There’s been quite a bit in the national press this week about the subject. Amanda Hess, writing for Atlantic Cities, penned an article called Race, class, and the Stigma of Riding the Bus in America. This article drew quite a bit of attention, including a somewhat scathing criticism from Jarrett Walker, who objected to the article’s focus on “white people” as a proxy for higher economic classes, particularly in a the context of a majority-minority city like Los Angeles (on which the Hess article focuses). Not to be outdone, the long-time transit and poverty advocates supporters at Reason (yes, that was sarcasm) penned an article called How Rail Screws the Poor, accusing LACMTA of building rail projects to wealthy communities while cutting bus service elsewhere. Zooming out, a new Brookings Institute report looks at employer access to labor via transit in major US cities. (Portland does quite well on this report, actually).
The issue of whether how well transit systems ought to be optimized to focus on the “transit-dependent”–those who generally don’t have reliable alternatives to public transit for journeys too long to walk or bike–is a major topic in local transit debates. OPAL got started as advocates for those in poverty (though they’ve expanded their focus to all transit riders), and service to the poor is still a major issue. TriMet, for its part, claims that “equity” is a big part of their mission, though many of the agency’s critics would likely dispute that (or argue they are failing to carry this mission out). Unfortunately, the subject is frequently a major source of heat rather than light.
On the transit dependent
Many transit advocates hate the term “transit-dependent”. Many of us dislike even more the term “captive rider”. Both terms, and especially the latter, frame transit as a bad thing–something to be avoided if possible. (This is one reason I like to talk about the auto-dependent; whether a car is liberating or limiting depends on where you live). The terms also promote the notion that there exists a class of rider whose needs not be considered when doing service planning–they will use the service no matter what, so a transit agency can safely go ahead and ignore (or screw) them. (The private sector gouges the poor enough already).
However, it is clear that there are some people who have greater dependency on public transit than others; and in virtually all of the country, land use and infrastructure investments favor motorists over transit riders. Pretending otherwise isn’t beneficial–it fails to reflect reality, and it can also lead to those who depend upon transit to have their needs neglected. Even if an organization cares deeply about the poor and makes equitable service part of its mission, the needs of the “transit-dependent” rider will often be different than the needs of the so-called “choice rider”; and failure to account for these difference will often look like gouging or neglect.
Transit elasticity of demand
One way to characterize a rider’s dependence upon something is via the economic concept of elasticity. Elasticity (assuming that it can be measured with some precision) is roughly defined as the marginal percent change in consumption divided by the marginal percent change in price (or some other parameter)–roughly speaking, it’s a normalized derivative of the demand curve. Price is the usual parameter for elasticity, but other parameters besides price drive consumer behavior. In the case of transit, patrons are also interested in things like coverage, service span, frequency, reliability, speed, comfort, safety, social status, and amenities. Many of these things are difficult to quantify, let alone subject to a rigorous elasticity analysis, but if transit is made more useful and/or pleasant to use, it will attract riders; conversely if it is made to suck more, than they will be driven away.
Overall, a commonly used “overall” price elasticity figure for transit is -0.3, which means that for a given percentage rise in price, one can expect a -0.3% change in ridership. (The -0.3 figure is somewhat disputed; here is a good reference). If one accepts -0.3 as a valid value, this means that raising prices will increase revenue–the additional fare revenue will only be partially offset by passengers leaving the service–an important point to remember when considering TriMet’s FY13 budget planning. (Interestingly enough, analysis of the Portland Streetcar indicates that raising prices does not raise revenue; implying an elasticity closer to -1.0. If this is correct, this would imply that either the Streetcar ridership is less dependent on transit in general; or many trips on the Streetcar are more easily substituted. Anecdotal evidence seems to suggest the latter; it is often used for short trips that could easily be made on foot).
Often times, elasticity measures are computed and reported for an entire population. In practice, different members of a given population will express different elasticities for something. In the case of transit, this manifests itself as a significant subset of the ridership whose personal elasticity is closer to 0–the transit dependent–and another subset whose elasticity is far greater (in the negative direction) than the mean.
What do the dependent need?
Some riders are completely dependent on the service–they have no access to a car (or other motorized transport) whatsoever. Others may have limited dependency–someone in a household with more adults than automobiles, or who uses a service like Zipcar may have part-time access to an automobile, but not have one available on demand. Others have a car in the garage, ready to run on a moment’s notice, and elects to use transit only to avoid a difficult commute (or for reasons such as reducing one’s personal carbon footprint). These groups of riders have very different needs.
The first group needs comprehensive service–that runs to all parts of the cities, at as many hours of the day as possible. Frequency and speed are important as well, but likely take a back seat to basic availability. If you have no option other than transit, and the bus or train doesn’t run to a place when you need it to, then you can’t go there.
