Right, because the air in Los Angeles is barely breathable, just as it was in the 70s. Wait, I got that wrong; the air in L. A. isn't too bad these days. What were we talking about? Oh, yeah, the effectiveness of regulations. On that topic, now I'm not saying there's direct cause-and-effect here, but...

Cleaner air in LA is primarily due to two technological breakthroughs: positive crankcase ventilation, and catalytic converters. That's the proper way to do regulations: once a cleaner tech is invented, make it mandatory. The other way around, mandating a steady decrease de-coupled from technological advances, is more likely to lead to cheating.

But how do you direct the industry to invent cleaner tech without the pressure of emission regulation?

A concrete solution which would have avoided "Dieselgate" is if the EPA had made urea NOx removal tech mandatory on all diesel vehicles. Saying "cars in the future must be X% cleaner" is needed, but it's not a complete solution.

Plus, more often than not, the technology to make cars cleaner is available, but expensive and so manufacturers don't use it.

Water-methanol injection, for instance, is well proven tech (since WW2), but not adopted at large scale because it requires changes to gas pump infrastructure etc. The only way to overcome a chicken-and-egg problem like that is governments making it mandatory.

Are you implying that we should abolish pollution restrictions on car manufacturers?

Free market is so well-known for handling negative externalities well...

Not necessarily, no.