Contesting Identity in Collaborative Knowledge Production: Evidence from Coming-Out Announcements and Wikipedia Biographies

Abstract

When a public figure announces a transgender or non-binary identity, every institution that curates knowledge about them must decide whether and how to update the record. We study this process at the moment it is sharpest: on English Wikipedia biographies in the week of a public coming-out or pronoun announcement. Combining hand-verified announcement dates for public figures with complete revision histories, we estimate stacked event-study models comparing each biography to pageview-matched control biographies of cisgender public figures. Editing conflict, measured by reverts, rises roughly eight-fold in the announcement week, an effect that survives wild cluster bootstrap and randomization inference, and is not driven by any single event, and decays within about a month. Total editing rises almost proportionally: the share of edits that are reverts is flat, indicating that conflict scales with attention rather than reflecting a shift to a more conflictual editing regime. Administrative closure is tied to the events themselves: a third of announcements draw a protection lock on the biography within eight weeks, against one lock across 80 attention-matched control windows, with “biographies of living persons” the leading justification. The results characterize how a self-governing epistemic institution absorbs a contested re-classification of a person: through a short, intense burst of decentralized contestation followed by administrative closure, rather than a durable shift in editing behavior. The data identify the existence, dynamics, and governance of contestation; they are silent on the motives of individual editors, which we leave to future work.