Need for Speed? Northwest High-Speed Rail Project is an Expensive Mistake


Cascadian high speed rail — 250 mph trains zipping between Portland, Seattle, and Vancouver — is one of those ideas that sits just far enough outside the streams of public discourse to avoid the scrutiny that it deserves. One the places that such scrutiny might have come from, the editorial page of the Seattle Times, has inexplicably joined the cheering section

A small backwater of the Washington State Department of Transportation, the Cascadia Ultra-High-Speed Ground Transportation Project, keeps churning out studies and reports, and the Legislature set aside $150 million as a potential match for federal funds to be used to keep the study machine humming. 

Two of these studies are quite informative. One is the 2018 “Ultra High‐Speed Ground Transportation Study” produced by the late great Northwest engineering firm CH2M-Hill. The other one is “Keeping it on the Tracks,” produced by University of Washington faculty and students. The analysis below is based on findings in the CH2M-Hill study.

But as substantive these studies may be, they never make a convincing case for an investment of this magnitude. The studies have lots of “how” but very little “why.” The dollar figures are enormous, and while there is discussion of recovering operating costs, little mention is made of the massive capital and environmental costs of these systems. The tone of the discussion seems to be that high speed rail (HSR) is self-evidently a Good Thing, and that not having such a system is a failure.

Where does this bias-to-build come from? 

High-Speed Rail Website page.

We are told we are behind the times compared to other countries, which is true. But the two pioneers in High-Speed Rail (HSR), France and Japan, are not exactly economic juggernauts, so it is hard to say that HSR provides some sort of economic juice. China has built a great deal of HSR, but infrastructure development in China does not involve any public evaluation of benefits and is often more of a jobs program than a spur to industrial growth.

The role of transportation infrastructure in spurring economic development is often exaggerated. Opening up brand new corridors can be totally transformative, as happened with the arrival of transcontinental railroads in Seattle beginning in 1893, or with the opening of a bridge across Lake Washington in 1940. But marginal improvements in the performance of an existing corridor have far fewer effects. Shaving some time off a journey that is already pretty straightforward might help existing businesses, but is unlikely to attract a lot of brand new business activity.

HSR ties back to the idea of “Cascadia.” Starting in the 1980s we saw a flurry of activity around various notions of Northwestern-ness, tying together Washington, Oregon, and British Columbia. Books like Joel Garreau’s Nine Nations of North America promoted the idea of commonality among states and provinces on the West Coast. The Discovery Institute’s Cascadia Center and the Pacific Northwest Economic Region sprung up around that time. Regional leaders – notably Seattle Mayor Paul Schell, Discovery Institute founder Bruce Chapman and leading venture capitalist Tom Alberg — embraced the idea.

The 2018 CH2M-Hill study evokes the idea of mega-regions and suggests that the Portland-Seattle-Vancouver corridor is a megaregion that needs to improve its connectivity. But the benefits of that improved connectivity are never spelled out. What economic activity will take place only if we add HSR to the Portland-Vancouver corridor? The study ducks this question, saying that any kind of economic impact analysis is premature.

The reality of economic geography is that metro areas like Seattle, Portland, and Vancouver don’t really connect at a high level. Every metro area has its own economic logic, tied to its location, its hinterlands, and its history. Each has an economic base that drives growth, and each city serves as the commercial center for its surrounding region, providing specialized business and personal products and services. There is, of course, business done between major cities, but not really at the scale that would change their fundamental economies.

Vancouver BC is the commercial center for Western Canada and an important node for business in Asia – and Seattle and Portland are just not going to be part of that. Microsoft has a couple of thousand employees in Vancouver BC, but that is about visa availability in Canada and not about some deliberate partnership. Underneath its hipster veneer, Portland is an industrial center serving national and international markets.

There may be a common ethos across the cities of Cascadia, but it’s hard to detect forces driving them closer economically. Businesses in these cities concentrate on serving their local area and the world, not neighboring provincial capitals.

