New data
The Commonwealth is investing $31 million to develop the National Disability Data Asset (NDDA). This will present many opportunities to improve outcome monitoring for people with disability in Victoria.
The development of large, person-level linked data assets means that now is the time to accelerate investment in research and analysis that aims to use these datasets to monitor outcomes for people with disability. This includes evaluating the effectiveness of social policy on improving the lives of people with disability in Victoria.
The NDDA will include a series of linked datasets from the Commonwealth and state governments. It will provide insights on service-use pathways and outcomes for people with disability across government service systems. In a secure data lab environment, it will provide detailed, unidentified individual-level information on interactions with government services (for example, NDIS, Disability Support Pension (DSP) and tax data), and key life events (for example, deaths and hospitalisations).
However, it is vital that we understand what can and cannot be done with such data.
The NDDA will have complete coverage of those who use a specific service. For example, NDIS data in the NDDA will accurately reflect everyone who uses the scheme at a given time, and since its roll out. Linking this information to hospital records, for example, will allow researchers to assess whether NDIS services and supports enable people to return home from hospital promptly.
The NDDA will allow more granular analysis of key events in people’s lives. For example, after a health event, such as hospitalisation due to a heart attack, what factors predict NDIS participants’ return to work and recovery? Further, how can NDIS services be better used to aid participants return to their income prior to the event?
However, the NDDA will not have total coverage of all Victorians with a disability. Because data assets, such as the NDDA, are largely made up of administrative datasets, the main sources of information on disability will be services (NDIS) and payments (DSP) people receive related to disability.
We know that not everyone with disability uses the NDIS or receives DSP. So, if we relied solely on NDIS and/or DSP payments for outcome monitoring in the NDDA, there will be sections of the population with disability who are missing.
There is potential, if the Census is linked to the NDDA, to use the Census question on disability to find people who have a disability but do not receive disability services or payments. But the limitation of this approach is the Census captures just people with severe and profound disability, and only a snapshot at one point in time.
This is not to say that the NDDA will not have immense value – it will allow researchers to answer policy relevant questions that, to date, they have not been able to. Rather, it should be used to add detail, nuance and policy-pertinent information that cannot be supplied by social surveys or identifying people with disability using medical information alone, which is not best practice.
Our recommendation is that when using linked data assets, such as the NDDA and PLIDA for outcome monitoring, it is made clear who is included in the analysis, and which groups are potentially missing.
The linked data can then be used alongside surveys (such as HILDA) to add richness to our understanding of how people with disability in Victoria are faring and how social policy can be optimised to improve outcomes.
Future work should also include engagement with Victorian Government departments to understand what data sources might be relevant and accessible for outcomes monitoring for Inclusive Victoria: state disability plan 2022–2026, and future iterations of the plan.
Intersectionality and diversity
Population data used in projects like this almost always contains information on key sociodemographic characteristics of individuals. For example, the main data source used in this project was the Household Income and Labour Dynamics of Australia (HILDA) survey. This contains information on gender, income, age, remoteness, education and other characteristics.
These identities and social positions intersect with disability, creating unique challenges in achieving the goals outlined in the current plan. In short, disability is not the only factor that determines good (or bad) employment outcomes, or good (or bad) health.
We know from the literature, our expert knowledge and discussions with VDAC that sociodemographic factors influence the measures.
Intersectionality and the diverse experiences of people with disability was a common theme in our discussions with VDAC.
For this project, we examined various intersecting social identities and characteristics.
For example, we found that psychological distress is significantly higher among young people with disability: 85% of young people with disability experience psychological distress.
Food insecurity is also more prevalent among people with disability living in outer regional Victoria (21%) compared with those in major cities (8%).
Investigating these intersections can help to better target and tailor the current plan and related government actions.
We recommend that future iterations of the outcomes framework adopt an intersectional approach by:
- including case studies in future State Government reports on specific intersections of interest
- incorporating intersectional reporting in the standard outcomes reports, possibly using icons to indicate associations with gender, income, age and other important intersecting factors, where data permits, such as Indigenous status and people from a diverse background. Please note we recommend that intersectional analysis of these groups should be developed closely with the communities the monitoring focuses on
- producing a special report examining on intersectionality.
Better measurement of progress
In this report, we have slightly adjusted how the results are reported to focus on change over time. Previously, in the old monitoring framework, it was noted that:
- Absolute measures describe the situation for people with disability only. Relative measures describe the situation for people with disability relative to people without disability. Relative measures allow us to determine if changes in the wellbeing of the population are equally shared by people with and without disability.
In most circumstances this is correct when you are considering just one point in time. However, there are 2 key problems with this approach that need addressing.
First, inequality changes (either widens or closes), but not because of a change (improvement or worsening) in the measure for people with disability. For example, if smoking rate drops among people without disability but remains the same for people with disability, the inequality may change.
Second, inequality widens according to the relative measure, but the difference in the percentage closes (or vice versa).
For example, if the outcomes go from 40% to 20% for people with disability, and from 25% to 10% for people without disability, this shows that there has been a larger reduction in the outcome for people with disability and a closing of the gap between people with and without disability.
However, if you calculate the relative inequality at the 2 time points, the relative measure has increased from 1.6 (40% / 25%) to 2 (20% / 10%).
The above paragraph is technical, and potentially hard to understand.
But it matters, because, using the same data, one inequality measure appears to show progress, however the other appears to show that things are worsening.
In this report we guarded against this by adding a text description of the result. VDAC also requested this to aid interpretation of the results by a broad audience.
There is clearly work to be done in how we assess progress. We recommend road-testing a range of presentation methods and inequality metrics for future outcomes framework reports.
Timing of reporting and data availability
One of the challenges of producing this report was the availability of data.
For example, the Survey of Disability Ageing and Carers (SDAC), was most recently completed during 2022 only became available to researchers in July 2024.
As a result, we could not include this data in this report.
We recommend that future reporting for the outcomes framework consider the timing of data releases.
One option could be for the Victorian Government to maintain data access to the main data collections required for this work, like the PLIDA, the NDDA and NDIS unit record data.
If researchers are then commissioned to help analyse results and produce reports, arrangements should be made to facilitate their access to the data via an amendment to the data sharing arrangements.
Updated