- Survey frame
- Completing the survey
- Inclusion Strategies
- Silent electors
- Aboriginal and Torres Strait Islander Communities, Rural and Remote Australians
- Australians overseas
- People with a disability
- People from culturally and linguistically diverse backgrounds
- People experiencing homelessness
- People in aged care facilities
- AMLPS Information Line and Online Enquiries
- Drop off strategy
- Capturing and processing your response
- Generating the survey results and ensuring anonymity
- Securing survey responses
On 9 August 2017 the Australian Treasurer directed the Australian Statistician to request statistical information from all Australians on the Commonwealth Electoral Roll, on whether or not the law should be changed to allow same-sex couples to marry. On 16 August 2017, the Minister for Finance, acting on behalf of the Treasurer, amended the direction to the Australian Statistician to confirm that Australians 18 years and older on the electoral roll as at 24 August would be included in the survey.
The Australian Marriage Law Postal Survey (AMLPS) was conducted by the Australian Bureau of Statistics (ABS) in response to this direction.
All people enrolled on the Commonwealth Electoral Roll (or those who had made a valid request for enrolment) by 24 August 2017, and being 18 years of age, and not serving a prison sentence of three years or longer were eligible to participate in the AMLPS.
Participation in the AMLPS was voluntary.
To administer the AMLPS, and ensure integrity of the responses received, the ABS created a statistical survey frame (hereafter referred to as the "frame"). The Commonwealth Electoral Roll (as at 24th August 2017) formed the basis of this frame, providing details of eligible electors (including names and addresses, date of birth and gender).
The frame was used to provide name and address for each eligible Australian to the print/mail house where survey materials were created and mailed. Date of birth and gender information from the frame was used to provide the demographic breakdown of participation information only.
The AMLPS collection period ran from 12 September 2017 (when the mail-out of survey forms began), to 7 November 2017.
Eligible Australians were able to provide a new address, or request replacement forms, via the Australian Marriage Law Survey website or the survey Information Line.
Nearly all eligible Australian electors received their survey materials by post. Survey materials included a survey form, a reply paid envelope and a letter with instructions on how to complete the survey form. The survey asked one question: should the law be changed to allow same-sex couples to marry? Respondents were asked to indicate either yes or no to the one question by marking the appropriate response box, and then return their survey form using the reply paid envelope.
Survey responses could also be provided via an online form and a telephony service. These options were available for eligible Australians who:
- were overseas throughout the collection period;
- had blindness, low vision or other disability that made the paper form a more difficult option; or
- were in a remote or other location throughout the collection periods where they could not reasonably access a form by post.
Eligible Australians in these categories were able to obtain a Secure Access Code (SAC) through the survey Information Line or the ABS website throughout the period 25 September 2017 to 20 October 2017.
The ABS engaged with a wide range of government departments and agencies, and external stakeholders to promote participation by all eligible Australians.
The AEC and ABS took measures to protect the privacy of silent electors and Australian Defence Force (ADF) and Australian Federal Police (AFP) personnel who were registered with the AEC as being deployed overseas. The AEC printed survey forms for these electors, with non-identifiable barcodes provided by the ABS (like with all other forms), using a different mail house to the one used by the ABS. The survey packs included an addressed letter from the Electoral Commissioner, and an unaddressed letter from the Australian Statistician. A designated telephone Hotline administered by the AEC was set up to assist these electors. Silent electors were also able to access replacement materials, and the online or telephony response service, where required. At no time was the ABS provided with the name, address, gender or age of any silent elector or registered ADF or AFP personnel. For this reason, the age and gender statistics of silent electors and registered ADF and AFP personnel are not included in the survey participation rates.
Survey forms were posted to all eligible participants' addresses within Australia including; Aboriginal and Torres Strait Islander communities, PO boxes and nominated identified mail addresses (shelters, hotels, workers camps, Australia Post offices (forwarded mail set up). In cases where there was no street address or household mail delivery, mail was held for collection at their local mail agent (like all other mail).