Planning an HSR corridor runs immediately into a big problem: getting in and out of major cities. Right-of-way costs, community destruction, and safety considerations make it difficult to put new surface corridors into developed areas, and tunneling and elevated track are massively expensive. The 70-mile stretch from about Spanaway to Marysville is heavily developed, and getting more so each year.

The UW study emphasizes avoiding compromises on speed, so would seem to suggest building the entire new right of way from a Portland station to a Vancouver station, which would mean 316 miles of new track. A less aggressive and costly approach would be to use existing tracks in dense cities and build new tracks in rural areas. For the Seattle-Portland route that might mean about 125 miles of HSR and about 45 miles of conventional rail. For the Seattle-Vancouver route, that might mean about 110 miles of HSR and 40 miles of conventional rail. Under this assumption the HSR program would build about 235 miles of new HSR right of way.

The Ch2M-Hill study estimated, on the high end, that the Portland to Vancouver corridor could be built for $42 billion, or $133 million per mile for the full 316 mile corridor. But rail projects nearly always come in at least twice as high as early estimates. Indeed, the latest estimates for the Los Angeles to San Francisco HSR project are now at $128 billion or $256 million per mile. At that figure, the full 316-mile corridor would come in at $81 billion and a scaled-back program of 235 HSR miles would come in at $60 billion. (You should, of course, try this at home with your own assumptions!) The per-mile cost of the two approaches and two segments would be different, of course, but let’s keep it simple.

The capital side of these projects is hard to nail down, since funding comes from a variety of places. For our purposes, let’s imagine financing the whole capital cost with bonds at a 4 percent interest rate over 30 years. The annual debt service on the 235 mile program would be $3.5 billion and the longer program would be $4.7 billion. To put these figures into some context:

  • The 2023 operating budget for the City of Seattle, including all utilities, is $5.9 billion.
  • The latest estimate for the total cost of the new SR-520 corridor is $5.1 billion.
  • Public transit agencies in Washington spent $5 billion in 2022 to provide 158 million trips.
  • $4.7 billion could fund 12,000 affordable housing units.

The CH2M-Hill study estimates the annual ridership on the Portland-Vancouver corridor to be around 2 million trips initially, rising to 3 million in 20 years. That means that the annual debt service on the capital side of the project in year one would be $1,700 per passenger trip for the low end, and $2,300 per trip for the high end. 

The study shows that annual operating costs might be recoverable from passenger fares. But when the study factors in what appear to be very optimistic assumptions about financing annual capital costs, total cost recovery drops to 15 percent in the early years. Using more realistic assumptions about capital costs, total cost recovery would drop to under 5 percent.

What would this investment get us? Again, to keep it simple, just think of 100 miles of train travel across open land. The 250 mph HSR train could cover that ground in 24 minutes and a faster conventional train could cover it in an hour. The UW study presents a strong case for not having intermediate stops, which a conventional train would continue to offer. So in the end for the Seattle-Portland and Seattle-Vancouver trips, HSR might save between an hour and an hour and a half, depending on the length of the new track.

In the context of a 150 or 175 mile door-to-door intercity trip, this is not a lot of time. A rail passenger still needs to get to and from the station and needs to arrive suitably early for security and boarding. And, of course, that extra time spent on a conventional train can be put to productive use, since there is internet connectivity on these corridors. 

Yes, the HSR trip would offer a smooth riding experience. But at $1,700 per passenger in capital cost to save an hour, and $2,300 (or “only” $1,100 under the CH2M-Hill study’s very optimistic assumptions) to save 90 minutes? And whose time is really worth that much per hour? To put it bluntly, rich people.

The cost and benefit pattern of the whole package is highly regressive. State and provincial taxpayers will pick up the capital cost (the Feds will dangle a few dollars on the front end, but make no mistake, this system would be funded almost entirely in-state) and high-earners, whose time is extra valuable, will reap the benefits. (Distributing capital costs between Oregon, Washington, and BC will be a huge political challenge in itself.)

And we need to admit just how meager the projected ridership is. The 2 million annual riders projected for the early years comes out to 5,500 per day. The sweet spot of cost recovery comes at 12 round trips per day (with trains only one third full initially). The model used in the CH2M-Hill study projects that even with luxury HSR, rail will not get above a 17 percent mode share. In other words, after spending upwards of $80 billion, the great majority of people traveling on the corridor will still be in a car, bus, or airplane.

