Models of disability support governance: a framework for assessing and reforming social policy.

Henman, Paul ; Foster, Michele

The problematic of public governance has occupied governments and academia for the last few decades, resulting in an immense and proliferating literature (Bell & Hindmoor 2009; Bevir 2011; 2007; Kjaer 2004; Newman 2005; Rhodes 1997). At the heart of this problematic is concern about the proper role of state and non-state actors in financing and delivering services in such a way that balances efficiency, effectiveness, equity, accountability, responsiveness, and democratic agendas. Three ideal typical modes of governance are widely delineated: state or bureaucratic governance; governance through markets; and governance through networks or partnerships. While each mode has various strengths, they also have their weaknesses and occasions of governance failure (Jessop 1998).

This intense examination of governance arises out of two evolving reform agendas. First, New Public Management (NPM) involves the installation of business approaches to public management, market models of service delivery, and management of outcomes and performance measurement (Lane 2000; McLaughlin et al. 2002; Pollitt 1993). Secondly, the citizen or service user rights movement has sought to reshape policy and service delivery around service users--not institutions--who are conceptualised as active individuals with rights, ideally resulting in more responsive, holistic and personalised service delivery (Clarke & Newman 2007; Needham 2011a; Yeatman et al. 2008). There are also a range of post-NPM public sector governance models, including neo-Weberian State, New Public Governance (Pollitt & Bouckaert 2011), and New Public Service (Denhardt & Denhardt 2007). The search for the optimal organising model for public services as supplied by these abstract models sits within specific policy, institutional, constitutional, business and non-government organisation (NGO) settings.

The disability domain has been especially challenging for governments seeking to identify a governance model that provides a cohesive, comprehensive and sustainable approach to financing and delivering services. The history of disability services is peppered with debates about the provision of disability services with explicit and implicit references to governance models. These include the campaign of several decades ago to end institutionalisation of people with disability and mental illness (MacKinnon & Coleborne 2003; Brunton 2003; Young & Ashman 2004; Emerson 2004; Mansell & Ericsson 1996), to more recent calls for direct payments for individuals requiring disability services (Blyth & Garnder 2007; Glasby et al. 2009; Lord & Hutchison 2010; Priestley et al. 2007; Prideauz et al. 2009; Glendinning et al. 2008), and articulations of co-production of services (Wilson 1994; Hunter & Ritchie 2007). Despite vigorous reform, people with disability and their families continue to experience poor access to services, and much poorer outcomes than the wider community in which they reside (WHO 2011), which is suggestive of continuing governance failure.

Australia's experience in seeking a governance model for disability is similar to the rest of the developed world. In the context of 'underfunded, unfair, fragmented, and inefficient' state-based systems (Productivity Commission 2011: 2), Australia's recent reform of a National Disability Insurance Scheme (NDIS1) embodies a national governance and funding framework. Apart from new and enhanced approaches to financing, allocation of disability support for people with high care needs is to be nationally coordinated, purportedly involving more integrated, holistic and individualised support through market-based solutions (Madden et al. 2014; Productivity Commission 2011). NDIS-financed disability services are also expected to be coordinated with mainstream services and other informal supports.

Given these recent, rapid and complex changes in disability services and the need to examine and evaluate such reforms, there is value in developing a conceptual and analytical framework to identify, delineate and systematically assess different models of governance for disability support. In order to provide conceptual and analytical insight into the operation of governance and the interaction of service users, service providers and funders, such a framework needs to be applicable at a lower level of disability service governance than the macro articulation of state, market and network governance. This paper offers such a conceptual framework. Accordingly, there are two main aims of this paper.

The first aim is to conceptualise, articulate and delineate different ideal types of disability support governance. In this paper we present four models of disability support governance: uncoordinated; caseworker coordinated; dwelling-based; and user-coordinated. Just as the three macro governance models--of state, market and network--have strengths and weaknesses and points of governance failure (Jessop 1998), the articulation of these four governance models of disability support enables the relative strengths and weaknesses of these models to be discerned. Such recognition has important applied policy and service delivery implications. This conceptual work is thus directed towards analytical and methodological innovation. Analytically, these models enable empirical data to be analysed and categorised according to the models, thereby assisting in the evaluation of the likely strengths, weaknesses and points of failure that may arise from a specific policy setting or service delivery arrangement.

The second aim is to present an approach to visualising the governance of disability support that can be applied to both the ideal types and empirical data on actual disability service arrangements. Using network diagrams created by NodeXL (Hansen et al. 2010), this visualisation method enables different disability support arrangements to be compared, contrasted and categorised, especially the application of social network analysis techniques. Visualisation of data is fast becoming an important new tool in analysing social networks (Ackland 2013). The complexity and diversity of disability support networks can make them difficult to comprehend, compare and analyse. By making relationships visual, relationships, patterns of relationships, and structural relationships can be more readily discerned.

Taken together, these conceptual and methodological developments provide a framework to navigate the empirically messy, often highly complex and confusing reality of disability support. Conceptual and analytical tools enable the empirical realities to be mapped and compared, patterns identified, and outcomes assessed. Thus this paper provides the intellectual infrastructure to analyse policy and service delivery arrangements systematically in real-life situations in order to assess the relative points of success for different types of disabilities and contexts.

