Consolidating the Australian Dream:
Reconfiguring the Multi-Unit Housing Network
SUMMARY OF THESIS SUBMITTED FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
Jasmine Samantha Palmer
PhD, Centre for Housing, Urban and Regional
Planning, Department of Geography,
Environment and Population, School of Social
Sciences, Faculty of Arts, University of Adelaide.
B.A. (Arch.), Deakin University
B.Arch. (Hons.), University of Adelaide
M. Design Science (Sustainable Design),
University of Sydney
Self- funded ‘strata’
project with or
with NFP Developer
funded by investors
funded by ethical
investors with limited
multi-unit housing and the Australian dream
The free-standing single-family home continues to embody the Great Australian Dream to the majority of the
population, despite multi-unit housing representing the majority of new dwelling construction in major cities
in recent years. Many Australians view multi-unit housing as a temporary housing arrangement on the path
toward achieving the Great Australian Dream, as a bespoke home for the urban elite, or as a housing option of
last choice (Randolph & Tice 2013). As Australian strategic urban plans promote more multi-unit housing
development in the interests of renewal, intensification, and sustainability, can the Great Australian Dream be
Australian homeowners have a tendency to personalise their domestic environments through the contracting
of original constructions and later renovation, remodelling, and extension. This familiar system
…reflects Australia’s distinctive form of ownership as it carries individualism to greater lengths
than in other ownership societies. Every Australian purchaser of a new detached dwelling has the
ability to mould it to their individual needs and tastes in a way that only the very affluent can do
in other societies. (Burke & Hulse 2010, p. 828)
In contrast, multi-unit provision is supply-led and, employing more industrialised construction and
development systems (Burke & Hulse 2010; Turner 1976), seldom offers opportunity for individualisation. As
speculative developers carry all financial risk, they hold substantial decision-making capacity. However, their
short-term interest in the buildings produced mean decision-making is informed by priorities and objectives
misaligned from those of potential owner-occupiers; generating “split incentives” on matters of use value,
sustainability, and maintenance (Easthope & Randolph 2016).
Pursuing infill development, Australia’s strategic urban plans indirectly ask households to forego the privilege
of housing personalisation so entrenched in the Australian psyche. They entrust the implementation of infill
housing to the speculative market, with its history of delivering commodified housing product attractive to
private property investors. This raises concerns regarding building quality and long-term built and social
legacies (Altmann 2015; City of Melbourne 2013) as market value is prioritised over use value, community, and
Strategic urban plans provide visions of “dynamic communities where people will want to live and work… each
[with] a unique character… a sense of belonging and connectedness, which will make them a drawcard”
(Government of South Australia 2010, p. 71). However, if speculative supply-led provision persists, it is more
likely consolidation areas will lack diversity of housing type, be comprised of buildings designed and built to
minimum standards, and be home to a high proportion of rental households with high mobility rates. To deliver
the strategic visions greater consideration of future occupants’ preferences is necessary.
Buy an Existing Detached
Buy an Existing Multi-
Initiate Design and/or
Construction of New
Initiate design and/or
construction of new
Home Purchase Options in existing Australian Provision.
based on Dolin et al. 1992.
Housing preference studies undertaken in Adelaide (Fischer & Ayturk 2011), Perth (State of Western Australia
2013), Melbourne, and Sydney (Kelly et. al. 2011) identified mismatches between household preferences and
the housing delivered via speculative provision. They concluded an owner-occupier market exists for quality
multi-unit development, but also show speculative multi-unit development falls short of meeting this demand
due to a lack of diversity and owner-occupant appeal. To consolidate the Australian Dream, it is necessary to
bridge the design and amenity gap between demand-led, free-standing dwellings and supply-led multi-unit
housing. One way of achieving this is to relocate multi-unit design decision-making responsibilities to future
residents. Enabling households to initiate the design and/or construction of new multi-unit dwellings
empowers then to construct, for themselves, the new urban communities strategic planners envisage.
The Research Question
In the interests of supporting households to initiate the design and/or construction of new multi-unit dwellings,
this research asks the primary research question:
What are the impediments to collective self-organised housing provision in Australian urban
Describes the existing system of multi-unit housing provision in Australia, focusing particularly on design
briefing and the determinants of dwelling function.
Investigates current examples of Australian multi-unit urban infill projects which seek an alternative to
existing provision systems.
Observes international housing sub-markets which enable user participation in the provision of multi-unit
urban infill housing.
Draws together information from both the international and Australian examples to identify opportunities
for reconfiguring the existing multi-unit housing provision system in Australia; redefining the role of
occupants/owners in the housing provision and urban consolidation processes.
In promoting urban consolidation, Australia’s strategic urban plans have the unintended consequence of
reducing, if not eliminating, an individual household’s capacity to directly engage with new dwelling
production. Contrasts exist between the production processes of a free-standing home, typically constructed
to contract, and that of a multi-unit building, typically speculatively designed and sold off-the-plan; with the
latter removing the capacity for future occupants to directly influence dwelling function or design.
Hence, in providing a vision of higher-density urban communities, Australia’s strategic urban plans arguably
contribute to a disjunction between the function of higher-density dwellings and the individual households
who seek to call them home. In response, a small number of Australian households and industry professionals
have proposed alternative production processes to realise multi-unit housing suited to future occupants’
This thesis examines the relational network of social and technical actors in existing multi-unit housing
development. Structures of housing provision are visualised through an actor-network lens, and network
relations are analysed to identify which (or what) network actors (or actants) influence design decisions,
particularly regarding dwelling function and cost. The existing actor-network is compared with those of four
alternative Australian cases. Employing both network analysis and primary interview data, impediments to
collective self-organised housing in Australia are identified. Comparison of the actor-networks of the
alternative cases recognises different types of network change in each case, reflecting participants’ motivations
and resources, with associated variation in outcomes. Means of addressing the impediments identified in the
Australian cases are drawn from two international cases, building groups in Berlin, Germany and collective
custom build in the United Kingdom.
