A New Gateway

A New Gateway

St. Louis’ Gateway Transportation Center is quite new. It was completed in 2008. It has four tracks, a dozen Greyhound bus bays, and connections with Metrolink and the St. Louis bus system. It’s a step up from the modular Amshack that previously inhabited the site. It won an award. But that doesn’t mean it’s very good.

Indeed, it’s crammed into some residual space under the I-64 viaduct. It’s a better entry to the city than what Cleveland still has (of course, Cleveland only gets its Amtrak trains at 2 AM), but it doesn’t hold a candle to St. Louis Union Station, a couple of blocks to the west.

That said, St. Louis Union Station — formerly a 42-track stub terminal — doesn’t match modern passenger rail needs particularly well, either. When the station was built, it was a true terminal: you could catch a train there in any direction, but all of the two dozen or so railroads that went to St. Louis either began or ended there. If you wanted to go from Kansas City to Indianapolis, say, you had to change trains at St. Louis: nobody ran one straight through.

The biggest problem with Gateway, though, is that it’s much like Providence station. There, the station was built to handle 1980s traffic; it’s already become congested as traffic volumes along the Northeast Corridor have risen. Here, too, the station was built around mid-2000s traffic — the facilities are extremely undersized for even modest traffic growth.

And St. Louis, which once had some two dozen railroads servicing it, is well-placed for more than just “modest” traffic growth. If even regular hourly or semi-hourly service to Chicago and Kansas City would overwhelm Gateway’s current facilities, then what would 4-train-a-day service to Little Rock do? Or new service to Indianapolis? Louisville? Oklahoma City? A dozen major destinations lie between 200 and 500 miles from St. Louis, and Amtrak currently only offers service to a quarter of them. St. Louis as intercity corridor hub demands a larger facility. And that’s without investigating the potential for commuter rail in the St. Louis region! (Clearly, with two dozen major railroads accessing the city, the opportunities for commuter service throughout the region are extensive.)

The First Problem

Buses can access most points in downtown St. Louis. Trains can’t. Even constructed as a run-through station, there are only so many places where a major intercity station can be built, simply because there are only so many places where adequate station boxes can be provided.

Screenshot 2016-06-21 at 1.46.54 AM
Potential station boxes in St. Louis. From left to right, counterclockwise: original Union Station box; new Union Station through-station box; Gateway box; Busch Stadium box; Gateway Arch box; and Edward Jones Dome box

It’s arguable that American intercity rail is so skeletal that it’s difficult to get a good sense of how large a station box at a 12-route nexus ought to be. This is only partly true — actually, there are a number of former stations, some still in use, scattered around the country. Buffalo Central Terminal, currently abandoned, had 14 tracks — but was noted for being too large for its traffic. Detroit’s Michigan Central Station, also abandoned, had 10 or 11 tracks, while Toledo’s Central Union Terminal — still in use — had 12 tracks historically and has 5 tracks today. Fort Wayne’s Baker Street Station’s box is in a condition that can charitably be called “decrepit” (seriously, Michigan Central’s box is in better condition), but it appears it historically hosted six tracks. Indianapolis Union Station had 12 tracks. Baltimore Penn has 8 tracks. And so on.

Generally speaking, a station track requires platform space of equivalent width (by loading gauge) to board and deboard. That is, if the track has a loading gauge of 15 feet, then 15 feet of platform width is required. The Association of American Railroad (ARRA)’s loading gauge is 10’8″; this suggests the platform width per track also needs to be 10’8″. We can then simplify by rounding to 11′; this yields a per-track platform width of 11’4. (We can see this in action at the current Gateway transportation center, where two tracks and the platform servicing them is a total of 44′ wide.) The upshot of this is that the station box needs to be between 132′ (6 tracks) and 264′ (12 tracks) wide.

There are six such sites available in St. Louis.

  1. The first site is, of course, the former Union Station box. The idea here would be to rip out the failing urban mall in the trainshed and convert it into a new station facility; another possibility would be to use the umbrella-platform annex to the west as the main facility. The stub nature of this box, however, is a significant detriment.
  2. The second site is the current Gateway box. This site is relatively narrow, bordering I-64 to the north and a Terminal Railroad Association (TRRA) yard to the south. This suggests it is expandable, but any such expansion would have to involve moving TRRA tracks around.
  3. A third possibility would be a transverse underground station box as part of a Union Station redevelopment (to replace the failing mall etc.). Redevelopment on the trainshed’s east side renders this difficult, as buildings stand where the throat would need to be.
  4. A large parking lot lies between the TRRA tracks and Busch Stadium. Using this space for a major train station is tantalizing, but it is also psychologically the furthest from downtown and has the poorest connection with Metrolink. It is also close to the end of the MacArthur Bridge approach.
  5. Gateway Park itself certainly has room to host a station within its limits. A station between the Arch and the Eads Bridge has excellent access to Metrolink here, and of course the station entrance would be one of the country’s most unique, evoking Chicago’s Millennium Station and not much else. The problem with this site, though, is that there is no good access to it from the MacArthur Bridge, forcing any traffic into Illinois across the Merchants Bridge.
  6. Finally, there’s the Edward Jones Dome. The Rams recently vacated the space, moving to Los Angeles, and while a major station would be an interesting use of a 20-year-old arena, and be convenient to the busiest part of downtown St. Louis, such a facility would also have to be stub-tracked with no connection to the MacArthur Bridge at all.

For this discussion, I’m focusing on improving facilities around the current Gateway site.

Gateway: The Challenge of Building Under a Bridge

Clearly, the dominant feature of the Gateway site is I-64. It sets roof heights between Poplar Street and the station box. The escalator ramp needed to solve this clearance challenge renders part of the station box unusable (part of why Gateway currently only has four tracks rather than six). And, for as much of an improvement as it offers, Gateway still feels less like a proper city entrance and more like something crammed under the Interstate. Which is, of course, exactly what it is.

But there is a difference between having your design fight your site and having it work with your site. Gateway’s design ultimately fails because it fights its site at every turn. It tries its level best to ignore the Interstate until it can’t anymore, and — partly because of that fight — runs its bus dock up the site’s north side, keeping the station from truly opening into the city.

