Experimentation and Public Policy Design: The Case of LA’s New Toll Lanes

We know that there is a lot that we don’t know.   This suggests that we should run more field experiments to pilot potentially good policy ideas.   Today’s UCLA Daily Bruin offers some good news from Los Angeles.

“A new toll lane opening on the 10 Freeway, which runs east-west from Santa Monica through Los Angeles, could ease traffic for some UCLA commuters this weekend.

Starting Saturday, solo drivers will be able to use the carpool lane on the 14-mile stretch from Los Angeles Union Station to the 605 Freeway if they pay a fee anywhere from 25 cents to $1.40 a mile depending on congestion and time of day, according to Metro.”

Some UCLA Urban Planner will write a very nice paper studying the benefits of efficiently pricing public property.  We will learn some good lessons about the Tragedy of the Commons problem and somebody will have a nice pot of revenue collected from this pricing policy.   Other cities will learn from LA’s experiment and will be likely to mimic this smart policy.   I would like to know why the policy is only being adopted on the 14 mile stretch and whether it could be extended if the policy is deemed to be “effective”.    For some specifics about this experiment click here.

Author: Matthew E. Kahn

Professor of Economics at UCLA.

27 thoughts on “Experimentation and Public Policy Design: The Case of LA’s New Toll Lanes”

  1. The funniest thing about the future and cars will be their total reliance on cooperation. That is once testosterone (competition) is no longer is in the pilot’s seat,– once driverless cars are ubiquitous, the sensors in cars will work with GPS and traffic computers to cause autos to cooperate and micro-adjust to keep traffic maximally moving. This is already being discussed and designed.

    Why do I say that is funny?

    Because I live in a culture that honors competition and markets beyond all common sense. It’s a veritable religion with a thriving priesthood.
    And where you prove your worth (and your individuality) by the type of car you drive.
    And looky: Naked competition amongst drivers leads to traffic jams. While designed cooperation amongst machines smooths the flow and works better.
    You’d think there might be some takeaway in that wouldn’t you?
    Not for this culture…

    Side note: I used the word “testosterone” above because I recently read that globally car crashes are the second leading cause of death in men (tied with AIDS). In woman automobile accidents are a negligible factor.

    Second side note: Do you know what the leading cause of death (globally) in woman is? Answer here: http://oecdinsights.org/2010/03/04/hivaids-kills-more-young-women-than-anything-else/
    You can blame that on testosterone too…

    1. In most countries men drive far more kilometers per year than women, so of course they die more often in road accidents. The more telling figure would be road accidents per thousand kilometers driven.

  2. This line:

    “We know that there is a lot that we don’t know. This suggests that we should run more field experiments to pilot potentially good policy ideas.”


    From your later comments it is clear you already know the outcome:

    “Some UCLA Urban Planner will write a very nice paper studying the benefits of efficiently pricing public property. We will learn some good lessons about the Tragedy of the Commons problem and somebody will have a nice pot of revenue collected from this pricing policy. Other cities will learn from LA’s experiment and will be likely to mimic this smart policy.”

    FWIW, I agree with your sentiment (err… the first one) and hope that this pilot project is an enormous success. How we know that this is “smart policy” or which “benefits” will accrue is less clear.

    1. You really can’t know the social welfare outcome. It may lead to increased total fuel consumption, if all the Teslas and Prii pay tolls and all the klunkers move even more slowly on the creep lanes.

      Of course, the outcome for rich folk obviously a good one. And that is the only one that counts.

  3. How does this fit in with the fact that the Orange County toll roads have gone broke and are only being propped up by the State and fiscal juggling? http://www.ocregister.com/articles/toll-380898-roads-road.html

    Where will the money from the tolls actually go?

    What does this effort to implement tolls mean for future public support for building more “freeways”?

    By the way, koreyel’s comment above is spot on. How will cars be sold if the blow the other guy off the road thing is gone?

    1. How will cars be sold if the blow the other guy off the road thing is gone?

      Number of seats
      Cargo space
      Amenities (high-speed internet, # of USB ports, HDTV, sound system, hot tub, etc.)
      Fuel efficiency
      Lower maintenance cost
      Big honking tailfins

  4. I wonder how long it will take for the HOV/toll lanes to fill up and become no more speedy than the regular lanes.

    It will be interesting to see what happens, especially because this is one way to get actual revealed preferences for the cost of commuting time.

  5. It’s not clear to me whether the lane itself is actually new – a physical expansion of the road – or whether just the idea of charging tolls on it is.

    1. I believe you are correct — these were the carpool lanes. They told us they were going to build more of them. Not a peep about how they would actually be used.

