bus-rail debates in a beautiful abstract city, and in los angeles

On a recent post, commenter Pantheon laid out the core idea that explains why I cannot be a full-time rail booster, even though I love riding trains as much as anyone:

The problem can be posed in the abstract in the following way. Let's say we have a city with 20 neighbourhoods, A-T. Our city has a big deficit in transit infrastructure, and limited resources for redressing it. We have X dollars to build infrastructure, which is enough to do one of the following things:

1. High-speed, premium rail serving neighbourhoods A-F.

2. Really good busways (a la Brisbane) serving neighbourhoods A-L.

3. Busway network with ROW on city streets and traffic priortization, a huge improvement over what exists currently, but with some compromises and thus not a 'perfect' system that emulates rail. Nonetheless, serves all neighbourhoods A-T.

I don't know if my assumptions on the financials are perfectly accurate, so feel free to correct them. But as you can see, this is really a political question. What represents the most just approach for the citizens of our city? Is it fair to ask the residents of neighbourhoods G-T to make due without transit so we can build a premium rail network in A-F? And what about the poor residents in neighbourhoods M-T? They don't get any transit unless we go with the cheapest option.

(UPDATE:  Important clarification thanks to commenter Des:  Let's assume that all these neighborhoods, A-T, are more or less equally deserving of transit in terms of their density and demographics and development potential.  We're not talking about a case where neighborhoods M-T aren't dense enough to justify service.  We're talking about a case where they are, but where a high-end technology choice means we can only serve a few neighborhoods any time soon, and they're last in the queue only because somebody has to be.)

This isn't a hypothetical, of course.  In fact, it's a pretty good scheme for understanding the last two decades of transit development in North America.  And like so many American stories, the most vivid, Technicolor example happened in Los Angeles.

For most of the 1980s and 1990s, Los Angeles was pursuing Pantheon's method #1:  Huge investments in a few high-quality rail lines.  I recall vividly the debates in the press around 1983-85 about the emerging heavy-rail subway project that eventually became the Red Line.  A key idea in the discourse was:  "Los Angeles is a world class city, every bit as wonderful in our own way as Paris or London.  And world class cities have subways."  A strong driving sentiment was that Los Angeles needed not just rail, but the model of rail being modelled by London and Paris, the underground third-rail metro.

To be fair, it was never that clear-cut.  Two light rail lines also opened in the 1990s, the Blue Line to Long Beach and the peculiar Green Line in the median of the Century Freeway, which some would argue was really about getting the freeway built.  But rail was definitely the focus, and for the densest part of the city, the plan was for underground rail transit.  There was no major investment in bus infrastructure in this period.  The El Monte Transitway already existed, but it dates from the 1970s.

Because underground rapid transit is so expensive, they could only build it very slowly, so a only a few neighborhoods were going to be benefit in the early years.  That was the key tradeoff. 

(When people in young New World cities ask why we can't build rapid transit as fast and cheaply as Paris or New York did a century ago, I remind them that a century ago there were no environmental laws and labor was treated like dirt.  That made construction orders of magnitude cheaper than it is now.)

Then came the Bus Riders Union, representing some the vast range of neighborhoods that weren't going to get rail for decades under the plan, and the consent decree that required MTA to give more attention to the bus network.  This was followed by the revolt against underground rail in 1998, right as the Red Line was opening.

So focus shifted to cheaper things that could be done on the surface, and that could therefore cover more neighborhoods with the same dollars.  Since then, we've seen an explosion in the diversity of transit modes, including not just light rail but also three different degrees of Bus Rapid Transit:

• Exclusive and grade separated like Brisbane.  (Harbor Transitway, though the aesthetic standard is far below Brisbane's)
• Exclusive but at-grade with signals.  (Orange Line)
• Non-exclusive, at-grade, in traffic, but with wide stop spacing and signal priority (Metro Rapid)

The Metro Rapid is similar to strategy #3 in Pantheon's scheme, or actually one step further down, because it doesn't have exclusive right-of-way but does influence signals.  One way to think about the Metro Rapid is that it answered the question:  "What is the best level of rapid transit that we can afford to offer all over the city, now?"  When anyone complains that the Metro Rapid isn't everything rapid transit should be, I agree, but in the same breath I point out that 700 route-miles of it came into existence in less than a decade, and it's rather precision-designed to be the best rapid transit that could be delivered at that pace.

That's always the core of these debates for me.  I like riding trains more than riding buses, but I like a big network that covers the whole city even more than I like riding trains.  If you compare the Metro Rapid to the citywide rail network of your dreams, sure, it's inferior.  But Metro Rapid exists, you can use it right now, and while I disagree with some of the compromises made, I give MTA a lot of credit for getting it done.

