Recording of the Stats NZ and Whaikaha webinar is now available
On Wednesday, 9 April Stats NZ and Whaikaha discussed insights from the Household Disability Survey 2023 and what it means for disabled people.
You can view the recording (with the option of closed captions) on YouTube external URL and a transcript is available.
More information is available on the Stats NZ website. external URL
We received more questions than we had the time to answer during the webinar. In the time since, we have worked with Stats NZ to provide responses.
Just wondering how you account for the discrepancy between previous data sets that showed 1 in 4 people as being disabled in NZ and it now being 1 in 6? Thanks
The way in which disabled people were identified in the 2023 Household Disability Survey (HDS) is different to how they were identified in previous disability surveys run by Stats NZ including the 2013 NZ Disability Survey which gave the ‘1 in 4’ estimate. The threshold for counting someone as disabled in the 2023 HDS was set higher than in previous disability surveys. Household Disability Survey 2023 – findings, definitions, and design summary external URL has more information; see section 10.2.
Additionally, two surveys were run in 2013: the HDS and the Disability Survey of Residential Facilities (DSRF) (these were jointly referred to as the NZ Disability Survey). The group of people living in residential facilities is small but they have a higher disability rate. The 2013 DSRF increased the overall disability rate by one percentage point (from 23 percent to 24 percent). In 2023, only the HDS was run.
Several questions on people living in residential care which have been grouped together:
- So does this mean that 1% of disabled people who are in residential care where not included in this survey?
- Are you saying people living in residential care are not part of a household?
- You mention that disabled people living in Residential care were NOT included in the 2023 HDS, but were they included in the 2013 disability survey?
People living in residential care were not included in the 2023 HDS. In this context, ‘residential care’ refers to aged-care facilities, long-stay-hospitals, and hospices.
It does not refer to community-based supported living (such as the homes run by IDEA Services) – people in community-based supported living were included in the 2023 HDS.
In 2013, people living in residential care were surveyed in the DSRF which ran alongside the HDS. However, the DSRF was not run in 2023 for several reasons which are explained here: Household Disability Survey 2023: Final content | Stats NZ external URL .
In terms of disabled people living in Residential Care not being counted - what about aged and disabled people living in aged care (Rest Homes and Retirement Villages), were they surveyed?
People living in aged-care facilities/rest homes were not included in the 2023 HDS. Most HDS questions are not relevant for people living in aged-care facilities. In previous years, people in aged-care facilities were surveyed in the DSRF which was a much shorter version of the HDS. However, the DSRF was not run in 2013 (for reasons explained here: Household Disability Survey 2023: Final content | Stats NZ external URL ).
People in retirement villages were included in the 2023 HDS.
Did this survey go to residential communities such as IDEA Services, or similar residential providers and also supported living?
Yes, people living in community-based supported living, such as the homes run by IDEA Services, were included in the 2023 HDS.
Did you survey the people who were in transitional housing, emergency housing etc?
People living in emergency housing such as motels or boarding houses were not included because these groups are harder to reach. This exclusion of people in emergency housing is consistent across all social surveys run by Stats NZ.
So not all populations have been included in the survey but you have somehow concluded that 1 in 6 people in New Zealand are disabled - and have been leading with this statement?
All estimates from the HDS are representative of people living in NZ households only. A household refers to people living in either private dwellings or ‘residential and community care facilities’ (community-based supported living). This information is included in all Stats NZ and Ministry of Disabled People - Whaikaha releases.
How could future disability surveys better capture those who are Autistic? This is highly relevant to provision of services and to better identify current prevalence? The Washington Short Set is inadequate and leaves out many. I'm aware one additional question was included but still inadequate.
The 2023 HDS used a much longer question set than the Washington Group Short Set, based on the Washington Group Extended Set, plus additional questions added by Stats NZ.
The HDS used a functional definition of disability, whereby a person was considered disabled if they had significant difficulty with at least one area of human functioning. Neurodivergent people were not considered disabled purely on the basis of being neurodivergent; they had to have functional difficulties that meet the survey’s criteria.
The HDS asked about several types of functional difficulties that may be linked to neurodivergence:
- For adults: learning, communicating, remembering, concentrating, socialising, and anxiety
- For children: learning, communicating, remembering, concentrating, anxiety, controlling own behaviour, making friends, and accepting changes to their routine.
70 percent of people diagnosed with autism met the HDS’s definition of disabled.
Why is autism not automatically counted as a disability yet it is an accepted diagnosis for Disability Support Services and this data set could potentially influence decisions around this?
As above, a person being diagnosed with autism did not mean they were automatically counted as disabled in the HDS; they needed to have functional difficulties. Not all people diagnosed with autism are eligible for government-funded Disability Support Services (DSS). An autistic person would need to complete a needs assessment first. Some autistic people, as a result of this needs assessment, will not quality for DSS.
