NDIA Accountability Project

Challenges in aligning NDIS ICT system with legislative requirements

‹ PrevPage 1 of 14 · Source p. 1Next ›

Submission to the Joint Standing Committee on the National

Disability Insurance Scheme

NDIS ICT Systems Inquiry

Dr Darren O’Donovan

Senior Lecturer in Administrative Law, La Trobe Law School

This submission addresses terms of reference items (c) and (d)

c. the appropriateness of the agency facing IT systems d. the impact of ICT infrastructure on the implementation of the NDIS Reflecting the expertise and lived experience of people with disability in relation to the participant facing portal, this submission focuses on making recommendations regarding the Agency facing systems. There are three main themes:

  • Challenges in aligning the ICT System with the legislative requirements

  • Evidence of significant gaps in the data collection and extraction capability in the current system.

  • The ICT system as a decision-making environment: NDIS computer-aided decision making has been being complicated by manual overrides, practical constraints due to the high volume of decision-making and under-inclusive field design and data capture.

These three broad themes interact with each other, and provide a framework for analysing the Agency facing elements of the system.

  1. Aligning the ICT System with the Legislative Requirements The NDIS ICT system has featured prominently in the existing oversight body reports into Agency decision-making. Both the Office of the Ombudsman1 and the ANAO2 reports on eligibility decisions have expressed concern at the failure to fully align the business rules of the system with the legislative requirements, particularly around time limits and notification.

The most concerning aspect of the Auditor General’s Report into NDIS eligibility decision making was the finding that:

“The ICT system provides computer-aided decision making, which the ANAO identified was being manually overridden in a large volume of cases, associated with a known

1 Office of the Ombudsman, Administration of reviews under the National Disability Insurance Scheme Act 2013, May 2018. Hereinafter “Ombudsman” 2 Australian National Audit Office, Decision-making Controls for Sustainability — National Disability Insurance Scheme Access, October 2017. Hereinafter “ANAO”.

misalignment between the NDIS Rules and the ICT system business rules. The ANAO also identified a discrepancy between the system business rules and other NDIA guidance.”3

The Auditor General’s report identifies the institutional challenges generated by the limited ICT project delivery time and the slow place by which the detail of the legislative framework was laid down. Rather than a product of individual intention, ICT flaws were a product of the need to create a full ICT system against an uncertain backdrop. The secondary legislation needed to “spell out” the decision-making under the Scheme took significant time to come into place. Given the unwieldy nature of the COAG process the implications of this negative cycle of governance lags and ICT delivery needs close examination. There is also a strong likelihood of future tribunal and court cases refining the scope of the scheme, the implications of which will have to be integrated on tight deadlines.4

Previous iterations of NDIS ICT systems, as decision-making environments, did not appear to

reflect the Australian Government Information Management Office’s Automated Decision-

Making: Better Practice Guide (2007).5 In relation to eligibility decisions, the Audit Office made the following findings:

Human Services advised the ANAO that there is no single source of information setting out the business rules that underpin computer aided decision making. The ANAO reviewed documentation provided by both Human Services and the NDIA. In addition to transition rules not being aligned, the ANAO identified a discrepancy between the business rules and other NDIA guidance, with the inclusion of mild intellectual disability as a List A condition. A recent review by KPMG of NDIA access decisions also identified problems with the business rules, and concluded that these create ‘a data integrity issue in that the data generated cannot be relied upon for the purposes of assessment or review.’6 (Emphasis added)

This finding raises concerns that the aligning of the legislation with the business rules of the system has not reached the standards outlined at pages 36 and 37 of the Better Practice Guide. The Committee should confirm the actions which have been taken to comply with the ANAO recommendation.

Recommendations

The Committee should consider the recent KPMG report referenced by the Audit Office. It should identify and evaluate any remedial actions taken to enhance data integrity across eligibility and planning decisions in the aftermath of the ANAO findings.

3 ANAO, at [20]. 4 A recent example of such a case is Mazy v National Disability insurance Agency [2018] AATA 3099 (9 August 2018), where Deputy President Constance found the NDIS responsible for the provision of community nursing support to people with intellectual disability with diabetes. The Agency argued that this support was not most appropriately funded by the NDIS. The Conmmittee should explore the types of changes required when such determinations are handed down. 5 Available at: https://www.oaic.gov.au/images/documents/migrated/migrated/betterpracticeguide.pdf note: the definition of automated decision-making includes systems which merely aim at guiding decision-makers. The NDIS is not a statutory framework in where substantive decisions will ever be automated. The focus will be on creating a suitable online decision-making environment for planners, which informs and structures their discretion. 6 ANAO, above n 1, at [3.54].

