Sunday, June 1, 2008

High Speed Rail isn't a Toy

NOTE: We've moved! Visit us at the California High Speed Rail Blog.

A growing phenomenon in media commentary on high speed rail seems to be an embrace of the necessity and value of the system, but an unwillingness to part with 20th century dogma on taxes and government spending in figuring out how to pay for it. We saw that with Bruce Reed's HSR op-ed last week, and we see it today with Thomas Elias' op-ed in the Long Beach Press-Telegram. Elias opens by recounting fast, efficient, and comfortable trips in France on the TGV, and then examines the California plan:

The question: In a day when the state may run a $15 billion deficit or more and when Gov. Arnold Schwarzenegger has tried for across-the-board 10 percent budget cuts, how can we afford a massive new toy? Many equate the high-speed train idea to a family that wants to buy a new Ferrari when it can only afford macaroni and cheese for dinner.

But high speed trains may be more than a mere luxury. They can make business travel between urban centers easier and more comfortable, without many of the restrictions and complications terror fears have brought to air travel. They can also restore California's aura of leading the way toward a better lifestyle for all Americans. The trains would be instant tourist attractions, with reservations booked months in advance.

Elias would be a stronger advocate of HSR if he more strongly discounted the "HSR is an expensive toy" claim - in fact, the price of gas is turning ALL cars into Ferraris, at least in terms of cost, and HSR is the affordable mac-and-cheese alternative. Yes, HSR is more comfortable, but it won't be just for tourists and businessmen - it will be the most affordable way for average Californians to visit mom and dad at Christmas, to go to their child's graduation in June, to get to work in the morning. All HSR advocates need to internalize this thinking - our cause is MUCH stronger if we show people that HSR is the affordable workaday transportation solution for the 21st century, instead of taking a "gee whiz" attitude that was outdated even in the 1970s.

That quibble aside, Elias' op-ed is primarily concerned with funding:

These realities suggest the financing plan now proposed, with a conventional bond to be paid off by all state taxpayers, might not be the best way to finance a massive project like high speed rail.

Why should all taxpayers pay for a toy that will be used only by a few? Why should taxes from people in Redding or Chico be used for a rail system that will never approach those cities, even as it serves the likes of Bakersfield, Fresno, Madera and Stockton?

The plain answer is that residents there should not be taxed for this. Nor should poor Californians in Los Angeles, the San Joaquin Valley or the San Francisco Bay area who will rarely if ever ride these trains. For it is reality that - just as in Europe or Japan - fares on high-speed trains would be considerably higher than on conventional ones.

So unlike dams or highways or public hospitals or sewers, high speed rail should be built neither with general obligation bonds, as now proposed, nor with general fund revenues on a pay-as-you-go basis.

Rather, revenue bonds are the answer. Many an American stadium, toll road, bridge, airport terminal and short-cut tunnel has been built this way, with borrowed money that is eventually paid back by users of the project.

In short, since high-speed rail is not as essential as freeways or ordinary passenger and freight trains, why finance it the same way? Just as air fares are now taxed to pay for increased airport security, high speed rail tickets could be priced to meet bond payments. Just as hotel guests are often taxed to pay for improvements and services intended for tourist use, why not make high speed train riders pay in full for the service they are enjoying?

The framing here is absolutely atrocious. HSR isn't as essential as freeways? Or ordinary passenger trains?! Most Californians will never use it?! While I welcome Elias' support of HSR his assumptions are completely wrong. Freeways are the true luxury - they are an astronomical cost in subsidies, pollution, global warming, congestion, and fuel. Elias wants us to believe that this is "reality" but instead it's a 1970s worldview passed off as fact. Ordinary Californians, including those in the working-classes, use intrastate air travel rather frequently, and will also use HSR frequently - especially as they won't be able to afford air travel or long-distance road travel for much longer.

The reality is that HSR is exactly like a dam, a highway, a public hospital, or a sewer. It will be an essential piece of public infrastructure without which society will have a hard time functioning. Just as the State Water Project in the 1950s enabled the state's economic prosperity of the late 20th century, so too will HSR enable growth and prosperity in the 21st century. Without it the state will be dependent on oil-based transportation that the rest of the world is already abandoning - leaving California uncompetitive and making it difficult to get around the state.