Being disabled is not about "functional difficulties” it is about the "barriers we face" full stop. Children only have mothers, fathers or support/staff. How will Stats NZ get rid of the Medical Model of Disability language for the next survey?
Stats NZ uses a functional definition of disability, and this is different from both the medical model of disability and the social model.
A medical model would ask what conditions or health problems a person has been diagnosed with whereas the HDS asked how much difficulty they had performing aspects of human functioning. The functional approach identifies people who are at greater risk than the general population of social exclusion (for example, not able to access education or employment) if their environment was unaccommodating.
The social model says people are disabled by barriers in society, not by their impairment. The HDS asked about the barriers faced by disabled people, but these questions were not used to classify someone as disabled.
This article, written by the Washington Group, explains why it is not suitable to identify disabled people in quantitative surveys (such as the HDS) by asking about barriers: How Are The Washington Group Questions Consistent With The Social Model Of Disability? - The Washington Group on Disability Statistics external URL .
Stats NZ is not aware of any social model question sets that could be used to identify disabled people in the HDS while still producing reliable data and allowing the HDS to meet its objectives.
Were any adults interviewed who were non-verbal with an intellectual disability, and where a welfare guardian was able to answer questions on their behalf?
Stats NZ can’t give information about specific individuals in the survey. However, the general protocol for any individual who found the interview difficult was to first attempt to complete the interview with the individual themselves and a support person of the individual’s choosing. If this wasn’t possible because the individual was unable to participate at all, Stats NZ asked that a person answer on their behalf (usually this would be individual’s primary carer).
How appropriate is it for Stats NZ to use the WG questions to determine prevalence - given that this is not what this instrument was designed for? Is there a better method for determining prevalence that Stats NZ is considering adopting?
The Washington Group was set up by the United Nations and they work on developing methods to improve statistics on disabled people globally. The Washington Group’s work has been tested extensively in many countries and cultural contexts. Given NZ’s multicultural society, cross-cultural suitability was an important requirement for the HDS.
It’s correct that the Washington Group Short Set (WGSS) is not suitable for disability prevalence. However, the HDS used a much longer question set based on the Washington Group Extended Set for adults, and the Child Functioning Module for children. These question sets are suitable for prevalence, and Stats NZ added additional questions to them to make the HDS more inclusive.
Do the other surveys mentioned (HLFS etc) also use this "a lot of difficulty" threshold for who counts as disabled?
All Stats NZ surveys use same threshold to identify disabled people (“a lot of difficulty” or “cannot do at all”). This is the threshold recommended by the Washington Group.
Were any disabled people involved in the decision to use the WG questions vs another set? And what impacts have Stats NZ considered on, in effect, now telling disabled people they are statistically not considered disabled anymore? And what impact does this have on any social services for disabled people who are not 'considered' disabled?
Stats NZ ran a consultation process in 2021 to determine what the 2023 HDS should ask about. Disabled people and people involved with the community (such as service providers or carers) were encouraged to make a submission, and the consultation was promoted extensively via disability networks. Stats NZ also met with some disability groups as part of the consultation process. There was some feedback received that the Washington Group questions are not viewed favourably by some in the disabled community, however, no viable alternatives for measuring disability prevalence in a statistically robust way were able to be identified. Stats NZ uses questions based on those developed by the Washington Group on Disability Statistics. This group specialises in disability measurement and their work has been tested extensively in many different countries and cultural contexts.
What is the end goal for government in using the Washington group - to justify more budget costs towards Deaf and Disabled people?
Disability data is used by government to plan services to better support Deaf and disabled people.
Why are there only binary genders represented?
In the 2023 HDS, people were able to report a gender other than male or female, so the survey did collect information about gender diverse people. However, Stats NZ have concerns about the quality of this data due to the very small number of gender diverse people in the sample. Because of these concerns, Stats NZ did not publish any estimates for gender diverse people on its website, nor were estimates for gender diverse people included in the Webinar.
If data on gender diverse disabled people is needed, this might be possible as a customised request (noting the data will be low quality). Contact Stats NZ to discuss.
In your result, there were higher percentages of disability among LGBTQIA+ community. When the prevalence is measured by binary sex, will it miss people who are gender diverse?
Although a prevalence estimate for gender diverse people was not published by Stats NZ (see previous question), gender diverse people were included in the survey and are included in the estimates, including the prevalence rate of 17 percent.
Why wasn't the 15-44 year group broken down further as it would be very useful to get data on adolescents and young adults as they transition to adulthood?
