Human Transit

"abundant access": a map of a community's transit choices, and a possible goal of transit

In my book Human Transit, I argued that the underlying geometry of transit requires communities to make a series of choices, each of which is a tradeoff between two things that are popular.  I argued that these hard choices are appropriate assignments for elected boards, because there is no technical ground for making one choice or the other.  What you choose should depend on what your community wants transit to do.  Examples of these choices include the following: 

• Ridership or Coverage?  Should transit agencies be maximizing their ridership, which is easily done by abandoning service to low-ridership areas, or should they spread out service throughout their service area so that everyone has some service, despite the predictably low ridership that results. (HT Chapter 10)
• Connections or Complexity?  If you try to design a transit network in which people won't have to transfer, you'll end up with a network with high complexity and low frequency.  I first explained this here, and later, and more rigorously, in HT Chapter 12.
• Peak-first or all day?  While it's widely assumed that transit is most productive during the peak, this is not always the case, because service running only on the peak, and especially only in one direction, can be massively expensive to operate.  Should the peak commuter be the primary focus of a transit network even given these costs, or should a transit agency seek to build an all-day network of services that encourages reliance on transit at all hours, while still meeting peak capacity needs?  (HT Chapter 6).
• How far do you assume people will walk?  Assuming higher walking distances allows transit lines and stops to be further apart, increasing both speed and frequency.  Research has found repeatedly that people will walk further to better service, though when you first propose to increase walking distance they will complain.  This is an especially tough tradeoff because frail seniors and disabled people are small in numbers but have a particular aversion to walking.  (HT Chapter 5).
• Should transit be protected from congestion?  Doing so, by giving transit a lane (or trackway) not shared with traffic will dramatically improve reliability and thus the degree to which people will find they can trust the service.  However, this requires a city to display a profound commitment to transit in the allocation of its streetspace.  (HT Chapter 8).

Note that while the framing of these choices tends to make them sound binary ("Ridership or coverage?") the choice is not between two boxes; it's of a position on the spectrum where the two terms define the extremes.  You don't have to choose between ridership or coverage, for example, but you have to choose a point on the spectrum between them, and you have to accept the mathematical fact that, as on any spectrum, moving toward one extreme implies moving away from the other.

In the Epilogue of HT I briefly introduced the idea that for each of these choices, one option seems to trigger a positive-feedback loop, while the other option does not.  Here is a new and expanded diagram of that idea.  Click to enlarge, or download here:    Download Abundant access diagram  (Tip: Print it on A3 or 11x17.  There's a lot here.)

What do I mean by positive feedback loop?  On each of these choices, the option that points toward the center leads to a particular kind of network that supports all of the goals implied by all the centerward choices.  For example, if you plan your network for maximum ridership rather than coverage, you also generate a network that encourages higher walking distances, that supports easy connections rather than complexity, and that tends to present the strongest case for facilities that protect transit from congestion.  

On the other hand, the options that point outward from the center tend not to support each other.   You can achieve the goal implied by these outer choices, but (arguably with some exceptions) each one will require a separate effort.  For example, if you believe walking distances to transit should be low, as seniors tend to advocate, then you need to put parallel transit routes -- and the stops on those routes -- very close together.  This does nothing to advance any of the outer "outward" goals, whether it be growing peak demand, or avoiding connections, or achieving a luxurious experience that will compete with a BMW.

Another way of saying this is that the choices that point inward all point to a single set of network design principles, and a single set of consistent priorities, while the choices pointing outward lead to many different kinds of network design and competing priorities.

What happens when you move toward the center?  These choices converge on a single, consistent goal that I propose to call Abundant access, which means:

The greatest possible number of jobs and other destinations are located within 30 minutes one way travel time of the greatest possible number of residents.  

[Why 30 minutes?  The question is tangential to my point, and other figures could be used, but if one has to pick a figure the most solid basis is Marchetti's constant [PDF], the idea that humans throughout history have tolerated about one hour of daily travel time.  Obviously, shorter or longer periods may apply to trips other than the commute.]

For any individual, abundance of access can be visualized using a map of isochrones, like the Mapnificent.com travel time maps, which show the area you can get to within a fixed amount of time on some combination of transit and walking.

 

Of course, any quantification of abundant access must consider not the area but the amount of stuff in it: jobs, retail, recreational opportunities, hospitals, all the things that form the universe of destinations.  As a matter of public policy, it must also maximize over the largest possible share of the population.  In the case of transit, the population over which abundant access is most cost-effectively maximized tends to be a more urban, high-density population, because among readily available measures, density is one that best predicts the intensity of ridership that will arise from a service investment.  (Fortunately, that's fairer to than it sounds, because people living at high densities use much less road infrastructure per capita than people living at lower densities.)

The goal of abundant access has several kinds of appeal.