The Committee should confirm the completion date for the promised “systems-based tool (NDIA Knowledge) to integrate NDIS business processes, policies and guidance to staff via a central repository”7

The Committee should confirm whether all the business rules of the system have been consolidated in one place and ensure they are accessible to Agency staff and those affected by their operation.

1.1 Compliance with Legislative Timelines

Both the Ombudsman and the Audit Report’s criticise the inability of the ICT system to embed or flag all legislatively required timelines. In a recent response to Senate Questions on notice, the Agency reiterated that:

‘The National Disability Insurance Agency does not currently systematically monitor the average length of time of a review.”8

The Agency has, in the past, slid significantly off compliance with express timelines mandated by the legislation. For instance it is a legislative requirement that the NDIA has 14 days on whether to conduct a plan review. In a recent question on notice response the Agency stated:

“The average time frame between lodgement and decision for plan review requests received through any channel between 26 October 2017 and 15 June 2018 was 30 days…[the

Agency] acknowledges that  this timeframe  is not being met in  all instances and  is

implementing several strategies to improve the response time.”9

Under the Act, the NDIA must also complete plan reviews and internal reviews ‘as soon as reasonably practicable’. The recent tribunal cases of Simpson10 and FJKH11 underline a past inability to log, track and triage the outcome of reviews. The Committee should consider what role the ICT system limitations played in the Agency’s limited response to the Ombudsman’s findings in relation to these issues:

“The NDIA has acknowledged timeliness standards are potentially valuable but advised, due to its current workloads and the limitations of its current data about review outcomes and timeliness, it is not yet in a position to decide or work towards defined service standards.”12

Later in the report, in relation to removing the manual workaround required to record internal review decisions, the Agency’s position was recorded as:

“NDIA is considering and prioritising ICT needs and requirements and will incorporate this recommendation into the consideration process”13

7 ANAO above n 1 at [3.57].

8 NDIA SQ18-000040

9 NDIA SQ18-000164

10 Simpson and National Disability Insurance Agency [2018] AATA 1326 (22 May 2018) 11 FJKH and the National Disability Insurance Agency [2018] AATA 1294 (15 May 2018). 12 Ombudsman, at [5.17] 13 Ombudsman, page 8 of the Agency’s response.

This raises issues of capacity and on what principles ICT work is prioritised. The primacy of legal compliance in any ICT reform process is underscored by the Better Practice Guide.14 Agency responses to many recommendations and questions on notice seems indicative of workflow or resourcing difficulties. The underlying reasons for any implementation delays should be explored by the Committee particularly as the Better Practice Guide states that “automated decisions should be designed so that changes in the business rules can be easily updated across systems”.15

Recommendations

Baseline timeliness requirements should be embedded into the system as a matter of priority.

The Committee should ensure that the Agency has a clearly stated policy of prioritising legal compliance in its ICT workflow, and seek explanation of how ICT work items are currently bundled, triaged or planned.

The Committee should, in particular, recommend that the Agency is adequately resourced to ensure ongoing capacity to promptly update reference package “amounts” and business rules in line with new court and tribunal rulings.

1.2 Reliance upon Manual Actions in Review Processing

The Audit Office report underlined that significant aspects of the decision-making process require manual workaround or manual data entry. The most high profile example relates to initial ICT system design requiring the plan review function to record reviews of statements of supports approvals:

“the NDIA advised us its ICT system currently requires staff to input a plan review decision to give effect to an internal review decision. The decision maker is required to draft and send a s 100 letter, advising the outcome of the internal review, and should intercept the s 48 letter automatically generated by the system. If one or both of the steps required of the decision maker does not occur, the participant may end up with only one (incorrect) letter or two (conflicting) letters.”

The distress and confusion caused by this entanglement of unscheduled plan reviews and internal reviews has been flagged heavily by the Senate Committee, the Audit Office, the Office Ombudsman and the Administrative Appeal Tribunal. This inquiry is however, uniquely placed to explore the design reasons for such a misalignment between the legislation and the ICT system. It should also ensure that the Agency now has the ability to separately report adequate data on the outcomes of these two distinct processes.

