Why are all of these people on the crowded freeway when they could be riding the light rail? Because, even in bumper-to-bumper traffic, cars are usually faster and more convenient than rail transit.
Moreover, the costs of each new light-rail line are rapidly escalating. The eastside line was built at a cost of about $14 million per mile. The westside line is now under construction at a cost of $55 million per mile. Costs for the south-north lines are expected to be around $100 million per mile--and, given the experience of the first two legs, will probably be much more.
As in many cities, Portland's original transit system consisted of streetcar lines built in the 1880s through the 1920s. All were privately built, often by real estate subdividers who wanted to entice people to buy lots in suburban estates by giving them the most modern transit available to those estates. By 1920, most of the lines had consolidated into one company, Portland Traction.
As auto ownership increased, streetcars became less profitable. By the late 1950s, the company had replaced all of the streetcar lines with buses and renamed itself Rose City Transit. In 1960, buses carried 11 percent of all Portland commuters. But transit use continued to decline so that by 1969 well under 5 percent of all trips used transit and the company was nearly bankrupt.
Starting in the mid-1960s, the federal government encouraged cities to take over local transit companies. The Urban Mass Transit Act of 1964 offered funds for capital improvements with the hope that they could reduce pollution and congestion by luring people out of their cars and back to transit.
With Rose City Transit threatening to go out of business, Oregon accepted the federal invitation and created transit districts to take over bus service in Portland and Eugene. Portland's district covered three counties, so it was named the "Tri-County Metropolitan Mass Transportation District" or Tri-Met for short.
During the 1970s, Tri-Met used federal funds to make significant improvements in service. These included buying a new fleet of buses, constructing a bus mall downtown, high-occupancy vehicle lanes on freeways, doubling of service on several routes, and offering park-and-ride stations in the suburbs. These service improvements, combined with increases in the price of gasoline, increased ridership by 12 percent per year to more than 250 percent of the 1970 levels. By 1980 Tri-Met was carrying 10 percent of the region's commuters--not quite as much as in 1960 but much more than in 1970. Each rider lured out of his or her car cost taxpayers about $1.
But many people argued that this was not enough. Streetcars--now called light rail--were not only cheaper and less polluting than buses, but they were said to be more likely to lure middle-class drivers out of their cars. In 1973, the state of Oregon did a study that seemed to confirm some of these ideas.
They study looked at running light-rail cars on existing rail lines. Three of the routes studied were little-used or recently abandoned rail lines that had once been streetcar routes. Two were heavily used freight lines. The study concluded that it would be feasible to restore light-rail service to the three former trolley lines and one of the freight lines. Moreover, the total capital cost, including the vehicles, would be just $53 million.
Portland's decision to cancel the Mt. Hood freeway freed up a lot more than $53 million. So, instead of restoring service to existing routes, the city decided to build an entirely new line. Since the federal government was paying for most of the construction, there was little incentive to contain costs, which soon ballooned to many times the initial estimates.
Tri-Met planned and built the first light-rail line, but Metro took over planning for later lines. Metro cared even less about costs than Tri-Met. In fact, Metro views light rail not as a form of transit but as a development tool: Light rail "is not worth the cost if you're just looking at transit," says Metro's John Fregonese. "It's a way to develop your community at higher densities."
When Metro was planning the second light-rail line, it estimated that building the line an extra five miles through farm land--at a cost of $175 million--would increase transit's share of travel by 0.35 percent. Metro decided to do it so it could develop the farms into high-density residential areas.
Metro estimates that extending the third light-rail line an extra five miles to Portland's most popular shopping mall--at a cost of $455 million--would increase ridership by only 600 people per day. Yet this is Metro's preferred alternative because it wants to redevelop the area around the shopping mall to New Urban specifications.
Because light rail will consume most of Portland's transportation budget, Metro's plans are more likely to bankrupt the region than they are to reduce congestion and auto use.
