The Tacoma News Tribune's August 23, 2003 editorial, "Tacoma Link's true potential lies in the future" reports, "As mass transit projects go, Tacoma Link's costs have not been scandalous - nothing like the $1 billion in overruns racked up by Central Link in Seattle."
Yes, the cost overrun for Tacoma Link is merely 34 percent which, by Sound Transit's standards, is an exceptional accomplishment.
However, the acknowledged cost overruns for Central Link total $2.74 billion, so far, announced in stages as described in the following table:
|Central Link Light Rail Cost Estimate Increases, 1996-2001|
|Projected cost increase revealed through 1999||$300 million|
|November/December 2000 increase announced||$1 billion|
|January 2001 increase disclosed||$560 million|
|October 2001 revealed increase for Initial Segment only||$880 million|
|Total cost increases through October 2001||$2.7 billion|
One would think that the editorial board of the News Tribune would have noticed. $1 billion is merely the specific overrun acknowledged in December 2000, which was preceded and followed by other overruns. Further, it's certain to grow far beyond $2.74 billion.
In September 2001, Deloitte & Touche issued its performance review of Sound Transit's cost estimating methods. In summary, it found the following shortcomings:
|"Development of estimates to match a budget|
|Overly optimistic estimates|
|Contingencies prematurely reduced|
|Inadequate/insufficient data (e.g. no soils data, ROW based on
|Inadequate soft costs...."|
Deloitte & Touche further reported:
"[T]he estimates were developed with too much optimism for the best case scenario to occur in all cases.... [E]stimates were prematurely reduced at the direction of Sound Transit management.... Also, the Agency estimates were lowered in 1999 based upon overly optimistic staffing plans. In addition, the Project schedule was based upon a completion date in the fall of 2006 but this was also discovered to be overly optimistic and aggressive."
The only segment that has been competently estimated, to date, is Initial Segment -- which is why its cost burgeoned $880 million in 2001. Whereas Initial Segment might be completed within its current budget (though Deloitte & Touche made a list of warnings of possible additional overruns), the remaining segments have not yet been competently and honestly cost-estimated. As the D&T team said above in the first bullet -- "development of estimates to match a budget" -- Sound Transit contrived costs to reflect the budget it desired, rather than make the budget fit the costs.
Cost History in Detail
The starting point is what Sound Transit represented to voters in 1996: $1.7 billion (in 1995 dollars), equal to about $2.3 billion in year-of-expenditure dollars ($YOE). There had been a nearly $300 million increase, prior to November 2000, when the agency admitted that its cheapest tunnel bid was $300 million over-budget, and then, in December 2000, it admitted some more stuff, $700 million worth, so the total overrun then was $1.3 billion, going from $2.3 to $3.6 billion.
At the January 2001 board meeting, King County Councilman and former Sound Transit Board Member Rob McKenna, with brilliant questioning, obtained the concession of Sound Transit executives Joni Earl and Hugh Simpson that the actual projected cost at that time was $4.16 billion, not $3.6 billion. The latter figure failed to include substantial costs. The project the USDOT Inspector General reviewed, in his April 4, 2001 preliminary report, was a $4.16 billion project, not a $3.6 billion project. It was then $1.86 billion over budget.
From there it gets harder, as Sound Transit stopped providing estimates for 21-mile Central Link. It gives numbers only for Initial Segment, which is neither University Link (the 7.2 miles from NE 45th to South Lander, which was to be the first segment, aided by the original $500 million full funding grant agreement), nor Airport Link (the 14 mile balance, from South Lander to South 200th).
Here, then, are the basic numbers needed to follow the story:
Central Link Light Rail As Approved by Voters in 1996
|University Link||$1.4 billion|
|Airport Link||$900 million|
Central Link During 1999-2000 in the First Federal Grant Application
|Airport Link||$1 billion|
Central Link in December 2000, After Tunnel Cost Expansion Revealed
|University Link||$2.3 billion|
|Airport Link||$1.3 billion|
Central Link, January 2001, For the Revised Federal Grant Applications
|University Link||$2.6 billion|
A Portion of Central Link, Summer, 2003
|Initial Segment||$2.44 billion|
Note that the 14-mile "Initial Segment," the starter rail to Central Link light rail, which was supposed to be the starter rail to the 125-mile regional system the News Tribune editors are counting on, will cost more at $2.44 billion than the entire 21-mile Central Link project was supposed to cost at $2.3 billion.