First, it can be measured objectively without recourse to psychology or culture.  Ridership estimates are based heavily on travel times that approximate the notion of abundant access, but they also add psychological factors that are less stable, such as observed preferences for particular technologies.  These factors may be emotionally vivid, but like many emotional factors they are likely to change with time and especially with generations -- just as emotional attitudes toward cars are changing now.  Abundance access measures a fact that is entirely objective -- travel times.  Unlike emotional reactions to technologies, the value of access has been constant across millennia of human experience.  

Second, abundance of access is literally a quantification of freedom, in the sense that matters to us in transportation.  Isochrone maps like Mapnificent's, in particular, show us our freedom in a very immediate way: here is where you are free to go, now.  Abundant access measures the transportation element of opportunity of all kinds, which is one of the main reasons people have moved to cities since their invention.

The concept of freedom is sadly undervalued in much urbanist discourse, and I am always looking for ways to reintroduce it.  Much urbanist writing, for example, is blatantly prescriptive ("you should want this kind of community"), which feeds conservative stereotypes of urbanism as manipulative or coercive.  We need to be able to talk not just about ideal communities but about freedom and personal responsibility, a frame in which all the great urbanist ideas -- and all the urgent environmental imperatives -- can be stated equally well.  In that frame, the key idea is not "the good" but "choice," where freedom of choices also implies responsiblity for your choices.

So again, all of this work is descriptive, not prescriptive.  I'm not saying that you should like the goal of abundant access; that would be value judgment, and you get to make those for yourself.  

I am saying, though, that abundant access as an idea has certain features and consequences, including a tendency to be self-reinforcing.  As mapped along the various axes of choice, abundant access is a single consistent vision whereas the opposite choices lead to many unrelated visions.  

Still, you don't have to like it.  My role, as always, is to describe consequences clearly, and help people choose.    

can "network primers" reduce grief about network design?

Greater Vancouver's transit agency TransLink has now published something called the Managing the Network Primer [Download PDF] and has a new webpage devoted to this and similar themes.  It's a simple explanation of why effective transit networks are designed a certain way.  It pushes back on the confused and confusing notion that "buses are flexible compared to rail," by emphasising the ways that all kinds of transit work or fail along the same geometric principles.  It can also be used to explain to a reasonable person why the bus can't just deviate to serve (or avoid) her house, business, or development project.  I'm honored to have made substantial contributions to it, in the context of my ongoing work with TransLink.

Subliminally present in this document -- and more explicit, I hope, in the next version -- is that this is also a primer for transit-oriented land use planning.  If you want good transit at your home or development or senior center or business park, you need to locate it where transit can be efficient, and therefore abundant.  So by understanding what efficient transit looks like, you can get a better grasp of what truly transit-oriented development looks like.

So like my book, and like my course, this document will be interesting for transit geeks but is really meant for people in land use planning, development, architecture, and other related fields.  These are the people who are deciding how effective transit can be in the future -- every bit as much as transit planners are.

Once again, TransLink is being careful not to tell any city what it should do with its land use.  Rather, this is part of a process of explaining the transit consequences of the choices a city might make. 

If your regional transit agency is giving its city governments the impression that it's telling them what to do, it may have something to learn from documents like this.  Getting this message right is something I've worked on for a while, and it was the subject of my recent keynote at the Canadian Urban Transit Association annual conference.   Let me know if I can help.

does suburban local service get cars off the road? (updated)

See new updates at the end, based on comments to May 4.

Ricky Leong in the Calgary Sun on why Calgary should spend more money serving far-flung suburbs:

Here’s the thing about providing transit to our far-flung suburban neighbourhoods. The benefits of new transit services will be felt by people everywhere in the city, not just those residents who choose to take the bus and train.

Fewer cars on the road should lead to better driving and cycling conditions for everyone, everywhere, for example.

Suburbanites choosing transit over private vehicles should also mean less wear and tear on roads, so the savings from reduced maintenance would be a net benefit to the entire city.

I'm not trying to embarrass Leung, because the error here is extremely common in journalism, as it is in public perceptions.  Last November Lisa Margonelli built a long New York Times article around the same mistake.

Buses circulating in low density suburban areas (as opposed to express to Park-and-Rides) can serve many valid purposes, but getting cars off the road generally is not one of them.

The universal fact that these people are missing is that buses circulating in low-density areas -- especially where street networks require transit to thread slow and complex labyrinths -- is a predictably low-ridership service.  Outside of a bit of rush-hour activity and sometimes a surge at school bell times, this is the fact of life about local service in low-density suburbs.  If you don't believe me, ask for your own transit agency's stop-level ridership data and compare it to the density and walkability of the area around each stop.  Apart from anomalies created by special land uses, you will find it is much higher in more urban parts of your community where these conditions are more favorable.

In 20 years of looking at  detailed transit data, I've never seen a local suburban bus route whose performance (ridership per unit of service cost) was anywhere near that of a frequent urban bus route operating in area whose layout is favorable to transit. 