We also need to factor in the incredible environmental destruction that will take place in building HSR. Think of all the streams and wetlands to be crossed, the hills to be cut, the valleys to be filled, the tunnels to be drilled, the wildlife corridors to be cut off. All the time, massive amounts of carbon are being pumped into the atmosphere from construction and from manufacturing and transporting all that concrete, steel, and ballast rock. Anyone proposing this kind of environment footprint for cars would be pilloried. Even the Sierra Club thinks this is not a good idea.

Furthermore, trains are not all that energy efficient. Trains must be heavy to stick to the track, and even a fully loaded train has maybe a ton of steel rolling along for each passenger – about the same as two people in a car. And with HSR, wind drag becomes a big issue. Wind resistance is an exponential function of speed, as anyone learns who watches their fuel mileage deteriorate badly as they cruise along at 100 miles per hour. Trains experience five times the wind drag at 300 kph as they do at 100 kph—and it’s much worse in tunnels. 

So while HSR has the advantage of less friction on its wheels, it has the disadvantage of the energy requirements of pushing through the air at high speeds. Yes, the trains will be electric, but they will still be pulling quite a bit of power out of the grid during times of peak electricity demand.

Looking forward, we are contemplating spending tens of billions of dollars for a 19th century technology when 21st century technologies are right on the horizon. Before any new Cascadia HSR project could possibly break ground, vertical takeoff electric air taxis will be flying those rich people to Vancouver or Portland, practically door to door in an hour. Several firms have these aircraft in the air, and while projections of their final approval by the FAA are certainly overly optimistic, it would be reasonable to bet that they will be flying passengers before any HSR project even gets started. A Cascadia HSR line will be obsolete long before it ever gets finished.

For the rest of us who can spare that extra hour or two, how about just improving the existing conventional rail systems to make the trips smoother? Far easier to electrify and double-track existing lines than to build entirely new corridors. The current train trips in the Cascadia corridor are quite scenic, and with improved reliability and a little extra speed, they could be much more popular than they are today, at a far, far lower cost.

So why does HSR keep rolling along when the costs and benefits are off by orders of magnitude? The stickiness of HSR probably comes from a number of places. Envy of sleek systems in other countries? Fear of being left behind? Mistaken notions of economic and environmental benefits? The old romance of rail travel? Glorious visions of luxury rail travel in the Hunger Games movies?

Maybe we just follow the money? The state is spending millions of dollars on studies, with more in the pipeline, and the transportation planning and engineering industry is eager to get its hands on that money, potentially hundreds of millions of dollars. Studies always recommend more studies, and the firms that perform those studies get paid whether the conclusions make any sense at all.

The back-of-the-envelope calculations above are based on the numbers generated by CH2M-Hill, a very highly regarded firm, but no one has the nerve to get to the bottom line of the costs and benefits. If the estimated capital cost per boarding turned up in the executive summary of some report, that could blow up the whole program and cut off future funding. 

The studies purposely avoid any normative judgements as to whether HSR is a good idea. Because the promoters of HSR are siloed in their transportation world, they don’t contemplate the opportunity costs of tying up that much money in what is really a low priority concern. Laid out against other priorities in the state, HSR would fall to the bottom.

Finally, we cannot ignore the role of Microsoft in all of this. Brad Smith, Microsoft’s President, has taken a personal interest in high speed rail between Seattle and Vancouver, and, through vehicles like Challenge Seattle (a gathering of CEOs, led by former Washington governor Chris Gregoire), Smith keeps the study money flowing. Why the taxpayers of Washington should subsidize Microsoft’s personnel strategies is a question that has not been addressed.

Given the radical imbalance of costs and benefits, under even the most generous assumptions, it is extremely unlikely that any Cascadia HSR system will be built. So why do we keep wasting time, energy and money studying it?

Michael Luis
Michael Luis
Michael Luis is a public policy consultant who has been wrestling with housing, growth and economic development issues around Washington State for over 30 years. He is author of several books on local history and served as mayor of Medina.


