Coyote Blog

The Anti-Planner argues that mass transit will never be energy efficient, mainly because it is virtually impossible to improve occupancy.  The arguments for transit saving money all tend to include the line "will be efficient when occupancies increase" but he shows pretty clearly why that is probably not going to happen.

Also note pages 2-15 and 2-16 of this report.  Compare the trends of auto and airline energy intensity with rail and bus.  While cars and planes have decreased their energy use per passenger mile by quite a bit, rail has been flat and buses have been getting worse.  In fact, auto transit became more energy efficient than buses twenty years ago and continues to get better.   Airline travel has become nearly as energy efficient as Amtrak.

We have all heard environmentalists and other American intellectual snobs lamenting that we just are not as smart as Europeans because we have so much less passenger rail.  But because freight and high-speed passenger rail service does not coexist well on the same tracks, urging more passenger rail on the US rail net is effectively asking for more freight to be dumped onto the highways.

Megan McArdle writes:

Moving freight by rail rather than by truck is an enormous carbon saving; one locomotive can haul as much as hundreds of trucks.  It also reduces highway congestion.  Unfortunately, it's hard for passengers and freight to share tracks.  In part, it's difficult simply because it's expensive to upgrade track to handle passenger speeds, but also because freight moves much more slowly, and on an irregular schedule.
I might well argue that if we were simply trying to maximize environmental benefit, we'd ignore passenger rail, and focus on upgrading our freight systems, which sorely need it.  Moreover, these upgrades could largely be made without the massive procedural obstacles that block new high speed rail lines.

But freight rail is not sexy.  It does not excite donors, and it does not excite most of the voters who are motivated by high speed rail.  Politicians win votes by delivering (or at least promising) highly visible improvements; not by silently enhancing the movement of goods from port to Wal-Mart.

One of the justifications for diverting highway money to ridiculously expensive light rail systems is that light rail supposedly reduces energy consumption.  Really?  This is via the most recent report from the DOE's Transportation Energy Book, as highlighted by the Anti-Planner (click to enlarge):

The figures for cars are from tables 2.12 and 2.13 of the same report.  Even the best light rail systems are not substantially more efficient than cars, and this gap will likely continue to close, as it has for years, as cars get more efficient.

A Note on Freight: By the way, passenger rail promoters in the US always point to the Europeans as having a better rail system.  But while the Europeans put more of their passengers on rail than does the US, they put less of their freight there.  I would argue that the US system is much more "green", as the differences in energy use between a ton mile of freight on road vs. rail is much larger than the difference in energy use of a passenger mile on road vs. rail.  And besides, from a lifestyle standpoint, would you really want more freight on the roads?  (This is a real tradeoff -- unless one spends the absurd amount of money to build two separate systems, a rail network can be optimized for freight or passengers -- the two do not coexist very well on the same tracks).

Postscript: Just to head off the obvious rhetorical battles -- the incremental energy efficiency of moving one driver to a light rail rider of an existing system is very high.  The car consumption goes away and the train does not incrementally increase its energy use much with one more passenger.  So at the margin, it is correct when someone tells you that it saves energy to shift your commuting to an existing light rail line.  However, it does not make sense, from an energy perspective, to build a light rail line in the first place.  The investment is too high, the energy savings are negligible or non-existent, and the operating cost are so high that light rail tends to crowd out bus operations that help the poor.  As I have written before, for every light rail system I have checked, the cost to build the system is enough to buy every daily rider a Prius and the operating deficit enough to keep every one of these Prius's filled with gas.

Update:  I further understand that cars in the city likely have lower gas mileages than these averages, particularly for commutes that might be substituted by light rail.  But light rail is sold as if it is substantially more energy efficient, and it really would have to be orders of magnitude more efficient to justify the capital costs that are so much higher than for an equivalent capacity of roadway.  The efficiency is just not there.

From the Anti-Planner:

Forty states have asked the federal government for a total of $102 billion for high-speed rail. This suggests that the Antiplanner's estimate of $90 billion for the cost of the Obama high-speed rail plan was low.

Secretary of Behavior Modification Ray LaHood says that this "shows that the country is ready for high-speed rail." Of course, all it really shows is that state bureaucrats are ready for free federal money.

