Strengths, Gaps, and Challenges in the Decision-Making Processes in the Design of the National Disability Insurance Scheme, Australia

Abstract

This study examines the decision-making processes that were used in the development of the National Disability Insurance Scheme (NDIS). It draws on research that highlights that decision-making can be improved if organisations follow specific decision-making processes. This study sought to examine whether formal decision-making rules and processes were used in the development of the NDIS. Semi-structured interviews were conducted with 58 senior public servants between 2016 and 2021 in the Commonwealth government departments that had policy oversight and responsibility for both the design and implementation of the NDIS. The data showed that there were examples of data-driven decisions and risk analysis used, but also considerable gaps in decision-making processes. These gaps potentially resulted in poorer decision-making quality, which may have negatively impacted the design of the NDIS and could potentially explain some of the current problems with the scheme. The paper concludes with recommendations for future research to improve decision-making processes within public administration.

1. Introduction

One of the leading academics in the field of human decision-making recommends that organisations and individuals be judged by the quality of their decision-making processes, not the outcome of the decision itself. This is due to outcomes being influenced by many factors outside of the control of any organisation (Kahneman, 2011). An interesting case study of where decision-making outcomes have been heavily criticised is the National Disability Insurance Scheme (NDIS). The NDIS has been under heavy criticism and focus for the increasing cost of the scheme, which was not expected at the design stage (Chrysanthos, 2023) and research has identified significant market gaps, delays, and equity issues, which are posing significant risks for people with disabilities, missing out on necessary services (Green et al., 2022b; Malbon et al., 2019). While the commentary has focused on the outcomes of decisions, Kahneman’s (2011) work suggests that what is important to understand is the actual decision-making processes and whether they followed decision-making guidelines or rules.

The NDIS represented a seismic shift from block-funded disability services to a personalised market-based model. It is considered the largest public administration reform in decades, reflecting changes internationally regarding a personalisation agenda (Dickinson & Glasby, 2008). The intention was that the NDIS would replace a $22 billion funding system that was considered inequitable and inconsistent (Collings et al., 2016; Productivity Commission, 2011) due to each State and Territory government having control of disability services and delivering different models (Fisher, 2010). A review of the existing arrangements found that the programmes and supports available were insufficiently funded, fragmented, and provided people with no choice or control (Productivity Commission, 2011).

The NDIS was conceived as a market-based social insurance scheme and was passed into legislation with bipartisan support in 2013 (Thill, 2015). The original intention of the NDIS was that 460,000 people with significant and permanent disability would receive a personalised funding budget and plan to fund their care (Collings et al., 2016; Productivity Commission, 2011). The speed of reform and implementation was rapid, with a seven-year time frame to full implementation, which compares to the 30 years it took for this reform process in the United Kingdom (Carey et al., 2018). There were three stages to implementation. A trial phase involving seven sites, a transition phase in which all eligible people were placed on provisional service plans, and then the ‘full scheme’, where everyone would receive a complete plan and funding support (Productivity Commission, 2011, 2017).

There is little research examining the formal decision-making processes within public administration (Joyce et al., 2024). Given the scale and significance of the development of the NDIS, this makes it an interesting case study in the decision-making processes. This study explored whether structured decision-making processes were being used in the development of the NDIS, based on interviews with key personnel conducted over a five-year period. The data reported in this paper comes from senior policy makers involved in decisions at all stages of the planning and implementation phases, from initial design through to implementation decisions. Interviews covered the processes of how key decisions were made, with results showing there were some formal decision-making processes used, but also many times when there was a lack of a good decision-making process. There were some formal decision-making processes related to risk management, data-driven decisions, and a range of committee structures to ensure a range of opinions were canvassed. However, there were many instances of a lack of process and decision-making rules. This highlights the risk of making poor decisions and could have increased the possibility of some of the major problems now facing the NDIS. This paper concludes with recommendations for addressing this gap in research and practice.

2. Theoretical Framework

Good decision-making processes and rules fall under the domain of policy capacity (Capano & Howlett, 2021; Lawrence et al., 2020; Wu et al., 2015). Good decision-making characteristics are viewed as fundamental to policy formation (Capano & Howlett, 2021). There are various levels to consider policy capacity, which include system capacity, organisational capacity, and individual capacity (Wu et al., 2015). At the organisational capacity level, having defined rules and processes for making decisions is seen as an important factor (Lawrence et al., 2020). Lawrence et al. (2020) developed a policy capacity tool, and one of the criteria points was the presence of a decision-making rule or process in place. Much of the foundation of what constitutes a good decision-making process or rule comes from many different strands of psychology research, including cognitive, affective, social, and organisational psychology.

