Wikipedia:Wikipedia Signpost/2013-10-02/News and notes

In what will be remembered as a game-changing week for Wikimedia grantmaking, the Foundation's executive director, Sue Gardner, published a forthright and in places highly critical statement, Reflections on the FDC process, and grantmaking staff revealed that the WMF will significantly strengthen its targeting of optimal impact in funding. These clear signs of the Foundation's intention to shift grantmaking strategies come just after 11 WMF entities placed on-wiki their applications for the October round of Funds Dissemination Committee allocations—the first of two rounds in the 2013–14 financial year in which the combined maximum budget will be US$8M. (The FDC is the volunteer committee, supported by a team of WMF specialist staff, that administers annual grants to chapters and other eligible entities.) The FDC also published the annual report for its first year of operation, 2012–13.

Straight talking from the executive director
Sue Gardner's statement emphasised the successes of the new system, but went on to express significant concerns about how the WMF's affiliated entities are developing: "too large a proportion of the movement's money is being spent by the chapters [whereas] the value in the Wikimedia projects is primarily created by individual editors: individuals create the value for readers, which results in those readers donating money to the movement. ... I am not sure that the additional value created by movement entities such as chapters justifies the financial cost".

While stating her confidence in all FDC members, Gardner wrote: "I am troubled by the FDC being disproportionately chapters-centric, and my concern increased rather than decreasing following the 2013 FDC member elections, which resulted in the two open FDC seats being filled by chapters Board members. ... the FDC process, dominated by fund-seekers, does not as currently constructed offer sufficient protection against log-rolling, self-dealing, and other corrupt practices. I had hoped that this risk would be offset by the presence on the FDC of independent non-affiliated members".

"With such a high proportion of [resources] now funding [chapters' staff and offices], we need to ask if the benefits are turning out to be worth the cost. It's possible that a well-managed shift to some staff support can help a volunteer community stay energized and enthused, but the risks and costs of setting up bricks-and-mortar institutions also dramatically increase, alongside sometimes difficult dynamics between staff and community".

In Gardner's view, we lack evidence that this spending "is significantly helping us to achieve the Wikimedia mission. I believe we're spending a lot of money, more than is warranted by the results we've been seeing. I am concerned by the growth rates requested by the entities submitting funding requests to the FDC." She challenged Wikimedians to ask themselves "whether there are more imaginative and agile ways of organizing our movement that will support our work better".

FDC staff: chapter allocations and the global south
Gardner's comments were, in part, reinforced and extended in presentations of unusual candour by FDC staff at the Foundation's regular Metrics Meeting last Thursday, streamed live and now available on YouTube (starting at about 37 minutes). WMF senior director of grantmaking, Anasuya Sengupta, said that analysing the impact of grants will be a priority in the coming year. One of her main points concerned the asymmetry in chapter allocations: "Last year, of the $5.65M we gave out, about $3M went to three chapters: Germany, France, and the UK. ... there are questions to be asked by all of us around where's the money going, for what, to whom, and how are our individual contributors being supported as best as we can".

Katy Love, the FDC's senior program officer, pointed out that the FDC guidelines generally allow a maximum annual increase of 20% in the funds allocated to any one chapter. Yet most of the 11 applicants are asking for considerably greater increases: Germany, the largest recipient—which already has access to generous levels of funding from German donors—is asking for 36% more ($2.43M). The UK's claim is up 32% from last year, Argentina's 34%, Switzerland's 38%, Austria's 41%, Serbia's 111%, Israel's 204%, and India's 401%. Only the bids from the Netherlands (up 20%) and Sweden (up 21%) were within the guidelines. This will present the FDC with "a very interesting process", Love commented.

Returning to the issue of asymmetry, the disparity between the allocations to the global south and north are more dramatic. Last year, Sengupta said, the global south received just 8% of WMF funding (from all sources, not just the FDC), with an average grant of $1554. The global north received 92% of the funds, and by the Signpost's calculation an average grant of $79,780. Despite this, reaching into the global south is a key priority for the Foundation. Asaf Bartov, head of WMF grants and global south partnerships, explained the rationale: "When our GS editorship is low, we are missing important voices, with different contexts, knowledge maps, hierarchies and categories." This leads to systemic bias from "the paucity of GS editors". More strikingly, Bartov pointed out that in terms of page visits "we are reaching 7% of the planet with our free knowledge – 7%! – so there's a way to go."

