For Release July 9, 2009
Citiwire.net
Touted coast to coast as the key to “transit-oriented development,” light rail systems are close to “accepted wisdom” as keys to 21st century metropolitan growth.
But it’s worth remembering that the Interstate Highway System, when first promulgated in the 1950s, was widely and uncritically hailed too. Sadly, we failed to take into count the price we’d ultimately pay as the massive roadways subsidized scattered development and simultaneously helped empty out many downtowns.
In short, there were winners and losers. Cities since have pumped many billions into reviving downtowns, but even to this day the phrase “inner city” still conjures unfavorable images of the bereft neighborhoods left behind to decay when the middle classes escaped to the fringes.
The spread phenomenon persists. The Cleveland metro region, for example, has about as many people as it did in the 1990s, but it’s expanded its urbanized footprint some 30 percent–mostly to overwhelmingly auto-dependent fringe areas.
It’s becoming clear the freeway-based system may not be sustainable. Metro-area auto congestion continues to worsen, urban sprawl continues unabated (while threatening some of our most productive farmland), concerns about greenhouse gas emissions are growing, and there’s the specter of steeply rising prices for “post-peak” oil.
But leaping to light rail could well trigger a new set of unintended consequences. In some locations, it may work well–it is a proven and popular transportation tool. But I’d argue it’s as mismatched to today’s American urban form as was the freeway-centric vision of the 1950s to the urban form of its day. And that we ignore this mismatch at our peril.
For good or for bad, the modern American city can be characterized mostly as a widely-scattered set of dispersed origins–essentially, where people live–and multiple destinations–those clusters of employment and retail where people travel to in a given day.
Light rail (and most bus rapid-transit systems) create, though, a “traditional” transit network, often likened to pearls-on-a-string. It’s a formula that works fine for relatively compact and dense cities. But when population is widely and often thinly dispersed, travel times often become excessive and even prohibitive. What’s more, the land area directly served by stations–that is, how far people are willing to walk–may be smaller than many think, particularly under the kinds of harsh environmental conditions most American cities will experience at least part of the year.
Depending on how rapidly energy prices price, real incomes may not be sufficient for many households to continue operating multiple motor vehicles. Yet if a high portion of suburban homes are nowhere near transit stations, and local transit systems (themselves cash-strapped) are unable to create enough bus feeder lines, neither the geography nor the economics of a rail-centered system may work.
Even for those with access to feeders, trips to destinations may be excessively lengthy, require too many transfers, and still not get people close enough at the end of the trip.
The end result? Pockets of affluence developed around light rail stations–but vast swaths of existing housing stock depressed in value because of transportation costs. These suburban zones can rapidly become the new slums, as declining values lead to deferred maintenance and families doubling or tripling up.
Does this mean that investments in transit are misguided? Hardly. Nor does it mean that there is anything wrong with rail-centered transit modes. What it does suggest, though, is that American metro regions, unlike their more densely packed European or Asian counterparts, have large swaths of suburban territory ill-fitted for rail.
Fortunately, there are other alternatives. Among the more promising are “Quickway” systems, such as the highly successful Busway network in Brisbane, Australia. Brisbane does have grade-separated runway for speedy, high quality bus service through its core. But it also permits a network of branching and express bus services, with routes fanning out to many neighborhoods, to provide service directly to key job sites inside the central system. Buses can move to and from the exclusive center city lanes from regular roads. The result? Many more people have access to effective rapid transit within a short walk of their home, trip times are often even faster than driving, many people get a “one-seat” ride, and operating subsidies are reduced or even eliminated (due to the gains in productivity that grade separation makes possible).
The best evidence for the strength of Brisbane’s model is in ridership, which has climbed 50 percent in just five years. It is also noteworthy that Brisbane’s strategy has produced gains in land value over a much larger land area than that typically associated with light rail (or even more traditional bus rapid transit lines).
The lessons are clear: regions need to think first not about rails versus buses but the kind of network will serve them best. The goal of a viable citistate-wide system should comes first and the preferred mode, fittingly, second.
Alan Hoffman’s e-mail address is alan@missiongrouponline.com.
Citiwire.net columns are not copyrighted and may be reproduced in print or electronically; please show authorship, credit Citiwire.net and send an electronic copy of usage to webmaster@citiwire.net.
