When I published an argument for removing I-5 from central Seattle on Monday, I hoped it would start a conversation, and it does seem to have gotten some traction. Charles Mudede added a think piece for The Stranger. Capitol Hill Seattle blog summarized my case here. Shrill KIRO talk radio host Jason Rantz even got wind of it and, true to form, worked it into a rant that isn’t worth a hyperlink. And below, even Knute Berger added a cryptic tweet. I’ll take it!

I also got a lot of negative feedback. I will respond to some of the more constructive bits in the following

Downtown I-5 Traffic Counts Are Flat

This may come as a shock to some people but Washington State Department of Transportation (WSDOT) data shows that Downtown I-5 annual average daily traffic (AADT) is actually down from where it was a decade ago. In 2005, the AADT was 214,000, while in 2015 it was 206,000.

AADT on I-5 Downtown according to WSDOT records. (Graphic by author)
AADT on I-5 Downtown according to WSDOT records. (Graphic by author)

This is all while the region added hundreds of thousands of residents over the last decade. We’ve grown rapidly, but traffic counts on I-5 in Downtown have actually decreased over that timeframe. An armchair transportation planner (perhaps even WSDOT professionals and their models) might expect adding residents would mean adding traffic Downtown but that doesn’t seem to be how it’s necessarily worked.

The Seattle metropolitan area has grown steadily. (Puget Sound Regional Council)
The Seattle metro area has posted big population gains since 2000. (Puget Sound Regional Council)

Seattle proper had a population of 563,374 in the 2000 census, 608,660 in the 2010 census, and the Census Bureau estimates the population reached 684,451 in 2015. Despite adding more than 100,000 new residents, Seattle had greater average daily traffic counts in 2004 than it does now.

The Seattle Department of Transportation estimated Seattle’s average daily traffic is on a downward trend for the decade. (City of Seattle)

Understanding The Lived Experience Of Traffic

Now, the lived experience of traffic is still terrible. I never disputed that and was reminded many times it was. But this is why restoring the street grid Downtown might actually make things better. Freeways are brittle. Too much traffic equals gridlock and no means of escape until the next exit. When you are driving on a grid of city streets, you can’t achieve the same top speed but there are options; when one route becomes congested, you simply make a couple turns and select another. You can even say the heck with it and find a place to stop and say order a pizza. Unlike a certain TV commerical by a certain freeway-fetishizing gubernatorial candidate, you can actually obtain a pizza when traveling on a city street. This is why–for me at least–the lived experience of rush hour on a city street grid is less suffocating than rush hour on a freeway.

Planning With Transit In Mind

One ever so helpful tweeter also reminded me that I-5 carries more traffic than Link does. After the U-Link extension, light rail carries 68,000 riders per weekday on average while as the graph shows above I-5 carries just over 200,000 vehicles, most of them single occupant vehicles, downtown. The added context though is that Sound Transit projects its system will carry 350,000 riders–the majority of them on the Central Link shadowing I-5–once ST2 is built out. That would put light rail on par with I-5 as a mover of commuters downtown. And if voters approve ST3 this November, Sound Transit projects 600,000 riders on light rail alone (likely a conservative estimate itself).

Changing The Social Contract

But I-5 also carries buses, some argue. Why disrupt those patterns? Here’s where reimagining I-5 gets really exciting. The land created would be a sort of blank slate. We can use some of the land to create a slimmer street with transit prioritized so that buses like the ones coming across SR-520 can still continue downtown efficiently. And in the process it would grow the usable boundaries of Eastlake, South Lake Union, and First Hill. Transit would be put next to destinations rather than on a closed freeway, and we could add protected bike lanes to boot to make a true multimodal “Compete Street,” a concept nearly every city adopts as a best practice but rarely actually uses. The possibilities are endless once we free ourselves from the hold of the idea of having freeways through urban cores.

Gaining More Than We Lose

Many people agreed with me that plowing I-5 through Downtown Seattle was a mistake (and that list includes President Dwight Eisenhower who approved the Highway Trust Fund in the first place). However, when it comes to correcting the mistake they stop short. People are used to the Downtown freeway connection now, and to some, the feeling of loss is stronger than the feeling of gain, even if a logical analysis suggests the gain is greater than the loss.

Other Cities Have Successfully Removed Freeways

Seoul has a beautiful linear park and river in place of an elevated expressway in its urban core. #winning (Visit Seoul)
Seoul has a beautiful linear park and river in place of an elevated expressway in its urban core. #winning (Visit Seoul)

Supporters were helpful with support of places that have taken urban freeways out of their cores. Rochester, New York is removing the freeway loop that hems in its downtown. San Francisco tore down the Embarcadero freeing its downtown waterfront. And Seoul famously removed the Cheonggyecheon Expressway and turned it back into a river (the Cheonggyecheon) surrounded by a linear park. Land values jumped, the park turned it a major destination, average temperatures near the park dropped 6.5 degrees in the sweltering summers, and city-wide traffic didn’t get noticeably worse. Removing the freeway also removed a source of induced demand.

