Beware of Rail Transit Oversell
A Occasional Paper
Michael D. Setty • Leroy W. Demery, Jr.• Revised April 5, 2006
Copyright 2003–2007,
Rail transit is not a miracle catalyst for desired changes in transportation and urban development patterns. Necessary supporting policies stem from well-defined goals, objectives and criteria. The paper considers the minimum transport traffic required to justify rail transit. Low cost rail lines can be justified on the basis of unit operating cost savings, given a weekday traffic density greater than 5,000 passenger-miles per mile of route, e.g., an average of 5,000 passengers travel over each mile of route per weekday. Nationwide, corridors having a weekday traffic density of 5,000 or more are relatively rare. (By contrast, there are probably several hundred with sufficient patronage potential to justify enhanced bus strategies). Given adequate service levels, this threshold traffic density can be created even on small-scale operations. The threshold is not an absolute prerequisite, for site-specific issues must be considered. An initial stage may carry less than the threshold, but may be justified by potential on-line development and other desired land-use changes. In cases where rail development costs are low, lower traffic densities may be justified, but no savings over bus operating expenses should be expected unless pre-existing bus service is particularly unproductive. Full-scale analysis of alternatives may not be required, but reference to a minimum traffic density criteria helps build credibility. New rail lines are more likely to be judged “successful” by stakeholders if minimum traffic thresholds can be met.
Many U.S. cities have sufficient population, downtown employment, and an existing base of transit patronage strong enough to justify major investment for improved transit service quality and capacity. Transit can be a tool to help create livable communities, but such communities do not arise as the result of transit investment alone. Other tools and policies are needed to help insure that transit provides a viable transportation choice. Such policies include well-defined goals, objectives and evaluation criteria. Land-use and development issues are important, but transportation-related criteria should not be ignored. Lack of clarity regarding objectives and criteria – and lack of realism about what light rail transit (LRT), and commuter rail technology used in similar ways, as a tool can accomplish – are symptoms of “LRT (and rail) oversell.” This in turn may require extensive “damage control” efforts by project promoters following the opening of an “oversold” facility.
Threshold criteria to justify urban rail transit are well defined but, in the U.S., not as well known as they should be. An LRT or heritage streetcar line can be justified on the basis of unit operating cost savings if traffic density is sufficiently high. If traffic falls short of this threshold, the line may still be justified with reference to specific development objectives, positive economic impacts and so forth. This paper outlines systematic threshold criteria for low-cost LRT, and dis cusses how the threshold weekday traffic density can be attained even on a short starter line.
Traffic Density
Traffic density refers to the number of people who travel over each mile of line, on average, during each weekday (or some interval). For example, if an average of 5,000 people travel over each mile of line during a average weekday, then that line carries a “weekday traffic density” of 5,000. Technically, this refers to “5,000 passenger-miles per mile of route per weekday.” (Conversion to metric units does not change the number: 5,000 pass-mi per mi of route becomes 5,000 pass-km per km of route.)
This concept is easy to grasp and is used extensively in other countries. It is less familiar in the U.S., where weekday boardings and peak-hour volumes are more familiar measures.
Traffic density may be calculated as follows from three statistics: boardings, line length and average travel distance:
Traffic Density = (Boardings * Average Travel Distance) / Line Length.
“Boardings” are counted during the interval of interest (e.g. annual, daily or weekday).
“Boardings * Average Travel Distance” gives passenger-miles (or passenger-km). Therefore, one may simply divide total passenger-miles by the line length over which this travel occurs.
One complication is that traffic density cannot be related to a boarding count without reference to average travel distance (ATD) and line length. If line length is held constant, 5,000 passengers per weekday, ATD two miles, represents exactly the same transportation workload as 10,000 passengers per weekday, ATD one mile. Another, often overlooked in the U.S., is that longer ATD requires a higher number of vehicle-miles (or greater capacity per vehicle-mile in the form of larger vehicles) to serve a given number of “passengers” or “boardings per route-mile.” If these are not provided, then average vehicle occupancy will increase (more passenger-miles per vehicle-mile), and by implication peak-hour crowding will also increase.
The Rail Transit Threshold
A minimum threshold traffic density of 5,000 weekday passenger-miles per mile of route for low-cost LRT and similar low cost rail is supported by several studies from the U.S. and other countries, illustrated in Figure 1 below. (The weekday density threshold of 5,000 is equivalent to 1.5 million annual passenger-miles per mile of route.)
