Two candidates appeared on ballots in the same election cycle. One was challenging for a seat on a five-member city council and had never held elected office. The other was running for a state legislature, and won. A year later, one of their Wikipedia articles was gone — folded into the page about the election itself, the name surviving only as a redirect. The other's article stood untouched, built in large part from coverage that started the day its subject was sworn in.
Wikipedia:Notability (people) treats political office the way it treats almost everything else: a handful of categories carry a presumption, and everything else has to prove itself. Members of national and state or provincial legislatures are presumed notable. So are heads of state and government, and judges on the highest courts of a jurisdiction. Nearly every other elected or appointed role — city council, school board, party chair, and any candidate for any of the above — carries no such presumption. It falls back to the general notability guideline, the same one that governs a musician, a company, or an unelected civil servant: substantial coverage from independent sources, not an inference drawn from the ballot.
Most of what a campaign generates is routine by the guideline's own definition. An announcement of candidacy. A joint appearance at a candidates' forum. A line in a roundup of who's running for what. A quote in a piece about polling two weeks out. These exist for nearly every name on the ballot, the losing candidates as reliably as the winner, because the local paper covers the race, not because any one candidate did something that set them apart from the others running. Coverage of this kind rarely moves a borderline case past the bar.
A legislative seat is evidence. A candidacy is an announcement.
What does move it is coverage that would have existed with or without an election on the calendar: an investigative piece examining the candidate's record, a feature profile unconnected to the campaign calendar, attention that kept appearing after the race ended and the news hook was gone. The test has less to do with the subject matter — politics generates plenty of ink — and more to do with whether the attention was earned by something other than a name appearing on a ballot.
Losing candidates without that kind of record rarely go through a contested deletion debate. The more common outcome is quieter: an editor proposes a merge, the content folds into the article about the election or into a list of candidates, and the name becomes a redirect pointing to the place where it still makes sense. Nobody disputes that the person exists. The article the campaign wanted just wasn't the shape the sourcing could support on its own.
For a campaign weighing whether a Wikipedia article belongs in its communications plan, the useful diagnostic runs backward from polling day: strip away the candidacy, and ask what coverage of this person would remain. A stack of campaign-season clippings and nothing else points toward a merge, win or lose. A profile, an investigation, a record that predates the race — and the office becomes almost beside the point. The sourcing was already doing the work.
