Public Interest Transportation Forum -

Sounder Commuter Rail From Everett to Seattle:
$3 Fare Doesn't Cover the $98 Cost Per Rider

by Emory Bundy and Tom Heller

Introduction from the Editor, Public Interest Transportation Forum

Sounder commuter railroad service on existing train tracks in western Washington State from Lakewood (south of Tacoma) to Everett is one of the three modes of high capacity public transportation that Sound Transit promised to establish in the central Puget Sound region in the ten-year period 1997-2006.  As of April 2004, Sound Transit -- through an agreement with the Burlington Northern Santa Fe (BNSF) Railroad, which owns the tracks -- has been able to establish three daily round trips between Tacoma and Seattle, and one daily round trip between Everett and Seattle.  The other two modes high capacity transit [HCT] modes are express bus on HOV lanes, and light rail.

As Sound Transit stated in the 1996 Sound Move Plan:

"Because commuter rail is a major regional investment, there must be continued, deliberate and careful consideration given to developing this new north-south HCT corridor. Communities, local jurisdictions and citizens will be part of the project from beginning to end."

In the spirit of continued, deliberate and careful consideration, two Seattle citizens, Emory Bundy and Tom Heller, applied methodology used by the Federal Transit Administration (FTA) of the U.S. Department of Transportation to examine the cost per rider of Sounder commuter rail service between Everett and Seattle, both present and future.

Assuming ridership growth continues as it has so far or even grows a little faster, Bundy and Heller found that the cost paid by taxpayers for each traveler boarding the train in 2004 would be $298 per trip. By comparison, the same-day, one-way price of a ticket for this train is $3 dollars or less.  Assuming ridership growth continues, the cost per ride will be $98 for the period 2004 through 2025.

Sound Transit (also known as RTA, Regional Transit Authority) made promises in 1996 for the future performance of Sounder commuter rail in the Sound Move plan approved by voters in November of that year.  Among the promises were these two:
bulletRTA will establish performance objectives to monitor and evaluate commuter rail service productivity related to cost, ridership and other relevant measures.
bulletBased on system performance and productivity monitoring, the RTA may choose to modify service (i.e. increase or decrease service; add or drop stations; add or delete segments; change days and times of service, etc.).

However, in the presence of the data (described in an editorial by Bruce Ramsey published in The Seattle Times on April 7, 2004) from Bundy and Heller of very low productivity and massive cost per rider in providing a daily train from Everett to Seattle that operates at about one quarter of its capacity, the response of Sound Transit leadership is to add more trains. The findings of Emory Bundy and Tom Heller were answered by Sound Transit Board Chairman John Ladenburg in a Seattle Times guest editorial on April 16:

We can all agree that starting Sounder service north wasn't cheap. But let's use some perspective. Consider, for example, that the cost for Sounder operating rights on this 36 miles of dedicated right of way is about $7.1 million per mile, which is a bargain compared with the cost of building roads. For example, the cost per mile for expanding Interstate 405 is estimated at between $43 million and $51 million a mile; expanding I-5 between Northgate and downtown Seattle is between $75 million and $100 million per mile.

Chairman Ladenburg is trying to change the subject with this comment. But while diverting attention from the cost-effectiveness  point of Bundy-Heller, he does inadvertently make a new point that is very telling -- not mentioning the performance that each kind of infrastructure expense provides.  He compares the cost of operating passenger trains on BNSF's existing freight railroad tracks ($7.1 million per mile) with the cost of expanding parallel interstate highways such as I-5 ($75 to 100 million per mile).  This comparison is incomplete without also describing what each of the two infrastructures contributes to transportation in the region, yet Ladenburg leaves out the volume of trips that each of these infrastructures is expected to provide.

On the one hand, Sounder North hugs the scenic Western Washington coastline with tracks and train stations that do not conveniently serve many people, and moves only a few hundred commuters per day in one direction during peak hours.