Returning to Pantheon's schema, there's no denying that there's a tradeoff between expensive and wonderful service to a few places vs. less pleasant but functional service that covers most of your city.  There's also a tradeoff between expensive and wonderful service for your grandchildren vs. less wonderful but functional service that can be there next year, for you.  The Metro Rapid is especially useful as a model of the latter extreme, which is one reason so many North American cities are now working with the concept.  Los Angeles wants it all, but if you feel MTA should focus most of its energy on getting subways built, remember that they tried that approach in the 1980s, and it caused a massive backlash from both bus riders and taxpayers.  Given the history, the politics, and the incredible technical challenges of building transit in Los Angeles, I think MTA is doing a good job.  The Metro Rapid is an especially important model that deserves to be understood, and appreciated, for what it is.  Sure, it's inferior.  Sure, it's compromised.  But it exists!

UPDATE:  Nothing in this post should be read as implying that the current or recent levels of funding for transit in the US are adequate, or that the US should not be aiming to build transit at a more European or East Asian pace.  A sudden dramatic expansion in the money available for transit would obviously change many corridor-level debates, shifting many current recommendations in the direction of higher-quality options.  On the other hand, even such a renaissance would not eliminate the crucial role of Bus Rapid Transit at least as interim solution, because in such a scenario, corridors that current do not justify any project would rise to the level where at least a Metro Rapid would make sense.

Comments

[Alon Levy]

Well, there are cities in more recent years that built subways fast, and not just in China. The Tokyo subway's second line only opened in 1954, and its third line only in 1960; last year, the city opened its 13th subway line. The Mexico City Metro first opened in 1969; by 1994 it had 10 lines, and today its 11 lines have marginally lower ridership than the New York City Subway. The Madrid Metro has had most of its track length constructed in the last 30 years, as far as I can tell. Even in the US, some cities opened subways at reasonable speed in recent decades, for example Washington.

The issue isn't environmental protection. It's that SoCal was unwilling to make rail its primary method of getting around. When it did build rail for rail's sake, as on the Blue Line, ridership per dollar was fairly high. The underperformers are the Green Line, which was built as a freeway mitigation measure, and the Gold Line, which is slow and serves suburbia.

[Peter Parker]

Jarrett, may I offer an option 1a: high speed transit network serving suburbs A-F, with a co-ordinated local bus network serving suburbs G-T?

Then every suburb benefits, and if the rail allows inefficient freeway running of buses to be removed and local bus routes to be doubled in frequency (as happened in many area along Perth's Mandurah line).

One interesting performance measure of the relative development of city's transit is the proportion of a city's homes within 10 minutes walk of a service (say) every 15 minutes or better over a given span.

We rarely if ever read about this - transport departments report statistics such as number of passenger boardings, route kilometres run or kilometres of track.

Admittedly this is a service-based indicator rather than an infrastructure-based one. The former requires recurrent spending, and unfortunately can be wound back, unlike already-built infrastructure.

Your option 3, though service levels aren't mentioned, I suspect, distributes the more frequent service over a larger area than option 1, so is more akin to more service rather than more infrastructure.

If one was to look at Australian cities and compare the proportions of population near a 15 minute service mentioned above, I suspect you'll find Adelaide with it's extensive go-zone frequent network (but its crumbling diesel rail network) near the top.
Melbourne would be quite high due to its trams in the inner suburbs, while Perth would be a bit lower as a lot of areas are away from trains and frequent buses. Away from the busways and BUZ buses, Brisbane would rate lowly terribly due to infrequent buses and trains - see Brent Palmer's frequent service map: http://brentpalmer.blogspot.com/

Unfortunately Adelaide (along with Sydney) has been one of the underperformers in public transport patronage, and hasn't had the passenger growth that Melbourne, Brisbane and Perth have had.

So it looks like that pure service alone won't deliver the highest growth (not that Go Zones are particularly great on evenings/weekends and maybe half of Adelaide's population is still outside one) and there does need to be some form of infrastructure improvement as well.

[Des]

You need some hypothetical context to go with your hypothetical city. Neighbourhoods A-T probably differ from one another in ways that are important to the likely success and utility of transit.

Let's characterize the neighbourhoods as follows:
Neighbourhoods A-F are very dense, mixed-use, pedestrian-oriented, with scarce and expensive parking.

Neighbourhoods G-L, meanwhile, are variably dense, and though mostly single use, have some focal points where employment and activities concentrate and to which nearby people will certainly walk, though parking's not hard - sometimes you even have to pay.

Neighbourhoods M-T, however, are super low-density, half-acre lots, power-centres by the freeway junctions, with no mixed uses or walkability whatsoever and the only outdoor spaces are vast free parking lots.

Taking these factors into account, I would argue that one could justify varying spending as follows:
Concentrating spending on a high quality system (some sort of high frequency/capacity rapid transit) in A-F, where transit is a sure bet.

Fund some good mid-range initiatives and solid future plans for G-L (maybe BRT or limited stop service progressing to LRT?), where transit will have some success and there's probably potential for future growth (with the right kind of land use planning and urban design).

