Democracy requires any ban be "mutual coercion, mutually agreed to"; a majority must elect, and re-elect, a government that will effect the ban. This seems impossible in the current car hegemony; but as with ailing dictatorships the cars' media and market image of universal approval is increasingly unreal (more 1). Surveys already show that public opinion on reducing car use is actually 1/3 support, 1/3 opposition & 1/3 contingent (note 1). Only local and partial bans are likely in the immediate future, but these can be bulwarks for extension and replication.
Public acceptance could be achieved within the first 2 decades of the century because fuel crisis economics will reduce the numbers driving and protract the average interval between voting age and car ownership, as could future anti-collision measures. But public compliance in abandoning cars is contingent on:
These are crucial to the ban. Any government contemplating it must establish these and make them unable to be degraded, defunded or debundled from the ban legislation. The universal ban itself brings non-disadvantage. More influential people on public transport ensures its standards are maintained. Security is improved by freeing police from the task of averting car traffic carnage and by increased outdoor life and walking traffic (more 2). Vested interests: private car manufactures, land speculators, and bloated government roads departments will never accede and would need be defeated in corridors of power and public relations.
Mental hegemony can be countered with education and urban bans on car advertising and moving motorsport to venues on urban fringes. Phased implementation would need be carefully timed with the full installation of good alternate transport; enforcement in itself is the least problem for unlike the bogus alternatives a ban is policeable; a car is a car, big and obvious (more 3).