The paper is structured in three parts. The first section outlines the rationale and process by which the distinct models of disability support governance for people with high care needs were identified. The second section--the body of the paper--progressively presents the four ideal typical models of disability support governance: uncoordinated; caseworker coordinated; dwelling-based; and user-coordinated. For each model, the relative strengths, weaknesses, and occasions for governance failure are discussed and illustrated with empirical data from research conducted by the authors. The concluding third section reflects on how the models of disability support governance and their visualisation can further policy research, evaluation and development in both national and international contexts.

Conceptualising models of disability support governance

The four models of disability support governance described in this paper, their characteristic features and their strengths and weaknesses, resulted from a three-staged conceptual and empirical development process. To begin with, four conceptual models and an initial set of characteristics were discerned from a broad review of the literature on disability services. To identify and delineate different governance models, the focal organising principle used was the locus of control, that is, where coordination of services, personnel and organisations occurs. The rationale for this focus was our goal to understand the operation of care and support services from the point of view of the recipient of care and support.

Consideration of the various care and support services that people with a disability may need, the individuals and organisations that provide such services, and the flows of financial resources that enable such services to occur, are critically important, but are secondary to their governance, which is denoted by the control and coordination of those services. In this regard, we depart from previous work by Breda and colleagues (2006), who previously delineated three 'models of long-term care for disabled people': individual; the family informally supporting the individual; and the organisational provider providing formal care. Whilst there are some similarities with our approach, these models are delineated by where the focus of the services lies, rather than the locus of coordination and control.

Stage two involved identifying the relative strengths, weaknesses and points of failure of each model by both directed reading of published research on disability reforms that empirically assess how these models function in practice, and empirical research conducted by the authors on the management and financing of lifetime care and support (Foster et al 2012). That project examined the funding, provision and coordination of lifetime care from the perspectives of service users, family and care service providers. It included case studies of 25 adults with high care and support needs as a result of an acquired brain injury, spinal cord injury, Huntington's disease, Multiple Sclerosis, or Motor Neuron Disease. Each case study involved separate interviews with the adult receiving care services (25), a nominated family or friend (22), and a care service provider (18). In addition, each individual receiving care and their family member were invited to complete a slightly modified Services Obstacles Scale (SOS). (2) The scale asks respondents to assess their experience against a range of typical service problems according to a seven-point scale.

In the third stage, a visualisation methodology of disability governance models was developed, road-tested and refined in the process of analysing the empirical case studies from our research. The visualisation technique involves the construction of a directed network between 'nodes' that represent an individual, organisation, or service type (see Figure 1 below for an illustration). Individuals are depicted as circles, and organisations and service types are depicted as squares. In each of our empirical case studies we identified up to six different services: Personal Care (PC); Domestic Help (DH); Respite services (Res); Accommodation or housing (Acc); Income ($); and Aids and Equipment (Aids). Additional or alternative service types could be used in subsequent research. Connections between these nodes were by a directed edge, represented by an arrow to indicate the direction of control in that service relationship. For example, if an individual identifies and purchases a wheelchair, then the arrow would point outwards from the individual to the wheelchair ('Aids'). Alongside each arrow we identify the source of funding for the service indicated by that arrow.

It is acknowledged that the real world of service delivery is far more complicated than these categories depict, and the direction of control in a service provider-user relationship is rarely uni-directional. For example, the person requiring and purchasing a wheelchair is likely to have input from an occupational therapist - who could be denoted in the visualisation--and there may be considerable diversity of 'Aids', their providers and their financiers. Similarly, informal care relationships between spouses, or parent-child situations are much more negotiated. Such relationships could be depicted by a two-directional line. However, to provide analytical strength an assessment was made about where the dominant control lies in a care provision relationship, and that was the direction depicted.

This three-stage process generated four ideal models of disability support governance: uncoordinated; case manager coordinated; dwelling coordinated; and user coordinated. Table 1 summarises the characteristics of each governance model, which are described in more detail below. These models are ideal types delineating the different modes of thinking in the literature and policy, and accordingly enable comparative analysis. Each governance model is illustrated with a real case study of an adult individual with a high level of long-term care and support purposively selected from the Lifetime Care project conducted. These empirical case studies, which use pseudonyms, illustrate the key features of each conceptual model, their relative strengths and weaknesses. They also record the great complexity of lived experience of people with a disability, and demonstrate how this empirical complexity can be captured with our visualisation method.

Four models of disability support governance

Uncoordinated governance model

The uncoordinated model of disability support governance is characterised by an individual receiving a mix of services through various channels that they are deemed to need--and to have eligibility--to receive. This is diagrammatically illustrated in the left-hand side of Figure 1, where the individual is the central circular node 'IND'. As explained above, the square nodes surrounding the individual represent the various service organisations providing distinct services, namely personal care (PC), domestic help (DH), respite (Res), accommodation (Acc), disability-related aids and equipment (Aids) and income ($). Other services could be included, such as transportation, rehabilitation and therapy, information and advocacy, and community engagement.

