If I completely misinterpreted your main point and instead you are saying that it wouldnt change the way things are currently run in practice, then you are probably right; HOWEVER [and I think this is what the parent comment was getting at], the local governments would be far, far better off, at least allowing for some good to come out of this system
They can, but by doing so, they'll pay a lot and enrich the state rather than parasites. On one hand, auctions cost them money; OTOH, they can't charge more for the lease than what a taxi driver can earn. Those are conditions where free market regulation works: if they provide little service, they can only make little profit.
> [some] would get more cheap licenses, and [some] would get exorbitant fees.
As soon as you create a bureaucratic regulation body enforcing byzantine rules, you've created a weak point in terms of corruption resistance. Such an incomprehensible administration is exactly what lobbyists can successfully tune to fit their patrons' interests.
More to the point, the only reason why you might want to regulate the number of taxis is to make sure that every driver can earn decent wages without behaving recklessly. With this respect, the medallion system has been broken, seemingly beyond repair, by parasitic investors. Your proposal doesn't sound like a focused answer to the original problem "how to ensure that taxi drivers earn decent wages without having to behave recklessly?"
We don't worry about reducing work to a living wage in any other field - we allow free entry into carpentry, construction, business & don't require them to buy medallions before starting to work.
It's a general problem of resource allocation - just like cancer. Some group or cell identifies a resource and starts diverting or blocking it for their own benefit, and this increased power/energy is enough to protect the problem from unmotivated defenders.
"It is the world's most demanding training course for taxicab-drivers, and applicants will usually need at least twelve 'appearances' (attempts at the final test), after preparation averaging 34 months, to pass the examination"
"In all some 25,000 streets within a six mile radius of Charing Cross are covered along with the major arterial routes through the rest of London.... The Knowledge includes such details as the order of theatres on Shaftesbury Avenue, or the names and order of the side streets and traffic signals passed on a route."
Overkill in the era of GPS? Probably. But pretty cool.
Actually, New York already offers owner/operator medallions at a reduced rate. They are transferrable and leasable to a certain extent.
Now imagine you're a fresh aspiring immigrant just coming to New York. Who in his sane mind would loan you $200K with no collateral if you could just decide next year it's not for you and they couldn't even sue you - you have nothing. Who would drive the cabs then - bankers? There's not "anyone else" that could get a loan even of couple of thousands with no collateral and no way to recover it. The best you get is credit cards companies - they charge 20%+ APRs for a reason. Imagine having to pay off a $200K loan with 20% APR.
On the other hand, if you make the license cheap, the corruption potential would be overwhelming - if you have somebody distributing very limited supply of licenses costing $100 but allowing income of, say, $100K per year, an incentive for bribery and other trickery would be impossible to resist. Pretty soon the price would get close to the actual cost, only the difference will be pocketed by corrupt officials.
http://www.calgaryherald.com/opinion/op-ed/Nenshi+Hailing+be...
What needs to happen is for passenger interests to be given weight. One way to do that might be to have the cities offer some short-term licenses for single shifts on a spot market, and some long-term (maybe quarterly or yearly) licenses with a fee inversely related to the driver's score over the last period (determined by average rating by his fares, and/or relative average tip amount for the route).
Applicants for long-term licenses with good ratings and relatively high tips for their routes for the previous period would get cheap licenses, those with poor scores would get more expensive licenses, and those without a track record or filed for the benefit of non-drivers would get exorbitant fees.
This would have the secondary benefit for municipalities of drivers being incentivized to declare their tips.