Symi 2014: The Challenge of Deliberative Democracy, by James Fishkin

strict warning: Declaration of views_plugin_style_default::options() should be compatible with views_object::options() in /home/symisymp/public_html/sites/all/modules/views/plugins/ on line 24.

Deliberative democracy is a practical answer to a philosophical question: what would the people think about an issue under good conditions for thinking about it? Why is that a philosophical question? How is it a practical answer? What does this add to democracy as we know it in real political systems around the world?

Democracy should have some discernible connection to the voice of the people. But how can we consult the people? How do we know what they really want? We hear mobilized voices from interest groups and angry portions of the public. But these may often be unrepresentative. Those who speak up are often just the people who feel strongly. The whole public can be represented in good public opinion surveys with scientific samples, but often the public is uninformed or disengaged. Polls may represent little more than the public’s impression of sound bites and headlines. A large literature demonstrates that usually public information levels are low. In the classic phrase of Anthony Downs, most voters are “rationally ignorant.” If I have one vote in millions why should I invest time and effort in attempting to become informed about complicated public policy or political issues? My individual vote or opinion will not make much difference. There are other areas of life where most people can make more of a difference.

Even voters who pay attention to public issues and discuss them are now very likely to talk mostly with people they agree with, to consult websites and publications they agree with, and to watch news sources they agree with. The like-minded convince each other and may miss the other side of the issue entirely. Furthermore, falsehoods spread virally via social media, and there is the persuasion industry which often intends to mislead, if not misinform the public in order to serve the interests of clients, whether they be candidates or parties or interest groups. Lastly, even when people report opinions in polls, they may be phantom opinions or “non-attitudes.” Members of the public do not like to admit that they don’t know. Asked a question about which they really have no opinion, they may nevertheless offer a response. The public offered answers in the US to polls about the famous “Public Affairs of 1975.” But it was fictional. There was no act to have an opinion about. They even offered responses when the Washington Post decided to celebrate the 20th un-anniversary of the Public Affairs Act of 1975 and asked survey respondents about its repeal. They got phantom opinion responses to that question as well.

Deliberative Polling (DP) is an attempt to study what views the public would have if it were effectively motivated to engage with public policy or political issues under good conditions. A carefully balanced briefing document vetted by an advisory committee provides the pros and cons of competing policy options. An initial questionnaire is used to recruit a sample, usually representative in attitudes and demographics. That sample deliberates for at least an entire day, sometimes two or three, in moderated small group discussions and in dialogue with competing experts who answer questions from the sample developed in the small groups. After this deliberation the sample takes the same questionnaire as on first contact. Usually the DPs are face to face, but sometimes a sample deliberates with this process online with voice and video. In any case, more than 2/3 of the questions asked in DPs change significantly with deliberation. When people think about an issue, their views are usually different than just a top-of-the-head impression of sound bites and headlines.

This process has been employed more than 70 times in 18 countries around the world, ranging from the US and Britain to China, Japan, Brazil, Italy, Greece, Poland, Bulgaria, Macau, and Australia. Deliberative Polling projects are now underway in Uganda and Ghana with the support of USAID. 

Why should policy makers pay attention to deliberative opinion when it is often not the opinion people actually have? It provides a route to responsible advocacy—it shows which difficult trade-offs the public would accept and which they would not, and for what reasons, all on the basis of good information. The process has often had major policy effects.  Policy makers, well aware of the limitations of standard public opinion research and also the unrepresentativeness of the often angry voices that can be mobilized in public meetings or on the internet, are at least interested in hearing about representative and informed opinion on key issues. It may give them “cover” to do the right thing.

Consider four examples: 

A) Texas. Beginning in 1996, there were 8 DPs in Texas on energy choices—how to provide electric power in each of the eight service territories of the state. Coal? Natural Gas? Wind? Conservation? The eight projects demonstrated that the public would be willing, after deliberation, to pay a little more on its monthly bill to support wind power. This percentage, averaged over the eight projects, increased with deliberation from 50% to 84%. In plans filed with the Public Utility Commission and with the legislature this had a substantial impact. Between 1996 and 2007, Texas went from last among the fifty states to first in the amount of wind power. Story is detailed in various writings including my book When the People Speak (see also

B) Japanese Pensions. In 2011 a national DP on pension reform with our partners at Keio University Center for DP demonstrated that the government’s proposed option of individual retirement accounts lost support (decreasing from 69% to 35%) after deliberation. The public in top-of-the-head polls thought it a good idea but the Japanese were too risk averse to put their retirements into market based investments. They wanted a guaranteed government run system—especially after they deliberated. But to maintain the viability of such a system in an aging society, they needed a tax that they could know everyone paid without any invasion of privacy (a major factor in Japan). Support for a rise in the consumption tax increased dramatically. Both the DPJ and LDP governments followed the logic revealed in the DP and the consumption tax rise has now been approved to make the current pension system viable (

C) The Roma Schools in Bulgaria. In 2007, we did a national DP project in Bulgaria on the condition of the Roma. At the time the Roma went to separate schools in a separate language and without teachers (older children teaching the younger children). In a three day deliberation about the condition of the Roma, there were dramatic changes of opinion— including support for closing the Roma-only schools and busing the children to regular schools (support rising from 42% to 66% after deliberation). The Roma schools have now been desegregated and our Bulgarian partners report on the role of the DP in catalyzing this major policy change. See the interview with the Prime Minister at the time in the New York Times on the CDD website to see how seriously he took the national deliberation (

D) Northern Ireland. In 2007, a Deliberative Poll in Omagh on education reform clarified the forms of cooperation that Protestants and Catholics would be willing to engage in, and the ones they would not. Even the willingness to view the other community as “trustworthy” or “open to reason” increased by 15 points with deliberation. The forms of cooperation that were approved are now embodied in a new campus for six schools on a former military base. Our colleagues in Northern Ireland have documented the extent to which the DP results were invoked in the planning. An overview of the project can be found at

By providing the reasoning that has weight with the public when they are informed (and by finding the reasoning that does not even if elites think it should), the deliberative process empowers the advisory group and policy makers to invoke, advocate and legitimate an option on the basis of good information. In these cases and others that we can cite, the long term basis for important reforms or policy changes can be established. Since random samples of the public are more amenable to dialogue than the organized interests who speak for them, this process can be used in situations of conflict or polarization such as Bulgaria with the Roma or Northern Ireland with Protestants and Catholics.

Deliberative democracy can be made practical, at least on selected issues. It would be a great improvement in the democratic systems we now have if public policy can be connected to what the people really want—when they think about it.


James Fishkin

Stanford University

June 2014