Today, the $2.44 billion Initial Segment number is understated, for several reasons. But accepting it, so as not to get bogged down, $2.44 billion holds profound implications for the current and ultimate price tag, because Initial Segment overlaps Airport Link considerably. Nearly 12 of the 14 miles of Initial Segment are identical to Airport Link--South Lander Street to South 154th Street. The portion that is part of Airport Link only--the approximately two miles from South 154th to the terminus at South 200th, including the stop at SeaTac Airport--is a considerably more-costly stretch than the portion that is part of Initial Segment only. That's the two miles from South Lander to Convention Place Station. The latter is all in public ownership, nearly half of it is the exclusive, level busway heading south from downtown, and the balance is the downtown Seattle transit tunnel, which is largely paid for, including the stations. Hence Initial Segment (now at $2.44 billion) should cost no more than Airport Link (last at $1.56 billion), and probably less.
Future Cost Escalations
Sound Transit has not announced any further cost estimates. They are working on them as of summer 2003. What follows is an independent estimate for completing Central Link Light Rail to the 21 miles promised in 1996, University District to S 200th below Sea Tac Airport.
Airport Link originally defined as S. Lander to S. 200th was always less than 40 percent the cost of the entire project. Assuming the Initial Segment to be a close surrogate to Airport Link as described above, using today's admitted cost for Initial Segment, $2.44 billion, and presuming that it's 40 percent of the cost of Central Link, then the full cost of Central Link computes to be $6.1 billion, which equals $2.44 for Initial Segment (40%) plus $3.44 billion (60%) for what's left to build. Compared to the 1996 promise of $2.3 billion, this $6.1 billion represents a cost escalation since 1996 of $3.8 billion in year-of-expenditure dollars. Subtracting the total cost overrun to date of $2.7 billion reveals the estimated additional cost overrun of $1.1 billion that we are now expecting on top of the $2.7 billion.
However, there are several compelling reasons why even this level of projected overrun is too optimistic:
1. Sound Transit has admitted that having a terminus in the University District is not viable, despite the plan through early 2001 to terminate there. Now, Sound Transit has its sights set on building all the way to Northgate in its northern extension phase. University Link has become North Link. While the cost of a fully grade-separated line to Northgate will be a billion dollars or more, we estimate $300 million as the additional cost of taking the line out to Ravenna, the minimum requirement for meeting 1996 commitment to get to University District, for the purposes of this analysis.
2. Years of inflation will be added to the project. If there was merely $3 billion more required for the project, which is very conservative, annual inflation at 3 percent would add $90 million in the first year, compounded in each successive year. If the mid-point of the project to continue to SeaTac and to North Seattle is 2011 (start in 2008, complete in 2014), another euphoric assumption, inflation would add nearly $500 million to the cost.
3. The costs of the debt will be vastly greater. Central Link was supposed to be built with prudent borrowing and modest debt service costs, on the order of $75 million, according to Sound Move. (We're speaking here of debt during the capital development phase, not long-term debt, after operations begin.) Taking Sound Transit's current figure, Initial Segment alone, the cheap, easy, first segment, will cost $202 million for debt service. But this segment will begin with at least $600 million in the bank, and a fair portion of it will be paid from current tax revenues, not bond revenues. If the rest of Central Link is to proceed, it will start with North King County 100 percent in debt, and with most of its annual budget already pledged to debt costs and operating subsidies. So a much higher portion of the post-Initial Segment project will have to be debt-financed -- and the ability to finance debt will be severely circumscribed by the debts and subsidies already incurred. We estimate $400 million debt financing cost beyond the $300 million that we estimate by extrapolating from the Initial Segment debt service.
In summary, our $5 billion projected cost overrun for Central Link is a conservative estimate made up as follows:
|Total Estimated Cost Overrun for 21 Mile Central Link Light Rail|
|Through October 2001||$2.7 billion|
|Estimated overrun extending Initial Segment north to U Dist, south to S 200th||$1.1 billion|
|Northern terminus at Ravenna||$300 million|
|Additional cost of debt.||$400 million|
|Total Estimated Cost Overrun||$5 billion|
If Sound Transit knew how it was going to cover an overrun this big, it would roll out its new regional light rail plan, with updated cost estimates, a plan of how to finance the project, and a schedule. It doesn't do so because it is all a fantasy, a fantasy originally founded in phony cost estimates, and voter credulity.
The problem Sound Transit faces now is intractable: Reality.
Return to the Public Interest Transportation Forum home page.
Last modified: February 07, 2011