Even if we take street congestion to be a tragedy of the commons, it's not obvious that more cabs imply more congestion, since as part of a network of public transportation, they may help reduce personal car usage.
But even if we take it for granted, I still think your approach is pernicious. The issue with your - and the city's - method is that while the claimed problem is congestion, the solution is to limit the number of registered cars per service. But the error in that reasoning is that it presumes that they are perfectly correlated, and so it prevents market players from coming up with better solutions that reduce congestion without reducing the number of cars.
For example, services like Uber which rely completely on calls via the Internet have the potential to cause much lower congestion compared to a service that relies on having idle drivers drive around looking for people waving their arms.
If one wants to reduce congestion, then the only good solution is to tax cars driving in congested areas.
But even if we take it for granted, I still think your approach is pernicious. The issue with your - and the city's - method is that while the claimed problem is congestion, the solution is to limit the number of registered cars per service. But the error in that reasoning is that it presumes that they are perfectly correlated, and so it prevents market players from coming up with better solutions that reduce congestion without reducing the number of cars.
For example, services like Uber which rely completely on calls via the Internet have the potential to cause much lower congestion compared to a service that relies on having idle drivers drive around looking for people waving their arms.
If one wants to reduce congestion, then the only good solution is to tax cars driving in congested areas.