Wikimedia Foundation elections/2021/Post Analysis

The Movement Strategy and Governance team prepared this report about the 2021 Board of Trustees election. The MSG team was the first team of facilitators to focus on supporting the Board election. Outreach, content translation, and conversations increased participation. The outreach, translations, and engagement were well-received. MSG succeeded in increasing the diversity of candidates and participation. We did not succeed in increasing diversity of the Board.

This report will help us look forward to the future. This report reflects what went well in this cycle while addressing some of those things that did not. This report follows the timeline of events in the election process. A summary of the major areas of concern and recommendations close the report.

Retrospective Insights

Planning and Outreach

The 2021 Board of Trustees election was unique in many ways. This was the first time using Single Transferable Vote. This was the first time with a staff team dedicated to supporting the election and outreach. The volunteer Elections Committee typically runs processes with limited support from staff.

The responsibilities were not clear between the Elections Committee and the staff team. Several unanticipated situations emerged during the election process. (For example: the Elections Committee did not expect so many questions about their actions.) This showed the community, staff and contractors do not fully understand the role of the Elections Committee. This is something we will need to address.

One innovation that did go very well was a new approach to outreach. The Board listed outreach as an important role for the Wikimedia Foundation during the election. The Movement Strategy & Governance facilitation team developed a plan for increasing the number of voters and diversifying regional participation. The MSG team created the Election Volunteer program to provide additional language and outreach support. Election Volunteers started as a pilot program to fill outreach gaps. It was so successful this program will continue.

Election Volunteers promoted the election by translating and sharing messaging in up to 61 languages. They hosted conversations about the election in more than 50 languages. They encouraged community members to take part in all areas of the election. The staff and volunteer outreach was successful.

We discovered some members from emerging communities felt uninformed about movement governance. They chose to not take part in the election. In the United States we encountered apathy and lack of engagement. Participation in the election increased by 1,753 voters over 2017. Turnout was 10.13%, 1.1 percentage points more than 2017. Two hundred fourteen (214) wikis took part in the election compared to 175 in 2017. While the MSG team missed the goal of 20% voter turnout, we did not hit that goal in part because of these issues. The 2021 Board election taught us about what we can do better for the next election cycle.

Areas to improve for 2022

  • A refreshed scope for the Election Committee, communicating their role and responsibilities.
  • Discuss and agree on policies and practices for Board of Trustees elections.
  • Motivate emerging communities to take part in governance conversations and practices all year. Consider self-selection and the cultures that are motivated to self-nominate.
  • Integrate translation planning and Election Volunteer participation into all public-facing governance communication. Consider how graphics can overcome language barriers.
  • Collect more data around Election Volunteer activities. This includes translations, events, and community engagement. This will help us make meaningful changes in the program.
  • Provide clearer information about translations so community members can take part. This includes distribution dates of messages and location of texts for translation.

Call for Candidates

Twenty candidates submitted their applications during the Call for Candidates. One candidate withdrew their application. Over half of the 19 candidates were from areas outside North America and Western Europe.

Previous elections listed candidates' applications in submission order. The MSG team tried to do something more equitable by creating a randomized image gallery. Some candidates expressed concerns of bias and privacy violation. They feared they would be judged based on their appearance. This concern was largely but not entirely voiced by white western men.

The Board of Trustees identified some desired skills for the incoming Board candidates. As part of the candidacy process, candidates completed the Evaluation form. MSG created a chart outlining the Board’s skills and experience assessment. The Evaluation form received mixed or negative responses from community members. Some community members had not supported the concept during the Call for Feedback. One person expected individual Evaluation forms from each trustee instead of a summary. Other people declined to consider the Board’s stated needs.

Each Board election the community has the opportunity to submit questions for candidates to answer. This usually lines up with the Call for Candidates. This year, the MSG facilitators supported the Elections Committee. The recommendation to the Elections Committee was to select about ten questions. This would allow for candidates to answer questions and have translations provided. We believe this provides a more equitable opportunity for candidates. Some community members posted the entire list of 61 questions for candidates to answer. Seven of the 19 candidates answered those questions. Three of those seven were elected. Four candidates who were not elected answered the entire list of questions. They were eliminated in calculation rounds 6, 8, 11 and 14. There is no strong correlation between answering all community questions and getting elected.

Areas to improve for 2022

  • Candidates need to know more about candidacy and being a trustee. Candidate Resources might fill this role. Build efforts to increase the pool of future Board candidates and their skills.
  • Reconsider the way of displaying candidates' applications. Wikidata-like statements may provide simpler processing of information by the voters.
  • Reconsider the use and implementation of the Evaluation form.
  • Structure the community question process with transparency. Involve more copyediting of the community questions.
  • Identifying and supporting women and non-binary candidates for the Board of Trustees.

Campaign

The MSG facilitators hosted eight campaign meetings. These meetings were for the community members to meet with the candidates. Some candidates could not attend all meetings. The MSG facilitators supported candidates in recording short video statements to post on Commons. Less than one-fourth of candidates recorded videos. The majority cited privacy, accessibility, and bias concerns.

