On the regresivity issue, the original discussion in this thread was why the tax is politically unpopular. And the answer is: because it's perceived as regressive by the voting public, and acts regressive within the driving segment. You point out to a widescale redistribution effect to the non-driving public that counteracts this, but I disagree that it makes a significant difference outside a very narrow intellectual public like this forum. Congestion charging will always be an uphill political struggle.

Thank you for the perspective on the voucher idea. Yeah, in a perfect world, it's equivalent with a congestion tax coupled with a tax deduction of equal amount, I've hinted to that. But you would need to dynamically change the amount of the tax to keep up with road supply and demand, and that will clearly be be an explosive political decision. Also, for cash in hand to be psychologically equivalent to public services, the voters have to have full trust in the public spending process, which seems unlikely. So in practice revenue generating vouchers could be much more palatable. The additional transaction costs and friction could be negligible compared to a traditional congestion tax: to drive more than your allotment, you load your driving wallet with cash that is deducted at prevailing market voucher prices; at the end of the year, if you generally drove less than the allotment, you will find a nice tax credit waiting for you inside the public tax internet gateway.

By the way, I'm a city councilor in my European town and that's why I might show a strong slant towards practical ways to achieve political consensus, as opposed to just going for the superior theoretical solution. Cities rarely move on theoretical optimum paths and recently, of all idiotic policies, I had to struggle for ending the public subsidy of free parking. No-brainer, but wait until a significant part of the public received if for decades and strongly believe it's some sort of natural right.