As recently as 1999, $2.5 billion was supposed to be enough to build two of the region's most ambitious transportation projects: an 18-mile highway across the Maryland suburbs and a 22-mile Metro line from Falls Church to Dulles International Airport.
Six years later, that amount isn't enough to finance either. The cost of Maryland's intercounty connector is $2.4 billion, and finance charges could push the price to $3 billion. It would cost as much as $2.4 billion to extend Metro just halfway to Dulles.
The steadily escalating price tags of the Washington region's two largest proposed transportation projects are at the heart of debates this summer as one, the connector, nears construction after 40-plus years of indecision and as managers of the other struggle to keep its costs down.
Are they still worth it?
The answer depends on how you look at it. Whether a project is worth building depends primarily on how much it costs, how much time people will save when they use it and how much economic development it will generate. It's also critical to consider whether the cost of one project saps too much money from others.
Then there are factors that are harder to tally, such as what impact the project will have on the environment, how the area around it will grow and even whether people will like its look and feel.
In 2010, according to projections, the connector would carry about 80,000 vehicles a day, equaling $37,500 per driver. In 2011, planners say, the rail extension would add 15,100 riders to Metro, or $158,940 for each.
Those are damning figures to opponents and on the high end of similar projects, transportation experts said.
But benefits of large projects aren't limited to time savings enjoyed by users. They often serve as the backbone of business expansion, and that benefits the entire state.
"As costs go up for these kinds of projects, generally the benefit picture still outweighs the costs," said Hani S. Mahmassani, director of the Maryland Transportation Initiative at the University of Maryland.
David L. Lewis, chief executive of HLB Decision Economics and a leader in analyzing projects, agreed: "Congestion-relieving projects, especially those that were already worthwhile 20 years ago and still haven't been done, simply become more and more worthwhile as time goes on."
But Lewis said rising costs also raise the question of whether a particular project "is the next best project to do."
That point has been raised by opponents of the connector, who contend that an inner suburban rail line would generate at least as much economic activity as the highway.
Escalating costs are nothing new in the Washington region, or across the country. The price of the Springfield interchange, for instance, has more than tripled, from $220 million to nearly $700 million.
The granddaddy of rising prices is the Big Dig, the depression of Boston's Central Artery that was originally estimated at $2.6 billion but has turned out to cost $14.6 billion.
Lewis also cited Dulles International Airport as an example of a project for which benefits did not justify costs -- at least for its first 25 years.
"Everyone says: 'We built it way ahead of the curve. Aren't we smart?' " he said. "In fact, if we had the money over 25 years, we could have had infrastructure projects with much bigger benefits relative to their costs than Dulles Airport."
Transportation experts said faulty figures are almost inherent to big projects, because initial estimates are calculated before detailed work is done.
Sam Carnaggio, manager of the Dulles rail project, said initial estimates are at best ballpark guesses. But he said when politicians use them to sell projects, "it becomes cast in concrete."
David Luberoff, a Harvard University researcher and co-author of "Megaprojects," said politics plays a significant role. "If you think about this as a political issue, you would expect that the advocates for a project will underestimate the cost and overestimate the benefits," he said. "That's clearly one factor."
On the Dulles rail project, "the budget estimated at $1.5 billion was based on drawings," Carnaggio said. "Those certainly aren't engineering drawings. There's no detail whatsoever. The costs obviously are going to be different as you get into the details."
Carnaggio said those details include bigger stations, $100 million for two power substations and inflating costs for the project's builder.
The financing plan for the rail project calls for the federal government to pay about half its cost, with toll collections on the Dulles Toll Road and other state money to cover a quarter and business taxes the rest.
A federal government assessment of the project last year rated its cost-effectiveness at "medium-low." That was based on its $1.5 billion cost vs. savings of an estimated 17,800 commuting hours each weekday. Carnaggio said costs have to come down to maintain that rating.
The intercounty connector first hit planning maps in the 1950s as a link in a proposed outer beltway. By the late 1970s, it was estimated at $262 million, or $781.5 million adjusted for inflation. When it was scrapped in 1999, its price was $1 billion. When it was revived in 2003, its cost had risen to $1.4 billion, which quickly became $1.7 billion and, now, $2.4 billion.
Maryland officials said $750 million of the connector's price would be paid by borrowing against future federal funds -- consuming about 20 percent of that money over the next 15 years -- and the rest would come from state money and tolls.
State officials said the higher cost is a reflection of project alterations, inflation and uncontrollable price spikes in steel and other materials.
For instance, longer bridges to protect waterways added $120 million. Enhanced storm water runoff techniques cost an additional $20 million, and $30 million was added for toll-collecting facilities.
William Buechner, chief economist at the American Road and Transportation Builders Association, said that the price of materials, which accounts for nearly half of project costs, jumped 8.5 percent last year, largely because of an increase in steel prices caused by demand from China, and that costs will go up an additional 8 to 10 percent this year.
"What it means is that in 2006, it'll cost 10 to 15 percent more to do the same thing as in 2003," Buechner said. "It's not just the ICC and Dulles connector whose costs have been going up. A lot of states are finding they can't do all the projects they want to do."
Staff writer Peter Whoriskey contributed to this report.