Bureaucratic Attachment Parenting

Great article in Salon. One of my most basic community design beliefs is that when you do commerce, you have to split it up into multiple parcels and sell them to different owners. A “town center” or “main street” that has a bunch of facades but one owner is always going to be a mall; only a hodgepodge of variously-owned properties is going to have a snowball’s chance in hell of evolving into something authentic.

High-speed rail is pro-sprawl

So the LA Times has this article up on two different visions for California, where they quote some HSR backers who say it’s all about “reducing the suburbanization of California” and “communities of dense apartments around stations,” and then they quote some teabaggers who say “YOU GONNA FORCE US INTO SOVIET APARTMENT BLOCKS, WHY DO YOU HATE AMERICA.”

This is all so much poppycock.


Say that to yourself five times.

Here’s the thing. You don’t need trains to have communities of dense apartments near urban centers. You don’t need cars, and you don’t even need streetcars. That’s… pretty much the natural order of things. The whole purpose of commuter transportation is and has been, historically, so we don’t have to live at high density.

Subways allowed easterners to move from tenements to rowhouses. Streetcars allowed westerners to switch from apartments to single-family detached. Interurbans let you move to the next city over. Commuter trains let you move fifteen, twenty miles out into the country, and freeways simply expanded that range. With a trolley, you could live on 50th and Hawthorne and have an easy ride into Downtown PDX. Trains let you live in Riverside, Illinois and hop an express into the City of Chi. Entire suburbs of low-density housing were built around train lines. Trains allowed Joe Biden to live in Delaware and work in Washington DC.

Trains will allow Cali to sprawl even more.

Suppose you’ve got a business with a client base in LA. Right now, your option is… to have an office in LA. And live in LA. Sure, you might be in Redondo, or Whittier, or Compton, but you’re pretty much stuck in the basin, because the customers dictate the office location and the office dictates the house location.

But now suppose you’ve got a bullet train. All of a sudden, Bakersfield is less than an hour away. You can stick your office in East LA, and you’ve got an hour commute counting the transfer to the Yellow Line. Or you can stick your office up in Bakersfield, and just make a bunch of client visits on the train. Maybe you stash a car in a garage near LAUPT so you don’t even need to use transit to get around at the other end.

So now you’ve got an office in Bakersfield. Your work commute just switched from bumper-to-bumper on the 405 to stunning mountain roads. You buy a Harley, or a Hayabusa, depending on how old you are. Life is sweeet.

Think 10,000 people wouldn’t make the same choice?

Want a compact city? Under-develop your transport network. Like Baghdad. The place is scarcely 12 miles across, before it fades to desert on all sides. Yet it holds seven million people.

Want a sprawling city? Build lots of trains. Like Kyoto-Osaka-Kobe, which sprawls for 100 miles. The three cities combined hold only 5.5 million people, but the extended metropolitan region holds three times as many. How do you get such a widely dispersed, sprawling population? Trains.

Of course, you can do it with roads too. Barker-Cypress to Baytown is about 45 minutes in the off-peak, thanks to Houston’s massive freeway capacity. This might not be the ideal a lot of the eco-boosters have in mind, but that’s the thing – transportation’s pretty agnostic.

And there’s an outer limit to highway sprawl. People get pissy when their one-way commutes start to tick above 45 minutes. Pretty much every urban center, then, can support an initial 45-minute radius of people who can commute to Downtown, and then another, lighter 45-minute radius of people who can commute to various suburban job centers. Since those centers are rarely on the outer edge of the first 45-minute circle, your total radius is about 75 minutes, which is typically about 30-40 miles. All of your fringe suburbs are in this range. Marysville to Seattle is 35 miles. Rosenberg to Downtown Houston is 36. Aurora to the Chicago Loop is 41. With an autocentric transport policy, this is as far as cities go.

But high speed rail expands this range by powering through those first 40 miles of auto-sprawl at neck-snapping speeds, then making periodic stops out in the hinterlands. Those hinterlands are then free to sprawl themselves. Hence, Wilmington.

So, by all means, build high speed rail. Fast trains are awesome. But understand that those trains will in fact promote more development even further out. EVERYTHING – whether we’re talking about the Katy Freeway or the TGV – is going to be used to carry out life at a lower density, to spread out, to sprawl. That’s how humanity works.

Why is there no Express Light Rail?

Why doesn’t anyone, anywhere in the US run express light rail?

No, seriously.

There’s nothing inherent about US LRT technology which says every train has to run the same boring route all the time. Sidings and pullouts and overtake points are entirely possible. Actually, US LRT has about the same approximate loading gauge as the Japanese 1067mm network. And those guys are the masters of this shizzle. Check it:

That’s a Keihan 7000-series overtaking what looks like a refurb 2400 series on the far side of the platform. Local shows up, express shows up a minute later, express takes off, local follows. I’ve actually seen a video of a four-train meet at this stop – NB and SB locals arrive within 30 seconds of each other, NB and SB expresses show up simultaneously, expresses depart, locals follow.