--A 1982 report by Pushkarev, Zupan and Cumella established 4,000 weekday passenger-miles per mile of route as the threshold to justify “a very low-capital light rail line” (Pushkarev et al. 1982, 256).
--About 1980, Japanese investigators established the traffic-density threshold to justify retention of rail passenger service at 4,000 daily passenger-miles per mile of route (Aoki et al. 2000, 168; Suga 1998, 3). This is roughly equivalent to 5,000 weekday passenger-miles per mile of route.
--Japan developed monorail and automated guideway transit (AGT) technology during the 1960s and 1970s. Criteria for application of various urban transport modes were then established. By 1980, the “domain of efficiency” for each mode was worked out, based on technical and economic factors. Nehashi (1988) implies that the minimum daily traffic density for rail transport is 4,000 passenger-miles per mile of route. Again, this is roughly equivalent to a weekday traffic density of 5,000.
--A report published in 1991 by the Berlin Senate (Berliner Senat, the city council) states that LRT is more economical to operate than buses if the traffic density exceeds 5,000 passenger-miles per mile of route per day (Ludwig 1991).
--The German firm Light Rail Transit Consultants GmbH (LRTC) is an engineering, operations and maintenance consulting firm organized by the engineering subsidiaries of four large German public transit authorities (Company Profile of LRTC). A handbook published by LRTC in 1993 states the minimum traffic density threshold for LRT at 4,000 passenger-miles per mile of route per weekday (Gerndt et al. 1993, 71).
--In 1948, as part of the litigation associated with the bankruptcy of the Pittsburgh Railways Company (PRCo), the City of Pittsburgh petitioned the court for an independent study on replacing all streetcar lines with buses. PRCo trustees objected but the court agreed. The study recommended replacement of about 25 percent of the streetcar network with buses (Lougee 1949). In response, the trustees proved that conversion of only one line could be justified on economic grounds(Tennyson 2004). The minimum utilization to justify rail retention implies about 2,700 passenger-miles per mile of track per weekday. This works out to about 5,000 passenger-miles per mile of route per weekday.
-The findings of Pushkarev et al. (1982) drew criticism, notably by Kain (1988) and Pickrell (1985). These are the only studies known to the authors that establish a threshold for low-cost rail transit significantly different from 5,000 weekday passenger-miles per mile of route (Figure 1).
The authors found that the Kain threshold is based on a calculation error (failure to convert from average weekday boardings to peak-hour, peak-direction passenger volume). The Pickrell threshold is based on a different analytical framework – benefit-cost analysis, eschewed explicitly by Pushkarev et al. (1982, 58). Results were not evaluated in the context of the original analytical framework and cannot be reconciled with results from the U.S., Canada and Western Europe. The authors reject the findings of Kain and Pickrell for these reasons.
Actual traffic density statistics for various U.S. and Canadian LRT systems are illustrated in Figure 2.1 – 2.2 below. The authors have chosen to present annual traffic density statistics in order to account for weekend ridership. Threshold weekday traffic density values (above) were annualized as appropriate. The authors chose to include heavy rail (HRT) systems in Los Angeles, San Francisco and Vancouver for the interest of those attending RailVolution 2004 in Los Angeles.
Building Weekday Traffic Density
It is possible to attain the threshold traffic density of 5,000 even on a short, small-scale rail transit facility such as Tacoma Link or the Portland Streetcar.
Tacoma Link, extending just 1.6 miles (2.6 km), shuttles passengers between the Tacoma Dome transit center and various destinations in downtown Tacoma. It was built as the eventual outer segment of a full-scale light rail line from Seattle. Early construction was justified by the need for a local connector between downtown and regional bus, Greyhound, Amtrak, “Sounder” commuter rail services and parking at the Dome. Another planning objective was to spur the ongoing revitalization of the Tacoma CBD.
Although Tacoma Link has just three cars, service is provided every 10 minutes throughout the day Monday-Saturday, and on Sunday afternoons. Service is provided free of charge. Weekday ridership has increased from 2,170 during the first full month of operation (September 2003) to 2,300 at January 2004. This implies a respectable annual growth rate of 18 percent.
The authors estimate that the average Tacoma Link passenger travels for about one mile (1.5 km). This implies a weekday traffic density of about 1,400, up from slightly more than 1,350 at opening. A weekday traffic density of 5,000, assuming a one-mile ATD, implies 8,000 boardings per weekday. This would occur after less than eight years given continuation of the 18-percent implied annual growth rate.