The thesis argues successful consolidation of the Australian Dream advocated by contemporary urban planning
policies requires that future residents be provided with greater capacity to influence design decision-making
within the multi-unit structure of housing provision. Drawing lessons from the alternative cases, it proposes
reconfiguring the actor-network of multi-unit housing to enable this to occur. The comparison of Australian
and international case studies builds knowledge applicable to the development of policies and programs
promoting collective self-organised housing in Australia, with relevance also to other dwelling densities.
The Thesis comprises four Parts, each consisting of two or more chapters.
Part One sets the research Context (Ch1), introduces
the theoretical frameworks engaged in the research
(Ch2), and details research methods (Ch3).
Part Two: The Existing presents existing Australian
multi-unit provision over three chapters divided into a
review of secondary data (Ch4), the introduction of
primary data (Ch5), and subsequent analysis (Ch6).
Part Two concludes having provided a unique view of
the existing multi-unit housing network and identifying
key design decision-making influences.
Part Three: The Australian Multi-Unit Innovators
comprises three chapters. Chapter 7 introduces the
Australian Multi-Unit Innovators, Chapter 8 presents
the experiences of stakeholders to date, and Chapter 9
compares the individual projects, observing
differences in their capacity to reconfigure the black-
box via network analysis. Part Three concludes by
identifying impediments to alternative multi-unit
Part Four: International Multi-Unit Innovators
comprises two chapters. Chapter 10 introduces the
international cases, detailing their selection and
relevant locational features. It also identifies variables
amongst international self-organised housing sectors
which facilitate initial comparisons between the two
cases. Chapter 11 then draws lessons from the
international cases in response to the impediments
identified in Part Three.
The Thesis then concludes with response to the initial aims and research questions.
Part One: Context
Developing without Developers – Collective Self-Organisation
The dominance of speculative multi-unit housing provision is not unique to Australia, with the vast majority of
privately owned multi-unit dwellings in developed nations realised speculatively. However, around the globe,
a long history of demand-led multi-unit innovations exists as a minority contribution to housing supply and has
been the subject of renewed interest from both housing researchers and practitioners over recent years. Past
examples include collectively owned and managed housing in Mumbai, India, the Turkish Republic, Hong Kong,
Israel, and Argentina. Each emerged in the context of housing shortages resulting from conflict, rapid urban
growth, or social inequality, and a dominant housing system failing to meet the needs of a segment of the
System failures continue today, with dwelling shortages in many urban locations, declining affordability, and
an ongoing prioritisation of housing’s role as investment and commodity. In this context, self-organisation of
multi-unit housing is increasing in many western locations, either through the collective action of households,
civil society groups, or policy change.
Reviewing literature on alternative multi-unit housing provision across the globe, the thesis navigates an array
of terminology which is often inconsistent (Tummers 2011), with terms frequently misappropriated across
language divides. Collective Self-Organised (CSO) Housing is used in the thesis to mean housing in which
a “collective of individuals organize, finance, plan, and commission their own project” (Brunoro 2013 p1).
Discussing CSO housing in Germany, Junge (2006) describes it as “developing without developers.”
Self-organisation of housing, particularly multi-unit housing in which households benefit from the power of
collective action and capital, contributes to diversity and affordability, can support community development
and environmental investment, and can facilitate urban regeneration. It enables citizen participation in
development, facilitating “an alternative outcome that would not otherwise have existed” (Jarvis 2015, p. 205)
and realising a “flexible pathway towards diverse futures” (Ache & Fedrowitz 2012, p. 410).
Importantly, self-organisation offers a useful tool in the implementation of urban consolidation policies; one
which can respond to specific household aspirations, and avoid the lack of innovation typical of speculative
development. Since the emergence of the urban consolidation agenda, the discussion has continued as to
who will inhabit the new dwellings proposed, what will that dwelling form be, and where is it to be located?
These questions are not easily answered in a private, speculative housing market. They are more directly
addressed when the prospective inhabitants themselves are free to speculate on their own behalf and “build
for their own use” (Millington 2000, p. 27). This thesis aims to assist Australian households to do just that.
Understanding, Conceptualising & Visualising Housing
Transcending disciplinary boundaries, housing eludes the application of a singular theoretical framework. This
research integrates knowledge sets, concepts, methods and analysis techniques across disciplinary boundaries
to generate a distinctive view of housing systems, their effectiveness, and their limitations. Initially, multi-unit
housing provision is conceptualised following Ball’s Structures of Housing Provision Thesis (1986), in
which all individual and institutional stakeholders are recognised as actants. This holistic approach to viewing
housing provision facilitates effective comparison of alternative housing provision systems within and across
locations. Additionally, it enables identification of structural and relational attributes transparent to
disciplinarily-compartmentalised or problem-specific research.
Secondly, concepts drawn from Actor-Network Theory are introduced, including understandings of
network stabilisation and the creation of black-boxes, processes of network translation, and recognition of
human and non-human agency. Both Actor-Network Theory and Ball’s Provision Thesis view agency and power
as relationally generated, acknowledging relations of domination and subordination cannot be assumed but
become known only following analysis of network relations. Flows of design information into and through the
housing provision networks are observed to identify key actants in design decision-making and identify
opportunities for network reassembly.
Thirdly, understanding and comparison of the socio-technical networks of housing provision is enabled by
visualisation and quantification using Social Network Analysis software. Together, these frameworks
provide a layered theoretical lens to examine multi-unit housing provision, comparing alternatives and allowing
previously silenced actors to emerge.
Example of the network mapping used to visualise relations between
stakeholders and determine key design decision-makers. Shown here
is existing multi-unit provision with stakeholders sized by the network
analysis measure of network centrality (betweenness).
The actor-network visualisations developed to represent the structures of housing provision under examination
in this thesis provide unique insights in these cases and, with further development, may prove equally useful
to other research examining the complex web of social and technical actors in housing systems.
Research Design & Methods
The research employs a system-embedded case study design following Thomas (2011), including the use of
key, nested, and outlier cases to abductively build knowledge. Thomas' case study approach is shown to
converge appropriately with Actor-Network Theory in comparative housing research pursuing transferability.
Understanding of existing
practices and impediments to
Cases compared both with each other and
with the existing to generate knowledge.