As for M-T, sorry, it's just not worth it in terms of transit's likely success (ridership, cost-recovery, general utility) and that perhaps they can organize carpools, jitneys, or hitchhiking schemes.

To make it less hypothetical, let's call A-F 'Parts of Vancouver', G-L, 'Most of Vancouver, Parts of the North Shore, Richmond, Burnaby, and New Westminster', and M-T 'Virtually everything South of the Fraser'.

Any thoughts?

[Tim]

I have to agree. In the best possible world I'd prefer to have a perfect bus network where no bus comes less than every 15 minutes. As nice as a train is, it's never going to get me everywhere I want to go without lots of walking. If I know that a transfer is never going to waste more than 15 minutes of my time I'm much more likely to take transit. And if I can go anywhere I'm much more likely to take transit. Buses are cheap and re-routable. They do make a lot of sense.

Of course, there are still a lot of people who will never ride buses because they smell worse than trains. Or something like that...

[Tim]

Err... more than 15 minutes. I'm fine with buses coming more frequently.

[EngineerScotty]

To make it less hypothetical, let's call A-F 'Parts of Vancouver', G-L, 'Most of Vancouver, Parts of the North Shore, Richmond, Burnaby, and New Westminster', and M-T 'Virtually everything South of the Fraser'.

Well, the US is south of the Fraser, so that kinda works. :)

[Alon Levy]

In Tel Aviv, the most frequent bus lines come every 3 minutes at rush hour and every 7 minutes outside rush hour. Many other lines come every 9-10 minutes. But they don't give good quality of service to city residents, since they are slow and bunching-prone, and often take circuitous routes.

[Jarrett at HumanTransit.org]

@ Peter Parker. Yes, Perth has been especially careful about spreading the benefits of the rail line far beyond the stations, through careful deployment of connecting bus services. We tried to do that in Portland, but the financial basis for bus service was, and still is, too volatile.

@ Des. Very important point. Actually, I'm assuming that all the neighborhoods mentioned ARE equally able to generate patronage, for the purposes of this example. Your point is important but different, namely that denser neighborhoods have to be front of the queue if ridership is the objective. I've added a clarification.

[Wad]

Jarrett wrote:

I agree, but in the same breath I point out that 700 route-miles of it came into existence in less than a decade

Time for a quibble.

Out of the 24 extant Metro Rapid lines, ( http://www.metro.net/riding_metro/bus_overview/bus700.htm ) , only four of them are "virgin" Metro Rapid lines: 705, 714, 734 and 741.

The remaining 20 all had antecedents as limited-stop bus services ( http://www.metro.net/riding_metro/bus_overview/bus300.htm ) .

What qualifies a local bus for a limited? Planners requires two conditions: high-frequency service and heavy transfer activity.

Before Metro Rapid, limited-stop buses had never been marketed for speed. Their primary purpose was to split trips to encourage passengers coming from or going to another bus to use the limiteds, while passengers needing only one bus to use the locals.

The higher speeds were merely a byproduct.

Speed only became relevant right around the time the FTA made bus rapid transit money available. Though I'm not saying the two are related. ;)

[Jarrett at HumanTransit.org]

@ Wad. Wad. You keep bringing up the limiteds, and I never quite get what you're after. Yes, there used to be limteds, but my understanding is that they didn't have the standardized frequency and service span, nor did they have the signal priority capabilities of the Rapid. Limiteds used to be thought of as supplements to locals only when demand warrants, and the Rapid program changed that priority and created standard service levels all day.

So although some existing service was logically folded into it, it was a new level of investment and a new level of mobility, where this is understood as incorporating speed, frequency, and span.

Please update me if I have this wrong.

[anonymouse]

The Rapids are generally the Limited renumbered from 3xx to 7xx, painted red, sometimes with signal priority and separate stops, and generally their introduction was accompanied with an upgrade of service by one level up the hierarchy of no service, rush hour only, weekday only, 7 day service.

Anyhow, the basic problem with transit in LA is that it literally takes 3 times longer than driving, fairly consistently, with the exception of the rail lines, which are maybe 1.5x slower than driving during off peak and often actually faster during the peak. Introducing rapid buses might change the 3x to 2.5x, but that still means that a trip that takes 45 minutes of driving takes an hour longer by bus rather than an hour and a half. Rapid buses are a good way to improve the efficiency of existing service and serving the existing ridership more cost-effectively, but in order to attract new riders, you need to get them out of their cars, and you just can't compete there unless you have the speed, and there's no way to get that in the densely built parts of LA except with subways.

[Jarrett at HumanTransit.org]

@ anonymouse. Not claiming for a moment that you don't need to be pursuing subways and other high-capacity options long term. But the initial study of the Wilshire-Whitter Rapid did show substantial new ridership, and even if it was the same people shifting from the locals, that meant they were served substantially faster. The signal priority and frequency upgrades were an significant improvement even for passengers who shifted from the limiteds.

