GGT’s bus ridership is sagging, but how to fix it?

Transit 005There’s no question about it: Golden Gate Transit ridership is in decline. But, as we cap off a year with two crazy days for transit (America’s Cup and the Giants parade), we should take a step back and look at where our ridership is going and, perhaps, how the situation might be improved.

A word about data

GGBHTD uses a July-June fiscal year, so we’ll be discussing projections for the rest of the 2013 fiscal year (last July to next June) as well as what has happened historically for the last few fiscal years. This confounds analysis. Federal data uses the federal fiscal year (October-September) while Census data uses the calendar year. If I switch into a different year, I’ll be sure to mention that in the text.

As well, it takes about one month for GGT to audit bus ridership numbers, so December’s numbers haven’t been released. Nevertheless, today seems like as good a day as any to discuss ridership.

The numbers

GGT’s ridership has been dominated by bus services for a long, long time, but its share has shrunk significantly since 2002.

In FY 2012, GGT carried 8.7 million passengers, roughly one quarter by ferry. Total ridership is down significantly from 2002, when GGT carried 10.8 million passengers – only one-sixth by ferry. All the losses have been borne by the bus segment of the system, which has seen annual declines in seven of the last ten years. In contrast, ferries have seen declines in only four of the last ten years, led by sometimes double-digit growth rates at Sausalito.

These trends look set to continue in the current fiscal year. Bus ridership for the first half of FY 2013 is down 1.6 percent over the same period in 2012, while ferry ridership is up 8 percent. Intriguingly, this could be the first fiscal year that total ridership increases entirely on the strength of the ferries.

Below is a normalized chart of ridership trends, with 2002 set as the baseline.

Transit ridership normalized to the 2002 fiscal year. Ridership for the current fiscal year is a projection.

Transit ridership normalized to the 2002 fiscal year. Ridership for the current fiscal year is a projection. Click to enlarge.


The most obvious trend is the decline from FY 2002 to FY 2004. In that time, San Francisco was still suffering the aftershocks of the tech bubble pop, which was compounded by the 2001-2002 recession. According to the Federal Transit Administration, ridership peaked in federal year 2001 at 11.6 million transit trips – 9.7 million by bus, 1.9 million by ferry. That’s also around when bridge crossings peaked, implying overall travel demand fell.

This is borne out by census data. Since calendar year 2000, commutes by Marinites have fallen by about 6 percent. But this isn’t enough to account for the shift away from buses, which are down 32 percent. There must be structural reasons as well.

Sure enough, that’s what we find. From 2002 to 2004, GGT dramatically restructured and cut its routing to cut costs, reducing its vehicle revenue miles (how far its buses travelled collecting fares over the year) by 32 percent. Over subsequent years, revenue miles increased only 1 percent. Though GGT projected only a 15 percent passenger decline, fare increases, competition from other non-car modes of transportation like bikes and ferries, and declines in commutes, overall took a toll.

Turning the situation around

If bus ridership is going to increase again, GGBHTD needs to see itself as a single transportation agency. At the moment, bus schedules don’t link well with ferry departures, bridge tolls are too low to push people to bus usage, and timing-point schedules are hindering the development of high-frequency corridors. Most of these are actually revenue-raising measures, which could be pumped back into the bus system.

1.      Loosen restrictions on bus riders

Anyone riding Muni or AC Transit knows it is used for every errand under the sun, from commutes to groceries to getting to the airport. The front of the buses have center-facing seats reserved for seniors and the disabled, and passengers can exit out the rear door. In Muni’s case, they can enter that way, too.

Yet, for whatever reason, GGT has chosen to keep its rear-door Clipper readers deactivated, forcing everyone to exit out the front. Navigating the system with anything that can’t fit on a lap is against bus policy. As well, the lack of level-boarding buses hinders the ability of seniors from using the buses.