Provide guidance in overcoming identified
impediments to future Australian CSO
Data Collection & Analysis Methods
Deliberative & snowball
Comparative analysis of ego-
Data triangulation through
Transcription & analysis of
Two distinct stages of analysis are
undertaken. The analysis of primary
and secondary data collected
generates network representations.
These are analysed and compared
with both visual observation and the
use of Social Network Analysis
software. Data is analysed both
qualitatively and quantitatively.
Part Two: the Existing
Strategic Urban Plans call for ‘medium-density housing’, ‘higher-density housing’, and ‘high-density housing’ to
transform Australian cities in the coming decades. In doing so, they do not question the existing ‘black-box’ of
speculative multi-unit provision and hence risk perpetuating historic mismatches between the multi-unit
housing produced and that desired by households attracted to the ‘Consolidated Australian Dream.’
Prior to pursuing alternative structures of provision (SoPs), this part of the thesis looks inside the existing black-
box. It asks ‘What influences the design and function of multi-unit infill housing currently being
constructed in Australia?’ and ‘What are the impediments to deviation from the existing multi-
unit infill housing supply system?’
An abundance of past research has described and interrogated existing multi-unit housing in production,
management, exchange, and/or consumption. Most research, however, addresses only one or two of these
sub-systems of provision described by Burke and others (Burke & Hayward 2000, Burke and Hulse 2010,
Burke 2012). Here, knowledge constructed by others (secondary data) is combined primary data from key
stakeholder interviews to construct a systemic view of the multi-unit SoP. Importantly, both humans and
non-human participants in production are recognised, with
texts (e.g. strategic plans, building codes, planning documents, legislation),
values & perceptions (e.g. market value, use value, risk perception, political risk),
artefacts (e.g. project design brief, building/dwelling design), and
organisations (as collections of humans – e.g. financial institutions, state planning authority)
conceived as capable of acting within the network and influencing resultant built outcomes.
The resultant actor-network mapping focuses specifically on the flow of design information between actants;
who (or what) provides input into design briefing, who (or what) sets limits or boundaries on design, and who
(or what) takes action, making design decisions.
Mapping of design information flows in existing multi-unit SoP.
Multiple levels of detail are visualised within the stabilised network, to understand why the black-box produces
the outcomes it does. In this case, why we have the multi-unit housing designs we have. The thesis uses the
actor-network mapping to observe existing provision from multiple stakeholders’ perspectives. For example:
Ego-networks identify who (or
what) influences an actant or
who (or what) an actant can
Mapping of flow paths of design
distance between actants and
barriers to information flow.
This flow path shows the
connection between Use Value
and the Project Design Brief to
be lengthy and indicates design
input from potential owner-
occupiers is mediated by
and external risks prior to
informing the design brief.
With the use of network-analysis software (UCINet), network attributes are used to identify focal actors,
mediators and obligatory passage points (OPPs) in design decision-making.
Those who acts to align the interests of others with their own
property developer, development profit, market value, urban design master plan,
local planner, selling agent, marketing consultant, financial institutions, risk
perception, local planning documents, tax legislation.
Those with capacity to transform, translate, distort, or modify design information
to suit their interests
urban design master plan, construction costs, property developer, design team,
financial institution, marketing consultant, local planning documents.
Those who define a situation which must be met; holds power of approval
market value, development profit, financial institution, development investors,
urban design master plan, local planning documents, property developer.
Simplified Mapping of existing multi-unit SoP showing only those
actants identified with capacity to influence design outcomes
using network metrics. Shows absence of design input from
consumption sub-system and dominance of non-human actants.
Key Findings from Part One: the Existing
Market value and development profit are the primary drivers of multi-unit dwelling provision, with these
monetary measures carrying greater influence in the SoP than the use value of dwellings.
A disconnection exists between the existing production and occupation subsystems, with limited opportunity for
occupants to engage in, or directly inform, the housing production process.
Potential multi-unit occupants not engaged with the available investment-purchaser driven product remain silent
within the existing actor-network.
Multi-unit owner-occupiers are not able to align the interests of others with their own, transform, translate,
distort or modify design information to suit their interests, or hold power of approval.
Key design decision-makers are concentrated in the production and management subsystems, only about half of
whom are human actants capable of negotiation and possible innovation.
Strategic urban plans are not focal actors in determining multi-unit design outcomes. In projects where one
organisation or individual carries multiple key roles (eg a developer assuming roles as development investor,
construction contractor, and design team leader) the influence of strategic urban plans is further dimished.
Strategic-level and project-level actants have very different understandings of the location of risk within the
network, of who in the network decides what types of dwellings to build, and the influence purchasers are able
to exert on dwelling design. The interviews supported the previous premise that multi-unit housing remains a
closed black-box to many strategic actants.
The co-location of decision-making with financial risk in the existing SoP leaves those actants exposed to long-
term risk (liveability, community, environmental) without influence and allows the motivations of actants seeking
short-term financial profit to translate information flows and mediate design decisions on their behalf.
In answer to the research question, ‘What influences the design and function of multi-unit infill housing
currently being constructed in Australia?’, this case has shown the key influencers to be:
financial profit, financial risk, and financial systems, (incl. banking institutions and tax legislation)
Equally, it has revealed who or what has the least influence on design and function of multi-unit infill housing,
that being the potential owners/occupants and use value. In particular, the SoP excludes potential multi-unit
owners/occupants not represented by previous multi-unit purchasers.
Impediments to deviation from the existing multi-unit SoPs identified include:
reluctance on the part of key actants to compromise their current network position for fear of financial
the resistance of existing network connections, in which those actants who may seek an alternative
network solution (such as use value, future occupants, or architects/designers) are not located in a
structural network position which enables them to modify the existing network.
the stability of existing network connections, meaning that despite significant reconfigurations
occurring between actants in the production subsystem, the network properties of actants in the
management subsystem (policy, planning, financial, institutional and contractual systems of
development) change very little in response.
Part Three: the Australian Multi-Unit Innovators
The stabilised multi-unit structure of provision (SoP) described in Part Two of the thesis has dominated infill
housing for some time, limiting industry and community exposure to alternatives (Martel et al. 2013a).