I don’t expect North American operators to have the ability to run the tight, exact headways of a suburban Japanese railway, but the basic principle is quite possible. The local just has to chill for maybe 5 minutes instead of 2.

Where would this work? Well, lots of places, but for a system that particularly cries out for express LRT, look no further than this proposal for Raleigh-Durham-Chapel Hill. Yonah has a very nice diagrammatic map up detailing their proposed LRT and commuter rail lines.


The Raleigh-Cary side is 18 miles, and a full 10 miles of those directly parallel the commuter rail, from West Morgan to Cary Parkway. Likewise, the Durham-Chapel Hill side is 17 miles, and 4 of those are parallel to the commuter rail. The distance between endpoints for the LRT is 15 miles.

So: 14 miles of redundant commuter rail to connect to a 15-mile “core” section.

But why not just build a single-track LRT track in between cities? Run LRVs on commuter rail frequencies in between the cities, run them on typical LRT frequencies within. Same service plan, but a single technology. They could save themselves from building a bunch of redundant systems. Not only that, but most transit agencies own their own LRT, fee simple, where the freight railroads keep the tracks. Transit agencies paying to improve freight railroads for commuter trains essentially amounts to giving some dude money to buy himself a house, in exchange for him agreeing to let you sleep there.


Suppose Raleigh and Durham were connected by continuous LRT. There might be a case for the Greenfield-Parkway-to-Downtown-Raleigh segment as a standalone DMU line. You know, grab some RDC’s from somewhere, or something. But there’s no way the single station in “West Durham” would justify its own commuter rail spur if LRT was continuous. There is, then, eight miles of proposed commuter rail on the Durham side which is duplicative, which brings us to 18 miles of redundant commuter rail to bridge a 14-mile gap.


So what should this thing look like instead?

The stupid, idiotic, dumb move would be to say, “well let’s just build one rail system, durrr,” construct LRT between the cities using the same doubletrack cross section as within them, and add in a bunch of extra stations in the RTP area because hey, that powerpoint handout you got says that “light rail” stops every 1/2 mile to a mile, as opposed to “commuter rail” which stops every 3-5 miles. Congratulations, you have now created a slow loris rail that takes two hours to go from end to end.

No, the planners have got it figured right that the demand within Durham and Raleigh is different than the demand between the cities, it’s got different req’d peak frequencies, different baseline service levels. But there’s no reason not to operate it over a single track network using a single contiguous technology. Like this:

Midday operations might work like so. Trains originate at Northeast Center every 15 minutes. Every other train becomes an express from Downtown Raleigh to Cary Parkway, stopping only at NCSU and Downtown Cary. Meanwhile a “shorty” local train originates in Raleigh one minute after the express leaves and runs the rest of the way to Downtown Cary. This keeps 15-minute service at all local stops while allowing a single-seat ride from Raleigh and points Northeast to Durham-Chapel Hill.

All the locals turn back at Cary Parkway, but the expresses continue on, through RTP, all the way to Durham/Alston where they revert to locals and run to UNC. Locals originate at Alston on a staggered 30-minute frequency so that there’s 15-minute coverage over the Durham-Chapel Hill line.

This gives you 15-minute service over the LRT segments and 30-minute service over the “commuter rail” segments, but Northeast Center to UNC is now a single-seat ride of about 75-80 minutes in length. Not at all bad, for a trip that’s 35 miles even via the most direct auto route.


Once you define UNC-Durham-Raleigh as a single continuous line, you realize that it’s crazy indirect for endpoint to endpoint trips – most notably the large swing to the north from Chapel Hill to Durham and back. For a starter line, that’s great… but with a single-technology system, you can add shortcuts and expansions later.

Yep, that’s bee-line UNC-NCSU service plus a direct airport connection. It’s really not any different than a big train set. Buy the basic kit with oval of track and sleek locomotive! Then add variations for an even more fun setup!

But you can’t do this if you’re starting out with some balkanized, Philadelphia-type system where everything is a different and incompatible mode. Nope. You need Express Light Rail. But… no one’s done it. Anywhere, in the entire US – at least in modern times.


The Melbourne Identity

What an awful attempt at putting a pun in a posting title. I’m like an elderly columnist trying to stay relevant, who doesn’t realize that they haven’t released one of those movies in like five years.

Anyway, check this out:

Two things stand out in this photo. Tram terminus, 50mph speed limit. Not bad eh?

I like pics like this because they’re a nice rejoinder to George Will-esque arguments that trains are just there to brainwash you into not driving, like a true individual. Nope. The two can, and should, exist side by side.

Melbourne has a crazy extensive tram network. That map shows lines as well as frequencies; note the preponderance of 9- and 12- minute headways throughout the system. And Melbourne seems to share a lot of the core values of most North American cities. Globally, a lot of places have extensive tram networks. Several exceed Melbourne. But to a certain extent there’s a feeling like “well, they’re Europeans, you expect them to have good trams.” The urban vibe is completely different.

