That's not a very useful statistic. It refers to whether the person was living in San Francisco when they most recently became homeless. But plenty of scenarios are captured in it that don't at all line up with the impression that the statistic is intended to give.
E.g. a homeless man comes to San Francisco and lives on the streets. He gets temporary assistive housing but is kicked out for criminal activities after a month, and is back on the streets. Do you think that statistic captures him as an outsider homeless, or an insider?
There's a big continuum here, but it's rare for someone to have been living here and paying their bills for a couple years and then become homeless because they're kicked out by an evil landlord or whatever. And even when it happens, San Francisco is pretty dang effective at getting and keeping them off the streets.
That's sort of all besides the point, because the actual homeless people that people get angry at--that homeless woman setting up a tent to block my fire escape, and that homeless man who punched me in the head in broad daylight on Market Street--don't respond differently to interventions regardless of whether they are an insider or an outsider.