10 Comments
It would be a shame to misread the real lesson of the highway boom: not the building of highways but the failure to build a complementary local road system. Suburban sprawl and inner city disinvestment are not just products of highway construction, but also of federal housing and mortgage insurance policy, racism, school funding formulas, federal sewage treatment construction grants, opportunistic local use of state and federal highway funds without matching local investment, disinvestment in transit, and a bunch of other things that, combined, created a distorting advantage for greenfield development at the edge. Rather than equating LRT to highways (another distorting proposal in an of itself) and intimating that the “solution” is to use buses as if they were cars, the real solution is to plan and develop LRT as part of an integrated, multimodal transportation system. It’s not about buses versus LRT, but buses and LRT, and cars and bikes and peds. The issue isn’t whether LRT replaces highways, but whether LRT both adds choices and increases accessibility. Sorry, but there are no silver bullets, the lesson that the highway era should have taught us and one that most proponents of LRT are not keen to replicate.
Mr. Seltzer raises some excellent points. The push to greenfield development in this country was not just the product of freeway development, but of the confluence of several factors, of which Seltzer cites many.
However, one can make the case that some of these same factors are working to push the LRT/BRT/TOD model today, including “smart growth” incentives and large subsidies for development taking place around transit stations. Don’t get me wrong: I’m not against the linking of transit, land use, related infrastructure, and public investment. But what I am warning against is the *unthinking* linkage of these elements, one that treats the transit station as the be-all and end-all of development, regardless of whether or not the actual transit linkages from that point are competitive or not.
Los Angeles is a great case study of this point. In 2007, the LA Times reported on its own study of TOD in that region. They found that, while residents of TODs *were* more likely to ride transit than were others living farther afield, the *absolute* numbers were still small enough as to be relatively insignificant. The reasons why residents didn’t ride were directly related to the Three Key Strategic Variables I had identified and published about in prior studies: the transit didn’t go to many of the places to which they needed to travel; if it did go there, the stations were still too far away from their actual destinations; travel times by transit were excessive and often required too many transfers; and they didn’t “feel” comfortable using the system for a variety of reasons.
So the issue comes down to *how effective is your transit network.* And in this I can only call into relief Mr. Seltzer’s admission about “proponents of LRT”: it is precisely because they are proponents of a single technology, and NOT because they are proponents of “effective transit” (they are *not* necessarily the same, the central thesis of my article), that I need to raise my warnings.
Finally, I apologize that in the space of a single column I was not able to adequately describe the Quickway model that Brisbane has been following. It is not about using “buses as if they were cars,” it is about creating a transit infrastructure that can be used for more than just “pearls on a string” transit. How much more useful would LRT systems be if the trains could “jump the track” and travel the last mile or two to get closer to where people are or where they’re going? How about if we so lowered the cost of operating a “train” that we could operate more express services, saving people time? How about if we direct TOD and infill development, not merely to 30, 40, even 60 relatively small station areas in a large metropolis and instead create ten times the effectively-served land area (think about this the next time you try to build relatively affordable unsubsidized housing)? But for those who’d like to see in greater detail what it is Brisbane is doing, you can download a copy of the study I wrote for the Federal Transit Administration in 2008, entitled “Advanced Network Planning for Bus Rapid Transit,” available at http://www.fta.dot.gov/documents/BRT_Network_Planning_Study_-_Final_Report.pdf.
Looking at the market driven evolution of moving information on a network brings some insight to moving people too. When networks were being used 10 or more years ago, the cost of access and servers were higher. Most data moved around fewer “hubs”. Token ring and ATM networks were well suited for these tasks. They provided reliable and high speed transport. Token rings are networks designed like the pearls on a string. The information stops at each stop and then moves to the next. But, there were only a few stops to stop at. Maybe only a dozen computers on an office token ring.
ATM networks are based on the bucket idea. Buckets are a certain size, like the number of seats on a train. When it is time for the train to leave the station, it leaves, regardless of whether the train seats are full or not. The trains are on time, but seldom full.
Today’s Internet has many more hubs and servers. With many points to inter-connect, the winning technology has been TCP/IP. TCP/IP can move both small and large amounts of information efficiently . It does this through a number of technologies, but the most successful ones have been queuing, dynamic routing, and prioritization. The ability to vary the size, route and importance in real time.
ATM is still the main protocol to move large amounts of data from on hub to another, like a high speed rail line from New York to Los Angles. But, the ability to dynamically allocate capacity for feeder and branching services is exclusive to buses and taxis. Grade separating in high capacity and congested areas gives similar throughput to rail systems without having to transfer. A win/win solution.
I believe that the greatest obstacle to effectively building bus systems that use quickways, is that they aren’t European/sexy. Not that there is any problem with the math. We need to re-brand this solution. We need to “sell” the public that “green” is the new “sexy”.
