Kling on free parking

Arnold isn't convinced:

If we abolished free parking, would parking spaces be scarcer? Keep in mind that if the price of parking went up, this would cause movement along the supply curve as well as along the demand curve. Maybe the total number of parking places would decline (it depends on elasticities), but the one result you can predict with certainty is that the number of unused parking places would go up. Is that necessarily welfare-improving?
The key is not to "abolish" free parking, but to a) abolish minimum parking requirements, and b) put prices or higher prices on congested municipal-owned parking spaces.  Both a) and b) will lower the demand for parking and a) will lower the supply of parking, so why should the number of unused parking spaces necessarily go up?  If you treat something as an appropriately scarce resource, it should be used more effectively.
There are plenty of DC restaurants which don't have their own parking lots, but they use paid valet parking and find ingenious ways to store cars more effectively.  The parking fee means that some people walk there or use the Metro, rather than driving and parking.  No one finds this arrangement especially objectionable and while valet parking is at a discount to market still it is priced.  At lunch time valet parking is less likely but still people pay to park, usually in nearby lots.  No one would suggest that these restaurants be forced to put in minimum parking.  Nor would anyone suggest that mandated minimums would be neutral with respect to parking efficiency.
I'm simply asking for the same switch in reverse, namely to do away with minimum parking requirements.  Very likely, such a change will have a bigger impact on future developments than on past developments (it can be hard to reconfigure a parking lot), although some malls might sell off or rent their now-liberated parking spots to other commercial ventures.
Pricing parking on busy residential streets is just common sense S&D and the price should vary with peak times.
Most of all, I am calling for "parking recalculation," so I am surprised Arnold is skeptical.  Maybe he thinks the recalculation won't bring much change (the Wal-Mart in North Dakota may never charge for parking), but in fact we find a wide variety of parking pricing practices around the world and even around the U.S., as laws and institutions and real net prices vary.
You could argue that politics already pushes us to somewhat efficient outcomes for policy, as indeed NYC does usually (though not always) treat parking spaces as more scarce than does Fargo, North Dakota.  Still, there is an obvious chain for political failure.  Development decisions are very often made on a one-by-one, sequential basis.  Other merchants, or nearby homeowners, fear parking overflow and they lobby as if this private cost were actually a social cost.  At each step of development, lots of parties are pushing for minimum parking requirements.  Some "once-and-for-all" parking policy decisions could limit this political incentive. 
Another simple public choice story is this: minimum parking limits the supply of land and boosts the returns to local homeowners.  It raises retail prices but many of the store's customers are from out of town, so that is a vote-winning strategy at the local level, namely scarcer land and higher prices for stores.
People who drive cars also have disproportionately more political power than people who do not, especially in most suburban areas.
Most of the time, legal quantity minimums have real effects on markets and they are not set efficiently at the political level.
Addendum: Here is Arnold's response to Robin, here is Robin on Arnold.