Establishment of standard 5-day or 7-day all-day frequencies substantially changes the attractiveness and usefulness of any service, as does the consistent branding that makes the network visible. So the fact that the Rapid was created out of pre-existing limiteds, to some extent, doesn't diminish the value of the idea or the effort involved. Obviously there are huge operating costs bound up in the Rapids and no agency could add that much service at once without streamlining what was already there.

[Pantheon]

Another good example of the third option in my example is the Sound Transit express buses in Seattle. Are they as fast as rail? No. Are they a fully protected busway a la Brisbane? No. But they get you way out into the suburbs really fast because they take the freeway and have no stops between downtown and wherever they're going. And most importantly, they exist! We don't have to wait until 2030 or 2050 - there are two dozen routes that are here now, serving all the suburban communities equitably.

Another point about comfort. A lot of people say buses will never be as nice as the train. But I would rather ride a ST Express bus over the Link light rail or MAX any day. Comfort is a subjective thing. I really like the plush seating and the inteior design that resembles a Greyhound coach service. The coach-style design probably reflects lower ridership numbers. But it facilitates a much more private, luxuriant ride than you would normally associate with public transit.

I don't know if a train could do the trip faster, but I can't imagine it would shave off very many minutes. Bottom line: I was out in the far flung suburbs of Seattle in 20 minutes, enjoyed a plush ride with near total quiet and privacy, and was dropped off within a block of where I needed to go. Hard to beat that.

[Jarrett at HumanTransit.org]

@ Pantheon. Yes, though there's an important distinction between frequent rapid express buses, like the Sound Transit 550 Seattle-Bellevue, and peak-focused express buses, which are actually very inefficient because their markets are so one-directional.

The 550 between Seattle and Bellevue is very close to being completely exclusive ROW, except for the last couple of miles into Bellevue.

[Wad]

You keep bringing up the limiteds, and I never quite get what you're after.

Here's the high concept: Metro Rapid is rapid transit with a s--- eating grin.

Yes, there used to be limteds, but my understanding is that they didn't have the standardized frequency and service span, nor did they have the signal priority capabilities of the Rapid.

The signal "priority" capabilities are pointless.

LADOT, not Metro, controls the signal priority. And you know what happens when you let a bunch of traffic engineers solve a transit problem.

Buses do not get priority. They get conditional signal extensions. First, the signal system must check to see that intersecting traffic is not backed up. Second, LADOT fixes signal priority to the lines' fixed headway.

Example: If a Rapid line runs every 15 minutes, and the first of two buses is 10 minutes late (with the trailing bus on time but 5 minutes behind the late bus), the late bus will get the alloted signal pre-emption, but the signal system won't give the on-time bus pre-emption.

Limiteds used to be thought of as supplements to locals only when demand warrants, and the Rapid program changed that priority and created standard service levels all day.

No, the BRT grant changed that priority.

Metro does a pole dance to deceive minds into believing that a long-standing operational solution is now considered the equivalent of a rail service.

Here's a simple syllogism.

1. Operationally, limited-stop buses and Rapid buses are identical.
2. Under BRT grant guidelines, a bus rapid transit application as the agency defines it is the service equivalent of a rail service. (I am reminded of the Michael Setty vs. Tom Rubin debate over modal equivalency.)

Therefore

3. Limited-stop buses and rail are identical.

This is not my logic. It's Metro's.

Let me make myself clear. I am fine with the delivery of Rapid service. I do caution people to avoid falling into the logic pit that I have outlined above.

[Jarrett at HumanTransit.org]

Wad.  So you're stuck on the fact that the BRT grant used for Metro Rapid mentioned comparability to rail?  People say what they need to say to get grants, and always will.  You're closer to the ground, but in the presentations I've seen on this, I've never heard LACMTA staff claim that the Rapid is "as good as rail" in any sense.  Rex Gephardt certainly doesn't describe it that way.  I would praise Metro Rapid even if it had been nothing but a rebranding of the pre-existing limiteds, because of the way it shifts focus to rapid-stop service and presents it as an integrated product.  And yes, I share your impatience with the compromises made in implementation, but as we saw on the Gold Line, rail can get compromised too.

[Russ]

Jarrett, I am not sure this works for me as an abstract idea. If we take a city to be of uniform density (and have a uniform distribution of destinations), then it stands to reason that it will have a uniform transit service as well. Which means, there are only two reasons why you might not want to cover the whole city simultaneously:
1) that the lowest level of service is inadequate to meet demand, and therefore will need upgrading immediately, or;
2) that there is such a pronounced network effect that it makes more sense to cover a contiguous segment of the city well, to recoup a higher farebox revenue (and lessen the funding burden for future upgrades.

The latter point may have some relevance here, but depends on the funding model.

But, because real cities are neither uniform nor unresponsive to changes in transport accessibility, I am not sure this type of thought process is even a good starting point for analysis. It implies that the main two measures of good service planning are the percentage of neighbourhoods with a service, and the quality of that service, when they are only measures of service provision.

