The northern terminus of the planned Issaquah-South Kirkland line in 2041, and one candidate for a future extension in ST4 (Image by author)

Although we are early in the ST3 program, some observers are already looking forward to extending Link light rail lines into the suburbs and adding more lines in Seattle. The ST3 plan funds several studies of suburban extensions. Current taxes do not support further expansions at the pace of ST3, however. Unless Sound Transit secures another large tax increase, capital spending beyond ST3 will be mostly squeezed out by the costs of managing what has already been built and financing the bonds accumulated in ST3.

The budget for future projects is constrained by Sound Transit’s tax authority. Sound Transit levies nearly all the taxes currently permitted by the Legislature; the only unused authority is a small rental car tax. Any prospect of further authority is hard to forecast. Certainly, it is difficult to imagine today’s Legislature granting more tax authority. Many legislators were unhappy about how the ST3 program far outran the smaller 15-year program they anticipated in 2015, and high car tabs remain unpopular. On the other hand, fifteen years is a long time in politics, and a new generation of legislators in the 2030s may take a sunnier view.

But let’s suppose we are limited by current law, or equivalently that voters resist new taxes. In that scenario, Sound Transit might ask voters in the waning years of the ST3 program to authorize more projects with an extension of current taxes. How much could Sound Transit build with voter approval if they just roll the current law taxes forward indefinitely? Less than you might expect. It turns out that a capital program extended to 2060 would have a run rate perhaps only a third as large as the 2016-2041 program.

Why is this? ST4 will face several constraints that were not present in ST3.

  • Sound Transit will have to operate rail and BRT systems as they are completed. The rail network will grow from 22 miles today to 116 in 2041, and bus hours will also grow. Operations and maintenance (O&M) are a significant expense for the mature system.
  • Sound Transit prudently committed in ST3 to keep the system in a state of good repair (SOGR). SOGR are capital expenditures to repair and replace existing network elements as they wear out. The Board took to heart the cautionary experience of systems where the initial outlay of building the network was followed by a period of under-investment. That requires about $400 million per year in 2018 dollars by 2040 and an even larger commitment beyond.
  • Sound Transit will be burdened by the debt overhang from ST3. By the 2030s, projected debt tops $17 billion and bumps up against statutory limits on debt that can be issued without 60% voter approval (1.5% of the property tax base in the RTA). Loading up on debt will super-charge the ST3 capital program, allowing more projects to be built more quickly than pay-as-you-go. It also means that debt servicing consumes 18% of tax revenues after the capital program peaks in the 2030s.
  • Additionally, the stretched ST3 financial plan is vulnerable to shocks that reduce revenues or raise costs higher than projected. Sound Transit has modeled scenarios where even a “small recession” early in the plan would push agency debt over the limit by 2032 unless the pace of the capital program is dialed back.

The graph demonstrates the implications in constant 2018 dollars.* Because we’re interested in potential capital investments, I’ve adjusted using the Capital Cost Index from the Sound Transit Financial Plan. This is the cost escalator Sound Transit uses to convert estimated capital project costs to future Year-of-Expenditure (YOE) dollars.

The black line is revenues from all sources (about 80% from taxes during ST3 with the balance from fares and grants). The line dips in some years as car tabs are reduced in 2028 and as grant income reduces with the completion of the program.

The capital program (in blue) pushes outlays well above revenues for most of the ST3 program. Sound Transit manages that by issuing bonds. By the mid-2030s, the bond capacity is nearly exhausted and further outlays cannot exceed available revenues by much. Only with the substantial completion of the capital program are revenues in surplus. Without fresh voter approval, those surpluses are available for future debt repayment and tax reductions.

In 2041 as the first trains roll to Issaquah, 82% of revenues from all sources will be spent operating and maintaining the network, keeping it in a state of good repair, and servicing the bond debts associated with the ST2/ST3 capital program.

Here’s the challenging math of new capital programs. The 2018-2035 capital program averages about $1.7 billion per year in current dollars. The 2041-2060 revenue surplus averages only $550 million. There’s some scope to expand revenues beyond that with grant revenue and by replacing maturing debt, but those would be offset by even more operations expense and more debt servicing costs.

Historically, most agencies go through a period of rapid expansion after which they must pivot to operating the system they have with minor extensions. Unless the Legislature and voters approve a large expansion of Sound Transit taxes, that is the future of Puget Sound light rail.

* The graph mimics one, sometimes referred to as the ‘blob diagram’, that has been shared at Sound Transit Board meetings. I redrew that chart (always shown in YOE dollars) in inflation-adjusted dollars to more realistically capture the funds available for future capital projects. The Capital Cost Index (CCI) is a forecast inflation factor for products used in building major capital projects. Historically, it grows somewhat more quickly than the Consumer Price Index (CPI), but not as quickly as Right of Way costs. All data is from the 2018 Sound Transit Financial Model.

68 Replies to “Sizing ST4: future rail expansions will be smaller”

  1. But… there will still be money available to make spot improvements. That is important.

    In 40 years, none of it will really matter anyways. If we don’t quickly make the pivot to a country that once-and-for-all ends fossil fuel use, a change that can only happen with a full embrace of public transit, we’ll be facing a world of famines, mass migration, and war. Every year that passes with no change (or currently a regression backwards) makes me more pessimistic about the future.

    1. At full ST3 buildout in 2040, ST is projecting a regional 400,000,000 reduction in annual vehicle miles traveled. In 2017, there were 31,000,000,000 annual vehicle miles traveled in the metro area. So, about 1.2% of miles traveled switch to transit.

      If the goal is carbon reduction (which I agree it should be), we’d be better off spending hypothetical ST4 monies on electric vehicle charging infrastructure (including buses) and direct incentives to encourage consumer adoption of EVs and PHEVs and address the emissions of the 98.8% of miles traveled that don’t switch to transit.