While the current bus configurations do keep the buses comfortable, it makes them less useful. Reforming these would make the bus more useful for everyday travel and speed boarding and alighting. Low-floor buses that provide level boarding could be the priority for all future bus acquisitions, rolling them into the regular capital replacement budget.

Cost: Marginal

2.      Coordinate ferry departures and bus arrivals

While GGT does a fantastic job timing different bus lines, it does a miserable job coordinating with ferries. Route 29 to Larkspur Landing, for example, arrives 40 minutes before the next ferry departure. Part of this is Marin Transit’s fault, which times Route 29, but GGT could easily fix the problem, too. A shuttle between the Transit Center and Larkspur Ferry, say, would encourage people to take the bus to the ferry rather than drive.

The total round trip, including layovers, would take about 25 minutes.

This leaves 20 minutes of dead time between cycles, so this service could be added on to certain Route 23 or 35 buses, rebranded as 23F or 35F, which would save costs.

Cost: $660,000 per year for new service, $340,000 for route extensions.

3.      Charge for parking at park & ride lots

I explored this concept in a previous post for ferry terminal parking, but it should apply to all park-and-ride lots that get full over the course of a typical day. This would accomplish two goals: to increase reliance on transit to get to transit, and free up spaces for midday travelers. Though some riders might abandon transit altogether, GGT would very likely see a net gain in ridership. As well, the parking fees collected could be pumped back into service, either for collector buses or for better frequency on selected corridors.

The principal barrier to implementation is Caltrans, which controls the park & ride lots along the 101 corridor. Sacramento would probably need to intervene to force them to charge and to pass the money along to GGT. The ferry terminal lots, however, are controlled by GGT and could be priced now. Ideally this would paired with the ferry shuttles described above.

Income: Variable, but likely in the low hundreds of thousands per year. Implementation would require a one-time capital investment for parking meters, ticket dispensers, etc.

4.      Increase tolls on the Golden Gate Bridge

I discussed this concept in the past as well. A driver should pay just as much to cross the bridge as a bus rider. If we raise the base FasTrak toll to $7.20 and the congestion toll to $8.80 – the cost of going to San Francisco and back from Southern and Central Marin, respectively – the bus becomes a much more attractive alternative. A commuter would pay the same no matter which mode she chooses, so why not choose a bus?

Cost: Political. Former San Rafael mayor Al Boro scuttled the last attempt at congestion pricing at the Golden Gate Bridge, and that was for a much more modest increase.

Income: Significant. When congestion pricing on the Bridge was last analyzed it was part of a broader pricing scheme that saw a 12 percent drop in traffic coming in to San Francisco from all sources. If this holds true for the Bridge on its own, toll revenues would still go up, to the tune of about $50 million per year. If any of those drivers convert to busing, the income would be even higher.

5.      Provide in-city pickup and drop-off for all-day routes in San Francisco and Richmond

It’s a bit of cheating to get numbers up, as it opens up an entirely different market, but it could provide a good source of income. In San Francisco especially, GGT provides great redundant express service. The all-day lines (basic and Route 92) could pick up riders in-city, making that part of their journey more profitable.

To do this, the boards of SFMTA and AC Transit would need to grant GGT permission to run routes through their territories.

Given how much this may slow the San Francisco routes, especially Route 92, GGT may want to wait until the Van Ness and Geary BRT corridors are completed.

Income: Unknown, but the alteration should only be made if the result would be revenue-positive.

In short, GGT should make the bus convenient (suggestion 1), easy (suggestions 2 and 5), and financially attractive (suggestions 3 and 4). It should reinvest new revenues into increased service and better infrastructure, part of a virtuous circle of rising ridership and declining congestion.

If Marin wants to be a green, environmentally-sustainable place, the bus must be part of the equation. Falling ridership isn’t a given, but it will take conscious steps to bring it back in step with ferry ridership.


About David Edmondson
A native Marinite working in Washington, DC, I am fascinated by how one might apply smart-growth and urbanist thinking to the low-density towns of my home.

