Human Transit

portland: balance the budget yourself

Portland's Tri-Met faces another horrible funding shortfall this year, but they've come up with a good survey tool to engage the public in their decisions about what services to cut.  It's one of those "balance the budget yourself" tools that's becoming increasingly necessary to bring voters into contact with reality about government budgets.  

If you live in Portland, you should definitely work through the survey and send them your own balanced budget and comments.  If you're not in Portland, is your transit agency communicating about its trade-offs this well?

los angeles: deleting some lines can be fair

The New York Times today bewails the loss of Los Angeles bus line 305, which soon will stop running diagonally across the city's grid, from Watts to Beverly Hills and Westwood. 

NYT reporter Jennifer Medina assumes this is purely a victimization-of-the-poor story, starting with this observation:

The 305 was one of several lines created under the consent decree, and it is the only direct route from the city’s impoverished southern neighborhoods to its affluent West Side, where legions of janitors, nannies and maids work each day.

Sounds sad, and it's easy to fill an article with interviews with 305 riders who will experience the deletion as a hardship.  But as that paragraph should warn us, 305 was a symbolic service.  It cannot have been relevant to very many people, not even to many people in the targeted demographic ("janitors, nannies, maids" according to the NYT).  Why?  If you explore the route and schedule [ Download PDF ] and look at how the route fits into the larger network ("System map overview" here), you'll notice:

• Line 305 is a diagonal shortcut across a high-frequency grid, where trips between anywhere and anywhere can usually be made on lines running every 15 minutes or better with some are far more frequent than that.  Meanwhile, the Line 305 frequency is every 40-60 minutes.  [PDF]  That means that the 305 is the fastest path between two points on the line only if it happens to be coming soon.  If you just miss one, there's another way to get there faster, via the much more frequent lines that flow north-south and east-west across this entire area. 
   
  
• The 305's low frequency exposes its riders to the risks of waiting for a single bus: you're basically making an appointment with one driver who may not show up for a variety of reasons.  Routing the same trips via the high-frequency grid means much higher reliability, because the abundance of buses along a line means you are less dependent on any one of them.
   
  
• Most important, the alleged target demographic -- trips from the "poor south" to the "affluent west" for domestic workers -- was mostly not served by the 305.  Both the "poor south" and the "affluent west" are enormous areas.  So no one bus line was ever going to connect all or even most of the "poor south" with all or even most of the "affluent west." 

These points, but especially the last, identify a public transit service as symbolic.  Symbolically, the 305 links the "poor south" and the "affluent west," and thus helps everyone feel good about having served domestic workers.  In fact, the 305 runs through a small part of the vast "poor south" and a small part of the vast "affluent west," but it's still useless for most of the people making that kind of trip, because both areas are so large that no one bus line, or even five, could link all of the likely origin-destination pairs between them.

(You could take other buses in each area and transfer to the 305, but the low frequency of the 305 makes this very risky.  Once you've accepted the need to connect, you might as well ride along the main grid and connect with a high-frequency line to take you where you're going.)

This problem is why frequency and connections were invented.  The governing principle of transit in these core parts of Los Angeles is the high-frequency grid, which allows everywhere-to-everywhere travel at high frequencies with at most one connection.  Yes, it may be sad that some domestic workers who are used to zero-transfer trips are now going to have a one-transfer trip, but that only means that 305 riders will have the same level of transit mobility that everyone else has, including most domestic workers.  It also means that Los Angeles transit will be treating all of this demographic equally, rather than arbitrarily preferring people whose path happens to lie along Line 305.

The other moral of this story is even simpler: If your mission is to serve a whole city or region, designing transit routes around any self-identified group of people is almost always a bad idea.  Most successful and attractive transit seeks maximum versatility, by serving the most diverse possible range of demographics, trip purposes, and origin-destination pairs.  You can make exceptions where a single demographic group produces sufficiently massive ridership, as in some commute markets.  But in general, the way people self-organize and self-identify politically is a bad guide to how to meet their transit needs efficiently.  Everyone can draw the perfect transit line just for their interest group, but such proposals tell you nothing about what a good transit system would look like.

