THE SAN FRANCISCO BAY BRIDGE

When commuters in Berkeley want to get to San Francisco (across the bay), they generally choose one of two options. The first is to drive across the Bay Bridge. The second is to ride on BART -- short for Bay Area Rapid Transit, the public transportation system.

When there is no traffic, driving the Bay Bridge is the quickest option, taking about 20 minutes. However, as traffic gets heavier, the narrow bridge becomes congested, and travel time increases. For calculation purposes, let's suppose that each additional 2,000 cars adds another 10 minutes to the trip. Therefore, 2,000 cars lengthen travel time to 30 minutes; 4,000 cars to 40 minutes, and 6,000 cars to 50 minutes.

BART, on the other hand, always makes the trip in 40 minutes, no matter how many people are riding along. During rush hours, BART simply adds more passenger cars to each train.

Of course, we can expect each commuter to act selfishly in the search for the shortest trip possible. If traffic on the Bay Bridge becomes too heavy, and travel time is extended to 50 minutes, a certain percentage of drivers are going to switch to BART and its faster, 40-minute ride. However, if too many people switch to BART, then congestion on the Bridge falls, and so does travel time -- to, say, 30 minutes. This will then attract others back to the Bridge. In time, an equilibrium is established, in which everyone makes the trip in 40 minutes. With 10,000 commuters, this equilibrium is reached when 4,000 drive over the bridge and 6,000 ride on BART.

But is the group really saving the most time in this solution? Not at all. Another arrangement saves the group much more time, but it requires group agreement and cooperation. Suppose they agree to reduce the drivers on the bridge from 4,000 to 2,000, cutting their travel time from 40 to 30 minutes. The other 2,000 would-be drivers agree to take BART, since its 40-minute commute is the same they would have faced on the Bridge anyway. In a single morning commute, the group would collectively save 20,000 minutes -- or almost two weeks -- of travel time.

So, how can the group arrive at this solution? One way would be to issue 2,000 licenses for the Bridge, with a way of rotating them among the 10,000 commuters to ensure fairness.

But for those who are opposed to government intervention in our lives, a more market-based solution suggests itself. A toll could be established on the bridge, essentially making drivers pay for their saved time. The toll could be raised or lowered until the desired number of drivers were taking the bridge. And the toll funds could be saved for the construction of a second bridge, alleviating the problem in the first place.

There are a few problems with this approach, however. Toll booths themselves are primary causes of congestion. If the city collected the toll, conservatives would lambaste this as another tax. A private owner could collect it, but the group has no incentive to allow this. Why? Because he could collect tolls indefinitely, even after a second bridge was built. (Compare that to city ownership, where the people can vote on how long the toll should be collected, and how the money should be spent on themselves.) Nor would a private owner have much of a financial incentive to build the second bridge. He could build it, but then he would have to cut his tolls to attract more drivers to drive it. Why bother in the first place?

Several points emerge from this example. First, the invisible hand does not always work; self-interest does not always result in group benefit. Second, the invisible hand works (after a fashion) only when a monetary price is added to a previously free commodity -- in this case, commuter time. Third, many fortuitous and complex factors can distort an equilibrium, such as the fact that BART's commuting time is always 40 minutes, or that toll booths help snarl traffic. Consequently, group action is sometimes needed to ensure maximum individual benefit.

Return to Overview