The thing is, building under a bridge is a solved challenge. It’s not uncommon to find the spaces under European railroad viaducts let to vendors, and the base of the Queensboro Bridge is home to a supermarket. If you let it, the I-64 viaduct can drive your design and make a better project than you thought possible. In order to do this though, you have to do more than just build around the bridge — you have to let your structure be part of the bridge.

Programming the Station

Screenshot 2016-06-21 at 1.45.42 AM
A new Gateway program. In green we have a new entrance park, a feature alluded to in my article on St. Louis greenspace. Abutting it are extensions of 15th and Spruce streets; just to its south in dark red are the main concourses — the North one with its park views, the West one wrapping around the 15th St. porte-cochere, the South one dramatically placed under I-64, and the Bus one servicing the western bus terminal. In yellow are the secondary concourses, accessing Metrolink, the city buses, and Amtrak furtherest south, while secondary station functions (offices, food halls, retail, etc.) are shown in blue and material handling is in purple alongside the tracks. In pale red, across the street from the bus terminal, is the main parking garage; in blue are four associated development parcels.

Here’s one way to do it.

First, the section under the higher span (~25′-63′ clearance) is separated into two hall spaces. The western block is the intercity bus terminal hall, with major entrances onto 18th under the viaduct, and at the corner of 16th and Poplar. The bus bays themselves lie under the eastbound viaduct, accessed via 16th, with a busway directly linking to 20th or 21st streets. This section is structurally separate from the rest of the facility, but its curtain wall will be in the same style as the main hall; above the bus entrance on 16th will be a large sign with a station logo on it.

The second hall span is the Grand Hall, the principal public feature of the whole facility. Extending from ~60′ east of 15th all the way to 16th, all under I-64’s westbound viaduct, the Grand Hall is a towering space, defined by a glass curtain wall on its north side with a narrow clerestory suspended between the eastbound and westbound viaducts. This gives a serene, gracious, sunlit space. The Grand Hall functions as the primary pedestrian path linking the bus terminal with the rest of the station; in addition, an adjacent food court spills out into it.

Just below the Grand Hall, under the eastbound viaduct (~38′ clearance), is the West Hall, home of the main food court. Keep in mind here that the food court is located between the two intercity termini — this makes it easy for weary laid-over travelers (and Greyhound sure does like their layovers) to find food. The ceiling, while significantly lower than the Grand Hall’s, is still spacious, but the space will most likely be organized into a cafeteria surrounded by small commissaries. Access to the food court can be had from the Grand Hall and the waiting room. Restrooms are also conveniently located in it.

East of the food court is the Amtrak waiting hall and concourse. The Grand Hall’s huge expanse means this facility doesn’t have to be as large as it otherwise might; even so, the main hall fills the space under I-64’s eastbound spaces from the columns west of 15th to those east of it. Locating the concourse entrance slightly west of where it currently is gives it slightly more room under the slip ramp to 14th Street, which in turn makes the facility’s most significant vertical pinch point less of one. In addition, once the pinch point is passed, access to the concourse is had by a pair of elevators flanking a a grand stair — grand stairs can be squeezed into less space than escalators, and every inch of space that isn’t involved in other station functions is an inch we can squeeze into the station box for wider platforms or more tracks.

East of the Amtrak concourse, the Grand Hall meets the West Hall. This is where the under-the-interstate portion of the facility meets the (mostly) freestanding portion. The West Hall runs along 15th Street from Spruce to Poplar, and midway along it lies the station’s dropoff porte-cochere, a feature that makes an unobtrusive overall imprint on the façade. This part of the station, which curls east into a North Hall along the 1400 block of Spruce, is an “entrance complex”, connecting the station with the city. It fronts a public park occupying the block bounded by Clark Avenue and 14th, 15th, and Spruce streets, covering the Metrolink tracks’ curve.

Next, we have the Metro Concourse. This concourse connects from roughly where the Grand Hall meets the West Hall east to the MetroBus terminal along 14th Street, with a north branch connecting to Metrolink’s Civic Center station. Between this concourse and the entrance halls, we have an internal square, which could be viewed as expansion space for station services.

Finally, we turn our attention to the back of house. A baggage facility lies alongside the station box, perhaps connecting to a baggage carousel next to the Amtrak concourse; under the viaduct to the east are station offices and administration. This can further extend out along the rooftop of the North Building, which may well offer an event space as well. And a parking deck along Poplar between 16th and 18th streets offers rental car and long-term storage for station users.

Be the Building, Be the Bridge

Screenshot 2016-06-21 at 2.14.08 AM
The I-64 viaduct presents a significant site challenge, but also an invitation to dramatic architecture. The trick is integrating the station facility into the existing viaduct structure.

Okay, what about the architectural treatment?

Well, we clearly have plenty of structural members already in place to work with — the bridge pillars and deck. This bridge is designed to take moving traffic; cleverly designed, most of the relatively light static load of the station can be integrated into this dynamic load. The idea here, then, is to develop the south building’s structural members as cantilevers, with only the parts closest to the ground being transferred onto their own foundations.

Overall, the site is split into two structures: the south and north buildings. The south building is wholly integrated into the I-64 viaduct: its structure is attached to the overlying bridge, and its building program is determined by the bridge’s constraints. The north building is freestanding; its roofline should be similar in height to the eastbound I-64 roadway deck, perhaps curving up to provide a more dramatic flair and strong boundary to the station park. The offices can also extend along the middle of the north building, especially if the north end of the structure gets an events hall — they can share the same elevator core, one in the freestanding north building.