  1. Well said. Big infrastructure projects frequently attract the support of potential beneficiaries, giving them a life of their own divorced (or at least separated) from considerations of public good.

  2. Finally — someone with the courage to call it out!

    “…The back-of-the-envelope calculations above are based on the numbers generated by CH2M-Hill, a very highly regarded firm, but no one has the nerve to get to the bottom line of the costs and benefits.”

    “We also need to factor in the incredible environmental destruction that will take place in building HSR. Think of all the streams and wetlands to be crossed, the hills to be cut, the valleys to be filled, the tunnels to be drilled, the wildlife corridors to be cut off. ”

    If, on the other hand, we put more buses into the mix, and made them free — we could see a real, tangible benefit with not nearly the environmental costs.

    High-Speed Rail is simply not working. (Not as large as a debacle of the Honolulu light rail project, but pretty bad.)

  3. Thanks, Michael. I’ve been intrigued by the prospect of high speed rail in the Cascadia corridor, but it’s never felt doable for the reasons you lay out so well… And another one: In Britain and, I believe, elsewhere on the continent, rail has become a public-private partnership. The government owns the rights of way (the tracks) and private companies run the railroads. That’s the logical split. We Americans keep trying to do this in reverse. The government owns and operates the trains and the private sector owns the tracks. It’s never worked. And it probably can’t work.

    • You are on the right track (sorry for that sappy allusion) Mr. A. Alas, the history of railroad route development in the western U.S. “Gilded Age” was big time gambling. Oh, the bankruptcies! Oh, the federal corruption! Stockholders lost their socks and underwear in those days – 150 yrs. ago. As for those “trans-continental” routes – it was necessary to get on and off in the mid-west to make it ‘way out here in the 1880s and 1890s from America’s centers of power, money and commerce: New York, Boston, Philadelphia and Baltimore. Until the BNSF (today that would be Warren Buffett) forks over to fix the various landside problems along the tracks they own, this is wishful thinking. This is not France. Or Germany. Or Switzerland. Or Britain. Or Japan.

  4. Thanks Michael for this great critique. Please also see the Joint Transportation Committee’s independent review of the UHSGT project, table 40 in particular has some good comparisons of rail options. It is good assessment of the UHSGT program that got buried as soon as it was published. We need better rail, but there are other choices. Also is the UW study “Keeping it on the Tracks…..” they have this gem of a quote about mega projects and warped policy decisions:

    2.2 Megaproject risks magnify financial impacts

    Similarly, there is a fundamental tension between the need to be politically united behind such a substantial public project and the motivation of public officials to use the typical tools of their office to modify the aim, focus, and physical outlay of the project in ways that reduce viability, delay the development, and unnecessarily increase costs for the taxpayer. Referred to as the technological sublime (Frick, 2016), this term describes how the desire to promote a megaproject can too easily become a bias in favor of a project irrespective of its consequences, followed by a bias in favor of changes that appear to elevate or convey grandeur, yet are inefficient for the scope, scale, and timing of the project. In this situation, political leadership is viewed as governing the scope of the project simply by virtue of their role in continuing the allocation of public funds, and persons willing to ‘speak truth to power’ with alternative points of view become difficult to find (see, generally, Ginsberg and Paschall, 2022; Wildavsky, 1979). This phenomenon also describes what occurs when leadership becomes vulnerable to moral hazards. This occurs when project leadership and consulting and/or engineering firms become locked into self-affirming narratives, to avoid bad news and the risk that funding will end.

    It turns out that there is an entire field of research on the perils of pursuing the “technological sublime”.

  5. A long read about the perils of Cascadia HSR and “earthquake” isn’t mentioned. Before I go rocketing across the Nisqually Valley at 300mph, I want to know how fast it can stop when the earth starts to move. Everything else comes second for me.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Comments Policy

Please be respectful. No personal attacks. Your comment should add something to the topic discussion or it will not be published. All comments are reviewed before being published. Comments are the opinions of their contributors and not those of Post alley or its editors.