      By the time ST3 is complete we’ll have as many track miles as the Chicago El or the DC Metro. I think it’s pretty clear that’s close to (or over) the point of diminishing returns for this metro area.

      1. There’s a big difference in utility between 1 mile of track connecting two dense urban neighborhoods and 1 mile of track extending further out into the low-density suburban fringe. Track miles are the wrong way to measure the size of the system.

      2. Can’t argue with that – if you’re looking to reduce emissions, the low-density suburban tracks are *more* valuable because they take the long-distance Everett-Seattle type commuters and their CO2 emissions off the road. Urban lines capture mostly short trips, many of which would be made by transit anyway.

      3. There are counterfactors. Fewer people live in Everett so the total number of trips is smaller. People who live in Seattle are more inclined to take transit on average, and if Seattle had great multi-line grid-like service running every 5-10 minutes you’d see ridership go up to a level like cities that have comprehensive subways. Alon Levy quantified this, an idea I’ve had for a long time, that if you double frequency it more than doubles ridership, and if you add destinations (e.g., a cross line) you get a similar network effect. A lot of those people are coming from buses, but the bigger issue is that it makes it more feasible for people to give up their cars completely, and that’s a dramatic carbon drop, like a passivhaus that rarely needs energy for heating/cooling so they can just not install a full-sized heater or air conditioner. And that level of transit/density/destinations allows more people to switch to transit on the margins. Everett can’t match that because it’s a smaller city, and the supermarkets and shops and evening activities people go to aren’t on Link or other transit paths. Everett will get some urban villages around Everett Station and downtown, but they won’t be large enough or comprehensive enough for people to do everything there and find the rest on Link. Downtown Bellevue is large enough for many of people’s needs, but the plans for the Spring District or Everett aren’t large enough for that, and it remains to be seen how big downtown Lynnwood will really grow.

      4. By the time ST3 is complete we’ll have as many track miles as the Chicago El or the DC Metro. I think it’s pretty clear that’s close to (or over) the point of diminishing returns for this metro area.

        Right, but as Mars pointed out, the problem is that we will have blown it all building the wrong things, while important, cost effective projects will not have been built. We will be stuck with the maintenance costs of a long distance, largely suburban line, as opposed to a comprehensive urban rail network. There will still be areas that are essentially part of downtown — like Belltown and First Hill — without rail. There will be connections (like Ballard to the UW) that will be faster via extremely slow buses or a moderate jog. No one can really look at our system — when fully built out — and say that it is anywhere near complete. Yet at the same time, it is far bigger (will have far more miles of track) than any city this big. This is one of those “Only in America” moments. Seriously, I can’t think of any other country that completely screws up their transit projects this way.

        The problem is that it leaves the area with very tough choices. One option is to bite the bullet, and keep spending more money. This will put us into weird territory (if we aren’t there already) as a region with extremely high transit spending without very high transit usage. Or we skimp a little bit on service to the under-performing areas. There is great precedent for that (just about every city that runs trains like the ones we are building runs them infrequently) but it runs contrary to the implied if not promised service levels. Are we comfortable telling folks from Tacoma that their train to Seattle will not only be slower than the existing bus most of the day, but also less frequent? The final option is the one that Dan alludes to. We muddle along with a poorly designed system and pretty good bus service, forever regretting our previous choices.

        My guess is their will be some sort of combination. I seriously doubt the system will ever expand any farther into the suburbs (for the reasons Dan mentioned). But Seattle itself still has cost effective mass transit needs that may be built. Nothing grandiose, like the vision of Seattle Subway, but a few miles of subway here or there. Seattle residents — like all urban residents — are comfortable with very high tax rates, and it wouldn’t shock me if they accept them to build things that are actually far more cost effective than anything in ST3. To get there though, we will likely have to wait a long time, as folks realize that much of what is being planned for, and built, is not very good.

        In that regard, it is kind of like BART. Fifty years after construction and they are just now considering expanding it where it should have been build in the first place. So I figure by 2060 or so we should be able to plan something similar.

      5. “it is kind of like BART. Fifty years after construction and they are just now considering expanding it where it should have been build in the first place.”

        They are? Where? All I’ve heard is mutterings about a second Transbay Tube somewhere, San Jose Diridon, and far east extensions.

      6. Mike, the second Trans-Bay Tube is blue-skyied to go through SoMa and then swing up past the Civic Center and out Geary. It’d be a two-seat ride to most places in the traditional downtown, but much faster than Geary BRT.

      7. “This is one of those “Only in America” moments. Seriously, I can’t think of any other country that completely screws up their transit projects this way. ”

        No other country gives the suburbs an effective veto over the region’s mass-transit decisions, land-use restrictions, revenues and budgets. No other country has autonomous suburbs with sacrosanct boundaries as close to a growing city as Shoreline, Burien, Tukwila, and Mercer Island, — boundaries which exist because the bulk of the white middle class and upper class fled city municipalities to escape school desegregation because it couldn’t cross school-district boundaries. And they took their political power with them and have great clout in the state government because they’re People Like Us and the actual majority of the region’s population. And being the more-affluent people originally, they’d drunk the strongest kool-aid for low-density single-use residential areas served primarily by highways — a lifestyle they could afford. (And now the suburbanization of poverty makes this especially cruel.) All this is the context that explains why Sound Transit and BART are the way they are.

