The $3.1 billion in new federal grants announced today for the California high-speed rail project still leaves the California High Speed Rail Authority (CHSRA) at least $7 billion short of having enough money to complete the initial operating segment from Merced to Bakersfield.
The $3.1 billion announced today is actually two separate grants – $1.03 billion from money already on hand, and $2.04 billion from moneys to be released over the 2024-2026 period. In total, this brings all federal grants for this project up to $6.8 billion: $3.5 billion from the Obama Administration in 2010 and 2011, then a total of $3.3 billion in various grants from the Biden Administration over the last three years:
|
|
Federal |
|
|
Grant (Million $) |
FY09 |
ARRA stimulus grant |
2,552.6 |
FY10 |
Regular appropriation |
928.6 |
FY16 |
EPA Brownfields grant |
0.6 |
FY21 |
RAISE grant |
24.0 |
FY22 |
RAISE grant |
25.0 |
FY22 |
CRISI grant |
201.9 |
FY23 |
RAISE grant |
20.0 |
FY22-23 |
IIJA FSP grant |
1,031.2 |
FY24-26 |
IIJA FSP commitment |
2,042.4 |
Total Federal Commitment |
6,826.3 |
Per CHSRA’s 2023 project update, aside from federal grants, the project has the $8.5 billion in proceeds from the Proposition 1A bonds approved by voters in 2008 (which was supposed to to pay for an entire San Francisco to Anaheim system), plus $5.4 billion in proceeds from state cap-and-trade carbon permit auctions received by the end of 2022, plus an estimate of $6.0 to $7.7 billion in future cap-and-trade proceeds from 2023 through the current sunset of the program in 2030.
All told, that is between $26.7 billion and $28.4 billion of funding available for the project.
But the cost of the Merced-Bakersfield segment by itself has now grown to around $33 billion, and there is also $3.2 billion in work outside the segment that has either already been done or is irrevocably committed, for a total CHSRA financial need of as much as $35.2 billion.
(The “Central Valley Segment” listed below is what was originally supposed to be funded by the FY 2009 and 2010 federal grants, the 119-mile right-of-way and track from Madera in the north to an orchard in Shafter in the south. The FY 2009 and 2010 grants do not pay for any electric catenary or high-speed trainsets for the CVS, so the CVS does not qualify as an operable HSR segment on its own.)
Estimated YOE Costs (Mil. $) |
30% |
50% |
65% |
|
Confidence |
Confidence |
Confidence |
Central Valley Segment (Madera-Shafter) |
17,311 |
17,806 |
18,262 |
34-mile Merced Extension |
3,780 |
4,140 |
4,462 |
18-mile Bakersfield Extension |
2,752 |
3,020 |
3,258 |
Stations |
1,044 |
1,147 |
1,237 |
Track, Systems, Utilities |
2,765 |
3,024 |
3,255 |
6 High-Speed Trainsets |
465 |
516 |
561 |
Maintenance Facility |
342 |
382 |
418 |
PD, Support, Contingencies |
1,376 |
1,463 |
1,523 |
TOTAL, Merced-Bakersfield |
29,835 |
31,497 |
32,976 |
|
|
|
|
Project Development bal. |
539 |
545 |
559 |
Program-Wide Support |
458 |
475 |
490 |
SF & LA Bookends |
1,298 |
1,298 |
1,298 |
|
|
|
|
TOTAL NEEDED |
32,127 |
33,815 |
35,323 |
That leaves CHSRA between $4.7 billion and $8.6 billion short of being able to pay for an operable HSR segment running from Merced to Bakersfield. The 50% cost confidence level, minus the average point of CHSRA’s range for total revenues, is $6.8 billion, which we round to $7 billion for headline purposes.
There’s only $3 billion remaining in the bipartisan infrastructure law’s Federal-State Partnership for Intercity Passenger Rail grant program for projects off the Northeast Corridor, so even if USDOT were to give California the entire amount (which they will not do), the project would still be at least $4 billion short. The project may receive some future aid from CRISI grants, Rail-Highway Crossing grants, and RAISE grants, but those tend to be much smaller grant amounts. (CHSRA had set an aspirational target of getting $8 billion in grants out of the bipartisan infrastructure law, but the total so far is $3.3 billion and not likely to grow much past $4 billion.)