Individuals and groups seeking to navigate an alternative multi-unit outcome are pursuing a small number of
innovative infill housing projects in Australia. In the spirit of Collective Self Organised (CSO) housing projects
they seek to enable owner-occupier input into design to varying degrees. This part of the thesis examines four
such innovative examples in the form of a nested, outlier case study (Thomas 2011a). At the time of data
collection (Feb 2015) one case was occupied, one was in development, and two were yet to be realised. All
seek an alternative to the developer-led, speculative SoP,
comprise 4-25 dwellings in a single development of 3 or more storeys,
are located in areas designated for consolidation in strategic urban plan, and
were initiated by future residents or professionals seeking improved higher-density living outcomes.
The four cases each have distinct approaches to navigating or reconfiguring the existing multi-unit SoP.
Publically available information was collected for each case and interviews conducted with key stakeholders,
including residents, future residents, architects, instigators, development partners, and government advisors.
The interviews focused on three main inquiries:
stakeholders’ motivations to pursue an alternative housing SoP,
barriers experiences, and
resources necessary to succeed in delivering their housing ambitions.
The actor-network mapping undertaken in Part Two: the Existing was also used to examine and compare the
Stakeholders are motivated by a general dissatisfaction with the choices provided by the existing multi-unit
SoP and the ongoing commodification of housing. Motivations are grouped into two main themes: financial
and ideological. Financially, stakeholders seek to either deliver housing at a lower price point, or to achieve
more desirable housing outcomes for equivalent cost. Ideological motivations include the pursuit of more
environmentally aware ways of living, a desire for a greater sense of community, or frustration with the existing
SoP in relation to realising ethical professional responsibilities. Ideological motivations inform expectations in
relation to design participation.
All stakeholders seek SoPs which enable greater design participation by future residents, although expectations
vary. Projects instigated by future residents seek the greatest degree of design involvement, commencing from
design briefing and including site selection. Those instigated by design professionals offer less design input and
do not engage with future residents until later in the development process. However, the ability for future
residents to become focal actants in the new SoP networks is seen to vary not with the process of instigation,
but with the location of risk and land ownership. It is concluded only future residents who collectively own the
land to be developed are guaranteed the power to approve design decisions (become OPPs). No cases increase
individualisation of dwellings beyond the existing multi-unit SoP.
Cases in which resident design agreement
is not essential for development to occur.
Projects able to default to existing SoP.
Cases requiring resident
‘approval’ of design to proceed.
Previous research by others has investigated barriers to infill housing development in Australia. This research
identifies barriers specific to implementing the innovative SoPs the cases propose. Barriers identified include:
Accessing knowledge: The bespoke nature of development means a process pioneered by one group to
accommodate their unique personal and locational circumstances is seldom replicable. This leaves each group
of innovators with the challenge of developing new processes and solutions (financial, legal, administrative)
relative to their own unique circumstances. This challenge is accentuated by a lack of project documentation
and information exchange across developments.
Time commitment: The uniqueness of the SoPs proposed can be exceptionally time-consuming and have
involved the pursuit of numerous ‘dead ends.’ The resulting lengthy process becomes a hurdle to progress.
Accessing professionals with appropriate experience and common interests: The proposed SoPs,
none of which include traditional developers, require a subtle reconfiguring of professional roles and
boundaries. To date, few industry professionals have experience in effectively negotiating such changes and
very few have experience in designing directly for (and with) a collective of multi-unit residents. Where
professionals hold the knowledge required to fulfil these altered roles they begin a brokerage role, providing
them significant power over design and livability outcomes.
Inflexibility of existing SoP: Each case has sought to reconfigure the existing, stabilised black-box of multi-
unit provision, which one interviewee suggests results from the vested interests in the current system which
“work against reforms”. Each case has experienced situations in which the existing SoP, and the processes it
has brought into being, have resisted the alternatives they propose. In particular, these relate to accessing
finances, accessing land, the existing relationships between risk, design and market values.
Many of these barriers are experienced differently by community instigated projects and project instigated
Comparing Innovators’ Actor-Networks
The thesis makes both visual and metric network observations, comparing the information flows between key
actants in each innovative SoP to identify variations. The comparison shows that each produces distinct
outcomes and opportunities for design. Most cases enable owner-occupiers to pass design information more
directly across existing subsystem boundaries. Most redefine the roles of the developer, the design team,
financial institutions, and future residents while engaging more human actants in the production subsystem.
All instigators are motivated to achieve improved multi-unit living environments for themselves or their clients,
and each disrupts the existing SoP to varying degrees. Community instigated projects involving a client group
of future residents in design are shown to facilitate more significant disruption than professionally-led projects.
However, the future residents’ capacity to collectively embrace and share financial risk is shown to have even
greater influence on a projects capacity to achieve its original ambitions. Similarly, those SoPs which remove
the developer (be they for-profit or not-for-profit) are more likely to achieve their desired housing outcomes.
Those SoPs which remove the developer are also shown to increase the influence of Strategic Urban Plans, a
desirable outcome of interrogating, rather than re-enrolling, the existing black-box.
Ego-network of Owner-Occupier
Integrated Developer SoP.
Ego-network of Owner-Occupier
Ego-network of Owner-Occupier
Ego-network of Urban Coup Inc as client group
Ego-network of Project Specific Development Collective as
client group and self-developer
The Nightingale Model
Ego-network of owner-occupier
Key Findings from Part Three: Innovative Cases & Constrained
This part of the thesis argues the capacity for network actants to alter the actor-network varies with their
motivations, their structural position in the existing actor-network, their ability to influence others, and their
capacity to access knowledge and resources. In concluding this part, Gulati and Srivastava’s (2014) framework
of constrained agency and network action is employed to identify the different types of network action being
undertaken by each of the innovative SoPs. Utilising this framework for comparison it is shown that innovative
SoPs utilising a combination of capability-based resources (technical knowledge) and symbolic resources
(cultural frames, world views and narratives) offer the greatest potential to successfully reconfigure the actor-
network of provision.
The greatest success in delivering built projects to date occurs when extensive network disruption is proposed
and additional actants become brokers of unique information essential to network operation. In the cases
observed here, this involved the proposition of an alternative finance model, not solely the proposition of an
alternative design solution.