A close look at the eastside line, however, raises serious doubts about the value of additional routes. First, Tri-Met can claim that the eastside line was "on time, under budget, and exceeded ridership projections" only by comparing results with projections made shortly before the line opened--projections that had changed dramatically from the original estimates.
At the time the decision was made to build the line, Tri-Met estimated it would take three years and cost $135 million to construct and that it would carry 42,500 people per day within five years of completion. The actual figures turned out to be four years, $214 million, and 21,000 riders per day after five years--55 percent higher costs yet only half the projected ridership. In 1996, ten years after opening the line, ridership is still less than 75 percent of the projection for five years ago. Operating costs also turned out to be significantly higher than originally projected.
One reason why ridership is poor is light rail's slow speed. The eastside line travels in the streets for all but 4 miles of its length, so its speed averages less than 20 miles per hour. Cars can easily beat that even in bumper-to-bumper traffic. Light-rail cars take nearly 45 minutes to go the entire 15 mile route. Tri-Met's express buses go 10 miles or more in 15 minutes. Since Tri-Met eliminated express buses parallel to the light-rail route, the line actually increased the trip time for many riders.
When the eastside line was being planned, Tri-Met estimated that building an exclusive bus lane instead would attract 13 percent more riders than light rail at a significantly lower cost--mainly because the buses could go faster. But Tri-Met preferred light rail's higher capacity; if ridership exceeded projections, the buses might not be able to cope. The planners never considered the possibility that ridership might fall short of projections.
It is likely that the strongest--though rarely spoken--argument for light rail was precisely that it was more expensive than other alternatives--an expense paid mainly by the federal government. The role of light rail as pork is underscored by an eight-page light-rail brochure currently distributed by Tri-Met. The brochure barely mentions ridership or congestion reduction. But it prominently features the number of jobs created in light-rail construction and the number of federal dollars flowing into the area--most of which, it proudly announces, stayed in the region.
Savings in light-rail operating costs were supposed to make up for its higher capital costs. Tri-Met claims that light rail's operating costs per rider are lower than for buses. But Tri-Met averages the costs of the poorly utilized feeder lines serving the light rail into other bus costs; if light-rail costs are compared only against the long-distance bus routes similar to those that light rail replaced, then buses are often less expensive. Plus, the major costs of maintaining the roadbed and overhead electric lines won't appear on Tri-Met's budget for several more years.
At the same time as the light rail was built, the state added two new lanes to 4 miles of a freeway that parallels the light-rail line. Though this complicates any calculations of the effects of light rail on congestion, the available evidence suggests that light rail has had a negligible effect on congestion.
Between 1977 and 1982, freeway traffic was stagnant at just over 100,000 vehicles per day. After the light rail was built and the freeway expanded, freeway usage grew steadily to more than 160,000 vehicles per day. In early surveys, Tri-Met was disappointed to find that three-quarters of light rail users had previously been bus riders--a much higher share than projected. It claims that increases in ridership since then represent people leaving their cars behind, but light rail's ridership gains have been slower than the population growth of the area it serves.
If the light rail carries an average of 26,000 riders per day, and if a third of those riders would otherwise have taken cars at an average of 1.2 riders per car, then the light rail keeps about 7,200 cars off the freeway. For half the cost of the light rail, the additional freeway lanes increased the freeway's capacity by a third to a half, while the light rail reduces freeway traffic by less than 5 percent. Metro and Tri-Met officials like to say that one light-rail line has the capacity to carry several lanes of freeway traffic. But capacity isn't relevant if people don't use it.
Moreover, since 1990 nearly 90 percent of the growth of light rail's ridership has been during the nonpeak periods--between 9 am and 3 pm or after 7 pm. Ridership during the peak hours increased by only 1 percent. Ridership actually declined between 7 and 8 am, 3 and 4 pm, and 5 to 7 pm. This is in spite of the fact that, according to Tri-Met reports, light-rail cars rarely, if ever, operate to capacity even during rush hour.
