Obfuscation Doesn't Change the Facts: Rail Transit Is Unreasonably Expensive

In a recent Wall Street Journal column, P.J. O'Rourke described the great expense and ineffectiveness of mass transit.[1] In the process, he referred to Heritage Foundation research on the exorbitant costs of urban rail (light rail, subways, and commuter rail).[2] Specifically, O'Rourke compared the cost of new rail systems to the cost of leasing riders luxury SUVs and found transit wanting. Transit advocates have challenged this point, but it still holds true.

For proof of the proposition, start by looking in almost any daily newspaper. The Newark Star-Ledger, for example, recentlycarried an advertisement for a new BMW X-5 luxury SUV, available for lease at an annual cost of less than $6,300.

Then compare that to the cost of rail. For example, the Federal Transit Administration reported in 2000 that each new trip on the Hiawatha light rail line in Minneapolis would cost a projected $18.57-and this was before the cost of the line escalated more than 50 percent. Using the pre-escalation number, the cost for each new annual commuter using the line two ways each work day (for 450 trips each year) would be approximately $8,400-or about $2,100 more than the cost to lease a BMW luxury SUV for the year. If that's not clear enough, see Chart 1 for a graphical comparison.

Responding to the O'Rourke column, Mr. Janek Kozlowski of Alexandria, Virginia, wrote a letter, which the Journal published, disputing the transit/luxury car comparison. Mr. Kozlowski bases his argument on the new Hiawatha line. His detailed calculations include the cost of leasing cars for all riders of the Hiawatha system, most of whom were previously bus riders, not drivers. Kozlowski adds in the cost of operating the cars-that is, gasoline and maintenance-and building new highway capacity-on the specious assumption that, without light rail, more highway capacity would be necessary. But if one thing is clear from the U.S. experience in light rail, it is that new rail systems have a minimal impact on traffic congestion. For example, Portland's expensive new light rail system has been accompanied by one of the largest increases in traffic congestion in a major urban area.

Mr. Kozlowski, like some other transit funding advocates before him, mistakes the transit/luxury car comparison as an actual policy alternative to building urban rail. But it is a rhetorical point, not a policy proposal. More often than not, the most prominent justification for building new rail systems is alleviation of traffic congestion, which requires getting drivers out of their cars. It is difficult to imagine a more absurd public policy approach than to spend more money to get a driver out of a car than to lease him or her a new one. Urban rail is an unreasonably expensive way to get cars off the road.

Mr. Kozlowski is not alone in attempting to refute the cost comparison. The same tactic has been used before in transit industry reports, perhaps to cloud the issue-and for good reason: transit's extravagant expense should be an embarrassment.

If, for example, the people of Minneapolis and St. Paul were to invest the same amount of money per driver-$8,400, as described above-to attract all drivers in the area to transit, they would have to spend $13 billion per year, almost as much as Minnesota's total tax revenue. Obviously, no urban area could afford such an expensive congestion reduction strategy.

And if the people of Minneapolis-St. Paul were to invest the same amount per daily trip to attract all trips from cars to transit, the cost would approach the gross personal income of the metropolitan area. If there were a Nobel Prize for irrational public policy, reducing traffic congestion with light rail would be the winner.

Of course, projections based on planning reports tend to be overly optimistic. In some cases, virtually no net transfer of ridership from cars to urban rail systems occurs at all. For example, the most recent expansions of the Dallas area DART light rail system have been accompanied by a reduction in transit ridership. Similarly, rail ridership is now lower after the doubling of the length of the St. Louis light rail line.

All of this demonstrates that it is extremely expensive to build urban rail systems. And more importantly, even the most grandiose systems have little or no impact on traffic congestion. That is because, whether in the United States or Western Europe, transit principally serves core urban areas and is not an affordable way to serve the non-core and suburban markets that represent the bulk of travel demand. This fact is well understood, if not readily admitted, by transit and urban planners, who blather on about transit and its ability to reduce traffic congestion. However, their actions speak louder than their words: no regional transportation plan anticipates transit improvements that would materially reduce automobile use. Case closed.

Wendell Cox is a Visiting Fellow at The Heritage Foundation and a visiting professor at the Conservatoire National des Arts et Metiers in Paris.