Integrated Property Developer
Part Two previously concluded design outcomes of the existing multi-unit SoP are predominately influenced by
property developers, development profit, financial institutions, market value and tax legislation. Part Three
demonstrates that by reconfiguring the actor-network the influence of these actants can be reduced and that
of previously uninfluential actants increased, including designers, owner-occupiers, and the project design
brief. SoPs in which future residents act as a group or collective to define a project vision provide future
occupants with the strongest capacity to influence design outcomes, although this varies with ability and
willingness to accept risk.
The stabilised nature of the existing actor-network limits the capacity for actants to realise their ideal SoP and
requires, sometimes undesirable, compromises. The thesis outlines challenges faced by the Australian
innovators, including those pertinent at project-level:
realising financial savings,
balancing individual and collective decision-making,
balance design input and individualisation with costs and risks,
and those relevant to the strategic development of a viable alternative multi-unit housing SoP over time:
information dissemination, and
role of government and policy.
Part Four: the International Multi-Unit Innovators
The Australian multi-unit innovators all sought to increase the influence of owner-occupiers in the multi-unit
SoP. They demonstrated that reconfiguring the actor-network can improve owner-occupiers structural
position in multi-unit provision, but also identified multiple challenges exist to realising the SoPs they propose.
This part of the thesis examines two international CSO housing sectors, one relatively mature, and the other in
formation. It draws lessons from these international experiences, reflecting on how they have sought to
overcome project-level and strategic-level challenges.
Berlin Building Groups (BG)
The contemporary forms of building groups (BG) in Berlin build upon a long history of self-help and community
living, and now represent approximately ten per cent of all new housing construction, predominately in the
form of infill housing. BGs can be professionally-led or community-led and ‘develop without developers’ to
provide housing to future residents at cost.
UK Collective Custom Build (CCB)
Speculative housing development plays a more significant role in the UK than in Australia, with the vast majority
of all dwellings constructed by for-profit developers at all densities. Custom Build (CB) and Collective Custom
Build (CCB) housing is promoted by current UK planning and housing policies, with reference to examples from
the European Continent, including German BGs. The experiences of policies aiming to support CCBs in the UK
therefore offer this research an example of previous attempts to transfer lesson from BGs to another
In addition to literature review and document analysis, semi-structured interviews were conducted with key
stakeholders in BGs and CCBs in Berlin and the UK in November 2014.
Drawing lessons from international cases
This part of the thesis aims to address the challenges faced by Australian Innovators at both project and
strategic levels. It discusses past experiences in each of the international locations to draw lessons applicable
to the Australian context, it is notable that greater interconnections between project-level and strategic-level
actants exist among the Building Group (BG) and Collective Custom Build (CCB) stakeholders interviewed than
in the Australian cases. In both international locations, multiple actors actively engage with sector advocacy in
addition to direct participation in a particular project or projects. Stakeholders undertake strategic-level
advocacy in both professional and non-professional capacities, with the majority of interviewees indicating a
sense of social responsibility to assist in developing CSO housing beyond the projects that benefit them
The thesis details how the project-level challenges have been experienced in each location and draws
together learnings relevant to each. Here, the strategic-level lessons are summarised together with
recommendations for policy intervention in the Australian context.
Intermediary bodies have emerged in each location, however, their capacity to bring together top-down and
bottom-up actants varies. Initial intermediaries are the professional consultants and industry actants who
provide services and support to innovators. Berlin architects interviewed argue CSO housing sectors need to
maintain such bottom-up roots, acknowledging that dissemination of information by authoritative agencies
reduces insecurities about an innovative approach to housing, but suggesting “sharing a meal at a friend’s
[building-group] residence builds more trust.”
In the UK case, the initial bottom-up intermediaries have partnered with top-down policy makers to act as a
public access information repository and promote both CB and CCB. While this organisation has the capacity
to influence top-down policy makers, they are charged with advocating on behalf of multiple housing types and
tenures, not just urban infill CSOs. Housing types which experience fewer barriers to implementation have
Neither case provides Australia with a definitive solution to establishing an ideal network of intermediaries,
but together they have provided some cautionary lessons and show:
Intermediaries emerging from bottom-up network action are essential to CSO project development
and remain so throughout the process of sector maturation.
Early intermediaries established for sharing knowledge can become unnecessary during sector
development and hence emerging SoPs should not be dependent upon their ongoing presence.
The most desirable intermediaries are those which maintain an active connection with residents and
communities and build trust.
Top down intermediaries must engage with multiple and diverse solutions, not only those that are
market dominant, preferred by industry stakeholders, or politically motivated.
Role of Government and Policy
The literature and interviews from both Berlin and the UK concur that government policy in this area has two
1- to remove barriers to innovative self-formed, bottom-up groups building for themselves,
2- to promote and facilitate expansion of CSO housing as a trustworthy alternative to supply-led provision.
Experiences in Berlin also indicate it is critical to continue to support self-forming groups over time as ongoing
innovators, ensuring opportunity for the resultant housing to continue to evolve with occupant needs and
providing ongoing impetus for professional facilitators to continually seek improvement.
The current UK policy approach is one of top-down enablement of bottom-up development which will take
time to filter through the national planning system before its impact is measurable. Whilst stakeholders
support the policy intentions, they identified a disjunction between the stated intentions and the conservative
programs implemented. Catalyst funding programs introduced to provide access to land and finance were
viewed as overly bureaucratic, more suited to developers than community-based groups, and failing to address
the barriers experienced by stakeholders. To date, UK policies rely on existing stakeholders’ knowledge and
capital without making changes to the existing SoP sufficient to build trust and social capital.
A substantial difference between the two cases is the time at which government intervention or assistance has
occurred. The late-stage interventions in Berlin sought to support a CSO sector which had emerged and
developed its own, revolutionary SoP. In contrast, no such SoP yet exists in the UK and government
intervention there seeks to nurture one while avoiding risk exposure; an approach which reinforces the roles
of current dominant actants in multi-unit provision and fails to support innovative alternatives.