Induced Demand

Speaking of induced demand, the evidence is strong to support this theory. Here’s what I wrote in a post from last year:

In 1968, mathematician Dietrich Braess formulated what became known as Braess’ paradox: in a congested road network, the addition of a new route will increase overall travel times. Seattle Urban Mobility Plan said the paradox can also be expressed as “the theory that direct routes often function as bottlenecks, and so reductions in total capacity can reduce congestion.” Cities have seen results that support the conclusion. Seoul saw traffic volumes decrease and property values go way up after it demolished Cheonggye Expressway, its downtown highway viaduct, and replaced it with a linear park. Stuttgart, San Francisco, Portland, New York, and Milwaukee have seen similar results when tearing down urban freeways.

It’s overarching reason is why we can’t use highways to build our way out of our traffic woes. People often cite geography as why Seattle can’t add more freeways, but actually, even if we could, new freeways would likely do more harm than good according to induced demand and the Braess paradox. Conversely, removing our central freeway (I-5 in Downtown) might actually improve traffic flows.

Taking Climate Change Seriously

Finally, I’ll conclude with perhaps humankind’s greatest challenge going forward: climate change. Transportation is the number one source of carbon emissions in Washington state. Plan Washington described the situation: “The majority (54.5%) of carbon emissions in Washington come from transportation via the consumption of gasoline and diesel in on road vehicles. Compare this to the national landscape, where a mere 33 percent of emissions come from transportation, and the electric sector takes up more then twice as large a portion.” Continuously prioritizing automobiles only furthers our dependence and fails to address our number one carbon pollution source. Electric cars would lower their carbon footprint, but even electric cars running on renewable energy still face the problem of embodied carbon, the carbon required to build the vehicle, and the carbon to transport it to the consumer. The Guardian suggested the embodied carbon of car may even be greater than its total tailpipe emissions.

The upshot is that – despite common claims to contrary – the embodied emissions of a car typically rival the exhaust pipe emissions over its entire lifetime. Indeed, for each mile driven, the emissions from the manufacture of a top-of-the-range Land Rover Discovery that ends up being scrapped after 100,000 miles may be as much as four times higher than the tailpipe emissions of a Citroen C1.

Now this study may be extreme, but even conservative estimates of embodied carbon suggest it’s equivalent to years worth of tailpipe emissions. And alas the embodied carbon of hybrids and electric cars is generally higher than conventional cars because building of the carbon intensive batteries. As Forbes reported, “The Union of Concerned Scientists did the best and most rigorous assessment of the carbon footprint of Tesla’s and other electric vehicles vs internal combustion vehicles including hybrids. They found that the manufacturing of a full-sized Tesla Model S rear-wheel drive car with an 85 KWH battery was equivalent to a full-sized internal combustion car except for the battery, which added 15% or one metric ton of CO2 emissions to the total manufacturing.”

We don’t have to outlaw single occupant vehicles, but we do have to reassess their priority in our transportation system if we want to grapple with climate change. Freeways through urban cores are counterproductive to any reasonable long-term carbon reduction strategy.

What’s Better Than A Lid? Remove I-5 Entirely From Central Seattle

We hope you loved this article. If so, please consider subscribing or donating. The Urbanist is a non-profit that depends on donations from readers like you.

Doug Trumm is The Urbanist's Executive Director. An Urbanist writer since 2015, he dreams of pedestrianizing streets, blanketing the city in bus lanes, and unleashing a mass timber building spree to end the affordable housing shortage and avert our coming climate catastrophe. He graduated from the Evans School of Public Policy and Governance at the University of Washington. He lives in East Fremont and loves to explore the city on his bike.


  1. It would be great to see some more detailed proposals for this with maps- maybe two or three options.

    Would there be a surface road following the present course? That would mean awful grid breaking diagonal streets and 6-way intersections between Spring and Denny, and also between Yesler and Dearborn, and reduced development area as a result of having more surface area devoted to street.

    Or have a diagonal tunnel (of four lanes or so) in the Denny to Spring section?

    Maybe it has been pointed out already, but 6th and 7th Ave ought to become full streets and between them full city blocks can be developed (though maybe one or the other gets the surface highway treatment, which isn’t desirable but is maybe unavoidable and it is worth thinking up what that would actually be like).

    Also higher level summary comparison between ‘Lid I-5’ and ‘Rid I-5’ and rebuild as-is would be nice.

  2. P.S. I realize that removing Mercer to I-90 is a smaller removal than Doug’s originally proposed 225 to I-90 removal. But I think it’s easier to defend and would be highly profitable.

    P.P.S. There will be this sunken underground space, really wide, through most of the Mercer to I-90 area. Build a new Link line in it. Put in stations. Cap THAT. (It’s narrower than the freeway. A lot narrower. Even with station platforms.) Connect it to the existing system through the East Link tracks at I-90. Where I-90 was elevated, you have a choice whether to put the Link line elevated, on the surface, or underground.

  3. Do it! Do it!

    You can re-sign I-405 as I5 for through traffic (and you should). For local traffic, from the south, terminate I-5 at I-90 (this means it continues to handle most of the freight traffic from SODO and the ports to the south). For local truck traffic, from the north, terminate I-5 at the Mercer exit and steer the trucks through that new super-expensive tunnel or along the overbuilt waterfront highway to the port.