The uncoordinated model of governance is characterised by individuals having services done to them, that is, they are passive recipients of services which have been deemed appropriate and managed for them by others (as illustrated by the ingoing arrows), and the notion that different organizations provide different types of services funded by different funding streams. However, the services are managed and controlled discretely by these organizations. This is typical of the block funding model, whereby organisations are provided with an amount of funding to provide specific services to a number of people with disability (Chenoweth & Clements 2009). It is a default model in that there is no one point of contact and coordination and therefore an absence of overall service coordination. The key strengths of this model are that the individual is the central focus of the services. The weaknesses of the model are that the individual has services done to them, with little autonomy, and that services are fragmented and uncoordinated. Accountability to both individual and government is diffuse and potentially problematic. Governance failure is manifest when the services fail to address the person's needs and can arise as a result of several interrelated factors. It can occur as a result of lack of financing of services, the segmented or complex service arrangement that can create gaps in addressing service needs, poor coordination of services, services driven by organisational rather than personal needs, or no autonomy to shape care for their own needs and life objectives.

[FIGURE 1 OMITTED]

The realities of this governance model can be depicted by an empirical case study (Figure 1, right-hand side). 'Anna' is a 47 year old single woman with Huntington's disease who has four children. Anna lives at home in public housing with her youngest child of 17 years who assists her mother constantly. As illustrated in the Figure, Anna has four carers; two carers are coordinated by a non-government organisation (square labelled PC-DH1) that provides both personal care and domestic help. These services are funded by the organisation and the state disability agency (as demonstrated by the tag 'NGO 1-DS' on the inwards arrow). A second organisation comes three times per week to assist with showering and one morning a week to provide domestic assistance (Carer 3). Anna's daughter is her fourth carer (Carer 4), and a third NGO provides respite for her daughter (Res). Anna's income derives from the means-tested Australian government Disability Support Pension, and her aids and equipment have been initiated by herself and variously funded by herself, the Queensland Department of Housing, and the government Medical Aids Subsidy Scheme (MASS).

In this example, Anna finds the NGO 1 carer coordinator 'helpful.' She is especially satisfied with her main carer:

If there is anything going on, she will fix it. She's good (Anna).

This carer decides for Anna because I can't make up my mind anymore (Anna).

However, the care is uncoordinated and involves many people. Anna's experience is there are 'lots of different people' coming to her home and it 'gets confusing' and Anna feels 'terrible.' The care and support arrangement in this individual example is seemingly precarious, as depicted by Anna's expression 'I'd be buggered' without her main carer or her daughter. She perceives that there are insufficient care hours from NGO 1 for cooking and washing, due to insufficient funds:

It's a big problem (Anna).

Anna does not get formal help on weekends, but relies on her daughter. Anna also lacks the financial resources to pay personally for medical rehabilitation and disability related services, despite the perceived need for physiotherapy for balance and back pain issues. In responding to the Services Obstacles Scale,

Anna reported agreement with the statements 'I am dissatisfied with the amount of professional help and services being provided'; 'Transportation is a major obstacle toward getting enough help', and strong agreement with the statement 'I don't know if there are good care and support resources in my community'. These responses can be interpreted as arising from the complexity and lack of autonomy of her situation. Anna did, however, express confidence in the quality of care she received.

Anna's case study illustrates a number of characteristics of the uncoordinated governance model of disability services. The complexity of multiple people and funding sources is problematic, confusing, and generates dissatisfaction. Because care and support is uncoordinated it fails to maximise the effects of the services to the benefit of the individual, and there is little autonomy to improve this. Funding clearly limits what level of service is achievable. As a result of multiple funding sources and lack of autonomy, the individual has very little capacity to reduce the precarious nature of her support infrastructure.

Caseworker governance model

One policy and service delivery response to the lack of coordination is the appointment of caseworkers to manage and coordinate an individual's case and support needs. The caseworker governance model is illustrated in multi-agency work (Abbott et al. 2005) and is depicted in the left-hand side of Figure 2. The arrows directed outwards from the caseworker (CW) to the care and support services indicate that the caseworker is initiating and coordinating them on behalf of the individual who receives them (arrow pointing to IND). This model is evident in different funding arrangements. Services can be financed through diverse funding streams for separate service types, as with the uncoordinated model. Alternatively, funding can be directly provided to the caseworker with which to 'purchase' services on behalf of the individual. Individual care packages or individual budgets managed by a caseworker are examples of this model (Lord & Hutchison 2010), as are disability insurance models whereby compensation funding is provided to a caseworker. A clear strength of this model is that the services are actively coordinated by someone with knowledge of the individual's needs and the care and support service sector, who can act as an advocate for the individual (Abbott et al. 2005; Challis et al. 2002). A key feature of this design is that it is more likely to be person-centred or individualised, with more holistic service provision. Such coordination is also likely to avoid the stress on the individual concerning the financing, organisation and adequacy of care. It also provides a locus of accountability to both the individual and funding bodies for the quality and cost-effectiveness of care. However, the quality of the services is partly based on levels of funding, the quality and capabilities of the caseworker, and the range of service options (Spall et al. 2005). Moreover, if the caseworker acts as a gatekeeper their role can undermine individual choice, autonomy and control. These are the situations that can lead to the failure of this governance model (see Table 1).

[FIGURE 2 OMITTED]

The right-hand side of Figure 2 illustrates the caseworker-coordinated model with the case study of 'Beth', aged 47 with a spinal cord injury. Beth lives in public housing, with care and support coordinated by a dedicated spinal cord injury NGO. Individualised funding for Beth is provided to and held by the organisation under a special spinal cord initiative of the state government. As illustrated in Figure 2, the NGO coordinates four rotating carers who come three times per day, for up to eight and a half hours at a time, and personal carers are arranged for holidays. Psychological counselling is also accessed; however, funding for this service is presumably capped as Beth is unable to access psychology services on an ongoing basis. Although Beth perceives a need for more therapeutic services, including psychology, she is not able to afford private physiotherapy, psychology and psychiatry services. By contrast, Beth's air-conditioner was funded from unused carer hours for one year, and an emergency grant of $2000 towards her fully adjustable bed was funded by the NGO. The government's MASS program and Beth personally jointly financed a power-assist wheelchair. Beth's personal income comes from a partial government Disability Support Pension and from private superannuation.

