Posts by Yonah Freemark

Perhaps unsurprisingly, the idea was perceived as heresy, both by Dan Malouff and MWAA board members. Mame Reiley, one board member, said, "I just don't think that's what we labored for... it is not rail to Dulles."

Board members raised concerns that the federal government might delay the program because the board was "starting over." And indeed the proposal appears to have been dismissed by the authority board as unacceptable.

But such a change could be a reasonable money-saver and may actually improve transit service for both commuters and air travelers. The question is immediately relevant to the Dulles Rail extension, but also equally valid to many cities, as the issue of extending rail networks out towards airports is frequently of concern for transportation planners in major metropolitan areas.

The question of how to reach Dulles by rail has been fraught with controversy since project development began. Originally, the concept was to connect the Metro line to an underground station about 550 feet from the main terminal, but after the project's price tag had exploded past $3 billion, cost-savings became necessary.

The MWAA, which runs Dulles Airport in addition to the Metro extension, eventually agreed in July 2011 to move the stop about 600 feet farther away and to elevate it above the ground. Riders wanting to get off at Dulles will have to make the more than thousand-foot walk from the station to check-in.

Brown's likely stillborn proposal to replace the direct rail link with a people mover reflects the fact that riders are likely to see this connection as inconvenient, especially compared with that at Reagan National Airport, where customers only have to walk about 150 feet between Metro platform and the terminal entrance.

Brown suggested rerouting the Metro line away from the airport (the existing plan is shown in orange below and would be about 4 miles from Route 28 to Route 606), so that it runs directly along the Dulles Greenway (in blue, about 2.5 miles from Route 28 to Route 606). A people mover (also in blue, about 1.5 miles) would connect the Route 28 station to the front of the terminal.

Though customers would have to transfer, they would now get a more direct journey, since it would be far easier to fit in front of the terminal the tracks and station for the people mover than it would have been for the Metro line (and in fact this explains why that latter possibility was never brought up).

This would save a total of $70 million, according to planner estimates, because it would replace about 1.5 miles of very expensive Metro infrastructure (readied for eight-car trains) with much lighter automatic people mover infrastructure, designed for one- or two-car trains.

We know this would save some money. How would this change affect customers?

Riders commuting in to Tyson's Corner, Arlington, or Washington from outer suburban destinations on the end of the rail line west of Dulles would save time: At the 35-mph average speed expected for Silver Line trains,* it will take about 6.9 minutes to get from Route 28 to Route 606 using the current plan. The more direct route proposed by Brown would reduce that journey to 4.3 minutes. That's almost half an hour in saved travel time per week per commuter.

Even better, those using the Silver Line to get to and from the airport might actually save time travelling too.** Though these customers would have to transfer between Dulles Metro and the people mover, if that connection were timed and across the platform (as is quite possible when two automated systems are linked and built at the same time), the time lost would be only two or three minutes.

Meanwhile, once they actually get off at the terminal, the experience of riders taking the people mover would be much superior: Rather than walking 1,150 feet to the terminal, which would take them about 4.8 minutes on average, they would walk something more like 150 feet, which would take them only 0.6 minutes.*** See this back of the envelope comparison:

	Arrive at Rt 28 station	Timed transfer to people mover	Time to Dulles Airport station	Walk to terminal	Total travel time
Existing proposal	0 min	--	2.5 min	4.8 min (or about 3 min with moving walkway)	5.5-7.3 min
People mover proposal	0 min	3 min	2.5 min	0.6 min	6.1 min

Though the use of the people mover raises questions about operating another rail system, it could be maintained with similar vehicles as those already servicing Dulles on the Aerotrain, which connects checked-in passengers to the terminals.

The Washington region would not be alone if it chose to make its airport rail link stop somewhat short of the terminal itself. In Phoenix, the new light rail system was built in coordination with airport officials, who are currently constructing an automated train between the rail station and the terminals. The San Francisco Bay Area is building an airport connector to the Oakland Airport that will link a BART station some miles away to the terminals.

And Miami's new AirportLink Metro Rail project will not actually stop at the airport, but instead at a new central station with transfers to a people mover.

Riders in these regions will not suffer; they may lose a few minutes transferring between trains, but if the connection is short and timed, that pain can be minimized. Avoiding the airport, paradoxically enough, could both save money and improve the situation for riders.

* 35 mph: PlanItMetro projects it will take about 22 minutes to travel the 12.8 miles between Dulles Airport and Tysons 7 Station.

** The only customers would would lose out with this change would be those traveling to and from Dulles from outer-suburban locations.

*** Assuming that people with bags travel at about 4 feet/second, a bit slower than the average walking speed of an elderly person.

Cross-posted at The Transport Politic.

