Reflection — Inclusive Organizing for Open Source Communities
This, the second in a series of posts reporting findings from three months of research into the state of D&I in Mozilla’s communities. The current state of our communities is a mix, when it comes to inclusivity: we can do better, and this blog post is an effort to be transparent about what we’ve learned in working toward that goal.
If we are to truly unleash the full value of a participatory, open Mozilla, we need to personalize and humanize Mozilla’s value by designing diversity and inclusion into our strategy. To unlock the full potential of an open innovation model we need radical change in how we develop community and who we include.
We learned that while bottom-up, or grassroots organizing enables ideas and vision to surface within Mozilla, those who succeed in that environment tend to be of the same backgrounds, usually the more dominant or privileged background, and often in tenured* community roles.
*tenured — Roles assigned by Mozilla, or legacy to the community that exist without timelines, or cycles for renewal/election.
Our research dove into many areas of inclusive organizing, and three clusters of recommendations surfaced:
1. Provide Organizational Support to Identity Groups
For a diversity of reasons including, safety, friendship, mentorship, advocacy and empowerment — identity groups serve both as a safe space, and as a springboard into, and out of greater community participation — for the betterment of both. Those in Mozilla’s Tech Speakers program, spoke very positively of the opportunity, responsibility and camaraderie provided by this program, which brings together contributors focused on building speaking skills for conferences.
“Building a sense of belonging in small groups, no matter focus (Tech Speakers, Women of Color, Spanish Speaking), is an inclusion methodology.” — Non-Binary contributor, Europe
Womoz (Women of Mozilla) showed up in our research an example of an identity group trying to solve and overcome systemic problems faced by women in open source. However, without centralized goals or structured organizational support for the program, communities were left on their own to define the group, its goals, who is included, and why. As a result we found vastly different perspectives, and implementations of ‘Womoz’, from successful inclusion methodologies to exclusive or even relegatory groups:
- As a ‘label’. ‘Catch-all’ for non- male non-binary contributors.
- A way for women to gather that aligns with cultural traditions for convening of young women — and approved by families within specific cultural norms.
- A safe, and empowered way to meet and talk with people of the same gender-identity about topics and problems specific to women.
- A way to escape toxic communities where discussion and opportunity is dominated by men, or where women were purposely excluded.
- An online community with several different channels for sharing advocacy (mailing list, Telegram group, website, wiki page).
- As a stereotypical way to reference women-only contributing in specific areas (such as crafts, non-technical) or relegating women to those areas.
For identity groups to fully flourish and to avoid negative manifestations like tokenism and relegation, the recommendation is that organizational support (dedicated strategic planning, staffing, time, and resources) should be prioritized where leadership and momentum exists, and where diversity can thrive. Mozilla is already investing in pilot for identity with staff and core contributors which is exciting progress on what we’ve learned. The full potential is that identity groups act as ‘diversity launch pads’ into and out of the larger community for benefit of both.
2. Build Project-Wide Strategies for Toxic Behavior
Research exposed toxic behavior as a systemic issue being tackled in many different ways, and by both employees and community leaders truly dedicated to creating healthy communities.
Insights amplified external findings into toxic behavior, that shows toxic people tend to be highly productive thus masking the true impact of their behavior which deters, and excludes large numbers of people. Furthermore, the perceived value or reliance on the work of toxic individuals and regional communities was suspected to complicate, or deter and dilute intervention effectiveness.
Within in this finding we have three recommendations:
- Develop regionally-Specific strategies for community organizing. Global regions with laws and cultural norms that exclude, or limit opportunity to women and other marginalized groups demonstrate the challenge of a global-approach to open source community health. Generally, we found that those groups with cultural and economic privilege within a given local context were also the majority of the people who succeeded in participation. Where toxic communities, and individuals exist, the disproportionate representation of diverse groups was magnified to the point where standard approaches to community health were not successful. Specialized strategies might be very different per region, but with investment can amplify the potential of marginalized people within.
- Create an organization-wide strategy for tracking decisions, and outcomes of Community Participation Guidelines violations. While some project and project maintainers have been successful in banning and managing toxic behavior, there is no central mechanism for tracking those decisions for the benefit of others who may encounter and be faced with similar decisions, or the same people. We learned of at least one case where a banned toxic contributor surfaced in another area of the project without the knowledge of that team. Efforts in this vein including recent revisions to the guidelines and current efforts to train community leaders and build a central mechanism need to be continued and expanded.
- Build team strategies for managing conflict and toxic behavior. A range of approaches to community health surfaced success, and struggle managing conflict, and toxic behavior within projects, and community events. What appears to be important in successful moderation and resolution is that within project-teams there are designed specific roles, with training, accountability and transparency.
3. Develop Inclusive Community Leadership Models
In qualitative interviews and data analysis, ‘gatekeeping’* showed up as one of the greatest blockers for diversity in communities.
Characteristics of gatekeeping in Mozilla’s communities included withholding opportunity and recognition (like vouching), purposeful exclusion, delay of, or limiting language translation of opportunity, and well as lying about the availability or existence of leadership roles, and event invitations/applications. We also heard a lot about nepotism, as well as both passive and deliberate banishment of individuals — especially women, who showed a sharp decrease in participation over time compared with men.
*Gatekeeping is noted as only one reason participation for women decreased, and we will expand on that in other posts.
Much of this data reinforced what we know about the myth of meritocracy, and that despite its promise, meritocracy actually enables toxic, exclusionary norms that prevent development of resilient, diverse communities.
“New people are not really welcomed, they are not given a lot of information on purpose.” (male contributor, South America)
As a result of these findings, we recommend all community leadership roles be designed with accountability for community health, inclusion and surfacing the achievement of others as core function. Furthermore, renewable cycles for community roles should be required, with designed challenges (like elections) for terms that extend beyond one year, with outreach to diverse groups for participation.
The long-term goal is that leadership begins a cultural shift in community that radiates empowerment for diverse groups who might not otherwise engage.
Our next post in this series ‘We See you — Reaching Diverse Audiences in FOSS’, will be published on July 17th. Until then, check out ‘Increasing Rusts Reach’ an initiative that emerged from a design sprint focused on this research.