        1. Then I don’t understand the whole thing.

          Is the point that the carpool lane is underutilized, so we want others to be able to use it, a little? OK, so some traffic shifts over to the carpool – excuse me, HOT – lane. Other lanes become slightly less congested. But don’t they fill up again, or at least show some increase after the initial reduction?

          IOW, I’m having a hard time grasping what the theory motivating this experiment is. By what mechanism is congestion going to decrease even though you have just as many, or more, cars using the same number of lanes?

          1. Well, I admit I have not followed every single move made by CalTrans and the MTA and so on. There has been construction to extend the carpool/HOT lanes, and I think they even snuck a second lane in there in many places. And they are trying to improve transit in this entire, ginormous Southern Cal region, so, while I enjoy b****ing about lack of voter input, well I can see why the powers that be don’t want to go that route. In fact, CalTrans is going to have some issues with expanding in parts of Orange County. But tough luck, I say. If we continue to undermine people’s faith in our elections, at some point it will be too much. I would rather have voters make bad decisions, and learn from them, as I believe is possible! than this nonsense we are getting now.

        2. At least around here (Minnesota) you don’t need any gizmo to use the HOV lanes if you’re carpooling. Motorcycles, buses, and cars with with at least two occupants can just go ahead and use them. There have been cases of drivers being pulled over and given a ticket despite having a mannikin in the passenger seat that they were hoping would fool the cops patrolling, looking for solo drivers without the gizmo.

  6. This doesn’t strike me as much of a solution to traffic congestion. The desirability of the carpool lane is premised upon it being proportionally less used than the other lanes. As usage of the carpool lane increases, it becomes less advantageous/fast. Thus, if the 10 Freeway has 5 lanes, with one being the carpool lane, as long as less than 20% of the traffic is in the carpool lane, in theory it should move faster (leaving aside the problem of a slow driver holding up traffic wanting to drive faster). Once the carpool lane has 20% or more of the traffic, it ceases to provide an advantage.

    Here are the problems:

    1. This seems to me to be a zero-sum game. If some traffic is diverted to the carpool lane because it moves faster, the remaining lanes become relatively more congested and, therefore, move even slower. Thus, the carpool lane provides a benefit to users while penalizing non-users.

    2. We end up with a caste or class system, since wealthier drivers–those who can afford to pay for a less congested lane–will be using the carpool lane, while drivers who can’t afford the fees are consigned to the regular lanes, which move more slowly. On public highways, do we really want to allow wealthier drivers to have a better class of service, particularly when it means worse service for everyone else?

    1. It is not a solution to traffic congestion.
      It is a solution to the problem that I, a rich person, want to get somewhere fast (using a road constructed with public funds) and I don’t want you peons slowing me down. It’s a solution that would only work, and only “makes sense”, in a society that is extremely stratified by wealth, so that the costs mean nothing to the target users, but are exorbitant to the loser peons.

        1. Yes why we are talking about peons, vassals and feudal lords….
          And getting a little ticked off about it….

          Ever hear of a fellow named Paul Singer?

          Elliott Capital Management, a vulture fund based in the tax haven Cayman Islands owned by conservative financier Paul Singer (a big donor to the Romney campaign), refused to accept the terms of the debt restructuring that was accepted by more than 92% of bondholders in 2005 and 2010. It has demanded payment in full, and has actively pursued its case in different courts across the world. A few months ago, the Argentine frigate Libertad, which ironically means freedom in Spanish, was seized in Ghana after a local judge ruled in favour of Elliott Capital Management. Judge Thomas Griesa has recently ruled in a district court in New York that the Argentinian government must pay $1.3bn to the same vulture fund – the full face value of their holdings plus accumulated interest starting in late 2001 – on the basis of an unusual interpretation of the pari passu clause in debt contracts.

          Why Argentina is now paying for its dangerously successful economic story | Jayati Ghosh and Matías Vernengo | Comment is free | guardian.co.uk

          I imagine Paul will pay to use fast lane…

  7. Here is one of my *other* beefs with economists — this idea that $1 means the same thing to one person as it does to another. And the idea that therefore, you can understand “revealed preferences” as a result of shenanigans like this one. All you are revealing is the underlying power relations of society, which determine wealth largely (yes, there is *some* role of individual effort/”merit” , just much much less than economists ever will admit).

    Who wants to go slower? If these people would just be honest, I would have more respect for them. Right now, the MTA is a bunch of grifters who rob busriders to help the subway, who lie by omission to get tax hikes extended, and they aren’t directly elected either. There is virtually no way, other than litigating, for anyone to influence them.

