January 18, 2018

The estimated cost of the initial Fresno to Bakersfield segment of the California high-speed rail system has soared another $2.8 billion, to $10.6 billion. The original estimate was just under $6 billion and the revised estimate in the system’s 2016 Business Plan was $7.8 billion. The increase in cost was actually over $3 billion, but the “contingencies” overrun line item in the budget was reduced by $323 million in the new estimate, offsetting the gross cost increase a tad.

(The segment actually runs from Madera, 25 miles north of Fresno, to an orchard in Shafter, just north of Bakersfield, but that won’t fit neatly into a headline.)

The Board of Directors of the California High Speed Rail Authority (CHSRA) was given the bad news at their monthly meeting on January 16. A consultant with the lead contractor a two-page document with a broad overview of the revised cost estimate. As usual, Ralph Vartabedian with the Los Angeles Times has the best coverage of how the meeting went – see here.

More details will be available when the Board considers a draft 2018 Business Plan next month. At that point, the Board will also have to discuss where to find another $2.8 billion in funding – do they ask the state legislature to drain most of the remainder of the $9 billion in bonds authorized by voters in Proposition 1A in 2008 (back when $9 billion in state funding was supposed to be enough to pay for 50 percent of the construction cost of the entire system from San Francisco to San Diego), or do they ask for more appropriations of cap-and-trade revenues, or something else?

At the original $6 billion cost estimate for the segment, funding was to come roughly 50-50 from state Prop 1A bonds and federal appropriations (which were appropriated in fiscal 2009 and 2010, before Congress got out of the high-speed rail business in 2011). When the cost estimate went up to $7.8 billion, the extra money came from state cap-and-trade revenues, so the federal cost share dropped to 38 percent. At a $10.6 billion cost estimate, the federal cost share will drop to around 28 percent.

(If the final cost eventually jumps to around $14.8 billion, the federal cost share will drop to 20 percent, the same level as all of those “incentive grants” in the forthcoming Trump Administration infrastructure package.)

The $2.8 billion in new overruns comes mostly from increased costs of right-of-way acquisition and utility relocation, as well as $315 million for building physical barriers to separate high-speed passenger tracks from adjoining freight train tracks. The biggest jump is in the northernmost subsection, where the cost has risen from the earlier estimate of $2.0 billion to $3.4 billion, but that may only be because that subsection was started first and is two years farther along than the southernmost subsection.

Eagle-eyed readers will have noticed the $0.000 for “vehicles” in the cost estimate. This is because the initial segment won’t actually have any high-speed trains. (Those won’t be purchased until the extension to San Jose is built, for the first operable segment.) The San Jose to Shafter operable segment, dubbed “Silicon Valley to Central Valley,” had an estimated capital cost of $20.7 billion in the 2016 Business Plan, and CHSRA planned to pay for it with the remainder of the $9 billion in Prop 1A bonds as well as issuing new bonds secured by future decades of dedicated cap-and-trade revenues.

Completing Phase I (which would extend the system south to downtown Los Angeles and also from San Jose into downtown San Francisco) was budgeted at an additional $41.5 billion in the 2016 Business Plan, but the cost of digging 15 miles of tunnels under the San Gabriel Mountains is expected to add to that greatly. CHSRA currently has no plan to finance the remainder of Phase I.

(Ed. Note: At this point, let me get on my soapbox yet again to point out that it is a terrible, horrible, no good, very bad idea for the federal government to give construction money for a new rail system before any of the environmental clearance work has begun, or before the route is selected, or before a good cost estimate is in hand, or before the state or local partners have their end of the financing nailed down, or for anything less than a minimum operable segment (MOS). The Federal Railroad Administration did all of that with the California project in 2010. If Congress had given the $10 billion in 2009-2010 high-speed rail appropriations to the Federal Transit Administration, instead of FRA, we wouldn’t be in this situation. Because whenever the FTA is financing a new subway or light rail system, they don’t give any construction money away for anything less than a MOS for which route selection and cost estimates and local revenue sources are set in stone and for which most of the permitting is done. If the federal government ever gets back in the high-speed-rail financing business, funding should only be given out through something akin to the FTA’s full funding grant agreement process, and never for anything less than a MOS.)

(Further Ed. Note: $7.25 billion for professional services. Damn.)