On the other hand, Interstate-5 moves hundreds of thousands of cars, vans, buses, and trucks in both directions all day. Looking just at the morning peak period, the 2010 forecast numbers from the Puget Sound Regional Council's Metropolitan Transportation Plan suggest that I-5 will carry a multiple of at least 200 times more rush hour travelers than Sounder will. Using Ladenburg's cost per mile numbers (not verified), we can see that this large multiple in performance is achieved at only 14 times higher cost.  This makes highway spending that supports transit buses, van pools, and car pools the real bargain in a comparison with hundreds of millions of dollars for a few rush hour passenger trains. 

That said, the remainder of this page is an essay on cost per rider from Emory Bundy and Tom Heller, released April 6, 2004. They were assisted by Jim MacIsaac, P.E. of Bellevue, Washington and Tom Rubin of Oakland, California.

The Per-Rider Cost of Sounder Commuter Rail, Everett to Seattle, 2004 - 2025 Using Sound Transit Data and FTA Methodology

"If ridership doesn't improve, Sound Transit's operating expenses this year for its new Sounder commuter train between Everett and Seattle could add up to more than $38 for each one-way passenger."  (Eric Pryne, The Seattle Times, April 1, 2004)

"We're doing the best we can with the hand we've been dealt."  (Lee Somerstein of Sound Transit, quoted in The Seattle Times, April 1, 2004)

The operating expenses of Sounder, Everett to Seattle, currently are $38.54 per rider, per trip, as Eric Pryne reported, and Sound Transit conceded.  But if one accepts all Sound Transit's cost data, and the Federal Transit Administration's methodology for computing cost, the actual, total cost per rider for 2004 will be $179.  That's assuming Sound Transit reaches its current target of 678 boardings per day for 2004--which it admits it won't do.  It is running less than half that, 315 per day.  Should it reach 60 percent of its goal for 2004, 407 average daily boardings, which is optimistic, it will cost $298 per boarding.

At $298 per boarding, each round-trip commuter, for the year, will cost $155,000.  If the ridership target is fully met, each round trip commuter will cost $93,000.  For the level of subsidy, subtract $1,300 for fares.  The user will contribute about one percent of the cost.

Sound Transit says costs will go down over time.  That's true.  If all Sound Transit data is accepted at face value, in 2025 the per-trip cost will be "only" $42.  Over the 22 years, 2004-2025, the per-trip cost will be $56.  The cost for each round-trip commuter, 2004 to 2025, will be $640,640.  If all other Sound Transit data is accurate--but ridership is at 60 percent, itself a generous assumption--the cost, and the subsidy, for each round-trip commuter, 2004-2025, will exceed $1 million.  (All numbers are in current, 2004 dollars.)

These calculations apply the Federal Transit Administration's official methodology for calculating the per-trip cost, and cost-effectiveness, of a new transit system, like Sounder commuter rail.  The methodology is posted on the FTA website.

The calculations rely entirely on Sound Transit's official data--even though that data is not credible, neither the cost figures nor ridership numbers.  By the end of the first quarter of the first year, Operating & Maintenance costs for 2004 had to be increased 24 percent, $3,274,000 to $4,065,000.  They are not credible, given actual O&M costs of Sounder Tacoma/Seattle.  As for ridership, originally daily ridership for the first full year of Sounder Everett/Seattle was projected at 1,373.  As it was just getting underway, Sound Transit cut its goal in half, to 678.  Daily ridership is less than one-quarter the first year's operation, projected when the project was selected and justified.   And it's less than one-half what Sound Transit predicted just a few months ago.  If one of Sound Transit's predictions is changed--and ridership is set at sixty percent of that forecast by Sound Transit--the per-trip costs are $298 this year, and $70 in 2025.  The subsidy for each round-trip commuter, over the 22 years 2004-2025, will exceed $1 million.  It's likely to be worse than that.