("Favorable to transit" means (a) higher density, (b) gridded local streets that provide easy direct access to stops for pedestrians, (c) a safe and pleasant pedestrian environment, (d) a mix of land uses and (e) simple linear paths, usually arterials, where transit can operate efficiently and effectively.  These features tend to be found in older inner cities, but they can be replicated and in some cases even retrofitted.)

Indeed, because transit's market is mostly the area within walking distance of a stop, low density and obstructed street patterns are the very definition of a poor transit market that will yield much lower ridership on investment, and that therefore justifies a poor level of service. Poor ridership out of low density isn't an empirical fact; it comes close to being a mathematical fact, because in most cases density around a stop is the size of that stop's market.

To claim that this predictably low-ridership service will result in "fewer cars on the road" is thus a geometrically incoherent claim.  "Fewer cars on the road" is the result of people riding transit, not transit existing.  So a service that is guaranteed to generate few riders per unit of cost will is guaranteed to get fewer cars off the road per unit of cost, compared to one that is guaranteed to generate many ridership per unit of cost -- such as a line in transit-favorable geography or a rapid transit corridor.

The only high-ridership form of public transit that can serve low-density suburbs with obstructive street patterns is based on Park-and-Ride.  Where land values permit, abundant park-and-ride and fast radial services can get cars off the road  efficiently.  Radial rapid transit (bus or rail) is also good because new transit-favorable neighborhoods can often be built later around its stations.   Park-and-Ride needs to be understood broadly as also accommodating dropoffs, Bike-and-Ride, etc.  But the key thing it does not require is for a bus to actually drive around inside a labyrinth of suburban local streets.

If a transit agency's objective is to get cars off the road, then like any business you start by focusing on your competition's weaknesses.  The car is least convenient in areas of high density and good walkability, and geometrically these also provide the the highest ridership per unit of investment.  The one other area is the suburban commute corridor -- the freeway into the city -- where congestion during peak periods makes the car a weak competitor.  That's why peak commute services -- to Park-and-Ride, not to people's front doors -- is also a high-ridership prospect, and one that gets cars off the road efficiently.

Again, there's nothing wrong with running bus routes around in low-density suburbs.  But it is a low-ridership proposition and therefore can't be justified by many sustainability outcomes; it's justification has to lie in social needs and perceptions of equity among neighborhoods.  If your city really wants to get cars off the road, however, it's not a good way to serve that purpose.

This kind of confusion is why elected officials should be asked to think more clearly about how they want to balance the conflict between ridership-related goals (including lower subsidy, fewer cars on the road, and resulting sustainability outcomes) or coverage-related goals (including lifeline access and "equity" across all arts of a community.  Both goals are noble, but don't pretend to be doing one if you're really doing the other. 

Better still, ask elected boards to adopt policies about how much of the budget should be spent pursuing ridership -- which means running services where high ridership is the predictable outcome -- and how much should be spent pursuing coverage, i.e. distributing service everywhere regardless of low ridership.  The Reno area's transit board did this in 2005, as the result of work I did with them, and as a result, this whole conversation is much clearer, less personal, and more clearly tied to the community's actual values.

For a more detailed exploration of this fundamental issue, see Chapter 10 of Human Transit.

UPDATES:  Comments on this one are often critical, but what I see in them is that:

• People have different definitions of "suburban".  It borders on being an unhelpful word but we don't have a good word for it.  What I meant was: landscapes whose geography is intrnsically hostile to transit, specifically the absence of all or most of the following:  (a) relatively high density, (b) gridded local streets that provide easy direct access to stops for pedestrians, (c) a safe and pleasant pedestrian environment, (d) a mix of land uses and (e) simple linear paths, usually arterials, where transit can operate efficiently and effectively.
  
  
• As in almost everything, we're talking about a spectrum rather than a pair of boxes.  When I talk about dividing a budget between ridership and coverage, it does sound like I'm dividing the world into two boxes.  In fact, many services are in the middle, contributing partly to a high-ridership outcome but with segments or periods that don't.  See Chapter 3 of Human Transit for more on this language problem ...
  
  
• Ridership can be good by the standards of a local subarea and bad by citywide standards, if the entire subarea is less transit-friendly in the terms outlined above.  Another way this issue is sometimes handled is to divide the region firmly into subarea, put budget walls between the services for each subarea, and then assess ridership by each subarea's standards.  This used to be how it worked in Seattle's King County Metro, and the effect was that a line in the suburbs with a certain performance could be called successful while a line with the same performance closer-in to the city would be judged a failure.  Not everyone could follow the "logic" of that and the agency has now let go of that rigidity. 
• I am making a point about geometry, not about culture or values or psychology or ethics.  I'm certainly not telling people where they should live, only about the consequences of those choices.  But concealing or ignoring the facts of geometry leads to policy based on denial, and that leads us in circles rather than forward.