The caseworker clearly plays an important role in the providing Beth's services, but Beth is also active in her own care governance. Beth and her caseworker negotiate hard for what Beth regards as necessary and that she 'deserves'. She interviews and selects her carers, although she perceives the quality of staff as not high, as it

isn't the top shelf job that you are getting (Beth)

The overall level of control was deemed to be located in the caseworker, as depicted by the arrow direction in Figure 2. While Beth is satisfied with the level of personal care and domestic support she receives, she dislikes the intrusion of carers--'it drives me insane'--and is not satisfied with having carers in the house as she feels a lack of control and 'railroaded' by some of them. High staff turnover of care workers is also an area of dissatisfaction. While Beth has an assessed package of care that is managed by her caseworker, the NGO has discretion over the expenditure of unused care hours at the end of the year. Interestingly, while personal care hours allocated to Beth might be unused at the end of the year, these are not automatically rolled over to her. Rather, this is decided by the organisation in consultation with their government funding agency. This is perceived as unfair by Beth:

If they are your hours, you should be allowed to have access to them for equipment and things that you may need (Beth).

In her responses to the Services Obstacles Scale, Beth strongly agreed that 'lack of money' for services is a major problem, and agreed that there were

very few [support] resources in [her] community (Beth).

However, transportation to access help was perceived to be satisfactory.

This case illustrates the importance of an active, capable and collaborative caseworker in making this model of governance operate effectively. Yet it also illustrates the significance of the role of caseworkers within organisations and how they can use their position to shift resources between several clients to achieve what they perceive as fair and equitable overall, even if not in the best interests of a specific individual. The case highlights an increasingly central issue of the balance of advocacy and brokerage in care coordination with gatekeeping (Chester et al. 2010). Such circumstances demonstrate that adequate resourcing is essential for any governance model of care and support services to operate successfully.

Dwelling-based governance model

The provision of care and support in specialised community based facilities demarcates the third governance model of care and support (see left-hand side of Figure 3). In many respects this model mirrors that of the caseworker governance model. However, in this model the responsibility for coordinating services lies with the dwelling in which the individual resides, not with a caseworker. This is illustrated by the arrows directed outwards from 'Acc' to the various care and support services and to the individual (IND). Empirically, the dwelling-based governance model occurs in traditional residential support for people with disabilities, whereby the institution provides all support services (French 1994; King et al. 1971), and also in more recent community based supported accommodation (Walsh et al. 2010). Funding in this model varies from a single stream dominant in traditional residential support, to multiple streams which would be pooled into the dwelling, either for each individual, or collectively for all residents in the facility. As with the caseworker model, strengths of this model include coordination of services on behalf of the individual, thereby reducing stress and worry, and a clear locale of accountability. Another benefit of this model is that it enables economies of scale through the coordination and provision of services to multiple people residing in the one dwelling. (4) While service coordination can be person-focused, a possible weakness is that this person-focus is balanced by the overall dwelling needs and those of other residents. In this respect, this governance model may involve reduced personal choice and autonomy, with the dwelling acting as a gatekeeper, and the capacity to change dwellings is typically difficult and onerous. The model is likely to fail when the dwelling and its management is of poor quality and resourcing is inadequate (see Table 1).

[FIGURE 3 OMITTED]

The right-hand side of Figure 3 illustrates the dwelling governance model with the case study of a 38 year old mother of four children with an acquired brain injury. 'Cath' has 24-hour care in a house with two other women and three support workers, as depicted in the top left-hand part of Figure 3. She lives close to her husband and children, where she returns 'home' on weekends, and where care is provided by her husband. While her dwelling coordinates her main care and support, Cath has several agencies involved in her care, as depicted in Figure 3 by the number of services for psychology and psychiatry (Psych), two sources for aids and equipment (Aids & Aids2), and physiotherapy (Physio). Importantly, there is referral and liaison between some of these varied organisations to reduce the problems associated with uncoordinated service delivery evident in the uncoordinated governance model. Cath is generally satisfied with her accommodation, which makes her

feel pretty comfortable (Cath).

and allows her choice about cooking meals and inviting visitors.

They make you feel like you're in your own home and you're a normal person (Cath).

However, Cath also perceives a lack of choice about carers and dislikes having a male carer:

I have to go to the bathroom even though I feel very uncomfortable because he is a male. And showering, I don't have to shower when he is on (Cath).

Overall, Cath is satisfied with the quality of her current care. Although she would prefer to reside permanently at home, this choice is not possible due to resources and family relationships. Financial resources are also a difficulty. Cath's husband has insufficient funds to manage the household and the additional expenses incurred to provide for her. As a result, Cath is now in receipt of the government Disability Support Pension, which she finds inadequate. This is reinforced in her strong agreement to the statement 'Lack of money to pay for medical, rehabilitation and disability related services is a major problem' on the Service Obstacles Scale. Cath also perceives that there were 'very few resources in the community' for her problems.