All of this, I know, is obvious. But when it comes to rail transportation, this formula has been avoided, especially in the U.S. The owner of railroad tracks usually is also the operator of trains along them. When other operators want to move their own trains in, conflicts typically erupt.

The frequent disagreements about acceptable service levels between national rail operator Amtrak and freight railroads on tracks that the latter owns (and which it isn't very happy to share) are indicative of this problem. But these disagreements are not irreconcilable.

Indeed, an infrastructure owner that is able to arbitrate between competing operators could be more effective in producing efficient service for everyone than might be an owner-operator, which discriminates against other operators.

In this context, last week's revealing of House Transportation and Infrastructure Committee Chairman John Mica's (R-FL) plan for the Northeast Corridor raises a number of interesting questions. Convinced of the value of private sector competition and promoting a pull-out of the federal government from every public service imaginable, Mr. Mica has submitted a proposal that attempts to re-imagine the Northeast Corridor, Amtrak's flagship route and the nation's most-traveled intercity rail line, as a place where, fundamentally, the rules of the road—but not the railroad—could apply.

The bill (draft text) would force Amtrak to abandon its control of (much of) the Northeast Corridor between Washington and Boston, handing it over to the Department of Transportation, which in turn would lease it to an "Executive Committee."

Amtrak would have to give up all of its assets and it would lose federal funding. The Committee, in charge of infrastructure and setting pricing policies, would then engage a public-private partnership (PPP) with a private group, which would commit to upgrading the line and then operating trains to offer two-hour trips between New York and Washington and 2h30 between New York and Boston—within ten years, twenty years more quickly than Amtrak has said it would be able to make roughly the same improvements.

Mr. Mica also claims that this could be done at a cheaper price than Amtrak's $117 billion proposal.

Outside of the Northeast, states would have to offer their rail corridors to competitive bidding; current subsidies to Amtrak would simply be redistributed to the winners of those operations bids.

Despite the wide-ranging proposed effects of the bill as summarized, the manner in which any of this would be implemented remains incredibly unclear. How would intercity rail operators interact with the freight and commuter railroads that also use the tracks, in the Northeast and elsewhere? If a PPP were implemented, how much would the government agree to commit to pay for improvements?

Unfortunately, the bill would not provide a realistic way to promote true operational competition. Nor does it would it offer a promise of actual federal support to fund an upgrade of the corridor, which seems unlikely to be sponsored by private entities alone. Most problematic would be the transfer of authority over the line's management to the currently non-existant Executive Committee, whose ability to make decisions about rail properties has yet to be tested, let alone proven.

Fortunately, the proposal is unlikely to make it through the Senate, where Democrats and other Republican supporters of Amtrak are likely to prevent the bill from passing even if it makes it through the House. The American intercity rail system and the governance bodies that oversee it at the federal and state levels are too underdeveloped to be able to guarantee that this semi-privatization wouldn't be a disaster.

But Mr. Mica's bill does articulate a number of policy changes that could play an important role in shoring up passenger service in the Northeast. The status quo, in which Amtrak operates relatively infrequent and slow passenger trains within the nation's most important megaregion, certainly is not ideal. If managed appropriately, the separation of track ownership and line operations could allow for a situation in which multiple operators offer competing services along the same routes, just as Megabus and Bolt Bus compete for the most customers on I-95.

In mainland Europe, E.U. regulations have mandated that national rail companies like France's SNCF or Germany's DB allow other operators (in many cases, SNCF and DB affiliates) to run trains between similar destinations. Though I am not convinced that this will produce universally positive results, it will at least likely result in lower fares for customers on the most heavily trafficked rail corridors.

And focusing on the most-used lines is clearly Mr. Mica's goal; according to the bill, the second-highest stated priority for potential investors are "activities that benefit the greatest number of passengers" (just after safety). Amtrak's current policies do not exactly fit that bill since they are designed to push lower-income individuals (like myself) onto slower and less comfortable intercity buses.

Yet the Mica proposal would not produce true competition in rail operations. It would encourage competition in rail operations contracts. Rather than invest in the infrastructure and then open up the rights to use tracks, the PPP structure as proposed would be a build-operate-maintain system in which one private group would invest in improvements and then have control over operations, which it would perform itself.

Mr. Mica has repeatedly referred to Amtrak as a "Soviet-Style" system because it has a monopoly over its services, but it is hard to see how a PPP extended over a long contract would be any different, except that it would charge even higher prices to make up for the initial cost of capital improvements and—even worse—it would be literally banned from cross-subsidizing other services with the profits, according to the proposed bill. Is this in the public interest?

The biggest question of all, though, is whether Mr. Mica is in complete denial about the extent of either the private sector's ingenuity or their collective willingness to invest in public infrastructure. While it may sound nice, asserting that corporations can rebuild the Northeast Corridor in 10 years at a far lower cost to the taxpayers than Amtrak has proposed could is a stretch. And even a $50 billion upgrade would be larger than any single private investment in infrastructure ever in the U.S. What evidence does Mr. Mica have that a plan like this could move forward?