Unlike Peter, therefore, I don't think a measure of service accessibility is very helpful. If used as a target, it implies putting in a lot of services of low potential, in preference to services between well frequented areas. I can see the political value of this type of measure, but from a planning perspective I tend to think it leads to waste.

[Jarrett]

@ Russ. I think this discussion has gotten detached from the original context of the post. As a means of analysis for network coverage in general I agree completely with your point. But this post was specifically about very expensive investments -- e.g. rail transit subways -- that can only be deployed gradually due to their expense, and which inevitably do constitute incomplete networks that do not cover all areas that could support them.

[Alon Levy]

You could still have bus-rail connections, though. One of the reasons Calgary's light rail succeeded is that it retooled its bus system to feed the new rail service. New York did something similar in the 1920s and 30s, and again in the 1980s, changing the bus schedules to converge at the outlying terminals of new subway lines.

[Jarrett]

@Alon. Yes, of course. A supporting bus network is always essential.

[anonymouse]

And by the way, Metro Rapid does not imply any uniform standard of service. On some lines, the buses run frequently every day and into the evenings. On others, it's peak only service with 15 minute headways. Sometimes, only part of the line will run on weekends. And there isn't always much information at the stops themselves either, especially about the Rapid network as a whole, which is one place where some marketing might actually be beneficial: showing people all the places they can get to with guaranteed frequent service.

As for ridership, I've been looking at LA ridership statistics every once in a while, and my impression is that the general trend seems to be a long but very slow decline on the bus network, with most of the growth coming from rail.

[Russ]

Jarrett, I'd argue that all investments are necessarily deployed gradually, but I see your point that some may be deployed significantly more slowly. I'd also argue though, that transport planners fail to adequately conceptualise the sort of transitional change that is required under those constraints.

To put it another way, you only get a clean slate once. Every future transport project will be assessed int he context of what is there, and I'd argue, that if your end goal is a subway system, that the existence of BRT lines will be an impediment to that conversion. It will be argued that you can't remove the old line because it will have developed a niche market of shorter trips that the subway will serve poorly; it will also be argued that the subway will not add much value, because a BRT already serves the area, and therefore it crowds out potential trips, and lowers the investment potential of the new line. So, while I see your point that something is better than nothing, the manner in which transport interacts with the city risks embedding something worse than what you might actually prefer (or need).

[bzcat]

In general, I agree with Wad on all points except when he said:

"1. Operationally, limited-stop buses and Rapid buses are identical."

I would think creating a whole new fleet of buses that cannot be used for local/limited service and having to maintain the dual fleet creates all kinds of operational and service dispatch problems. The lack of flexibility and the capacity redundancy required to maintain a dedicated Rapids fleet the size of Metro's must be pretty wasteful. Instead of planning for peak capacity for one fleet of buses, you now need to plan for peak capacity on 2 parallel fleets.

[Pantheon]

To Russ:

You make an interesting point, which I think merits some discussion. Clearly, there are tradeoffs no matter which path we take. These tradeoffs involve choices which may benefit one neighbourhood or generation over another. But even if we assume for the sake of argument that busways are an inferior form of rapid transit, the tradeoff you advocate represents a more severe inequality. There are ethical problems with asking one generation to do all the sacrificing so that another will reap all the rewards. A city is, after all, composed of real people who live with the consequences of these decisions.

Your argument reminds of of "political correctness" in its orginal, Bolshevik incarnation. According to this meaning, actions were deemed to be correct or incorrect solely based on their political consequences, rather than the intrinsic morality or immorality of the actions themselves. The implicit assumption was that promoting and maintaining a communist state was of such overarching importance that an action could be politically justifiable even if it were inherently immoral.

Under this line of reasoning, it was politically incorrect to feed starving peasants loaves of bread - even if loaves were freely available - because to ameliorate their suffering would only dull their revolutionary fervour. This was not merely theoretical - decisions such as this were made repeatedly, and many suffered the consequences.

What you propose is a logically identical form of reasoning. In your view, it is not justifiable to provide busways to the citizens of our imaginary city on the grounds that they might actually be happy with those busways, and will therefore be less likely to advocate for a rail system. Better that they should suffer, so they will be on their knees begging for their rail salvation.

This brand of ruthlessly utilitarian thinking is something I cannot countenance, and its very existence shows that rail has become a kind of irrational faith among transit planners. Like any faith - whether spirtual, philosophical, or political - the ideal is cherished so completely that nothing is too worthy to be sacrificed in its service. And as with any faith, tragedy piles upon tragedy, ending in folly.

On a more pedestrian note, if rail is so great then it should have nothing to fear from busways. And if it does fear them, then maybe it isn't so great.

[Russ]

Pantheon, your first mistake is to assume I am some sort of irrational rail booster. On the contrary, that categorisation would come as a great surprise to the people I've offended by saying a lot of rail spending is a waste of money. The best transit is probably the most cost effective, and if that is bus-ways then by all means use them.