Part-time transit users can often afford to give up comprehensive service, and are often more interested in transit service optimized for specific trips, most often the commute. Many in this camp will use transit when it’s convenient, and not use it otherwise–and may find themselves preferring transit solutions that optimize the most common trips over the less-utilized routes and corridors. To the extent that “choice riders” can easily abandon the bus or train, they may demand amenities and virtues beyond basic service parameters like span, coverage, frequency, speed, or reliability.
The problem? Reconciling these two goals can be difficult.
The paradox of rapid transit
A major source of objection to capital-intensive rapid transit projects is that rapid transit tends to correlate more with the needs of the choice rider. This is not to say that it isn’t usable by the transit-dependent–of course it is. All riders benefit from fast, frequent service. Many capital projects which have been maligned by demagogues (whether on the right, like the Reason editors, or on the left, like the LA Bus Riders Union) in fact provide a great deal of service to the poor. But rapid transit, by its nature, focuses on corridors–corridors which are then optimized for fast, efficient, and reliable service. This is great if you live/work in the corridor. It’s good if you can drive to a park and ride, or take good connecting transit to reach the corridor. It’s bad if you are carless, and the connecting transit is poor or nonexistent–then the corridor is of little benefit to you.
Doing it right
Does this mean that building rapid transit is a mistake? Certainly not. Many large cities, including virtually all of the largest metropolises on the planet, would not function without quality rapid transit. The key, however, is to build rapid transit on top of a an existing, thriving, and functional basic service network–which in the US is typically local bus.
The problem comes when a city builds rapid transit lines, but without a core network of comprehensive basic (typically bus) service–or cuts basic service to fund rapid transit. There are plenty of examples of boondoggle light-rail or streetcar projects being built in US cities with poor bus service. These lines tend only to serve the destinations within walking distance (or park-and-ride users), and generally have poor ridership as a result. The network effects which make transit transformative, simply are not there.
TriMet has avoided this problem, for the most part. The Blue, Red, and Yellow lines were laid on top of an existing, high quality bus network. Bus service leaves something to be desired out in Hillsboro or Gresham, but is good in the core. However, the recent opening of the Green Line was more problematic, While the proximate causes of Portland’s recent service cuts are factors other than MAX–TriMet didn’t plan the Green Line with the intent of sharply reducing the frequent bus network, the fact that its opening coincided with the recession and the subsequent service cuts, did not look good. There’s a significant danger that the opening of PMLR will have similar issues–TriMet isn’t planning on reducing bus service to fund Milwaukie MAX (other than redundant lines like the 33 and the 99), but if the financial picture doesn’t improve, PMLR may be an additional expense that will have repercussions elsewhere.
Additional complications arise.
- Rapid transit lines (unlike freeways) are frequently regarded as a valuable amenity to live near, which can have the perverse long-term effect of pushing the poor away from the transit service they are dependent on. Much of greater Portland’s lower-income population lives outside the inner city. While Portland has built MAX lines into lower-income communities like Rockwood or south of Lents, the strongest concentration of service is in the “core” between the West Hills and I-205, Johnson Creek and Killingsworth–a region which contains some pockets of poverty (particularly in the eastern corners) but is mostly middle-class in nature.
- In times of economic difficulty, the lines which are easiest to cut are the social service lines–as they are the ones which lose the most money. No matter how much an agency cares about equity, it has to balance its books at the end of the day.
- Exacerbating the problem, many grant-funded capital projects include operational commitments as part of their funding conditions. Many proposed delaying the Green Line opening as a way of dealing with the recession, on the grounds that it was mostly redundant with the 72 bus and the existing Red and Blue lines into downtown. The trouble with that proposal was that TriMet had to run it. This is especially a problem when a capital project turns out to be a boondogggle.
For TriMet: I would get the operational house in order prior to any more major capital expansions on rapid transit corridors. (I’m not talking about Portland-Milwaukie, which is already in progress, but things beyond that). There are plenty of ways that capital dollars can be spent (assuming local politicians will continue to have a goal of winning federal grants to goose the local economy) that don’t place all the eggs in one geographic basket.
For the USDOT/FTA/Congress: Funding of capital projects (both urban rapid transit and urban freeways) ought to be conditioned upon having a thriving and useful bus service in place–before laying track or pouring concrete, efficient use of existing infrastructure should be encouraged. Transit projects without good bus service lack the network effects to be successful. The best transit projects are those that are built to relieve overcrowding in an existing bus system, not ones dependent on speculative new riders, even if one suspect that there are large numbers of potential patrons who will ride trains but not bus. And likewise, freeway construction should be discouraged until more efficient utilization of existing streets is made. (Transit project is subject to unfair funding competition with freeways as it is; it would be undesirable for a proposal design to reduce boondoggle projects to have the effect of diverting funds to highways instead).