For years, everyone entering Tel-Aviv from the east (e.g. from Jerusalem or from the airport) during the morning rush hour got stuck in the usual traffic jam for about an hour or so. Every day. An hour or so. About a week ago a new toll “fast-lane” has been opened, bypassing the traffic jam. Part of the intention is to reduce traffic into Tel-Aviv: public transportation as well as carpools use the lane for free and a huge parking lot has been set up at the entrance with free public transportation to Tel-Aviv provided. (There are also various criticisms of this lane, such as the fact that a lane that was once “free for all” is now “for the rich”.)
What is most interesting and novel, certainly for this blog, is the pricing mechanism: The price is set dynamically and may vary from 6 NIS to 75 NIS according to traffic (that’s from less than $2 to over $20) — and is displayed prominently at the entrance to the toll way. Automatic systems monitor the number of vehicles on the lane and their speed and adjust the price as to ensure a driving speed of 70 km/h. It seems that the designers take it for granted that an equilibrium will be reached: as prices go up, less drivers will elect to use the fast lane, reducing congestion and increasing the speed, and when it rises above 70 km/h, prices will decrease again, presumably boosting demand, etc. This makes sense, but a skeptical engineer may question each of these logical steps. Will drivers even be sensitive to the prices? Maybe they will interpret a high price as a signal that traffic outside of the lane is really awful, encouraging more drivers to take the fast lane? I would tend to view the consistent emergence of an equilibrium as a non-trivial validation of economic theory. Let’s wait and see. The last that I’ve checked, there weren’t enough people using the fast lane even at the lowest price. I certainly enjoyed that when I traveled to Tel-Aviv yesterday, reaching my destination almost an hour earlier than I’m used too.
Turing's Invisible Hand 
Hi Noam,
Nice post!! I’ve always wanted something like this on American roads (and for that matter on Indian roads)- but I guess political will can outweigh efficiency. That said, just fyi, Singapore’s congestion charging system automatically adjusts to traffic & time of day, or so I’m told.
This is interesting indeed. It would indeed be interesting to see what kind of equilibrium is reached. Indeed in an idealized world, there will be a clear relationship between the commute time on the free lanes, and the price of the toll lane, and it would be neat to see a plot of that to validate the assumptions.
This comment brings an interesting divide to light. When I see the word “efficiency” as used here, I imagine some function that is being optimized at an efficient outcome. To me, say when talking about how to use public resources such as highways (built on public land, using public capital), the choice of this function itself is a political question, with no obvious correct answer. I would view the political process itself as some kind of imperfect negotiation to agree on the function itself.
These are generally referred to as “HOT” (High Occupancy Toll) lanes in the US. If you have several people in your car they are free,else you pay the toll. Several states have implemented them, a list of which can be found at this wikipedia page http://en.wikipedia.org/wiki/List_of_HOT_and_ETL_lanes_in_the_United_States
I think that I was interested in the more specific congestion pricing toll roads.
i don’t understand what you mean by `emergence of equilibrium’. I mean, what combination of price and speed is an equilibrium ?
Well, I suppose that what I really would want to theoretically see is a graph of the actual speed as a function of the price, and indeed have it be monotonically increasing and continuous. That would ensure in particular that there exists some price which makes the speed 70 km/h.
Practically speaking, what I would call an equilibrium is no more than the actual speed being always around 70 km/h with prices varying according to congestion as to make it so.
But if they adjust the price according to the speed, then the shape of this graph is determined by their adjusting mechanism more than by any economic theory, so i wouldn’t say it validates the theory.
Actually, after thinking about it again I realize I don’t really understand the adjusting mechanism since it is not clear to me what actually happens when the speed in the tall lane reaches 70. But given a formal definition of the mechanism, my guess is that economic theory will not really predict an equilibrium but would just define the speed that is generated by the mechanism as the equilibrium outcome
You are obviously right that the details of the price adjustment mechanism matter and I’d really be interested if they were disclosed. Still, to me that seems like a standard “control theory” task, and the more mysterious part is the human behavior part of the loop.
Kamal Jain sent me a related preprint of a research paper: The Design of Everyday Mechanisms: A Toll Bridge Pricing Example by Nikhil R. Devanur, Kamal Jain, and Sven Seuken.
This is indeed an interesting price mechanism. Of course the precise details are interesting. Is the price only a function of the speed at the fast lane or is it also a function at the speed at the slow lanes. And how the price increases as a function of the (70 – the speed).
Taking toll for a lane that was once free as a vehicle to increase overall social welfare while not is acceptable. I hope Israeli policymakers will adopt a similar policy to the investors who are behind the huge gas discoveries near the Israeli seashores. In the process of negotiating the future terms for these guys it was assumed that they will get the right to produce the gas and the only issue is the precise terms. As a matter of fact, there is no apriori reason to believe that those who discovered the gas are the best players to produce it and it seems that the only real commitement for them is to guarantee them very good gains on the investement they already made (which is a tiny percentage of future investements). Why not open the production stage via some auction to competition? The government can put a lower threshold on what it expects and more on from that point.
singapore public speaking…
[…]Speed-Price Equilibrium « Algorithmic Game-Theory/Economics[…]…
Current XFINITY Cable TV Bundled Packages…
[…]Speed-Price Equilibrium « Algorithmic Game-Theory/Economics[…]…
Best Cable Deals…
[…]Speed-Price Equilibrium « Algorithmic Game-Theory/Economics[…]…
should the number of vehicles entering and the speed be the functions of pricing? But that’s a formula from business perspective and consumers can be easily distracted by reverse psychology end up with choosing a wrong gate.