What I was trying to note, is that the a first order analysis is insufficient, because there are ongoing opportunity costs (on the order of decades) if a particular service is chosen now, instead of waiting to build something superior.

Let's look at it mathematically though. Suppose a subway provides a slightly better service over a busway (if it doesn't why would it be built?), but won't be built for 20 years. Hypothetically then, the busway might serve 10% of the population now, rising to 20% in 100 years time. The subway, by contrast might serve 15% of the population when it is built, rising to 25% in 100 years. In this scenario, the extra few percent of riders a subway will serve would never be justification for a subway. Yet, multiply those figures out, and the busway comes to 15x100 = 1500, and the subway 20x80 = 1600. In other words, building a busway benefits this generation at the expense of an unserved future generation. More importantly, over its life-span, the subway serves more people in total.

So, yes, it is a utilitarian calculation, but not a political one because I'm a rail-fan, but a practical one because it is reasonable to suppose a better service will serve more people. Cities don't change very much, because requisitioning a right-of-way is a really hard sell politically (and financially). But I am not claiming giving everyone a service is a bad idea, merely proposing that the alternative might have hidden merits, and they are worth discussing.

[Alon Levy]

Pantheon, can you please take yourself seriously? The communists starved people to death; rail advocates think that mediocre bus service is going to fail to achieve its goals while reducing political pressure to build good transit don't propose anything of that sort.

And where did that crack at political correctness come from? William "Slavery was no big deal, but Reconstruction was terrible" Lind?

[M1EK]

This ignores political reality. In most US cities, it really operates like this:

The transit system in Our City gets almost all its funding from a local sales tax. Most middle-class people never set foot on the transit system today. The transit system is woefully underfunded compared to European models, and is critically dependent on voters for its very survival - voters that do not, today, use the system (in aggregate - those middle-class voters turn out in higher numbers than do the transit dependent).

Rail may not gather a majority of the voters as passengers even when implemented perfectly, but it garners a majority of the voters as "potential passengers". IE, the voters can see themselves riding that train, or something like it. They end up supporting expansions of the rail system AND continued support of the bus system (a la Dallas and Houston). Eventually, you get to a point where a lot of middle-class people really DO use the transit system - and some even migrate to buses, at which point your system is secure from the road warriors who would have raided it for freeway dollars.

If you don't build any good rail at all, you lose those voters. Then, you lose the sales tax revenue for your existing bus service. Then, the transit-dependent people end up suffering even more than they would with the presumed under-investment in their bus service in the "all rail" case.

(Of course, it really isn't under-investment; what ends up happening in reality is that a good LRT line ends up supporting more bus service off the corridor as existing, already owned, buses can be pulled off that line and used elsewhere).

[Dan Wentzel]

M1EK,

You've just described the difference between what happened in Los Angeles County where Measure R passed (raising the sales tax a half-cent in the middle of a recession) and Orange County where 8% of all bus service is being cut while freeways continue to be expanded.

[Pantheon]

To Alon:

Two arguments can be identical in kind, even if their consequences differ in degree.

To Russ:

If a busway is significantly less expensive than the subway, it may be a more rational investment despite the lower ridership. Sure, it would have to make compromises. But I would rather have something than nothing. And I don't think there is anything about busways that preclude future rail investments, but that is a separate point.

On further reflection, there is an additional ethical problem with using the tax dollars from our entire city to fund projects that benefit only some neighbourhoods. If you live in neighbourhood T, how would you feel about your tax dollars being used to fund rail lines to neighbourhoods A-F, while you get nothing?

This is an ethical problem regardless, but it gets even worse when transit is used for "development" purposes. The Portland Streetcar is utterly appalling - a so-called transit investment that is really a thinly veiled government subsidy for the purpose of enriching multimillionaire developers. But I digress.

[EngineerScotty]

Pantheon...

Ensuring an equitable distribution of tax revenues to infrastructure has long been a tricky problem. In the Portland context, the suburbs (particularly Washington County) love to whine about subsidizing inner-city transit (which, according to rumor, is one reason WES was built).

Overall, Portland does a pretty good job--MAX serves quite a few poor (or even blighted) neighborhoods, and currently the most exclusive parts of town (Lake Oswego, Dunthorpe/Riverdale, other parts of the West Hills) have no MAX service.

Transit-oriented development is an interesting issue, and you bring up a big problem that I wish Jarrett would blog more about (hint hint!)--many consider speculative transit development, where service is provided to a greenfield or brownfield (or even an existing neighborhood targeted for gentrification) in anticipation of high-density, transit-friendly development, to be corruption. Developers do, after all, benefit from such things, as do many other powerful actors (construction interests, construction unions). On the other hand, the charge that politicians are doing transit project for political patronage purposes, rather than mobility improvements, is probably made far more than it is actually justified. In the specific case of the Portland Streetcar, there isn't any credible evidence that I'm aware of that anything illegal or below-board occurred--simply the fact that a developer made money isn't enough (in my mind) to allege corruption.