Cross-posted on The Transport Politic.

Thus Washington's network is relatively new: Extensions continue to open every few years; a major new line running to and beyond Dulles Airport, in fact, is in construction.

This means that many of the changes that have been hypothesized to accompany heavy rail service, like densification, may not have yet appeared. Nonetheless, in some places, such as along the Rosslyn-Ballston Corridor in Arlington County, Virginia, significant urban redevelopment has occurred. Similarly, in cities like Charlotte, Denver, and Minneapolis, major new construction has begun after the completion of light rail lines.

Just how widespread are these effects? Have similar changes happened everywhere where new Metro stations have opened in the Washington region?

To examine this question, I have delved into recently released Census 2010 data to consider what has changed since 2000. By considering the alterations in development patterns near stations that opened about ten years ago, we can better understand what has occurred.

On first evaluation, there is no clear connection between the opening of a new station and increased construction—at least on a ten-year timeline.

Between 1997 and 2001, nine Metro stations opened, two of which were in the heart of the city on the Green Line (Columbia Heights and Georgia Avenue) and the rest of which were at the termini of the Red (Glenmont), Blue (Franconia-Springfield), and Green Lines (Congress Heights, Southern Avenue, Naylor Road, Suitland, and Branch Avenue).

Compared to their host jurisdictions, only three of the nine stations saw higher growth in adjacent Census tracts: Columbia Heights, Franconia-Springfield, and Branch Avenue. In the areas around these stations, densification was significant, promoting the theory that transit can be an effective tool for urban regeneration and growth.

These changes were particularly interesting at Columbia Heights, where an already pretty dense neighborhood only became more so thanks to rapid replacement of low-lying building stock with taller buildings. Around the other two stops, largely vacant land was replaced with new construction.

Around two other stations—Georgia Avenue and Glenmont—growth was also positive, but it was slower than in Washington and Montgomery County, respectively.

Finally, four of the studied stations saw a decrease in population in the surrounding Census tracts. Each station is on the southeastern branch of the Green Line, which runs through arguably the region's weakest area from an economic perspective. The presence of transit did not appear to be of any help here: Though Washington and Prince George's County saw population growth between 2000 and 2010, the specific neighborhoods around these stations did not.

Changes appear to be quite context-dependent. The population of the area around the Columbia Heights station expanded significantly, likely not only because of the presence of Metro, but also because of a growing interest nationwide in living in urban cores. On the other hand, the poor attractiveness of Prince George's County, just east of the District of Columbia, likely reduced developer interest in building around stations there.

This analysis indicates that the presence of a transit station cannot provide alone for the kind of urban redevelopment planners often hope to produce when they allocate funds to new rail lines. This does not mean that the opening of the new Metro stations was not an important element of regional growth in Washington, but rather that that infrastructure in itself is not enough to encourage developer interest. In the case of many of these stations, land was not available, zoning was not free enough, and the neighborhoods were not attractive enough to see substantial change, at least over the past ten years.

Transit systems like the Washington Metro are very expensive to construct, so public authorities must make a greater effort to coordinate planning efforts to allow for the creation of more transit-oriented districts to take advantage of such investments.

I would like to note several important caveats: The use of Census tract data in this analysis was meant to provide a neighborhood-level glimpse into development changes. Residents (or potential residents) are likely to see Metro stations as assets, even if their homes are not in immediate proximity. Yet development changes are likely to be unusually affected by that proximity: It may be useful to reconsider these questions at the block level. It is possible, for instance, that the areas directly adjacent to the southeast Green Line stations did see growth, even when surrounding neighborhoods did not.

Opening Day	Place/Station (# Census tracts)	Pop 2000	Pop 2010	Density 2010	Change in Pop	Change in Pop	Change in Hsg. Units
2001-01-13	Congress Heights (3)	11,964	11,221	6,080.85	-743.00	-6.21%	-5.35%
2001-01-13	Southern Ave (3)	12,826	11,730	6,624.12	-1,096.00	-8.55%	-2.25%
2001-01-13	Naylor Rd (6)	22,775	22,262	5,114.41	-513.00	-2.25%	0.31%
2001-01-13	Suitland (4)	17,272	16,833	3,788.74	-439.00	-2.54%	-3.51%
2001-01-13	Branch Ave (1)	3,425	4,696	2,582.80	1,271.00	37.11%	83.13%
1999-09-18	Georgia Ave/ Petworth (5)	20,490	21,351	25,104.06	861.00	4.20%	6.70%
1999-09-18	Columbia Heights (4)	16,434	17,646	44,015.96	1,212.00	7.37%	19.99%
1998-07-25	Glenmont (6)	26,866	28,678	4,606.17	1,812.00	6.74%	1.52%
1997-06-29	Franconia/ Springfield (3)	11,443	13,293	3,772.78	1,850.00	16.17%	16.15%
	Montgomery County	873,374	971,777	1,978.15	98,403.00	11.27%	12.33%
	Prince George's County	801,476	863,420	1,788.76	61,944.00	7.73%	8.54%
	Fairfax County	969,840	1,081,726	2,766.78	111,886.00	11.54%	13.51%
	DC	572,059	601,723	9,856.49	29,664.00	5.19%	7.96%

Cross-posted at The Transport Politic.

