I wish I were making this up. The Washington State Department of Transportation continues to insist that traffic volumes on the SR-520 bridge across Lake Washington are going up up up—even though actual traffic volumes have been flat or declining for more than a decade!  Here’s a chart that makes the point.


In a charitable mood, you could forgive the 1996 projections.  Back then, rapid traffic growth on SR-520 was a recent memory: up through about 1988, traffic growth was both steady and rapid.

By 2011, however, it should have been perfectly obvious that the old predictions were proving inaccurate. Yet WSDOT just kept doubling down on their mistakes—insisting that their vision of the future remained clear, even as their track record was looking worse and worse.  So now they’ve wound up with an official traffic forecast, in the final Environmental Impact Statement no less, that doesn’t even pass the laugh test.

It would be funny—if the state weren’t planning billions in new highway investments in greater Seattle, based largely on the perceived “need” to accommodate all the new traffic that the models are predicting will show up, any day now.

  • In case you don’t believe me about the numbers, feel free to check out the sources directly.  I’d be happy to be corrected.

    The data on recent traffic volumes—the dark green dots—come from three sources.  I start with WSDOT’s biennial Ramp and Roadway Report.  Then, to add in the missing years I factor in data from the Annual Traffic Report series and Seattle’s Traffic Flow Data. The blue trend line is just the basic linear regression of the blue dots, as calculated by Excel.

    The light orange line is based on a projection that dates back to 1996, which was mentioned in WSDOT’s 1999 Trans-Lake Washington Study.  Since that report is only available on CD-ROM, but not online, I’ll quote the report directly:

    Under the No Action solution set, Trans-Lake travel [on SR-522, SR-520, and I-90] is expected to increase by about 168,000 daily person-trips…in the year 2020…Because capacity is limited on SR 520, only about 20,000 additional vehicle…trips are expected there.

    (If current trends hold, the projections of 20,000 additional vehicle trips will be off by about 20,000 additional vehicle trips.)

    The green line is from the 2002 Trans-Lake Washington Project report, also available on CD-ROM, which projected that traffic under the “No-Action Alternative” would grow by 20 percent through 2030.

    The dark orange line—127,400 cars by 2030—is from the recently released Final Environmental Impact Statement for the SR-520 bridge replacement project.  It’s based on the projections described in part 1 of the “transportation discipline report” (see Exhibit 5-3 on p. 92 of the pdf, and the projections for SR-520 at Midspan under the 2030 No Build Alternative: 127,400 vehicles by 2030).

    I could have included another projection from the 2006 Draft Environmental Impact Statement—127,860 vehicles per day by 2030, as claimed in Exhibit 10-8 in Appendix R part 6. But it was getting hard to fit all the wrongness on a single chart.

    Now, I know that total traffic volumes aren’t the only traffic trends worth paying attention to. The traffic models make projections about peak-hour delays as well, which are probably what commuters care most about.  But given that the models have proven so stubbornly and preposterously wrong about traffic volume trends, it’s hard to believe that they have much of value to say about future traffic delays.

    [Thanks to Jake Kennon and Pam MacRae for help with the numbers!]