That said, the steel structure used to suspend the Grand Hall’s glass curtain will almost certainly decide the entire station’s ornamentation and branding regime. Every other façade, even the clerestory between the interstate’s decks, will have to relate to it in some way. I suspect a sunburst pattern for the main girders will do the job, which can then be extended in a sort of spiky Gothic-like pattern all around the north building …


Screenshot 2016-06-21 at 2.20.59 AM
A simple diagram of potential regional corridors that hub at St. Louis

The basic plan here involves a modest improvement to throughput capacity — where the current station box is only four tracks (three through and one stub), this plan has a six-track (all through) station box; there would be about a dozen intercity bus bays. One can certainly argue this is not enough — St. Louis is the natural hub for all travel from the Midwest to points southwest (e.g. in Oklahoma or Texas), and about a dozen natural corridors radiate from the city. Beyond that, any commuter rail in the St. Louis region will run through Gateway, as will high-speed rail, both taking up slots of their own.

It is worth noting, however, that the minor TRRA yard immediately to the south of the box offers significant expansion space, though perhaps one with the price that TRRA will have to reconfigure its traffic patterns — as it stands, this TRRA yard appears to be an arrivals facility for traffic off of the ex-Frisco and ex-MoPac lines to the southwest.

The plan for Gateway Terminal that I’ve outlined is instead meant to provide major improvements for site circulation and enough available space to handle increased demands, one that can grow along with the terminal’s needs. This is perhaps the greatest advantage of its current site — this ability to grow from a small facility to a large one, perhaps modularly over time.

While looking into old Midwestern intercity stations, I discovered that many of them, ca. 1950, handled 3,000-5,000 passengers a day; this is roughly the amount Los Angeles Union Station currently serves. For Gateway Terminal, handling that many passengers should be the natural end goal.

This is not unattainable, especially not for a well-designed transportation hub. The city’s main north-south and east-west light-rail corridors intersect at this site; any reasonable commuter rail schema would converge on it; and St. Louis sits at one of the Midwest’s largest corridor hubs, with paths stretching to Chicago, Columbus, Cincinnati, Kansas, Oklahoma, and Texas.

Hopefully the ideas I’ve laid out here can grow into a hub St. Louisians find handsome, eminently usable, and able to take them wherever their hearts’ desire takes them.


It’s the End of the United Kingdom, And I Feel Fine

It’s the End of the United Kingdom, And I Feel Fine

So the UK has decided to leave the EU. What are the ramifications? The scope of their decision is enormous. On one hand, it is wise: the EU, after all, has become home to stifling bureaucracy, and if one sees its construction as the great neoliberal project of the last century, then leaving it is a rejection of neoliberalism. On the other hand, idiotic and even insane: the UK is itself a confederation, and by leaving the EU, the seeds of its own undoing have been cast.

Applying my prognostication powers, I see that the long-term Brexit effect will be, bluntly, catastrophic for Britain — but it will also begin an EU-wide internal crisis. In the long run, the major effects seem to be (a) an unwinding of the UK, and (b) at minimum, some badly-needed reorganization within the EU.

(It is also interesting to note that the Leave vote spread’s demographics seems to correlate closely with the American far right’s.)

Some of the effects are very short-term and we can suspect them being played out within the next few years.

Decamping of London’s financial sector. Much like New York, London is a financial services city, first and foremost. And also much like New York, it is a city that functions as a financial command and control center on a continental scale.

By voting to leave the EU, the British have effectively shorn themselves of the advantages London offered. While the city will remain as a center of local finance, the global financial sector will be forced to move elsewhere on the continent (Paris, Frankfurt, and Dublin all seem to be natural choices here — Paris because it is the EU’s largest city; Frankfurt because it’s the largest financial center on the mainland; and Dublin because it’s the largest city in the Anglophone EU and a financial center in its own right).

It is the very presence of the global financial services industry that bids up London housing prices into the stratosphere, which means that anybody, and I mean anybody, who owns even a scrap of land in London will be looking to get out now. Which brings us to…

British recession. While this is unlikely to hit the already-economically-depressed British Rust Belt, aka its North, all that much, prices in the London property market will massively fall as the financial services sector leaves, in turn taking a large suite of ancillary services and hangers-on with it. I’d say that prices throughout the city will fall by 25%-50%, and the economic reordering that comes with it will plunge Britain into a protracted recession — unitary states function as overlarge city regions; the country as a whole is dependent on its dominant city.

But wait!, you say. The UK is not a unitary nation — it’s a confederation. Well, that leads us to…

The end of the UK. Note not “as we know it” — the UK as a sovereign nation will be ended. Why? Well, leaving the European Union is politically divisive throughout the UK — even more so when one considers that, while the Brexit heartland was Northern England, the Scottish and Northern Irish both voted to stay in (and every part of Scotland voted to remain).

The UK’s EU membership was a key part of why Scotland’s recent independence referendum failed, and as such, Scotland has already signaled it wants to pursue another independence referendum — one now tied with accession to the EU in its own right. Similarly, over in Northern Ireland reunificationist politicians (like Sinn Fein) are signaling that they want a referendum on whether to secede from Great Britain and unify with Ireland.

Of course, the Troubles still scar the latter, but in Scotland the political calculus has changed to such a degree that most will be in favor of independence (remembering that last year’s referendum barely failed). The bottom line here is that the UK is toast, the catalyst for its long-term disintegration laid.

An interesting subplot here is that Scotland may seek to be named legal successor to the UK’s EU membership (contingent on an independence referendum, of course). If this does indeed turn out to be the case, then Scotland finds itself in the advantageous position of being able to secede from the UK and accede to the EU in one swift stroke. Let the curmudgeonly Tories run their own damn country into the ground!

The overall effect of all of this is a shifting of political and economic power away from London; London has no say in an independent Edinburgh, nor will it have anywhere near the financial clout it used to. In fact, in the Scottish succession scenario, Scotland may well turn out to be the most advantageous position for the EU’s financial muscle to congregate, flipping the British Isles’ economy on its head!

We can also see some medium-term Brexit effects brewing, most of them manifesting in deeply ironic ways.

Greater European Regulatory Control. Even though England is out of the EU, that does not mean they are free of EU regulatory meddling. Quite the contrary, in fact. England’s economy is so interdependent on Europe’s and the UK had amassed so many EU opt-outs that — in their natural next step, EFTA membership — they will have to renegotiate quite literally everything.