29 Responses to GGT’s bus ridership is sagging, but how to fix it?

  1. Dan Lyke says:

    So I only have experience up in Sonoma, although second-hand I hear tales from my wife who takes developmentally disabled, some of whom are physically disabled, young adults on GGT buses.

    On #1, if GGT goes to uncomfortable low buses, they lose one of the two advantages they have over Sonoma Transit. The kidney-jouncing rattletrap low buses are no competition for the higher coaches. Totally with you on the rear exit, though (though I can’t use it, ’cause I’m generally getting a bike). (The second advantage is that the routes of #80 etc is slightly faster than the 48X, though see below…)

    A big thing in my morning decision over which system to take, though, is simply driver demeanor. I’ve been berated for not wearing lights because I allegedly wasn’t terribly visible at the stop on the wide sidewalk at 7:14 in the morning. I’d consider this seriously, except this is the driver who at least twice has missed the Cotati Railroad Ave exit, gone up to the 116 exit, turned around and gone all the way back down to Petaluma Blvd N. to get back to Railroad Ave. So much for faster.

    My wife reports all sorts of hassles with drivers, “accidentally” pulling up too far so that the wheelchair lift is blocked by a sign and then claiming they’re unable to back-up without supervisor approval, so sorry, no wheelchair this stop. Lots of other struggles in drivers dealing with her students as riders. Which is tough, I understand the frustration, but as a user of transit I understand that the developmentally disabled are a large portion of the user base, and if I as a fellow rider am made to feel uncomfortable about how they’re being accomodated, then I’m less likely to choose that option.

    But I think the real issue is what comes up with considering #5: Unless there’s some way for fare recovery plus per-rider subsidies to actually cover operating costs, why would any bus system want to increase ridership? Service the routes so you can get that checkbox on your State and Federal funds, but riders just slow service and cause hassles.

    Or at least that’s the impression I get, from pretty much all bus services I use, as a transit user.

  2. dw shelf says:

    ^force them to charge and to pass the money along to GGT.

    Something is wrong with this kind if thinking.

    If bus service makes sense, the government doesn’t need to be forcing anything.

    If bus service doesn’t make sense, then we would expect fewer people to use it.

    • Dan Lyke says:

      one of the problems we have in “bus service makes sense” is untangling all of the ways in which we subsidize the automobile. The problem is that we have all of these wacky ways in which we spend tax dollars and government resources intertwined in ways that… really… nobody’s got verifiable models for “makes sense”.

      And, likely, nobody ever will.

      • dw shelf says:

        Just because it’s tangled doesn’t imply any of:
        1. We can assume whatever we’d like.
        2. That highway subsidies should be offset by forced transfers from automobile drivers to bus riders.

        Let’s get rid of highway subsidies.
        And airport subsidies.
        And bus subsidies.

        Let the money flow where it flows. Then, we will know that things are making sense.

        • Dan Lyke says:

          Absolutely agreed. A VMT would probably be the best way to start to untangle those subsidies, but barring that raising the gas taxes to the point where they actually start to cover the direct costs of roads and infrastructure (let alone the indirect costs of health and pollution) would be a good start. Isn’t a politician alive, though, who’d campaign on raising the gas tax by several dollars a gallon. Sigh.

    • In the context, it’s an issue of Caltrans being forced to charge for parking, which they would probably hesitate to do because they tend to yield to the demands of current drivers rather than the needs of potential ones. But charging would help people that want to drive to the bus but can’t by freeing some spaces for use in the middle of the day. The charge simply puts a price on a scarce good so it can be allocated more efficiently.

      • dw shelf says:

        ^The charge simply puts a price on a scarce good so it can be allocated more efficiently.

        If the money were to be used exclusively to supply highways or parking lots, you might be logically proposing such an equation. But that’s utterly not what you originally proposed.

        • I don’t follow. Charging for a scarce resource (parking spaces) will allow the more efficient allocation of that resource (transferring the use of the space from a driver who places a low value on it to a driver who places a higher value on it). That’s the efficiency gain.

