U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590

Skip to content

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number:  FHWA-HRT-12-053    Date:  November 2012
Publication Number: FHWA-HRT-12-053
Date: November 2012


Casual Carpooling Scan Report

San Francisco, CA, December 7–8, 2010

The scan group arrived in San Francisco, CA, on the evening of Monday, December 6, 2010, and stayed in Berkeley, CA, on the East Bay. On Tuesday morning, the group visited the casual carpool  lines at the North Berkeley Bay Area Rapid Transit (BART) station. There are only two morning destinations for San Francisco casual carpoolers (the equivalent term for slugs in San Francisco)—Downtown and Civic Center. The Downtown drop-off point is near the intersection of Fremont and Howard Streets.

Six cars are lined up along multi-lane street waiting for passengers at a train station parking lot. It appears to be early morning: the sun is rising in the background and there is no traffic on the street aside for the queuing cars. Some of the cars have their lights on. In front of the queuing cars a crosswalk is painted on the street. A full parking lot lined with trees and street lights can be seen behind the line of cars.  A few people are milling around the first car in the line.

Figure 10. Vehicles queuing along Sacramento Street, waiting for casual carpoolers adjacent to the North Berkeley BART station in Berkeley, CA.

The Civic Center drop-off point is in the general area of 9th and Market Streets. As shown in figures 10 and 11, carpool traffic at the North Berkeley BART station was rather brisk. All scan team members got a ride into the city without incident. The group met  to debrief, discuss, and compare their trips. Fellow passengers commented that they like casual carpooling because they are guaranteed a seat, which is not always the case on BART, where trains are frequently “standing room only.” The same trip on BART from North Berkeley to Civic Center costs $3.70.

Early morning on a fall day. A line of cars are parked along the curb of a tree-lined street. The cars are all facing away from the camera on the right side of the street. Aside from the queuing cars, there is no traffic and there are no other parked cars on this side of the street. A woman is standing on the curb, opening the passenger-side door of the third car in line. The woman is wearing a backpack and is carrying a red bag. Just past the woman there is a brown trash can, a yellow newspaper dispenser, a parking sign, and a crosswalk sign. There is a grassy median in the middle of the street; tall street lights line the median but they are not on. Cars are parked on the other side of the street.

Figure 11. A casual carpooler confirms the driver's destination before entering a vehicle outside of the North Berkeley BART station in Berkeley, CA.

The BART system has a clever way of ensuring that only their paying passengers park at the station. It only costs $1 to park, but the payment must be made inside the station. To enter the station, one must swipe a fare card at the turnstile. Entering and exiting the same station incurs the maximum fee of $5.40. So someone parking at the station but not intending to ride BART, perhaps casual carpooling instead, will pay $6.40 per day to park there. The team later found out that free, on-street parking without time limitations is abundantly available in the neighborhood west of the station.

San Francisco is different from Washington, DC, and Houston, TX, because in July 2010 it began charging carpools (HOVs) to cross the San Francisco–Oakland Bay Bridge (SFOBB). Since that change, protocol has evolved to dictate (generally, though not universally for all destinations) that casual carpool riders offer a dollar to their drivers. Most of the scan group members followed this protocol, and most of the drivers accepted the token reimbursement. (An HOV now pays $2.50 to cross the SFOBB; thus, if a driver picks up two passengers, and each offers $1 in reimbursement, the driver is only paying $.50 out of pocket to cross the bridge.) The HOV discount on the SFOBB is available from 5 to 10 a.m. and from 3 to 7 p.m. The equivalent toll for an SOV is $6; thus, carpooling can save a driver $3.50 per day. The toll on the SFOBB is only for westbound traffic; there is no toll in the eastbound direction. One of the major time-saving motivators for casual carpooling is that HOVs get to skip long queues leading up to the toll plaza and instead merge into traffic right at the plaza, which can save up to 30 minutes of commuting time.

In the late morning on December 7, the scan group met with Rick Hutchison and Carlo Latasa of San Francisco City CarShare, a non-profit carshare company (comparable with ZipCar on the East Coast) that is trying to integrate ride-sharing into its Web interface. The group had a productive conversation with Hutchinson and Latasa about the psychology of carpooling and how best to sell the idea to their carshare members. As a carshare company, City CarShare can take a slightly different approach to ride-matching, because users are already using their Web site for carsharing transactions (whereas companies that only offer ride-matching services must attract users to their sites for that specific purpose). Past research has revealed that online dynamic ride-matching services can have a very difficult time luring users to their sites, which makes the integration with carsharing all the more valuable. In addition, carshare members have a financial incentive to split rental fees by sharing rides.