The Anti-Planner has more on the California high speed rail proposal I wrote about earlier.  My guess was that the first $9 billion bond issue, on the ballot this fall, would not get the train out of the LA metro area.  Well, I was right and wrong.  The smart money thinks the line will start at the other end, in San Francisco.  But the betting is that for $9 billion the line won't even get out of the San Francisco metro area, making it perhaps as far as San Jose. 

But we have a second data point -- there is a proposal on the table to extend BART from Fremont to Santa Clara for $4.7 billion, a distance (as shown on the map below) about a third of that from San Francisco to San Jose.

I am not sure what high-speed rail technology that they are considering, but a true high-speed line requires special alignments, track, and signaling that should make it FAR more expensive per mile than a BART line (just as an example, a true high-speed line could take miles to make a 90 degree turn, eating up land and reducing alignment flexibility in a very congested and hilly area).  And remember, the BART cost estimate is probably low.

No way these guys get to San Jose for $9 billion, much less to LA for $40 billion.  Just what Californians need with their massive budget deficit:  a brand new white elephant.

The Anti-Planner links an absolutely scathing article in the Miami Herald on the absolute disaster they have made of their mass transit system.  This is a great summary:

Miami is just one more example of the points the Antiplanner keeps making about rail transit:

1. Transit agencies might run excellent bus systems. But when they
start building rail, they quickly get in over their heads by optimistic
forecasts, unforeseen costs, and the sheer humongous expense of
building dedicated transit lines.

2. Though all rail systems require periodic expensive maintenance,
few transit agencies set aside any money for this because it is easier
to spend the money now and let future managers worry about the future.

3. Though the rail systems are usually built to serve downtown
white-collar workers, in the end it is the transit-dependent people who
rely on buses who pay the cost.

4. There is only one thing rails can do that buses can't do better,
faster, and more flexibly, and that is spend a lot of your money.

I would like to observe one other thing at work in the Miami example that looks to be exactly what we are facing here in Phoenix in the next election.  Miami offered up a transit tax referendum for something like $800 million.  They promised a mix of highway improvements and rail.  In several cases, including the upcoming referendum in Phoenix, I have tried to warn people that the people who put these referendums together are rail-ophiles.  They have learned, however, that rail alone won't sell a bond issue or tax, so they throw in a bunch of highway improvement promises, which people really will pay for, as window dressing.  Often, however, these improvements never get done, as they are empty promises to sell the tax.  We see exactly this in Miami:

But five years and more than $800 million later, the county has spent more
than half the new money on routine Transit operations and maintenance while adding 1,000
jobs to the payroll.

   There were initial achievements. The county added 11 million miles of bus service, gave
free rides to seniors, and briefly experimented with 24-hour rail. It spent $40 million on
hundreds of tiny public-works projects....

   For example, here is the cost estimate that was attached to the 44 road projects that
county commissioners asked for: $0. The projects have since been estimated to cost
$428.2 million.

   Nor was any money earmarked for an unspecified number of flyover intersections on the
list of promised improvements. Such projects, which involve raising an existing road to
pass over another, cost as much as $18 million apiece today. None have been built.

So this tax was sold in part as a highway improvement tax, but $0 was actually budgeted.  The highway piece was a lie to sell the tax.  Beware Phoenicians.

The Anti-Planner links an article by a Federal Reserve Bank economist on the housing market in Houston and how it is affected by zoning:

"Given that Houstonians had access to the same new types of
mortgages as the rest of the country and that Houston has had greater
population growth than other large metros, we might expect price
appreciation to be stronger in Houston than elsewhere," says the
article. "However, the opposite has been true."

The reason? Houston's lack of zoning and its large supply of land
available for development allowed builders to respond to easy credit by
increasing the pace of construction. Slow and unpredictable permitting
processes prevented builders in many other regions, including Florida
and the Pacific Coast states, from similarly stepping up production.

While some cities and regions have further delayed construction by imposing adequate public facilities or concurrency ordinances, Houston allows developers to create their own municipal utility districts.
Through these districts, the developers install the sewer, water, and
other facilities needed by their developments and charge the property
owners over time.

