Submission to the Joint Standing Committee on the NDlS
Impact of the “NDIA Price Lie”
On Friday, 9th August 2024 The Australian published an article by
political reporter Sarah Ison titled *Calculation on NDIS price ‘totally
broken’*. An editorial piece by the same paper the next day argued that
the care economy matters too much to be unsustainable.
The key points are:
- The NDIA made material mistakes in its cost model. These are generally errors; they are not “differences of opinion”.
- The direct impact of these errors is a price limit difference of the order of ten percent or more.
- The NDIA has repeatedly ignored these issues and made no correction to its pricing model or outcome, instead continuing to cut prices in real terms, including in its 2024 Annual Price Review.
- The NDIA continues to promote¹ that price limits are determined by the 25th percentile in provider sector cost data, which the NDIA knows is false. The NDIA is acting without transparency and honestly and is in breach of its own Code of Conduct; in doing so is doing significant damage to its relationship with the provider sector.
While I have questioned the NDIA’s integrity over the cost model issue, the far bigger issue is that providers
– especially registered providers - are at financial breaking point.
The sector has provided a wide range of benchmark data demonstrating that NDIS price limits are not
sustainable, including surveys by Ability Roundtable and StewartBrown. In spite of the fact that this data is
more recent and more relevant than any work undertaken by the NDIA, the NDIA has acknowledged but
then ignored these submissions.
There is an urgent need for price relief for the provider sector. It is now clear that this will not come
voluntarily from the NDIA, which is conflicted in setting prices while trying to manage Scheme cost.
Required Action
A refreshed approach to pricing is needed and work on this has commenced².
Until a rigorous, evidence-based price model is introduced and/or price determination is overseen by an
independent authority, the NDIA must introduce temporary price relief.
I propose that this price relief be directed to registered providers, who have a higher cost of registration and
compliance than other providers do; in addition, this cost is increasing significantly over time.
The amount of this temporary payment must be at least 5%. This temporary payment should be a monthly or
quarterly payment to registered providers based on their NDIS claims for specified support categories
over the previous period.
Implemented in this way, the payment is targeted to providers with high quality (and the highest costs), and
the payment does not affect participants or the price guide.
Failure to implement temporary price relief while errors in the NDIA’s Cost Model are addressed will result in
providers leaving service provision and turmoil for participants.
Submission author: Dr Andrew Young
I am the CEO of a registered NDIS provider and am on/have been on Boards for others. I am a former CEO of the
University Centre for Social Impact, in which role a key focus was systemic delivery of better social outcomes.
¹ For example, there are references in the NDIS 2023-24 Annual Pricing Review Report, and in presentations about the 2023-24 price review made to providers in July 2024.
² The NDIS Review (Dec 2023) recommended pricing be overseen by an independent authority. The NDIA also announced on June 28 2024 that “In coming months, we will appoint an independent expert to oversee a review of the NDIS pricing structure ahead of next year’s APR.”
1
Submission to the Joint Standing Committee on the NDIS
Impact of the “NDIA Price Lie”
Attachment: What are the Errors in the DSW Cost Model?
This attachment summarises key errors in the NDIA’s Cost Model.
**I believe that all of the issues below are known to the NDIA and the NDIA has “not refuted” any of
these claims.** If the reader is uncertain about the accuracy of the claims below, the NDIA could be asked to
confirm or refute them.
1. Overall Model Approach
The Disability Support Worker Cost Model (DSW Cost Model) calculates the cost of a support worker hour,
based on the Award wage of a worker, and building up to a total cost per hour incorporating various oncosts.
The overall structure of the DSW Cost Model is an accurate reflection of real-world provider costs, and
many of the on-cost calculations are correct.
The DSW Cost Model incorporates key variables from the Deloitte “Financial Benchmarking Survey” (NDIA
Benchmark Survey) of around 1,000 providers (repeated annually over a few years). The key variables
from the survey that are used in the Cost Model include:
- Span of Control (number of Support Workers per Supervisor): 15:1 (page 8³)
- Utilisation (percentage of the Support Worker’s paid hours that are “billable”): 92% (page 10)
- Organisation Overheads (finance, IT, HR, “quality and compliance” etc): 12% (page 10)
- Percent of Staff that are “Permanent” (not casual): 70% (page 9)
- “Operating Overheads” including Workers Compensation: 1.7% (page 8) Employee Allowances: 1% (page 8) and Payroll Tax: 0% (deliberately omitted from the cost model – top of page 8).
The key issue is the methodology the NDIA has used to select values from the survey for use in the
DSW Cost Model.