Nobody should be happy about the severe cuts being imposed on many US transit agencies that urgently need to move in the opposite direction.  But as in San Francisco in 2009, cuts are sometimes an opportunity to delete services that have passionate, well-connected defenders, but that simply don't make sense if your goal is a complete network that people can use to go wherever they're going.

san francisco: may 2010 service cuts mostly restored

As I said at the time, the first round of service cuts in San Francisco, implemented in 2009, actually did some good, by deleting some segments that for geometric reasons were always going to be ineffective.  But the second round implemented in May 2010 were mostly just painful.  Now, after a long struggle, that destructive second round is being reversed, mostly on September 4 with the remainder to come in December.  I often criticize journalists for featuring bad news but missing the corresponding good news, so it's only fair to do the same myself. 

willingness to pay for transit improvements

Do your city's political leaders understand what funding sources people would support if they knew what they were buying?  A few weeks ago, the Source (a blog by the Los Angeles transit agency Metro) reported on a survey showing that current riders would pay 50c more in fares for a doubling of their frequency of service.  This isn't as encouraging as it sounds, because a doubling of frequency, even with significant ridership increases as a result, will cost a lot more than 50 cents per new rider.  But it's a useful soundbite.  These questions, broadly called "willingness to pay" questions, need to be asked more, and more probingly.

In the recent Sydney Morning Herald Inquiry into Sydney public transport, we did some deep probing of public opinion on willingness to pay.  Our survey experts at the University of Sydney's Centre for the Study of Choice took 2400 randomly selected people through a series of questions designed to observe the value they attach to different kinds of public transit improvements, and what taxes or fees could be higher, and by how much, if the money were clearly going to those improvements.  The methodology, called a discrete choice experiment, is designed to avoid some of the usual biases in these questions.  The method and its results explained toward the end of Chapter 1, here.

Ultimately, these findings determined the shape of our recommended 30-year investment plan.  The most fundamental recommendations are (a) that Sydney needs  a major transit investment program, funded largely by a range of increased state taxes and user fees, and (b) the plan must be dominated by transit, not roads. 

We made these recommendations not just because of our own analysis and desires, but because that's the overwhelming preference of the public.  Throughout the project, we recommended only things that could be funded using sources that, according to the survey, would get majority support, if the funding was clearly tied to the improvements.  We concluded that if voters could vote up or down on that package of fees and taxes, to be used exclusively to build the specified transit lines, a majority would say yes.

For better or worse, Australians are not used to voting on specific initiatives.  There have been national referenda on major national issues, including famous ones such as the granting of full rights to Aboriginals (passed in 1967) and the creation of a republic in place of the British crown (defeated in 1999).  But there is no precedent for voting on state and local issues like how to fund the construction and operations of transit.  Most Australian transit agencies are not separate governments entitled to certain funding, as they typically are in North America.  Instead, transit is managed by state government direction.  It's just one of many things funded out of the state's general revenues, and the elected government decides how that revenue is spent.

During the last year, watching the state and local budget crises across the United States, I've come to appreciate the stability of the Australian approach.  Popular votes on specific taxes and spending have tied the state of California in knots, to the point where the task of government is almost impossible.  American transit agencies that rely on a specific tax, usually some form of sales or payroll tax, experience huge swings in revenue, so at least once a decade they find themselves slashing service.  This doesn't happen in Australia, or at least not as visibly.  Service doesn't improve dramatically in better times, and isn't slashed as badly when times are bad.

But the Australian system is not so good at expressing a public consensus to raise taxes and fees for a specific purpose.  Such a consensus clearly exists in Sydney to move forward on major transit investment.  If the voters could set up a funding stream to do this and an agency tasked with doing it, as would happen in many American states, then we would be on our way.  Instead, we have to wait for one of the competing major parties to embrace such an investment plan.  The parties, however, aren't focused on the whole of public opinion; they're focused quite narrowly on marginal seats, the relatively small number of voters whose judgments will turn the election, and this is a very different polity than the city as a whole.

Sooner or later, I suspect Sydney will need to have some sort of referendum to put the funding sources in place to fund a decent transit expansion program.  Based on our work in the Inquiry, which assessed the public's willingness to pay and then crafted an infrastructure plan to fit it, such a referendum could pass.   But I certainly don't see the way to that in the Australian political system, nor does either of the major parties seem ready to take the lead.

seattle suburbs: the silence of sundays

Community Transit, which serves most of the northern suburbs of Seattle, is shutting down completely on Sundays.  This wouldn't be unusual in a small-city transit system, but CT's service area (most of Snohomish County) is a big suburban expanse with about half a million people.  It has enough transit demand to support a low-end Bus Rapid Transit line, called Swift, which will presumably not run on Sundays either.