The result is that housing prices did not bubble, and they are not
significantly declining today. As of the fourth quarter of 2007, in
fact, they were still increasing. Anecdotal evidence from local
realtors and developers indicates that the tightening credit market has
soften the demand for homes under $200,000, but homes above that price
are still selling well.

Whatever correction Houston faces, says the article, "takes place in
the context of prices that are squarely in line with local construction
costs and without the painful supply-induced downturn under way in many
other markets." This leaves Houston relatively immune to the ups and
downs of housing prices experienced in regions with planning-induced
housing shortages.

I need to think a bit about how that relates to this.

The Anti-Planner has an absolutely fabulous article about a Wisconsin passenger rail proposal, but in fact what the article really is about is how government decisions get made.

According to RTA's latest newsletter,
the KRM would cost about $200 million to start up and would require a
$6.3 million annual operating subsidy. For that it would carry about
1.7 million trips per year, which translates to 6,700 per weekday.

In other words, RTA wants to spend $200 million to take 3,350 people
to and from work each day. The Milwaukee-Racine-Kenosha urbanized areas
have about 750,000 commuters, so RTA's proposal would take less than
half a percent of them to work. But they would all have to pay for it
in the form of some local taxes plus a diversion of a share of federal
and state gasoline taxes to fund the rail line.

By the way, though this post isn't meant to be entirely about rail itself, let's use Coyote's test on this rail proposal.  As a reminder, here is Coyote's test:

Take the total capital charge and compare it to the cost of buying every projected rider at $22,000 Prius.  Then, take the operating subsidy (which is always higher than projected) and see how it compares to the average gas consumption in a year of said Prius's.  If the projected capital charge and subsidy could have bought every rider a car and all the gas they need to drive it, then the rail line is not only an average run-of-the-mill government boondoggle, but a total and complete ripoff.

And, the KRM... FAILS.  And fails miserably.  The $200 million charge would have bought every rider TWO Prius's and still have some money left over, and the operating subsidy, sure to be larger in reality, would buy each rider about 627 gallons of gas a year, which at 30mpg would get them 19,000 miles per year.  But don't worry, KRM, every single new rail system to which I have applied the test has failed (Phoenix, Houston, LA, Albuquerque).

But lets continue:

The planned commuter line would run 14 round trips per day, which
means each train would have about 240 people on board. That's about
five bus loads. So why not just buy five buses for each planned
trainset and move people by bus instead?

The newsletter explains that RTA considered a bus alternative, but
it would attract only a third as many people as the rail line. It would
also cost only an eighth as much to start up, so I always wonder why
don't they just invest three-eighths as much in buses and carry as many
people as the rail line.

But then I noticed that the rail line was projected to have seven
stops between Milwaukee and Kenosha, while the bus line would stop 27
times. As a result, the bus would take almost twice as long as the
train. No wonder it attracted so few people!

The train would average just 38 miles per hour and RTA admits that
it would not go significantly faster than motor vehicles, so there is
no reason why buses could not be run on schedules similar to the train.
So why didn't they consider an alternative in which buses stopped only
seven times?

It turns out they did. The report
from the consultant hired by RTA included a bus-rapid transit
alternative that stopped fewer times than the regular bus alternative.
It included some exclusive busways, so it cost a lot more than the
regular bus alternative, but it would cost only half as much as the
train. Moreover, it was projected to carry as many riders as the train.

Naturally, RTA told the consultant to drop this alternative from further consideration.

The Anti-Planner shoots back what to me looks like a really good proposal:

The consultant had also estimated that the bus-rapid transit
alternative would disrupt traffic more than the trains. But if the
busways (which would move no more than about 5 buses per hour) were
opened to low-occupancy vehicles that pay a toll, they would actually
relieve congestion. Plus, the tolls would pay for most if not all of
the new lanes, and by varying the toll, the lanes would never get
congested so the buses could meet their schedules. This would result in
transportation improvements for both auto drivers and transit riders,
and at a very low cost to taxpayers

We don't need more highway funds because right now, as estimated by the Anti-Planner, about 40% of Federal highway funds go to non-highway projects.   In particular:

Over the past fifteen years alone, America has spent well over $100
billion on rail transit construction projects but has little to show
for it. As mobility advocate John Semmens pointed out a few days ago in
a recent Washington Times op ed, transit's share of urban travel has actually declined since 1995.