This isn't an easy topic.  Many of the terms are emotive.  But there's an important fact here.  Again, see Chapters 9 and 10 of Human Transit for much more thorough discussion.

basics: the "new route" problem

When people think of a new transit need, they often jump prematurely to the idea that they need a new route.  This new article of mine -- linked to in Chapter 7 of my book -- explains why this can be a bad idea.  

basics: operating cost

At several points in the book, I include links to articles on the blog that expand more fully on a topic.  The article on operating cost is now online, here.   Remember: If you don't understand operating cost, you don't understand transit!

network design for high ridership, a dense city example

How do transit network designers go about their task? Surprisingly little has been written about this.  You can pick up books that appear to cover the "network planning" process and find examples of good and bad networks but rarely a description of how to do the design thinking itself.  EMBARQ's recent manual for network planners in India, for example, provides great detail about how to analyze demand and evaluate results, but show no awareness of the really challenging task of network design, which sits in between those tasks.

I'm thinking my next book will probably be a little e-book talking through the process in more minute detail. For now, let me talk through a quick example, just to capture the outlines of how a network designer might approach a problem. 

My example is the core of Halifax, Nova Scotia, a dense peninsula built out mostly before 1945 and therefore highly conducive to transit.  It isn't the only transit-conducive place in the metro region, but it's the largest, and its peninsular location helps us isolate it as a design problem, so I can discuss it in a brief example.

The peninsula's developed area is about 6 km (4 mi) long and 3 km (2 mi) wide at its widest point. 

First, we take stock of the strategic focal points, especially major destinations and chokepoints.

Major destinations are part of why you'd focus on the Halifax peninsula as the core of any effective regional transit strategy.  The strongest transit destinations are:

• intense, in terms of number of people coming and going.
• all-day, so capable of using relatively efficient two-way, all-day service instead of just the more expensive one-way peak commuter express service.

Those are usually places with lots of people coming and going, not just employees.  What kinds of places are those?  Large scale:

• Government institutions
• Universities
• Medical facilities
• Retail

They're all on the peninsula.  Provincial government, two universities, the major medical centers serving all of Atlantic Canada, and retail centers both downtown and on the west edge.

The other reason you focus on the peninsular core is that it's already dense and walkable, because most of it was built before 1945.  So much of Halifax is here, and is so amenable to transit, that a strategy trying to optimize transit mobility must take full advantage of these opportunities.  If you care about mobility to an outer suburban community, you must care about circulation within the peninsula, because many of your citizens are going there and they may need to move about within the peninsula while there.  What's more, peninsula residents who choose not to own cars (which for many could be a viable and liberating choice) will sometimes need to travel outward, and a transit-intensive peninsula will generate outward transit demand further supporting the radial services that outer suburbs need.

So what do we have on the peninsula? 

All access is via just six chokepoints (red circles), plus ferry lines east across the harbor.  These will be ideal locations for transit connection points (yellow rectangle with T) because many lines, offering travel to many origins and destinations, must converge there anyway.  So the transit agency has done a good job here.  Except for the northernmost bridge and the northernmost exit to the west, all the chokepoints have transit connection facilities nearby.

The whole peninsula is dense and there are tall buildings here and there throughout, but the big concentrations are in the areas I've shaded.  Downtown is the blue area on the waterfront, while the parallel blue box further west is the medical zone.  The magenta zones are the major universities.  The southern one, St. Marys, is a Catholic school with about 7000 students.  The northern one, Dalhousie, is a major public university (16,000 students) that spills eastward along the green axis toward downtown, mixing with major medical facilities, office buildings, and dense, often student-oriented housing.  This green area is the core of a potential sustainable-transport paradise, because all the elements that make Halifax in general such a strong market (listed above) are mixed especially closely here, while the dominance of the universities, medical facilities, and shopping guarantees all-day two-way flow, the best situation for highly efficient transit.

I hope you've followed all this even if you don't know Halifax, because the same kind of assessment needs to be made of any city.  This doesn't mean that you have to have all of Halifax's features, but you need to be looking for those features you do have, with particular focus on density patterns and chokepoints, becuase these are the best starting point in defining a strategic outline of the network, within which you can then develop local ideas in more detail.

Now we look at the network of major transit-operable streets:

The grid pattern means that some kind of high-frequency grid is almost certainly in order, so that people can get anywhere to anywhere on the peninsula with at most one connection.  Obviously, since all trips within the peninsula are short, connections are a bigger disincentive, so we'll try also to link major destinations with direct services.  However, if we made direct service everywhere to everywhere our primary goal, we'd end up with lots of overlapping lines and would struggle to afford enough frequency on them all.

Notice, however, the way the southwest edge of the peninsula runs on a diagonal compared to the street grid.  That's an invitation to draw L-shaped lines, an east-west grid element connected to a north-south grid element.  Like this:

 

Grid lines that "bounce" off of diagonal edges, like this one, tie together more destinations without a connection while remaining complete grid lines (i.e. running all the way across the grid)  so as to maximize the number of other connections that can be made. 

