Graham Street Station and the Failure of the Enterprise Initiative

This struck me as strange from the get-go. For one, a center platform on an existing rail alignment like this is highly unusual, particularly for a system with overhead catenary electrification like Link. It costs a lot of money to realign tracks and to move the catenary to accommodate a center platform, and while there are safety and comfort benefits for center platforms, this choice is a significant factor in the current price tag being almost unfathomably high for an infill light rail station. In the 2010s, it was common for infill stations such as the MAX (Portland) station at Civic Drive to cost between $1 million and $3 million ($1.5 to $4.5 million in 2026 dollars). That’s 0-2% of Graham Street station’s cost, as we’ll see next.

There are other mitigating factors at Graham Street, including the fact that the alignment is in the middle of a busy arterial with significant sewer impacts, but as advocates we shouldn’t accept a $200 million price tag as some sort of normal thing. Indeed, when you dig into The full ST Alternatives Analysis 2025 (as I did in the more detailed article), you won’t really find in-depth discussions about costs and trade-offs with different options.

Here’s what I mean by this. Only one of the six studied alternatives even considers reducing any space for cars on MLK; the “Split Near-Side option”. In that one the platforms would be where the current left turns are, so the southbound platform would be on the north side of Graham Street in the center of the road, and the northbound platform would be on the south side of Graham Street. (See ST’s alternatives analysis for a diagram and its other alternatives.) Even before we consider the fact that a new station at Graham Street would definitionally improve access for a whole host of transit riders in Rainier Valley, there are operational reasons for Sound Transit to prefer this approach. Given that the 1 Line corridor along MLK is unsafe by any metric (there’s a crash involving Link on MLK every 40 days) removing a left turn or shrinking space for cars would actively improving the safety of the least-safe part of the 1 Line corridor. That seems like a win-win.

These options that remove car lanes are less expensive, improve safety, and improve operations. But Sound Transit dismisses them essentially out of hand, either by not considering it at all, or putting their thumb on the scale of a deeply flawed alternatives analysis to remove the only alternative that had it. The stated justification for this is that it would shift left turns to already-busy alternative intersections (Othello and Orcas Streets), but it’s not clear if any in-depth analysis was actually done on this front, or if that is a base assumption made by the technical analysts responsible for the report.

In either case, this kind of reasoning is deeply problematic for a transit agency to engage with. While there are common-sense considerations for making sure traffic isn’t unduly impacted, if we take for granted that Graham Street Station deserves to be built for ridership and access reasons, then we should frame any additional spending above the minimum viable station that removes car access as spending done to preserve car access. There is no specific estimate I can provide for this, but considering that the two properties most likely to be impacted by the selected alternative will cost north of $5 million (parcels #8113100322 and #3333001680), that’s a good first-order guess. And recall that is more than a typical light rail infill station cost in Portland adjusted for inflation in the 2010s.

Realigning the tracks and catenary also comes with significant operational risk, which drives up costs. Sound Transit does not adequately cover in the alternatives analysis. While building a side-platform infill station (such as Pinehurst) also comes with construction challenges, they pale in comparison to rebuilding active tracks and moving the overhead catenary wires. This is likely why a survey of all light rail infill stations in US history reveals just two center platforms – both in Salt Lake City, and both planned in advance to preclude the need for alignment changes – out of about 30. It may be true that there is a community preference for a center platform, but given that is a significant cost driver, and part of the community preference is a desire for a less expensive option, we need to have a critical eye towards this.

The community is right to prefer a center platform – the concerns voiced are valid – but Sound Transit has some sort of responsibility to clearly articulate to the public the cost drivers associated with both options. To me, it is very clear they did not do this as the only closely studied side platform option was the one which maximally impacted the nearby sewer infrastructure. You can find more of the details in this companion post, but in essence the selected option should have initially rated low on constructability (a criteria related to construction timelines and impacts) based on their own analysis, but was not actually rated that way, and this lead to Sound Transit advancing incongruous options.

And we should return to look at the safety impacts as well. The only mention being that center platforms reduce risk that comes from crossing two tracks at once. But any survey of crashes and injuries on the 1 Line corridor along MLK will reveal that the primary safety risk is being hit by a car, and the safety improvements that accompany reducing car throughput are not considered, while the safety benefits from reducing turn conflicts are considered equal to crossing two tracks at once. This all leaves a very narrow set of safety improvements which are not properly weighted to consider within the context of station alternatives. The net result is that the suite of safety improvements we are likely to get will do little or nothing to improve the real safety issues that exist for transit riders, pedestrians, and cyclists on MLK stemming from too many cars driving too fast.

Parting Thoughts

To some extent, this is a bit of an exercise in futility, as there are few alternatives analyses out there that provide enough depth and context into the reasonable alternatives. And once an agency has made its choice, the purpose of the document really becomes closer to a post-hoc justification for the decision made than an unbiased look at the universe of options (this dynamic is not isolated to transit projects as anyone who has followed the Interstate Bridge Replacement project is surely aware). For an agency like Sound Transit that makes decisions by exhaustion that attempt to please as many people as possible while heading off any potential controversy first and reasonable choices that may save money second, it’s not exactly shocking to see this dynamic play out at Graham Street.

But once upon a time, Sound Transit promised that the Enterprise Initiative would do exactly this. It would reexamine the underlying assumptions, find creative cost-cutting solutions, and do everything to avoid the typical Sound Transit option of “deferring whatever we can’t build to the ST(N+1) ballot measure”. Instead what we’ve gotten is a bunch of options relating to the least-bad way to defer projects. This shouldn’t be acceptable to anyone, but it’s particularly galling for Seattle riders who are fairly obviously getting the short end of the stick. Even with its ridiculously bloated cost, Graham Street is both the cheapest project in absolute terms and among the best by rider-per-dollar measures. But most riders for least money isn’t something anyone on the Sound Transit board other than Dan Strauss has publicly considered.

This isn’t to say that we won’t have to defer something, and it’s not to say we should defer a specific suburban project in favor of a specific Seattle project. It’s to say that Sound Transit has not made an honest effort in attempting to reduce costs, despite a public initiative claiming to. Instead of re-evaluating if spending $100 million to preserve left turns from MLK to Graham is a worthy investment for an at-grade train that constantly gets hit by cars turning left, the answer is that Sound Transit was right all along and we don’t have to reconsider any fundamental decision already made. This attitude is why West Seattle will still end up with an expensive tunnel, even after the elevated alternative has reduced its biggest impacts (elevated tail tracks extending to the heart of the Junction). It’s why Ballard still costs three arms and four legs.

If I’m honest, I find it difficult to be in this position. I care deeply about transit expansion working in my adopted home, and it feels like critiquing transit agencies can lend itself too much to the reactionary anti-transit voices still present in our community. But nothing I’ve seen from the planning of ST3 projects has really impressed me. It all feels like capitulating before anyone will even consider asking a difficult question, and relying on future voters to bail them out. I don’t think this is an acceptable line to take, and transit riders in the Seattle area deserve better. It’s a shame that Sound Transit seem unwilling or unable to make that better future we were promised happen.