By Steven Pearlstein
Friday, December 29, 2006
It's time to tackle the hot topic on everyone's mind this New Year's weekend -- congestion pricing.
Maryland and Virginia are moving ahead with a form of congestion pricing on the Beltway and Interstate 95. In exchange for building additional lanes, contractors will be allowed to collect tolls that will vary by the minute. When traffic is heavy, the price will rise to whatever level is needed to keep the express lanes flowing. When demand is low -- presumably at times when traffic is flowing smoothly in the normal lanes -- the price will fall to near zero. Under most scenarios, buses and carpool vehicles will travel free.
Recently, the staff at Metro used congestion pricing to design a fare increase meant to raise an additional $64 million a year for the bus and subway system. On average, that would work out to a 14 percent fare increase. The plan would impose the biggest increases on peak-hour subway commuters who use the 19 busiest stops in the downtown core.
Economists love these schemes because they use "market mechanisms" to allocate "scarce resources." But a lot of non-economists, who don't spend much time worrying about allocative efficiency, think them unfair and unnecessarily complicated.
In the case of the highway schemes, many people are uncomfortable with the idea that basic public services, once available to everyone regardless of income, will now be allocated on the basis of ability to pay. If more capacity is needed, why not add lanes the old-fashioned way, by having government pay for them and letting everyone use them?
Several problems there.
First, history is pretty clear that adding lanes doesn't relieve congestion for long. More capacity simply invites more cars, either by stimulating additional housing development or luring transit and carpool riders back into their cars.
Second, pigheaded voters in Maryland and Virginia have been unwilling to raise gasoline or other taxes to pay for the highways they claim to want. So the next-best alternative is to get private contractors to finance highway expansion in exchange for a stream of future toll revenue. At the least, these "Lexus lanes" will provide some temporary relief while speeding the commutes of bus riders and carpoolers at no cost to them. And given the alternative, isn't it better to impose tolls on those who can best afford to pay them -- along with ordinary folk who occasionally have the urgent need to get home in time to drive Tammy to soccer practice?
In fact, congestion pricing has proven successful not only for managing highway congestion but also for relieving crowded downtown areas like those of London and Stockholm. Mayor Michael Bloomberg of New York has flirted with this idea for Manhattan. And it's something this region needs to consider for downtown and even Tysons Corner, once the Metro extension is completed.
The criticism of Metro's fare plan is that it would penalize people who can't readily change when and where they work. Skeptics ridicule the notion that people are going to commute before dawn or get off the subway at Rosslyn and hoof it downtown.
In fact, Metro planners have no illusions about subway fares as an instrument of behavior modification. They realize that most people won't change their commuting patterns in the short run.
Instead, Metro approached the problem as any business would, looking for the best way to raise money to cover higher operating costs while causing the fewest customers to defect to the competition -- that is, to drive.
One calculation showed that even if parking fees and fares go up, the cost of parking at a Metro lot and commuting in would still be about 70 percent of what it would cost to drive downtown and park.
Experience further shows that the passengers who are least willing or able to defect to other forms of transportation are those who commute downtown from the suburbs at peak times.
Based on those analyses, Metro planners concluded that the best strategy would be to skew the fare increases to their captive suburban peak-hour commuters. Fare hikes would be lower for the more price-sensitive, off-peak passengers. They'd also go easy on "reverse" commuters who work in the suburbs, where lower parking fees make commuting to work by car a more attractive alternative. Bus fares would go up only modestly on average, in the hope that some passengers who now ride the subway for short distances might switch to buses during peak hours.
The idea of charging a premium for peak-hour commuters to crowded stations also makes economic sense if you consider the capital costs associated with adding capacity.
The subway system is approaching its limits during peak commuting hours on certain lines (Red and Orange) as they pass through the downtown core. What that means is that adding capacity soon won't be a simple matter of buying a few more cars and hiring a few more train operators. Rather, it will involve the enormous costs of making stations bigger to accommodate longer trains, or digging tunnels for express trains.
This raises a different fairness issue. Is it fair to ask all passengers to pay for expensive new capacity that is required only to handle downtown commuters for a few peak hours? Or is it fairer to put the bulk of that burden on those who will benefit from the new capacity?
There is no right answer to these questions -- it depends on whether your goal is to maintain market share for public transit, align fares with costs, or try to use fares over the long run to alter people's behavior. The best approach is probably a mix of all three.
As I see it, Metro planners are on the right track, though perhaps they ought to use a narrower window for peak hours. In the future, perhaps they should set fares according to the actual cost of service, rather than distance traveled. And it may be worth considering different fare levels for each line, with lower fares to boost ridership on underused lines and higher fares to dampen peak-hour demand on lines nearing capacity.
Then again, we could just follow the New York City model and charge a flat rate, say $2.25, for any ride at any time. But what fun would that be?
Steven Pearlstein can be reached email@example.com.