But Melbourne is a sprawling, suburban city, with low-density suburbs and a downtown defined by large office towers. When the Wachowski brothers needed a “generic North American city” for The Matrix movies, they shot in Melbourne. Here’s a couple more spots on that same tram line:

As you get closer in, the tracks switch from LRT-style separated running to mixed traffic, and low-density single-family gives way to your basic “Goldilocks urbanism” with a mixture of detached houses and apartment blocks, walkable commercial streets and parkable strip centers.

Not bad at all.

And while the highway system doesn’t really come close to Florida-Texas-Californian levels of buildout, it is pretty new. 15 years ago Melbourne was following the eastern European model, where radial freeways all slow out into surface streets. But with CityLink they brought motorways into the core and created a proper crosstown expressway network. Check the 1960’s World’s Fair architecture on that “sound tube”, which supposedly reduces traffic noise for the benefit of some nearby housing projects. I’m not sure I buy the stated rationale, but then, I think it’s worth it to make freeways look cool for the sake of it. Like this column detailing on Moses’s BQE – straight outta Popular Mechanics – or this incredibly cool sign arch on Seattle’s Alaskan Way Viaduct.

Induced Demand isn’t an argument against highways.

Watching the Coogs pwn Pedo State here, it’s a good feeling.

Anyway, Induced Demand (which I’ve also heard referred to as “Triple Convergence”) is the phenomenon whereby traffic rapidly expands to fill newly-created highway lanes.

From where I sit the arguments look a lot like those on global warming. There’s broad consensus across disciplines that it exists, though some claim otherwise. And it often gets used to justify really, really crappy policies. Like not building highways, or even tearing them down.

So here’s a few problems I see with the arguments.

1: You’re still carrying more people.

One of the more common arguments against widening goes like “such and such a highway averaged just 28mph at rush hour in [year]. Five years after they finished widening it, it was still clogged, averaging just 31mph in [later year]. Clearly, building more highways won’t fix congestion.”

The problem here lies in defining congestion as speed irrespective of capacity. If you have a four-lane highway moving at 30mph, you’re maybe carrying 80,000 cars a day. Going to a ten-lane highway while retaining the 30mph jam-up means you’re carrying north of 200,000 cars a day. That’s, at a minimum, 120,000 people who get to benefit from that capacity, even if it’s not moving super-fast. 120,000 people who got to move closer to where they want to live, got to take a different, better job, or just eat dinner somewhere on the far side of town.

2: You can narrow the rush hour

There’s a post over on Greater Greater Washington about “Myths about highways”, and under one of them the writer says “Neither Atlanta nor Houston’s multiple Beltways have erased congestion.” Right. Beltway 8 doesn’t have magical powers. But what Houston’s capacity increases have done is narrow the rush hour. Think about it.

Everywhere in the US (except, perhaps, Toledo) is jammed at 4:45 in the afternoon. But what’s it like at 6? 7? 8? In Portland you can leave downtown at 7:30, head north, and still hit a slowdown when you get to the Interstate Bridge. The Schuylkill Expressway in Philadelphia has one of the widest rush hours I’ve ever seen; that road jams up on Sundays. Everywhere, always. But in a city with massive capacity like Houston, the rush hour starts and ends quickly. I’m accustomed to doing 70+ on IH 10 into the mid-afternoon, and I’m likewise accustomed to free-flow on US 59 as early as 6:00, 6:30pm. Congestion shows up, and then it leaves. It doesn’t linger.

Looking at “rush hour” conditions creates a blinkered view. It’s almost impossible to build a non-toll highway system which will operate at LOS A at 4:30pm. But it’s quite possible to build one that will clear up within the hour. And it’s the difference between that highway and the one that stays clogged until 9pm that controls whether people eat dinner across town, how much they socialize with people in other places, whether a given metro area or region is truly connected.

3: You enabled the decisions which led to the induced demand

On that same “myths about highways” post, the author argues that, rather than take traffic off Lee Highway or Arlington Boulevard, a wider I-66 would have lead to “More people … living west of Manassas and working in downtown DC.” And what, exactly, is wrong with that? Given that some people want to live in the ‘burbs regardless, would you rather have a completely fragmented economy, where ‘burbians work in the ‘burbs and city people work in the city?

I recall reading in a book somewhere that San Jose’s West Valley Freeway, originally intended to provide a smooth bypass around downtown, instead attracted so much new housing development around it that it was congested within a year or two. But I don’t really see that as a negative.

Transportation networks enable long-term land use decisions. That’s why cities like Portland are so gung-ho on Streetcars, the permanance of the rails leads developers to build big mixed-use condoblocks. Freeways work the same way with lower-density uses, like single-family detached homes. 85 opened up the southwest flank of San Jose to development. Certainly if I was one of the developers building those neighborhoods, I’d be very happy that they filled up so fast that the freeway was jammed within a year.

