ICANN's Global Elections: On the Internet, For the Internet

A Study of the ICANN At-Large Elections by
Common Cause and the Center for Democracy and Technology

March 2000

4. Options for Change

This section of the report provides an overview of the primary options available to ICANN in building an At-Large membership and election process. Those consulted as a part of this study offered a wide variety of thoughtful and creative ideas to address the problems and tensions described above.

• 4.1 Election Structure
• 4.2 Membership
• 4.3 Putting candidates on the ballot
• 4.4 Campaign Rules and Regulations
• 4.5 Voting systems
• 4.6 Other Important Suggestions

4.1 Election Structure

As described above, there is widespread dissatisfaction in the ICANN community with the current indirect election structure, under which members would elect an 18-person At-Large Council with the very limited responsibilities of choosing the nine At-Large board directors. Election systems and voting experts from outside the ICANN community also expressed strong concerns about this structure.

Many participants in this study stressed that direct elections are a baseline precondition for the development of an engaged and watchful electorate that will hold the At-Large board directors accountable. There will be much less incentive to join and energetically participate in ICANN if the role of the membership is limited only to selecting an intermediary body of electors.

Indirect elections could also facilitate capture, depending upon the voting system used. If the voting system used by the Council is similar to the winner-take-all system used in the DNSO Names Council elections, for example, victory in a simple majority of indirect At-Large Council races could allow an interest to handpick all nine At-Large board directors. Conversely, under direct elections, a would-be captor would need to win every election to gain all nine of the board seats.

Regardless, an elected council voting for the board is much more susceptible to political deal-making and gamesmanship, outside of the eye of the public and without ultimate accountability to the electorate.

Opposing viewpoints: One of the primary arguments in support of indirect elections is that they provide a check on potentially destabilizing influences. Presumably, a deliberative intermediary body could prevent the election of "fringe" candidates - without grounding in ICANN or its mission - to the board. And without such a body, some fear that candidates' attempts to excite and engage voters will by their nature push ICANN into policy-making arenas in which it does not belong.

A second concern raised is that California state law would give those who directly elect board members certain legal rights that would jeopardize ICANN's stability.

Either the At-Large board seats (in the case of a direct election), or the At-Large Council seats (in the case of an indirect election), could be allocated to subgroups of the At-Large membership.

While ICANN's bylaws already include specific geographic diversity requirements for the At-Large Council and the At-Large board seats, ICANN could set up the At-Large elections so that all elected seats would represent geographic districts and would be filled exclusively by voting members living within those districts. Others have suggested that the elected seats be divided among two or more existing Internet organizations (e.g. the Internet Society), or among two or more sub-constituencies of the At-Large membership (i.e. an individual domain name holders' organization, an e-mail address holders' organization, etc.).

Supporters of this option argue that dividing the elected seats in these ways would disaggregate power in the At-Large process, just as the Supporting Organization structure spreads out power across three separate entities. This would greatly reduce the risk of capture. The subgroups could also serve to educate and disseminate information among the membership, leading to a more engaged and active membership.

Opposing Viewpoints: Some argue that the ability of like-minded individuals around the world to coalesce behind candidates and/or platforms would be lost if candidates are selected by strictly regional electorates to fill strictly regional seats. Others point out that geographic allocation of seats might facilitate capture, because it will be easier for an interest group to organize majority voting blocks in regions where there is currently little native interest in ICANN.

There was widespread skepticism of the practicality of using any other scheme to divide elected At-Large seats among different types of members or organizations. Even some that favored this option in principle admitted the difficulty of finding a fair way to determine which groups merit guaranteed representation and which do not.

If the indirect election structure is maintained, some support the addition of a requirement that any candidate for election to the At-Large Council must bind him/herself to vote on a first ballot for some identified (and qualified) slate of nine At-Large board candidates. Such an approach would force council candidates to make clear who they will try to elect to the board. It would lead interested groups towards the constructive process of finding complete sets of board members that are sufficiently qualified and balanced to attract support from a wide range of constituencies.

Enlarge the At-Large Council: Some who opposed indirect elections nonetheless suggested that a larger At-Large Council would provide more opportunities for elected representation to various segments of the Internet community (including those holding minority viewpoints). A larger Council will also make capture more difficult by increasing the number of elections that must be won to achieve a majority.

