Disability Identification - the Australian Experience

Jennie Widdowson
Director, Welfare section
Australian Bureau of Statistics

Abstract The Australian government was interested in identifying people with disabilities as early as 1911. As time goes by, with greater awareness of the community and insights from other countries, the way and categories for identification have been changed. In this paper, it does not only attempt to draw the Australian experience by listing out the different means of identification, but also examines the interacting factors contributing to results of the identification and the impairment trends in Australia.

Governments in many parts of the world have realised the need to be able to identify people with a disability so that policies and programs can be developed to assist them. In Australia, the International Year of Disabled Persons in 1981 served to increase awareness of the needs of people with a disability. Since then there has been an increasing focus on this population group, especially since their inclusion as one of the designated Equal Employment Opportunity groups. More recently,the Disability Services Act of 1986 and the Disability Discrimination Act of 1992 have further highlighted the need to understand the size of this group and to recognise their rights and support needs. The issue of identification of people with disabilities was also seen as important enough for the Council of Australian Government to set up the Commonwealth State Disability Strategy in 1991. This is making sure that standard legislations and entitlements are available in all Australian States and the Commonwealth and dividing the provision of services to people with a disability between State and Commonwealth responsibility.

Historical Identification

As long ago as the 1911 Census, the Australian government was interested in identifying people with particular disabilities. In this Census people were asked in terminology which was relevant at that time to state if they were blind or deaf and dumb. In the notes that accompanied the Census it was explained that only those people who were totally unable to see should be entered as blind and only those who were dumb as well as deaf should be included in the deaf and dumb category. The notes also stated that "care should be taken that no such cases go unrecorded, as the possession of reliable information on these points is of no little value and importance in the interests of the unfortunate persons who are so afflicted." This seemed to imply that government had a special interest in this group of people.

The 1911 Census identified 3,112 people who were blind and 1,826 who were deaf and dumb, which gave a rate of 1.1 people per 1,000 population for the combined group or 1.8 per 1,000 when standardised to the 1993 population for comparative purposes. In the 1993 Australian disability survey, 0.8 people per 1,000 population would have been identified with these impairments. However, the rate of disability was reported from the 1993 survey as 180 per 1,000 population due to the broader range of impairments which were included.

Census vs Survey Identification

There are no standard methods to measure disability. Worldwide rates of disability have varied from 2 persons per 1,000 population (in Peru in 1981) to 209 per 1,000 population (in Austria in 1976). This variation does not necessarily show a real variation in the prevalence of disability but rather a variation in disability identification in much the same way as the difference shown between the 1911 Census and current rates. It is difficult to make any comparison in prevalence rates across the multitude of census questions used to identify disability and the differing sets of screen questions and methodologies used in household surveys. There are, however, some general conclusions that can be made. All census questions appear to identify lower rates of people with disabilities than household surveys. This is partly due to the nature of census questions which must be relatively simple in concept and must rely on everyone in the population being able to easily understand them and answer them in the same way. Household surveys, on the other hand, are interviewers who can explain the questions and prompt for answers.

Not only is there a problem with the way people understand census questions, but also the recognition or admission of disability by the person responding is not always consistent. Some people do not feel that they have a disability even though their disability may be obvious to an observer. In other cases people may realise they have a restriction of some functions but feel that they are no different from their peers and so fail to identify themselves. In addition, some people feel that they should fit in category described by a census question on disability, whereas the more rigorous definitions and different styles of questions of a household survey would exclude them. It is therefore necessary to try to identify disability by using measures which are not subjective or dependant on a person's view. Canada has both census and survey rates available for comparison, while Australia has the results from a census test and its household survey (see table 1).

It has also been observed, even in population surveys,that disability rates in more developed countries tend to be higher than in other countries. This may also be due to the difference in interpretation of questions relevant to a particular environment and culture. For example, the Philippines which uses similar screening questions to Australia has recorded a disability rate of 44 per 1,000 population.

Identification in Household Surveys

Disability is defined in the current version of the International Classification of Impairments, Disabilities and Handicaps (ICIDH) as "any restriction or lack of ability to perform an activity in a within a range considered normal for a human being." It is generally recognised that people who have any functional limitation or restriction in activity have a disability. However, disability tends to be a continuum where there is no absolute measure that can be universally used to decide whether a person has a disability.