A close-up photo of a white traffic sign with red letters mounted on a metal pole. The sign reads: “No Passenger Zone; 3 p.m. thru 7p.m. Mon-Fri”.

Figure 12. Sign prohibiting parking along Beale Street, designating the area as a "passenger zone" for carpool formation during the afternoon commute hours.

At lunch time, the scan group was joined by Mark Evanoff and Jessica Scorpio. Evanoff is president of the AlterNetWays Company, a long-standing ride-matching service based in the Bay Area. Scorpio is affiliated with GETAROUND, a new startup company that offers peer-to-peer carsharing. The premise of peer-to-peer carsharing is that instead of a standalone company like ZipCar or City CarShare owning the shared vehicle, private individuals share their own personal vehicles with one another, thereby saving the capital expense of purchasing or leasing new vehicles that would otherwise be required to make carsharing possible. This concept has some similarities in logistical constraints, user psychology, and technical requirements to organized dynamic ridesharing.

A daytime photo of a wide downtown city street. A long line of people is standing on the curb on the right side of the photo. Street signs line the curb; they are blurry, but appear to be parking signs, carpool signs, and signs that direct drivers to highway on-ramps. Behind the queue of people a chain link fence lines the sidewalk. A few cars are driving down the street, approaching a green traffic light. Beyond the traffic light there are buildings on both sides of the street, and a pedestrian overpass connects two of them.

Figure 13. Casual carpoolers queuing for afternoon rides along Beale Street in downtown San Francisco, CA.

In the early afternoon, the scan group met with Jerry Robbins of the San Francisco Municipal Transportation Agency (MTA). Robbins has worked in various positions with the city of San Francisco for more than 25 years and has been following casual carpooling for much of that time. Robbins shared with the group a valuable file of newspaper clippings about casual carpooling (some many years old).

In the late afternoon, the scan group visited the (home-bound commute) casual carpool lines on Beale Street between Howard and Folsom Streets. Signage associated with no parking restrictions to allow for casual carpooling in this location is depicted in figure 12. This was a very impressive casual carpooling site, as can be seen in figure 13, because of the length of several carpool lines and the speed with which they moved. It is interesting to note that the carpool lines that served locations furthest from downtown San Francisco had the highest number of riders queuing, and those lines also moved the fastest.

For example, Vallejo is 52 km (32 mi) from San Francisco, and the line had 45 people in it at 4 p.m. Vallejo commuters cross two toll bridges (SFOBB and Carquinez) so the norm for that line is that passengers are expected to pay the driver $1.25 in each direction. Like SFOBB, the Carquinez Bridge is only tolled in one direction, but the tolls are in effect in the opposite direction. Those commuting into San Francisco from Vallejo pay the SFOBB toll in-bound and the Carquinez toll out-bound (see figure 14). Vallejo drivers also often (but not always) took three rather than two passengers, especially at times when the passenger lines were long. Even with three passengers, all are still expected to pay the $1.25; thus, the driver may actually recoup more money than the cost of the tolls (but not more than the total cost of gas, parking, mileage, depreciation, etc).

Outgoing highway traffic on a cloudy day. Traffic is bumper-to-bumper as it approaches digital tolling signs indicating which lanes to be in. There appear to be eight lanes of traffic. The digital signs indicate which lanes are for carpools, cash, FasTrak (an electronic toll collection system used in California), and buses. Carpools are directed to the two outer lanes; people paying with cash are directed into the second, third, sixth, and seventh lanes. The two middle lanes are for FasTrak-only drivers (FasTrak can also be used in the second thru eighth lanes). The far right lane is for buses and carpools only. Just before the digital tolling signs there is an exit sign on the right. An overpass with its own exit signs can be seen in the distance.

Figure 14. View of tolling signage on the westbound lanes of the San Francisco–Oakland Bay Bridge. Note that HOVs and buses are directed to reserved lanes at the far right and left sides.

The scan group casual-carpooled to the Richmond BART station, because afternoon casual-carpooling to North Berkeley is virtually non-existent. The reason for the lack of casual carpooling to North Berkeley appears to be because drivers pay no toll westbound on the SFOBB and thus save no money by picking up riders or time by bypassing tolling queues. In addition, North Berkeley is close enough to the city that the time savings offered by the HOV lanes are not worth the time required to stop and pick up passengers. From Richmond, the team took BART back to the North Berkeley Station where they had begun the day. The casual carpool line for Richmond was much shorter (about 10 people) than the line for Vallejo but also took much longer. Some members of the scan waited as long as 45 minutes for a ride. Because the line moved so slowly, those in line asked each driver if he or she would take a third passenger.