One of these, hilariously, is border control. While the UK and Ireland had identified their open border across Ireland as the reason for Schengen opt-outs (with eurozone Ireland likely to join Schengen should the border across the island cease to be a problem), every other EFTA member — Switzerland, Norway, Liechtenstein, and Iceland — is part of the Schengen community, and indeed, common Schengen membership may well be now seen as an implicit part of any EFTA member’s EU relations …

The long and the short of this is that England just bungled away a status quo that actually gave it the strongest border control in all of Western Europe. There are very few universes in which Brexit does not lead to England joining Schengen, mostly out of economic necessity.

The same holds true with the way EFTA broadly applies the EU acquis, the shared law that enables the implementation of the common market. Which, of course, has to be applied for any European good imported into England, or any English good exported to the EU.

Diminution of English influence, both in Europe and around the world. This happens on two fronts — first of all, the diminution of influence one would naturally expect, given that it’s now only a matter of time before Scotland leaves for Europe’s bluer pastures; and second, because England just rid itself of the representation it had within the EU.

There’s also the question of the military costs of Scottish independence, which — again — is all but inevitable, given that Scotland wants in even more badly than England wants out. Much of the UK’s ability to project military power was and still is based in Scotland … and other areas which the Brits own due to military reasons, such as Gibraltar, are themselves intimating they’d rather be in the EU.

In short, the next few years will determine whether or not the EU is indeed Europe’s hegemon. The cards are stacked against England: London’s might as a financial center? Gone. Its ability to use EU membership to carve out opt-outs hither and yon? Gone. The Brexiters’ own stated goals of greater British autonomy? Gone in the most ironic way possible.

What about the EU, anyway? Here there is no easy answer. A number of European states have strong independence movements, some led by charismatic idiots; members of this bloc have cheered Farage’s apparent success.

For them, Brexit is just the first step in dismantling the EU. Nationalist politics — Britain for the Brits, Denmark for the Danes, Holland for the Dutch, Czechia for the Czech, etc. — come to dominate the European continent again. But the problem of this is that it flies in the face of the reason the EU came about in the first place. Sovereign states need to be bigger in a smaller world to survive; states like Denmark or Czechia are deeply interdependent on their neighbors, their blocs, and Europe as a whole.

But Brexit is also a grand experiment in whether this politics is even viable. (Spoilers: It probably isn’t.) Obviously, Brexit and its ancillary effects — devaluation of the pound (perhaps long-term as kicking the financial sector out may well be a loss the country can never recover from), recession (what, other than the financial sector, does Southern England have that makes it wealthier than Northern England?), disintegration of sovereignty (the surety of Scottish secession), and so on all make the UK look like an early-2000s’ Austria-Hungary.

However, there are other real effects elsewhere in the EU. Obviously, this reawakens the whole question of European sovereignty; recall that its greatest structural problem is that it is a pooling of sovereignties but not a fully independent sovereignty in its own right. A second major problem is military projection: With the UK out, France is now the EU’s strongest military player. For nationalists like Farage or Wilders, a collapse of the European project is to be applauded; for most workaday Europeans, such a scenario, with internal trade subject to punitive tariffs, decreased continental mobility, and so on, is terrible.

My Own Prediction: Brexit is tantamount to England kicking London’s financial sector out. This sector has been what has powered the British economy; with it gone, Southern England starts to look an awful lot like Northern England. Obviously, the European financial sector will reestablish itself elsewhere on the continent — where, exactly, is unknown.

Meanwhile, Scotland will begin working on its next independence referendum. Behind the scenes, Scotland will lobby to succeed the United Kingdom as an EU member. If successful, Scotland would secede from the UK and accede to the EU the same day the UK’s EU membership ends. This begins the UK’s breakup.

With its financial sector gone, England sinks into a deep recession. Unable to receive EU monies, England cannot subsidize Wales and Northern Ireland — and the increasing economic gap between the latter and Ireland leads to calls for them, too, to pursue seceding from England.

In order to access the European market in any meaningful way, England must join EFTA. Not doing so enforces punitive tariffs on a sluggish economy; doing so forces England to submit to Schengen and the EU acquis despite it not being a EU member.

The EU is forced to reexamine itself — in particular, its central bureaucracy. Perhaps Brexit is a significant enough crisis of sovereignty to force it to a constitutional convention.

Meanwhile, the economic malaise — coupled with inflated living costs — leads to a migration reversal. Opportunity is no longer found in Britain for Britons, and this in its turn leads to a re-accession movement, as the full scale of what the EU offers England is gradually realized.

This is probably going to be the biggest long-term Brexit effect: Breaking Britain’s belief it can go it alone, giving way to a generation that realizes England needs Europe even more than Europe needs England.

The Nation

The Nation

Europe stands on a knife’s edge between conflicting desires for greater unification and nationalism. Most European nations now have far-right parties calling for dissolution of the European Union; in some — such as Britain — they have grown quite powerful. The idea of “Brexit,” of a nation leaving the European Union, seemed ludicrous just a decade ago. Yet here we are.

The fact that it’s being done via votes and referenda, by ballot questions like “Should Scotland declare its independence from the United Kingdom?” or “Should the United Kingdom leave the European Union?” rather than visible public protests and demonstrations, however, should not hide the fact that what we’re really seeing in Europe is a resurgence of its own — unique — political problem: that of nationalism.

In the world’s other large and diverse states, be they Russia, the U.S., Brazil, or China, the politics are nationalism are minimal. There is nobody (sane) who would put the sovereignty of Pennsylvania over that of the U.S. in general, for example; even in (somewhat less sane) Texas, the secession movement is very much a minority. When the Soviet Union fell apart, its more restive nationalities all gladly took the chance to secede; as a result, modern Russia is composed of the various oblasts, republics, and krais that decided, a generation ago, that they wanted to be Russian. And even in China, where Tibet and Xinjiang are often interpreted (by outsiders) as annexed nations, everything east of the Gobi is interpreted as being Chinese in every sense of the word; the People’s Republic and the Republic of China may disagree on politics and administration, but they both agree they are Chinese — and both claim the same territory.