I think Harry’s comments are on the mark. Light rail is perceived as sexier than transit, although generally much more expensive and difficult to implement. For trips under a mile, walking is still the simplet and best solution for reasonably fit people, assuming sidewalks are present. I’d like to see incentives employer/insurance incentives for walking: saving fuel, improving air quality and individual health.
It is clear that the increase in bus ridership in Brisbane has more to do with faster and more frequent service, as opposed to Hoffman’s theories about “Quickways.” Transit systems that have high ridership per capita depend far more on frequent service than “one seat rides” as shown by Australian author Dr. Paul Mees in his book, A Very Public Solution” and much of the work by Dr. Vukan Vuchic. Also see http://mailer.fsu.edu/~gthompsn/garnet-gthompsn/my_web/Fundamentals%20of%20Successful%20Transit.htm
Yes, Michael, but the Quickways enable the faster transit and allow them to afford the higher frequencies.
No known relation to Alan Hoffman.
Mr. Seltzer represents the argument for LRT very well: it is not a silver bullet but part of a broader multi-modal system that increases transportation choice. I also believe he did a good job at showing the “distorting proposal” of LRT to Highway building – including citing the three most important factors of why they are different and why the Highway plan turned out how it did: the federal housing, mortgage insurance policy and racism – it was not prudent for Mr. Hoffman to omit these factors in his comparison, regardless of space.
I think Mr. Hoffman’s warnings to the Light-Rail community are not needed. If ‘how’ something is done determines ‘what’ the outcome is, factions or interest groups must become the advocate of that one most single salient issue. In many communities you’ll find the same advocates for Light-Rail are the same advocated for a smarter, more comprehensive and efficient transportation system. The warning that “it is precisely because they are proponents of a single technology” might be more of Mr. Hoffman’s subjective reality than what is actually said, written and done in the transportation planning community.
The proverbial 1-seat ride is counter-productive. No transit system can operate without transfers. When whole transit systems are arranged to enable 1-seat rides, the whole system suffers.
Light rail systems require transfers. This ’splitting’ or ‘combining’ of transit modes can improve bus service.
Think of it like a ‘matching supply to demand’ equation. Light rail lines cover long-distances and serve a high demand. Bus lines that run short distances to light rail stations have a measurable demand that can be more practical to meet with suitably frequent bus service.
We should think of mass transit as a “fundamental” transportation mode, no less fundamental than walking and bicycling. Once these modes of travel are considered fundamental, the land-use and development patterns evolve to serve their optimal function.
The question isn’t bus vs rail. The question is how may all urban/suburban modes of travel function to produce within metropolitian regions dozens of functioning neighborhood economies where long-distance travel is no longer required to meet most daily needs.
Through both his post and comment, Mr. Hoffman raises many points I’ve been making about the ineffectiveness of at-grade light rail in Los Angeles. I’ve been making pretty much the exact same statements you made in the two paragraphs in your column, which accurately diagnosed L.A.’s problem. If you only knew how I’ve been chastised and demonized by “light rail advocates” for it. Simply, what has become clear is that there are, as you said in your column, “light rail advocates” and they are NOT necessarily “transit advocates.”
Second, in a city as dispersed as L.A. travel times are very important, and most of the right-of-ways looked at for rapid transit implementation (because the ROWs already exist, most are owned by the MTA, and are therefore cheaper to build on) are too far from actual destinations to show any travel time benefit. To even attempt to show significant benefit, 55-70 mph travel times between stations on the ROW are needed. Such speeds can only be obtained through grade separation for trains. (Unless the train is flying between stations, how can the light rail trip compensate for the additional travel time required for a transfer from the train station to the actual “doorstep” of the destination?) Furthermore, because the roadway network is already beyond capacity, branch lines aren’t feasible, and neither are high frequencies, which even at 3 minutes terribly screws up traffic. All of these elements point to the need for a system that doesn’t interfere with cross traffic (too slow) or is grade separated. But to the LRT advocates grade separation (which obviously has countless other benefits aside from increased travel times) is “too expensive” and to implement such would reduce the size of the “network” we need to build. No mind the end network they seek will be too slow to attract most choice riders, which in L.A. really is anybody with enough money to NOT ride transit. (Our real target should be the $10/hr working poor who drive broken down cars because our bus network is simply unreliable).
I have in many ways said to the at-grade advocates that most of their arguments is advocacy not for at-grade LRT, but rather BRT. Simply, if there’s no concern about implementing a line/system with the capacity and design to solve transportation challenge 50 years in the future, no concern about vehicular impacts, community impacts, operational limitations, there is a desire for a “network,” and grade separation is “too expensive,” in L.A. that translates to advocacy for a BRT network.
