Parts 1 and 2 of this series looked at the public side of the DC Alternatives Analysis process that took place between 2002 and 2005. Several newsletters were published, public meetings were held, and the study team met with civic groups and maintained a presence at various community events. The widely distributed documents only tell a small fraction of the story, and if one wants to understand why the final report had such disappointing recommendations, one needs to delve into the more technical study documents, which weren’t widely distributed. The contrast between that which was published publicly and the technical documents kept internally is instructive for anyone following a similar engineering study of similar scale.

Broadly speaking, these technical documents attempt to quantify the decision-making process in order that every subsequent decision have justification. The process obscures the study biases by shifting them into the methods of quantification, and ultimately confuses quantifiability with importance.

Setting the Stage

The formal program of the study was documented in the Project Work Plan, in January 2003. One of the first of the study documents was the short Quality Assurance Program, an eight-pager released in November 2003. It establishes the tedious tone in which all further study documents will be written with empty management-speak such as “All DMJM Harris staff performing tasks on the project will utilize the appropriate implementing procedure for the work being performed.”

Two reports were finished in August 2004: The Needs Assessment and the Evaluation Framework. These followed the extensive series of community meetings in late 2003. The Needs Assessment was the only technical document that was published on the (now-defunct) study website. It examined population, employment, and overall destination patterns across the city in relation to existing transit service. The Evaluation Framework brought together all the input–from DC agencies and from the community–about routes and goals and needs and defines what sort of analysis is to be done. A structure of seven routes is proposed, two of which have alternative routings, but the stops along those routes are not defined yet. The project goals are laid out, and the measures and criteria used to evaluate choices in terms of those goals are defined. The general work plan for several documents that follow is laid out.

Route and Mode evaluation:

Screen 1, released September 2004, evaluates seven potential transit modes (streetcars, “bus rapid transit,” light rail transit, diesel multiple units, automated guideway transit, monorail, and heavy rail), and ends up recommending only streetcars and “bus rapid transit” for further evaluation.

The Definition of Alternatives, released in November 2004, analyzed the routes given in the Evaluation Framework for the two chosen modes. Station locations were assigned and propulsion technologies are considered. For each route, a “service plan” was developed, including the headways between successive runs and calculations for route travel times. Although there are separate calculations for streetcars and for “bus rapid transit,” no details are given about the assumptions that went into the calculation of the travel times.

Screen 2, released March 2005, takes the service plans and route structure from the Definition of Alternatives and tries to evaluate how well each would fulfill the project goals by applying a set of “Measures of Effectiveness,” which are defined in the Evaluation Framework. Claiming that “the operational characteristics of BRT and Streetcar are similar at the level of detail” under study, it lumps both into a “premium transit service option” to decide whether a particular corridor should have “premium transit,” or whether it should only receive some bus service enhancements. Corridors were ranked (high, medium or low) based on a few criteria for each of the four project goals, leading to a composite score. Further analysis on meeting corridor deficiencies and operational considerations, and concluded by recommending some routes for “premium transit” and relegating some to get only “local bus enhancements.”

Screen 3, released May 2005, takes the “premium transit” corridors of Screen 2 and applies further “Measures of Effectiveness” to determine whether each corridor should be Streetcars or  ”bus rapid transit.” Each corridor is broken into segements, and the effectiveness criteria are applied to the segments individually. Where applicable–which isn’t as frequently as one might think–Streetcars and “bus rapid transit” are evaluated separately. The scores from these evaluations are totaled, to come up with proposals for streetcar routes, “bus rapid transit” routes, and “rapid bus” routes.

And onwards

Further study documents, released May–September 2005, looked at the finances of the proposed system and put forward the timetable. All the study findings were summarized in a final report, published in October 2005. Future posts in this series will look in detail at some of these technical reports.

Part 1 of this series looked at the beginnings of the DC government’s effort to expand the transit network. We left off in the Spring of 2005, having been to several meetings and having received several newsletters.

The study finishes

The final project newsletter, Fall 2005, and an “Executive Summary“ of the whole project were presented to the public at a final meeting, held September 29, 2005. For transit enthusiasts following the project, the end results were disappointing and frustrating. Instead of a visionary transformation of mobility in the District, the final recommendations proposed a meager streetcar buildout that, despite its modest size, would take 25 years to build. The report was frustrating because it relied on tortured reasoning that bordered on downright dishonesty, it used self-contradictory and mutually inconsistent reasoning, and offered little more than poorly-defined chimeras wrapped up in wishful thinking.

Added to the project was “Rapid Bus,” as a lower-class technology mode, joining streetcars and “bus rapid transit.” Modes were assigned to routes. The newsletter used separate streetcar and “bus rapid transit” assignments, while the executive summary lumped these together as “premium transit.” In the newsletter, streetcars got a handful of routes: the crosstown Georgtown to Minnesota Avenue route; the north-south Georgia Avenue route, which would end at K street; a Union Station to Anacostia via Eastern Market route; an M Street SE/SW route, and a short Bolling AFB–Pennsylvania Ave route. A bit of “bus rapid transit” was added: mainly Woodley Park to Eastern Market via Florida Avenue, while the rest of the 50-mile route structure developed over the course of the study was designated “rapid bus.”

