Revising Sound Transit 3

The following paragraphs are the comment I will be sending on the proposed Sound Transit 3 package. The comment period closes Monday May 2nd, 2016 at 5:00pm PST

Dear Sound Transit:

At a high level, I don’t really see the value (compared to the cost) of much of the investments being made in ST3. The vast majority of rail proposed is either mostly redundant with existing Link (the second downtown tunnel, the second SODO segment, and the Factoria to Wilburton segment), or follows freeways (West Seattle, the spine segments, Issaquah and Redmond light rail) out to far-flung areas distant from much actual or potential dense, walkable, ridership generating areas.

With regards to the freeway rail, these are corridors where simply increasing restrictions on HOV lanes to ensure reliable fast grade separated free flow for buses would have huge impacts at little to trivial cost, especially given that Sound Transit has already built (or will soon be building) much of the capital needed to make that system work very effectively. When WSDOT reports that 48% of the people traveling on I-5 southbound during peak period at Northgate travel in the HOV lanes (and 35% on transit), it is simply nonsensical that we are planning to spend billions of dollars of capital over two-plus decades to get the speed and reliability benefits that transit riders deserve on I-5 today given their relative usage of the roadway. I understand that it’s not in Sound Transit’s scope to reallocate road resources, but this is the context in which we can evaluate the benefits and costs of these proposals.

Simply put, extending the spine further south, north and east from its currently funded termini is not worth the cost. The subsidy per rider (after operations and fares) is roughly on the order of $10 per rider after fares ($6 to Federal Way, $8 dollars to the Tacoma Dome, $13 dollars to Everett and $16 for Redmond), which is hard to justify in any context, let alone one in clearly superior non-complex counterfactuals exist and one in which opportunities for effective transit orientated development appears minimal. A similar claim can be made about Issaquah rail and arguably West Seattle rail as well.

But it’s not reasonable to let a perfect plan be the enemy of a good one so the onus is on Sound Transit to make the suburban lines and urban lines as strong as possible to negate the inherent wastefulness of freeway-oriented rail spending in the deep suburbs.* As the plan stands, I won’t vote for this. However, I think the plan is salvageable with some plausible changes.


The middle-level issue with what is proposed is that the insistence on building rail to particular places and completing the spine forces other alignments to be suboptimal. The negative externalities of alignment decisions ought to more carefully considered and if Sound Transit cleaned up one of the following two glaring issues in their final proposal (along with insisting on an I-5 alignment with a Paine Field spur, stub or BRT in Snohomish County) I would probably get on board with the proposal.

The first alignment with huge negative externalities is the completion of the spine south of the funded Highline CC station. Currently, the East King rail projects grade as generally awful with subsidies per rider between $15 to $20 per rider after fares, a preposterously high number. A seemingly more efficient use of East King resources would be rail in Renton as part of an east-west line between Burien, TIBS, Southcenter and Renton. Preliminary studies of that alignment had an order of magnitude lower cost per rider projections (something closer to $7 to $10) and it would also serve some of the region’s poorer residents as well as areas seemingly more ripe for large-scale development and growth. Renton, in particular, is fairly centrally located, mostly has an existing street grid, and room to grow and working to spur TOD in that area would go a long way towards improving the economic fortunes of the more southern, poorer parts of King County.  Simply put, the best place to invest East King County resources is in Renton.

But of course, South King County lacks the resources to both build the spine towards Federal Way and build their half of a Burien/Renton line. So the insistence on building the spine in South King County comes at the expense of a far more efficient use of resources in East King County and thus a better package overall.

It’s worth adding here that Pierce County would also be better off with additional urban service over completing the spine. The Tacoma link extension to Tacoma CC grades out as the best investment outside of Seattle in terms of cost effectiveness** and Tacoma is a relatively dense urban area that would attract a lot more local jobs and resources by enhancing connections within the city instead of attempting to be a bedroom community for Seattle. Moreover, it is unambiguous that urban service generally helps poorer people relative to commuter service. Urban service is also more conducive to TOD, the promotion of walkability and ultimately the attraction of the types of young talented individuals and companies that make Seattle and certain other American cities so robust economically.

