing policy.

Recent Opportunities and Threats
In both cities, the framework we have developed highlights specific ways to improve processes and policies towards desperately needed outcomes. In the two years since this research ended, the need for better housing policy regimes has only intensified. In 2019, the Canadian federal government adopted the National Housing Strategy Act (Government of Canada, 2019), which enshrined the right to housing at the federal level. However, the continuing primary emphasis on provinces and territories to steer housing policy raises concerns about the capacity of the federal government to enforce the right to housing. There are, as of yet, no national definitions for affordable housing, nor are there nationally consistent mechanisms to define need by economic or social groups, let alone a transparent costing of various policies and programs that might ensure that housing rights become reality. In fact, there are remarkably few requirements attached to agreements between the national and second-tier governments, which will make joint actions between all three levels of government almost impossibly challenging.
Australia has seen much less movement towards a national housing strategy. However, both countries are currently in the process of dealing with the global COVID-19 pandemic, which has both highlighted and exacerbated housing inequalities. On the one hand, the fact that it is impossible to self-isolate without an adequate home has led to rapid, albeit temporary, direct government provision of housing for those homeless households in greatest need in both countries (Farha, 2020). On the other hand, it is clear that predatory real estate investment trusts who scooped up properties in the wake of the 2008 Global Financial Crisis, exacerbating affordable housing need and stress, are already gearing up to take advantage of this latest economic and health crisis (Keil, 2020). It has become increasingly clear that without strong steering by all levels of governance, millions of households around the world are at risk of eviction and a new stage of housing crisis beyond previous comprehension (Rogers and Power, 2020). The scaffolding on which adequate affordable housing outcomes might be built requires stronger foundations, especially in this era of increasing need for genuinely collaborative partnerships between all levels of government, the private sector, and community organisations to provide for basic human needs.