This isn't too surprising. If the goal is to reduce car ownership (and trips), then my first thought would be that adding ride sharing services could only help if it fills few remaining trips that a robust public transit system can't easily accommodate. Even if on a per trip basis ridesharing has greater externalities, that might be ok if only a small minority of trips need a car in the first place. Of course, that's obviously not true for already car-dependent cities.

Other aspects that the linked policy brief mentions also follows, such as ridesharing displacing transit more among more wealthy populations. To me, the most surprising finding is that ridesharing did not lower car ownership even in cities that aren't car-dependent. If I had to hypothesize about why that might be the case, I would say it would probably be due to ridesharing replacing transit, making it worse in turn. Especially since U.S. transit is so bad in general, making it much worse would render it useless in many if not most U.S. cities.