Review of 2021 Census Topics

Census Need for Assistance
It is essential that the disability topic continues to be part of Censuses going forward. The ability to disaggregate data by disability (and especially by people who need assistance with basic daily activities) is essential to monitor and evaluate major national reforms to disability policy and service delivery.
Expand the output
The AIHW proposes that, if the current disability measure (the ‘Core activity need for assistance’ topic) is maintained as is, that the ABS expand the output.
The ‘Core activity need for assistance’ topic has been included since the 2006 Census to measure the number of people needing help in one or more of the three activity areas of self-care, mobility and communication (so-called ‘core’ activities). The output for the ‘Core activity need for assistance’ topic is a derived total number of people needing help with core activities (‘Core Activity Need for Assistance’ variable), and does not detail which of the three activities resulted in the need for assistance. Modifying the current output to detail each individual activity would allow separate identification of the number of people needing help with each activity.
Expand the topic coverage
In light of recent major reforms to disability policy and service delivery, the AIHW suggests the ABS consider expanding the need for assistance topic.
The Census questions about core activity need for assistance were designed to be conceptually comparable to questions about profound or severe core activity limitation in the ABS Survey of Disability, Ageing and Carers (SDAC). This does not cover all ‘life domains’ as conceived in the International Classification of Functioning, Disability and Health (ICF), and it does not cover people with milder forms of disability.
Expanding the breadth of activities (to go beyond the simply the core activities) along with expanding the depth of the level of need categories (to enable enumeration of ‘disability status’ beyond profound or severe) would allow for a more detailed assessment of a person’s level of functioning and need for support in everyday activities. For example, something like the below matrix-type approach could be used within a Census to capture a wider disability population. Such a set of questions is broadly consistent with the short disability questions that the ABS uses in a number of its social surveys and is grounded in the ICF. This matrix could be followed by additional questions about specific education and employment restrictions to ensure comprehensive coverage of the ICF domains. The matrix below was designed by the AIHW for the Standardised Disability Flag, intended for broad application in a variety of settings.



An expansion of the need for assistance topic is not intended to replace the highly valued ABS SDAC, which collects more detailed information on people with disability, much like other ABS surveys collect more detailed information on other key population groups than is available in the Census.
If the Census need for assistance topic is to be changed, the AIHW does not support the Washington Group questions as a replacement for this topic (see ‘Additional comments’ section for detail).

Data on people with disability inform the development of federal and state/territory government policies and community-based programs, including both disability-specific and mainstream services. In particular, major reforms to disability policy and service delivery over the next few years, such as the completion of the national roll-out of the National Disability Insurance Scheme (NDIS) and the release of the new National Disability Strategy (NDS), will need to be monitored and evaluated using comprehensive and high quality data about people with disability, especially people who need assistance with basic daily activities. If the topic was not included in the 2021 Census, there would be no alternative data sources that could be used for such important national measures and indicators, particularly for groups not well enumerated elsewhere.
In particular, the ability to articulate a broader story—with reforms such as the NDIS included in the context of the wider disability population—will be critical to governments’ understanding of the costs of, and pressure on, key systems and supports, as well as of the impact of these reforms on the experiences of, and outcomes for, the wider population of people with disability (not just those with severe or profound core activity limitation). This cannot be achieved without obtaining good estimates of a wider population of people with disability, both those who are eligible for the NDIS (including the subset who, for whatever reason, do not use the NDIS) and those who are not eligible (such as many with mild or moderate disability). The current Census core activity need for assistance topic does not enumerate all of this population.

Detailed data on the need for assistance topic are highly policy relevant and essential for estimates of disability and the potential population for disability services for small geographic areas and small population groups, such as Aboriginal and Torres Strait Islander peoples, people from culturally and linguistically diverse backgrounds, and the population in a particular small area. These groups are either not well enumerated in other key data sources (such as the ABS SDAC), or output from those sources is precluded because of reliability concerns.

