New U.S. light rail transit starter systems — Comparative total costs per mile

LEFT: LA Blue Line train emerging from tunnel portal. (Photo: Salaam Allah.) RIGHT: Norfolk Tide LRT train on single-track railroad roght-of-way. (Photo: Flickr.)

LEFT: LA Blue Line train emerging from tunnel portal. (Photo: Salaam Allah.) RIGHT: Norfolk Tide LRT train on single-track railroad right-of-way. (Photo: Flickr.)

This article has been updated to reflect a revision of the LRN study described. The study was revised to include Salt Lake City’s TRAX light rail starter line, which was opened in late 1999.

What’s been the been cost per mile of new U.S. light rail transit (LRT) “starter systems” installed in recent years?

The Light Rail Project team was curious about this, so we’ve reviewed available data sources and compiled a tabulation comparing cost-per-mile of “heavy-duty” LRT starter systems installed in or after 1990, all adjusted to 2014 dollars for equivalency. (“Heavy-duty” distinguishes these systems from lighter-duty streetcar-type LRT projects.)

This is shown in the figure below, which presents, for each system, the year opened, the initial miles of line, the cost per mile in millions of 2014 dollars, and comments on significant construction features. (“RR ROW” refers to available railroad right-of-way; “street track” refers to track embedded in urban street pavement, almost invariably in reserved lanes or reservations.)


Major data sources have included TRB/APTA 8th Joint Conference on Light Rail Transit (2000), individual LRN articles, and Wikipedia.

Averaging these per-mile cost figures is not meaningful, because of the wide disparity in types of construction, ranging from installation of ballasted open track in railroad right-of-way (lowest-cost) to tunnel and subway station facilities (highest-cost). These typically respond to specific conditions or terrain characteristics of the desired alignment, and include, for example:

Seattle — While Seattle’s Link LRT is by far the priciest system in this comparison, there are explanatory factors. Extensive modification of existing Downtown Seattle Transit Tunnel (and several stations) previously used exclusively by buses; tunneling through a major hill, and installation of a new underground station; extensive elevated construction to negotiate hilly terrain, major highways, etc.

Dallas — This starter system’s costs were pushed up by a long tunnel beneath the North Central Expressway (installed in conjunction with an ongoing freeway upgrade), a subway station, a new viaduct over the Trinity River floodplain, and significant elevated construction.

Los Angeles — The Blue Line starter system included a downtown subway station interface with the Red Line metro and a short section of subway before reaching the surface of proceed as street trackage and then open ballasted track on a railroad right-of-way.

St. Louis — While this system’s costs were minmimized by predominant use of former railroad right-of-way, a downtown freight rail tunnel was rehabilitated to accommodate the system’s double-track LRT line, with stations; an existing bridge over the Mississippi River was adapted; and significant elevated facilities were installed for access to the metro area’s main airport.

Hopefully this cost data may be helpful to other communities, in providing both a “ballpark” idea of the unit cost of new LRT, and a reality check of any estimated investment cost already rendered of such a new system. ■

Cases where voters okayed rail transit after first rejecting

Rail transit ballot measures are critical events. But if one is rejected, is it a "catastrophic" for the community? Graphic:

Rail transit ballot measures are critical events. But if one is rejected, is it a “catastrophic” setback for the community? Graphic:

Voter rejection of a rail transit project is almost always unfortunate.

But is it catastrophic? Does it signal that the majority in a community will persistently and permanently reject any rail project, or does it represent a more temporary setback, with remaining hope that a better plan, a better presentation to voters, at a better time, could have a chance to win approval?

This issue often arises not only in communities where a rail transit project has unified support from transit advocates, but even in cases where an official plan has faced strong opposition from rail transit supporters. In an effort to mobilize support, proponents of the given project may argue that it may be the community’s “only chance for rail”, that, no matter its deficiencies, a given plan cannot be allowed to fail, because it would be a “disaster”, setting back rail development for decades, perhaps forever.