There has been considerable research on cognitive biases and decision-making heuristics that can produce suboptimal decisions, which was popularised in Kahneman’s (2011) book. Research has demonstrated how people routinely overattend to some pieces of information (e.g., novel and particular pieces of information) and neglect others in decision-making (e.g., base rates), and we are less accurate at probabilistic type decisions than we realise. As another example, our confidence in a decision is impacted by how easily the information is retrieved and if events appear to be causally related. Decision-making processes related to reflective question prompts and imagining failure to determine risks are examples of approaches to improving decision-making to address cognitive biases at an organisational level (Bergen & Bressler, 2018; Klein et al., 2019).

As reviewed by Lerner et al. (2015), there are a number of ways in which emotions drive and influence decisions. Different emotional states can lead to different types of biases and how information is analysed. For instance, the emotion of sadness has been shown to be related to increased deliberation and less reliance on heuristics relative to feeling angry or happy. Reducing the effect of unhelpful emotions in decision-making has proven difficult, and at present, using choice architecture methods has the most promise. This involves presenting the choices in such a way that it increases the chances of positive decisions (such as the placement of healthy food in a cafeteria). How this relates to complex policy choices is not clear.

There is also a large body of research examining how social processes influence decisions (Larrick, 2016). This research has led to a number of recommendations for improving the decision-making process within organisations. For example, to avoid groupthink, a potential solution is to enable people to think about a problem individually before doing so as a group. There are also many other tactics to ensure that different perspectives are considered. This can include making sure there is cognitive diversity among the group, which means a variety of thinking and problem-solving styles rather than just assuming that demographic diversity can be used as a proxy. Over time, people who have worked together start to think alike, and it is important to bring in new perspectives while still balancing some sense of group cohesion. There are also tactics to encourage debate and other specific group roles. Much of this research has been applied in organisational business settings.

This psychology research has been used in fields such as medicine, engineering, and business to guide decision-making processes and rules (Atanasiu, 2021; Kahneman, 2011). These techniques are particularly common in corporate environments at a board and managerial level where there are specific decision-making processes (Marnet, 2011). These decision-making considerations and tactics include the number of people to include in decisions and particular roles and tasks people can perform in a decision process (Atanasiu, 2021). There is a range of specific decision-making rules and strategies that organisations can use to improve the decision-making process (Atanasiu, 2021; Bergen & Bressler, 2018; Kahneman, 2011; Klein et al., 2019). These strategies include: question prompts encouraging reflection on where biases may be occurring (Bergen & Bressler, 2018), a process of hindsight thinking as a novel way of assessing risks (Klein et al., 2019), and approaches to deliberately elicit counter viewpoints where there is the possibility of similar thinking of group members (Atanasiu, 2021; Bielby, 2000).

The World Bank produced a report which, within a policy context, uniquely analysed decision-making biases of their staff and the decision-making processes they used to address these biases (Parkhurst, 2017; World Bank, 2015). The report highlighted four decision contexts where biases may occur: confirmation bias (seeking information to confirm existing beliefs), complex decision contexts (increases the potential biases in logical reasoning to occur), sunk cost bias (continuation of a failing project), and how context and social environment influences decisions (making assumptions based on own personal experience that do not necessarily apply to people in different circumstances) (World Bank, 2015). The decision-making processes used to address these biases included deliberate processes to produce counter viewpoints (Arinder, 2016; World Bank, 2015). However, it is unknown the extent to which these types of decision-making processes are being used in public administration more broadly (Joyce et al., 2024).

Applying specific decision-making rules may be much more difficult in public administration relative to corporate environments, engineering, and medicine, given the complexity of the policy process (Tiernan, 2011). The process of policy decision-making can be more complex than decision-making in other organsational contexts, as it is less structured and ordered (Clavier & de Leeuw, 2013). Policy theories illustrate the numerous factors that influence policy decisions, such as the role of institutions, values and belief systems, and a myriad of actors across different levels of government and community (Clavier & de Leeuw, 2013; Hill & Hupe, 2006; Sabatier, 2019). There are a number of theories that conceptualise the complexities of the policy process. These theories examine different components of the policy process. They include focusing on institutions (institutional rational choice), ideas and belief systems (the advocacy coalition framework), interactions between problems, political issues and policy solutions (multiple streams theory) (Sabatier, 2019) and the interrelationship between actions and policy decisions between levels of government (Hill & Hupe, 2006).

