Citizens' Assemblies: Wise Democracy from the Minipublic

(Originally published on September 6th 2008)

Politicians should take note; there is a new answer to some of the toughest questions of our times. When presented with an issue with no obvious popular and sensible solution, or a situation where a legislature is unable to make progress on an important topic, 100 random citizens can be called on to solve the political puzzle, as they did in the Canadian provinces of British Columbia and Ontario (my home province).

Following the 2001 election, the newly elected premier of BC followed through on a promise to create a citizens' assembly to consider changes to the provincial electoral system. In 2006 the Ontario government followed suit as part of their democratic renewal efforts. Both citizens' assembly projects followed an innovative model designed by former BC politician, Gordon Gibson, and were given a clear and independent mandate by an all party committee.

Each assembly process began with tens of thousands of written invitations sent out to random citizens all across the province. Through several stages of positive responses and further lottery selection, the members of the assembly were narrowed down to 158 in BC and 103 in Ontario. Members came from every electoral riding. Their ranks included equal numbers of women and men, and spanned the demographic spectrum in rough similarity with census data. While not absolutely perfect, this was a more representative sample of folks than I have ever seen at any town meeting or campaign rally.

Central to the citizens' assembly model is the learning phase. In Ontario and BC, members spent six weekends learning about the topic from panels of experts, custom educational materials, and a staff of adult educators selected and trained to present a range of perspectives in a way that avoids biasing the process. By the end of this learning phase these assorted bus drivers, home makers, blue-collar managers and school teachers were able to debate election reform at a Masters level.

Following this learning phase, the assembly members took part in a series of public meetings and opportunities for comment from the public, giving members a greater understanding of the varied views and opinions within the population.

Finally each assembly went through an exhaustive six weeks of facilitated consensus driven deliberations and structured decision-making. Members talked in small groups and large groups, debated, researched, weighed options, heard concerns and voted step-by-step through each of the key decisions required to find a common answer. In the end both BC and Ontario citizen assemblies ended with over 90% of their members voting in favour of a common final recommendation. As the third party evaluations and academic reviews have come to prove, these staggering majorities were not the result of charismatic manipulation, authoritative coercion, or exhausted frustration. These results represent over 100 random people approaching full agreement on an open ended question—on an issue as complex as election reform. This was achieved by a thorough understanding of the options and respectful discussion with the stated goal of seeking the best solution that would be in accord with the commonly recognized values of the people. This was an example of the wise and practical democracy most of us assume is impossible. As Gordon Gibson expressed it "For someone with a faith in democracy, this was like seeing God."

To put this demonstrated model of the citizens' assembly into context let's quickly look at some more traditional methods of hearing the 'voice of the people' on public policy:


  • Elections: Candidates often win less then 50% of the votes cast, (but still more than their multiple competitors). Voters are generally poorly informed by combative media campaigns and are unable to recall much detail about the policy positions of their favourite candidates. Once elected, politicians are driven by short term public perceptions and party rivalry in order to secure a re-election.

  • Expert Panels: In formal committees, politicians and government bureaucrats are informed by select experts. The members of these committees are often well informed about their subject matter, but without any necessary grasp of public values. The selection of experts may bias the advice.

  • Opinion Polls: These telephone surveys are a result of top of mind reactions to yesterdayís sound bites and newspaper headlines. They superficially reflect public values, but without the educated, deliberated, and reasoned conclusions one would want to steer a society by.

  • Focus Groups: Focus groups typically have a small number of people at the table who are usually not informed about the issue at hand. Depending on the facilitation, focus groups may yield results that are uninformative, and not highly representative of the values of the population as a whole.

  • Town Halls & Hearings: Comments from the floor in a public hall have always been abused by the loudest and most charismatic speakers who are first to speak their complaints and accusations to the room. While iconic of our early democracy, the self-selected public speakers who tend to participate are often driven by personal or interest group agendas and are quickly situated in Us-VS-Them debates. These are not well informed, representative, or consensus-driven events.

assembly_hall250.jpgIn comparison, the citizen's assembly model is what deliberative democracy theorist Archon Fung calls a "minipublic," that is " educative forum that aims to create nearly ideal conditions for citizens to form, articulate, and refine opinions about particular public issues through conversations with one another." It is one of few processes where the shared values of the public are directly applied to policy recommendations, rather than guessed or assumed by privileged individuals—sometimes with their own agenda. That said, the citizens' assembly model it is not a perfect system. It is susceptible to manipulation or corruption by incompetent staff, or can be directed by a biased chair, possibly appointed for political reasons. According to the third party evaluations, this was not the case in Ontario or BC.

Both the BC and Ontario Citizens' Assemblies on Electoral Reform ended with referendums (similar to U.S. ballot initiatives) that were carried out as an addendum to the provincial elections. That is, the thoroughly debated, close-to-consensus recommendation of over 100 random citizens (who had been highly educated on the topic at hand), was subject to 60% approval by a general public that was overwhelmingly uninterested and uninformed about the subject matter. In BC the proposal won 57.7% of the votes, but did not pass the 60% threshold required. In Ontario the proposal only received 37% support. One theory for the difference between the two is the much higher level of media coverage of the citizens' assembly process that occurred in BC, i.e. the more people learn about the citizens' assembly process, the more likely they are to support its recommendation. In any case, referendums are dependent on expensive media campaigns and commercial news coverage with often trivial, controversy seeking, and superficial rhetoric. Without a complete overhaul of the media system, like public opinion polls, referendums are not appropriate mechanisms for wise policy decisions. In short, the citizens' assembly model works to produce useful recommendations to government and like any legislative commission or committee, should not be required to pass a referendum.

Beyond these two Citizens' Assemblies on Electoral Reform, Canadians have and continue to use similar random selection, educated and deliberative citizen panels to inform various government decisions, such as the newly starting Ontario Public Drug Programs Citizens' Council, the ongoing independent Canada's World project, or some of the many citizen dialogues conducted by the Canadian Policy Research Networks. Based in Toronto, a young firm called Mass LBP is aiming to make a business out of citizens'-assembly-inspired public consultation.

Internationally, many governments and non-government organizations have conducted similar processes under many different names: Consensus Conferences, Study Circles, Planning Cells, National Issues Forums, 21st Century Town Halls, Citizen Juries, and Citizen Panels (among others). Each model varies in the number of members, the amount of time given to education and deliberation, and the facilitation process, but as Matt Leighninger of the Deliberative Democracy Consortium suggested at the BC When Citizens Decide conference, "We should avoid 'modelitis' that focuses on the difference between models rather than the similarities. The larger context is more important then the specifics of the model." That larger context most importantly includes the political will of the government to listen to recommendations from its citizens.

Looking to 2009, we will see elections in Canada, the Unites States of America, Germany, Mexico, India, Japan and over 50 other counties. These politicians will all be facing such challenging issues as climate change, public education reform, strains on health care, improving child care, supporting minority rights, addressing aboriginal land claims, fresh water protection, demographic shifts, sustainability and development. When looking for direction on such complex issues, there will be many that seek advice from business leaders and experts, some that carry out traditional consultations with the usual suspects, but only a courageous few that will take the political risk to champion citizens' assembly like process that will have actual influence based on the deliberation of informed random citizens. These few pioneers will be the examples for future democratic leaders and we should give them our support.

You can learn more about deliberative democracy processes at the National Coalition for Dialogue & Deliberation's Learning Exchange.

Jason Diceman is a stakeholder engagement consultant with LURA Consulting and author of the popular Dotmocracy Handbook for large group decision-making.