This is a fairly dramatic step by North American standards.  Local transit in suburban areas generally appeals to people with few choices, but many, many of these people work in low-wage jobs in the service sector, such as restaurants and big-box retail.  These business are open seven days a week and often are often busiest on weekends, so most of their employees have to do some weekend shifts.  A transit system that doesn't run on Sundays will no longer be useful to these people.  Based on what I've seen elsewhere, most of them will find other arrangements, and CT is likely to lose them on all five days a week that they travel, not just Sunday.  Some, those without any good transport options, may lose their jobs. 

I hope CT or some other local government researches what happens to these riders when Sunday service ends.  The best approach might be to survey the Sunday riders before the service stops, asking them for follow-up contacts so that they can be interviewed again a few months in the future.  This would not only provide good data for other agencies facing the need to cut service, but would also be a nice way for the agency to convey some concern for the well-being of these customers.

new york: free student fares and the budget crisis

Today's New York Times editorial on New York's transit funding crisis goes beyond general warnings about the need to fund transit, and digs into specifics: Raise bridge tolls, resist election year tax-cutting, hold the line on pay raises, and this:

What it should not mean is doing away with free passes for needy students. Both the state and city will have to contribute more to help pay $214 million a year to help keep these students in school.

Perhaps they read this.

portland: 2010 service cuts announced

Ouch.  Here is TriMet's proposal for the next round of service cuts, due to take effect on September 2010.  As the agency says on its budget: the big picture page:

A recent reassessment of our income projections compared to actual income shows that payroll receipts are down $15 million and passenger revenues down $8 million. To maintain fiscal stability, we again need to reduce expenses—this time to the tune of $27 million. The cuts we've already made ($50 million since 2001) have reduced our options for future cuts—we've already cut the easy things. Our remaining expenses are largely determined by factors beyond our control, including contract commitments, fuel prices and Portland's rate of economic recovery.

The 21% share paid by fares (called farebox return or Farebox Recovery Ratio [FRR]) is not bad for a city of Portland's size and density.  More critically, by the agency's numbers the decline in fares received is 30% of the total shortfall to be covered.  That suggests a real danger of cuts continuing to drive down ridership, leading to a downward spiral of cuts and ridership losses that would gradually destroy the agency's relevance.

This risk may have given some impetus to the proposal to cut low-ridership "Coverage" services even where the few riders have no alternative.  Four routes are to be eliminated completely, and three of these (27, 154, and 157) are outer-suburban feeders into Gateway, Oregon City, and Clackamas, respectively.  (See Tri-Met system map here.)  All serve relatively low density areas but not especially affluent ones, a reminder that density determines ridership much more than wealth does.  There's not much of an alternative for residents of the areas served, especially Willamette and Happy Valley, but if good transit service were really important to you, you wouldn't live there.

They're also taking the opportunity to eliminate some awkward deviations caused by transit-hostile development design.  A dramatic example is the Cornell Oaks employment area on Line 67.  It's on the south side of the US 26 freeway on the east side of 158th Avenue.  Have a look:

The north-south street on the far left is 158th Avenue.  The core street of the development, Greenbrier Parkway, comes off of 158th, does a few S-curves that we're supposed to find sensuous or relaxing, and then ends in a landscaped turnaround loop, a violation of the "Be on the Way" rule.  (Actually the cul-de-sac has a connection to a parking lot, which connects to another street, all of which looks like it's designed to permit fire trucks but obstruct buses.) 

Currently, some Line 67 trips, which are running past on 158th, actually turn down this street, drive all the way down, and then, having no alternative, turn around and drive all the way back.  They do this while carrying through-passengers who just want to ride along 158th.  This is called a "backtracking deviation," and psychologically it's the very worst kind because the through-riders can't even pretend they're going in roughly their desired direction.  It's especially bad here because we're at a point where the bus is relatively full (by the low standards of suburban local routes.)  There's a MAX station just to the south off the map, and a lot more suburbia to the north, so everyone wanting travel between MAX and that suburbia gets hauled through the Cornell Oaks detour.