Clavier and de Leeuw (2013) summarise policy change as multifaceted, with change occurring at different moments in the policy process (i.e., following elections or after a public crisis such as an epidemic). Policy change can occur due to a change in the balance of power between actors involved in a policy or through the influence of new knowledge. The extent of public policy change can also vary from a dramatic paradigm shift or, alternatively, as a result of small steps. Thus, at the broad level of policy change, it may be difficult to implement a formal decision-making process. While having decision-making rules is seen as an important component of policy capacity (Lawrence et al., 2020), to what extent decision-making rules can be applied in a policy context is unclear. To address this gap, this study adopted a case study approach, focusing on the NDIS, to examine decision-making processes for NDIS policy development and implementation. The aim of the study was to ascertain whether there were any structured decision-making processes or rules used in the design of the NDIS, and if so, the quality of those decision-making processes.

3. Methodology

3.1. Research Design

This study aimed to investigate the decision-making processes that were used by policymakers when making decisions about the design of the NDIS. This includes decisions regarding overall scheme design, particular features of the NDIS, and implementation phases and considerations. This study adopted a case study research design to enable an understanding of how decisions were made in real-life contexts and how these decision processes changed over time (Yin, 2013). Semi-structured interviews were conducted with 58 senior public servants between 2016 and 2021 in the Commonwealth government departments that had policy oversight and responsibility for both the design and implementation of the NDIS, specifically the Department of Prime Minister and Cabinet and the Department of Social Services.

3.2. Data Collection

Purposive sampling was conducted of people in key positions involved in the NDIS across the following policy and programme areas: Governance and Stakeholder Relationships, Policy and Legislation, NDIS Transition, NDIS Quality and Safeguards, NDIS Financial Policy and Performance, Financial Modelling, Budget Development, Market Oversight, Market Reform, Disability Employment and Carers, and Mental Health. Interviews were either conducted in person or online. This study draws on four rounds of participant interviews. Most participants took part in one or two interviews, and seven participants took part in at least three interview rounds.

The semi-structured interview questions covered both decision-making processes and the policy details related to those decisions. The process questions are the main element of this research paper which concentrated on how decisions were made: who was included, what information was used to guide the decisions, what processes were used to guide the decision (for instance how did they choose between different policy options), and what reflection was undertaken about the decision. In respect of the content, each round of interviews focused on something slightly different, dependent on the particular challenges and opportunities at the time. The data from this paper is drawn from all the interview rounds. Informed consent was obtained from all participants, and the study was approved by the UNSW Human Ethics Committee (HC16396).

3.3. Data Analysis

Given the limited research on decision-making processes at the organisational level in public administration, there was no pre-existing analytic frame to use, and hence a thematic analysis was undertaken (Braun & Clarke, 2006). Following best practice qualitative research, the data analytic steps have been described (Levitt et al., 2018). The data was first analysed using thematic analysis to uncover themes related to describing decision-making processes, some of the barriers and facilitators of good decision-making processes, and any reflection on improvements to the decision-making process (Blaikie, 2009; Hansen, 2006; Saldana, 2016) This involved a process of reading the data line by line and grouping the data into meaningful categories (Corbin & Strauss, 1990; Minichiello et al., 2008). The next analysis stage involved drawing connections between the key themes (Minichiello et al., 2008), such as how these different contextual factors influenced the decision-making processes. Further themes were developed related to the differences in decision-making processes at a policy and practice level. The themes were discussed and refined with all study authors.

4. Results

The results begin with some of the broader factors influencing NDIS policy decision-making processes. This establishes the context for how policy decisions were made. This is followed by more specific insights into formal decision-making processes, including a description of these processes and participant perceptions regarding their success or otherwise. The final section covers participant suggestions for how these decision-making processes could be improved. This section also highlights the differences in the application of decision-making rules at a policy level compared to a service level.

4.1. Context—Factors Influencing the Decision-Making Processes

This section presents key factors influencing decision-making processes outside of the control of the people involved in the decision-making processes, particularly, lack of time and political pressures.

4.1.1. Lack of Time

One of the common factors identified by participants was the lack of time to implement good-quality decision-making processes:

Then the whole thing, you know the timelines so the whole thing was brought forward, ludicrous implementation deadlines.

(Senior Public Servant P20, R1)

You can’t implement 94% of the scheme in a three-year period, you just can’t do it.