Expand the powers of the At-Large Council: Others suggested that becoming a member in an indirect system would become more meaningful and attractive if the At-Large Council were given a formal and ongoing policy role in ICANN, like its Supporting Organization counterparts. Some roles that were mentioned include the power to propose matters for board consideration and to review board decisions once they are made. Those who oppose direct elections might similarly oppose this option because as the powers of the office increase, candidate appeals to voters will become more aggressive, and the potential increases for a broader policy-making role for ICANN.

We believe the electorate should directly elect the At-Large board directors. Having an At-Large Council as a mediating body diminishes both the reality and appearance of an open, inclusive, representative, and democratic election.

While issues of stability are a concern, the Supporting Organization structure already provides a great measure of stability and expertise to the board. An intermediary council will make it quite difficult to engage a broad electorate, and will be susceptible to unpredictable and unaccountable deal-making that disenfranchises voters. Further, depending on the voting system used and other considerations, indirect elections could facilitate, not diminish, the risk of capture.

Our sense is that the election could be structured under California law to deal with the concerns raised should ICANN choose to do so, and if not there is a significant sentiment that providing members with some rights under law will improve the quality and engagement of the membership.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

4.2 Membership

ICANN's current proposal would open membership eligibility to any individual who possesses an e-mail address and can be reached at a verifiable physical address. ICANN believes that possession of an e-mail address is perhaps the most inclusive measure of an individual's participation in the Internet today, and a verifiable physical address serves as a necessary check on an individual or organization creating massive numbers of fictitious e-mail address owners, while also providing ICANN a means to investigate irregularities in its membership rolls. While many support this basic definition of the ICANN At-Large electorate, some consider a membership of e-mail address holders too inclusive, while others consider it too exclusive.

Many believe a broad electorate of e-mail address holders will likely be uninformed, susceptible to fraud and capture, and made up of individuals without a real stake or interest in ICANN's decision-making. One of the more frequent suggestions made would allow only domain name holders to be eligible to vote for At-Large board members.

According to this point of view, domain name holders are likely to possess a well-developed understanding of the Internet and the issues faced by ICANN, and will be more likely to elect technically competent representation.

In addition, because domain name holders must register and provide proof of identity, this limitation would create a clearly defined electorate that could be more easily monitored for potential fraud. The costs generally involved in obtaining a domain name will also make it more difficult for a would-be captor to enroll massive numbers of voters.

Finally, since ICANN's primary mandate is to set technical standards for the Internet and coordinate the assignment of domain names, domain name holders are the constituency most directly affected by ICANN policy.

Opposing viewpoints: Contrary to the arguments supporting this option, many have pointed out that some of the most sophisticated members of the Internet community - those involved in IP numbers and technical standards - are not necessarily domain name holders and would be excluded by this option. Obtaining a domain name requires money more than it requires Internet sophistication.

In addition, some argue that an electorate of domain name holders will make it only marginally easier for ICANN to approach the goal of 'one person, one vote,' particularly as the number of individuals with multiple Internet names and numbers continues to grow.

And, among those who see this as too exclusionary and undemocratic, there is a belief that ICANN already is captured by institutional interests through the Supporting Organizations. To these participants in our study, limiting membership to domain name holders would further tilt ICANN in favor of corporations, which possess a large percentage of all domain names, and against individual consumers and Internet users. Since corporate interests already largely control the election of nine board seats through the Supporting Organizations, making corporations a disproportionate segment of the electorate for the remaining nine At-Large seats violates the intended balance of the board.

Additional ideas regarding the electorate:

Those who oppose fees point out that this is tantamount to imposing a "poll tax," with all the undemocratic and discriminatory connotations that carries. To these commentators, any such "poll tax" would publicly discredit ICANN as an open and inclusive decision-making body. In addition, a fee might dampen participation by a broader electorate and make capture easier for an interest with great organizational resources and a passionately held agenda.

Create other barriers to membership: Other barriers to membership which are intended to have an effect similar to a fee have been suggested, such as requiring members to participate in certain ICANN activities (e.g. on-line candidate debates), or to demonstrate substantive knowledge of ICANN and its current issues. Variants of the arguments against a membership fee also apply to these options to a greater or lesser extent.

Requiring members to register a reasonable but meaningful time in advance of an election, and to re-register annually after each election [10], are less vulnerable to the "con" arguments stated above. If implemented in a fair and democratic way, these barriers could help ICANN properly prepare for the elections and audit the electorate for fraud.

