If you use the fourth power law to calculate road stress, a typical car is on the order of five orders of magnitude more damaging to roads than a typical bicycle.
There's no way to sensibly scale this to licensing costs - if, like in my jurisdiction, car registration is ~$100/yr, proportional cost for a bike would be on the order of a tenth of a cent per year - and that's before you take into account all the other negative externalities of cars that bikes don't have.
I drive much more than I bike, but I'd happily pay increased car tax to actively subsidize bike users and to get more cars off the road.