TRB FANTASIES:Light Rail Costs vs Road
Larwin provided no backing for this claim. He cant. Its ridiculous, yet it is precisely the kind of unfounded assertion that has many good people supporting construction of new light rail lines in US cities and thinking they will be a substitute for road improvements.
Light rail is generally vastly more expensive than an enlarged highway. USDOTs last HPMS database (Status of the Nations Surface Transp Systems Condition & Performance p289) shows new highway costs as $0.9m to $2.3m/lane-km. What USDOT calls high-cost additions average up to $3.2m/lane-km. 91-Express, a recent major highway enhancement in southern California cost $125m and consists of 65 lane-km so it cost about $2m/lane-km. 80 lane-km of HOT lanes in the median of US101 in Sonoma Co CA has been costed at $128m or about $1.6m/lane-km (TRnl#28 June 1998 p1). In San Diegos I-15 plans for expansion and lengthening of the current HOT lanes involve an additional 72 lane-km to form 32km of 3-lane with a moveable barrier. The project director expects the work to cost $200m to $300m, so theres a maximum of $4m/lane-km in Larwins backyard. (TRnl #25 March 1998 p1)
Wendell Cox who tracks light rail projects says they generally run above $12m/km and can easily run to $30m/km. Most of them have trouble attracting a maximum load but they are physically capable of carrying maximum hourly passenger loads of 3,000, about the same as a general purpose motorway lane. HOV/HOT lanes of course can carry more. With a decent volume of buses, as in Houston, 8k pas/hr is achieved and the Lincoln Tunnel busway has carried over 20k pas/hr.
Not only is highway expansion usually a fraction of the cost of light rail on the mainline but it costs far less to support since the same vehicles can peel off the mainline and go into local streets closer to peoples origins and destinations, reducing the need for feeder services, shuttles and transfer stations, that are such an expensive and user-unfriendly addition to any rail system.
A rail system is usually a riskier investment than an extra roadway lane because it is such an inflexible single purpose investment for hauling commuters, whereas a highway lane serves all kinds of traffic. The extra highway lane is a fractional addition to capacity and thus more easily used productively. Any surplus capacity in a dedicated road lane for bus and vans can always be sold off to single occupant motorists, while any surplus capacity on rail just lays there unused.
Cox writes: Light rail tends mainly to move former bus passengers. Studies have generally shown that only a quarter or so of rail riders were former auto drivers... This means that the best new US light rail systems remove 750 cars at peak hour from the road.
The only reasonable way to judge the cost effectiveness of either transport investment is the cost per auto capacity created (by diverting autos in the case of rail, or by adding capacity).
At 2,500 cars per hour, the capital cost per auto capacity created of two new motorway lanes is $2,100/km, Cox says. At 750 autos removed per hour, the capital cost of a light rail line would be $16k to $40k/km, by the same measure (6 to 16 times the cost of comparable highway capacity.) And thats just the beginning, Cox points out. Light rail fares will cover 25% to 60% only of operating costs and the vehicles that ride the rails are virtually custom-built and high cost. Roads and their vehicles by contrast can be self-financing.
COMMENT: TRB as a purportedly objective professional body, a part of the National Academies of Science, should not be a party to spreading such misconceptions about the relative cost of different transp modes. This certainly serves the special interest of the rail operators and system providers, but is a serious disservice to decisionmaking officials and the public. Larwin incidentally runs the light rail system in San Diego. One wonders about the credibility of TRB when it appoints such a misinformed, self-interested person to such a top position. Whats next: Philip Morriss top guy to the National Institutes of Health? Rupert Murdoch for FCC chairman? (see Wendell Cox at www.publicpurpose.com)