Elias' awful framing aside, I'm not opposed in theory to using a revenue bond to pay for HSR. I have no doubt that HSR will attract enough riders to pay back the bonds. One advantage of a general obligation bond, however, is that it makes it more likely to attract the private investment that state politicians are demanding before they sign off on HSR. Investors see a general obligation bond as a safer risk than a revenue bond, which is especially important against the backdrop of a global credit crunch.

And despite my criticisms of the way Elias is selling HSR, as with Bruce Reed's op-ed I am again pleased that we seem to be moving away from a debate over whether HSR is necessary to one of how we should pay for it. It's a sign of real progress.


Rafael said...

Old habit die hard. Many Californians have become so accustomed to driving and flying cannot really see themselves hopping onto a train - any train. To their mind, that mode of transportation belongs to a different era and/or social stratum. That could prove a very expensive conceit.

In Europe, by contrast, intercity passenger trains are considered a perfectly normal means of transportation. For example, Austria has 8.2 million inhabitants - fewer than LA county. Yet in 2006, its state railways (OeBB) delivered 196 million passenger trips on rails, of which 29 million were long distance on 267 daily trains. OeBB owns all of the country's hydroelectric dams, so its electric trains consume no fossil fuel at all.

To put these numbers into context: the capital Vienna has ~1.6 million residents, with next largest - Gov. Schwarzenegger's native Graz - home to just ~250,000. Moreover, because of its demographics and alpine terrain, OeBB cannot even offer full-fat high speed trains - their fastest ones use tilting ICE-T rolling stock at 230kph (144 mph).

California, of course, is much more like Spain than Austria in terms of size, population and how it is distributed. In 2006 - i.e. even before the Madrid-Barcelona line was completed - a whopping 43% of all passenger-kilometers delivered were on the dedicated high speed (300-350kph) AVE network. Average seat capacity utilization was 65%, well above local and regional trains at 38%. In other words, operating profits from AVE are used to subsidize operations of the slower trains.

The story is much the same in France, Germany and other European countries: wherever intercity trains offer a fast, clean and convenient alternative to driving or flying, customers flock to them. Of course, prices do need to be competitive with the alternatives, i.e. driving and flying. Diesel now costs over $7.50/gallon in most of Europe, gasoline and kerosene even more.

Competing on price isn't particularly difficult: according to Deutsche Bahn, an ICE train with only 50% of seats occupied requires the equivalent of just 2 liters of gasoline per passenger and 100km. In the US, you'd say it achieves ~117mpg per passenger. Plus, electricity can be produced from any number of primary sources in power plants whose criteria emissions - if any - are easily managed.

Besides, every railway operator in Europe offers a wide range of discount fares. For example, an economy class eForfait for daily long-distance commuting from Reims to Paris (~94 miles) comes to EUR 423 ($656) per month for the first two years. Given the French working week and vacation time, that works out to ~220 annual trips each way, at an average cost of just $18 each. Plus, customers can choose to travel the same route on the weekend for pleasure at no extra charge. No frequent train traveler pays full fare in Europe, nor will they in California.

The eForfait compares very favorably with the total cost of ownership of a diesel commuter car, given that fuel cost alone comes to $12/trip - depreciation, registration, insurance, maintenance, motorway tolls and parking fees all come on top of that. Plus, commuting by train is faster, safer, less polluting and a lot less stressful.

Conclusion: modern passenger trains aren't toys at all. Rather, they are how Californians can avoid paying $4/gallon at the pump and, ever-larger fuel surcharges and other fees when booking an airline ticket.

The way things are going, where do you think fuel prices will be in 2018 - the earliest date California's HSR system could possibly be in operation?

Anonymous said...

In contrast to your view on the "framing" of the Elias column, I find the "framing" of the article perfect. ( It appeared in print in the local Palo Alto paper this AM)

At the heart is indeed the financing of this project, and whether the State should be obligating the taxpayers to fund a project that will be used by only a fraction of the population.

However, the notion that fare box collections could ever pay for this project is pure fantasy. We don't have the population density to ever begin to cover the capital and operating cost of the system.

I find it amusing the proposition that completion of this project will have any real effect on the price of gasoline. Even the most optimistic forecast by CHSRA, that HSR would take 6% of auto traffic on long hauls off the highways, the effect on gasoline prices would be negligible.

Air travel in the US always has been much more efficient and cost effective than in Europe. Train service in Europe has been the tradition and is where Europe put it resources, whereas in the US we concentrated on air and highways.

The future for auto travel I don't find bleak. Much more efficient autos have been appearing and are certainly on the horizon. The same is true for newer aircraft.