Together the international cases provide some cautionary lessons and show:
ideally, a combination of both hard and soft policy measures is required to address the range of barriers
experienced by stakeholders, with different needs existent at project and strategic-levels;
policy measures must recognise the interactions between barriers to CSO development, addressing
these from an integrated, network wide perspective;
it is essential to avoid reinforcing the dominance of existing actants and ensure future residents are
key actants in new SoPs;
government interventions must be more than modifications to the existing SoP, but avoid prescribing
a fixed, ideal alternative;
in encouraging alternative housing provision, policy should employ cautionary approaches to avoid
unintentional constraints; ideally achieving policy neutrality over time;
policies incentivising or encouraging CSO projects need to recognise the unique challenges faced in
inner urban areas to realise the strategic aims of consolidation.
Key Findings from Part Four: International Innovators
The international CSO housing cases reinforce previous observations, identifying the need to reconfigure, not
just alter, the existing black-box of speculative multi-unit provision to facilitate a viable CSO housing sector.
The case studies provide insights to establishing a viable CSO housing sector where none currently exists.
Network Action Resources
In the three case study locations of Australia, Berlin, and the United Kingdom, CSO instigators identify a number
of failures or controversies in the dominant multi-unit SoP. Failure to meet the living requirements of a portion
of the market, failure to deliver quality multi-unit homes at affordable prices, and more. Such failures result in
controversies between actants; between use value and market value, home and commodity, desire for the
“new forms of housing” and “new urban form” promised by strategic urban plans and the desire for profit.
Controversies have motivated actants in all locations to reopen the black-box; to alter the meaning of relations
with and between other actants, reorienting the actor-network to a different world view.
In both locations, those seeking multi-unit innovation have been more likely to achieve their objectives when
they possess (or are able to access) the capability-based resources to manoeuvre themselves into a network
position which increases their capacity to act. From such a position they become focal actants, enrol others in
their proposed alternative SoP, align interests, and enact translation (Callon 1986a).
The relatively mature Berlin CSO housing sector demonstrates the capacity for multiple actor-networks of
provision to co-exist. Berlin CSOs enrol professionals experienced in the existing SoP actor-network alongside
new actants; engaging stakeholders previously unassociated with (possibly failed by) speculative multi-unit
provision. As Hamiduddin & Gallent state: “[m]aximum benefit will be derived from this model if it is seen as
one route to delivering the homes that communities need amongst a jigsaw of alternatives” (2015, pp. 17-18).
In contrast, most CSO housing actor-networks in the UK enrol existing actants such as for-profit developers and
Housing Associations. Enrolled in both speculative and CSO housing actor-networks, these institutional actants
struggle to move between the different action frames and world views of the two networks. As such, CSO
housing groups find themselves competing with conflicting world views of powerful actants enrolled in their
own SoP. The Berlin CSOs have created a new actor-network of multi-unit provision which circumvents the
need to engage housing institutions or profit-seeking actants from the existing black-box.
UK CSO projects tend to enrol institutional actants due to an entrenched reluctance among existing
stakeholders to deviate from known risk profiles. In both UK and Australian projects, CSO groups willing to
assume development risk are often prevented from doing so by the constraints of the existing SoP. Sharam et
al. (2015) show reluctance on the part of Australian financial institutions to fund alternative development due
to associated business risk, with one lender stating:
If something goes wrong and it all gets totally stuffed up and we lose a million dollars
we don’t want to be on the front page of the [newspaper] suing couples because they’ve
tried to do the right thing and we’re the bad guys. (Sharam et al. 2015, p. 5)
The Berlin actants have succeeded in shifting both risk and risk perception. This reflects a revaluing of risk due
to the absence of profit-seeking in the CSO housing SoP.
It is advisable to exercise precautionary principles, to attempt to anticipate unintended consequences when
implementing change in complex systems and actor-networks. However, in UK and Australian CSO housing
excessive caution on the part of multiple actants is inhibiting progress. In both cases intervention from an
influential actant is required to break the cycle of risk reluctance and demonstrate CSO infill housing viability.
However, interventions in early phases of sector maturation can unintentionally limit future options by defining
parameters before alternative combinations of alternatives are explored. For example, the processes of group
formation, constitution, decision-making, financing, and ownership structures have incrementally developed
in Berlin over time as experience and knowledge have increased. The Berlin Building Groups are highly diverse,
meeting the needs and ambitions of households, and continue to evolve. The actor-network(s) which enable
their provision are sufficiently flexible to accommodate groups with differing objectives. The case studies
emphasise that there is no single or ideal CSO housing SoP and any land or finance interventions by government
should avoid prioritising one CSO solution over others. An SoP prioritised by legislation or eligibility restrictions
risks becoming locked-in, resisting controversy from rivals and generating an additional multi-unit black-box.
Collective Self Organised (CSO) housing seeks to disrupt the existing Australian multi-unit housing Structure of
Provision (SoP). For most participants, Australian and international, CSO housing seeks to address the
mismatch between speculative multi-unit housing design and their households’ needs or preferences.
Instigators proposing alternative SoPs must problematise (make controversy around) the existing black-box of
provision; modify existing actants or relations, recruit others to share their vision, and negotiate resistance
from existing network actants and relations.
Examining the existing Australian multi-unit SoP in comparison with innovative alternatives in Australia, Berlin,
and the UK, the thesis identifies multiple impediments to collective self-organised housing provision in
Australian urban consolidation. Collating the findings of the preceding thesis parts through an Actor-network
lens identifies five primary impediments:
Agencement and asymmetry
(mis)Alignment of interests
Mediation and participation
Controversy and competition
Current Australian housing policy is interested in liberal individualism and supporting capitalist development
(Burke & Hulse 2010; Jacobs 2015). The majority of policies informing housing outcomes are managerial,
targeted at ameliorating housing challenges both generated and problematised by existing, dominant housing
SoPs. Policy typically reinforces existing structures of provision in support of existing influential actants
economic interests through targeted subsidies and taxation interventions. The asymmetry of knowledge in the
existing actor-network provides key actants with the power to override the interests of others, to problematise
housing in line with their own interests, and offer certainty. In doing so they hold greater capacity to act, to
influence policy to maintain existing black-boxes.