In Australia, a number of questions concerning impairments whcih have either implicit or explicit restrictions connected to them are used to operationalise the identification of disability by the Australian Bureau of Statistics (ABS) in household surveys. Detailed testing in 1980 was used to determine which questions were most representive of a broad spectrum of impairments which were likely to lead to disabilities and would produce reproducible results. Similar sets of screening questions are used in a number of other countries. Twelve screening questions were used in the 1981 and 1988 ABS surveys of disability while three more questions were added in 1993, to try to cover additional areas of impairments which may have been previously missed (or included people who might not have identified themselves in more specific areas of impairment). (see table 2)

The disability rate in 1981 was found to be 73.8 per 1,000 population while in 1988 it rose to 155.2 and to 180.2 in 1993. If the data are age standardised to account for changes in population structure and only the 12 identical screen questions were included for 1993, then the disability rate would have been 78.2 in 1981, 159.1 in 1988 and 165.6 in 1993. This means that the disability rate has changed very little in the five years between 1988 and 1993. However, the three additional questions in 1993 identified an additional 14.6 people with a disability per 1,000 population. This is one example of how sensitive the measured level of disability can be to the questions asked.

Loss of sight not corrected by glasses

Loss of hearing

Speech difficulties

Blackouts, fits or loss of consciousness

Slowness at learning or understanding

Incomplete use of arms or fingers

Incomplete use of feet or legs

Treatment for a nervous or an emotional condition

Restriction in physical activity or physical work

Disfigurement or deformity

Mental illness requiring help or supervision

Long term treatment of medication for a condition , and still restricted

* Difficulty gripping or holding things

* Long term restricting effects of head injury , stroke or other brain damage

* Any other long term condition resulting in a restriction

* Only included in 1993 survey

Canada uses screening questions that are based on a person's abilities and support needs to identify its disability population. This has resulted in a disability rate of 132 per 1,000 for Canada. However, if Australia used similar questions for their identifications, then the rate in Australia would be 135 per 1,000 population (similar to our handicap rate of 142 per 1,000), again showing the effect that the method used for identification can have on the resulting rate.

Trends in Impairments 1981-1993

The screening questions which are used to determine disability cover a broad spectrum of impairments and restrictions. While the disability rate derived from these questions changed only slightly between 1981 and 1993, there appeared to be a rapid increase in the disability rate between 1981 and 1988. As the questions used were identical for the the 1981 and 1988 surveys, age standardisation does not account for the differences. It has previously been assumed that people were more likely to identify themselves as having an impairment in 1988. This may have been due to the greater acceptance of disability in the Australian community resulting from the education and advocacy programs put in place following the international Year of Disabled People in 1981. However, the underlying trends shown by the impairments and restrictions used as screening questions for the disability surveys show quite diverse patterns and not the expected consistent increase in identification. The population identified with a disability is the result of the interaction of all these patterns (see table 3)

The trends evident in the impairments can be grouped. Rates for loss sight have changed little over the last 12 years, while the rates for those who identified difficulties using arms and fingers, nervous or emotional conditions or deformities appear to have decreased slightly over the same period. There are two areas of impairment which might have been expected to have been sensitive to identification and, therefore, less well identified in 1981. Rates of identification of people with speech difficulties and those who are slow at learning have both shown steep rises since 1981. Either there has been a real rise in these two impairments or there has been a steadily increasing ability for people with these impairments to identify themselves.

Two impairments which showed little change in rate between 1981 and 1988 but a noticeable rise from 1988 to 1993 were for those who were identified by loss of hearing and those who had fits or blackouts. The more recent increase in rate may be due to the effects of occupational health and safety education which has made people more aware of the need to identify these impairments. The increased loss of hearing rate may also reflect increased exposure to noise in the workplace, environment (e.g. aircraft noise) and home (entertainment systems).

Four screen questions (difficulties using feet and legs, restrictions in physical activity, long-term treatment for a condition and mental illness needing supervision) all showed large increase in rates from 1981 to 1988. However, they also all show an anomalous pattern where the 1988 rate is higher than that for 1993 which may be a genuine pattern of change for these impairments and restrictions. On the other hand, these four screening questions are all broad in their interpretations of restriction and encompass a great variety of underlying conditions. This may, therefore, have led to the inconsistent patterns shown by the rates. Another explanation may be that the true rates for these restrictions may have actually fallen between 1981 and 1993 due to improved management and lifestyle, but the self-identification in 1981 was low.

Conclusion

Although it may be expected that there has been consistently better identification of disability since 1981 because of the greater acceptance of disability in the community, this does not seem to be wholly the case. There are a number of paradoxical results for which further research will be necessary to analyse the trends in more details and to determine other possible interaction factors. It may be that results from the next disability survey in 1998 will also help throw some light on the complexities of these impairment trends. However, it can be seen from the example in this paper that disability identification is complex and subject to many interacting factors and is very sensitive to the type of method used.

ASIA & PACIFIC JOURNAL ON DISABILITY

Vol.1, No. 1, September 1997

Published by the Asia and Pacific Regional Committee of Rehabilitation International (RI) and the Regional NGO Network (RNN)