Recommendations

The Committee should ensure that the promised 2018 system update to remove the requirement to use the plan review function to record internal reviews has taken place.

14 See page 34. 15Better Practice Guide, page 24.

The Committee should recommend that the Agency ensure that it can track internal review outcomes by disability category (including activity domains), support type and appeal grounds as well as geography.

1.3 Recording ‘Interim’ Plans

A further example of a seemingly inefficient ICT solution relates to the handling of cases where there is a “gap” between the person’s NDIS plan. The cumbersome process for resolving situations is outlined in a recent response to a question on notice:

“After a new plan is approved, the NDIA business system will remediate any plan gap, allowing participants and providers to claim for supports provided in that period, and mitigating the problem in future plans. In this process, the lapsed plan end date will be extended to the day before the new plan was approved and funded supports in the expired plan will be pro-rated appropriately to ensure there are adequate funds to cover this additional period.”

The author is concerned about the approach to recording such “interim plans”. As recently was underlined by Deputy President Constance of the Administrative Appeal Tribunal, legally an NDIS plan functions until it is replaced by a new plan.16 An ad hoc pro rata extension of an old plan may raise issues under section 37(2) of the Act. As Justice Reeves of the Federal Court underlined in SSBV v National Disability Insurance Agency:

“Under the Scheme, each participant has a plan. Under the Act, a plan commences on the date prescribed by s 37(1) , and it continues in force until one of the events specified in s 37(3) occurs. It is therefore incorrect to refer to a plan expiring on a specified date. A plan may be reviewed from time to time under s 33(2)(c) of the Act, but such a review does not result in a plan ceasing to operate.”17

There is no legislative requirement for an NDIS plan to have an “expiry date” after which no support is delivered. Arguably, expressing an NDIS plan in timeless terms (e.g weekly, monthly rates of support) would better reflect the principle of certainty of funding contained in section 4(3) of the NDIS Act. The Committee should confirm the administrative priorities or ICT limitations that drove the creation of “expiry dates” to sit alongside review dates.

The author acknowledges that the Agency has made extraordinary and largely effective steps following recent media reports regarding these expired plans. Future “gaps” in plans will, however, continue occur where an individual struggles to submit evidence needed to complete their annual review.

The legal status of “commitments” given to providers to cover gaps in funding are dubious if they are not accompanied by the required s 34 analysis of whether such funding is reasonable and necessary. The process of inserting them after the creation of a later plan is not a process mapped by the Act. The author submits that the best and most certain approach is to avoid expiry dates, regulate possible gaps when making the initial plan and work to comply with the

16 Subsection 37(2) of the NDIS Act provides that a plan cannot be varied after it comes into effect, but can be replaced under Division 4. SHGH and National Disability Insurance Agency [2018] AATA 674 (27 March 2018). 17 SSBV v National Disability Insurance Agency [2018] FCA 1021.

review process envisaged within the legislation. Alternatively, the agency needs to more clearly record that a new “short term” plan has actually been created to cover the interim period.

Recommendation:

The Committee should identify the underlying reasons why plans are “backdated” within the current system.

Rather than backdating support though a confusing IT process, the Agency should design NDIS plans to function up until the date they are replaced. It may be preferable for NDIS plans to be expressed in a timeless terms such as “weekly”, “monthly”, “yearly” rather than tied to a listed expiry date.

1.4 Records of Phone Interactions

The case of Simpson underlines that the contents of a phone call are vitally important – in that case the individual lodged an appeal by phone and received an unclear communication back by phone. The Tribunal struggled to parse the interaction which had occurred:

“One may reasonably view the telephone call itself as the decision; it was, at the very least, evidence that the decision had been made. The content of the call dealt with finality with her request for review, even though – as summarised very briefly in the agency’s internal note – it appeared to be saying that it could not consider that request for review because it was out of time.”18 (emphasis added)

Given the reliance of many participants on phone communication the NDIS should consider the recording of phone calls where an individual consents to doing so. It should clarify whether the current contract with Serco would limit this. The author notes that the Department of Human Services recently moved to the default recording of all calls.19

Recommendation:

If it is not already occurring, the Committee should recommend that Agency consider the consensual recording of calls. The Committee should check the current procedures around the mandatory documentation of the purpose of any phone calls made to or by a participant.

