High-Speed Rail in France

TGV as our train departs the Gare de Paris-l'Est, 1 September 2014

TGV as our train departs the Gare de Paris-l’Est, 1 September 2014

Coming from the land of Amtrak – even in the Northeast Corridor, where at least intercity rail more or less exists – to ride the TGV in France was an absolute delight and an experience utterly unlike any I’ve ever had before in the United States. Rail is affordable, efficient, fast, and omnipresent. One leg of our journey required a bit of a detour; even then we rode four trains in six hours without a single hitch.

Continue reading


Observations: Paris

Place Saint-André-des-Arts, Paris

Place Saint-André-des-Arts, Paris

Paris is, of course, a magical city. We stayed in a lovely but somewhat out-of-the-way hotel in the XVIIe Arrondissment, near the Place du Maréchal Juin and Line 3 of the Metro. Connected to the Pereire stop was an RER station, Pereire-Levallois on the RER C. More on the RER later. The food was also spectacular, naturally.

We ended up buying a three-day “Paris Visite” card, with unlimited rides on buses, trams, and trains in zones 1-3 for three days. The cost of this? €26.50 each, or roughly $34. The card itself is quite tiny, and I’m surprised more Parisians don’t lose theirs. But on y va.

The Metro was everything one could want: fast, frequent, relatively uncrowded, well-lit and clean. Seriously, though, frequencies made this system so much more usable than many others. Headways across most lines were generally around six minutes – even on a Sunday – and I think the longest wait I saw was 8 minutes. The longest wait we ever had was 7 minutes, and that was after transferring and just missing the connecting train. Overall, waiting for a train was an unusual and brief activity. Continue reading

Streetcars: “Good” Enough?

Lots and lots and lots and lots of discussion sparked today by David Alpert’s CityLab article, “Hey, Streetcar Critics: Stop Making ‘Perfect’ the Enemy of ‘Good’.” I’d Storify it all if I had the time but the conversations went down so many branches there isn’t really a linear thread. At any rate, I suggested that maybe we should all just sort of summarize our arguments. So here’s why I disagree with David, in long-winded rambling form.

Put simply, as configured, streetcars – here using the definition of mixed-traffic single or articulated railcars operating on the surface – provide such a miniscule upgrade to the actual quality of transit that their construction must absolutely be questioned by even the most diehard of transit advocates (and I would hope not to be mistaken for anything but). And to my mind that is for basically one reason alone: a lack of dedicated lanes.

Without those dedicated lanes, you can’t have consistency of service. Streetcars in mixed traffic are constrained not just by the flow of traffic, fast or slow as it may be, but by obstructions: predominantly cars in the tracks, but also cyclists, people in the street, et cetera. Any delays then have knock-on effects – you can’t easily run an extra streetcar or two to try and make up the schedule, as it will just get caught in the same traffic or end up in the same delay.

Now, the argument is often made that just because a streetcar is built as mixed-traffic doesn’t mean it cannot be converted to a dedicated right-of-way later in the future, after some nebulous combination of political willpower and rider outrage and ___??? come together to justify it. That in itself is an all too-rare occasion, and even worse is that the way many modern streetcars are constructed means that many of them can never be given dedicated lanes. The H Street streetcar in DC is a prime example of that. By putting in two curb-running lanes, the District Department of Transportation has ensured for all time that this streetcar line will be held up by traffic jams, turning vehicles, and parked cars themselves, because there’s no way to isolate the streetcars without completely blocking off the parking lanes (not a bad thing in my eyes, but extremely unlikely).

The curb-running H Street Streetcar, beholden to the mercy of traffic for all time

The curb-running H Street Streetcar, beholden to the mercy of traffic for all time

And those are just the physical limitations. Streetcars are the entry point to transit for many smaller cities, and their respective agencies seem to treat the lowly streetcar as some kind of high-capacity commuter rail, with scheduled headways of 15 minutes or more. This is often justified by the increase in capacity over a bus, meaning that fewer streetcars can move the same number of people as more buses. But this defeats the whole purpose of good transit. Regardless of mode, less frequent transit is less useful transit. And while you might find people willing to wait an extra minute or two for rail over bus, usually the cuts made to frequency result in 5-10 minute longer waits.