TriMet's position on the streetcar seems to be that if a neighborhood wants to build streetcar lines, TriMet is happy to operate them; otherwise the 40' bus is the standard local service. For busy local corridors, streetcar may offer gains in operational efficiency, and attract riders who wouldn't otherwise ride--otherwise, it doesn't offer significant advantages over the bus.

[Pantheon]

To EngineerScotty:

Transit-oriented development is one thing, but perhaps the more appropriate term for this is development-oriented transit.

I don't have any specific evidence of political corruption. However, I believe that if it smells fishy, there's probably a fish around somewhere. The Portland Streetcar just happens to link one huge condo development area (south waterfront) to another (pearl). It's an awfully fortuitous route for the developers.

The Pearl district in general is clearly awash in government subsidy - from boardwalks to lighted trees, to manicured parks with their fake ponds and streams. It's like an urban mall. Did the developers pay for any of this, from which they so clearly benefit?

The Streetcar is just another piece of this ugly picture. Even if we cannot prove corruption, there should be a simple rule that developers and retail businesses should pay for things which clearly benefit them.

Another rule: whenever transit service is proposed to an existing brownfield, a private investigator should be automatically appointed to find out who owns that brownfield, and if he has recently purchased any black leather suitcases or made large cash withdrawals. Especially if that brownfield is in Portland.

[Jarrett at HumanTransit.org]

@ Scotty and Pantheon.  Yes, I agree with Scotty.  Urban development involves public-private negotiations all the time.  Ideally, these negotiations happen in the context of a shared ideology about what kind of city the parties are trying to create.  But it's still a negotiation between two parties with bottom lines, who would each prefer that the other contributed more.  Pantheon, for the record, by the time something as complicated as the Pearl gets done there are all kinds of cross currents of public and private money.  But please remember that a large part of the Streetcar, and I'm sure a lot of the other Pearl improvements, were done by a Limited Improvement District (LID) which is an assessment on property owners (i.e. developers or their clients).  It's not corruption when a government makes a deal with a private investor to do something they both want to do, and that requires pooling their resources.  You can accuse the government of poor negotiating and thus paying too large a share, but even poor negotiating is not corruption. 

[Alon Levy]

Two arguments can be identical in kind, even if their consequences differ in degree.

Okay, so you're insane. You think that a government that due to incompetence causes 5 million people to starve to death is really making the same argument as people who argue against compromising on good transit.

What socialists do say on the model of what transit advocates argue is that piecemeal reforms on issues like health or social security are likely to neglect the poor while making the middle class stop worrying about the poor. This has nothing to do with what Lenin or Stalin or Mao actually did - they caused mass starvation either through incompetence or as a means of genocide rather than as a way of furthering the revolution - and the people who make those arguments have never called for killing people.

But please, keep digging.

[Russ]

Pantheon, sure. I am not arguing there aren't circumstances when a busway isn't a better investment. It often is. Merely that the consequences of pursuing that investment in terms of future planning needs to be part of the calculation. Which they aren't, in the original formulation of the problem.

The argument that there is an ethical problem of an inequitable distribution between neighbourhoods is specious. Why choose neighbourhoods as your point of distribution, why not cities, or states, since tax dollars are regularly flowing from federal authorities, or in reverse, why not streets, or individual houses? Transit varies greatly in cost between different areas, and has widely different levels of use, and that will be reflected in the funding it receives in different places. Yes, it is more equitable to fairly distribute transit, but I would put it somewhere down the list of goals a transit planner should be concerning themselves with, not least because equality of transit distribution is likely to encourage decentralised planning and sprawl, which will correlate with higher auto-use.

It isn't even an easy question what would count as an equitable distribution of transit. Having a service near your house is a significantly different proposition to having a service go where you want it to go, and it is the latter that should determines what counts as equality of service provision.

[Pantheon]

To Russ:

My argument is far from specious. Again, you are disentangling my argument from its original context. I suggest you re-read Jarrett's post. We are not talking about forced equality amongst neighbourhoods with vastly different transit needs as part of a comprehensive plan. Rather, about a situation where one mode of transit is so expensive that its deployment results in an incomplete network serving only a small fraction of the city for decades on end.

[EngineerScotty]

Here's another way to look at it. Portland's first branch of its MAX system connected downtown with the suburb of Gresham. In reality, the project was somewhat constrained to go there due to the fact that it sorta replaced the Mount Hood Freeway, which have built, would have served Gresham as well.

Since then, four additional major MAX projects have been undertaken--westside, airport, Interstate, and the recently completed Green Line/Transit Mall project. Now, significant parts of the city have MAX service. The Milwaukie line will start construction next year, and a few others are in the pipe.

What if, instead of doing that, the city of Portland and Tri-Met spent the money on bus infrastructure projects, throughout the town--and each succeeding MAX project was likewise replaced with a comprehensive set of further improvements. What would the result look like?