Now, let’s say you don’t like the very idea of single-family housing, you think we should live in high-density apartment blocks, preserve open space, etc. OK, I’m receptive to that. I’ve read a bunch of stuff that says Dutch kids are happier than American kids, score lower on measures of dysfunction (teenage pregnancy, drug use, whathaveyou) and they certainly live denser than we do. But that’s not an argument against the freeway’s effectiveness. You’re not saying “The freeway doesn’t work,” you’re saying “the freeway isn’t conducive to the kind of development I prefer.”

Rhetorical Gimmicks

That list bit is, let’s be honest, a rhetorical gimmick. At the urban-suburban level, freeways promote low-density housing development. Most of the people arguing against freeways don’t really like low-density housing development. But if you make a facial argument that “low density development is bad,” you’ll get a lot of responses like “I like my backyard” or “what do you have against being able to find a parking space,” maybe even a backlash group or two.

On the other hand, if you can argue “freeways don’t work, they’re always jammed up,” you might very well get that same person who likes her yard, her two car garage, to say “well, I get stuck in traffic a lot, you might be right.” Might even get them to sign on to the idea that we shouldn’t build any more freeways, because traffic.

But it’s still a gimmick.

PDX and the Sanctity of the Bungalow

Hypothesis: The common denominator between all Portland development and infrastructure projects is the need to preserve the Bungalow, the archetypical “craftsman” single-family detached residence. Official literature touts PDX as being a center of “smart growth,” but E. Kimbark MacColl was writing about a city of “backyards and barbeques” back in the 70’s, and wherever smart growth runs into established neighborhoods of single-family, 1.5-story dwellings, “preservation” wins out over “density” or “mixed use” every time.

The first Portland freeways escaped criticism because they bypassed bungalows. The Banfield clung to a freight rail corridor, the Baldock repurposed an existing interurban right-of-way, and Canyon was a piecemeal improvement of an existing highway. I-5 north took out houses, but those were largely black neighborhoods, and blacks in the 50’s didn’t have a lot of political sway over highway alignments. The first freeway to require bungalow demolition in white neighborhoods was the Mount Hood. It was a thick swath, with feeder roads and a wide median reserved fur future transit corridors, very similar to Houston’s SH 288. The Mount Hood, not conincidentally, is the one that got canceled, the “turning point” in the Portland myth where the city repented to pursue alternative transportation, smart growth, whathaveyou.

Since then, every big Portland project has avoided the bungalows. The headline urban districts – first the Pearl, now the South Waterfront – were both constructed on former industrial land. Portland built a new freeway through industrial sections as late as the 80’s (the US 30 – Yeon connector), and the Sunrise Corridor will likewise take out some warehousey stuff. So we know warehouses are fair game. What about other uses?

Portland has three major groupings of residential zoning (the headline districts use EXd, which is sort of a “hack”, since it’s nominally an “employment” district but allows unlimited residential units subject to design review). On the low-density end you have R7’s and R10’s which are sixth- and quarter-acre lots, respectively. You really only find it on the outskirts, near where Portland runs into unabashedly suburban places like Milwaukie or Gresham. On the high-density end you have various R1 and R2 zones and their friends, which are used for everything from inner-city apartment blocks to suburban garden units. And in the middle is R5, the sacred bungalow zone. So check out a zoning map:

The first thing you’ll notice here is strip zoning. Pretty much everything that fronts on an arterial is commercial or mixed-use. But if you get even 150′ from that arterial you’re into protected R5. Now, you might hear that this is to promote transit development and whatnot. But transit accessibility is determined by walking distance, which is the same on a sidestreet as it is along the arterial. Here, the sidestreets don’t allow transit development, which tells you the strip zoning is there not to encourage transit use but to push development to the fringes of bungalow neighborhoods. Arterial-frontage lots in southeast Portland are mostly single-story retail, which is fair game for redevelopment.

The second thing you’ll notice is how there are big chunks of higher-density zoning to the east, around 82nd. Now, one might ask, why? The inner neighborhoods are the more desirable ones, they’ve got more parks and retail within any given walking distance, rents are higher, commutes are shorter. But here all the allowed density is out along 82nd. Doesn’t make sense if you’re thinking in terms of urban services, etc. But it makes shitloads of sense when you realize that the inner ring is bungalows, the outer ring is ranchers. 40’s and 50’s ranch houses aren’t sacred the way bungalows are sacred. They don’t exude that “Craftsman charm.” So it’s OK to tear them down for redevelopment.

Some might argue that this zoning is there because of Green Line MAX, that this is actually transit-oriented development. But this isn’t the case. For one thing, these higher densities continue well east of any MAX station walk circle:

For another thing, closer-in MAX stations, like 60th Street, have R5-protected bungalows within 100 yards of the damn platform:

So the list of building typologies it’s okay to demo and redevelop are:

-Heavy Industrial

-Light Industrial


-Postwar residential

Stuff that’s off-limits?


