Interregional High Speed Rail: the myth of the 400 mile cap

Recently, we began to consider interregional high speed rail, or, in other words, high speed rail spanning more than the current corridors proposed. Before we delve in to details, it’s time to dispel some myths. The first one is that high speed rail is not competitive over distances of 400 miles.

No, I’m not making that up. Obviously, as distances become longer, air travel becomes more competitive, since when they are flying at cruise level, planes are faster than trains. However, making up a number, in this case 400 miles, is just not true. The problem is that very important economists writing for very important newspapers (in this case, Ed Glaeser for the Times and Robert Samuelson for the WaPo) make stuff up, and because they have degrees from places like Harvard, people believe them.

Both writers pieces have been thoroughly discredited (and there are many more such posts, like this one), but no one has mentioned one of Samuelson’s rather-blatant misrepresentations. In his piece, he states as fact (without any source, of course), that

Beyond 400 to 500 miles, fast trains can’t compete with planes.

. This is rather interesting. Why? because not only does he fail to mention places where trains compete comfortably with planes in a 400-500 mile corridor, but he doesn’t mention either a 500+ mile corridor where a train line doesn’t compete or offer any rationale about why they couldn’t.

So, I’ll do his dirty work for him. First of all, let’s find a city pair with high speed rail of greater than 400 miles. Say, Paris to Marseille. By air, it’s 406 miles, by road, it’s about 482. Either way, it’s in Samuelson’s not-really-competitive range. Here’s the interesting thing. Of the air-rail market on the Paris-Marseille route, the TGV has taken 69% of the traffic. That’s up from 22% before completion of the line. I think that’s competitive.

In fact, it’s time, not distance, that governs competitiveness, and the time is definitely more than three hours. According to SNCF’s Guillaume Pepy

High-speed rail has historically captured the major share of combined air/rail traffic along routes where train journeys are under 3 hours. But this is changing, says SNCF’s Pepy: “With air travel becoming more complicated and increasing airport congestion, high-speed rail now wins 50% of the traffic where rail journeys are 4.5 hours or less,” he said. On the Paris-Perpignan route (5 hrs by train), TGV has 51% of the air/rail market, on Paris-Toulon (4 hrs) 68%.

It seems that, even for trips of four or five hours, high speed rail can be competitive. In that amount of time, a train averaging 160 mph could cover 640 to 800 miles. If that is the case, then a lot more corridors are plausible for consideration for high speed rail including a route between the East Coast and the Midwest. Especially between cities with congested airports. In other words, New York and Chicago.

Viaducts: in with the old

Recent travels have taken me to Chicago, San Francisco and Paris (and explain the lack of activity on this page). It is the latter of these cities which I am going to use to explore the above-city landscape.

What is the above-city landscape? Well, in the late 1800s and through much of the 1900s, cities realized that it was generally quite easy to build transportation networks above street level. The first of these took the form of steam-powered, elevated railroads. In most cases, these were built on metal structures above the street, but in a few, they were built as stone or masonry structures instead. The next generation were electrically-powered elevated railroads, which were mostly built in the early 1900s and, in many cases, torn down during the latter half of the 20th century, which were followed by, after 1950, mostly concrete elevated road structures.

Most of these structures, especially the narrower, non-road ones, were built over existing roads. (Road structures are often several lanes wide and required significant property takings, as there were no existing rights of way wide enough to carry them.) Thus, when they fell in to disuse or when they were made redundant by paralleling surface or underground routes, most were seen as a blight to the landscape and torn down. Metal, over-street elevateds are easy targets: they are ugly, they block light, they generally carry noisy traffic and their supports impede the flow of traffic. If they no longer serve a purpose (such as carrying passengers) there is usually little debate as to their fate. Abandoned elevateds are a rare sight indeed.

In a some cases, however, elevated railroads were not built over a street, but next to it, or in between streets. Examples of this type of construction include some active lines, such as the Park Avenue Viaduct in New York, the Reading Viaduct in Philadelphia and various elevated lines in Chicago (the Red Line north of the Loop and the Blue Line east of Logan Square). Quite often, however, segments of urban, elevated lines have been abandoned, for various reasons: a new at-grade or (more often) underground segment opened, their need was made redundant by a parallel line, or the need they served ceased to exist. Once this occurs, cities are left with long, grade separated rights-of-way, and no clear procedure for what to do with them.