          The money could go to road travel, increasing demand for road space and decreasing demand for park & ride use, or it could go to transit, decreasing demand for road space and keeping neutral demand for park & ride use, or it could go to something entirely unrelated. None of these uses change how much more (or less) efficiently parking spaces are allocated under a pricing scheme.

          • dw shelf says:

            ^Charging for a scarce resource (parking spaces) will allow the more efficient allocation of that resource (transferring the use of the space from a driver who places a low value on it to a driver who places a higher value on it). That’s the efficiency gain.

            There’s the allocation of resources, using supply and demand pricing.
            And there’s “tax the choke point”.

            They’re not at all the same, although those who are looking for money to fund their pet project would conflate them.

            The tenuous relationship between transit and automobile usage is for another day.

  3. Pingback: Today’s Headlines | Streetsblog San Francisco

  4. dw shelf says:

    ^nobody’s got verifiable models for “makes sense”.

    When it comes to transportation systems, we have an entirely verifiable model. We analyze the question, “does the system compete effectively with the American automobile based transportation system”.

    When it competes (as airlines sometimes do), it makes sense.
    When it doesn’t compete (as government buses and trains seldom do in the USA), it doesn’t make sense.

    There do exist some private bus lines, which seem to be making some sense, as verified by “they seem to be showing a profit”. Now that’s a verifiable model.

    • Alai says:

      Well, automobile-based transportation systems don’t compete effectively with the American automobile based transportation system, either: it’s hard to compete when your competitor is subsidized. For example, see bankrupt and bailed out toll roads, and the Yankee stadium parking garages (bankrupt despite being also subsidized).

      You’ll say we should remove all the subsidies. And I agree. But if it’s not happening, what do we do in the meantime? Subsidize autos and leave everything else to fend for itself?

      • dw shelf says:

        ^You’ll say we should remove all the subsidies. And I agree. But if it’s not happening, what do we do in the meantime? Subsidize autos and leave everything else to fend for itself?

        The way to rid ourselves of the scourge of subsidies it to work to end subsidies. Surely it’s illogical to suggest that one subsidy means we need another?

        Maybe we could start with General Aviation airports, which are massively subsidized, and used almost exclusively by the very wealthy.

        • Nathanael says:

          If you’re going to change anything, you have to start with the big-ticket items; no point in screwing around with small change.

          That means you start with cutting the general-fund subsidies for road-building. Good luck.

        • Nathanael says:

          And in fact, it IS logical to say that one subsidy (which we can’t get rid of) means that we need another.

          Good luck figuring out how to remove the road subsidies. Maybe it’s possible.

    • V_Taylor says:

      The primary goal of public transit is to transport people, not to make a profit. It’s a public service, like the public library or police and fire service. The assumption is that is won’t make any money at all. Public transit actually does recover some of the cost through the farebox, so in this regard it’s unique among public services.

  5. david says:

    The proper way to implement GG bus local rides in SF or the East Bay is to start by honoring Muni and AC passes via Clipper. The real stumbling block will be the driver unions who will fear their fraternal union members “poaching” their work. Of course, we the rider/taxpaying citizens don’t care what color the bus or driver’s uniform is, we just see a less than full bus passing by while we wait..

    • Definitely, though a wrinkle is that GGT charges more for fares than the other buses. Perhaps GGT could deduct the typical Muni or AC Transit fare from its base fare ($4 within SF, $3.60 in the East Bay) and charge for the balance. Keeping a higher price point would help manage the potential crowds.

    • John Murphy says:

      If a bunch of riders started taking the various Marin/Sonoma routes to get from downtown to the Marina, it would be a disaster. The 72 leaves SF nearly full as it is. Now there is a nice express route to the Marina to substitute for the crappy 30/45? The bus would fill downtown with Marina District bound riders and cross the bridge empty.