Information about the AMLPS was advertised through radio, print media, social media and television. Specific material was translated into twelve Aboriginal and Torres Strait Islander languages and was available on the ABS website and at regional and remote pick up locations. In addition to this in-language support, a trusted person process could be used by Aboriginal and Torres Strait islander persons to assist them to understand and participate in the survey.
While most residents of Aboriginal and Torres Strait communities used the paper form, people in remote areas were also given the option of participating online or by telephone.
ABS staff travelled to over two hundred locations across remote locations, regional centres and capital cities in order to provide form pick-up locations, where eligible Australians could collect or return forms, or receive assistance on the survey.
These staff were all experienced in remote work, most having undertaken similar field work through the Census of Population and Housing or elections. Staff received further specialist training for assisting Aboriginal and Torres Strait Islander peoples including cultural awareness training, engagement with communities and training in protocols and customs.
Separate to ABS pick-up locations, the ABS partnered with the Department of Human Services (DHS) to provide additional opportunities for people in remote locations to contact the ABS at one of their 600 access points across Australia. Each service point provided self-service computers/phones where people could freely access the survey Information Line and the ABS website. This allowed for a person to request new materials from the ABS and to participate in the survey using online or telephony options.
People who had an overseas postal address listed on the Electoral Roll were posted a letter containing a SAC to respond online or by phone. No paper forms were posted overseas. Australians temporarily overseas may have been able to complete their form before they left or after they returned. Forms were delivered to Australian nominated addresses by 25 September 2017, including people that notified the AEC that they would be temporarily overseas. Eligible Australians overseas for the entire collection period could also ask a trusted person to respond to the survey on their behalf. Otherwise, they were able to request a SAC and complete the survey online or by phone.
Support was available for people with an illness, injury or disability who found completing the survey difficult. The ABS worked in partnership with National Relay Service (NRS) to assist people, particularly those who have a hearing or speech impairment to complete the survey. People could also ask a trusted person (a family member or friend) to help them complete the survey, or to complete the survey on their behalf. People with an illness, injury or disability could also request a SAC to respond online or by phone. All survey materials and the ABS website were designed to maximise accessibility for people with vision, hearing or intellectual disabilities. The AMLPS Information Line was also available to provide assistance to people.
Disability Discrimination Commissioner, Alastair McEwin, commended the Australian Bureau of Statistics on their approach to ensure people with disabilities were included in this survey. He noted, "the Bureau pro-actively engaged with the sector in order to determine best practice supports", and that "appropriate adjustments were also made available within a relatively short period of time. In my view, this set the Bureau apart as a leader in implementing accessible alternatives on a national scale."
Eligible Australians from culturally and linguistically diverse backgrounds were able to complete the survey independently through the postal service with the assistance of a trusted person (a family member or friend) or through the support of Translating and Interpreting Services (TIS). There were instructions in 15 languages on the reverse side of the letter sent with the survey form on how to contact the TIS.
People experiencing homelessness who were unable to access their mail, were able to participate in the survey by collecting a survey form from an ABS pick-up location, requesting a survey form to be posted to an address of a trusted person (a family member or friend) to hold the form for them to collect or requesting a SAC and complete the survey online or by phone.
Survey materials were posted to aged care facility residents' addresses on the Electoral Roll or to an alternative address advised to the ABS. People in aged care facilities were able to complete the survey independently through the postal service with the assistance of a trusted person (a family member or friend) or by requesting a SAC and complete the survey on online or by phone.
The AMLPS Information Line was open from Monday 14 August until Friday 24 November. This service utilised a free call 1800 number for callers within Australia, and also a phone number that overseas callers could access. Between 14 August 2017 and 7 November 2017 the Information Line was open to the public from 8am to 8pm (local time) seven days a week. From 8 to 24 November 2017 the opening hours were reduced to 9am to 5pm (local time) Monday to Friday as a result of decreasing number of calls.
From 14 August 2017 until 7 November 2017 the Information Line received 206,828 calls.
The busiest day for the Information Line was 24 August 2017 (the close of the Electoral Roll) which resulted in 26,912 telephone enquiries.