(Senior Public Servant P5, R1)

The rushed timelines, described as ‘insanely ambitious’ (Senior Public Servant P20, R3), also included those associated with the establishment of the NDIA. The speed with which the agency had to establish itself and then implement the scheme was considered by some participants as ridiculous and a key factor that contributed to proper decision-making processes not being well established:

So after having a very bad start to the commencement of transition, which damaged (publicly) the reputation of the scheme, and the scheme remains, I think, dogged by those issues, and in my view they were because of decisions made in haste and inadequate preparation…I don’t want to slate this all home to the agency, but part of that went (to me) to their decision-making processes…. That’s why I think it’s still quite miraculous it’s all happening. I can’t think of another example of an organisation that’s had to establish itself and then grow at such scale and pace—like to be a national $22 billion organisation that’s implementing stuff across the country—within two to three years timeframe.

(Senior Public Servant P21, R3)

The results suggest that time pressures prevented the ability to establish proper decision-making processes in many instances. However, as will become clear, even with these time pressures, there were still opportunities that were missed for good decision-making processes to be used.

4.1.2. Political Pressure and Path Dependence

Another major contextual factor to consider in relation to decision-making processes is the political process itself. Participants discussed how, once decisions were made, there was a great reluctance to deviate from those decisions, even when it was blatantly obvious to those making decisions that changes needed to take place:

So political attention means that failure is not an option… I would characterise it as, one of the consequences of the political attention is that every milestone, every agreement has to be met… It’s the decisions that were made some years ago, actually, that no-one was prepared when we signed bilateral agreements over that 18 month period, that no-one was prepared to step away from.

(Senior Public Servant P33, R3)

This has implications for the quality of decision-making processes, as decisions were rushed and continuously adhered to, thus contributing to a range of policy development and implementation problems.

4.1.3. Policy Alignment

One of the most common factors impacting NDIS decision-making processes, identified by participants, was related to the alignment of interests across different departments and between State and Federal governments. From the outset of the decision-making processes, alignment across different Australian Public Service (APS) departments and agencies was a key consideration in advancing the policy. Early decisions focused on considering which departmental secretary and minister would be best placed to progress the policy idea and thus in what portfolio it should initially be placed:

And then it was a matter of which secretary was going to be best placed to nurture it and the one working to [Minister Name] I think is probably the answer to that.

(Senior Public Servant P1, R1)

A key consideration was also aligning the interests of the Federal and State governments. The initial proposal from the Productivity Commission was that the scheme should be funded and run by the Federal government, but there were concerns with this proposal as it would mean that States would abdicate their current financial and programme commitments to disability. Given that States are responsible for the portfolios of health, education, and transport, it was argued that they needed to be part of the scheme. While it was considered a remarkable achievement to have the States and Federal governments jointly agree to the scheme, it did mean that aligning the interests of different levels of government dominated the decision-making process:

Through the negotiations with states and territories it evolved from being the model of the productivity commission to a sort of more traditional eligibility assessment social support model.

(Senior Public Servant P5, R1)

A specific example of aligning to State interests was the decision to locate the NDIA office in Geelong; a decision made based on the Victorian Government’s desire to promote regional employment. Some participants argued this was an insufficient rationale for this decision and had major consequences for staffing:

Then the decision was made, purely a political decision to move to Geelong… Well he lost almost all of his staff.

(GP20_R1)

These results show that key decisions related to which department had responsibility for the NDIS were influenced by the alignment of interests across different departments and agencies. Decisions regarding the scheme were also shaped by efforts to align Federal and State government interests. It is evident that these processes focused on who needed to make decisions from a political perspective; however, how these decisions were made from a process point of view seemed to receive less attention.

4.1.4. Policy History

One of the interesting observations of participants was how past experience of government service provision shaped the decision that the NDIA would not be a service provider:

Look I think that probably they were pretty influential in not having the NDIA be a service provider. So Employment National was an Australian government arm of employment services provision and it went broke, and there were a whole lot of Australian government employees who lost their jobs and it was a pretty big scandal, and so that was probably pretty front of mind.

(Senior Public Servant P4, R1)

And I can remember, interestingly, at the time it was not long after the pink batts drama [This refers to the deaths of four home insulation installers, attributed to the poor design and implementation of the Rudd government’s home insulation program, see (Grattan, 2014)]. And so there was some members of the taskforce who were very, very focused on that.

(Senior Public Servant P6, R1)

Our results highlight that historical factors shaped an unwillingness for the government to have a direct role in service provision and influenced the decision to adopt a market-based approach to the NDIS instead. This is interesting from a decision-making perspective where a past event influences how people make decisions regarding future events, even though the context and situation are somewhat different. The context of the decision-making has revealed the complexities and time pressures that may have made it difficult to use specific decision-making rules. The next section will uncover whether there were any formal decision-making processes used.