Light rail also does almost nothing to reduce pollution. Although a third of its riders may have otherwise driven a car, more than a third of its riders drive or are driven to the light-rail station. Since most auto pollution is emitted when engines are cold, starting your car to drive to the station pollutes just about as much as driving all the way to work.
Construction of the light rail was accompanied by an overall decline in Tri-Met ridership from 40 million riders per year in 1980 to 35 million in 1986. By 1990, ridership had rebounded to 40 million, but the share of commuter and other Portland-area trips using mass transit had fallen from 10 percent in 1980 to 6 percent in 1990. It fell even further by 1995.
Part of the reason for the decline is the general fall in gasoline prices during the 1980s. But Tri-Met's preoccupation with light rail--which replaced only three of about 100 Portland bus routes--rather than making improvements to a broad number of transit routes probably contributed. Transit's share of commuter traffic also fell in other cities, such as Seattle, that didn't build light rail--but not by as much as Portland's.
In the 1970s, Tri-Met got cars off of the highways at a cost to taxpayers of less than $1 per car. The eastside light rail costs taxpayers at least $10, and probably much more, for each auto trip it saves. Portland's westside line, now under construction, and south-north lines, now in the planning stages, will cost considerably more than that.
In 1979, Metro estimated that a westside light-rail line would cost $175 million (about $350 million in today's dollars) and carry more than 92,400 riders per day. Metro also examined the exclusive busway option, and estimated it would carry 92 percent as many riders at 60 percent of the capital costs. Since the federal government was paying 75 percent of the capital costs, Metro decided to go for light rail.
The 17.6-mile line is now under construction and will open for service in 1998. Costs have ballooned to $944 million, 2.7 times as much as the original projections. Part of the increase came from a decision to bore a three-mile-long tunnel--actually, two tunnels, one for each direction of traffic--through the Tualatin Hills just west of downtown Portland. This option was made even more expensive by a station in the middle of the tunnels, 260 feet below the surface, to serve the Portland zoo.
Ray Polani, long the loudest voice for light rail in Portland, believes that the twin tunnels and zoo entrance were bad decisions. "The tunnel has a 5 percent grade, which is extremely steep for rail," he says. "Projections indicate that only a few people will use the zoo entrance, yet it is adding enormously to the cost."
Another added expense was a decision to extend the line from Beaverton to Hillsboro. This was expected to add $175 million to the cost but add only 420 transit riders per weekday. This would increase transit's share of total traffic by just 0.1 percent--at a cost to taxpayers of at least $170 for each auto trip saved by the light rail (calculated by amortizing construction costs over 50 years at 10 percent and assuming 290 weekday trips per year and 1.2 people per auto).
Planners had to strain to justify the expense of the Hillsboro extension. Without building the extension, they projected that 27.8 percent of Hillsboro commuters to Portland would ride the bus to Beaverton and take the light rail from there. By making less capital intensive improvements to bus service, they estimated they could increase this to 28.2 percent. The light-rail extension, at an estimated cost of $175 million, would boost this all the way to 28.3 percent. Metro documents proudly note that this is "the highest of any of the alternatives," without mentioning that it also requires, by far, the highest subsidy per rider.
These and other costly changes to the original plan led urban studies professor Kenneth Dueker to describe Portland's light rail as "gold plated." It is certainly far removed from the original conception of a low-cost alternative to buses that would use existing railroad tracks.
Metro also projected that total ridership on the westside line, including the Hillsboro extension, would be less than 40,000 per day by 2005. That's well under half the 1979 projections, made without the Hillsboro extension. Since the highway paralleling this line carries about 86 percent as many vehicles as the eastside line, actual ridership will probably be the same as, or a little less than, the eastside's 26,000 passengers per day.
Unlike the eastside light rail, which passed through urbanized areas, the Hillsboro extension will have several station stops in areas that are farmers' fields today. This is giving Metro and local planners the opportunity to design and zone for neotraditional neighborhoods that will be built from scratch. This opportunity is probably why Metro decided to spend another $175 million on a line that would carry only a few people. Several such developments have been planned or are under construction.