This case study illustrates some of the strengths and weaknesses of the dwelling-based governance model. The coordination of care and support through the organisation means that Cath has few concerns. Indeed, the quality of the facilities and how the dwelling is run make it a satisfying experience. However, Cath perceives a lack of control in not being able to influence or dertermine which carers are appointed to her care, specifically in relation to having a male carer on some shifts, and also in not being able to determine where she lives. The issue of finances was also raised as a critical element in ensuring success.

User-coordinated governance model

In response to the disability rights movements (Shakespeare 2006) and ideas about active users of public services (Hunter & Ritchie 2007), a fourth model for disability governance has been widely promoted, namely, services being defined and managed by the person receiving care. The user-coordinated model is illustrated in the left-hand side of Figure 4 with the arrows of coordination moving outwards from the individual, thus reversing the uncoordinated governance model. Alternatively, a family member can act on behalf of the individual--where this occurs the user-coordinated model begins to blur with the caseworker governance model. The rationale and strength of this model is that the individual has control and autonomy over their care and support needs because s/he is regarded as the one who knows best about their needs and preferences. As a consequence, they can shape their services, their service providers, and their service scheduling to reflect their personal preferences and aspirations. To enable the user-coordination model, funding is ideally provided to the individual to cover a range of services. This is evidenced in policies of direct payments in England (Blyth & Garnder 2007; Glasby et al. 2009; Priestley et al. 2007) and Australia (Fisher et al. 2010), self-directed care in England (Prideauz et al. 2009), and the planned NDIS in Australia (Productivity Commission 2011). Alternatively, an individual budget may be provided for the individual, though managed by a third party, but the individual retains control over its spending, perhaps within dictated guidelines, such as in the model of English personalised packages of care (Needham 2011b; Leece 2007; 2004).

[FIGURE 4 OMITTED]

Another outcome of this governance model is that it promotes a market for care and support, which in idealised market conditions promotes service innovation, responsiveness and choice (Glendinning 2008). In the alternative, markets can fail to develop or provide services, especially where profits are not viable (Greener 2008). Such market failure is particularly pertinent for people living in regional areas where a threshold of services users is unlikely to be reached. Coordination challenges may also lead to difficulties. While the allocation of management control to the individual enables autonomy, it also means that the individual assumes responsibility for establishing the care arrangements and funding usage.

Individual accountability for finances and quality of care can be an unwelcome burden. Evaluation of the English direct payment scheme has revealed that the successful acquisition, coordination and management of an appropriate direct care package depends substantially on the availability and capacity of the individual and their informal support networks (Blyth & Gardner 2007). If inadequate resourcing is provided to the individual, this model can individualise systemic failure and lead to blaming the individual for their situation. In short, governance failure in this model will result from market failure in care services, an inability of the individual to manage their funds or to operate as a rational actor, perhaps due to a lack of awareness of available services from which to choose, and indeed a lack of funding with which to purchase adequate care (Table 1).

The user-coordinated model of disability services governance is illustrated by the case of 'Dave' a 57 year old male, diagnosed with motor neuron disease three years previously (see Figure 4, right-hand side). Dave is divorced and has a daughter who contributes to his personal and domestic care (Carer1). He lives in his own home with a live-in 24/7 carer who provides the majority of his personal care and support needs (Carer 2) and receives personal care for showering three days a week from an NGO financed by a government personal care package (Carer 3). He personally employs Carer 2 directly from Thailand. He pays for Carer 2's return flight to Australia, wages, board and lodging for three months, then flies Carer 2 home again, then employs a new Thai Carer 2. Dave specifically chose these care arrangements after finding the options available on the Australian market inappropriate for his needs. His previous experience of carers provided by an NGO was that care was not provided at a time he desired:

I like to have a shower before about ten every day.... I don't really want to sit, especially in summer, until three o'clock for someone to come and shower me that day, and then seven o'clock one morning, and then half past four the next day. (Dave).

After deciding he wanted a live-in carer, Dave found that the three carers required on a daily basis to meet his needs was an expensive option and ultimately there was

just the lack of available services and the willingness of people to come in (Dave).

Self-funding care provides Dave with flexibility to meet his own care needs, as well as calmness and companionship. He is in control. This is also reflected in his decision to purchase an electric wheelchair rather than borrow from an NGO:

we chose for convenience to buy ... I probably have about five thousand dollars' worth of equipment I purchased myself, only for the ease of picking something that was just right (Dave).

However, this would not be possible without significant financial resources. Dave purchases his services with a disability pension from private insurance, savings obtained by selling his business, and remortgaging his house, which he describes as

eat[ing] my house (Dave).

This particular case study demonstrates clearly that the user-coordinated governance model of care and support provides good outcomes for people who have the capacity to organise, direct and finance their service needs. They are in control and determine what they want, when and how. However, the example also illustrates the possibility that the market does not necessarily provide these services, in that Dave has decided to import carers, raising questions for employment and immigration policies.