4.2. Formal Processes Guiding the Decision-Making Process

This section focuses on some of the structured and formal processes that guided decision-making regarding the NDIS. Participants identified some formalised processes, including risk management deliberations, structured consultation processes, and data-driven decision-making through a series of research and evaluation reports used to inform decisions.

4.2.1. Risk Analysis

Firstly, some risk assessments were undertaken, although this process was not undertaken consistently, and participants involved in the risk assessment process provided mixed feedback on its utility:

And that was quite an interesting forum because it was totally focused on risk management and so what we had to do was, we’d sent our risk framework to them and the mitigations that we’d planned or had in place. And they then challenged us on those assumptions, on have we thought of everything, is the world going to tend, that type of thing. <laughs> It sort of went to quite extreme.

(Senior Public Servant P2, R1)

This particular quote was in relation to risk assessment on how the scheme would operate at the ground level with respect to assessments and service delivery. One of the interesting themes to emerge from this data was that decision-making processes, where they existed, often related to service delivery considerations rather than broader policy design considerations. This will also be covered later in the results.

Many participants were concerned about the implementation aspects of the NDIS and highlighted that there had not been an opportunity to discuss these concerns in a decision-making forum. Participants outlined several risks which had not been canvassed thoroughly within a decision-making process:

I think that there is very substantial risk around the implementation of the NDIS…I have heard it said that the NDIS could make the pink batt scheme look like a walk in the park … … it is an absolute credit to the staff and the agency that they’ve been able to do it. Me and my colleagues who are involved in it, we were expecting a train wreck by now.

(Senior Public Servant P5, R1)

So huge number of potential risks for the agency, just building scale and workforce and capability and training and outsourcing and so the list goes on. From the commonwealth perspective making sure they’re ready and that we are satisfied that any big risks are being managed or there’s contingency plans. Market sector workforce, market transition obviously is huge risk. I’m quite worried about.

(Senior Public Servant P22, R1)

Individuals were aware of some of the implementation risks and how areas such as mental health, autism and developmental delay may impact the scheme. While some of these risks were considered, the speed of implementation, ‘I’m not comfortable that it’s happening so quickly’ (Senior Public Servant P22, R1), meant that not all the different risks and contingencies could be properly analysed. It was often up to individual workers to consider particular risks rather than a formal process for decision-making.

4.2.2. Consultation Processes

Formal decision-making processes focused on who needed to be included in the decision-making process. There were formal agreements in place between the Commonwealth and State governments, but these were not always followed. Another decision-making mechanism was organised consultation processes with providers, consumers, and various government departments, but, again, this was not consistently utilised:

This was supposed to be a Commonwealth decision, it wasn’t done with a whole heap of consultation.

(Senior Public Servant, P20, R1)

Q: But the States were comfortable with having the NDIA look the way that it’s now looking?

A: It wasn’t an issue discussed with the States.

(Senior Public Servant P16, R1)

When these consultation processes were not used, participants stated that senior staff had to ‘retrofit the governance and the policy’ (Senior Public Servant, P25, R2) to match what decision had been made.

4.2.3. Data Driven Decisions

Participants did provide positive examples where decisions were guided by research and data at a Cabinet and departmental level:

When you give them good evidence-based databased analysis and say ‘this is a risk, we’ve got to change it’, they do.

(Senior Public Servant P22, R1)

Well there’s a formal evaluation, and as well as that there’s sort of ongoing, I guess feedback and reporting from the trials that should be coming through the commonwealth state apparatus and the ministerial council.

(Senior Public Servant, P4, R1)

The results revealed some examples of structured decision-making processes for risk analysis, consultation processes for seeking feedback, and data-driven decision-making. For example, participants identified committee structures, inclusive of consumer and carer groups, which were put in place to ensure a diverse range of people were involved in key decisions. There were also instances where data and research were presented to assist the decision-making process. However, there were many instances of key decisions that were rushed, without data or consultation, and not made at a committee level. This means there was the possibility of less-than-optimal decisions.

4.3. Gaps in Decision-Making Processes

Our results highlight that there were several key documents and reports that guided decision-making, consultations did take place, and there was evidence of formal decision-making processes, such as risk management forums and meetings. However, as already described, these processes were inconsistently applied, and decisions were made without any formal analysis, consultation, or defined decision-making process. The lack of processes around certain decisions was noted by participants, particularly with respect to human resourcing decisions:

A: As I say I have never seen anything that actually sets it out in terms of implications or pros and cons. There may have been something, but I didn’t see it. It would be very interesting to have seen it.