Furthermore, because of the aforementioned delays and holdups, streetcars are hands-down, full-stop, slow. Like, slower-than-walking-speed slow. With a long wait for a slow train, you’d be better off walking. And if walking is a better option than transit, what is even the point?

Now, what of the benefits? Streetcars are “better than nothing,” especially on a corridor without existing transit. But other improvements are mode-agnostic. The DC Streetcar will have 3 doors and feature all-door boarding and unloading. But that’s not because it’s a streetcar; it’s because we’ve decided that proof-of-payment is acceptable. Why can’t we do the same for buses? (Short answer: there is no reason we can’t do the same for buses.) In general, if buses are already at capacity, including headways of 1-2 minutes (e.g., 16th Street in DC), the initial solution should be dedicated lanes, which allow for a much more consistent throughput. Once you max out buses in a dedicated lane, then the extra capacity calls for a switch to rail. But at that point you’ve already got a dedicated right-of-way so you’re basically building a light rail anyways.

Essentially, streetcars are a needless step.

But also do not think me part of the faction that David is trying to argue against. He references Jarrett Walker’s opposition to further mixed-traffic streetcars in Toronto and later says that if you are against streetcars, flawed as they might be, you fall into the hands of the anti-transit crazies. But as Walker himself says – and continues to say, despite the Toronto Star‘s misinterpretation of his opposition, that those streetcars are overrated. But opposition to a bad project doesn’t mean opposition to all projects. It in fact means we have to push for more. The car-centric attitude of “no way will we remove a traffic lane for transit” needs to end up in the dustbin of history. We can and we will repurpose lanes for transit because with the state of vehicle traffic and sprawl in this country, dedicated lanes are the only means of making transit actually work (and that’s leaving aside questions of equity).

“So what are the alternatives?” many might ask. I would say it’s the above. Firstly increase the roadway capacity, then increase the vehicle capacity. The Germans have an oft-quoted expression: ” Organisation vor Elektronik vor Beton” (organization before electronics before concrete). Organization to mean sorting out interagency turf wars and that sort of thing. Electronics refers to signalling and other equipment that can increase the capacity of existing infrastructure. Building new infrastructure is last. I’d argue that dedicating lanes is a form of “electronics,” in this sense.

And as Dan finally convinced me, there is, once in a while, sufficient justification for a streetcar. I would suggest that this is true assuming that all of the following conditions are met:

  • Vehicle traffic on the planned route isn’t so intense as to ensure speeds of less than 10mph
  • The streetcar is center-running, to permit eventual dedication of the right-of-way
  • Whatever transit agency is running the streetcar will not use a streetcar’s slightly higher capacity as an excuse to increase headways
  • If the streetcar is on an existing transit route, that existing transit will not suffer service cuts (unless the streetcar runs the entire length of the route)
  • The line constructed connects two actual activity centers/destinations
  • Or, TL;DR: if the route is sufficient to support full light rail/trams (in the Swiss sense of the word), then a streetcar could be built as the light rail version of a premetro.

I will also add that this ignores the “economic development” purpose of streetcars (e.g., the justification for the Columbia Pike streetcar). But that’s intentional. Supposedly what differentiates bus and rail is the development they generate because of the “permanence” of rail. But for just a few thousand dollars’ worth of concrete and some political capital, it’s possible to build an equally “permanent” set of dedicated lanes, which will do much more to help people get around the region. And since “development” seems to inexorably lead towards “gentrification” these days, if dedicated lanes didn’t drive an increase in development, you’ll have achieved a significant service upgrade for an affordable area.

But perhaps the best thing we can do to ensure transit’s continued relevance and utility is to ensure that it connects a) where people are and b) where they’d like to go. This means focusing on underserved but sufficiently dense areas, in order to provide the greatest amount of transit improvement to the greatest number of people. This will in turn drive a harmonious cycle of development regardless of mode.

(Thanks to Dan Malouff, Tim Krepp, Navid Roshan, Michael Hamilton, David Edmonson, David Alpert, and anyone else I’m forgetting for engaging in said discussion).