A Low-Cost High-Speed Rail Plan?

New York University’s Transit Costs Project has been asking the question, “Why do U.S. transit projects cost so much more than similar projects in Europe, Asia, and South America?” A recent report from the project proposes to significantly speed up trains in Amtrak’s Northeast Corridor for only $12.5 billion (plus another $4.5 billion for new rolling stock), which is far less than Amtrak’s proposal to spend at least $110 billion in the same corridor.

Amtrak’s plan called for building a new line for the entire 457 miles between Boston and Washington. The Transit Cost Project’s plan instead proposes to fix specific bottlenecks on the current line. The result, says the project, would be trains that could travel between New York and either Boston or Washington in less than two hours, as opposed to 3 hours (New York-Washington) to 3-2/3 hours (New York-Boston) today.

The most important lesson from this report is that it inadvertently answers the original question: why are U.S. costs so high? That answer is that U.S. transit agencies, including Amtrak, always seek the high-cost solution to any problem. They have no incentive to make a profit because doing so would lead Congress and the states to reduce their operating subsidies. They have no incentive to find affordable solutions to any problem because they regard the federal government and state taxpayers as bottomless pits of money.

Instead of building a whole new line as Amtrak proposes, the project suggests some modest improvements such as $1 billion for modernized catenary that can support faster trains, $1.7 billion for capacity improvements in New Jersey, $600 million for a fourth track through Baltimore, $1.1 billion for improvements at several junctions, and $100 million to replace a worn-out bridge that Amtrak wanted to spend $4.25 billion on just to have less steep grades.

That all sounds reasonable at first glance, but the project also proposes to spend $5.0 billion on 78 miles of new line between New Haven and Kingston Station, Rhode Island. This immediately raised a red flag for me because that cost averages $64 million per mile, including bridges across several rivers. California is currently spending close to $100 million a mile building rail in the Central Valley, which doesn’t have any significant rivers to cross.

The Transit Costs Project says it bases its cost estimates on “typical western European costs.” Those are the costs that it thinks American transit agencies and Amtrak should aspire to instead of the amounts they are currently paying, which are much more. In other words, it won’t be enough for Amtrak to limit itself to the specific improvements in the project’s report; it will also have to make significant reforms to implement those improvements in a cost-effective manner.

The project’s case study of New York transit costs found three areas that made U.S. projects cost much more than European ones. First, U.S. transit agencies tended to overbuild their projects, providing much more capacity and using higher-cost techniques than necessary. Second, U.S. labor costs are much higher than in Europe. Third, poor management within the transit agencies leads to much higher planning and design costs and forces contractors tend to bid more on U.S. projects.

The project’s high-speed rail proposal addresses the first of these issues but not the other two. Since these are unlikely to be fixed any time soon, the costs projected by the Transit Costs Project should be approximately doubled. While the resulting $25 billion (plus around $5 billion for new trains) is still a lot less than Amtrak’s proposal, it is questionable whether the benefits of such expenditures can justify this cost.

This suggests that there is a fourth reason for high transit project costs in the U.S., and that is political. The political system thrives on delivering high-paying jobs to labor unions, high-cost contracts to planners, designers, and builders, and high-overhead projects to government bureaucracies. Nothing the Transit Cost Project has proposed would fix this problem.

The Transit Cost Project itself has a major blind spot, which is that it takes for granted that passenger rail is the most appropriate tool in many different situations. In fact, planes are faster and less expensive on routes such as New York-Boston, New York-Washington, and Los Angeles-San Francisco. Buses are slower but are far less expensive and cars are more convenient and less expensive over shorter distances such as New York-New Haven or Washington-Baltimore. Within urban areas other than New York, buses are superior in almost every way to rails, and even in New York the costs of new rail construction and rehabilitation of existing rail are so high that the real question should be whether the region can afford to maintain the rails it has, not whether it should build more.

I appreciate that the Transit Cost Project is examining some of these questions. If it is to be truly effective, however, it needs to take an even deeper look.