The meetings presented challenges. Candidates did not know what to expect at meetings as this was new for 2021. Some candidates cited privacy, equity, and workload concerns. Recorded meetings, chat logs, and transcripts were uploaded to Commons. Two candidates expressed their concerns about this. They did not know their performances would be recorded and uploaded.

The Board of Trustees Community Affairs Committee met with candidates. Candidates mentioned this discussion came too late in the election process. They said it would have been more beneficial for them during the Call for Candidates phase.

Areas to improve for 2022

  • Candidates need clear written expectations and documented responsibilities of being a candidate. This includes privacy considerations.
  • Examine the design and interest for campaign activities. Consider the various stakeholders.
  • Meeting facilitation should support diverse needs of communities. It should provide a consistent quality of experience for attendees and candidates.

Voting

Voting eligibility for the election was defined for most voter situations. The complications came from several pieces of the process. One was defining voter eligibility criteria for several groups of community members. The other was creating the list of eligible voters to load into SecurePoll. Voting criteria for the developer community was established. This happened through discussions between developer community members and the Elections Committee. Affiliate volunteers not meeting the edit count threshold were unable to vote. Processing and loading the list of eligible voters into SecurePoll is a semi-manual process. It can take a series of days to complete.

A bug in SecurePoll exacerbated by a software update caused a two-week delay. After that, the voting process went without complication. Community members commented on the tool interface and discussed Single Transferable Vote.

Areas to improve for 2022

  • Find a solution to enable community organizers and affiliate volunteers to vote.
  • Create a way to know the home-wiki of voters instead of the wiki where they first created their account.
  • Provide clearer documentation about STV and voting. STV is a foreign term in many parts of the world where no institutions conduct this type of election.
  • Examine the voter interface for SecurePoll and the user experience.
  • Investigate reasons why eligible voters choose not to vote.

Post-election

While diversity was a goal of this election, the desired regional diversity did not occur. Community discussions after the election were about regional diversity and Single Transferable Vote. We believe Single Transferable Vote did help narrow the diversity gap but not enough. Candidates from regions outside of North America and Western Europe were closer to being elected than in prior elections.

Not all eligible voters had their votes counted in the election. The Elections Committee struck some valid votes. A quick review showed this was approximately seven of the 78 struck votes. This seems to have affected Wikimedia Foundation staff. The rationale is not obvious. They were not, for instance, staff who voted with both their staff and volunteer accounts. There is no clear solution or process to reinstate eligible votes. The Elections Committee has not provided the rationale for the removal.

The community began discussion and review of the election results after the data was available. SecurePoll provides a data dump as soon as results are tallied. As a result, The Kurier announced the election results before the MSG team could make the official announcement. This was to take place after the Board had a chance to review the results. This situation resulted in a rush to share the official announcement.

Areas to improve for 2022

  • Staff across departments are considering regional diversity beyond those areas identified above. They will be presenting a more thorough proposal later in the fiscal year.
  • Develop a workflow for contesting struck votes.
  • Share the preliminary results as soon as the data is publicly available.
  • Examine the possibility of providing more metrics about voting outcomes.

Summary of sizable areas of concern and possible solutions

This is a summary of the sizable areas of concern combined with some of the "Areas to improve" sections. These concerns listed below are the main concerns that emerged during the 2021 Board election. We should seek to address these in the next election cycle.

  • Missing regional diversity of trustees.
    • Some community members suggested designated regional seats. A brief (~4 week) feedback period before the next Board election can discuss ideas to improve regional diversity. Underrepresented regions are key stakeholders in this process.
  • Undefined roles and responsibilities between staff, trustees, and the Elections Committee.
    • The MSG facilitators will host discussions to better understand the roles and responsibilities of the Community Affairs Committee, the Governance Committee, the Elections Committee and staff members.
  • New processes helped, but did not close the information gaps.
    • Connect the needs of the Board and the Evaluation Form with the skills and experience that each candidate offers.
    • Reconsider the way of displaying candidates' applications. Wikidata-like statements may provide simpler processing of information by the voters.
    • Experiment with a voting advice application to help voters make informed decisions.
    • Be considerate about the amount of initiatives on-going in the community to help prevent community fatigue.
  • Lack of policies about Board election processes.
    • Documentation should be established about:
      • Expectations and limitations of the candidates during the election
      • Board election processes, like the community question process and the events during the campaign
      • Board of Trustees involvement
  • Challenging voting interface when ranking long lists of candidates through Single Transferable Vote.
    • Consider usability improvements for SecurePoll if STV is the voting system used.
    • If improvements to SecurePoll are not possible, provide voter support using other methods, like a Voter Guide.
  • Needed growth of community engagement in governance
    • Diversifying and increasing people involved in governance discussions is important. Investigating what people know about governance and what they need to become more involved will inform this work. It is critical to consider diversity, equity, and inclusion.