Q: No one’s mentioned such a document to me at any point.

(Senior Public Servant P20, R1)

Someone’s making a decision and saying, ‘well, this is the way we’re going to go’, without what I would have thought you would want in an organisation which is a lot of discussion and debate amongst the very senior people, to say, can we get this as right as possible?

(Agency Staff Member, P1, R2)

While it was generally acknowledged by participants that, at times, having to obtain agreement across Commonwealth and State governments and various agencies slowed the pace of decision-making, there were also instances where implementation decisions were made in haste, which had considerable implications for policy development and implementation:

But at the same time, there’s a piece where under all those radars, implementation decisions are being made that have policy consequences. I’m not necessarily sure that’s always a good thing…I think there’s this pocket of stuff happening, for very pragmatic and good-hearted reasons, because you’ve just got to get on with it. But they have quite far-reaching policy implications.

(Senior Public Servant, P33, R3)

There seemed to be no guidance at an agency level on which decisions required further deliberation from a policy perspective and which decisions could be made without further scrutiny. These judgements seemed to rely on the experience of the staff involved, and, given the speed at which implementation was required, processes were not developed or followed. There were also particular problems raised with the decision-making processes within the NDIA and the degree to which decisions made at senior levels were not followed through at an implementation level:

The translation of that into action often didn’t happen because someone else in the organisation thought, either it was a bad idea or they had a different priority… I’m not sure what structures were in place to make sure it even happened in the first place.

(Senior Public Servant P26, R2)

The results reveal that there was no consistent application of decision-making rules and processes. The lack of decision-making rules and processes may have had adverse consequences related to the design elements of the NDIS. As mentioned in the first theme, to some extent, there were political and time pressures that prevented good decision-making processes. However, the participants themselves recognised that even within these constraints, better decision-making processes could have been implemented at departmental and agency levels.

4.4. Future Plans for Decision-Making Processes

Future reforms about the decision-making processes were focused on trying to more clearly delineate the roles and responsibilities of the different agencies and key personnel. Participants acknowledged challenges within the decision-making environment for the NDIA:

I mean that a private corporation CEO responsible to a board, a private board, does not have somebody—a minister sitting above that food chain and then a prime minister above—and a cabinet and then a prime minister above that, that they have to be conscious of—in decision-making.

(Senior Public Servant, P53, R4)

Participants also highlighted how future deliberations on changes to the decision-making process are focused on changes to who needs to be included in the decision processes, but less focus on changes to how the decisions will be made:

Our minister is very keen to have a greater level of streamlining in decision-making. So for example rather, than unanimity across jurisdictions for all board members appointments where that took 18 months to achieve, just basically going with the majority as long as the Commonwealth is part of majority.

(Senior Public Servant, P33, R3)

It is an interesting observation that while there were gaps in robust decision-making processes and rules at a policy level, clear decision-making processes were seen as necessary at a client implementation level. One of the great ironies present in the data was the high level of focus and debate on the decision-making structures and processes required for client assessments and reviews. There was considerable thought given to having a structured procedure by which to make client assessments:

But the more you see the right clinical decision-making, and the right parameters around that, the fewer outlies you have.

(Senior Public Servant, P6, R1)

We’ve also asked them to make sure they’ve got consistency across the planners and how decisions are made.

(Senior Public Servant, P14, R1)

This was an area of extensive review and risk analysis, as previously mentioned, focused on how to improve the client-level planning decision-making process. A large consultation process took place, involving over 100 meetings in various city and regional locations, about this topic. The findings revealed:

There was a universal agreement that it’s not necessarily about how quickly things happen it’s about the transparency of decision-making…Another key thing from the Tune Review was not only transparency but consistency with decision-making.

(Senior Public Servant, P26, R4)

The reference to decision-making transparency in this quote relates to a lack of understanding and information provided to clients on how their planning decisions were made and the lack of consistency in decisions regarding the level of support provided. The information behind planning decisions was not often apparent to clients. This led to considerable effort put into the process and standardisation of planning decisions, with arguments that there is a need for a good decision-making architecture that provides a degree of standardisation without being too restrictive:

I think it’s trying to look at creating more defensible and robust framework for decision-making around planning…. The ultimate driver is transparent, defensible, carefully delivered supported decision-making.

