Saturday, January 26, 2008

Extending the Orange Line to Dulles

Washington is abuzz because the Federal Government looks like it will refuse to sign off on funding for an extension of the Orange Line (the Metro line that goes west into Virginia) to Dulles airport. While there are legitimate issues about the lack of transparency in how the Federal Transit Adminstration made its decision, it almost certainly made the correct decision.

Heavy rail is not a good transporation option in the absence of high levels of density. John Kain showed years ago that inapporpriate heavy rail cannibalizes funding for other forms of transit, and tends to reduce public transportation ridership.

A rail line to Dulles is particularly inappropriate because there is already a dedicated four-lane highway to Dulles--a highway that is vitually never congested. This means a viable public transportation option--nice express buses that run frequently--could be put into place at relatively little expense. The problem now is that the buses don't run frequently enough to enough places, and so they are not an attractive option relative to driving. But $5 billion (the cost of the proposed extension) placed in treasury securities could fund an awfully large number of bus routes.

I am actually a little more sympathetic to heavy rail than was Kain--I am not sure his work captured all the general equilibrium benefits of rail under certain circumstances. But I am pretty sure that the proposed extension to Dulles is economically a non-starter.


Katherine Helen Anne said...

Richard: Thanks so much for this interesting blog.

On the Dulles rail line, I would think that extending rail to Dulles would draw in a lot more passengers than increasing the frequency of the buses for two reasons. First, people are highly averse to shifting modes partway through a trip. I try to take the 5A bus when I go to Dulles, but it adds a lot of hassle from most starting points: going from e.g. Woodley Park to Metro Center, switching to the orange line, riding to Rosslyn and then switching again onto a bus, all with luggage. Second, wouldn't the trip length be shortened substantially from most starting points?

I tend to choose whenever possible to fly out of National airport rather than Dulles for precisely this reason - it's a quick, easy rail ride away and so I don't need to add in cab fare to the total cost of the journey. Similarly, if I need to go to Raleigh, NC, and I can't find a good fare from National, I would consider renting a car and driving there instead because it's less hassle than going all the way out to Dulles. (If I owned a car I would certainly do so.)

I would think others would substitute out of National or other modes of long-distance transportation into flying from Dulles as well as those currently driving on the Dulles toll road, suggesting that the traffic on the road isn't evidence of all the latent demand.

What do you think?

Richard K. Green said...

Thanks for the thoughtful comment. Let's work this through. The capital cost of the system is estimated at $5 billion (although these things almost always come in other budget), and the forecast ridership is 60,000 per day. Remember, this is gross ridership, and so it doesn't net out cannibalized bus ridership.

Suppose the opportunity cost of capital and depreciation is 5 percent per year (this is too low, but..). That means that the capital cost of the system is $250,000,000 per year. $250 million divided by (60,000*365) means this capital subsidy per rider per day is $11, even under the most generous calculations. Then you need to add the operating subsidy. It's hard to see how this is economically sensible.

Pantograph Trolleypole said...

You're looking at it all wrong. You're argument that density begets heavy rail is why people think these investments are so expensive, you don't look at the actual economics of the investment. Much like the highways around the country have spurred sprawl, rail lines around the country have spurred density.

Just look a few miles away at the Rosslyn Ballston corridor. 32% of the Counties tax base comes from 7% of the land along the metro line. That development wasn't there when they built the subway. Yet the value was created by looking at transit and land use together. By looking at them separately, density will never happen in this highly car subsidized country and transit will always be called "too expensive", its a self fulfilling prophecy of sorts.

Also, most of the people who live on the RB corridor walk to shops, walk to the grocery store and don't use their cars. The County has measured traffic on the corridor and it has hardly increased during the time period, yet density has. This means there is a great reduction in Vehicle Miles Traveled that comes from building these lines and building density around them. How much money and resources does that reduction save?

Now lets look at your calculations for the efficiency of the project. You're just using the riders to calculate the economics of the project itself, but it doesn't happen in a vacuum. What about the reduction in the need for sprawl subsidies for each new housing unit's road and sewer infrastructure? What about the household savings for folks that take transit instead of driving? What about the increased tax base? What about the increases in efficiency and certainty of the commute for each of those people that use the line? What about the operational cost savings per passenger? (It costs around 40 cents per passenger mile vs 53 cents for operating a car per mile) This doesn't even count buying the car or the roads.

There are many other advantages but for the most part, the whole auto-centric system we've set up in this country is a huge waste of resources which has misallocated our post war wealth towards an unsustainable end.