Attempts to artificially control prices for a scarce asset in the names of protecting equality will always fail. Prices will inevitably rise to their market level, the only difference is whether the owner of the good (the city) will reap the full benefit of the transaction, or whether the fruits will be split between (initially) secondary sellers and (later) law enforcement efforts.

The most important effect is that the incentives that might have favoured innovation to improve the efficiency of the use of the good and offer affordable alternatives are distorted in favour of incentives which favour unscrupulous competition for arbitrage gains.

Well - ignoring the fact that city maintained street parking is a public resource - I believe you are ignoring the cost of time (of the person searching for parking) in your analysis which, in effect, creates a perfect market equilibrium....or something.

A scarce resource is a scarce resource. It doesn't matter what intentions it was made with, or what rhetoric is used to justify a clearly defective pricing strategy.

time spent looking for parking is waste, aka deadweigt lost. Better to transfer money (zero sum) than waste time (negative sum)

[Citation Needed]

If the issue is fairness, there are two things people generally have they can give up: time and money. Right now finding parking is based on random chance. A better system, if we decide time is the thing people pay with, would involve people having to wait a pre-determined amount of time before they can take ownership of a parking spot, which would increase as demand did and vice versa. Additionally this duration could be determined by the amount of time they spend in the spot.

The status quo is obviously broken. The question is if there is an easy and effective way to improve fairness.

Forcing people to waste time is just wasteful. Everyone should pay, and certain desrving people (however that is decided) can get financial support or nontransferable credits (bit these have practical challenges)

It wouldn't be a matter of wasting time, it would be a manner of a person having to plan ahead before parking. If you want to park in a busy neighborhood, you may find that you can't actually do so for several hours, if others ahead of you in line have reserved those spaces. This is completely fair since everyone is treated equally, there is no 'wasted time' in this system that wouldn't have been wasted in the current system provided the spots are maximally full at all times. The current system results in some people getting lucky, others driving around in circles trying to find a spot, and others completely giving up. All of that time is being wasted. Beyond that the folks driving around are congesting the roadways, resulting in additional negative effects on the overall efficiency of allocating parking.

The main issue with designing a reservation based system is how efficient it is with regards to spots being continually filled, along with the constraint that it be considered fair by all participants. It seems like a design challenge but the current greedy algorithm probably sets the bar pretty low. It would be interesting to simulate.

Fwiw, SF is actually moving to market-rate pricing for meters and garages, with a pilot in a subset of spaces currently underway: http://sfpark.org/