But Oxford Street sort of peters out at the north end, so where do we go?  We "tie off" or "anchor" the line by going to a nearby major destination where many other connections will be available.  Fortunately, one is nearby.  The "T" just southwest of where Oxford ends is also a major shopping center. 

Meanwhile, at the other end, we notice that many connections throughout the east-of-harbour Dartmouth area are available at a "T" just over the bridge.  A single frequent line to that point, covering many of the major destinations of the peninsula, can plug into all of those connections, so that the whole catchment area has easy one-transfer access to most of these destinations.  So we have this:

And in fact, Halifax's Metro Transit has already had the same thought.  Their most frequent transit line, Line 1, looks like this:

It's on Spring Garden instead of South, but that's only about 400m difference.  The university and most medical facilities in the area are adequately served from either.  But Spring Garden is more of a continuous "mainstreet" through this area, although major institutions front on both. 

So what have we missed?  Well, the next big corridor is Robie, which (along with nearby parallel streets like South Park) is really the medical axis.  It would have been tempting to turn our first line north along Robie instead of Oxford, because while Oxford is perfectly fine lowrise density Robie has the highrise intensity of institutions.  At the south end, Robie bounces off the diagonal grid-edge as Inglis, which gets us to the smaller university, St. Mary's.  If we bounce again at the east end of Inglis we head north along Barrington into downtown.  Once we're past Spring Garden we're duplicating the first line, but we're virtually downtown by this point, so duplicative service on one street, adding up to very high frequencies, makes sense and can be useful for internal downtown demand.  (Remember, the shorter the trip, the higher the frequency you need to compete for it.)  So that suggests something like this:

This is not what existing service does.  Robie now has an overlay of several lines doing slightly different things, several of which end in a very odd one-way loop at the south end, like this one, rather than going downtown. 

Perhaps it's just trying to turn around.  One-way loops are a good idea at outer low-density ends of lines, but they introduce needless confusion and circuitousness in dense areas, where it should be possible to serve all markets two way.

Meanwhile, back in our ideal grid, our Robie-Inglis route encountes a perplexity at its northwest corner.  At Chebucto Road, Robie seems to branch into the continuation of Robie and the larger Windsor Road, which takes on a more car-oriented character as it heads for the chokepoint at the far northwest corner of the peninsula, near Clayton Park.

This situation is tricky, and I wouldn't make a recommendation at the high-level scale of this exercise; much deeper knowledge of existing travel patterns and land uses would be needed.  For now, let's assume that the smart thing to do is follow Windsor, because it completes the grid, serves a chokepoint, and thus is useful both as an internal corridor on the peninsula and a gateway path to the larger region.

Now, you can see how some east-west frequent lines would largely complete the grid.  To work best they'd have to aim for the chokepoints too, so that while offering intense local circulation within the city they'd also provide access to much of the peninsula from the surrounding suburbs.  Metro Transit has already drawn most of these lines, and they seem to work pretty well except that many aren't frequent enough to really function in a high-frequency grid, or even to compete for the very short east-west trips that they would serve within the dense peninsula.

The dark blue line, existing lines 2 and 4 combined, is already every 15 minutes, but the other two key grid lines here, orange Line 6 on Quinpool and and green Line 9 via the far north, are still every 30, not enough to really function as elements of a grid. Notice too how 6 and 9 bounce off edges of the grid, completing linear corridors before turning to find a major destination and connection point at which to terminate.  (Lines always want to terminate at big destinations, called anchors in the business, because otherwise they tend to empty out near the end, leaving permanent wasted capacity.)

Yes, there are still gaps, but we've hit all the major denstinations and most of the peninsula's density.  As we turn to those gaps, network design starts to get fun and subtle, as we have to dig into more detailed data to find the clues for how we should patch routes together.  And here I'll stop, as I've reached the end of what I can do in such a "high level" view.

But here's my point:  I'm almost certain that the lines drawn here, or something very, very like them, would be part of a successful network for Halifax, because they are big-picture responses to issues that are obvious at this big-picture scale.  Network designers sometimes fail to take this high-level sketch planning step, and instead wade too quickly into the million possibly interesting details.  If you focus on too much detailed data too soon, though, you end up wandering around inside the data, unable to get any big picture "structure" that you can hang onto while you consider the subtler questions. 

Also, while it's very important for many people to be able to follow this kind of thinking, designing optimized networks, especially in difficult geography, is a bit of an art.  Like any inductive thought, it involves deeply understanding the data but then being open to ideas that come in unpredictable bursts of inspiration, much the way scientific theories come about.  The "having ideas" part isn't in many manuals because it's not really teachable; it's to some extent an innate talent.

