As we have discussed repeatedly on this blog (e.g. “Middle-aged White Guys“), one of the most puzzling issues in analyzing Wikipedia is its continuous decline in active editors since 2007, shortly after a period of exponential growth:

The number of active editors (>=5 edits/month) in English language Wikipedia (Halfaker et al. 2013)

Active editors (>=5 edits/month) in English language Wikipedia (Halfaker et al. 2013)

Aaron Halfaker, together with R. Stuart Geiger, Jonathan Morgan and John Riedl, has now published results of their research efforts to understand the reasons behind this editor decline in American Behavioral Scientist under the title “The Rise and Decline of an Open Collaboration Community: How Wikipedia’s reaction to sudden popularity is causing its decline” (see Preprint PDF).

One of Halfaker et al.’s core findings is that, while the proportion of desirable newcomers entering Wikipedia has not changed since 2006, the proportion of them being reverted in their first session has increased (“good_ faith & golden” refers to sub-groups of desirable newcomers):

Proporation of reverted newcomers (Halfaker et al. 2013)

Proporation of reverted newcomers (Halfaker et al. 2013)

One of the main reason for this increase in reverts even of good faith contributors lies in algorithmic responses to Wikipedia’s enormous success, as Halfaker writes in his summary of the paper:

In order to maintain the quality of encyclopedic content in the face of exponential growth in the contributor community, Wikipedians developed automated (bots) and semi-automated tools (HuggleTwinkle, etc.) to make the work of rejecting undesirable contributions waste as little effort as possible. These tools are apparently effective at their job.

In other words, it was the successful implementation of algorithmic bureaucracy in form of bots that turned away larger portions of potential future editors. As a result, Halfaker et al. (2013) suggest that

Wikipedia has changed from “the encyclopedia that anyone can edit” to “the encyclopedia that anyone who understands the norms, socializes him or herself, dodges the impersonal wall of semi-automated rejection and still wants to voluntarily contribute his or her time and energy can edit”.

What I would be interested in is a systematic comparison with other language versions such as the German version, where newcomers have never been greeted by bots. However, also in the German Wikipedia bots are of increasing importance in terms of edit counts (see the paper “Work-to-Rule” on algorithmic governance in Wikipedia by Claudia Müller-Birn, Jim Herbsleb and myself). In any case, comparisons of different language versions would help to more clearly differentiate between algorithmic and other reasons for Wikipedia’s editor decline.