        Vancouver, Toronto, and German and other European and Asian cities start with “Where can high-capacity transit be the most effective, serve the largest number of people, and bend the curve on car use?” Then they Just Do It, putting high-capacity transit in those corridors, and planning the largest TOD concentrations around those stations, and give them adequate budgets to build and operate and maintain them, and don’t give suburbanites and NIMBYs lots of ways to veto the process, especially via environmental reviews that presume that transit amenities are negative impacts. And those other countries also have frequent buses throughout the suburbs and all-day transit to surrounding rural towns, and (except Canada) all-day commuter rail throughout the metropolitan area. And they don’t have regulations and business traditions that drive up the cost of construction so high. All these differences explain why Pugetopolis and Bay Area transit are what they are. It’s not just bad decisions about Link corridors and stations; that’s just the tip of the iceberg. And we shouldn’t single it out as if it exists in a vacuum. The blame lies at the totality of NIMBY power, oversized suburban clout, a state legislature that generally sees cities as an evil to contain and a gold mine to mine, and public attitudes that were shaped by Futurama and Le Corbusurian propaganda of the mid 20th century.

      8. None of the transit investments will directly reduce carbon emissions, because existing car capacity will always induce demand. The only ways to reduce emissions in total are electrification, congestion pricing, and reducing lane miles. Transit allows us to get around without a car, but it must be paired with reducing car capacity to reduce emissions.

  2. Very nice, thorough analysis. It would be nice to see the system expanded, but it is clear from this analysis that the bonding will be nearly maxed out and something like changing the car tab fees without maintaining a revenue neutral position will reduce the speed of the system build-out.

  3. Is the system expected to be operated at less than full capacity after 2042? There is a dip in the O&M curve that year, and it’s flat thereafter. Also, it’s a pretty complacent projection that a no-capital expenditure system can have flat maintenance costs. Trains get damaged and they get older every year.

    Or does “State of Good Repair” include railcar replacement?

      1. Okay, so it’s not what agencies usually mean by SOGR, which is maintenance backlog, but rather ongoing maintenance and rolling stock replacement.

      2. The term is, I think, used somewhat interchangeably to mean different things. As an accounting matter, SOGR is capital and O&M is expense.

        So periodically changing the oil or replacing the tires on the bus is maintenance and is expensed. Replacing the bus at the end of its economic life is capital and therefore SOGR.

        But if an operator didn’t replace the worn out tires, the bus would be unsafe and not in a state of good repair. You can find analogs anywhere a maintenance backlog could shorten the life of under-maintained assets.

        There is a document ‘Seattleite’ linked to which has a list (on page 14) of 2018 SOGR projects.

  4. And why are “Revenues” rising after 2040 in Constant Dollars? Does Sound Transit expect sales taxes and — if they still exist — car tabs to rise faster than inflation? That should set Republicans’ hair on fire.

      1. That’s a nice, comforting and aesthetic assumption. But if we lose global seignorage and have to buy the State of Good Repair and maintenance materials from overseas in China’s money, you can invert that little “curve”.

  5. By reaching out as far as ST3 does, it pretty much disincentives voting until 2028 or later too. Plus, the expansion ridership increase is so high that it’s hard to predict how that will play out in terms of operations and overcrowding. Fixes will eventually be needed. I think too that ST3 will end up underfunded — so I think a pitch for more funding to complete ST3 will be needed.

    Perhaps overshadowing all of this is a awareness that most productive corridors that remain will have service and so geographic expansion will be less motivating.

    I would not be surprised if future expansion takes a different funding approach — subarea-only measures, city measures, state support (our region will represent 60 or 70 percent of the state by then). I’m rather amazed that development fees are not yet part of the mix.

    Finally, technology is changing transit significantly,so that our who definition of investment preferences could easily shift.

    1. I’ll do a side-by-side of the financials you point to when not at my day job desk, but I expect the 2018 variances vs budget should be entirely captured in the ST financial model. The data in the model is from Fall 2018, so changes over the course of the 2018 budget cycle were mostly known by then and captured in the model. Maybe some minor shifts by the time Q4 closed, but not significant.

      There is a general shift in the 2018 projections toward both higher costs and higher revenues than previously expected. So far, it’s roughly washing out on the long term impact. The big imponderable is whether an eventual cooling in the economy hurts revenues more or less than it helps on costs.

      The lower than budget spend rate in 2018 came from delays in projects, so not helpful in the long run except very indirectly via some debt and operations savings.

  6. Seattle-only package could be Ballard-UW, Ballard-Northgate, Aurora, and maybe W Seattle – Burien. A King County ST4 expansion could include Burien – Renton – Bellevue.

    South Sound would be electrifying Sounder or adding a third track in spots along Link for more express Link service between Tacoma, Federal Way, the Airport, and Downtown. Then we’d need a bypass between TIBS and downtown. So many dreams lol.

    1. Some rail linkages to First Hill and Belltown (western side) should also be in that mix. The densities in these areas are significantly higher than any other area unserved by Link.

      1. I am SO pissed that they’re filling both sides of the Battery Street tunnel. It points DIRECTLY at the Denny Way station which will be only three blocks from the curve to Aurora. They should have kept half of it for a transitway.

    2. You’ll never get electrification on the Seattle Sub.

      Even though Thurston County is NOT apart of the ST taxing zone, I’d like to see Sounder extended to Olympia, with 3 new stations. Would stop at the existing station at Centennial Amtrak station, use the rebuilt East Olympia line to Downtown Olympia. Use the existing trains and run 4 a day. Could be a joint venture with BNSF, WSDOT, and ST to provide an alternative to driving. May potentially need to triple track Nisqually to East Olympia to reduce bottleneck at Nisqually Jct.

      I’d like to see Sounder increased to weekend service, 4 on a Saturday and Sunday in line with the current game schedules, preferably running the full 82-mile corridor instead of having to change trains at Seattle. Could just run the Everett set, except for game days obviously.

      I’d also like to see Sound Transit complete the North Sounder corridor upgrades (double tracking, additional crossovers) and increase to 6 trains a day.

      Lastly, adding a third main in Kent and from Auburn to Tacoma (TR Jct/L Street)

      1. “Even though Thurston County is NOT apart of the ST taxing zone, I’d like to see Sounder extended to Olympia, with 3 new stations. Would stop at the existing station at Centennial Amtrak station, use the rebuilt East Olympia line to Downtown Olympia. Use the existing trains and run 4 a day. Could be a joint venture with BNSF, WSDOT, and ST to provide an alternative to driving.” Ditto for getting to Marysville. You’d stop somewhere in the south part of town and at the Kruse Junction area where the shopping center is, close enough for a shuttle to the outlet mall and the casino. Sounder would seem like the only practical option to serve the area with rail, and the tracks already exist. People could continue on Sounder or switch to Link. It really is surprising how much traffic there is between Everett and Marysville, presumably most people are going to/coming from points south.

      2. Electrification of both freight and passenger rail is probably the most cost effective (if not simply the most effective) way to reduce rail related global warming emissions. If we are going to have a “Green New Deal” than it should include that well before we talk about expanding commuter rail or building a high speed national train network.

      3. Bingo, Ross, and Bravo for saying so, at least for the dozen or so primary routes nationally. However, it means that most tunnels on those lines would have to be deepened or raised by a couple of feet to clear the pans.

      4. I’m guessing there are instances (such as the one you mention) where battery/electric would be more cost effective. Run on wire most of the way, but run on batteries for short segments.

      5. Excellent idea. Most locomotives have tons of ballast to maintain tractive effort. That weight could be batteries, even cheap lead-acid ones. The pans would have to be reliably retractable, but I can’t imagine that would be a difficult engineering problem.

        The batteries would also be proof against the odd breakage of the catenary that didn’t also block the tracks.

  7. I have to think that any ST4 would have to break the subarea model, which ST3 already strained to the limit. Seattle would likely vote for many potential projects (Ballard-UW, Ballard-Crown Hill, an Aurora line, West Seattle-White Center, Duwamish bypass, etc) but unless we’re talking about many billions for Sounder capital to make it a frequent all-day service, I just don’t see which suburban projects would be plausible and likely to carry a vote. Tacoma Mall and North Everett? What could you possibly build in East King in ST4? South Kirkland to Totem Lake?

    Allowing subareas to vote and tax themselves separately might be the only way further packages are possible. Would the legislature have to change the enabling law for this to happen?

    1. The subareas’ interests will probably diverge after ST3, with Seattle wanting several large projects and the other subareas not wanting much. I’ll address the other subareas first because they’re less complicated.

      Snohomish: Stated goal is extending Link to downtown Everett and Everett CC, which would be Link’s final terminus. That’s inexpensive because it’s a short distance and presumably elevated. I haven’t heard Snoho ask for anything else. The biggest thing it wanted was Paine Field but that’s already in ST3. … My wishlist would be extending Swift Green to UW Bothell (and dt. Bothell and 405 Stride), expanding 405 Stride to multi-line service, and finishing the six planned Swift lines. Possibly a Tacoma Link line or two. Don’t expand Sounder North; consider retiring it. … Snoho has a problem with major growth in Marysville, Maltby, Highway 9, and Monroe which are outside the ST district. It may make sense to annex Marysville and extend STEx(press) or Sounder to it, or augment the planned Swift line (201/202) or a third Stride line.

      Pierce: Stated goal is extending Link to Tacoma Mall, which would be the final terminus. This also is inexpensive because it’s short and presumably elevated. Secondary goals are a few Tacoma Link lines, DMU/STEx in Orting, STEx something around Puyallup, maybe something Spanaway-ish. Some residents have asked for Tacoma-Bellevue STEx but I haven’t heard boardmembers request it. More Sounder is always popular but we’ll have to see whether ST gets a good ST3 deal from BNSF. If it does it will be almost hourly with “some evening and weekend service”, so what would they want beyond that? Some transit fans have suggested extending Link a little further to South Tacoma and Lakewood. I’m assuming that only the Tacoma Mall extension is definite.

      South King: It’s unclear what South King wants if anything. The West Seattle – Burien – Renton Link study found it to be low-ridership for the cost, and there’s been a loud silence from the cities since then so they may have cooled on it. Renton-Bellevue is the least likely because ST believes it will be many years before sufficient ridership develops there. More Sounder is always a hit, but that again depends on how well this next round of expansion goes with BNSF. Metro Connects will significantly improve South King’s circulation, including all-day expresses on Seattle-FW, Seattle-Kent-Auburn, WSJ-Burien-SeaTac-KDM-Kent, Renton-Enumclaw, and Auburn-Snoqualmie. (Some of these may end up being peak only.) What would South King want from ST beyond that? Also, South King is the poorest subarea so it can’t afford that much. Seattle-Renton has come up officially, either from Rainier Beach or from a “Metro 8” line at Garfield High School.

      East King: I don’t see strong consensus on Kirkland Link service (Totem Lake-Issaquah, Kirkland-520-UW-Ballard, Kirkland-Sand Point-UW-Ballard, or Kirkland-Bothell-UW) or on upgrading 522 Stride to Link. Possibly expanding 405 Stride to a multi-line service. (To address downtown Kirkland, Bothell, and Woodinville.) I don’t see any one idea getting strong traction at this point, or a great willingness to pay for it.

      North King: Here’s where there’s strong demand for several large projects. The most needed are Ballard-UW and “Metro 8” (a vague concept connecting Uptown, First Hill, and the Central District). The most regional (i.e.,multi-subarea) support would probably come from 522 Link, Aurora Link, and the Georgetown Bypass idea. But the Georgetown Bypass would primarily benefit South King and Pierce, and they have not requested it nor offered to contribute to it at all. They were so uninterested in it they let it be deleted from ST’s long-range plan in 2014. I don’t see a lot of support for a Ballard-Northgate segment. The long-range plan has a 522 line to Northgate, but when they get around to solidifying it they may reroute it to 145th, or down 35th-ish to the U-District.

      In the past ST1/2/3 were held together by Seattle wanting lots of lines and tunnels and the other subareas wanting long extensions to Tacoma, Everett, and Redmond. Now that that’s budgeted I can see Seattle strongly wanting more lines and the other subareas strongly wanting only a little bit and weaky wanting some other things. So it would make sense to allow each subarea to decide its own tax rate and vote timing. But that would require splitting the tax district, because “equal representation” in a tax district means everybody having the same rate, the same vote, and equitable benefits. Splitting the tax district might mean splitting the board, or allowing only subarea boardmembers to vote on subarea issues. How much could ST remain a single agency under those conditions?

      1. I think for anything beyond ST3, the city would have to go it alone. Ballard-UW built as a P3 project using EMU trains like the Canada Line paid for with the city’s dormant monorail tax authority?

      2. The monorail authority is estimated at $1 billion, so Ballard-UW would be just barely below or above that. The main cost is tunnels, not the vehicle type. 45th is arguably too narrow for an elevated line, and the neighbors would be out with pitchforks, and Pacific Street would be too far out of the way for the U-District pedestrian center, so there’s no alternative to a tunnel.

      3. I agree with both of you — it is highly likely that any rail expansion would occur only within the city. The suburban areas and smaller, less densely populated cities (like Everett and Tacoma) are simply inappropriate for additional rail (although that didn’t stop them before). You could build a massive “buses in the suburbs, trains in the city” system, but that requires way more leadership than we’ve had in a while (and its failed in the past). It would be hard for a leader to suddenly pivot and say that Everett, for example, should build a system of express buses to Link along with a big frequent network within Snohomish County when it runs contrary to what they have been saying for years. Oh, and they are building it anyway. Once Swift is built out — and ridership is higher than Link in the same region — it will be tough to tell those areas that they need to spend a bunch of money on transit when they will obviously have more important concerns. By then the U. S. will continue with the more globally common pattern of poverty moving to the suburbs. When the last vestiges of red-lining becomes a distant memory, and yet another generation has to explain that Oakland, Brooklyn, or the Central Area “really isn’t that way anymore Grampa”, I doubt that people there will want to spend billions on trains within the city instead of what every city prioritizes (police, schools, health and human services, etc.).

        But Seattle — even after the billions upon billions that will be spent serving secondary priority places like West Seattle — will still have large unmet transit needs, and will have to choose between muddling along with bus service or building it themselves.

      4. I wasn’t suggesting that vehicle type was a major cost factor, just that you can get something that provides the functionality of light rail without falling afoul of the prohibition of using the tax authority on it.

        Shorten the tunnels by running elevated west of 3rd and coming out under the Ship Canal Bridge to a station on Campus Pkwy and you might get under $1b.

      5. Probably the most suburban place that would be good to get light rail is Lake City, as it does have a bit of density to it. Also, there are an awful lot of trains going to Northgate that are predicted to be quite empty north of there, so splitting the line into a Lake City branch wouldn’t be too far fetched. Getting the line onto Lake City Way seems like it would be challenging though.

      6. I know DMUs as having a diesel engine in each car. How are EMUs different from Link? Do they still get power from an overhead wire or third rail? Doesn’t Link have a motor in every car? What else would the interior cab ends control?

      7. “Oh, and they are building it anyway.”

        Right, this is what I meant that Metro Connects will be a game changer in South King. Community Transit has a similarly robust plan relative to the county’s size and tax tolerance, and Pierce Transit’s plan is a large step in the right direction. So the issue will be, what ST additions will people want from ST when these are running? The increased local services — both RapidRide, Swift, all-day expresses, and frequent local grid routes — will lessen the need for additional ST services.

        The Everett and Tacoma extensions were predicated on the fact that they’re the largest cities in their county and historically two of the three largest cities in Pugetopolis (“the Seattle-Tacoma-Everett area”). There will be no more “Everetts” or “Tacomas” to go to.

      8. The main difference would be third rail power vs overhead catenary. So, no real practical difference at all, other than being “not light rail”

        Plus if you go full grade separated, with platform doors, you could go fully automated like Skytrain.

      9. But “Pacific Street” goes directly to Lower Fremont and has more employment than central Ballard by a significant degree. And if you don’t like a tramway — which would be perfect for a line four miles long — large parts of a line going that way can be elevated, certainly west of Gasworks Park. And east of there, as far as the University Bridge there is enough land to run at-grade. If you are worried about missing the U-District dig a stretch of tunnel up to 43rd and east through campus to U Village. That’s only a mile and a quarter at most of tunnel.

        You can come down to the surface along Market west of Eighth. It’s only a mile to 24th from there. If Ballard-Downtown is elevated, dip 15th under Market and put a center platform station right under the Green Line station with direct platform-to-platform escalators to and from both sides.

        This is like Ross’s idea for a Ballard-UW subway via Lower Fremont, but mostly on the surface and elevated. It doesn’t have a station in central Wallingford, but it could have one for the high rises going up along Lake Union.

      10. Yeah, here is my idea: It makes several assumptions:

        1) The Ballard to downtown line runs elevated on 14th, then curves around to serve Ballard. This is where a lot of the cost savings come in (shared track and two shared stations).

        2) The station in Fremont has the entrances shown, as well as the changes to make it work. The idea is that you have one side that opens out onto Lower Fremont (Fremont Avenue and 36th) while the other side opens up the hill, close to a bus stop. But the bus stop would have to be added on Aurora, and there would have to be an easier way to get from the troll area to bus stop. The latter doesn’t sound that difficult, but adding a stop for the E next to the bridge might be hard.

