19 April, 2023

Innovating democratic lotteries: The Canberra Forum

The Canberra Forum was the first of what we hope will be a series of events helping to bridge the gap between elected politicians and the people they represent. Research by the Museum of Australian Democracy (MOAD) showed that while most Australians support an increase in deliberative democracy, politicians, as a group, were sceptical about its usefulness. Our partners, DemocracyCo, saw an opportunity to help drive democratic innovation by demonstrating to politicians the usefulness of deliberation in the one area that they all have a stake regardless of their party affiliation - relations with their constituents.

On 18 November 2022, we handed over the details of 40 people living in the electorate of Canberra to DemocracyCo. These 40 people will be part of the Canberra Forum, working together to both set an agenda for their discussions and come up with recommendations for their Member of Parliament, Alicia Payne. They will be meeting for six sessions, four face-to-face and two online, prompted by the question:

What are the most important problems or opportunities for Canberrans and how could they be addressed?

You can find more information and follow the proceedings here.

Recruitment was done following our standard two-stage sortition template, aligning with the OECD’s good practice principles.


Stage One

For the random selection of addresses, we usually use a postal database. For this job, however, working with the office of Alicia Payne, we used the electoral roll, allowing us to send out individually addressed invitations. However, due to privacy considerations, we used a proxy database with no identifying data to run the selection, which was then matched to the original data by the Member’s office.

To make sure that we were able to reach young people who were not eligible to register to vote, we emphasised that the invitation applied to anyone in the household and that we particularly encouraged young people to register. This appeal worked, and we had a high percentage of young people registering, as you can see from the pie charts below.

The invitation card was designed by Alicia Payne’s office, and so is different, in look and feel, from our usual invitation package. It included a link to register and a link to a survey, available to all Canberrans, which will be used to help set the agenda for the sessions.

Unusually, no remuneration was offered to participate in this event. Nonetheless, overall we had a good response rate, but this could have affected the responses in some groups as discussed below. Overall, we sent out 8,000 invitations and received 223 responses.


Stage Two

As part of the sign-up procedure all potential participants were required to share some basic information about themselves: their address, date of birth, ancestry and qualifications.The questions on ancestry and qualifications followed the census questions, and we were able to use the 2021 ABS Census data for the first time. In contrast to the census, however, we narrowed the options for ancestry to “Aboriginal or Torres Strait Islander; Asian; European (British); European (Other); Mixed; Other”. The census allows people to choose multiple options, so does not allow the option of ‘Mixed’ - we thus assigned a proportional number of people to the ‘Mixed’ category (9.9%).

We then used this information to input into a “sortition algorithm”; randomly selecting 40 people from the pool of 223 potential participants. The algorithm we use has been mathematically proven to be the fairest possible

The selection data can be summarised by the following pie charts:

The way to understand these pie charts is as follows:

  • Column 1 (Target): These pie charts give information about the Canberra’s population as a whole, using primarily information from the 2021 ABS Census. The information on geography was calculated with assistance from Alicia Payne’s office. 
  • Column 2 (Respondents): These pie charts summarise the information that was provided to us by the 223 people who signed up as potential participants. There is some skewing in statistics here compared with our target: for instance, we found that there were comparatively fewer people with ‘No Qualifications’ than in the target population. This result was unsurprising, given that there was no remuneration for the process.
  • Column 3 ( Confirmed Selected): These pie charts summarise information about the 40 people who were finally confirmed to participate in the assembly. Notice that, thanks to our use of a sortition algorithm, the pie charts in this column are similar to the target charts in the first column. As part of our recruitment process all of these people were contacted by telephone to confirm that they were still willing and able to participate -- in the event that this was not the case, we used the sortition algorithm to replace people who dropped out with others who shared similar characteristics. You may notice that the 16-24 age group was over-represented in the final sample: there was a large overlap between this category and those with no qualifications, so, in consultation with DemocracyCo, we made the decision to increase the target number of people recruited in this category. When we make the selection, we give each target a narrow band of people to choose. Thus, for example, the target for 60-69 year olds was a minimum of six people, and a maximum of seven, given the percentage in the population for a sample of 40 would be ‘6.9’. We were thus able to go to the lower target for other populations, and keep the sample reasonably representative beyond this skewing.

A note about the gender pie charts is important: you will see that 4% of respondents had a gender identity other than "male" or "female". Since the ABS only allows two gender options in the census, our target pie chart only has values for "male" and "female". To ensure that this 4% of respondents were treated fairly we manually adjusted the algorithm so that this 4% of respondents had the same likelihood of being selected as the remaining 96%. On this occasion, our random selection resulted in a large number of these people being selected, primarily because the overlap with ‘no qualifications’ was high. We very much regard this method of operating as a stop-gap measure -- our hope is that the ABS will soon reflect Australia’s gender diversity in its questions.


We use cookies on our websites. You are free to manage this via your browser setting at any time.