Create expanded opportunities for member participation in ICANN: As some emphasize, providing members with opportunities for meaningful participation in ICANN activities, as opposed to requirements of participation, would have the healthy effect of encouraging a more engaged and watchful membership more likely to hold its elected leaders accountable for their decisions. By making membership more attractive, a larger and broader membership that is less susceptible to capture will develop.

We believe that ICANN is heading in the right direction in defining the electorate to include any individual with an e-mail and verifiable physical address. Given the rapidly changing nature of the Internet and the impracticability of determining today who will claim a legitimate stake in ICANN tomorrow, it is unwise for ICANN to define its At-Large membership more narrowly than it is currently. A more exclusive membership would seriously damage the legitimacy ICANN seeks as an open and democratic body, and is unlikely to yield any benefit in return, particularly since a narrower membership will not necessarily prevent capture or lead to a more informed electorate. [11]

In addition, because corporate and other institutional interests already select half the board through the Supporting Organizations, there is both a legitimate need for individual Internet users to have a voice in ICANN and a level of built-in stability that will insulate ICANN against the outcomes feared by those who see an open and democratic At-Large process as potentially dangerous.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

4.3 Putting candidates on the ballot

To date, ICANN has not established detailed procedures governing how candidates become eligible for the At-Large ballot. Several participants in this study stressed the importance of such procedures in creating fair and effective elections.

If a proper set of strict criteria is established for At-Large candidates, unqualified candidates will be weeded out. Among others, these criteria could include: age, experience, demonstration of Internet knowledge, willingness to provide information on issue positions and to appear in on-line debates, or proof that a candidate does not have a conflicts of interest (including documentation that the candidate is not employed, or in some other way compensated, by a Supporting Organization or other group with an interest in ICANN policy).

While there is general agreement that some objective criteria for candidates would be beneficial, there are few specific criteria that are widely considered fair and democratic. In particular, some potential criteria conjure up public images of elitism (i.e. a test of knowledge of ICANN issues would resemble a literacy test) that are arguably inconsistent with ICANN's mandate.

Option 6: Create a clear nominating process

Few, if any, interested parties to whom we spoke disagree with the need to establish a clear process for nominating candidates to the ballot. Some believe that the dangerous potential consequences of an uninformed or captured electorate could be avoided by the creation of a strong and distinguished nominating committee possessing ultimate authority over who appears on the At-Large ballot.

Others argue that determining who is on the ballot is nearly as important as choosing among candidates at election time, and that a strong nominating committee removed from the electorate is undemocratic on its face. Individuals sharing this point of view generally prefer a member-centered process.

A hybrid approach has also been suggested in which a nominating committee would be established to actively recruit high-quality, Internet-savvy candidates, but in which other candidates could also gain direct access to the ballot through nomination by members, such as by meeting some threshold requirement of electronic petition signatures from At-Large members. This hybrid approach would ensure that at least a full slate of qualified candidates - endorsed as such by the nominating committee - would appear on the ballot, but that the ballot remains open to access by others as well.

Additional ideas regarding candidates and ballots:

Slates or endorsements would help voters make sense of the field of candidates, by allowing them to incorporate in their decision-making the opinions of organizations whose values they share or do not share. This would help prevent the election of candidates who misrepresent their positions to voters, as organizations would have the resources to vet candidates on their past record and viewpoints, particularly before allowing them to appear on the slate.

Some have argued, however, that slates or endorsements would facilitate capture by providing on-the-ballot instructions to voters enlisted by a would-be captor.

We recognize that, at least in the short-term, it will be difficult for voters to develop the knowledge of candidates and their positions necessary to make informed choices. We believe ICANN can respond by creating a candidate slate process and/or endorsement procedures by which organizations can attach their names to candidates. These mechanisms will provide voters with important clues about candidates.

A nominating process which provides an opportunity for institutional players in ICANN to place a group of candidates on the ballot will have a similarly healthy effect, as long as it does not deny ballot access to other candidates backed by members. In addition, while we do not believe that establishing strict criteria for candidates will by itself ensure qualified directors, we do support requirements that candidates provide information on their positions regarding key issues, and believe that candidates should be strongly encouraged to participate in on-line debates or discussions.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

4.4 Campaign Rules and Regulations

ICANN's current policy documents are also silent on campaign rules and regulations. Such rules and regulations could play a significant role in who is elected through the At-Large process and whether or not that process is fair and democratic. ICANN should consider what campaign rules and regulations are appropriate for the At-Large elections.