        3) We can afford the tunneling. This would require a lot of it (the train would pop out of the ground around Leary). The main savings come from sharing the stations.

        The key really is the first point. The idea is that we build a bridge over 14th (to please the port) but then send the train curving over to 15th. Eventually it is extended to 15th, and eventually the other line ties into it.

        Without that, it just doesn’t make sense. I think it is far more likely that they tunnel straight across (with a station at 46th and Aurora). That could mean running cut and cover or elevated starting at 3rd NW, which would result in significant savings (that is what Ron was getting at).

      11. A branch line to Lake City should do so at Roosevelt, with a short tunnel extension through the choke point south of 80th – with a station at 15th – and then elevated into Lake City proper as far as NE 145th. As you say, trying to get a line from Northgate down to Lake City would be challenging and almost assuredly can’t be done on the surface. An alternate would be extending a Ballard-UW line past U Village and up 25th to LCW, although there is very little but SFH along that route save the little Ravenna business district. The advantage to either is it leaves a line pointing further north towards Kenmore/Bothell/Woodinville should that ever be a need or should East King just want to throw money at something.

        Lake City doesn’t just have “a bit” of density – despite its post-war origins I believe it has the highest density census tract between the U District and the Canadian border (RossB mentioned this before, IIRC). Not only does it have some existing density, it has a decent amount of flat land that is underutilized for in-city property, no views for NIMBYs to complain about blocking, and single-family housing stock that is neither historic nor particularly expensive in Seattle terms. It would be one of the easier upzones to get through the Council, although without great transit that’s unlikely.

      12. But “Pacific Street” goes directly to Lower Fremont and has more employment than central Ballard by a significant degree.

        Not from what I can tell — greater Fremont and Ballard look almost identical (*. Things have changed since the last census, with new offices built close to Stone Way (Tableau moved there) but there are plenty of new offices being adding in Ballard as well (right now it is mostly retail and medical).

        The main point, though, is that lower Fremont/Wallingford has more employment than the upper area (and roughly the same population). That is a good point, which is why a lower route might make more sense. A combination of surface and elevated or maybe even cut and cover (although the water table may be an issue) might save a lot of money, while providing for good stops.

        The biggest drawback to such a route is that it doesn’t tie in to the bus network very well. It greatly improves the rail network (making trips like Northgate/Lynnwood to Fremont/Ballard much faster) but much of the city will still depend on buses. Likely the great majority or riders, including the great majority of riders in that region. In other words, no matter where you build a Ballard to UW subway, the vast majority of people west of I-5 and north of the ship canal will have to take a bus to get anywhere. Building the Ballard to UW line without good connections to the north-south buses would greatly reduce the value of the line. Someone trying to get to work at the UW** from somewhere on Aurora would likely just continue to transfer to the slow 44. The only other option would require Metro to run an additional bus route from Aurora down to lower Fremont (which seems unlikely). Same with Phinney Ridge/Greenwood (the folks on the 5). In that case I could see Metro running a bus to Fremont, but that would make a ride to the UW or Ballard much slower, and water down the system (i. e. have less frequent buses as Metro stretches the system to serve Link).

        That is a trade-off though, and one that doesn’t have an obvious solution. But my guess is unless the lower route was a lot cheaper, tying in to Aurora buses would be better.

        * I can’t link to a particular employment map (the way I can with population data). But if search for “Seattle” and then select the city and get the default numbers, you can see how Ballard and Fremont are pretty similar in terms of employment.

        ** As you can see from that map, the UW is a major employment center (along with being a major cultural and educational center).

      13. Scott, you can’t have a diversion at Roosevelt because ST didn’t plan for it. It’s the same problem as what embargoes Ballard-UW-Downtown using the recently built tunnel. ST will never put a level crossing for a revenue line in the Northgate Link tunnel, so a diversion to Lake City would have to be a flying junction for the southbound trackway. It would either under- or over-run the Northgate Link tunnel then descend or ascend to the same level just north of the Roosevelt Station and join at that point.

        While TBM’s can produce such a convoluted tunnel, the killer to this is that both of the Northgate tubes would have to be broken into with all the engineering problems that creates because the rings would no longer be solid to transmit soil pressures, but weren’t designed not to be.

        Also, Roosevelt isn’t going to be a busy enough station to warrant a diversion there. That would be tantamount to reducing service to Northgate to every eight to minutes during off-peak operations.

        There is a tail track at Northgate which might be taken out of turnback service and made into a connection for the southbound track, but it would be have to be very steep unless the separation between the tracks north of Northgate Way were widened so it could descend over several blocks.

        Any of these retro-fittings would be extremely disruptive for a good long while, if indeed they could even be engineered.

        No, the only way Lake City is going to be served is by an extension of a Ballard-UW line that swings north from U-Village or Children’s OR by extending the Green Line through Crown Hill and Licton Springs as Seattle Subway has suggested.

      14. Yes, the E is a problem going through Lower Fremont but if there is so much travel between the E north of 46th and the U-District, just have an overlay line that uses Stone Way north to the 62nd Street ramps. If there are ever bus lanes on 45th it would be “natural” to have a RapidRide using them too.

      15. Thanks, Tom – good points all. Appreciate the thoughtful comment and breakdown. None of it will happen in my lifetime anyway, so best of luck to those of you still here! :)

      16. Nobody planned the entirety of the London Underground or any other such heritage system 50 years in advance. Sometimes stuff has to be added, and there are ways of doing this that are less disruptive than others.

        A flying junction such as the Chicago Transit Authority has between the green and orange line could be added to each side of the existing line with disruptions only during actual rail welding.