To evaluate the validity of this argument, and assess the actual delay between the failure of rail ballot measures and the ultimate passage of support for a subsequent rail transit ballot initiative, the LRN Project team examined available cases since 2000 where an initial rejection of rail was followed by a successful later vote. LRN’s approach has examined this issue strictly from the standpoint of attracting voter support — in other words, if the issue of rail transit is re-voted, how long does it take to win approval?

It should be noted that this study has examined the sequence of events only in cities where, after the failure of an initial measure, a new measure for rail transit (often with a somewhat different plan) was offered to voters. In other cases, poorly prepared or presented rail plans were rejected by voters, but rail planning was subsequently dropped (e.g., Spokane, Columbus) or has proceeded without needing a public vote (e.g., San Antonio).

Thus this study has sought to address the question: If rail has previously been rejected by voters, but a new rail measure is subsequently presented for a vote, how long does it take to achieve successful voter approval for rail?

Since 2000, there have been six cases where such re-votes have occurred:

Austin — A plan for a light rail transit (LRT) system was very narrowly defeated in 2000; rail transit was subsequently repackaged as a light railway using diesel-multiple-unit (DMU) rolling stock, and passed in 2004 (now branded as MetroRail). Delay between votes: 4 years.

Kansas City — An officially sponsored LRT plan was defeated in 2001; a different LRT plan initiated by a citizens’ referendum was subsequently approved in 2006. (However, the successful vote was annulled by the city council; implementation of an officially sponsored streetcar project is now underway without a public vote.) Delay between votes: 5 years.

Cincinnati — An LRT plan was rejected in 2002. Rail transit was subsequently repackaged as a streetcar plan that was forced to a public vote, and ultimately was approved in 2009. (A re-vote, forced by opponents’ referendum, was held in 2012, and the streetcar project again passed.) Delay between votes: 7 years.

Tucson — An LRT plan was rejected in 2002; rail transit was subsequently repackaged as a streetcar plan, then submitted for a public vote and approved in 2006. (The new system, branded as Sun Link, is due to open later this year.) Delay between votes: 4 years.

Seattle — A multi-modal transportation proposal, Roads and Transit, including LRT expansion, was defeated in 2007 (with opposition from environmental organizations and other traditional pro-transit groups, dissatisfied with the plan’s heavy highway element). A new package, Sound Transit 2, was prepared, with much heavier transit emphasis, and presented and approved by voters in 2008. Delay between votes: 1 year.

St. Louis — Proposition M, including funding for the region’s MetroLink LRT system, was defeated by voters in 2008. A new package, Prop. A, aided by an improved campaign, and including funding to improve and expand LRT, was subsequently approved in 2010. Delay between votes: 2 years.

From these experiences, it’s plausible to conclude the recent re-votes on rail transit have taken from one to seven years to succeed. This would not seem to suggest that initial loss of a vote results in a “catastrophic” delay of “decades” before a rail transit project can muster approval.

On the contrary, the average delay, on the basis of these cases, is 3.8 years. However, the data seems to suggest a pattern, whereby the delay before a successful rail transit re-vote is less in cities already operating some form of rail transit (Seattle, St. Louis), in contrast to cities where rail would be a totally new addition to the transit mix (Austin, Tucson, Kansas City, Cincinnati). This differential in average delay is illustrated graphically in the chart below:

Left bar: Average years of delay in cities already operating rail transit. Right bar: Average delay in cities with no current rail transit.

Left bar: Average years of delay in cities already operating rail transit. Right bar: Average delay in cities with no current rail transit.

Other than to infer that the loss of a vote does not inevitably represent a “catastrophic” setback for rail transit in a given city, this study with its very small data set does not offer a basis for strong conclusions. However, there is opportunity for plausible speculation:

• Conditions for a more speedy re-vote and approval of a rail transit ballot measure may be more propitious in communities that already have experience with successful rail transit systems.

• The process of re-submitting a rail transit measure to a vote may depend not so much on public attitudes but on the determination of sponsoring officials, their responsiveness to public input, and their willingness to re-craft specific project details to more closely conform to public needs and desires.