NO JOY IN CARPOOLING; DIAMOND LANES ON FREEWAYS CAUSE MORE TROUBLE THAN THEY'RE WORTH.

Byline: Gary M. Galles Local View

Over a period in which California virtually abandoned building regular freeway lanes, it has engaged in a $2.3 billion experiment in High Occupancy Vehicle, or diamond, lanes. And the 925 miles of HOV lanes constructed so far are scheduled to double in the next two decades.

One would expect such an extensive, expensive undertaking in thecongestion capitol of America would reflect evidence that HOV lanes actuallyreduce rush hour freeway congestion, thus reducing pollution (vehicles pollute more in slower, congested driving). But beyond backers' repeated, undocumented claims of success, the only real ``proof'' cited is that HOV lanes typically carry more people during rush hour than adjacent mixed-use lanes.

As the Legislative Analyst's Office just concluded, ``On average,California's HOV lanes carry 2,518 persons per hour during peak hours - substantially more people than a congested mixed-flow lane and roughly the same number of people as a typical mixed-flow lane operating at maximum capacity.''

Unfortunately, the fact that most HOV lanes carry more peak hour passengers than adjacent lanes does not demonstrate their success (though it does demonstrate the failure of those that do not), because the number of people carried in HOV lanes during rush hour does not accurately indicate its effect on congestion in other lanes. Even if an HOV lane carries more people than a mixed-use lane, it can increase how many cars are forced into the other lanes.

More rush hour HOV riders doesn't demonstrate their success at reducing congestion in other lanes primarily because such lanes funnel those already riding together into the HOV lane. While that can raise the number of persons in HOV lanes, it typically increases rather than decreases congestion in the other lanes, worsening pollution.

Say that a four-lane freeway was carrying 25 percent of the cars in each lane during rush hour, with 8 percent of the cars sharing rides (above the 1998 Southern California average). Converting a lane to HOV use would shift those cars onto the HOV lane, resulting in both fewer cars and riders in the HOV lane than the previous mixed-use lane, but compress the remaining 92 percent of cars into the three remaining lanes. That would also sharply increase the congestion and therefore pollution in those lanes (as when it was first tried on Interstate 10 in Santa Monica), creating an all-around disaster.

Where the car-pool percentage is twice the Southern California average, and 15 percent of the cars share rides, converting a lane to HOV use would shift those cars into the HOV lane. Since each of those cars would carry two or more riders, that lane would carry more riders than before - 15 percent times the 2.3 riders estimated as the California HOV average.

Even though the HOV lane carries more people, unless it increased carpooling among the other drivers, it still forces the remaining 85 percent of the cars into the remaining lanes. This increases the backup in those lanes, adding congestion and reducing the number of cars and passengers carried - hardly a success.

In even this optimistic case, we can see how many existing drivers would have to begin carpooling just to keep HOV lanes from increasing congestion in other lanes. To leave no more than 75 percent of the cars in theother three lanes, 10 out of each 85 drivers not now carpooling would have to share rides in five cars, which move to the HOV lane. This would require a one-third increase in the fraction of carpoolers. But while limited regional surveys indicate that HOV lanes increase carpooling to some degree (no statewide evidence on this crucial determinant of HOV success exists after a quarter-century), there is little to suggest that so many solo rush hour drivers will switch to car pools as a result (the Legislative Analyst's Report found that ``surveys have found that many carpoolers would ride together whether or not the HOV lane existed'').

The number of rush hour HOV riders also overstates their successin reducing congestion in other lanes because many in HOV lanes would not otherwise have driven separately on the freeway at that time, so their ride-sharing does not reduce regular lane freeway congestion. This includes those driving children or families, the infirm, or other nondrivers. It also includes those ride-sharers who would have traveled at nonrush hour times or who would have driven on surface streets to avoid congestion, but who shift to rush hour freeway driving because of the uncongested HOV lane.

These increase the number of rush hour HOV riders, but do not reduce rush hour use of regular freeway lanes.

As the Legislative Analyst's Office concluded, it is past time for a comprehensive examination of HOV lanes. But that means the 24 percent of HOV lanes that, even during peak demand, are so underutilized that they fallshort of Caltrans' usage criteria for HOV lane construction already worsen congestion and pollution, and need to be abandoned (as with Interstate 580 and New Jersey's Interstate 287). And it means going beyond simple ridership counts to determine the real effects of the ``successful'' HOV lanes. If the car-pooling and congestion experience elsewhere (as with New Jersey's Interstate 80) is a guide, we will also end up abandoning even many of these supposed successes as expensive failures.

CAPTION(S):

drawing

Drawing: no caption (cartoon of road and cars)