Conclusion: Deploying governance models as analytical tools

Changing modes and practices of financing, delivering and coordinating disability services sits within a wider reform of public service governance. While macro governance models of hierarchy, market and network contribute to disability service reform, this paper has posited that more specific, meso-level models of governance provide greater utility in designing, applying and assessing disability service settings. Accordingly, this paper presented four ideal typical models of governance using visual illustrations based on empirical examples. These ideal types are broadly representative rather than prescriptive of different disability governance models. We argue that the conceptual and methodological approach to governance developed herein provides a way of describing, exploring and assessing the complex nature of governance of disability services and, more specifically, how service user, service provider, and funding interact and coordinate in practice. To that end, the articulation and visualisation of the above conceptual models of disability support governance is designed as an analytical resource and a basis for further development. This concluding section clarifies the objective of this model development and signals possible directions for the use of this work for policy and empirical analysis.

Notably, the four ideal typical models of governance proposed here comprise a useful analytical tool with which to cut through the complexity and messiness of disability governance in the real world--a messiness that often challenges systematic analysis and evaluation. In particular, the models act as ideal types to highlight significant differences and their relative merits. Real situations can easily involve elements from more than one of the four governance models. This is not necessarily a weakness of the models, as they can aid analysis of an actual situation to identify potential challenges and occasions when governance failure may occur. Moreover, the models assist policy researchers and developers to avoid the hubris of seeking a single perfect governance model. All approaches have pitfalls. Rather, it is about balancing competing elements. Indeed, Jessop (1998) acknowledges as much. Governance models involve tension between cooperation and competition, openness and closure, governability and flexibility, accountability and efficiency. These competing tensions are evident in the governance models outlined. Too much flexibility and choice for the individual may result in too much complexity, uncertainty and unmanageability on behalf of care providers, and vice-versa. In the case of direct funding, for example, public concern about risk led to care coordinators and managers being mandated to scrutinise and potentially reject individual choice on the grounds of risk to the person, to others, and to the public purse (Stevens et al. 2011). On the other hand, too much accountability by delivery organisations to government may undermine efficiency in services, or responsiveness to service users' preferences. Too much competition in a market environment can result in lack of communication and coordination between different service providers servicing an individual. However, one thing is common across all models. If there are not sufficient financial resources to fund the necessary care and support, regardless of source, no model will work. A related finding of this analysis is that pooling of financial resources is much more likely to facilitate disability service governance than separate fragmented streams.

There are, however, limitations to the conceptual analysis of disability service governance above. Consideration has not been given to the nature of the organisation or person providing care, whether it be informally provided by family and friends, or formally by organisations, or whether the organisations are state, not-for-profit, or for-profit. The quality of these service providers and the relationship with the service user is of critical importance in the quality of the provision and governance of disability services. The intention of the NDIS is to coordinate support funded under the scheme with appropriate mainstream and informal support options. As such, it will be critical to consider the balance and interaction between these different mechanisms and their various strengths and limitations.

While it is acknowledged that no one perfect governance model exists, it is nevertheless important to consider that some models may be more suitable for specific individuals. People with high-level cognitive abilities, for example, are much more likely to be willing and able to manage their own care using the user-coordinated governance model (Glendinning et al. 2008; Blyth & Garnder 2007).

Apart from the conceptual analysis of governance models, the models and their visual representations also provide analytical tools for empirical work. Clearly, empirical case studies could be categorised into the dominant model in operation, or using some hybrid where necessary. Such categorisation can seek to determine to extent to which each model is in existence. More importantly, actual case studies can be linked to measures of governance success and failure, such as the use of Service Obstacles Scales or user satisfaction measures. Further research might consider measurements of unmet need, quality of life, poverty or social exclusion, and personal satisfaction and happiness. By linking measured outcomes to empirical arrangements, research can systematically and empirically evaluate the relative success of different governance models. Furthermore, empirical research may seek to examine how individual characteristics and situations are related to governance success/failure and satisfaction with care, thereby advancing our understanding of the fit between particular circumstances, individuals, and governance models.

The visual representation of disability governance is also an analytical tool that can be further developed. In the case studies presented in this paper, a limited list of service types was depicted. Transportation, advocacy, information provision, and support services to (informal) carers were not visualised, but could be included. Visual representations could also aid analysis by generating metrics, such as the number of organisations (i.e., square nodes) and individuals (i.e., circular nodes) involved in providing care and support, the number of service coordinators, the number of services/providers being coordinated for each case, and the number of edges connecting the individual. Significantly, these visualisations are depictions of networks that can be numerically analysed with social network analysis techniques (Borgatti et al. 2013; Easley & Kleinberg 2010; Scott 2012). Such techniques can characterise the roles different nodes play within a network. For example, an 'authority' node is one to which many others point (e.g., uncoordinated model), while a 'hub' node is one that points to many others (e.g., user-coordinated model). Such network metrics can further advance quantitative analysis of empirical cases by identifying patterns in disability governance and correlating these with outcomes. Mapping care and support arrangements over time is another potential use of the visualisations of care.

Moreover, these methods are also applicable to regional and international comparative research to compare and contrast the patterns of disability governance in different locations, which can be related to macro-level outcomes, such as poverty, unmet need, satisfaction, and social, political and economic engagement.

In the end, conceptual focus on governance returns to questions of metagovernance, which Jessop describes as 'the organisation of self-organisation'. It involves 'the design of institutions and generation of visions which can facilitate not only self-organisation in different fields but also the relative coherence of the diverse objectives, spatial and temporal horizons, actions, and outcomes of various self-organising arrangements' (1998: 42). Conceptual engagement with the governance of disability must return to ways in which wider processes and practices can generate disability governance that functions as defined within those very processes. This is indeed a task not only for government policy, but for service providers, people with disabilities, their families and friends, and others with a vision of what might be possible.