- The NDIA’s stated assumption is that the overall cost of an efficient (25th percentile) provider can be estimated by using the 25th percentile value for each key variable in the Cost Model.
- This assumption is simply wrong. I have demonstrated directly to the NDIA that rather than the 25th percentile in overall cost, this approach results in overall cost around the sixth percentile – that is, 94% of organisation are unsustainable at the price set using this methodology, based on the NDIA’s own benchmark data. I have provided mathematical evidence of this to the Agency on multiple occasions.
One way of explaining this error is through this analogy: in an Olympic decathlon field of 20 athletes, the
“25th percentile” is the 5th-best athlete.
- If the same athlete came 5th in each of the ten decathlon disciplines, the NDIA model says that athlete would come fifth overall.
- In fact, that athlete would win the gold medal (ie would be top five percent in the field of 20).
- This is because athletes are good in different disciplines – some are strong, others are fast. The same is true of NDIS providers – for example, a provider employing casual staff or contractors can have high Utilisation (they only pay for hours worked), but may not be able to achieve low Overheads (they have high turnover and training costs). The NDIA modelling assumes a provider can do both.
2. Span of Control
The DSW Cost Model uses a “span of control” figure of 15:1 – that is, one Supervisor is provided for every
15 Support Workers. This figure is from the 25th percentile figure in the NDIA Benchmark Survey.
- The Benchmark Survey ratio is in terms of headcount – for every 15 Support Workers there is one Supervisor.
- In the Cost Model, the ratio is in “Full Time Equivalents” – FTE. The reality is that most Support Workers are part time, so 15 FTE is (on average) 25-30 headcount.
³ Page number references refer to the 2021-22 DSW Cost Model, which is the most recent detailed cost model published.
Apart from increases and decreases in temporary loadings, prices have moved in line with Award movements since.
2
- The NDIA Benchmark Survey 2021-22 also has data for the Span of Control Ratio in FTE terms – this figure (at the 25th percentile) is 7.6:1.
- The NDIA applied the wrong figure in its cost model, and the impact is that the cost allowed in the DSW Cost Model for Supervisors is half what it would be if the correct figure from the NDIA Benchmark Survey was used.
3. Annual Leave
The DSW Cost Model includes a calculation for the cost of annual leave, based on a Support Worker
entitlement of four weeks annual leave a year.
Under the Fair Work Act s87 deals with annual leave entitlements and provides that “an employee is entitled
to 5 weeks of paid annual leave, if: a modern award applies to the employee and defines or describes the
employee as a shiftworker for the purposes of the National Employment Standards.”
The National Employment Standards defines a shift worker as “an employee:
- who is required to work over a roster cycle that includes any of the 7 days of the week
- who is regularly rostered to work Sundays and public holidays.”
4. Organisation Overheads
The NDIA Benchmark Survey indicated that Organisation Overhead was 19.8% (2020-21 survey) or 21.8%
(2021-22 survey) at the 25th percentile.
The NDIA has in its Cost Model explicitly stated that it chose to ignore the benchmark survey data for
overhead, on the basis that it believed that “more efficient providers didn’t fill out the survey”. The NDIA has
made up an arbitrary figure for overhead (12%) and has never provided evidence to support this number,
even in the face of contemporary benchmarking surveys showing the true cost to be significantly higher.
In addition, one of the drivers of overhead cost for Registered Providers is the cost of registration and
compliance; for example, My Supports has analysed its costs for these activities finding that costs are of the
order of 6-7% of NDIS income.
5. Payroll Tax
The DSW Cost Model explicitly “does not provide for payroll tax as most jurisdictions exempt not-for-profit
and smaller organisations from payroll tax”.
However, the NDIA’s Benchmark Survey indicated nearly half of providers are not exempt from payroll tax.
Payroll tax adds around 5% to DSW costs in all States.
As a footnote, the NDIA may respond that it considers other factors in determining its annual price review outcome.
First, this response is not a defence of the way in which the NDIA has ignored the errors in its cost model. There are (literally) hundreds of references stating that price limits are determined by the cost model. The NDIA is still today telling providers that the price is set to be aligned with the 25th percentile of provider costs. The NDIA must return to an honest and transparent approach if it is to begin to repair the sector relationship.
Second, the “other factors” the NDIA might refer to are not transparently communicated with the sector and are likely hotly contested. The provider sector has presented to the NDIA a large amount of data demonstrating the price limits determined by the cost model are unsustainably low.
The temporary payment proposed in this paper must be implemented while the cost model approach to price-setting is re-built.
3