What would be the manifestations of these different approaches? How can we weigh the advantages and disadvantages of upzoning the center city for more office space? Is our goal to produce vital, walkable, and dense downtown districts, or simply to expand new construction there, no matter the use?

The missing ingredient in this discussion is transportation. When we discuss the demand in downtowns like Washington's for more office space, we sometimes make an assumption that the transport network will be able to handle whatever is thrown at it.

In fact, there is a direct relationship between a downtown's growth and the transportation provided to it. In general, businesses want to locate their offices in places that are accessible and that provide the benefits of agglomeration, and this sometimes means downtown, but not always.

If the trip to and from the center—by whatever mode—becomes too arduous, there are significant reasons to locate outside of it. How does this fact apply to a place like Washington?

Once a downtown—which I will define as a traditional single-use American CBD—reaches a certain size, once it provides employment for a certain number of people, it has three basic options:

One, it can do nothing to its transportation network, in which case the downtown has no capacity to absorb increasing growth. In these cases, residential uses become more important since the relative land values demanded for office space decrease (as it is harder for more people to enter into the downtown from elsewhere and there is more interest in walking to and from work).

This is arguably what has happened to places like Chicago's West and South Loop, where almost all recent development there has been in the form of residential towers despite the close proximity to the downtown core.

Two, it can expand or improve transportation through the highway network, in which case parking lots become increasingly valuable and may displace existing buildings. This was the choice cities like Houston took since 1950, sacrificing what had once been walkable neighborhoods for an automobile-dominated core.

Three, it can expand or improve transportation through the transit network (bus and/or rail), in which case higher densities become increasingly valuable, and taller buildings may replace shorter ones or parking lots. This has happened in DC since the construction of Metro, beginning in the 1970s.

The discussion in Washington has hinged around the opposite side of the conversation, focusing on land use instead of transportation. The argument, asserted by people like Stephen Smith, suggests that the problem is that the government is exerting inappropriate control over densities by limiting heights and the result is that rents in the office core are increasing far higher than they would were there to be skyscrapers.

The problem is compounded by the fact that downtown Washington's growth is limited, notes Ryan Avent, by the fact that outlying neighborhoods are stuck to one- or two-story buildings (and there is little push to challenge that condition), so the Paris approach, in which the entire city is made up of 6 to 10 story buildings, is not much of an alternative, either.

These arguments are compelling: mini-downtowns in the suburbs, such as along Arlington's Rosslyn-Ballston corridor, can absorb some of the growth, but there is clearly strong demand for continued concentration in the center city.

Whether this is a long-term phenomenon, however, depends on the transportation provided into the downtown. Imagine that the height limits in Washington were lifted—or, at least, buildings twice as high could be built. In the short-term, this would surely produce the desired effects, allowing downtown to absorb more of the region's job growth, reduce office rents, and aiding in the continued gentrification of the city as a whole.

In the longer-term, however, as the city's downtown building stock is gradually replaced, the worker density in the center of the city would roughly double. Would this be sustainable?

If the city's transportation network remains as it is, mostly relying on the existing Metro network and a functioning, if not great, bus system, this would cause significant problems.

Here's why: Much of the Metro system is already at capacity during peak hours. In essence, today's transportation network is designed with a capacity roughly equivalent to what is generated under the current height limit.

Moreover, road expansion is simply not an option, not only because there is no room for new highways into downtown but also because, as already stated, a focus on roads-based transportation encourages downtowns to be transformed into automobile-based neighborhoods.

As the transit system becomes more congested, because of job expansion and a lack of transportation improvements, the cost of transportation into the core—in terms of time and money—will increase. This will reduce the appeal of locating offices downtown and encourage new construction to be residential rather than office-based. Is this desirable for Washington? Does the city want a mixed-use core or a office-based one?

The alternative is allowing an increase in zoning along with an improvement in the transportation network. This may seem obvious, but Washington has not yet committed the funds to an expansion of the Metro network or serious improvements to the bus corridors, putting in question the viability of a lifting of the height limits. The downtown's growth must be approached by considering transportation and land use in complement with one another.

Cross-posted on The Transport Politic.