The Cornell Oaks deviation is a typical example of how transit agencies are forced to wear the costs of poor design, so bravo to TriMet for refusing to do it anymore.  While some people who work in these business parks probably want service, their developer did everything possible to prevent bus service from working.  One solution you'll hear proposed for places like this is for the businesses to run their own employee shuttle to the nearby MAX station (not far south off of this image).  Doing so would reflect the fair costs of their decision to locate in a place where efficient public transit is impossible.  

Still, these cuts are a small part of the damage.  The main damage is further loss of frequency on the former "Frequent Network," now including midday MAX light rail frequencies as well as buses.  This is really where the downward spiral of cuts and ridership could play out.  I have one strong suggestion about these cuts: Given that you've already given up on the 15-minute frequency standard, try to hold the line at 20 -- exactly 20.  TriMet is currently being vague about what the exact post-cut frequencies would be, saying, for example, that on certain routes the elapsed time between buses "would increase by 2-10 minutes at various times throughout the day."  It's impossible to know what this means, but it could mean that we'll see lots of important lines running every 17, 18, or 19 minutes.  My view is that 20-minute frequency is actually better than many of these compromises, because it yields a schedule that you can remember because it repeats the same way every hour.

The cuts will also undermine the historic usefulness of transit for intra-downtown travel, which was at one time an important feature of downtown Portland as a place to work, shop, and play.  While the old transit mall and fare system made it easy to grab just any bus that came along, the new transit mall has very widely spaced stops, so you can no longer easily walk to the stop of whichever bus you see coming if you want to travel a short distance along the mall.  More critically, the downtown free fare zone now applies only to rail, and you're expected to use light-rail for these short trips.  Midday frequency of light rail will have an impact on this.  More on the peculiar mall shuttle situation in another post.

unemployment and the transit imperative

As American leaders suddenly pivot to focus on unemployment, Emily Garr at The Avenue picked up on a line of President Obama's recent speech in Lorain County, Ohio.

You can't get to work or go buy groceries like you used to because of cuts in the county transit system.

She goes on to describe cuts in the transit services to this suburban county that are definitely not "trimming the fat" but more like multiple amputations.

Transit advocates need to be picking up this line.  Back in the mid-90s, when welfare reform was timely, I routinely ran passenger surveys on various transit systems as part of planning projects.  The surveys had many other purposes, but I made a point to ask both "what is your trip purpose?" and "if transit had not been available, how would you have made your trip?"  A common answer to the second question was that the person would not have been able to make the trip.  Cross-tabulate that with a trip purposes of "work" and you get a count of people who could not hold their jobs without public transit.  It's an easy thing to do in any customer survey, and every transit agency and advocate should know this number. 

It's also important to notice that the people who are on the verge of not being able to hold their jobs are mostly in relatively low-wage jobs in the service sector -- restaurants, fast food, big box retail, etc -- and these people are commuting all day and much of the night.  Transit that supports high employment is all-day service, not just peak service aimed at the generally better-off 9-to-5 commuter.

new york: cutting service without cutting mobility

The U.S. service cut epidemic -- the result of transit agencies relying on recession-sensitive funding sources -- is about to touch New York again.  But as in the first round of San Francisco cuts, the New York MTA planners are doing their best to reduce service but not mobility.

As with everything in New York, the details are complicated, but here's a simple example.  The G subway line is the only line running directly between Queens (the top of this map) and Brooklyn (the bottom).  It has long included a segment from Court Square to Forest Hills where G trains run alongside two other subway lines, the E and V which connect Queens directly to Manhattan.  (Full subway map here.)

In happier times, a transit agency could indulge the preferences of people from Forest Hills who want to get to Brooklyn without having to change trains.  But when service needs to be cut, it's far better to cut this duplicative segment than to cut a segment that is somebody's only service.  And that's usually the real-world choice.

So in future, the G will run only as far as Court Square, and passengers who want to get to/from Forest Hills and Brooklyn will need to use the E or V between Court Square and Forest Hills, then change.  But as a result, frequencies and unique services will not be cut as much as they would have been otherwise.  That's the only way to get through a service cut without reducing mobility.