I'm not sure what it would look like, and if well-done, it might have turned into something truly useful. OTOH, I doubt it would have turned into anything like the current result, which is LRT serving a large part of the area, with significant additional service in the works.

Why not?

My concern with the development strategy you outline, is that if you are building a system in pieces, doing a little bit for everyone at each step may involve significant amounts of rework. If the end goal is "high quality transit throughout the region", and you don't have enough money available do to that in one step--it's often more efficient to build the system in pieces, then it is to build a lower quality but comprehensive system, and then upgrade. I'm not too concerned about the possibility that the populace will decide that good enough is good enough (if they do, then so be it)--but the upgrade strategy may result in more money spent to achieve the same end result.

If the choice is between quality transit for a few vs acceptable transit for many, and only one project will ever get done, then yeah--the more comprehensive solution will (in the abstract case) be more attractive in my opinion. If, on the other hand, the initial project is indeed the first of many--and there's a good chance that the entire system will get built, then high-quality service to a part of town may be the more economical first step.

Obviously, this generic analysis assumes the abstract case outlined in the lead. Given more specific data about a particular region, I might come up with a different answer.

[Alon Levy]

Rather, about a situation where one mode of transit is so expensive that its deployment results in an incomplete network serving only a small fraction of the city for decades on end.

Or, a situation where a fraction of the city gets a one-seat ride on rail, and everyone else transfers.

I'm afraid to mention Calgary again - it's not LA, so LA can't learn from it.

[Jarrett at HumanTransit.org]

Scotty. Here's what I think it would have looked like:

(a) Fully separated busways, Brisbane style, Goose Hollow to Beaverton and Lloyd Center to Gateway. Extensive branching at Beaverton and Gateway to provide direct if not-so-frequent to major outer corridors like TV Highway or outer Stark.
(b) At-grade busway with signals, Beaverton to Hillsboro, but with capacity for buses to branch off and serve other major destinations like Tanasbourne. A similar busway done by now along McLoughlin, Hawthorne to Milwaukie.
(c) On-street median busways Gateway to Gresham, and along Interstate, but probably also in many other corridors where space permits but where rail hasn't gotten yet, certainly including Barbur, outer McLoughlin, Lake/Harmony, perhaps even Macadam and TV Highway.
(d) The transit mall in its pre-rail form, but with wider stop spacing to increase capacity. Possibly east-west transit mall in a similar configuration.
(e) Rumblings about now that the volume of buses on the surface downtown is getting a bit much, though we're certainly glad we gave the buses room to move unlike Ottawa or pre-2009 Minneapolis. Discussions that maybe it's time to imitate Brisbane and build one or two busway subways under downtown.

Just thinking out loud. I'm not saying Portland would be a better place, or as transit-oriented as it is now, and not that I advocate it. But as a thought experiment, I suspect it would look more or less like that.

[Alon Levy]

Jarrett, re: e), is there a safe way to build a tunnel for diesel buses?

[Jarrett at HumanTransit.org]

I'm not an engineer, but I know they exist. Brisbane's busway has a two-station underground segment; the fleet is part diesel but converting to CNG.  Significantly, sliding doors are used to separate platform from roadway, as shown here:  http://www.humantransit.org/2009/04/brisbanes-new-downtown-subway.html

 I have personally witnessed ordinary diesel buses running through the
Seattle Transit Tunnel, though it's not standard procedure.

[Nathanael]

"1. High-speed, premium rail serving neighbourhoods A-F.

2. Really good busways (a la Brisbane) serving neighbourhoods A-L."

Correct that to "really good busways servicing neighborhoods A-C ", and you have your pricing correct.

That's one of the problems with this entire exercise. Bad assumptions.

[Nathanael]

Oh, and I have an even better point which hasn't been considered yet.

We want to encourage efficient land use. Less expense in terms of road/water/sewer/electric/telecom construction & maintenance, police/fire operations, damage to ecosystems, etc.

This means we do *not* want to spread mass transit out to as many neighborhoods as possible. We want to hit the densest neighborhoods and those with the most potential for density first, and encourage people to move there. So (1) may actually be a *BETTER* option from a land-use perspective.

[Nathanael]

"I like a big network that covers the whole city even more than I like riding trains."

Well, but you like taking random walks through cities too!

I realized that the Wilshire Corridor (future Red Line) really does contain most of the places most people want to go in LA. So a fair number of people may well prefer a dense spinal line to a widespread network which serves that high-density line poorly....

The Your Mileage May Vary factor seems to be very large.

[TransitPlanner]

The Metro Rapid in Los Angeles does not function as a useable network because service span, frequency and routing varies dramatically from route to route. There isn't enough money to completely overlay the Metro Rapid network over the Metro Local network, so it is done picemeal. Very complicated and confusing onfusing to potential passengers. The rail network is very simple to unsderstand by comparison. One way to solove this problem would be to consolidate and have every bus line be a Rapid Bus !