On the face of it, we’re being asked to believe that there is a significant difference between “bus rapid transit” and “rapid bus,” yet no clear description of the features that differentiate “bus rapid transit” from “rapid bus” is given. “Rapid bus” is supposed to have: limited stops, fancier shelters, real-time arrival information, and signal prioritization, as is “bus rapid transit.” They’re both supposed to use “large vehicles,” with those for “rapid bus” being “distinct” and those for “bus rapid transit” supposedly “recalling the design of streetcars.”  Even though “rapid bus” vehicles “could be” 60-foot articulated buses, the largest buses used anywhere in the United States, “rapid bus” vehicles “tend to be smaller” than “bus rapid transit” vehicles. One feature that was mentioned for “bus rapid transit” but not for “rapid bus” was off-vehicle fare payment.

“Bus rapid transit” is said to run “either on mixed traffic or on dedicated rights-of-way,” and several of the “bus rapid transit” examples cited do make use of dedicated rights-of-way. Boston, Ottawa, and Pittsburgh are explicitly called out as examples of “bus rapid transit,” and shows pictures of the system in Rouen, France. Both Ottawa and Pittsburgh rely on massive, exclusive and grade separated rights of way for their systems, built on abandoned railroad rights-of-way. Rouen’s TEOR uses several sections of exclusive right-of-way with optical guidance. No similar opportunity for constructing such facilities exists along the corridors proposed for DC, and surely such rights of way would be as incompatible with neighborhood scale as light rail would be. The documents make no mention of the heavy criticism that Boston’s Silver Line has received. Incidentally, neither Boston nor Pittsburgh use off-vehicle fare payment. I find it a little dishonest to cite systems whose success (such as it is) relies on features that are not under consideration for DC.

The proposed timetable for transit buildout is galling. All lines are supposed to start as “rapid bus.” Of the meager streetcar and “bus rapid transit” network that’s proposed, the newsletter states “Naturally it will take time–two decades, in fact” to get it built. The system would not be finished until 2030, even though the need is immediate and growing, as was illustrated in the Needs Assessment. And the executive summary mentions Mayor Williams’s goal of adding 100,000 residents by 2013 is cited: presumably, we don’t want to make them wait 17 years for adequate transit.

Similarly frustrating is the lumping together of streetcars and “bus rapid transit,” as if it’s possible to create a bus-based system that could be considered equivalent to a rail based system. Although it’s been quite clear in all the public outreach that there are lots of people who are excited about streetcars, there is no evidence that anyone is excited about more buses. This is not to say that everyone is excited about streetcars, but the opposition to streetcars does not come from an enthusiasm for buses. Streetcars offer better ride quality than buses, and the investment in infrastructure required for streetcars gives other investors confidence in the civic commitment. Indeed, while there are dozens of examples of rail-based urban revival, there are no examples of bus-based revival. Rail has shown itself to be far more attractive to riders: whenever a bus service is replaced by rail, ridership grows; whenever rail is replaced by bus, ridership drops. 

If the final output of the study was underwhelming, what happened next was breathtaking. Once the study was over, absolutely nothing happened. The study documents seemed destined to gather dust on someone’s shelf. It did not help that the DDOT Mass Transit Administrator, Alex Eckmann, left in February 2005, before many of the reports would be released and decisions would be made. The position would remain vacant until the study was nearly complete. The DDOT director who pushed to get the study underway, Dan Tangherlini, left DDOT in February 2006 to become interim general manager of WMATA. Because this was near the end of then-mayor Williams’s term, DDOT was left in the hands of the deputy director; no high-profile appointment was made. This left a leadership void from which no push to move forward with the transit project was made.

So how did the study team arrive at its conclusions? Part 3 of this series will examine the detailed technical documents produced by the study team.

Prologue

Bringing a 50-mile streetcar network to Washington DC is the top priority for the DC Chapter of the Sierra Club. I have been following this issue with the Sierra Club since 2002, and it was recently suggested to me that I write down a brief history of the effort, to provide context for those new to the subject. Current progress on the issue is blogged at streetcars4dc.org. 

The DC Transit Improvements Alternatives Analysis gets underway.

The last time streetcars ran in DC was the early morning of January 28th, 1962, after which all lines were converted to buses. Such was the state of public transit in the District until March 27, 1976, when Metrorail opened. Metrorail, of course, has been a tremendous success, but it does not serve all areas of DC, and was designed primarily to move suburban commuters to their jobs in downtown DC.

The District government has, in principle, been planning to bring streetcars back to DC for some time now. My involvement began in September 2002, when I testified on behalf of the  at a joint oversight hearing of the DC City Council. A relatively small, two-year study had recently been completed (DC Transit Development Study), and then-DDOT director Dan Tangherlini, and then-DDOT Mass Transit Administrator Alex Eckmann went before the council (read their presentation) to ask that a more expansive study be funded. Plans to expand transit in the District stretch back further than that, and are generally said to have begun with the Barry-era DC Vision Study of 1997, itself 2 years in the making. And after more than ten years of talk and study, there are still no streetcars.

