Tuesday, January 23, 2007

Cascade On Rail: If Facts Are Oxygen, Get the Ventilator

Carla's obviously braver than me to visit the Cascade Policy Institute on a recurring basis; I've been there once and ran screaming from all the prejudicial inquiry and shoddy misuse of statistics. She knew I'd get a kick out of the latest "Quick Point" from the Institute's President and CEO, John Charles. I'm sure he has quite a busy schedule running a think tank with so much damned science to cast aspersions on, which is why there's apparently no time left for things like references, stated facts or honest reporting of the numbers in his own work.

Start reading and shout out "says who?" when you get to the parts Charles concocts from whole cloth (if you can't shout where you are, furrow your brow and cock your head in a gesture of skepticism at the monitor instead):
The current transit mall is highly productive, moving a lot of bus passengers in a small amount of space. When rail construction is finished, the mall will actually have lower passenger throughput then it has today, because light-rail is a low-speed, low-capacity system.
Brrrp! Stop right there. First of all, let's confront physics, my good man. A standard two-car train will hold 200 people comfortably, a stuffed bus no more than about 50. That four buses weaving in and out irregularly will crowd the transit mall much more than one smooth-moving, predictable-path train seems demonstrably obvious, does it not? Regardless, mere recognition of the fact that buses and trains will share the mall, and that the number of bus lines moved off it will be few, renders bogus Charles' claim that "the mall will have lower throughput." Bus + train > bus only. Duh.

But let's not just toss away the lame attempt to pretend that rail service is inefficient. On the contrary, rail actually not only creates its own ridership, it also has been shown to contribute or at least coincide with an increase in bus service as well. At least that's what the Transportation Research Board of the NRC had to say just as MAX began to take hold in Portland:
In Portland, as post-WWII conversion of rail lines to bus accelerated, ridership dropped 14 percent per year—one of the sharpest declines in the nation. The exception was ridership on the suburban rail lines before conversion. After the less-severe decline that occurred when Saturday was phased out as a workday, ridership began to grow again, paralleling the experience in Shaker Heights, Ohio. Then the highway department closed the bridge into downtown for repairs, truncating rail service short of downtown and requiring a shuttle bus to complete the trip. This depressed ridership severely (although not as severely as Portland’s bus ridership) until a 33 percent fare increase was applied. Service was then discontinued in defiance of the Public Utility Commission. An appeal to the court was fruitless because the highway department had rebuilt the bridge without rails. Bus ridership continued its sharp decline, and by 1958, ridership was down 74 percent (14).

A 15-mi eastern radial rail line opened in Portland in 1986. Declining ridership on the all-bus system became increasing ridership on the new combination system. The cost per passenger declined. Light rail is now carrying 11 percent of the passengers on 4 percent of the service. The cost per rail passenger-mile is only 20 cents, compared to 40 cents by bus (36). The synergistic effect has now increased the number of bus riders.
Care to have it put into ubergeeky abstractese, so that Charles can understand it? Here are two colleagues of the TRB presenting their findings at an APTA conference:
Rail modes attract greater utilized capacity per unit of offered capacity during peak period than bus modes, and this aspect of consumer choice may be quantified by regression analysis. Data and observations fail to support alternative hypotheses to a consistent and observable consumer preference for rail.
The mall will handle greater capacity, not less. And I believe it will actually make buses more efficient as peak-hour commuters out of town dominate that mode, while short-hoppers and tourists will stick to the train. But that's just my analysis. I don't have a reference. Thenagain, neither does Charles when he says this:
The mall will also be much more dangerous. TriMet bus drivers have warned public officials that mixing rail, buses, cars and bicyclists in tight quarters will lead to fatalities, especially since the rail cars will weave back and forth in a serpentine pattern.
I've watched that movie he links to a dozen times or more, and it tends to look safer and more natural every time I watch it. The number one enemy of traffic is irregularity that throws off flow. The pattern of bus/rail alternation as they move down the mall is the very same one done bus-to-bus now. The only change is instead of three bus slots--any of which may (and do) pull out into the middle traffic lane at any time--there's a single train slot that always pulls into traffic starting at the intersection, and always has the right of way compared to the center lane. The left lane is for cars all the way down; no reason ever to stray into other lanes.

Am I going to stand up and say there will be no accidents? Of course not. It could be pretty confusing at the beginning (although people said the same thing about the move to 3rd and 4th, and I'll be damned but I honestly think it runs BETTER that way so far), but just like when the original mall was built, confusion will fade. Of course I'm not forced to prove it won't be dangerous; what I'd like to know is what--besides fearful, unnamed bus drivers--proves it will be, as he so categorically states.

One of the best features of light rail is where riders come from. The Tuscon project linked above indicated a fairly consistent 30% share coming from car drivers choosing to use rail (whereas bus riders often have no other choice). Portland is actually kind of spoiled; in other towns I've been in it's nowhere near as socially acceptable to be a bus rider by choice. It's true for Portland Metro in spots too; the service for Lake Oswego's 30,000 people is among the most dismal of any area TriMet covers--and they say it's because no one will ride it. But you bet your ass the downtown Oswegans will line up for a smooth streetcar ride into town. Actually, I was a little disappointed to see Charles fail to tee up one of the most popular canards of the anti-rail railers, that rail increases congestion. Seat for seat, nothing takes cars off the road like rail, so you do the math.

Let's go back one more time to what he did tee up:
TriMet will be spending more than $200 million for this project, despite the fact that rail is rapidly losing market share. According to a recent employer survey, light-rail’s share of downtown commute trips dropped by 29% over the past 5 years, despite the addition of two new rail lines during that period.
That last bit is correct as far as it goes, although you can see from comparing the years that the ridership share fluctuated back and forth, hitting the exact same rate in 2004 as in 2001, before dropping in 2005. Even then, the figure used is for the downtown commute share only, which is hardly indicative of the full extent of rail usage (I've done enough MAX rides to PGE Park to prove that).

You know Charles was fishing for a convenient number, because he used such a bizarrely narrow way to view ridership comparisons. There are all kinds of comparative measures, and TriMet made a nice little graph so you could compare them {pdf}, on page four. And you know what? From '01 to '06, the "market share" for rail vs buses was up in each of these categories:
  • vehicle hours
  • revenue hours
  • vehicle miles
  • revenue miles
  • passenger miles
  • originating riders
  • and boarding riders--usually by 50% or more.
Ah, but that's just flimsy data--actual recorded ridership, mileage and revenue. Charles has a business survey that debunks all that in one tangentially worded question!

C'mon, John. Don't waste our time with that premature ejaculate of a column. You should know better after (correctly, to an extent) calling out Portland's Kyoto numbers. That was an honest mistake in a still-remarkable result. What you write in your latest column is misleadingly stated at best, totally unsupported and contradicted by the actual literature at worst. Booo.