The broader effects of designing the world for car-efficiency wipe out the individual benefits, for most people. Exception for those who can afford to be driven around by others.

I was skeptical too the first time I read this kind of argument. I ran the numbers for my case, which was sitting around the median (commute duration) or significantly better than the median (household income, car cost) for relevant numbers, for my car-dependent middling-costs US city, and it was still roughly break-even without even factoring in not being able to make commutes double as exercise time.

I had to have a car. My life would have fallen apart without it, that's how big a benefit it apparently was. Yet if I actually examined what was going on, it wasn't providing any real benefit to me at all, just negating harm done by designing my city around cars. That's how the numbers worked out, much to my surprise. For most residents of that city it was worse, the city being designed for cars was making their lives worse.

The problem is that all the infrastructure that cars need (roads, parking lots) makes everything WAY further apart. For example, downtown Houston is literally like 25% parking lots by area. And that's not even counting other car infrastructure like roads. So to some degree, cars are just satisfying a demand for transportation that they themselves create

Denser, less car-centric areas are more economically productive than less dense areas. Car infrastructure prevents density. So I would argue that, at least in some cases, cars decrease economic efficiency