January 23, 2005
The Federal Transit Administration has posted draft 2002 transit data to the National Transit Data Base web site. You can download these data tables in Excel format.
I have extracted some of the more interesting data for rail transit lines and put them in a downloadable Excel file. These data include:
Although I included both miles and directional route miles, directional route miles is the one to use. A mile of rail line that has travel in both directions, either on one track or two tracks, represents 2 directional route miles. Caution: When you hear that a particular rail line cost $40 million a mile, the cost usually includes track in both directions, so it is really only $20 million a directional route mile.
The file also includes calculations of the following:
Finally, the file includes averages of most of these numbers by mode (light rail, heavy rail, etc.).
Don't forget that directional route miles often only represent half of a bidirectional rail line. Rail passenger miles as a percent of a freeway lane only counts rails in one direction. If freeway lane construction costs $10 million a mile and rail construction costs $40 million a mile, halve the rail construction cost when comparing costs and passenger miles.
For example, each mile of San Diego's light-rail system carries 15 percent of a freeway lane mile. San Diego's light-rail lines cost $30 million a mile to build, meaning they cost $15 million a directional route mile. If freeways cost $10 million a lane mile, then they are (15/10)/0.15 = 10 times as cost effective at moving people than light rail.
Here are a couple of tentative conclusions based on a comparison of these data with the 2001 numbers:
Light-rail systems in Boston, Buffalo, Cleveland, Philadelphia, Pittsburgh, Sacramento, San Diego, San Francisco, and San Jose all carried fewer riders in 2002 than in 2001. In some cities, such as Dallas, light-rail ridership grew only because they opened new lines. In Dallas, Denver, and other cities, the growth in light-rail ridership was nearly balanced by a decline in bus ridership.
In addition, ridership dropped on commuter-rail systems in Chicago, Dallas, Ft. Lauderdale, and San Francisco and on heavy-rail systems in Atlanta, Chicago, Cleveland, New York, Philadelphia, and San Francisco. In many cases the declines were small, but highway driving increased in nearly all of these cities, indicating transit is losing more market share.
Transit agencies that build rail lines can be particularly hard hit by a recession because they usually must make huge fixed annual payments on bonds. They also feel compelled to offer frequent rail service even if ridership drops. Thus, they often respond to recessions by raising fares or making major cuts to bus service.
An October, 2003, report from the American Public Transportation Association says that the recession has had "substantial negative impacts on the transit industry." A copy of this memo can be downloaded from americandreamcoalition.org/Transitdeclines.pdf.
In 2001, New Jersey's Hudson-Bergen rail line was the nation's worst-performing light-rail operation. Ridership on Hudson-Bergen increased, but San Jose's light-rail ridership declined more than 15 percent in 2002. This allowed San Jose to surpass Hudson-Bergen as the nation's worst-performing light-rail line as measured by cost per rider, vehicle occupancies, and passenger miles per route mile.
San Jose's transit agency, the Valley Transit Authority, is funded by one sales tax dedicated to operations and another sales tax dedicated to capital improvements. Due to a funding crisis, the agency decided to postpone rail construction and divert capital improvement funds to operations.
A mile of San Jose's light rail carries only 5 percent as many people as a San Jose freeway lane mile. San Jose also has a commuter rail line that carries only 2 percent as many people as a freeway lane mile. Even considering that these numbers might improve with economic recovery, any further investments in rail transit in San Jose are pretty foolish.
Much more information can be gleaned from these tables. Feel free to email me if you have any questions.