    1. There are actually solutions to this problem. One that has been suggested in the field of law is fines based on fractional income rather than straight dollar amounts. So park in a handicapped space, or whatever, and rather than a flat $300 fine, you pay 4 days income.

      For what I think are obvious reasons, this scheme has gained no traction in US law. However if it were applied to preferential access to publicly-provided infrastructure (eg fast lanes) I’d have no problem with the concept of these lanes, and in fact would then actually agree with the arguments of economists re efficiency and revealed preferences. (Arguments that today are just so much BS because, as you say, of the differential value of a dollar.)

      (This scheme is, IMHO, actually quite feasible technically. The Fastrack device could be tied in to your IRS info for example. It’s not perfect; it doesn’t tackle the wealth vs income issue, or even the larger fraction of disposable income after necessities issue, but it’s a start.)

      1. Very interesting. No doubt better than what we have now, though I can also imagine, as only a somewhat numbers literate person, that if my income fluctuated, plus the cost of the toll fluctuating by time of day, I’d be too confused to use the lane!!! Yeesh.

  8. An easy solution would be to tie fines and fees to the car’s registration value. Not perfect, but very easy.

    1. Didn’t some guy in Sweden have to pay a six-figure speeding fine a few years ago, because Sweden uses that approach?

  9. As I’m a computer geek, I know a lot of people on good incomes who are very carefully with a dollar. They probably wouldn’t use the pay lanes. But since I’m also an armchair economist, I think they’re a wonderful idea so long as the revenue is used for transportation infrastructure (particularly public transit, of which I’m a huge fan).

    If I could pay $5 to get home 15 minutes sooner and thus work 15 minutes more overtime while still going to bed at the same time, I’d more than recoup the cost. If I’m late for something really important, I’d still probably pay. If I’m in no rush, I wouldn’t. If someone can’t possibly spare $5, that’s unfortunate for them, but not really my problem. They still get to use the highway, just more slowly.

  10. 25 responses and no one seems to have read at the link.

    The plan affects both the 10 going east from downtown, a particularly congested road, and, I believe the 110 going south from downtown toward the harbor. Both sets of HOV lanes are being made into double-lanes — the work has been completed on the 10, with a net widening of the total road. I haven’t seen the 110, to see where they are at.

    The plan, as I understand, is to vary the toll with traffic congestion. Signs advise you what the rate per mile will be for a vehicle occupied by a single person. The transponders allow the driver to set the occupancy to 1, 2 or 3 persons, and the toll varies accordingly; a transponder set to “3” doesn’t pay a toll.

    There’s been careful consideration given to both privacy issues and the position of poor people. Poor people can get a discount and avoid the account maintenance fee of $3 a month (which is also forgiven if you make a low minimum number of trips, even if you never pay). And, you can get an anonymous account. (Motorcycles have to have a transponder, but are allowed to set it to 3, to avoid the toll.)

    The existence of two lanes is critically important. A single HOV lane seldom does anyone any good, and a toll would not help with that problem, at all. You can figure this out with a little basic queuing theory.

    I’m actually impressed with the thoughtfulness, with which this has been approached. Just going from 1 HOV lane to 2, proves that someone has matched experience with 1 lane, against simple, basic analysis. Hooray. I don’t know why we needed years of experiment, but some people are just stupid, I guess.

    The transponders have only been available at AAA, which discounts them 20% (compared to 10% off at CostCo), for a couple business days, they are short of supply. So, it will be a while, before there’s any traffic. There was none — zip, zero, nada — when I passed down the road on Saturday.

    Economists are always going on and on about the incentive effect of price. Personally, I think feedback is often the missing channel of information. No one would get on the Hollywood Freeway in Hollywood, if you could know before getting on, that it was fatally congested. But, naturally, most entrances give you no information; there’s no view of the freeway, prior to committing to the entrance ramp, in lots of places. Caltrans transmits real-information on congestion on freeways and even major streets, but lots of GPS-driven guides don’t communicate this information well. Congestion is a significant cost, even if there’s no price being charged. Lots of people, with discretion, will avoid adding to it, if they can, but they have to learn this information in a timely way.

    I’m fine with cheap, expeditious toll collection systems. Yay! Gas taxes are a pretty efficient toll collection system, despite the tiny number of electric cars. Higher gas taxes, please! But, private financing of tollroads is incredibly stupid and economists advocating for them should branded like Cain. Congestion becomes something the rentier demands from his cronies in government, destroying efficient administration. That experiment has been tried: don’t do it!

    1. Actually, I did read the page, long enough to notice that no criteria were given for what would be considered “success.” I did miss the part about poor people though. It has also received absolutely no attention in the news articles on the program.

Comments are closed.