Full coverage of the service cuts is at Second Avenue Sagas, which often features thoughtful and detailed commentary on New York transit issues.  Thanks to Alon Levy for the tip.

san francisco: "the fuse has been lit"

Updated Jan 16

The next round of San Francisco service cuts have been announced, or as commenter Ted King puts it, "the fuse has been lit."  For local coverage see the SF Chronicle and Streetsblog SF. 

Here are some of the most interesting points from the budget summary (via Streetsblog, not the Chronicle):

Although the budget hole to be closed is $16.9m, the service cuts are only $4.8m.  That's impressive.  They achieve so much non-service savings by a whole pile of cuts to other things, designed to have wide but manageable impacts.  Labor takes a ping: not just 0.7m in "concessions," but also charges for parking at the workplace.  (Since a huge share of the drivers report to work around 4:30 in the morning, many don't have good transit options.) 

The service cuts are all frequency cuts, and they affect every single line.  Unlike the first round of cuts, which pruned and shaped the network by removing weak branches and segments, this round is intended to spread the pain without cutting anyone off entirely.  They're even cutting peak-hour commuter services, with the exception of rail.  Here's how they describe the increase in headways (elapsed time between consecutive trips on a line) by service type:

"Peak AM & PM"

Rapid & Express Bus: 1-2 minutes
Rapid Rail: No change
Local Bus: 1-5 minutes
Community Bus: 5-10 minutes

"Midday & Late Evening"

Rapid & Express Bus: 1-5 minutes
Rapid Rail: 1-5 minutes
Local Bus: 2-10 minutes
Community Bus: 5-10 minutes

Virtually everyone will wait a little longer, not just for their trip but also for any connections.  Crowding and pass-ups will get a little worse. 

There's a range of fare changes, of which the most interesting is the expansion of the range of "premium" fares. 

There has long been a premium cash fare for cable cars (currently $5 instead of the standard fare of $2), because they are very expensive to operate and cater to tourists, but as compensation to the locals who rely on them, the cable cars accepted monthly passes at the standard rate.  Now, a premium monthly pass will be required.  The same more expensive pass will be required on "express" routes, which are mostly peak-only, one-way, and thus relatively inefficient.  This, like the frequency cut on those same services, will hit an especially high-end demographic. 

Most fascinating, to me, is the proposal to charge the premium cash fare ($5 instead of $2) to ride the F-Market & Wharves streetcar.  (Confusingly but understandably, they do not propose imposing the premium pass charge, so monthly pass users will see no impact.)  When the F was created, it was made very clear that this was not a tourist ride like the cable cars; it was to be an integral part of the network.  It replaced two bus lines (8-Market and 32-Embarcadero), and that was acceptable only because it was understood to be providing the same or better mobility at the same cost. 

Now that deal's changing. Cash passengers, at least, will pay something extra to ride a streetcar instead of a bus or subway, so we'll start to see what value people attach to the streetcar's distinctiveness and amenity.  (If you don't know why I'm curious about this, see here.)

As it happens, many F riders have credible alternatives: most can walk to the metro subway stations along Market Street, and a few can walk to Haight Street buses.  But with the deletion of bus service along Battery/Sansome Streets in the last change, there's really no good alternative for riders on the Embarcadero segment, including anyone traveling between the Financial District and Fisherman's Wharf.   So people who need to make these trips will be hit with the higher fare whether they want to ride a streetcar or not.

If this change prevails it's a major conceptual shift in how San Francisco treats its historic streetcars.  And it will be interesting, too, to see what a dramatic fare penalty does to the 20% of the streetcar's ridership that uses cash fares, especially along Market where there are alternatives. 

This may well have large impacts on the viability of any extensions to the historic streetcar network.  With this fare change, the agency will be saying that the streetcars are conceptually like the cable cars.  The cable cars' fundamental justification is the experience of riding them, not the transportation, and for this reason there is little serious prospect of extending them.  Is the same true of the historic streetcars now?

Finally, It's important to say, over and over, that these huge cuts aren't San Francisco's doing; they're the result of the combined effect of an overall economic downturn (which hits the sales taxes on which much California transit funding relies) and California's profound budgetary and governance problems.  But nothing will be done about those problems until enough people feel the pain.