      And slowing the return trip home would shift people to cars. Right now there is an overlay vs a single occupancy car because of the carpool lane in San Rafael. But the buses take a long time to get across San Francisco – I can ride my bike from Caltrain 4/K to the Golden Gate Bridge 10 minutes or so faster than the buses. If the exit of SF gets too onerous, and potentially a full bus – might as well drive.

      Charging for the Park and Ride in Rohnert Park would just result in people parking in the neighborhoods at the stops 1-2 upstream from the Park and Ride.

      • Oh definitely. The peak-only lines shouldn’t be used for anything but North Bay commuting given how full they already are, but the basic routes and Route 92 could be pressed into service.

    • V_Taylor says:

      We used to be able to take the ferry and get a special transfer which allowed a free ride on Muni. That was eliminated about two years ago, but Clipper is programmed such that a Muni trip taken to or from the ferry is discounted by $0.50.

  6. FDW says:

    Actually, I’ve got a really simple idea: Lobby to put some kind of tax for Marin and Sonoma Counties that would pay for the complete restoration of the 2002-04 cuts on the ballot.

  7. V_Taylor says:

    First, note that as bus ridership has sagged, ferry ridership has soared, at least in part because of the faster ferries. I don’t know, but suspect, that bus riders shifted to the boats when pier-to-pier travel time dropped from 45 minutes to 30 minutes(Larkspur). Once a commuter finds that the ferry works for them, there’s almost nothing that will make them move back to the bus, because the user experience is so far superior to the bus.
    The other most important factors for bus ridership are the cost of gasoline and levels of employment. As gas prices go up, people switch to transit. When employment is down, there are fewer commuters. The graph above is consistent with these three factors: ferry improvements, gas prices, and the recession/depression since 2008.

    That being said, a few comments…

    1. Activate rear-door Clipper readers – Yes. Why install them and then not use them?
    Low-floor buses – No. These are commuter buses with rides of up to an hour. They definitely have the right vehicles for this use. You want to raise ridership, right?
    2. Coordinate buses with ferries – absolutely. Why this hasn’t been done is a mystery, but I think it has something to do with buses to the ferry being considered inter-county or something. That is, Marin Transit is somehow prevented from actually serving the ferry. Needs more research. Buses to the ferry should be tightly timed, and should run straight up and down Sir Frances Drake Blvd and on 101 as far as Rowland Plaza, stopping at parking lots and Park and Rides. This service could be paid for through charging for parking at Larkspur Landing.
    3. Charge for parking at Park and Rides – No. We want people to park there and take transit. There is NO transit to transit. The cost to run buses up every hill and dale is prohibitive. Charge at the Ferry – yes, and use the money for feeder buses.
    4. Raise the bridge toll – No. It’s already excessive and presents a hardship for those who have to drive into the city (tradesmen, for instance). The gap between the cost of taking the bus vs. the total cost of driving in (gas, wear and tear, lost time to do other things while commuting, and parking charges) does not need widening.
    5. Picking up in-city – No. That would put GGT in competition with Muni, so it will never happen, but besides that, the riders wouldn’t stand for all the stopping to let people off. The ride is long enough as it is. Again, you want to raise ridership, right? Degrading the passenger experience is not the way to do it.

    The selling points of the bus are the low cost and reasonably fast trip into parts of the city not reachable from the Ferry Building. It will never compete with the ferry for those who can use the boat to get to work, because of the difference in the user experience between the two modes.

    • Re: 1 – The commuter buses are doing fine and shouldn’t be changed, save for large package rules (while currently banned, could be changed to say luggage, etc, can only be in the one-seat rows toward the middle of the bus & can’t block the aisle). I’ve been hassled by drivers for taking transit from SFO, for example. There’s no reason for a local Marin-only bus to be high-floor, however, or to ban large loads.

      Re: 2 – Duplicating Route 101 to Novato would do the trick, though without much schedule padding. God help the driver who must endure passengers who missed their ferry b/c of traffic! Rowland might be close enough to ensure punctuality, though that would limit the usefulness of the route outside ferry shuttle service.