Calculating the Cost and Cost Effectiveness of Sounder Commuter Rail,
Everett to Seattle, 2004 - 2025

The present per-trip operating costs for Sounder Everett/Seattle is $38.64, as The Seattle Times recently reported and Sound Transit conceded.  But to illustrate how deficient it is to use only Operating & Maintenance costs to measure the cost of a trip, consider that this year's O&M is budgeted at $4.065 million--while the capital development cost of Sounder Everett/Seattle is 97-times greater, $393 million.  If the huge, primary cost is ignored--as though it didn't exist--the true per-trip cost is not even approached.  The public pays all the bills, not just the annual operating costs.

To fail to account for the capital development cost is like pretending the "cost" of one's housing is merely the annual cost of utilities, insurance, and maintenance--and failing to include rent, or mortgage payments and down payments, which are the predominant costs.  The capital cost of the house is its primary cost, many multiples of the annual operating cost.  The same is true for Sounder commuter rail.

Accountants, economists, and financial experts are well-versed in proper methods to calculate the true cost of a new project, over its lifecycle.  Business people, banks, and investors rely on the efficacy of such exercises.  That's how they measure whether an investment is worth the money.  There are projected costs of operating and maintaining the facility or service, and projections for the capital cost to develop it.  The former can be projected on a yearly basis, but the latter must be "annualized."  That is, the costs are properly distributed over the useful life of the train, building, roadway, whatever.

The Federal Transit Administration has an official methodology for calculating the cost effectiveness of a new transportation system.  It expresses the result in terms of Cost Per New-Rider.  Or, more precisely, "Incremental Cost per Incremental Passenger in Forecast Year"--which typically is twenty years after operation begins.

Relying solely on Sound Transit's official data--even though those data are unreliable -- we have applied the FTA methodology.  The capital investment costs have been annualized in the most conservative fashion.  Sound Transit's annual O&M and ridership projections are taken at face value.  Every Sounder patron is treated as though he or she is a new transit patron.  With those caveats, each boarding in 2004 will cost $179.  In 2025, $42.  The average cost of each boarding over the 22 years, 2004-2025, will be $56.  The average cost of a daily commuter, 2004-2025, $640,640.  The average fare over 22 years, $28,000.

Here are the myriad ways the assumptions underlying the costs, and accompanying calculations, are exceedingly generous to Sound Transit:

1.  Ridership numbers for Sounder, Everett/Seattle, aren't even holding up for the first year.  Sound Transit admits, "No, we're not going to hit it."  And well into its fourth year of operation, Sounder, between Tacoma and Seattle, has scarcely one-third the ridership goal projected for 2004, 3,128 of the projected 9,318 daily boardings.  (Sound Transit, official ridership numbers for the week of March 28-April 2, 2004, and c/f Sound Transit ridership projections for Sounder, Tacoma/Seattle, 2001 Financial Plan)  At its rate of growth over the past three-plus years, Sounder Tacoma/Seattle won't reach 50 percent of its "worst case" ridership target in 2010, 10,200, much less its goal of 11,000.

2.  The current O&M cost of a Sounder train running between Tacoma and Seattle, 30+ miles, is approximately $4.5 million.  Sound Transit predicted that the 2004 O&M cost of one train running 30+ miles between Everett and Seattle would be $3.274 million, with no explanation why it would be so much cheaper. (Sound Transit, 2003 Financial Plan)  Indeed, before the end of the first quarter, projected O&M for 2004 was increased by 24 percent, to $4.065 million.  But Sound Transit's O&M predictions for the future remain unchanged, riddled with obvious understatements.  For example, in 2025 it claims it will run four daily trains between Everett and Seattle, and serve nearly 3,400 average daily boardings--with an O&M budget of merely $9.24 million. That sum, at best, is sufficient to operate but two trains.  This year, the O&M budget for three trains operating between Tacoma and Seattle, a similar distance, is in excess of $14.5 million.  When Sound Transit operated two daily trains between Tacoma and Seattle, the annual O&M cost exceeded $10 million.

