
Imagine a system which offered drivers something in return: a chance to drive on roads free of traffic congestion. Drivers would pay, but in return they would get a reasonably smooth drive to their destination. We have only so much road space, and more people want to drive than the roads can accommodate. Normally, when we have a commodity where demand outruns supply, we have some form of an auction. Essentially, it works like this. You tell (via a kiosk or website) a central computer you want to drive to a destination, when you want to get there, and the maximum amount you will agree to pay for the privilege. The computer then matches the number of people who want to drive in a given area at a given time. If more people want to drive than the roads will accommodate, the computer block those who offered to pay the least from driving in that area at that time. Then if more people still want to drive than the roads will hold, the computer repeats the process. The system bills everyone who drives in a particular area at a particular time the minimum accepted price; in other words, if you bid twenty dollars to drive downtown at nine am, but the system accepted bids for fifteen, then you pay fifteen dollars for your drive.
Auction-based congestion pricing involves significant technical problems,and it would require a significant investment. However, the city would probably recover money equal to a significant proportion of what we currently waste on traffic congestion, and according to the OECD, we currently waste over two billion dollars. If the city could recover even half of those losses, we could fund transit expansion easily.