Fanis Grammenos’ follow-up to his Portland-grid diss ends with a praise of protected bungalow zoning in Ladd’s Addition:

Having a strong sense of community identity and an appreciation of its valued attributes, residents fought and achieved a down-zoning of its future density. Though by no means urban at 7 dwelling units per acre (18 per ha), it seems to produce a satisfying milieu. The residents have embraced the result and the APA lauds their strong attachment.


The primary problem with bungalow preservation uber alles is that it’s a mismatch with all the rest of the Portland region’s policies regarding growth and transport. The specific issues that arise from this mismatch depend on which perspective you address it from.

From a “smart growth” standpoint, zoning based on bungalow preservation acts to stunt growth. It prevents east and north Portland from experiencing the kinds of density increases that have spread throughout Houston’s inner loop. There is, fundamentally, no reason why the entire swath of Portland north of Woodstock, west of 50th, and south of Fremont could not be zoned EXd tomorrow. Certainly the demand is there. And Portland’s older neighborhoods have more going for them than the brand-new from-scratch places like the Pearl or South Waterfront.

From a “bungalows uber alles” standpoint, Portland’s transport setup is poorly designed. Longstanding decisions to run MAX light rail at grade in Downtown render it useless for crosstown trips. When I lived in PDX I was known to hop off at Goose Hollow and ride a bike to Lloyd Center, jumping 2 or 3 trains ahead in the schedule by doing so.

If the bungalows are the heart of Portland than it’s important to make them accessible, and that means better connections to the region’s employment centers. In this respect I call out the Sellwood Bridge reconstruction plan as particularly bad. The Sellwood redo should have been part of a major east-west capacity addition – a 4- or 6-lane bridge, conversion of Tacoma Street to half of a one-way couplet, and widening of portions of Taylor’s Ferry, Terwilliger, and Multnomah Boulevards.

Then again, perhaps no one wants a six-lane Sellwood. Either method is a legitimate growth strategy. You can accept that auto traffic will get worse, and instead focus on upzoning and building out transit infrastructure and upzone. New York City follows this approach; they haven’t added any new auto capacity in well over fifty years, and the only transport improvements being seriously planned all run on rails. You can also build a city full of single-family zoning, and focus on building the highway infrastructure this requires. Phoenix comes to mind. The vast majority of Maricopa county is zoning-restricted to a sprawling, low-rise form. But the freeways are new and smooth, arterials are being expanded, and there’s even a light rail system that I hear is doing pretty well.

My personal preference hews towards allowing everything, like Houston does. Houston just passed an extension of the “urban” area (which allows higher densities), and they’re building craptons of light rail that will support a denser, more urban form. At the same point, highways have not been neglected, whether you’re talking about rebuilds of inner-loop radial freeways or outer beltways like the always-controversial Segment E. You can have your cake and eat it too.

But the common denominator between all these cities is that their land use plans and transport plans are in harmony with each other. Phoenix’s highways support its sprawl. New York’s rail lines support increased urban densities. And Houston, which has no zoning at all, is building a bunch of highways *and* trains so pretty much whatever happens, they’re down. But Portland is building bike lanes and slow-loris light rail while simultaneously prohibiting urban redevelopment in almost all of its myriad low-density neighborhoods. What’s up with that?

How it’s done.

Check out this suburban intersection in Calgary:

Obviously, not my personal favorite style of development. But if you are going to build this sort of low-density, segregated-use, single-family (and a lot of us prefer this stuff), this is how to. The right-of-way banking at this intersection is agnostic – it could support a grade separation with uninterrupted movement on either the east-west or the north-south arterial. On the non-road side, there’s a good heirarchy of green space – improved parks/ballfields within the neighborhoods, and the much-vaunted “open space” with trails outside.

A lot of the big MPCs in Houston do a good job with this sort of park heirarchy. The Woodlands comes to mind. But the Woodlands absolutely fails at moving the massive east-west traffic flows which attempt to use its arterials every morning. It can take longer to get from the back of the Woodlands to the front than it does to get the rest of the way downtown via the Hardy or the HOV.

People tend to brush past this when they talk about Calgary being transit-centric. Certainly, it’s got a great system – the C-Train provides 21-22 hours of a coverage a day and runs every 15 minutes from the moment the trains start, every 3 minutes during the rush. But when you hear people talk up Calgary’s transit, they’ll say stuff like “the city chose not to build freeways connecting Downtown.” That’s technically true, in the sense that there’s no “10th Avenue Expressway” spitting out traffic right at the heart of the core. But Calgary didn’t and doesn’t exactly neglect auto infrastructure.

For one thing, Downtown has a dense arterial network connecting it to those freeways which don’t quite go all the way there. From the west, a combined 14 lanes are available via Memorial, Bow Trail, and the 11th/12th Couplet. From the east, there’s 12 lanes between Memorial and 9th alone, more if you count some of the other arterial routes in.