3.  There are proper procedures for "annualizing" capital investments, spelled out by FTA's methodology.  Many questions arose as to which costs should be assigned a 100-year annualization--producing the lowest cost, and which should be assigned 30 years, or 25.  Unless the reviewers knew, without question, that a cost should be assigned a shorter period, it was assigned 100-years, the most generous.  Accordingly, 76 percent of the total costs were assigned to the 100-year category.  The only exceptions were Structures (the stations), 30 years, and Rail vehicles, 25 years.

4.  The latest, agreed capital cost of Sounder, Everett to Seattle, is $393 million, according to press reports.  But the latest Sound Transit budget we had on hand contained only $377 million of the costs.  So we used the lower figure, $377 million

5.  Sound Transit's contract with BNSF identifies other liabilities Sound Transit has accepted, for environmental permitting, off-site environmental mitigation, construction of bridges and trestles to replace culverts, and to reimburse BNSF for certain taxes.  It's not known whether, or to what degree, this will further increase the capital costs--but the fact of Sound Transit's additional liabilities is established.  There's additional risk, beyond $393 million.

6.  There is a very large reduction in value, and increase in per-trip cost, that we did not include at all, for lack of sufficient data.  The FTA cost effectiveness methodology is geared to new transit patrons.  That is, it is not a benefit to create another transit system merely to serve riders that already are being served--if the new rides are more costly than before.  All Central Puget Sound transit patrons switching from buses to trains will result in markedly increased costs to serve them.  In FTA's formulation, the cost of the system is spread only among "new riders."   To fully estimate FTA's "Incremental Cost per Incremental Passenger" requires data on how many Sounder patrons are switching from buses to trains, and how much money, if any, is being saved by reduced or eliminated bus operations.  Since we lacked the necessary data, every boarding of Sounder was treated as an incremental contribution to the region's transit patronage.  Even though some share--30 percent?  50 percent?  70 percent?--is not.  Had all the required data been applied, the per-trip cost would have been substantially higher, and the cost-effectiveness of Sounder further downgraded.

A More Realistic Calculation of the Per-Ride Cost
of Sounder Commuter Rail, Everett to Seattle

To get closer to a reasonable calculation of the per-trip cost, and cost-effectiveness of Sounder commuter rail, Everett to Seattle, one sole variable was changed:  ridership.  For the reasons detailed above--including actual ridership in the first months of operation, actual ridership experiences in the corridor between Tacoma and Seattle, and the contrast between Sound Transit's predictions and what is occurring--we reduced predicted ridership to 60 percent the official projection.  We left every other variable the same, including capital costs, generous annualization of those costs, O&M costs, even though they are dubious, and we treated every Sounder patron as though s/he is a new transit user.

With those conservative caveats, the per-trip costs are $298 per boarding in 2004, $70 per boarding in 2025, and $93 per boarding for the entire period, 2004-2025.

A Summary of Sound Transit's Promises vs. Performance, to Date

In Sound Move: The Ten-Year Regional Transit System Plan, approved by voters in 1996, the capital cost of 82-mile Sounder was $650 million.  (Sound Move, page A-1, $539 million--but that was in 1995 dollars.  The conversion factor to contemporary dollars is 1.2.)  The current capital cost projected by Sound Transit is $1,200,000,000--a $550 million, 85 percent cost overrun in constant dollars.  It may go higher.

In Sound Move, the ridership projection in 2010, a benchmark year, was 13,550.  Sounder Tacoma/Seattle (including Lakewood) was to serve 11,000 average daily trips, and Sounder Everett/Seattle, 2,550.  The "worst case" was 12,600--10,200 on the Tacoma/Seattle segment, and 2,400 for Everett/Seattle.  (Sound Move, C-19)  To reach its goals, Sound Transit projected 9,318 for the Tacoma/Seattle segment in 2004 (currently, it's 3,128, 33 percent), and 2,158 for Sounder Everett/Seattle (currently 15 percent).  However, Sounder Everett/Seattle has been two years late getting underway.  Its first full year of service was supposed to be 2002, when it was projected to average 1,373 daily boardings.   A few months ago, Sound Transit cut the first year's goal in half, from 1,373 to 678.  Daily ridership is averaging about 315.  So it's about 15 percent of the initial projection for 2004, 23 percent of the initial projection for the first full year of service, and 46 percent of the lowered level that Sound Transit recently projected for 2004.  In short, to accept 60 percent of Sound Transit's projected ridership in the future is generous, given what is known at present.

