Are streetcars-in-traffic skeptics sacrificing goodness for perfection?

That's David Alpert's frame in a piece in the Atlantic Citylab today (links added):

Jaffe, Walker, Yglesias, and Capps have no duty to support Team Transit [sic!] no matter what. They should speak their minds. And anyone who supports mass transit expansion should want it to be as close to perfect as possible.

I worry about streetcar criticism that states that a streetcar without every desirable feature is worse than nothing.

But streetcars also have another set of opponents: Those who simply don’t want to fund any transit at all, regardless of its specifics. They seize on any flaw to stop projects that might change their street or interfere with their driving.

So I worry about the effects of this latest trend in streetcar criticism. While streetcar projects can and should be better, many of these articles go further and either imply or outright state that a streetcar without every desirable feature is worse than nothing.

That’s not right. Perfect transit is absolutely a goal, but the perfect must not be the enemy of the good. There are plenty of reasons why a streetcar might be worth supporting, even if it isn't as long, frequent, or speedy as we might like. 

I have spent my whole career helping people value what's really good-but-not-perfect in transit choices.  Our difference is that in Alpert's framing of the question, the fundamental good to be defended at all costs is the streetcar technology, while to me the fundamental good is the liberty of large numbers of human beings, and their access to both happiness and economic opportunity.

Let us take Alpert's perfect-vs-good frame and deploy it differently. Many earnest American leaders visit places like Bordeaux and Strasbourg and agree their cities should look just like that. This looks perfect to them, but they realize they'll have to start with something that's good-but-not-perfect, an imperfect good.

Well, which "good" element should we start with? In Bordeaux and Strasbourg, the streetcar (never mixed with traffic) is a result rather than a cause of a whole bunch of other things: policies that limit car access, for example, so that transit of any mode can run reliably and so that it delivers people into a rich pedestrian space. The Bordeaux and Strasbourg streetcars also began with the "imperfect good" of bus services, which were used to build robust lines with actual existing markets that would support the future rail service.

Why should the "imperfect good that we start with" be the streetcar instead of a really liberating transit system run, for now, by buses?  Why must we start with  a hunk of decontextualized technology rather than our liberty and opportunity to go where we want to go?  

Alpert goes on to make other points about why "imperfect but good" streetcars are worth supporting:

Imperfect transit can still be good for cities.

Millennialsempty nesters, and others want walkable, livable urban places. Unfortunately, there aren’t enough of those in the United States, which is why they’re increasingly expensive.

There are plenty of places on the edges of cities that could become more walkable, more urban, and have more of a sense of place. To do that, they need better transit, more amenities, and more residents—which generally means more density. When such a place achieves greater walkability and urbanization, the factors making it so strengthen over time. … It’s a momentum game, and even an expensive, sub-optimal transit solution—such as a less-frequent streetcar with no dedicated lane —can push the cycle in the right direction.

The sheer abundance of places that need to be made more walkable is actually the strongest argument against the streetcars-in-traffic campaign.  In transit, if it doesn't scale, it doesn't matter.  Streetcars-in-traffic have helped enrich a few superdense districts, but they are far too slow, unreliable, and expensive to scale to the size of our urban mobility problem — at least not as long as they remain stuck in traffic.  (Once they get out of traffic, they are essentially light rail.)  Nor are streetcars remotely necessary for the development of walkable, urban places.  

If you want to see how a city massively expands the usefulness of transit, and thus the potential for transit-oriented lives, look to what is happening in Houston.  Massive, scalable, high-frequency bus grids that are useful for getting all over the city, and that can be created now.  

An imperfect streetcar might be all your city can afford—for now.  …  Your city can make it better later, and may even plan to.

The frame here is: "The streetcar technology is the essential good, and people's ability to access their entire city is a nice-to-have that we hope to add in the future."   

But even if you accept that frame, what's the track record of claims that modern streetcars-in-traffic, first built in compromized ways, have led to later efforts to improve them?  Perhaps you should study Portland, which has been living with this product for longer than any other US city.

The streetcar has been extended up to the limits of usefulness for such a slow-by-design service (about 3.5 miles).  But there are no serious proposals for taking cars out of its lanes for enough distance to matter, nor is there much energy behind extensions.  Why?  

In Portland, support for streetcar spending has collapsed.  A recent Bureau of Transportation poll found that only 38% of Portland residents would assign a more-than-neutral priority to further expansions of the streetcar.  The same number for more frequent bus service is 67%.   (Light rail, in exclusive lanes by definition, is at 59%)

The Portland Streetcar has taught Portland residents a lot about what's really matters as you define an "imperfect good."  Listen to what they've learned:  Frequent, useful, reliable transit — using tools that scale to the scale of the whole city —  is the "imperfect good" that matters.

24 Responses to Are streetcars-in-traffic skeptics sacrificing goodness for perfection?

  1. Alan Robinson October 1, 2014 at 3:38 pm #

    Jarrett, please include a link to the original article.
    David Alpert’s article certainly reinforces Michael’s point of view in the previous article that public discussion of transit in the US forces a false choice of rationing expensive projects vs. planning a complete transportation network. It’s quite sad to see and needs to be forcefully rebuffed. Thank you.

  2. EngineerScotty October 1, 2014 at 4:07 pm #

    In recent years in Portland:
    * The LO Streetcar has been cancelled, and with fairly good reason.
    * The SW Corridor will not be “streetcar” of any sort, whether “rapid streetcar” (Streetcar-sized vehicles in an exclusive ROW) or mixed-traffic; LRT and BRT remain the options on the table for this.
    * The Powell/Division project will be BRT; LRT and Streetcar were both taken off the table for that project.
    A few proposed streetcar extensions are still talked about–north up MLK, east on Broadway to Hollywood, and most bizarrely, a short streetcar line serving the Amberglen development in suburban Hillsboro. But none of these seems to have much traction, let alone detailed planning or funding.

  3. Peter Laws October 1, 2014 at 4:40 pm #

    We’re gearing up for a streetcar here in OKC and I hear the same rationale: “it’s a good starter and we’ll expand it (to two tracks) and more routes later”. No talk of expanding existing transit (which was just re-arranged by N\N). Nope, buses aren’t sexy, but they move a lot more people. I’m worried that the anti-transit (and, surprise, there are lots of them in Oklahoma!) will seize on any stumbles made by the streetcar as an excuse to reduce or eliminate any other non-private-car transportation.
    If you want to build an economic development tool, fine, that’s what Classen and Shartel did 100+ years ago with their OKC streetcar companies, but bill it that way. Don’t pretend it is filling some otherwise unfilled transit need.

  4. EngineerScotty October 1, 2014 at 5:04 pm #

    It’s worth noting that Portland Streetcar was built on top on an existing and functional transit system, and largely pitched as a development tool. It’s basically sexier bus service, and hasn’t been any more than that.
    Building a streetcar service in the absence of a decent existing network, is probably not a wise use of funds in any case.

  5. Rob October 1, 2014 at 7:28 pm #

    Jarrett – Hurray! Thank you, well said.

  6. Jonathan Hammond October 1, 2014 at 8:09 pm #

    I think Alpert is being disingenuous; the interest in streetcar projects in DC is primarily for their real-estate development potential. He’s fallen into a suite of ill-considered policy objectives on that front.

  7. Michael Andersen October 1, 2014 at 9:25 pm #

    I’m not sure how well Streetcar service would ever have polled in Portland; quite possibly it would have been even worse in 1999 before anybody knew what the word meant. Like many Portland transportation ideas, Streetcar may well have been built in spite of public opposition; I’m not sure and I don’t even know if anybody checked.
    In any case, TriMet’s Baby Boom-fueled payroll tax feast of the 90s and early 00s helped Portlanders temporarily avoid the “false choice” Alan describes above, since there was money to invest in bus, light rail and streetcar service alike.
    Since the recession, transit funding in Portland is bumping up against its political constraints and Jarrett’s general interpretation that public opinion does not favor Streetcar investment is pretty persuasive to me.

  8. Marc October 2, 2014 at 5:57 am #

    Mixed-traffic streetcars have made headway because they’ve procrastinated on a much-needed, but difficult, discussion: the need for urban streets to start dedicating lanes and traffic signals for transit just as they already do for cars.
    As argued here…
    … there is more than enough space for transit on most US city streets, but the political/DOT will to dedicate lanes and initiate TSP isn’t there. If we can’t initiate a discussion on proper prioritization for transit now (so it’s actually competitive with cars), then when?!
    Each mixed-traffic streetcar proposal – or mixed-traffic watered-down “BRT” or trunk bus route, for that matter – only keeps pushing this discussion back and back and back…

  9. Angie Schmitt October 2, 2014 at 6:12 am #

    Here’s the thing. You’re only comparing streetcars with other types of transit. Ok, yes other types of transit may be better. What if it was a streetcar or a highway interchange? Or a streetcar vs no additional money for transit? I’m sorry, but if you guys think Cincinnati was at some sort of decision point between expanding bus service and adding a streetcar, you are mistaken. That is just not the case. And furthermore the suggestion that these areas of Cincinnati could densify and redevelop without the streetcar. Yes, theoretically they could. But that has not been the pattern for decades and decades — the opposite has occurred. It takes something big for people in places like that to make a cultural shift.
    I get where you guys are coming from, but I think that is the real flaw in this whole discussion. With Mayor Cranley, it’s not streetcar vs enhanced bus service, it’s streetcar vs more police officers, or literally streetcar vs expensive new interchange. Tucson and Cincinnati are not Portland and DC. And Portlanders might not see streetcar expansion as a priority but thousands, literally thousands of Cincinnatians took to the streets in protest when their streetcar line was under threat. They don’t have light rail.
    In a perfect world, you are right, Cincinnati would weigh the costs and benefits of a streetcar with expanded bus service and decide on expanded bus service. In the real world, that would have to be funded by continuing local dollars, only an extremely small fraction of Cincinnati uses transit and they’re almost all extremely poor and thus have very limited political power. Where would the impetus to do that come from? Wherever the source, the sad fact is greater Cincinnati is a long way from it. If anything, the whole streetcar discussion got many more people interested in transit in Cincinnati, and how it can be a tool for economic development.
    I am hopeful the city of Columbus will gets its act together and copy Cincinnati. If there’s a city perfectly suited for streetcars, it’s Columbus. Everything is centrally located around high street in Columbus and it’s impossible to park there and getting denser all the time.

  10. Jim D. October 2, 2014 at 7:23 am #

    Angie, you make some fair points but there is a troubling aspect to your argument as well. If it is true that a streetcar line is one part of a zero-sum game for transit, then you are quite possibly making a city’s entire transit system worse off in order to divert funding to a relatively small percentage of riders. This is especially true if the streetcar line never scales up to be big enough to take advantage of the mode’s economies of scale – a poorly-designed streetcar line can easily require a higher subsidy per passenger than a bus system. The answer is to fight for all modes and not declare victory when a streetcar/light rail proposal is approved.

  11. Jarrett October 2, 2014 at 7:30 am #

    Angie. If you put a streetcar just on the downtown-OSU segment of High St, what would the all the buses coming in on High St from further north do?

  12. Ben G October 2, 2014 at 7:40 am #

    I was in Toronto several days ago, and there was a stark contrast between the streetcar lines that had dedicated track vs the ones in mixed traffic. I’d imagine the ones in mixed traffic were easily 3x slower than the ones with dedicated track — and probably much worse than that at rush hour when people actually need to get where they are going.

  13. Angie Schmitt October 2, 2014 at 9:14 am #

    I’d take the streetcar all the way through Clintonville maybe even farther north, ideally, through german village on the south side, maybe farther, until it gets suburban. I don’t know, maybe buses could run express down a parallel route from farther to serve the suburban areas. 23 becomes a highway at some point. Can’t remember if it’s before or after the outer belt.
    If Cincinnati’s streetcar expansion is coming at the expense of bus transit, yes you’d have a point, and I wouldn’t necessarily defend that, but that’s not that case here, pretty certain. It’s the most significant transit expansion Cincinnati’s seen in decades. Give them a break. Is it as good as Portland would do after all its experience? No. Fine. Fair critique. Still progress for Cincinnati, IMO.

  14. Jarrett October 2, 2014 at 10:06 am #

    Your “I don’t know” is a problem. A starter streetcar line won’t be bigger than downtown-OSU, and that means enormous numbers of interrupted trips for the bus riders that now flow through this segment to points beyond. This is a great example of how too-small streetcar starter lines that obstruct the bus system yield a net reduction in people’s ability to get where they’re going.

  15. Jim D. October 2, 2014 at 1:03 pm #

    If Cincinnati’s streetcar expansion is coming at the expense of bus transit, yes you’d have a point, and I wouldn’t necessarily defend that, but that’s not that case here, pretty certain.

    I do hope that Cincinnati’s leadership is prepared to support additional funding for the streetcar line. Even a barebones rail line is typically two to three times more expensive to operate per hour than a bus line.

  16. Wanderer October 2, 2014 at 1:29 pm #

    I’m in Oakland, and I know how utterly distorted transit planning politics can get. Still, it seems like transit advocates would want to fight to reframe the choices, rather than accept minimal and lousy ones. Transit advocates should be the ones saying (in many cases) “This slow, short streetcar will do almost nothing for our city’s transit system. We should spend this money on _________, which will be really effective.”
    We want to support transit, and we don’t want to play into the hands of pro-car politicians and groups. But at what point do we say this proposal is so bad, at least as likely to do harm than good, that we don’t want it.

  17. Jarrett October 3, 2014 at 7:36 am #

    Wanderer. Fairly early in the game, I think. Jarrett

  18. Alex October 3, 2014 at 7:59 am #

    In Minneapolis, a streetcar is proposed for about half of the length of Nicollet Ave, where a bus currently runs the entire length. The rump bus service is proposed to run express to downtown after the segment where the streetcar begins. Bizarrely, there is also a limited-stop bus service proposed for the segment where the streetcar will be (and even more confusing, a local bus as well). They don’t explain how the limited-stop bus will ever be able to pass the streetcar on two-lane Nicollet Ave. This service plan was developed by URS in conjunction with the City of Minneapolis.

  19. Robert Wightman October 3, 2014 at 8:55 am #

    Bem G says:
    “I was in Toronto several days ago, and there was a stark contrast between the streetcar lines that had dedicated track vs the ones in mixed traffic. I’d imagine the ones in mixed traffic were easily 3x slower than the ones with dedicated track — and probably much worse than that at rush hour when people actually need to get where they are going.”
    True they are slower than autos but when you carry over 50,000 passengers per day on King Street they are the only thing that will work on Toronto’s 4 lane downtown streets. The bus routes in the downtown are all slower than the mixed traffic streetcars.
    I agree though that a lot of the lines being built in some US cities are not a wise investment.

  20. Angie Schmitt October 6, 2014 at 9:55 am #

    Actually Mayor Cranley has insisted that streetcar funding NOT come from regular transit funding and is promoting a plan to pay for it with increased street parking fees. Part of the funding in the early years will actually be private, it will come from foundations.
    Clintonville to German Village is only 8.8 miles at the longest. Dismiss it if you’d like. But the business as usual approach in Columbus is developing 500 additional square miles outside the outer belt — something that definitely won’t be cheap or good for transit riders. I used to ride the 2 down high street. Not many suburban commuters are going to choose a route like that that stops every two blocks anyway. It just wouldn’t be efficient. It’s a short distance route, primarily. An express route would be better for them anyway.
    Also, did you guys see the study from Reid Ewing that found a 3 to 1 multiplier around Portland’s Max light rail in miles diverted from driving thanks to transit oriented development? That’s sorta what pro streetcar folks have been arguing. Anti-streetcar folks have sort of dismissed that as a transportation benefit, but it’s a real thing. I suppose Columbus could build TOD around bus lines, but let’s be real, they will not. I do think more BRT would be good there for sure.
    Meanwhile Mike Coleman has proposed light rail to the airport, which makes no sense to me. But it would be fast! Even though it’d serve entirely low density, suburban communities and have almost zero potential for TOD.

  21. M1EK October 7, 2014 at 8:00 am #

    Jarrett, somewhat off topic, but what would it take to get you to offer an opinion on Austin’s Project Connect proposition on the ballot in a month?

  22. DC Bus Rider October 8, 2014 at 6:00 pm #

    What’s maddening about David Alpert is that – unlike Mr. Walker – he refuses to air dissent on the counterviews of streetcars on his blog. This most recent article proves once and for all that he views foes of streetcars as enabling the anti-transit enemy. Period. Accordingly, he does not run articles or even links to articles – ever – that are skeptical of streetcars, whether from anti-transit right wingers, The Economist, or a man as pro-transit as they come, Jarrett Walker. (Unlike this blog, there is no link to this post on Mr. Alpert’s blog, He claims that pro-transit writers “are under no duty” to support his preferred mode, yet he cannot be adult enough to air the debate on his blog. So when it comes to editors, contributors and presumably readers of his blog, they are, in fact, under a duty to support streetcars.

  23. Wanderer October 10, 2014 at 4:04 pm #

    Angie, The Portland MAX has, for the great bulk of its route, dedicated right of way. It does run on-street through Downtown Portland. So MAX can offer a transportation time savings that the Portland streetcar simply can’t. There’s no reason to expect that a streetcar would have the impact that MAX does (assuming that Ewing’s analysis is correct).
    The roadway analogy would be saying that the effects of a new freeway would be replicated by roadway improvements on a local arterial.
    Oregon, the Portland metro government, and the city of Portland have all had strong policies seeking to limit sprawl and support transit-oriented development and Downtown development. Those are good things, but they mean that TOD is likely to be stronger in Portland than elsewhere. I’m not familiar with Columbus, but it doesn’t sound like they’re going to do that.
    I don’t know the route, but transit planners often work on spreading out bus stops that are too closely spaced. To me, this is an example of an apples-oranges comparison. A new rail route will have properly spaced stops vs. bus stops that are too close. But bus stops can be relocated, Portland for one made a huge effort at that when it rebuilt the Transit Mall.

  24. Hobokun October 11, 2014 at 11:52 pm #

    I definetely agree that a city should flesh out its existing bus system and designate right-of-way before laying down the rail, and thankfully thats just what we did here in Los Angeles. In addition to the hundreds of miles of “local” routes that take you literaly anywhere within a 20-mile radius of central LA, all the major streets have “rapid” limited-stop buses that have designated lanes in a few parts. When this didnt meet the citys transit needs (I take the Western Ave rapid regularly, & it takes 5 minutes to go a mile around 4:30 pm) we began to work on a massive rail building campaign to expand our tiny system, which had previously served as a comparison to the bus lines, to see which was better. And we’re building it the right way, with only one street-running segment built and one other short one planned, for what will be a 7-line system spanning 100s of miles.