1.5 Correcting Basic Errors in Plans

The Agency informed the Ombudsman of its plans to implement:

18 Simpson and National Disability Insurance Agency [2018] AATA 1326 (22 May 2018) Deputy President Humphries underlined that “in reality, what was before the Tribunal was indicative simply of Mrs Simpson’s application falling between the cracks, and being overlooked. The Tribunal is familiar with far too many cases where the statutory pathway for internal review has not been observed in a timely way” at [26] 19 Evidence of DHS at Additional Estimates, 1 March 2018, at page 155. The period of retention for these recordings has not been published, however, such transcripts can be obtained by FOI.

“Changes to its ICT system to allow staff to make straightforward corrections to plans without the need for a full plan review.”20

The concept of a “straightforward correction” should be carefully defined, and restraints imposed within the ICT system. This also ensure that participants and advocates know what can be easily and quickly fixed, and hold the Agency to that expectation.

It is also important that there is clear communication about the difference between an ad-hoc correction and a review. The ANAO report found that original decision-makers were fielding requests for reviews and resolving them in person. It correctly underlined that:

“If conducted in a timely manner, reconsideration by the ODM may result in a speedier resolution of the review request, particularly where the review application is accompanied by additional supporting evidence, as the ODM is already familiar with the circumstances of the case. However, ODM reconsideration has no legal status and cannot supplant the legal requirement for review by an internal review officer.”

While a reformed system is laudable, a “minor changes” system needs to delivered in a way that ensures the ability of the ODM to correct mistakes is curtailed. Without proper on the ground communication any such a power may confuse participants as to whether an appeal has occurred.

Recommendation:

The Agency should publish a definition of a “straightforward correction” and confirm that the system does or will restrain planners from making corrections of a broader scope.

20 Ombudsman, at [2.7].

  1. Improving Data Collection, Recording and Extraction There are, in the author’s view, a large number of documented instances where the ICT’s design appears to have resulted in lost data, frustrating or complicating the proper policy evaluation of NDIS implementation. A most concerning example was in the Productivity Commission Report on Costs:

“The Commission was unable to assess the effect of adding learning or social interaction to the eligibility criteria, because the NDIA does not collect data on which (or how many) of the six activity domains are relevant to each participant when they enter the NDIS. Speech Pathology Australia, however, said that their members who are NDIS providers are not providing services to children whose only disability relates to learning and literacy.”

The Commission was forced to rely upon the anecdotal experience of a provider, rather than gathered data. This is especially frustrating as, unlike other aspects of NDIS decision-making, the “activity domains” that were relevant to the individual could have been recorded through a drop down menu or field.

The author has collated a sample of answers to questions on notice and responses to oversight bodies which indicate a range of important data points about the Scheme which are in either an unstructured format or are unrecorded:

  • Reasons for exiting the scheme are “not recorded as structured data”21

  • The Productivity Commission (October 2017) found that “the NDIA does not collect data on which (or how many) of the six activity domains are relevant to each participant when they enter the NDIS.”22

  • Reasons for the underutilisation of plan funding are not systematically recorded.23

  • The lack of data collected on where and in what circumstances participants have been allowed to use paid family carers under the scheme.24

  • Data on the number of people who are found eligible for the NDIS on appeal (as distinct from first instance) is not recorded.25

  • Following the introduction of the first stage of a new ICT system on 1 July 2016, the NDIA was “unable to gather data on the number of requests received for internal review of access decisions (as at 30 June 2017)”.26

    • The system does not record the nature of available Specialist Disability

Accommodation. The Agency was recently unable to report to the Senate on how many enrolled dwellings are single participant, mixed tenancy, up to three participants, up to five participants or are legacy stock.27

21 NDIA SQ18-000174. 22 Productivity Commission report on NDIS costs at page 21. 23 See pages 20-21 of the McKinsey Review of NDIS pricing. The McKinsey Review contains excellent recommendation for more granular and focused data collection at page 60. The author strongly supports the use of exception reports informing planners and LAC of underutilisation anomalies, this is an excellent initiative. See Question on Notice NDIA SQ18-000091. 24 Productivity Commission, page 352. 25 NDIA SQ18-000167. 26 ANAO, at [4.4]. 27 NDIA SQ18-000139.