That evening, the scan group met with Dr. Betty Deakin of the University of California (U.C.), Berkeley, to discuss her empirical research of San Francisco casual carpooling. Dr. Deakin has worked with the San Francisco Metropolitan Transportation Commission (MTC) on a study of the effect of adding tolls for HOVs on the SFOBB.

On the morning of December 8, the scan group visited the Grand Avenue casual carpool location in Oakland, CA. This line is located under a highway overpass, sheltered from the elements. The group spent about an hour there observing the riders and drivers. For the duration of the observation period, queuing vehicles outnumbered queuing passengers. At its peak, the line was about 20 vehicles long. The team timed the wait for several vehicles, which was about 17 minutes; vehicles arriving earlier may have waited even longer. Having more drivers than passengers made it very easy for riders. They had no wait and could hop in a car as soon as they arrived. This also enabled scan team members to talk with waiting drivers, which they had difficulty doing in most other cases. Two notable observations were (1) a man riding up to the line on a foldable bike and placing it in the trunk of the casual carpool vehicle at the front of the line; and (2) a woman walking up to the general area of the line and then getting on the phone to call her co-worker with whom she typically casual carpools (both as passengers) to find out how soon she would be there.

The scan group casual-carpooled to the downtown drop-off location on Howard Street between Fremont and First Streets. San Francisco has installed signs in this area indicating that it is a “Carpool Drop-Off Zone 7–10 a.m.” The scan team observed that this drop-off location was especially active.

At mid-morning on the same day, the scan group had a meeting with representatives from AVEGO, a company that markets dynamic ridesharing software for smart phones and is working on dynamic ridesharing pilot projects in Seattle and elsewhere, as well as representatives from the San Francisco MTC and the Climate Protection Campaign of Santa Rosa. AVEGO is currently implementing a Washington State funded dynamic ridesharing pilot project on State Route 520 (http://go520.avego.com/st-pilot/) in the Seattle area. The project involves recruiting 250 regular drivers and 750 regular passengers for a 6-month pilot by using an iPhone application to help facilitate ride matches. The representatives from San Francisco MTC and the Climate Protection Campaign were interested in meeting the scan group because of plans (currently in the works) to fund and implement an organized dynamic ridesharing pilot project (probably with an electronic component) in the Bay Area.

Summary of Lessons Learned and Analysis of Practices

There is somewhat of a natural pricing experiment that takes place on the Vallejo casual carpool line that merits a particularly close observation. Since the implementation of the carpool toll, passengers are generally expected to offer the driver $1.25, such that two paying passengers would fully cover the costs of the $2.50 carpool toll. The scan group was told by many riders that even if the driver accepts more than two passengers, the passengers are still supposed to each offer $1.25. Thus, although the HOV-3-lane benefits apply to drivers with only two passengers, drivers can receive additional reimbursement if they take more passengers.

If further research showed that this small payment dramatically affected the average number of passengers a driver might pick up, it could be worthwhile for governments to provide such a small incentive to drivers in other circumstances in which this “natural incentive” does not exist in an effort to relieve roadway congestion. (It is probably much less expensive than the typical bus passenger subsidy that governments provide). The MTC is currently conducting an evaluation of the toll changes, including the new carpool toll. A survey and counts were conducted in April 2010 prior to the July 1, 2010, implementation of the carpool toll, and a follow-up count and survey was to be conducted during the same timeframe in 2011. MTC plans to include survey questions to explore these pricing matters further.

Many HOV lanes in the United States have become congested, and raising occupancy requirements to use such lanes may raise objections. As a result, determining how to persuade casual carpool drivers to take more passengers than what is required to use HOV lanes or to qualify for HOV toll discounts—whether by somehow providing financial incentives, hiring someone to personally coax drivers, or by other means—could be very beneficial.

San Francisco MTC conducted a follow-up survey of casual carpoolers in 2011 , the results of which can be found by accessing the following link:   http://www.mtc.ca.gov/library/CasualCarpool_summary_2011.pdf.  The survey findings provide additional insights on user behaviors and motivations beyond those garnered directly from the scan.

Total HOV volume has declined since the new toll went into effect in July 2010. Casual carpooling in San Francisco, however, still moves a very high volume of people.


Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101