Urban viaducts are often seen as a blight, and while they do represent significant infrastructure, there is often pressure to tear them down. In Boston, no one could wait to get rid of the Central Artery—there was almost no discussion of keeping it for any reason. Although it could have been used as an elevated park or a means to connect North and South Stations, consensus was to remove it and reconnect the city to the waterfront. This was likely the correct approach; the structure was close to 100 feet wide and ran between the city and the harbor, casting an ominous shadow.

In some cases, however, disused structures are less abhorred and there is not such swift pressure to demolish them. This, in particular, is the case with railroad structures. Few, if any, highway structures in cities are less than six lanes wide—if you are going to bother building an elevated highway and the various accoutrements which go with it (exits, entrances, underpasses, and such), it makes little sense to build it as a two-lane roadway. A two lane roadway can not handle much traffic, and the marginal cost of adding a few extra lanes is relatively small. Thus, highway bridges tend to be at least six lanes wide, and with shoulders, barriers and supports, they are often 100 feet wide (add a couple of exit ramps and they are even more intrusive). Furthermore, because the roadways need to be accessed from below, these structures are usually built at a minimum height above other streets, often providing less than 20 feet of clearance. Thus, highway structures tend to create large and dark spaces underneath, which are almost universally disliked.

Railroad structures, however, often are built differently. Height is less of a consideration, although elevated structures are usually not built any higher than necessary. However, width is much less of an issue. Highways need to be built to a considerable width because the capacity of a highway lane is only about 2000 persons per hour. One railroad track, however, can carry ten times that many people (trains carrying 1000 passengers and operating at three minute headways are commonplace), so in most cases, no more than two tracks are needed. In a few cases, three tracks are built to allow for extra capacity, and sometimes even four—although since the entirety of Grand Central Station can be served by four tracks, wider structures are rarely necessary. And since railroads don’t need breakdown lanes, exit ramps or barriers, elevated railroad structures are rarely wider than about 40 feet, and often only 20 feet from side to side. These structures are not as often seen as the “Chinese Walls” that highways (or railroads built entirely on fill) are compared to, and therefore not universally torn down when they are no longer in use.

While aerial structures have been abandoned for some time, there is not yet a definitive protocol for what to do with them. Some, of course, are torn down and, often, the rights of way are used for new structures, all but obliterating the previous use (except to the well-honed eye). For instance, the CTA in Chicago demolished several short elevated segments, such as the Humbolt Park Line (the only visible traces of which lie in buildings which end suspiciously short of nearby alleys) and the north end of the Paulina Connector (redundant once the State Street Subway was built), which is only visible where the structure is still used for railroad signals.

More recent closures, however, have not necessarily been followed with demolition. As cities have transformed, planners and residents have realized that there is potential to use old viaducts to create unique urban spaces. Demolishing such structures often leave narrow and sometimes-bizarre plots of land which are not conducive to new development (especially when they are less than two dozen feet wide), so the land does not have much intrinsic value. However, the structures are often quite sound (having often been overbuilt) and seen as opportunities to bring green space in to the city—without demolishing the structure. The two most significant examples of this type of reuse are the Viaduc des Arts / Promenade Plantee in Paris and the High Line, which very recently opened in New York City. We’ll explore both of these in an occasional series.

Upper Midwest High Speed Rail

(This post was originally written as a comment on Richard Florida’s Blog at The Atlantic in response to a comment about what new markets would be opened with a high speed rail link from the Twin Cities to Chicago)

Time-competitive rail service from Minneapolis to Chicago would transform the market dramatically. Currently there are three main options:

* By car, which is 400 miles each way and takes seven hours, without traffic. With traffic, it’s quite a bit longer. The cost is dependent on gas prices, but if you are going to Chicago you have to worry about parking, and tolls

*By train/bus, which takes about the same time as driving. By bus, it’s seven hours in a airplane-legroom seat (hell on wheels); by train it’s more spacious and more expensive

* By plane, which is pretty quick (1:20 in the air plus an hour-or-so on each end) but has extremely variable costs. Since Southwest is flying the route and demand is low, tickets with advanced purchase are currently about $100 roundtrip. Last summer, however, when only Northwest, American and United were “competing,” tickets cost on the order of $400. And without 14 days advanced notice, tickets on the route are over $200 even now.

