At the Internet Governance Forum (IGF) in Brazil, a 2015 deliberative poll was initiated in order for experts from corporations, governments, civil society, and academic institutions from around the world to debate how internet access should be facilitated.
Problems and Purpose
By 2015, forty percent of the world’s population had access to the internet, compared to one percent just twenty years earlier. However, this dramatic growth has disproportionately occurred in developed countries, causing a ‘digital divide’. One reason that the narrowing of this divide is important, is the internet’s educational opportunities, which allow the accumulation of skills, thus potentially reducing economic inequality.
The rate of internet access growth is believed to be slowing due to factors such as: an absence of physical infrastructure, low wages, and politically imposed barriers. There are various proposals for expanding access to the internet. However, there are drawbacks that must be considered, with concerns regarding net neutrality, privacy and surveillance.
In November 2015, a pilot Deliberative Poll was carried out at the Internet Governance Forum (IGF) in Brazil, to stimulate considered debate, and to prompt decision-makers globally to act upon the outcome. The focus of this Deliberative Poll was to rank proposals on how the expansion of internet access should be facilitated. This Deliberative Poll differs from previous cases as its participants were not members of the public from a single country or region alone. Instead the representative sample was of experts: from corporations, governments, civil society and academia from across the world.
Background History and Context
The Tunis Agenda at the World Summit On The Information Society (2005) called upon the United Nations (UN) to establish the IGF, which would be: “multilateral, multi-stakeholder, democratic and transparent” . The UN extended the IGF for a further ten years in 2015. This international organization is comprised of different types of actors from corporations, governments, civil society, to academia from across the world. Therefore, each of these has differing levels of independence, influence and power. Professor Fishkin, Professor Diamond and Doctor Senges sought to remedy these imbalances of power and enhance the quality of dialogue by conducting a Deliberative Poll at the 2015 IGF in Brazil. Since Professor James Fishkin developed the method of Deliberative Polling in 1988, the technique has been used across the world to better inform participants on policy issues, with the process concluding in democratic decisions suggesting ways forward.
Organizing, Supporting, and Funding Entities
This Deliberative Poll was organized by Professor James Fishkin and Professor Larry Diamond from the Center for Deliberative Democracy (CDD), and Doctor Max Senges from the Centre on Democracy, Development and the Rule of Law (CDDRL) at Stanford University. This Deliberative Poll was in part sponsored by The Internet Corporation for Assigned Names and Numbers (ICANN).
Participant Recruitment and Selection
Emails were sent in waves to publicly available addresses to potential participants with desired demographics. Participants were selected through a stratified sample which aimed to reflect the community within the IGF as it was in 2015, rather than what it would ideally look like. No incentives were paid to respondents due to the low budget nature of this pilot project.
As a pilot scheme, there were just 61 participants who took part in the deliberations, which is between three to four times less than in other Deliberative Polls. Of those who did participate fully in the Deliberative Poll: 24% were employed in academia, 34% percent from either technology or the private sector, 22% from civil society and 16% from governments. It was expected by the organizers that the confidential nature of the polls would reflect the sincere opinions of the individuals, rather than the interests of their employer. In terms of gender: 62.3% of participants were male and 37.3% were female. 31% of participants came from North America, 15% from South America, 15% from Europe, 12% from Asia and 12% from Africa.
Methods and Tools Used
Deliberative Polling involves a random sample confidentially answering a set of questions before and after they engage in a considered debate on that topic. Therefore, a comparison between the raw and informed responses will clearly illustrate the extent to which greater knowledge impacts upon participants’ preferences. Also, it reflects how the population that the participants represent would respond if they were more knowledgeable. Qualitative data is collected to consider why participants’ preferences shifted following deliberation. Professor James Fishkin developed this democratic innovation in 1988.
What Went On: Process, Interaction, and Participation
As a pilot, the length of the deliberation was half a day, which is half of the time that is usually allocated for a Deliberative Poll. Deliberation occurred at fourteen small face to face group discussions with moderators, and two discussions took place before the IGF using Google Hangouts. Two plenary sessions took place, allowing participants to question panels of experts with opposing viewpoints. An advisory committee and a wide range of experts with diverse opinions scrutinized the information provided to participants to ensure balance and accuracy.
Before and after the deliberation, participants were asked in a confidential questionnaire to rank which of the proposals should be prioritized on a scale of one to ten. Respondents to the questionnaires before and after the deliberation answered six questions that measured the extent of their gains in knowledge. The respect between participants was also questioned to measure the tone of the deliberation. The comparison between the scores for pre-deliberation and post-deliberation highlight shifts in policy attitudes, extent of knowledge, the equal of participation and the reasoned nature of the deliberation that occurred.
Thirteen proposals seeking to address the ‘digital divide’ were considered. These proposals were covered by four themes. The first of these considered leaving the problem to the markets, where access would be granted in exchange for advertising. The second theme suggested that NGOs, local businesses, and communities could expand the provision of Wi-Fi. The third theme proposed national and international actions, such as a Digital Solidarity Fund, connecting funders to projects, and the enshrining of the internet as a right. Finally, it was proposed that limits on intellectual property costs could emulate the distribution of medicines in developing nations.