By law, CHSRA gets 25 percent of the annual proceeds from the cap-and-trade permit auctions, but that program expires in 2030. CHSRA has long sought an extension of the cap-and-trade program (to 2050, if possible), as well as special authority from the legislature to issue bonds guaranteed by future cap-and-trade proceeds so they can borrow to build in the 2020s and 2030s and then pay back the bonds in the late 2030s and 2040s. But so far the legislature has been unwilling to go along. Without bonding, CHSRA has to wait for the $750 million to $1.0 billion in cap-and-trade proceeds to be deposited each year before spending the money, which (even if cap-and-trade is extended) would postpone completion of the Merced-Bakersfield segment from its currently scheduled 2030 start to around 2040.
But, even assuming that California can somehow find another $7+ billion to complete the Merced-Bakersfield section, that gets you…from Merced to Bakersfield, and back, at 200+ miles per hour. A godsend for those who live in that area, but that’s not that many people, and few of them were clamoring for a bullet train just to get from Fresno to Bakersfield and back. The project is doomed to white elephant status with extensive annual operating subsidy costs unless extended to a higher population area, either in the north or south or (as originally intended), in both directions.
If California politicians choose to continue the project, they can turn northwest towards the Bay Area, which the 2023 project update says will cost between $21 billion and $36 billion (though there is a lot of tunneling though the Diablo range with the possibility of enormous cost overruns there). If they turn south to L.A. and Anaheim, that’s an additional $32 billion to $53 billion (but again, massive tunneling through the San Gabriels, which the U.S. does not do at all well compared to other nations).
All in all, including systemwide costs, CHSRA now says that the total cost of the promised Phase 1 system (San Francisco to Anaheim) is between $89 billion and $128 billion. If you can assume that California can somehow find money to pay for the remaining costs of Merced-Bakersfield listed above, they would then have to find an additional $56 billion to $93 billion to complete Phase 1 (and, if history has taught us anything, the costs will be closer to the high estimates than the low estimates).
|
30% |
50% |
65% |
|
Confidence |
Confidence |
Confidence |
NoCal Preliminary Design |
213 |
213 |
213 |
SF “Bookend” Investments |
798 |
798 |
798 |
San Francisco to San Jose |
3,936 |
4,967 |
6,407 |
San Jose to Gilroy |
4,075 |
6,020 |
8,733 |
Gilroy to Carlucci Road |
10,316 |
13,627 |
16,762 |
Wye to Madera Balance |
1,842 |
2,240 |
2,601 |
Subtotal, SF to Central Valley |
21,180 |
27,865 |
35,514 |
|
|
|
|
Merced to Bakersfield |
29,835 |
31,497 |
32,976 |
|
|
|
|
Bakersfield to Palmdale |
13,712 |
17,140 |
20,740 |
Palmdale to Burbank |
12,635 |
16,775 |
24,428 |
Burbank to Los Angeles |
2,201 |
2,935 |
3,405 |
Los Angeles to Anaheim |
2,478 |
2,918 |
3,352 |
LA “Bookend” Investments |
500 |
500 |
500 |
SoCal Preliminary Design |
382 |
382 |
382 |
Subtotal, Central Valley to Anaheim |
31,908 |
40,650 |
52,807 |
|
|
|
|
Project Development & Support |
1,049 |
1,049 |
1,049 |
Heavy Maintenance Facility (Balance) |
248 |
275 |
301 |
Trainsets (Balance) |
4,161 |
4,643 |
5,084 |
Solar Power Generation (Balance) |
166 |
184 |
202 |
Subtotal, Program-Wide Capital Costs |
5,624 |
6,151 |
6,636 |
|
|
|
|
TOTAL PHASE 1 CAPITAL COST |
88,545 |
106,163 |
127,933 |
Minus Merced-Bakersfield & Already Spent |
-32,127 |
-33,815 |
-35,323 |
FUTURE FUNDING NEEDED |
56,418 |
72,348 |
92,610 |
To put this in historical perspective, when Golden State voters approved Proposition 1A in 2008, they had been presented with CHSRA’s 2008 Business Plan, which presented a Phase 1 cost estimate of $33 billion, but those were constant dollars which, while useful for engineers, don’t translate into the amount of money actually required to build something in the real world, which are year-of-expenditure dollars.
It was CHSRA’s 2009 report to the legislature (now scrubbed from their website, but still preserved on archive.org) that first converted the $33 billion constant dollar Phase 1 cost estimate to year-of-expenditure dollars – $42.6 billion in YOE dollars. The current 50% confidence base case estimate is 2.5 times that amount, and the (more realistic) 65% confidence high end case is an even 3.0 times the original cost estimate.
And, to give a picture of the blithe self-confidence with which CHSRA of that era planned its finances, here is their original plan for paying for the complete Phase 1 system, from page 93 of the 2007 report to the legislature: