Action Letter: Sound Transit Board Can Save Billions and Years

In a somewhat technical writeup in the Urbanist, a public sector transportation consultant from Accenture weighed in on how Sound Transit could save hundreds of millions of dollars by bringing its practices more closely in line with the European approach. (Given time and inflation delays, I’d argue that it is billions.) He also offers a petition to Sound Transit’s board, asking them to look at the findings of experts that show why US projects are so expensive, and get rid of the unnecessary, expensive, and unhelpful construction choices that so bedevil Sound Transit. 

I signed it, and recommend that you do so as well. 

Reed’s argument is about more than saving big bucks - although he does show how to do just that. It is about ensuring we get our promised system at all, and on a timeline where any of us will be alive to see it. His work is based on the renowned “Transit Costs Project,” an effort out of NYU to understand why North American fixed rail is so much more expensive than other high-wage countries like France, Spain, South Korea, Switzerland, Italy, Sweden, Austria, Japan, Germany and Norway. 

Norway is an extremely high wage country, more so than the US, and has above-average transit construction costs. Not only this, but their projects often blast through mountainsides, or involve water crossings. Still, they build much more effectively than Sound Transit.  

Norway’s cost of $531 million per mile (in 2020 dollars) is dwarfed by US average costs, at $1.007 billion per mile. In fact, Sound Transit’s West Seattle extension is expected to come in at $1.39 billion per mile (assuming Sound Transit’s current estimate of $1.68 billion per mile is in 2024 dollars). Cost per mile in Norway is 62% lower than in West Seattle. Imagine getting 2.5 times as much transit for our money! Or imagine building the same amount of transit and saving billions for green infrastructure projects, highway lids, and active transportation. 

There are many reasons why US projects are expensive, such as an excess of white collar workers and design costs, or permitting problems (something Maritza Rivera recently tried to make worse; thankfully her colleagues rejected her attempt to add red tape). All of this is made worse by meddling from lobbies and their political pawns like when the Chamber of Commerce got Bruce Harrell to try to move the South Lake Union station to the middle of nowhere. Thankfully that one failed too. 

But, according to Reed (and the NYU Transit Cost Project Folks), one of the biggest and most fixable problems is the stations themselves. Building them is a huge slice of the price of construction. Since every cubic foot is so damn expensive, affordable countries keep them close to the surface, don’t branch out to the side when building access to the stations, and keep the stations narrow and close to the length of their train platform. 

But Sound Transit currently plans to splurge like a cold-war defense department on all these items. 

ST3 excavation areas used to access the stations branch out away from the stations in a way that is wildly worse than ST2 or prior, and obscene compared to international standards. High performers use their vertical space efficiently, rather than branching out to the side on the way up to the surface like Sound Transit does. 

Sound Transit also plans wide stations, on average 34 feet. ST2 stations were 24 feet. No need for such bloat–keep the stations narrow.

High performance countries keep their station length similar to their platform length –just slightly longer. An excess of 5% is typical for the best countries; 20% is high. Keep in mind these stations are often nicer than ours - station length does nothing notable for the rider experience. But the stations for ST3 are on average 79% longer than their platform! That’s an obscene waste of money. Mined construction is especially expensive, and a couple of our stations have to be mined, so long stations in midtown and Westlake will be woefully wasteful.

Station depth matters much, as costs increase exponentially as you go deeper, because the reinforcement methods required become much more onerous. Sound Transit’s first stations were 60 feet deep. The average for ST3 (remember, most of these are new stops and won’t be built underneath existing stations) is 108 feet! This is a mistake.

Reed’s article highlights several other areas –the use of mezzanines, overinterpretation of codes, excessive deference to stakeholder input, and lack of standardization. He recommends adopting the “reference class forecasting” project management framework, which is considered best practice and is good at holding planners’ feet to the fire of reality. I’d like to rename this method to “appropriate benchmarking without assuming you can do magic.”

Among other things, Reed recommends that the board “systematically work to reduce station size, depth, and complexity based on best practice applications of codes and design standards in the creation of standard station designs,” “establish clear, objective criteria to decide between elevated, at-grade, trenched, or tunneled alignments” and for “construction methods employed, prioritiz[e the] quickest and least expensive/risky methods.” 

The letter advocating for reduced station depth, width and length and appropriate benchmarking is here. Please sign it and share it with your friends. Billions of dollars and years of crappy transit access are on the line.