But this is not the case in Europe. Nor has it been since the fall of Rome. Indeed, the whole concept of the “nation” and “nation-state” developed in Renaissance Europe. In most of world history, any given region is dominated by a single hegemon. In East Asia, this has always been China; in the Middle East, the Mesopotamian states gave way to the Persian states gave way to the Arab caliphate; in Peru, the Inca state had assumed that role by the time the Spanish came; and so forth. Historic Rome and modern Russia both follow that same model — as does the modern U.S., whose constituent “nations” (which are largely unaligned with individual states but occasionally aligned with blocs of states) latent and incipient. (One suspects the same of Brazil.)

The reason for this development should be fairly clear by now. There was no clear hegemon in Europe, no state that comfortably say they embodied what it meant to be “European”. The last state that was, was Rome, and though other states (Russia and the “Holy” “Roman” “Empire”, most notably) sometimes put on airs, as more localized ethnic identities became more important, the very idea of a single state that could span the continent began to look increasingly archaic.

This is all a very long way of saying that the modern idea of the nation is driven by the fact that the aristocracies of England, Spain, Portugal, France, Austria, and so on all had to treat with each other as equals; doing otherwise would trigger some military mire somewhere or another. And while the relationship of Ireland and Wales, say, to England; or Norway, the Faroes, and Iceland to Denmark; or Brittany, Aquitaine, and Provence to France  (I can keep going with this) was inherently unequal, these unitary bodies had roughly equal diplomatic and military power. And, as people in various subject regions began to assert their own independent cultural identities in the 19th century, the idea of the “nation” being synonymous with one of Europe’s sovereign entities began to break down further.

It is worth noting here that the nationalist movement that reached its peak in the early 20th century most likely did so then because of a combination of improved communications technologies, development of the middle class even in the sovereignties’ peripheries, and relative economic slowdowns fueling dissatisfaction with the status quo. Stop me if this sounds at all vaguely familiar.

This means that nationalist movements, as a form of political identity, are inherently reactionary — reactionary, that is, to a system that had been working for a while but does not appear to be now. We can see this, again, in Brexit: a reaction to the Euro’s troubles (a currency Britain does not use), to the Syrian migration crisis fueled by the Islamic State in the Middle East (where the British Isles are the only place north or west of Trieste not in the Schengen Area), and especially to the sense that Britain is a net donor to the EU but hardly gets anything out of it (mostly because they keep negotiating themselves opt-outs).

Reactionary movements aren’t always bad; indeed, they are a useful barometer because their rise means things aren’t working the way they used to. Reactionary movements in the arts, for example, develop when the previous dominant movement has lapsed into cliché and unintentional self-parody. Similarly, the Strong Towns movement can be said to be a reactionary movement against a failed planning and civil engineering paradigm.

But these are technical sorts of reactionary movements. What is happening with Brexit and other European nationalist movements now, and was happening with the nationalist movements that birthed Germany and Italy and later Poland, Finland, Iceland, and a dozen new nations out of Turkey and Austria-Hungary a century ago, is more a sort of political reaction — a retreat into ethnicity when things are going wrong with no clear cause why. This type of reactionary movement reflects a circle-the-wagons sort of mentality — weather the storm and we will be whole and hale.

Don’t get me wrong, this is — again — not all bad. Nationalist movements in themselves are valuable for preserving and even seeding language and literature. Nationalist movements have rescued languages like Hebrew, Manx, and Cornish from the dead; they help shine lights on dialect continua normal unitary modes of education shuffle under the rug. Nationalist movements popularize folk art and other localized forms of education. Sometimes they are needed — Mexico badly needs Nahuatl and Yucatec nationalist movements to help resurrect its Aztec and Maya cultural heritages, and the same can probably be said of Guatemala and K’iche (also Maya), and Peru (and Ecuador and Bolivia?) and Quechua (i.e. the Inca language). But, here too, such movements are a sort of cultural nationalism rather than a political one (though the latter is often tied into the former).

In the end, though, the early 20th century nationalist movements, to a large extent, failed. Yes, several left behind sovereign states, but they failed because — in the aftermath of WWII — it was widely realized across the European continent, for the first time in some fifteen hundred years, that Europe does not have the resources necessary to support a bajillion small nation-states with the standard of living the average European expects, a problem only exacerbated by their competition being other continental-scale nation-states (i.e. the U.S., Russia, and China … and, one suspects, soon enough India, Indochina, and the Mercosur bloc).

That kernel is that the E.U. has grown out of. It held true in 1500, when conquests of the New World papered the lack of adequate domestic resources over; it held true in 1800, when the Industrial Revolution drove colonialism’s second wind; it held true in 1945, when the great powers had just finished reducing each other to smoking husks; and it still holds true today.

And this is probably the greatest problem vexing Europe, the reason Brexit is a thing. It needs to be a unified hegemon in its own sphere, a continental-scale nation-state, to compete in the world today … it just doesn’t know yet whether it wants to be.

The Vacuum

The Vacuum

From the perspective of somebody who doesn’t know anything about the history of passenger railroading in America, it must seem strange that there is so little of it in the Midwest. The region is much like the heart of the European continent in quite a few measures: it’s a galaxy of small and midsize cities, anchored by a few larger metropolises, all close enough together to afford a vibrant intercity travel market.

Yet passenger service is wildly variant, between “nonexistent” and “inadequate”, throughout this region. This is, of course, a function of the history of passenger service in the region, particularly in the postwar era, and how its abandonment leads to a blindness to the fact that large chunks of the infrastructure remain. We can still see the many railroad stations, underutilized (Indianapolis, Pittsburgh, Cincinnati, Toledo) or abandoned (Detroit, Buffalo, Ft. Wayne) or repurposed (Cleveland, St. Louis), but only rarely demolished (Columbus), empty halls today that served, on average, about 5,000 passengers daily 60 years ago.

More can be done with this embodied infrastructure. More should be. We can see this simply by observing peer networks in Europe, in Japan, in Russia and Australia and China. Indeed, with a proper investment program (as Michigan has already shown, with its investment in the Wolverine corridor), there is plenty of available infrastructure for more intensive passenger flows throughout the Midwest, without impinging on freight flows.