  • The NDIA answered that it would be an unreasonable diversion of resources to identify how many new SDA dwelling have come online over last year (excluding public housing stock that is deemed to have met the standards).28

  • The Agency was unable to identify how many people currently living alone in SDA approved accommodation have had the amount of SDA funding in their plan reduced.29

  • ”The National Disability Insurance Agency does not systematically collect data on the employment type of all participants.”30

  • Young People in Residential Aged Care who were found ineligible for the NDIS as a result of not meeting the disability criteria, with a percentage breakdown of why they did not meet the disability requirements.31

  • Young People in Residential Aged Care who sought a review of their plan with a breakdown of outcomes.32

  • The Office of the Auditor noted that Integrated Performance Reporting Framework “Since July 2016, the NDIA has not reported on a number of measures due to data limitations.” This observation was not further explored given the narrow nature of the audit’s terms of reference.33

  • The practice of reporting the outcome of internal reviews appears to have been discontinued since June 2016 Quarterly report, when he was recorded that 66% of internal reviews were successful. The Productivity Commission found that “there is no publicly available data on the impact that internal reviews of decisions have had on the scheme costs”. The author notes that the recent quarterly report the Agency change its table format – removing the previous practice of disclosing the outcomes of settlements or resolved conciliations.34

  • The NDIA is not reporting detailed data on the proportion of participants who attain the goals outlined in their plans.35

  • “The National Disability Insurance Agency registers providers by state or territory, not by region.”36

  • “Data is not collected on the reason for changes in supports, nor review requests, including the number of plan reviews that reduce supports as a response to plan underutilisation.”37

28 NDIA SQ18-000130

29 NDIA SQ18-000143

30 NDIA SQ18-000035

31 The author notes the response of the Agency to the Standing Committee’s inquiry , Question Reference No: SQ18-000078 which stated: “An Exploring Housing Options package consists of Support Coordination and Capacity Building (daily activities) categories of funding. As many participants have funds in their plans in these categories, it has not been possible to separately identify those who specifically have an Exploring Housing Options package. In relation to younger people living in residential aged care with an Exploring Options package, anecdotally the numbers are very small. Work is underway to explore restructuring data and plan preparation to allow this reporting in the future.”

32 NDIA SQ18-000090

33 ANAO at [5.25]. 34 Productivity Commission at page 420. Compare table E.31 of the latest quarterly report with table 1.33 in the previous report. 35 Productivity Commission page 479. 36 Senate Standing Committee, Answers to questions taken on notice by the NDIA at a public hearing in Kalgoorlie on 18 April 2018, Reference No: SQ18-000081.

37 NDIA SQ18-000045

  • “The National Disability Insurance Agency holds no data about the number of National Disability Insurance Scheme-eligible people in hospital settings.”38

  • “It is not possible to provide data about the number of mainstream interface issues arising from individual participant plans.”39

  • “As at 30 September 2017, there are 52 participants under 18 who have had Supported Independent Living (SIL) or Shared Supported Accommodation (SSA) in their plan. $3.9 million has been committed in current plans. Data is not available on the number of requests for SIL or SSA that have not been granted.”40

The above list is only a sample drawn from the past year. In a system which is as avowedly data driven as the NDIS, there is a danger that what is not counted does not count.

The Committee should examine what system limitations or administrative emphases led to the occlusion of these important data points and what recording options or solutions might exist. The IT system is delivering core “line management” type data – based on decision-maker, broad geography and review/appeal KPIs. Based on the sample above however, it does not adequately deliver the necessary cross-sectional data to support best practice policy evaluation.

Any lack of balanced, structured data for frontline decision-makers is inevitably going to lead to issues of quality and accuracy. There is evidence that the Tribunal is struggle to identify or reconstruct the reasons for first instance decisions. For instance, it is of central importance that

decision-makers  analyse and  record  the  reasons why money went  unspent,  prior  to

implementing a pro rata reduction funds as occurred in the tribunal decision of LMNT v National Disability Insurance Agency.41 The adverse frontline consequences of key data points being trapped in an unstructured format, was underlined in the Productivity Commission’s criticisms of current ICT capability. The Commission found that the system:

(i)    “Cannot  transfer  free  text  to  the  data warehouse, so  details on  individual

circumstances that could impact on reasonable and necessary supports are not readily available” (ii) “Does not transfer all required information on participant assessments to the data Warehouse – for example, when assessing participants’ level of function, the data warehouse is provided a standardised functionality score, but information on the screening tool used for the assessment is not transferred to the data warehouse. This compromises the ability of the NDIA to assess the appropriateness of screening tools being used or to assess how consistently a screening tool is operating in practice.”42

Alongside missing or unstructured data, there is also a concern that reporting on e.g. expended funds is a costly and unwieldy process where there is a broad funding pool which multiple providers may have access for multiple reasons.