    Removing I-5 from Mercer to I-90 would free up a huge amount of very valuable real estate. Restore the street grid. I count at least 26 developable blocks.

  4. Wow, this great! Even better than the first article! We need to keep making these rational arguments and someday they will break through.

    • I don’t believe the I-5 traffic counts cited in the article.
      I have been driving on I-5 for a number of years and traffic seems much worse to me in terms of # hours where it’s congested.

      • I’ve heard that 2000 cars per hour is the per lane max. capacity. I think that assumes the vehicles in the lane are traveling at 40mph, so if you slow that to 20 for more hours of the day, less cars get through.

        • “Not believing” the #s is partly a manner of speaking.
          But they do seem off.
          Maybe I should have said “Based on my lived experience…”
          Can’t argue with “lived experience”, can you?

          • I think I see what you’re getting at.

            Erik offers one possible explanation that certainly makes sense. Vehicles moving slower reduces capacity overall. The longer peak congestion lasts, the less throughput you get. Other impacts along the corridor such as overtaxed exits and added weaving contribute. Time of day travel changes or using alternative modes offer other possible reasons. I’d also point out that the SR-520 toll started at the end of 2011, which definitely pushed many to start using buses so that could partially be at play.

          • Which, if I’m tracking correctly, suggest that I-5 is really at capacity & it is crowding itself which is limiting usage.
            It’s not that all of a sudden people don’t want to drive their cars, and they’re rushing to the bus, but simply that there’s only some number of cars can fit on the road.

            So wishfully thinking about voluntary car use is not so.

  5. Here’s a proposal that will probably be unpopular with this crowd:
    We could do the same thing to I-5 we are doing to the viaduct.
    We could have buried 2 lane in each direction highway for through traffic with no exits downtown. It would literally just be to carry I-5 traffic that passes through Seattle.

    Then we could have a surface state route with stoplights similar to aurora where I-5 currently is. This would have the advantage of completing the street grid, and solve the problem where offramps lack the capacity to handle traffic. There would just be a stop light at the intersection of the highway and mercer and other major roads.

    Traffic would be slow (25mph speed limit), but there would be more places to get off the road, so it would lead to fewer bottlenecks.

    We could even repair and reuse Bertha.

    • If I understand the situation correctly, if we did something like this we really wouldn’t need Bertha at all–a deep tunnel really isn’t needed when you can do a cut and cover.

      Either way, I’m all for getting rid of the I-5 scar in one way or another. It is always frustrating to hear people complain about lack of freeways or access when they don’t live anywhere near the city–believing that their mobility from the suburbs is more important than the lived experience of tens of thousands of people who live within a few thousand feet of a cancer-causing freeway. WSDOT should not put up a single freeway noise barrier anywhere in the state until they start providing some relief for residents in the most dense part of the region. Stop catering to the suburbs and exurbs.

    • It’s not a bad idea, but what’s the point? Why prioritize through traffic? What’s the impact on quality of life? I-5 might need to be fixed, but Aurora and “stroads” like it have way more deaths/mile traveled than I-5 if I’m not mistaken.

    • This is a much more realistic solution than the one proposed by the author. This article has so much in it that could be picked apart, but I’ll have to ignore most of it for now. One major issue that the author doesn’t seem to realize is that a lot of traffic on I-5 is through traffic from, for example, Tacoma or Portland to places like Bellingham, Everett, or Vancouver BC. What would he do with all those people? A through-traffic tunnel is a good solution. It takes care of those folks as well as almost all the issues with a downtown freeway (space, noise, some of the pollution, traffic, etc.). Pretty much all the traffic TO downtown could be handled well with expanded transit, biking, and walking (including using some of the newly freed-up space for buses and rail). Users coming from the suburbs could drive to a park-and-ride (as many of them now do) and continue downtown from there. I would prefer if we could avoid turning all that new space we gain from tunneling the through traffic into a surface highway, but a normal street (plus transit facilities) could work. I think we could pull that off.

  6. Doug – great 2nd article. I’ve loved the debate this has started.

    The one key issue that I don’t think you have addressed is the importance of I5 for the regional economy. I5 is an important state & national freight corridor – simply diverting the through traffic to 405 is not only a non-solution for local industries, it’s unfair to the East King cities that are built around 405.

    Induced demand is a valid argument for local & commuter traffic, but I don’t think it holds up for freight.

    Capping I5 and removing most or all of the downtown exits achieves many of the benefits you seek. But before removing I5 completely – unlocking even more benefits – the region needs to re-imagine how commercial traffic will move through King County.

    • Shouldn’t be a problem. Through freight traffic should take I-405 (should already be doing so), and through freight traffic should be a small minority of traffic (since by definition it’s going to British Columbia or northern Washington).

      For local traffic, retain I-5 south of I-90 — this accounts for the majority of freight traffic — and retain I-5 north of Mercer. The route Big Dig Tunnel – Mercer – I-5 will handle the residual freight traffic. Yeah, Mercer will have a lot of trucks for a few blocks; that’s the only downside.

Comments are closed.