This map is representative of one possible service breakdown–in terms of city pairs linked and convenient transfers available–centered around Michigan, Indiana, and Ohio. Of what’s shown, three (WolverineBlue Water, and Hoosier) are currently available. Yet the infrastructure exists for much, much more.

Ohio is particularly bad in this regard. Passenger trains don’t pass through Ohio during the daytime (unless they’re horribly late); yet a dense web of corridors linking Cleveland, Columbus, Cincinnati, Detroit, Pittsburgh, Buffalo, and eventually Chicago existed historically and can still be developed today. It’s a question, as people like Cap’n Transit regularly say, of priorities and will.

But that is the heart of the problem, is it not? For every successful demonstration like the Wolverine, ten more corridors lie fallow, awaiting reactivation. It will take time, dedication, and patience, but in the long run, the Midwest can rebuild its rail network, and in so doing, offer an amenity that Northeasterners and Californians are–like Europeans, Chinese, and Japanese–so accustomed to that they don’t even notice is an amenity anymore.

Headline Left By The Wayside

This is an interesting post. Let’s suss out the ramifications.

Recall here that (while it’s making the news these days for its leasing terms), Uber’s long-term plan still involves fleet automation. Walmart partnering with Uber and Lyft for grocery delivery represents an obvious angle for suburban automation.

In this model, which we’ll call a “showroom and delivery” retail model, most suburban retail needs are reduced to warehouses that send out orders via fully- or semi-automated fleets. (What’s likely is that drivers become minimum-wage workers whose main skill is customer service at the point of delivery–not even glorified delivery boys.)

One of the things I mentioned in my first post on automation was increasing efficiency of the road system, and this is part of that. The automated fleet is able to be deployed and redeployed, or heuristically deployed, more like Fed Ex or UPS than anything else. It also gradually eliminates errand needs, as instead of going to the store one can simply order online and have comestibles delivered to you. Even without automation, this is an ongoing change.

I am skeptical of how far-ranging the change will be. After all, this isn’t the first time this sort of sea change has happened. That honor would be the late 19th century, when the Sears catalog was first distributed.

Consider! That was the Amazon of its day. With a Sears catalog, you could buy a house (yes, quite literally a house) and everything for it, from furniture to clothes to a variety of, yes, foodstuffs. You didn’t need to leave the house to buy damn near anything anymore — just send the order off to Sears and, in 6-8 weeks, the package’d arrive. With the Sears catalog, how could the town carpenter stay in business? the clothier? haberdasher? dry goods seller?

Yet they did. Indeed, a lot of those small mercantile businesses became bundled in the grand department stores of the early 20th century, but enough of them remained on Main Street to keep the merchant class fueled. No, what eventually killed the merchants were the malls — and more specifically the chains that filled them. And Sears itself moved from being purely a mail-order business to a department store chain, first on Main Streets of middling prominence — and later in those same malls. For almost a century, Sears was the country’s largest, most successful, and most prominent retailer (even if it’s going the way of other early-20th-century mainstays like Woolworth’s nowadays).

The reason is, of course, that retail prognosticators continually undervalue the experience of shopping, and the meaning it still has in American life. Shops are a parade of third places, and part of the point of shopping, especially in the era when households had only one breadwinner, was to be seen out and about. The reason why the mall was what eventually killed the Main Street haberdasher is that it was, in essence, a third place made up of other third places, designed to be more attractive than Main Streets, with their traffic and weather, could ever be (at least by 1950s sensibilities). It, not the Sears catalog or even the department store, was a true 1-to-1 replacement for Main Street.

This is why we’re likely to see a rise of showroom-type retail, places to go first and foremost to browse and be seen. While automation + just-in-time deliveries will likely make staples trips (like those to the grocery store) obsolete, retail for the experience of retail has already begun to come alive again in American urban cores. This is the core essence of the boutiques that now line Main Streets in fashionable cities, and why low-margin retailing is dying even as small purveyors are offering handcrafted knickknacks behind every other window you walk past. These places are showrooms as much as stores.

Automation will speed up that process. The question is: How will this process play out in autocentric suburbs, where driving is a requirement to go much of anywhere? The critical mass of the Main Street environment has proven necessary for showroom retailing to work, and in the most dispersed places that critical mass just doesn’t seem to be there.

“Walking Distance to the Grocery Store”

“Walking Distance to the Grocery Store”

I am lucky enough to live in walking distance of a grocery store (at least, when I can actually afford to shop there). It’s five or six blocks away and I have no problem carrying a half a week’s worth of groceries home. Even in the urban core of a city like Philadelphia, this is a nice trick: the supermarket I go to is also the closest supermarket to the corner of Broad and Fairmount, a corner that’s a lot closer to Center City than where I am — only there, it’s a mile or so away instead.

Making it possible for everyone to walk to a grocery store essentially requires that they are spaced, on average, a mile apart from one another. This means that the closest grocery store is no more than half a mile from any given residence (i.e. every grocery store has a 1 square mile service area).

The problem with this should be self-evident. Grocery stores are low-margin businesses, which means that they need a lot of purchases in their respective service areas to survive. This in turn means that, as their service areas constrict, the population densities of those service areas must rise. If, as this post suggests, a population of 15,000 to 20,000 is necessary to sustain a 35k-50k sf full-service grocer, and these grocers are placed 1 square mile from each other, then they would need to have populations of 15,000 to 20,000 ppsm to be self-sustaining at that locational density.

Keep in mind here that South Philadelphia has a population density of 16,771 ppsm, itself some four times the median American urban population density of ~4,000 ppsm. For grocery stores to be close enough for everyone to walk to, then, living arrangements would have to be more akin to what we see in the big Northeastern cities than they are in Cincinnati or Seattle or Salt Lake or Denver. And of course you have to ensure that convenient walking paths to the grocery store exist within your service area.