In other words, travel by air is relatively fast but has a very variable costs, and is quite dependent on fuel prices. Travel by road or rail is cheaper, but the time cost of at least half a day makes it very unattractive. There is currently no middle ground–a relatively fast service which has relatively low prices and does not have draconian fees for booking at the last minute. This is one market which currently does not exist.

The other markets which do not currently exist are for intermediate city pairs. The most logical route (despite what various Minnesota politicians continue to argue) is via Saint Paul, Rochester (Mayo Clinic), Winona (Winona State Univ.), La Crosse (Univ. Wisconsin campus), Madison and Milwaukee. Looking at it in a similar manner to a recent post on this blog:

Route, Gmaps driving time, distance, 155 mph HSR time, daily flights
Saint Paul – Chicago via Milwaukee,     7:05, 422 miles,     HSR: 2:43,     Flights: 50,
Saint Paul – Chicago direct Chicago – Madison,     6:33, 401 miles,     HSR: 2:35,
Saint Paul – Rochester,     1:33, 78 miles,     HSR: 0:30,     Flights: 6*,
Saint Paul – Winona,     2:21, 113 miles,     HSR: 0:44,
Saint Paul – La Crosse,     2:42, 150 miles,     HSR: 0:58,     Flights: 6*,
Saint Paul – Madison,     4:22, 262 miles,     HSR: 1:41,     Flights: 5,
Saint Paul – Milwaukee,     5:19, 328 miles,     HSR: 2:07,     Flights: 17,
Rochester – Madison,     3:30, 211 miles,     HSR: 1:22,
Rochester – Chicago,     5:41, 350 miles,     HSR: 2:15,     Flights: 6,
La Crosse – Madison,     2:30, 143 miles,     HSR: 0:55,
Madison – Milwaukee,     1:25, 79 miles,     HSR: 0:31,     Flights: 4*,
Madison – Chicago,     2:34, 147 miles,     HSR: 0:57,     Flights: 11*,
Milwaukee – Chicago,     1:42, 92 miles,     HSR: 0:36,     Flights: 12*

I threw in daily flights as an afterthought, and it’s not all it might seem to be, since both Minneapolis (Northwest) and Chicago (United, American, Southwest) are hubs. So a lot of people on those flights are going somewhere else and just making the first leg of their trip. I put an asterisk (*) where it seemed most of the flyers were in this group; flights where it wouldn’t make sense to fly for such a short leg. Some of these routes, particularly Madison to Milwaukee, could easily be replaced by rail service if it existed (Milwaukee has an airport train station). In fact, I was once on a Midwest flight which had a cracked windshield and we needed a new plane, so they pulled one off the Milwaukee-Madison route and, presumably, put the passengers in a couple of cabs to Madison.

Routing some trains via O’Hare could potentially eliminate a lot of short, inefficient feeder plane trips, which the airlines might actually want to drop. You can already buy a plane ticket and travel portions on TGV or even Amtrak (anywhere between New Haven and Wilmington to Newark, for example). If you’re on Continental flight 94XX, you’re on a train.

The other market this opens are the various intermediate markets along the route (and similar markets exist along other routes, for sure). First of all, several cities become suburbs. Rochester (home of the esteemed Mayo Clinic) and Winona (two colleges) become suburbs of the Twin Cities, and the Mayo Clinic becomes a quick trip from Saint Paul or Minneapolis. On the other end of the line, Madison and Milwaukee become suburbs of each other, and both become suburbs of Chicago.

La Crosse would be a bit more than an hour from the Twin Cities and Chicago, wuoldn’t quite be a suburb of either, but its location in between the two could be quite advantageous. The same can be said for Madison, which would be less than two hours from the Twin Cities. Each city would be linked with several major Creative Class-type economies (Chicago, Madison, Minneapolis).

Finally, some other intermediate trips link rather interesting pairs. Milwaukee-Madison-La Cross link three of the campuses of the University of Wisconsin. Rochester, which is currently an hour, by road, from Winona and La Crosse, would be linked in less than half that time. And Winona and La Crosse, now a 45 minute drive (with no real public transport) would be a 15 minute trip by train.