Wow, money well spent, huh?  I have written many times on commuter rail follies in Phoenix and other western cities that are utterly unsuited to rail transit.  The most recent news here in Phoenix is that design flaws are appearing, even before the first train is run.

I have chastised our city on many occasions (more here) for spending enormous amounts of money on a new light rail / streetcar system for Phoenix.  These light rail systems can be twenty or more times as expensive, per mile or passenger carried, than a similar bus system.  But what really, really makes light rail nuts for Phoenix is the lack of flexibility.   Our hugely expensive new light rail system serves just one corridor, in a city that really does not even have a downtown.  Phoenix is characterized by a nearly infinite number of commuting routes that don't overlay nicely on a suburbs to city-center pattern as they might in, say, Chicago.  Further, the current route arguably follows the least congested route of any in the city!

The incremental cost of light rail over bus systems has been justified to us by our government overlords by economic development.  The argument goes that light rail creates more business development along their routes than a bus system.  Now, I am skeptical of this, given the region justified building a billion dollar stadium for the hapless Cardinals on the same justification (not to mention numerous subsidies of a couple of college bowl games that add little to an area that is going to get holiday tourists because of its climate whether there is a football game or not.

But what about Portland?  Supposedly Portland light rail is the go-by which all we unplanned cities should emulate.  But the Anti-Planner brings this helpful observation about Portland's experience with light rail and development:

Streetcar advocates often say that 7-mile-per-hour streetcars aren't about transportation, they are about economic development.
But they expect the Department of Transportation to pay for them out of
highway user fees. Why didn't they ask the Department of Housing and
Urban Development for the money?

Of course, the Antiplanner doesn't believe
that streetcars catalyze economic development. Instead, they merely
catalyze more tax subsidies for economic development. Portland spent
$90 million on a streetcar line and $665 million on subsidies to
development "” then credited the development to the streetcar line.
Yeah, right.

The unfortunately named and horrendously ill-conceived and over-priced Seattle trolley takes to the streets today.  The Anti-Planner has an overview in the third in his series on light rail follies.

The Anti-Planner has a series of posts of late on light rail that in total point to a perverse moral hazard in public transportation funding that helps to explain why states and cities are building so many rail projects, when the numbers almost never make any sense (as I blogged for LA, Phoenix, and Albuquerque).  Though the Anti-Planner does not state these rules, from his recent posts I have inferred three rules:

• A city can get capital construction dollars from the feds, but you can almost never get maintenance or operations money (similar story in recreation)
• The feds will fund big, expensive, sexy rail projects.  They will not fund purchases of buses and are unlikely to fund something so prosaic as a bus stop or terminal  (general rule of thumb:  federally funded projects must be large enough to justify being named at some future point after the local Congressman or Senator who earmarked the project.)
• It is very easy to de-fund bus systems -- you just don't replace aging buses and cut routes over time.  It is hard to de-fund, or, god forbid, abandon a rail line, since the thing sits out there so visibly.  Sunk costs can also be a political issue if rail lines were to be closed.

For most public transportation goals, particularly in spread out western and southern cities, buses are a cheaper and higher service solution than rail.  They can carry the same passenger traffic for far less total dollars (capital plus operating costs) and they can cover far more routes.  In fact, one can argue that rail lines are inherently regressive, as they tend to serve commuting corridors of the middle and upper classes rather than the typical routes of the poor, for whom the systems are nominally built.

So what can one expect by the application of these three rules?  Well, we would expect local authorities to favor large, expensive capital rail projects rather than refurbishment or expansion of bus systems.  As operating costs rise for the trains, we would expect bus service to be cut back to pay for the rail operating deficit.

Which is exactly what happens.  In fact, rail tends not to increase total ridership at all, at best shifting ridership from inexpensive buses to expensive trains, and at worst decreasing total ridership as rail lines with just  a few stations and routes replace more extensive webs of bus transport.  And, in twenty years, when these rails systems need extensive capital overhauls, we find cities with huge albatrosses on their hands that they are unable to maintain or update.