To make a long story short, then, while Abbotsford’s residents desired the ability to walk to their grocery stores, unless they’re willing to make significant spatial tradeoffs, it will never happen. If 15k people are needed to sustain a grocer, and they have just over 130,000, then the city can support 8.67 grocers (let’s call it 9). That’s 9 over 145 square miles, or one grocer per 16 square miles.

If the service area is 16 square miles, then the radius of the service area is 4 square miles — some eight times that of our “walkable” grocers! (We can see this by realizing that the total trip distance for our outlying resident is effectively the service area’s diameter.)

This is an excellent example of how population densities matter. With a population density of 921 ppsm, Abbotsfordians are asking for something that neither the city nor interested grocers really can deliver on — they’d have to get a lot denser first, about as dense as Vancouver proper (13.6k ppsm). Even though Vancouver has Canada’s least affordable real-estate market, this is unlikely.

Indeed, the population density sheds some light into why Abbotsford is Canada’s most car-dependent city. It’s built to be that way. Its population density doesn’t even clear the threshold for running bus routes (1,000 ppsm). While it might be clear Abbotsfordians want a walkable city (and, I might add, more walkable, connected cities in the Fraser’s alluvial plain would help alleviate Vancouver’s housing crunch), the question is: are they willing to make the necessary choices to become a walkable city? Are they willing to redevelop their center and make it a proper urban place?

Amtrak’s Place in RRNY

Screenshot 2016-05-28 at 2.42.55 PM
Current Amtrak paths through New York’s core

As we have developed the RRNY system, we’re finding that there is one last part to consider: Amtrak. Compared to the main RRNY routes, which is generally scaled reach a peak throughput of 32 tph (and can easily double that if 20 tph per track signalling is implemented), Amtrak’s services seem, well, minuscule.

Currently there are 4 tph total across Hell Gate in the peak hours, and the Keystone Service is scheduled around the 8:00 crunch through the North River Tubes. As daily trains (such as the Lake Shore Limited, Silver Meteor, and Vermonter) can be assumed to depart off-peak, they will not be counted towards Amtrak’s peak runs–only hourly trains will.

RRNY has two potential Amtrak paths: a modest-growth route that fits Amtrak around other services, and a high-growth route that gives Amtrak more dedicated infrastructure. Both programs, however, have some shared features.

Key Amtrak Features

Clearly, East Side Access is not going to be particularly useful under RRNY. Most LIRR trains (2/3) are diverted down the Atlantic Branch, heading to Midtown via Lower Manhattan if on Line 4, or bypassing Midtown altogether if on Line 5. The remaining trains use the East River trunk. But it represents a lot of emplaced concrete and should not just be left abandoned in the ground.

The key shared features are these:

  1. A Penn Station-Grand Central access tunnel that utilizes the ESA cavern
  2. Medium-speed Amtrak trains use ESA and the Hell Gate line
  3. High-speed Amtrak trains use the New Haven Line, on extra slots in the core and a pair of tracks between Mt. Vernon and New Rochelle that Line 2 leaves superfluous

The idea here is to split Amtrak services along different paths, particularly in the core. Only Line 2 and Quadboro services use Hell Gate at peak, and this leaves more available slots for more-frequent medium-speed runs; a slight gap between services and available frequency on the Manhattan Spine’s upper part gives high-speed trains a faster, more direct run out of Manhattan than by Hell Gate.

Meanwhile, the West Side Line remains the Empire Service’s home.

A Modest-Growth Scenario

Screenshot 2016-05-28 at 2.44.18 PM
Amtrak routes through the core under a modest-growth RRNY scenario. This plan allots Amtrak 16 slots (8 services) across the Hudson

This scenario is limited by potential Amtrak movements through the North River Tubes complex (which includes Gateway). This is because of arbitrage between them and the East River Tubes: between Lines 1 and 2, 32 tph cross the North River Tubes at peak, while, due to the 8 tph Line 1 receives from the West Side Line at Penn Station, 40 tph cross the East River Tubes.

That means that the East River Tubes are the most congested part of the entire network, congested to the point where they set 12 tph as the upper per-track throughput limit of, again, the entire network. (If you don’t think that’s enough slots to handle the demand, recall here that these estimates are scalable, and can easily double due to the way the system is structured: the only limit is the signalling system’s throughput capacity.)

But it also means there are 16 slots available through the North River Tubes–4x the amount Amtrak currently uses. Or: there are enough slots for 8 distinct hourly services of any type heading west from New York. (Recall that a service needs two slots, one for the train coming and a second for the train going.)

Finally, because the LIRR vacates the Hudson Yards, Amtrak can also take over that space. This means that corridor services can develop between New York and various New England points, with the Hudson Yards being their terminal facility.

Under the modest-growth scenario, 8 Northeast Corridor trains per hour, reckoned directionally, meet 4 Empire Corridor trains at Penn Station. These 12 trains (i.e. 24 tph) are then routed through the Penn Station-Grand Central Connector, a graceful spiral helix linking into Grand Central’s East Side Access station cavern.

From here, trains like the Empire and Keystone corridor services that terminate in New York are deadheaded through ESA to the Sunnyside Yards. Regional services are likewise sent along the Hell Gate Line (may change if Quadboro vacuums up all the slots Line 2 isn’t using), while high-speed trains use spare slots along the Manhattan Spine to the New Haven Line. Line 4 only needs 2 tracks between Mt. Vernon and New Rochelle for its New Canaan branch; the two tracks it doesn’t use mark the start of the Boston-New York HSL.

Developing a More Extensive Service Portfolio…

Screenshot 2016-05-28 at 3.16.06 PM
A significantly less modest block of services. Here, 11 services (22 slots) are needed across the Hudson. This compliments 5 services (10 slots) along the Hudson Line and 4 services (8 slots) terminating at Penn Station

But…where can corridor Amtrak trains go? And how many of them do we need?