But the situation is complicated. In relation to grantmaking, he said, "it's actually hard to spend effectively in the global south. We are very eager to fund work in the global south, but it has actually been hard finding fundable projects that align with our global mission ... So that is a major component in the low number of funds that make it to the global south – it's not like we say 'no' a lot."

Lessons learned
So just what has the Foundation learned about the criteria for funding successful activities? Bartov first talked about a fundamental prerequisite:
 * "The sine qua non of most programs is a core of self-motivating active editors. Most of the things you want to do, most of the things chapters in the global north are doing, depend on this core. It can be as small as four or five people, but those people need to be actual active editors ... who edit because they like it, they enjoy editing Wikipedia or Commons, they get it, and they are inherently committed to our principles like NPOV; not people who are editing because there's a contest on and they want to win the laptop. "Where that core doesn't exist, it's very hard to deploy any other type of program. If you want [a GLAM partnership] with the National Museum of Cameroon ... how are you going to deliver on what you promise the museum if you don't have local editors who will do the work – write the articles, show up to meet the curators. So this is the big, big challenge for which we don't have an answer: how do you grow such a core ... in a certain country? ... we're now cautious about active investment where there is not an active community—although it's still possible if you give us a really great idea."

The WMF's primary formula, Bartov said, is now that "growth happens when community and outside resources come together", although he pointed out that "we are in disagreement with some parts of the community, with some chapters, about this conclusion: some people think we should still do work where no community exists."

What has been learned more specifically, then? He listed six critical advances in the Foundation's understanding that will be of interest to all Wikimedians who are engaged in practical projects:
 * "Single-session, general-audience outreach has negligible impact everywhere (for example, just giving a single talk about Wikipedia to whoever shows up—the conversion rate to editors is tiny, and yet we keep doing it". This, he said, is distinct from multi-session programs, or very carefully selected audiences, which yield slightly better results. "WMF contractors operating "on the ground" are too complicated and not effective enough ("we now only partner with grantees"). "Sustained attention to local communities yields actionable plans (a major focus in India and Brazil at the moment, which "has yielded actionable plans in those countries"). "WP Zero is effective, but it's still a challenge to get people to use the resource you make available". "Just making offline resources [like software] available in the GS is not enough; distribution is the key".

What the WMF has been learning from its role in grantmaking appears to be leading to definite changes in its funding priorities and methodologies. While this is likely to be met with controversy in the movement, there will be inevitable implications for affiliated entities in terms of how they shape their own priorities and the way they achieve their goals. During the Metrics Meeting, Katy Love encouraged all interested members of the community to review and comment on the FDC applications.

Chapter governance under the spotlight
Recent news concerning a related issue—the standards of governance required of FDC grant recipients—was first raised in the Signpost two weeks ago. In that edition we linked to an anonymous tip-off on the Wikimedia India mailing list on 15 September. The post claimed that two returning members of the chapter's executive committee, Pranav and Karthik Nadar, were in the paid employment of a third, Moksh Juneja, who joined them as a new member at the August election. This information did not appear to have been disclosed to voters, despite the obvious potential for conflict-of-interest in having a block of three members in mutual employment relationships out of a total of nine members.

The incident took a new turn yesterday when Moksh Juneja confirmed the employment relationships "have ended / [are] in the process of ending", and denied that this had any connection with the anonymous tip-off. Former president of Wikimedia India Arjuna Rao Chavala served on the chapter's election committee for the August election; his current membership of the FDC means that this information is likely to be discussed on the Committee during the current round, in which the Indian chapter has applied for US$178k for a predicted total budget of $217k.

An additional point of debate may involve another matter raised in the same edition of the Signpost: the allegation by one candidate, Santosh M. Shingare, that the chapter's election committee had failed to release the list of eligible voters 21 days before the election as required by the chapter's rules. In making the allegation that he would no longer participate in the election, Shingare declared he would no longer participate as a candidate.