I am opposed to this project as formulated. If I am wrong and if private capital wants to take up the project and fund it as proposed in the Elias article, I would drop my opposition.

AB-3034 was supposed to mandate that 2/3 of the cost of the project would come from matching funds from Federal and private sources.

That has not happened. Those matching funds are nowhere to be seen. I even wonder how AB-3034 can become law, since it must pass the State senate by June 26th to qualify, and it is not on the agenda for the next three weeks. Strange it seems to me.

Rafael said...

@ Harold -

by your logic, every single road in California should be a toll road - because only part of the population will use it. Roads and road bridges are usually paid for out of the general fund, exceptions proving this rule. Did you think fuel excise taxes really cover the cost of maintaining all of that infrastructure?

Also, there is significant value in paying taxes that support obviously useful services, even if you yourself never use any of them. For example, HSR will reduce congestion at the airports and freeways taht you will continue to rely on. It will also help California meet its self-imposed AB32 targets, which would otherwise require similar contributions from other sources. Etc.

Your assertion that fare collections from HSR could never ever support ongoing operations appears to be based on your personal gut feeling - and frankly, not much else. HSR is not Amtrak, it is a completely different animal. It will have high-quality tracks dedicated to passenger service at very high speed. You must not extrapolate from Amtrak's financial performance, let alone that of passenger service prior to Amtrak.

California is roughly comparable in both size and population to Spain, where the AVE network does generate an operating profit that is used to subsidize regular-speed service. It's on-time performance is so good it can afford to offer customers a refund if their train is 5 minutes late. Good luck getting that guarantee from Amtrak or indeed, any airline. Why do you think railways all over the world are now building substantial HSR networks?

Finally, note that CHSRA actually does intend to fund the construction of the spurs to Sacramento and San Diego using revenue bonds. Once those are done, any operating revenue left over after obligations to third parties are met will go into the state's general fund. The November bond measure is only required to kick-start the process of getting the federal and private contributions needed to construct the trunk line.

You are right about one thing: HSR will not make gasoline any cheaper. No-one here ever claimed that it would. Rather, those individuals that choose to use HSR will reduce their annual mileage and hence, save on their annual transportation cost. Perhaps you think switching to a car that gets 35MPG is good enough, but even at $4/gallon, there are many who can no longer afford long-distance trips. At European prices, their mobility would be curtailed even further.

Wrt for the Senate, it may wait until the last moment to vote on AB3034 but it is not going to let it lapse. There is broad agreement that this proposition should now finally be put to voters in November. It should be theirs to approve or reject.

Anonymous said...


You write:

Finally, note that CHSRA actually does intend to fund the construction of the spurs to Sacramento and San Diego using revenue bonds. Once those are done, any operating revenue left over after obligations to third parties are met will go into the state's general fund. The November bond measure is only required to kick-start the process of getting the federal and private contributions needed to construct the trunk line."

This is news to me. Could you give me a link? How in the world does this jive with statements from Diridon and Kopp that this $10 billion bond measure is going to the the only funds from California voters needed to build the project?

On the Senate, I believe it will pass also. I think this kind of public process on a bill of such importance, deserves much more than an appearance at the last moment on the agenda. Lowenthal was supposed to have a comprehensive committee investigation in June, yet that appears to have disappeared. government at it worst.


Rafael said...

@ Harold -

in a recent half-hour interview with KCBS radio, CHSRA board member Rod Diridon stated clearly that:

(a) government financing would only be required for the SF-Anaheim starter line, which carries a price tag of ~$30 billion

(b) like HSR systems all over the world, he expects the starter line to generate an operating profit. In addition to the experience overseas (see my initial post), he bases this on two studies by Cambridge Systematics and Charles Ryder Associates that were commissioned by CHSRA.

IFF the system can attract 117 million passenger trips per year, annual operating profits could be as high as $1.3 billion on revenue of roughly $4 billion - at roughly $50 (in todays money) for a one-way ticket from SF to LA, far less than the price of an airline ticket (except if there's a brief price war).

The ridership study predicted around 95 million passengers for the whole system by 2030, based on cost increases (net of inflation) of 50% for both air and auto transportation from their year 2005 base case. Given the observed 25% increase in auto operations cost between 2000 and 2005 alone, that 95 million appears to be a conservative estimate.

Of course, the starter line by itself will not attract quite as many passengers, and certainly not immediately after it goes into service. However, there is every reason to believe it will become profitable within say, a couple of years.