It is difficult to set rules for the financing and conduct of campaigns for which the level of interest and participation remains unknown. As a starting point, concerns that not enough information on candidates will be available to the electorate - that campaigns will not be sufficiently vigorous - were expressed more often than concerns that campaigns will spiral out-of-control with huge contributions, massive expenditures, and unethical tactics.

Two sensible ideas that have been suggested in this regard are: 1) to encourage formal opportunities for voters to engage candidates through on-line debates and discussion, and 2) to establish disclosure rules for the source and size of campaign receipts and disbursements, as well as for other information like sources of personal income that candidates receive from entities with potential interests in ICANN decision-making.

Option 8: Encouraging Campaign Activities

As previously indicated, many participants in this study expressed concern that voters will lack familiarity with candidates and will find it difficult to choose among them. Some mentioned the importance of activities like on-line debates, candidate question-and-answer sessions, issue surveys and straw polls, and other activities that allow voters and candidates to express themselves and communicate with each other. In addition to helping voters navigate their ballots, this could also make membership more interesting and encourage a more engaged At-Large electorate.

Of course, there are those who fear a dynamic campaign and election process for its potential to politicize the At-Large elections and threaten ICANN's stability.

We believe that vibrant campaigns are critical considering that candidates and voters will not be well-known to each other initially. We thus support the facilitation of campaign opportunities by ICANN. With regard to campaign rules, we also support requirements that candidates disclose campaign receipts and disbursements of significance, along with other important information like sources of income that might create conflicts of interest.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

4.5 Voting systems

ICANN has made no formal decisions on the system of voting to be used for the At-Large elections. Some assume that the intention, at least for the elections of the board members by the At-Large Council, is to use a system similar to the one in the Names Council elections of the DNSO board directors. That system was a winner-take-all system that involved successive run-offs and allowed a simple majority to select all three board directors. There is very little, if any, support among those to whom we spoke for using a winner-take-all voting system in the At-Large elections.

The ICANN bylaws now provide for staggered elections. According to the current plan, the At-Large membership would elect six of the 18 total At-Large Council members in the first round of elections. These initial elections would be reviewed by the ICANN Board for their fairness and effectiveness before subsequent elections would proceed. This notion of staggered elections also was widely criticized in the course of our study.

For Council elections or board elections, or both, ICANN could employ a voting scheme that yields electoral outcomes in which the proportion of the total seats in the elected body held by representatives of a given constituency approximates that constituency's proportion of the total voting electorate. Cumulative, limited, and choice voting (also known as single transferable vote) are three such systems.

Compared to winner-take-all schemes in which 50 percent (plus one) of the electorate chooses all or nearly all the elected representatives, proportional representation voting systems increase the number of voters whose ballots help elect victorious candidates. Under such systems, a broader and more diverse range of voices would be represented in ICANN's policy-making bodies, and the risk of capture would be reduced as the seats would be spread out among different constituency groups.

Some people have expressed concerns, however, that proportional representation voting schemes facilitate the election of representatives from constituency groups with viewpoints that are held by rather small percentages of the electorate. Proportional representation voting schemes may lead to the election of board directors or council members with marginally-held views that may be aimed at moving ICANN outside its narrow policy scope.

Option 10: Hold all elections at one time

Many interested parties feel that all open seats should be filled in one round of elections, as opposed to the two-stage process currently outlined in ICANN's bylaws for its initial At-Large elections. This argument applies to member elections for the At-Large Council or the At-Large board seats. According to this viewpoint, member interest in the elections will wane if voters have to vote on two occasions, and the ability of people holding minority viewpoints to coalesce and achieve representation will decrease as the number of seats at stake in an election decreases.

Others note that it is important for ICANN to have a "test run" for its elections, so that improvements can be made before the remainder of the At-Large board members are elected.

Option 11: Deliberative Polling

Deliberative polling in an ICANN context would involve selecting a random sample of the membership, educating that sample thoroughly on the issues, and allowing them to select the At-Large board directors (among other potential roles).

Deliberative polling almost completely obviates the risk of capture as the board electors are selected randomly and are unknown prior to their selection. It also ensures that those selecting the board directors are a representative mix of the entire electorate, while proactively working to educate electors and providing them with a deliberative decision-making setting.

Deliberative polling requires a significant time commitment from those selected randomly to serve as electors or policy-makers - probably an unrealistic expectation of most prospective members. In addition, it is unlikely that At-Large membership would appeal to people if it carried only a slim prospect of being selected to serve as an elector or member of the "deliberative poll." This could well result in a disengaged membership that serves its board accountability role poorly.

We share the basic values behind proportional representation voting systems - increased opportunities for minority voting blocks to win representation and a diffusion of power among different constituencies. We recognize that the direct elections which we advocate will result in a total of only nine seats up for election. This limits the opportunities and potential efficacy of proportional representation voting systems, but ICANN should nonetheless work to incorporate some form of proportional representation voting in implementing direct elections.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

4.6 Other Important Suggestions

Many who participated in our study support limiting language for ICANN that delimits ICANN's powers in a very public way - stating the authority that ICANN does and does not possess.

Particularly since there is wide recognition that capture of ICANN by a narrow special interest cannot be avoided to an absolute certainty, it is critical that protections of ICANN's narrow mandate be constructed around ICANN's political processes. Such limiting language would discourage candidates and interest groups from campaigning on issues beyond ICANN's policy-making scope.

Detractors of limiting language point out that ICANN's bylaws already lay out its limited authority, and it is nearly impossible to further delimit ICANN's authority in enough detail to clarify it in every real-life case. In addition, the rapidly changing nature of the Internet counsels against drawing ICANN's powers too rigidly.

Option 13: Policy review role for the At-Large membership

Some have suggested that the At-Large membership, perhaps jointly with the Supporting Organizations, could play a role in reviewing board policy to ensure that it remains within ICANN's narrow mandate. This could be done indirectly through the selection of a review committee or directly through a referendum process.

Supporters argue that expanded roles and responsibilities for the membership will lead to a more engaged membership that will better hold the board accountable for its decisions. The At-Large membership, which is to represent the broad and diverse global Internet community, is an appropriate policing body of board action.

A sunset clause would provide for a periodic review of the election process and clarify formal procedures for making changes to that process.

Some participants in this study recognized that no matter how well ICANN constitutes its At-Large elections, mass elections for an international organization of this sort are unprecedented. Because ICANN is entering untested grounds, it cannot expect with a high degree of certainty that it will succeed in its first attempt to build elections that will produce the desired fair and democratic results over time. This dynamic is compounded by the rapidly changing nature of the Internet, which leaves open questions like who will desire and deserve representation in ICANN in the future.

It must be recognized, however, that a sunset clause should not create a complacency that results in initial decisions regarding the At-Large elections being made without full and careful consideration. Certainly, election procedures that seated those holding power will then be more difficult politically to change.

Option 15: Creation of an independent election authority

The election experts who we consulted in this study advocated for an independent election authority with broad powers to implement the At-Large election plan developed by the board and to monitor the elections as they are conducted. These experts argued that such an authority will enhance the public credibility of the elections and the trust of the membership in the fairness of the results, and they pointed out that no matter how open and transparent the process, an election organized by the ICANN staff with direction from the board will lack this public credibility and member trust.

We believe that limiting language, a sunset clause, and an independent election authority are all important steps for ICANN. Limiting language is critical to clarifying for potential voters and candidates the nature and scope of the organization they are participating in; a sunset clause recognizes the experimental nature of the At-Large elections and the rapidly changing nature of the Internet, as well as the corresponding need for regular review of the election process; and an independent election authority is fundamental to the public legitimacy of the elections.

[Top] [4.1 Election Structure] [4.2 Membership] [4.3 Putting candidates on the ballot] [4.4 Campaign Rules and Regulations] [4.5 Voting systems] [4.6 Other Important Suggestions]

[Next: 5. Recommended Principles for Moving Forward]

[Return to Table of Contents]

10: These ideas were also included in the MAC final recommendations, "Membership shall expire 30 days afer the annual election of Directors and must be renewed annually." (Full text available at : http://www.icann.org/macberlin.htm)

11: These arguments apply as well to fees and other substantive barriers to membership. Such barriers work against ICANN's legitimacy and cannot be counted on to achieve other goals like building an engaged membership or making capture more difficult. We do, however, support rules that specifically improve ICANN's ability to audit the electorate and prevent voter fraud, such as requiring members to register a reasonable period of time before the election.

Free Speech | Data Privacy | Government Surveillance | Cryptography | Domain Names | International | Bandwidth | Security | Internet Standards, Technology and Policy Project | Terrorism | Authentication | Right to Know | Spam

Our Mission / Get Involved / Staff / Publications / Links / Search CDT / Jobs / Action! Previous Headlines | Legislative Tracking | CDT's Privacy Policy