Where Are My Cars: SR-167 HOT Lanes

Revenue has fallen far short of expectations.

We’ve written before about the “high occupancy/toll” lane experiment on Washington’s SR-167. But for those unfamiliar with the concept: HOT lanes are special highway lanes that transit and carpools can travel in for free, but are also available to solo drivers who are willing to pay a toll. When the regular lanes start to back up, the HOT lane tolls increase. That way, the HOT lanes never get clogged, even when the regular lanes are full.

Besides keeping carpools and transit moving, the SR-167 HOT lanes have an additional value: they give researchers more nuanced understanding of how much people are willing to pay for a quick trip. And when we took a look at the SR-167 HOT lane data last year, the numbers surprised us: apparently, drivers really aren’t willing to pay much for a faster commute. Few drivers on SR-167 opted to the free-flowing HOT lanes, so HOT lane traffic volumes typically stayed low.

University of Washington PhD students Austin Gross and Danny Brent have taken our initial research several steps further. I’ll probably mention more about their work in a subsequent post. But in the short term, I was struck by their data showing how badly WSDOT transportation planners misjudged demand for the HOT lanes:


That’s right: actual HOT lane revenue in 2012 was about one-third of the “low case” projection that WSDOT made before the lanes were opened. That was likely due to two separate effects: fewer cars than anticipated using the HOT lanes; and lower toll rates needed to keep HOT lanes from clogging up. Both effects flow from flawed early assumptions and beliefs about how much demand there would be for driving in general, and for the HOT lanes in particular.

These findings don’t prove that HOT lanes are a bad idea! But they clearly have a bearing on today’s transportation debates. For well over a decade, the Washington and Oregon departments of transportation been planning massive highway megaprojects—wider urban highways, higher-capacity bridges, expensive tunnels, and more. And they’ve pursued these plans in the belief that bigger highways will ease congestion—and that drivers put a high value on congestion-free trips.

But the SR-167 HOT lane experiment shows that most drivers on that stretch of road simply aren’t willing to pay much for a fast commute. Which raises a question: given that drivers may not be all that willing to pay for a quicker trip, does it really make sense for taxpayers to invest so much in trying to give them what they won’t pay for themselves?

Note: just to be clear, the interpretations of Gross’s and Brent’s numbers are my own, not theirs. And their research goes much, much deeper into the specifics of pricing and driver willingness to pay; the chart above is just the tip of the iceberg. I’m very much looking forward to diving into their findings! And hat tip to GoodMeasures.Biz for the graphic!

We are a community-supported resource and we can’t do this work without you!



  1. norse says:

    Often debated question – especially in the context of the 520 bridge tolling. Morale of the story: if you don’t get the revenue you were targeting because your customers aren’t willing to pay your price point, the knee jerk reaction of raising prices to meet a target achieves the exact opposite of what you want.

  2. yvrlutyens says:

    The real moral of the story is that the demand for new road infrastructure is far less than people imagine. The willingness to pay a toll is after all a much better reflection of actual desire than a vote for asphalt on election day. If the toll revenue cannot cover the cost of the project, then people don’t really want the project.

    That said, 100% cost coverage is not reasonable because all transportation projects are subsidized and have large externalities. There is no free market that generates accurate prices. But as a rough measure, projects ought to be able to carry half their costs on tolls.

    The toll revenue imbroglio with the Deep Bore Tunnel says something pretty honest about it. No one really wants it. Yes they will take it if someone just gives it away for free, but when they actually have to pay for it, no thanks. Can you imagine a restaurant that had to price its $20 dinners at $2 to get people to eat there, and if they bumped the price to $3 demand would fall way off? It would be a clear indication that the vast majority of folks thought that the restaurant was crap.

    The Port Mann revenue totals for the first year of operation until March 31st 2013 are due out soon. The current projections are:

    2012-2013 50m (only tolled for part of year)
    2013-2014 185m
    2014-2015 209m

    These toll revenues are supposed to generate 100% cost recovery on the project (forgetting about externalities of course), but there has been a lot of skepticism about them. They have already been revised down, and the projections do not appear to fully incorporate declining traffic on the bridge. We shall shortly see where this is going.