The federal government paid 83 percent of the capital costs of the eastside line and 75 percent of the westside line. The state paid 12 percent of the costs of the eastside and 10 percent of the westside. Thus, Portlanders only had to pay about $10 million for the eastside line and about $125 million for the westside line.
So when Metro proposed a new, 29-mile line from Vancouver, Washington to Oregon City, many Portlanders responded with enthusiasm. The Oregonian wrote that no light rail would mean "Los Angeles-style and Seattle-style congestion, high-occupancy lanes, more car-pooling, more buses, more and more roads." In 1994, when Metro asked Portland-area voters if they wanted to help pay for a south-north light rail, 63.5 percent said "yes."
Portland State University's Kenneth Dueker contends that light rail achieved this landslide support because it "is the perfect `feel good' vote." People who don't expect to ever use it vote for it to demonstrate their "commitment to do something good about sprawl and congestion."
Voters also probably failed to understand that the south-north light rail would cost them much more than even the westside project. Initial estimates--which are always low--were that it would cost $2.85 billion--almost $100 million per mile. Moreover, the federal government would only pay half the costs, leaving Portlanders with a larger share of a much larger bill.
The plan voters approved in 1994 was to have Portlanders pay a third--$475 million--of the non-federal costs while the state paid another third and Vancouverites paid the rest. (It is called "south-north" rather than "north-south" because Oregon politicians figured that those who were paying more of the costs should be more prominently featured in the title.)
Two roadblocks appeared immediately. First, Vancouver voters turned down the chance to pay their share in a 1995 election. While Metro believed that a viable light-rail line could stop at the river dividing the two states, it decided to make the south line its first priority.
The second problem was that the Republican-dominated Oregon legislature wasn't interested in asking downstate Oregonians to pay for a Portland transit system. The 1995 legislature adjourned until 1997 without agreeing to the project.
Federal funding for Portland's light rail has been made possible mainly by Mark Hatfield, the chair of the Senate Appropriations Committee. When the legislature adjourned, Hatfield had not yet announced his retirement from the Senate. But he called Oregon's governor, John Kitzhaber, and let him know that the senator might not be around much longer, so Kitzhaber had better get the light rail approved while he had a chance.
Kitzhaber called a special session of the legislature and agreed to go along with several "anti-environmental" bills that he had previously vetoed if the legislature would fund light rail. As a further bribe, Metro agreed to give several hundred million dollars of Portland highway funds to downstate cities and counties. After much controversy and debate, the legislature agreed.
Light-rail opponents collected enough signatures to bring the state funding question to all Oregon voters in November, 1996. Portlanders will probably support it, but they make up less than half the state's voters. If voters reject it, federal funding may evaporate since Hatfield will be out of the Senate in 1997 and Congress is reducing funding for urban mass transit.
The south-north line could be built at substantially lower cost by staying on the east side of the river and avoiding downtown Portland--saving both the cost of a bridge and more than $250 million in construction costs downtown. Riders who wanted to go downtown could transfer to the eastside line, which crosses the river on a bridge that still had trolley tracks on it when the eastside line was constructed. But downtown interests wanted the new line to go through their area.
Ironically, major downtown businesses lobbied to not have the line routed their direction. Hotels, bookstores, breweries, and restaurants all argued that they would be better off if the line went somewhere else. So the line is being routed down Portland's existing bus mall, where businesses have already "adjusted" to transit--meaning buildings along the mall have among the highest vacancy rates in downtown Portland.
The routing of the south line is also plagued by politics. Rather than send the line to Oregon City, as originally planned, Metro now wants to build the line from Milwaukie to Clackamas, the area's most popular shopping center. At just 600 people per day and a projected construction cost of $455 million, that means a tax subsidy of $315 per rider--not counting operating costs.