Furthermore, Sounder Tacoma/Seattle is in its fourth year of operation, so it can be assessed for the relationship between Sound Transit's projections and performance.    In its first full year of operation, 2001, Sounder Tacoma/Seattle met 58 percent of its ridership goal.  In the second year, 47 percent.  In 2003, 33 percent.  Currently it serves one-third its daily goal for 2004.  Its increase in daily ridership is between 400 and 300 average daily boardings per year, and falling.  If there are 350 boardings added each year, there will be 5,350 daily boardings in 2010.  That will be 49 percent of the goal, and 52 percent of Sound Transit's "worst case" projection.

It is reasonable, nay conservative, to calculate that Sounder Everett/Seattle, will reach but 60 percent of its ridership goal. The record being compiled by Sounder Tacoma/Seattle is:  In 2001, average daily boardings for 2001 were 2,164.  In 2002, 2,587, a gain of 423 average daily boardings.  In 2003, 2,889, a gain of 302 average daily boardings--with one additional daily train in service.  For the first quarter of 2004, compared with the first quarter of 2003, the gain was merely 219 average daily boardings.  Sounder is experiencing diminishing returns--more riders, but with slower and slower rates of growth, year-to-year.  Hence, a projection of annual increases of 350 average daily boardings, 2004-2010, is generous.  At that, Sound Transit would fall short of 50 percent of its goal for Sounder Tacoma/Seattle.

As recently as its 2001 Financial Plan, Sound Transit projected Sounder's O&M for 2004 at $28,556,000.  Now, it's projected at $18,594,000--$14,529,000 for Tacoma/Seattle, and $4,065,000 for Everett/Seattle.  The good news is that's merely 65 percent as much as formerly projected.  The bad news is, the 2004 O&M projected in the 2001 Financial Plan was for 15 daily trains (nine Tacoma-to-Seattle, and six Everett-to-Seattle).  And the $18,594,000 is for only four trains (three Tacoma-to-Seattle and one Everett-to-Seattle).  O&M was projected at $1.9 million per train per year, and now it's $4.65 million--a $2.75 million per-train, 145 percent cost overrun.

In sum, Sounder is way over-budget on the capital side, way over-budget on the operating side, way behind schedule, and way under-performing on ridership goals, which means farebox revenue will be far short, too.  Why is it so bad?

"We're doing the best we can with the hand we've been dealt"

Sound Transit casts itself in the role of victim.  But the disaster--which will plague this region into future generations--is entirely of its own making.  It's the citizens, taxpayers, transit patrons, and travelers who are the victims.  Sound Transit is the culprit.  If the agency had told voters the truth in 1996, they never would have approved Sound Move:  The Ten-Year Regional Transit System Plan.  And if it had told the state legislature the truth, it would have been referred to the applicable statute, and told that Sounder commuter rail was not qualified to proceed, as it was not "a reasonable alternative" with costs equal to or less than a bus, as the law required. 

Sound Transit committed what the scholars call "strategic misrepresentation" to get the public's money.  It understated Sounder's costs.  It overstated Sounder's benefits.  And it rigged the cost comparisons with buses.  Now the public is stuck with the bills, while Sound Transit board members and managers lavish the money on appreciative vendors, contractors, union labor, and themselves.  The board members, all elected politicians save one, solicit favors in return--support for their election campaigns.  The vendors comply.  The information can be found in campaign contributions reports, and the election activities of the benefited trade associations and unions.