Extension of Sounder service south to Lakewood, scheduled for the end of 2007, and ongoing development in downtown Tacoma suggest that traffic potential for 8,000 Tacoma Link passengers per weekday may develop by 2011-2015. However, the current service could not accommodate the implied peak-hour volumes: 1,000 – 1,100 passengers per hour. This is so because prospective passengers are not likely to tolerate peak-period crowding at levels implied by 170-180 passengers per car. The authors estimate that additional peak-period capacity will be required once rideship grows to 4,000 – 5,000 per weekday. This will require either more frequent service or operation of two-car trains during peak periods. Either measure would require expansion of the Tacoma Link fleet from three to five cars.
One might argue that the traffic-density “threshold” for Tacoma Link should be significantly higher than 5,000 because the construction cost was not “low” – $80.4 million, or more than $50 million per mile. This, however, ignores the obvious – the line would have cost considerably less had it not been built to accommodate multi-car trains of full-scale light rail vehicles, with the overhead contact system designed for conversion to 1,500V dc.
The Portland Streetcar, which extends 2.5 miles (4.0 km) between the Good Samaritan Hospital and Portland State University, was built as a downtown circulator to connect major destinations and to stimulate development. Annual ridership increased from 1,365,000 during the first year of operation to 1,623,000 during the second year, an increase of 19 percent. Average weekday ridership is currently stated at 5,800.
The authors estimate that the average Portland Streetcar passenger travels a distance of 1.0-1.5 miles (1.5-2.5 km). This implies weekday traffic densities ranging from 2,300 to 3,500.
An extension from Portland State University to River Place, 0.6 mile, was completed and opened in late 2005. The authors estimated that this extension would attract sufficient ridership to bring average weekday boardings to about 7,000; actual results as of February 2006 were 9,000 total daily Portland Streetcar riders, implying a weekday traffic density of 3,500- 4,000. However, an increase in service frequency from the current 14 minutes to 10 minutes, as provided by Tacoma Link, would attract additional ridership. The authors estimate this increase in the range of 30 to 40 percent, which would generate a weekday traffic density in the range of 3,000 – 5,000. These estimates do not consider the effect of annual ridership growth, which occurred at a remarkably high rate during the initial operating period (above). As in Tacoma, Portland Streetcar vehicles can be coupled and operated as two-car trains should peak-period volumes require additional capacity. However, the current seven car fleet is not large enough to permit either 10-minute headways with sufficient maintenance reserve (at least two spare vehicles) or two-car trains on the busiest peak-period schedules.
The Portland Streetcar could be built as a low-cost project because tracks were not designed to accommodate full-scale LRT trains. The shallow (12-inch) slab minimized utility relocations, and use of relatively small, lightweight vehicles permitted the track to be fitted into existing grades, minimizing street and sidewalk reconstruction. The initial line cost $54.6 million ($22 million per mile), and the River Place extension is estimated to cost $18.2 million ($30 million per mile), including two additional cars.
The authors emphasize that the success or failure of a small-scale rail transit facility cannot be evaluated with reference to traffic density alone. In cases where rail development costs are low, lower traffic densities may be justified, but no savings over bus operating expenses should be expected unless pre-existing bus service is particularly unproductive.
In Memphis, the average passenger carried by the Main Street Trolley and Riverfront loop travels a very short distance – 0.75 mile. Weekday traffic density is therefore very low – less than 1,100. However, the Trolley’s significant contribution toward a major planning goal, downtown revitalization, is widely recognized. Perhaps less well known are the remarkable operating results achieved by the Trolley. During 2002, the Trolley accounted for 17 percent of all Memphis Area Transit Authority (MATA) boardings but required only seven percent of total operating expense. On a per-mile basis, the average Trolley passenger paid more than double the amount paid by the average bus passenger. Trolley operating expense per revenue service hour was 18 percent less than for MATA bus services in spite of lower speed (and lower labor productivity in terms of revenue miles per revenue service hour). In addition, the Trolley carried more than twice as many boarding passengers per revenue service hour than MATA buses. This is a remarkable record for a downtown circulator system, and contrasts sharply with results provided by capital-intensive facilities in Detroit, Jacksonville and Miami. On a revenue service hour basis, MATA Trolley services attract 10-55 percent more boardings and cost 40-70 percent less to operate than the three U.S. downtown peoplemovers.
