Where I live, I don't own a car largely because the "jobs" for which having access to a car is critical is handled by a reliable, efficient and safe cab/taxi on demand system (you need to get somewhere quickly, carry a lot of stuff, take your family/friends to A&E at night etc). So then I can rely on the very good Public Transit system for the bulk of my regular commute needs confident that I can get a cab when I need to.
If the reliable cab/taxi on demand didn't exist (which was indeed the situation before Uber and it's ilk), I would buy a car irrespective of how good the public transport system was (it was good even then) and irrespective of taxation dis-incentives. Once you own a car, for many use cases it becomes a lot more convenient and has low incremental cost to just drive.
As you may note, these "jobs to be done" vary by geography, lifestage and income - so while general principles can be drawn up, they need to be locally customized.
It was an epiphany I had in relation to my staunch pro-bike/public transport internal discourse. It occurred to me that we need to focus on the main motivations for driving and public transport in order to make them both better. Road networks should focus on freedom and personal autonomy, and public transport should focus on throughput efficiency. Making PT better at throughput by expanding the networks would give people more freedom on the road network too by reducing traffic.