Acknowledgements

The authors acknowledge the Australian Research Council Linkage Project grant (LP0883377), the Queensland Motor Accidents Insurance Commission (MAIC) and the Queensland Public Trustee for financing the research on which this paper is based. The contribution by research participants and research staff to the research project is gratefully appreciated.

Endnotes

(1) National Disability Insurance Scheme, see www.ndis.gov.au

(2) Service Obstacle Scale, see http://tbims.org/combi/sos/

(3) The use of the open-source program NodeXL to generate original figures for this paper is acknowledged. Details about its use can be found at http://nodexl.codeplex.com/ and in Hansen and colleagues (2010).

(4) Such economies of scale are also observable in Beth's case, in which a single organisation case manages several clients with individualised funding.

References

Abbott, D., Townsley, R. & Watson, D. (2005) 'Multi-agency working inservices for disabled children: what impact does it have on professionals?', Health and Social Care in the Community, 13, 155-163.

Ackland, R. (2013) Web Social Science, London, Sage.

Bell, S. & Hindmoor, A. (2009) Rethinking Governance, Cambridge University Press, Cambridge.

Bevir, M. (2007) Public Governance, London, Sage.

--(ed) (2011) The SAGE handbook of governance, London, Sage.

Blyth, C. & Garnder, A. (2007) "'We're not asking for anything special": direct payments and the carers of disabled children', Disability and Society, 22 (3), 235-249.

Borgatti, S.P., Everett, M.G. & Johnson, J.C. (2013) Analyzing Social Networks, London, Sage.

Breda, J., Schoenmaekers, D., Van Landeghem, C., Claessens, D. & Geerts, J. (2006) 'When informal care becomes a paid job: the case of Personal Assistance Budgets in Flanders', in C. Glendinning & P. Kemp (eds), Cash and care, Bristol, Policy Press.

Brunton, W. (2003) 'The origins of deinstitutionalisation in New Zealand', Health and History, 5 (2), 75-103.

Challis, D., von Abendorff, R., Brown, R, Chesterman, J. & Hughes, J. (2002) 'Care management, dementia care and specialist mental health services: an evaluation', International Journal of Geriatric Psychiatry, 17 (4), 315-325.

Chenoweth, L. & Clements, N. (2009) Funding and service options for people with disability, Final Report, (eds), Brisbane, School of Human Services and Social Work, Griffith University.

Chester, H., Hughes, J. & Challis, D. (2010) 'Patterns of commissioning, contracting and care management in social care services for older people in England', British Journal of Social Work, 40: 2523-2537.

Clarke, J. & Newman, J. (2007) Creating Citizen-Consumers, London, Sage.

Denhardt, J.V. & Denhardt, R.B. (2007) The new public service: Serving, not steering, Armonk, NY, ME Sharpe.

Easley, D. & Kleinberg, J. (2010) Networks, Crowds and Markets: Reasoning about a highly connected world, New York, Cambridge University Press.

Emerson, E. (2004) 'Deinstitutionalisation in England', Journal of Intellectual & Developmental Disability, 29 (1), 79-84.

Fisher, K.R., Gleeson, R., Edwards, R., Purcal, C., Sitek, T., Dinning, B., Laragy, C, D'aegher, L. & Thompson, D. (2010) Effectiveness of individual funding approaches for disability support, Australian Government Department of Families, Housing, Community Services and Indigenous Affairs, Canberra.

Foster, M., Henman, R, Fleming, J., Tilse, C., Allen, S. & Harrington, R. (2012) A study of financing and management of lifetime care and support for adults with acquired disabilities and high care needs in Queensland, UQ Social Policy Unit Research Paper No. 3, St Lucia, The University of Queensland.

French, S. (1994) 'In whose service? A review of the development of services for disabled people in Great Britain', Physiotherapy, 80 (4), 200-204.

Glasby, J., Le Grand, J. & Duffy, S. (2009) 'A healthy choice? Direct payments and healthcare in the English NHS', Policy & Politics, 37: 481-497.

Glendinning, C. (2008) 'Increasing choice and control for older and disabled people: a critical review of new developments in England', Social Policy & Administration, 42 (3), 451-469.

Glendinning, C., Challis, D., Fernandez, J., Jacobs, S., Jones, K., Knapp, M., Manthorpe, J., Moran, N., Netten, A., Stevens, M. & Wilberforce, M. (2008) Evaluation of the Individual Budgets Pilot Programme: Final Report, Social Policy Research Unit, York, University of York.

Greener, I. (2008) 'Markets in the public sector', Policy & Politics, 36 (1) 93-108.

Hansen, D., Shneiderman, B. & Smith, M.A. (2010) Analyzing social media networks with NodeXL: Insights from a connected world, Burlington MA, Morgan Kaufmann.

Hunter, S. & Ritchie, P. (2007) Co-production and personalisation in social care: changing relationships in the provision of social care, Jessica Kingsley Publishers Limited.

Jessop, B. (1998) 'The rise of governance and the risks of failure', International Social Science Journal, 50 (1), 29-45.

King, R.D., Raynes, N. V. & Tizard, J. (1971) 'Patterns of residential care. Sociological studies in institutions for handicapped children', London, Routledge & Kegan Paul.

Kjaer, A.M. (2004) Governance, Cambridge, Polity.