      Re: 3 – If the parking lot is full, then it’s being used for storage more than access. Those who can take transit to transit probably will if the lot charged. If the parking lot isn’t full, parking pricing will only be a net loss to ridership, so shouldn’t be applied there.

      • V_Taylor says:

        1. OK, so to untangle a little, GGT buses are all commuter buses on 101. The local service is Marin Transit. Sorry if I misunderstood your original post, but it sounded like you were referring solely to GGT service. I think Marin Transit does use primarily low floor vehicles.

        2. Shuttles to the Ferry must be able to run on time. If it was a few minutes late, I could see holding the boat for it. The other logistical issue is a bus coming in with a ferry-shuttle load of 30 passengers. To insure they can get on the boat, dockside management would have to hold boarding at [capacity-30] to insure that those who took the bus to the boat (behavior we want to encourage) are guaranteed a seat.

        3. I’m not sure which lot you’re referring to. The Larkspur lot is packed to overflowing by 8:15 am, because there is no way to get to the ferry except by car, unless you’re a hale and hearty biker and the weather is at least decent. Why GGF hasn’t been sued for social inequity, I don’t know.
        As for Park and Ride lots, I don’t think they are full, but I’m not certain, and it probably varies by lot. The very small lots on the 101 corridor at the bus pads do fill up pretty quickly, as does Manzanita, while Marin City will never fill as it’s just huge.
        The primary mode in Marin, as you know, is the car, and our land use pattern reflects that. Until we can get some infill housing going, I believe we need to have an interim system which accommodates people driving to intercept lots and then taking transit on main corridors. It’s fine, and it’s better than trying to provide transit service to sprawl.

        • re: 3 – Yes, I was referring to the park & ride lots along the freeway. Larkspur and, I’m sure, Sausalito, fill up quickly, as do a number of other lots. But something like 68% of Marin’s population lives within a quarter mile of a bus stop, and that goes up into the high 70s for a half-mile. Some of those people could take a local bus to the trunk line and free up space for their transit-bereft fellows.

          • Valerie Taylor says:

            I personally live about 1/4 mile from a bus stop, but unfortunately service at that stop is not useful in getting me to a transit center or to the ferry. As someone commented above, the bus needs to compete with the car. That bus line takes 30 minutes to make what is a 5 minute drive. I suspect many others are in the same situation. Some of GGT’s most successful routes start on collector roads in neighborhoods and then get out onto 101 to SF, thus those passengers have a one-seat ride to the city.

          • Alai says:

            And why would they take the local bus?

            You get a choice: wait for the bus, spend more time, and pay $4 for the round trip– or drive, and park for free. Only the foolish would take the bus, unless they had no other choice. The ridership numbers reflect this.

            Transit can work very well– in situations where it’s more advantageous to take transit than to drive. Often, this is a result of space being at a premium at one or both ends of the journey, such that parking a car is expensive.

            If Marin is committed to providing ample and free parking everywhere, transit will never be competitive. Simple as that.

  8. guess says:

    As a traveler heading to Mill Valley for an upcoming event (not my choice), I must say, your transit system completely sucks. We’re forced to rent a car, because your fixed-schedule bus routes are slow and very poorly timed. There are some events that we absolutely would not be able to get to because the buses simply don’t run (even though said events are on a bus route). And the taxi fare, just for a 2 1/2 mile trip, is $10. Highway robbery, especially in a much less congested suburban area (relative to SF). So, we’re forced to rent a car. And drive to a park and ride from our hotel, so that we can take a bus to SF for sightseeing. You know, since parking isn’t feasible in SF, and the bus service to our hotel (Holiday Inn Mill Valley) is lousy. Funny, even the transit service in my hometown (outer-ring suburb 40 miles from Seattle) is better than your transit service – and I consider that service pretty poor.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: