Monthly Archives: December 2017

What I learned in 3 months of obsessively tracking my WMATA trips

A few months ago, I was curious how long it would take me to ride in the same Metro car. Turns out, the answer was about a week.

I’ve since continued to log each and every Metro trip, and after three months of doing so, I have a few observations.

First, some overall stats: in that time, I’ve taken 220 unlinked trips, riding in 195 different railcars. I’ve taken the same car for two trips 23 times. The most rides for a single car is 3 – I’ve somehow managed to ride in car 3245 three times.

Assuming the sequential numbering of cars is consistent with their actual time in service, the oldest car I’ve ridden is the oldest one still in service: 2000. The newest car I’ve ridden is 7438 (WMATA’s recently hinted that cars up to 7456 have been delivered).

All together, in three months I’ve ridden about 16% of the fleet.

B8707DD9-B30A-44CE-920D-C8DCD8E55851

24B9969F-48AD-4537-8CF0-5E98CE5D63D7

For the actual rides, about ⅓ have been on the newer 7000 series, ⅓ on the combined fleet of 2000 and 3000 series cars, and the remainder split among the 5000 and 6000 series.

4D850F88-8FC2-4D8E-B8C4-1437B00883A5
Some observations:

WMATA’s fleet isn’t evenly distributed across the system, nor are my trips.

My trips are concentrated along my regular commuting route, which mostly sticks to the DC trunk line shared by the Orange, Blue, and Silver line trains.

05B58C52-EB80-427D-84BB-467B0670615C

Among just my trips on OR/SV/BL, however, the distribution of trains isn’t exactly even (even if that’s what the schedule suggests). The death of the Blue Line has been greatly exaggerated!

DC0BBC80-1E6E-48C2-8DF9-0BEA93299EC7

Since these are my most heavily used lines, it helps to show how the trains are unevenly distributed through the system. I seldom rode in older, 2000 series trains on the Orange line, despite lots of those trains on the Blue line.

I’d hypothesize that the 2000-series are based in a particular rail yard (probably the Alexandria yard). Likewise, the relative lack of 6000 series trains on OR/SV/BL is probably because the 6000 series are based out of Greenbelt (it’s a small sample, but 4 out of 7 Green Line trips were on 6000 cars; the remainder were on 7000 series trains).

I still have a long way to go to catch Matt Johnson.

Getting my kicks on I-66

The Virginia Department of Transportation started tolling Interstate 66 this week, and boy, people are pissed.

A few thoughts after two weeks of dynamic tolling:

Lots of people were cheating the old HOV rules: The shock over paying the new tolls (which wouldn’t apply for those who were driving the road in compliance with the HOV rules) shows how many people had been cheating the system.

Some drivers feel wronged by changing policies, such as the increased HOV hours and the loss of the HOV exemption for hybrid vehicles, but those folks clearly don’t account for the huge portion of cars driving the tolls up.

Faster may not be the most efficient: VDOT was boasting that the average morning rush-hour speed was 57 mph. The speed limit is just 55 mph; and the old average was something like 37 mph.

The $40 toll isn’t the cost of reducing congestion; instead, it shows the marginal cost of keeping an urban expressway flowing at rural traffic densities, enabling free-flow conditions at the speed limit. It’s a useful reminder for all drivers of just how expensive it is to ensure the kind of speeds too many of them expect.

It’s not clear at all that this is the optimal policy for VDOT to set. They haven’t released any information about changes in traffic volume. The law requires maintaining average speeds of at least 45 mph; if VDOT were to accept lower speeds in the morning without introducing congestion, they might avoid some of the more exorbitant tolls and allow for more drivers to use the road.

There’s precedent for this: Minnesota experimented with a complete shutdown of their freeway ramp meters in 2000. The end result was a determination that they could dramatically improve the system’s user experience without sacrificing the main benefits. But, it took a legislatively mandated shutdown experiment to get MnDOT to make the changes in policy.

Clear communications matter: This isn’t Virginia’s first HOT lane project, nor is it the first time tolls have spiked (though most previous events could be traced to some kind of incident – poor weather, a bad crash, etc). So what explains the backlash?

A big part of the problem appears to be a misunderstanding about the toll rates. During the approval process, lots of folks had a $17 round-trip figure in their head. That’s obviously a lot less than $40 for one-way.

Fredrick Kunkle dug into this in the Washington Post:

“The bottom line is this is very different from what we briefed people it would be,” Del. John J. Bell (D-Loudoun), an opponent of tolling on I-66, told my colleague Luz Lazo.

Others have been blunter in saying the McAuliffe administration misled people. The Republican Party of Virginia accused McAuliffe’s administration of ensuring that the tolls would be switched on only after the gubernatorial election to choose his successor. Loudoun County Supervisor Ron Meyer (R-Broad Run), who is also a member of the Northern Virginia Transportation Commission (NVTC), urged the NVTC to pass a resolution demanding that the tolls be lowered or suspended.

The defense seems like an honest response, but it might as well be included in a reprinting of How to Lie with Statistics.

But Brian Coy, a spokesman for the governor, said the administration never misled anyone. He said that when transportation officials talked about a $17 average daily toll during peak hours, they meant what they said, an average — all short trips and long trips along that section of highway, and with peaks and valleys of demand.

So, that single number was averaging both the dynamic toll rates as well as the different potential routings. An average of an average. But most drives don’t take an average route, they take a specific one. And since the initial communications didn’t discuss a per-mile rate (or an estimated range of rates), it’s not hard to see why people might feel surprised.

Clear communications matter because policies like this have real promise. They depend on political support, and that will be harder and harder to find if people think they’ve been deceived. Tolling, particularly when perceived as a solution to congestion, can be a political winner. Whether the $17 expectation was intentionally misleading or not is beside the point; those expectations have to be managed or a program like this could lose support.

The challenges of adding housing in single family neighborhoods

Too often, news articles on housing prices fall into easy traps and cliché, whether in discussing gentrification or city vs. suburb tropes. But Conor Daugherty’s piece in the New York Times (The Great American Single Family Home Problem) hits all the right notes.

In it, he tells the tale of a modest redevelopment proposal to redevelop a single family home into three units on the same lot. The political opposition is fierce, leading to years of delay and legal proceedings. And this is for a parcel already zoned for additional density; this particular saga doesn’t even touch on the challenges of rezoning an area currently occupied by single-family homes.

A couple of things stand out to me:

The missing middle: The author frames the cost trade offs well. Lots of cities allow downtown and highrise development, but this requires expensive construction techniques, and thus requires pricey rents to pencil out. Smaller-scale development (low-rise apartments, duplexes, townhomes, etc) can pencil at much lower prices – the thorny issue is the politics of building in existing single-family neighborhoods.

The problem is that smaller and generally more affordable quarters like duplexes and small apartment buildings, where young families get their start, are being built at a slower rate. Such projects hold vast potential to provide lots of housing — and reduce sprawl — by adding density to the rings of neighborhoods that sit close to job centers but remain dominated by larger lots and single-family homes.

Neighborhoods in which single-family homes make up 90 percent of the housing stock account for a little over half the land mass in both the Bay Area and Los Angeles metropolitan areas, according to Issi Romem, BuildZoom’s chief economist. There are similar or higher percentages in virtually every American city, making these neighborhoods an obvious place to tackle the affordable-housing problem.

“Single-family neighborhoods are where the opportunity is, but building there is taboo,” Mr. Romem said. As long as single-family-homeowners are loath to add more housing on their blocks, he said, the economic logic will always be undone by local politics.

Capital-A Affordable, vs. affordable: The three units proposed for the lot wouldn’t be cheap, but (crucially) they’d be cheaper than a re-habbed SFH on the same lot – and there’d be more of them.

They are estimated to sell for around $1 million. But this is an illustration of the economist’s argument that more housing will lower prices. The cost of a rehabilitated single-family home in the area — which is what many of the neighbors preferred to see on the lot — runs to $1.4 million or more.

The “economist’s argument” might be sound, but it’s a hard sell for the neighbors.

This kind of evolutionary redevelopment would’ve been completely natural and non controversial before the advent of zoning.

It’s always worth remembering how different the Bay Area’s housing market dynamics are. Daniel Kay Hertz notes that many of the same issues are in play in weaker regional markets, though the way things play out is quite different:

Aaron Renn doesn’t think much of the Bay Area’s strategy of generating affordability through redevelopment of single-family housing:

https://twitter.com/urbanophile/status/937159416847175680

First, it’s hard to say this is a cogent strategy; the vast majority of single family homes aren’t going to be rezoned anytime soon.

Second, Renn is correct, historically – at least since the advent of zoning. This was true for the Bay Area, too – suburban development offered a then-cheap and cost-effective way to add housing to the region’s supply. But that was decades ago (the NYT article includes maps showing the expansion of the suburbs over the recent decades), and the region has run out of room for new/expanded suburbs within a reasonable commuting distance.

Renn’s implied regional strategy isn’t going to work well in the Bay Area, either. Consider the recent articles on Bay Area super commuters. Relying on Stockton to be San Franscisco’s bedroom community has severe costs, after all.