Lane, J.E. (2000) New Public Management, London, Routledge.

Leece, J. (2004) 'Money talks, but what does it say? Direct payments and the commodification of care', Practice (UK), 16 (3), 211-221.

--(2007) 'Direct payments and user-controlled support', Practice (UK), 19 (3), 185-198.

Lord, J. & Hutchison, P. (2010) 'Individualised support and funding', Disability and Society, 18 (1), 71-86.

MacKinnon, D. Coleborne, C. (2003) 'Introduction: Deinstitutionalisation in Australia and New Zealand', Health and History, 5 (2), 1-16.

Madden, B., Mcllwraith, J. & Breil, R. (2014) The National Disability Insurance Scheme Handbook, Sydney, LexisNexis.

Mansell, J. & Ericsson, K. (1996) Deinstitutionalization and community living: intellectual disability services in Britain, Scandinavia and the USA, (eds), London, Chapman & Hall.

McLaughlin, K., Osborne, S.P. Sc Ferlie, E. (2002) New Public Management, (eds), London, Sage.

Needham, C. (2011a) Personalising public services, Bristol, Policy Press.

--(2011b) 'Personalization: from story-line to practice', Social Policy and Administration, 45 (1), 54-68.

Newman, J. (2005) Remaking Governance, (ed), Bristol, Policy Press.

Pollitt, C. (1993) Managerialism and the Public Services, Oxford, Blackwell.

Pollitt, C. Sc Bouckaert, G. (2011) Public Management Reform: A Comparative Analysis--New Public Management, Governance, and the Neo-Weberian State, Oxford, Oxford University Press.

Prideauz, S., Roulstone, A., Harris, J. Sc Barnes, C. (2009) 'Disabled people and self-directed support schemes', Disability and Society, 24 (5), 557-569.

Priestley, M., Jolly, D., Pearson, C., Ridell, S., Barnes, C. Mercer, G. (2007) 'Direct payments and disabled peole in the UK', The British Journal of Social Work, 37(7), 1189-1204.

Productivity Commission (2011) Disability Care and Support: Inquiry Report, Melbourne, Productivity Commission.

Rhodes, R.A.W. (1997) Understanding Governance, Buckingham, Open University Press.

Scott, J. (2012) Social Network Analysis, London, Sage.

Shakespeare, T. (2006) Disability rights and wrongs, New York, Routledge.

Spall, P., MacDonald, C. Zetlin, D. (2005) 'Fixing the system? The experience of service users of the quasi-market in disability services in Australia', Health and Social Care in the Community, 13, 56-63.

Stevens, M., Glendinning, C., Jacobs, S., Moran, N., Challis, D., Manthorpe, J., Fernandez, J., Jones, K., Knapp, M., Netten, A. & Wilberforce, M. (2011) 'Assessing the role of increasing choice in English social care services Journal of Social Policy, 40 (2), 257-274.

Walsh, P.N., Emerson, E., Lobb, C., Hatton, C., Bradley, V., Schalock, R.L. & Moseley, C. (2010) 'Supported accommodation for people with intellectual disabilities and quality of life: An overview', Journal of Policy and Practice in Intellectual Disabilities, 7 (2), 137-142.

WHO (World Health Organization) (2011) World Report on Disability, Geneva, WHO.

Wilson, G. (1994) 'Co-Production and Self-Care: New Approaches to Managing Community Care Services for Older People', Social Policy & Administration, 28 (3)236-250.

Yeatman, A., Dowsett, G.W., Fine, M. Sc Gursansky, D. (2008) Individualization and the Delivery of Welfare, Basingstoke, Palgrave.

Young, L. Sc Ashman, A.F. (2004) 'Deinstitutionalisation in Australia Part I: Historical perspective', British Journal of Developmental Disabilities, 50 (98), 21-28.

Table 1. Four models of disability support governance

Model             Uncoordinated             Caseworker

Characteristics   Separately provided       Services are
                  services are provided     coordinated by a
                  to the individual         caseworker
                  uncoordinated

Funding           Various                   'Pooled' into
                                            caseworker

Example           Default                   Individual budgets;
                                            disability insurance

Strengths         Individual at centre;     Coordinated;
                  nominally in control      person centred;
                                            holistic; advocate;
                                            reduces stress; clear
                                            accountability

Weaknesses        Individual has services   Depends on
                  done to them; little      caseworker; lack of
                  autonomy; fragmented;     choice or control;
                  diffuse accountability    gatekeeper

Governance        Segmented; complex;       Lack of choice or
failure           no autonomy; lack of      control; lack of finances
                  finances

Model             Dwelling-based             User-coordinated

Characteristics   Services are coordinated   Services are coordinated
                  by accommodation           by the individual (or
                  facility                   family member)

Funding           Single source or varied    Single from individual
                  into facility              budget

Example           Institutional or           English direct payments
                  community based care

Strengths         Coordination is somewhat   Personal autonomy
                  person centred; reduces    & choice; Innovative
                  stress; economies of       markets of care; clear
                  scale                      accountability

Weaknesses        Depends on dwelling;       Individual risk &
                  dwelling centred; lack of  responsibility;
                  choice, autonomy; hard to  inadequate financing;
                  change; gatekeeper         market

Governance        Poor dwelling              Inability to manage
failure           arrangements; lack of      funds; market failure;
                  choice or control; lack    lack of awareness of
                  of finances                possible services