Overall the insistence on completing the spine probably reduces the package’s utility for the South King, East King and Pierce County subareas combined by 30% to 50% verse the counter-factual of the rail options discussed above plus spending on whatever existing projects had enough funding. For example, if the Renton/Burien line had a $9 dollar per rider subsidy and capital costs similar to the Federal Way and Issaquah rail extensions combined, the Renton/Burien line would be about 45% better than the Federal Way and Issaquah lines combined. And the benefit for Pierce County would be the relative value of an additional urban line over the completion of the spine. Even if that latter figure is negligible, the 45% improvement is huge and represents a billion dollars of value or more.


The second externality issue comes with the West Seattle rail alignment in Seattle. Beyond redundant rail in SODO, building a West Seattle line requires building an expensive second downtown tunnel because only two of the three lines going south from the International district can use the existing tunnel. But in terms of inherency, the Ballard line doesn’t have to go all the way to the ID and could instead continue east from Madison (or Westlake) towards the dense job center of First Hill, the densifying neighborhood of the CD (in particular around Jackson and 23rd) and Judkins Park station for transfers to East Link.

Consider the counterfactual of building that extension of the Ballard line instead of the West Seattle line. If you assume net costs of $900 million (1 net mile tunnel plus 0.6 net miles elevated) over the proposed Ballard to Downtown line and an additional 24,000 riders for that additional segment, the cost effectiveness of the Seattle investments would increase by about 11%. Moreover, such a change would free up funding for bus improvements for West Seattle (including direct off-ramps to the SODO busway, which are desperately needed yesterday), for building the Interbay segment elevated, and for automating the line because it is not interlined with Link, with the resulting frequency and operational improvements. These additional benefits could easily improve the utility of the Seattle investments by 20% or more.

As with the Burien/Renton counterfactual, this is an improvement worth potentially billions of dollars. Indeed, given that First Hill is one of the biggest and densest employment centers in the state and contains a key regional asset in Harborview Medical Center, it is somewhat perplexing that HCT is not even being considered there. Regardless, at minimum, Sound Transit should design Madison station so an extension east to First Hill and the CD can be easily built in the future.***


As it stands the costs of these projects are in many cases not worth the benefits. Sound Transit should push to get its subsidy per rider values closer towards at maximum $7 dollars per rider, and ideally better. Roughly speaking, that cost/benefit is in the ballpark of being worthwhile. In a similar vein, Sound Transit should stop focusing on describing the reality of congestion in the region  and instead focus on how Sound Transit’s specific ideas will actually improve the mobility and well-being of people in this region. Because while we all know that the Puget Sound region has substantial mobility problems, this particular package doesn’t constitute a solution to this problem and instead represents a waste of tax payers dollars.


Alex Bailey


*Beyond the counterfactual of free-flowing transit on freeways, rail to the deep suburbs generally does not support high ridership. While transit is great for urban to urban trips and good for suburban to urban trips, it’s not very helpful for suburban to suburban trips because cars are almost always superior for such trips and suburban car ownership rates are very high. This means that all the potential trips from stations in the deep suburbs are long trips of at least 30 minutes because that’s how long it will take to reach the urban areas from the deep suburbs riding Link. Outside of commuting, people rarely make trips this long and so it’s hard to generate all that much ridership (and in turn value) from HCT in the deep suburbs.

**To be clear, cost per rider is not the only metric on which projects should be evaluated. Hours saved, development opportunities, long-term ridership growth, and social equity also matter. But there is little evidence that most of the freeway segments will spur particularly large amounts of development or score particularly highly in these other areas compared to potential counterfactuals. Overall when one proposal has substantially better cost effectiveness than another proposal then we should consider it substantially better unless there are extremely compelling reasons not to.

***On the topic of the Ballard line, I also support the west is best alignment in Interbay, as rail on 15th will have serious construction disruptions and reliability issues.