As usual, Yonah Freemark at the Transport Politic was much more thorough and mathematically ruthless in a topic I once considered.  Today he too wrote about the cost of high speed rail to the passenger (part 2) and how much fares should cost.  While I will provide excerpts of his post here, I highly recommend clicking the link to read his full essay for the mathematical detail and his graphical comparisons.  The following are some of the juiciest excerpts regarding Amtrak, Acela, the future of high speed rail and comparison to foreign high speed rail services.

On cost per hour:

But Amtrak’s problems are not due to the fact that it is particularly inefficient in the Northeast. The fares it demands per hour of travel are roughly on par with those charged by foreign rail operators. The American rail system’s problem, rather, is that its trains are too slow in general, increasing labor, maintenance, and operations costs. If U.S. rail services are to be successful in attracting customers at reasonable prices, in other words, one way to do so would be to offer services at higher speeds.

… Peer experience on specific high-speed routes demonstrates just how expensive Amtrak’s Acela trains are. Acela rides cost more than $0.35 a kilometer, compared to $0.20 for Milan-Bologna trains in Italy or $0.08 for Paris-Lyon trains in France.

On passenger capacity:

These issues are compounded by the relatively low capacity of the system’s existing trains — Northest Regional trains offer 5-9 passenger cars and Acela trainsets all have 5 seating cars. In 2007 and 2008, Amtrak was frequently selling out trains, and with no room to expand, the organization had an incentive to increase prices, especially since commuters on the Northeast Corridor subsidize rides on other parts of the system. Amtrak’s Acela trains have a capacity of 303 riders. Two French TGV Duplex trains coupled together can carry 1,024 passengers. The ability to move more people in one trainset allows for operational efficiencies — and, as a result, cheaper tickets for those who make it onto the train.

On future transit consumption:

If American high-speed services offered similar prices for time traveled as Amtrak does today — at $45 per hour of running time for standard fares and $15 at reduced prices — on faster trains, U.S. commuters would switch to rail in droves. The San Francisco-Los Angeles route being planned by the State of California, with a travel time of 2h40, would cost $40 for reduced-price tickets and $120 for standard fares; those costs seem perfectly acceptable for just about everyone. A renewed Northeast Corridor, offering travel between New York and Washington in 1h40 (at an average of 220 km/h), would cost $25 for customers buying reduced-price fares. People currently driving their own cars or riding buses between the cities would take a second look at those prices.

All of these points are valuable.  Any American who justifies any opinion regarding  high speed rail based on Acela experience is deluded.  Acela has a multitude of problems technologically (see the pantograph problem), logistically (sharing tracks), with funding and with the basic direction of the line (not enough straight track to reach speed).  However, that does not mean high speed rail in America cannot be efficient, appropriately priced for the American transportation consumer, and a boon to Americans and cities.  Freemark has masterfully shown how better systems can actually keep prices down while attracting more passengers.