The Hypothetical Hypotheticals of U.S. High Speed Rail

A recent blog post at the Daily Kos by BruceMcF (10/4/2009) takes on Randal O’Toole’s critiques of high speed rail (HSR). Some of the critiques are worth noting, but a surprising number fall into the same trap the author accuses Randal of falling into.

For example, BruceMcF criticizes Randal for being “hypothetical” and scenario driven, yet virtually all the studies of passenger costs and traffic outside the Northeast corridor are hypothetical and scenario-driven in the sense he uses the term. They aren’t based on hard numbers or experience. The United States simply has no meaningful experience with high-speed rail as it exists in Europe and Asia (and that includes the Northeast Corridor which doesn’t get above 100 mph for most of the trip). U.S. experience with intercity rail is also dismal. So, any forecasts of ridership, potential fare revenues, or per passenger mile operating costs are largely hypothetical, assumption driven, and based on “best guesses.”

In the worst cases, such as the California HSR intiative, the assumptions are wildly optimistic and one could be excused for viewing the commissioned reports as exercises in political gamesmenship. Reason Foundation analyzed the California initiative extensively in its HSR Due Deligence Report and we have pointed out other weaknesses in a U.S. system through our commentary on its job creation potential, relevance to European approaches, its potential cost, and its inability to serve broad-based transportation needs.

The Daily Kos author also engages in some of the same statistical slight of hand he accuses Randal of doing, mixing years and estimates and costs. He talks about operating cost/revenue ratio’s but not total costs. He also implies that the only real relevant criteria is whether rail covers its operating costs, even claiming that a 2.5 million riders on the Ohio 3-C (Cincinnati-Columbus-Cleveland) corridor would be a “runaway financial success.” No, the 3-C corridor would still not cover its total costs. Nor does he discuss meaningfully the very narrow segment of the traveling market that HSR attracts along this corridor. Based on the projections of an oustide consultant for the Ohio Rail Development Commission, 97 percent of the travel along the corridor would still be by car (2 percent by train, and the remainder by air and bus). See the Ohio Hub ridership and cost forecasts here.
On the whole, I think BruceMcF’s post validates many of the arguments made by HSR critics–it reaches a tiny part of the traveling public and will be a huge financial black hole–and tends to talk past the legitimate criticisms.
Nevertheless, I also believe he makes several reasonable arguments about the weaknesses of the HSR critiques. Critics, for example, tend to mix too casually data on intracity passenger rail performance, slow and unreliable conventional intercity rail, and HSR. It would be far more productive to take the best of the Amtrak services–the Downeaster, Keystone Corridor, Hiawatha lines, plus the Acela–to make the points that HSR is unlikely to be successful in the U.S. One of the few examples of this kind of benchmarking analysis was also commissioned by the Ohio Rail Development Commission as part of its due diligence for the Ohio Hub.
In addition, I think BruceMcF is spot on in his critique of the way HSR critics use density. It’s not the density of the corridor that matters. Its the employment/population densities at the station stops. The Acela is relatively successful because it links NY, Philly, and DC, not because the corridor is densely populated.
His characterization of the estimates of the potential ridership of the 3-C corridor in Ohio is accurate and more critics should spend time going through these reports. In fact, I generally agree that the Ohio Hub is probably one of the most viable HSR routes (although that should not be implied as a personal or institutional endorsement of the project). The simulations generated for the ORDC, which I believe are about the best we can do in the U.S., suggest that if the entire Midwest Corridor is built out, and various feeder systems provide support for HSR, it might cover its operating costs. None of the reports suggest that system will come anywhere close to covering a significant portion of the capital costs, and Ohio is relying on re-dedicating some existing funds to the corridor’s development and the feds pumping the lionshare of the capital funds into the system. The reports are available from the Ohio Rail Development Commission’s web site.
The ridership and cost estimates for the 3-C corridor were made by Transportation Economics and Management System (TEMS) and they use an approach that is different from standard methodologies (also critiqued in a report commissioned by the ORDC) which uses a more comprehensive approach to benefits by analysis land use and economic development impacts.
Anyway you slice it, however, IMO HSR benefits are underwhelming under the best scenarios. They certainly don’t justify making HSR the cornerstone of US DOT policy.
For those interested, the post in the Daily Kos was targeted at a study Randal wrote for the Show Me Institute in Missouri, but replicated with state specific data for a number of other states including Indiana (Indiana Policy Review), Ohio (The Buckeye Institute), Colorado (The Independence Institute, and Florida (American Dream Coalition). Most of this work is based on Randal’s work for the Cato Institute and the two most relevant (and widely cited) studies can be found here and here.