Consultants like me can help with the subtler thinking of network design, and of course many professional transit planners are adept at it.  But you don't have to have those skills, or want them, you can be a more effective advocate if you understand the kind of thinking I've been demonstrating here.  If you want to influence transit in your city, you have to understand the basics of the network design problem, as it arises from facts of geometry, facts of transit, and the unique geography of each city.  That way, even if you're not ready to do network design yourself, you can assess whether the designers of your own network have done a competent job, and make suggestions that they could actually use.

basics: dead running

The British/Australian term "dead running" means "running out of service, unavailable for passengers."  I like the term because it could be the title of a zombie movie.  I look forward to seeing if it attracts hits.

The North American term is "deadheading," which may remind you of rock fans of a certain era.  In both cases, the word "dead" is apt.  It's a waste, a loss.  It's unavoidable up to the point, but transit agencies are always trying to turn dead time into live, useful time.

The Brisbane Times has a piece by Marissa Calligeros announcing that '28 per cent of bus services are "dead running", where passengers are unable to board.'  That's a confusing way of putting it.  The reporter must mean that Brisbane buses are dead running 28% of the time. 

That sounds like a lot, but "dead running" is the result of two different issues that need to be kept separate.

• All transit vehicles must travel between their operating bases, where they are stored and maintained, and the beginning and endpoints of service.  Rail services usually have bases directly on the rail line, but may still have to dead-run through the rail system to reach their trip's starting point.  Bus bases can be anywhere.  The location of bases (called depots in Britain/Australia, often divisions in the US) is a major issue.  It's often worth spending capital money to save operating money, and careful investments in bases, reducing dead running, can do that.

• Transit agencies that run extensive one-way express incur massive amounts of dead running.  Brisbane, for example, is a very, very centralized city, with a downtown far out of scale to anything else.  That means a huge demand for one-way trips into downtown in the morning and out in the evening.  All those services that are needed in only one direction usually have to get back in the other direction so that the driver's shift can end where it began.  (The other alternative is to pay the driver to hang around downtown all day, which is even more expensive.)

There's no way to understand Brisbane's "dead running" issue, or that of any other city, without separating these two causes.  The Brisbane Times article talks only about the first, but the Brisbane bus system has a massive one-way peak due to its single dominant downtown and relatively lack of direct rail paths for much of the city.  Can dead running be better addressed by a rigorous review of whether these one-way peak services can be combined, replaced by links to rail, or otherwise made more efficient?  GIven the higher cost of dead running for one-way peaked service, could some of it be converted into two-way, all-day service at less expense than it would first appear?

That's an important question for any city.  Details of driver costs vary, but in general, dead running is one of the main reasons that one-way express service (bus or rail) can be more expensive than it looks.

stuff you need to know

If you want to do things with transit, you need to understand some basics about its geometry and costs, facts that may raise interesting questions about your own goals. 

To make this blog more useful as a reference, I've assembled the most important posts on these topics in the far right column under "Stuff You Need to Know."

If you haven't read them all, have at it!

Let me know if you think any should be deleted or others that should be included.  The intent is not a list of the "best" posts, but of posts that best lay out fundamentals of network design.

Also, note the new travel schedule under my photo.  I'll be in Madison for the Congress for the New Urbanism conference, and in Wellington (NZ) and Halifax (Canada) for project work. 

basics: walking distance to transit

The question of walking distance in transit is much bigger than it seems.  A huge range of consequential decisions -- including stop spacing, network structure, travel time, reliability standards, frequency and even mode choice -- depend on assumptions about how far customers will be willing to walk.  The same issue also governs the amount of money an agency will have to spend on predictably low-ridership services that exist purely for social-service or "equity" reasons.

Yesterday I received an email asking about how walking distance standards vary around the world.  I don't know the whole world, but in the countries I've worked in (US, Canada, Australia, New Zealand) the view is pretty consistent:

• If you have to choose a single walking distance standard for all situations, the most commonly cited standard is 400m or 1/4 mi.  Europe tends to be comfortable with slightly longer distances.
  
  
• However, people walk further to faster services.  (Rail advocates are more likely to phrase this as "people walk further to rail".)  This doesn't have to be a sociological or humanistic debate, though urbanists often frame it that way.  If you are a rational and informed actor seeking to minimize travel time, it often makes sense to walk more than 400m to a rapid transit station rather than wait for a bus to cover such a short distance. 
  
  
• Although the common standard is 400m or 1/4 mi, we all know that this cannot possibly be a hard boundary.  It makes no sense to assume that if you live 395m from a bus stop you'll be totally happy to walk that distance while your neighbor, who lives 405m from the same stop, will be totally unwilling to.  Obviously, the relationship between distance and willingness to walk is a continuous curve without sharp breaks.  This has to be said because our language often forces us to create the illusion of sharp breaks, e.g. when we say something like "people are generally willing to walk up to 400m to transit."

Finally, it's remarkably hard to sift data into a form that produces unequivocal guidance on the question.  For example, the leading US guide on transit planning, the Transit Capacity and Quality of Service Manual, offers only this:

(Source: TCQSM Chapter 3, Appendix A, p. 3-93.  Discussion and version in US units is on p. 3-9.)

