Clintonian Rail Follies:By Randal O'Toole

* $6.3 b on mass transit (a huge increase from past years);

* $1.6 b on congestion mitigation/air quality (vs. $1.0b last year);

* $719 m on "community transportation enhancements";

* $468 m on passenger rail;

* $52 m on a "Transportation and Community and System Preservation pilot program."

The administration also wants to give $25 m to Housing and Urban Development to promote ‘smart growth.’

Specific rail projects that the administration wants to fund include:

* Double-tracking Baltimore light rail;

* 6.6 miles of new heavy rail in Chicago;

* Extending a commuter rail line in Chicago;

* Denver's Southeast Corridor light rail;

* Extending Memphis's streetcar line;

* The Minneapolis Hiawatha Corridor light rail;

* The Hudson-Bergen light rail in New Jersey;

* Stage II of Pittsburgh light rail;

* Portland's North light rail;

* Operating Salt Lake City's light rail

* Seattle light rail; and

* Extending heavy rail in Washington, DC.

The amounts proposed for 2001 for many of these projects are small but they will commit the federal government to bs of dollars in the next few years. Recently, I was curious to compare the productivity of rail lines with highways. One best measure of productivity is passenger miles per mile of rail line or passenger miles per lane-mile of highway. Of course, if the rail lines did not exist, most of the passengers on many rail lines would ride buses. But passenger mile/mile is still a fair measure. These calculations are similar to ones done by Wendell Cox for light rail.

I started with data from the 1997 Federal Transit Administration's Transit Profiles for urban areas over 200,000 people http://www.ntdprogram.com/NTD/Profiles.nsf/Docs/1997over/$File/1997OverSet.pdf). This publication lists 48 light, heavy, and commuter rail lines and includes the miles of each line or set of lines plus the passenger miles carried. All but one of the 48 lines are in one of

the nation's fifty largest metropolitan areas.

For comparison I used the 1997 Highway Statistics, table HM-72, which lists freeway lane miles and vehicle miles traveled for each of the nation's 392 urbanized areas. This table is available at http://www.fhwa.dot.gov/ohim/ohimstat.htm. I assumed that each vehicle contained an average of 1.6 riders, which is the average estimated by the National Passenger Transportation Survey.

My results, summarized in the table below, indicate that only heavy rail comes close to achieving the productivities of a lane mile of freeway-and then only in New York City. Not counting New York City, the average heavy-rail line produced only 12,700 passenger miles of

travel per mile, about half what a typical freeway lane produces.

Table One: Productivity of Freeways and Rail Transit

(passenger miles per mile of rail line or lane-mile of freeway)

Light Rail Commuter RailHeavy RailFreeway

Average (1) 4,290 3,445 21,634 21,255

Average (2) 4,290 3,497 21,634 24,755

Average (3) 4,281 2,256 12,693 23,769

Highest 8,519 8,909 38,869 35,715

Highest city Portland New York New York Los Angeles

Second highest 7,323 4,183 16,280 28,323

Second city St. Louis Chicago Atlanta Chicago

1. 392 urbanized areas.

2. Fifty largest urbanized areas.

3. Fifty largest urbanized areas minus the highest.

Light-rail productivities are a small fraction of either heavy rail or motorways (freeways, expressways). The average motorway lane is three times as productive as the most productive light-rail line and six times as productive as the average light-rail line. Commuter rail is even less productive than light rail. If New York City commuter lines are excluded, the average commuter rail produces only 2,200 passenger miles per mile of rail line.

When cost effectiveness is considered, rail pales even further. The average light-rail line now being built or considered for construction in the US is expected to cost around $46m per

mile. But building a single lane-mile of motorway costs as little as $5m. In short, one dollar spent on motorway construction can produces as many benefits as $50 spent on light rail.

Heavy rail costs much more than light rail, commuter rail somewhat less depending on the state of the rails before being put into commuter service. And the cost of motorways can rise significantly in urban centers, where land costs are highest and mitigation of noise

and other impacts is most needed.

In general, then,

* Cities should spend money on heavy rail only if a mile of heavy-rail line costs less than half as much as a mile of motorway lane;

* Cities should spend on light rail only if a mile of light-rail costs less than one-sixth as much as a mile of motorway lane; and

* Cities should spend on commuter rail only if a mile of commuter-rail costs less than one-eleventh as much as a mile of motorway lane.

Anyone interested in my rail database in tab-delimited format should email me.

Note: People on the Urban Mobility email list who want to discuss this or other topics may want to subscribe to the transport-policy discussion group. This group was started by Wendell Cox but includes people on both sides of the rail-highway issue. To subscribe, go to http://www.publicpurpose.com.( Contact Randal O'Toole atThe Thoreau Institute rot@ti.org http://www.ti.org)

o Jonathan Richmond of Harvard’s Kennedy School who did an extensive study of recent urban rail projects said in no case was there evidence of a new rail line attracting sufficient ridership to have any measurable effect on road congestion.