Influence, Outcomes, and Effects
75% of participants who took part in the deliberative event found that it was valuable. Of the thirteen policy proposals that were discussed, the shift in preferences on seven of these was interpreted as being significant. Participants saw their knowledge increase from their pre-deliberation questionnaire score by an average of 10.6%. Both the gains in knowledge and attitudes to policy proposals imply the success of the deliberative event, considering that the attendants arrived as experts.
Firstly, support decreased from 5.26 to 4.32 for leaving internet access to the markets, a result viewed by the researchers as being significant . Secondly, support for action from NGOs, local businesses and user communities increased from a mean of 7.39 to 8.41. Thirdly there was decreased support from a mean score of 6.85 to 5.48 for an international digital solidarity fund, interpreted as highly significant by the researchers. The proposal consisting of connecting funders with projects for internet access declined in support from a mean of 7.62 to 6.5. The classification of internet access as a right waned in popularity from a mean of 8.14 to 6.94, due to concerns that it would be a distraction from other more urgent needs that still need to be addressed such as the provision of clean water, medicine and food . Finally, there was increased support for limits to intellectual property costs with the mean score rising from 5.4 to 6.23, which has been interpreted as significant by the researchers . The two most popular policy proposals were the provision of internet at government funded services and at local businesses or communities.
Analysis and Lessons Learned
The pilot was criticized by Marilyn Cade, a member of the Multi-stakeholder Advisory Group appointed by the UN Secretary-General to advise on the development of the IGF. She stated that the: “sample is neither representative nor in any way broad enough…the sample is very, very small” . Therefore, her statement reinforces that the size and makeup of the sample selected for a Deliberative Poll must be adequate in order to command legitimacy and respect from decision makers. Julia Reda, a representative from the European Parliament added that the qualitative responses will provide more insight into why people changed minds than the final deliberation scores . This implies that international actors are more likely to pick and choose selected quotes from the deliberation that justify their existing preferences, rather than act upon the final democratic outcome. Furthermore, Eileen Donahoe’s comment that: “No one is suggesting that the outcome should dictate policy outcomes”, highlights concerns that deliberative democratic exercises fail to make a difference and result in any real policy change.
A common criticism of deliberative events, as voiced by Sanders (1997) is that they simply cause participants to conform to the views of the dominant groups involved. Fishkin et. al. (2016) addressed this criticism by measuring whether participants’ views shifted towards the opinions expressed by men and those from developed countries following the deliberative stage. When Fishkin et. al. (2016) conducted a test of the data, they discovered that this was not the case.
A further criticism of Deliberative Polling has been made by Ladd (1996), whose theory of the Hawthorne Effect, where participants change their behavior because they are aware that they are an object of study, is not relevant in this case. The participants at the 2015 IGF Deliberative Poll were not simply members of the public. As experts and potentially representatives of larger interests, they were less susceptible to expressing attitudinal change. Furthermore, this Deliberative Poll did not provoke any media attention that would distort any findings.
Another lesson that has been learned from this Deliberative Polling is that participants would rather take part in face-to-face discussions than those carried out online (Fishkin et. al., 2016). Also, whilst the pilot was short in length, the scope of potential solutions was broad (Fishkin et. al., 2016). Therefore, a longer and more focused deliberation would result in deeper understanding.
 Fishkin, J. (1988). The Case for a National Caucus: Taking Democracy Seriously. Atlantic Monthly (August), pp.16-18.
 Sanders, L. (1997). Against Deliberation. Political Theory, 25(3), pp.347-376.
Fishkin, J., Senges, M., Donahoe, E., Diamond, L. and Siu, A. (2017). Deliberative polling for multistakeholder internet governance: considered judgments on access for the next billion. [online] Centre For Deliberative Democracy. Available at: http://cdd.stanford.edu/2017/deliberative-polling-for-multistakeholder-internet-governance-considered-judgments-on-access-for-the-next-billion/ [Accessed 1 Dec. 2018].
Fishkin, J., Senges, M., Donahoe, E., Diamond, L. and Siu, A. (2016). Executive Summary. Piloting the use of Deliberative Polling for Multistakeholder Internet Governance: Considered Judgments on Access for the Next Billion. [online] Centre For Deliberative Democracy. Available at: http://cdd.stanford.edu//mm/2016/06/IGF-DP-Report-Executive-Summary.pdf [Accessed 1 Dec. 2018].
Fishkin, J. and Senges, M. (2016). The public voice in internet governance. [online] Free Speech Debate. Available at: https://freespeechdebate.com/discuss/the-public-voice-in-internet-governance/ [Accessed 7 Dec. 2018].
Fishkin, J., Senges, M., Donahoe, E., Diamond, L. and Siu, A. (2016). Piloting the use of Deliberative Polling for Multistakeholder Internet Governance. [online] CDD. Available at: http://cdd.stanford.edu/2016/piloting-the-use-of-deliberative-polling-for-multistakeholder-internet-governance [Accessed 7 Dec. 2018].
Ladd, E. (1996). Magic Town: Jimmy Stewart Demonstrates the ‘Hawthorne Effect. ‘Public Perspective’ 7(3). p. 16.