After 2500 hours of research it became clear to me that the need in LA is for a grade separated network with light rail vehicles, so that travel times can be improved by providing station portals close to the existing destinations. Such also allows a system to be created where interlining, high frequencies and speed provide crucial travel time savings, where even if there is an eventual bus connection it is still competitive with an automobile trip. And most importantly it increases the capacity of the transportation grid, which is the only way “smart growth principles” can be implemented here without further worsening vehicular traffic beyond what is already expected. You’d be amazed how difficult it is to get folk to understand that if you want to widen a sidewalk, drop a lane, and/or add bike lanes, about the worst thing that one can do to the transportation grid is add a 25-foot right-of-way, coupled with multiple street-closures and at-grade crossings every 150 seconds during rush hour.
Alas, “capacity” and “travel times” are not typically words you see mentioned by at-grade advocates.
Bravo Mr. Hoffman!
The two paragraphs in your comment regarding L.A.’s problem are spot on. Through both your post and comment you have raised many points I’ve been stressing, when pointing out the ineffectiveness of at-grade light rail in Los Angeles. Yet, if you only knew how I’ve been chastised and demonized by the area’s “light rail advocates” for it. Simply, what has become clear is that, as you said, “light rail advocates” are NOT necessarily “transit advocates.”
In a city as dispersed as L.A. travel times are very important, and most of the right-of-ways looked at for rapid transit implementation (because the ROWs already exist, most are owned by the MTA, and are therefore cheaper to build on) are too far from actual destinations to show any travel time benefit. To even attempt to show a savings with light rail on most of these these ROWs, 55-70 mph travel speeds are need between stations, and such speeds can only be obtained through grade separated. Simply, unless the train is flying between stations, how can the light rail trip compensate for the additional travel time required for a bus transfer from the train station to the actual “doorstep” of the destination? At-grade makes branch lines and high frequencies infeasible because the roadway network is already beyond capacity. (At-grade crossings are already messing up traffic in the suburban outlays at just 3-5 min. rush hour frequencies, let alone 120-150 second frequencies.) All of these elements point to the need for a system without any traffic signal preemption (no traffic interference), which is too slow, or a system that is primarily grade separated. But to the LRT advocates grade separation (which obviously has countless other benefits aside from increased travel times) is “too expensive” and to implement such would reduce the size of the “network” we need to build. No mind the end network they seek will further worsen traffic and still be too slow to attract most choice riders, which in L.A. really is anybody with enough money to NOT ride transit.
I have in many ways argued that all of the arguments for at-grade rail are in actuality arguments for BRT/bus-only lanes, something they really hate, since they typically dispise BRT. But take a look at the at-grade arguments in LA and you’ll see that:
-there’s no concern about implementing a line/system with limited capacity and design to appropriately address transportation needs 30-50 years in the future;
-no concern about vehicular traffic impacts;
-no concern about community impacts;
-no concern about operational/frequency limitations;
-no concern about operating costs;
-there is a desire for an expansive “network” (regardless of the above); and
-there is a belief that upgrades, specifically grade separation is too costly and the capital dollars need to be stretched as far as possible.
Well in L.A. that translates to advocacy for a network of bus-only lanes, NOT at-grade LRT.
After 2500 hours of research it became clear to me that the need in LA is for a grade separated network, possibly with light rail vehicles (greater flexibility and uniformity with existing network), so that travel times can be improved by providing station portals close to the existing destinations. Such also allows a system to be created where interlining, high frequencies and speed provide crucial travel time savings, and where even if there is an eventual bus connection it is still likely to be competitive with an automobile trip. And most importantly, it increases the capacity of the transportation grid, which is the only way smart growth principles can be implemented without further worsening congestion beyond what is already expected. You’d be amazed how difficult it is to get folk to understand that if they want to widen a sidewalk, drop a lane, and/or add bike lanes, about the worst thing they can do to the transportation grid is add a 25-foot right-of-way, coupled with multiple street-closures and at-grade crossings every 150 seconds during rush hour.
Alas, “capacity,” “travel times,” “traffic impacts,” lasting impacts are not words you see mentioned by at-grade advocates. Nor do they like discussing the amount of government subsidies that go into building smart-growth properties around their station locations. Those subsidies could be used regardless of whether a light rail train is built at-grade near a station/on a corridor. The question is whether the transportation grid will see an increase in capacity and the transportation modes are actually attractive to allow those new residents/additional commuters/increased density to function without worsening the situation.