(Senior Public Servant P38, R4)

The interview data suggest that those involved in NDIS policy development are conscious of the need for good decision-making processes and structures. Participants expressed strong intent for having a robust procedure by which to make good decisions, but it is clear this was only consistently applied when thinking about how to make decisions at a service level.

4.5. Comparing Decision-Making Processes at a Policy and Practice Level

Table 1. Strengths, gaps, and challenges in the decision-making processes of the NDIS.

	Policy Level	Service Level
Who needs to be involved in the decision process	Strengths: Involving Federal and State Governments Consultation with Community Organisations Involving various government departments Gaps/Challenges: Consistently involving the stakeholders identified	Strengths: Participants Local Area Coordinators Support Coordinators Plan Managers Advocacy organisations Service providers Gaps/Challenges: Shifting roles and responsibilities between those involved
How they need to be involved in the decision process	Strengths: Some use of risk analysis Some evaluation and data-driven decision-making Gaps/Challenges: No apparent processes for uncovering and addressing biases No clear understanding of decision-making authority between and within agencies and government departments for key NDIS design decisions No clear roles assigned to people in the decision-making processes	Strengths: NDIS Independent Assessment Framework NDIS standardised assessment tools Suitability assessment process for NDIS providers Structured NDIS assessment process Gaps/Challenges: Lack of resourcing and funding for individual advocacy Assessment based on ‘worst day’ Lack of consistency in understanding key terms Lack of experience of those conducting plans Delays in planning process

distils some of the strengths and limitations of the decision-making processes. Table 1 illustrates the major gap identified in the results of a lack of consideration of how decisions should be made at a policy level. There was considerable attention about who needs to be involved in decisions at both a policy and service level, albeit there have been some challenges with shifting roles and responsibilities at a service level (Malbon & Carey, 2021). At a service level, there was also significant deliberation and attention given to the process by which decisions were made. Again, while implementation at a service level has been criticized for lack of resources (Green et al., 2022a) and problems with delays and experience of planners (Joyce et al., 2025), there was at least an acknowledgement of the need for a structured process by which service-level decisions would be made. The processes by which policy decisions should be made seem to have received less attention.

There was no indication throughout any of the interviews that there were any plans for future reform or improvement on decision-making rules or procedures by which NDIS policy decisions would be made.

5. Discussion

The results of this study demonstrate that there was considerable opportunity to improve the decision-making processes involved in the development and implementation of the NDIS. While policy work is very complex, interview participants reflected that at an organisational and committee level, there could have been improved decision-making processes. There was time allocated for risk analysis, which indicates there was scope for the implementation of other decision-making processes within these government departments and agencies. The results did indicate that a strength of the decision-making processes was the strong consideration of who needs to be involved in decisions. Senior public servants put a lot of thought into establishing decision-making processes that involved all the key relevant governments and agencies across Australia, although they were not always included as desired. The results revealed another strength as being the decision-making skill of policy decision makers (Klein, 2013) in understanding where alignment was possible between State and Federal Governments and their departments, and pursuing this direction.

The results are consistent with previous research on policy development, where one of the key factors involved in the NDIS decision-making process was alignment, which is one of the central factors in policy success (Joyce et al., 2019). In this study, alignment across the State and Federal governments was one of the critical elements in shaping NDIS policy direction. At times, alignment was achieved through specific processes at a committee level that enabled discussion and negotiation of interests; at other times, alignment was achieved through direct communication between senior politicians, including between the Prime Minister and Premiers. There were also examples of political values shaping decisions and particular interest groups having a role in advocating for policy positions, which correspond with research on how political values influence policy decisions, and the role of interest groups in shaping policy direction (Baumgartner et al., 2014; Kingdon, 1995; Schneider et al., 2014). Thus, it is clear that a range of people were included, which was a deliberate focus on who needs to be included in the decision-making processes.

The results revealed a main gap in decision-making processes—a lack of purposefulness regarding what roles key people should have played in decision-making. This highlights a potential direction for future research. This study has illustrated an absence of decision-making rules, which is a key component of policy capacity (Lawrence et al., 2020). There is a dearth of research on specific decision-making processes and rules in public administration (Joyce et al., 2024). It is clear that the policy-making environment was complex, as would be expected; however, there were examples of data-driven decision-making and risk assessment as examples. This illustrates that there was scope for decision-making processes, but they were not consistently used, and there may have been other techniques that could have been used. For example, risk analysis as a decision-making process has been shown to be ineffective in countering optimistic bias, and what is instead required is a process to elicit hindsight thinking, which alters the way risks are assessed (Bergen & Bressler, 2018; Klein et al., 2019).