The scholarly literature reports that urban rail projects are the most misrepresented of all transportation infrastructure projects.  The reason is simple:  Often they are terrible investments, and wouldn't be approved if sponsors told the truth.  "Cost underestimation cannot be explained by error, and seems to be best explained by strategic misrepresentation, i.e. lying."  (Bent Flyvbjerg et al, "Underestimating Costs in Public Works Projects: Error or Lie?," American Planning Association Journal [Summer 2002], and Don H. Pickrell, "A Desire Named Streetcar: Fantasy and Fact in Rail Transit Planning," American Planning Association Journal [Spring 1992].) 

Lying about projected costs is typically exceeded by lying about ridership--that is, projected ridership is overstated, on average, to a degree even greater than costs are understated.  (Flyvbjerg et al, Megaprojects and Risk:  An Anatomy of Ambition, [Cambridge University Press: 2003], pages 25-26, 37, and 44-45).  Sound Transit already admits a projected cost of Sounder of $1.2 billion, an 85 percent overrun.  The prospect of reaching less than half the ridership targets--a short-fall of 100 percent--is growing by the year.

In 1990-91, anticipating the possible development of commuter rail, and concerned about excessive costs or inadequate benefits, the state legislature passed a statute with the following requirement:

*RCW 81.104.120 Commuter rail service--Voter approval.  (1) Transit agencies and regional transit authorities may operate or contract for commuter rail service where it is deemed to be a reasonable alternative transit mode. A reasonable alternative is one whose passenger costs per mile, including costs of trackage, equipment, maintenance, operations, and administration are equal to or less than comparable bus, entrained bus, trolley, or personal rapid transit systems.

The only way Sounder could qualify under the statute was for the agency to "deem" it a "reasonable alternative."  And the only way it could do that was to apply phony cost and ridership numbers.  So it did.  But even having done so, it did not have to proceed, and sign the contracts with Burlington Northern Sante Fe.  Truer costs and ridership performance were manifest prior to that signing.  When Sound Transit inked its deal with BNSF for Sounder Everett/Seattle in 2003--and crowed about it at a press conference--it knew the capital cost was 3.7 times higher than the figure it represented to voters. It knew that Sounder Tacoma/Seattle, in its fourth year of operation, was reaching merely one-third of its ridership targets.  It knew that its ridership projections between Everett and Seattle were exaggerated--it even moved to cut the first year's prediction in half.  There was no question its original representations were bogus, both costs and benefits.  It still had the opportunity to act responsibly, honor the law, and its responsibilities to the public, and back out.  It didn't have to sign. 

Instead, the board and management of Sound Transit knowingly committed the region to a massively wasteful, terribly misrepresented project.  It did so even in the face of a state statute that required it to deem the "passenger costs per mile equal to or less than comparable bus" service--which it isn't, by an enormous margin, and Sound Transit knew it.  Nobody "dealt" Sound Transit that hand--it sought it, with the utmost mendacity, and an utter lack of care or concern for the public trust or interest.

"Underestimating costs and overestimating benefits for a given project leads to a falsely high ratio between benefits and costs for that project, which in turn leads to two problems:  First, the project might be started despite the fact that it is not economically viable.  Or, second, it may be started instead of another project that would have shown itself to yield higher returns than the project started, had the actual costs and benefits of both projects been known.  Both cases result in the inefficient use of resources and, for public projects, in waste of taxpayers money."  (Flyvbjerg et al, Megaprojects and Risk, page 47.)

"[V]irtually every project this article reviews represented the largest investment in public works ever undertaken by the local area, often by a considerable margin... [L]ocal officials continue to choose--almost always in favor of rail lines--among alternative transit projects on the basis of narrow margins among their projected costs and ridership. A wide margin of forecasting error may signal analysts' complicity in demonstrating the purported technical superiority of projects that could not prevail in an unbiased evaluation, but are favored by influential local officials for other--often unspoken--reasons."  (Pickrell, "A Desire Named Streetcar," page 158-59,)

Return to the Public Interest Transportation Forum home page.

Last modified: February 07, 2011