The threshold traffic density for low-cost rail transit, about 5,000 weekday passenger-miles per mile of route, is supported by several studies from the U.S. and other developed economies. This concurrence of results reflects analysis based on relationships between physical quantities: space, time, labor, energy and land. Although prices for inputs may vary among regions and countries, the underlying relationships are similar.
Even a small-scale rail transit facility may carry 5,000 weekday passenger-miles per mile of route given adequate service. This in turn requires a vehicle fleet sufficient to operate frequent headways and to provide additional peak-period capacity as needed. However, a successful downtown circulator system may carry a low traffic density if the average passenger travels for a very short distance.
It is also important to remember that LRT and other forms of low cost rail are not the “transit service” per se, but the means of providing that service. A corridor without sufficient traffic density to justify the investment required for rail may have sufficient density to justify enhanced bus or bus rapid transit (BRT) facilities. Nationwide, corridors having a weekday traffic density of 5,000 or more are relatively rare. By contrast, there are probably several hundred with sufficient patronage potential to justify enhanced bus strategies.
The authors express sincere gratitude to E. L. Tennyson, P.E., former Transit Commissioner, City of Philadelphia and former Deputy Secretary of Transportation, Commonwealth of Pennsylvania, for information and insights provided during preparation of this paper.
American Public Transportation Association. Transit Statistics.
Aoki, Eiichi, Mitsuhide Imashiro, Shinichi Kato and Yasuo Wakuda. 2000. A History of Japanese Railways 1872-1999. Tokyo: East Japan Railway Culture Foundation.
Apel, Dieter. 1990. Straßenbahn. Stadtbahn. Light Rail Transit. Archiv für Kommunalwissenschaften Band II. Berlin.
APTA Heritage Trolley and Streetcar Site.
Company Profile of LRTC.
Gerndt, Helmut, J. Hansen and W. Kokot. 1993. Light Rail Transit. Düsseldorf: Light Rail Transit Consultants GmbH.
Kain, John F. 1988. “Choosing the Wrong Technology: Or How to Spend Billions and Reduce Transit Use.” Journal of Advanced Transportation, 21, 3 (Winter 1988): 196-213.
Light Rail Central.
Lougee, N. A. & Company. 1949. Report on “Advisability of Substituting Bus Service In Whole or In Part for the Street Car Service Presently Being Undertaken” for W. G. George and Thomas Fitzgerald, Trustees, Pittsburgh Railways Company, Debtor [and] Pittsburgh Motor Coach Company, Subsidiary. New York.
Ludwig, Dieter. 1991. Integration der Tram in der Stadtstruktur Berlins. Straßenbahntagung Berlin 1991. Berlin: Berliner Senat.
National Transit Database (Federal Transit Administration, U.S. Department of Transportation).
Nehashi, Akira.  "New Urban Transport Systems Reconsidered" Japan Railway & Transport Review No. 16, June 1998.
North American Vintage Trolley Systems (John Smatlak).
Pickrell, Don H. 1985. Urban Rail in America: A Review of Procedures and Recommendations From the Regional Plan Association Study. Cambridge, MA: Transportation Systems Center, Research and Special Programs Administration, U.S. Department of Transportation.
Pushkarev, Boris S.; Jeffrey M. Zupan, and Robert S. Cumella. 1982. Urban Rail in America: An Exploration of Criteria for Fixed-Guideway Transit. Bloomington, IN: Indiana University Press.
Rode, Phillipp. The Case for Light Rail Transit on Manhattan’s East Side. 1999. New York: Institute for Rational Urban Mobility, Inc., and Department of Track and Railway Operations, Technical University Berlin.
Stadtbahn in Deutschland / Light Rail in Germany. 2000. Düsseldorf: Verband Deutscher Verkehrsunternehmen (VDV; Association of German Transport Undertakings), VDV-Förderkreis e. V. (VDV Promotional Group) and Bundesministerium für Verkehr, Bau- und Wohnungswesen (BMVBW; Federal Ministry of Transport, Building and Housing).
Suga, Tatsuhiko. 1998. “Are Trains Almighty?” Japan Railway & Transport Review, 18 (December 1998).
Tacoma Link Light Rail (Sound Transit).
Tacoma Link: The Little Tram That Could. Light Rail Progress.
Tennyson, E. L. 2004. E-mail correspondence dated January 22, 2004 and January 24, 2004.