(c) The network will be operated by one or more private companies. Part of their franchise deal will be that they must finance and execute the construction of the rest of the system, i.e. the spurs to Sacramento and San Diego.

It is precisely this mechanism that AB 3034 threatens to undermine by allowing all segments of the proposed network to compete for the initial funds. The saving grace is that proponents of each segment must raise matching funds from federal and/or private investors before a penny of the state bond can be spent on it. Chances are these outsiders will insist on the construction of a starter line - most likely SF - Anaheim - rather than entertain separate entreaties from various local bigwigs.

In all likelihood, the operators of the starter line will take out private revenue bonds to meet their obligation to construct the two spurs included in the original design. This has been the pattern in other countries. If, as expected, the network only becomes more profitable as a result, they should have little trouble raising the finance on the commercial market. The private bonds for these two spurs will presumably be subordinate to the original state bond, but they will be serviced before the state receives any net contributions to its general fund.

In other words, the state will not be able to siphon off profits from HSR operations for any other spending programs until the entire 800 mile network has been completed, probably sometime in the mid-to-late 2020s.

Diridon also talks about a possible further extension from Sacramento to Redding. Since that is not part of the original plan, my understanding is that the state will not be obliged to forgo its share of total operating profit to support its construction.

Anonymous said...


I want to now comment on this part of your post:

"It is precisely this mechanism that AB 3034 threatens to undermine by allowing all segments of the proposed network to compete for the initial funds. The saving grace is that proponents of each segment must raise matching funds from federal and/or private investors before a penny of the state bond can be spent on it.

Chances are these outsiders will insist on the construction of a starter line - most likely SF - Anaheim - rather than entertain separate entreaties from various local bigwigs.

There is nothing in AB-3034 as currently written that mandates any matching funds. The bill simply encourages, invites Federal and / or private funding.

Thus I disagree with your statement that each segment must raise its own matching funds. In point of fact as AB-3034 is written, no matching funds might be raised, and a segment that could be funded totally from the bond monies, could be built.

Also as written, CHSRA can spend $950 million on studies. If raising outside funds proves impossible, the authority can keep itself running by spending this considerable sum of money.

Finally the ridership study, clearly shows the greater number of passengers by far would be between stations in the two (La and Bay areas), and not traveling between LA and SF. That would indicate to me outside funding would much prefer to fund a small segment like say, LA to Anaheim, really a commuter segment or something that might gain ridership numbers on the order of what BART is able to command today. I only use this as an example, I don't think it is practical; I think it is too short.

Robert Cruickshank said...

Congress has repeatedly signaled its intention to provide funding for our line, an assurance that has been given by both Republican and Democratic members of Congress from our state. They are waiting on us to lay down the initial funding before they commit, but as I reported here before, support in Congress for this is strong.

As to Harold's other points, I don't think he is interested in being convinced, or interested in assessing the evidence with an open mind. rafael has accurately noted that California's population density is comparable to that of Spain, where HSR has been a runaway success, generating surpluses. The notion that "only a fraction" of the state's population would ever use HSR is flatly absurd, especially in the face of ever-rising gas prices and the resulting reductions in airline capacity and affordability on intrastate routes.

Then again, Harold may be a representative of a small group of HSR deniers, whereas most Californians seem to be embracing the concept of HSR and are now rightly debating the proper funding mechanism. That's exactly where the discussion now ought to be.

Rafael said...

@ Harold -

the full text of AB 3034 with amendments as passed by the Assembly is here.

1) From the Legislative Counsel's Digest:

"The bill would require that not more than 10% of bond proceeds be
used for environmental studies, planning, and preliminary engineering
activities, [...]"

Since only $9 billion of the bond are available for work on HSR proper, this means $900 million for the tasks described above.

2) Also from the Digest:

"The bill would require the authority to give priority in selecting segments for construction to those segments that are expected to require the least amount of bond
funds as a percentage of total cost of construction, among other

This is a problematic issue because it would give priority to those segments that cost least to construct rather than those likely to generate the ridership needed to generate operating profits. Those profits are critical to financing the construction of phase two.

From Section 2:

"the following high-speed train system corridors:
(A) Sacramento to Stockton to Fresno.
(B) San Francisco Transbay Terminal to San Jose to Fresno.
(G) Altamont Corridor connecting the central valley to the east
bay Central Valley to the East Bay.
(2) Nothing in this section shall prejudice the authority’s
determination and selection of the alignment from the Central
Valley to the Bay Area in its certification of the environmental
impact report.

