Proposal 6 for ICANN: Enhance Decision-Making Legitimacy by Experimenting with Innovative Voting Techniques

This is the sixth of a series of 16 draft proposals developed by the ICANN Strategy Panel on Multistakeholder Innovation in conjunction with the Governance Lab @ NYU  for how to design an effective, legitimate and evolving 21st century Internet Corporation for Assigned Names & Numbers (ICANN). 

Please share your comments/reactions/questions on this proposal in the comments section of this post or via the line-by-line annotation plug-in.

From Principle to Practice

The legitimacy of the decisions made by ICANN depends upon whether those affected by the decisions have been able to give their input. Therefore, it is incumbent upon ICANN to institute structures that lower barriers to meaningful engagement for netizens and the ICANN Community. ICANN could, therefore, experiment with innovative voting techniques, which allow for a community to identify both issues that are important and individuals who are best suited to speaking on those issues.

The use of topically-based voting models may also provide a means for ICANN to test the effect of engaging its community to organize around specific issues rather than around specific constituencies.

What Does it Mean to Experiment with Innovative Voting Techniques?

Around the world and in different contexts, a variety of voting models exist. All voting techniques have some set of criteria for what defines an eligible “voter,” which may be considered membership-criteria.

In thinking about how to introduce innovative voting techniques at ICANN that would a) make decision-making at ICANN more accessible, and b) empower members of the ICANN community to take thought-leadership roles, the Panel recommends that ICANN consider experimenting with two “voting models” in particular: Liquid Democracy and Ranked-Choice Voting.

Liquid Democracy

  • Liquid Democracy is an instance of “proxy” or “delegative” voting. Delegated votes are transitive and can be revoked at any time. Delegated votes create “voting blocs” where certain individuals carry the aggregated votes of others. Delegations are recursive in that a proxy (who has aggregated others’ votes) can also delegate their bloc of votes. Voting is also “alive,” in that people can change their votes through redelegation or revocation.[1] The video below by German designer Jakob Jochmann provides an introduction to Liquid Democracy.
YouTube Preview Image

Ranked-Choice Voting

  • Also known as “instant-runoff voting” or “alternative voting,” ranked-choice voting techniques let voters rank candidates (or issues) in order of preference rather than voting on a single candidate or issue. If a candidate or issue receives more than 50% of the vote (the majority), they win the election. When there is no majority, the candidate or issue with the least votes has their voters’ second choice votes counted. This process continues until one candidate or issue has the majority. These techniques prevent separate run-off elections and minority rule. The video below by C.G.P. Grey provides an introduction to ranked-choice voting.
YouTube Preview Image

Why Does This Proposal Make Sense at ICANN?

ICANN is neither a direct democracy nor a proxy voting system. ICANN’s multistakeholder process is meant to give voices, not votes, to stakeholders.[2] Instead of having direct control, stakeholders influence outcomes through their input, discussion, and advocacy for their point of view. Different Supporting Organizations (SOs) and Advisory Committees (ACs) have different processes by which they vote and deliberate. Usually, a Working Group (WG) or Task Force will present recommendations to an SO/AC Council (up to this point, everything is deliberative); then the Councils take a vote to approve or reject the WG’s recommendations, and then the Board of Directors takes a vote and makes final decisions based on input from stakeholders ranging from governments to Internet end-users to domain name registrars.[3]

ICANN is full of narrow, complicated issues and few people are really knowledgeable on more than a couple at a time.[4]

At ICANN, there are many different kinds of “policy” and “policy-development.” There is also a lack of clarity about what the difference is between “policy” vs. “implementation.” ICANN deals with a diversity of issues through diverse “constituency” structures, for example the Business Constituency (BC) or the Intellectual Property Constituency (IPC) under the Generic Name Supporting Organization (GNSO).[5] It has been argued that a lack of procedural clarity coupled with ICANN’s dizzying structural complexity causes a significant amount of ICANN’s decision-making to happen in ad hoc ways that are not methodologically rigorous.[6] These arguments highlight challenges to ICANN’s legitimacy, in particular where ICANN must ensure the stability of the global Internet while being able to innovate and take risks.[7]

Innovative voting methods such as Liquid Democracy or ranked-choice voting could:

  • Allow people to organize around topics and issues rather than around their constituencies. For example, in the GNSO, there are a host of different “constituency groups,” and people may not always agree with their constituencies on each issue. In this way these methods account for a multiplicity of priorities.
  • Remedy the fact that those responsible for casting votes (often volunteers) do not have enough time and knowledge to vote meaningfully on every issue.
  • Allow people to bypass the constituency-level vote by choosing not to vote for a Council member but to vote directly on the issue.
  • Allow people to delegate their votes on issues or on people to others who they trust to vote for them, creating “chains of trust.”[8]
  • Allow for vote aggregation around specific people, who may differ depending on the issue at hand. This means that delegated votes can identify “thought leaders” within the ICANN community on specific issues, and those “thought leaders” then make decisions on a specific issue with the support of the community.
  • Maintain the privacy of individual voters.
  • Lower barriers to participation because the cost of becoming a delegate is small.
  • Drive a diversity of candidates because expertise is based in subject-matter.
  • Limit the concentration of power:
    • For instance, in Liquid Democracy, people can revoke their delegation, break the “chain” of delegations, and take away many votes from the final representative at any time.[9]
    • In ranked-choice voting, a plurality of options are chosen in order of general voter preference, mitigating issues-based polarization of the Community.

Implementation Within ICANN

In ICANN’s case, many of the existing stakeholder groups and constituencies (“structures”) can be used as platforms to organize people’s votes. For example, Elliot Noss submitted to the MSI Panel’s engagement platform that the Regional At Large Organizations (RALOs), which make up the At Large Advisory Committee (ALAC) may be appropriate existing structures for finding “empowered leaders” through an innovative voting process. [10] Mikey O’Connor has also suggested that the Non-Commercial Stakeholder Group of the Non-Contracted Party House of the GNSO may be able to use Liquid Democracy techniques to enhance their ability to conduct and make use of broad-based outreach.

Here are some initial pilot ideas with which ICANN could experiment:

Liquid Democracy

  • Membership & Eligibility to Vote
    • People may become eligible to vote on issues or people at ICANN through two basic eligibility criteria: being the registrant of a domain name (therefore having interests by default) or by expressing interest (for example, by submitting a “statement of interest” much like how current Nominating Committee members or Working Group members submit “statements of interest”).
  • Voting Structures
    • New voting methods could be applied wherever voting currently occurs in ICANN, e.g., at the Council or the Board level. For example, where the GNSO makes selections to fill Seats 13 and 14 of the ICANN Board,[11] Liquid Democracy principles may be employed to allow GNSO members to either vote for a Board member directly or to delegate their vote to existing Council members or non-Council members. Alternatively, where the GNSO Council votes on an issue, a GNSO member may choose to vote directly on that issue, or to delegate their vote to a Council member or outside thought-leader who they trust to vote on the issue for them.
  • Voting Procedures
    • As Liquid Democracy voting occurs, vote-accumulation should be observable. Vote-accumulation may be used to identify “thought-leaders” and to clarify how different issues polarize the ICANN community, and which positions have majority and minority support. Individuals can also vote directly on issues.
  • Voting Outcomes
    • For piloting purposes, issues or candidates that go through the Liquid Democracy process should arrive at the Board as is current practice, and could provide a more comprehensive picture of the will of the community, to supplement recommendations or advice formally submitted through traditional channels.
    • Notably, if ICANN were to agree to formally adopt this technique in relation to a specific issue or election, the Board’s role should be confined to adopting/ratifying/rejecting the outcome of the vote, consistent with ICANN’s Bylaws.

Ranked-Choice Voting

  • Membership & Eligibility to Vote
    • These criteria should follow the same prescriptions as described above for Liquid Democracy implementation.
  • Voting Structures
    • Where SO/AC Councils or ICANN’s Board of Directors must take a vote, it makes sense to use ranked-choice voting to quickly determine which issues or positions win (for example, where the Board has the power to appoint the Nominating Committee Chair [12]). Furthermore, where Council members or Board members are to be elected, candidates may be selected by ranked-choice voting. The voting members of the Nominating Committee could also use ranked-choice voting techniques to select new ICANN Board members.
  • Voting Procedures
    • Ranked-choice voting could be applied not just to candidates for election but also to issues and objectives, for example strategic objectives as laid out in ICANN’s 5-year Strategic Plan.
  • Voting Outcomes
    • These criteria should follow the same prescriptions as described above for Liquid Democracy implementation.

Examples & Case Studies – What’s Worked in Practice?

The case studies that have applied these emerging voting techniques have notably left many open questions that beg for further testing and studying to figure out what works in what contexts. In comparison with case studies of ranked-choice voting, there are remarkably few case studies of Liquid Democracy voting techniques in practice.

Liquid Democracy

  • Pirate Party – Founded by Rick Falkvinge in 2006 in Sweden, the Pirate Party has gained influence and visibility especially in Europe. Some Pirate Party “chapters” are experimenting with Liquid Democracy techniques, notably the one in Berlin. In 2011, the Berlin Pirate Party drew 8.5% of the vote in the Berlin state election.
  • The World Parliament Experiment (WPE) –The WPE is a “generic simulation of a working Global Democracy on the Internet,”[13] which uses a combination of direct and representative democracy techniques. In particular, it uses a delegative voting method which offers voters three options: no delegation, random delegation, and delegation to a chosen person. [14]

Ranked-Choice Voting

  • In the United States – As of July 2012, ranked-choice voting elections had been held in a statewide election in North Carolina and for local elections in San Francisco, California; Oakland, California; Berkeley, California; San Leandro, California; Burlington, Vermont; Takoma Park, Maryland; Aspen, Colorado; Minneapolis, Minnesota; Pierce County, Washington; Telluride, Colorado; St. Paul, Minnesota; Portland, Maine and Hendersonville, North Carolina.[15]
  • Academy of Motion Picture Arts and Sciences – The final ballot for the 2009 Oscar Best Picture award used a preferential voting system to determine which of 10 contenders would win.

Open Questions – Help Bring This Proposal Closer to Implementation?

  • How should membership and eligibility criteria for voting be defined? How can individuals be certified or authenticated once these criteria are defined?
  • Could best practices from ranked-choice voting techniques be integrated with Liquid Democracy or proxy voting techniques?
  • How can demands for broader inclusion and more global participation at ICANN be met with tools that enable highly scalable participation?
  • Some have suggested housing a pilot within the RALOs or the Non-Contracting Party House? Are there other structures that would be good venues for testing the effect of these voting techniques?
  • What metrics could ICANN use to test how these alternatives might function in comparison to current decision-making processes?
  • How can innovative voting techniques be used more broadly, for example by “citizen juries” or to consider issues or their impacts retroactively (e.g., outside of a formal PDP?)
  • How could innovative techniques be applied in a manner that enables changes in delegation or rank over time (to respond to changing conditions) while ensuring an appropriate cut-off point at which votes are final/decisions get made?


1. “LiquidDemocracy.”
2. N. Steins and V. Edwards. “Platforms for Collective Action in Multiple-Use CPRs.” Conference of the International Association for the Study of Common Property, Vancouver, British Columbia, Canada.) (June 10-14, 1998): 10.
3. See “GNSO Policy Development Process (PDP).” ; “ccNSO PDP Process in Graphics – ICANN.”; and ICANN Primer. The GovLab. (October 2013): 11 – 13.
4. Noss, Elliot. Ideascale Idea Submission: “Liquid Democracy to Reinvigorate Decision-Making.” The GovLab MSI Panel Ideascale. (January 2014).
5. “GNSO Appointees to ICANN Board.”
6. Noss, Elliot. Ideascale Idea Submission: “Liquid Democracy to Reinvigorate Decision-Making.” The GovLab MSI Panel Ideascale. (January 2014).
7. For example, the new gTLD program is an instance of ICANN taking a calculated risk, which has destabilizing effects on the DNS (e.g., with name collisions) but is in line with ICANN’s mission to ensure consumer choice and competition on the Internet. A good example of this “calculated risk” is the “Fast Track” IDN-ccTLD process, which enabled countries to apply for an internationalized version of their ccTLD (i.e., apply for non-Latin script versions of ccTLDs) while avoiding a formal Policy Development Process. See also: ICANN Primer. The GovLab. (October 2013): 14.
8. “Transitive Delegations in Liquid Feedback.” (July 2012).
9. Ibid.
10. Noss, Elliot. Ideascale Idea Submission: “Liquid Democracy to Reinvigorate Decision-Making.” The GovLab MSI Panel Ideascale. (January 2014).
11. GNSO Council.
12. ICANN Nominating Committee.
13. “World Parliament Experiment FAQ.
14. Ibid.
15. “Instant Runoff Voting in the United States.”

The Tags . . .

No comments yet.

Leave a Reply