Wow. I just uncritically accepted when my local friends told me their fleet is old. It sounded logical... also the cars in that town really seemed 5+ years.
Parking is not a good argument because it depends where. My whole point is that ideal policy would be cutting down on cars in big cities but leaving them alone in rural areas where they are far more necessary, there are hardly any parking problems, and so on. This is entirely uncontroversial AFAIK it is generally accepted by urbanism experts that the feasibility of public transport depends on population density and especially on people commuting between well defined residential areas (suburbs) to working areas (downtowns usually, or industrial or commercial parks). Bicycling is more of a downtown thing too, beyond 5km distance it is less ideal. People who live in one small town and work in another one 20km away should drive, at best, car-share. So policies should be location based, such as London style entry taxes / congestion charges or Vienna style no free parking.
On a more meta note, on how to design a policy, in so far that it is possible, they should imitate the logic of the markets, that there should be an emerging, dynamic balance generated by different needs. Jack up the price of parking in every location where you really don't want cars. Or even reduce parking spaces. Doing it consistently would also cut down on the number of cars, because when people realize everywhere they want to drive to, mostly at work, is expensive, they will get rid of them, which is exactly what I did. However those people who mainly drive to places where parking is free will keep them, and this is a good thing, because it suggests they are sticking to low-density locations and it is a good choice there.
One of my favourite Less Wrong articles is Politics is the mindkiller. Part of the reason that political discussion so bad is the poor incentives - if you have little chance to change the outcome, then there is little reason to strive for truth or accuracy - but a large part of the reason is our pre-political attitudes and dispositions. I don't mean to suggest that there is a neat divide; clearly, there is a reflexive relation between the incentives within political discussion and our view of the appropriate purpose and scope of politics. Nevertheless, I think it's a useful distinction to make, and so I applaud the fact that Eliezer doesn't start his essays on the subject by talking about incentives, feedback or rational irrationality - instead he starts with the fact that our approach to politics is instinctively tribal.
This brings me to Joseph Bottum's excellent recent article in The American, The Post-Protestant Ethic and Spirit of America. This charts what he sees as the tribal changes within America that have shaped current attitudes to politics. I think it's best seen in conjunction with Arnold Kling's excellent The Three Languages of Politics; while Kling talks about the political language and rhetoric of modern American political groupings, Bottum's essay is more about the social changes that have led to these kinds of language and rhetoric.
Video of a related lecture can also be found here.