At first glance, this also appears to be problematic in that it appears to reopen the Altamont vs. Pacheco discussion.

The final Bay Area to Cental Valley EIR/EIS that has already been sent to the FRA explicitly favors Pacheco Pass plus an ill-defined "overlay" in the East Bay. If the FRA renders a positive "Decision of Record" on that, the Authority is expected to certify it.

Personally, I hope the incoming FRA director insists on clarification of just what exactly that overlay is supposed to be and if it is even a legitimate part of the HSR project.

After all, the map of maximum speed by location suggests that the overlay will have to be completely grade separated and require tunneling at both Altamont and Tolman Peak north of Niles. Do statewide traffic patterns justify the construction of not one but two full HSR alignments out of the Bay Area? If the overlay will only be used for commuter service, why is it part of the HSR project?

Given that routing all trains via north San Jose rather than constructing a new rail bridge at Dumbarton would appear to have only a minor impact on express line haul times between SF and LA, the FRA ought to ask why the Authority decided on Pacheco-plus-overlay instead of on Altamont only. According to the cost study, the difference is at least $6 billion and, federal matching funds do not grow on trees.

In particular, not one of the many variations considered bypasses San Jose Diridon station in favor of nearby Santa Clara station. How is this omission justified in the larger context of a statewide project?

From Section 4:

"2704.08. (a) Proceeds of bonds authorized for high-speed train
purposes pursuant to this chapter shall not be used for more than
one-half of the total cost of construction of track and station costs of each segment of the high-speed train system."

(emphasis added)

In other words, if folks in the Central Valley want the Sacramento-Stockton-Fresno segment to be prioritized, they must cough up at least 50% of the construction cost or persuade the feds and/or private investors to do so.

From Section 6, which details how the $950 million reserved for HSR feeder systems will be allocated:

"(f) In order to be eligible for funding under this section, an
eligible recipient under paragraph (3) of subdivision (a) shall
provide matching funds in an amount not less than the total amount
allocated to the recipient under this section."

This is what I meant by "saving grace": the requirement for external matching funds will impose the phasing discipline on the whole HSR project that the legislature was unable to impose itself - perhaps because of the requirement to approve the bill by a 2/3 supermajority.

3) Page 5 of the ridership study illustrates the markets for interregional trips expected for 2030. However, the slide covers all modes of transportation - not just rail.

The relatively low number of trips between the Bay Area and Los Angeles is indeed surprising, considering the desire to prioritize this section of the network and run almost a quarter of all trains in express mode between these locations.

Instead, the slide suggests that for the subset of corridors served by HSR, trips within the Central Valley, between the Central Valley and SF/LA and between Los Angeles and San Diego will dominate.

In other words, many trips will involve only a fraction of the route served. A seat occupied by one passenger between LA and Bakersfield may be used by another between Fresno and San Francisco. This is a key difference between seats on intercity train and those on short-hop flights. Seen in this light, an SF - LA starter line does make sense.

Another way to look at this is what fraction of HSR trips will likely be served by any given segment of the network. This level of detail is contained in Table 3.2-12 of this document. Note that this table is based on the base case in the ridership study. If you assume that auto and aviation costs net of inflation will be 50% higher than in 2005, the HSR total sums to around 95 million annual trips.

Anonymous said...


You wrote
From Section 4:

"2704.08. (a) Proceeds of bonds authorized for high-speed train
purposes pursuant to this chapter shall not be used for more than
one-half of the total cost of construction of track and station costs of each segment of the high-speed train system."

I admit to missing this constraint on funding. Thank you for pointing this out. So AB-3034 as written will require at least 50% of construction funds be provided from funds other than the fall 2008 bond measure.

Anonymous said...

Just a quick follow-up on where the rest of the funds for the project will originate.

Since this is at least, even by CHSRA estimates, a $30 billion project for just the SF to LA segment, AB-3034 does not require the the remaining $20 billions come from Federal or private equity. As Rafael point out, other bonds can be floated, which would be additional obligations on the taxpayers of California.

The public is certainly led to believe that only the original $10 billion bond issue, would come from California taxpayers. That was supposed to be the main purpose of AB-3034, and it certainly does not provide that mandate.

But as so often happens when the politicians get involved, even the supporters can end up not getting what they wanted and expected.

Rumor has it that the senate is going to make its own set of revisions.