I remember reading something about how if they increased the number of medallions in NYC by 20% then the additional traffic would actually cause a decrease in total passenger throughput. I don't remember where I read it, though.

Here's one take on it (though I'm not saying this is the only one).

In New York, taxis are part of the public transportation network. Taxi drivers go through extensive training to know where pretty much any address in the five boroughs is, and the fastest routes to get there. In order to incentivize taxi drivers to go through this training, and to incentivize owners to buy and maintain vehicles, the city limits the number of vehicles on the road at a given time to incentivize drivers and owners to go into the taxi business. According to the 2014 NYC Taxicab Fact Book, there are 13,437 medallions in circulation -- assuming that 70% of them are on the road at any one time, and taking the figure that 94% of pick-ups occur in Manhattan or at JFK or LaGuardia, that implies that there are over 200 taxis operating per square mile in Manhattan ... that's A LOT !!! It's not uncommon to grant a limited monopoly to a public utility to ensure consistent service.

Why lease medallions? Utilization. It would be really hard (and unsafe, and unpleasant) for an owner-operator to be in the car for more than 80-90 hours per week, whereas a leased car can be in operation for almost the entire 168 hours.

All that said, driver knowledge is much less important than it once was, with GPS (though taking routing advice from a mapping service in Manhattan could be a bad idea). Safety is probably less of a concern now too, since Uber tracks its drivers' every move. And, it's clear that medallion owners are reaping a lot of profit at the expense of drivers willing to work for sub-minimum wages. So, it's probably time to renegotiate the contract.

I think that this regulation also helps to cater for people with disabilities. For purely profit-oriented service, there's no point in adjusting cars for their needs, because they're a tiny subset of client base and therefore don't matter.

There shouldn't be incentives for drivers to go into the taxi business other than the market. Taxi rates should not be regulated, supply and demand ought to be enough. What broke the market was the extremely expensive medallion racket. They don't limit hairdresser licenses or medical licenses, why taxi medallions? It's a highly distorted market often controlled by highly corrupt mob related interests. How is that a good idea? The free market is surprisingly efficient when not encumbered by Byzantine attempts to regulate everything.

The free market is surprisingly efficient - in other words, ruthlessly effective - at eliminating money sinks. Like passengers with disability, which are only tiny part of client base (and not too wealthy a part) but require special adjustments to cars. I don't see how fully deregulated taxi business would do anything but ruthlessly eliminate them from the client base.

Yes, maybe we could use little less regulation. My personal issue with Uber is not the concept of less-regulated taxis. It's that it's managed by a bunch of sociopathic assholes, who openly try to exploit drivers and the community for profit, painting themselves as white knights while using every underhand tactic they can get away with to sabotage competition.

Well if deadlyfoxgrandpa's post is true that might make sense. You have to consider the additional load on the infrastructure of having more taxis on the road in an already congested area. He says passenger experience might go down with more taxis!

Last I checked there is no government regulation for the taxi business when done on privately owned property (either that you own or that you have permission from the owner to use). I wonder how many hairdressers are currently operating on a New York City street? Probably no where near the number of taxis.

What's the limit on limo medallions? Those use public streets.