The Information Line utilised the Translating and Interpreting Service for callers who needed language assistance. Between 14 August 2017 and 7 November 2017 there were 244 calls to the Information Line through this service.
From 25 August 2017, people were also able to submit enquiries through online forms. Between 25 August 2017 and 7 November 2017 there were 133,936 online enquiries.
Of these, 125,819 were requests for survey forms or secure access codes which were automatically processed and responded to. Overall, over 97% of online enquiries were responded to within 48 hours.
In the final weeks of the survey, ABS offices were set up as drop off points enabling members of the public to hand in their survey form before the 7 November deadline. The drop off locations were advertised on the Marriage Survey website and were promoted in the media. All forms received were stored securely until close of business on the 7th of November when they were securely transferred by an ABS officer to the nearest processing site. Nearly 1,000 forms were received at ABS offices across the country (with over 50% of these received on 7th November) and all forms received were processed and included in the final count.
Paper survey forms were mailed back to five scanning centres across Australia. At each centre, the forms were removed from their envelopes and scanned. The scanned images were securely uploaded to a central system where Optical Mark Recognition software was used to capture the barcode and the marks on the form in digital form. Where the marks on the form were simple (i.e. where one response box only was marked and no other marks were made on the form) the form was automatically coded to the appropriate response. Where the marks on the form were more complex (i.e. both response boxes were marked, or a combination of response boxes and other marks existed) the form was manually coded. Manual coding involved a human operator making a determination on the appropriate response. Coding staff followed a set of guidelines that were developed in consultation with Parliamentary Committees and the Australian Electoral Commission.
Quality of coding (please refer to Quality and Integrity Statement for detailed quality information)
- 12,056,512 (95.0%) of paper responses were automatically coded.
- 636,961 (5.0%) of paper responses were manually coded; 4.3% coded to a Yes or No, and 0.7% coded to Blank or Response not clear
- 24.0% of manually coded records were quality assured (separate to the External Observer process) and in 99.7% of cases the records were verified as correctly coded.
A total of 34,447 responses to the survey were received through online and telephony forms. Unlike paper survey forms, only a valid response could be given through online and telephone channels.
The barcode on the survey form was a single-use, anonymous code and provided the mechanism for ensuring that forms could not be counterfeited or duplicated. They also served to ensure that only one response was counted for each eligible person. If a person requested a new form, they were issued a new barcode and their original barcode was marked as invalid. If a person tried to submit more than one response using the same barcode, only one of the responses was counted.
Where barcodes were not able to be read (damaged or defaced) the form was not counted.
Once verified, a list of valid barcodes and their associated electoral divisions was sent from the ABS to the Processing site. The Processing site used this list to update each response record (Yes, No) with their associated electoral division. A file was then generated listing only responses (Yes, No) and electoral divisions, and no barcode, and sent back to the ABS. It was this method of operation that enabled the ABS to ensure all responses were both legitimate and completely confidential (refer to diagram below). At no time was it possible to connect the identity of the elector to their response, as the Processing site did not hold names and addresses and the ABS only received responses (Yes, No) associated with electoral divisions.
Results from the Australian Marriage Law Postal Survey have been produced showing;
- the number of Yes, No and unclear responses and their percentages at the National, State/Territory and Federal Electoral Division levels, and
- the level of survey participation classified by age, gender (both numbers and percentages) at the National, State/Territory and Federal Electoral Division levels.
Survey forms were returned through Australia Post's Priority mail service directly to each of the five scanning sites. At each site, survey forms were protected throughout the various stages of handling (including storage, opening the mail and extracting the forms, scanning and ultimately destruction) by 24/7 security guards. No unauthorised access was attempted.
Once the forms were scanned, the images and coded responses were stored in encrypted files. The ABS securely and separately held the keys to decrypt these files. The ABS worked with relevant government security advisors to implement high quality data protection measures to ensure the privacy and integrity of the data was maintained.
All survey forms, both physical and digital, will be securely destroyed within 60 days of the release of results.