One way of looking at it is to develop a corridor/service breakdown. Since we know that four high speed lines converge on New York, two linked together end-to-end, we can say that each has a high-speed medium-speed service pair. In other, similarly dense, parts of the world, there are often crack “super-high-speed” expresses on top of that. So in the New York market, we have:

Northeast Corridor

  1. Boston-DC super-express
  2. Boston-DC HSR
  3. Boston-DC medium-speed
  4. New York-Philadelphia medium-speed

Keystone Corridor

  1. New York-Pittsburgh super-express
  2. New York-Pittsburgh HSR
  3. New York-Pittsburgh medium-speed

Empire Corridor

  1. New York-Toronto super-express
  2. New York-Buffalo HSR
  3. New York-Niagara Falls medium-speed

for the major corridors. We can add to that three secondary blocks of (mostly) medium-speed services, essentially supercommuter corridors reaching out from Manhattan to snare the parts of its vast commute shed RRNY can’t reach. These include:

New England Service Area

  1. New York-Portland via Worcester
  2. New York-Boston via Springfield
  3. New York-Amherst
  4. New York-Albany via Pittsfield

Mid-Atlantic Service Area

  1. New York-Buffalo via Binghamton
  2. New York-Binghamton
  3. New York-Roanoke
  4. New York-Harrisburg

Champlain “Corridor”

  1. New York-Montréal HSR
  2. New York-Montréal medium-speed

From all of this we get 16 distinct hourly runs that either originate or terminate in New York, and three that pass through the city. Breaking out the part that requires a Hudson River crossing, we find 11 services, i.e. 22 slots required. This is obviously a fair bit more than the North River Tubes’ available capacity under the more modest assumptions allowing Amtrak to use RRNY’s spare capacity.

…And The Core Infrastructure To Go With It

Screenshot 2016-05-28 at 2.46.48 PM
As Amtrak service density goes up, the need for dedicated infrastructure becomes unavoidable.  This example uses a two-track HSL tunnel to bypass the North River Tubes bottleneck; the drill is continued to the far side of the Hudson-Harlem convergence

There are two potential solutions available here–and remember, at this point, we’ve added enough capacity demands to make seeking a new solution worthwhile. The first is to do more or less the same thing that we are already doing with Gateway, i.e. making the line between Penn Stations Newark and New York six-track to handle Amtrak’s new capacity demands.

The second solution is a bit more interesting. This comes about as a reflection that the Penn Station-Grand Central connection, at two tracks, is also nearing capacity, handling nine services (18 slots) between the Empire/Champlain and New England groups alone. Essentially, Amtrak now needs four New York tracks of its own.

(Aside: This also pushes RRNY closer and closer towards a true S-Bahn, even if it’d be a very large example by German standards: S-Bahn routes are generally decoupled from intercity ones through the core.)

To provide this, I propose using the now-disused Hoboken Station throat (remember, services that used to call at Hoboken are now mostly bundled into Lines 2 and 5) as a launch box for a wide-bore two-track tunnel, where the use of a wide bore futureproofs the tunnel. This tunnel gently curves until it runs under Park Avenue, passing through Amtrak’s deep-level Grand Central station box and continuing up to the Bronx, where it surfaces just past the junction with the Port Morris Line.

Under this plan, Grand Central becomes Amtrak’s main Manhattan hub: Mid-Atlantic, Keystone Corridor, and most Northeast Corridor trains stop only there (the lone exceptions are Grand Central-Penn Station-30th Street Clockers). Penn Station continues to be important, however, as the terminus of the New England service block as well as one of the Empire Corridor’s two Midtown stations.

Capacity Breakdown

Under the service portfolio outlined above, Amtrak needs:

  • 22 slots under the Hudson
  • 18 slots off the Empire Corridor
  • 14 slots into New England

The HSR components of the Northeast and Empire corridors do not begin until one passes Newark, Mt. Vernon, or Yonkers. This allows Amtrak to use its available slots in the core for every service. But it does also mean that there may be significant chokepoints on the Hudson and New Haven lines as well–on the former, between Spuyten Duyvil and Yonkers; the latter, New Rochelle and Stamford. These may be worth investigating.

Service Allocation

The general allocation of service is thus:

  • Northeast Corridor – Super-Express, HSR, Regional: ← Newark Penn – New Hudson River Crossing – Grand Central – Amtrak Manhattan Spine – Mt. Vernon →
  • New York-Philadelphia Clockers: Newark Penn – North River Tubes – Penn Station – Penn-Grand Central Connection – East Side Access – Sunnyside Yard ⊣
  • Keystone Corridor – Super-Express, HSR, Regional: ← Newark Penn – New Hudson River Crossing – Grand Central – East Side Access – Sunnyside Yard ⊣
  • Mid-Atlantic – Harrisburg, Roanoke: ← Newark Penn – New Hudson River Crossing – Grand Central – East Side Access – Sunnyside Yard ⊣
  • Mid-Atlantic – Buffalo, Binghamton: ← Paterson – New Hudson River Crossing – Grand Central – East Side Access – Sunnyside Yard 
  • Empire Corridor – Super-Express, HSR, Regional; Champlain – HSR, Regional: ← West Side Line – Penn Station – Penn-Grand Central Connection – Grand Central – East Side Access – Sunnyside Yard 
  • New England – Portland, Boston, Amherst, Albany: ← Mt. Vernon – Amtrak Manhattan Spine – Grand Central – Penn-Grand Central Connection – Penn Station – Hudson Yard 

where and → means “service continues” and ⊣ means “end of service”, i.e. the yard where the train is stored. Note that the Mid-Atlantic, Keystone, and Empire trains are stored in Sunnyside Yards (now given over entirely to Amtrak and most likely decked over and redeveloped), while New England trains are stored in Hudson Yards. Note also that (a) the Hell Gate Line has been bypassed entirely by the new Amtrak tunnel, allowing RRNY and Quadboro (Crossboro?) services to occupy all its slots, and (b) that, due to the Penn Station-Grand Central Connection’s congestion, the Clocker is the only Amtrak service still using the North River Tubes. Any further service growth along the Empire and Champlain corridors, like, say, an hourly train to Burlington via Rutland, would require shifting the Clocker down through the New Hudson Crossing (which, remember, is futureproofed to be quad-tracked!).


RRNY is turning out to be a deeper series than I first expected. There is still one more subject I need to explore: the core tunnels proper. After that, I have some things to talk about.