DC Vision had proposed four light rail corridors and a new crosstown Metrorail line. The Transit Development Study had looked at nine potential routes, and from that, DDOT proposed what amounts to five corridors for further study. Three were crosstown routes, converging in the east at the Minnesota Avenue Metro and crossing the river together. Two of these followed a northern route, heading west to Union station, from which one would go northwest to Woodley Park and another would head to Georgetown. The third crosstown route headed south to the Navy Yard, and continued to Waterfront. A north-south route followed Georgia Avenue/7th Street from Silver Spring to Waterfront, and an east-of-the-river route that went from Minnesota Avenue, to Anacostia, and finally to National Harbor.

The study that did go forward, formally the DC Transit Improvements Alternatives Analysis, was indeed quite extensive. A joint effort between DDOT and WMATA, most of the actual work of the study was done by transportation engineering firm DMJM+Harris. Numerous public meetings were held, fancy full-color newsletters and brochures were produced, and scores of meetings with community groups were held, including one with the Sierra Club in October 2003. Tangherlini and Eckmann both spoke enthusiastically about the plans for the 2002 Rail-Volution conference, which was held in DC that year. 

Throughout 2003, a series of semi-public meetings were held: nobody would be excluded, but they were advertised primarily to civic leaders. In July I made it to the meeting in the Mt. Pleasant neighborhood, at which, it turned out, I was the only attendee to show up. From a public outreach standpoint, this is not so good, but for a transit wonk like myself, a detailed, hour-long conversation with the project leaders was something of a treat. In the fall, the project released the first of what would be five project newsletters (Fall 03, Spring 04, Fall 04, Spring 05, Fall 05).

By Spring 2004, the list of routes under consideration had grown, and in addition, the route lines were drawn as narrow lines following actual streets, instead of broad bands indicating a general route. A route up Wisconsin Avenue was added, as was a Brookland-Columbia Heights-Woodley Park crosstown route. Alternate alignments for the southern route from Woodley Park were added: one that headed to New York Avenue and then went south to Eastern Market, and the other going to Union Station and snaking across the Capitol to Eastern Market. By the Fall of 2004, a Foggy Bottom-Waterfront route, hugging the Potomac, had been added, a a spur East of the River was added. The map presented in the Spring of 2005 backed off of showing lines along particular streets to re-work the routes into nine “proposed corridors.”

The newsletters, and other project documents, were available on a well-publicized, but regretfully, now defunct website called dctransitfuture.com (archived on the Internet Wayback Machine), and many of them are now on DDOT’s website. In conjunction with the public meetings and widely distributed newsletters, some substantial analysis analysis of transportation services and demand was being conducted, which culminated in the Needs Assessment, released in June 2004. Not surprisingly, it found a great need for expanded transit in the city. One of the most telling parts (Fig. 6, p. 17) shows a map illustrating the density of households with no car available: this quite clearly shows the need for high-capacity transit along Georgia Avenue and along H Street NE. Nearly all of the corridors from the 2004 newsletters were identified in the Needs Assessment as “recommended priority transit corridors” (Fig. 20, p. 43).

The project team met with the Sierra Club for the second time in June 2004. I recall asking if there was some thing we could focus on that could help move the project forward; the project team demurred, assuring us they had everything under control. The Fall 2004 newsletter announced that the technologies chosen for further study were streetcars and “bus rapid transit.” Through spring 2005, all the route maps presented in every facet of the study treated all the routes equally–they were drawn with the same thickness of lines with no distinguishing features–and no association of routes with technology had been yet indicated. The graphics had emphasized streetcars, modern and historic. Small features in the newsletters highlight the benefits of bringing a streetcar–and not a bus–to a specific neighborhood. 

The point made in Fall 2004 about streetcars was first that they were not considering more Metrorail, but also that it would not be the type of light rail one sees in Sacramento or Dallas or Denver. That is, it would not get dedicated rights-of-way, and would probably not be designed for long, multi-car trains. This makes sense: in these cities, the light rail is the only rail transportation and performs both suburban commuter and urban mobility functions. In DC, the transit is to be overlaid on top of Metrorail, which performs the suburban commuter function quite well.

The Fall 2004 discussion of “bus rapid transit” uses the subtle dishonesty which I’ve found to be typical of bus advocacy: it’s said to “combine the best of rail and bus,” begging the question of whether buses can ever provide service equivalent to rail (to which I answer “no”). Also mentioned is the fact that buses running on dedicated rights-of-way can be much faster than ordinary buses, which is a red herring: it’s already apparent that a dedicated right-of-way, for any sort of vehicle, is not being considered.

At any rate, those following the study at this point had reason to be optimistic.

Part 2 of this series will look at the conclusion of the study, in the second half of 2005. Part 3 summarizes the technical documents produced as part of the study. Further parts will look at the Anacostia Starter line and the issue of overhead wires.