This survey-based graph shows the breakdown of local bus passengers by the distance they walked to get to the service.  As you'd expect, few people walk more than 200m in downtown Washington, DC because in such a densely served area, few people would need to.  In low-density Calgary, at the opposite extreme, many people have to walk fairly long distances.

But extrapolating opinions from behavior is a tricky business.  It's hard to reason from how far people walk to how far they're willing to walk.  To do that, you'd have to determine whether each rider would be willing to walk further than he actually has to walk.   You'd also have to speculate about each rider's available options.  If 1/10 of Calgary's bus riders walk 600m or more, does that mean they're willing to?  Or does it mean that these people are so lacking in good alternatives that they feel forced to walk that far?  (The difference between "high income" and "low income" Washington DC suggests that range of options does have something to do with it.)  Sociologists and demographers can have a field day parsing this question, but they're unlikely to come up with an answer of such statistical certainty that it definitively sweeps the question aside.

So we approximate.  We generally assume that 400m is a rough upper bound for slow local-stop service, and that for rapid-transit (usually rail) we can expect people to walk up to 1000m or so.

But when we try to apply these rules of thumb, we hit another hard issue (or at least we do if we're willing to acknowledge it).  Are we talking about true walking distance, or just air distance?  Over and over, in transit studies, you'll see circles around bus stops being used to indicate the potential market area, as though everyone within 400m air distance is within 400m walk distance.

Remember this graphic?

In both images, the red dot is a transit stop and the red circle is an air-distance radius.  If you draw 400m circles around stops based on the assumption of a 400m walking distance, you're implying that the whole circle is within walking distance.  In fact, even with the near-perfect pedestrian grid in the right-hand image, the area within 400m walk (outline in blue) is only 64% of the red air-distance circle.  With an obstructed suburban network like the left-hand image, it can be less than 30%.

Obviously, the market area around each stop should really be defined by the walkable area, which requires a knowledge of the local pedestrian network.  That requires a complete GIS database of every walkable link in the community -- an extremely detailed task that few jurisdictions have been willing to attempt until recently.  Even in Canberra, Australia, which is known in the business for the extreme richness of off-street pedestrian connections, no reliable database of them was available for modelling purposes as recently as last year.

Still, if you don't have such a database, it should be easy to adjust the walking distance standard to reflect the problem.  If you know you have a good street grid, then you can just adjust the radius to reflect the area within 400m walk.  In the right-hand image above, do the math and you'll figure out that if the radius of the red circle is 400m, then a circle whose area is the same as that of the blue diamond -- the actual area in walking distance -- would have a radius of 319m.  So if you want to roughly model the actual radius that arises from a 400m walking distance, and you have a well-connected street grid, draw a circle 319m in radius.  That doesn't give you the correct boundaries of the area -- it's a circle rather than a square -- but it's a far better approximation than just drawing a 400m circle.   I have never actually seen this done, and I'm not sure why.

One reason might be that secretly, we transit planners all want people to walk further.  After all, most transit planners don't want to just passively respond to current behavior.  If they did, they'd all be highway engineers.  Most transit planners believe in the importance of shifting behavior in more sustainable directions, and see both transit ridership and walking as deserving encouragement through intervention.  They are also aware of the public health benefits of walking.

But we have a more vivid motive to encourage walking.  The nature of the transit product is such that if we could stop less often, assuming longer walk distances, we could achieve both better running times and reduced operating cost; the latter could be reinvested as higher frequency.  So the two most fundamental determinants of transit travel time -- running time and frequency -- both depend on our assumption about walking distance.

With such basic things at stake, it's understandable that planners are always looking for ways to push walking distance wider.  That may be the real reason that generations of planners have chosen to approximate a 400m walk with a 400m circle, even though every pedestrian knows how absurd that is.

I prefer to just have the argument in simpler terms.  In Canberra, we pushed the walking distance standard from 400m to 500m, not because people were calling us demanding to walk further, but rather because we looked at how much more frequency and speed we would achieve, and the ridership that could attract, and decided that 100m of radius was a small price to pay for such benefits.  It comes back to that graph near the top of this post, showing how far people walk to transit in different cities.  There's no definitive authority for a 400m standard as opposed to 300m or 500m or even 600m.  Yes, if you pick a bigger radius you'll lose riders from the outer edges of the radius, but on the other hand, you may buy so much travel time and frequency that your ridership goes up.  As with everying else in transit, it depends on what you're trying to do.

how universal is transit's geometry?

Suppose that somewhere else in our universe, there’s another planet with intelligent life.  We don’t know what they look like, or what gases they breathe, or what they eat, or whether they’re inches or miles tall.  We don’t know whether they move by hopping, drifting, or slithering.  We don’t even know if their lived environment is largely two-dimensional, like the surface of the earth, or freely three dimensional, perhaps a cloud-city full of cloud-beings who drift up and down as easily as they drift left or right.  We don’t know what they call themselves, so let’s call them borts.