        Various other methods have been used to add lines in various other places. There are obstacles to overcome but nothing impossible.

  8. It will be interesting to see how the suburb communities will react to transit investments around 2025. Northgate Link, Lynwood Link, East Link, and Federal Way Link will be done. It may be easier to get communities to come together to find a tax solution when more of them get to use the most expensive part of the Sound Transit network. Right now, my friend’s biggest complaints are paying for higher taxes and only seeing blue busses. Explaining the tax structure doesn’t help. They need to see the reward for their sacrifice. In 2025, St2 will be done with one year to make things smooth. I will guess more people will look for solutions then. Unfortunately it really isn’t soon enough.

    1. Public mood will probably go two diverging ways after 2024.

      1) “Wow, I love the Lynnwood, Redmond, and Federal Way extensions! We so needed them!’ And after ten years, “How did we ever get along without them?”

      2) “That’s a lot of money for Issaquah, Everett, and Tacoma, and it won’t have nearly as much benefit as Lynnwood, Redmond, and Federal Way. And we’ve still got fifteen more years of construction, and it’s crowding out other possibilities, and I can’t stop the taxes. I think enough is enough.”

      522 Stride will clearly be a hit because there’s a lot of built-up demand for it and it connects walkable trip pairs (or at least a one-sided trip with a P&R at one end). 405 Stride is less clear because it dumps you off at a freeway exit in the outskirts of Bothell, Kirkland, and north Renton, and at the edge of downtown Redmond. The only stations with good walkability are Bellevue, Lynnwood, and Burien. That’s the problem with the existing 535 and 560 and the previous 560 — they stop in the middle of nowhere so they don’t attract many people. 405 Stride will still be somewhat popular for Lynnwood-Bellevue, Canyon Park, a long walk go Kirkland Google, and the like, but it probably won’t reach high ridership or mode share, so will people think, “Meh, we spent a lot and it doesn’t do much; we don’t want more of that.” Issaquah Link is in a similar situation. It’s way overkill for Bellevue-Issaquah or South Kirkland – Bellevue, but its Bellevue College service will be a sleeper hit.

      1. previous 560 340. That was the Metro express from Shoreline P&R to Bothell, Bellevue, Renton, SeaTac, and Burien. It stopped at the 160th, 70th, and Coal Creek freeway exits, which are not exactly near the built-up parts of Kingsgate, Totem Lake, Kirkland, or Factoria that people are going to.

      2. I hope #1 is the thinking.

        As far as # 2 goes, I find it interesting what kind of different people might support transit after they have a good experience with it. During Husky season, I met 2 guys heading home from the game. They were going to Angle Lake. One of them was stoned. Get to that later. They both lived near the east side of Sumner. They told me they would love the link to go there. They seemed to be conservative, and would not get rid of their cars, but like the ride enough to support more expansion. One of them offered me free bag of weed after talking about it. Had to turn it down. Almost wish it was 25 years earlier. Anywaaaay, you never know what type of resident will support transit. Their support was based on a fun, cheap, relaxing ride. (We did not get into car taxes).
        Instead of telling people how much of a NIMBY or polluting SOV driver they may be, I would love to actually see what things would get them on transit more often. That means more support, more votes, more money. More Link.

      3. I used to think that too. But I am finding out that who you vote for is not necessarily a generic lifestyle stamp.
        There are plenty of 4:20 friendly people in the Bible belt. But this is off topic so I will stop.
        I hope that your original #1 thought catches on in 2025.

      4. @Jimmy — You turned down a free bag of weed? My, how times have changed. :) I hear you though. So many of my friends say things like “I only vape” or “I’m mostly into edibles”. Weed has gotten so good and so cheap that giving away a bag of weed is like offering a stranger several slices of pizza (“It’s good pizza, really, but I’m walking and don’t want to carry the box…”).

        Yeah, Anywaaaay. I don’t think the two views that Mike portrayed are the least bit contrarian. Someone in South Everett, for example, could get really excited about Link’s expansion to Lynnwood. Community Transit is building a very good transit system for an area that size. The addition of Link to Lynnwood will just add fuel to the fire. Not only will it mean that trips to Seattle during commute hours (both directions!) will be much faster, but it will unleash thousands of hours of service that has been spent slogging to and through downtown Seattle. This means very good frequency for much of the region, as well as consistent, dependable service to areas that have nothing right now. Trips within the county will be great, and trips to the more common places in Seattle will be much better.

        But then what? Unless you are right along the line, you really don’t get anything with an expansion to Everett. Most of the trips to Seattle require a bus than a train either way (or a drive to a crowded park and ride). There aren’t that many people who live close to a station, while there are plenty who live close to a connecting bus line (and those numbers will increase greatly). For example, let’s say you live between the future Mariner and Ash Way stations (along Ash Way). Lynnwood Link makes that trip to Seattle much faster (for commuting or an evening event). With Everett Link, things have changed only a little bit. Instead of taking a bus to Lynnwood (201/202) you take a bus to Ash Way. Big deal, really — you save at most a couple minutes. Getting to downtown Everett in the evening might be faster, but getting to downtown Everett in the morning would be the same (since the bus would be faster). Oh, and the 201/202 goes through Everett, which means that even in the evening (when traffic is so bad that taking a train is faster) you might still take the bus just for the one seat ride. You have other stops, I suppose, but none of those are very big attractions, and will likely be served well enough with buses anyway (especially as the bus system improves). ST3 will be an improvement, but it is definitely diminishing returns for just about everyone. I could easily see someone looking at Swift and wondering if it would have been fine to just run those buses to Lynnwood — or at most Ash Way — instead of running trains all the way to Everett. If the trains to Everett are largely empty most of the day (which I expect) and run less frequently than the Swift buses, then it would be hard for someone to get excited about building more train lines in Snohomish County.