The interests of Australian CSO housing innovators differ from those of key actants in the existing SoP,
prioritising housing and social benefits equally with economic benefits. Following K.Jacobs et al (2003), for CSO
advocates to progress their interests in alternative housing solutions it is necessary to construct a convincing
problematisation, coalesce support, and advocate the implementation of institutional measures for change.
However, while actants in all three case study locations would appreciate government assistance in negotiating
barriers to change, most express some reluctance toward government/state intervention which may bring
prescriptive solutions and compliance restrictions.
Strategic-level actants supporting alternative housing SoP are shown to have the dual roles of enabling
innovation and normalising alternatives. Australian CSO housing innovators are currently constrained by
existing actor-networks and the industry practices, policies, and legislation that have evolved to suit them.
Ideally, a multi-unit infill CSO housing system would operate in Australian in parallel with the dominant
speculative model, without preferential treatment provided by the state.
Initial encouragement is required to overcome existing impediments to CSO housing in Australia. At a
minimum, this requires recognising the existence of a non-speculative multi-unit housing sector through
appropriate legislation. Legislative definition would protect the emerging CSO housing sector from the
unintended consequences of future policies or legislation introduced in response to problematisations in the
speculative multi-unit SoP, and possibly exempt them from existing provisions which impede change.
Recognition would also avoid speculative developers co-opting future interventions designed to encourage a
CSO housing sector.
To enable ongoing innovation over time, SoPs should remain open and interactive, able to accommodate the
multiple variables identified in CSO housing sectors. The legislative differentiation of multiple SoPs would
enable implementation of targeted interventions, and concurs with a recent call for Australian “policy-makers
to tailor policies to a more diverse audience, and in doing so, improve future adaptive capacity” (Shearer et al.
2016, p. 16).
By identifying and addressing impediments to groups of households building for their own
use, the Great Australian Dream CAN be consolidated and the future visions of strategic plans
achieved. However, a reconfiguration of the actor-network of provision is required which
moves beyond existing stakeholders interests to both recognise future residents in provision
and provide them with the capacity to act.
This research contributes to the existing and ongoing multi-unit housing debate in Australia. It provides a
unique representation of the existing SoP and peers into the currently locked-in black-box of multi-unit
provision in Australia. Through comparison with actual and proposed alternatives it identifies opportunity for
future change to diversify housing outcomes and increase housing choice.
More specifically, it:
examines Australian alternatives to the existing multi-unit SoP(s) to explain why innovations sought by
instigators are not always achieved and identify barriers to change,
identifies a set of variables in international Collective Self Organised (CSO) housing which has the potential
to inform project specific planning across multiple locations, and to provide a basis for further research or
draws lessons from international examples to provide project-level and strategic-level insights to encourage
a collective self-organised multi-unit SoP in Australia, and
experiments with a unique means of utilising Actor-Network Theory (ANT) in system-embedded
comparative housing research, including the provision of seven factors to address in the design of ANT-
informed housing research and eight factors to address in visualising actor-networks of housing provision
with Social Network Analysis (SNA) software which can be adapted for use in other housing locations, types,
Self-organisation in the built environment is currently being researched internationally by scholars focusing on
public space, housing, urban design, and planning. For housing, self-organisation by civil society “blur[s]
traditional boundaries between housing production and consumption” (Stone 2015, p. 102).
Building specifically on the outcomes of this research project, further research is required to:
Test the working hypothesis generated from the case studies that CSO participants’ agencement in design
is directly linked to land ownership through the investigation of further international case studies.
Investigate the role of new professionals (agents/CSO developers/brokers) emerging internationally to
serve CSO housing groups, examining the diversity of services offered, location of risk, and perceived
effectiveness. This would inform the future development of such roles in Australia.vestigate the
professional roles of architects, project managers, and others in mature CSO housing sectors internationally
to determine how they differ from existing professional roles and, hence, what additional professional skills
are required to progress CSO housing in Australia.
Undertake research into unsuccessful CSO housing groups in Australia who have abandoned their collective
housing ambitions to determine if the reasons for their lack of success correspond with the challenges
identified by current instigators.
Re-examine these case studies, together with others as appropriate and over time, to determine which of
the CSO housing variables identified here have the greatest potential to enable design disruption, providing
target points for intervention in the actor-networks which leverage maximum gain.
Continue to develop and test the ANT/SNA mapping methods introduced here to analyse other housing
systems and other network intermediaries.
And, most important to the progress of collective self-organised multi-unit housing in Australia, commence
investigations to inform a proposed legislative definition of this housing sector, realising differentiation and
ANT/SNA Mapping Contribution
In combination with concepts from SNA and network analysis, ANT has provided this research with a
conceptualisation of housing as a heterogeneous socio-technical system. ANT’s capacity to combine with, and
advance, long-established theoretical constructs of housing is demonstrated. Ball observes “… contradictions
between the spheres of consumption, exchange and production [are] important causes of change in structures
of housing provision” (Ball 1986, p. 162). Viewed from an ANT perspective, such contradictions constitute
controversies, triggering problematisation and translation by powerful (focal) actants. The understanding of
change in actor-networks via translation provides a means of progressing Ball’s Structures of Housing Provision
beyond the static description of existing practices, or black-boxes, to the comparison of alternative futures.
Housing researchers have previously proposed the use of network analysis to identify key players in production
(Nicol 2013) and consumption (Heitel et al. 2015) subsystems, and to compare typical and alternative cases
(Nicol 2013). However, no completed precedent exists as guidance. The mapping provided the research with
five unique observations:
1- it enabled the combination of an extensive literature on Australian multi-unit housing provision and design into
a single, visual representation of design information flows.
2- using ego-networks, it highlighted the different views held by different actants and how these influence their
capacity to act. In particular, the mismatch between ones sphere of concern and ones sphere of influence.
3- both visual and metric analysis identified key actants in the SoPs, with SNA literature providing an
understanding of properties afforded to actants by their structural position.
4- mapping within the context of the Australian subsystems of housing provision, provided by Burke and
colleagues, ensured the analysis remained focussed on the multiple subsystems of provision rather than
focusing on the resultant architectural artefact, as is often the tendency in architectural research.