38 NDIA SQ18-000069

39 NDIA SQ18-000063

40 SQ17-000189 41 LNMT and National Disability Insurance Agency [2018] AATA 431 (6 March 2018). This case was complicated by the failure of the internal reviewer to provide a justification for a substantial reduction in support co-ordination, at [34]: “Somewhat surprisingly, there was no allowance for this at all in the revised plan of 12 May 2017. The only explanation given for eliminating this support in the very brief reviewable decision of 9 May 2017 was that “Support Coordination is not most appropriately funded by the NDIS”.” 42 Productivity Commission, page 476.

Recommendations

The Agency should commit to a Charter of Data Capture and Reporting for the NDIS, which delivers key priority data requested by representative and advocacy bodies in the disability sector. Parliament should consider creating a Schedule of required data in future legislative amendments.

The Agency should clearly identify which NDIS data points currently exist only as unstructured data requiring the resource intensive collation of “free text” within the ICT system.

The Committee should assess whether the Agency has met the Productivity Commission recommendation to “collect and publicly release granular data, feedback and reports on thin markets, including when Provider of Last Resort arrangements are used”. A thin market support data collection process should begin, identifying available vacancy and provider data in greater detail.

Annual reviews should include the recording of reasons for under-utilisation of funds as structured data through provision of a mandatory fields such as “late activation”, “lack of understanding”, “could not find a service provider” etc.

Annual reviews should include a mandatory field outlining the employment type the person is currently occupying or is searching for.

The Committee should confirm that the Agency has modified its eligibility data collection process to ensure that it can provide a breakdown of the qualification and rejection rate for each eligibility field.

Committee should recommend that the eligibility data is further broken down in line with the NDIS rules – with subfields or free text designed for the more specific examples included in the NDIS Becoming a Participant Rules.

The rate and reasons for successful internal reviews should be reported on regularly.

The ICT System as a Decision-Making Environment

While this inquiry can usefully examine past system outages and unauthorised access to support funding, it is also important to assess and make recommendations relating to the best practice design of the decision-making platforms.

Given the individualised, deeply situational nature of most NDIS planning decisions, the ICT system for planners can only involve the flexible channeling of planners’ judgments rather than more prescriptive forms of business rules. The aim will be to achieve a balance between ensuring consistent approaches to applicable legal and policy principles and valuing the specific evidence and findings regarding the individual person’s context and future plans. While process elements can be automated, the ICT system must resource decision-making effectively through the provision of:

“…useful commentary, including about relevant legislation, case law and policy for the decision-maker at relevant points in the decision-making process”

The Impact of Embedding Reference Package Amounts on the Decision-Making

Environment

The clearest statement of principle regarding the future shape of a reformed planning platform was supplied in an estimates hearing this year:

“It is still new, but our intention is to ensure that we utilise the system for what you would always expect within a customer relationship system to have a bit of business intelligence around it—providing flags or indicators to the staff members on the extent to which what they’re proposing for that person is consistent with what a typical person in those typical circumstances would typically require. They have to justify and put the right evidence in the system to explain their reasonable and necessary judgments where they are different from what you might have as your base.”43

The Agency then outlined a process of pre and post decision quality assurance, which the author strongly supports.

The author is concerned however, that the current system may place an increased evidence threshold on departure from “reference package” amounts inbuilt into the system. While this is intended only as a guide, it is important for the Agency to recognise the manner in which an ICT system can create a dominant or “elephant” path. To ensure a balanced exercise of discretion, significant questions would include:

  • Does the system require any process of confirmation that a person is “typical” or does it simply assume this?

  • What is the expected level of detail in a free text entry or file which is expected to accompany a “typical” planning decision?

  • What audit mechanisms are there for tracking the median and variability within plans, to ensure situations are not rounded to fit an overall numerical average?