Let’s make just two assumptions about them.

Let’s assume (1) that the borts tend to form large assemblages, which enable trade, creativity, ritual, and whatever other activities give value to their lives.  Let’s call these places cities.  Since cities are places where borts are relatively close together, it follows that they have relatively little space per bort.  Cities, by definition, are places where space is experienced as scarce.

Assume (2) that these cities are large enough that a bort can’t easily hop, drift, or slither around the city fast enough to reach all the needs and pleasures of daily life.  Given this reality, they must have invented vehicles of some kind that carry them faster and further; if they hadn’t, their cities could not have grown so large.

Do we have to make an assumption about their communications?  If the borts had either perfect telepathy or perfect virtual reality, then they would never need to move for any of the purposes of interaction.  But in this case, why would they have cities?  Let's assume [and I think this arises from (2)] that their communications are not so perfect, and that they do need to move around to do whatever borts do that constitutes their lives and economies.

Perhaps the borts have tried using a personal locomotion vehicle in their cities.  It gives a bort freedom to move at high speeds, but it’s much bigger than the bort’s body, so it takes a lot of space inside the city.  Given the city’s limits on space, these vehicles – which require a lot of space per bort – would have collided with those limits, causing something like congestion.

So (regardless of whether they’ve tried the personal vehicle) they must have invented a vehicle that can carry many borts at once for travel within their cities -- not to mention between them.  Call it a bortmover. 

But then comes the “network design” challenge: in a city where borts need to move around freely, from anywhere to anywhere else, what exactly should the bortmovers do?  In what patterns should they move? 

Well, they need to move in a known pattern, so that borts can predict them.  That means some kind of routing and schedule, and a schedule implies both a frequency and a span of service.  Borts need to get places soon, so that means that these vehicles must stop close to desired points of origin and destination, they must not require too much waiting, and they must be reasonably direct and also move at an adequate average speed.

So with only two basic assumptions about bort civilization, we can infer that they have concepts of stop spacing, frequency, span, directness, and speed.

Obviously, we can’t know what bortmovers cost to operate, but like all planets, the bort world is finite, so bort society must have a concept of scarcity.  Even if bortmovers cost nothing to run, there is still that defining scarcity of all cities: the relative shortage of space per bort.  Most likely there are other costs as well, reflecting the energy required to manufacture and operate the bortmover.  So there’s some limit to the number of bortmovers, which motivates borts to use them efficiently.

Presumably, some parts of the city have a greater density of bort activity than others, so even if a bort’s likelihood of boarding a bortmover is constant regardless of location, there’s more bortmover ridership in these denser areas.  So the borts experience a relationship between transit and density: if there are more borts within slithering/hopping/drifting distance of a bortmover stop, that’s a larger potential market, which means a busier bortmover.

So bortmover companies have to decide whether to route their bortmovers to maximize their ridership – which would mean no service to some of the more sparsely populated parts of the bort city – or to spread bortmover service across the whole city even though they’ll be crowded in some areas and lightly used in others.  They’ve faced the choice between planning for ridership and planning for coverage.

They’ve also discovered that if they try to run a direct bortmover from every part of the city to every other, they get poor frequency and high complexity.  Maybe the borts are all geniuses who love keeping complex networks in their minds, but poor frequency will still be an obstacle to getting where they’re going.  So borts must have discovered the connection.  That means they’ve faced a choice between connective networks (which are simple and frequent) and direct-service networks (which are complex and infrequent).  And if they've tried to optimize a truly connective network, they've probably discovered the grid, regardless of whether their cities are grid-shaped.

Again, because space is scarce, they probably experience something like congestion.  This problem will be worst where there are the most borts coming and going, but that’s exactly where the bortmovers are needed most, in high volume and with high quality.  So perhaps some of the busiest bortmovers have been given exclusive linear spaces that they can operate in, so that they can run fast and reliably.  Call them exclusive rights-of-way. 

We don’t know much about the borts or their world, but with just two basic assumptions, we can infer that if they have invented transit, they have discovered the concepts of stop spacing, frequency, span, speed, directness, travel time, connections, and even right of way. 

So bort transit must face the realities that we must face:  Closer stop spacing, for example, means slower speed.  Branching divides frequency. Straight lines are more likely to offer good travel time than winding ones, so the “be on the way” principle applies to bort communities as it does to ours.

In fact, almost all of the basic concepts of transit are realities of bort urban life, because they’re facts of  geometry.  They are true of anything we would recognize as transit, everywhere in the universe.  They are not negotiable.  So they should be the basis of our thinking about transit in cities, not a detail that can be left to the engineers.  Cities must respond to the intrinsic geometry of transit, just as suburbia as responded to the intrinsic geometry of roads.  And yes, you can use a cute, fun, or exciting transit vehicle, but that doesn't change any of these geometric facts.  If you don't understand transit's geometry, technology won't save you.

Updated 18 Nov 2011.