5- the mapping interacted successfully with primary data collection, both informing interview questions and
providing structural network explanations of interviewees’ observations.
In comparing alternative or proposed SoPs the ANT/SNA mapping provided analytical insights unavailable via
other means. Mapping multiple alternative SoPs, the researcher can identify actants (and their relations) which
remain unaltered, identifying the human and non-human actants which act to restrict network change. Using
SNA metrics allows alternative networks to be compared, identifying network changes more likely to achieve
desired outcomes or address identified challenges. The combination of actor-network mapping and SNA tools
has identifying opportunities for network analysis and intervention which are of value not only to multi-unit
housing provision, but also to any other design arena which requires the un-locking of a stabilised ‘black-box’.
Ache, P & Fedrowitz, M (2012), 'The Development of Co-Housing Initiatives in Germany', Built Environment, vol. 38, no. 3,
Altmann, E (2015), 'Policy Implications for Governing Australia's Apartment Communities: Tenants, committees of
management and strata managers ', in R. Dufty-Jones & D. Rogers (eds), Housing in 21st-Century Australia: People,
practices and policies, Ashgate, Surry, England, pp. 121-136.
Ball, M (1986a), 'The Built Environment and the Urban Question', Environment and Planning D: Society and Space, vol. 4,
Ball, M & Harloe, M (1992), 'Rhetorical Barriers to Understanding Housing Provision: What the "Provision Thesis" is and is
not', Housing Studies, vol. 7, no. 1, pp. 3-15.
Brunoro, S (2013), 'The Collective Self-Organized Housing Approach: Improving the quality of life towards nearly zero
energy strategies', in Sustainable Building 13 Portugal, Guimaraes, Portugal.
Burke, T (2012), 'The Australian Residential Housing Market: Institutions and actors', in R. Tomlinson (ed.), Australia's
Unintended Cities, CSIRO Publishing, Collingwood, pp. 35-49.
Burke, T & Hayward, D (2000), Housing Past. Housing Futures., Technical Report 4, Planning Melbourne for the 21st
Century, Swinburne Institute for Social Research, Melbourne.
Burke, T & Hulse, K (2010), 'The Institutional Structure of Housing and the Sub-prime Crisis: An Australian case study',
Housing Studies, vol. 25, no. 6, pp. 821-838.
City of Melbourne (2013), Understanding the Quality of Housing Design, City of Melbourne, Melbourne.
Easthope, H & Randolph, B (2016), 'Principal–agent Problems in Multi-unit Developments: The impact of developer
actions on the on-going management of strata titled properties', Environment and Planning A, May 19 2016, viewed May
19 2016, DOI 10.1177/0308518x16650894,
Fischer, H & Ayturk, DG (2011), Residential Density (Liveable Communities) Market Perceptions, Final Report, 5 August
2011, Adelaide Thinkers in Residence Program, Adelaide.
Government of South Australia (2010), The 30-Year Plan for Greater Adelaide: A volume of the South Australian planning
strategy, Department of Planning and Local Government, Adelaide.
Gulati, R & Srivastava, SB (2014), 'Bringing Agency Back into Network Research: Constrained agency and network action',
Contemporary Perspectives on Organizational Social Networks, vol. 40, pp. 73-93
Hamiduddin, I & Gallent, N (2015), 'Self-build Communities: The rationale and experiences of group-build (baugruppen)
housing development in Germany', Housing Studies, 9 October 2015, pp. 1-19.
Heitel, S, Kämpf-Dern, A & Pfnür, A (2015), 'Integration of Stakeholder Interests in Housing Companies’ Strategic
Management', Property Management, vol. 33, no. 3, pp. 224-244.
Jarvis, H (2015), 'Community-led Housing and 'Slow' Opposition to Corporate Development: Citizen participation as
common ground?', Geography Compass, vol. 9, no. 4, pp. 202-213.
Junge, C (2006), 'Developing without Developer: Practice and potential of building cooperatives in urban development -
the German case', London School of Economics and Political Science, MSc Regional and Urban Planning Programme
Kelly, J-F, Weidmann, B & Walsh, M (2011), The Housing We'd Choose, Grattan Institute, Melbourne.
Martel, A, Whitzman, C, Fincher, R, Woodcock, I & Tucker, D (2013), 'Getting to Yes: Overcoming barriers to affordable
family friendly housing in inner Melbourne', in K. Ruming, B. Randolph & N. Gurran (eds), State of Australian Cities, State
of Australian Cities Research Network, Sydney, pp. 1-11.
Millington, AF (2000), Property Development, Estates Gazette, London.
Nicol, LA (2013), Sustainable Collective Housing: Policy and practice for multi-family dwellings, Routledge, Oxon.
Randolph, B & Tice, A (2013), 'Who Lives in Higher Density Housing? A study of spatially discontinuous housing sub-
markets in Sydney and Melbourne', Urban Studies, vol. 50, no. 13, pp. 2661-2681.
Sharam, A, Bryant, L & Alves, T (2015), 'Identifying the Financial Barriers to Deliberative, Affordable Apartment
Development in Australia', International Journal of Housing Markets and Analysis, vol. 8, no. 4, pp. 471-483.
Shearer, H, Coiacetto, E, Dodson, J & Taygfeld, P (2016), 'How the structure of the Australian housing development
industry influences climate change adaptation', Housing Studies, 17 March 2016, pp. 1-20, viewed 20 April 2016, DOI
State of Western Australia (2013), The Housing We'd Choose: A study for Perth and Peel, Department of Housing, Perth.
Stone, W (2015), 'Reshaping Housing Consumption and Production from the Bottom Up: Insights from interpretivist
housing research', in R. Dufty-Jones & D. Rogers (eds), Housing in 21st-Century Australia: People, practices and policies,
Ashgate, Surry, pp. 89-104.
Thomas, G (2011), How to do your Case Study: A guide for students and researchers, Sage, London.
Tummers, L (2011), 'Intentional Communities: Methods for reviewing the rise of citizens' housing initiatives in a European
perspective', paper presented at European Network of Housing Researchers Conference, Toulouse, France, 5-8 July 2011.
Turner, JFC (1976), Housing by People: Towards autonomy in building environments, Marion Boyars, London.