    Report links for transportation nerds:




    • Clark Williams-Derry says:

      Great restaurant analogy. And thanks for the links!!

      This one in particular has a great chart of traffic on the Port Mann bridge, with a very familiar shape: traffic peaked in 2004-2006, and declined steadily thereafter.

  3. Elaine Livengood says:

    Out here in the predominately blue-collar and diverse south portion of King County, people find the extra toll is not worth it on the 167 HOV lanes. Maybe the more upscale 520 commuters will respond differently- maybe not.
    Opening up the 167 HOV to all traffic would get things moving better on 167 and I-405, where the daily commuter congestion backs up way too much at the I-405/167 interchange. Of course, we still would have the bottleneck where 167 becomes Rainier Ave. South. Millions are being spent to make this a grandiose boulevard. What a waste of money.

  4. Andres Salomon says:

    This brings up an interesting analogies to biking. Would I be willing to pay a toll (not a tax on bicycle registration, but an actual bike path toll) for safe, critical bicycle infrastructure?

    Let’s assume it was a reasonable price. Ie, not $2-$5 a trip, as the wear-and-tear from my bike and the space required do not compare to a car, and bicycle infrastructure simply isn’t as expensive as car infrastructure.

    So throw a $0.25 toll on the Dexter cycle track, or on an imaginary future downtown cycle track in a few select locations. Make it RFID/camera-powered, so you don’t need to actually stop to pay the toll. Make sure 100% of the revenue collected from the toll goes to support additional bicycle safety projects, or maintenance of existing bike transportation routes (and the cost of tolling overhead, of course).

    Would I be in favor of that, and be willing to pay? Absolutely, though the 100%-of-revenue-going-towards-bicycle-road-safety-improvements bit is essential towards my acceptance of such a thing. Of course, there would be obvious enforcement issues, but that’s already the case with the Link Light Rail right now. I can ride the light rail for free most of the time, as they only check that I’ve paid maybe 1 out of 20 trips that I take. I still continue to pay, though, because I appreciate the service and would prefer to see it funded.

    However, put the toll on a non-protected bike lane shared with fast moving cars, or on a mixed-use trail, and I’d be much less willing to pay it.

  5. seonaidh says:

    there are multiple stimuli. the incentive compared to the status quo isn’t that much of a gain. The fundamental fact is that the pain of NOT paying for the HOT lane must be greater–like parking in the Netherlands. it wasn’t just making cycling or busing easier, it really came down to the fact that the cost and “pain” of driving became so much greater: petrol costs, and the gradual elimination of a lot of parking spaces, the rising cost of using the remaining parking. The two have to work together. Until we really push the actual cost of driving onto drivers, the perceived benefit of alternative transportation is lower.

    • seonaidh says:

      To put it more bluntly: because the cost of gas is still relatively low, it is the cost of PARKING that matters and guess what? a lot of parking is still “free” at the end of the drive.

  6. John Niles (@JN_Seattle) says:

    Smart pricers in the private sector know you have to test, test, test prices to find the sweet spot. “Slide down pricing” is the main strategy. Start high to make sure no consumer revenue is left on the table, and then move the price down to see what happens to Volume multiplied by Unit Price which yields total product revenue.

    The “sale price” is a popular tactic as well, a temporary price reduction for a limited time to see who steps up to buy. (Macy’s is temporarily cutting prices on a number of items by 50% today and Saturday, in case you have’t heard.)

    The public sector is often unable by law and frequently unwilling by temperament to change prices to find the revenue sweet spot.

  7. Steve C. says:

    The problem is that the cost of using the HOT lane, while not expensive on a daily basis, becomes expensive on a monthly or yearly basis. Perhaps making monthly passes available for frequent drivers would be one way of boosting usage (if not necessarily revenues).

    On the other hand, it is still HOV usage of HOT lanes that matters most, when it comes to reducing SOV usage, which should be the real goal. Why that isn’t greater should be the real consideration.

Leave a Comment

Please keep it civil and constructive. Our editors reserve the right to monitor inappropriate comments and personal attacks.


You may add a link with HTML: <a href="URL">text to display</a>