      5. Ross, there’s a heck of a lot of peak hour traffic in the Lynnwood area for a bus to get through, and there is unfortunately no reason to expect that it won’t be congested all the way up to Mariner and beyond by the time Everett Link is finished. Also, if you lived between Ash Way or Mariner station, taking an Uber or on demand van or whatever these morph in to by then will be much more feasible than taking it to Lynnwood. Biking would also be much more tractable esp. if you have access to the interurban trail. In the case of Mariner itself, there are quite a few apartment complexes in walking distance–at least as many as there are at Lynnwood–and also the connection to SWIFT.

        Now Mariner to Everett with the Paine Field detour–that stretch does deserve to be picked on….

      6. So extend Link to Mariner or Ash Way. It doesn’t have to go all the way to Paine Field and Everett.

  9. 2060 is such a long time away that we really can’t predict with any accuracy what the population size, housing distribution, job types and locations, energy availability, inequality or social conditions will be. Who in 1978 would have predicted that:

    – People would commute from Everett and Tacoma. (Then it was Bothell and Kent.)
    – The region’s population would be 3.8 million, (Then 1 million, 90% of it in King County.)
    – The second-largest employer would have 45,000 jobs in South Lake Union doing e-commerce and web services (“What’s that?”)
    – Urban villages would become popular again.
    – Thousands of homeless people on the streets. (Then it was rare because SRO hotels existed, rents were cheap, inpatient mental-health services were widespread, the minimum wage had more purchasing power, and healthcare and college were inexpensive.)
    – Water shortages are a thing due to changing snow melts. (Then we didn’t have water shortages.)
    – Climate change may bring millions of refugees to the region from the rest of the US and internationally. (“What’s climate change?”)

    The most we can do is identify the trends: population will probably increase, climate problems will become worse. But we can’t say for certain it will be tenable to live in a low-density residential-only neighborhood where you have to drive even to the supermarket or a P&R. If it’s not tenable then infill housing will have to happen and the population will be more compact. But we’re in a race against time because we don’t know how long overseas shipping will remain viable in light of climate change or possible wars, and we depend on Asia for a lot of our steel and building materials and products. So we need to build enough housing and of a sustainable type and ramp up local agriculture and manufacturing before overseas supplies and markets get cut off and energy becomes too expensive. Maybe solar-powered buses will work, maybe solar SOVs will work at scale, but we can’t be certain.

    And 2060 is so far off that a person who’s 18 now will be 59 then. The people who decide what to do after ST3 will be a different generation, and they’ll have different needs and priorities. And if inequality continues to increase, we’ll be like Mexico thirty years ago with a handful of ultra-rich people and a lot of dirt-poor people who can barely pay a bus fare much less taxes for a new light rail, and they’ll be living in slums because they can’t afford regular apartments.

    1. No, we can’t predict what will happen in 2060, but we can certainly predict things based on certain trends. Many of the remarkable changes you mentioned were predicted. For example, the idea of suburban sprawl creating huge mega-cities was predicted back in the day (I remember studying it in high school in the 1970s). The Seattle metropolitan area (which includes Tacoma and Everett) has seen high growth since the 1950s, so it isn’t really shocking that we saw that continue. Long distance commuting may have increased, but it is still rare. Homelessness (and other social ills) were predicted after Reagan took power. Even South Lake Union as an extension of downtown isn’t too hard to predict given its geography. It is both close to the old downtown as well as the UW. Downtown also expanded to the west (towards First Hill) and the northwest (towards Belltown and Lower Queen Anne). The only area that hasn’t seen much expansion is to the south, and that is largely because it is still reserved for industry.

      We can make similar assumptions. It sure looks like the trend is towards greater urbanization. That could change, of course, but much of what drove suburb growth is unlikely to return. Driving everywhere is simply less attractive, as folks realize that traffic is not going to get better, and even the best transit system will still require a long trip to the city. White flight is generally a thing of the past as well. You can still get more house for your money in the suburbs, but more and more people would rather settle for less house than settle for a tougher commute (or just more driving in general).

      I also don’t buy the “climate refugee” argument. Seattle isn’t really that special as far as the climate is concerned, and people manage to live anywhere (Phoenix is way too hot for me know, a few degrees won’t matter). I would expect employment to drive population trends. I wouldn’t expect a huge move back to the Great Lakes region even though they obviously have plenty of water for example. Besides, if things really do become that bad — if millions of Americans (who are relatively well to do) move because of the climate — then the human race is screwed, and we will have much bigger concerns than a good rail system.

  10. BTW, excellent pic. Train is about to cross 108th- South Kirkland Park and Ride a short elevator ride to the right. Passing Downtown Kirkland in about five minutes. Twenty more to Totem Lake, the rest future history. It’ll be a beautiful train ride, and huge favorite with the grand-childrens’ children on homeowners presently fighting the project.

    Whose political presence justifies every fare-break and other inducement to be sure everybody age six (personal experience here) rides maximum number of trains possible. Whose interface will spread beneficial attraction to buses as well. For funding projections, develop formula for calculating how many years it’ll be until every child passenger can vote, and try to time elections accordingly.

    Get them early, and you’ve got them for a lengthening lifetime of pro transit votes. Source of whatever revenue this posting postulates. And remember: most six year olds are responsible for ever widening number of future ones.

    Mark Dublin

  11. And South Kirkland is a short elevator ride to the left. At age 73, we don’t stay six forever, let alone anywhere near that cute. Thanks for helping me find the elevator. There’s a great trolleybus running on Bellevue-Kirkland Way in its own signal-pre-empted lanes.


  12. Just a small suggestion: Maybe we should not be thinking about ST4 until ST3 is done or just about done.

Comments are closed.