An expanded need for assistance topic can be accurately collected in a form which the household completes themselves. A similar approach to that proposed for expanding the need for assistance topic is currently used in the AIHW’s Standardised Disability Flag for mainstream data collections. This flag is designed for use in collections covering a wide range of sectors, many of which have little experience with disability. To ensure that it could be accurately and effectively collected by people with limited exposure to this topic, the AIHW consulted with experts in disability concepts and measurement during the development of the Flag. Draft versions were then tested and refined through a series of focus groups, cognitive interviews and pilot testing. Pilot testing confirmed that the Standardised Disability Flag module was suitable for completion by the general public in a range of settings. Each of these development components also included input from people with lived experience of disability.
Versions of the AIHW need for assistance matrix have been successfully implemented in the AIHW’s Specialist Homelessness Services Collection and National Prisoner Health Data Collection, and are in the process of being implemented within other AIHW collections.
While it is acknowledged that an expanded matrix will add to the respondent burden, this is justified by the vastly improved depth and breadth of the collected data.

The need for assistance topic has been included in a number of Censuses and appears to be well accepted. More comprehensive measurement of health, functioning and disability have been used in ABS surveys for many decades, and appear to be well tolerated.
The topic of disability and disability support services engenders strong community interest generally, and typically has bipartisan support when considered by the Australian Parliament.

The expanded matrix as designed for the Standardised Disability Flag can be presented efficiently on a printed page. However, this was designed for a single respondent, so some consideration of how to format this effectively for multiple household respondents would need to occur. (For the online form, this is not so much an issue because the matrix can simply be served up for each respondent.)
The coding has been designed into the question, so there is a simple mapping of the response categories to the output data items.
The AIHW would suggest that for the topic to function well—to accurately measure disability as described by the ICF—there should be a short preamble to explain that the need for assistance is in the context of a health condition.
The number of questions and response categories is no more than required to fully cover the activity domains classified in the ICF.

The need for information on people with disability at a population level, especially for small population groups and small areas, will become increasingly important as major national reforms to disability policy and service delivery are rolled out. There will be a need to evaluate the impact of these reforms over time.

The ABS SDAC is the gold standard for measuring disability prevalence. However, the coverage of the SDAC does not include people living in very remote areas and discrete Aboriginal and Torres Strait Islander communities. Further, the SDAC does not support direct estimates at a small area level, or for small population groups of policy interest. The Census need for assistance topic allows analysis of the characteristics of these important groups.

The AIHW is aware that various stakeholders have advocated to include the Washington Group questions on disability. The AIHW does not support the inclusion of these questions in the Census. Factors to consider are:
• The resulting data are not comparable with other Australian data from the ABS SDAC or surveys which use the ABS short disability module in terms of scope, conceptual definition, severity scales, and the levels and types of disability. This will in turn create confusion in defining disability prevalence. In the current environment, it is critical that a definition of people with disability is consistent and transparent.
• The data are inadequate to identify the potential target population for disability services: the current Census questions provide information about the number of people who need help with basic daily activities. These people have been considered as the main potential target population of disability-related services. The four categories of response thresholds/scales for the Washington Group questions (No—no difficulty; Yes—some difficulty; Yes—a lot of difficulty; Cannot do at all) do not provide information about whether people who have ‘some difficulty’ or ‘a lot of difficulty’ need help with basic daily activities; and whether people who ‘cannot do at all’ will be able to do basic daily activities with help. Consequently, the information gathered using these questions would not meet current service planning and policy development requirements.
• The Washington Group’s questions are designed for a different purpose—they are useful for countries without a comprehensive disability survey (such as the SDAC) or for a component of surveys that are not specifically designed for disability collections.
• Non-inclusion of the short set questions in the Australian Census would not compromise Australian’s ability to report—data for international comparisons using the Washington Group questions may be derived from the SDAC, or potentially the Census if the breadth and depth of the topic is expanded.