For another thing, Calgary is big on grade separations. Streets have a few (Blackfoot) or maybe five or six (Memorial) or even twelve (16th Avenue). You can actually get pretty good performance out of an arterial network this way. A 45mph arterial grid with grade separations (Calgary) can provide the same travel times as a 60mph freeway network which interfaces with a 35mph arterial grid with stoplights everywhere (Houston).

So give these guys credit for planning. Give ’em credit for running an LRT system that doesn’t degrade into suburban bus-type headways after 10pm like Trimet does. But don’t drop any narratives about “showing a way forward with less dependence on cars” or some sort of thing like that. Calgary’s got a solid highway network, and they’re planning for more.

All good Libertarians are pro-transit

It constantly blows me away that rank and file (L)’s don’t like trains. I can understand the REASON foundation or someone else who’s getting a frackload of money from the oil companies, contractors, etc. But the rank and file?

To me it’s really simple.

Once upon a time, there were trains. They were private. And it was good. A Krugman– or Yglesias-style argument in favor of stimulus will point out that the transcontinental railroads were made possible by ginormous land grants from the federal government, a clear-cut “in kind” capital cost subsidy. And of course, they’re right. But there was no operating subsidy. The Rainhill Trials, that was all private financing. And more important, all the streetcar and urban rail was private too. The BMT was private (and tasty), the Key System was private, and the Galveston-Houston Electric Railway, that was definitely private.

Sometimes the trains were tied up with land development interests (private) or freight railroads (private) or utility companies (TRUSTS!), and once in a blue moon they were even public from the start. But for the most part it was all capitalism, “free enterprise” in the preferred GOP newspeak.

Meanwhile, governments large and small set about using YOUR TAX DOLLARS to build FREE competitors to the trains. In 1926 we came up with numbers and then in 1935 we dumped scads of money into roads. We also made it ILLEGAL for an electric company to own a streetcar company, which, if you think about it, is really stupid, since the primary infrastructure of an electric company is a bunch of wires and the primary infrastructure of a streetcar network is also… a bunch of wires, and the primary nonlabor input to streetcars is electricity, which electric companies sort of by definition have in spades… but hey, it was the 30’s, Americans had been screwed by Wall Street, they were tired of monopolies, shit happened.

Now you’d think that this would kill the trains. Certainly, more than a few small-town trolleys folded once the electric company umbilical cord was cut. But actually what happened was the trains came back stronger than ever. Basically, the “free enterprise” railroads decided to take their subsidized government competition head-on. And the result was this:

But, the government just kept on building highways. And not only that, they passed a law that said trains could only do 80mph unless the railroads installed a ridiculously expensive signaling system. So what did the railroads do? They all went and bought these:

In fact, the 50’s turned into the greatest decade for passenger rail. Smoothsides and E-Units were cranked out by the hundreds. Even commuter trains got streamlined. So what did the feds do? They went and did this:

And what did the cities and states do? Well, they spent a whole frackload of money building these:

And against that two-front assault, the trains folded. Most of the interurbans and streetcars got tore out in the 50’s. Railroad petitions to abandon passenger service really took off in the 60’s. In 1971, Amtrak took over most of the passenger trains and immediately cut over half the routes. In fact, far from being the national savior of passenger rail which some now ascribe to it, Amtrak was originally intended to facilitate an orderly wind-down of passenger service in the US. That’s why it doesn’t have a guaranteed funding source, why it’s constantly subject to the whims of who’s appropriating this year, why it squirrels away money like a battered housewife and why it maintains a staunch institutional commitment to incrementalism. It simply wasn’t intended to be a permanent entity.

Nowadays you propose a train system and (L)’s and (R)’s scream “socialism.” For me, that’s a serious WTF moment. Because railroads were PRIVATE ENTITIES THAT SUCCESSFULLY FOUGHT FREE PUBLIC COMPETITION FOR NEARLY FORTY YEARS.

So let’s posit an alternate history eh?

In 1926, the federal government nationalizes the core, trunk national rail system. (They leave the little branch lines and whatnot to private operators). They don’t make huge improvements at first, but they do create standardized train names, lots of reliable clockwork long-distance routes, and publish it all in a big national timetable.

Soon after, the depression hits. In 1935 the WPA and the PWA began sinking money into railroads in earnest. The biggest programs are electrification – where the US is still using steam in the 30’s, by the time the program wraps in 1943 you can travel under wire nonstop from NYC to LA, from Miami to Seattle.

In 1946 the federal government passes the “60mph rule,” which installs a maximum speed limit of 60mph on all public and private highways unless continuous rubberized guardrails are installed on both sides of the road. The Pennsylvania Turnpike installs these, and later the New Jersey Turnpike follows, but highway authorities elsewhere find the cost to be prohibitive.

Nevertheless, the nation’s automobility won’t be tied down. Throughout the 50’s, more people are driving cars then EVER. So in 1956, the federal government passes the “national system of interstate and defense railways.” From whole cloth, an entirely new system of high-speed railroads is constructed. These are fully grade separated, allowing speeds of 130mph throughout the system. Meanwhile, states and cities embark on subway programs throughout the 60’s and 70’s, replacing slow surface streetcars with fast municipal trains.

What do you think such a system would look like?

Well, it’s not even a hypothetical question, because I’ve just described Japan.

OK, so there was never any 60mph turnpike speed limit. But the basic outline – government nationalizes railway, government spends lots of money on electrification, government then constructs brand-new parallel high-speed system out of whole cloth – is exactly the history of the JR Group.

Know what? JR is private. Well, it’s private in the sense that GM is private. The Japanese National Railways had a whole chunk of debt from building the Shinkansen system, so the Japanese government spun it off into two companies – one that just held the primary rail assets, one that held the debt and other assorted frivolities. All the union contracts were renogiated. Pretty much like “old GM” and “new GM.”

But if that doesn’t quite meet your definition of “free enterprise”, you can’t really argue with the 100% privately owned suburban railways, which have survived by successfully competing with the government railway for 80 years.

That’s the Moonlight Echigo, doing about 55 by my estimate.

Other trains are faster. The Hokuhoku line runs at 100 (160km/h):

Should we try to emulate Japan? Naw. For one thing, that rail network is supported by a density that doesn’t exist here. And I’m cool with that. No, if I had to pick a country to emulate, it’d be Germany, ‘cuz you get ICE-T‘s and unrestricted Autobahns. But the truth is I don’t want to emulate any country. We build kickass roads here in the US of A. We used to build kickass trains, too. In fact, back in the 60’s, Japanese train companies licensed American train technology. Think about that for a minute. The Japanese… paid for our train knowhow.

Blows your mind.

So what I want, is for us to take our proven national badassery as regards highways and apply it to fast trains. And I want the (L)’s and the (R)’s to stop acting like roads that are half-funded by gas taxes are some sort of capitalist paradise while trains are an evil government plot to brainwash you into collectivism.

Trains were the original private enterprise. Roads were pretty much always socialism. That’s why the stuff on your N scale layout is all painted different eye-catching colors, whereas the highways all look pretty much the same. And (R)’s wanna call Amtrak a “Soviet-style operation”? Please commit Seppuku.

Then let’s build some more trains.

No, not really.

I think this is probably the third article I’ve read in the last month asking: “Are Freeways Doomed?” “Is THIS the post-freeway age?” “Are Urban areas moving on?”


All of these pieces work like any “bogus trend” piece – string together a few anecdotes, posit a trend, quote a couple authoritative-sounding people, call it a day. And indeed, more than one freeway has been removed in this country. But there’s no trend toward de-freewayization; quite the opposite in fact. What’s missing, then, is the underlying reasons for the changes.

Fundamentally, there are two reasons for US freeway closures:

(i) The freeway was replaced by a newer and bigger freeway, built to better design standards, at which time the old facility was abandoned.

(ii) The freeway  was part of a link in a grand master plan, that was truncated by the freeway revolts of the 70’s. In other words, it was pre-obsolesced by non-completion of the network.

Some examples:

Tom McCall Waterfront Park, Portland, Oregon

Berkeley’s Preservation Institute says: “When Portland decided to tear down the Harbor Drive freeway, the city made one of key decisions that transformed it into a national model for effective city planning.” Well… maybe. What actually happened was that they had one freeway built to 1942 standards, and in 1964 they opened up another freeway half-a-mile away built to 1964 standards. That was I-5 – the Eastbank Freeway – and it’s still truckin’ almost 50 years later.

Now it’s true that some traffic engineers raised their eyebrows at the idea. Even if Harbor Drive only had 24k ADT (which is well down into arterial territory), it was still predicted the city would grow. And considering how slowly traffic crawls across the Marquam Bridge today, perhaps there was probably a grain of truth in the forecast. But what the engineers didn’t predict was that Portland would soon enact a strict downtown height and FAR ordinance in an effort to ward off further skyscrapers in favor of the existing Glazed Terra Cotta building stock. This substantially slowed office development downtown and pushed the region’s employment base into a more suburban, office-park-dominated form. In fact, low-rise office parks are the very first thing you see when you cross the UGB into Greater Portland, whether you’re coming in on 26 East or I-5 North.

Midcentury traffic engineers thought Downtown office space would keep expanding, like any American city. Instead the downtown office market was almost frozen in time. But what really cinched the deal was when they went and built yet another freeway less than a mile away. Sandwiched by parallel north-south freeways of (then) modern design, serving a downtown whose skyline would forever be anchored by the same two buildings, there would never be a need for the widened and straightened Harbor Drive.

Park East Freeway, Milwaukee, Wisconsin

Milwaukee is a case where the infrastructure was obsolesced by the freeway revolts. In the original plan for Milwaukee’s freeway system, there were two north-south trunk highways – one inland, and one along the lake. While the inland route got built as planned (and is now signed as I-94 and I-43), the Lakefront route was only half finished. Thus the Park East Freeway – which, as designed, would’ve been an important connector distributing traffic between Lakefront and Inland routes – was rendered a fairly truncated spur. Not really necessary in its original form. And while Milwaukee gets New Urbanist props for killing the spur, it’s instructive to note what they replaced it with.

A brand-new surface street, striped for four lanes but obviously designed for six, got put right in its place. Now, from my perspective, as an infrastructure guy, I think this is pretty sweet. The original freeway was designed primarily as a connector (with distribution functions secondary), so it didn’t utilize a lot of the Milwaukee grid. A proper downtown highway spur should crap traffic out onto every surface street in sight, like 527 does. Thus the new surface street does a better job at fulfilling its primary raison d’etre, since it was actually designed for that purpose. It’s also more amenable to condos than an elevated highway is, which can be good for property values – and good for the local government, if they don’t blow it all on 20- and 30-year tax abatements like PDX does.

But a green eco-symbol this is not; it’s just the engineers replacing a middling facility with a better one.

Claiborne Expressway, New Orleans, Louisiana

This one is actually still there,  although there’s a good chance it’ll disappear in the next decade. If you’ve read Divided Highways you’ve read the tales of Claiborne’s vibrant business and music scene before the coming of the elevated. The pictures I’ve seen show a mostly auto-oriented strip of gas stations and buy-here-pay-here lots. But these also have their charm, and I’m sympathetic to the argument. Perhaps it shouldn’t have been put there.

What we do know for certain is the Claiborne didn’t last ten years before it had been supplanted with I-610, which cut several miles off the route for through-traffic. At this point the Claiborne became essentially just a spur, albeit one masquerading as a through route.

Even just as a spur, there would be a pretty decent argument for the Claiborne’s continued existence… except that the downtown NOLA office market isn’t exactly booming. In fact the consensus is, during times when a surface-street Claiborne would be slow, all the extra traffic could just be routed up the Ponchartrain, which is a solid eight lanes with full-width shoulders and feeder roads. Even CNU proposes that they add a direct connector for this purpose.


You can find more examples wherever you look. The Embarcadero Freeway in San Francisco was supposed to have been a vital shortcut between the Bay Bridge and the Golden Gate, providing a downtown through route to complement the east bay’s 580. In fact they only got it about 1/3rd built before they ran into affluent neighborhoods and the rest of it got canceled. The truncated version lasted until an earthquake killed it, at which point it was deemed not worth saving. But what if they’d finished it?

It’s not too hard to figure out what would’ve happened, since basically the same freeway got constructed in Seattle – the Alaskan Way Viaduct. When that freeway got wounded in a quake, they just patched it up with duct tape and JB Weld and set about planning Seattle’s Big Dig as a replacement. If the Viaduct had been cut off halfway – say, if it never went north of the Seneca exit – well, it’d probably have been torn down by now and replaced with a tourist trolley. Conversely, if the Embarcadero had been completed as designed, San Francisco would have almost certainly embarked on its own “Big Dig.”

The West Side Highway in New York doesn’t really count, since New York was broke in the 70’s. The highway collapsed because there wasn’t enough money to do even preventative maintenance. And certainly not enough to rebuild. In fact, they didn’t even tear it down for another 15 years – it just sat up there, closed to traffic. Yet even this gets spun as some sort of “cities transcending the freeway” narrative.

The Freeway Revolts Are Over

Assuming our economy doesn’t completely implode in the next five years, we’ll continue to build newer and better highways that obsolesce old ones. And when that happens, those old ones will make great spots for redevelopment. If I was Houston, I’d seriously be looking at I-10 between Crockett and Jensen – which has, by far, the worst geometry of any of the downtown freeways – and moving it about a half a mile north, opening up more of the north side of the Bayou to development.

What’s not going to happen anymore are the truncated spurs, the freeways rendered obsolete by revolts. It’s not because the concerns over freeways have gone away. It’s just that engineers have become sensitive to them.

The master freeway plans of the 40’s and 50’s were models of rationality and efficiency. But they didn’t really account for anything besides rationality and efficiency. Houston largely followed theirs and it’s one of the reasons the place is so easily navigable today. But the original plans also sliced right through parks, forests, wetlands, rich people neighborhoods. And thus the revolts.

It’s arguable that we’ve lost something. Newer highway alignments are no longer quite the paragons of scientific virtue they were in the drafting easel era. In a smaller, newer city like Tulsa, you can see the difference between 50’s and 60’s alignment studies versus modern ones.

I look at the alignment for SH 130 south of Austin and I’m amazed at the number of squiggles needed to put a highway through flat, relatively undeveloped terrain. But while this design methodology doesn’t necessarily result in better highways, it does result in highways that will be built.

And this is where the post-freeway era ends. There’s a very limited supply of highways that are “overbuilt” as a result of their connections never materializing. New construction isn’t going to provide us with any more because they’ll detour and slosh around anything that might have put up a fight 40 years ago. As time goes on, the pace of freeway removal will *slow*, not increase.

Clickbait article writers, take note.