The HDS is a sample survey and sample survey data about small population groups (like disabled young adults) can be subject to high sampling error. Stats NZ used the age group 15–44 years to avoid small sample sizes and poor-quality data.
The only data published on the Stats NZ website for young adults was the disability prevalence rate: 13 percent of people aged 10–19 years were identified as disabled and 14 percent of people aged 20–29 years. If additional data on disabled young people is needed, this might be possible as a customised request (although the data might be low quality). Contact Stats NZ to discuss.
Could this disability data be done regionally? For local councils to use
The disability prevalence rate is available for each region on the Stats NZ website. If additional regional data is needed, this might be possible as a customised request (although the data might be low quality). Contact Stats NZ to discuss.
I just don't understand how we can be comparing the findings of this survey with the findings of the previous considering they use two different methodologies?
Due to changes in the way disabled people were identified, results from the 2023 HDS are not comparable with any previous disability surveys run by Stats NZ. Stats NZ considers these changes to result in a more useful and more current measure of disability than previous NZ disability surveys.
Results from the 2023 HDS are available on the Stats NZ website and are accompanied by a message stating that they can’t be compared with previous results.
Are you intending to do a disability specific survey, so we can get a true comparison for the 2013 findings?
The 2023 HDS was a disability-specific survey. Stats NZ has no plans to return to the methods used in the 2013 Disability Survey.
Why did the survey "cherry pick" neurodiversity? What about dyslexia and FASD for example?
When Stats NZ ran a public consultation to determine what the HDS should cover, a large number of people requested that the survey ask about autism and ADHD. Because these two conditions were requested so heavily, they were included. But, as is the case for all surveys, there was only a finite number of questions that could be included in the HDS due to respondent burden and interview resources . Adding more questions to ask about other diagnoses would mean other questions (i.e. about the experiences of disabled people) would have had to be removed. Judgements on what to include in the survey were based on a) the survey’s objectives and b) how requested the data was by the community.
Deafness differs and is vast and individualistic too … were the hard of hearing excluded? Like the manner of those in autism and ADHD?
The 2023 HDS asked how much difficulty a person had hearing [if applicable: while using a hearing device]. If a person who was hard of hearing reported, “a lot of difficulty hearing” or “cannot hear at all”, they were counted as disabled in the HDS.
Do you have data on Deaf people?
Stats NZ does not have data on Deaf people. The 2023 HDS collected data on people who had either “a lot of difficulty” hearing or who “couldn’t hear at all” but we are unable to separate the two due to small sample sizes.
Is there additional context to the stats? What we are seeing are different views of different topics (i.e. one slide break down male/female i.e. binary). Another slide talks about LGBTQIA+ as a 'total' community, other slides talk about Māori, Pacific, European etc. It's hard to see how these stats make sense when there isn't an intersectional view on all the slides
Intersectionality in a sample survey can be difficult because the sample size can limit the reliability of estimates about small groups within a population. Stats NZ has published some intersectional disability prevalence data (e.g., disability rates by ethnic group and male/female, disability rates by ethnic group and age group). If additional intersectional data is needed, this might be possible as a customised request (although the data might be low quality). Contact Stats NZ to discuss.
You collect data for Māori and Pasifika yet you don’t include an analysis of equity and unmet need. Why?
This article has more information on unmet need, including the rate of unmet need for Māori and Pacific disabled people: Disabled people need more care and support external URL .
Would we be able to have links to those surveys
The full HDS results are available here: Disability statistics: 2023 | Stats NZ external URL .
Do you have much data around wellbeing and leisure?
Data on wellbeing and leisure is available here: Disability statistics: 2023 | Stats NZ external URL .
Under Download data, please download the Excel spreadsheet called “Disability statistics: 2023 – experiences and outcomes of disabled people – updated”. Table 9 has data on leisure and table 10 has data on wellbeing.
I may have missed it, but is there stats on employment
Employment data is available here: Disability statistics: 2023 | Stats NZ external URL .
Under Download data, please download the Excel spreadsheet called “Disability statistics: 2023 – experiences and outcomes of disabled people – updated”. Table 6.1 and 6.2 have data on employment.
ASD has sharpy risen across the globe. CDC in the U.S report one in 36 children. Will Stats NZ be reporting on NZ and if it’s increasing here? Will Stats NZ be capturing data on severity also? ASD severity 1, 2 and 3?
Information on autism is not collected in any Stats NZ survey other than the Household Disability Survey. The HDS runs every ten years so frequently updated data on autism is not held by Stats NZ.
However the NZ Health Survey, run by the Ministry of Health, collects data on autism diagnoses in children and this survey has annual updates: Ministry of Health | NZ Health Survey explore indicators external URL .
Neither Stats NZ nor the NZ Health Survey have information on autism severity.