43 Agency Representative at Supplementary Estimates, Wednesday, 25 October 2017 at page 47.

The key danger to embedding reference packages as the baseline for the new system is that financial sustainability is one variable in a much broader value for money test required by legislation. As Deputy President Humphries stressed in the recent AAT matter:

“It should be observed  that the enquiry demanded by  s 3(3)(b)  is flavoured with

methodological uncertainty. If the section is to be construed as saying that any decision which adds significantly to the cost of the Scheme is to be eschewed, then the Tribunal would have little difficulty in finding for the Agency. However, financial sustainability surely entails the making of value judgements about the cost of widening the Scheme’s scope versus the benefits so conferred. Significant additional cost may be justified if the benefits thus conferred are also significant. Adopting the construction urged on the Tribunal by the NDIA proceeds on the assumption that the NDIA has already made decisions pitching the level of support for disabled Australians at the right level, and that to supplant those decisions in favour of more generous ones, irrespective of the merits of doing so, is ipso facto wrong. It is doubtful that this was the intention of the legislature. It might also be suggested that the notion of a scheme’s financial sustainability is itself a function of the nation’s overall liquidity and its priorities, matters over which the Tribunal may lack competency to make findings.”44 (emphasis added)

The inclusion of “target” or “expected” amounts in the new transport operational guideline has not led the analysis of the likely funding in the two published AAT decisions involving transport.45 In neither case, the Agency did not provide an explanation of what cost-benefit logic underpins those guideline amounts. Outside of these transport amounts, the “typical” levels of funding for specific types of support have generally not been published or otherwise opened for public scrutiny or discussion. In addition, the Agency has been inconsistent in tabling evidence of projected costs in AAT matters.

The author underlines that the question of how best to approach value for money analysis under the Act is an extremely complex one. The submissions made should not be collapsed or exaggerated into ideas that the NDIS works towards fixed targets or that planners do not retain discretion. It is however, very important to have a sound evaluation of the Agency’s approach to cost-benefit analysis and to make the variables involved fully and equally visible to planners and participants.

Recommendations

The Agency facing material which describes how reference package amounts are to be followed or overridden should be published. The reference package amounts which set the baseline expectations of what planners should “typically” approve should be made publicly available.

The Committee should obtain a clear statement of timelines for the carry through of policy adjustments following Tribunal (or future court) rulings which modify the Agency’s preferred approaches. The ICT system must be responsive to changes in the policy settings.

44 BIJD v NDIA

45 In both cases, the tribunal funded the support at a level four times above the stated highest level in the current operational guideline.

Mandating the specific recording of reported outcomes, anticipated benefits/outcomes and the cost of alternative support

There is a danger that the benefit of NDIS supports is relegated to the unstructured, “free text” realm, while the system is structured around expected costs. NDIS funding decisions are based on insurance or investment logic – which is centred on identifying the future savings if a

support is not provided. Even  if the Agency has clear financial sustainability data,  it is

important, in the light of legislative requirements, to work equally hard at quantifying the benefits of support. As the Productivity Commission noted this is a more difficult task than measuring costs against cost expectations. Nevertheless, the test for supports under the NDIS Act requires an individualised value for money test. Even if a person’s level of need is the same, the benefits to delivering support across a lifetime may not be.

In the author’s view, the NDIS Short Form Outcomes Framework is underdeveloped relative to its cost projection framework. This may affect the quality of the frontline reasoning if planners are not provided with strong guidance on how to quantify the cost saving or benefit in a particular circumstance. The Committee should evaluate whether the current ICT system properly highlights the importance of identifying the estimated saving of providing a support. It is also important to systematically record the reasons for underutilisation: this will allow the planner to assess whether a funded support simply failed to produce the hoped for outcome or whether the participant was simply not able to e.g. find a provider.

Many of the key aspects which determine the quality of NDIS reasoning will inevitably exist only in free text. The Agency’s recent commitments to ensuring pre and post decision auditing should be praised. It is also important however, that it also commit to external auditing and be responsive to advocacy sector’s views on where possible poor quality reasoning might be occuring. The Agency also needs to be properly resourced to carry out this type of high quality auditing –the individualised nature of NDIS decision-making stands apart from other schemes.

Recommendations

The Agency fix a calendar and frequency for “deep dive” analysis and auditing of sampled free text reasoning on the “reasonable and necessary” and eligibility criteria.

The Independent Advisory Council should have the ability to require the completion of a set number of free-text analyses on its chosen issues each year.

The Committee should recommend that the Department ensure sufficient funding and capability exists for the Agency to carry out this activity.