Two weeks ago, we also reported on a conflict-of-interest issue in Wikimedia UK—that the Chartered Institute of Public Relations (CIPR—a professional body for public relations practitioners in the UK) has appointed current Wikimedia UK Secretary Alastair McCapra as its new chief executive. WMF general counsel Geoff Brigham has just published an opinion that "if a substantive issue arises at CIPR with respect to Wikimedia or its projects, I understand that Alastair intends to recuse himself from discussion and decision and appoint someone else at CIPR to act as the final decision maker on that issue within CIPR to avoid any appearance of a conflict of duty of loyalty. Alastair would also recuse himself on the WMUK board from a discussion and decision on any issue concerning CIPR. If an issue is particularly contentious and critical to the very fabric of either organization, Alastair may need to make a decision of resignation to address the potential conflict on that question, but such decisions can be handled on an issue-by-issue basis. The mere possibility of such a scenario does not necessitate resignation today."

Chris Keating, Chair of the WMUK board, has posted this message: "If there is a widespread view that, even with the steps we've outlined, it's not in the charity's best interests for Alastair to continue, then we will listen to that. However since Alastair posted the details of how he will handle this situation, only 5 people (myself included) have taken part in the resulting discussion. Some have posted at some length and in strident terms, but I don't yet see the picture I would need to see to be persuaded we are taking the wrong course of action here."

Discussion continues at the WMUK water cooler, beneath Brigham's statement.

In brief

 * Data security: On 3 October 2013 the WMF's deputy director and VP of engineering and product development, Erik Möller (Eloquence), revealed that there had been a four-month window between 29 May and 1 October 2013 in which the private user information (user email addresses, password hashes, session tokens, and last login timestamp) of 37,000 Wikimedia users had been accessible to some 228 individuals who had LabsDB access.
 * The LabsDB program, initiated in May, is intended to allow tool builders access to certain data from Wikimedia databases, data which is&mdash;or, in this case, should have been&mdash;automatically redacted before reaching tool builders. According to Möller, redaction failed in 30 databases owing to a schema incompatibility, allowing this user data to be read by anyone with a LabsDS account. Since this was caused by the addition of new wikis with different schemas (a column unused in earlier wikis was removed in the newer ones), affected wikis are only those added after May 2013. This means many smaller wikis, as well as Wikidata and Wikivoyage, were affected; the English Wikipedia and Commons were not.
 * There is no vetting process for who can become a LabsDS user. According to Krinkle: "[//meta.wikimedia.org/w/index.php?title=Talk:October_2013_private_data_security_issue&diff=5926114&oldid=5925940 anyone can sign up for an account and start developing software (within their own space)]"; this means that anyone could also have accessed a read-only copy of the database, and therefore the personal information. The upshot is that the data could be used for illegitimate means (although, as Möller emphasises, there is no evidence that it has been).
 * Access to the data was revoked within 15 minutes of the bug report, and affected editors have been informed via email and required to change their passwords. People who use their Wikipedia password on multiple websites are urged to change their password on these other websites as well, preferably to a different password than the new one they choose here. According to staffer [//meta.wikimedia.org/w/index.php?title=Talk:October_2013_private_data_security_issue&diff=5926509&oldid=5926331 Marc-André Pelletier], in the future, the Wikimedia Foundation will add a layer of redundancy to the checks to prevent this from happening again.


 * Wikimedia France: The chapter has announced that it has hired a new executive director, Nathalie Martin.
 * Arbitrator resignation: English Wikipedia arbitrator NuclearWarfare has [//en.wikipedia.org/w/index.php?title=Wikipedia:Arbitration_Committee/Noticeboard&oldid=575600826#Resignation_of_NuclearWarfare resigned] with immediate effect. He is the third arbitrator to depart in this year.
 * Quarterly update: The Quarterly update of the changes to the English Wikipedia's content policies has been released.
 * Openings: Several positions related to the Wikimedia movement have opened or remain open, including a Software engineer for the mobile team, an Operations engineer for the WMF Labs, a Fundraising coordinator, and a Tester for Visual Editor.
 * Australian politician copies Wikipedia: The Canberra Times has profiled an Australian representative who copied and pasted from Wikipedia and other online sources to report on a four-day trip to the United Kingdom.
 * Editing Wikipedia for credit: The New York Times (among other sources) has reported on the University of California, San Francisco's offering of course credit to students for editing articles on diseases. The effort comes as a result of WikiProject Medicine's outreach to medical communities.
 * Engineering report: The Foundation's engineering report for September has been published on-wiki, on the WMF's blog, and in summary form.