Future research could examine whether public administration organisations are using techniques such as: reflections on where biases may be occurring (Bergen & Bressler, 2018), creative ways to uncover potential failure (Klein et al., 2019), processes to avoid group think by first asking people to reflect on the policy challenge individually before discussing as a group (Larrick, 2016), or processes like encouraging counter viewpoints where there is a risk of similar thinking of group members (Atanasiu, 2021; Bielby, 2000). In this study, none of these types of decision-making processes were evident, and interview participants themselves commented on the lack of any decision-making process or rules being in place for key decisions. This contrasts with the service level provision as there was great thought into both who needs to be involved in service planning decisions and deliberate processes and tools for how service decisions should be made. While the implementation of service planning has been problematic (Joyce et al., 2025), there was recognition of the need for a structured process. Deliberate decision-making processes at the policy level were often absent.

Whether policy makers could have predicted the rapid increase in cost (Chrysanthos, 2023) and some of the other market challenges at programme design is difficult to tell (Green et al., 2022b). However, if we are to judge not on the decision outcomes but rather on the process of decision-making (Kahneman, 2011), then these results illustrate serious shortcomings in the decision-making process. This seems an unfortunate gap, given the scale of the reform of the NDIS, one of the largest administration reforms in decades (Dickinson & Glasby, 2008).

This study highlights the need to document the types of decision-making processes that are being used in public administration across different departments and in different jurisdictions. If future research confirms the findings of this study, that there is a gap, then the next step will be to ascertain what types of decision-making processes and rules best suit public administration organisations. As previously mentioned, the World Bank examined which decision biases may be present within their organisation and addressed these biases, including deliberate processes to produce counter viewpoints (Arinder, 2016; World Bank, 2015). Another direction for future research is to consider how the workplace environment may influence decision-making quality within public administration. Moss et al. found that if the workplace environment provided a sense of meaning and purpose in people’s lives, it could reduce cognitive biases (Moss et al., 2016). Further, the presence of these biases is linked to increased workload and burnout (Moss et al., 2016). Burnout is a significant problem in public administration and many workplaces (Golembiewski et al., 1998). Thus, further research could explore how a particular type of workplace environment can improve decision-making capability in public administration. It is not just the presence of a decision-making rule or process that is important, but also the workplace environment in which those decisions are made. This study highlighted the problems with decision-making quality when there were considerable time pressures.

The major limitation of this paper was that questions were semi-structured about how decisions were made and what processes were used, and did not specifically ask whether decision-making processes used in management science research were being applied. There may have been instances where such processes were used, but this is very unlikely, as interview participants did provide detailed descriptions of what occurred during meetings and the decision-making structures in place. Further, we did not survey participants nor seek to quantify the number of times that they experienced the issues addressed in this paper. Future research could more specifically survey public administration staff on the use of different decision-making structures and rules. Another limitation was that we did not specifically examine which decision-making biases may be present, which future research could address. Future research could also explore how the workplace environment in public administration influences decision-making biases and the use or otherwise of decision-making rules and processes.

6. Conclusions

This paper has revealed a lack of formal decision-making processes and rules during NDIS policy development. Many of the current problems the NIDS is experiencing could potentially have been avoided with better decision-making processes. There are a number of decision-making processes and rules, such as deliberate processes of dissent and alternative problem framing, which have been recommended for policy decision-making (Schippers & Rus, 2021). There is also a range of strategies and decision-making rules used in corporate environments that were not evident in this case study. Interview participants explicitly noted the absence of formal decision-making procedures or guidelines for key decisions. This is contrasted to service-level planning, where careful consideration was given to designing structured tools and processes to guide decisions.

The challenge at this point in applying decision-making processes and rules is that working within a policy environment can be far more complex than other fields (Tiernan, 2011). However, this data did reveal that there were instances where data-driven decision-making and risk assessments were applied, suggesting that structured decision-making processes were possible, even if inconsistently employed. If there is time for risk analysis, then this demonstrates that there is scope for structured decision-making processes within a public administration context. These findings underscore the need to document and analyse the decision-making approaches currently used across different departments and jurisdictions in public administration. Should further research confirm the presence of this gap, the next step will be to determine which decision-making models and frameworks are most appropriate and effective for public sector organisations. This study has revealed that there needs to be a far greater focus on the implementation of quality decision-making processes to ensure better policy outcomes. Further research is required to determine whether this lack of decision-making process is evident across other government departments at senior levels and, if so, how to address this gap.