The highway from Portland to Clackamas carries less than a quarter of the traffic than the highway paralleled by the eastside light rail. So a Portland-Clackamas light-rail line is not likely to carry much more than a quarter as many passengers as the eastside line. Even Ray Polani, Portland's first light-rail supporter, thinks that the Clackamas line is a mistake.
An alternative route from Clackamas to the airport, crossing the eastside line in east Portland and with possible extensions to Oregon City and Vancouver, was rejected because it didn't go downtown. Most of the necessary right of way is already public, and a 1987 study estimated total right of way costs of just $1 million and total construction costs of under $100 million. The highway paralleling this route carries three times as many vehicles as the one paralleling the planned route.
The north line would carry more riders than the south line, but not as many as the east- or westside lines. The highway that it parallels carries about 75 percent as much traffic as the highway paralleling the eastside line and just over 90 percent of the traffic as the westside line. A push to build the line into Vancouver may be particularly misguided: Fewer than 5,000 people commute each day from Vancouver to downtown Portland.
In sum, the south-north line is a huge boondoggle. Just how huge can be seen by what the region won't be able to afford to build if it builds the light rail. That information, however, is buried in an appendix of Metro's Regional Transportation Plan.
The document is a regional transportation plan that Metro prepared in response to the Intermodal Surface Transportation Efficiency Act of 1991. The law links federal transportation spending to the Clean Air Act Amendments of 1990. Regions failing to meet clean air targets (which include Portland) must gear their transportation planning to improving air quality as well as mobility.
Metro's transportation plan --which is an "interim" plan until the 2040 plan is done--is naturally based on New Urban principles, including (in the order in which they are listed in the plan):
The plan doesn't explicitly say so, but it seems that congestion is actually Metro's goal in residential and commercial areas. From Metro's view, congestion not only promotes transit but encourages denser development as people move closer to work to minimize commute time.
The transportation plan includes a "preferred" or wish list of all the projects that Metro would like to do over the next two decades. But it also has a second, more realistic "financially constrained" list, which is "based upon funds that can reasonably be expected over the 20-year life of the plan." The wish list would cost a total of $4.6 billion, while the constrained list would cost just $928 million. These numbers do not include the south-north light-rail line, which Metro says is "already funded."
Counting the Oregon portion of the south-north light rail line, 50 percent of the the preferred budget is dedicated to transit, 2 percent to pedestrian and bike facilities, and 48 percent to roads and streets. The constrained budget puts 73 percent into transit--mostly the light rail--1 percent in walkways and bikeways, and just 26 percent in roads and streets. Yet Metro says that even after all these transit investments transit will carry well under 5 percent of Portland's person trips.
The budgets demonstrate that the huge cost of the south-north light-rail line is measurable in more than just dollars. Projects on the preferred list but not the constrained list include nearly $1 billion of transit improvements, $130 million in bikeway and pedestrian improvements, and nearly $1 billion in road work that could significantly expand highway capacities. All of these could be funded with the cost of the Oregon portion of the south-north line, leaving several hundred million dollars for congestion pricing and other transportation improvements.
For example, just $66 million could begin a new bus service on twelve major routes that Tri-Met calls "FastLink"--faster, frequent, more comfortable service that Tri-Met describes as "the bus equivalent of light rail." By extension, for less than a fifth of the cost of the south-north line, Tri-Met could run light-rail-like schedules and service on all of its 85 major bus routes. The transportation plan explicitly says that the expense of just operating the south-north light rail will "limit future bus expansion."
Of course, federal light rail funds might not all be reprogrammable to bus transit, pedestrian improvements, and highways. But just the state and local share of the south-north line would pay for all of Tri-Met's preferred bus improvements, including FastLink service on numerous routes.
This is not to suggest that all of the new freeway expansions and other facilities on the wish list are necessarily worth building. But it is clear that construction of the south-north light rail will force the region to forego a huge amount of activities that could greatly relieve congestion and pollution in the entire area. Yet Metro does not want to consider this trade off and has specifically excluded it from consideration in its regional transportation plan.