Amtrak

An editorial (reg req or no reg here) yesterday in the Seattle Times by Neal Peirce talks about Amtrak.

Peirce starts talking numbers. He had to do that. Nothing gets me more interested in a story than giving me numbers to average or manipulate. So here we go.

Peirce says:

"...an abrupt cutoff of federal support, currently $1.2 billion a year..."

"...last year increased its ridership 4.3 percent, to 25 million passengers..."

For fun, let’s assume that no one riding on Amtrak ever makes multiple trips. At 25 million unique passengers, the government spends $48 each. Seattle to Portland today is running at $31 one way taking 3 and a half hours, not a profitable route. Driving takes just under 3 hours to cover 180 miles. Seattle to Eugene runs $50 one way and takes just over 6 hours. Driving takes approximately 4 ½ hours to cover 280 miles. Seattle to Whitefish, Montana costs $67 and takes nearly 14 hours. Driving takes approximately 9 hours to cover 530 miles. If you assume 17 miles per gallon in a car at $2.29 per gallon, the cost to drive is $24.25, $37.72, and $71.39 respectively. With the added flexibility of being able to leave whenever you want, a car is much more economical.

Well, what about urban areas. New York Penn Station to Newark, New Jersey, obviously a commuter route (17 minutes on the train), cost $27, running every thirty minutes during the day and every 17 or so during the rush. This pair is served by light rail also. On the subway system, the trip takes approximately 33 minutes. But there are many more options as three to four subway trains leave in the same time span during the rush. Additionally, the fare for the subway is $1.50. Amtrak costs 18 times as much for only twice the speed and one quarter of the flexibility

How bad does it get? The White House states "...it would literally be cheaper for Amtrak to buy each passenger a plane ticket to the next destination. For example, a round trip ticket for direct flights between San Antonio and Chicago (the Texas Eagle route) can be purchased for as little as $216." Per passenger loss on that route is $258.

Peirce also goes on to mention:

"...Amtrak shutdown would be a body blow to the crowded Northeast Corridor (Boston-Washington)..."

Amtrak runs daily commuter service in the Northeast. So it’s safe to assume that ridership is definitely not 25 million unique passengers. So we can find the other bound of the per rider expense to our wallets. 25 million passenger trips, 730 daily trips per year per passenger gives us a ridership of 34,247 people. At $1.2 billion, that works out to $35,040 per passenger.

Somewhere between $48 and $35,000 is the actual cost. From Amtrak’s own website

"Each day, approximately 68,000 passengers travel on Amtrak."

This number is simply 25 million passengers spread evenly over 365 days in a year. Obviously not accurate, but we’ll run with it. If we assume these are the same people every day, the cost is roughly $17,650 per person.

So let’s assume that half the trips are made by people commuting 5 days a week to and from work. That’s 520 trips per commuter per year. That’s 24,039 people using Amtrak to commute and 12.5 million that use it only once. At $48 per trip, means we spend nearly $25,000 per commuter.

Peirce states that, "Other advanced nations are now building world-class rail systems." While this is true, many of them are implementing high speed rail. The infrastructure requirements of high speed rail are far beyond our current system. 97% of the track used by Amtrak was designed to transport freight.

If you’ve ever ridden Amtrak, you know how bumpy the ride is. You also know how slow the trains move the majority of the time. The freight lines that own the rail don’t care whether or not the wheat or oil or potatoes on board get thrown around a bit. Their trains can fly down the tracks at 80 miles per hour because a few bumps doesn’t hurt anything. Amtrak has humans on board that don’t care for being bounced against the sides of the car.

To have a truly functioning passenger rail service modeled on Europe and Japan, a huge investment in infrastructure must be made. A high speed national rail would have to move away from renting right of ways to owning and building its own tracks.

Other concerns would still exist, however. Most Americans don’t really care for trains rolling through their neighborhoods. They’re noisy and unsafe to people near the tracks. Trains are actually forced to slow down in towns, as well as blow their horns at crossings, making more noise. A system of tunnels would need to be constructed to allow trains to pass at speed under small towns like Sumner, WA. A few miles of tunnel for the Alaskan Way Viaduct is going to cost somewhere between $6 billion and $11 billion.

Europe and Japan are not forced to deal with the expanse or extremes of our geography. A drive from Lynden, WA to Key West, FL is just under 3,600 miles. Add to that two major mountain ranges, multiple major rivers, myriad temperature extremes, the costs to build an efficient line begins to build.

Peirce cites Senator Kennedy, "Japan and Germany, notes Sen. Edward Kennedy, D-Mass., spend about 20 percent of their national transportation budgets on trains, compared with the 2 percent Amtrak has requested."

Light rail costs approximately $23 million per mile on a relatively flat surface. The Colorado Rapid Transit Alliance estimates their proposed rail running between 80 and 110 mph will cost $40 million to $60 million per mile. (By comparison, highway construction runs $5 million to $10 million.) Even at the low end, which would be to assume no elevation change, no tunnels, no right of way purchases, etc., the cost of a rail from Lynden to Key West would be $144 billion. That one line is nearly three times the total budget for the Department of Transportation.

If we were to build a high speed rail system comparable to the heavy rail passenger system already in place, we would need to build roughly 22,000 miles of track. Total cost just to actually put rails in: $880 billion dollars.
$880,000,000,000
By 2036 at Peirce and Kennedy’s 20%, we will have paid for the rails.

That number assumes $40 million per rail mile. It doesn’t include rights of way, extraneous infrastructure such as tunnels and bridges, or any other possible nit that could drive up the cost. If we consider those and government cost performance, we should probably double the number. That pushes us out to 2047. This project would cost more and take longer than fixing Social Security.

Amtrak is a dinosaur. It is inefficient in time and expense. Implementing high speed rail will be a huge expense for a small, if even existent, benefit. Until rail can get you from Seattle to LA in two and a half hours, most people will fly.

Let Amtrak die.